NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE. Single Technology Appraisal (STA)

Transcription

1 NATIONAL INSTITUTE FOR HEALTH AND CLINICAL EXCELLENCE Single Technology Appraisal (STA) Fluocinolone Acetonide Intravitreal Implant for Diabetic Macular Oedema Alimera Sciences, Inc. Manufacturer/Sponsor Submission of Evidence 30 March 2012

2 Contents List of tables and figures... 1 Section A Decision problem Description of technology under assessment Context Equity and equality Statement of the decision problem Section B Clinical and cost effectiveness Clinical evidence Cost effectiveness Section C Implementation Assessment of factors relevant to the NHS and other parties References Appendices List of tables Table 1: Base-case cost-effectiveness results... 7 Table A1: Unit costs of technology being appraised Table B1: Summary of Disallowed Treatments for DMO in Subjects with Chronic DMO (Integrated FAME Studies) Table B2: Summary of Off-protocol Treatments for DMO by Type of Therapy in Subjects with Chronic DMO (Integrated FAME Studies) Table B3: Eligibility criteria used in search strategy Table B4: List of relevant RCTs Table B5: Eligibility criteria in the RCTs Table B6: Characteristics of participants in the RCTs across randomized groups Table B7: Primary and secondary outcomes of the RCTs Table B8: Summary of statistical analyses in RCTs Table B9: Summary of Disposition of Subjects in the FAME Studies with Chronic DMO Table B10: Summary of Relevant Outcomes for Subjects with Duration of Chronic DMO in FAME A+B Table B11: Summary of the trials used to conduct the indirect comparison 53 Table B12: Summary of Outcomes for Subjects with Chronic DMO in FAME A+B Compared to DRCR Protocol B Table B13: Ocular Adverse events across randomised groups Table B14: Incidences of Cerebrovascular Adverse Events in the Subjects with Chronic DMO in the FAME Studies Table B15: Baseline characteristics of FAc implant treated FAME patients and laser treated patients from the DRCR Protocol B trial Table B16: Key features of analysis Table B17: Summary of variables applied in the economic model Alimera Sciences, Inc. Submission of Evidence 1

3 Table B18: Average Percentage of Subjects Experiencing a Change in Vision State Table 19: Utility Weights for BCVA Health States Considered by the Model 99 Table B20: Relationship between health states and HRQL Table B21: Unit costs associated with the technology in the economic model Table B22: Monitoring costs following treatment Table B23: List of adverse events and summary of costs included in the economic model Table B24: Variables explored in sensitivity analysis Table B25: Assumptions included in PSA Table B26: Model outputs by clinical outcomes Table B27: Summary of % time in each health state over 15 years Table B28: Change in healthcare cost of blindness Table B29: Summary of predicted resource use by category of cost for FAc intervention against OSC Table B30: Summary of predicted resource use by category of cost for FAc intervention against laser only treatment Table B31: Base-case results Table B32: Probability of cost effectiveness Table B33: FAc implant treatment compared with OSC Table B34: FAc implant treatment compared with laser only Table C1: Estimated Patient Population ( ) Table C2: Total budget impact for non-fac implant market basket Table C3: Total budget for FAc implant treatment of DMO Table C4: Search strategies for RCT evidence Table C5: List of studies for further consideration Table C6: List of studies for further consideration Table C7: Details of HRQL studies identified in the literature search Table C8: Additional Treatments in the Sham Group by Duration of DMO 232 List of figures Figure A1: Typical Clinical Pathway for DMO (Retinal thickening within 500 microns of centre of macula) Figure A2: Impact of the FAc implant on the Clinical Pathway for DMO (Retinal thickening within 500 microns of centre of macula) Figure B1: Flow Chart for Studies Considered for Direct Comparison Figure B2: Percent (ASE) of Subjects with Chronic DMO with 15 Letter Increase from Baseline BCVA Figure B3: Mean (SEM) Change from Baseline BCVA in Subjects with Chronic DMO Figure B4: Mean (SEM) Change from Baseline in Excess Centre Point Thickness in Subjects with chronic DMO Figure B5: Flow chart for Indirect Comparisons Figure B6: Mean Change from Baseline Excess Centre Point Thickness over Time (Integrated Chronic DMO Population) Alimera Sciences, Inc. Submission of Evidence 2

4 Figure B7: Flow diagram for search for cost-effectiveness studies Figure B8: Economic model structure Figure B9: Treatment arm matrices Figure B10: Percent of subjects with 15 letter improvement in BCVA by number of FAc implants (post-hoc analysis) Figure B11: Percentage of Patients with 2-Step Improvement in ETDRS Score Figure B12: Association between VFQ-25 and Driving Difficulty Figure B13: Flow diagram for HRQL literature search Figure B14: Tornado diagram of deterministic sensitivity analysis, FAc implant vs. OSC Figure B15: Tornado diagram of deterministic sensitivity analysis, FAc implant vs. laser only Figure B16: Cost-effectiveness plane for the FAc implant vs OSC (10,000 iterations) 130 Figure B17: Cost-effectiveness acceptability curve for the FAc implant vs. OSC 131 Figure B18: Cost-effectiveness plane for the FAc implant vs. laser only 131 (10,000 iterations) Figure B19: Cost-effectiveness acceptability curve for the FAc implant vs. laser only 132 Figure B20: FAc Scenario analysis of FAc implant compared with OSC 134 Figure B21: FAc implant treatment compared with laser only Figure Appendix : Percent of Subjects with 15 Letter Improvement from Baseline BCVA over Time by Duration of DMO (FAME A and FAME B, with LOCF) Alimera Sciences, Inc. Submission of Evidence 3

5 Executive summary The UK approved name, brand name, marketing status and principal mechanism of action of the proposed technology. Fluocinolone acetonide (FAc) intravitreal implant in applicator (ILUVIEN ) is an intravitreally injected, sustained release implant designed to improve visual acuity in certain diabetic patients who have chronic diabetic macular oedema (DMO). The MAA was submitted using the Decentralised Procedure (DCP) with the UK MHRA as the Reference Member State (RMS). A positive outcome from the DCP was delivered on 27 February 2012 with the issuance of the Final Assessment Report (FAR) from the RMS and the agreement of the six Concerned Member States that the FAc implant is approvable. The expected approval date by the MHRA is March The active component is the corticosteroid FAc, which has anti-inflammatory and anti-vegf properties. The formulation(s), strength(s), pack size(s), maximum quantity(ies), anticipated frequency of any repeat courses of treatment and acquisition cost. The intravitreal implant in applicator contains 190 micrograms of FAc. The frequency of repeat courses of treatment will be at physician discretion based upon ongoing evidence of efficacy. The recommended dose is one FAc implant in the affected eye every 36 months with an acquisition cost of The indication(s) and any restriction(s). The FAc implant is indicated for the treatment of vision impairment associated with chronic diabetic macular oedema considered insufficiently responsive to available therapies. Alimera Sciences, Inc. Submission of Evidence 4

6 The recommended course of treatment. The recommended course of treatment is one FAc implant in the affected eye every 36 months. The main comparator(s). Given the variation across Trusts in current practice for chronic DMO patients who are insufficiently responsive, optimised standard of care can broadly be defined as no treatment, i.e. watching and waiting or registering for blindness services low vision aids when appropriate, with occasional additional maintenance laser treatments and in some cases intravitreal injections of triamcinolone or VEGF antibodies (typically bevacizumab). Considering this definition, optimised standard of care is represented in this submission through a direct comparison to the control arm of the FAME studies. The submission also includes an indirect comparison to laser monotherapy. The immediate laser treatment arm of the DRCR Protocol B Study provides a means of comparison to the standard first line therapy, laser photocoagulation. Although this population is not comprised of subjects with chronic DMO and 40% of the subjects were laser naive, this data set is the only published, peer-reviewed, multi-year data set available studying laser monotherapy in center-involved DMO. Whether the key clinical evidence in the submission comes from head-tohead randomised controlled trials (RCTs), from an indirect and/or mixed treatment comparison, or from non-randomised studies. Key clinical evidence in the submission comes from the FAME A and FAME B trials comparing the 0.2 µg/day FAc implant directly to optimised standard of care (sham injection plus rescue laser and Off Protocol Therapies (OPT) such as intravitreal triamcinolone and VEGF antibodies at physician discretion) for the treatment of chronic DMO in subjects who have received prior macular laser. Alimera Sciences, Inc. Submission of Evidence 5

7 The main results of the RCTs and any relevant non-rct evidence. For the integrated subgroup with chronic DMO, the peak effect, when measured as the difference between sham and treated in the proportion of subjects who had a 15 letter increase from baseline in BCVA, for the 0.2 μg/day FAc implant was 27.1%; and the effect was significantly maintained at Month 36 (0.2 μg/day, difference of 20.6%). The proportion of subjects who achieved a 15-letter improvement in BCVA at Month 36 was comparable to that observed at Month 24 for the integrated subgroup with chronic DMO. The most common adverse events associated with the FAc implant are typical of corticosteroids: increased intraocular pressure and acceleration of cataract formation. The impact of these events, which are predictable and mitigatable, is included in the model. In relation to the economic evaluation, details of: the type of economic evaluation and justification for the approach used the pivotal assumptions underlying the model/analysis the mean costs, outcomes and incremental cost-effectiveness ratios (ICERs) from the evaluation. In line with the NICE reference case, a cost-utility analysis was conducted. A Markov model was created in which health states were defined by 5-letter changes in BCVA in the treated eye. In years 1 through 3, disease progression reflected the treated populations in the FAME and DRCR.network protocol B studies: thereafter natural history progression rates were used, and assumptions were made regarding the extrapolation of visual acuity outcomes. Utility weights were taken from a published, peer-reviewed study conducted by Brown et al. (2000) that employed the time trade-off method to assess the utility associated with varying levels of visual acuity in patients with age related macular degeneration (AMD). Alimera Sciences, Inc. Submission of Evidence 6

8 Resource use and unit costs were estimated from clinical trials, taking the perspective of the NHS and PSS. The time horizon of the analysis was 15 years and future costs and outcomes were discounted at 3.5% per year. Base case results were calculated in a probabilistic analysis; the ICER was calculated as the ratio of the mean incremental cost and the mean incremental QALY from 10,000 iterations. Tabulation of the base-case results as follows: Table 1: Base-case cost-effectiveness results FAc implant Optimised standard of care Laser monotherapy Technology acquisition cost 13, Other costs 6, ,466 Total costs 20,270 8,939 9,283 Difference in total costs N/A 11,331 10,987 QALYs QALY difference N/A ICER N/A 22,655 16,463 QALY(s), quality-adjusted life year(s); ICER, incremental cost-effectiveness ratio The probability of the FAc implant being cost-effective in direct comparison to optimized standard of care at a willingness to pay (WTP) threshold of 20,000 was xx%, and at a WTP threshold of 30,000 was xx%. The probabilities for the FAc implant being cost-effective in indirect comparison to laser monotherapy were xx% at a 20,000 WTP threshold and xx% at a 30,000 WTP threshold. Subgroup analyses considered and clinical- and cost-effectiveness results. No sub-group analyses were conducted. Alimera Sciences, Inc. Submission of Evidence 7

9 Section A Decision problem 1 Description of technology under assessment 1.1 Give the brand name, approved name and, when appropriate, therapeutic class. For devices, provide details of any different versions of the same device. Brand name: ILUVIEN Approved name: Fluocinolone acetonide intravitreal implant in applicator Therapeutic class: Anti-inflammatory agents, corticosteroids, plain ATC code: S01BA What is the principal mechanism of action of the technology? The technology is an intravitreally injected, sustained release implant designed to improve visual acuity in certain diabetic patients who have chronic diabetic macular oedema. The active component is the corticosteroid fluocinolone acetonide (FAc), which has anti-inflammatory and anti-vegf properties. 1.3 Does the technology have a UK marketing authorisation/ce marking for the indications detailed in this submission? If so, give the date on which authorisation was received. If not, state current UK regulatory status, with relevant dates (for example, date of application and/or expected approval dates). The MAA was submitted using the Decentralised Procedure (DCP) with the UK MHRA as the Reference Member State (RMS). The six Concerned Member States (CMS) were Austria, France, Germany, Italy, Portugal and Spain. Start of the procedure was 6 September 2010 (reference UK/H/3011/01/DC). A positive outcome from the DCP was delivered on 27 February 2012 with the issuance of the Final Assessment Report (FAR) from Alimera Sciences, Inc. Submission of Evidence 8

10 the RMS and the agreement of the CMS that the FAc implant is approvable. Expected approval date by the MHRA is March The manufacturer is a small biopharmaceutical company and does not have the infrastructure needed to support a Centralised Procedure and marketing across all member states. The decision to submit via the DCP was not based on a difference in clinical practices between the UK and the other countries or trial design/data package. 1.4 Describe the main issues discussed by the regulatory organisation (preferably by referring to the [draft] assessment report [for example, the EPAR]). If appropriate, state any special conditions attached to the marketing authorisation (for example, exceptional circumstances/conditions to the licence). The main clinical issue cited in the Day 70 RMS assessment and Day 105 CMS assessments was that the benefit to risk ratio for the full patient population at 24 months was not robust. In response to this issue, 36-month data on a pre-planned subpopulation with chronic diabetic macular oedema (DMO) in whom a significant benefit to risk ratio was demonstrated was submitted in the Day 106 response. The FAR concluded that in the chronic DMO population considered insufficiently responsive to current therapies, a positive benefit to risk ratio was demonstrated. Copies of the FAR and the Summary of Product Characteristics (SPC) are provided in section 9.1, Appendix 1, and respectively. 1.5 What are the (anticipated) indication(s) in the UK? For devices, provide the (anticipated) CE marking, including the indication for use. The fluocinolone acetonide intravitreal implant (FAc implant) is indicated for the treatment of vision impairment associated with chronic diabetic macular oedema considered insufficiently responsive to available therapies. Alimera Sciences, Inc. Submission of Evidence 9

11 1.6 Please provide details of all completed and ongoing studies from which additional evidence is likely to be available in the next 12 months for the indication being appraised. No additional evidence will be available in the next 12 months. 1.7 If the technology has not been launched, please supply the anticipated date of availability in the UK. October Does the technology have regulatory approval outside the UK? If so, please provide details. There are no regulatory approvals outside the UK at this time however as previously noted, approvals are pending in the six CMS countries. 1.9 Is the technology subject to any other form of health technology assessment in the UK? If so, what is the timescale for completion? No. We anticipate a review from the SMC but no final submission date has been provided For pharmaceuticals, please complete the table below. If the unit cost of the pharmaceutical is not yet known, provide details of the anticipated unit cost, including the range of possible unit costs. Alimera Sciences, Inc. Submission of Evidence 10

12 Table A1: Unit costs of technology being appraised Pharmaceutical formulation Fluocinolone Acetonide (FAc) Intravitreal implant in applicator Acquisition cost (excluding VAT) 5500 Method of administration Intravitreal injection Doses 190 micrograms Dosing frequency Once every 36 months Average length of a course of treatment The course of treatment lasts three years Average cost of a course of treatment 5500 Anticipated average interval between courses of treatments Within 1 month of completion of previous treatment (Month 37) Anticipated number of repeat courses of treatments Will be at physician discretion based upon ongoing evidence of efficacy Dose adjustments None possible 1.11 For devices, please provide the list price and average selling price. If the unit cost of the device is not yet known, provide details of the anticipated unit cost, including the range of possible unit costs. Not applicable. This is not considered a device Are there additional tests or investigations needed for selection, or particular administration requirements for this technology? There are no additional tests or investigations needed for selection or particular administration requirements for this technology Is there a need for monitoring of patients over and above usual clinical practice for this technology? No additional monitoring will be required in clinical practice What other therapies, if any, are likely to be administered at the same time as the intervention as part of a course of treatment? None Alimera Sciences, Inc. Submission of Evidence 11

13 2 Context 2.1 Please provide a brief overview of the disease or condition for which the technology is being used. Include details of the underlying course of the disease. Diabetic retinopathy (DR), a complication of diabetes mellitus, is the leading cause of blindness in the working-age population of developed countries (ages 20 to 74) (Ciulla et al. 2003). At any time during the progression of DR, subjects can develop DMO, the principal cause of impaired visual acuity and the main cause of blindness among patients with diabetes. DMO involves thickening in the macular area due to breakdown of the blood-retinal barrier (Ciulla et al. 2003) and is considered a multifaceted disease in which hyperglycaemia leads to vascular abnormalities, leakage of the blood retinal barrier and macular oedema. The pathogenesis includes increased VEGF-induced permeability changes (Whitmire et al. 2011), inflammatory changes such as leukostasis (Erhlich et al. 2010), microglia activation (Zeng et al. 2008) and neuronal changes (Barber et al. 2011). It is now widely accepted that an inflammatory condition is present in DMO. In an animal model of diabetes, (Brucklacher et al. 2008) demonstrated that changes in permeability and apoptosis occur in concert with changes in genes involved in inflammation, vascular leakage, and neurodegeneration. 2.2 How many patients are assumed to be eligible? How is this figure derived? 53,235 patients are assumed to be eligible. The International Diabetes Federation has forecasted 3,064,000 diabetics in the UK in Ninety percent of diabetics live in England and Wales (Diabetes.co.uk website) and thus we assume there are 2,757,600 diabetics eligible. Of those diabetics, 6.6% can be expected to have clinically significant diabetic macular oedema (CSMO) (Coscas et al. 2010). This results in 182,002 diabetics with CSMO. Because diabetes is a disease that Alimera Sciences, Inc. Submission of Evidence 12

14 has many other co-morbidities and these conditions become progressively disabling over time, we understand that not all CSMO patients will be under the regular care of a retina treating ophthalmologist. We have compared epidemiological data (number of patients with the disease) (Kantar Health 2011) to patient flow through offices (patients actively being treated) and found that 22% of CSMO patients do not regularly see their retina treating physician (SciMedica Group 2011b). Additionally, the proposed therapeutic indication for the FAc implant is for the treatment of vision impairment associated with chronic diabetic macular oedema considered insufficiently responsive to available therapies. This would preclude any new patients from accessing the FAc implant. It is our understanding that existing patients represent 75% of the CSMO patients (SciMedica Group 2011b). To derive an estimate of the number of existing patients who are considered insufficiently responsive, we reviewed the Diabetic Retinopathy Clinical Research Network Protocol I. In this study, 50% of patients in the laser + sham arm were insufficiently responsive to laser treatments, as defined by a gain of up to +4 letters or any loss of visual acuity (DRCR 2010). 2.3 Please give details of any relevant NICE guidance or protocols for the condition for which the technology is being used. Specify whether any specific subgroups were addressed. Ranibizumab was reviewed as a treatment for DMO. Final guidance was issued in November That review did not address the inadequately responsive chronic DMO patients covered in this application. 2.4 Please present the clinical pathway of care that depicts the context of the proposed use of the technology. Explain how the new technology may change the existing pathway. If a relevant NICE clinical guideline has been published, the response to this question should be consistent with the guideline and any differences should be explained. The current NICE guidance on Type 2 Diabetes addresses timing and procedures for follow-up of patients with diabetes with respect to diabetic eye Alimera Sciences, Inc. Submission of Evidence 13

15 disease, but a guideline is not currently available for the treatment of DR or DMO. The existing clinical pathway is highly individualized based on the progression of DMO, response and acceptability of the various treatments, the guidance of individual trusts, and the practices of each ophthalmologist. The clinical pathway of care is not expected to change; patients will continue to be seen for check-ups and treatment in eye clinics as in current practice. The FAc implant is expected to be administered as an intravitreal injection on an outpatient basis. The following section addresses the impact of the FAc implant on the most common clinical pathway. Common Clinical Pathway Laser photocoagulation therapy is currently considered the standard of care treatment for DMO in the U.K. For patients who are insufficiently responsive, therapies available to retinal specialists vary. Some Trusts allow only laser photocoagulation and with an insufficient response, provide support via low vision aids and blind registration. On rare occasions, in other Trusts, insufficiently responsive patients are given therapies such as triamcinolone acetonide or bevacizumab on a case-by-case basis. These patients then receive support from blind services. Thus, the definition of optimised standard of care for patients who are insufficiently responsive varies on a Trust-by- Trust basis. Alimera Sciences, Inc. Submission of Evidence 14

16 Figure A1: Typical Clinical Pathway for DMO (Retinal thickening within 500 microns of centre of macula) Laser Photocoagulation globally Optimised Standard of Care Impact of the FAc implant on the Existing Clinical Pathway The FAc implant will be used in subjects with vision impairment due to chronic DMO who are considered insufficiently responsive to other therapies. Insufficiently responsive is defined as an inadequate visual function outcome. Thus, the FAc implant will provide a therapeutic option for patients who have been previously underserved. Alimera Sciences, Inc. Submission of Evidence 15

17 Figure A2: Impact of the FAc implant on the Clinical Pathway for DMO (Retinal thickening within 500 microns of centre of macula) Laser Photocoagulation (Vision impairment and retinal thickening involving the centre of macula in patients considered insufficiently responsive to available therapies) Fluocinolone acetonide 2.5 Please describe any issues relating to current clinical practice, including any variations or uncertainty about best practice. If patients do not respond to one or more focal/grid laser treatments, some Trusts may permit the use of intravitreal injections of VEGF antibodies or triamcinolone acetonide on a case-by-case basis. However, in the absence of published literature, there is uncertainty surrounding the degree of use of these therapies. 2.6 Please identify the main comparator(s) and justify their selection. Since the FAc implant is not indicated as first line therapy in the context of the clinical pathway in the UK, the main comparator is optimised standard of care. Given the variation across Trusts in current practice for chronic DMO patients who are insufficiently responsive, optimised standard of care can broadly be Alimera Sciences, Inc. Submission of Evidence 16

18 defined as no treatment, i.e. watching and waiting or registering for blindness services and low vision aids when appropriate, with occasional additional maintenance laser treatments and in some cases intravitreal injections of triamcinolone or VEGF antibodies (typically bevacizumab). Considering this definition, optimised standard of care is represented in this submission through a direct comparison to the control arm of the FAME studies. The control arm of the FAME studies was comprised of a sham injection and allowed for rescue laser at physician discretion after the Week 6 assessments. A percentage of patients in the study also received off-protocol therapies (OPT), including intravitreal triamcinolone and VEGF antibodies, as deemed medically necessary by a masked assessing physician (see section 5.2.1). Thus, this population approximates optimised standard of care in the UK for patients who are insufficiently responsive to available therapies. The submission also includes an indirect comparison to laser monotherapy. The immediate laser treatment arm of the DRCR Protocol B Study provides a means of comparison to the standard first line therapy, laser photocoagulation. Although this population is not comprised of subjects with chronic DMO and 40% of the subjects were laser naive, this data set is the only published, peer-reviewed, multi-year data set available studying laser monotherapy in centre-involved DMO. 2.7 Please list therapies that may be prescribed to manage adverse reactions associated with the technology being appraised. The most common adverse reactions associated with the FAc implant are acceleration of cataract formation and increased intraocular pressure (IOP). Patients having cataract surgery are generally prescribed topical ocular antibiotics, corticosteroids or non-steroidal anti-inflammatories (NSAIDS) for 1-2 weeks following surgery. The following classes of IOP-lowering agents were used in the FAME studies: Beta blocker; Prostaglandin; Alpha agonist; Carbonic anhydrase inhibitor; Osmotic agent; Cholinergics. Alimera Sciences, Inc. Submission of Evidence 17

19 2.8 Please identify the main resource use to the NHS associated with the technology being appraised. Describe the location of care, staff usage, administration costs, monitoring and tests. Provide details of data sources used to inform resource estimates and values. It is expected that the administration of the FAc implant will take place largely in an outpatient setting of an ophthalmic unit. The administration is assumed to take place under BZ23Z (Vitreous Retinal Procedures Category 1: outpatient procedures) at a cost of 150. It is believed that the quarterly follow up will take place during normal DMO patient monitoring under code B130 (Ophthalmology Consultant Led: Follow up Attendance Non-Admitted Face to Face) at a cost of 73 and may include RA23Z Diagnostic imaging code: Ultrasound scan at a cost of 53. The administration procedure is essentially the same as for administration of an intravitreal injection and will have similar related costs for monitoring of visual acuity, IOP and cataract. The data sources used to inform resource estimates and values include the Department of Health NHS Costing Manual, NHS Reference Costs, review of published literature and previous NICE Single Technology Assessments (NICE TA237). See section 6.5 for further details regarding the resource estimates and values. 2.9 Does the technology require additional infrastructure to be put in place? The use of the FAc implant does not require additional infrastructure to be put in place. Alimera Sciences, Inc. Submission of Evidence 18

20 3 Equity and equality 3.1 Identification of equity and equalities issues Please specify any issues relating to equity or equalities in NICE guidance, or protocols for the condition for which the technology is being used. None, as there is no NICE guidance relating to the treatment of DMO Are there any equity or equalities issues anticipated for the appraisal of this technology (consider issues relating to current legislation and any issues identified in the scope for the appraisal)? The sponsor is not aware of any issues concerning equity or equalities in the appraisal of the FAc implant How have the clinical and cost-effectiveness analyses addressed these issues? Not applicable. Alimera Sciences, Inc. Submission of Evidence 19

21 4 Statement of the decision problem Population Final scope issued by NICE Adults with visual impairment due to chronic diabetic macular oedema which has had an inadequate response to prior treatment Decision problem addressed in the submission SAME Rationale if different from the scope Intervention Fluocinolone acetonide intravitreal implant SAME Comparator(s) Outcomes Optimised standard care without fluocinolone acetonide intravitreal implant (including corticosteroids, laser photocoagulation, and anti-vegf therapy). The outcome measures to be considered include: Best corrected visual acuity (the affected eye) Best corrected visual acuity (both eyes) Mortality Optimised standard care (including corticosteroids) and laser monotherapy Best corrected visual acuity (affected/treated eye); Proportion of people receiving subsequent photocoagulation Central retinal thickness measured by optical coherence tomography Adverse events associated with We believe two control populations should be considered: 1. A direct comparison to optimised standard of care, represented by the control arm of the phase 3 clinical trials for the FAc implant 2. An indirect comparison to laser monotherapy The rationale for including an indirect comparison to laser monotherapy is that this is the standard first line therapy and in some locales, the only therapy utilized for treatment of DMO. We are satisfied with the outcomes listed in the draft scope with the exception of the following comments: 1. BCVA in both eyes: treatment in only the worse eye was assessed in the FAME trials; however, we do recognize that a proportion of patients in clinical practice do require bilateral treatment. Alimera Sciences, Inc. Submission of Evidence 20

22 Final scope issued by NICE Proportion of people who receive subsequent laser photocoagulation Central retinal thickness measured by optical coherence tomography Adverse effects of treatment, including cataracts, raised intraocular pressure, retinal detachment and stroke Health-related quality of life Decision problem addressed in the submission the treatment including cataract/raised intraocular pressure/retinal detachment/stroke Rationale if different from the scope Thus, in our economic analysis we intend to employ assumptions about bilateral treatment in line with the precedent set by the NICE s recent review of ranibizumab in DMO (i.e. 35% of patients treated bilaterally) with a resulting 25% uplift in HR-QoL). 2. Mortality: we do not believe this is an appropriate outcome measure to be assessed. Mortality was not evaluated as a treatment outcome in the FAME trials and thus there are no data to support this analysis. However, a heightened relative risk (RR) of mortality for the general diabetic population will be included in the economic model analysis, in line with the precedent set by NICE s recent review of ranibizumab in DMO (i.e. RR of 2.45 for DMO vs in the general population). 3. We do not propose to use HRQOL results as an outcome due to the nature of the design Alimera Sciences, Inc. Submission of Evidence 21

23 Final scope issued by NICE Decision problem addressed in the submission Rationale if different from the scope of ocular clinical studies. The phase 3 trials for Fluocinolone acetonide were based on the treatment of only one eye, which was considered to be the worst eye from the perspective of visual acuity. Monocular subjects or bilaterally treated subjects were not included in the patient population because the FAc implant is not approved by any regulatory authority and thus their inclusion would be considered unethical by an ethics review committee. Because of this design feature, treatment related quality of life differences cannot be detected because of the bilateral vision of all subjects enrolled in the studies. This phenomenon has been seen in other phase 3 trials for retinal diseases and is not a unique finding. In the cost effectiveness model, we do employ validated utility values which have been associated with various levels of vision. Alimera Sciences, Inc. Submission of Evidence 22

24 Economic analysis Subgroups to be considered Special considerations, including issues related to equity or equality Final scope issued by NICE The reference case stipulates that the cost effectiveness of treatments should be expressed in terms of incremental cost per qualityadjusted life year. The reference case stipulates that the time horizon for estimating clinical and cost effectiveness should be sufficiently long to reflect any differences in costs or outcomes between the technologies being compared. Costs will be considered from an NHS and Personal Social Services perspective. None None Decision problem addressed in the submission SAME SAME SAME Rationale if different from the scope Alimera Sciences, Inc. Submission of Evidence 23

25 Section B Clinical and cost effectiveness Element of health technology assessment Defining the decision problem Comparator(s) Reference case Section in Guide to the methods of technology appraisal The scope developed by NICE and Therapies routinely used in the NHS, including technologies regarded as current best practice and Perspective costs NHS and PSS to Perspective benefits All health effects on individuals to Type of economic evaluation Synthesis of evidence on outcomes Measure of health effects Source of data for measurement of HRQL Source of preference data for valuation of changes in HRQL Discount rate Equity weighting Cost-effectiveness analysis and Based on a systematic review 5.3 QALYs 5.4 Reported directly by patients and carers Representative sample of the public An annual rate of 3.5% on both costs and health effects An additional QALY has the same weight regardless of the other characteristics of the individuals receiving the health benefit HRQL, health-related quality of life; NHS, National Health Service; PSS, Personal Social Services; QALY(s), quality-adjusted life year(s) 5 Clinical evidence ILUVIEN is a sustained-release intravitreal drug delivery system that releases submicrogram levels of fluocinolone acetonide (FAc), a glucocorticoid, in the vitreous humor for 36 months. The positive outcome of the Day 210 Assessment Report indicates that the MAA has been deemed approvable for the following indication: Alimera Sciences, Inc. Submission of Evidence 24

26 ILUVIEN is indicated for the treatment of vision impairment associated with chronic diabetic macular oedema considered insufficiently responsive to available therapies. This indication is based on a subgroup from the phase 3 clinical trials (FAME A and FAME B) assessing the FAc implant for the treatment of diabetic macular oedema (DMO). Use in patients with chronic DMO derives from the results obtained from a preplaned subgroup analysis which was based on assessing the primary outcome as a function of the median duration of DMO at baseline. Upon unmasking of the dataset, it was determined that the median duration of DMO in FAME Studies was 3 years. A significant treatment effect was found in those subjects treated with the FAc implant having been diagnosed with DMO for 3 years. 5.1 Identification of studies Describe the strategies used to retrieve relevant clinical data, both from the published literature and from unpublished data that may be held by the manufacturer or sponsor. The methods used should be justified with reference to the decision problem. Sufficient detail should be provided to enable the methods to be reproduced, and the rationale for any inclusion and exclusion criteria used should be provided. Exact details of the search strategy used should be provided in section 9.2, appendix 2. The strategy employed to retrieve relevant clinical data involved identifying all trials in which the FAc implant has been investigated in the target population specified in the MAA. Alimera Sciences is the only source of the ILUVIEN FAc implant and the only clinical trials in subject with chronic DMO performed with ILUVIEN are the phase 3 FAME studies. This was confirmed with a literature search. Appropriate data for comparison of the FAc implant to other treatments for chronic DMO considered insufficiently responsive to other therapies would be: Laser monotherapy followed by observation Alimera Sciences, Inc. Submission of Evidence 25

27 Optimised Standard of Care: laser monotherapy with occasional laser treatments and in some cases intravitreal injections of triamcinolone or VEGF antibodies (typically bevacizumab) 5.2 Study Selection for Direct Comparison to the FAc Implant Describe the inclusion and exclusion selection criteria, language restrictions and the study selection process. A justification should be provided to ensure that the rationale is transparent. A suggested format is provided below. In order to provide an accurate assessment of the clinical and costeffectiveness of the FAc implant in its indicated target population, the comparator groups should provide data relative to this target population. Two important criteria are the determination that the patient has chronic DMO and that s/he is considered insufficiently responsive to available therapies. As there is no generally accepted definition for either of these criteria, a detailed discussion of each is provided in Appendix The indication for chronic DMO stemmed from a pre-planned sub-group analysis that was conducted on the basis of the median duration of DMO in the FAME trials. The median duration of DMO was 3 years, and as a result this has been included in the inclusion criteria for study selection. Further information regarding the sub-group analysis can be found in section The FAME studies enrolled subjects with DMO who had received prior laser photocoagulation who had increased central retinal thickness ( 250 microns per time-domain OCT) and baseline visual acuity of 19 and 68 ETDRS. By definition, these subjects were not responding adequately to the standard of care, laser photocoagulation, and are considered insufficiently responsive for the purposes of the study selection. Summary of Off-Protocol Medications in Subjects with Chronic DMO During the FAME studies, some subjects received treatments that were not permitted in the study protocol at the discretion of the investigator. These Alimera Sciences, Inc. Submission of Evidence 26

28 subjects were retained in the study and included in the full analysis population analysis. In subjects with chronic DMO, the percentage of subjects who received at least 1 off-protocol treatment for DMO in the study eye was significantly (p<0.001) higher in the sham group (35%) compared with the 0.2 μg/day FAc (13%) group. Of these, most received 1 or 2 off-protocol treatments. The mean number of such treatments was also significantly (p<0.001) higher in the sham group (1.0) compared with the FAc implant treatment group (0.2). Table B1: Off-protocol Treatments Number of Off-protocol Treatments, n Summary of Disallowed Treatments for DMO in Subjects with Chronic DMO (Integrated FAME Studies) Sham (N = 112) Treatment Group 0.2 μg/day FAc (N = 209) Mean (SD) 1.0 (2.0) 0.2 (0.7) Minimum, Maximum 0.0, , 4.0 P-value <0.001 Number of Subjects, n (%) At Least 1 Treatment 39 (34.8) 28 (13.4) Number of Off-protocol Treatments by Frequency, n (%) 0 Treatments 73 (65.2) 181 (86.6) 1 Treatments 15 (13.4) 16 (7.7) 2 Treatments 7 (6.3) 5 (2.4) 3 Treatments 7 (6.3) 6 (2.9) 4 Treatments 3 (2.7) 1 (0.5) 5 Treatments 0 (0.0) 0 (0.0) 6 Treatments 2 (1.8) 0 (0.0) 7 Treatments 1 (0.9) 0 (0.0) 8 Treatments 1 (0.9) 0 (0.0) >8 Treatments 3 (2.7) 0 (0.0) The percentage of subjects with chronic DMO who received intravitreal steroids was statistically significantly higher in the sham group (24%) Alimera Sciences, Inc. Submission of Evidence 27

29 compared with the active treatment group (8%, 0.2 μg/day FAc, p<0.001), as was the percentage of subjects who received posterior sub-tenon s steroids (5%, sham; 1%, 0.2 μg/day FAc; p 0.007) The percentage of subjects who received anti-vegf therapy was also statistically significantly higher in the sham group (15%) compared with the active treatment group (3%, 0.2 μg/day FAc, p 0.002). The percentage of subjects who received vitrectomies was not statistically significantly different between the sham group and the 2 active treatment groups. Table B2: Summary of Off-protocol Treatments for DMO by Type of Therapy in Subjects with Chronic DMO (Integrated FAME Studies) Type of Therapy Treatment Group Sham (N = 112) 0.2 μg/day FAc (N = 209) Intravitreal steroids 27 (24.1) 17 (8.1) P-value 1 <0.001 Posterior sub-tenon s steroids 5 (4.5) 1 (0.5) P-value Anti-VEGF therapy 17 (15.2) 7 (3.3) P-value 1 <0.001 Vitrectomies 9 (8.0) 9 (4.3) P-value P-value based on Cochran-Mantel-Haensel chi-square test stratified by baseline BCVA. Specific Inclusion/Exclusion Criteria for Selection of Studies for Direct Comparison The criteria used to select the studies which provide a direct comparison to the FAc implant are summarized in Table B3. The comparator of shaminjection plus laser photocoagulation and other therapies as needed is defined as the optimised standard of care. Alimera Sciences, Inc. Submission of Evidence 28

30 Table B3: Eligibility criteria used in search strategy Clinical effectiveness Inclusion criteria Population: Patients with DMO 3 years duration Centre retinal thickness 250 microns, BCVA 19, 68 letters, At least one prior macular laser photocoagulation Interventions: FAc implant vs. sham injection plus laser and other therapies as needed. Outcomes: % with 15 letter improvement from baseline BCVA, Mean change BCVA, % receiving subsequent laser photocoagulation, central retinal thickness, adverse events associated with treatment including cataract, increased IOP, retinal detachment and stroke. Study design: randomised, double masked, multi-centre prospective parallel group study Language restrictions: English Exclusion criteria Population: Retinal oedema not due to diabetic retinopathy Interventions: None Outcomes: None Study design: Not randomised Language restrictions: Non-English A flow diagram of the numbers of studies included and excluded at each stage should be provided using a validated statement for reporting systematic reviews and meta-analyses such as the QUOROM statement flow diagram ( The total number of studies in the statement should equal the total number of studies listed in section Alimera Sciences is the only source of ILUVIEN FAc implants and the FAME studies are the only randomized clinical trials comparing the FAc implant to a control group performed by the manufacturer. This was confirmed by a literature search using PubMed. Therefore, the combined FAME studies are the only appropriate direct comparison group. (Figure B1) The sham treatment group is considered optimised standard of care and the subgroup with duration of DMO 3 years is the basis of the indication: subjects with Alimera Sciences, Inc. Submission of Evidence 29

31 chronic DMO who are considered insufficiently responsive to available therapies. It should be noted that the ILUVIEN FAc implant (which is injected into the inferior vitreous) is not identical to the RETISERT FAc implant (which is surgically implanted near the front of the eye). As a result the efficacy and safety data for these two products is not equivalent. Figure B1: Flow Chart for Studies Considered for Direct Comparison When data from a single RCT have been drawn from more than one source (for example, a poster and a published report) and/or when trials are linked (for example, an open-label extension to an RCT), this should be made clear. Not applicable. Alimera Sciences, Inc. Submission of Evidence 30

32 Complete list of relevant RCTs Provide details of all RCTs that compare the intervention with other Table B4: Trial no. (acronym) C A (FAME A) and C B (FAME B)= the FAME studies therapies (including placebo) in the relevant patient group. The list must be complete and will be validated by independent searches conducted by the Evidence Review Group. This should be presented in tabular form. A suggested format is presented below. List of relevant RCTs Intervention Comparator Population Primary Study Reference 0.2 μg/day FAc implants Sham injection (+ rescue laser and off protocol treatments at physician discretion = optimised standard of care) Subjects with DMO who have received 1 prior macular laser Month 24 CSR: C A Month 36: Synoptic Report C A Month 24 CSR: C B Synoptic Report C B Month 36 ISE Month 36 ISS Please highlight which of the RCTs identified above compares the intervention directly with the appropriate comparator(s) with reference to the decision problem. If there are none, please state this. FAME A and FAME B compared the 0.2 µg/day FAc implant directly to optimised standard of care (sham injection plus rescue laser and OPT at physician discretion) for the treatment of chronic DMO in subjects who have received prior macular laser. The MAA approval for the FAc implant (and thus the decision problem) is based on a subgroup of patients within the FAME studies with duration of DMO 3 years. This subgroup was derived from a pre-planned analysis (see section 5.3.6). The results of the 0.2 μg/day FAc dose group are the subject of this submission When studies identified above have been excluded from further discussion, a justification should be provided to ensure that the Alimera Sciences, Inc. Submission of Evidence 31

33 rationale for doing so is transparent. For example, when studies have been identified but there is no access to the level of trial data required, this should be indicated. None of the relevant studies have been excluded from further discussion. List of relevant non-rcts Please provide details of any non-rcts (for example experimental and observational data) that are considered relevant to the decision problem and a justification for their inclusion. Full details should be provided in section 5.8 and key details should be presented in a table; the following is a suggested format. There are no relevant non-rcts considered relevant to the decision problem. 5.3 Summary of methodology of relevant RCTs As a minimum, the summary should include information on the RCT(s) under the subheadings listed in this section. Items 2 to 14 of the CONSORT checklist should be provided, as well as a CONSORT flow diagram of patient numbers ( It is expected that all key aspects of methodology will be in the public domain; if a manufacturer or sponsor wishes to submit aspects of the methodology in confidence, prior agreement must be requested from NICE. When there is more than one RCT, the information should be tabulated. Methods for the FAME Study FAME A and FAME B were identical prospective phase 3 studies performed under the same protocol. The studies were multicentre, double-masked, sham injection controlled, parallel-group studies conducted in North America, Europe and India (101 sites). Eligible participants were adults age with DMO who had received at least 1 prior laser treatment whose vision was 19 to 68 ETDRS letters and centre retinal thickness was 250 microns at baseline. Important exclusion criteria were IOP >21 or glaucoma, and systolic Alimera Sciences, Inc. Submission of Evidence 32

34 blood pressure >180 or diastolic blood pressure >105. The studies took place under the same study protocol between September 2005 and September Patients were randomly assigned to receive either sham injection or intravitreal injection of 0.2 μg/day or 0.5 µg/day FAc implants. Retreatment was permitted anytime after the Month 12 assessments provided the subject meet retreatment criteria. Laser photocoagulation was permitted as needed after Week 6. Subjects who received other intravitreal treatments for DMO (off-protocol treatments, OPT) were considered in violation of the protocol; however, they were not withdrawn from the study and were included in the full analysis population (see section 5.3.7). A total of 956 subjects were randomized in a 1:2:2 ratio stratified by baseline BCVA and site, and enrolled in FAME A and FAME B (N=185 sham, N= μg/day FAc and N= μg/day FAc) using a computer generated schedule and an integrated voice recognition system. Subjects were stratified based on their BCVA at baseline (<49 and 49 ETDRS Letters). The primary endpoint was the proportion of subjects with an improvement of 15 letters from baseline BCVA at Month 24 and the study continued through Month 36 to assess persistence of effect. For the primary efficacy variable, pair-wise comparisons were made between active drug and sham using a Cochran-Mantel-Haenszel (CMH) chi-square test stratified by baseline VA. Mean change from baseline in BCVA was analyzed using an analysis of variance model with treatment group and baseline VA strata as fixed effects. FAME A and FAME B each demonstrated a statistically significant effect of the FAc implant on vision. Because the studies were identical in design and execution, the integrated results are used for the cost-effectiveness model. Alimera Sciences, Inc. Submission of Evidence 33

35 5.3.2 Describe the RCT(s) design (for example, duration, degree and method of blinding, and randomisation) and interventions. Include details of length of follow-up and timing of assessments. The following tables provide a suggested format for when there is more than one RCT. FAME studies A and B were performed under a single protocol (C ) as randomised, double-masked, sham injection-controlled, parallel-group, multi-centre studies. The studies consisted of 18 planned visits over 36 months. The studies were randomised and double-masked to eliminate bias. Subjects in FAME A were located in the northern locations of North America, Europe and India while FAME B subjects were located in the southern locations. Three treatment arms included 956 randomised to sham injection control, 0.2μg/day FAc and 0.5 μg/day FAc in a 1:2:2 ratio. To preserve masking, 2 investigators were used. One investigator performed the treatments and the other masked investigator performed all assessments and determined retreatment eligibility. Following screening and drug administration, visits were performed at Week 1, Week 3, Week 6, Month 3 and quarterly through Month 36. Participants Provide details of the eligibility criteria (inclusion and exclusion) for the trial. The following table provides a suggested format for the eligibility criteria for when there is more than one RCT. Highlight any differences between the trials. Alimera Sciences, Inc. Submission of Evidence 34

36 Table B5: Trial no. (acronym) FAME A and FAME B Studies Eligibility criteria in the RCTs Inclusion criteria Males and non-pregnant females at least 18 years of age BCVA of 19 and 68 letters (20/50 or worse but at least 20/400) in the study eye by an ETDRS chart. BCVA of the non-study eye must be no worse than 20/400 Diagnosis of diabetes mellitus (type 1 or type 2). Any one of the following will be considered to be sufficient evidence that diabetes is present: Use of insulin for the treatment of diabetes for at least the 3 months prior to screening Use of oral antihyperglycemia agents for the treatment of diabetes for at least the 3 months prior to screening At least one macular laser treatment more than 12 weeks prior to the screening visit DMO based on investigator s clinical evaluation and demonstrated on fundus photographs, fluorescein angiograms, and optical coherence tomography (OCT) Mean foveal thickness of at least 250 µm by OCT in the study eye Ability and willingness to comply with the treatment and follow up procedures Ability to understand and sign the Informed Consent Form. No expectation that subject will be moving out of the area of the clinical center to an area not covered by another clinical center during the next 36 months. Exclusion criteria Pregnant, lactating females or females of childbearing potential (unless using reliable contraception, i.e. double barrier, surgical sterilization, oral contraceptives, Norplant, intrauterine device (IUD) Laser treatment for DMO within 12 weeks of screening or judged to be necessary within 6 weeks following enrollment Any ocular surgery in the study eye within 12 weeks of screening Yag capsulotomy in the study eye within 15 days of screening Prior intravitreal, subtenon, or periocular steroid therapy within 3 months prior to enrollment (e.g., triamcinolone) or prior treatment with intravitreal anti-vegf treatment within 2 months of enrollment (Lucentis, Avastin, Macugen.) Systemic treatment with Avastin is also not allowed within 3 months prior to screening. Any change in systemic steroidal therapy within 3 months of screening Glaucoma, ocular hypertension, intraocular pressure (IOP) > 21 mmhg or concurrent therapy at screening with IOP-lowering agents in the study eye Retinal or choroidal neovascularization due to ocular conditions other than diabetic retinopathy Alimera Sciences, Inc. Submission of Evidence 35

37 (e.g., presumed ocular histoplasmosis, high myopia (spherical equivalent > 8 diopters), macular degeneration) Any active viral, fungal or bacterial disease of the cornea or conjunctiva or any history of a potentially recurrent infection which could be activated by treatment with a steroid, (e.g., ocular herpes simplex virus). Known or suspected hypersensitivity to any of the ingredients of the investigational product or to other corticosteroids History of vitrectomy in the study eye History of uncontrolled IOP elevation with steroid use that did not respond to topical therapy History or presence of any disease or condition (malignancy) that in the investigator s opinion would preclude study treatment or follow-up Any lens opacity which impairs visualization of the posterior pole or significantly impairs vision, in the opinion of the investigator. Peripheral retinal detachment in prospective area of insertion Participation in another clinical trial within 12 weeks before the screening visit or during the study Resting Systolic blood pressure at the screening visit of greater than 180 or diastolic blood pressure greater than 105. Adapted from Pharmaceutical Benefits Advisory Committee (2008) Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra: Pharmaceutical Benefits Advisory Committee Alimera Sciences, Inc. Submission of Evidence 36

38 5.3.4 Describe the patient characteristics at baseline. Highlight any differences between study groups. The following table provides a suggested format for the presentation of baseline patient characteristics for when there is more than one RCT. The treatment groups with chronic DMO in the FAME studies were comparable at baseline: mean age years, % male, predominantly white. Most of the subjects had Type 2 diabetes and the mean duration of DMO for the groups ranged from years. Table B6: Baseline characteristic Characteristics of participants in the RCTs across randomized groups C (FAME A+B, Subjects with Chronic DMO) Sham 0.2 μg/day FAc 0.5 μg/day FAc FAME A + B (n = 536) (n = 112) (n =209 ) (n =215 ) Mean Age 62.9 years 63.7 years 63.9 years Gender 60.7 % male 39.3% female Race (%) White Black/African American Asian Other Mean Duration of DMO years Diabetes Type (%) Type 1 Type % 8.0% 22.3% 0.9% 57.4% male 42.6% female 73.2% 6.7% 20.1% 0% 63.7% male 36.2% female 74.9% 7.4% 17.2% 0.5% 5.4 years 5.1 years 5.1 years 8.9% 91.1% 8.6% 91.4% 6.5% 93.5% Alimera Sciences, Inc. Submission of Evidence 37

39 Outcomes Provide details of the outcomes investigated and the measures used to assess those outcomes. Indicate which outcomes were specified in the trial protocol as primary or secondary, and whether they are relevant with reference to the decision problem. This should include therapeutic outcomes, as well as patient-related outcomes such as assessment of health-related quality of life, and any arrangements to measure compliance. Data provided should be from pre-specified outcomes rather than post-hoc analyses. When appropriate, also provide evidence of reliability or validity, and current status of the measure (such as use within UK clinical practice). The following table provides a suggested format for presenting primary and secondary outcomes when there is more than one RCT. The percent of subjects with 15 letter improvement from baseline BCVA is the standard assessment for phase 3 studies in DMO because it is considered a clinically meaningful outcome for measuring efficacy. This parameter was the protocol-specified primary endpoint. In addition to assessments of the anatomical changes (mean change in excess retinal thickness and % with 3 step worsening in ETDRS grade, the Visual Function Questionnaire-25 (VFQ-25) was specified as an exploratory outcome to assess changes in the health-related quality of life. Alimera Sciences, Inc. Submission of Evidence 38

40 Table B7: Trial no. (acronym) FAME A and FAME B Studies Primary and secondary outcomes of the RCTs Primary outcome(s) and measures % with 15 letter increase in BCVA Reliability/validity/ current use in clinical practice Current standard for clinical trials for DMO and AMD Secondary outcome(s) and measures Mean change from baseline BCVA Mean change in excess retinal thickness % with 3 step worsening of ETDRS grade % requiring laser Reliability/validity/ current use in clinical practice Reliable, standardized method for visual acuity Reliable, retinal thickness is a standard assessment for DMO Standard assessment by Reading Centre Current standard of care Alimera Sciences, Inc. Submission of Evidence 39

41 Statistical analysis and definition of study groups State the primary hypothesis or hypotheses under consideration and the statistical analysis used for testing hypotheses. Also provide details of the power of the study and a description of sample size calculation, including rationale and assumptions. Provide details of how the analysis took account of patients who withdrew (for example, a description of the intention-to-treat analysis undertaken, including censoring methods; whether a perprotocol analysis was undertaken). The following table provides a suggested format for presenting the statistical analyses in the trials when there is more than one RCT. Table B8: Trial no. (acronym) FAME A and FAME B Studies Summary of statistical analyses in RCTs Hypothesis objective To determine whether either dose level of the implant is superior to the control group with respect to the proportion of subjects who have an increase of 15 or more letters of BCVA (VA responders) at Month 24 compared to baseline. Statistical analysis For primary (% subjects with 15 letter improvement from baseline BCVA): Pearson chisquare test, with Hochberg- Bonferroni adjustment for multiple comparisons Sample size, power calculation For Study A or Study B: 180:180: 90 subjects expected to provide 89% power to detect a difference of 16% including a projected 10% dropout rate. Data management, patient withdrawals Full Analysis Population includes all subjects randomised with LOCF for missing data. 29% of subjects discontinued before Month 36 assessment Provide details of any subgroup analyses that were undertaken and specify the rationale and whether they were pre-planned or posthoc. The decision problem relates to the subjects with chronic DMO (duration of DMO 3 years) which was the basis of the labeling. This subgroup was derived from a preplanned assessment of the primary efficacy variable of 15 letter responders. Upon unmasking of the dataset, it was determined that the Alimera Sciences, Inc. Submission of Evidence 40

42 median duration of DMO in this population was 3 years. A significant treatment effect was found in those subjects treated with FAc implant having DMO for 3 years. Participant flow Provide details of the numbers of patients who were eligible to enter the RCT(s), randomised, and allocated to each treatment. Provide details of, and the rationale for, patients who crossed over treatment groups and/or were lost to follow-up or withdrew from the RCT. This information should be presented as a CONSORT flow chart. For the FAME studies, 1279 subjects were screened of whom 323 were screen failures. The majority were excluded because they did not meet inclusion/exclusion criteria (264 subjects). Less than 10% of those excluded (29) did not choose to give consent or withdrew consent prior to treatment. The remaining 30 subjects did not participate for other reasons. Across the treatment groups 72-76% of subjects completed the 3 year study. The most common reasons for discontinuation were: subject withdrew consent, was lost to follow up and death (Table B9). Alimera Sciences, Inc. Submission of Evidence 41

43 Table B9: Summary of Disposition of Subjects in the FAME Studies with Chronic DMO Chronic DMO Category Sham 0.2 μg/day FAc 0.5 μg/day FAc Total randomised (N) Randomised, not treated (n, %) Randomised and treated (n, %) Total completed (n, %) Total discontinued (n, %) Subject withdrew consent Lost to follow-up Death Adverse event Unsatisfactory therapeutic effect Protocol violation Unknown Critical appraisal of relevant RCTs The validity of the results of an individual study will depend on the robustness of its overall design and execution, and its relevance to the decision problem. Each study that meets the criteria for inclusion should therefore be critically appraised. Whenever possible, the criteria for assessing published studies should be used to assess the validity of unpublished and part-published studies. The critical appraisal will be validated by the ERG. The following are the minimum criteria for assessment of risk of bias in RCTs, but the list is not exhaustive. The design of the FAME studies was appropriate to the decision problem in that the comparator group consisted of a sham-injection and laser photocoagulation as needed. In addition some subjects received off-protocol therapies at the discretion of the investigators. As a result the comparator group represents the optimised standard of care. Alimera Sciences, Inc. Submission of Evidence 42

44 Randomisation was managed by a computerized schedule and the study supplies were identical in appearance. The treatment groups were comparable at baseline and care providers, participants and outcome assessors were masked to treatment allocation. All assessments were performed by a masked assessing physician while treatments were administered by an unmasked treating physician. There was no indication that the masking was inadequate in that equal percentages of subjects were retreated with study drug across the groups. There were no significant imbalances in drop-outs. All assessments were reported in the clinical study reports. A classical intention to treat analysis was utilized for the primary endpoint and missing data were imputed by last observation carried forward Please provide as an appendix a complete quality assessment for each RCT. See section 9.3, appendix 3 for a suggested format. See section 9.3 for the quality assessment of the FAME studies If there is more than one RCT, tabulate a summary of the responses applied to each of the critical appraisal criteria. A suggested format for the quality assessment results is shown below. Not applicable. 5.5 Results of the relevant RCTs Provide the results for all relevant outcome measure(s) pertinent to the decision problem. Data from intention-to-treat analyses should be presented whenever possible and a definition of the included patients provided. If patients have been excluded from the analysis, the rationale for this should be given. If there is more than one RCT, tabulate the responses. From Week 3 through Month 36, both doses of FAc were statistically significantly (p 0.022) favored over sham in terms of the proportion of subjects with chronic DMO who had a 15 letter increase from baseline in BCVA (Figure B2). At Month 24, the primary end point, the proportion of Alimera Sciences, Inc. Submission of Evidence 43

45 subjects with chronic DMO who had a 15-letter increase from baseline in BCVA was 13.4%, 34.4%, and 28.4% in the sham, 0.2 μg/day FA, and 0.5 μg/day FAc groups, respectively, representing statistically significant treatment differences of 21.1% (0.2 μg/day FAc vs. sham, p<0.001) and 15.0% (0.5 μg/day FAc vs. sham, p=0.003). Peak treatment effect was observed at Month 30 for both doses of FAc (0.2 μg/day, difference of 27.1%; 0.5 μg/day, difference of 19.5%), and the effect was significantly maintained at Month 36 (0.2 μg/day, difference of 20.6%; 0.5 μg/day, difference of 15.4%). In each active treatment group, the proportion of subjects who achieved a 15-letter improvement in BCVA at Month 36 was comparable to that observed at Month 24. For the integrated subgroup with chronic DMO, the peak treatment effect (difference between sham and treated) for the 0.2 μg/day FAc dose was 27.1%; and the effect was significantly maintained at Month 36 (0.2 μg/day, difference of 20.6%). The proportion of subjects who achieved a 15-letter improvement in BCVA at Month 36 was comparable to that observed at Month 24 for the integrated subgroup with chronic DMO. Mean improvements in BCVA were consistently observed in each active treatment group through Month 36 in the integrated FAME studies among subjects with chronic DMO (Figure B3). Mean increases were numerically greater in each active treatment group compared with the sham group at all evaluations through Month 36. Statistically significant (p 0.046) treatment differences were observed in favor of the 0.2 μg/day FAc group over the sham group at each time point through Month 36 except Months 12 and 18. Differences between active therapy and sham ranged from -2.3 letters to -6.6 letters during the first 6 months, -2.7 letters to -6.3 letters between Months 9 and 18, -4.1 letters to -5.7 letters between Months 21 and 27, and -4.5 letters to -6.1 letters between Months 30 and 36. Mean improvements in excess centre point macular thickness were consistently observed in each active treatment group through Month 36 in the integrated FAME studies among subjects with chronic DMO (Figure B4). Statistically significant (p 0.043) treatment differences were observed in favor Alimera Sciences, Inc. Submission of Evidence 44

46 of 0.2 μg/day FAc over sham treatment through Month 18. The magnitude of the treatment effect diminished during the third year of the study due to improvement in the sham group and not due to loss of effect in the active treatment groups. Further, retreatments were not permitted after Month 33 and subjects who received only 1 implant may have been experiencing reduced levels of FAc as drug was depleted The information may be presented graphically to supplement text and tabulated data. If appropriate, please present graphs such as Kaplan-Meier plots. A graphical representation of the primary variable of the FAME studies is provided in Figure B2. The FAc implant treatment group demonstrated statistically significant differences from the sham treatment group at all time points. Figure B2: Percent (ASE) of Subjects with Chronic DMO with 15 Letter Increase from Baseline BCVA The mean change in BCVA was greater at all time points throughout the study for subjects with chronic DMO. (Figure B3) Alimera Sciences, Inc. Submission of Evidence 45

47 Figure B3: Mean (SEM) Change from Baseline BCVA in Subjects with Chronic DMO Likewise the change in excess retinal thickness was greater throughout the study for subjects with chronic DMO. (Figure B4) Figure B4: Mean (SEM) Change from Baseline in Excess Centre Point Thickness in Subjects with chronic DMO Alimera Sciences, Inc. Submission of Evidence 46

48 5.5.3 For each outcome for each included RCT, the following information should be provided. The unit of measurement. The size of the effect; for dichotomous outcomes, the results ideally should be expressed as both relative risks (or odds ratios) and risk (or rate) differences. For time-to-event analysis, the hazard ratio is an equivalent statistic. Both absolute and relative data should be presented. A 95% confidence interval. Number of participants in each group included in each analysis and whether the analysis was by intention to treat. State the results in absolute numbers when feasible. When interim RCT data are quoted, this should be clearly stated, along with the point at which data were taken and the time remaining until completion of that RCT. Analytical adjustments should be described to cater for the interim nature of the data. Other relevant data that may assist in interpretation of the results may be included, such as adherence to medication and/or study protocol. Discuss and justify definitions of any clinically important differences. Report any other analyses performed, including subgroup analysis and adjusted analyses, indicating those pre-specified and those exploratory. The results of the FAME studies demonstrate an important therapeutic effect compared to optimised standard of care. At Month 36, the proportion of subjects with a 15-letter increase in BCVA was 18.9% and 28.7% in the sham, 0.2 μg/day FA, respectively, representing a treatment difference of 9.8% (0.2 μg/day FAc vs. sham, p=0.018). The percentage of subjects with 15 letter improvement in BCVA is a standard endpoint in DMO studies. An improvement from baseline of 15 letters is clearly a clinically relevant Alimera Sciences, Inc. Submission of Evidence 47

49 outcome, and the statistical significance of the change proves the FAc implant s efficacy. In the integrated FAME studies (FAME A + FAME B), median change from baseline in BCVA during the third year ranged from 7 to 8 letters in the active treatment groups and 2 to 3 letters in the sham group. A median difference of one line compared to the sham treated group is relevant because the sham group is not a completely untreated control, but represents the optimised standard of care. The increase of 1 ETDRS line was achieved with fewer laser and intravitreal steroid treatments in the FAc-treated groups than in the sham group. The mean change from baseline in excess centre point thickness was reduced consistently throughout the study in subjects with duration of chronic DMO despite the fact that 85% of the subjects in the FAc implant group had cataract surgery which can exacerbate macular oedema. This demonstrates the persistent reduction in retinal oedema with the FAc implant. The final secondary outcome was the percentage of subjects with 3 step worsening on the ETDRS Multi-step Eye Scale for Diabetic Retinopathy. There were very small in all groups with this degree of worsening during the study and no difference was noted; however, there was a small but statistically significant difference in favor of the FAc implant with respect to 2 step improvement on the ETDRS Multi-step Eye Scale. The subjects with chronic DMO treated with the FAc implant required less laser photocoagulation than sham-treated subjects. In the integrated analysis, the percentage of subjects with chronic DMO who received at least 1 laser treatment for any reason in the study eye was 61% and 41% in the sham and 0.2 μg/day FAc, respectively. Of these, the majority received only 1 or 2 laser treatments. The mean number of laser treatments received was higher in the sham group (1.4) compared with the active treatment groups (0.8 and 0.2 μg/day FAc). The differences between each active treatment group and the sham group were statistically significant (p 0.003). Alimera Sciences, Inc. Submission of Evidence 48

50 The percentage of subjects with chronic DMO who received at least 1 offprotocol treatment for DMO in the study eye was significantly (p<0.001) higher in the sham group (35%) compared with the 0.2 μg/day FAc (13%) and 0.5 μg/day FAc (16%) groups. Of these, most received 1 or 2 off-protocol treatments. The mean number of off-protocol treatments for DMO was also significantly (p<0.001) higher in the sham group (1.0) compared with the active treatment groups (0.2, 0.2 μg/day FA; 0.3, 0.5 μg/day FA). The majority of subjects with chronic DMO in the FAME studies received only 1 FAc implant or sham injection as study treatments. Table B10: Summary of Relevant Outcomes for Subjects with Duration of Chronic DMO in FAME A+B Integrated FAME Studies Sham 0.2 μg/day FAc 0.5 μg/day FAc % with 15 Letter Improvement in BCVA N=112 N=209 N=215 Month 36, n (%) 15 (13.4) 71 (34.0) 62 (28.8) Difference % C.I. (-29.6, -11.6) (-24.2, -6.7) P-value < Mean Change in BCVA Month 36, n (%) N=112 N=209 N=215 Observed n (SD) 55.9 (17.93) 59.8 (19.18) 58.5 (18.96) Change n (SD) 1.8 (18.59) 7.6 (17.40) 6.2 (16.27) Difference % C.I. (-10.2, -2.0) (-8.8, -0.6) P-value Mean Change Excess Centre Point Thickness Month 36, n (%) N=111 N=207 N=210 Observed n (SD) (169.53) (147.59) (176.54) Change n (SD) (193.66) (188.91) (208.62) Difference % C.I. (-8.9, 88.7) (-5.7, 92.0) P-value Alimera Sciences, Inc. Submission of Evidence 49

51 Integrated FAME Studies Sham 0.2 μg/day FAc 0.5 μg/day FAc Any Laser Treatments Month 36, N=112 N=209 N=215 No. of Treatments % of Subjects with 1 Laser Treatment Mean Treatments/ Subject 68 (60.7%) 85 (40.7%) 75 ( 34.9%) P-value <0.001 Study Drug Treatments Month 36, N=112 N=209 N=215 No. of Treatments % of Subjects with 1 Study Treatment Mean Treatments/ Subject 74 (66.1%) 159 (76.1%) 148 (68.8%) Alimera Sciences, Inc. Submission of Evidence 50

52 5.6 Meta-analysis Not applicable. 5.7 Indirect and mixed treatment comparisons Describe the strategies used to retrieve relevant clinical data on the comparators and common references both from the published literature and from unpublished data. The methods used should be justified with reference to the decision problem. Sufficient detail should be provided to enable the methods to be reproduced, and the rationale for any inclusion and exclusion criteria used should be provided. Exact details of the search strategy used should be provided in section 9.4, appendix 4. The strategies employed to retrieve clinical data for indirect comparison that are relevant to the decision problem take two key considerations into account: (1) the definition of the comparator arm, and (2) the target population in which the comparator was studied. The FAc implant is indicated for patients with chronic DMO considered insufficiently responsive to available therapies. Thus, the current comparator based on U.K. practice would be observation; however, data does not exist which would allow us to make an economic comparison. Well controlled, randomised, double masked, multicentre studies in subjects with DMO which included a laser treatment arm with treatment duration of at least 2 years were considered Please follow the instructions specified in sections 5.1 to 5.5 for the identification, selection and methodology of the trials, quality assessment and the presentation of results. Provide in section 9.5, appendix 5, a complete quality assessment for each comparator RCT identified. Alimera Sciences, Inc. Submission of Evidence 51

53 An indirect comparison to a laser photocoagulation monotherapy comparator is considered the closest comparator to UK clinical practice. To identify relevant studies a PubMed literature search was performed using the terms diabetic macular oedema, randomized clinical trials, laser photocoagulation and 2 years. Figure B5 summarizes the outcomes of the search. (see section 9.4 for details.) Figure B5: Flow chart for Indirect Comparisons Studies were only included if the % of subjects with 15 letter improvement from BCVA was reported in the intent to treat dataset. This resulted in the identification of 2 studies (Gillies et al and DRCR 2008). The report by Gillies et al. was a small study (84 eyes of 54 patients). The results may be biased because both eyes of some patients were included in the study while for others only one eye was randomized. Further, the study was not reported to have been performed in accordance with Good Clinical Practices. Alimera Sciences, Inc. Submission of Evidence 52

54 Therefore, we have elected to use the DRCR Protocol B (2008) dataset which is currently available from a published, peer-reviewed source for laser treatment. The laser arm of this study assesses the effect of laser in patients with centre involved DMO. This approach is considered to be a very reasonable and conservative approach as the DRCR population included almost 40% of laser naive subjects and there was no stipulation on duration of DMO at randomization. Thus, the overall severity of DMO in this population is not considered to be as severe as that studied with ILUVIEN. (It should be noted that the 3 year data for DRCR Protocol B has been published; however, the data are based on subjects completing 3 years of follow-up. As a result the analysis is not appropriate for use in the model.) Provide a summary of the trials used to conduct the indirect comparison. A suggested format is presented below. Network diagrams may be an additional valuable form of presentation. The laser photocoagulation treatment arm of the DRCR Protocol B Study is the most relevant and credible laser monotherapy comparator data available. It is available as a published, peer-reviewed article of a phase 3, multicentre clinical trial. (DRCR 2008) It should be noted that the results of a comparison to a general group of subjects with DMO will provide a more conservative assessment of the cost-effectiveness of the FAc implant compared to the standard of care. Table B11: Summary of the trials used to conduct the indirect comparison No. trials References of trials Intervention Comparator DRCR Protocol B (DRCR 2008) Sham + Prompt Laser FAME Study (FAME Study ISE) FAc implant Alimera Sciences, Inc. Submission of Evidence 53

55 5.7.4 For the selected trials, provide a summary of the data used in the analysis. The percentage of subjects with 15 letters increase from baseline BCVA is available for the studies; however, only two years of data for the ITT population have been published for the DRCR data set. In addition, the mean change from baseline BCVA was reported for both studies Please provide a clear description of the indirect/mixed treatment comparison methodology. Supply any programming language in a separate appendix. As described in the review of available comparators in section 5.7.2, there is a paucity of information in the target patient group. The only appropriate source found was the DRCR Protocol B publication (DRCR 2008). In the absence of other suitable comparators there was no requirement for blending data from multiple sources by mixed treatment methodology. The primary variable for the FAME studies (% subjects with 15 letter improvement from baseline BCVA) was used as the primary comparison and therefore the corresponding data from the DRCR publication was used directly in the model Please present the results of the analysis. Table B12: Summary of Outcomes for Subjects with Chronic DMO in FAME A+B Compared to DRCR Protocol B DRCR Protocol B Month 24 1 LASER PHOTOCOAGULATION % with 15 Letter Improvement in BCVA FAME Studies Month 36 FAc implant n (%) N=115 N=209 18% 34.0% Mean Change in BCVA n (%) N=115 N=209 Observed n (SD) N.A (19.18) Change n (SD) 1 (17) 7.6 (17.40) Alimera Sciences, Inc. Submission of Evidence 54

56 DRCR Protocol B Month 24 1 LASER PHOTOCOAGULATION FAME Studies Month 36 FAc implant Mean Number of Laser Treatments N=272 (completers) N=209 Mean (SD) 2.9 (1.4) A straight line extrapolation was used to obtain Month 36 data for the DRCR data Please provide the statistical assessment of heterogeneity undertaken. The degree of, and the reasons for, heterogeneity should be explored as fully as possible. No statistical assessment of heterogeneity was undertaken because the data set for the DRCR Protocol B study was not published. Numerical differences clearly favor ILUVIEN for percentage of subjects with 15 letter improvement from baseline BCVA and mean change in BCVA. In addition, on average, three times as many laser treatment were administered in the DRCR study (where laser treatment was required if criteria were met) compared to the FAME studies If there is doubt about the relevance of a particular trial, please present separate sensitivity analyses in which these trials are excluded. Not applicable Please discuss any heterogeneity between results of pairwise comparisons and inconsistencies between the direct and indirect evidence on the technologies. In both the indirect and the direct comparisons, the FAc implant provided a greater benefit than the comparator in terms of % of subjects with 15 letter improvement from baseline BCVA; however, the laser treatment group of the DRCR Protocol B study experienced a slightly better outcome, i.e. 18% Alimera Sciences, Inc. Submission of Evidence 55

57 responders versus 13.4% responders in the FAME studies. The most likely explanation for this is that Protocol B included subjects with and without prior laser at baseline. Thus, some of the subjects in Protocol B who may have experienced an improvement due to laser would not be included in the target population for the FAc implant. In addition, the subjects in the laser treatment group of Protocol B received more laser treatment during the 2-year study than subjects in the FAME study sham group received in 3 years. This is likely because some subjects in the FAME studies had already received maximal laser prior to entering the study. The mean change in BCVA was similar in Protocol B and the sham-treatment group of the FAME studies. The direct comparison using the FAME studies sham-treatment (optimised standard of care) subjects with chronic DMO is considered the most accurate comparison because it includes the target population and optimised standard of care. 5.8 Non-RCT evidence If non-rct evidence is considered (see section 5.2.7), please repeat the instructions specified in sections 5.1 to 5.5 for the identification, selection and methodology of the trials, and the presentation of results. For the quality assessments of non-rcts, use an appropriate and validated quality assessment instrument. Key aspects of quality to be considered can be found in Systematic reviews: CRD s guidance for undertaking reviews in health care ( Exact details of the search strategy used and a complete quality assessment for each trial should be provided in sections 9.6 and 9.7, appendices 6 and 7. No non-rct evidence was considered. 5.9 Adverse events If any of the main trials are designed primarily to assess safety outcomes (for example, they are powered to detect significant differences between treatments with respect to the incidence of an adverse event), please repeat the instructions specified in Alimera Sciences, Inc. Submission of Evidence 56

58 sections 5.1 to 5.5 for the identification, selection, methodology and quality of the trials, and the presentation of results. Examples for search strategies for specific adverse effects and/or generic adverse-effect terms and key aspects of quality criteria for adverseeffects data can found in Systematic reviews: CRD s guidance for undertaking reviews in health care ( Exact details of the search strategy used and a complete quality assessment for each trial should be provided in sections 9.8 and 9.9, appendices 8 and 9. No studies were designed primarily to assess safety outcomes or were powered to detect significant differences between treatments with regard to an adverse event. The phase 3 FAME studies provided the safety data base for MAA approval and provide safety data on 376 subjects exposed to the marketed product. Consistent with ICH guidance, this sample size allows for the detection of any adverse event with a true incidence of at least 1%, with 95 % confidence Please provide details of all important adverse events for each intervention group. For each group, give the number with the adverse event, the number in the group and the percentage with the event. Then present the relative risk and risk difference and associated 95% confidence intervals for each adverse event. A suggested format is shown below. Table B13 summarizes the adverse events of importance reported in subjects with chronic DMO during the studies. Retinal detachment is a known risk of intravitreal injections; however there were no cases of retinal detachment in the target population in the FAME studies. Alimera Sciences, Inc. Submission of Evidence 57

59 Table B13: Ocular Adverse events across randomised groups Ocular adverse events % of patients (n) Day 1 Through Month 36 Subjects with Chronic DMO FAc implant Sham Relative risk (95% CI) Cataract-Related Events (PHAKIC SUBJECTS) N=121 N=66 Cataract Surgery 85.1% (97) 36.4% (24) 2.34 (1.69, 3.25) Other Adverse Events (All Subjects) IOP Increased/Ocular Hypertension AE FAc implant (N = 209) Sham (N = 112) 34.4% (72) 14.3% (16) 2.41 (1.48, 3.94) IOP-Medication 35.9% (75) 15.2% (17) 2.36 (1.47, 3.80) IOP-Lowering Surgery Vitrectomy (for any reason) 5.3% (11) 0% (0) % (7) 7.1% (8) 0.47 (0.17, 1.26) Endophthalmitis 1.0 (2) 0% -- Retinal Detachment 0% 0% -- An additional class of adverse event which may be of interest is cerebrovascular events. The incidence of cerebrovascular events was the same for the sham and the FAc implant treatment groups and none of these events were considered related to the FAc implant treatment. The incidences are provided in Table B14 for completeness sake only, and are not accounted for in the cost-effectiveness model. Table B14: Incidences of Cerebrovascular Adverse Events in the Subjects with Chronic DMO in the FAME Studies Cerebrovascular Events FAc implant N= 209 Sham N= 112 Brain stem infarction 1 (0.5) 1 (0.9) Cerebral haemorrhage 0 0 Cerebral infarction 1 (0.3) 0 Cerebral ischaemia 0 1 (0.9) Cerebrovascular accident 5 (2.4) 4 (3.6) Total 7 (3.3) 6 (5.4) Alimera Sciences, Inc. Submission of Evidence 58

60 5.9.3 Give a brief overview of the safety of the technology in relation to the decision problem. Treatment of DMO with the FAc implant is associated with two significant adverse events which impact on the decision problem: formation/progression of cataract and increased IOP. In the FAME studies, 58% (121/209) of the subjects with chronic DMO were phakic at baseline. Cataract surgery was required by 85.1% of phakic subjects treated with the FAc implant; however, it should be noted that cataract surgery is common in diabetic patients. In the subjects with chronic DMO in the FAME studies, 36.4% of phakic subjects in the sham-treatment group required cataract surgery, and 51.5% developed a cataract during the study. Further, cataract surgery is a very common procedure with a low rate of complications. The risk of IOP elevations has been clearly defined by the FAME studies. The mean IOP of the FAc implant-treated group was increased by less than 3 mmhg during the first year of the study and decreased in the second and third years. Over the life of the study, 35.9% of the FAc implant-treated subjects with chronic DMO required IOP-lowering medication for at least 7 days compared to 15.2 % in the sham-treated subjects. Most of these subjects responded to IOP-lowering medications while surgery was required in 5% of subjects overall Interpretation of clinical evidence Please provide a statement of principal findings from the clinical evidence highlighting the clinical benefit and harms from the technology. Summary of Benefits of the FAc implant This section will summarize the benefits of the FAc implant for the target population, subjects with chronic DMO. The integrated FAME studies results for this subgroup will be the primary focus. Alimera Sciences, Inc. Submission of Evidence 59

61 Three year efficacy of the FAc implant is greater than the standard of care in patients considered insufficiently responsive to available therapies The results of the primary efficacy endpoint, percent of subjects with 15 letter improvement from baseline BCVA at Month 24 demonstrated the efficacy of both doses of the FAc implants compared to sham treatment. A preplanned subgroup analysis revealed that the majority of the efficacy observed in the Full Analysis Populations was due to the subjects with duration of DMO 3 years (the median for each of the FAME Studies). In both FAME A and FAME B the efficacy outcomes of the FAc implant groups were statistically significantly better than sham at nearly every time point throughout the 3 year study. The difference between the sham and the FAc implant treatment groups was dramatically and statistically greater for subjects with longer duration DMO (p<0.001 for each FAc group vs. sham). The target population has been defined in the labeling as patients with chronic DMO considered insufficiently responsive to available therapies. For clarity, subsequent sections will summarize the results using the integrated FAME studies Full Analysis Population and the integrated subgroup with DMO 3 years. Consistently reduced retinal thickness over 36 months An important advantage of the FAc implant is that the anatomical improvements occur rapidly and are consistently maintained for three years. The excess retinal thickness was reduced in subjects with chronic DMO (Figure B6). The mean change from baseline in excess centre point thickness was reduced consistently throughout the study in subjects with chronic DMO despite the fact that 85% of phakic subjects (49% of all subjects with chronic DMO) in the FAc implant group had cataract surgery which may exacerbate macular oedema. Alimera Sciences, Inc. Submission of Evidence 60

62 Figure B6: Mean Change from Baseline Excess Centre Point Thickness over Time (Integrated Chronic DMO Population) This is supported by the finding that the area of fluorescein leakage was reduced to a greater extent in subjects with chronic DMO treated with the FAc implant compared to those in the sham group throughout the study. This may also relate to the improvement in the anatomical signs of DR observed in the target population. A greater proportion of subjects with chronic DMO in the FAc implant -treated group had 2 step improvement in ETDRS Multi-step Eye Scale of DR (sham, 8%; 0.2 μg/day FA, 17%). The consistent improvement in retinal thickness, reduction in area of leakage and improvement in ETDRS classification support the concept that sustained treatment with FAc is more effective in addressing the anatomical changes in the retina than the sham treatment. Fewer laser treatments, fewer anti-vegf treatments, fewer vitrectomies In the integrated analysis, the percentage of subjects with chronic DMO who received at least 1 retreatment of study drug was 34%, 24% in the sham, 0.2 μg/day FA, respectively. The FAc implant-treated subjects with chronic DMO required less laser photocoagulation than the sham-treated subjects. In the integrated analysis, Alimera Sciences, Inc. Submission of Evidence 61

63 the percentage of subjects with chronic DMO years who received at least 1 laser treatment for any reason in the study eye was 61% and 41% in the sham and 0.2 μg/day FAc groups, respectively. Of these, the majority received only 1 or 2 laser treatments. The mean number of laser treatments received was higher in the sham group (1.4) compared with the active treatment groups (0.8, 0.2 μg/day FAc). The differences between each active treatment group and the sham group were statistically significant (p 0.003). In the integrated analysis, the percentage of subjects with chronic DMO who received at least 1 off-protocol treatment for DMO in the study eye was significantly (p<0.001) higher in the sham group (35%) compared with the 0.2 μg/day FAc (13%) and 0.5 μg/day FAc (16%) groups. Of these, most received 1 or 2 off-protocol treatments. The mean number of off-protocol treatments for DMO was also significantly (p<0.001) higher in the sham group (1.0) compared with the active treatment groups (0.2, 0.2 μg/day FA; 0.3, 0.5 μg/day FA). Advantages of Sustained Release Dosage Form The FAc implant provides important advantages over laser photocoagulation, IVTA and anti-vegf antibodies. First, the sustained delivery of FAc from the implant provides the lowest daily exposure to corticosteroid which provides efficacy while minimizing the risks of multiple intravitreal injections. Secondly, the consistent treatment of DMO provided by the FAc implant will reduce the burden of vision loss and frequent office visits for the patient and their family. (Since a large portion of the population with DMO is still of working age, this may be especially important.) In the FAME studies, after the first quarter, follow-up visits occurred quarterly. This level of monitoring was sufficient to control IOP changes and is in keeping with normal clinical practice for the subjects with DMO. Summary of Risks of the FAc implant This section will summarize the risks of the FAc implant for subjects with chronic DMO. Alimera Sciences, Inc. Submission of Evidence 62

64 Risk of cataract is clearly defined The risk of formation or progression of cataract is a well known side effect of glucocorticoids. The results of the FAME studies demonstrated that 85% of phakic subjects treated with the FAc implant for DMO experienced cataract and require cataract surgery. In view of the incidence of cataract in the phakic subjects in the sham group, 50% of whom experienced cataract surgery during the 3 year FAME Study, the use of the FAc implant accelerates cataract and the need for cataract surgery; however, as seen with the sham treatment group in the FAME study, diabetes is associated with an increased risk for cataracts. The risk of earlier cataract surgery is more than offset by the benefit of immediately improved vision. This is supported by the observation that the visual outcomes in the third year of the study in subjects who were pseudophakic at baseline and phakic subjects who became pseudophakic were similar in the FAc implant group. Thus, the long term benefit of the FAc implant in subjects who require cataract surgery is essentially the same as those who do not require surgery. Cataract surgery is a very common procedure with a low rate of complications. Based on the information that cataracts are very common with the FAc implant, it is likely that clinicians will remove cataract as early as feasible to reduce the impact on the patient s vision. Risk of IOP increase is clearly defined The risk of IOP elevations has been clearly defined by the FAME studies. The mean IOP of the FAc implant-treated group was increased by less than 3 mmhg during the first year of the study and decreased in the second and third years. Over the life of the study, 38% of FAc implant-treated subjects required IOP-lowering medication for at least 7 days compared to 14% in sham-treated subjects. Most of these subjects responded to IOP-lowering medications while surgery was required in 5% of subjects overall. The most concerning aspect of elevated IOP is the increased risk of glaucomatous optic neuropathy. The investigator s and reading centre s Alimera Sciences, Inc. Submission of Evidence 63

65 assessments did not indicate a clinically relevant increase in the mean cup/disc measurement in the overall population or in the subjects with IOP elevations versus control subjects treated with the FAc implant. These findings indicate that most subjects who experience elevated IOP elevations have been adequately treated and anatomical changes indicative of glaucoma did not occur. The incidence of glaucoma/open angle glaucoma based on adverse event reports was 6% in both treatment groups compared to 2% in the sham group. Per the current definition of glaucoma, the diagnosis of glaucoma was almost certainly incorrect in some cases because neither anatomical changes nor visual field changes were reported for most subjects. Based on the IOP of the subject whose FAc implant was removed and the overall trend for mean IOP, the ocular hypertension caused by the FAc implant is reversible. It can be expected that IOP will return to near-baseline levels when FAc levels are decreased. Risk of steroid class events Glucocorticoids can cause increased risk of infections and delay wound healing; however, notable increases of these events in the FAc-treated groups compared to the sham group were not observed in the FAME studies. The rate of endophthalmitis at any time in the study eye of all subjects who received FAc implants in the FAME studies was 0.5% (4/768 subjects). This includes a case which occurred after cataract surgery and was not considered related by the investigator. On a per injection basis, the rate of study drugrelated endophthalmitis is 0.2% (2/977 injections). It should be noted that the majority of the subjects in this treatment group experienced cataract surgery and intravitreal administrations of off-protocol therapies, which on their own, have the potential to increase the risk of endophthalmitis. Thus, the rate of 0.2% is considered acceptable. Delayed ocular wound healing can be related to the administration of glucocorticoids. In the FAME studies, there was no indication of a notable Alimera Sciences, Inc. Submission of Evidence 64

66 increase in the incidence of ocular adverse events related to wound healing or cerebrovascular events Please provide a summary of the strengths and limitations of the clinical-evidence base of the intervention. The phase 3 studies were conducted at 101 sites in North America, Europe, and India. All sites in the FAME studies used the same protocol and performed the studies in accordance with current Good Clinical Practices and local requirements. The studies were randomised and double-masked to eliminate bias. To preserve masking, 2 investigators were used. One investigator performed the treatments and the other performed all assessments. A sham injection control was selected because injection of a placebo implant, while viewed as the best control, was not ethically acceptable to the investigators and ethical committees. This resulted in a potential for unmasking since the intravitreal implant may be visible to the assessing physician during an ophthalmic examination and may be perceived by the subject as a floater. The impact of this potential unmasking was viewed as small because the primary variable (15-letter improvement in BCVA) is a robust endpoint and was assessed by masked, certified assessors and not by either of the investigators. In addition, several anatomical assessments were included to provide corroboration for the BCVA, including retinal thickness measurements. An important consideration in evaluating the design of the phase 3 studies is that subjects (including those who received active drug) were allowed to have additional laser treatments at the discretion of the investigator. This was viewed as necessary because the duration of the studies (3 years) was too long to deny treatment to any subject who did not respond to treatment. Subjects were not to receive any non-approved treatments in the study eye for DMO or systemic treatments for DMO. However, 35% of sham-treated subjects and 13% of FAc implant-treated subjects did receive these offprotocol therapies. The differential use of laser treatments and off-protocol Alimera Sciences, Inc. Submission of Evidence 65

67 drug therapies may have confounded differences between the control and treated groups, favoring the control group. A final important issue with the design is that subjects were eligible for retreatment with the masked study medication to which they were randomised any time after the Month 12 assessments if they experienced vision loss or retinal thickening per optical coherence tomography (OCT). As a result, subjects received various numbers of treatments (1 4) during the studies. The protocol was designed to allow flexible timing of retreatment because the duration of therapeutic effect was not known prior to the start of the study Please provide a brief statement of the relevance of the evidence base to the decision problem. Include a discussion of the relevance of the outcomes assessed in clinical trials to the clinical benefits experienced by patients in practice. The primary endpoint (percentage of subjects with 15-letter improvement in BCVA at Month 24) was chosen based on input from regulatory authorities because it represents an indisputably clinically relevant change in vision. The Month 36 follow-up confirmed the durability of the drug effectiveness and safety. The safety database, which was developed in accordance with standard clinical trial methods, is sufficiently large to provide a clear safety profile. In summary, the FAME studies provided a reliable clinical evidence basis for assessment of the decision problem. Alimera Sciences, Inc. Submission of Evidence 66

68 Identify any factors that may influence the external validity of study results to patients in routine clinical practice; for example, how the technology was used in the trial, issues relating to the conduct of the trial compared with clinical practice, or the choice of eligible patients. State any criteria that would be used in clinical practice to select patients for whom treatment would be suitable based on the evidence submitted. What proportion of the evidence base is for the dose(s) given in the SPC? The FAME studies permitted the use of laser photocoagulation after Week 6. In addition some subjects received off-protocol therapies at the discretion of the investigator. In clinical practice, it is anticipated that the FAc implant will be used in patients who are considered unresponsive to other therapies so there would rarely be a rationale for treatments other than the FAc implant. Further, it is anticipated that patients will receive 1 implant per 3 years. When the FAME trial was initiated, the duration of in vivo drug release was not known. It was estimated to be 2 3 years, based on in vitro kinetics, but was not actually known. During the phase 3 study, a phase II trial was initiated which included sampling of intraocular fluid from the human subjects participating in the trial, over 36 months. Based on these results, it was seen that drug released from the FAc implant could be measured over the entire 36 month period. In addition to this, in the phase 3 trials an effect on the anatomy of the retina was consistently seen for the entire 36 month period, and finally, a significant improvement in visual acuity was ultimately demonstrated. Thus, it was not possible to inform investigators of the in vivo release of the FAc implant while they were participating in this double-masked trial. This lack of understanding was the only reason that re-treatment after 12 months was allowed in the study. Compounding this lack of knowledge regarding in vivo duration of release was a lack of understanding of the timing or magnitude of the effect of cataract in the FAc implant treated population. For some patients, there was initial vision loss as the cataract was beginning to form, but, before it could be clinically Alimera Sciences, Inc. Submission of Evidence 67

69 detected. A temporal relationship between administration of off-protocol therapies and cataract formation was seen for several patients. In clinical practice, the in vivo duration of release and the expectation of cataract development in phakic subjects will be understood, thus, in clinical practice more than one FAc implant is unlikely to be administered per 36 month period. 6 Cost effectiveness 6.1 Published cost-effectiveness evaluations Identification of studies Describe the strategies used to retrieve relevant cost-effectiveness studies from the published literature and from unpublished data held by the manufacturer or sponsor. The methods used should be justified with reference to the decision problem. Sufficient detail should be provided to enable the methods to be reproduced, and the rationale for any inclusion and exclusion criteria used should be provided. The search strategy used should be provided as in section 9.10, appendix 10. A search for cost-effectiveness studies in diabetic macular oedema was performed using Medline, Medline (R) In-Process, Embase, EconLIT, NHS- EED, DARE, and HTA databases. Details are provided in Appendix A flow diagram of the search for cost-effectiveness studies related to the decision problem is provided in Figure B7. Alimera Sciences, Inc. Submission of Evidence 68

70 Figure B7: Flow diagram for search for cost-effectiveness studies Of the four articles retrieved for full-text review, two were not full reports of cost-effectiveness studies (one was a review article, one was an abstract), and the two full study reports were based on studies of patients with early DMO (i.e., not patients refractory to other treatments). Description of identified studies Provide a brief overview of each study, stating the aims, methods, results and relevance to decision-making in England and Wales. Each study s results should be interpreted in light of a critical appraisal of its methodology. When studies have been identified and not included, justification for this should be provided. If more than one study is identified, please present in a table as suggested below. No relevant cost-effectiveness studies were identified. Alimera Sciences, Inc. Submission of Evidence 69

71 6.1.3 Please provide a complete quality assessment for each costeffectiveness study identified. Use an appropriate and validated instrument, such as those of Drummond and Jefferson (1996) 1 or Philips et al. (2004) 2. For a suggested format based on Drummond and Jefferson (1996), please see section 9.11, appendix 11. No relevant cost-effectiveness studies were identified. 6.2 De novo analysis Patients What patient group(s) is(are) included in the economic evaluation? Do they reflect the licensed indication/ce marking or the population from the trials in sections 1.4 and 5.3.3, respectively? If not, how and why are there differences? What are the implications of this for the relevance of the evidence base to the specification of the decision problem? For example, the population in the economic model is more restrictive than that described in the (draft) SPC/IFU and included in the trials. The economic evaluation includes a cohort of patients with chronic DMO, based on those enrolled in the FAME trial. It also includes a cohort of DMO patients with more recent disease, based on those enrolled in the DRCR Protocol B trial. In the direct comparison, patient groups analysed in both the intervention and comparator arms reflects the licensed indication for the FAc implant detailed in sections 1.4 and 5.3.3, i.e. patients with chronic DMO who have received at least one prior treatment with laser photocoagulation. In the indirect comparison, the patient group analysed in the intervention arm is the same as the direct comparison; however, the patient group analysed in the comparator arm (from the DRCR Protocol B trial) is somewhat different 1 Drummond MF, Jefferson TO (1996) Guidelines for authors and peer reviewers of economic submissions to the BMJ. The BMJ Economic Evaluation Working Party. British Medical Journal 313 (7052): Philips Z, Ginnelly L, Sculpher M, et al. (2004) Quality assessment in decision-analytic models: a suggested checklist (Appendix 3). In: Review of guidelines for good practice in decision-analytic modelling in health technology assessment. Health Technology Assessment 8: 36. Alimera Sciences, Inc. Submission of Evidence 70

72 than the group for which the FAc implant is indicated (Table B15, see section 5.7.2). Table B15: Baseline characteristics of FAc implant treated FAME patients and laser treated patients from the DRCR Protocol B trial Baseline characteristics FAc implant treated chronic DMO population (FAME) Laser monotherapy treated population (DRCR Protocol B) BCVA score (ETDRS) Between 68 and 19 Between 73 and 24 Location of disease Prior photocoagulation for DMO Duration of diabetes (years) Centre involved DMO 100% 60% Mean: 17.9 (SD: 8.3) Definite retinal thickening resulting from DMO involving the centre of the macula assessed to be the main cause of vision loss 15 (median); (25 th percentile: 9; 75 th percentile: 21) Duration of DMO Mean: 5.1 (SD: 3.12) Not reported Centre point thickness on OCT (µm) Mean: (SD: ) Age (years) Mean: 63.7 (SD: 8.9) 398 (median) (25 th percentile: 329; 75 th percentile: 505) 63 (median) (25 th percentile: 58; 75 th percentile: 69) Key differences between the FAc implant treated population in FAME and the laser treated population in the DRCR trial include: All FAc implant treated patients received prior laser photocoagulation compared to 60% of laser treated patients in DRCR Protocol B Duration of diabetes was longer in the FAc implant treated group compared to the laser treated group in DRCR Protocol B Duration of DMO is associated with duration of diabetes, and while this metric was not reported in DRCR Protocol B, one may presume that the duration of disease was shorter for laser treated patients in DRCR Protocol B as compared to FAc implant treated patients Centre point thickness at baseline was greater in the FAc implant treated group compared to the laser treated group in DRCR Protocol B Alimera Sciences, Inc. Submission of Evidence 71

73 There are no RCTs assessing the use of laser photocoagulation as monotherapy in a population directly comparable to the chronic DMO population in the FAME trials (section 5.7). The DRCR Protocol B trial was selected based on the conclusion that the population was most similar to the population in which the FAc implant was studied, namely that the population had center involved DMO. However, the differences in the population illustrate that the DRCR Protocol B patient group represents an earlier stage of DMO than the chronic DMO patients in the FAc implant group. As a result, responses to treatment between the two groups may be different as a result of differences in disease progression. The implication of these differences for the economic evaluation is that in the indirect comparison, patients treated with laser monotherapy in the DRCR Protocol B trial may respond better to laser photocoagulation than would a group of chronic DMO patients who had all shown an insufficient response to prior laser photocoagulation. Thus, the indirect comparison between the FAc implant treated patient group from the FAME trials and the laser monotherapy treated group from the DRCR Protocol B trial likely reflects a more conservative estimate of differences in outcomes between the therapies than if we were to do a true comparison to laser monotherapy in the exact same group. Model structure Please provide a diagrammatical representation of the model you have chosen. Figure B8 illustrates the structure of the Markov model with outcomes represented by ETDRS 5-letter-score changes in BCVA. After each cycle (3 months), patients VA may improve, worsen or stay the same. Alimera Sciences, Inc. Submission of Evidence 72

74 Figure B8: Economic model structure Please justify the chosen structure in line with the clinical pathway of care identified in section 2.4. The structure of the model is consistent with the current management of the disease for patients who are chronic and insufficiently responsive to available therapies. In the first three years of treatment patients in the direct comparison are treated as per the FAME study protocol and results. The only non-trial based cost incurred is, when appropriate, the provision of blindness support when the patient falls below a visual acuity level of 35 letters or less. After the first three year cycle in the absence of the FAc implant whether by initial treatment strategy decision or by withdrawal from the active treatment arm, the standard of care will be regular monitoring of the patient s condition and again, when appropriate, the provision of blindness support when the Alimera Sciences, Inc. Submission of Evidence 73

75 patient falls below a visual acuity level of 35 letters or less. The indirect comparison arm of the study illustrates the impact if it is decided to use laser therapy as an intervention of last resort Please define what the health states in the model are meant to capture. There are 14 health states in the model. Of these, 13 are defined by patients visual acuity (VA) on the basis of 5 letter increments on the ETDRS scale; a distinguishable change in health states is therefore a change of 5 letters in the total number letters that either can or cannot be read on an eye chart. Utility values associated with the 13 VA-related health states therefore capture the impact of varying degrees of visual gain or loss on patients quality of life. The 14 th health state is death, and the utility associated with this is zero. (see section and 6.4.9) How does the model structure capture the main aspects of the condition for patients and clinicians as identified in section 2 (Context)? What was the underlying disease progression implemented in the model? Or what treatment was assumed to reflect underlying disease progression? Please cross-reference to section 2.1. As described in section 2.1 the presence of the condition is characterized by the status of the macula and the progress of the disease takes place as this degenerates due to oedema. However, the patient status and the progress of DMO are manifested to both the patient and the treating clinician by a progressive decline in visual acuity caused by the damage to the retina. The health states of the model are directly built upon the accepted standard for categorization of visual functioning (i.e. change in 5 letter scales) and therefore reflect directly any positive or indeed negative change to the patients disease condition. Alimera Sciences, Inc. Submission of Evidence 74

76 6.2.6 Please provide a table containing the following information and any Table B16: Factor additional features of the model not previously reported. A suggested format is presented below. Key features of analysis Chosen values Justification Time horizon 15 years Patients with diabetes and in particular DMO have a higher relative risk of death than the general population and given the mean age of the FAME study patients (approx. 62) this was felt appropriate. A prior NICE STA in DMO also used 15 years. Cycle length 3 months A health technology appraisal report for age-related macular degeneration (AMD) treatments published by NICE suggested that a 3-month interval is appropriate for assessing visual acuity changes in AMD. This was repeated in the NICE STA of ranibizumab. Half-cycle correction Were health effects measured in QALYs; if not, what was used? Discount of 3.5% for utilities and costs Perspective (NHS/PSS) Yes QALYS were used in the model A discount rate of 3.5% was used for both The model was designed from an NHS perspective The model used the health state distribution at the midpoint between cycles for the entire duration of each cycle. Utility values were not measured in the trial but were obtained directly from the efficacy data using the method of Brown et al As defined in the reference case The cost of blindness has been included but only from a direct health care perspective. Reference Ranibizumab for the treatment of diabetic macular oedema (NICE technology appraisal guidance 237) Colquitt et al., 2008 Ranibizumab for the treatment of diabetic macular oedema (NICE technology appraisal guidance 237) NICE Reference case Brown et al NHS, National Health Service; PSS, Personal Social Services; QALYs, quality-adjusted life years Technology Are the intervention and comparator(s) implemented in the model as per their marketing authorisations/ce marking and doses as stated in sections 1.3 and 1.5? If not, how and why are there Alimera Sciences, Inc. Submission of Evidence 75

77 differences? What are the implications of this for the relevance of the evidence base to the specified decision problem? The FAc implant is implemented in the model in accordance with the marketing authorisation and dose as stated in sections 1.3 and 1.5. Direct comparison The model employs data from the FAME trial to compare the FAc implant with optimised standard of care. Per the trial protocol, patients in the active arm received the FAc implant and were allowed rescue laser treatment at physician discretion any time after week 6. The control arm received a sham injection and was similarly allowed rescue laser treatment at physician discretion any time after week 6. Although not allowed in the protocol, offprotocol therapies (OPT) were additionally used in a percentage of patients in each arm (see details in section 5.2.1). OPT included IVTA, Macugen (fellow eye only), Lucentis, Avastin and Ozurdex. At the time of use in the trial, none of the OPT had marketing authorisation for DMO, and currently only Lucentis has a marketing authorisation in the UK for DMO. In the model, the use of OPT is included in the intervention and control arms per the frequency and doses used in the trial (see section 5.2.1). For Lucentis, this reflects the dose per its current market authorization for DMO (Lucentis Summary of Product Characteristics). Of note, the use of OPT is included in the intervention and comparator arms derived from the FAME trials for the first 36 months (for which trial data are available). However, the use of OPT is not included in treatment extrapolations for the intervention of comparator arms over the remaining horizon of the model. Indirect comparison The model employs data from the FAME trial for the intervention arm and efficacy data from the DRCR protocol B trial for the comparator arm (laser monotherapy) and the patients in the treated arm who are withdrawn from Alimera Sciences, Inc. Submission of Evidence 76

78 therapy. FAME data for the FAc implant intervention arm are implemented as described above. For the laser monotherapy comparator arm, 24 month BCVA data were available, and these data have been extrapolated to 36 months assuming that levels stayed the same from 24 to 36 months. Baseline visual acuity data for the comparator group were not published, and so the levels are assumed from the FAME control population. Adverse event data from the FAME control arm are also implemented for the comparator arm in the indirect comparison Please note that the following question refers to clinical continuation rules and not patient access schemes. Has a treatment continuation rule been assumed? If the rule is not stated in the (draft) SPC/IFU, this should be presented as a separate scenario by considering it as an additional treatment strategy alongside the base-case interventions and comparators. Consideration should be given to the following. It is assumed that in clinical practice, one FAc implant will be injected every 36 months; this assumption is reflected in the base case for the model (see section ). There are no treatment continuation rules assumed within the 36 month course of treatment with the FAc implant. However, an assumption has been made about the proportion of patients who would be retreated with an additional FAc implant at the end of the 36 month treatment course. The SPC does not state any specific criteria for retreatment after the first 36 month treatment course, but the assumption made in the model is described below. At the end of the first 36 month treatment course, patients in the intervention arm are separated into two groups: those who are retreated with the FAc implant and those who are not. In order to qualify for retreatment, patients must have gained 5 ETDRS letters of VA compared to baseline within the initial 36 months of treatment. The rationale for the FAc implant retreatment assumption stems from the result reported in the DRCR Protocol I, which assessed the effectiveness of Alimera Sciences, Inc. Submission of Evidence 77

79 laser, triamcinolone and ranibizumab in treating DMO (DRCR 2010). For the comparisons between prompt laser and the two ranibizumab arms, a difference in change in mean BCVA at 1 year versus baseline of 5 letters was considered clinically significant. It is therefore plausible to assume that patients experiencing a gain of 5 ETDRS letters of VA over 36 months would be candidates for retreatment with the FAc implant. In fact, in real-world practice, an improvement of 10 or 15 ETDRS letters is generally accepted by retina specialists as a clinically significant improvement in VA. Therefore, the retreatment criterion applied in the model is a conservative estimate, considering that some retina specialists would likely reserve FAc implant retreatment for a smaller group of patients who gained 10 or 15 letters from baseline after 36 months of treatment. The VA endpoint upon which the retreatment criterion is based is a gain of 5 ETDRS letters, this is considered a clinically relevant improvement which would cause a physician to retreat. Furthermore, the retreatment criterion defined in the assumption can be reasonably achieved for many patients; indeed, xxxx% of chronic DMO patients treated with the FAc implant in the FAME trials gained 5 ETDRS letters over the course of the trials. There are no additional costs or health consequences that result from the retreatment criterion assumed for the FAc implant in the model. Assessing patients change in VA from baseline at the end of the 36 month treatment cycle is considered appropriate timing given the duration of FAc therapy. Implementing this criterion would not introduce any changes to routine clinical practice, considering that DMO patients are already monitored every 3 to 4 months. Because the retreatment assumption applies to the end of the first full course of the intervention, it is unlikely to predict patients for whom the initial FAc implant is particularly cost-effective. However, it does identify patients for whom an additional FAc implant would be clinically justified and cost-effective, and also identifies patients that could be considered insufficiently responsive Alimera Sciences, Inc. Submission of Evidence 78

80 to the FAc implant and for whom the cost-effectiveness of additional treatment would be unlikely. There are no issues with respect to withdrawal of treatment from patients who do not meet the retreatment criterion, as the retreatment decision is taken at the conclusion of the first full course of treatment with the FAc implant. There are no other equity considerations associated with the retreatment assumption. 6.3 Clinical parameters and variables Please demonstrate how the clinical data were implemented into the model. The clinical efficacy data were used directly to build a quarterly matrix of the number of patients in each of the levels of visual acuity represented by the 5 letter categories. This was repeated in each quarter for the 3 years of the trial giving a transition matrix to be built up. The appropriate utility value was assigned to each of these visual acuity levels according to the technique of Brown et al. (2000). Data on rescue laser treatments, off-protocol interventions (additional short acting DMO treatments, see section 5.2.1), adverse events and side effects (see section 5.9.2) were included in the model specifically for each intervention Demonstrate how the transition probabilities were calculated from the clinical data. If appropriate, provide the transition matrix, details of the transformation of clinical outcomes or other details here. The clinical efficacy data were used directly to build a quarterly matrix of the number of patients in each of the levels of visual acuity represented by the 5 letter categories. This was repeated in each quarter for the 3 years of the trial allowing a transition matrix to be built up. A copy of the treatment arm matrices for the FAc implant, OSC and laser only arms is included in Figure B9. Alimera Sciences, Inc. Submission of Evidence 79

81 xxxxxx xx: xxxxxxxxx xxx xxxxxxxx Alimera Sciences, Inc. Submission of Evidence 80

82 6.3.3 Is there evidence that (transition) probabilities should vary over time for the condition or disease? If so, has this been included in the evaluation? If there is evidence that this is the case, but it has not been included, provide an explanation of why it has been excluded. Whilst a true placebo based assessment of patient visual decline was not measured in the FAME studies, there is evidence in clinical trials that patients do decline with time (discussed in section 6.4.2). In the post 3 year phase of the model it has been assumed that the vision of patients with active treatment and no further active intervention will decline at differing rates per quarter. This is discussed in section Were intermediate outcome measures linked to final outcomes (for example, was a change in a surrogate outcome linked to a final clinical outcome)? If so, how was this relationship estimated, what sources of evidence were used, and what other evidence is there to support it? No intermediate outcome measures were linked to final outcomes If clinical experts assessed the applicability of values available or estimated any values, please provide the following details 3 : No clinical experts were engaged to assess the applicability of values or estimate any values employed in the economic model. However, three consultant ophthalmologists currently practicing in the NHS and experienced in treating patients with DMO were informally engaged via through a third-party agency during the scoping process to provide guidance on the definition of optimized standard of care for chronic DMO patients who are insufficiently responsive to available therapies. 3 Adapted from Pharmaceutical Benefits Advisory Committee (2008) Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra: Pharmaceutical Benefits Advisory Committee. Alimera Sciences, Inc. Submission of Evidence 81

83 Clinical experts were selected from a national group of ophthalmologists who have been involved in development of ophthalmology guidelines and are widely published in the area of DMO. These experts had previously participated in market research commissioned by Alimera Sciences; however, this informal exchange was separate from the market research previously conducted. Iteration was not used in the collation of opinions. responses were reviewed and employed to inform the selection of the appropriate comparator to represent optimized standard of care in this submission. The general consensus from clinical experts was that optimized standard of care for this group of patients varies on a Trust-by-Trust basis, but can broadly be defined as no treatment, i.e. watching and waiting or registering for blindness services and low vision aids when appropriate, with occasional additional maintenance laser treatments and in some cases intravitreal injections of triamcinolone or VEGF antibodies (typically bevacizumab). (See sections 2.4 and 2.6). Summary of selected values Please provide a list of all variables included in the costeffectiveness analysis, detailing the values used, range (distribution) and source. Provide cross-references to other parts of the submission. Please present in a table, as suggested below. The model was originally designed so the majority of inputs could be analysed as variables with sensitivity analysis to determine the key parameters of interest. However, the following key variables were included in the model analysis. Alimera Sciences, Inc. Submission of Evidence 82

84 Table B17: Summary of variables applied in the economic model Variable Value CI (distribution) Visual acuity change per quarter after 3 years FAc implant RTx arm NRTx arm 5% improvement in 5-10 letters 3% decrease 0 to 8% %n/a in 5-10 letters Relative risk of mortality as more conservative value Cataract surgery Glaucoma surgery (post year 3) Other adverse events Number of FAc implant injections Number of laser therapies Number of ranibizuma b injections OSC arm DRCR laser only FAc implant arm OSC arm Rate for first 3 years (all arms) Rate post year 3 (control arm) Number of non-ranibizumab OPT in FAc implant and OSC arms Rate of FAc implant retreatment after Year 3 Bilateral treatment FAc implant arm Per FAME trials 50% of FAME values for all arms 1.79% p.a 1.20% % p.a. Reference to section in submission Cost Rate for 1.18% p.a. 1.71% % Control arm Rates per FAME trials +/- 20% rates from FAME Year , Year Year Year Year Year Year Year Year Year Year Year Year Year Year Rates per +/- 20% rates FAME trials from FAME xxxx% 25% - 100% xx% 0% - 35% Control 35% 0% - 35% arm Utility weights , to < to < Values and 60 to < % CI from 55 to< Brown et al. 50 to < Alimera Sciences, Inc. Submission of Evidence 83

85 Variable Value CI (distribution) Reference to section in submission 45 to < to < to < to < to < to < < Utility improvement for bilateral 25% 0 50% treatment Drop-out rate xxxx% 0% % Cost of blindness /- 20%, or not included Discount rate for Costs and QALYs 3.5% 0 and 6% Reference case Administration costs FAc implant Laser therapy OPT CI, confidence interval, RTx: patients retreated after 3 years, NRTx: patients not retreated after 3 years with either control treatments or the FAc implant Although the model has been designed such that the majority of inputs are treated as variables rather than fixed assumptions, other variables in the model do not significantly impact the uncertainty of the model and therefore have not been reviewed separately in the formal deterministic review Are costs and clinical outcomes extrapolated beyond the trial follow-up period(s)? If so, what are the assumptions that underpin this extrapolation and how are they justified? In particular, what assumption was used about the longer term difference in effectiveness between the intervention and its comparator? For the extrapolation of clinical outcomes, please present graphs of any curve fittings to Kaplan-Meier plots. Definition of retreated population At the end of the first three-year treatment cycle with the FAc implant, it is assumed that treating physicians will want to continue treatment for those patients who were tolerant to the medication and have responded sufficiently to the FAc implant. In the FAME trials, after 36 months, 76.1% of the subjects Alimera Sciences, Inc. Submission of Evidence 84

86 who were randomized as chronic DMO patients to the FAc implant treatment arm completed the last visit. This rate of drop-out (23.9%) is consistent with other large, phase 3 trials studying patients with DMO and other retinal diseases such as wet AMD. The timing for patients to leave the study varied, and occurred throughout the 36 month follow up period. The reasons included death, geographic relocation and changes in systemic disease which precluded continued participation in the study. A 5 letter improvement in VA is used as the threshold for defining sufficient response justifying retreatment with the FAc implant after the first 36 month treatment cycle. This threshold is based on the result reported in the DRCR Protocol I assessing the effectiveness of laser, triamcinolone and ranibizumab in treating DMO (DRCR 2010). For the comparisons between prompt laser and the two ranibizumab arms, a difference in change in mean BCVA at 1 year versus baseline of 5 letters was considered clinically significant. Within the FAME trials, xxxx% of FAc implant treated chronic DMO patients gained 5 letters in VA at 36 months. Based on the above, it is assumed that xxxx% of chronic DMO patients originally treated with the FAc implant would be retreated at the end of the first 3 year treatment cycle. This is reduced by the drop-out rate of xxxx%, leading to a total of xxxx% of patients qualifying for retreatment with the FAc implant in the second 3 year treatment cycle. In further 3 year FAc implant retreatment cycles it is assumed that all patients would be considered sufficiently treated and thus qualify for retreatment; however, the drop-out rate of xxxx% continues to apply to the later retreatment cycles. As a result, the percentage of patients qualifying for retreatment with an FAc implant in the third, fourth and fifth treatment cycles are approximately xx%, xx % xxx xx%, respectively. Extrapolation of visual acuity outcomes To estimate the quarterly changes in vision states for patients with chronic DMO treated with the FAc implant who either respond and continue treatment, or do not respond and stop treatment, we used the only reliable dataset for Alimera Sciences, Inc. Submission of Evidence 85

87 such patients, i.e. data from the FAME studies, studying the FAc implant in the treatment of chronic DMO, as follows: Based on the best corrected visual acuity (BCVA) at month 36, in the intent to treat population, we divided the population into those subjects who experienced an improvement from baseline of 5 letters ( responders ) and those subjects who experienced an improvement of <5 letters or who had a worsening in vision relative to baseline ( non-responders ). Subjects identified as responders represent those chronic DMO subjects who would be considered eligible for retreatment with the FAc implant, whereas, those subjects identified as non-responders would stop treatment with the FAc implant. The changes in vision states over the 36 month course of the FAME studies for these two populations was then determined on a quarterly basis. The percentage of subjects whose change in BCVA from the previous quarter was characterized as a) an improvement of 5 letters, b) a change of less than 5 letters (i.e., between -4 and +4 letters), or c) a loss of 5. The average percentage across all 12 quarters of the FAME studies for each change in vision state was determined. Because of the potential confounding effect of cataracts on vision during the first two years of the FAME studies, average percentages were also calculated based solely on the quarterly data derived from the third year of the FAME studies. This produced the results in Table B18. Alimera Sciences, Inc. Submission of Evidence 86

88 Table B18: Average Percentage of Subjects Experiencing a Change in Vision State Month 36 Change from Baseline in BCVA Change in Vision State From Previous Quarter xxxxxxx xx xxx xxxxx xxxxx xx xxxxxx xxxxxxxxxxx xx xxxxxx xxxxxxxxx xxxxxxx xx xxxxx xxxxr xxxx xx xxxxxx xxxxxxxxxxx xx xxxxxx xxxxxxxxx Responders (x=xxx) xxxx% xxxx% xxxx% xxxx% Non-responders (x=xx) xxxx% xxxx% xxxx% xxxx% If the percentage of subjects in the responder group experiencing improvement is adjusted by the corresponding percentage in the nonresponder group, then the adjusted percentage of responders experiencing a transition to a higher vision state ranges from x% (xxxx% xxxxx xxxx%) based on data from the entire three years of the FAME studies to xxx% (xxxx% xxxxx xxxx%) based on data solely derived from the third year of these trials. Likewise, the adjusted percentage of non-responder subjects experiencing a transition to a lower vision state, adjusted by the corresponding percentage in the responder group, ranges from xxx% (xxx% xxxxx xxxx%) based on data from the entire three years of the FAME studies xx xxx% (xxxx% xxxxx xxxx%) based on data solely derived from the third year of these trials. Alternatively, if the percentage of subjects experiencing a transition to a higher vision state is adjusted by the percentage experiencing a transition to a lower vision state, then the adjusted percentage of responders experiencing any transition in vision state ranges from xxx% (xxxx% minus 15.4%) based on data from the entire three years of the FAME studies xx xxx% (xxxx% xxxxx xxxx%) based on data solely derived from the third year of these trials. Alimera Sciences, Inc. Submission of Evidence 87

89 Likewise, the adjusted percentage of non-responders experiencing any transition in vision state, based on the same methodology, ranged from xxx% [xxxx% xxxxx xxxx%] based on data from the entire three years of the FAME studies to xxx% [xxxx% xxxxx xxxx%] based on data solely derived from the third year of these trials. Based on these two approaches, which use the actual data collected during the 3 year FAME studies, the following values were considered appropriate for extrapolation of vision outcomes in further treatment cycles in the base case analysis. For the FAc implant, 5% of the subjects who were considered responders at month 36 (i.e. those who are retreated in subsequent treatment cycles) will experience a transition to a higher vision state per quarter, and 0% will experience a transition to a lower vision state per quarter. For those subjects who were considered non-responders at month 36 (i.e. those who are not retreated after 36 months), 0% will experience a transition to a higher vision state per quarter, and 3% will experience a transition to a lower vision state per quarter. The population being continually treated with the FAc implant is not expected to make any transitions to lower vision states, and 5% of the population is expected to transition to the next higher vision state each quarter. Populations receiving no additional treatment after the first 36 month treatment cycle, i.e. those who gained <5 letters of VA with the FAc implant at 36 months and all patients in the direct and indirect control arms, are not expected to make any transitions to higher vision states, and 3% of the populations will transition to the next lower vision state each quarter. Extrapolation of costs At the end of the 36 month FAME data based treatment cycle it is assumed that patients who are not retreated with the FAc implant revert to observation, which does not involve any further expense except for support of those entering a classification of blindness. The change in cost blindness for patients having a visual acuity score lower than 35 letters is assigned at a discounted rate. Alimera Sciences, Inc. Submission of Evidence 88

90 The cost of AEs associated with the ongoing treatment with the FAc implant are incurred at the rates recorded in the clinical trials with appropriately discounted costs. In the case of cataract surgery, this only occurs until the patients are all pseudophakic (i.e. their lens has been replaced with an artificial one which is not susceptible to cataracts). In the control arms receiving no further treatment patients are assumed to continue to develop cataracts with the progress of their disease. The best source of data obtained was the paper by Klein et al which reported a 17.9 % incidence over 10 years in a diabetic population. These patients thus incur additional costs for cataract surgery in subsequent treatment cycles. Routine monitoring costs are incurred in all treatment arms of the model. It could be argued that as they are incurred at the same rate post the first treatment cycle, these costs could be excluded from the analysis. There are no additional visits required for FAc implant monitoring or maintenance injections associated with the use of other intravitreal therapies. However, the differential use of OPT in the FAME trials results in a differential cost in the first cycle; therefore the cost item as a whole has been accounted for in the first but not subsequent 3 year treatment cycles in the model Provide a list of all assumptions in the de novo economic model and a justification for each assumption Assumption: Retreatment with FAc implant within the first treatment cycle The FAME trial protocol included pre-specified criteria for retreatment with the FAc implant related to decline in visual acuity or exacerbation of DMO that allowed for additional implants to be administered during the post 12 month visit period at the discretion of a masked investigator. As a result, some chronic DMO patients in the FAME trial received multiple FAc implants over the 36 month treatment period (see section 5.3). However, the model assumes that all patients will receive only one FAc implant every 36 months. The rationale for this assumption is explained in section Alimera Sciences, Inc. Submission of Evidence 89

91 Assumption: Use of data from FAc implant treated patients including those who received multiple injections As explained in section , while the FAME trials were on-going, the duration of release in vivo, as well as the timing and magnitude of the effect of cataract, were not understood. Because so many of the FAME trial subjects had severe disease, the investigator making treatment recommendations was masked, and there was no cap on the number of FAc implants that could be administered during the post 12 month visit period, some subjects with very severe disease received multiple FAc implants during the 36 month period in this double-masked, investigational drug study setting. Consistent with this scenario, post-hoc analysis of visual acuity outcomes for subjects receiving single vs. multiple FAc implants shows that subjects receiving one FAc implant had a more favourable visual acuity outcome at 36 months compared to those who received multiple FAc implants (Figure B10, Alimera data on file). Alimera Sciences, Inc. Submission of Evidence 90

92 xxxxxx xxx: xxxxxxx xx xxxxxxxx xxxx xxx xxxxxx xxxxxxxxxxx xx xxxx xx xxxxxx xx xxx xxxxxxxx (xxxx-xxx xxxxxxxx) In order to maintain consistency with the assumption that patients will receive one FAc implant per 36 months treatment cycle (section ), one option would be to employ data from the post-hoc analysis of visual acuity outcomes for the subjects who received one FAc implant during the trial. Given the greater improvement in visual acuity outcomes for this group, employing this post-hoc analysis in the model would favourably impact the cost-effectiveness outcomes for the FAc implant compared to control. However, given the lack of statistical validity of post-hoc analysis, it was determined that the most conservative approach would be to employ efficacy and safety data from all chronic DMO subjects including those who received multiple FAc implants, but, based on the reasons outlined, assume only one FAc implant would be administered per the 36 month treatment. Alimera Sciences, Inc. Submission of Evidence 91

93 Assumption: Retreatment after the first three year treatment cycle The FAME studies permitted the use of laser photocoagulation after Week 6. In addition some subjects received off-protocol therapies at the discretion of the investigator. In clinical practice, it is anticipated that the FAc implant will be used in patients who are considered insufficiently responsive to other therapies so there would rarely be a rationale for treatments other than the FAc implant (see section ). As a result, the model assumes use of all treatments per the trial for the first 36 month treatment cycle. However, in extrapolation to subsequent treatment cycles, patients who are identified as candidates for retreatment with the FAc implant are assumed to only receive this therapy; no laser therapy or additional intravitreal injections. Assumptions made for the rate of retreatment with the FAc implant are discussed in section For the population that is not retreated with the FAc implant and the control arms in the direct and indirect comparison, the first 36 month treatment cycle assumes use of therapies as per the trial. However, for subsequent treatment cycles patients are assumed to receive no further treatment. Consistent with the population for which the FAc implant is indicated, these subjects have been treated with the available therapies considered appropriate for their disease, and are insufficiently responsive. Therefore, if this population is not receiving the FAc implant, there would be no other treatment options and they would simply be observed Assumption: Change in visual acuity status after the first three year treatment cycle These assumptions are discussed in Assumption: Bilateral treatment Whilst the FAME studies were based upon the accepted clinical practice of treating one eye first, generally the worst functioning, and awaiting outcomes Alimera Sciences, Inc. Submission of Evidence 92

94 prior to retreatment, it is assumed that a number of patients will receive treatment in the second eye. In the NICE appraisal of ranibizumab, it was assumed that 35% of DMO patients would be candidates for bilateral treatment; given that the committee agreed this to be a plausible assumption, it has been adopted in the current model. In addition, it has been assumed that patients treated in the first eye with the FAc implant who showed a subsequent rise in IOP xxx xxxx would not be considered candidates for bilateral treatment. This is based on an assumption that physicians would not be willing to treat the second eye due to concerns about IOP in the second eye. The rate of IOP rise xxx xxxx was xxxx% in the chronic DMO population within the FAME trial, and this has been applied to the 35% bilateral treatment assumption discussed above. Thus, for the control population, it is assumed that 35% of patients are candidates for bilateral treatment, whereas for the FAc implant treated patients it is assumed that xx% are eligible for bilateral treatment. The assumed rates of bilateral treatment for the FAc implant arm and the control arms have been simply applied to all costs that could be assigned to one eye only, but it is assumed that some costs such as monitoring visits would apply for both eyes without additional costs. These have therefore not been increased Assumption: Bilateral benefit With the treatment of the second eye detailed above it is assumed that the patient will accrue improvement in visual acuity and overall health related quality of life. The data from the FAME studies and thus the model have been built upon the basis of treating a single eye, generally accepted to be the eye with the worst vision status. The precise impact of treating the second eye is not clear as it could not be assumed to be as dramatic as any improvement seen in the first treated eye; however, it is assumed that there would be both an improvement in visual acuity and other sight related functioning such as stereoscopic vision, perception of depth of field etc. In the absence of trial Alimera Sciences, Inc. Submission of Evidence 93

95 based data therefore we have adopted the same 25% uplift in utility values as used in the ranibizumab review (NICE TA237) Assumption: Mortality rate Even with the current improvement in diabetes control and care, the relative risk of mortality for patients with Type 2 diabetes is considerably higher than that of the general population. In the NICE Appraisal of ranibizumab a value of 2.45 derived from the work of Mulnier et al. (2006) and Hirai et al. (2008) was applied and this has been adopted in this analysis. The base mortality data was derived from the Office of National Statistics. There was no differential mortality rate assigned to deteriorating vision or lack of response to FAc implant therapy and the rate employed was consistently applied to all arms of the model Utility Weights It was not possible to obtain an alternative source of utility weights for the model. However, given the importance of the health state utility values to the ICER of the intervention and comparators, a range based on the confidence interval published by Brown et al was used for the sensitivity analysis Number of Monitoring Visits Due to the long term effect of the FAc implant there is no need for repeated procedure-related healthcare visits. Any monitoring of the FAc implant is assumed to take place at the quarterly DMO monitoring visits that currently occur in routine clinical practice (section 2.8). The small number of OPT interventions in the first 3 year treatment cycle are also assumed to take place associated with these visits although additional injection related charges are assumed to be incurred. Alimera Sciences, Inc. Submission of Evidence 94

96 6.4 Measurement and valuation of health effects Patient experience Please outline the aspects of the condition that most affect patients quality of life. DMO typically results in a gradually progressing loss of visual acuity. Ferris and Patz (1984) have reported that 53% of untreated DMO subjects lost 2 lines of visual acuity over a 2 year period. In the Early Treatment of Diabetic Retinopathy Study (ETDRS), over 3 years, 33% of the untreated group experienced a moderate vision loss ( 15 letters) (Early Treatment Diabetic Retinopathy Study (ETDRS 1987). Davis et al. (2009) reported that subjects treated with placebo in Study PKC-DRS2 experienced a loss of 0.67 letters per month over the 3 years of the study. Thus, while DMO can improve without treatment in some patients, the general trend is for the disease to progress and for vision loss to increase with time. In studies of DMO the standard outcome is the improvement of 3 lines from baseline to ensure that the endpoint is clinically relevant. Loss of 3 ETDRS lines is considered a moderate vision loss. The loss of 2 lines of vision results in a loss of function that is clear to the patient. A recent study has demonstrated that greater severity and laterality of DR (which includes DMO) were associated with lower general and vision-specific HRQOL. (Mazhar et al. 2011) For the patients this means that as DR progresses in one eye, then involves both eyes which progress, the patient s quality of life is reduced. As a result, a treatment that reverses moderate vision loss may allow the patient to avoid this reduction in quality of life by permitting them to continue reading without low vision aids, drive a car and participate in physical activities that require unimpaired vision. Alimera Sciences, Inc. Submission of Evidence 95

97 Patients (%) Please describe how a patient s HRQL is likely to change over the course of the condition. The patients HRQL is likely to be improved through reversal of moderate vision loss which improves the patient s health status and will allow greater independence and productivity. The results of the ETDRS Multi-step Eye Scale assessment in the FAME studies support this view. Subjects with chronic DMO in the FAME studies who were treated with the FAc implant experienced a greater incidence of 2 step improvement in ETDRS DR Eye Scale than the sham-treated subjects (Figure B11). This demonstrates that the anatomical changes in the FAc implant-treated eyes were reversed by a clinically relevant amount over the 3 year treatment period. Figure B9: Percentage of Patients with 2-Step Improvement in ETDRS Score 18 Sham (N=109) 0.2 g/d FAc (N=198) * p= % p= Months * 8.3% The mean ordinal score for subjects in this treatment group at baseline was 5.7 (equivalent to >47and <53 on the ETDRS Scale). A two-step improvement indicates that some patients DMO will be improved sufficiently to reduce driving difficultly based on the association between the VFQ-25 driving difficulty score and the DR severity score reported by Mazhar et al. (2011). (Figure B12) Alimera Sciences, Inc. Submission of Evidence 96

98 Figure B10: Association between VFQ-25 and Driving Difficulty Relationship between national Eye Institute-Visual Function Questionnaire-25 (NEI-VFQ-25) driving difficulty subscale and composite score and severity of diabetic retinopathy (DR) based on concatenated scale with a locally weighted regression plot (LOWESS). ETDRS scores correspond with the following clinical severity levels: No DR (levels 10-13); unilateral or bilateral nonproliferative DR (NPDR; levels 14-53); unilateral of bilateral proliferative DR (PDR; levels 60-85). The DR severity concatenated scale step 9 (43/43) corresponds with bilateral moderate NPDR. The NEI-VFQ-25 scores have been adjusted for covariates. Source: Copied from Mazhar et al HRQL data derived from clinical trials If HRQL data were collected in the clinical trials identified in section 5 (Clinical evidence), please comment on whether the HRQL data are consistent with the reference case. The following are suggested elements for consideration, but the list is not exhaustive. The FAME studies included an assessment using the VFQ-25 (and VFQ-39 at selected sites) at baseline and Months 24 and 36. Subjects in the FAME Alimera Sciences, Inc. Submission of Evidence 97

99 studies received study treatment in 1 eye (typically the worst eye), while the VFQ-25 assesses the overall visual function quality of life which is driven by the best eye. As a result the VFQ-25 in the FAME studies is not an accurate source of the impact of treatment with the FAc implant in both eyes or in the best eye. As with other clinical trials with monocular treatment using the worst eye at baseline, the VFQ-25 and VFQ-39 assessment did not provide useful data in the FAME studies because it is primarily dependent on the function of the best-seeing eye. The cost-effectiveness model incorporates utilities from a published, peer-reviewed source rather than from the direct comparison study (Brown et al. 2000). Mapping If mapping was used to transform any of the utilities or quality-oflife data in clinical trials, please provide the following information. As described in section 6.4.3, utility values were not obtained directly in the FAME studies. In the absence of any peer-reviewed values for DMO that could be assigned to the trial derived visual acuity levels (see section 9.12), the utility weights were obtained from a study of AMD patients by Brown et al. (2000). The study utilized the time trade-off method. Because the Brown study did not report the utility weights associated with BCVA in patients between the ranges of 35 and 50, the un-weighted average of the utility weights above and below this range of BCVA have been assumed as the utility weight associated with this BCVA range (Table 19). Alimera Sciences, Inc. Submission of Evidence 98

100 Table 19: Utility Weights for BCVA Health States Considered by the Model BCVA Health States ETDRS letters (Snellen equivalents) Utility Weights 95% Confidence Intervals At least 75 or greater ( 20/32) At least 70 and less than 75 (20/32-23/40) At least 65 and less than 70 (20/40-20/50) At least 60 and less than 65 (20/50-20/63) At least 55 and less than 60 (20/63-20/80) At least 50 and less than 55 (20/80-20/100) At least 45 and less than 50 (20/100-20/125) At least 40 and less than 45 (20/125-20/160) At least 35 and less than 40 (20/160-20/200) At least 30 and less than 35 (20/200-20/250) At least 25 and less than 30 (20/ ) At least 20 and less than 25 (20/320-20/400) Less than 20 (< 20/400) Assumptions Matched to the utility reported for 20/20 to 20/25 in Brown et al., /30 to 20/50 20/60 to 20/100 Brown et al., 2000 did not report the utility weight associated with this range of BCVA We assumed the average of the utility weights above and below this range of BCVA 20/200 to 20/400 Counting fingers to light perception HRQL studies Please provide a systematic search of HRQL data. Consider published and unpublished studies, including any original research commissioned for this technology. Provide the rationale for terms used in the search strategy and any inclusion and exclusion criteria used. The search strategy used should be provided in section 9.12, appendix 12. The literature review focused on utility values tied to visual acuity. Early searches suggested limited evidence on health-state utilities in this area, with no known studies in DMO. Thus, the search parameters were broadened to include other disorders affecting visual acuity. A flow diagram of the search for HRQL studies reporting utility values associated with visual acuity is provided in Figure B13. Alimera Sciences, Inc. Submission of Evidence 99

101 Figure B11: Flow diagram for HRQL literature search Final results included 22 articles for review, 4 HTA assessments, and 9 conference abstracts. Details of the search are provided in Appendix 9.12, along with a summary of the articles reviewed Provide details of the studies in which HRQL is measured. Include the following, but note that the list is not exhaustive. Population in which health effects were measured. Information on recruitment. Interventions and comparators. Sample size. Response rates. Description of health states. Adverse events. Appropriateness of health states given condition and treatment pathway. Method of elicitation. Method of valuation. Mapping. Uncertainty around values. Alimera Sciences, Inc. Submission of Evidence 100

102 Consistency with reference case. Appropriateness for cost-effectiveness analysis. Results with confidence intervals. Appropriateness of the study for cost-effectiveness analysis. The selected articles were reviewed and a shortlist of potential sources produced namely Sharma et al 2000, Brown et al 2000 and the NICE Review of ranibizumab (NICE TA237), containing the utilities produced by Brazier et al. It should be noted that the Sharma and Brown groups are the source of utility values for the majority of studies carried out since. On review of the articles and HTA reviews, the most credible source of utility values would appear to be the Brazier data (qtd. in NICE TA237), due to the patient group assessed, the fact that time trade off (TTO) was used and the data had face validity. However, the full details of the study and data have not yet been made fully available in the public domain and therefore are not suitable for use in the model for the current submission. In the Sharma study, the assessment technique used was TTO, but there were a few key issues with the design that make the data less suitable for use in the current model. These were specifically that patients with a wide range of ocular conditions were included in the study, there were variations in the standard visual acuity assessment procedure, and equations were used to modify the directly assessed values. In conclusion therefore the study by Brown et al. (2000), which studied the related AMD patient group and used the TTO technique, was selected for the model. Brown et al. (2000) did also use the standard gamble (SG) technique in the study, but these data were not used in the current submission, as the data reported are higher as compared to the TTOgenerated data, which was attributed by the authors to a poor understanding of the SG technique being used by the respondents. Alimera Sciences, Inc. Submission of Evidence 101

103 6.4.7 Please highlight any key differences between the values derived from the literature search and those reported in or mapped from the clinical trials. As demonstrated by the literature search described in section 9.12, there are little relevant published data on utilities in this patient group. In the ranibizumab NICE review (NICE TA 237) the following data from Brazier et al. (2006) were included. Source: Ranibizumab Appraisal (NICE TA237) ERG Report Table 11 The range of utilities stated is considerably narrower than used in the current model, has higher values on average and is based on 10 letter increments. This greater range could have a positive impact on the benefit accrued by the FAc implant over optimised standard of care as patients reach the higher extreme of the VA range and non-treated patients decline. This would be offset by the higher values accrued by patients not changing their VA. However, in the absence of a peer-reviewed publication of this data or more information on the way that the ranibizumab RESTORE trial data were used to generate the assumptions employed in their submission, it is not possible to comment further on the relative benefits of the data sets. Alimera Sciences, Inc. Submission of Evidence 102

104 Adverse events Please describe how adverse events have an impact on HRQL. The model did not consider utility decrements due to the adverse events or procedures and interventions for the adverse events. As the utility values for the model were obtained directly from the quarterly BCVA values, the impact of adverse events such as cataract formation on patient vision during FAc implant exposure is incorporated in the overall utility benefit assessment. The impact of cataract on the FAc implant visual acuity outcomes may be observed in particular on the visual acuity levels used for the transition matrix for year 2 of the model. The resolution of the cataracts through surgery and consequent return of improving visual acuity may be further observed through this data. The impact of any treatment-related adverse events that do not impact the visual acuity directly and the therefore have an impact on the estimates of HRQL are not accounted for. Quality-of-life data used in cost-effectiveness analysis Please summarise the values you have chosen for your costeffectiveness analysis in the following table, referencing values obtained in sections to Justify the choice of utility values, giving consideration to the reference case. Values chosen for the cost-effectiveness analysis are summarized in section As stated in 6.4.8, AEs that impact visual acuity, namely the formation of cataracts, were included in the trial derived utility values; the impact of other AEs on quality of life is not included in the utility values If clinical experts assessed the applicability of values available or estimated any values, please provide the following details 4 : Four consultant ophthalmologists currently practicing in the NHS and experienced in treating patients with DMO and three UK-based HTA experts were invited to participate in a face-to-face advisory board with the aim of 4 Adapted from Pharmaceutical Benefits Advisory Committee (2008) Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra: Pharmaceutical Benefits Advisory Committee. Alimera Sciences, Inc. Submission of Evidence 103

105 obtaining input on, and validating key assumptions, being employed to develop the health economic model used to support this submission. Clinical experts were selected from a national group of ophthalmologists who have been involved in development of ophthalmology guidelines and are widely published in the area of DMO. HTA experts were selected based on previous roles within either NICE or the Scottish Medicines Consortium (SMC) and/or current professorship in Health Economics in UK based universities. An honorarium was offered for participation in the advisory board, and participants travel and accommodations were additionally reimbursed at cost. Of the seven experts invited, one clinician and one HTA expert participated in the advisory board. The advisory board also included clinicians and HTA experts from other European markets not relevant to this submission. The advisory board comprised one day, and included several sessions including an overview of the clinical data from the FAME trials, a review of the economic model structure and a discussion about key model inputs and assumptions. Prior to the advisory board, participants were sent background information that included an overview of the disease landscape, a summary of data from the FAME trials, an overview of the cost-effectiveness model, a summary of utility studies informally collected from the literature and a summary of the NICE review for ranibizumab in DMO (NICE TA237). One topic of discussion during the advisory board involved the definition of health states and the selection of utility values to be employed in the model. Opinions were solicited from the group regarding the use of utility values from Brown et al. (2000) and the applicability of these data for the current submission. The use of the Brazier data in TA237 was discussed, and UK participants indicated that if these data were available, they may be most appropriate for the economic model submitted to NICE. However, given the fact that the Brazier data are not available in the public domain, UK participants confirmed that data from Brown et al. (2000) were likely the most appropriate source for utilities for this submission (see section for further details). Iteration was not used in the collation of opinions. Alimera Sciences, Inc. Submission of Evidence 104

106 Please define what a patient experiences in the health states in terms of HRQL. Is it constant or does it cover potential variances? The health states are designed to capture HRQL related to patients visual acuity in the treated eye as measured by the number of ETDRS letters they are able to read on an eye chart. Table 20 shows the relationship between the health states (BCVA by increments of 5 ETDRS letters), the range of vision loss defined by the ICD-9, statistical estimates of reading ability and estimated skills for orientation and mobility (Duane s Clinical Ophthalmology, Vol. 3, Chapter 51). Since the visual system alone provides as much input to the brain as all other senses combined, it is not surprising that vision loss can have a devastating impact upon patients lives; this is illustrated by the relatively low utility values associated with poorer BCVA values (ETDRS letters). Table B20: Relationship between health states and HRQL Health state ETDRS letters Ranges of vision loss (ICD-9-CM) Statistical estimates of reading ability Estimated skills for orientation and mobility 1 75 Range of normal vision Normal reading speed Normal reading distance Reserve capacity for small print Normal visual orientation Normal mobility skills 2 70 to to to 65 Near-normal vision Normal reading speed Reduced reading distance No reserve for small print Normal orientation and mobility Needs more scanning Occasionally surprised by events on the side 5 55 to to to to 45 Moderate low vision Near-normal with reading aids Uses low power magnifier or large print books Near-normal performance Requires scanning for obstacles 9 35 to to to to 25 Severe low vision Slower than normal with reading aids Uses high power magnifiers Visual mobility is slower than normal Requires continuous scanning May use cane as adjunct for identification Alimera Sciences, Inc. Submission of Evidence 105

107 Health state ETDRS letters Ranges of vision loss (ICD-9-CM) Statistical estimates of reading ability Estimated skills for orientation and mobility 13 <20 Profound low vision Near-blindness Marginal with reading aids Uses magnifiers for spot reading, but may prefer talking books No visual reading Must rely on talking books, Braille or other non-visual sources Must use long cane for detection of obstacles May use vision as adjunct for identification Visual orientation unreliable Must rely on long cane, sound, guide dog, other blind mobility skills Were any health effects identified in the literature or clinical trials excluded from the analysis? If so, why were they excluded? No additional health effects were identified If appropriate, what was the baseline quality of life assumed in the analysis if different from health states? Were quality-of-life events taken from this baseline? The baseline quality of life was taken directly from the health state of the patient measured in visual acuity. The changes in patient quality of life for the model were taken from this baseline Please clarify whether HRQL is assumed to be constant over time. If not, provide details of how HRQL changes with time. A patient with chronic DMO is expected to experience a change in their visual acuity with time which will be associated with a change in their HRQOL Have the values in sections to been amended? If so, please describe how and why they have been altered and the methodology. The data obtained from Brown was deficient of the utility weights associated with BCVA in patients between the ranges of 35 and 50. As a result, we assumed the un-weighted average of the utility weights above and below this range of BCVA as the utility weight associated with this BCVA range. No other modifications were made. Alimera Sciences, Inc. Submission of Evidence 106

108 6.5 Resource identification, measurement and valuation NHS costs Please describe how the clinical management of the condition is currently costed in the NHS in terms of reference costs and the payment by results (PbR) tariff. Provide the relevant Healthcare Resource Groups (HRG) and PbR codes and justify their selection. Please consider in reference to section 2. The HRG description for laser (including administration) and administration of intravitreal injections used for NHS costing is as follows: HRG code HRG name BZ23Z Vitreous Retinal Procedures category 1 It has been selected because the OPCS codes fall within HRG codes and are known as: Laser photocoagulation to lesion of the retina NEC (C82.6) vitreous retinal (VR) banding of 1(NHS Information Centre). Injection of therapeutic substance into posterior segment of eye NEC (C89.3) VR banding of 2 (for the FAc implant, bevacizumab, ranibizumab, IVTA and dexamethasone intravitreal implant) (NHS Information Centre). Monotherapy with laser, the FAc implant, bevacizumab, ranibizumab, IVTA and dexamethasone intravitreal implant would be costed as a Category 1 procedure (sum of VR bands from 0 to 2). (NHS Information Centre) Please describe whether NHS reference costs or PbR tariffs are appropriate for costing the intervention being appraised. The NHS Reference Costs and the British National Formulary (BNF) will be used to cost resources associated with the FAc implant (Joint Formulary Committee). The NHS Reference Costs are more appropriate because they represent the actual national average costs that have already been incurred as a result of Alimera Sciences, Inc. Submission of Evidence 107

109 delivering care. These costs take into account staff time, event-based time and standard equipment time. Thus, they include opportunity costs, whereas the PbR Tariffs are prices (or prospective costs) which are prone to adjustment in the future. Thus at point of use in the model, costs based on the PbR Tariffs will not reflect opportunity costs of delivering care. Resource identification, measurement and valuation studies Please provide a systematic search of relevant resource data for the UK. Include a search strategy and inclusion criteria, and consider published and unpublished studies. The search strategy used should be provided as in section 9.13, appendix 13. If the systematic search yields limited UK-specific data, the search strategy may be extended to capture data from non-uk sources. A total of 136 hits were identified from the literature search in databases from Medline, EconLit, NHS EED and Embase. However, after a series of filtering processes, no study was identified that reported resource use in the management of DMO patients in the UK. When DMO-related studies were identified, the articles were not of UK population and the perspectives taken by the authors did not allow resource use estimation to be applied to the UK. Alimera Sciences, Inc. Submission of Evidence 108

110 6.5.4 If clinical experts assessed the applicability of values available or estimated any values, please provide the following details 5 : As discussed in section , an advisory board involving a consultant ophthalmologist and an HTA expert, both based in the UK, was conducted to obtain input on the key assumptions employed in the economic model. The use of resources was a topic of discussion during this advisory board, and UK participants were asked to provide guidance and insight into the medical resource costs relevant to DMO patients and the appropriate sources for treatment costs in the UK. Questions regarding resource use included the treatment setting (outpatient, daycase or inpatient care), frequency of visits for treatment and monitoring with laser photocoagulation and off-label products (e.g. IVTA, bevacizumab), the use of specialist equipment (e.g. OCT, fluorescein angiography), drug costs, and costs associated with the use of low vision aids, blindness registration and additional services provided for these patients. Iteration was not used in the collation of opinions. Given the small sample size, the responses were used to guide assumptions regarding resource use as opposed to generating average values. Intervention and comparators costs Please summarise the cost of each treatment in the following table. Cross-reference to other sections of the submission; for example, drugs costs should be cross-referenced to sections 1.10 and Provide a rationale for the choice of values used in the costeffectiveness model discussed in section Adapted from Pharmaceutical Benefits Advisory Committee (2008) Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra: Pharmaceutical Benefits Advisory Committee. Alimera Sciences, Inc. Submission of Evidence 109

111 Table B21: Unit costs associated with the technology in the economic model Items Intervention & comparator FAc implant Ref Laser Ref Technology cost 5,500 Sec. 0 1 Sec. 2.6 Admin. cost Total treatment cost 5, OPT used in both arms of FAME studies Items Avastin Lucentis IVTA Ozurdex Ref Technology cost Sec. Admin. cost Total treatment cost The full cost of the laser is assumed to be accounted for in the NHS Reference Cost for administration. No acquisition or maintenance cost has been applied 2 The average price of a single dose of intravitreal bevacizumab 1.25mg in a pre-filled syringe is estimated at 105. This was ascertained by communications with Moorfields Pharmaceuticals in 2010 (Horsley et al 2010) 3 Lucentis 10mg/ml, 1 x 0.23ml vial = (British National Formulary) - Lucentis single injection = 0.5mg/0.05ml (DRCR 2010) 4 Kenalog (IVTA) 40mg/ml vial, 5 x 1ml vial = 7.00 (British National Formulary) 5 Ozurdex 700microgram implant + applicator = 870 (British National Formulary) 6 BZ23Z - Vitreous Retinal Procedures Category 1: outpatient procedures Outpatient tariff for intravitreal injections was chosen based on NICE s position in the review of ranibizumab in DMO (NICE TA 237). Note that the model does not include a comparison between the FAc implant and Avastin, Lucentis, IVTA or Ozurdex. These treatments are included in Table B21 due to the fact that they were used off-protocol in a percentage of patients in both arms of the FAME studies (see section 5.2.1). The model incorporates the use of these therapies per the FAME studies over the first 36 months, and thus their costs must be considered (see section 6.2.7). Monitoring costs following treatment are presented in Table B22. Alimera Sciences, Inc. Submission of Evidence 110

112 Table B22: Monitoring costs following treatment Items Intravitreal injections Laser Ref. in submission Follow up attendance Section 2.8 OCT at follow-up Section 2.8 Fluorescein Angiography (FA) Section 2.8 VA assessment Section 2.8 Total monitoring cost B130-Ophthalmology Consultant Led: Follow up Attendance Non-Admitted Face to Face 2 OCT costs are applied only to non-administration visits. RA23Z-Diagnostic Imaging code: Ultrasound Scan less than 20 minutes 3 Patients receiving intravitreal injections do not receive a Fluorescein Angiography 4 FA costs are applied only to non-administration visits. In lieu of the diagnostic code, the cost of FA is sourced from Cruess et al. (2008). The prices have been adjusted for inflation using a 3% rate 5 A routine check that is included in the cost for outpatient procedure or attendance The unit costs selected for the model were considered to be the most relevant and comprehensive taking into account staff time, standard equipment, nonhigh cost drugs and tests (Department of Health NHS Costing Manual), whilst avoiding potential double counting. The use of these values, based on NHS Reference Costs, will also allow more explicit comparison with previous UK economic evaluations (see sections and 6.5.2). It is assumed that patients requiring treatment in both eyes will receive treatment in both eyes during the same year. The additional cost of bilateral treatment is limited to treatment (drug and procedure) costs and adverse event costs, on the basis that monitoring of both eyes can be carried at a single outpatient treatment. Health-state costs Please summarise, if appropriate, the costs included in each health state. Cross-reference to other sections of the submission for the resource costs. Provide a rationale for the choice of values used in the cost-effectiveness model. The health states should refer to the states in section The health state costs include the combined costs of treatment and monitoring (section 6.5.5). There are four health states in which patients VA Alimera Sciences, Inc. Submission of Evidence 111

113 is <35 ETDRS letters, representing severe vision loss (Meads and Hyde 2003; NICE TA 237). These four health states include additional costs associated with the management of patients with partial sight or blindness. This assumption is consistent with previous UK models that have estimated costs associated with severe visual loss (NICE TA 237). It is consistent to assume that these costs are incurred when one eye reaches health states <35 letter score. The method used to derive the estimate for the cost of severe visual loss is described in section Adverse-event costs Please summarise the costs for each adverse event listed in section 5.9 (Adverse events). These should include the costs of therapies identified in section 2.7. Cross-reference to other sections of the submission for the resource costs. Provide a rationale for the choice of values used in the cost-effectiveness model discussed in section The resources associated with adverse events were estimated based on previous NICE submissions in DMO. The costs associated with managing them are presented in Table B23. Alimera Sciences, Inc. Submission of Evidence 112

114 Table B23: List of adverse events and summary of costs included in the economic model Adverse events (per procedure/case) Cataract Items Value Reference in submission Phacoemulsification (daycase) Section 5.9 Antibiotics Section 2.7 NSAID Section 2.7 Total Endophthalmitis Non-elective admission for vitreous biopsy (1-5 days stay) 1,783 4 Section 5.9 Retinal detachment Non-elective admission for vitrectomy (overnight stay) 1,339 5 Section 5.9 Vitrectomy Non-elective admission for vitrectomy (overnight stay) 1,339 5 Section 5.9 Glaucoma Trabeculectomy 1,128 6 Section 5.9 Elevated IOP Beta blocker + prostaglandin Section BZ02Z- Phacoemulsification Cataract Extraction and Lens Implant: Daycase 2 Average cost of Ciloxan (ciprofloxacin), Exocin (ofloxacin), and Moxivig (moxifloxacin). (British National Formulary) Dosing assumptions and costs: Ciloxan 0.003, 5 ml bottle = 64 drops over 12 days. MSP = 4.11; Exocin 3 mg/ml, 5 ml bottle = 66 drops over 12 days. MSP = 1.63; Moxivig 5 mg/ml, 5 ml bottle = 42 drops over 14 days. MSP = 8.58 All MSP prices assume the cost of the entire bottle 3 Acular (ketorolac) 0.005, 5 ml bottle = 66 drops over 22 days. MSP = MSP price assumes the cost of the entire bottle (British National Formulary) 4 BZ23Z - Vitreous Retinal Procedures Category 1: Non-elective long stay where presentation/diagnosis & treatment occur same day (based on clinical expert advice) 5 BZ22Z - Vitreous Retinal Procedures Category 2: Non-elective short stay where presentation/diagnosis & treatment occur same day 6 BZ17Z - Glaucoma Procedure Category 3 7 Assumes chronic treatments are comprised of 50% beta blockers, 25% Xalatan, and 25% generic latanoprost. Dosing assumptions and costs: Betopic solution (betaxolol) 0.005, 1 drop daily for chronic treatment. 5 ml bottle = 60 drops over 60 days. MSP = 1.66 x 6 bottles per year = 9.98 annual cost x 50% of market basket = 4.99 annual cost of Betopic solution in the model. Xalatan (latanoprost) 50ug/mL, 1 drop daily for chronic treatment. 2.5 ml bottle = 60 drops over 60 days. MSP = x 6 bottles per year = annual cost x 25% of market basket = annual cost of Xalatan in the model. Latanoprost generic , 1 drop daily for chronic treatment. 2.5 ml bottle = 60 drops over 60 days. MSP = 3.27 x 6 bottles per year = annual cost x 25% of market basket = 4.91 annual cost of latanoprost in the model. Hence, annual cost for elevated IOP is = (British National Formulary) Miscellaneous costs Please describe any additional costs that have not been covered anywhere else (for example, PSS costs). If none, please state. Costs associated with partial sight and blindness (severe vision loss) due to DMO were sourced from the NICE review of Lucentis in DMO (Ranibizumab Manufacturer NICE submission; Royle et al. 2011; NICE TA237). The single technology appraisal of Lucentis in DMO estimated the cost of blindness to be 6,046 in the first year and 5,871 in the following years (Ranibizumab Manufacturer NICE submission; Royle et al. 2011). As it is difficult to discern which patients are in their first year of blindness and which are continually Alimera Sciences, Inc. Submission of Evidence 113

115 blind, the following year s figure of 5,871 was chosen. The cost was inflated to 6, using a 3% inflation rate compounded annually for 2 years. 6.6 Sensitivity analysis Has the uncertainty around structural assumptions been investigated? Provide details of how this was investigated, including a description of the alternative scenarios in the analysis. The precise definition of the product label and consequent scope of the decision problem for the FAc implant have restricted the uncertainty in this case, as there are a limited number of parameters that require assessment. However, scenario analysis was carried by both deterministic and probabilistic techniques as detailed in section and Health state utility values Whilst not derived from the identical patient population, it is believed that the utility values used in the model (section 6.4.9) are a reasonable representation of the impact of DMO on HRQL. Alternative sources of utility values were not tested in the model, although sensitivity analysis was performed on the utility weights used based on the confidence intervals published in the source document (Brown et al. 2000). Time horizon The time horizon of the model was chosen to best reflect the age of the patient population and their increased mortality. Any time horizon assessed is required to match the treatment cycle duration to ensure that the appropriate level of benefit was obtained for each once per cycle drug cost. Discount rate As the model has a time horizon of 15 years and a significant proportion of the benefit is accrued in the future, the rates of discount were investigated. Sensitivity of the model to the discount rate is explored in deterministic but not probabilistic sensitivity analyses. Alimera Sciences, Inc. Submission of Evidence 114

116 Number of FAc implant injections The base case assumes one injection of the FAc implant every 3 years. To investigate the impact of this assumption rather as compared to data from the FAME studies in which some patients received additional injections of FAc implant in years 2 and 3 of the first treatment cycle (section , 5.3), the assumptions of additional in cycle retreatment were explored. Change in VA after year 3 The extrapolation of efficacy and therefore generation or loss of HRQL after the 3 year trial period is potentially very impactful on the ICER and therefore investigated in the sensitivity review. Bilateral treatment Although in clinical trials it is standard to treat a single eye, it is likely that in practice the successful treatment of one eye would be followed by the treatment of the second, less severely impaired eye in patients experiencing DMO bilaterally. However, the HRQL benefit from vision gain in the second eye is not expected to be directly proportional to the HRQL benefit associated with improving vision in the first eye. The sensitivity analysis investigates the assumption made in the model surrounding the impact of bilateral treatment on HRQL. Alimera Sciences, Inc. Submission of Evidence 115

117 Table B24: Variables explored in sensitivity analysis Variable Base Case Scenarios tested Discount Rate 3.5% Number of FAc injections in treatment cycle Number of laser therapies Number of off protocol treatments Rate of FAc RTx after year 3 Change in VA after year 3 (per quarter) Relative risk of mortality Cost of blindness / inclusion of cost Cataract surgery, years 1-3 Cataract surgery after year 3 Cost of cataract surgery Glaucoma surgery post year 3 (control arms) Administration costs 1 in year one only Per FAME or DRCR trial Data from FAME Studies 0% and 6% for both costs and QALYs 1 in year 1, 0.21 in year 2 and 0.11 in year x rate of laser per year in trials xxxx% 25% - 100% FAc implant RTx: 5% of patients gain 5 and <10 letters; 95% remain the same as previous quarter NRTx: 3% of patients lose 5 and <10 letters; 95% remain the same as previous quarter Dependent on beta distribution for each variable x%, x% xxx x % xxxxxxxx xxx xxx xxx xxxxxxx xxx x% xxx x% xxxxxxxxx xxx xxxx 6295 annually Absence of cost or +/- 20% Data from FAME studies FAc RTx: rate from FAME study for phakic patients NRTx: 1.79% p.a 50% of rates observed in FAME studies 50% of rate observed in FAME studies 1.20% % p.a % p.a. 0.71% % Data obtained as described in section Utility weights Per Brown et al Bilateral treatment rate Bilateral benefit accrual xx% for FAc 35% for control 25% increase in total utilities Drop-out rate xxxx% x xxxx% +/- 20% of published costs 95% CI from Brown et al % and 35% for FAc 0% for control 0% and 50% increase in total utilities RTx: patients retreated after 3 years, NRTx: patients not retreated after 3 years with either control treatments or the FAc implant Alimera Sciences, Inc. Submission of Evidence 116

118 6.6.2 Which variables were subject to deterministic sensitivity analysis? How were they varied and what was the rationale for this? If any parameters or variables listed in section (Summary of selected values) were omitted from sensitivity analysis, please provide the rationale. All of the assumptions in Table B25 were tested using deterministic techniques. After initial sensitivity analysis of model parameters, those that did not contribute significantly to the uncertainty in the model were removed from the analysis. These included administration costs, adverse event costs (with the exception of cataract surgery) and monitoring Was PSA undertaken? If not, why not? If it was, the distributions and their sources should be clearly stated if different from those in section 6.3.6, including the derivation and value of priors. If any parameters or variables were omitted from sensitivity analysis, please provide the rationale for the omission(s). A PSA was undertaken using those assumptions included in the Table B25. Variables omitted from PSA included discount rate and those assumptions such as the decision on inclusion of bilateral treatment which are not suitable for such analysis due to their non-continuous nature. There were a number of additional variables from the clinical trial that were analyzed in preliminary deterministic sensitivity analysis and were deemed to be of minor contribution to uncertainty in the outcomes of the model; as a result, these have not been included in the PSA. Alimera Sciences, Inc. Submission of Evidence 117

119 Table B25: Assumptions included in PSA Assumption Post 3 year change in VA per quarter FAc implant RTx Post 3 year change in VA per quareter NRTx Utility weights Relative risk of death Bilateral treatment rates DMO treatments Adverse events Base case value 5% change in 5-10 letter improvement, balance stable 3% change in 5-10 letter decrement, balance stable Brown published data 2.45 xx% FAc 35% comparator Observed FAME trials data Observed FAME trials data Distribution used Triangle xxx% Triangle xxx% Normal distribution 95%CI Triangle Triangle 24-40% -FAc 29 40% Control Triangle Trial values +/- 20% Beta +/- 20% Beta 1.20% 2.49% Source of data FAME data set FAME data set, literature Brown et al 2000 NICE TA237 (Hirai 2009, Mulnier 2006) NICE TA237 (Hirai 2009, Mulnier 2006) Assumption FAME trial FAME trial Cataract surgery (post 3 years) 1.79% p.a. Klein 1995 Glaucoma Beta surgery 1.18% p.a. 0.71% 1.76% (post 3 years) Klein 1997 Costs associated Beta 6295 annually with blindness +/- 20% Royle et al NRTx: FAc implant and control arm patients who are not retreated after 3 years RTx: retreated Where suitably robust data were available, complex distribution were used. For example, the utility weights were obtained from a published source, and a normal distribution was used with 95% confidence intervals. PSA for FAME trial AE data employed beta distributions including sample sizes. However, where the basis for such distribution was not apparent, a simple triangle distribution was adopted. Alimera Sciences, Inc. Submission of Evidence 118

120 6.7 Results Clinical outcomes from the model For the outcomes highlighted in the decision problem (see section 4), please provide the corresponding outcomes from the model and compare them with clinically important outcomes such as those reported in clinical trials. Discuss reasons for any differences between modelled and observed results (for example, adjustment for cross-over). Please use the following table format for each comparator with relevant outcomes included. The data from the FAME studies was used directly to build the transitional matrix of the model and therefore a comparison is not applicable Please provide (if appropriate) the proportion of the cohort in the health state over time (Markov trace) for each state, supplying one for each comparator. Alimera Sciences, Inc. Submission of Evidence 119

121 xxx xxxxxxx xxx Alimera Sciences, Inc. Submission of Evidence 120

122 xxxxxxxxx xxxxxxxx xx xxxx (xxx) Alimera Sciences, Inc. Submission of Evidence 121

123 xxxxx xxxxxxxxxx xxx Alimera Sciences, Inc. Submission of Evidence 122

124 6.7.3 Please provide details of how the model assumes QALYs accrued over time. For example, Markov traces can be used to demonstrate QALYs accrued in each health state over time. The Markov model multiplies the BCVA utility value for each health state by the appropriate utility weight to accrue the utility value. These are summed per cycle and then discounted using the discount rate of 3.5% Please indicate the life years and QALYs accrued for each clinical outcome listed for each comparator. For outcomes that are a combination of other states, please present disaggregated results. For example: Table B26: Model outputs by clinical outcomes Outcome QALY Cost ( ) FAc implant therapy for 15 years ,269 Optimised standard of care ,939 Laser therapy using DRCR data ,283 QALY, quality-adjusted life year Alimera Sciences, Inc. Submission of Evidence 123

125 6.7.5 Please provide details of the disaggregated incremental QALYs and costs by health state, and of resource use predicted by the model by category of cost. Suggested formats are presented below. The design of the transition matrix in the model precludes disaggregation of the undiscounted QALYs from the data outputs. Therefore, the Table B27 illustrates the model change through percentage occupancy at each Markov health state Table B27: Summary of % time in each health state over 15 years Health state % in HS FAc arm % in HS Sham arm Change % in HS Laser only Change >75 xxxxx% xxxx% xxxxx% xxxx% xxxxx% xxxxx% xxxx% xxxx% xxxx% xxxx% xxxxx% xxxxx% xxxx% xxxxx% xxxx% xxxxx% xxxxx% xxxxx% xxxxx% xxxxx% xxxx% xxxxx% xxxxx% xxxxx% xxxxx% xxxx% xxxxx% xxxxx% xxxxx% xxxxx% xxxx% xxxx% xxxxx% xxxx% xxxxx% xxxx% xxxx% xxxxx% xxxx% xxxxx% xxxx% xxxx% xxxx% xxxx% xxxx% xxxx% xxxx% xxxxx% xxxx% xxxxx% xxxx% xxxx% xxxxx% xxxx% xxxxx% xxxx% xxxx% xxxxx% xxxx% xxxxx% <20 xxxx% xxxx% xxxx% xxxx% xxxx% QALY, quality-adjusted life year Adapted from Pharmaceutical Benefits Advisory Committee (2008) Guidelines for preparing submissions to the Pharmaceutical Benefits Advisory Committee (Version 4.3). Canberra: Pharmaceutical Benefits Advisory Committee In clinical practice for DMO, decisions to treat are not principally based on patients VA, and therefore, the model structure has not been developed to reflect differential treatment decisions by VA. Interventions are rather based on change in relative states i.e. losing a certain number of levels of vision. The only exception for this are the healthcare costs associated with blindness. These were not disaggregated by health state, but rather pooled as a cost for the health states 30-35, 25-30, and <20. Alimera Sciences, Inc. Submission of Evidence 124

126 Table B28: Change in healthcare cost of blindness Health states Change in cost ( ) FAc arm Change in cost ( ) FAc arm Change ( ) Change in cost ( ) laser arm Change ( ) <35 xxxxxxx xxxxxxxx xxxxxxxxx 2, , Table B29: Summary of predicted resource use by category of cost for FAc intervention against OSC Item Cost intervention (FAc implant) Cost comparator (OSC) Increment Absolute increment % absolute increment Technology cost 13, xxxxxx xxxxxx 76.7% Mean total drug treatment xx 0 xx xx 0% cost Mean laser treatment % cost Administrati on cost % Monitoring cost 4,919 5, % Costs of AEs % Healthcare cost of ,673-2,126 2, % blindness Total 20,270 8,939 xxxxxx xxxxxx 100% Cost of blindness is change in contribution to healthcare costs associated with VA score of <35 ETDRS letters Alimera Sciences, Inc. Submission of Evidence 125

127 Item Table B30: Summary of predicted resource use by category of cost for FAc intervention against laser only treatment Cost intervention (FAc implant) Cost comparator (Laser only) Increment Absolute increment % absolute increment Technology cost 13,961 0 xxxxxx xxxxxx 72.2 Mean total drug treatment xx 0 xx xx 0% cost Mean laser treatment cost % Administration cost % Monitoring 5,844 4,919 cost % Costs of AEs 1, % Healthcare cost of , , % blindness Total 20,270 9,283 xxxxxx xxxxxx 100% Cost of blindness is change in contribution to healthcare costs associated with VA score of <35 Alimera Sciences, Inc. Submission of Evidence 126

128 Base-case analysis Please present your results in the following table. List interventions and comparator(s) from least to most expensive and present ICERs in comparison with baseline (usually standard care) and then incremental analysis ranking technologies in terms of dominance and extended dominance. Change in mortality was not a measure in the FAME trial programme and therefore life years gained has not been included. Table B31: Base-case results Technologies Total costs ( ) Total QALYs FAc comparison with standard of care Incremental costs ( ) Incremental QALYs ICER ( ) versus baseline (QALYs) ICER ( ) incremental (QALYs) Optimized standard of care 8, na na na FAc implant 20, xxxxxx ,655 FAc comparison with laser therapy only Laser only 9, na na FAc implant 20, xxxxxx ,463 ICER, incremental cost-effectiveness ratio; QALYs, quality-adjusted life years Sensitivity analyses Please present results of deterministic sensitivity analysis. Consider the use of tornado diagrams. Alimera Sciences, Inc. Submission of Evidence 127

129 xxxxxx xxx: xxxxxxx xxxxxxx xx xxxxxxxxxxxxx xxxxxxxxxxx xxxxxxxxx xxx xxxxxxx xxx xxx Alimera Sciences, Inc. Submission of Evidence 128

130 xxxxxx xxx: xxxxxxx xxxxxxx xx xxxxxxxxxxxxx xxxxxxxxxxx xxxxxxxxx xxx xxxxxxx xxx xxxxx xxxx Alimera Sciences, Inc. Submission of Evidence 129

131 6.7.8 Please present the results of a PSA, and include scatter plots and cost-effectiveness acceptability curves. The analysis for the two comparisons are shown below along with associated scatter diagrams and cost effectiveness acceptability curves. At the national UK WTP threshold of acceptability of 30,000 per QALY there is a 64% and 70% chance of the FAc implant being a cost-effective replacement for optimised standard of care and laser therapy only, respectively. Table B32: Probability of cost effectiveness Intervention FAc implant vs. OSC FAc implant vs. laser only % chance of being cost effective WTP = 20k WTP = 30k WTP = 50k WTP = 100k xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxx xxxxxx xxx: xxxxxxxxxxxxxxxxxx xxxxx xxx xxx xxx xxxxxxx xx xxx (xxxxxx xxxxxxxxxxx) Alimera Sciences, Inc. Submission of Evidence 130

132 Xxxxxx xxx: xxxxxxxxxxxxxxxxxx xxxxxxxxxxxxx xxxxx xxx xxx xxx xxxxxxx xxx xxx xxxxxx xxx: xxxxxxxxxxxxxxxxxx xxxxx xxx xxx xxx xxxxxxx xxx xxxxx xxxx (xx,xxx xxxxxxxxxx) Alimera Sciences, Inc. Submission of Evidence 131

133 xxxxxx xxx: xxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxx xxxxx xxx xxx xxx xxxxxxx xxx xxxxx xxxx Alimera Sciences, Inc. Submission of Evidence 132

134 6.7.9 Please present the results of scenario analysis. Include details of structural sensitivity analysis. Table B33: FAc implant treatment compared with OSC Scenario Incremental Incremental cost ( ) benefit ICER ( ) Base case 11, ,655 Relative risk of mortality = 1.93 xxxxxx xxxxx 21,900 Bilateral 35% in FAc implant arm xxxxxx xxxxx 19,934 treatment None xxxxx xxxxx 14,630 Bilateral No utility gain xxxxxx xxxxx 20,358 benefit 50% utility gain xxxxxx xxxxx 25,535-20% xxxxxx xxxxx 23,505 Cost of +20% xxxxxx xxxxx 21,804 blindness Cost not included xxxxxx xxxxx 26,905 FAc implant 100% retreated after Y3 xxxxxx xxxxx 24,334 retreatment 25% retreated after Y3 xxxxx xxxxx 18,824 FAc = no change NRTx = 3% deteriorate 11, ,323 FAc = no change NRTx = no change 11, ,192 Variation in VA after 3 years* Drop-out rate Utility Weights Utility Weights FAc implant retreatment within 3 yrs Administration FAc = 5% improve NRTx = no change 11, ,580 FAc = 8% improve NRTx = no change 11, ,465 FAc = 8% improve NRTx = 3 % deteriorate 11, ,149 FAc = 8% improve NRTx = 5% deteriorate 11, ,090 0% xxxxxx xxxxx 30, % xxxxxx xxxxx Low end of CI for all xxxxxx xxxxx utility values 18,883 High end of CI for all xxxxxx xxxxx utility values 28,387 Yr 2 = Yr 3 = xxxxxx xxxxx 26,809 20% lower than xxxxxx costs published xxx 22,684 Administration 20% higher than xxxxxx costs published xxx 22,625 Discount rate 0% costs, 0% QALYs xxxxxx xxxxx 19,049 6% costs, 6% QALYs xxxxxx xxxxx 25,530 Cataract formation, 50% of FAME xxxxxx xxxxxx observed rates 22,192 *FAc: FAc implant retreated after 3 years; NRTx: FAc implant and OSC patients who are not retreated after 3 years Alimera Sciences, Inc. Submission of Evidence 133

135 xxxxxx xxx: xxx xxxxxxxx xxxxxxxx xx xxx xxxxxxx xxxxxxxx xxxx xxx Alimera Sciences, Inc. Submission of Evidence 134

136 Table B34: FAc implant treatment compared with laser only Scenario Incremental Incremental cost ( ) benefit ICER ( ) Base case xxxxx xxxxx 16,465 Relative risk of mortality = 1.93 xxxxx xxxxx 15,962 Bilateral 35% in FAc implant arm xxxxx xxxxx 15,173 treatment None xxxxx xxxxx 10,936 Bilateral No utility gain xxxxx xxxxx 15,467 benefit 50% utility gain xxxxx xxxxx 17,596-20% xxxxx xxxxx 17,243 Cost of +20% xxxxx xxxxx 15,682 blindness Cost not included xxxxx xxxxx 20,362 FAc implant 100% retreated after Y3 xxxxx xxxxx 18,706 retreatment 25% retreated after Y3 xxxxx xxxxx 12,041 FAc = no change NRTx = 3% deteriorate xxxxx xxxxx 20,785 FAc = no change xxxxx xxxxx NRTx = no change 25,245 Variation in VA after 3 years* Drop-out rate Utility Weights Utility Weights FAc implant retreatment within 3 yrs Administration FAc = 5% improve xxxxx xxxxx NRTx = no change 19,177 FAc = 8% improve xxxxx xxxxx NRTx = no change 17,306 FAc = 8% improve xxxxx xxxxx NRTx = 3% deteriorate 15,056 FAc = 8% improve xxxxx xxxxx NRTx = 5% deteriorate 13,909 0% xxxxx xxxxx 22, % xxxxx xxxxx 19,327 Low end of CI for all xxxxx xxxxx utility values 13,712 High end of CI for all xxxxx xxxxx utility values 20,649 Yr 2 = Yr 3 = xxxxx xxxxx 19,757 20% lower than xxxxx xxxxx costs published 16,538 Administration 20% higher than xxxxx xxxxx costs published 16,387 Discount rate 0% costs, 0% QALYs xxxxx xxxxx 14,066 6% costs, 6% QALYs xxxxx xxxxx 18,345 Cataract formation, 50% of FAME xxxxx xxxxx observed rates 16,116 *FAc: FAc implant retreated after 3 years; NRTx: FAc implant and laser monotherapy patients who are not retreated after 3 years Alimera Sciences, Inc. Submission of Evidence 135

137 xxxxxx xxx: xxx xxxxxxx xxxxxxxxx xxxxxxxx xxxx xxxxx xxxx What were the main findings of each of the sensitivity analyses? Relative Risk of mortality An increased risk of mortality would be expected to have an impact on the number of patients available for therapy; however, as there is not a differential rate between treatments and the on-going cost of treatment is small compared to the original treatment costs, the impact on the cost effectiveness of the FAc implant is relatively minor. Bilateral treatment The treatment of both eyes imposes a significant extra cost on the treatment regardless of the type of intervention. The costs have been assumed to be Alimera Sciences, Inc. Submission of Evidence 136

138 only those that would require intervention per eye and therefore costs such as monitoring have not been increased. As the treatment with the FAc implant is a major cost component, and the major cost expended after the first three years, the cost effectiveness is sensitive to additional treatment. Utility gain with bilateral treatment (bilateral benefit) The ICER of the FAc implant is sensitive to changes in the utility gain, but the response and direction is dependent on the relative rates of retreatment compared with the comparators in the model. However, even at an unlikely rate of 50% improvement in utilities with bilateral treatment, the FAc implant treatment option remains cost-effective against OSC at approx. 25,500 per QALY. Change in cost of blindness (inclusion) The inclusion of blindness as a cost is an important driver in the cost effectiveness valuation, as it is directly driven by the efficacy of the intervention. In the model, the cost included is the change in cost caused by the movement of patients in or out of the classification of blindness (i.e. <35 ETDRS letters) from the baseline of the study, rather than the overall cost to the NHS of supporting the patient. The level of cost is important, but within the +/- 20% parameters used in the sensitivity analysis, the FAc implant remains cost-effective over both comparators. The FAc implant also remains cost-effective if the contribution of change in cost of blindness is removed from the model. FAc implant retreatment As described in section , the rate of retreatment with the FAc implant after the first 3 year treatment cycle is based upon the rate of drop-out observed in the FAME studies and the percentage of patients showing response to therapy, defined by 5 letter gain in BCVA. To test the model for sensitivity to the percentage of patients retreated, a range of 25% to 100% was evaluated. The ICER was significantly impacted by reducing the rate of retreatment with the FAc implant to 25%, due to the reduction in the FAc implant drug costs which is a significant proportion of the overall treatment Alimera Sciences, Inc. Submission of Evidence 137

139 costs. The upper range (100% retreated) increased the ICER in both comparisons, but still to a point where the FAc implant remained cost effective at a WTP of 30,000 per QALY. Utility weights The model is quite sensitive to variation in the source utility weights as would be expected as it is a key component of the model. Due to the disposition of patients in the transition matrices of the two arms, when the source utility values are decreased to the higher limit (95% CI), the ICER values drop to 18,883 and 13,712 for FAc against OSC and FAC against laser only, respectively. Conversely, increasing the source utility values to the lower limit (95% CI) causes the ICERs to rise to 28,387 and 18,883 as the incremental utility drops between the FAc implant and comparator arms. An alternative source of suitable utilities would have provided more insight, as this would be expected to vary the exact range rather than just relative magnitude. However, this was not possible as discussed in section Drop-out rate While it is believed most appropriate to employ the rate of drop-out from the trials in the base case, scenario analyses have assessed the impact of no drop-out and half the rate of drop-out as compared to what was seen in the FAME trials. Both scenarios increase the ICER for the FAc implant compared to OSC and laser monotherapy; however, even in the scenario with greatest impact (0% drop-out in the comparison to OSC), the ICER increases to 30,102, which is only slightly above a WTP threshold of 30,000 per QALY. Variation in visual acuity after 3 years The long term extrapolation of efficacy is extremely important to the benefits accrued by treatments in the model, and this is accentuated as the cost is approximately constant whilst the benefit accrued changes. A number of scenarios have been tested because of the sensitivity of the model to assumptions around visual acuity after 3 years. In extreme scenarios, e.g. no change in VA from year 3 to year 15 in any group, the FAc implant becomes significantly less cost-effective at xxxxxxx Alimera Sciences, Inc. Submission of Evidence 138

140 per QALY against OSC and xxxxxxx against laser only. However, it is accepted that without treatment, DMO patients visual acuity will deteriorate (Davis 2009) and even in the most conservative scenario of no change in VA for patients retreated with the FAc implant and x% of patients in the control and non-fac implant retreated groups losing x -xx letters per quarter, the FAc implant is cost-effective at approximately xxxxxxx per QALY. It is not however felt that this conservative scenario is reasonable, as a gain in VA of 5 letters would be required to justify for retreatment with the FAc implant, and in all scenarios where VA is assumed to improve with retreatment of the FAc implant, the technology is cost-effective. Further, Figure B11 shows the underlying trajectory of change in VA between the FAc implant and OSC groups, which is predictive of long-term VA status. Based on the trajectory of change shown in this figure for the overall retinopathy state, one would predict that the FAc implant-treated group would continue to improve and those not treated with any therapy would continue to decline. FAc implant retreatment within the first treatment cycle For the reasons discussed in section , it is believed unlikely that patients would be retreated with the FAc implant within a 3 year treatment cycle. However, during discussions with NICE in the scoping process a sensitivity analysis around dosing assumptions was requested. Therefore, we have used the rates observed in the physician masked FAME trials for scenario boundaries. As expected, the FAc implant becomes less cost effective against both comparators, but remains below 30,000 per QALY with other base case parameters remaining constant. Discount rate The cost-effectiveness of the FAc implant against both comparators is sensitive to variations in discount rate due to the fact that costs of therapy are principally accrued early in the treatment process, and benefit is accrued in primarily in later phases. This is particularly the case for the FAc implant; however, the FAc implant remains cost-effective over the range of discounts tested for both comparators. A differential, higher rate of discounting for cost compared with benefit would reduce the cost-effectiveness of the intervention, Alimera Sciences, Inc. Submission of Evidence 139

141 whereas the opposite would improve it. Neither has been formally tested as they are deemed implausible. Administration costs Due to the small number of therapeutic interventions expected in this patient group and therefore included in the model, the ranges explored in the scenario analysis have little impact on the ICERs obtained. There is a slightly greater improvement in ICER for the FAc implant compared to laser treatment, as there are slightly higher number of uses of laser in this comparison (as compared to the OSC comparison). However, this variable would only become significant with greater use of short acting treatments or repeated therapy over the time horizon of the model, neither of which are plausible in this patient population (see section 6.3.8). Cataract Formation As discussed in section 5.9, the principal adverse event for the FAc implant is the formation of cataracts, which in the majority of cases requires cataract extraction and replacement of the affected lens. Whilst contributing a cost for its removal and a negative impact on utilities until it is removed, the influence of cataract and its extraction on cost-effectiveness is constrained by the fact that each eye can only develop cataract once, and that cataract removal cost is relatively low. The high rate of cataract formation in the study meant that this was used as a high value for the sensitivity range and a rate of 50% of that as a low to reflect the clinical impact. The ICERs calculated were approximately 22,200 for the FAc implant against OSC and 16,100 against laser. Other variables There were a number of additional variables from the clinical trial that were analyzed in preliminary deterministic sensitivity analysis and were deemed to be of minor contribution to uncertainty in the outcomes of the model. These have not been therefore not been reviewed separately here. Alimera Sciences, Inc. Submission of Evidence 140

142 What are the key drivers of the cost-effectiveness results? The key drivers of the cost effectiveness of the FAc implant are those associated with the additional use of the intervention over the model time horizon. This includes assumptions about the number of FAc implants patients receive within the first treatment cycle and the percentage of patients who are retreated with the FAc implant after the first treatment cycle, as well as the change in visual acuity after year 3 for both the FAc implant and the comparators. The latter is the key driver of the results, and as demonstrated in the sensitivity analysis (section ), the only scenario where the FAc implant is significantly not cost-effective is when there is no differential benefit assumed for the FAc implant with further injections over the time horizon. This scenario is counterintuitive, as there would not be continued use of the FAc implant in the absence of demonstrable efficacy or deterioration of patients VA. With any of the differential levels of VA change supported by trial data and literature, the intervention is cost effective at the generally established WTP levels. The decision on including bilateral treatment for a patient has a significant impact on the CE values as there is only a proportion (25%) of HRQOL gained for the additional FAc implant and associated costs. This means the ICER values go down significantly if only one eye is treated, to 14,600 for the FAc implant against OSC and 10,900 against laser only. 6.8 Validation Please describe the methods used to validate and quality assure the model. Provide references to the results produced and crossreference to evidence identified in the clinical, quality of life and resources sections. The economic model was developed by a third party agency and was reviewed by a health economist with responsibility within the organisation for quality assurance. The suitability and plausibility of the assumptions were approved by ophthalmology experts with the sponsoring organisation. Alimera Sciences, Inc. Submission of Evidence 141

143 The trial data from the FAME studies were validated against the clinical study report submitted to the European regulatory authorities. In addition, clinical experts have been involved in validating the model in two settings. First, as discussed in 6.3.5, three consultant ophthalmologists were informally engaged to provide guidance on the definition of optimised standard of care for this patient population. In addition, as discussed in section , an advisory board involving a consultant ophthalmologist and an HTA expert based in the UK was conducted to obtain input on the key assumptions employed in the economic model. After receiving an overview of the FAME clinical data and the structure of the economic model, clinical and HTA experts were asked to provide opinions on the following topics: Cost-effectiveness model structure, including cycle length, time horizon and data extrapolation Data comparison techniques, focusing on methodology for direct vs. indirect comparisons The way that effectiveness (utilities) is considered in the model Epidemiology and demographic data used to inform the model Existing use of approved and non-approved treatments for DMO patients Clinical expectations for positioning and use of the FAc implant Issues surrounding bilateral treatment (frequency, HRQL benefit) Use of the FAME data to support development of the model Appropriateness of sub-group analyses of patients and treatments Applicability of the FAc implant data to the real world (e.g. impact of adverse events, generation of utility data) Alimera Sciences, Inc. Submission of Evidence 142

144 Options for comparators that are included in the model, and which are most relevant for the FAc implant Costs associated with the DMO patient journey Review of specific cost and treatment assumptions and data sources referenced Iteration was not used in collating opinions. Expert opinions were employed to guide development of the model and validate key assumptions used throughout. 6.9 Subgroup analysis No subgroup analyses were conducted. This section is not applicable Interpretation of economic evidence Are the results from this economic evaluation consistent with the published economic literature? If not, why do the results from this evaluation differ, and why should the results in the submission be given more credence than those in the published literature? No comparable data were obtained from the literature reviews (see section 6.1.2) Is the economic evaluation relevant to all groups of patients who could potentially use the technology as identified in the decision problem in section 4? The economic evaluation is relevant to all groups of patients who could potentially use the FAc implant as identified in the decision problem What are the main strengths and weaknesses of the evaluation? How might these affect the interpretation of the results? Strengths As the available options for therapy in the target patient group are quite limited in the NHS treatment paradigm, the model allows a fairly simple description of added cost for added benefit that is intuitive to follow. Alimera Sciences, Inc. Submission of Evidence 143

145 Although the direct in trial measurement of patient quality of life is the accepted gold standard, the direct correlation of change in visual acuity and utility values in this model is an advantage. Issues with sensitivity and responsiveness are often the weakness of patient reported outcomes measures, and the sensitivity to change and responsiveness of the visual acuity scores mean that any change in the patient condition related to their sight is directly reflected in their health status. Whilst this does not take into account the patients experience as a whole, the benefit of the intervention and its significant AEs are sight related and therefore this is not perceived as an issue. The model is based on data from two phase III, randomized, multi-center, multi-continent clinical trials, that assessed the efficacy of the FAc implant in patients with DMO, and revealed a unique benefit to patients with chronic DMO. This dataset is the first such dataset in this uniquely identified, but, growing population with DMO. As such, while data from hundreds of subjects is available for inclusion, and while the effect was reproduced in both clinical trials, the true long-term impact on vision loss in such patients, previously having no treatment options, is most likely significantly underestimated. This is demonstrated by the significant cost-effectiveness associated with the comparison to laser monotherapy, in a population where 40% of the subjects were laser naive, and thus, clearly not in a chronic state in their disease. Additionally, while the incidence of cataract surgery is high in the FAc implanttreated group, and the associated costs have been included, it is important to note that diabetes is an established risk factor for cataract. Thus, even without the administration of the FAc implant, the cost associated with treating this condition would be borne by the health system. That the cost effectiveness is maintained over a range of sensitivity analysis including different vision benefits, different rates of bilateral treatment, differential rates of FAc implant retreatment, and as compared to a population expected to clearly benefit from laser monotherapy much more so than a chronic population that has received maximum laser, speak to the robustness Alimera Sciences, Inc. Submission of Evidence 144

146 of the results, and the significant benefit expected to be gained by the population with such a treatment option available. Weaknesses Utility data directly from the trial, or at least from the specific patient group (i.e. DMO patients), would remove a source of uncertainty from the model confirming the impact of visual acuity on HRQL on patients with chronic DMO. However, the use of the data from Brown et al was deemed the best possible scenario matching a condition with a similar level of impact and patient type. An alternative option would be the direct generation of utility values in the specific chronic and insufficiently responsive patient group. This was not a viable option available for resource and logistical reasons. The model is sensitive to the projected change in visual acuity beyond the data seen in the clinical trial, and a source of data to directly base these assumptions on rather than extrapolation of the trial data would significantly improve the model. However, as stated in , whilst the precise values may be debated, the presence of some differential change would be the basis for any clinical intervention decision and therefore the FAc implant should remain cost effective even if the precise value is not known What further analyses could be undertaken to enhance the robustness/completeness of the results? Further information on the precise current standard of care of patients with chronic DMO who are insufficiently responsive to treatment in the UK health care system would inform the model in the creation of an even more robust comparator arm to the FAc implant. As this standard of care may involve the use of interventions such as approved and unapproved short acting drugs, laser photocoagulation or support of living with blindness, it would be difficult and costly to gather definitive data. Alimera Sciences, Inc. Submission of Evidence 145

147 Section C Implementation 7 Assessment of factors relevant to the NHS and other parties 7.1 How many patients are eligible for treatment in England and Wales? Present results for the full marketing authorisation/ce marking and for any subgroups considered. Also present results for the subsequent 5 years. The eligible patient population for the FAc implant is established based on the methodology described in Section 2.2. Per the described methodology, Table C1 illustrates the estimated patient population from 2011 through Table C1: Estimated Patient Population ( ) UK Diabetic Population 3,064,000 3,095,000 3,126,000 3,157,000 3,188,000 3,219,000 England and Wales 90.0% 90.0% 90.0% 90.0% 90.0% 90.0% % Patients under active treatment 78.0% 78.0% 78.0% 78.0% 78.0% 78.0% % Existing Patients 75.0% 75.0% 75.0% 75.0% 75.0% 75.0% % pop with CSMO 6.6% 6.6% 6.6% 6.6% 6.6% 6.6% Insufficiently responsive 50.0% 50.0% 50.0% 50.0% 50.0% 50.0% Total 53,235 53,774 54,313 54,851 55,390 55, What assumption(s) were made about current treatment options and uptake of technologies? Budget impact calculations assume that current treatment of chronic DMO patients who are insufficiently responsive to available therapies is restricted to quarterly monitoring and registration for low vision aids and blindness services when applicable. Some Trusts may additionally provide funding for intravitreal injections of unlicensed products such as IVTA and bevacizumab for this group of patients. However, in the absence of data on the real-world use of Alimera Sciences, Inc. Submission of Evidence 146

148 these treatments, it is assumed that the volume is negligible and they are therefore excluded from budget impact calculations. It is assumed that a portion of incident cases are treated with the FAc implant as last-line therapy. The remainder receive no treatment. It is also assumed that the FAc implant treatment will be administered at baseline as monotherapy. Budget impact calculations are based on the assumptions regarding resource use and treatment duration described in section 6.5. In addition, assumptions about the rate of bilateral treatment mirror those made in the costeffectiveness model for FAc implant treated and non-fac implant treated patients (see section ). 7.3 What assumption(s) were made about market share (when relevant)? Market shares for the FAc implant are based on market research commissioned by Alimera Sciences and conducted by SciMedica Group. Per the defined population of patients with chronic CSMO considered insufficiently responsive, we estimated market shares of xx%, xx%, and xx% in years 1, 2 and 3 respectively (SciMedica Group 2011a). Please note, per the budget impact model, the shares described represent cumulative shares. 7.4 In addition to technology costs, please consider other significant costs associated with treatment that may be of interest to commissioners (for example, procedure codes and programme budget planning). The primary costs of treatment are presented in section In addition, the costs of managing adverse events arising from treatment (Table B23 in section 6.5.7) and the cost savings from vision loss (Section 6.5.8) avoided will be different depending upon the treatment administered. It is assumed that patients requiring treatment in both eyes, will receive treatment in both eyes during the same year, The additional cost of bilateral treatment is limited to treatment (drug and procedure) costs and adverse event costs, on the Alimera Sciences, Inc. Submission of Evidence 147

149 basis that monitoring of both eyes can be carried at a single outpatient treatment. Assumptions about the rate of bilateral treatment in the budget impact calculations 7.5 What unit costs were assumed? How were these calculated? If unit costs used in health economic modelling were not based on national reference costs or the PbR tariff, which HRGs reflected activity? Unit costs are the same as in the cost-effectiveness analysis (sections 6.5.5, and 6.5.7). The costs of treatment and management of adverse events are based on national reference costs. The calculation of costs associated with severe vision loss is presented in section Were there any estimates of resource savings? If so, what were they? Patients treated with the FAc implant experience better vision than those who are not treated. Thus, although the cost of treatment and consultations for patients receiving the FAc implant exceeds the treatment and consultations costs for non-treated patients, this is partly offset by resource savings from vision loss avoided; greater savings are made through the FAc implant treatment than through no treatment. Over a short time horizon of 5 years, the cumulative saving of vision loss avoided through the FAc implant is xxxxxx per patient. 7.7 What is the estimated annual budget impact for the NHS in England and Wales? Budget impact is calculated relative to the FAc implant target population, i.e. chronic DMO insufficiently responsive to available therapies (Table C2, C3). Note that calculations assess the impact of three years of uptake over five years to capture the benefit of the FAc implant relative to the three year treatment duration. Alimera Sciences, Inc. Submission of Evidence 148

150 Table C2: Total budget impact for non-fac implant market basket Year 1 Year 2 Year 3 Year 4 Year 5 Total DMO drugs DMO procedures Adverse events Monitoring 15,544,757 15,544,757 15,544,757 15,544,757 15,544,757 77,723,783 Blindness 51,461,957 94,346, ,231, ,231, ,231, ,504,533 Total 67,006, ,891, ,776, ,776, ,776, ,228,317 Table C3: Total budget for FAc implant treatment of DMO Year 1 Year 2 Year 3 Year 4 Year 5 Total DMO drugs xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxx xx xx xxxxxxxxxxx DMO procedures Adverse events 1,029,108 2,058,216 2,264, ,351, ,294 3,367,348 6,687,306 7,078,627 6,336,407 24,172,983 Monitoring 16,673,349 18,930,532 21,413,435 21,413,435 21,413,435 99,844,185 Blindness 50,165,877 76,823,168 83,427,831 82,734,810 81,887, ,039,471 Total xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxx xxxxxxxxxxx 7.8 Are there any other opportunities for resource savings or redirection of resources that it has not been possible to quantify? As shown in section 7.6 there will be further savings arising from vision loss avoided after year 5 that would only be fully reflected in a longer term analysis. However, even in the short term there are other costs to patients, care givers and society that have been excluded from the analysis to ensure consistency with the NICE reference case. The costs of partial sight and blindness, taken from a societal perspective, would include the costs of lost productivity, tax relief and benefits. These costs would be accrued to patients themselves but also to unpaid care givers; family members and friends who provide care to patients with DMO. In addition, the indirect costs of unpaid caregiving to DMO patients who need personal care but do not access government funded health or social care is not included in this analysis. The indirect costs of blindness in Alimera Sciences, Inc. Submission of Evidence 149

151 UK adults has been estimated as being more than double the direct cost (direct health care system costs of 2.14 billion vs. indirect costs amount to 4.34 billion) (Minassian and Reidy 2009), which suggests that the FAc implant therapy would be highly cost effective from a societal perspective. Alimera Sciences, Inc. Submission of Evidence 150

152 8 References Aspinall PA, Hill AR, Dhillon B, Armbrecht AM, Nelson P, Lumsden C, Farini- Hudson E, Brice R, Vickers A, Buchholz P. Quality of life and relative importance: a comparison of time trade-off and conjoint analysis methods in patients with age-related macular degeneration. British Journal of Ophthalmology : Print. Barber A, Gardner T and Abcouwer S. The Significant of Vascular and Neural Apoptosis to the Pathology of Diabetic Retinopathy. Investigative Ophthalmology & Visual Science : Print. British National Formulary. Joint Formulary Committee. London: BMJ Group and Pharmaceutical Press. n.p. Web. Accessed on 20 Mar < Brown GC. Vision and Quality-of-Life. Transactions of the American Ophthalmological Society : Print. Brown GC, Sharma S, Brown MM, Kistler J. Utility Values and Age-related Macular Degeneration. Archives of Ophthalmology (1): Print. Brown MM, Brown GC, Sharma S, Landy J, Bakal J. Quality of Life With Visual Acuity Loss From Diabetic Retinopathy and Age-Related Macular Degeneration. Archives of Ophthalmology (4): Print. Brucklacher RM, Patel K, VanGuilder HD, Bixler GV, Barber AJ, Antonetti DA, Lin CM, LaNoue K, Gardner TW, Bronson S and Freeman W. Whole genome assessment of the retinal response to diabetes reveals a progressive neurovascular inflammatory response. BioMed Central Medical Genomics :26 doi: / Print. Ciulla TA, Amador AG and Zinman B. Diabetic Retinopathy and Diabetic Macular Edema: Pathophysiology, screening, and novel therapies. Diabetes Care : Print. Alimera Sciences, Inc. Submission of Evidence 151

153 Clarke PM, Cull CA, Simon J, Holman RR. Assessing the Impact of Visual Acuity on Quality of Life in Individuals With Type 2 Diabetes Using the Short Form-36. Diabetes Care (7): Print. Colquitt, J.L., Jones, J., Tan, S.C., Takeda, A.L., Clegg, A.J. and Price, A. Ranibizumab and pegaptanib for the treatment of age-related macular degeneration: a systematic review and economic evaluation. Health Technology Assessment (16): doi: /hta Web. Coscas G, Cunha-Vaz J, Loewenstein A, Soubrane G. Macular Edema: A Practical Approach (Developments in Ophthalmology). Switzerland: S. Karger AG :(1): 80. Print. Cruess AF, Zlateva G, Xu X, Soubrane G, Pauleikhoff D, Lotery A, Mones J, Buggage R, Schaefer C, Knight T, Goss TF. Economic Burden of Bilateral Neovascular Age-Related Macular Degeneration: Multi-Country Observational Study. Pharmacoeconomics (1): Print. Davis MD, Sheetz MJ, Aiello LP, Milton R, Danis R, Zhi X, Girach A, Jimenez M and Vignati L. Effect of Ruboxistaurin on the Visual Acuity Decline Associated with Long-standing Diabetic Macular Edema. Investigative Ophthalmology & Visual Science :1-4. Print. Diabetes.co.uk. n.p. Web. Accessed 21 Mar < Diabetic Retinopathy Clinical Research Network (DRCR). A Randomized Trial Comparing Intravitreal Triamcinolone Acetonide and Focal/Grid Photocoagulation for Diabetic Macular Edema, Ophthalmology : Print. Diabetic Retinopathy Clinical Research Network (DRCR). Randomized Trial Evaluating Ranibizumab Plus Prompt or Deferred Laser or Triamcinolone Acetonide Plus Prompt Laser for Diabetic Macular Edema. Ophthalmology : Print. Alimera Sciences, Inc. Submission of Evidence 152

154 Duane TD, Tasman W, Jaeger EA, eds. Duane s Clinical Ophthalmology, Vol. 3, Chapter 51. Philadelphia: J.B. Lippincott Print. Early Treatment Diabetic Retinopathy Study (ETDRS) Research Group. Photocoagulation for Diabetic Macular Edema: Early Treatment Diabetic Retinopathy Study Report No. 4. International Ophthalmology Clinics :( ). Print. Earnshaw SR, Moride Y, Rochon S. Cost-Effectiveness of Pegaptanib Compared to Photodynamic Therapy with Verteporfin and to Standard Care in the Treatment of Subfoveal Wet Age-Related Macular Degeneration in Canada. Clinical Therapeutics (9): , Print. Erhlich R, Harris A, Ciulla TA, Kheradiya N, Winston DM and Wirostko B. Diabetic macular oedema: physical, physiological and molecular factors contribute to this pathological process. Acta Ophthalmologica : Print. Espallargues M, Czoski-Murray CJ, Bansback NJ, Carlton J, Lewis GM, Hughes LA, Brand C, Brazier J. The Impact of Age-Related Macular Degeneration on Health Status Utility Values. Investigative Ophthalmology & Visual Science (11): Print. FAME A (C A) Clinical Study Report. Alimera Sciences, Inc Data on File. FAME B (C B) Clinical Study Report. Alimera Sciences, Inc Data on File. Ferris FL III and Patz A. Macular Edema. A Complication of Diabetic Retinopathy. Survey of Ophthalmology Suppl: Print. Gillies MC, Sutter FKP, Simpson JM, Larsson J, Ali H, Zhu M. Intravitreal Triamcinolone for Refractory Diabetic Macular Edema. Ophthalmology : Print. Alimera Sciences, Inc. Submission of Evidence 153

155 Hernandez-Pastor LJ, Ortega A, Garcia-Layana A, Giraldez J. Cost- Effectiveness of Ranibizumab Compared with Photodynamic Treatment of Neovascular Age-Related Macular Degeneration. Clinical Therapeutics (12): Print. Hirai FE, Knudtson MS, Klein BE, Klein R,. Clinically Significant Macular Edema and Survival in Type 1 and Type 2 Diabetes. American Journal of Ophthalmology (4): Print. Hopley C, Salkeld G, Mitchell P. Cost utility of photodynamic therapy for predominantly classic neovascular age-related macular degeneration. British Journal of Ophthalmology (8): Print. Horsley W. Anti-vascular endothelial growth factor therapies (bevacizumab and ranibizumab) in the management of macular oedema secondary to retinal vein occlusions. NETAG (North East Treatment Advisory Group) Web. Accessed 20 Mar < reports/avegf%20for%20rvo%20secondary%20to%20mo%20- %20NETAG%20appraisal%20report%20-Jan2011-%20- %20web%20version.pdf/view>. Integrated Summary of Safety (ISS) FAME A and FAME B Studies. Alimera Sciences, Inc Data on file. Javitt JC, Zlateva GP, Earnshaw SR. Cost-Effectiveness Model for Neovascular Age-Related Macular Degeneration: Comparing Early and Late Treatment with Pegaptanib Sodium Based on Visual Acuity. Value in Health (4): Print. Kantar Health. Epidemiology of Diabetic Macular Edema. Alimera Sciences, Inc. Research Data on File. Klein BE, Klein R, Moss SE. Incidence of Cataract Surgery in the Wisconsin Epidemologic Study of Diabetic Retinopathy. American Journal of Ophthalmology (3): Print. Alimera Sciences, Inc. Submission of Evidence 154

156 Klein BE, Klein R, Moss SE. Incidence of self reported glaucoma in people with diabetes meilitus. British Journal of Ophthalmology (3): Print. Lloyd A, Nafees B, Gavriel S, Rousculp MD, Boye KS, Ahmad A. Health utility values associated with diabetic retinopathy. Diabetic Medicine (5) : Print. Lucentis (ranibizumab) for the treatment of visual impairment due to diabetic macular oedema (DMO). Single technology appraisal (STA). Manufacturer Submission Web. Accessed 20 Mar < Lucentis Summary of Product Characteristics. 18 Jan Web. Accessed 05 Apr < +solution+for+injection/>. Mazhar K, Varma R, Choudbury F, McKean-Cowdin R, Shtir CJ, Azen SP on behalf of the Los Angeles Latino Eye Study Group. Severity of Diabetic Retinopathy and Health-Related Quality of Life Ophthalmology. 118: Print. Meads C and Hyde C. What is the cost of blindness? British Journal of Ophthalmology : Print. Michaelides M, Kaines A, Hamilton RD, Fraser-Bell S, Rajendram R, Quhill F, Boos CJ, Xing W, Egan C, Peto T, Bunce C, Leslie RD, Hykin PG. A Prospective Randomized Trial of Intravitreal Bevacizumab or Laser Therapy in the Management of Diabetic Macular Edema (BOLT study) 12-Month Data: Report 2. Ophthalmology (6): e2. Print. Minassian D and Reidy A. Future Sight Loss UK (2): An epidemiological and economic model for sight loss in the decade 2010 to Royal National Institute of Blind People Web. Accessed 20 March Alimera Sciences, Inc. Submission of Evidence 155

157 Mulnier HE, Seaman HE, Raleigh VS, Soedamah-Muthu SS, Colhoun HM, Lawrenson RA. Mortality in people with Type 2 diabetes in the UK. Diabetes Medicine : Print. National Institute for Health and Clinical Excellence - Final appraisal determination - Ranibizumab or the treatment of diabetic macular oedema. National Institute for Health and Clinical Excellence Web. Accessed 20 Mar < pdf/english>. NHS Costing Manual Department of Health April Web. Accessed 20 Mar Web. Accessed 20 Mar < digitalasset/dh_ pdf>. NHS Payment by Results National Tariff Information. NHS Information Centre Web. Accessed 20 Mar < Raftery J, Clegg A, Jones J, Tan SC, Lotery A. Ranibizumab (Lucentis) versus bevacizumab (Avastin): modeling cost effectiveness. British Journal of Ophthalmology : Print. Royle P, Cummins E, Henderson R, Lois N, Shyangdan D, Waugh N. Evidence Review: Ranibizumab for the treatment of diabetic macular oedemal. Aberdeen Health Technology Assessment Group Web. Accessed 20 Mar < Sahel JA, Bandello F, Augustin A, Maurel F, Negrini C, Berdeeaux GH, and the MICMAC Study Group. Health-Related Quality of Life and Utility in Patients with Age-Related Macular Degeneration. Archives of Ophthalmology (7): Print. Alimera Sciences, Inc. Submission of Evidence 156

158 SciMedica Group. ILUVIEN Patient Identification Study. Alimera Sciences, Inc. Research. 2011a. Data on file. SciMedica Group. ILUVIEN Patient xxxxxxx Study. Alimera Sciences, Inc. Research. 2011b. Data on file. Shah VA, Gupta SK, Shah KV, Vinjamaram S, Chalam KV. TTO utility scores measure quality of life in patients with visual morbidity due to diabetic retinopathy or ARMD. Ophthalmic Epidemiology (1): Print. Sharma S, Brown GC, Brown MM, Hollands H, Shah GK. The costeffectiveness of grid laser photocoagulation for the treatment of diabetic macular edema: results of a patient-based cost-utility analysis. Current Opinion in Ophthalmology. 2000b. 11(3): Print. Sharma S, Brown GC, Brown MM, Hollands H, Shah GK. The Cost- Effectiveness of Photodynamic Therapy for Fellow Eyes with Subfoveal Choroidal Neovascularization Secondary to Age-related Macular Degeneration. Ophthalmology. 2001a. 108(11): Print. Sharma S, Brown GC, Brown MM, Shah GK, Snow K, Brown H, Hollands H. Converting visual acuity to utilities. Canadian Journal of Ophthalmology. 2000a. 35(5): Print. Sharma S, Hollands H, Brown GC, Brown MM, Shah GK, Sharma SM. The cost-effectiveness of early vitrectomy for the treatment of vitreous hemorrhage in diabetic retinopathy. Current Opinions in Ophthalmology. 2001b. 12(3): Print. Sharma S, Oliver-Fernandez A, Bakal J, Hollands H, Brown GC, Brown MM. Utilities associated with diabetic retinopathy: results from as Canadian sample. British Journal of Ophthalmology (87): Print. Smith DH, Johnson ES, Russell A, Hazlehurst B, Muraki C, Nichols GA, Oglesby, A, Betz-Brown J. Lower visual acuity predicts worse utility values among patients with type 2 diabetes. Quality of Life Research : Print. Alimera Sciences, Inc. Submission of Evidence 157

159 Tranos PG, Topouzis F, Stangos NT, Dimitrakos S, Economidis P, Harris M, Coleman A. Effect of laser photocoagulation treatment for diabetic macular oedema on patient s vision-related quality of life. Current Eye Research (1):41-9. Print. Wasserman Y. Physicians' Fee Reference th Edition. Wisconsin: Wasserman Medical Publishers, Ltd Print. Whitmire W, Al-Gayyar M MH, Abdeksaid M, Yousufzai BK and El-Remessy AB. Alteration of growth factors and neuronal death retinopathy: what we have learned so far. Molecular Vision. 2011; 17: Print. Zeng H, Green WR and Tso MOM. Microglial Activation in Human Diabetic Retinopathy. Archives of Ophthalmology. 2008; 126: Print. Alimera Sciences, Inc. Submission of Evidence 158

160 9 Appendices 9.1 Appendix 1 Alimera Sciences, Inc. Submission of Evidence 159

161 9.1.1 RMS Final Assessment Report Decentralised Procedure RMS Day 210 Final Assessment Report OVERVIEW AND LIST OF QUESTIONS Iluvien Fluocinolone Acetonide UK/H/3011/01/DC Applicant: Campharm Limited Reference Member State: UK Start of the procedure: 6 September 2010 Date of this report: 27 February 2012 Alimera Sciences, Inc. Submission of Evidence 160

162 TABLE OF CONTENTS I RECOMMENDATION 164 II EXECUTIVE SUMMARY 164 II.1 PROBLEM STATEMENT 164 II.2 ABOUT THE PRODUCT 164 II.3 GENERAL COMMENTS ON THE SUBMITTED DOSSIER 165 II.4 GENERAL COMMENTS ON COMPLIANCE WITH GMP, GLP, GCP AND AGREED ETHICAL PRINCIPLES. 165 III SCIENTIFIC OVERVIEW AND DISCUSSION 166 III.1 QUALITY ASPECTS 166 III.2 NON CLINICAL ASPECTS 167 III.3 CLINICAL ASPECTS 168 IV BENEFIT RISK ASSESSMENT 173 V LIST OF QUESTIONS as proposed by RMS 174 V.1 QUALITY ASPECTS 174 V.2 NON CLINICAL ASPECTS 174 V.3 CLINICAL ASPECTS 174 VI RECOMMENDED CONDITIONS FOR MARKETING AUTHORISATION AND PRODUCT INFORMATION 175 VI.1 LEGAL STATUS 175 VI.2 PROPOSED LIST OF FOLLOW-UP MEASURES AND SPECIFIC OBLIGATIONS IN CASE OF A POSITIVE BENEFIT RISK ASSESSMENT 175 VI.3 SUMMARY OF PRODUCT CHARACTERISTICS (SPC) 175 VI.4 PACKAGE LEAFLET (PL) AND USER TESTING 175 VI.5 LABELLING 175 Alimera Sciences, Inc. Submission of Evidence 161

163 ADMINISTRATIVE INFORMATION Proposed name of the medicinal product in the RMS INN (or common name) of the active substance(s): Pharmaco-therapeutic group (ATC Code): Pharmaceutical form(s) and strength(s): Reference Number for the Decentralised Procedure Reference Member State: Member States concerned: Iluvien Fluocinolone acetonide S01BA15 Intravitreal implant, 0.19 mg UK/H/3011/01/DC UK AT, PT, ES, FR, DE, IT Applicant (name and address) Campharm Limited, 671 Chemin De Dayan, Callian, F , France Names and addresses of manufacturers of dosage form Names and addresses of manufacturers responsible for batch release in the EEA RMS contact person: Names of the assessors: Alliance Medical Products Inc 9342 Jeronimo Road, Irvine CA92618 USA Brecon Pharmaceuticals Ltd Wye Valley Business Park, Hereford HR3 5PG, UK Name MR/DC Team Tel: Quality: Name(s) Dr Valentine Ibekwe Tel: Non-clinical: Name(s) Dr Rachel Hawkins Tel: Clinical : Name(s) Dr David Silverman Tel: david.silverman@mhra.gsi.gov.uk Statistical Name(s) Dr Andrew Thomson Tel: andrew.thomson@mhra.gsi.gov.uk RMP Name(s) Dr Bridget King Tel: bridget.king@mhra.gsi.gov.uk Alimera Sciences, Inc. Submission of Evidence 162

164 Alimera Sciences, Inc. Submission of Evidence 163

165 I RECOMMENDATION Based on the review of the data on quality, safety and efficacy, the RMS considers that the application for Iluvien, in the treatment of vision impairment associated with chronic diabetic macular oedema, considered insufficiently responsive to available therapy is approvable provided that the applicant complies with the recommendations for product information and commits to perform a number of specific obligations to be reported back to the RMS and CMS within predefined timeframes. A list of such measures is in section VI of this report. II II.1 EXECUTIVE SUMMARY Problem statement Diabetic macular oedema is the result of retinal microvascular changes that occur in patients with diabetes. Thickening of the basement membrane and reduction in the number of pericytes is believed to lead to increased permeability and incompetence of retinal vasculature. This compromise of the blood-retinal barrier leads to the leakage of plasma constituents into the surrounding retina, resulting in macular thickening due to fluid accumulation, resulting in significant disturbances in visual acuity. Prolonged oedema can cause irreversible damage resulting in permanent visual loss. Current proven treatments for diabetic macular oedema (DME) include laser therapy and tight diabetic control; at present there are no approved pharmacologic therapies for DME, although centralised Type II variations are ongoing to extend the indications of anti-vegf drugs, currently licensed for age-related macular degeneration, to include treatment of DME. In addition, Ozurdex (Allergan Pharmaceuticals Ireland), a sustained release, biodegradable dexamethasone implant was licensed earlier this year for treatment of macular oedema following central or branch retinal vein occlusion, with an effect lasting up to 3 months. II.2 About the product Corticosteroids, such as fluocinolone acetonide (FA), are postulated to reduce vascular endothelial growth factor (VEGF) and reduce the inflammatory aspects of DME. Iluvien is a sustained-release intravitreal drug delivery system that releases submicrogram levels of fluocinolone acetonide (FA), a glucocorticoid, in the vitreous humour for months. It is inserted into the eye via the pars plana through a 25-gauge needle attached to an inserter device. The drug product proposed for the market (0.25 μg/day initial release) consists of a drug core containing FA in a polyvinyl alcohol (PVA) matrix encased in a polyimide tube. One end of the tube is coated with PVA and the other end with silicone adhesive. The PVA end is permeable and controls the release of FA into the vitreous chamber. In the clinical studies, a higher dose (0.5 μg/day initial release) was also used. This dose is identical to the low dose with the exception that the silicon adhesive is not utilised and drug is released through both ends of the tube. FA, the drug substance in Iluvien, is also the active agent in Retisert (Bausch & Lomb, Rochester, NY), an intraocular delivery implant approved in 2005 by the Food and Drug Administration (FDA) (but not elsewhere) for treatment of non-infectious posterior uveitis. The insertion procedure is less invasive as it does not require surgery and allows placement of the insert in a more posterior position than Retisert (which is sutured at the pars plana). Due to the small gauge needle used, the insertion site is a self-sealing wound. It was also anticipated that the lower exposure of FA in the anterior segment would provide a better safety profile while maintaining efficacy. Alimera Sciences, Inc. Submission of Evidence 164

166 The claimed indication is for the treatment of chronic diabetic macular oedema (in adults aged 18 years and over), considered insufficiently responsive to available therapies. Insertion of the implant should be performed under sterile conditions, and an additional implant may be administered after 12 months if the patient experiences decreased vision or an increase in retinal thickness secondary to recurrent or worsening diabetic macular oedema. II.3 General comments on the submitted dossier This is a decentralised application, submitted under Article 8(3) of Directive 2001/83/EC, as amended, known active substance. With the UK as RMS, the applicant also seeks a licence in Austria, Portugal, Spain, France, Germany, and Italy. A risk management plan has been submitted as part of this application. The applicant sought scientific advice from several National regulatory agencies in the EU. UK: May 2006, initial protocol review; July 2008, statistical analysis plan; March 2010, pre-marketing authorisation application meeting Spain: June 2006, scientific advice meeting Portugal: August 2006, protocol and risk management program discussion Germany: December 2006, scientific advice meeting France: April 2006, clinical program discussion A rabbit PK study was performed in conjunction with the 24-month ocular toxicity study. This study evaluated FA levels in plasma and ocular tissues at multiple time points over 24 months. The pharmacokinetics of FA release in human subjects with DME was assessed in a 36-month human PK study involving plasma and aqueous humour sampling. Two phase 3 clinical studies were performed under an identical protocol to determine the safety and efficacy of 0.2 μg/day and 0.5 μg/day FA inserts in the treatment of DME. The submitted dossier includes reports for the primary readout of these studies (Month 24) and for the Month 18 readout of the human PK study. All of the studies are ongoing; final results will be reported when all subjects have completed Month 36. A product-specific waiver was granted by the Paediatric Committee in accordance with Article 13 of Regulation (EC) No 1901/2006 for the paediatric population to less than 18 years of age, on the grounds that the disease for which the specific medicinal product is intended does not occur in the paediatric subset of less than 12 years of age, and in the subset from 12 years to less than 18 years of age the product does not represent a significant therapeutic benefit as clinical studies are not feasible. II.4 General comments on compliance with GMP, GLP, GCP and agreed ethical principles. The manufacturer responsible for QC and batch release, Brecon Pharmaceuticals Limited holds a Manufacturing and Importers Licence (MIA 11724) issued by MHRA (UK). The QP for Brecon Pharmaceuticals Ltd (batch release site) has also provided a declaration of compliance of the API manufacturer (Farmabios Italy) with EU-GMP requirements. The QPD is based on a site audit conducted by the QP for Brecon Pharmaceuticals Ltd. The drug product manufacturer, Alliance Medicinal Products Inc. has been inspected by the MHRA (UK) inspectorate in December 2010 and a copy of the GMP certificate is provided. Alimera Sciences, Inc. Submission of Evidence 165

167 GCP No issues regarding GCP or GLP aspects have been identified during the review of this dossier. III SCIENTIFIC OVERVIEW AND DISCUSSION III.1 Quality aspects Drug Substance The EDQM Certificate of Suitability procedure has been used to submit data for fluocinolone acetonide drug substance. The CEP states a retest period of 5 years and limits for supplementary tests for related substances and residual solvents. The drug substance data provided by the applicant is considered generally satisfactory. Drug Product It is acknowledged that the implant in applicator drug product is of an innovative design and presents a practical and effective means of implant administration. The development package is comprehensive. The development of the product is based on other similar drug products developed by psivida and Alliance Medical Products Inc. The manufacturing process was initially developed at psivida through to Phase 3 clinical studies and licensed out to Alimera Sciences. Further optimisation of the manufacturing process was then performed, while keeping it essentially unchanged, first at Avail Medical Products Inc. and then at the proposed commercial manufacturer, Alliance Medical Products Inc. Sivida Corp. and Alliance Medical Products Inc. has experience in development of similar dosage forms (Vitrasert and Retisert ). The only excipients are polyvinyl alcohol and water for injections. These are adequately controlled. PVA is commonly used in ophthalmic preparation as ocular lubricant. The safety of the components of the delivery system is assured as they are known materials. The Polyimide tube is used in intraocular lens and its biocompatibility has been characterised. The clinical and pre-clinical study batches were manufactured at psivida using active ingredient from another source. The drug release rate from current registration batches have been compared with the clinical study batches and found to be essentially equivalent. The proposed commercial manufacturer is Alliance Medical Products Inc, USA. Alliance is yet to be inspected by an EU Competent Authority. This was discussed during validation and agreed that evidence of satisfactory inspection must be provided before assessment phase II of the procedure. The manufacturing process is adequately described and validated on three production scale batches. The fully assembled product is terminally sterilised by gamma irradiation and it has been shown that with exception of some degradation of fluocinolone acetonide, there is no adverse impact of gamma irradiation on the drug product. The specifications are generally acceptable subject to clarification of a number of outstanding points. Analytical procedure for control of the drug product are adequately described and validated. Transfer of analytical methods to the site for re-analysis on importation into the EU has been completed. Details of the container closure have been provided including specifications/inspection routine and technical diagrams. Suitability as well as usability of the devices was verified during development. Stability investigation of three primary batches of the drug product has been commenced at ICH long term and accelerated storage conditions and results are available for up to 12 months. This is Alimera Sciences, Inc. Submission of Evidence 166

168 supported with 36 months stability data for the clinical study batches. There was a general increase in the level of individual degradation impurities and consequently, total impurities which is reported to be as a result of the gamma irradiation process. There was no significant change in any of the other parameters tested. A shelf life of 24 months is proposed and is supported by available stability data. III.2 Non clinical aspects Pharmacology DME is characterised by the development of intraretinal oedema, due to leakage from retinal vessels. VEGF is the primary mediator of the permeability changes and angiogenesis-related progression of diabetic retinopathy. Intraocular administration of a corticosteroid has shown to reduce intravitreal VEGF levels by turning off the gene for production of VEGF and causing regression of active neovascularization by direct inhibition of VEGF-producing cells. Published data confirmed that FA inhibits expression of VEGF in ARPE- 10 cells. Intravitreal glucocorticoid administration (triamcinolone acetonide [Kenalog ]) has been shown to reduce intravitreal VEGF levels, decrease macular thickness, and increase visual acuity in man. The primary pharmacology data presented in the non-clinical dossier is taken from the literature and mainly reports efficacy of FA in uveitis using intravitreal implants. The pharmacological effect of FA will be restricted to the eye therefore no interactions systemically are anticipated. As there is clinical experience with intravitreal FA, the lack of safety pharmacology studies and pharmacodynamic interaction studies with FA is acceptable. Pharmacokinetics A general discussion comprising a review of published data on the pharmacokinetics of FA has been provided. The only PK (TK) data available through use of the proposed poduct has been generated in the 24-month repeat-dose rabbit study. The distribution of FA following intravitreal administration of evidence of dose proportionality was observed. It is noted that FA concentrations in the aqueous humour in rabbits were generally below the limit of quantification. No concerns relating to dose-dumping have been identified. Toxicology The lack of single-dose, carcinogenicity and reproductive and developmental toxicity studies with the proposed product is acceptable, given the lack of systemic exposure following intravitreal administration and clinical experience with FA intra-occularly. Focal degenerative lesions which affected fibers in the posterior polar and posterior cortical regions of the l -month and 24month toxicity studies in rabbits. These findings are not surprising, as lens fiber degeneration/cataract development in the posterior subcapsular region of the lens has been reported following intravitreal dosing of corticosteroids, including FA (Retisert ). It should also be noted that cataract formation was seen clinically (see Clinical AR). Histopathologic examination revealed focal retinal scarring in the 9- and 24-month toxicity studies in both placebo-group and Iluvien-treated animals. This effect was seen more frequently in the area of the injection site in the eyes of the rabbits treated with inserts than in rabbits subjected only to the sham injection and was more apparent in the rabbits from the groups which received multiple inserts. Discussion of the device part of the product has been provided and the lack of phototoxicity studies adequately justified. Alimera Sciences, Inc. Submission of Evidence 167

169 An appropriate environmental risk assessment has been submitted in line with relevant guidance. No Phase II assessment is required. There are no non-clinical objections to the granting of an MAA to the proposed product. III.3 Clinical aspects Pharmacokinetics The pharmacokinetic information provided by the applicant is adequate. An in vitro study of the implant shows that the low dose formulation has preferable release rate characteristics, with a sustained release close to the predicted level through 36 months. In addition a Phase 2b pharmacokinetic study has been conducted in humans. This demonstrates that intravitreal insertion of Iluvien implants does not result in a measurable systemic exposure to the active substance, and that following a single implantation there is a sustained release of FA within the eye (measurable in the aqueous humour) for 36 months. However, no dose proportionality was found between the two dose levels and a dose-response relationship was not observed for the measured efficacy parameters. This initially raised a concern regarding the dose levels selected for the pivotal Phase 3 studies, but the benefit-risk balance of the low dose implant was considered on its own merits. Pharmacodynamics The active ingredient of this medicinal product is fluocinolone acetonide (FA). FA is a glucocorticoid that has been used dermally as an anti-inflammatory product for more than 30 years. The subnanomolar affinity of FA for recombinant human glucocorticoid receptor was confirmed using a standard receptor binding assay. FA is also the active agent in Retisert (Bausch & Lomb, Rochester, NY), an intraocular delivery implant developed for treatment of non-infectious posterior uveitis. Although approved in the USA in 2005, this product was declined a licence in the EU due to safety concerns. The proposed product is claimed to provide a sustained release of FA within the vitreous prolonging its effect on the macula. Pharmacodynamic data were collected from a phase 2b human PK study; dose-response proportionality was not evident for efficacy outcomes between the low and high doses, a finding that was confirmed in the pivotal Phase 3 studies. In addition a dose-response relationship has been shown with regard to safety outcomes, which initially raised a concern that the lower dose of 0.2 μg/day selected for registration was too high a dose. As mentioned above however, the clinical data for the low dose implant are considered in detail below and in the detailed clinical assessment report. Clinical efficacy At Month 24 (the primary endpoint of the studies) approximately 29% of patients treated with Iluvien (low dose) had gained >15 letters, compared with 16% of sham-treated subjects. This represents a 13% difference between active treatment and control. The result is statistically significant, as are the individual results of FAME A and B. Maximal efficacy is seen at Month 30, with 38% of treatedsubjects gaining 15 letters (16% difference vs. control), though a slight decrease in efficacy is seen at Month 36 to 34% (10% difference vs. control), possibly because the majority of implants were reaching the end of their predicted 3-year lifespan, and subjects were not to be retreated within the last 3 months of the study. The outcome is better in patients with poorer baseline vision in both FAME studies, though the results were not statistically significant (the studies were not powered for this). The biggest gain in vision for subjects treated in the FAME studies appears to come from patients Alimera Sciences, Inc. Submission of Evidence 168

170 with DME duration of 3 years. In both FAME A and B statistically significant differences were seen between the low-dose and sham treatment groups at Months 24, 30, and 36. In this subgroup, at the end of the study 32% of subjects in FAME A and 36% of subjects in FAME B had gained 15 letters of vision, compared to around 13% of control patients. This subgroup analysis also reveals that subjects with more recently diagnosed DME (<3 years duration) do not demonstrate additional benefits from treatment with Iluvien over standard of care. The demographics and baseline characteristics of these two subgroups are similar, other than certain notable differences. At baseline, subjects with shorter duration of DME were both more likely to be phakic, and to have a smaller area of cystoid macular changes. Although there were more phakic subjects in the subgroup with duration of DME <3 years, this does not account for the reduced visual response of the subgroup because the confounding effect of cataract on vision was resolved by the third year of the study (the majority of subjects with significant cataract had been operated on by Month 24). With regard to the secondary endpoint of mean change from baseline in BCVA, the results for the entire population of subjects in the FAME studies are less impressive. In FAME A the mean change from baseline in vision for treated subjects was 3.7 letters at Month 24 (vs. 3.2 in control group), and 4.9 letters at Month 36 (vs. 3.3 in control group). The results were slightly more impressive in FAME B: mean change from baseline in vision for treated subjects was 5.1 letters at Month 24 (vs. 0 in control group), and 5.7 letters at Month 36 (vs. 0.7 in control group). However, in the subgroup of patients with duration of DME 3 years, the results for this secondary endpoint are more convincing. In both studies, from Month 24 onwards the mean change in vision was over 5 letters. The difference from the control group results was impressive in FAME B, reaching 8.5 letters at Month 30. For the anatomic secondary endpoint of change from baseline in excess centre point thickness, the results show that treatment with Iluvien reduces retinal thickness in a sustained fashion over the course of the studies. The difference from the reduction seen in the control groups is not marked in the latter stages of the study, and contrary to the findings for the visual function endpoints, the results of FAME A were more impressive than those of FAME B. As discussed in the Day 70 report, a difference for the primary endpoint of 13% over placebo is not of certain clinical relevance. However, the control group is not a true placebo comparison, since around a third received a disallowed treatment for DME (eg, steroid or VEGF inhibitor) and over a half received macular laser. With this in mind, the 13% increase in the proportion gaining 15 letters at Year 2 looks more impressive, though still needs to be balanced against the risks. The difference to control group increases for the subgroup of patients with poorer baseline vision (18% at Year 3 in FAME A, 22% at Year 3 in FAME B). Since subjects in the studies were not treatment-naïve and were to have received macular laser prior to enrolment, this subgroup of patients with poor baseline vision is reflected in the amended indication (ie, considered insufficiently responsive to available therapy). The subgroup of patients with chronic DME ( 3 years duration) displayed the best outcome: nearly 35% of subjects gained 15 letters (21% more than control group), and the mean increase in vision was 6 letters at Year 2 and 7.6 letters at Year 3. These are clinically relevant improvements in visual function, and demonstrate a meaningful benefit over sham treatment (standard of care). Furthermore, three-quarters of patients in the low-dose arm only required a single treatment during the three year study period, demonstrating the long-term efficacy of a single implant. The results of the FAME studies demonstrate a comparable degree of efficacy for Iluvien to other therapies for DME. In the DRCR Network study the primary outcome was measured at 1 year. Ranibizumab plus prompt or deferred laser was compared to sham plus prompt laser. Ranibizumab produced a mean improvement in vision of around 9 letters (a difference of 6 from sham), and around 30% of patients gained 15 letters of vision (a difference of 15% from sham). In the RESTORE study ranibizumab alone or with laser was compared against laser alone, with a primary endpoint again at 1 year. The mean improvement in vision was 6 letters for ranibizumab (a difference of 5 from laser), and around 22% of patients gained 15 letters of vision (a difference of 14% from laser). Alimera Sciences, Inc. Submission of Evidence 169

171 The response rates for the control groups in the FAME studies was higher than expected, but these subjects received laser and other disallowed therapies. Even despite this, treatment with Iluvien has been shown to produce meaningful and persistent improvements in vision over three years. Clinical safety The safety review of Iluvien has been prepared from data collected from the open-label, phase 2b pharmacokinetic study (the FAMOUS study) through Month 18, and the two phase 3, pivotal studies (FAME A and FAME B) through Month 24. Final 36 month safety reports for all studies were submitted at Day 106. The updated safety data do not highlight any new concerns in addition to those raised at Day 70. Of the phakic subjects (ie, those with a natural, crystalline lens at the start of the study) who received the dose intended for marketing, 80% had developed cataract within 2 years, and 94% of these underwent cataract surgery. This is approximately twice as high as the incidence of cataract in the sham group or in the non-study eye of those in the active treatment groups. A dose response relationship was observed between the high and low strength implants, and between those receiving single and multiple implants. However it is encouraging that the visual outcome was good in those subjects who did require cataract extraction, perhaps due in part to the anti-inflammatory effect of the corticosteroid. To put this in context, 71% of the phakic subjects in the low dose group entered the study with some degree of cataract, which is not surprising since diabetes is a risk factor for cataract. Therefore phakic subjects who receive an implant are very likely to experience worsening of their cataract, and to require surgical treatment. The rate of IOP elevation considered an adverse event (including adverse event reports of ocular hypertension and IOP increased) rose from 33.3% at Year 2 to 37.1% at Year 3 in the low dose group. The proportion of subjects requiring pressure-relieving medication or a surgical procedure also rose proportionately. Whilst it is clear that the risk of a subject requiring some degree of pressure-relieving intervention are high, the IOP does appear controllable with the mean IOP of all subjects treated with the low dose implant being around 2 mmhg above baseline at Year 3. Furthermore, the rates of complications of increased IOP appear low. Five cases of endophthalmitis (including one of fungal eye infection) were reported during the studies. One occurred in the non-study eye, and the other 4 in the study eye of subjects treated with low dose implants, though only two of these appear drug-related. The general risks in the new target population (those with chronic DME insufficiently responsive to available therapy) were no higher than in the general population. In summary, the potential risks associated with Iluvien appear generally manageable. Diabetic patients develop cataract earlier than non-diabetics, and therefore Iluvien can be seen to accelerate this inevitable, but treatable, complication. Furthermore, the increase in IOP is manageable in most patients with medication (though more than one medication is often required), with only a small proportion requiring an invasive intervention. Pharmacovigilance system The RMS considers that the Pharmacovigilance system as described by the applicant is sufficient. The Member States may consider that the Pharmacovigilance system will fulfil the requirements. The applicant must ensure that the system of pharmacovigilance is in place and functioning before the product is placed on the market. Alimera Sciences, Inc. Submission of Evidence 170

172 Risk Management Plan Summary of the Risk Management Plan (as revised by the applicant) Safety concern Proposed Pharmacovigilance (PV) Activities (routine and additional) Proposed Risk Minimisation Activities (routine and additional) Important Identified Risks Increased intraocular Routine pharmacovigilance pressure / development of glaucoma Product Information: See Section 4.2, 4.3, 4.4 and 4.8 of the SmPC and Section 2 of the PIL Formation or progression of cataract Endophthalmitis Retinal complications Vitreous complications 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance Physician training DVD/printed guide for administration of implant and details of important adverse events See warnings in Section 4.4 and 4.8 of the SmPC and Section 2 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Sections 4.2, 4.3 and 4.4 of the SmPC and Section 2 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Sections 4.2 and 4.4 of the SmPC and Section 2 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Sections 4.2 and 4.4 of the SmPC and Section 2 of the PIL Haemorrhagic events occurring with the concurrent use of anticoagulant and antiplatelet agents 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Important Potential Risks Systemic corticosteroid Routine pharmacovigilance effects Physician training DVD/printed guide for administration of implant and details of important adverse events See Section 4.4 of the SmPC and Section 2 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Sections 4.2, 4.4 and 4.6 of the SmPC and Section 2 of the PIL Procedural complications Retinitis secondary to reactivation of latent viral or other ophthalmic infections 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study See Sections 4.2, 4.3 and 4.4 of the SmPC and Sections 2 and 4 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Section 4.3 of the SmPC Alimera Sciences, Inc. Submission of Evidence 171

173 Important Missing Information Use in paediatric population Routine pharmacovigilance See Section 4.2 of the SmPC and Section 2 of the PIL Use in pregnant women Routine pharmacovigilance See Section 4.6 of the SmPC and Section 2 of the PIL Use in lactating women Routine pharmacovigilance See Section 4.6 of the SmPC and Section 2 of the PIL Long-term safety data Routine pharmacovigilance See Sections 4.2 and 4.8 of the SmPC Repeat use Implant removal Off-label use Significant retinal ischaemia 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance 5 year post authorisation registry study Routine pharmacovigilance See Sections 4.2 and 4.8 of the SmPC and Section 3 of the PIL Physician training DVD/printed guide for administration of implant and details of important adverse events See Section 4.4 of the SmPC Physician training DVD/printed guide for administration of implant and details of important adverse events See Sections 4.1 and 4.4 of the SmPC and Section 1 of the PIL See Section 4.2 of the SmPC Discussion on RMP The first version of the RMP submitted by the applicant was considered to be deficient in a number of areas as no Important Potential Risks were identified and the Important Identified Risks and Important Missing Information required additional information. Furthermore no enhanced pharmacovigilance activities or additional risk minimisation activities had been proposed. The main areas of concern were the lack of long term safety data and data on repeat administration. To address these the applicant was requested to conduct a five year post-authorisation observational study to collect data on repeat use, explantation, off label use and the important risks associated with Iluvien. To minimise the risk of procedural complications the applicant was requested to produce educational materials in the form of a technical leaflet and training DVD. These enhanced risk minimisation activities (educational materials) are intended to provide guidance on the correct administration technique for use of this product and to highlight the important risks associated with Iluvien and the monitoring requirements to aid in their diagnoses. The applicant subsequently revised the RMP taking into account revisions requested in the first LOQ. Following a review of the applicant s response a further LOQ was proposed at Day 120 to correct information provided within certain sections of the RMP and to clarify some minor points. All requested revisions have now been implemented. Periodic Safety Update Report (PSUR) The PSUR submission scheme should follow Volume 9A of The Rules Governing Medicinal Products in the European Union starting with 6-monthly PSUR. Alimera Sciences, Inc. Submission of Evidence 172

174 IV BENEFIT RISK ASSESSMENT Iluvien is a sustained-release intravitreal drug delivery system that releases submicrogram levels of fluocinolone acetonide in the vitreous humour for months, and is intended to treat chronic diabetic macular oedema. In the wider population of all patients with DME the potential benefits of Iluvien over the control group are not sufficient to outweigh the risks, particularly when considering that those patients with more recent onset DME (<3 years) gained no additional benefit over sham. However, in the subgroup of patients with chronic DME ( 3 years duration) who have not responded sufficiently to available therapies, significant benefits have been demonstrated which are sufficient to outweigh the potential risks. Therefore the benefit-risk profile of Iluvien in the treatment of vision impairment associated with chronic diabetic macular oedema, considered insufficiently responsive to available therapies is considered positive. Alimera Sciences, Inc. Submission of Evidence 173

175 V LIST OF OUTSTANDING QUESTIONS as proposed by RMS V.1 Quality aspects Potential serious risks to public health None Points for clarification Drug Substance Drug Product V.2 Non clinical aspects Potential serious risks to public health None Points for clarification None V.3 Clinical aspects Potential serious risks to public health None Points for clarification None Pharmacovigilance system None Risk Management Plan None Alimera Sciences, Inc. Submission of Evidence 174

176 VI RECOMMENDED CONDITIONS FOR MARKETING AUTHORISATION AND PRODUCT INFORMATION VI.1 Legal Status POM VI.2 Proposed list of follow-up measures and specific obligations in case of a positive benefit risk assessment Follow-up measures: Specific Obligations: Area Description Pharmacovigilance The applicant commits to a 5-year post-authorisation registry study An open label, registry study of the safety of ILUVIEN (fluocinolone acetonide 190 micrograms intravitreal implant in applicator) in patients with chronic diabetic macular oedema and should submit interim and final reports at 3 years and 5 years respectively, as well as updates within the Periodic Safety Update Report (PSUR) submissions every 6 months. The full study protocol should be submitted within three months of marketing authorisation approval. Pharmacovigilance The applicant commits to providing a technical leaflet and a training DVD for healthcare professionals in order to provide guidance on the correct administration technique and safety risks. These will be submitted for approval to health authorities prior to launch. Quality The applicant commits to investigate the effect of gamma irradiation on one additional batch and to report the data if any change in release rate is noted following gamma irradiation. VI.3 Summary of Product Characteristics (SPC) None Package Leaflet (PL) and User Testing None VI.4 Assessment of User Testing The package leaflet has been user-tested and assessment of the user test report is provided in a separate attachment. VI.5 Labelling See attached t/c QRD Label text. Alimera Sciences, Inc. Submission of Evidence 175

177 9.1.2 Summary of Product Characteristics (SPC) SUMMARY OF PRODUCT CHARACTERISTICS 1. NAME OF THE MEDICINAL PRODUCT ILUVIEN190 micrograms intravitreal implant in applicator. 2. QUALITATIVE AND QUANTITATIVE COMPOSITION Each implant contains 190 micrograms of fluocinolone acetonide. For a full list of excipients, see Section PHARMACEUTICAL FORM Intravitreal implant in applicator. Light brown coloured cylinder, approximately 3.5mm x 0.37mm in size. Implant applicator with 25 gauge needle. 4. CLINICAL PARTICULARS 4.1 Therapeutic indications ILUVIEN is indicated for the treatment of vision impairment associated with chronic diabetic macular oedema, considered insufficiently responsive to available therapies. 4.2 Posology and method of administration Posology The recommended dose is one ILUVIEN implant in the affected eye. Administration in both eyes concurrently is not recommended (see Section 4.4). An additional implant may be administered after 12 months if the patient experiences decreased vision or an increase in retinal thickness secondary to recurrent or worsening diabetic macular oedema (see Section 5.1). Retreatments should not be administered unless the potential benefits outweigh the risks. Only patients who have been insufficiently responsive to prior treatment with laser photocoagulation or other available therapies for diabetic macular oedema should be treated with ILUVIEN. Paediatric population There is no relevant use of ILUVIEN in the paediatric population in diabetic macular oedema (DMO). Alimera Sciences, Inc. Submission of Evidence 176

178 Special populations No dosage adjustments are necessary in elderly patients, or those with renal or hepatic impairment. Method of administration FOR INTRAVITREAL USE ONLY. Treatment with ILUVIEN is for intravitreal use only and should be administered by an ophthalmologist experienced in intravitreal injections. The intravitreal implant insertion should be carried out under controlled aseptic conditions, which include the use of sterile gloves, a sterile drape, and a sterile eyelid speculum (or equivalent). Adequate anaesthesia and a broad-spectrum microbicide should be given prior to the insertion. 1. Preoperative antibiotic drops may be administered at the discretion of the treating ophthalmologist. 2. Just prior to insertion, administer topical anaesthesia over the insertion site (inferotemporal quadrant recommended) as one drop followed by either a cotton-tipped applicator soaked in anaesthetic or as subconjunctival administration of adequate anaesthesia. 3. Administer 2-3 drops of adequate topical antiseptic into the lower fornix. The lids may be scrubbed with cotton-tipped applicators soaked with an adequate topical antiseptic. Place a sterile lid speculum. Have the subject look up and apply a cotton-tipped applicator soaked with an adequate antiseptic to the insertion site. Allow seconds for the topical antiseptic to dry prior to insertion of ILUVIEN. 4. The exterior of the tray should not be considered sterile. An assistant (non-sterile) should remove the tray from the carton and peel the lid from the tray without touching the interior surface. Visually check through the window of the applicator system to ensure that there is a drug implant inside. 5. Remove the applicator from the tray with sterile gloved hands touching only the sterile surface and applicator. 6. To reduce the amount of air administered with the implant, administration requires a two step process. Before inserting the needle in the eye, push the button down and slide it to the first stop (the curved black marks). At the first stop, release the button and it will move to the UP position. 7. The optimal placement of the implant is inferior to the optic disc and posterior to the equator. This can be achieved by directing the needle toward the lower aspect of the optic disc. Measure 4 millimeters inferotemporal from the limbus with the aid of calipers. 8. Remove the protective cap from the needle. 9. Gently displace the conjunctiva so that after withdrawing the needle, the conjunctival and scleral needle entry sites will not align. Care should be taken to avoid contact between the needle and the lid margin or lashes. Insert the needle in the eye. To release the implant, while the button is in the up position, push the button forward to the end and remove the needle. 10. Remove the lid speculum and perform indirect ophthalmoscopy to verify the placement of the implant, adequate central retinal artery perfusion and absence of any other complications. Alimera Sciences, Inc. Submission of Evidence 177

179 Following intravitreal insertion indirect ophthalmoscopy examination in the quadrant of insertion should be performed to ensure successful placement. Scleral depression may enhance visualisation of the implant. Examination should include a check for perfusion of the optic nerve head immediately after the insertion. Immediate IOP measurement may be performed at the discretion of the ophthalmologist. Following the procedure, patients should be monitored for potential complications such as endophthalmitis, increased intraocular pressure, retinal detachments, and vitreous haemorrhages or detachments. Biomicroscopy with tonometry should be performed between two and seven days after the implant insertion. Thereafter it is recommended that patients are monitored at least quarterly for potential complications, due to the extended duration of release of fluocinolone acetonide, of approximately 36 months (See Section 4.4). 4.3 Contraindications An intravitreal implant with ILUVIEN is contraindicated in the presence of pre-existing glaucoma or active or suspected ocular or periocular infection including most viral diseases of the cornea and conjunctiva, including active epithelial herpes simplex keratitis (dendritic keratitis), vaccinia, varicella, mycobacterial infections, and fungal diseases. ILUVIEN is contraindicated in patients with hypersensitivity to the active substance or to any of the excipients listed in Section Special warnings and precautions for use Intravitreal injections have been associated with endophthalmitis, elevation in intraocular pressure, retinal detachments and vitreous haemorrhages or detachments. Patients should be instructed to report without delay any symptoms suggestive of endophthalmitis. Patient monitoring within two to seven days following the injection may permit early identification and treatment of ocular infection, increase in intraocular pressure or other complication. It is recommended that intra-ocular pressure be monitored at least quarterly thereafter. Use of intravitreal corticosteroids may cause cataracts, increased intraocular pressure, glaucoma and may increase the risk of secondary infections. The safety and efficacy of ILUVIEN administered to both eyes concurrently have not been studied. Concurrent treatment of both eyes is not recommended until the patient s systemic and ocular response to the first implant is known. In the FAME studies, the incidence of cataract surgery among all phakic subjects was approximately 3 fold higher in the ILUVIEN treated group (80.0%) than the sham treated group (27.3%). The median time to cataract reported as an adverse event was approximately 14 months, and phakic subjects experienced vision impairment from the development of cataract from approximately month 9 through to month 18 after implantation, prior to cataract extraction (see Section 4.8). In the overall clinical trials population, which excluded subjects with baseline IOP>21 mmhg, the proportion of ILUVIEN treated subjects requiring treatment with IOP-lowering medication was 38% compared to 14% in the sham treated group. This proportion increased to 47% in those subjects with Alimera Sciences, Inc. Submission of Evidence 178

180 greater than median IOP at baseline ( 15 mmhg). Surgical interventions for the treatment of ocular hypertension were required in 4.8% of subjects treated with ILUVIEN compared to 0.5% of subjects treated with sham (see Section 4.8). Therefore, ILUVIEN should be used with caution in patients with high baseline IOP, and IOP must be monitored closely. In the event of IOP increases that do not respond to IOP-lowering medications or IOP-lowering procedures, the ILUVIEN implant can be removed by vitrectomy. A limited number of subjects with Type 1 diabetes were investigated in the FAME studies, though the response to ILUVIEN in these subjects was not significantly different to those subjects with Type 2 diabetes. There is limited experience of the effect of ILUVIEN in eyes following vitrectomy. It is likely that drug clearance would be accelerated after vitrectomy, though steady state concentrations are not expected to be affected. This may shorten the duration of action of the insert. In the FAME studies there were 24% of subjects in the sham treated group who were treated at any time with either anti-coagulant or anti-platelet medications as compared to 72% in the ILUVIEN treated subjects. Subjects treated with ILUVIEN concomitantly or within 30 days of cessation of treatment with anti-coagulant or anti-platelet medications experienced a slightly higher incidence of conjunctival haemorrhage versus the sham treated subjects (0.5% sham and 2.7% ILUVIEN treated). The only other event reported at a higher incidence rate in the ILUVIEN treated subjects was eye operation complication (0% sham and 0.3% ILUVIEN treated). 4.5 Interaction with other medicinal products and other forms of interaction No interaction studies with other medicinal products have been performed. 4.6 Fertility, pregnancy and lactation Pregnancy There are no adequate data from the use of intravitreal administered fluocinolone acetonide in pregnant women. Long-term systemic treatment with glucocorticoids during pregnancy increases the risk for intra-uterine growth retardation and adrenal insufficiency of the newborn child. Studies in animals have shown teratogenic effects following systemic administration (See Section 5.3). Therefore, although the systemic exposure of fluocinolone acetonide would be expected to be very low after local, intraocular treatment, ILUVIEN is not recommended during pregnancy unless the potential benefit justifies the potential risk to the foetus. Breast-feeding Fluocinolone acetonide is excreted in breast milk. No effects on the child are anticipated due to the route of administration and the resulting systemic levels. However ILUVIEN is not recommended during breast feeding unless clearly necessary. Fertility Alimera Sciences, Inc. Submission of Evidence 179

181 There are no fertility data available. 4.7 Effects on ability to drive and use machines Patients may experience temporarily reduced vision after administration of ILUVIEN and should refrain from driving or using machines until this has resolved. 4.8 Undesirable effects Summary of the safety profile ILUVIEN was evaluated in 768 subjects with diabetic macular oedema across the FAME clinical trials. The most frequently reported adverse drug reactions included cataract operation, cataract and increased intraocular pressure. In the Phase 3 studies, 38.4% of subjects treated with ILUVIEN required IOP-lowering medication and 4.8% required IOP-lowering surgeries. The use of IOP-lowering medication was similar in subjects who received two or more treatments with ILUVIEN. Two cases of endophthalmitis were reported in subjects treated with ILUVIEN during the Phase 3 studies. This represents an incidence rate of 0.2% (2 cases divided by 1,022 injections). While the majority of subjects in the FAME clinical trials received only one implant (See Section 5.1), the long-term safety implications of retention of the non-bioerodable implant inside the eye are not known. In the FAME clinical trials, 3-year data show that events such as cataract, increased intraocular pressure and floaters occurred only slightly more frequently in subjects receiving 2 or more implants. This is considered a function of the increased exposure to the drug rather than an effect of the implant itself. In non-clinical studies, there were no indications of an increase in safety issues other than lens changes in the rabbit eyes with 2-4 implants over 24 months. The implant is made of polyimide and is essentially similar to an intraocular lens haptic; it is therefore expected to remain inert inside the eye. Tabulated list of adverse events The following undesirable effects were assessed to be treatment-related and are classified according to the following convention: very common ( 1/10); common ( 1/100 to < 1/10); uncommon ( 1/1,000 to < 1/100); rare ( 1/10,000 to < 1/1,000); and very rare ( 1/10,000). Within each frequency grouping, undesirable effects are presented in order of decreasing seriousness. Infections and infestations Nervous system disorders Eye disorders Uncommon: endophthalmitis Uncommon: headache Very Common: cataract operation, cataract 1, increased intraocular pressure 2, floaters (myodesopsia) Common: glaucoma 3, trabeculectomy, eye pain 4, vitreous haemorrhage, conjunctival haemorrhage, blurred vision 5, glaucoma surgery, reduced visual acuity, vitrectomy, trabeculoplasty Uncommon: retinal vascular occlusion 6, optic nerve disorder, Alimera Sciences, Inc. Submission of Evidence 180

182 maculopathy, optic atrophy, conjunctival ulcer, iris neovascularisation, retinal exudates, vitreous degeneration, vitreous detachment, posterior capsule opacification, iris adhesions, ocular hyperaemia, sclera thinning, removal of extruded implant from sclera, eye discharge, eye pruritus Injury, poisoning and procedural Uncommon: extrusion of implant, implant in line of sight, complications procedural complication, procedural pain 1 Includes MedDRA terms for cataract (NOS), cataract subcapsular, cataract cortical, cataract nuclear and cataract diabetic. 2 Includes MedDRA terms for intraocular pressure increased and ocular hypertension. 3 Includes MedDRA terms for glaucoma, open angle glaucoma, borderline glaucoma, optic nerve cupping and optic nerve cup/disc ratio increased. 4 Includes MedDRA terms for eye pain, eye irritation and ocular discomfort. 5 Includes MedDRA terms for vision blurred and visual impairment. 6 Includes MedDRA terms for retinal vein occlusion, retinal artery occlusion and retinal vascular occlusion Description of selected adverse reactions The long-term use of corticosteroids may cause cataracts and increased intraocular pressure. The frequencies stated below reflect the findings in all patients in the FAME studies. The observed frequencies in patients with chronic DMO were not significantly different to those in the overall population. The incidence of cataract in phakic subjects was approximately 82% in ILUVIEN treated subjects and 50% in sham treated subjects in the Phase 3 clinical trials. 80% of phakic subjects treated with ILUVIEN required cataract surgery by Year 3 compared to 27% of the sham treated subjects, with most subjects requiring surgery by 21 months. Posterior subcapsular cataract is the most common type of corticosteroid -related cataract. Surgery for this type of cataract is more difficult and may be associated with greater risk of surgical complications. In the FAME studies subjects with a baseline IOP of > 21 mm Hg were excluded. The incidence of increased intraocular pressure was 37%, and 38% of subjects required IOP-lowering medication, with half of these requiring at least two medications to control the IOP. The use of IOP-lowering medication was similar in subjects who received retreatment with an additional implant during the study. Additionally, 5.6% (21/375) of subjects who received an implant required a surgical or laser procedure to control the IOP (trabeculoplasty 5 (1.3%), trabeculectomy 10 (2.7%), endocycloablation 2 (0.5%), and other surgical procedures 6 (1.6%)). In the subset of subjects with greater than median IOP at baseline ( 15 mmhg), 47% required IOPlowering medication and the proportion of surgical or laser procedures increased to 7.1%. In this subset, there were 5 (2.2%) subjects treated with trabeculoplasty, 7 (3.1%) with trabeculectomy, 2 (0.9%) with endocycloablation and 4 (1.8%) with other glaucoma surgical procedures. 4.9 Overdose No case of overdose has been reported. 5. PHARMACOLOGICAL PROPERTIES Alimera Sciences, Inc. Submission of Evidence 181

183 5.1 Pharmacodynamic properties Pharmacotherapeutic group: ANTIINFLAMMATORY AGENTS, corticosteroids, plain ATC code: S01BA15 Corticosteroids inhibit the inflammatory response to a variety of inciting agents. They inhibit the oedema, fibrin deposition, capillary dilation, leukocyte migration, capillary proliferation, fibroblast proliferation, deposition of collagen, and scar formation associated with inflammation. Corticosteroids are thought to act by the induction of phospholipase A inhibitory proteins, collectively called lipocortins. It is postulated that these proteins control the biosynthesis of potent mediators of inflammation such as prostaglandins and leukotrienes by inhibiting the release of the common precursor arachidonic acid. Arachidonic acid is released from membrane phospholipids by phospholipase A 2. Corticosteroids have also been shown to reduce levels of vascular endothelial growth factor, a protein which increases vascular permeability and causes oedema. The efficacy of ILUVIEN was assessed in two randomized, multicenter, double-masked, parallel studies enrolling subjects with diabetic macular oedema who had previously been treated with laser photocoagulation at least once, each involving three years of follow-up. There were 74.4% of subjects treated with 1 implant, 21.6% with 2 implants, 3.5% with 3 implants and 0.5% with 4 implants and 0% > 4 implants). The primary efficacy endpoint in both trials was the proportion of subjects whose vision improved by 15 letters or greater after 24 months. In each of these trials, the primary endpoint was met for ILUVIEN (see Figure 1 for the integrated results of the primary efficacy endpoint). Alimera Sciences, Inc. Submission of Evidence 182

184 Figure 1: Percentage of Subjects with 15 Letter Improvement Over Baseline, Integrated FAME Studies When efficacy was assessed as a function of duration of disease, those subjects with a duration of DMO greater than the median ( 3 years) had a significant beneficial response to ILUVIEN, whilst those with shorter duration DMO did not show an additional benefit over control treatment with regard to visual improvement (Figures 2 and 3). These subgroup data support the indication in Section 4.1, of use in patients with chronic DMO (ie, duration of at least 3 years). Alimera Sciences, Inc. Submission of Evidence 183

185 DMO 3 Years Figure 2: Comparison of Percent of Subjects with 15 letter Improvement from Baseline BCVA and Mean Change from Baseline Excess Center Point Thickness by Duration of DMO Subgroup 3 years Alimera Sciences, Inc. Submission of Evidence 184

186 DMO < 3 Years Figure 3: Comparison of Mean Change from Baseline Excess Center Point Thickness and Percent of Subjects with 15 letter Improvement from Baseline BCVA by Duration of DMO Subgroup < 3 years The European Medicines Agency has waived the obligation to submit results of studies with ILUVIEN in all subsets of the paediatric population for the treatment of diabetic macular oedema. See Section 4.2 for information on paediatric use. Alimera Sciences, Inc. Submission of Evidence 185