The legally binding text is the original French document TRANSPARENCY COMMITTEE OPINION. 1 February 2012

Transcription

1 The legally binding text is the original French document TRANSPARENCY COMMITTEE OPINION 1 February 2012 Examination of the dossier for proprietary medicinal products included for a 5-year period starting on 11 July 2007 (Official Gazette of 11 July 2007) GARDASIL suspension for injection, Human Papillomavirus vaccine [types 6, 11, 16, 18] (recombinant, adsorbed) vial (glass) 0.5 ml B/1 (CIP code: ) GARDASIL suspension for injection in a pre-filled syringe, Human Papillomavirus vaccine [types 6, 11, 16, 18] (recombinant, adsorbed) pre-filled syringe (glass) 0.5 ml Box of 1 prefilled syringe + 2 needles B/1 (CIP code: ) Applicant: SANOFI PASTEUR MSD Vaccine against papillomavirus ATC code: J07BM01 List I Date of Marketing Authorisation: 29 September 2006 (Centralised procedure) Amendment 18 August 2010 and Amendment 1 August 2011 Reasons for request: - Renewal of inclusion on the list of proprietary medicinal products refundable by National Health Insurance. - Further to the submission of new data and new vaccination recommendations from the Haut Conseil de la Santé Publique [High Council for Public Health (HCSP)], reassessment of the improvement in actual benefit, in compliance with Article R of the French Social Security Code. - Inclusion on the list of medicines refundable by National Health Insurance and on the list of medicines approved for hospital use and various public services in the extension of indication: premalignant genital lesions of the vagina. Medical, Economic and Public Health Assessment Division 1/25

2 1 CHARACTERISTICS OF THE MEDICINAL PRODUCT 1.1. Active ingredient 1 dose (0.5 ml) contains about: Human Papillomavirus Type 6 L1 protein, 20 micrograms Human Papillomavirus Type 11 L1 protein, 40 micrograms Human Papillomavirus Type 16 L1 protein, 40 micrograms Human Papillomavirus Type 18 L1 protein, 20 micrograms L1 protein in the form of virus-like particles produced in yeast cells (Saccharomyces cerevisiae CANADE 3C-5 (Strain 1895)) by recombinant DNA technology, adsorbed on amorphous aluminium hydroxyphosphate sulphate (225 micrograms) as adjuvant Indications 1 Gardasil is a vaccine for use from the age of 9 years for the prevention of: - premalignant genital lesions (cervical, vulvar and vaginal) and cervical cancer causally related to certain oncogenic Human Papillomavirus (HPV) types. - genital warts (condyloma acuminata) causally related to specific HPV types - the use of Gardasil should be in accordance with official recommendations Dosage The primary vaccination series consists of 3 separate 0.5 ml doses administered according to the following schedule: 0, 2, 6 months. If an alternate vaccination schedule is necessary, the second dose should be administered at least one month after the first dose and the third dose should be administered at least 3 months after the second dose. All three doses should be given within a 1-year period. The need for a booster dose has not been established. Paediatric population: Gardasil has not been used in children under the age of 9 years. The vaccine should be administered by intramuscular injection. The preferred site is the deltoid area of the upper arm or in the higher anterolateral area of the thigh. It is recommended that individuals who receive a first dose of Gardasil complete the 3-dose vaccination course with Gardasil 1 Note that section 5.1 of the GARDASIL SPC of 1 August 2011 refers to indications that the HCSP does not currently recognise in women aged 24 to 45 or in men aged 16 to 26 (genital warts). 2/25

3 2 SIMILAR MEDICINAL PRODUCTS 2.1. ATC Classification (2011) J: Antiinfectives for systemic use J07: Vaccines J07B: Viral vaccines JO7BM: Papillomavirus vaccines J07BM01: Papillomavirus (type 6, 11, 16, 18) recombinant 2.2. Medicines in the same therapeutic category Not-strictly-comparator medicines: CERVARIX human papillomavirus types 16 and 18 recombinant, adsorbed In contrast to GARDASIL, CERVARIX only includes two genotypes in its composition and is not indicated in the prevention of premalignant vulvar and vaginal lesions or the prevention of external genital warts Medicines with a similar therapeutic aim Not applicable 3/25

4 3 REMINDER OF PREVIOUS TRANSPARENCY COMMITTEE OPINIONS Improvement in actual benefit (opinion of 18 April 2007) GARDASIL is a primary means of prevention designed in the short and medium terms to prevent morbidity associated with high-grade cervical and vulvar dysplasia. It will also prevent external genital warts. Its preventive effect on the occurrence of cervical cancer, which has not yet been demonstrated and must be limited to approximately 70% of cases (since human papillomavirus types 16 and 18 are only involved in 70% of cervical cancer cases), will only be shown over the long term, since it is estimated that some years elapse between infection with human papillomavirus and the appearance of invasive cancer. A vaccination campaign must be accompanied by an organized screening programme for precancerous and cancerous cervical lesions (secondary prevention) by means of cervical smears. Bearing in mind on the one hand: - that GARDASIL is the first vaccine against human papillomavirus 6, 11, 16 and 18 - the efficacy of this vaccine in preventing precancerous cervical lesions (high-grade cervical dysplasia (CIN 2/3), high-grade vulvar dysplasia (VIN 2/3) and external genital warts due to Human Papillomavirus types 6, 11, 16 and 18 and on the other hand: - the absence of data on the duration of the protection provided by the vaccine beyond 5 years - the safety profile to be confirmed under actual conditions of use - the potentially damaging effects that may be induced by vaccination: a reduction in condom use, since the vaccine may be perceived as protection against all sexually transmitted diseases reduced take-up of screening for cervical cancer, with the risk of seeing a rise in the number of cases of cervical cancer due to oncogenic HPV types other than 16 and 18 a shift in the incidence of cervical cancer to an older age group if a booster were to prove necessary and if some women failed to have it the selection of other oncogenic HPV types. The Committee considers that the GARDASIL vaccine provides a moderate (level III) improvement in actual benefit within the strategy of preventing precancerous and cancerous cervical lesions, high-grade vulvar dysplasia and external genital warts in the populations recommended by the Technical Committee on Vaccinations and the French High Council for Public Health (opinion of 9 March 2007). In any case, the Committee believes that a nationwide programme of screening for precancerous and cancerous cervical lesions by means of cervical smears (secondary prevention) ought to be organised and implemented before a vaccination campaign is introduced for oncogenic human papillomavirus. It also points out that the explanations to be given to girls and young women by the doctor prior to vaccination should be compiled in writing in a document validated by the competent authorities. 4/25

5 4 ANALYSIS OF AVAILABLE DATA 4.1. Preventive efficacy The new available data relate to efficacy data after a median 3.6-year follow-up of the four randomised, controlled, double-blind clinical studies versus placebo in young women aged from 16 to 26 years:. 2 phase II studies (Study 005 Study 007). 2 phase III studies (Study 013 Study 015) - Study 005 (phase II) assessing the reduction in the risk of onset of persistent infections caused by papillomavirus type 16 (N = 2409). - Study 007 (phase II): assessing the reduction in the risk of onset of persistent infections, cervical dysplasias of all grades (CIN 1 to CIN 3), AIS (cervical adenocarcinomas in situ) and/or cervical cancers, external genital lesions: external genital warts, vulvar dysplasia, vaginal dysplasia and/or vaginal or vulvar cancers caused by papillomavirus of type 6, 11, 16 and/or 18 (N = 1158). - Study 013 (phase III): assessing the reduction in the risk of onset of cervical dysplasias of all grades (CIN 1 to CIN 3, AIS or cervical cancer, dysplasias of the vulva and vagina VIN 1, VIN 2, VIN 3, VaIN 1, VaIN 2, VaIN 3, vulvar cancer, vaginal cancer or genital warts caused by papillomavirus infection of type 6, 11, 16 and/or 18 (N = 5455 women). - Study 015 (phase III): assessing the reduction in the risk of onset of high-grade cervical dysplasias (CIN 2 and CIN 3) or AIS or cervical cancer caused by papillomavirus infection of type 16 or 18 (N = 12,167 women). The primary vaccine efficacy endpoints were the reduction with respect to placebo in the onset of: - persistent infections at 12 months due to HPV (Human Papillomavirus) of type 16, i.e. presence of the same HPV type over two cervical samples at least 12 months apart (Study 005), - persistent infections due to HPV of types 6, 11, 16 or 18 (Study 007), - cervical dysplasias (or cervical intraepithelial neoplasias) CIN 1/2/3, vulvar and vaginal dysplasias VIN 1/2/3 and VaIN 1/2/3 and genital warts due to HPV of types 6, 11, 16 or 18 (Studies 007 and 013), - high-grade cervical dysplasias or cervical intraepithelial neoplasias (CIN 2/3) and histologically confirmed AIS due to HPV types 16 or 18 (Study 015). The efficacy of GARDASIL was assessed on the basis of the combined data from four studies with a 2-year follow-up (interim analysis) and at the end of the study after a median follow-up of 3.6 years. The efficacy of the vaccine was assessed in various populations: - per protocol population, - intention to treat populations: women not infected with one or more types of papillomavirus targeted by the vaccine and having received at least one injection (MITT2) women whether or not infected on inclusion with one or more types of papillomavirus from the vaccine and having received at least one injection (MITT3). This population is close to the general population of women aged from 15 to 25 years women not infected with the 14 types of HPV tested and having received at least one injection (RMITT2). This population is close to the population to be vaccinated. 5/25

6 The efficacy results are expressed in terms of the reduction in the relative risk of onset of premalignant genital lesions and genital warts. High-grade cervical dysplasias (or cervical intra-epithelial neoplasias): histologically proven CIN 2/3 and AIS were used as substitute markers for cervical cancer. The expression premalignant genital lesions of the cervix, vulva and vagina corresponds to high-grade cervical intra-epithelial neoplasias (CIN 2/3), high-grade vulvar intra-epithelial neoplasias (VIN 2/3) and to high-grade vaginal intra-epithelial neoplasias (VaIN 2/3). From this point on, only data from studies carried out in the populations recommended by the HCSP i.e. young girls aged 14 years and young women aged from 15 to 23 years will be examined. Preventive efficacy with respect to HPV 6/11/16/18 The results for preventive efficacy in the final analysis are given along with a reminder of the results at 2 years from the interim analysis, when data from the interim analysis were available. 1 Per protocol population This population included women not infected on inclusion by one or more types of papillomavirus targeted by the vaccine and for the 7 months following the first injection (anti- HPV serology negative on D1, PCR negative) and having received the three injections. Table 1: Efficacy against premalignant cervical lesions (CIN 2/3 or AIS CIN 3 or AIS) associated with papillomavirus of type 16 and/or 18 (per protocol population) at end of study GARDASIL Placebo GARDASIL Placebo Efficacy endpoints CIN 2/3 or AIS associated with HPV of types 16/18 CIN 3 associated with HPV of types 16/18 AIS associated with HPV of types 16/18 Number of cases Number of subjects* Number of cases Number of subjects* % Efficacy at 2 years [95% CI] [ ] 100 [ ] 100 [ ] Number of cases Number of subjects* 2** ** Number of cases Number of subjects* % Efficacy at 3.6 years [95% CI] 98.2 [ ] 96.9 [ ] 100 [ ] *Number of subjects with at least one follow-up visit after the 7th month **On the basis of the virology results, one case of CIN 3 in a patient chronically infected with HPV 52 seems to be due to HPV 52. HPV 16 was found in only 1 sample out of 11 (in month 32.5) and was not detected in the specimen excised by loop diathermy. In the second case of CIN 3 observed on a patient infected with HPV 51 on Day 1 (in 2 samples out of 9), HPV 16 was detected in a biopsy in month 51 (in 1 sample out of 9) and HPV 56 was detected (in 3 samples out of 9) in month 52 (in 1 sample out of 9) in the specimen excised by loop diathermy. Note: The values and Confidence Intervals are adjusted per person-years at risk. At study end, the efficacy in preventing premalignant cervical lesions (CIN 2/3 or AIS) associated with papillomavirus of type 16 and/or 18 was 98.2% (95% CI [ ]) versus 100% (95% CI [ ]) at 2 years of follow-up (see Table 1). At study end, the efficacy in preventing premalignant vulvar lesions (VIN 2/3) associated with HPV of types 6, 11, 16 and 18 was 100% (95% CI [ ]) versus 100% (95% CI [ ]) at 2 years of follow-up. 6/25

7 At study end, the efficacy in preventing premalignant vaginal lesions (VaIN 2/3) associated with HPV of types 6, 11, 16 and 18 was 100% (95% CI [ ]). This indication was not taken into account in the initial analysis. At study end, the efficacy in preventing genital warts associated with HPV of types 6, 11, 16 and 18 was 99.0% (95% CI [ ]) versus 98.9% (95% CI [ ]) at 2 years of follow-up. In addition, in a sub-group of Study 013 including 3578 women aged from 16 to 24 years, the efficacy at study end in preventing persistent infections at 6 months caused by HPV of type 16 was 98.7% (95% CI [ ]) and caused by HPV of type 18 was 100.0% (95% CI [ ]) and at 12 months for infections caused by HPV of type 16 was 100.0% (95% CI [ ]) and caused by HPV of type 18, was 100.0% (95% CI [ ]). The efficacy of GARDASIL in preventing CIN of grades 1, 2, 3 or AIS associated with HPV of types 6, 11, 16 and 18 was 96.0% (95% CI [ ]). The study-end analysis of the results confirmed the efficacy of GARDASIL in preventing diseases caused by each of the four types of HPV targeted by the vaccine in the per protocol population. 2 MITT2 population This population included women not infected (anti-hpv serology negative on D1, PCR negative) with one or more types of papillomavirus targeted by the vaccine and having received at least one injection. Study-end efficacy in preventing: - premalignant cervical lesions (CIN 2/3 or AIS) associated with papillomavirus of type 16 and/or 18 was 97.4% (95% CI [ ]) versus 98.8% (95% CI [ ]) at 2 years of follow-up. - premalignant vulvar lesions (VIN 2/3) associated with HPV of types 6, 11, 16 and 18 was 96% (95% CI [ ]) versus 100% (95% CI [ ]) at 2 years of follow-up. - premalignant vaginal lesions (VaIN 2/3) associated with HPV of types 6, 11, 16 and 18 was 100% (95% CI [ ]). - genital warts were 96% (95% CI [ ]) versus 93.4% (95% CI [87 97]) at 2 years of follow-up. 3 MITT3 population This population, close to the general population of women aged from 15 to 25 years, included women infected or not infected on inclusion with one or more types of papillomavirus targeted by the vaccine (positive or negative serology/positive or negative PCR) and having received at least one injection. 7/25

8 Table 2: Efficacy against premalignant cervical lesions (CIN 2/3 or AIS CIN 3 or AIS) associated with papillomavirus of type 16 and/or 18 at study end Efficacy endpoints CIN 2/3 or AIS associated with HPV of types 16/18 CIN 3 associated with HPV of types 16/18 AIS associated with HPV of types 16/18 Gardasil Placebo Gardasil Placebo Number of cases Number of subjects* Number of cases Number of subjects * % Efficacy** at 2 years [95% CI] 39.0 [ ] 34.3 [ ] 54.3 [< ] *Number of subjects with at least one follow-up visit after 30 days after Day 1. **The percentage efficacy is calculated using the combined protocols. The values and Confidence Intervals are adjusted per person-year at risk. Number of cases Number of subjects * Number of cases Number of subjects * % Efficacy at studyend** [95% CI] 51.8 [ ] 46 [ ] 60 [< ] In the population of studies 005, 007, 013 and 015, the study-end efficacy of GARDASIL in the prevention of: - premalignant cervical lesions (CIN 2/3 or AIS) associated with papillomavirus of type 16 and/or 18 was 51.8% (95% CI [ ]) versus 39% (95% CI [ ]) at 2 years of follow-up. - premalignant vulvar lesions (VIN 2/3) associated with HPV of type 6, 11, 16 and 18 was 73.3% (95% CI [ ]) versus 68.1% (95% CI [ ]) at 2 years of follow-up. - premalignant vaginal lesions (VaIN 2/3) associated with HPV of type 6, 11, 16 and 18 was 85.7% (95% CI [ ]). - genital warts associated with HPV of types 6, 11, 16 and 18 was 80.3% (95% CI [ ]) versus 68.5% (95% CI [ ]) at 2 years of follow-up. Cross-protective efficacy against non-vaccine oncogenic HPV types The efficacy of GARDASIL in preventing CIN 2/3 or AIS due to 10 types of HPV not targeted by the vaccine (HPV 31, 33, 35, 39, 45, 51, 52, 56, 58, 59) and structurally related to HPV types 16 and 18, were assessed on the basis of the combined data from 2 phase III studies (N = 17,599) after a median follow-up of 3.6 years (end of study). The studies did not have the necessary power to assess separately the efficacy against diseases caused by a given type of HPV. The main analysis was performed in the MITT2 population in women who were negative for the type analysed, but who could be positive for other types of HPV (96% of the total population). The EMEA stressed that persistent infection can be used as an intermediate cervical cancer marker for the non-vaccine types, provided that high efficacy figures are obtained that are statistically significant and consistent with those obtained with clinical cervical lesions, and in addition that HPV types 16 and 18 are the most oncogenic. 8/25

9 Table 3: Vaccine efficacy against CIN 2/3 or AIS and persistent infections at 6 and 12 months associated with HPV 31 in the MITT2 population at end of study % Efficacy against CIN 2/3 or AIS HPV serotype [95% CI] HPV [ ] % Efficacy against persistent infections at 12 months [95% CI] 36 [ ] % Efficacy against persistent infections at 6 months [95% CI] 49.0 [ ] Efficacy in preventing CIN 2/3 or AIS premalignant lesions was established only for HPV type 31 among the 10 types of non-vaccine HPV tested. Preventive efficacy was not demonstrated for the other types of non-vaccine HPV (in particular HPV 33 and HPV 45). According to the various epidemiological data, HPV types 31, 33 and 45 could be involved in 10 to 14% of cervical cancers. Overall preventive efficacy The impact of GARDASIL on the overall risk of a cervical disease occurring (whatever the type of HPV in the lesion) was assessed on the basis of 30 days after the first dose during the course of two phase III clinical studies (studies 013 and 015) including 17,599 women. The RMITT2 population included women not infected by the 14 types of HPV tested (negative smears, anti-hpv serology negative for the vaccine-related types on D1, PCR negative for the 14 types of HPV) and having received at least one injection (one dose of the vaccine). This population had not been analysed at the time of the intermediate analysis at 2 years, as only data on vaccine-related HPV types were available. Data on infection with the 10 non-vaccine types were analysed in particular to assess cross-protective efficacy. This population is the closest to the population of young girls to be vaccinated in accordance with HCSP guidelines. In the RMITT2 population, administration of the vaccine reduced the incidence of CIN 2/3 and AIS caused by types of HPV whether targeted or not by the vaccine by 42.7% (95% CI [ ]) and genital warts by 82.8% (95% CI [ ]) at study end. In the MITT3 population (women whether infected or not on inclusion by one or more types of papillomavirus included in the vaccine) administration of the vaccine reduced the incidence of CIN 2/3 and AIS caused by any type of HPV whether or not targeted by the vaccine by 18.4% (95% CI [ ]) and that of external genital warts by 62.5% (95% CI [ ]). In this population, the overall incidence of CIN 2/3 and AIS caused by any type of HPV and that of external genital warts was smaller than that observed in the population of HPV-naïve women (RMITT2). The vaccine has no effect on the course of infections or diseases which are ongoing at the time of vaccination. Further data have been provided by the company. They are not taken into account by the Committee. This relates to a published study 2 as a follow-up at 8.5 years (extension of Study 005) on vaccine efficacy against premalignant lesions associated exclusively with HPV 16 and two unpublished studies: - one Scandinavian long-term follow-up study (extension of Study 015) in 2700 women vaccinated on the basis of cancer registers (interim results) - one long-term follow-up study (Study 018) in adolescents aged from 9 to 15 years, assessing tolerability and immunogenicity (study submitted to EMEA). 2 Rowhani-Rahbar A, Mao C, Hughes JP et al., Longer term efficacy of a prophylactic monovalent human papillomavirus type 16 vaccine. Vaccine 2009; 27 (41): /25

10 4.2. Impact on conisations The impact of the vaccine on conisation rates, regardless of the types of HPV involved, was assessed in studies 007, 013 and 015, including a total of 18,150 subjects. In the RMITT2 population of women naïve to 14 types of current HPV and who had a negative cervical smear on Day 1, GARDASIL reduced the proportion of women who had undergone conisation by 41.9% with respect to the placebo group (95% CI [ ]) at study end. In the MITT3 population of women whether infected or not on inclusion by one or more types of papillomavirus included in the vaccine, this reduction was 23.9% (95% CI [ ]) Impact of GARDASIL vaccination on genital warts and premalignant CIN 2+ lesions in Australia Observational studies were included with the dossier assessing: - the impact of vaccination in reducing the incidence of cases of genital warts - the impact of vaccination in reducing premalignant cervical lesions CIN 2/3 in young girls aged 18 years, since the introduction of the vaccination programme. These data on the impact of HPV vaccination under real conditions of use of the quadrivalent vaccine were published based on the experience with HPV vaccination in Australia. In Australia, the quadrivalent HPV vaccine was provided free of charge in April 2007 for girls aged from 12 to 13 years as part of a school vaccination programme and a vaccination catchup campaign aimed at girls aged from 13 to 18 years. Starting in July 2007, the quadrivalent vaccine was made available to young women up to the age of 26 years through general practitioners and vaccination centres. Impact data on genital warts In a retrospective study comparing the proportion of new patients consulting for genital warts at eight healthcare establishments between January 2004 and December 2009, the proportion of new patients with genital warts decreased, having been 11.7% in July-December 2007 and 4.8% in July-December 2009 in women under 27 years of age in 2007 and resident in Australia (59%, 95% CI [54 61], p < 0.05). 3 A more recent study conducted in a health centre reinforced this observation. 4 Impact data on premalignant lesions The aim of this study was to determine the impact of a quadrivalent HPV vaccination programme on the onset of premalignant cervical lesions in the State of Victoria by analysing the trends in cervical anomalies in women, before and after the introduction of the vaccination programme. The before period was between 1 January 2003 and 31 March 2007 and the after period was between 1 April 2007 and 31 December The main result was a decrease in the incidence of high-grade cervical lesions in young girls under the age of 18 years, falling from 0.80% before to 0.42% after. In women aged from 18 to 20 years, the decrease between the two periods was not significant. 5 It is not possible from these observational studies to attribute these decreases in incidence with certainty to the vaccination programme. 3 Donovan et al. Quadrivalent human papillomavirus vaccination and trends in genital warts in Australia: analysis of national sentinel surveillance data. Lancet Infect. Dis 2011; 11: Read et al. The near disappearance of genital warts in young women 4 years after commencing a national human papillomavirus (HPV) vaccination programme. Sex Transm Infect Dec; 87 (7): Epub 2011 Oct 4 5 Brotherton et al. Early effect of the HPV vaccination programme on cervical abnormalities in Victoria, Australia: an ecological study Lancet 2011; 377: /25

11 4.4. Immunogenicity Summary: The immunogenicity of GARDASIL has been assessed in: women aged 18 to 26 years (GARDASIL n = 4666; placebo n = 4249), girls (GARDASIL n = 1471; placebo n = 583) and boys (GARDASIL n = 1071; placebo n = 275) aged 9 to 15 years. The minimum level of protective antibodies has not been defined for HPV vaccines. The immunogenicity and safety were demonstrated in boys aged 9 to 15 years. The vaccine efficacy in girls aged 9 to 15 years was extrapolated from these immunogenicity data. New data: - Persistence of the immune response to GARDASIL in clinical studies: At the end of the phase III clinical studies in women aged 16 to 26 years in the per protocol population, 90%, 95%, 98% and 60% of the subjects having received GARDASIL were seropositive to anti-hpv 6, anti-hpv 11, anti-hpv 16 and anti-hpv Studies of co-administration with other vaccines: Co-administration studies have shown the administration of GARDASIL did not significantly interfere with the antibody response to the components of each vaccine: combined booster vaccine against diphtheria (d) and tetanus (T) as well as whooping cough [acellular] (ca) and/or poliomyelitis [inactivated] (P). A tendency towards smaller geometric mean anti-hpv antibody titres was observed in the group receiving a concomitant administration. The clinical significance of this observation is unknown Adverse effects The most commonly observed adverse effects were reactions at the injection site (77.1% of the subjects vaccinated in the 5 days following each vaccine dose) and headache (16.6% of the subjects vaccinated). These adverse effects were generally of mild to moderate intensity. The adverse events reported during the clinical studies and/or after placement on the market are listed in Table 4 below. 11/25

12 Table 4: Adverse events reported after vaccination with Gardasil during the clinical studies and/or after placement on the market Organ system class Frequency Adverse effects Adverse events Disorders of the blood and lymphatic system Indeterminate Idiopathic thrombocytopenic purpura*, adenopathy* Immune system disorders Indeterminate Hypersensitivity reactions including anaphylactic/anaphylactoid reactions Nervous system disorders Very common Indeterminate Headache Dizzy feeling 1 *, Guillan-Barré syndrome*, syncope occasionally accompanied by tonic-clonic movements* Gastrointestinal disorders Common Nausea Indeterminate Vomiting* Musculoskeletal and systemic Common Pain in the extremities disorders Indeterminate Arthralgia*, myalgia* General disorders and administration site conditions Very common At injection site: Erythema, pain, swelling Common Fever At injection site: ecchymosis, pruritus Indeterminate Asthenia*, shivering*, fatigue*, feeling faint* *Adverse events observed after placement on the market (their frequency cannot be estimated from the data available). 1 During the course of the clinical trials, the dizzy feeling was observed as a common adverse event in women. In men, the dizzy feeling was not observed more commonly in vaccinated subjects than in subjects receiving placebo. Syncope was occasionally accompanied by a fall and/or tonic-clonic movements, which necessitates close monitoring of vaccinated persons for about 15 minutes following the administration of the vaccine. A third report on the European and French national risk management plan was circulated by AFSSAPS on 11 July Taking all the available data into account, it finds that the risk/benefit ratio for this vaccine remains favourable and close to the safety of use profile as it was defined at the time it received its marketing authorisation. A pharmacovigilance follow-up report on HPV vaccines of 22 November 2011 (see Appendix 2) circulated by AFSSAPS states that: The analysis of the pharmacovigilance data reported in mid-september 2011 by the CRPV [Regional Pharmacovigilance Centre] in Bordeaux, as well as the analysis of the pregnancy cases reported on at the end of May 2010 by the CRPV in Lyons, did not find any new safety issues flagged against these vaccines, following the administration of over 4 million doses of anti-hpv vaccines. The possibility of an increased risk of onset of various auto-immune diseases following anti-hpv vaccination was examined on the basis of data from two epidemiological studies conducted in France, one a cohort study with vaccinated/non-vaccinated subjects performed by AFSSAPS on the basis of the SNIIRAM database (a French national health insurance information system), and the other a case-control study carried out by the LA-SER Group and conducted on the basis of data from the network of PGRx pharmacoepidemiological studies into auto-immune diseases /25

13 The preliminary results from these studies are in agreement and do not show any significant link between this vaccination and the risk of autoimmune diseases. These two studies were judged by the members of the CNPV (French National Pharmacovigilance Committee) to be powerful enough to validate their relevance. The National Pharmacovigilance Committee has however proposed maintaining an increased level of pharmacovigilance targeted on autoimmune conditions and continuing with pharmacoepidemiological studies Conclusion regarding the new available data The vaccine efficacy of GARDASIL in preventing premalignant CIN 2/3 and AIS lesions associated with HPV 16 and/or 18 assessed versus placebo at the time of the interim analysis at 2 years was confirmed and upheld in the final analysis after a median follow-up period of 3.6 years in all the analysed populations, including: - in the per protocol population: 98.2% (95% CI [ ]) - in the population of non-infected women: 97.4% (95% CI [ ]) - in the total vaccinated population, whether infected or not: 51.8% (95% CI [ ]). In the population of women not infected by the 14 types of HPV tested for on inclusion and having received at least one injection (population close to the target population for vaccination), the overall impact of GARDASIL reduced the incidence of CIN 2/3 and AIS caused by types of HPV whether or not targeted by the vaccine by 42.7% (95% CI [ ]). In the population of women whether infected or not on inclusion by one or more types of papillomavirus included in the vaccine, administration of the vaccine reduced the incidence of CIN 2/3 and AIS caused by any type of HPV whether or not targeted by the vaccine by 18.4% (95% CI [ ]). In light of these data, the Committee emphasises the low relative risk of onset: - of CIN 2/3 with HPV 16/18 observed in the total vaccinated population, whether infected or not: 51.8% (95% CI [ ]) compared to that found in the per protocol population: 98.2% (95% CI [ ]) - of CIN 2/3, regardless of the HPV type observed in the lesion compared to that found in the CIN 2/3 with HPV 16/18; indeed, in the total vaccinated population, whether infected or not, it was 18.4% (95% CI [ ]) versus 51.8% (95% CI [ ]). The vaccine efficacy of GARDASIL in preventing premalignant vulvar lesions associated with HPV 6, 11, 16 and 18 at 2 years was confirmed at study end in the various populations analysed. The vaccine efficacy of GARDASIL in preventing premalignant vaginal lesions associated with HPV 6, 11, 16 and 18 was demonstrated at study-end: - in the per protocol population: 100% (95% CI [ ]) - in the population of non-infected women: 100% (95% CI [ ]) - in the total vaccinated population, whether infected or not: 85.7% (95% CI [ ]). The efficacy in preventing external genital warts associated with HPV 6, 11, 16 and 18 assessed at 2 years was confirmed at study end in the various populations analysed. It was 82.8% (95% CI [ ]) at study end in the population close to the target population to be vaccinated. Efficacy in preventing CIN 2/3 or AIS premalignant lesions was established only for HPV type 31 among the 10 non-vaccine HPV types tested. Preventive efficacy was not demonstrated for the other types of non-vaccine HPV tested. 13/25

14 In the population of women who were not infected with any of the 14 types of HPV tested, by the end of the study GARDASIL had reduced the proportion of women undergoing conisation by 41.9% (95% CI [ ]). In the population of women whether infected or not on inclusion with one or more of the papillomavirus types included in the vaccine, by study end GARDASIL had reduced the proportion of women undergoing conisation by 23.9% (95% CI [ ]). No clinical study has compared GARDASIL and CERVARIX in terms of vaccine efficacy in preventing premalignant cervical lesions. In women aged from 16 to 26 years, the longest follow-up of immune response was that in Study 007, in which a peak geometric mean titre (GMT) of anti-hpv 6, 11, 16 and 18 antibodies was observed in the 7th month. The GMTs decreased up to the 24th month, then stabilised until at least the 60th month (5 years). At the end of the phase III clinical studies in women aged 16 to 26 years in the per protocol population, 90%, 95%, 98% and 60% of the subjects having received GARDASIL were seropositive to anti-hpv 6, anti-hpv 11, anti-hpv 16 and anti-hpv 18. The safety data obtained in the clinical studies did not show any significant increase in adverse effects in the group of women vaccinated with GARDASIL or in the placebo group. A third report on the European and French national risk management plan was circulated by AFSSAPS on 11 July The report concludes that the risk/benefit ratio of this vaccine is favourable and that its safety of use profile is close to that identified at the time it received its Marketing Authorisation. This is confirmed by the absence of any particular pharmacovigilance issue against it (see report by CPRV in Bordeaux from mid-september 2011). As the dossier stands at present, the following data have yet to be established: - the efficacy in terms of preventing cervical cancer, even though additional data have been presented specifically regarding premalignant CIN 3 and AIS lesions, immediate precursors of cervical cancer, - how long the cross-protective efficacy lasts beyond 3.6 years is not known, - the immunogenicity in immunosuppressed populations (study in progress), - the assessment of a possible modification in the viral ecology associated with the introduction of the vaccination Guidelines by the HCSP On 21 October 2011, 8 the Haut Conseil de la Santé Publique [High Council for Public Health] recommended: Continuing to improve screening for cervical cancer and rapidly expanding the system of mass screening in accordance with HAS guidelines. In accordance with the HCSP opinion of 17 December 2010, continuing with the papillomavirus vaccination of 14-year-old young girls and catch-up vaccination of young women up to the age of 23 years who are not sexually active or have been sexually active for less than a year. This vaccination may be carried out with either of the two existing vaccines HCSP opinion on GARDASIL vaccine and the global cervical cancer prevention strategy (21 October 2011) <}0{> 14/25

15 The HCSP is of the opinion that: - there are no data currently available that could call the efficacy of these vaccines into question; - a follow-up of the side effects notified following HPV vaccination has not raised any pharmacovigilance issues; - there are currently no data available to suggest that the vaccine could possess a potentially deleterious nature in certain population categories, especially as regards the onset of cancer in previously infected women. It points out that: - This vaccination cannot be expected to have a significant impact on the incidence of cervical cancer in the long run unless there is sufficient vaccine coverage; - The likelihood is that the young girls who come forward for vaccination are the ones that will later come forward for screening. - Vaccination would have all the greater impact if it reached the women who do not come forward for screening. It therefore recommends: - Firstly, an improvement in vaccination coverage through better access to vaccination and by optimising the way it is organised, as is done in some other European countries, where they have obtained coverage of 80% or more. - Secondly, all available means should be deployed to reach the populations among which screening is likely to be least often carried out even though vaccination could never replace screening. On 17 December the HCSP, after a reminder of the guidelines already formulated by the Technical Committee on Vaccination and the CSHPF (French national high committee on public hygiene), their section on transmissible diseases in their opinion of 9 March 2007 recalled: - the need to organise nationwide screening for premalignant and malignant cervical lesions, as vaccination against HPV 16 and 18 could never be a substitute for it; - the need to oblige firms producing or required to produce an HPV vaccine at the same time, in their publicity material, to promote the use of this vaccine and screening for cervical lesions and to mention the lack of efficacy in preventing all cervical cancers. and concludes: - that both vaccines have shown their protective effect against CIN 2 and other lesions associated with genotypes 16 and 18; - that the quadrivalent vaccine has additionally shown efficacy in the prevention of lesions caused by HPV of genotypes 6 and 11 (including genital warts and CIN) and premalignant vulvar and vaginal lesions of grade 2 and more (VIN 2 or more and VaIN 2 or more); - that the available data are in favour of the bivalent vaccine having a greater capacity to provide for cross-protection against certain oncogenic types of HPV other than HPV 16 and 18; - that the clinical safety data relating to the ASO4 adjuvant in the bivalent vaccine are satisfactory. It considers that in the current state of knowledge, and as part of the vaccination strategy against human papillomavirus infection in young girls aged from 14 to 23 years, there is no longer any reason for preferentially recommending one of these two vaccines. It recalls that the two vaccines available against HPV infection are not interchangeable, and that any vaccination course started with one of them should be carried through to completion with the same vaccine. 9 HCSP opinion on the vaccination of young girls and women aged from 14 to 23 years against human papillomavirus infections (17 December 2010) 15/25

16 On 5 May 2008, 10 the HCSP expressed the opinion that immunosuppressed persons (those with HIV infections, transplant patients or those constitutionally immunosuppressed) form a group at risk of developing HPV-related cancer. There are no immunogenicity, safety or protection data relating to vaccination against HPV in patients undergoing immunosuppressant treatment. The HCSP recommends that vaccination against HPV should be offered to young girls who are due to receive a transplant before the age of 14 years while still remaining within the age bracket set for these vaccines in the Marketing Authorisation. 5 PRESCRIPTION INFORMATION According to IMS data (moving annual total August 2011), 720,000 prescriptions were issued for the proprietary medicinal product GARDASIL, most of them (86%) by general practitioners. 10 HCSP opinion on the age of vaccination against human papillomavirus (HPV) infections in young girls due to undergo transplantation (5 May 2008) 16/25

17 6 TRANSPARENCY COMMITTEE CONCLUSIONS 6.1. Actual benefit GARDASIL is a vaccine against human papillomavirus 6, 11, 16 and 18 for: - the prevention of premalignant genital lesions (cervical, vulvar and vaginal) and cervical cancer causally related to certain potentially life-threatening types of oncogenic Human Papillomavirus (HPV). - the prevention of genital warts (condyloma acuminate), which are recurring non-life threatening non-malignant tumours. This proprietary medicinal product falls into the category of a preventive therapy (primary prevention). The efficacy/adverse effects ratio for this medicinal product is high. Public health benefit: The incidence of invasive cervical cancer in France is estimated at 2810 new cases per year (InVS projections ). It is therefore the 10th most common cancer in women. The number of deaths from cancer is estimated at 998 in 2011, which puts cervical cancer in 13th place as the cause of cancer deaths among women in The epidemiology of vulvar and vaginal premalignant lesions is poorly documented in France. Their progression towards a carcinoma is rare. Vulvar and vaginal cancers remain rare and account for respectively 3% and fewer than 2% of gynaecological cancers. The prevalence and incidence of genital warts are difficult to assess. According to estimates, in France the annual incidence of genital warts is around 107 new cases per 1000 inhabitants. 12 These lesions, whilst not life-threatening for the patients, have a considerable impact on their quality of life. The public health burden that cervical cancer, the premalignant lesions of the vulva and vagina and the genital warts represent is therefore substantial. Reducing the incidence of cervical cancer is an established public health priority (objective 48 of the Law of 9 August 2004 relating to public health policy to continue reducing the incidence of cervical cancer by 2.5% a year, and in particular by hitting the target of 80% screening coverage among women aged from 25 to 69 years, Cancer Plan Measure 15: To improve the structuring of the action plan for national mass cancer screening programmes ). Vaccination against oncogenic human papillomavirus (HPV) could constitute a response to this need, in addition to optimising nationwide cervical cancer screening. Even though improvements in the levels of screening have been taking place since 1995, it had only reached 58.5% in , 13 which is still a long way short of the 80% target, especially in certain social and occupational categories Lyon civil hospices / Health Monitoring Institute / National Cancer Institute / Francim / National Institute of Health and Medical Research. Projections of the incidence and mortality from cancer in France in Technical report. June [Retrieved ]. 12 E. Lukasiewicz et al., Incidence et prise en charge des condylomes acuminés externes en médecine générale, Ann Dermatol Venereol 2002; 129: DRESS. The state of health of the French population Monitoring the objectives appended to the Public Health Law 2011 Report (in publication). 14 Public health objectives. Assessment of the objectives of the Law of 9 August Proposals. HCSP. April /25

18 In France, vaccination coverage (full vaccination schedule) was estimated from the national health insurance system s reimbursement data on 31 December On average it was 33.3% among young girls born in 1993 and 23.7% for those born in The majority of young girls start vaccination at age 15 or later. The coverage rate for the catch-up population (aged years) on 31 December 2009 is at its highest for the three doses at 35.6% in the cohort of young girls born in 1992 (aged 15 in 2007) and falls steadily with age (14.6% for 18-year-old girls in 2007 and 1.8% for women aged 23 years in 2007). Given the poor vaccination coverage rate at the launch of the campaign, in particular among young girls at 14 years of age, and the inadequate individual screening rate in the absence of a national system of mass screening, the public health need is still present. In light of the results of studies showing the vaccine efficacy of GARDASIL in highgrade cervical dysplasias, high-grade vulvar and vaginal dysplasias and external genital warts associated with papillomavirus types 6, 11, 16 or 18 and 31, GARDASIL is expected to have a considerable impact on the reduction of short-term morbidity (CIN 2+), especially among young girls not infected with HPV. However, given the development of the natural history of HPV infections, 3.6 years is insufficient time to be able to assess the impact of GARDASIL in terms of morbidity and mortality (CIN 3 and carcinoma in situ) owing to the limited number of events observed in the trials. Moreover, the updated safety data are reassuring and, after five years of follow-up, do not suggest that there was any increased risk of autoimmune diseases within the context of long-term conditions associated with anti-hpv vaccination. 16 The transferability of the results into practice is no longer guaranteed in the long term, given that: - cross-protection was not demonstrated with every one of the oncogenic HPV types; between 10% (simple cases of infection) 17 and 20% 18 of cervical cancers could be associated with other oncogenic types of HPV, which is why the practice of screening should be maintained, even in previously vaccinated young girls. - the duration of vaccine protection is as yet unknown, even though we have more retrospective data. The need for a booster dose has not been established. - the consequences of vaccination on the viral distribution and ecology of HPV are not yet known (risk of emergence phenomena). - the consequences of vaccination for screening practice will not be known until the first cohorts of vaccinated young girls reach screening age, especially bearing in mind that vaccination coverage rates in the age brackets concerned are still low. If fewer vaccinated women came forward for screening, the risk of an increased incidence of and mortality from these cancers could not be ruled out. Arrangements need to be made, therefore, for regular monitoring of screening practices among vaccinated and unvaccinated subjects. 15 Fagot J-P, et al. HPV vaccination in France: Uptake, costs and issues for the National Health Insurance. Vaccine 2011; 29 (19): National follow-up of the adverse effects of the human papillomavirus vaccine Gardasil. National Pharmacovigilance Committee of 22 November Available at f.pdf consulted on 19/01/ Prétet J-L et al. Human papillomavirus (HPV) genotype distribution in invasive cervical cancers in France: EDITH study. Int. J. Cancer 2008; 122, Sanjose S, on behalf of the Retrospective International Survey and HPV Time Trends Study Group. Human papillomavirus genotype attribution in invasive cervical cancer: a retrospective cross-sectional worldwide study. Lancet Oncol. 2010; 11 (11): Epub 2010 Oct /25

19 In addition, the company has submitted models to estimate the impact at the population level of GARDASIL vaccination accompanied by individual screening on the morbidity and mortality associated with premalignant and malignant cervical lesions caused by oncogenic HPV types. However, the hypotheses tested are somewhat unrealistic, especially those used for vaccination coverage in France (> 70% in the projections), and the parameters the company adopted for adapting the American dynamic model 19 to the French situation are insufficiently detailed (e.g. whether to take male vaccination into account not recommended in France). Consequently, the hypotheses used in these models make it impossible to estimate the impact of the vaccination programme that has been in place in France since The impact in public health terms of vaccination against papillomavirus is dependent on reaching a sufficient level of vaccination coverage in the recommended populations which, according to the available data, does not seem to be the case in France at the present time. Indeed, according to the model produced by Cresge/InVS 20 in 2007, the epidemiological impact of individual screening associated with vaccination of 14-year-old girls with a 30% vaccination coverage rate would help reduce by 8% the incidence of diagnosed cases of cervical cancer and by 6% the associated mortality as compared to individual screening by itself. However, these figures do not take the catch-up programme into account. According to that same model, organising cervical smear screening in France, whose introduction nationwide prior to bringing in vaccination against oncogenic human papillomavirus had been recommended by the Transparency Committee in its opinion of 18 April 2007, would by itself help reduce the incidence of diagnosed cases of cervical cancer by 16.1% and the associated mortality by 19.5%, compared to individual screening by itself. According to experts at Cresge/InVS, for vaccination coverage rates lower than 60%, vaccination has a lesser impact on the number of deaths than organising screening based on the Alsace model. Consequently, with the inadequate vaccination coverage as it is at present in France, and despite the expected considerable potential public health interest in vaccination against papillomavirus, the public health benefit for GARDASIL at the present time, with the scant retrospective information at our disposal, is regarded as slight. In order to optimise the public health benefit of this vaccine and to meet the public health need, it would seem necessary for measures to be rapidly put in place for: - stimulating the vaccination programme and increasing the vaccination coverage rate especially among HPV-naïve young girls, among whom vaccination efficacy is highest, - and also for improving access to, information about and interest in cervical smear screening, particularly among young women in poorer neighbourhoods. Based on updated data, the question of the efficacy of the various possible prevention strategies (mass screening, individual screening plus vaccination, etc.), deserves to be reassessed. There is an alternative to vaccination for the prevention of premalignant cervical lesions and cervical cancer. Screening, which relies on a cytological test cervical smear is an effective secondary means of preventing cervical cancer. There is no way to screen for vulvar or vaginal lesions. 19 Elbasha EH, Dasbach EJ. Impact of vaccinating boys and men against HPV in the United States. Vaccine (42): Dervaux B, Lenne X, Lévy-Bruhl D, Kudjawu Y. Modélisation médico-économique de l impact de l organisation du dépistage du cancer du col utérin et de l introduction de la vaccination contre les HPV dans le calendrier vaccinal - March Saint-Maurice (France): Institut de veille sanitaire, November 2008, 25 pp. Visit: 19/25