This electronic thesis or dissertation has been downloaded from the King s Research Portal at

Transcription

1 This electronic thesis or dissertation has been downloaded from the King s Research Portal at Non-motor symptoms in advanced Parkinson s The natural history and response to advanced therapies Reddy, Papari Prashanth Awarding institution: King's College London The copyright of this thesis rests with the author and no quotation from it or information derived from it may be published without proper acknowledgement. END USER LICENCE AGREEMENT This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International licence. You are free to: Share: to copy, distribute and transmit the work Under the following conditions: Attribution: You must attribute the work in the manner specified by the author (but not in any way that suggests that they endorse you or your use of the work). Non Commercial: You may not use this work for commercial purposes. No Derivative Works - You may not alter, transform, or build upon this work. Any of these conditions can be waived if you receive permission from the author. Your fair dealings and other rights are in no way affected by the above. Take down policy If you believe that this document breaches copyright please contact librarypure@kcl.ac.uk providing details, and we will remove access to the work immediately and investigate your claim. Download date: 24. Dec. 2018

2 Non-motor symptoms in advanced Parkinson s: The natural history and response to advanced therapies Papari Prashanth Reddy MD (Res) King s College London Page 1

3 Abstract Background and Method Non-motor symptoms (NMS) are a key determinant of health related quality of life (HrQoL) in Parkinson s disease (PD) (Martinez-Martin et al., 2011) and are present in virtually every PD patient. Many aspects of NMS are treatable by dopaminergic drugs (Chaudhuri and Schapira, 2009) and treating NMS remains a key unmet need. (Chaudhuri et al., 2011) Currently treatment choices in advanced PD are Apomorphine (Apo), intra-jejunal levodopa infusion (IJLI) and Deep brain stimulation of the subthalamic nucleus (DBS-STN) and are solely recommended based on motor symptoms. This thesis addresses a range of comparative multicentre, international collaborative studies of real life dopaminergic therapies in PD with a focus on non-motor effects. Non-motor effect has been measured using the validated PD non-motor symptoms scale (NMSS). (Chaudhuri et al., 2007) The largest real life worldwide study with a long-term longitudinal follow-up design addresses a key unmet need with over 40 patients in each arm. Using regression analysis and comparative measures, effects on motor function, quality of life were compared and contrasted with non-motor effects. A patient reported outcome tool was developed using the same cohort of patients. Results and Conclusion This project has shown the beneficial effects of Apo, IJLI and DBS-STN therapies on aspects of NMS in advanced Parkinson s that translate to a robust improvement in the HrQoL. For the first time we show that there are differences between Apo, a dopamine agonist based therapy versus IJLI (a levodopa therapy) on aspects of NMS and this would need to be explored further. The effects could be explained by a central dopaminergic effect and effective Page 2

4 treatment of NMS arising from non-motor fluctuations. (Storch et al., 2013) Choice of advanced therapies in PD should therefore take into account their non-motor effects and patient expectations along with improvement of motor symptoms. Page 3

5 TABLE OF CONTENTS 1 TABLE OF FIGURES 7 2 TABLE OF TABLES 8 3 ACKNOWLEDGEMENTS 10 4 CHAPTER 1: INTRODUCTION TO NON-MOTOR SYMPTOMS (NMS) IN PD HYPOTHESIS BEING TESTED AND RELATED OBJECTIVES: METHODS STUDY DESIGN THESIS STRUCTURE PROJECT CONCEPTION AND EXECUTION 25 CHAPTER 2: THE ROLE OF CONTINUOUS DOPAMINERGIC STIMULATION (CDS) IN THE MANAGEMENT OF NON-MOTOR SYMPTOMS IN PARKINSON S DISEASE INTRODUCTION INTRAJEJUNAL LEVODOPA INFUSION (IJLI) APOMORPHINE INFUSION (APO) DEEP BRAIN STIMULATION OF SUBTHALAMIC NUCLEUS (DBS-STN) EFFECTS OF CDS ON INDIVIDUAL NMS 34 Restless Leg Syndrome, Periodic Limb Movements and non-motor symptoms related to early morning akinesia Pain Gastrointestinal Disturbances and Dysphagia Orthostatic Hypotension Peripheral Oedema Visual Hallucinations Insomnia Depression and anxiety 39 Page 4

6 5.1.9 Cognition and attention Urological symptoms Sustained-release dopaminergic therapy CONCLUSION 41 CHAPTER 3: CHRONIC SUBCUTANEOUS INFUSION THERAPY WITH APOMORPHINE IN ADVANCED PARKINSON S DISEASE COMPARED TO CONVENTIONAL THERAPY: A REAL LIFE STUDY OF NON-MOTOR EFFECT ABSTRACT: INTRODUCTION PATIENTS AND METHODS STATISTICAL ANALYSIS RESULTS DISCUSSION 46 CHAPTER 4: INTRAJEJUNAL LEVODOPA INFUSION IN THE UK: COMPARATIVE REVIEW OF MOTOR AND NON-MOTOR EFFECTS IN TREATED VERSUS UNTREATED PATIENTS WITH ADVANCED PARKINSON S DISEASE ABSTRACT INTRODUCTION PATIENTS AND METHODS STATISTICAL ANALYSIS DISCUSSION 59 CHAPTER 5: BILATERAL SUBTHALAMIC STIMULATION IMPROVES ASPECTS OF NON- MOTOR SYMPTOMS IN PARKINSON S DISEASE ABSTRACT INTRODUCTION MATERIAL AND METHODS 65 Page 5

7 1.26 STATISTICAL ANALYSIS RESULTS DISCUSSION 70 9 CHAPTER 6: EUROINF: A MULTICENTRE COMPARATIVE OBSERVATIONAL STUDY OF APOMORPHINE AND LEVODOPA INFUSION IN PARKINSON S DISEASE ABSTRACT: INTRODUCTION: METHODS: Design and patient selection: Assessments: Data Analysis: RESULTS: DISCUSSION: CHAPTER 7: PERCEPTIONS OF SYMPTOMS AND EXPECTATIONS OF ADVANCED THERAPY FOR PARKINSON S DISEASE: PRELIMINARY REPORT OF A PATIENT-REPORTED OUTCOME TOOL FOR ADVANCED PARKINSON S DISEASE (PRO-APD) ABSTRACT: INTRODUCTION: METHODS: RESULTS: DISCUSSION: CHAPTER 8: CONCLUSIONS AND INTEGRATED MESSAGES ARISING FROM WORK PRESENTED IN THIS THESIS: REFERENCES: APPENDICES PUBLICATIONS ARISING FROM THE WORK PRESENTED IN THE THESIS 150 Page 6

8 1 Table of Figures Figure 1: Flow chart illustrating the studies discussed in the thesis Figure 2: Flow chart to describe the work presented in the thesis. The interlink between the chapters is illustrated with arrows Figure 3: Mean change from baseline in UPDRS and Non Motor Scale scores in the IJLI versus the control group Figure 5: Scatterplots showing relation between perceived severity and expected improvement for each of the 4 domains on the PRO-APD Figure 6: Profiles of perceived severity for illustrative cases X and Y (gaps indicate zero ratings), showing broadly similar severities across individual symptom profiles using the PRO-APD Figure 7: Profiles of expectations of improvement ratings for illustrative cases X and Y (gaps indicate zero ratings) showing clearly different profiles of expectation following treatment using the PRO-APD Page 7

9 2 Table of Tables Table 1: Non-motor symptoms questionnaire (NMS-Q) Table 2: Non-motor symptom scale (NMSS) Table 3: Parkinson s disease Questionnaire (PDQ) Table 4: Showing the individual NMS and their responsiveness to dopaminergic therapy Table 5: Changes following Apomorphine infusion and continuing conventional therapy (comparator) in motor, non-motor, and quality of life dimension Table 6: Change in the items of Non-Motor Symptoms Scale (Apo infusion group) Table 7: NNT in patients treated with Apomorphine Table 8: Baseline patient characteristics in the IJLI group and Controls Table 9: Significant improvement of all non-motor outcomes with DBS-STN Table 10: RC, effect size and NNT with DBS-STN therapy Table 11: Spearman s rank correlations between outcomes with DBS-STN Table 12: Correlations between PDQ-8 SI and NMSS domains after DBS-STN Table 13: Supplemental table 1 Significant improvement of UPDRS-III and -IV Table 14: Supplemental table 2 RC, ES and NNT of UPDRS-III and -IV Page 8

10 Table 15: Supplemental table 3 Spearman s rank correlations between UPDRS-III and -IV, NMS, and QoL outcomes Table 16: Descriptive and comparative statistics at baseline and follow-up for each group of treatment Table 17: Magnitude of the change from baseline to follow-up for each group of treatment Table 18: Number needed to treat for improving ½ SDBaseline Table 19: Reported side effects of IJLI and Apo therapy at six months follow up period Table 20: Illustrative case vignettes X and Y Table 21: Patient demographics, motor and non-motor severity, Quality of Life, Mood and Cognition Table 22: Domain and total PRO-APD scores of problem severity and expectation of change Table 23: Correlations (Pearson r) between clinical variables and PRO-APD severity and expectation domain scores Page 9

11 3 Acknowledgements I would like to thank my supervisor and mentor, Prof. K Ray Chaudhuri, for his guidance, support and encouragement through the years. The experience and exposure I gained presenting at international meetings and preparing abstracts and papers under his guidance has been invaluable. This work would not have been possible without his support and faith in me. I would like to thank Prof. Stephen Jackson for supporting me and helping me finish my research. I would like to acknowledge Prof. Pablo Martinez-Martin for his help with the statistical analysis required for my studies. I would like to thank my wife for encouraging me to enrol in a research degree and for supporting me through tough times. I am also extremely grateful for my father for helping me get to where I am today and for always pushing me to achieve more. Page 10

12 ABBREVIATIONS Addenbrooke s Cognitive Examination-Revised (ACE-R) 6 months follow-up (6MFU) Activities of daily living (ADL) Apomorphine Infusion (Apo) Continuous dopaminergic stimulation (CDS) Deep brain stimulation- subthalamic nucleus (DBS-STN) Dementias and Neurodegeneration (DenDRoN) Effect Size (ES) Food and Drug Administration (FDA) Health related quality of life (HrQoL) Hospital anxiety and depression Scale (HADS) Hoehn and Yahr (HY) Impulse control disorder (ICD) Intrajejunal Levodopa Infusion (IJLI) King s College Hospital (KCH) Levodopa equivalent dose (LED) Levodopa-equivalent daily dose (LEDD) National Institute for Health and Care Excellence (NICE) Non-motor International Longitudinal study (NILS) Movement disorders society (MDS) Non-motor symptoms (NMS) NMS Scale (NMSS) NMSS total score (NMSS-T) Non-motor symptom Questionnaire (NMSQ) Non-motor symptom Questionnaire total score (NMSQ-T) Page 11

13 Number needed to treat (NNT) Parkinson s disease (PD) Parkinsons disease Questionnaire (PDQ) Parkinsons disease Questionnaire summary index (PDQ SI) Patient Reported Outcome (PRO) Patient Reported Outcome in Advanced Parkinson s disease (PRO-APD) Percutaneous endoscopic jejunostomy (PEJ) Quality of life (QoL) Randomised control trial (RCT) Relative change (RC) Serious adverse event (SAE) Standard deviation (SD) Standard error of the mean (SEM) Standardised response mean (SRM) Scales for Outcomes in Parkinson s disease (SCOPA) Short Parkinson s Evaluation Scale (SPES) Subthalamic nucleus (STN) Summary index (SI) Total electric energy delivered (TEED) Unified Parkinson s disease Rating Scale (UPDRS) Page 12

14 4 Chapter 1: Introduction to Non-motor symptoms (NMS) in PD James Parkinson described Parkinson s disease (PD) in his famous essay An Essay on the Shaking Palsy in (Parkinson, 2002, Parkinson, 1817) The non-motor symptoms (NMS) of PD were recognized by James Parkinson himself and he referred to sleep disturbance, constipation, dysarthria, dysphonia, dysphagia, sialorrhoea, urinary incontinence and, at the last, constant sleepiness with slight delirium. Since then, numerous studies have indicated that NMS are frequent accompaniments of PD, affecting memory, bladder and bowel, and sleep, among others. (Chaudhuri and Martinez-Martin, 2008) However, although common, the NMS of PD are not well recognized in clinical practice. One study in the USA showed that existing depression, anxiety and fatigue are not identified by neurologists in over 50% of consultations, and existing sleep disturbance in over 40%. (Shulman et al., 2002) Researchers and clinicians have been able to better identify NMS as a whole using the Non-motor symptom questionnaire (NMS-Q) (Chaudhuri et al., 2006b) and grade rate Non-motor symptoms using the Non-motor symptom scale (NMSS). Since the availability of the NMS-Q (Martinez-Martin et al., 2007) several studies in different populations, in over 2500 patients have been published looking at the prevalence of non-motor symptoms using the NMS-Q and their impact on the quality of life. (Barone et al., 2009, Martinez-Martin et al., 2007) NMS are treatable and dopaminergic non-oral therapies have shown promise in the management of some but not all NMS. (Chaudhuri and Schapira, 2009) However, no studies have systematically explored these options and a novel set of studies which address individual effects of these advanced therapies (Apomorphine infusion, intrajejunal levodopa infusion and Deep Brain Page 13

15 stimulation) on NMS in advanced PD using validated tools are presented in this work. In addition the first ever patient rated outcome (PRO) tool was developed in a pilot study and serves as a cross cutting instrument to explore non-motor issues in relation to the advanced therapies. Currently the indications (based on NICE guidelines and Scottish guidelines) (Conditions, 2006, Grosset et al., 2010) for advanced therapies are based on the motor symptom severity. From studies so far there is a better correlation between NMS than motor symptoms with quality of life. (Chaudhuri and Odin, 2010, Chaudhuri and Martinez-Martin, 2008) Hence there is a requirement for an instrument to measure a patient perception of severity of their symptoms and thereby their expectations from proposed therapy. This would be in line with the FDA recommendations to the pharmaceutical industry. (Health et al., 2006) Why is this work novel?: It is important to note that very few centres in the world are able to deliver advanced therapies for PD under one roof and use NMS measures as one of the primary outcome variables. The International centre of excellence in Parkinson s care and research allows this at King s College Hospital (KCH) where we have one of the largest cohort of patients on Apomorphine, IJLI therapies and also an established Deep brain stimulation service. The comparative work presented in this thesis addressing NMS outcomes of Apomorphine infusion and IJLI as well as development of a relevant PRO are all part of a novel dataset, which may open up new avenues of research. It is not likely that there will ever be a randomised trial comparing Apo and IJLI owing to costs, logistic and ethical issues. The studies proposed therefore have Page 14

16 the constraint of being non randomised but are the next best option and carried out in real life population likely to generate good external validity. 1.1 Hypothesis being tested and related objectives: 1. Advanced therapies for PD (Apo, IJLI and DBS-STN) will show holistic improvement in NMS in PD as judged by Non-Motor Symptom Scale (NMSS) and translate to improvement in health related quality of life in advanced PD 2. Comparative case control studies (non randomised) with individual therapies (using a control group on standard/conventional therapy) will outline specific effect on NMS and may flag up a differential effect of Apo vs. IJLI 3. NMS may be differentially affected by a dopamine agonist therapy such as Apomorphine compared to intrajejunal levodopa infusion. An European multicentre study led by KCH is presented which address a real life comparative study (Apo vs. IJLI) to address any differential effects as observed in 2 4. In relation to above, we also present data for a third treatment arm which address effects on NMS with DBS-STN therapy using similar methodology from KCH testing the hypothesis: DBS-STN has a beneficial effect on aspects of non-motor symptoms in PD 5. A cross cutting arm will explore the development and use of a Patient Reported Outcome (PRO) questionnaire of patients undergoing all 3 treatments with a specific focus on NMS 6. Patient expectation from advanced therapies (Apo/IJLI/DBS-STN) in Parkinson s disease (PD) is influenced by the severity of the symptom experienced 1.2 Methods 1. Two separate pilot studies were undertaken comparing Apo with standard conventional therapy (oral therapies) and IJLI with conventional therapy in PD. Page 15

17 Non motor symptom scale (NMSS) scale was used to measure the non motor symptoms burden in both the studies to see if there was a holistic improvement in the NMS with Apo and IJLI compared to conventional therapies after 6 months of initiation of therapy compared to the patients on conventional therapy. Alongside the NMSS PDQ-8 and UPDRS scales were used comparing conventional therapy with Apo/IJLI 2. To compare the non-motor effect of the advanced therapies (Apo and IJLI) in PD an open label design was chosen. A larger sample size (n>40) was chosen to minimize selection bias and to make sure patient characteristics were similar in both groups with relation to age, disease duration and severity of symptoms before initiation of therapy. NMSS was used to measure the severity of nonmotor symptoms before initiation of therapy and at 6 months post intervention 3. Effect on non-motor symptoms with DBS-STN was explored using an open label study design 4. Patient reported outcome in advanced Parkinson s (PRO-APD) tool was devised. A pilot study was undertaken to look at the practicality of using the PRO-APD tool for advanced Parkinson s disease. The PRO-APD measured patient reported severity of individual symptoms in these domains -motor, nonmotor, psychological and social. These scores were then compared to the patient reported expectation form the proposed therapies (Apo, DBS-STN and IJLI) The novel aspect of work presented in this thesis focuses on non-motor outcomes measured individually and in a holistic manner using a selfdeclaration tool (NMS-Q) as well as a validated grade-rating tool the NMSS and QoL measure using PDQ-8. The clinimetric details of these scales are presented below in tables 1,2 and 3 which are reproduced with permission from Page 16

18 Prof. Pablo Martinez-Martin and Prof K Ray Chaudhuri (Martinez-Martin et al., 2014b) which were published in the Guide to Assessment scales in Parkinson s disease, Springer Healthcare Page 17

19 Table 1: Non-motor symptoms questionnaire (NMS-Q) Scale: Non-motor symptoms Questionnaire (NMS-Quest) Reference: Chaudhuri KR, Martinez-Martin P, Schapira AHV, Stocchi F, Sethi K, Odin P, et al. International multicentre pilot study of the first comprehensive self-completed nonmotor symptoms questionnaire for Parkinson s disease: the NMSQuest study. Mov. Disord. 2006;21(7): Martinez-Martin P, Schapira AHV, Stocchi F, Sethi K, Odin P, MacPhee G, et al. Prevalence of non-motor symptoms in Parkinson s disease in an international setting; study using non-motor symptoms questionnaire in 545 patients. Mov. Disord. 2007;22(11): Description of It is a screening questionnaire revealing the range of NMS in PD scale The NMS-quest is a self-completed questionnaire featuring responses as yes and no to each item. It is composed of 30 items grouped into 9 domains: I, Digestive (7 items); II, Urinary tract (2 items); III, Apathy/Attention/Memory (3 items); IV, Hallucinations/Delusions (2 items); V, Depression/Anxiety (2 items); VI, Sexual function (2 items); VII, Cardiovascular (2 items); VIII, Sleep disorders (5 items); IX, Miscellaneous (pain, weight change, swelling, seating, diplopia) (5 items). Time frame: past month. Time for administration: 5-7 min. The screening questionnaire is filled out by the patient/caregiver while waiting to be seen in clinic. It is specific to identify the NMS in PD. Copyright? The Movement Disorders Society How can the scale The scale can be obtained from the original publication: Chaudhuri KR, Martinezbe obtained? Martin P, Schapira AHV, Stocchi F, Sethi K, Odin P, et al. International multicentre pilot study of the first comprehensive self-completed non-motor symptoms questionnaire for Parkinson s disease: the NMS-Quest study. Mov. Disord. 2006;21(7): Clinimetric properties of scale in Patients with PD Feasibility Specifically designed for PD patients. Used in all stages of PD to identify if the patient has any NMS. Vocabulary avoiding medical jargon and adapted to a seventh-grade level. Designed to be applicable to PD patients across various levels of disabilities. Scores significantly increase with disease duration and HY stages. Dimensionality NMS-Quest has 9 domains, although its structure has not been tested. Acceptability Parameters observed were satisfactory and included a low ceiling (0.8%) and floor (2.4%) effect, closeness of mean to median (9.5 vs. 9), and an almost complete range of scores (0-28). Reliability Internal consistency: satisfactory for total score and most domains. Item-domain corrected correlation were high. Test-retest and inter-rater: not tested. Validity Convergent: NMS-Quest was highly correlated with NMSS total score and corresponding domains. Association of total NMS-Quest with HY stage was moderate (rs = 0.31). A lower correlation was found with disease duration (rs = 0.22). Known-groups: Total score significantly increased with increased age, disease duration and severity of disease. Internal: Inter domain correlations were poor to moderate (range: 0.06 to 0.37). Responsiveness & Interpretability Not tested. Cross-cultural adaptations Translated and validated into many languages. Overall impression Advantages: It is a quick and easy screening tool used by the patient/caregiver to flag up problematic NMS as an aid to clinical management. Over 75% of patients and 80% caregivers felt that the NMS-Quest will improve the doctor s ability to treat PD better, and 90% of patients and caregivers felt that the issues raised in the NMS-quest were relevant to day-to-day life. Disadvantages: It does not assess severity of symptoms or effect of treatment. Page 18

20 Table 2: Non-motor symptom scale (NMSS) Scale: Reference: Description of scale Copyright? How can the scale be obtained? Non-motor symptoms Scale (NMSS) Chaudhuri KR, Martinez-Martin P, Brown RG, Sethi K, Stocchi F, Odin P, et al. The metric properties of a novel non-motor symptoms scale for Parkinson s disease: Results from an international pilot study. Mov. Disord. 2007;22(13): It is a tool to quantify NMS, covering the range of NMS each scored by the physician for symptom severity and frequency. It is composed of 30 items grouped into 9 domains: I, Cardiovascular (2 items); II, Sleep/fatigue (4 items); III, Mood/Apathy (6 items); IV, Perceptual problems/hallucinations (3 items); V, Attention/memory (3 items); VI, Gastrointestinal tract (3 items); VII, Urinary (3 items); VIII, Sexual function (2 items); and IX, Miscellaneous (4 items). Time frame: past month. Time for administration: min. The NMSS is rated by health professionals and obtained through clinical interview. Score for each item is based on multiple of severity (from 0 to 3) and frequency scores (from 1 to 4). The Movement Disorders Society The scale can be obtained from the original publication listed above. (Note: the correct denomination of the Domain 3 is Mood/Apathy ). Clinimetric properties of scale in Patients with PD Feasibility Specifically designed for PD patients. Used in all stages of PD to identify the severity and frequency of a patient s NMS. Designed to be applicable to PD patients across various levels of disabilities. Scores significantly increase with severity of disease based on HY stages, NMSQuest, and health-related quality of life assessments. Dimensionality An exploratory factor analysis supported the prior nine domain structure, explaining 63% of the variance. Acceptability The overall floor and ceiling effect of the total NMSS score were lower than 1%. Skewness was 1.2. The domains showed variable floor effect. Reliability Cronbach s α-coefficient ranged from 0.44 to 0.85and item homogeneity from 0.16 to The multi-trait scaling reached a success, and probable success rate was higher than 95% for all domains, except the miscellaneous domain (47% success rate), which contained wide ranging, unrelated questions from diplopia to weight change. Most of item-total correlations were higher than the criterion 0.30 ( ), the lowest values corresponding to the Miscellaneous domain. With exception of an isolated domain (cardiovascular or sexual), test-retest was satisfactory (>0.70) in both validation studies. Validity NMSS total score reached high correlation with SCOPAAUT (rs=0.64), PDQ-39 (rs=0.70), and EQ-5D Index (rs=0.57). Correlation coefficient values with other measures were: Hoehn and Yahr staging, 0.38; SCOPA-Motor, 0.44; SCOPA-Psychiatric complications (PC), 0.51; SCOPA-Cognition, 0.44; Clinical Impression of Severity Index (CISI)-PD, 0.49; PD Sleep Scale (PDSS), 0.53; and EQ-VAS, NMSS domains showed a tight association with other measures for similar constructs: sleep/fatigue with PDSS (rs=0.56), perceptual problems/hallucinations with SCOPA-PC (rs=0.53), and attention/memory with CISI-PD cognition (rs=0.51). A moderate to high association was found between attention/memory and SCOPAPC (rs=0.49). There were weak correlations between the corresponding domains and scales for mood and frontal function assessment. The correlation with domains of the NMS Questionnaire ranged from 0.44 to Responsiveness Standard error of the measurement for the NMSS is satisfactory ( ) & Interpretability (<1/2SD). The scale has been found sensitive to changes induced by advanced therapies. Cross-cultural Translated and validated into many languages. adaptations Overall impression Advantages: explores a wide range of NMS that may occur in a patient with PD and provides a holistic assessment. The scale captures symptoms that are severe but relatively infrequent (e.g., hallucinations) and those that may be less severe but persistent (e.g., constipation, fatigue or low mood). This method increases the weight of symptoms simultaneously that are persistent and severe. Disadvantages: Miscellaneous domain showed poor clinimetric attributes, The overall floor and ceiling effect of the total NMSS score were lower than 1%. Does not allow more in-depth evaluation of individual symptoms. Page 19

21 Table 3: Parkinson s disease Questionnaire (PDQ)-8 Scale: Parkinson s disease Questionnaire-8 (PDQ-8) JENKINSON, C., FITZPATRICK, R., PETO, V., GREENHALL, R. & HYMAN, N. 1997b. The PDQ-8: Development and validation of a short-form Parkinson s Disease Questionnaire. Psychol Health 12, Overview The questionnaire assesses subjective health status, although it is classified as health-related quality of life instrument. The eight items are scored on a likert-type scale from 0(never) to 4 (always). Subscale scores are transformed into a scale by summing the items raw scores, dividing them by the maximum possible raw score, and then multiplying by 100 a Summary index (SI) is calculated. Higher scores mean lower quality of life. Time frame: the month prior to assessment. Specific for PD patients. Copyright? Owned by Isis innovation Limited How can the scale For obtaining the scales, the manual and the licence: University of Oxford (Isis be obtained? Innovation Limited) Clinimetric properties of scale in Patients with PD Feasibility Older patients and those with more severe impairments could have difficulties with the response options. Dimensionality Factor analysis has identified a single factor Acceptability Observed range scores were almost coincident with the possible range. NO floor or ceiling effects were detected. Reliability Internal consistency: suitable. Inter-rater and test-retest reliability appropriate. Reference: Validity Content validity not tested for PDQ-8. Convergent: close correlations with other quality of life and clinical scales. Known-groups: significant differences by H&Y staging scale stages. Responsiveness & Interpretability Widely applied as an outcome measure in clinical trials and have been proved to be sensitive to changes in health status. Minimal important difference (MID) has been calculated for the scale. Scale is applicable to PD patients of both sexes and at all ages. Available in several languages and used in different cultural settings. Cross-cultural adaptations Overall impression Advantages: Includes dimensions relevant to PD patients: widely used and extensively analysed across different settings and countries; adequate psychometric properties; responsive to changes; recommended by the Movement Disorder Society (MDS) task force. Disadvantages: Lack some relevant areas for PD; some limitations in reliability; dimensionality not well established and some clinimetric properties need further analysis. Page 20

22 Experiments+ Apo$Vs.$conven,onal$therapy:$Effect$ on$nms$ IJLI$Vs.$Conven,onal$therapy$:$Effect$ on$nms$ DBS$Vs.$conven,onal$therapy$:$ Effect$on$NMS$ explore$the$use$of$a$pa,ent$ Reported$Outcome$tool$across$the$ advanced$therapies$ real$life$ compara,ve$ study$(apo$vs.$ IJLI)$to$ address$any$ differen,al$ effects$$ Figure 1: Flow chart illustrating the studies discussed in the thesis Page 21

23 1.3 Study design 1. Open label multicentre case control study The reason for opting for the second best study design (open label) instead of a randomised study was due to the following factors: a. Randomisation of patients comes with a variety of practical and ethical difficulties b. Real life data collection was opted for, as that would be clinically more relevant as opposed to cherry picking patients (ideal patient) for the study c. A study involving a double dummy model involving the 3 therapies has a lot of practical problems given the different approaches/modalities used (DBSSTN- brain surgery, IJLI- PEG tube, Apo- subcutaneous infusion) d. Open label studies are useful for comparing similar treatments to determine which is most effective e. Clinical Indications for these treatments are based on patient characteristics and therefore randomisation would not be possible. DBS-STN and IJLI are more invasive compared to Apo hence blinding would not be practical Advantages of case control studies are (Mann, 2003): 1. Multiple outcomes can be studied 2. Less affected by ethical issues than RCTs especially when data collected is from routine clinical practice 3. Suitable for complex situations where surgical procedures are involved and RCTs are not feasible 4. Less expensive compared to RCTs Page 22

24 Disadvantages are: 1. Potential for bias from lack of randomisation 2. Confounding variables may be an issue Efforts were made in the studies performed to eliminate confounding variables by selecting patients who were very similar at baseline in the groups that were being compared. It is also important to note that even randomised trials have drawbacks and they are not always possible or ethical, especially when the treatment options involve invasive procedures. Another drawback of randomized trials is that they usually involve patients who are considered fit to enter, are likely to finish the trial, and have very strict inclusion and exclusion criteria compared to a real life scenario. The cohort of patients that are usually recruited to randomised trials are quite different from the patients in a real life setting. (Rosendaal, 2001) Mann C.J. states Cohort, cross sectional, and case-control studies are collectively referred to as observational studies. A study that provides estimates and examines association of events in their natural settings without recourse to experimental intervention. Often these studies are the only practicable method of studying various problems, for example, studies of aetiology, instances where a randomised controlled trial might be unethical. (Mann, 2003) Black N. states The view is widely held that experimental methods (randomised controlled trials) are the "gold standard" for evaluation and that observational methods (cohort and case control studies) have little or no value. This ignores the limitations of randomised trials, which may prove unnecessary, inappropriate, impossible, or inadequate. Many of the problems of conducting randomised trials could often, in theory, be overcome, but the practical implications for researchers and funding bodies mean that this is often not Page 23

25 possible. The false conflict between those who advocate randomised trials in all situations and those who believe observational data provide sufficient evidence needs to be replaced with mutual recognition of the complementary roles of the two approaches. Researchers should be united in their quest for scientific rigour in evaluation, regardless of the method used. (Black, 1996) 2. Recruitment from clinics clinician and patient agreed on the type of therapy based on existing guidelines (NICE, Scottish guidelines) and then consented for data collection using the Non motor symptom scale (NMSS), Unified Parkinson s disease Rating Scale (UPDRS) and Parkinson s disease Questionnaire (PDQ-8) which are validated scales. 3. NMSS, UPDRS and PDQ-8 were used to measure severity of symptoms before and after the proposed treatment (6 months) 1.4 Thesis structure 1. Introduction: Non Motor Symptoms (NMS) in Parkinson s disease 2. The role of continuous dopaminergic stimulation in the management of nonmotor symptoms in Parkinson s disease: Oxford University Press (Reddy et al., 2014a) 3. Chronic subcutaneous infusion therapy with Apomorphine in advanced Parkinson's disease compared to conventional therapy: a real life study of nonmotor effect: Journal of Parkinson's disease (Martinez-Martin, 2011) 4. Intrajejunal levodopa versus conventional therapy in Parkinson disease: motor and non-motor effects: Clinical Neuropharmacology (Reddy et al., 2012) 5. Bilateral subthalamic stimulation (DBS-STN) improves aspects of non-motor symptoms in Parkinson s disease Page 24

26 6. EuroInf: A multi-centre European comparative study of Apomorphine versus intrajejunal levodopa infusion in Parkinson's patients: Movement Disorders Journal (Martinez-Martin et al., 2014a) 7. Perceptions of symptoms and expectations of advanced therapy for Parkinson's disease: preliminary report of a Patient-Reported Outcome tool for Advanced Parkinson's disease (PRO-APD): Health and Quality of Life Outcomes (Reddy et al., 2014b) 8. Discussion 1.5 Project conception and execution Working as a specialist trainee at King s College Hospital I was fortunate to meet Prof Chaudhuri in He was at the forefront of researching non-motor symptoms and raising awareness in the international community when I met him. I was at this stage attending the movement disorders clinic as part of my training. I started out with helping collect data on the non-motor symptoms using the non-motor symptoms questionnaire and the non-motor symptom scale in his clinics. During that time I was struck by the lack of evidence in treating nonmotor symptoms. At this stage apomorphine was well established in treating motor symptoms and Intrajejunal levodopa infusion therapy was introduced at King s, both achieve continuous dopaminergic stimulation. I then discussed with Prof Chaudhuri about embarking on a project looking at the effect of these continuous dopaminergic strategies on the non-motor symptoms. I was able to secure the ethics application for the study of non-motor symptoms and the effect of continuous dopaminergic strategies on them. The ethics application also covered devising a Patient reported outcome tool to study Page 25

27 patient perception of severity and their expectations from these advanced therapies. The study was approved by the Regional ethics committee of South East London: ref number: 10/H0808/46. The funding source was a grant from King s Health Partners, obtained through open competition. This grant funded my research fellow post for 3 years. During that time I was responsible for patient recruitment and consent for the experiments described in this thesis. The patients were recruited from the Movement disorders clinics at King s College Hospital I devised the Patient reported outcome tool, administered the scales and questionnaires and collected the data from the experiments described in this thesis I was in charge of co-ordinating the experiments with other movement disorder units across Europe; collate data, ensure uniformity in study design and recruitment I liaised with Prof. Pablo Martinez-Martin for the statistical analysis for the experiments I prepared the manuscript for submission to the relevant journals, shared it with the other authors and amended the manuscript based on their comments and published 5 papers in peer reviewed journals based on the experiments described in this thesis I had the opportunity to present my findings at various international meetings and conferences and was awarded 3 travel grants Page 26

28 Structure'of'Thesis' Apo+Vs.+conven)onal+ therapy:+effect+on+nms+ Introduc)on:+Non+ Motor+symptoms+ in+pd++ Role+of+ Con)nuous+ Dopaminergic+ S)mula)on+in+ trea)ng+nms+ IJLI+Vs.+Conven)onal+ therapy+:+effect+on+nms+ + DBS+Vs.+conven)onal+ therapy+:+effect+on+nms+ real+life+ compara)ve+study+ (Apo+vs+IJLI)+to+ address+any+ differen)al+effects++ Percep)ons+of+ symptoms+and+ expecta)ons+of+ advanced+therapy+for+ Parkinson's+disease:+ Pilot+study+exploring+ the+use+of+a+pa)entl Reported+Outcome+ tool+for+advanced+ Parkinson's+disease++ Figure 2: Flow chart to describe the work presented in the thesis. The interlink between the chapters is illustrated with arrows Page 27

29 5 Chapter 2: The role of continuous dopaminergic stimulation (CDS) in the management of non-motor symptoms in Parkinson s disease This Chapter is based on: REDDY, P., ANTONINI, A. & ODIN, P The role of continuous dopaminergic stimulation in the management of non-motor symptoms in Parkinson s disease. Non-motor Symptoms of Parkinson's Disease, Oxford University Press, 2nd edition, p Introduction Non-motor symptoms (NMS) in PD have a significant impact on the quality of life of patients and yet they are under-recognised and under-treated. (Antonini et al., 2008, Barone et al., 2009, Chapuis et al., 2005, Chaudhuri et al., 2006a, Chaudhuri and Odin, 2010) This is mainly due to poor awareness and insufficient evidence base for treating the NMS. (Antonini et al., 2012) The motor features of PD largely arise from dopamine deficiency in the substantia nigra and related dysfunction in the striato-cortical loops. Other neurotransmitter systems are also involved in PD, with depletion of acetylcholine, norepinephrine, and serotonin levels (Winogrodzka et al., 2005). However current treatment has largely focused on replacing endogenous dopamine via the administration of L-dopa or other dopaminergic drugs. (Poewe, 2008, Agid et al., 1999, Rascol et al., 2002, Nutt and Holford, 1996, Fabbrini et al., 2007) Some, but not all, non-motor features (sleep disorders, autonomic dysfunction and fatigue) have a dopaminergic basis (Table 1). (Wolters and Braak, 2006, Braak et al., 2003) The term continuous dopaminergic stimulation (CDS) is used for treatments, which can provide relatively steady plasma levels of dopaminergic drugs and thereby continuous dopamine receptor stimulation. Options for such treatments are subcutaneous Apomorphine infusion, Intrajejunal Levodopa/carbidopa gel Page 28

30 (IJLI) infusion, Rotigotine skin patch and Deep brain stimulation (the later not being dopaminergic, but based on continuous stimulation). CDS is a very effective option for patients with severe motor fluctuations poorly controlled by oral therapy. (Antonini et al., 2010) Delivering drugs through nonoral approach has made possible achieving steady plasma levels of the dopaminergic drugs thereby achieving CDS. These approaches have in turn reduced the side effects and increased efficacy of the drugs. (Gershanik and Jenner, 2012) Motor fluctuations complicate levodopa treated PD. Similarly a non-motor aspect of these fluctuations have been described by several workers such as Witjas and classified them into three subtypes: dysautonomic, cognitive and psychiatric, and sensory or pain. (Witjas et al., 2007) NMS such as pain and excessive daytime sleepiness may result from off periods. Witjas and colleagues reported that anxiety (66%), drenching sweats (64%), slowness of thinking (58%), fatigue (56%), and akathisia (54%) were the most common nonmotor fluctuations and these correlated with motor disability and pulsatile levodopa treatment. (Witjas et al., 2002, Storch et al., 2013) While there is significant evidence on the efficacy of long acting dopamine agonists in reducing some of these NMS, there is still a paucity of controlled trials focusing on the effect of CDS on NMS. Currently no consensus exists as to whether there is a difference in the effect of the various CDS therapies on the NMS. No randomized head to head trials exist comparing the effects of the various therapies offering CDS. (Volkmann et al., 2013) Concerning the mechanisms of the effects of CDS on NMS, these probably vary depending on the individual NMS, one of these certainly being that the patients spend more time in "on" and have less NMS in "on" compared to "off". Page 29

31 Table 4: Showing the individual NMS and their responsiveness to dopaminergic therapy. (Table reproduced with permission from Prof. Chaudhuri) (Chaudhuri and Schapira, 2009) Page 30

32 1.7 Intrajejunal levodopa infusion (IJLI) IJLI (trade name: Duodopa) represents a more physiological option of dopamine replacement by providing a more stable L-dopa plasma concentration compared to oral therapy. Striatal dopamine terminals normally store dopamine and buffer the plasma dopamine fluctuations seen with oral L-dopa therapy. However, these terminals degenerate in advanced PD exposing dopamine receptors to the fluctuating dopamine concentrations, resulting from per oral L-dopa therapy (Honig et al., 2009). By administering L-dopa directly into the proximal part of the small intestine, the drug bypasses the erratic gastric emptying and can be quickly and continuously transported over the intestinal wall. This results in more stable plasma concentrations and most likely a more stable stimulation at the dopamine receptor level. Moving from per oral L-dopa to IJLI therapy thus reduces off periods and peak-dose dyskinesias and provides more predictable clinical benefits for the patients (Antonini, 2007). There are also results indicating that also the non-motor symptomatology does improve with IJLI. A recent 6-month open label study of IJLI demonstrated improvement in six of the nine domains of the Non-motor Symptom Scale and for the total NMS-burden score, in addition to improvements in Parkinson s disease Sleep Scale (PDSS) and PDQ-8 quality of life scores. (Honig et al., 2009) Results from my study are presented in chapter 4, the mean change from baseline in UPDRS and Non Motor Scale scores in the IJLI versus the control group (Reddy et al., 2012) is illustrated in figure 3 (chapter 4). 1.8 Apomorphine Infusion (Apo) Apomorphine is a dopamine agonist with effects on the D2 and D3 receptors. It is the only available dopamine agonist with an antiparkinson effect comparable to L-dopa. Apo is normally administered as subcutaneous injections at demand Page 31

33 or as a continuous subcutaneous infusion with portable pumps. It has shown to be effective in reducing "time in off" compared to oral therapies but reports suggest that can reduce dyskinesias (severity and duration) if monotherapy is achieved. (Katzenschlager et al., 2005a, Tyne et al., 2004, Manson et al., 2001, Nutt and Carter, 2000, van Laar et al., 1998, Pfeiffer et al., 2007, Colosimo et al., 1998, Trosch et al., 2008, De Gaspari et al., 2006) In my study Apomorphine has also shown to have an effect on NMS, particularly mood and urinary symptoms (Martinez-Martin, 2011) similar to IJLI therapy (Reddy et al., 2012). However, it must be considered that unlike LCIJ, Apo is commonly used in combination with oral levodopa therapy. Change in the items of Non-Motor Symptoms Scale in patients treated with Continuous Apomorphine infusion is shown in Table 6 (Chapter 3). The symptoms, which improve significantly on Apomorphine therapy, are highlighted in bold in table Deep Brain Stimulation of Subthalamic Nucleus (DBS-STN) While Deep Brain Stimulation (DBS-STN) is not really a form of continuous dopaminergic therapy, it does continuously modulate dopaminergic transmission. DBS-STN is currently most commonly administered via bilateral subthalamic nucleus (STN) stimulation. DBS-STN reduces off time, dyskinesia time and intensity, as well as, the amount of dopaminergic medication. It may also reduce fluctuation-related anxiety (Witjas et al., 2007), depression (Goetz et al., 2008), off period-associated pain and pain-intensity. (Gierthmuhlen et al.) A recent paper by Ashkan et al 2012, (Ashkan et al., 2012) showed that DBSSTN reduced overall NMS-load as measured by the Non-motor Symptom Scale, along with the motor improvements. Improvements have been seen concerning sleep, pain and urinary problems, with a few studies also Page 32

34 demonstrating improved bladder control in PD patients on DBS-STN. (Seif et al., 2004, Winge et al., 2007, Ashkan et al., 2012, Borgohain et al., 2012) The outcome of NMS after DBS-STN in PD varies across studies. Some symptoms improve (sleep disorders, pain or sensory complaints, obsessivecompulsive disorder) and other aspects worsen (verbal fluency) or appear, (apathy, body weight gain). Isolated studies note mild improvements in working memory, visuomotor sequencing and conceptual reasoning, some gastrointestinal, urogenital, sweating and olfactory disturbances; whereas other studies have reported declines in verbal memory (long delay recall), visuospatial memory, processing speed and executive function with orthostatic hypotension remaining unchanged. The reasons for such a variability in effect on different NMS is probably due to the multifactorial etiology of the NMS, including preoperative vulnerability, changes in dopaminergic medications, surgical and stimulation effects, and underlying PD-related factors and psychosocial effects. Specific patient subgroups may be at greater risk of cognitive deficits, e.g. patient s older than 69 years, or with cognitive impairment prior to surgery, or preoperative psychiatric symptoms. (Sevillano-Garcia and Manso-Calderon) The suggestion of a differential effect of target (Globus Pallidus (GPi) vs. Subthalamic Nucleus (STN) on cognition and mood was seen in the COMPARE trial, particularly when more ventral stimulation was used in STN. (Okun et al., 2009) The findings from my study on the effect of DBS-STN on NMS are presented in chapter 5. Page 33

35 1.10 Effects of CDS on individual NMS Restless Leg Syndrome, Periodic Limb Movements and non-motor symptoms related to early morning akinesia Sleep disturbances are very common in patients with PD, and include Restless Legs (RLS), Periodic Limb Movements, Insomnia, Excessive Daytime Sleepiness (EDS) and REM-sleep Behaviour Disorder (RBD). The prevalence/occurrence of RLS in PD is controversial but observational studies suggest a two-fold increase. (Chaudhuri and Behan, 2003) Neuroimaging has pointed to a presynaptic striatal pathology underlying Idiopathic RLS and PLMS. (Ruottinen et al., 2000) However, RLS has not been associated with cell loss in the substantia nigra or presence of Lewy body pathology. (Connor et al., 2003) RLS symptoms can fluctuate along with motor status, being particularly worse during off periods. A study by Tan and colleagues showed that 0.8% of PD patients had RLS-like symptoms during wearing-off (Tan et al., 2002), while Witjas and colleagues observed that 22% of patients experienced symptoms consistent with RLS relating to off periods, although 18% experienced them independent of treatment. (Witjas et al., 2002) RLS occurs in PD in several forms. Typical RLS is uncommon while an RLSlike syndrome is common. The latter may either be non-responsive to dopaminergic therapy or result from nocturnal motor fluctuations, which may benefit from extended/continuous release dopamine agonists. RLS is largely regarded as a non-motor phenomenon in treated PD. Dopamine agonists are currently the first line choice for the treatment of RLS and PLMS. (Hening et al., 1999) In comparative studies, Cabergoline, a longacting dopamine agonist that is no longer used due to the risk of heart valve Page 34

36 fibrosis (Antonini, 2007) and Rotigotine transdermal patch have been shown to provide more effective relief from nocturnal painful dystonia, akinesia and early morning dystonia than controlled-release levodopa. (Rinne et al., 1998, Pal et al., 2001) The recent RECOVER study (Trenkwalder et al., 2011) was the first large-scale, double blind trial to investigate sleep as a primary outcome measure. It demonstrated that 24-hour transdermal Rotigotine treatment significantly improved nocturnal disturbances and reduced night time disabilities including pain, immobility, cramps and limb restlessness. Continuous dopaminergic stimulation can improve overall sleep quality (Nyholm et al., 2005) and sleep/fatigue symptoms, as measured by the NMSS and PD Sleep Scale. (Honig et al., 2009) These observations have also been reported after the use of Ropinirole slow release, Rotigotine patch, Apomorphine infusion and intra-jejunal levodopa, suggesting that nocturnal NMS may occur as a result of inadequate dopaminergic stimulation, which might be alleviated by CDS. (Chaudhuri et al., 2013) Pain Pain is a major component of the NMS complex in PD (Defazio et al., 2008), with one large study showing it to be reported by 29% of patients. (MartinezMartin et al., 2007) Although a number of neurotransmitters and tracts control and transmit the perception of pain, dopamine modulates pain at various levels including the spinal cord, thalamus, periaqueductal grey, basal ganglia and cingulated cortex. (Chudler and Dong, 1995, Shyu et al., 1992) This may explain why PD patients have a lower pain threshold compared to a healthy control population. (Djaldetti et al., 2004) Further Tinazzi et al have shown that the nociceptive input processing is abnormal in pain-free PD patients Page 35

37 independent of motor signs and is not affected by dopaminergic drugs. (Tinazzi et al., 2008) The majority of pain in PD is caused by motor fluctuations and dyskinesias secondary to dopaminergic therapy. (Quinn et al., 1986) Fluctuation-related pain can be further categorised into dyskinetic pain, off period dystonia-related pain, and off period generalised pain. (Chaudhuri et al., 2006a, Chaudhuri and Schapira, 2009) RLS or akathisia can also be a cause of pain in PD. There are no trials specifically investigating the effect of CDS on pain. However, in a small overnight study Apomorphine infusion alleviated pain in six patients with RLS. (Reuter et al., 1999) Studies have shown that the pain threshold in PD-patients may increase with the administration of levodopa. (Djaldetti et al., 2004) Continuous dopaminergic stimulation, either by subcutaneous Apomorphine infusion or duodenal L-dopa infusion may be more effective. A 12month prospective study of Duodopa infusion in 7 patients with PD showed a significant improvement in bodily discomfort. (Antonini, 2007) The RECOVER study reported improvement of pain with Rotigotine compared to placebo, as measured by a Likert scale. (Trenkwalder et al., 2011) RECOVER study is actually the first controlled trial that addresses pain as an independent secondary outcome variable (Trenkwalder et al., 2011) confirmed in a post hoc analysis. (Kassubek et al., 2014) Gastrointestinal Disturbances and Dysphagia Gastrointestinal symptoms such as constipation, anorexia, nausea, dysphagia and reduced gastric motility are very common NMS and can also occur as a side effect of dopaminergic therapy. Saliva aspiration does not correlate with the severity of PD, reflecting that autonomic phenomena in PD dysphagia are likely to be unrelated to motor Page 36

38 progression. (Ali et al., 1996) There is not much evidence that the classical drugs used to treat Parkinson s disease can decrease or post-pone silent laryngeal penetration/silent aspiration and there are controversies regarding the dopamine capacity to decrease swallowing disturbances in PD patients. (Monte et al., 2005, Hunter et al., 1997) Patients suffering from constipation during off-periods may relieve their symptoms with subcutaneous Apomorphine injections, suggesting a role for dopaminergic treatment in improving gastric motility. (Merello, 2008) One noncontrolled study has described the efficacy of CDS in reducing gastrointestinal symptoms in PD. (Honig et al., 2009) Orthostatic Hypotension Orthostatic hypotension (OH) is one of the most disabling NMS. Noradrenergic and neurocirculatory degeneration results in reflex cardiac vagal anomalies and altered sympathetic reflexes with decreased levels of serum norepinephrine. OH in PD has been observed independent of levodopa, (Goldstein, 2006, Goldstein et al., 2005, Pursiainen et al., 2007) while vasodilation with hypotension and bradycardia can also result from dopaminergic therapy. (Polakowski et al., 2004) This suggests that the autonomic dysfunction may be caused by the disease itself, by the effect of medication, or both. Furthermore, studies have shown varying supine and standing blood pressure measurements in on-off states, suggesting that more stable, prolonged-release dopaminergic therapy may be beneficial in some patients. Anecdotal observations suggest that in some advanced PD patients the use of levodopa infusion and concurrent withdrawal of oral therapies may lead to improvement of OH. Page 37

39 5.1.5 Peripheral Oedema Peripheral oedema may be seen in advanced PD, particularly as a side effect of dopamine agonist therapy. It most commonly presents as a unilateral painful swollen leg, mimicking thrombosis or a complex regional pain syndrome. Levodopa significantly reduces venous tone in the lower extremities and nonergot dopamine agonists including Pramipexole and Ropinirole cause a dosedependent oedema. However, 10% of patients may experience fluctuations in oedema as off-period phenomena, suggesting that CDS may be beneficial. (Tan and Ondo, 2000) Visual Hallucinations Visual hallucinations (VH) are common in patients with PD, affecting one third of patients with fluctuations. VHs may result from a combination of reduced perceptual input and lowered attention. (Collerton et al., 2005) Some studies have shown that subcutaneous infusion of Apomorphine may not worsen VH s (Di Rosa et al., 2003, Morgante et al., 2004) and may in fact improve contrast sensitivity (Geerligs et al., 2009) and VHs in PD. (Ellis et al., 1997, MartinezMartin, 2011) Some study results indicate that hallucinations might also improve when moving from per oral therapy to IJLI infusion Insomnia Sleep problems are present in 60-98% of patients with PD. (Mondragon-Rezola et al., 2010) In a study by Gerstad et al reported a prevalence of insomnia in 5460%. (Gjerstad et al., 2007) It also fluctuates over time and the cause is often multifactorial. Coexisting depression is a common cause.(caap-ahlgren and Dehlin, 2001) Garcia Ruiz showed that overnight use of Apomorphine infusion improved sleep architecture. (Garcia Ruiz, 2006) Zibetti et al. showed improvements in nocturnal sleep in patients started on intrajejunal levodopa Page 38

40 infusion. (Zibetti et al., 2013) All these studies show that there is a direct link between the dopaminergic pathways and sleep. Therefore improving dopaminergic stimulation at night can play a major role in improving insomnia and quality of sleep Depression and anxiety The effect of dopaminergic therapy in mood disorders and apathy in PD can partly be explained by the knowledge that levodopa is taken up and decarboxylated in serotoninergic neurons, which also converts levodopa to dopamine. (Chaudhuri and Schapira, 2009) Anxiety often coexists with depression and can respond to dopaminergic therapy. (Lemke et al., 2005) In clinical practice anxiety is seen to occur as a dopamine-dependent event as part of wearing off (presenting as panic attacks) and would therefore respond to dopaminergic strategies aimed to prevent wearing off which is best achieved by the CDS strategies. (Chaudhuri and Schapira, 2009) Similarly, depressionrelated anxiety might also respond to dopaminergic treatment, although, in some patients, anxiety can remain a constant underlying problem that is independent of dopaminergic state and that might not respond to dopaminergic therapy. (Richard et al., 1996) A randomised but unblinded study showed a significant improvement in anxiety after Deep brain stimulation of the subthalamic nucleus compared with conventional and best medical treatment at 6 months of follow-up. (Goetz et al., 2008) In an observational study by Witjas and co-workers, fluctuation-related anxiety was reported to be significantly lowered after Deep brain stimulation of subthalamic nucleus in advanced PD. (Witjas et al., 2007) These finding once again support that CDS is useful in treating anxiety in PD. Page 39

41 5.1.9 Cognition and attention The early cognitive changes of PD might involve the caudate and corticostriatal pathways. (Emre, 2003) Abnormalities of dopamine uptake and brain metabolism in cortical targets of striatal dopaminergic fibers have been reported. (Lewis et al., 2003, Rinne et al., 2000) A gradual increase in dopamine levels in the striatum improves motor function and memory when there is saturation of the dorsal striatum. The ventral striatal dopaminergic stimulation however has a detrimental effect on cognitive function, particularly impairment in learning new stimulus associations. (MacDonald et al., 2011) In view of these finding CDS strategy to optimize dopaminergic stimulation and preventing peaks with oral therapies hereby preventing excessive ventral striatal effects would be most useful Urological symptoms Dopaminergic stimulation has a varied effect on micturition. Striatal D1 receptor stimulation inhibits micturition and D2 stimulation activates it. (Seki et al., 2001) Hence urge incontinence and urgency can be worsened by dopaminergic therapy while voiding difficulties are improved with dopaminergic therapy. (Uchiyama et al., 2003) Seif et al showed with Deep brain stimulation of the subthalamic nucleus a beneficial effect on bladder function in PD. These improvements were thought to be due to improving bladder capacity and reflex volume; which is possibly secondary to improved sensory motor integration and improvement afferent activity. (Seif et al., 2004) However recent studies looking at the non-motor effects of IJLI and Apomorphine infusion showed improvement in the urinary domain with both the therapies. (Martinez-Martin, 2011, Reddy et al., 2012) Page 40

42 Sustained-release dopaminergic therapy The recently introduced dopamine agonists with sustained drug delivery mechanisms, such as extended release Ropinirole XL, Pramipexole prolonged release and the Rotigotine transdermal patch may reduce NMS burden in patients with PD. Although they do not provide a strictly continuous form of dopaminergic stimulation, they may be particularly effective in relieving sleeprelated NMS including RLS, RBD and nocturia, and improving Parkinson s disease Sleep Scale (PDSS) scores. (Poewe, 2008, Pahwa et al., 2007) However, the benefits of these dopamine agonists have to be balanced against some sleep-related adverse events, including excessive daytime sleepiness (EDS), hallucinations and the risk of impulse control disorders Conclusion Treating NMS remains a significant unmet need in PD. Some NMS, especially those associated with off periods, may be relieved using continuous dopaminergic stimulation. Both the pump-based therapies and DBS-STN can achieve considerable improvements of many non-motor symptoms, beside improvements in the motor symptomatology. While there are a few studies assessing the efficacy of CDS in treating NMS, there are none that directly compare them. The Rotigotine transdermal patch may offer an alternative CDS strategy for those unsuitable to invasive therapies. Further research is needed to elucidate the specific benefits of each therapy and their role in treating nonmotor symptoms in Parkinson s disease. Page 41

43 6 Chapter 3: Chronic subcutaneous infusion therapy with Apomorphine in advanced Parkinson s disease compared to conventional therapy: a real life study of non-motor effect This Chapter is based on: MARTINEZ-MARTIN, P., REDDY P., ANTONINI A. ET AL Chronic Subcutaneous Infusion Therapy with Apomorphine in Advanced Parkinson's Disease Compared to Conventional Therapy: A Real Life Study of Non Motor Effect. Journal of Parkinson's disease, 1, Abstract: Background: Apomorphine infusion therapy remains under-used and there are no comparative studies of motor and non-motor effects of Apomorphine infusion Methods: In this paper we report preliminary results from a clinical observational real life surveillance-based study focused on effects of this therapy on nonmotor symptoms and health-related quality of life in a group of patients on Apomorphine. Results: Apomorphine infusion led to highly significant improvements in UPDRS 3 (p=0.0003), UPDRS 4 (p=0.0003), PDQ-8 (Parkinson s disease questionnaire, p=0.001) and NMSS total (non motor symptoms scale, p=0.0003). Furthermore, Apomorphine was tolerated in patients with visual hallucinations, illusions and paranoid ideations while significant improvement in specific non-motor symptoms such as hyperhidrosis, nocturia, urgency of micturition, and fatigue was recorded. Levodopa equivalent dose decreased significantly ( ± to ±282.29, p< ) and a large effect size of intervention was noted. In an untreated group no such improvement was noted. The number needed to treat (NNT) for improvement >1 Standard error of the mean (SEM) in the Apo group was calculated and was lower than 2 for >1 SEM improvement of UPDRS 3, NMSS, and PDQ-8 total scores. Page 42

44 Conclusions: This pilot observational study suggests that non-motor effects are evident with Apomorphine therapy and patients suitable for Apomorphine deteriorate in the absence of therapy Introduction Subcutaneous administration of Apomorphine (Apo) has been used as treatment for advanced Parkinson s disease (PD) with refractory on-off periods and dyskinesias and is nationally recommended in many countries. (Lees and Turner, 2002, Poewe and Wenning, 2000, Chaudhuri and Clough, 1998, Conditions, 2006, Grosset et al., 2010, Katzenschlager et al., 2005b, Manson et al., 2002, Garcia Ruiz et al., 2008, Stocchi et al., 2001, Poewe et al., 1993) We have reported the beneficial effect of Apomorphine infusion on aspects of nonmotor symptoms (NMS) of PD as assessed by the validated PD NMS scale (NMSS). (Martinez-Martin, 2011, Chaudhuri et al., 2007, Martinez-Martin et al., 2009) There are no studies comparing non-motor effects of Apomorphine infusion in two groups of PD patients deemed suitable for Apomorphine therapy yet where one group is treated by conventional (oral and patch therapy) treatment only because of local funding issues Patients and Methods Between , 17 PD patients (all satisfying the UK PD brain Bank criteria) (Hughes et al., 1992b) with severe dyskinesias (as rated by patients, carers and clinician) and on-off fluctuations unresponsive to changes in existing therapy were started on Apomorphine infusion (12-16hrs/day) and followed up in centres across Europe. All had baseline assessment ( 7 to 10 days pretreatment of motor function (Unified Parkinson s Disease Rating Scale, UPDRS 3 and 4) (Goetz et al., 1995), Hoehn and Yahr staging (HY) (Hoehn and Yahr, 1998), Non-Motor Symptoms Scale (NMSS) (Martinez-Martin et al., 2009, Page 43

45 Chaudhuri et al., 2007) and health-related quality of life (HrQoL) (Parkinson s disease questionnaire, PDQ-8) (Jenkinson et al., 1997b). Follow-up assessments were carried out in a manner similar to that recently reported. (Garcia Ruiz et al., 2008) For analysis, baseline data was compared to the last available follow-up visit of each patient at on stage. Seventeen PD patients were selected for comparison (C) as these patients were not funded for Apomorphine. These patients also had severe dyskinesias and on-off fluctuations unresponsive to further modifications in their existing therapy. These patients were followed up while on best conventional therapy using the same clinical protocol. As this was a clinically dictated study, matching for C and Apo cases from individual centres was not possible. This study is part of a five year global study (NILS) which is evaluating progression and incidence of motor and NMS in PD and the effect of therapy on the motor and non-motor aspects of PD. The study (a global one) has thus far recruited 1088 patients with baseline data ranging from untreated PD to advanced, the latter group (from Europe and those suitable for advanced therapies) being screened for the purpose of this report. The study was approved by individual research ethics committees of all participants and is led by a national portfolio adoption of the study in UK (DenDRoN, portfolio number 10084) 1.15 Statistical analysis Chi-squared and Fisher s exact tests were used for comparison of proportions, while non-parametric tests were applied for ordinal and continuous variables. Multiple comparisons were corrected by the Benjamini-Hochberg method (Benjamini and Hochberg, 1995) and for the association of changes Spearman rank correlation coefficient was used. (Wechsler, 1997) For each measure, the Page 44

46 relative change [RC= mean (testt2 testt1) x 100 /mean testt1)], the effect size [ES= mean (testt2 testt1)/sd testt1] and the standardised response mean [SRM= mean (testt2 testt1)/sd (testt2 testt1)] were calculated to determine the magnitude of change. For both effect size indices, ES and SRM, the standard values are: , Small effect; , Moderate effect; and 0.80, Large effect. (Kazis et al., 1989, Husted et al., 2000, Crosby et al., 2003) The number needed to treat (NNT) for improvement >1 SEM in the Apo group was calculated. (Guyatt et al., 2002, Schünemann et al., 2006, Wyrwich et al., 2005, Rejas et al., 2008) 1.16 Results Seventeen Apo (mean age 59.5±11.7 years, disease duration 12.05±4 years, median HY score 4) were started on Apomorphine infusion using a standardised local initiation protocol and the data compared to 17 C patients (mean age 66.4±7.0 years, disease duration 13.23±4.7 years, median HY score 3.9) over a period of 12.5±11.5 months. At baseline, there were no significant differences between Apo and C with respect to age, gender and PD duration and both groups were on comparable levodopa and dopamine agonist treatments and had resistant on-off fluctuations and dyskinesias. The levodopa equivalent (LED) doses in both groups were similar between Apo ( ± mg) and C ( ± mg) at baseline. At baseline, the UPDRS 3, NMSS-Total and PDQ-8 showed significant differences between C and Apo group, the C group being lower in scores (indicative of better state). During follow-up, the C group showed worsening in PDQ-8 (Table 5) while Apo showed highly significant improvements in UPDRS 3 (p=0.0003), UPDRS 4 (p=0.0003) PDQ-8 (p=0.001) and NMSS total (p=0.0003, Table 5). The LED increased significantly during follow-up in the C group ( ± to Page 45

47 ± mg, p=0.001) while in the Apo, LED decreased significantly ( ± to ±282.29, p<0.0001) with improvement affecting more than 90% of patients and large effect size. In the Apo group, moderate to large effect size on the NMSS domains of sleep, mood/apathy, attention, gastrointestinal, urinary, and miscellaneous was evident (Table 5). In the Apo group, there was a high correlation between change in UPDRS 3 and PDQ-8 (rs= 0.85) and moderate between change in NMSS and PDQ-8 (rs= 0.44). The NNT was lower than 2 for >1 SEM improvement of UPDRS 3, NMSS, and PDQ-8 total scores. The average NNT for NMSS domains was 3.95 (Table 7) 1.17 Discussion Our key observations from this study are as follows: 1. There was a beneficial effect of Apomorphine infusion on the whole of NMSS observed over one year average follow-up. (Table 7) 2. Patients with visual hallucinations, illusions and paranoid ideations, typically thought to worsen after dopamine agonist treatment, reported no worsening after Apomorphine infusion. 3. Apomorphine infusion resulted in significant improvement in some specific NMS such as hyperhidrosis, nocturia, urgency of micturition, and fatigue. Our real life study included patients who may be excluded in randomised clinical trials and so we used a pragmatic post-treatment surveillance method. The follow-up period is therefore variable as clinical appointments were based on the discretion of the clinicians or local arrangement methods as also reported from a large multicentre study of Apomorphine infusion. (Garcia Ruiz et al., 2008) Page 46

48 Apomorphine demonstrated benefit with a moderate to large effect size in all domains of NMSS (Table 5) along with UPDRS and PDQ-8. These data for the first time indicate efficacy of Apomorphine infusion on NMS and improvement in HrQoL similar to the effect of intra-jejunal levodopa gel infusion in advanced PD as shown in a previous report from our group. (Honig et al., 2009) The NNT confirmed a beneficial effect for most patients, both for motor and non-motor manifestations (Table 7). We also report non-worsening of the perceptual and attention domain of NMSS (Table 5) following Apomorphine, contrary to what is to be expected with the use of a potent dopamine agonist. In part this may be related to the apomorphine structure containing a piperidine moiety, which has a potent antipsychotic action. (Poewe and Wenning, 2000, Chaudhuri and Clough, 1998, Manson et al., 2002, van Laar et al., 1998) Improvement of psychosis was reported in 1978 in a placebo controlled study of Apomorphine in 18 chronic schizophrenic patients. (Tamminga et al., 1978) In PD, open label studies had suggested safety of Apomorphine infusion in those with neuropsychiatric problems on oral therapy. (van Laar et al., 1998, Ellis et al., 1997) Tolerability of Apomorphine infusion in our study may also be partly due to a significant reduction in LED and cessation of therapy with other oral dopamine agonists, except the use of Rotigotine patch in some cases. The NMSS has been validated in over 600 patients and allows assessment of specific NMS within one instrument. (Chaudhuri et al., 2007, Martinez-Martin et al., 2009) Using NMSS, we report significant beneficial effects on sleep disturbances such as onset insomnia and restless legs while excessive daytime sleepiness was not worsened. Other NMS effects included a significant improvement in fatigue, motivation, anxiety, flat mood, anhedonia, attention Page 47

49 deficit, dribbling of saliva, urinary dysfunction, particularly urgency and nocturia, and hyperhidrosis (Table 6). Some of these NMS may have an underlying dopaminergic basis while in others the improvements indicate successful amelioration of subtypes of non-motor off periods. (Witjas et al., 2002) Our data also supports data from previous open-label studies reporting a beneficial effect of Apomorphine infusion on sleep (notably severe insomnia), anhedonia and night time pain and cramps suggestive of restless legs syndrome. (Reuter et al., 1999, van Laar et al., 1998) This is not a randomised controlled study and we do not have a true control group. However, data from randomised controlled trials may occasionally have limited external validity. (Relton et al., 2010) We studied small number of patients although by comparison, the only other comparative study addressing Apomorphine and Deep brain stimulation included 13/12 patients only. (De Gaspari et al., 2006) Similarly, the only comparative study of Apomorphine versus levodopa infusion included 4 patients only. (Nyholm et al., 2009) In our study, a moderate to large effect of Apomorphine was evident with the current sample size. In conclusion, this observational post-treatment surveillance based study indicates the non-motor effects of Apomorphine infusion. Specific non-motor effects of Apomorphine are highlighted, calling for controlled studies with nonmotor primary endpoints. Page 48

50 Table 5: Changes following Apomorphine infusion and continuing conventional therapy (comparator) in motor, non-motor, and quality of life dimension Control UPDRS-Motor Baseline Follow-up p Baseline Follow-up p (9.68) (11.42) (8.21) exam. UPDRS- Apomorphine (12.80) 7.93 (5.43) 7.00 (4.46) (6.43) 3.53 (3.52) (2.97) 1.18 (2.90) (5.63) 2.76 (3.51) (9.58) (9.32) (11.47) (9.63) (10.33) 8.06 (8.78) (19.85) Complications NMSSCardiovascular Sleep Mood/apathy (13.04) Perceptual 2.23 (5.03) 2.59 (6.26) (6.92) 1.88 (3.35) 0.04 Attention 6.00 (8.40) 7.18 (7.76) (9.62) 8.71 (7.75) (5.97) 7.12 (6.49) (7.35) 4.41 (5.11) Urinary 4.29 (3.57) 6.23 (4.26) (8.93) 5.71 (6.72) Sexual 3.12 (6.58) 3.29 (6.12) (5.96) 2.00 (3.94) 0.42 Miscellaneous 4.12 (5.67) 4.29 (5.55) (14.54) 9.47 (9.70) (65.43) (43.40) (37.65) (23.10) (17.57) Gastrointestinal NMSS-Total score PDQ-8 (45.39) (19.80) (21.54) Benjamini-Hochberg correction: p<0.027 UPDRS: Unified Parkinson s disease Rating Scale. NMSS: Non-Motor Symptoms Scale. PDQ-8: Parkinson s disease Questionnaire-8 items Page

51 Table 6: Change in the items of Non-Motor Symptoms Scale (Apo infusion group) Items & Domains Baseline Follow-up p* Cardiovascular 1 Light-headedness 3.53± ± Fainting 1.23± ± Sleep/ Fatigue 3 Daytime sleeping 4.12± ± Fatigue 7.29± ± Difficult falling asleep 5.12± ± Restless legs 5.53± ± Mood/ Cognition 7 Lost interest surroundings 2.94± ± Lack motivation 3.65± ± Nervous 6.12± ± Sad 5.53± ± Flat mood 2.23± ± Difficult experiencing pleasure 2.76± ± Perceptual problems 13 Hallucinations 1.41± ± Delusions 1.53± ± Double vision 1.65± ± Attention/ Memory 16 Problems with concentration 5.47± ± Forget recent events 4.06± ± Forget doing things 3.29± ± Gastrointestinal 19 Dribbling saliva 2.23± ± Swallowing 2.00± ± Constipation 3.12± ± Urinary 22 Urgency 3.71± ± Frequency 2.59± ± Nocturia 4.41± ± Sexual Function Page 50

52 25 Altered interest in sex 1.59± ± Problems having sex 0.94± ± Miscellaneous 27 Unexplained pains 1.76± ± Lost taste/smell 4.41± ± Change in weight 3.00± ± Excessive sweating 7.59± ± * Wilcoxon test Significant after Benjamini-Hochberg correction, p<0.025 Page 51

53 Table 7: NNT in patients treated with Apomorphine % of patients SEM NNT improving >1 SEM UPDRS-Motor exam UPDRS-Complications NMSS- Cardiovascular Sleep Mood/apathy Perceptual Attention Gastrointestinal Urinary Sexual Miscellany NMSS-Total score PDQ NNT= (1/ % Improved) x 100 (n= 17) Page 52

54 7 Chapter 4: Intrajejunal levodopa infusion in the UK: comparative review of motor and non-motor effects in treated versus untreated patients with advanced Parkinson s disease This chapter is based on: REDDY, P., MARTINEZ-MARTIN, P., RIZOS, A., MARTIN, A., FAYE, G. C., FORGACS, I., ODIN, P., ANTONINI, A. & CHAUDHURI, K. R Intrajejunal levodopa versus conventional therapy in Parkinson disease: motor and nonmotor effects. Clinical Neuropharmacology, 35, Abstract Seventeen patients with advanced Parkinson s disease (PD) were treated with intrajejunal levodopa infusion (IJLI) and compared with a matched group of nine patients (termed comparator (C)) not given IJLI due to funding restriction by primary care trusts (PCT) in the UK, although considered to be clinically eligible for IJLI). Assessments were baseline and follow-up (six months) with Hoehn and Yahr (HY) staging, unified PD rating scale (UPDRS-III and IV), Parkinson s disease questionnaire (PDQ-8, quality of life (QoL)) and non motor symptom scale (NMSS). Baseline characteristics were comparable between the groups. The IJLI-treated group showed highly significant improvements in UPDRS-III (p=0.005), UPDRS-IV (p=0.0004), total NMSS score (p=0.004) and QoL (p=0.01) while the C group showed no change in these parameters. A large effect size of IJLI was seen in treated patients for UPDRS-III (1.13), IV (1.52), NMSS score (0.82) and QoL (1.12) while continuing conventional treatment registered no effect in C. This study confirms the robust effect of IJLI on motor and non motor symptoms, and QoL in advanced PD as described in open label studies, but additionally points to need for such treatment in those denied this therapy due to centrally dictated funding policies leading to inequalities in health care. Page 53

55 1.19 INTRODUCTION Advanced Parkinson s disease (PD) is a complex, fluctuating condition characterised by motor and non-motor complications: wearing off, unpredictable off periods and dyskinesias that occur despite combinations of per oral therapies and remains a clinical challenge. At this stage, treatment options include Deep brain stimulation of subthalamic nucleus (DBS-STN), Apomorphine infusion (Apo) and more recently, intrajejunal levodopa (IJLI) infusion. Experience with Apomorphine and DBS-STN is extensive, while, in the UK, intrajejunal levodopa infusion has been introduced relatively recently and clinical use is consequently more limited often due to its expense. IJLI delivers continuous levodopa via a portable delivery system with a percutaneous endoscopic gastrojejunostomy (PEG-J) tube, which reduces on/off fluctuations, thereby resulting in a more stable clinical control than oral levodopa.(nyholm et al., 2003) In a prospective 12-month assessment of clinical and quality of life (QoL) changes in seven patients with advanced PD, severe motor fluctuations and dyskinesia (the DIREQT study), duration of off periods and time with disabling dyskinesia was significantly shortened (p<0.01). (Antonini et al., 2007) There were significant improvements in the UPDRS-II (activities of daily living) and -IV (motor complications) in the on condition (p<0.02) and in four of PDQ-39 domains. (Antonini et al., 2007) Furthermore, in a prospective, open-label, observational study, 22 advanced PD patients (mean age 58.6 years, duration of disease 15.3 years) showed a statistically significant beneficial effect in six of the nine domains of the non motor symptom scale (NMSS) at six months with improvements in cardiovascular, sleep/fatigue, attention/memory, gastrointestinal, urinary, and miscellaneous (including pain and dribbling) domains and for the total score of NMSS scale, paralleling Page 54

56 improvement in motor symptoms (UPDRS-III motor and IV complications in best on state). (Honig et al., 2009) However, there are no studies of IJLI in PD that are comparative with a group on conventional therapies in a real-life setting. Furthermore, in the UK, the funding of these treatments is cost-driven and based on decisions made by Primary Care Trusts (PCTs) where the panel consists of general practitioners, pharmacists and managers. For each case that is clinically thought to be eligible for IJLI by the King s team, an individual funding request needs to be lodged with the relevant PCT to satisfy the panel that the money spent on the therapy is cost effective. While unique to the UK market at the moment, such approaches may become more common place across Europe in future in relation to expensive therapies for advanced PD. Consequently, there is a real need for more comparative data to demonstrate the benefits of the advanced PD therapies relative to conventional treatments on patient-orientated outcomes and support the economic case for a particular treatment choice. King s College Hospital, London, has the largest cohort of IJLI treated patients in the UK while also registering a number of patients denied IJLI treatment by PCTs on the basis of funding restrictions. Clinically all these patients are deemed suitable for IJLI while the funding refusal is unpredictable thus providing a serendipitously matched and randomly selected group of patients who can be observed in relation to continuing conventional therapy as a comparator group to those given IJLI PATIENTS AND METHODS Participants Between 2009 and 2011, 32 patients were referred to King s College Hospital, London, for IJLI therapy. Of these patients, 26 were considered clinically eligible Page 55

57 for IJLI therapy (based on levodopa responsiveness, dyskinesias and unsuitability/intolerance to Apomorphine/DBS-STN therapies), 17 were granted funding for this treatment by relevant PCTs while nine, considered eligible for IJLI by the clinician were denied funding by the local PCT and six were considered unsuitable for IJLI by the clinician. As described above, patients who were eligible for IJLI but refused funding for this treatment constituted the control (C) group. Medications Patients on IJLI were given no additional medications, other than Rotigotine patch or oral levodopa at night to avoid a night time rebound off state as infusion therapy was usually sustained for daytime hrs. The dose of IJLI was optimised to suit the individual patient and included levodopa equivalent doses of pre-existing oral and other therapies. The C group were continued on best optimised oral therapies in the face of no other alternatives. Evaluations Assessments were conducted at baseline and follow-up (six months) and included Hoehn and Yahr (HY) staging, UPDRS-III and IV, PDQ-8 and NMSS score (subdomains and total) as previously published. (Honig et al., 2009) All assessments were performed by a single clinician. For analysis, baseline data were compared with the last available follow-up visit at six months of each patient in the on condition. All clinimetric details of the instruments used in this study are well established and discussed in the paper by Honig et al. (Honig et al., 2009) 1.21 Statistical analysis Chi-squared and Fisher s exact tests were used for comparison of proportions. Wilcoxon test was applied for comparison between baseline and follow-up Page 56

58 scores, while Mann-Whitney test was applied for comparison between groups. Multiple comparisons were corrected by the Benjamini-Hochberg method (Benjamini and Hochberg, 1995) and, for the association of changes, the Spearman rank correlation coefficient was used. For each measure, the relative change (RC=mean [testt2-testt1] x 100/mean [testt1]) (Deyo and Centor, 1986) and the effect size (ES=mean [testt2-testt1]/sd [testt1]) was calculated to determine the magnitude of change. (Wyrwich et al., 2005, Norman et al., 2007) For the effect size index, ES, the standard values are: , small effect; , moderate effect; 0.80, large effect. (Wyrwich et al., 2005, Husted et al., 2000) Ethical approval: This study is part of a five year global study (NILS) which is evaluating progression and incidence of motor and NMS in PD and the effect of therapy on the motor and non-motor aspects of PD. The study (a global one) has thus far recruited 1088 patients with baseline data ranging from untreated PD to advanced, the latter group (from Europe and those suitable for advanced therapies) being screened for the purpose of this report. The study was approved by individual research ethics committees of all participants and is led by a national portfolio adoption of the study in UK (DenDRoN, portfolio number 10084) Patients All patients could be assessed at the 6-month time point from the start of therapy and there were no discontinuation of IJLI. Analyses are therefore based on 17 patients who received IJLI and nine who did not. We have continued to follow up these patients, with the longest follow up on IJLI being 3 years. Of these 2 patients have died from causes unrelated to IJLI therapy. Baseline variables Page 57

59 Patients in the IJLI and C groups were of similar age (p=0.13; Table 8), no significant differences for gender, PD duration, HY score, UPDRS-III or IV and PDQ-8 at baseline (all p>0.2; Table 8). However, NMSS score was higher (worse) in the IJLI group than in the C group (p<0.02; Table 8). The mean dose of IJLI was 1996 ± 675 mg/day (range 1100 to 3204 mg/day). The mean daily duration of treatment was 16 hours. The C group was on a combination of per oral therapies: dopamine agonist, levodopa and entacapone and these were continued after IJLI treatment was refused. Baseline to follow-up: intra-group comparison There were no significant changes in any parameters from baseline to follow-up in the C group, with a trend towards worsening score in PDQ-8, while in the IJLI group highly significant improvements in UPDRS-III (p=0.006); UPDRS-IV (p=0.0004); NMS score for sleep (p=0.02), gastrointestinal tract (p=0.0012), urinary (p=0.0015), sexual function (p=0.026) and total score (p=0.004); PDQ-8 (p=0.017) were observed (Figure 3). Baseline to follow-up: inter-group comparison The difference in the change from baseline to follow-up between the IJLI and control groups was statistically significant for UPDRS-III and IV, PDQ-8, NMSS total and the following NMSS subdomains: cardiovascular disorders, sleep, gastrointestinal problems, urinary difficulties, sexual function and miscellaneous (all p<0.027 for the IJLI versus the C group, Figure 3). Group treated with IJLI: effect size The percentage of patients in the IJLI group who improved by >½ a standard deviation from baseline to follow-up were 82.35% (UPDRS-III), 88.24% (UPDRS-IV), 70.60% (NMSS total score) and 70.60% (PDQ-8). A large effect Page 58

60 size of IJLI on UPDRS-III (1.13), IV (1.52), NMSS score (0.82) and PDQ-8 (1.12) was seen in treated patients DISCUSSION A robust effect of IJLI on both motor and non motor symptoms and QoL in advanced PD is demonstrated in this small review, while a similarly affected PD population denied this therapy showed no beneficial effects on these parameters when continuing conventional therapy. Our real life study included patients who may be excluded in randomised clinical trials and so we used a pragmatic post-treatment surveillance method. Given the small sample size in this study, effect size gives the most robust indication of the magnitude of response observed in IJLI treated patients; for motor, non motor and QoL measures this was considered large. (Wyrwich et al., 2005, Husted et al., 2000) The greater improvements observed among IJLI treated patients compared with the C group in NMSS scores at follow-up may, at least in part, be due to the more severe NMSS scores recorded among patients in the IJLI group at baseline. However, the two groups were matched in relation to UPDRS III and IV domains as well as PDQ-8 in all of which a statistically significant and greater beneficial effect was observed with IJLI. To our knowledge, this is the first real-life study comparing the effects of treating eligible advanced PD patients with IJLI with those continuing standardised medications in similar patients. The C population is, in effect, naturally selected since it comprises patients considered to be clinically suitable for IJLI treatment, but for whom funding for this treatment has been refused based on the funding issues associated with individual PCT s. It is therefore a reasonable comparator population for the purposes of comparison of the impact of treating versus not treating comparable Page 59

61 patients with IJLI. Furthermore, the study was carried out in a single centre with assessment of patients done by a single clinician at a single centre to minimise the potential for variation. Peripheral neuropathy, largely axonal in nature, related to levodopa infusion has been recently reported, however in our cohort so far no such event has been observed on surveillance. We noted normal vitamin B12 and folate level in all patients except borderline reduction of B12 level in one patient. Two patients have died after 2 years on IJLI from unrelated causes. Although a small study, the findings presented help us to understand the benefit of treating with IJLI in carefully chosen patients. Further larger scale comparative studies are needed to examine this effect in greater detail. Page 60

62 Table 8: Baseline patient characteristics in the IJLI group and Controls IJLI Control ± ± 6.08 Male 11 7 Female ± ± (3-5) 3.7 (3-5) UPDRS-III (mean) ± ± 9.27 UPDRS-IV (mean) 9.23 ± ± 5.99 NMSS (mean) ± ± PDQ-8 (mean) ± ± Age (mean) Gender PD duration (mean) HY score (stage) (n) Page 61

63 10 0 Mean change from baseline * * * -10 * * * * * * Control IJL * *p<0.027 vs control; p<0.027 vs baseline -50 CV= cardiovascular GIT=gastrointestinal tract PDQ-8= Parkinson s disease questionnaire (quality of lilfe) UPDRS-III= unified Parkinson s disease rating scale (motor) UPDRS-IV= unified Parkinson s disease rating scale (motor complications) Figure 3: Mean change from baseline in UPDRS and Non Motor Scale scores in the IJLI versus the control group. The Y axis represents the mean change from baseline in the UPDRS scores, NMSS scores and the PDQ-8 scores. (Reddy et al., 2012) Page 62

64 8 Chapter 5: Bilateral subthalamic stimulation improves aspects of nonmotor symptoms in Parkinson s disease This chapter is based on: Haidar Salimi Dafsari, K. Ray Chaudhuri, Mark William Kellet, Pablo Martinez-Martin, Christiane Herchenbach, Alexandra Rizos, Prashanth Reddy, Monty Silverdale, Keyoumars Ashkan, Mike Samuel, Angelo Antonini, Lars Timmermann. Bilateral subthalamic stimulation improves th aspects of non-motor symptoms in Parkinson s disease. 17 International Congress of Parkinson s disease and Movement Disorders Late-Breaking Abstracts, MDS Study Group Abstracts. p Abstract BACKGROUND Subthalamic nucleus Deep brain stimulation (DBS-STN) is well established to improve motor symptoms and quality of life (QoL) in patients with Parkinson s disease (PD). While non-motor symptoms (NMS) are crucial for QoL in these patients, only neuropsychiatric and neuropsychological symptoms have been systematically studied in a longitudinal design thus far. However, these are only a part of NMS in PD. We hypothesised that DBS-STN is associated with a beneficial effect on a range of NMS. METHODS In this multicentre, open, prospective, international study we investigated nonmotor effects of DBS-STN in real-life use. We evaluated using NMS Scale (NMSS) and Questionnaire (NMSQ), PD Questionnaire-8 (PDQ-8), Scales for Outcomes of PD (SCOPA) motor examination and complications, and activities of daily living (ADL) preoperatively and at 6 months follow-up (6MFU) in 60 consecutive patients (35 male, mean age: ±7.84 years, mean disease duration: ±4.22 years). RESULTS All outcomes improved significantly at 6MFU (PDQ-8, p=0.006; ADL, p=0.012; all others, p<0.001; Wilcoxon signed-rank, respectively paired t-test; Bonferroni- Page 63

65 correction). Post-hoc analyses of NMSS domains showed a significant reduction of sleep/fatigue and miscellaneous domains (p 0.001), perceptual problems/hallucinations (p=0.036), and urinary (p=0.018) scores. Effect sizes were moderate effect for NMSS, and motor complications, large for motor examination, and small for other outcomes. CONCLUSIONS This study provides evidence that bilateral DBS-STN improves NMS burden in patients with PD and opens the door to a more balanced evaluation of DBSSTN outcomes. Further randomized studies are needed to confirm these findings and compare DBS-STN effects on NMS to other invasive therapies of advanced PD Introduction Subthalamic nucleus Deep brain stimulation (DBS-STN) is well established for the symptomatic treatment of Parkinson s disease (PD) improving motor symptoms, activities of daily living (ADL), and quality of life (QoL) (Deuschl et al., 2006, Follett et al., 2010, Weaver et al., 2009). Non-motor symptoms (NMS) play a crucial role for QoL in patients with PD (Chaudhuri et al., 2011, Barone et al., 2009). Long-term effects of DBS-STN on neuropsychological (Frankemolle et al., 2010, Jahanshahi et al., 2000) and neuropsychiatric symptoms (Witt et al., 2008, Castelli et al., 2006) have been studied. However, these symptoms contribute only to a part of NMS in patients with PD. Previously published studies on a wider range of NMS have methodological limitations due to a lack of objective clinician-based assessment (Nazzaro et al., 2011) and small cohort sizes of 10 subjects. (Reich et al., 2011, Hwynn et al., 2011) In this study, we therefore investigated prospective data using validated nonmotor clinician-based and self-assessment outcome measures collected on 6 Page 64

66 months follow-up (6MFU) of a multicentre registry trial. We hypothesized that DBS-STN is associated with a reduction of NMS burden in patients with PD. Furthermore, to investigate the relationship between changes of NMS, motor and QoL outcomes from baseline, an exploratory aim of our study was to analyse their correlation Material and methods Design This was a multicentre, open, prospective, European registry study (Cologne, London, and Manchester) of a subgroup of the Non Motor Symptoms study group of the International Parkinson s disease and Movement Disorders Society with a longitudinal follow-up (EuroInf study). Centres were chosen on the basis of experience in using motor and non-motor scales as well as performing DBS-STN surgery and therapy. Subjects All patients were diagnosed according to British Brain Bank criteria (Hughes et al., 1992a) and were screened for treatment with DBS-STN in accordance with consensus criteria of the International Parkinson s disease and Movement Disorders Society (IPMDS) due to an insufficient medical control of motor symptoms. All patients responded to levodopa with >30% improvement of motor symptoms, assessed by Unified Parkinson s Disease Rating Scale (UPDRS)-III. Preoperatively, neuropsychiatric and neuropsychological assessments of patients were performed by consultant psychiatrists and neuropsychologists. Exclusion criteria were clinically significant psychiatric diseases and Minimal Mental State Examination scores < neuropsychological impairment. Ethical approval Page points as an indicator of

67 The study was approved by the local ethics committees (Master vote: , Cologne; United Kingdom: National Research Ethics Service South East London REC 3; 10/H0808/141; NIHR portfolio (UKCRN) number 10084) and carried out in accordance with the Declaration of Helsinki. Clinical assessment Motor symptoms and NMS were assessed preoperatively in clinical MedON (On state whilst on medication) state and postoperatively on 6MFU in clinical MedON/StimON (DBS-STN stimulation on) before an adjustment of stimulation parameters thus reflecting a real-life state. (A) Motor impairment was assessed with the Short Parkinson s Evaluation Scale (SPES) / Scales for Outcomes in Parkinson s disease (SCOPA) motor examination which has been shown to highly correlate with the MDS-UPDRS motor examination. (Martinez-Martin et al., 2005) Motor complications and ADL were also assessed with the SPES/SCOPA scale which correlates with corresponding parts of the UPDRS scale. (Marinus et al., 2004) Additionally, for a subset of patients, we examined UPDRS directly. (B) Non-motor symptoms were examined with two tests: (1) We collected data of the Non-Motor Symptom Scale (NMSS), a clinicianadministered scale which tests for nine domains of NMS with 30 questions. (Chaudhuri et al., 2007) These questions are assessed with weighted scores of symptom severity and frequency. Severity of symptoms are rated by a range of 0 (none) to 3 points (major source of distress and disturbance for patients) and frequency is assessed by a range of 1 (<once per week) to 4 (daily or all the time) points. The theoretically possible maximum NMSS total score (NMSS-T) is 360, the minimum score 0. Page 66

68 (2) Furthermore, we also collected data of the Non-Motor Symptoms Questionnaire (NMSQ), a patient-based self-assessment questionnaire with dichotomous answers for the presence of NMS (Chaudhuri et al., 2006a) with a maximum NMSQ-T of 30 corresponding to the number of questions. (C) Patients QoL was investigated with the self-assessment rating scale Parkinson s Disease Questionnaire-8 (PDQ-8). (Jenkinson et al., 1997b) The PDQ-8 was designed as a shortened version of the PDQ-39 and quantifies the frequency of eight aspects of daily living with impact on the QoL. PDQ data is provided as PDQ-8 Summary Index (SI). (Jenkinson et al., 1997a) (D) In addition, we recorded the therapeutic medical regimen and stimulation parameters. The levodopa-equivalent daily dose (LEDD) was computed according to a method previously published by Tomlinson et al. (Tomlinson et al., 2010) and the total electric energy delivered (TEED) was estimated according to a method previously published by Koss et al. (Koss et al., 2005) 1.26 Statistical analysis All outcome parameters were checked for normality distribution with the Shapiro-Wilk test or, when necessary, Kolmogorov-Smirnov test with Lilliefors correction. For a longitudinal analysis of these parameters Wilcoxon signedrank test was computed, respectively Student s paired t-test when criteria for parametric tests were fulfilled. The Bonferroni method was used to correct Type I errors for multiple comparisons. All values are stated as mean ±SD, when the aforementioned criteria were fulfilled, unless stated otherwise. To investigate the size of DBS-STN effects, in addition to the difference between baseline and follow-up scores, the relative change (RC = [(mean Testbaseline - mean Test6MFU) / (mean Testbaseline]) and the Cohen s effect size (Cohen, 1977) were calculated. Effect size values 0.80 are considered large Page 67

69 effect, moderate effect, and small effect. (Honig et al., 2009) Half a SD at baseline (½SD TestBaseline) was used in a method previously applied to QoL outcomes to detect treatment responders. (Honig et al., 2009) The number needed to treat (NNT) for an improvement ½SD TestBaseline was calculated for each outcome. (Martinez-Martin and Kurtis, 2012) Also, to evaluate changes of specific aspects of NMS, a post-hoc analysis of NMSS domain scores was used with the above mentioned methods. Furthermore, to investigate the relationship between all outcome parameters, Spearman s rho was computed, respectively Pearson s correlation analyses when criteria were fulfilled, for changes of values from baseline. To examine the relative importance of specific NMSS domains for QoL, correlations between all domains and PDQ-8 SI were calculated. These relationships are also reported as changes from baseline Results Data of 60 consecutive patients (35 male) aged years (±7.84) with long histories of PD (10.45 ±4.22 years) and moderate to high LEDD ( ±475.93) at baseline are presented. The median Hoehn & Yahr score was 2.75 (interquartile range: 2-3). The assumption of normal distribution was violated for PDQ-8 SI (p=0.015), and SPES/SCOPA motor examination (p=0.019), ADL (p=0.006) and motor complications scores (p=0.006), but not for NMSS-T (p=0.066), and NMSQ-T (p=0.200). Bilateral DBS-STN in patients with PD significantly improved all outcome parameters applied in this study (see table 10). In particular, DBS-STN significantly reduced NMS in the clinician-based NMSS-T (p<0.001) as well as Page 68

70 in the patient-based NMSQ-T (p<0.001). Using NMSS-T as a main outcome parameter the statistical power was 0.88 (α=0.05; two-sided test). Post-hoc analyses of NMSS domains, also reported in table 9, showed a significant reduction problems/hallucinations of sleep/fatigue (p=0.036), urinary (p<0.001), symptoms perceptual (p=0.018), and miscellaneous domain (p=0.001) scores. An illustration of NMSS domain scores is included in figure 4. In the latter domain questions regarding excessive sweating (p<0.001) and change in the ability to smell and taste (p=0.001) were significant. Additionally, trends were observed for cardiovascular (p=0.096) and gastrointestinal (p=0.082) symptoms. Furthermore, PDQ-8 SI (p=0.006), ADL (p=0.012), and motor outcomes (motor examination and complications, p<0.001 respectively) improved significantly. The magnitude of improvement is indicated in table 10. DBS-STN had a moderate effect size on NMSS-T (0.50) and SPES/SCOPA motor complications (0.66), a large effect size on SPES/SCOPA motor examination (0.81), and a small effect size on NMSQ-T (0.48), PDQ-8 SI (0.47), and SPES/SCOPA ADL (0.43). Mean RC of all stated outcome parameters was % (±7.16). The mean effects size of all stated outcome parameters was 0.57 (±0.14) resulting in a NNT of 2.12 (±0.24). Around 47% of patients treated with DBS-STN improved ( ½SD TestBaseline or more) their QoL and around 42% improved NMS as indicated by the NNT values. UPDRS-III and -IV were available for a subset of patients (n=42 and n=43) and improved significantly on 6MFU (see table 13) with large effect size (0.90 and 0.91) resulting in a NNT of 1.62 and 1.60, respectively (see table 14). Page 69

71 LEDD reduction was 43.22% from (±475.93) to (±337.25) and reached statistical significance (Student s paired t-test: p<0.001). Mean TEED was 88.51µJ (±83.78) at 6MFU. There was no significant inter-hemispheric difference as the median and inter-quartile range of the right STN were 36.30µJ and µJ and of the left STN were 63.47µJ and µJ (Wilcoxon signed-rank test: p=0.904). To explore the relationship between changes of outcome parameters from baseline to 6MFU correlation analyses (table 11) was computed which showed a significant relationship between the improvements of NMSS-T and PDQ-8 SI (p=0.001) as well as NMSQ-T (p=0.026). Interestingly, however, the reduction of NMSS-T was not significantly correlated to SPES/SCOPA motor examination and complications or ADL scores. The correlation analyses on the aforementioned subset of patients for which UPDRS data was available showed a significant relationship between the improvements of NMSS-T with UPDRS-IV (p=0.012), but not with UPDRS-III (table 15). Further correlation analyses between PDQ-8 SI and NMSS domains (table 12) indicated a significant relationship with sleep/fatigue (p=0.016), mood/cognition (p<0.001), and attention/memory (p=0.001). Noteworthy, there was no correlation between PDQ-8 SI and improvements of urinary and the miscellaneous NMSS domains in our cohort, although these domains significantly improved from baseline. During the study period, no significant adverse effects were observed Discussion This multicentre European study provides evidence that bilateral DBS-STN improves NMS burden in patients with PD. Also, in accordance to previous studies investigating changes from baseline to 6MFU, in our cohort DBS-STN Page 70

72 significantly improved motor outcomes and QoL. (Deuschl et al., 2006, Martínez-Martín et al., 2002) Studies investigating changes of motor examination from baseline MedON to 6MFU MedON/StimON report comparable results for motor examination improvement. (Varma et al., 2003, Chou et al., 2013) Incorporating good clinical practice we assessed NMS, QoL and motor symptoms using validated scales in a multicentre approach. Our main observations indicate that continuous bilateral subthalamic DBS-STN significantly improves NMS, in particular its following aspects: - Sleep/fatigue: Our results of a significant improvement of this domain are in accordance with previous studies reporting subjective and objective improvements of sleep efficiency, quality, and architecture after continuous bilateral DBS-STN. (Iranzo et al., 2002, Monaca et al., 2004) - Urinary symptoms: Previous studies have reported immediate effects of DBS-STN on bladder control, (Seif et al., 2004) most likely mediated through a modulation of information transfer between the periaqueductal grey area and the cortex. (Herzog et al., 2008) To our knowledge this is the first report of long-term effects of DBS-STN on urinary symptoms. - Perceptual problems/hallucinations: Our results are in line with previous studies which have shown that DBS-STN may lead to an improvement of hallucinations in patients with PD. (Yoshida et al., 2009) Although no correlations were found between LEDD reduction and an improvement of this NMSS domain (data not shown), a possible mechanism seems to be an amelioration of these symptoms depending on a reduction of LEDD below patient-specific individual thresholds. Further studies are needed to Page 71

73 investigate the relationship between DBS-STN and LEDD effects on perceptual problems and hallucinations. - In the miscellaneous domain: We found beneficial effects on olfactory symptoms and excessive sweating. While previous studies have shown a beneficial immediate (Wolz et al., 2012) and long-term effect (Trachani et al., 2010) of DBS-STN on sweating probably due to a reduction of ON/OFF fluctuations, the effect on olfactory symptoms may seem more surprising. However, a longitudinal study investigating DBS-STN effects on olfaction showed a significant improvement of odor identification thresholds, but not detection thresholds, on 6 and 12MFU, (Guo et al., 2008) possibly indicating an improvement of cognitive odor information processing. Our results support these previous findings suggesting that DBS-STN may have a beneficial effect on olfaction. Additionally trends were observed for a beneficial effect of DBS-STN on the following aspects of NMS: - Cardiovascular symptoms: Previous studies have shown an immediate improving effect of DBS-STN on orthostatic regulation. (Stemper et al., 2006) In a study on immediate effects of DBS-STN on a range of NMS the severity of dizziness improved in the StimON-state. (Wolz et al., 2012) To our knowledge no longitudinal data of this kind have been published. The trend of improvement of cardiovascular symptoms observed in our study could indicate long-term effects of DBS-STN on cardiovascular symptoms; however, further studies are needed. - Gastrointestinal symptoms: A study by Arai and co-workers provided evidence for a long-term improvement of gastrointestinal dysfunction after DBS-STN. (Arai et al., 2012) The observed trend of improvement of Page 72

74 gastrointestinal symptoms in our study may support the aforementioned findings. Also, a meta-analysis by Stowe and colleagues has shown an improvement of a range of NMS, including hallucinations, cardiovascular, gastrointestinal, and sleep and fatigue symptoms associated with a reduction of LEDD (Stowe et al., 2011). The observed improvements of aforementioned aspects of NMS may therefore also reflect an indirect response to an LEDD reduction. However, DBS-STN and LEDD reduction may also simultaneously exert effects, synergistic or competing, on NMS. A separation of these two seems difficult in a real-life study. Here we merely present data of net outcome of DBS-STN and LEDD change following DBS-STN therapy initiation. Further studies may help to further elucidate the interplay of effects and the weight of the individual components. Understanding this issue may also help to answer a closely connected question, as to how DBS-STN effects on NMS can be explained, when LEDD reduction is not responsible. In theory, at least two ways of action seem possible: Firstly, a direct modulation of basal ganglia-thalamo-cortical loops has been discussed by which, e.g., autonomic centers of the thalamus, lateral frontal, and anterior cingulate cortex could be modulated thus leading to improvements of symptoms like sweating and bladder control. (Trachani et al., 2010, Herzog et al., 2008)Secondly, a spread of electric current to regions in proximity of the STN, by which, the PPN (pedunculopontine nucleus) could be modulated thus resulting in an improvement of sleep architecture. (Alessandro et al., 2010) An interesting finding of the current study is that the improvement of NMSS-T was significantly correlated with the improvements of QoL (PDQ-8 SI) and motor complications (UPDRS-IV) while the correlation between PDQ-8 SI and Page 73

75 motor examination (SCOPA/SPES motor examination and UPDRS-III) did not reach statistical significance. This confirms previous studies that NMS (NMSST) may be a greater determinant of QoL than motor impairment. (Chaudhuri et al., 2011) As a real-life study the current work has weaknesses. As a registry study this was not a randomized or placebo-controlled study with sham stimulation. Patients were selected and recruited in a consecutive fashion as per clinical routine and underwent a standard protocol for DBS-STN initiation in accordance to published international selection criteria for DBS-STN as well as standard agreed follow-up plans. Patient support provided by DBS-STN companies (e.g. an introductory training to patient programming and recharging devices) was not systematically assessed in this study, but support was available for all patients in participating study centres. Our study of non-motor effects of DBS-STN, as an invasive symptomatic therapeutic option, has one of the largest patient numbers for studies of this kind. The multicentre set-up of the study may ameliorate the bias caused by a single centre design. However, using clustering and stratification methods in cohorts with further extended patient numbers may allow a characterization of treatment responses of specific NMS subtypes. (Erro et al., 2013) The aim of this process is to tailor individual therapeutic approaches for patients with PD based on their profile of NMS and motor symptoms. To conclude, DBS-STN ameliorates NMS burden in a range of aspects of NMS. In our cohort around 40% of patients treated with DBS-STN improved their NMS. NNT results were consistent with RC and effect size results for all outcome parameters. Reports of these parameters are needed to better understand responses to different treatment strategies like DBS-STN, Page 74

76 Apomorphine, Intrajejunal Levodopa and conventional pharmacotherapy. Further studies on treatment responses of specific NMS subtypes to different treatment strategies may help to eventually provide a basis for individualised medicine for patients real-life requirements. Page 75

77 Table 9: Significant improvement of all non-motor outcomes with DBS-STN Baseline Follow-Up Mean SD Mean SD <0.001 Cardiovascular Sleep/fatigue * <0.001 Mood/cognition hallucinations * Attention/memory Gastrointestinal Urinary * Sexual function Miscellaneous * NMSQ total score <0.001 PDQ-8 SI motor examination <0.001 ADL motor complications <0.001 NMSS total score p NMSS domains Perceptual problems / SPES/SCOPA Post-hoc analyses of NMSS domains showed a significant improvement of these domains. Page 76

78 Table 10: RC, effect size and NNT with DBS-STN therapy Relative Effect Number change (%) size needed to treat Cardiovascular Sleep/fatigue * Mood/cognition hallucinations Attention/memory Gastrointestinal Urinary Sexual function Miscellaneous NMSQ total score PDQ-8 SI motor examination ** ADL motor complications * NMSS total score * NMSS domains Perceptual problems / SPES/SCOPA * Moderate effect size ** Strong effects size Page 77

79 Table 11: Spearman s rank correlations between outcomes with DBS-STN NMSS-T NMSQ-T Correlation Sig. (2-tailed) N PDQ-8 SI Correlation Sig. (2-tailed) N SPES / SCOPA Correlation motor examination Sig. (2-tailed) N SPES / SCOPA Correlation ADL Sig. (2-tailed) N SPES / SCOPA Correlation motor complications Sig. (2-tailed) N NMSQ-T PDQ-8 SI SPES / SCOPA SPES / motor examination SCOPA ADL.290* **.371** * **.311* ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). There were no missing data for NMSS-T, PDQ-8 SI, and SPES/SCOPA motor examination. Missing data for SPES/SCOPA ADL and motor complications were acceptable and for NMSQ-T negligible. Page 78

80 Table 12: Correlations between PDQ-8 SI and NMSS domains after DBS-STN Perceptual problems / PDQ-8 SI Cardio- Sleep / Mood / halluci- Attention / Gastro- Sexual vascular fatigue cognition nations memory intestinal Urinary function Miscellaneous Correlation *,496**.199,410** Sig. (2-tailed) N * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed). There were no missing data for PDQ-8 SI and NMSS domains.

81 (a) 6MFU (b) Cardiovascular 1,50 Sleep 1,25 1,00 0,75 0,50 Sexual function Mood 0,25 0,00 / cognition Urinary Perceptual Gastrointestinal Attention / (a)box plots and (b) radar chart of NMSS domains. Significantly improved domains are marked Figure 4: Box plots and radar chart of NMSS domains. with a black star. In (b) NMSS domain scores are normalized with respect to baseline values per subject. Blue area: baseline, copper area: 6MFU data. A bigger copper area reflects an improvement of the NMSS domain (computation: 2-6MFU/baseline). This chart is reproduced with permission from Haidar Salimi Dafsari and Lars Timmermann. Page 80

82 Table 13: Supplemental table 1 Significant improvement of UPDRS-III and -IV Baseline Follow-Up Mean SD Mean SD p UPDRS-III <0.001 UPDRS-IV <0.001 Table 14: Supplemental table 2 RC, ES and NNT of UPDRS-III and -IV Relative Effect Number change (%) size needed to treat UPDRS-III UPDRS-IV Page 81

83 Table 15: Supplemental table 3 Spearman s rank correlations between UPDRS-III and -IV, NMS, and QoL outcomes NMSS-T NMSQ-T Correlation Sig. (2-tailed) N PDQ-8 SI Correlation Sig. (2-tailed) N UPDRS-III Correlation Sig. (2-tailed) N UPDRS-IV Correlation Sig. (2-tailed) N NMSQ-T PDQ-8 SI UPDRS-III.290* **.371** * * * Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed). Page 82

84 9 Chapter 6: EuroInf: a multicentre comparative observational study of Apomorphine and levodopa infusion in Parkinson s disease This Chapter is based on: MARTINEZ-MARTIN, P., REDDY, P., KATZENSCHLAGER, R., ANTONINI, A., TODOROVA, A., ODIN, P., HENRIKSEN, T., MARTIN, A., CALANDRELLA, D., RIZOS, A., BRYNDUM, N., GLAD, A., DAFSARI, H. S., TIMMERMANN, L., EBERSBACH, G., KRAMBERGER, M. G., SAMUEL, M., WENZEL, K., TOMANTSCHGER, V., STORCH, A., REICHMANN, H., PIRTOSEK, Z., TROST, M., SVENNINGSSON, P., PALHAGEN, S., VOLKMANN, J. & CHAUDHURI, K. R EuroInf: A Multicentre Comparative Observational Study of Apomorphine and Levodopa Infusion in Parkinson's Disease. Movement Disorders Nov 10. doi: /mds (Epub ahead of print) 1.29 Abstract: Background: Subcutaneous Apomorphine infusion (Apo) and intrajejunal levodopa infusion (IJLI) are two treatment options for patients with advanced Parkinson s disease and refractory motor complications, with varying cost of treatment. There are no multicentre studies comparing the effects of the two strategies. Methods: This open label, prospective, observational 6-month multicentre study compared 43 patients on Apo (48.8% males, age 62.3±10.6 years; disease duration 14 ±4.4 years; median Hoehn & Yahr stage 3; interquartile range 3-4) and 44 on IJLI (56.8% males, age 62.7±9.1 years, disease duration 16.1±6.7 years; median Hoehn & Yahr stage 4; interquartile range 3-4). Results: Cohen s effect sizes ( 0.8 considered as large) were large with both therapies with respect to total motor, non-motor, and quality of life scores. The non-motor symptoms scale with Apo showed moderate improvement while sleep/fatigue, gastrointestinal, urinary, and sexual dimensions of the non-motor symptoms scale showed significantly higher improvement with IJLI. 75% on IJLI improved in their quality of life and non-motor symptoms while in the Apo group, Page 83

85 a similar proportion improved in quality of life but 40% in non-motor symptoms. Adverse effects included peritonitis with IJLI and skin nodules on Apo. Conclusions: Based on this open label non-randomised comparative study we report that in advanced Parkinson s patients, both IJLI and Apo infusion therapy appear to provide a robust improvement in motor symptoms, motor complications, quality of life and some non-motor symptoms. Controlled randomised studies are required Introduction: Treatment options for advanced Parkinson s disease (PD) and refractory motor complications are limited and include device-aided, invasive therapies. Broadly three different strategies are available, continuous subcutaneous Apomorphine infusion (Apo) and intrajejunal levodopa gel infusion (IJLI) in addition to Deep brain stimulation of the subthalamic nucleus or internal globus pallidus. In some countries, treatment options are limited by the cost of therapies assessed by Quality Adjusted Life Years as well as Incremental cost-effectiveness ratio and an evidence base to support the use of one therapy over another is essential. (Chinthapalli, 2013, Dodel et al., 2014) There are, for instance, considerable differences in the cost of initiating Apo versus IJLI. Yet in many countries one therapy is favoured over the other based on national reimbursement policies. Level 1 evidence is now available for IJLI (Olanow et al., 2014) but still lacking for Apo pump and the latter is therefore not always included in treatment guidelines despite its widespread use and long clinical experience. Non-motor symptoms (NMS) in PD have remained a key unmet need both in terms of recognition and treatment. (Chaudhuri et al., 2006a, Chaudhuri and Odin, 2010) Here, we report for the first time a prospective observational study Page 84

86 comparing Apo with IJLI using validated motor, non-motor and quality of life outcome measures captured at six months follow-up across 13 European centres Methods: Design and patient selection: This was a prospective, open label, non-randomized multicentre 6-month comparative observational study. Patients were included in the study if they had already been selected to receive treatment with IJLI or Apo by their clinical teams as part of their routine clinical care after discussion regarding suitability for advanced therapies including Deep brain stimulation and informed choice of patients. Suitability was based on published criteria focused on motor fluctuations, disabling off periods and dyskinesias not optimally controlled on oral and transdermal therapies. Patients had to be levodopa responsive, nondemented and were regarded as not optimally managed on oral and transdermal therapies to be suitable for advanced therapies. Centres were chosen on the basis of previous track record of performing infusion and advanced therapies (able to provide Apo and IJLI and Deep brain stimulation under one roof), as well as infrastructure so as to support clinical data collection and experience in using motor and non-motor scales. All patients satisfied the UK PD brain bank criteria for diagnosis on PD (Lees et al., 2009) and consecutive cases were included Assessments: Patients were started on either IJLI or Apo at baseline following above criteria and re-assessed at six months. In addition to demographic data, the following assessments were applied by the same nominated researcher at each site during on state when evaluated in clinic: Page 85

87 Hoehn and Yahr (HY) staging. As recommended by the ad hoc Movement Disorder Society task force(goetz et al., 2004) the original HY(Hoehn and Yahr, 1967) was used to assess the global motor dysfunction of PD. Unified Parkinson s disease Rating Scale (UPDRS) sections 3 and 4 were applied to evaluate motor impairment and complications.(goetz et al., 2008) Non Motor Symptoms Scale (NMSS) was used to assess the burden of non-motor symptoms. NMSS is composed of 30 items grouped in 9 domains. The NMSS total score is obtained from the sum of the 9 domains (theoretical range: 0 360). (Chaudhuri et al., 2007, MartinezMartin et al., 2009) PDQ-8, a self-applied, PD-specific questionnaire for assessment of health-related quality of life (HrQoL). It is composed of 8 items scoring from 0 (never) to 4 (always or cannot do at all). The summary index (SI) is expressed as percentage of the total score from all items on the maximum possible score. (Jenkinson et al., 1997b) Conversion of daily doses of medicines to levodopa equivalent dose (LED) was carried out following Tomlinson et al. (Tomlinson et al., 2010) Follow-up assessments using above instruments were performed as part of clinic visits at six months following Apo or IJLI initiation. All patients were subsequently followed up at 6-monthly intervals to monitor for side effects. Ethical Approval: This study is part of a five year global study (NILS) which is evaluating progression and incidence of motor and NMS in PD and the effect of therapy on the motor and non-motor aspects of PD. The study (a global one) has thus far recruited 1088 patients with baseline data ranging from untreated Page 86

88 PD to advanced, the latter group (from Europe and those suitable for advanced therapies) being screened for the purpose of this report. The study was approved by individual research ethics committees of all participants and is led by a national portfolio adoption of the study in UK (DenDRoN, portfolio number 10084) Data Analysis: After checking of normality for continuous variables (Shapiro-Wilk test), comparisons were carried out intra- and inter-group, applying parametric (Student t-test for paired or unpaired data) or non-parametric tests (WilcoxonMann-Whitney test). Proportions were compared by chi-squared test. For balanced correction of Type I and Type II errors, multiple comparisons were corrected with the Benjamini-Hochberg method (modified p= p (m+1) / 2m where p=0.05 and m=number of tests). (Benjamini and Hochberg, 1995) For evaluating the magnitude of the change, in addition to the difference between baseline and follow-up, the relative change [RC=mean(TestFollow-up TestBaseline) 100/mean TestBaseline)](Deyo and Centor, 1986) and Cohen s effect size [ES=mean(TestFollow-up TestBaseline)/SD TestBaseline] (Cohen, 1977) were calculated. ES values are considered small effect ; moderate effect, and 0.80 large effect. (Kazis et al., 1989) In addition, analysis also addressed the following points: 1) fixing a threshold of beneficial response; 2) determination of the proportion of patients in each group of treatment surpassing this threshold for the main measures; 3) the number needed to treat (NNT) for obtaining one patient improving the threshold or more for each outcome variable of interest; The selected threshold was ½SDBaseline, which is one of the most widely used benchmarks for interpretation of change, related to the minimal important difference and the effect size. (Connor et al., Page 87

89 2003, Norman et al., 2003, Sloan et al., 2005, Wyrwich et al., 2005).Half a standard deviation at baseline (½SD TestBaseline) has also been proposed as a threshold for a clinically significant change in patient-reported outcomes(sloan et al., 2005, Kazis et al., 1989, Norman et al., 2003) and was used for the NMS and HrQoL outcomes.(norman et al., 2003) The number needed to treat (NNT) to improve ½SD TestBaseline was calculated for each treatment group. Adverse effects on IJLI and Apo were either flagged by patient/carer self report or evaluated during six months clinical follow up and are shown in Table 19. Safety of the procedure was judged by a clinician and clinical team led assessments at follow up visit and any discontinuation or hospitalisation due to adverse event was considered a serious adverse event Results: The IJLI group included 44 patients, 56.8% males, aged (mean±sd) 62.7±9.1 years, with 16.1±6.7 years of disease duration and median HY (interquartile range) 4 (3-4). The Apo group was composed of 43 patients, 48.8% males, with age 62.3±10.6 years; duration of disease 14 ±4.5; and HY 3 (3-4). Table 13 shows the scores of the variables used in the study at baseline for patients on IJLI or Apo. Differences between both groups for sex, age, duration of disease, HY-based severity level, UPDRS sections 3 and 4, NMSS, and PDQ-8 total scores were not statistically significant. In the IJLI group, pre IJLI, 9.1% were taking dopamine agonists, 9.1% were on amantadine while the rest were managed on combination of oral levodopa and decarboxylase inhibitor alone or in combination with catechol-o-methyltransferase inhibitor. In the Apo, group, 48.8% were on varying types of dopamine agonists and 30.2% were on levodopa with a catechol-o-methyltransferase inhibitor before initiating Apo infusion. Page 88

90 Mean daily dose of IJLI was ±771.5 mg/day (LED= ±857.2 mg/day) and of Apo 105.9±23.2 mg/day (LED=1059.3±231.5 mg/day) (t-test, p<0.0001). The Apo group needed additional oral levodopa therapy (although with a 30% reduction of the pre-apo dose) and night time Rotigotine patch in 4. Total LED combining oral, transdermal and Apo therapy was 1934 ± 374 mg and was not significantly different from the IJLI group. Continuous infusion was maintained for 17.3±3.6 h/day with IJLI and 15.9±3.5 h/day with Apo, and this difference was not significant. Table 16 shows an improvement compared to baseline in all variables of the study for both treatments, reaching significant levels in motor signs and complications, as well as in NMS as a whole and HrQoL. NMS domains for Cardiovascular, Sleep/Fatigue, Gastrointestinal, Urinary and Miscellaneous showed greater improvement with IJLI, whereas Apo mainly improved Mood/Apathy, Perceptual problems/hallucinations, Attention/Memory, Gastrointestinal, and Urinary domains. The magnitude of change in each group is shown in Table 17. Mean relative change for UPDRS sections 3 and 4 NMSS total score, and PDQ-8 index was 44.0% in the IJLI group and 36.4% in the Apo group (non-significant difference). The NMSS domains Sleep/fatigue, Gastrointestinal, Urinary, and Sexual functioning showed a significantly greater improvement (significantly higher relative change and larger effect size) with IJLI than with Apo (Table 17). Large effect sizes ( 0.80) were obtained with both therapies on the total scores of the four scales, except for NMSS with Apo, which showed moderate improvement (effect size=0.53). The proportions of patients, who improved by ½SDBaseline for IJLI vs. Apo, respectively, were: NMSS total score, 68.2% vs. 39.5% (p=0.0073); and PDQ-8 Page 89

91 summary index, 65.9% vs. 69.8% (non-significant difference). Table 18 shows the number of patients needed to treat for obtaining one patient who improves by ½ SD of the baseline score. In the Apo group no serious side effects were observed although severe somnolence complicated three cases and impulse control disorders occurred in four, none requiring discontinuation of therapy. In the IJLI group, problems such as stoma site irritation and abdominal bloating were relatively common. Tube dislocation required hospital visit for repositioning of tube while one case developed peritonitis and these were considered serious adverse events (Table 19) Discussion: Both Apo and IJLI are recognized treatment options for advanced PD and open label studies have suggested that both IJLI and Apo may have a beneficial effect on several key NMS such as sleep dysfunction, pain, mood, anxiety attacks and non-motor fluctuations along with motor benefits. (Martinez-Martin, 2011, Reddy et al., 2012) These observations ideally need to be investigated in a randomised controlled study, perhaps with a double dummy design. However, logistically this is unlikely to happen and therefore, we attempted a multicentre investigator-led, non-sponsored study incorporating prospective measurements of motor disorder, NMS and quality of life using validated scales. Within the limitations of an observational non-randomised open label study, the key observations from this work suggest: 1. Both Apo and IJLI, in suitably selected patients, lead to beneficial effect (as judged by effect size of intervention) on HrQoL as well as motor dysfunction and motor complications. Page 90

92 2. There appears to be some differences in the effects on NMS although this needs to be investigated further in controlled studies before any formal conclusions can be drawn. IJLI therapy showed a significantly greater beneficial effect on NMSS total score compared to Apo. However, Apo had a significantly better effect on mood and apathy scores in NMSS compared to IJLI. Our study has several weaknesses and it is important to address these first. Firstly, this was not a randomized or a double-dummy placebo-controlled study and is an open label comparative study. Secondly, being an observational study, the calculation of sample size as well as the study design was pragmatic and as such we can only provide results which may be indicative of a trend but not conclusive results. However, a randomised placebo controlled study comparing IJLI and Apo is likely to be very expensive and there is little chance of pharmaceutical funding addressing a real life population of PD comparing Apo and IJLI. Thus, we focused on a real life population using NMS and HrQoL as the main outcome measures where high level data is lacking, The numbers of patients in each arm (44 IJLI vs. 43 Apo) which happened to be matched in baseline parameters, represent, to our knowledge, the largest study of this kind reported and the multicentre nature is likely to reduce the bias introduced by single centre studies. Moreover, this was a real life study including some patients who would be otherwise excluded in clinical trials, e.g. due to cognitive impairment. Hence the observed results could be interpreted as being meaningful to clinical practice with good external validity although, not conclusive. Finally, the patient support provided by individual drug companies marketing IJLI and Apo could be variable and different in several European countries and could also influence some of the outcome variables, particularly Page 91

93 side effects issues and we could not control for this. For instance, the regular post discharge home monitoring provided by IJLI manufacturers could have reduced rates of hospitalization due to tube dislocation and detachment. We ensured that the process of initiation of either Apo or IJLI was as uniform as possible. All patients underwent a standard protocol (as provided by the manufacturers of IJLI and Apo) regarding both Apo and IJLI initiation as well as follow-up plans to maintain conformity. Our results indicate that both treatments have similar efficacy on motor scores, as evidenced by the large effect size seen in UPDRS 3 and 4 domains (Table 17). The key focus of this study was a possible effect on NMS, which remains among the biggest challenges in the management of PD and comprise dopaminergic and non-dopaminergic NMS as well as NMS manifested during motor fluctuations. (Storch et al., 2013) In our study, both Apo and IJLI showed an effect, moderate (0.53) for Apo and large (0.83) for IJLI, on the NMSS total score (Table 17). While sleep dysfunction (onset and maintenance insomnia, symptoms of night-time restless legs) improved significantly (with no worsening of daytime sleepiness) with both IJLI and Apo, the effects were moderate and significantly better with IJLI compared to Apo. The beneficial effects on sleep were seen with an average of hour infusion time and not necessarily 24hour infusion (only four patients had 24 hr Apo). Similarly, a significantly better beneficial effect on some autonomic symptoms such as gastrointestinal and genitourinary function (bladder, dribbling of saliva) was noted with IJLI compared to Apo (Table 2). Conversely, Apo showed a significantly greater effect on mood and apathy symptoms compared with IJLI. The mechanism of the improvements in NMS observed on these pump treatments may be variable Page 92

94 and may be explained by satisfactory treatment of non-motor fluctuations as well as a direct effect on dopaminergic NMS of PD. (Chaudhuri and Schapira, 2009) In terms of infusion time, a small difference (1.4 hr) between IJLI and Apo was noted, and whether this had any bearing on the nocturnal symptoms would need further investigation. In addition, our finding may be noteworthy that even in these advanced PD patients, some with pre-existing hallucinations; no worsening of these neuropsychiatric symptoms was noted with Apo, an observation reported previously. (Ellis et al., 1997) Patient related outcomes are recognized as a key outcome measure of modern clinical trials. In the whole sample, we addressed this issue using the measure of NNT in relation to HrQoL and NMSS measures. The between-group comparisons of the proportion of patients who improved their total NMSS and HrQoL scores ½SDBaseline and the NNT were consistent, as a whole, with relative change and effect size. In summary, around two thirds of patients treated with IJLI consistently improved ( ½SDBaseline) their HrQoL and NMS, whereas in the Apo group a similar proportion improved on HrQoL but only 40% expected improvement in NMS (Table 22). Safety issues in relation to IJLI and Apo merit discussion although it must be emphasized this was not a study to address safety and tolerability. At six months, there were minor problems such as stoma site irritation and abdominal bloating while tube dislocation (usually requiring a short hospital visit for resiting the tube) was common in the IJLI group and is considered as serious adverse event. These figures are similar to the adverse events reported by Olanow et al. (Olanow et al., 2014) In our study one patient on IJLI developed peritonitis (and recovered) in the observed six months period. However, relatively high rates of Page 93

95 gastrointestinal complications including infection have been reported in some studies of IJLI and in studies of PEG safety. (Wollman et al., 1995) In one study, seven patients out of 52 developed gastrointestinal SAE, all manifesting as acute abdomen. (Klostermann et al., 2012) At six months no cases of polyneuropathy were observed although both polyneuropathy and malabsorption have been described with IJLI therapy. (Santos-Garcia et al., 2012, Urban et al., 2010, Muller et al., 2013) As discussed previously, we cannot make any formal recommendations based on this study owing to its pragmatic nature and open label design. However, some clinically relevant points can be stressed. These include the recognition of the need for having a well experienced gastroenterology team, working closely with the IJLI team, to ensure correct placement of the intestinal infusion system and prompt response to complications. Apo appears safer in this regard although skin changes and somnolence may be a concern particularly in advanced PD patients many of whom may have pre-existing daytime somnolence. (Bliwise et al., 2013) Future studies may need to address what could be the best advanced therapy option in sleepy PD patients, as our preliminary observation suggest that IJLI may appear to be a better therapeutic option in this scenario. Finally, impulse control disorders (ICD) and dopamine dysregulation syndromes have emerged as key clinical challenges in the last decade, ICD s being mostly attributed to dopamine agonist use. (Weintraub et al., 2010) Interestingly in our study, IJLI therapy led to resolution of ICD s in four out of 44 patients with preexisting ICD's and no new cases of ICD were observed. This observation is in line with a recent open label study also reporting similar observations. (Cilia et al., 2014) Interestingly, however, only four new cases of ICD, none requiring Page 94

96 discontinuation of therapy, developed in the Apo arm, despite the fact that this is a potent dopamine agonist, possibly suggesting a low rate of development of ICD with infusion therapies. Further prospective studies addressing occurrence of incident ICD cases with Apo versus IJLI are, therefore, warranted. The possible cause of potentially a low rate of ICD with CDS strategy, particularly non-oral therapies is unclear. Evidence suggests that aspects of ICD may be driven by pulsatile dopamine release in the ventral striatum, and one possibility is that tonic dopamine delivery utilising non-oral CDS, may avoid such pulsatile dopamine release in the ventral stratum. However, this hypothesis needs to be tested in clinical studies. To conclude, based on data from this open label comparative study in advanced PD patients, both IJLI and Apo infusion therapy appear to provide a robust improvement in motor symptoms, motor complications and HrQoL. There is an emerging trend that IJLI appears superior in terms of reduction of some dopaminergic NMS of PD, particularly sleep/fatigue, gastrointestinal and urinary autonomic symptoms. Procedural complications were common in the IJLI arm and safety aspects of this treatment require close monitoring. This study provides the platform and need for future controlled studies towards developing an evidence-based rationale for a preferential choice of treatment with IJLI or Apo. Page 95

97 Table 16: Descriptive and comparative statistics at baseline and follow-up for each group of treatment UPDRS- Part 3* UPDRS- Part 4* NMSS Domains Cardiovascular Sleep/Fatigue Mood/Apathy Perceptual/Hallucinations Attention/Memory Gastrointestinal Urinary Sexual functioning Miscellaneous NMSS Total score PDQ-8 * Intrajejunal Levodopa Infusion Baseline Follow-up Mean SD Mean SD p < < Apomorphine Infusion Baseline Mean SD < < < < Page Follow-up Mean SD p < < < Paired t-test; the rest of comparisons, Wilcoxon test. Significant if p< 0.027, after correction for multiple comparisons. UPDRS Unified Parkinson s disease Rating Scale; NMSS Non-Motor Symptoms Scale; PDQParkinson s disease Questionnaire; SD Standard deviation.

98 Table 17: Magnitude of the change from baseline to follow-up for each group of treatment Relative change (%) Effect size IJLI Apo p* IJLI Apo UPDRS- Part UPDRS- Part NMSS Domains Cardiovascular Sleep/Fatigue Mood/Apathy Perceptual/Hallucinations Attention/Memory Gastrointestinal Urinary Sexual functioning Miscellaneous NMSS Total score PDQ-8 Summary index * Chi-squared. Significant if p< 0.027, after correction for multiple comparisons UPDRS Unified Parkinson s disease Rating Scale; NMSS Non Motor Symptoms Scale; PDQ- Parkinson s disease Questionnaire

99 Table 18: Number needed to treat for improving ½ SDBaseline NNT IJLI Apo NMSS Domains Cardiovascular Sleep/Fatigue Mood/Apathy Perceptual/Hallucinations Attention/Memory Gastrointestinal Urinary Sexual functioning Miscellaneous NMSS Total score PDQ-8 Summary index NNT: Number needed to treat NNT= [1 / % of patients who improved 1/2 SDBaseline ] x 100 NMSS Non Motor Symptoms Scale; PDQ- Parkinson s disease Questionnaire

100 Table 19: Reported side effects of IJLI and Apo therapy at six months follow up period Apo group IJLI group Subcutaneous/local site discomfort Minor: Persisting nausea x3 Stoma site irritation x 8 Severe somnolence x 3 Abdominal bloating x 7 ICD = 4 Serious: Hyper sexuality x 2 Tube dislocation x 9 Compulsive shopping x 1 Peritonitis x 1 (not requiring discontinuation of IJLI) Compulsive internet use x 1 ICD = impulse control disorders Serious= Clinically estimated serious side effects

101 10 Chapter 7: Perceptions of symptoms and expectations of advanced therapy for Parkinson s disease: preliminary report of a Patient- Reported Outcome tool for Advanced Parkinson s disease (PRO-APD) This Chapter is based on: REDDY, P., MARTINEZ-MARTIN, P., BROWN, R. G., CHAUDHURI, K. R., LIN, J. P., SELWAY, R., FORGACS, I., ASHKAN, K. & SAMUEL, M Perceptions of symptoms and expectations of advanced therapy for Parkinson's disease: preliminary report of a Patient-Reported Outcome tool for Advanced Parkinson's disease (PRO-APD). Health and Quality of Life Outcomes, 12: Abstract: Background: What do patients expect from a treatment? A patient-centred approach to treatment is becoming necessary given the choices for invasive treatments for Parkinson's disease. Patient's perception of severity and expectations from complex therapies has not been studied. We describe the concept of developing a Patient-Reported Outcome (PRO) tool and report preliminary findings of the perceived severity, expectations from therapy prior to initiation, their inter-correlations, association with clinical factors, and illustrate their potential use in individual patients awaiting therapy. Methods: Patient symptoms were grouped into four domains, with 8 motor, 7 non-motor, 7 psychological and 4 social questions. For each question, symptom severity was rated on a Likert scale scoring from 0 (no problem) to 7 (perceived as a severe problem). Similarly, the expectation for each symptom to change after therapy was rated on a Likert scale: score -3 (expected to be very much worse) to + 3 (expected to be very much improved). Results: 22 consecutive patients, routinely planned to receive one of DBSSTN/IJLI/Apo, were recruited (13 male, mean (+/-sd) age: 65.6 (+/-9.5) years, mean (+/-sd) disease duration: 14.3 (+/-5.7) years. Subjective severity scores are reported as mean (+/-sd) / maximum possible score: (i) motor 23.5 (+/-7.5) /

102 56, (ii) non-motor 15.5 (+/-5.6) / 49, (iii) cognitive - psychological 12.4 (+/-5.8) / 49, (iv) social 9.3 (+/-4.1) / 28. Expectation of change (improvement) scores are reported as mean (+/-sd) / maximum possible score of: (i) motor 14.0 (+/5.6) / 24, (ii) non-motor 8.5 (+/-4.1) / 21, (iii) cognitive - psychological 7.4 (+/4.4)/ 21, and (iv) social 5.5 (+/-2.8) / 12. For each domain, Pearson's correlation coefficient showed significant associations between severity and expectation within-domain. Conclusion: This tool (PRO-APD) provides a description of problem severity and expectation of treatments encompassing a holistic patient-driven view of care. PD patients about to receive complex therapy have moderately high perception of symptom load in multiple domains, and expect substantial improvements in multiple domains. These findings may be useful in documenting multi-domain symptoms, as well as counselling patients to help them reach realistic expectations and reduce potential dissatisfaction following therapy Introduction: Parkinson s disease (PD) is a progressive neurodegenerative disorder affecting a wide range of motor and non-motor functions, and leading to marked disability in its later stages. At a population level, subjective health status (quality of life) is associated with both motor and non-motor symptoms (NMS). (Forsaa et al., 2008, Martinez-Martin et al., 2011, Hinnell et al., 2012) However, for individual patients, the perceived significance of different motor and non-motor symptoms is likely to be influenced by the extent to which they interfere with particular aspects of their life, some of which will be specific to the individual (e.g. in relation to occupation and recreation)

103 Once a treatment is suggested, the question of how do we know if that treatment works? remains a complex one. The ability of treatment to manage individually important symptoms may influence a patient s preferences for and expectations of treatment, for example a watchmaker affected by tremor. Particularly pertinence applies when using complex, invasive, expensive therapies, such as Deep brain stimulation (DBS-STN), Intrajejunal Levodopa Infusion (IJLI) and subcutaneous Apomorphine (Apo) infusions, because these can offer marked benefits for some but not all symptoms of PD, but require high levels of expertise, high costs and demand significant time requirements from health care providers. (Volkmann et al., 2013) Understanding the individual patient s perceptions of the importance of specific symptoms and expectations of treatment is important clinically for a number of reasons. Firstly, where a choice of treatments is available it may be possible to choose or tailor the treatment better to suit the patient s requirements. For example, the optimal dose of dopaminergic stimulation or DBS-STN parameters for motor control may not necessarily be optimal for cognition or other NMS. Secondly, it will help patients and those treating them make better-informed decisions amongst treatment options where those treatments vary in how well they deal with specific problems. Thirdly, identifying possibly unrealistic expectations may be useful in preparing patients for the most likely clinical outcome and minimize adverse emotional reactions following treatment. Ultimately, the patient s perception of the ability of a treatment to improve the personally relevant (patient centred) aspects of their condition is the primary indicator of that treatment s success. Patient-Reported Outcomes (PRO) were defined nearly a decade ago by the US Food and Drug Administration (FDA) as a patient s report of a health

104 condition and its treatment. (Acquadro et al., 2003) PRO are subjective assessments by the patient of any aspect of health, e.g. symptoms, functional status, psychological well-being, quality of life, preferences, perceptions, and satisfaction with care. (Martinez-Martin et al., 2008) Although there are a wide range of PRO s available and used in Parkinson s disease, none to date capture patient perception and expectation from treatment. There is one study to date which explores patient perceptions and expectations to treatment in PD, however this only focuses on the oral therapies and does not focus on the advanced therapies in PD. (Nisenzon et al., 2011) The present study sought to assess patient perceptions of the severity and importance of a range of symptoms and their personal impact, and their expectation of improvement from the proposed therapy for those symptoms. This preliminary pilot work assessed the practicality of application of such a tool. Secondly we wished to investigate correlations between symptom severity and expectation of change before therapy, as this has not previously been explored for these therapies in PD Methods: The study was approved by the Regional ethics committee of South East London: ref number: 10/H0808/46. The funding source was a grant from King s Health Partners, obtained through open competition. A local PRO scale was designed, titled provisionally the Patient Reported Outcomes in Advanced Parkinson s disease (PRO-APD). The PRO scale was designed taking into consideration our experience with motor, non-motor, social and quality of life issues in Parkinson s patients. The questions were designed to encompass the spectrum of problems/issues encountered by Parkinson s patients. We have a large local and regional clinic at our hospital and the idea

105 of the questions was based on our clinical impression, coupled with currently available contemporary views from the literature. Further, a patient group was consulted with regards to the questions and was asked if they could add any other relevant questions. The final version was decided after input was sought from the patient group, a psychologist and clinicians dealing with PD. There were 8 motor, 7 non-motor, 7 cognitive/psychological and 4 social questions, giving a total of 26 questions (Questions are listed at the end of this chapter). Within each domain, the individual symptom severity was first rated by the patient on a Likert scale from 0 ( I do not have the problem ) to 7 ( I have a very severe problem ). Next, expectations of therapy for each item were rated on a second Likert scale. A bipolar scale was used with 7 points from -3 (expected to be very much worse) to +3 (expected to be very much improved), with 0 as expected to show no change post-therapy. The research fellow directly administered the questionnaires on paper to the patients after explaining the purpose of the questionnaire and informed consent was obtained from all the patients who agreed to take part. The patients rated their symptoms and expectations from therapy using the visual likert scale explained above. Additional, participants were assessed on a range of other measures, the new Unified Parkinson s Disease Rating Scale (MDS-UPDRS), (Goetz et al., 2007) Non-motor Symptoms Scale (NMSS), (Chaudhuri et al., 2007) Parkinson s Disease Questionnaire (8-item) (PDQ-8), (Peto et al., 1998) Hospital Anxiety and Depression Scale (HADS) (Zigmond and Snaith, 1983) and Addenbrooke s Cognitive Examination Scale-Revised (ACE-R). (Mioshi et al., 2006) PD patients were included in the study if they had already been selected to receive treatment with one of DBS-STN or IJLI or Apo by their clinical teams as

106 part of their routine care, at hospital-based clinical neurology settings at King s College Hospital NHS Foundation Trust, London UK. Separate informed consent was obtained for entry into the PRO-APD study. The PRO-APD severity and expectation scales were applied following the clinical decision to proceed but before the delivery of that therapy. In our institution, this waiting time can range from 6 weeks to 6 months. Data analysis and statistics For each patient, items within each domain were summed to give a domain score. The maximum possible scores for severity domains were: motor 56, nonmotor 49, psychological, 49, social 28. A total score was also calculated to quantify the patient s overall perceptions of their PD-related problems. Whole group domain and total means were then calculated. For the expectation components, the ranges of possible scores for change were motor -24 to +24, non-motor -21 to +21, cognitive/psychological -21 to +21, social -12 to +12. Negative values indicated that a patient was expecting a worsening of symptoms after therapy. In practice, no patient expected worsening on any item but the scales included the possibility of worsening symptoms after intervention. For each patient, the scores for expectations of improvement from each question in a domain were summed to give domain scores of expectations of improvement. Whole group domain and total means were then calculated. We also calculated an overall mean total expectation score for improvement, combining all domains. Because of the different numbers of items in each domain, standardized scores were also calculated expressed as a percentage of the maximum score possible for each domain, for both severity and expectation

107 Finally, we investigated total and domain-specific associations between severity and expectation. The Shapiro-Wilk W-test showed that data of interest/variables were compatible with a normal distribution. Pearson correlation coefficient was used for determining association and values higher than 0.60 were considered high. The statistical significance was accepted at p< 0.05, without correction for multiple comparisons as this was a pilot exploratory study Results: A total of 22 consecutive patients were recruited for the study (8 patients for DBS-STN, 11 patients selected for Apo, 3 patients for IJLI) of which 13 were male and 9 female. The demographics are shown in Table 21. Table 22 shows the mean scores for each domain for severity and for expectation, along with total scores. Severity scores for the motor domain were rated highest and cognitive/psychological the lowest. The expectations from therapy showed the same pattern across the sample. Within domains, significant associations were found between perceived severity and expectations for therapy in each domain, with higher severity associated with higher expected improvement (Table 22). Although significantly associated in the group as a whole, there was evidence of variability at the individual level between ratings of severity and expectation. The scatterplots for each domain are shown in Figure 5 A-D. Two illustrative patients ( X and Y ) both scheduled for subcutaneous Apomorphine infusion are indicated, on the plots. Despite showing comparable levels of perceived severity in most of the problem domain, patients X and Y show very different expected outcomes within three of the four domains. Schematic views of profile scores across each of the PROAPD items are shown in Figure 6 (severity) and Figure 7 (expectation), to illustrate this further. Two brief cases vignettes are provided in table

108 Across domains, the perceived severities of the motor problems were associated with the severity of non-motor problems (r=0.69, p<0.001) and social problems (r=0.56, p<0.01) but not the cognitive/psychological problems (r=0.19, p>0.05). No significant association was found between any of the other pairs of domains (0.25 < r < 0.39, p>0.05). For the expectation scores, associations were higher and significant for all domain pairs (motor: non-motor, r=0.68, p<0.001; motor: cognitive/psychological, r=0.45, p<0.05; motor: social, r=0.69, p<0.001; non-motor: cognitive/psychological, r=0.48, p<0.05; nonmotor: social, r=0.67, p<0.001; cognitive/psychological: social, r=0.59, p<0.01). The detailed associations between the perceived severity and expectation domain scores and total scores, with a range of clinical measures (UPDRS-III total score, NMSS totals score, ACE-R total score and HADS total score) are given in Table 23. Clinician rated motor severity was moderately associated with patient perceptions of motor problems but not with any other domain. Total NMSS score was associated with all domains, most strongly with the perceived severity of cognitive and psychological problems. The level of depression and anxiety as measured by the HADS was strongly associated with perceived severity of cognitive and psychological problems, with cognition function as measured objectively by the ACE-R being unrelated to any of the perceived severity scores. For expectation scores, UPDRS-III and NMSS were associated with expected motor change. NMSS, HADS and ACE-R were all associated with greater expectation in the cognitive and psychological domain, with worse NMSS, greater depression/anxiety and better cognition associated with greater expected improvement

109 1.38 Discussion: The aim of this pilot study was to design a multidimensional PRO measure assessment tool for PD patients scheduled for an invasive therapy and then to assess its practicality in application and potential clinical utility at both group and individual level. The measure is novel in combining subjective perceptions of symptom difficulty with expectations of treatment outcome. Despite the small numbers of patients assessed we have shown that such a scale can be pragmatic to use, even in a complex multimodal disease, such as PD, and that it is able to provide clinically meaningful information to supplement routine clinical assessment. This approach allowed us to determine for each individual a map of the patient s perceptions of their disease and their expectation from treatment, and offer the potential to help manage patients in preparing them for therapy. Traditionally, the measurement of improvement of Parkinsonism by therapy has been by measuring motor signs and broad measures of outcome, such as health-related quality of life. (Williams et al., 2010),(Deuschl et al., 2006) The multiplicity of symptoms in PD means that motor scores are not sufficient to describe overall quality of life. (Hinnell et al., 2012, Chaudhuri et al., 2007, Schrag et al., 2000) The impact of therapies on the other aspects of PD is generally less well understood and has been explored less than motor state. (Schrag et al., 2000, Reich et al., 2011, Fasano et al., 2012) This is in line with data from others and our group that suggests that standard PD severity scales are an insufficient measure of disease burden. Indeed, the gold standard Unified Parkinson s Disease Rating Scale (UPDRS) has been modified to the new UPDRS (MDS-UPDRS) for this reason. (Goetz et al., 2007). Despite an improvement as measured by standard scales, patients may describe being

110 unhappy after DBS-STN. (Reich et al., 2011, Agid et al., 2006). From our department, we have noted an adult patient whose main desire for treatment of her PD was the ability to sit still and sew. (Hulse N, 2008). Her degree of satisfaction following DBS-STN ( very pleased ) was not matched by the minor improvement (< 10%) in the UPDRS score. It is debatable which outcome measure was the more clinically relevant. Further, the assessment of success from a therapy can become more complex if one were to extend analyses to include a more complete assessment of stage of life, which would include social dynamics including relationships/marriage. (Brown, 2002) Measuring treatment effects on one symptom (e.g. tremor), or set of symptoms (e.g. the ability to move), is insufficient to determine if a therapy is multi-dimensionally beneficial for a patient, cohort or population, unless it is known that it is that symptom (or set of symptoms) which is personally significant to the patient. For example, we have observed that despite successful DBS-STN as determined by the patient, neurologist, neuropsychologist, neuropsychiatrist and family practitioner, a carer remained unhappy because the partner had changed, no longer needing such intensive support from the carer, leading to a mutually acceptable marriage breakdown. So was the treatment successful? Should it be offered to similar patients? The PRO concept informing the present study is a move away from fixing the symptom / sign and towards selecting treatments that are best able to address a patient s priorities across the range of PD-related problems and their impact. The concept is powerful and yet pragmatic, and is in line with trends that encourage patient inclusion in treatment decisions. This is a simple means of assessing the results of severity and expectation for individuals and its meaning to the patient. (Martinez-Martin et al., 2008) Utilization of any PRO does not

111 move away from the need to use standard severity scales, nor generic or specific quality of life scales. However, they can help clinicians and researchers, jointly with the patient, to understand better the impact of symptoms and proposed treatments (rather than the physiological effect of treatment on the sign). PRO-APD draws on three decades of experience in the assessment of other chronic multifactorial disorders (e.g. dementia (Selai, 2001), schizophrenia (McCabe et al., 2007) and neurorehabilitation (Edwards et al., 2002)). Most importantly, there is now guidance from both the FDA and European Medicines Agency for the use of PRO measures in drug development (Health et al., 2006) and so one expects to see their increasing use over the next generation of treatments. As a first step, we report here the preliminary use in assessing severity and expectation in a common disease, even if it is as complex as PD. Further study is required to extend this to other disorders, and to outcomes after therapy in order to assess if expected outcome is matched to measured outcome, and if not what the impact of the discrepancy might be. With further extrapolation, the measure should be equally applicable to other adult or pediatric chronic movement disorders, like essential tremor and dystonia with adjustment where necessary of the problems assessed. Unless a PRO is intended as a surrogate for traditional conventional measured (e.g. self-report version of a clinical scale for use in surveys) strong association are not necessarily expected or even necessary. In the present case, patients were asked to assess the severity of the PD-related problems and expectations of change. Given the difference in the various scales, item content and methods of assessment, we would expect only moderate associations between self-rated problem domain severity and standard scales. This was observed for motor

112 problems (with MDS-UPDRS part 3), NMS problems (with NMSS) and psychological and cognitive problems with (HADS but not ACE-R). This latter result suggests that the Cognitive/Psychological domain may be more sensitive to measuring mood problems in the present sample. Subjective ratings of cognition have also typically been found to be unrelated to objective measurements across the age span. (Ponds RWHM, 2000) At a group level, there was evidence for a robust association between the perceived severity of the different problems domains and the expectations for outcome, with all patients expecting improvement or, at worse, no change. We found that the associations between severity and expectations were strongest and most reliable for the motor and non-motor symptom domains, and less robust for the cognitive and psychological problems and weakest for the social and ADL domain (See Figure 5 A-D). This may reflect the patient s model of their disease and the degree to which they linked different problems to the underlying PD, and therefore amenability to treatment. For example, one can speculate that a patient who attributes memory problems to his/her age may have different expectations from a treatment targeting PD compared with a patient who sees the memory problems as an integral feature of the disease. Helping patients to understand and categorise their symptoms to this level of clarity may be difficult for some patients and could be time consuming on health care staff, yet it is likely to be an important part of helping them assess the possible impact of any future treatment. Low mood might conceivably have had an impact on expectations, with depressed and anxious patients expecting a less favorable outcome. In practice, depression and anxiety as measured by the HADS was associated

113 positively with degree of expected change, although the effect was significant only for the cognitive and psychological problems. Clinical outcomes from advanced treatments such as DBS-STN, Apo and IJLI can produce dramatic improvements in some patients and for some symptoms, both motor and non motor. However, not all patients show a marked improvement and not all symptoms show similar responsiveness. An assumption that the most severe (and personally significant) problems are going to improved, and by the largest extent, raises the risk that patients may be dissatisfied with the outcome. (Agid et al., 2006, Schupbach et al., 2006) Identifying such misperceptions in advance offers the opportunity to provide the patient with the facts to inform and possibly adjust their expectations of treatment ahead of time. Although lack of accurate information may contribute to unrealistic expectations, individual characteristics (such as optimism) may also play an important role. Optimism is recognized as a dispositional trait that is largely independent of factors such as mood. (Hurt et al., 2013) Evidence from patients with other physical conditions suggest that low levels of optimism is associated with poor outcome, perhaps because patients fail to engage with treatment or to be proactive in coping with the challenges of their condition. However, high levels of optimism may also be unhelpful if they set up expectation that cannot be met or encourage a passive everything will be alright attitude. Practically, moderate (rather than excessively high or low) levels of optimism may be the best for outcome in chronic disease such as multiple sclerosis and Parkinson s disease (De Ridder et al., 1998) and something that clinicians can encourage when discussing potential treatments. The two cases described in this study (table 20) were chosen to illustrate somewhat polarized expectations with case X having extreme levels of positive

114 expectation for the large majority of problems experienced (which may not have been achievable), while case Y had lower expectations for some and no positive expectations for other problems. Both patients might benefit from a detailed discussion about what the planned treatment can and cannot reasonably offer. Patient Y was more depressed and had a significant degree of cognitive impairment that may have biased their judgment. However, a more pessimistic and limited set of expectations is not necessarily bad and may even be closer to the likely clinical reality. A danger of low expectation is when they cause a patient to decline available treatment options. We note several important limitations to our study. Our study is a small pilot and further research is required to demonstrate the scale validity and utility. We appreciate the current lack of accepted consistency in meaning of the term holistic, and of uniform guidelines on the specific selection of one treatment modality over another, although some guidance exists, e.g. significant cognitive impairment may exclude DBS-STN, but not always exclude IJLI or Apo.(Volkmann et al., 2013) None-the-less, this is a pragmatic first step which exemplifies the benefits of a PRO approach, and which if developed further could help guide individual patients to be offered and accept individual therapies. In conclusion, a PRO approach encompasses a patient driven view of care. We present a pilot study using PRO-APD as a demonstration to provide a simple and pragmatic assessment of the severity and expectation of treatment, and which is in line with increased patient participation in their management. Patients opting for invasive therapy for PD have moderately high multi-domain symptom load, and also expect substantial improvements in multiple domains. These need to be considered when assessing patients for therapy, so that

115 individual expectations can be realistic, since unexplored expectations are more likely to lead to overall dissatisfaction following therapy. Our findings are preliminary. Further validation could lead to the use of PRO-APD as an adjunct to existing scales in clinical management

116 Table 20: Illustrative case vignettes X and Y Patient X: 55 year old male, with a 9 year history of PD, receiving treatment with Ropinirole 24mg, and levodopa/carbidopa (100mg/25mg) 6 per day. His off-state MDS-UPDRS-3 score was 83. There was no evidence of cognitive impairment (ACE-R=100) and only moderate levels of depression/ anxiety (HADS total 12). His perceived severity ratings indicated high levels of motor symptomatology. He reported a range of marked non-motor problems including sleep disturbance, bladder problems, pain and fatigue. Cognitive and mood complaints were mild, but significant impulsive behaviour reported. Apart from self-care, aspects of occupation and social function were subjectively markedly affected. The patient expected majority of the problems reported across all domains rated to be very much improved by advanced treatment. Patient Y: 56 year old male with a 13 year history of PD, receiving treatment with Rotigotine 16mg, levodopa/carbidopa (100mg/25mg) 18 per day, and levodopa/carbidopa controlled release (200mg/50mg) at night. His off-state MDS-UPDRS-3 score was 60. There was evidence of significant cognitive impairment (ACE-R=73) and significant depression/anxiety (HADS total=18). He reported significant motor symptoms included problems with swallowing and balance, plus fatigue, cognitive difficulties, and depression. With the exception of impact on socializing, however, his PD was not felt to have significant impact on other activities. Despite the range and severity of problems reported, he had only moderate expectations for the impact on treatment on the motor and nonmotor symptoms, with no expectation of change for the cognitive or mood problems

117 Table 21: Patient demographics, motor and non-motor severity, Quality of Life, Mood and Cognition Variable Mean (SD) Range Age (years) 65.6 (9.5) Duration of disease (years) 14.3 (5.7) 5 23 MDS-UPDRS (15.8) MDS-UPDRS (6.4) 0 20 NMSS total 74.9 (28.9) PDQ-8 total 37.3 (16.3) HADS total 16.3 (6.6) 5 28 ACE-R total 90.1 (7.6) Abbreviations: (MDS-UPDRS= Revised Unified Parkinson s Disease Rating Scale, NMSS= Non-motor Symptoms Scale, PDQ-8= Parkinson s disease quality of life Scale, HADS= Hospital Anxiety and Depression Scale, ACE-R= Addenbrooke s Cognitive Examination Scale-Revised)

118 Table 22: Domain and total PRO-APD scores of problem severity and expectation of change Variable Mean (SD) Range Mean (% Maximum) Correlation of (SD) severity with P expectation Motor severity 23.5 (7.5) % (13.9) (max 56) Motor expectation 0.79 < < < <0.001 of 14.0 (5.6) % (22.9) 15.5 (5.6) % (11.2) improvement (max 24) NMS severity (max 49) NMS expectation of improvement 8.5 (4.1) % (19.1) 12.4 (5.8) % (11.1) (max 21) Cognitive / psychological severity (max49). Cognitive / psychological expectation of 7.4 (4.4) % (20.8) 9.3 (4.1) % (14.6) improvement (max 21) Social severity (max 28) Social expectation of 5.5 (2.8) % (23.1) 60.5 (16.7) % (9.4) improvement (max 12) Total severity (max 182) Total expectation improvement of 35.5 (14.1) % (17.6) (max 78) NMS = non-motor symptoms, SD = standard deviation

119 Table 23: Correlations (Pearson r) between clinical variables and PRO-APD severity and expectation domain scores Domain MDS- NMSS Total HADS Total ACE-R Total UPDRS-3 Correlation Correlation Correlation Total (p) (p) (p) Correlation (p) Severity of problem Motor 0.48* 0.56** NMS * ** 0.67** * Cognitive Psychological Social Expectation of change (improvement) Motor 0.46* 0.52* NMS ** 0.61** 0.44* Cognitive Psychological Social NMS = non-motor symptoms; MDS-UPDRS-3 = Revised Unified Parkinson s Disease Rating Scale (Part 3); NMSS = Non Motor Symptom Scale; HADS = Hospital Anxiety and Depression Scale; ACE-R = Addenbrooke s Cognitive Examination (Revised) * p<0.05, ** p<

120 Figure 5: Scatterplots showing relation between perceived severity and expected improvement for each of the 4 domains on the PRO-APD. Solid line represents the regression line and the dotted line the 95% confidence interval. Arrows indicate case examples X and Y. For the motor, non-motor and cognitive / psychological domains, the severities in patients X and Y are comparable but expectations differ

121 Figure 6: Profiles of perceived severity for illustrative cases X and Y (gaps indicate zero ratings), showing broadly similar severities across individual symptom profiles using the PRO-APD

122 Figure 7: Profiles of expectations of improvement ratings for illustrative cases X and Y (gaps indicate zero ratings) showing clearly different profiles of expectation following treatment using the PRO-APD