For peer review only. BMJ Open. For peer review only -

Transcription

1 Ten-year follow-up of human papillomavirus vaccine efficacy against the most stringent cervical neoplasia endpoint registry-based follow-up of randomized controlled trial Journal: BMJ Open Manuscript ID bmjopen-0-0 Article Type: Research Date Submitted by the Author: 0-Jan-0 Complete List of Authors: Lehtinen, Matti; Tampereen Yliopisto, School of Health Sciences; Karolinska Institutet, Department of Laboratory Medicine Lagheden, Camilla; Karolinska Institutet, Department of Laboratory Medicine Luostarinen, Tapio; Karolinska Institutet, Department of Laboratory Medicine Eriksson, Tiina; Tampereen Yliopisto, School of Health Sciences Apter, Dan; VL Medi Harjula, Katja; Tampereen Yliopisto, School of Health Sciences Kuortti, Marjo; Tampereen Yliopisto, School of Health Sciences Natunen, Kari; Tampereen Yliopisto, School of Health Sciences Palmroth, Johanna; Tampereen Yliopisto, School of Health Sciences Petäjä, Tiina; Tampereen Yliopisto, School of Health Sciences Pukkala, Eero; Cancer Society of Finland Siitari-Mattila, Mari; Tampereen Yliopisto, School of Health Sciences Struyf, Frank ; GlaxoSmithKline Biologicals SA Site de Wavre Nieminen, Pekka; Helsingin Yliopisto Paavonen, Jorma; University Central Hospital, Department of Obstetrics and Gynecology Dubin, Gary; Pennsylvania University Dillner, Joakim; Karolinska Institutet, Department of Medical Epidemiology and Biostatistics BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from <b>primary Subject Heading</b>: Oncology Secondary Subject Heading: Infectious diseases Keywords: HPV vaccines, long-term follow-up, vaccine efficacy, CIN+, cervical cancer, randomized controlled trial on January 0 by guest. Protected by copyright.

2 Page of BMJ Open Ten-year follow-up of human papillomavirus vaccine efficacy against the most stringent cervical neoplasia end-point registry-based follow-up of randomized controlled trial Matti Lehtinen,, Camilla Lagheden, Tapio Luostarinen, Tiina Eriksson, Dan Apter, Katja Harjula, Marjo Kuortti, Kari Natunen, Johanna Palmroth, Tiina Petäjä, Eero Pukkala, Mari Siitari-Mattila, Frank Struyf, Pekka Nieminen, Jorma Paavonen, Gary Dubin, Joakim Dillner University of Tampere, Tampere, Finland; Karolinska Institute. Stockholm, Sweden; Family Federation Finland, Helsinki, Finland; Finnish Cancer Registry, Helsinki, Finland; GSK Biologicals, Wavre, Belgium; University of Helsinki, Helsinki, Finland; Pennsylvania University, PA, USA Address for correspondence: Matti Lehtinen, Department of Laboratory Medicine, Karolinska Institute, Stockholm, Sweden, matti.lehtinen@uta.fi Funding GlaxoSmithKline Biologicals SA (Belgium), Academy of Finland, Finnish Cancer Organizations and the Swedish Cancer Society BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

3 Page of Objective Due to long lag time between infection/cancer diagnoses human papillomavirus (HPV) vaccination programs will deliver vaccine efficacy (VE) estimates against cancer end-points late. Cancer registry follow-up of population-based, randomized trial cohorts of vaccinated and unvaccinated women speeds up the VE estimation. Methods We report interim results with person years of Finnish Cancer Registry -based follow-up of individually and/or cluster randomized cohorts of HPV-/ vaccinated and unvaccinated adolescent women enrolled in June 00/00, and between May 00 and April 00, respectively. The cohorts comprised - to -year-old unvaccinated women (NCT0), and and - to -year-old HPV-/ vaccinated women participating the PATRICIA (NCT00) and HPV-0 (NCT00) trials, respectively. The agealigned passive follow-up started months after the clinical trials end. Results During the follow-up of. to 0 years post enrolment we identified cases of cervical intraepithelial neoplasia grade (CIN) and cases of invasive cervical cancer (ICC) in the unvaccinated cohort, and CIN cases in the HPV-/ vaccinated women. Diagnostic blocks were available for HPV typing from % of the cases. CIN+ lesions were detectable in cases. HPV was found in of 0 unvaccinated CIN+ cases, and in CIN+ cases in the HPV-/ vaccinated women. The latter were all baseline positive for cervical HPV DNA. Baseline data was not available for the unvaccinated women. Intention-to-treat VE against any CIN+ was % (% CI, ). Conclusions Ten years post vaccination the AS0-adjuvanted HPV-/ vaccine shows continued efficacy against CIN+ irrespectively of HPV type. Vaccine efficacy was not observed in baseline HPV DNA positive subjects. All the trial cohorts were commenced before July 00 BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

4 Page of BMJ Open Strengths and limitations of this study *Country-wide cancer registry follow-up of sizeable randomised cohorts of HPV vaccinated and unvaccinated women for person years (up to0 years post vaccination) provides most reliable vaccine efficacy estimates against cancerous end- points. *Retrieval of most diagnostic histopathological blocks and state-of-science identification of the causal HPV type in the lesion enable identification of HPV type-specific vaccine efficacy estimates. *The per protocol defined interim analysis has limited statistical power. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

5 Page of Introduction High-risk (hr) human papillomaviruses (HPVs) cause approximately 0 and percent of cancers in females and males., Bivalent, quadrivalent and nonavalent vaccines against HPV types /, ///, and ////////, respectively, have an acceptable safety profile and are highly efficacious against a number of infections with hrhpvs and associated precancers. - Proof of vaccine efficacy (VE) against HPV associated cancers is, however, not easy to reach. The long lag time between exposure to hepatitis B-virus (HBV) and diagnosis of HBV-associated hepatocellular carcinoma, which delayed the determination of vaccine efficacy against cancer for to years,, is now delaying this in the HPV vaccine-cervical cancer context. Proving the concept of vaccine induced protection against (one of) the major HPV-associated cancers, i.e., in situ and invasive cervical carcinoma is not only of conceptual but also of practical importance. It would guarantee the impact of primary cancer prevention via prophylactic HPV vaccination, guide targeting this prophylaxis, and provide the scientific basis for understanding how and when such proof will be available for other HPV-associated cancers such as non-cervical anogenital cancers, and head and neck cancers. Our objective is to determine VE against cervical cancer. To accomplish this we identify invasive cervical cancer (ICC) and intraepithelial neoplasia grade (CIN+) incidence in passive, population-based cancer registry follow-up of a randomized clinical HPV vaccine trial cohort (PATRICIA), and a cluster-randomized control cohort enrolled in The follow-up of originally adolescent females has previously proven to be feasible. With the observed CIN+ incidence of./ in the control cohort, which equals that of the Finnish female population of similar age./00 000, it is well powered to verify % VE against CIN+ and ICC, 0 and years post vaccination, respectively. - We report interim results on the efficacy of the bivalent HPV/ vaccine against overall and HPV type-specific CIN+ end-points. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

6 Page of BMJ Open Methods Study design and Ethics Our cluster-randomized follow-up involves separate birth cohorts of clinical trial participants,, aged - years and non-participants,0 aged - years assigned according to the start of FUTURE and PATRICIA trials in August 00 and May 00 by ML (the principal investigator for phase III HPV vaccination trials in Finland). The former were also individually randomized., The trials, establishing the unvaccinated control cohort and their Finnish Cancer Registry (FCR) based follow-up were approved by the Finnish National Ethical Review Board (TUKIJA: /00 and /00), respectively. Patient involvement Patients (with cervical neoplasia / condyloma) were not involved in the design of this study in 000. A study on the feasibility of population-based enrolment was done in. Enrolment Starting in May 00 originally all 0 (Q/-Q/ born) - year old Finnish females resident in trial communities were sent invitations to participate to the PATRICIA trial (HPV-00, NCT00) on the immunogenicity, safety and efficacy of the AS0-adjuvanted HPV-/ vaccine against HPV/ positive CIN+ (fig. ). By June 00 a total of 0 women participated. They were randomly assigned to HPV-/ or hepatitis A-virus (HAV) vaccination in : ratio to receive three doses of the HPV-/ vaccine or the HAV vaccine at months 0, and, followed by active follow-up visits with month interval up to years., Following the end of the active follow-up in May 00, approximately 0% of the HAVvaccine recipients chose HPV-/ cross-vaccination during HrHPV DNA positives at the last PATRICIA trial visit ( from the HPV-/ and HAV cohorts) continued active clinical follow-up (HPV-0 protocol) for approximately. years after the end of the PATRICIA trial. They had annual HPV DNA testing and cytology until HPV DNA negative or exit colposcopy after years positive.. In addition, - year olds received three doses of the HPV-/ vaccine by May 00 in a concomitant HPV-0 trial which ended years later. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

7 Page of The unvaccinated control cohort of - year olds was recruited by inviting 0 and in June 00/00 based on the Finnish Population Register Centre (PRC) as described (NCT0).,0 All the FUTURE and PATRICIA, trial participants were excluded from the control cohort. There were no health care interventions targeted to the control cohort. Follow-up Both the vaccinated and unvaccinated control cohorts responded at the age of - years to a questionnaire on life habits with special emphasis on sexual health, i.e., at the beginning of the passive registry-based follow-up., Opportunistic vaccination by Gardasil or Cervarix vaccines after their licensures in 00 and 00 was considered based on questionnaires in 00 and 00, and among the vaccinated PATRICIA and HPV-0 trial cohorts in 00. In addition, vital status and emigration were updated until the end of 0 by a PRC linkage in 0. Invitations to cervical screening invitations were sent to all study participants at the age of years, also in communities which organize the screening from age 0 onwards. With over-lapping timewindows we age-aligned the passive follow-up for the different cohorts over. year period. The study outcomes were incident invasive cervical cancer and cervical intraepithelial neoplasia grade (CIN+) incident during the passive follow-up. According to local standard of care, women with cytological abnormalities were referred to colposcopy biopsy for histopathological diagnosis within a month period following cytology. Thus, the passive, Finnish Cancer Registry (FCR) - based follow-up was started months after the end of active clinical follow-up of the PATRICIA (and 0) trial and 0 trials. For the control cohort the passive follow-up was age-aligned by a comparable time-period. Based on age-specific incidence of CIN+ in the Finnish female population ( the enrolled cohorts of HPV-/ vaccinated women and control women exceed the numbers required for at least 0% power to identify statistically significant % vaccine efficacy against CIN+.,0 An interim analysis for CIN+ was planned to take place after years of passive follow-up, and the final analysis for ICC after 0 years of passive follow-up. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

8 Page of BMJ Open The FCR is population-based and receives cancer notifications from the entire country with 00 % coverage, and 0% coverage for CIN ( Individually, the passive follow-up of the different cohorts extended until 0 years post vaccination or receipt of informed consent (unvaccinated women) up to the end of 0. Following Finnish national ethical committee clearances in 00 and 00, registers of HPV-/ vaccinated and unvaccinated cohorts were established and have since been maintained at the University of Tampere.,0 Permission to link these registers with the FCR for the identification new cancer cases until 0 was obtained from the Finnish Institute for Health & Welfare in 00. For the interim analysis the age-aligned HPV-/ vaccine (N = ) and the HAV-vaccine (N = ) cohorts of the clinical trials, and the control cohort (N = ) were linked using personal identifiers with the FCR to determine the incidence (per person years) of CIN and ICC (CIN+) during the overlapping.-year follow-up periods of passive follow-up. Histopathological block retrieval and re-analysis Diagnostic, formalin-fixed histopathological blocks were identified by the permission of Valvira, a department of the Finnish Ministry of Health and Social Welfare. An experienced pathologist confirmed that the retrieved archival diagnostic block contained a CIN+ lesion. All eligible blocks were sectioned according to a PCR-proof manner as described. Extraction, amplification and typing of the lesional HPV DNA was performed as previously described. Statistic analysis Vaccine efficacy (VE) was calculated on the intention-to-treat (ITT) principle including all individuals regardless of baseline HPV status receiving at least one HPV-/ vaccine dose in the arm of HPV vaccinated using statistical software SAS. software (SAS Institute, Cary, NC, USA) according to Ewell 0 and Chan. The % confidence intervals were based on exact binomial distribution of number of vaccinated cases conditional on total number of cases. 0, BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

9 Page of Results Between May 00 and June 00 a total of 0 - year-old Finnish women participated the PATRICIA (HPV-00) trial (fig. ). Concomitantly, - year old Finnish females received the AS0-adjuvanted HPV-/ vaccine in an HPV-0 immunogenicity trial. In May to June 00 and 00 respectively 0 and year old non-hpv vaccinated women responded to a health questionnaire and consented to the passive registry-based follow-up. Only the women who were willing to participate in an HPV vaccination trial provided that they were of appropriate age, and retained their consent for 0 years were eligible to the control cohort of unvaccinated women (fig. ). The demographics of the HPV-/ vaccine and control cohorts did not differ except for the birth cohort (table ). The proportions of ever users of oral contraceptives and the number of sexual partners were slightly higher, and the time of sexual debut was slightly lower in the vaccinated women as compared to the unvaccinated women. The sizeable cohorts of unvaccinated and HPV- / vaccinated women for ITT-analysis resulted in years of follow-up. Mixture of crossvaccination and continuation of active follow-up in the HAV vaccine arm precluded it from the passive long-term follow-up. Re-review of all the CIN+ cases was performed in percent of the cases. The presence of CIN+ was confirmed in percent of the diagnostic blocks available (table ). Three of the CIN cases identified during the passive follow-up among the HPV-/ vaccinated individuals could be confirmed in the re-review. All were baseline (pre-vaccination) positive for cervical HPV DNA, and HPV DNA was identified also in the diagnostic blocks containing the CIN lesion (table ). The fourth CIN case was baseline HPV DNA positive but no diagnostic block was available. One CIN cases was identified during the 0 follow-up. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

10 Page of BMJ Open All the CIN+ cases were found in the FCR follow-up between. to 0 years post vaccination. Identification of and cases yielded overall CIN+ incidence rates of / and / women years in the HPV-/ vaccinated cohort and the unvaccinated cohort, respectively. This resulted in % (%CI,) overall VE against CIN+, irrespectively of HPV type (table ). For the re-reviewed material the corresponding overall VE against CIN+ was % (%CI -, ). In the two cohorts, HPV was found in all the three HPV-/ vaccinated CIN cases and in % () of the unvaccinated CIN+ cases that were available and eligible for HPV DNA typing (table ). VEs against HPV or HPV/ associated CIN+ were low (table ). The VE estimate against other than HPV clade A HPV type, including ///, associated CIN+ increased from % to 00% when CIN+ lesions with HPV co-infection were excluded from the analysis (table ). Numbers for clade A or other non clade A HPV types were small. There were no health care interventions targeted to the control cohort, and the cancer-registry follow-up was passive. No harm was caused in this study. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

11 Page 0 of Discussion Ten years post vaccination we found statistically significant VE of % for the HPV-/ vaccine against any CIN+ in passive cancer registry -based follow-up of our population-based cohorts comprising more than 000 originally - to -year old women. Our interim ITT estimates of VE against any CIN+ VE are in line with what we reported about the -year follow-up of the total vaccinated cohort (TVC) from the PATRICIA trial: VE was.%. With the close to follow-up years and FCR based follow-up, our 0% power, 0.% cumulative CIN+ incidence and 0% VE assumptions,0 were conservative. Retrieval and review of the histopathological blocks were important quality control steps for the HPV typing. Even with the reduced number of cases the overall CIN+ VE estimate of % was comparable with the above Finnish Cancer Registry information-based estimate,. Unfortunately, the lack of baseline data for the unvaccinated cohort precluded the TVC (i.e., baseline HPV negative) naïve type of analyses which earlier demonstrated a very high (.%) VE against any CIN, irrespectively of HPV type in the PATRICIA trial years post vaccination. One limitation was that some HPV-/ vaccinated women (.%) also participated in the HPV- 0 study, the effects of which may have been contradictory. In one of these women annual HPV DNA screening may have led to an earlier detection of the CIN lesions, identified in the FCR follow-up, due to the high sensitivity of HPV DNA screening compared to conventional opportunistic cytology. On the other hand, removal of an earlier CIN lesion in an HPV-0 participant could theoretically have led to excision of a lesion that might have surfaced as CIN+ in the FCR follow-up. ITT analysis of the entire cohorts of HPV-/ vaccinated and unvaccinated women participating in the passive follow-up does not allow distinguishing between these two alternatives but is a conservative approach. Most importantly for the validity of the ongoing long- BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

12 Page of BMJ Open term follow-up, all our vaccinated and unvaccinated study subjects were invited to organized cytological screening visits at the age of years. Moreover, opportunistic HPV vaccination among the unvaccinated controls following the licensure of the quadrivalent Gardasil and bivalent Cervarix HPV vaccines in 00 and 00, respectively has been negligible (data not shown). We found only a relatively low long-term vaccine efficacy against HPV-/ positive CIN+. This was because several baseline, pre-vaccination HPV positive vaccine recipients, developed HPV positive CIN during the 0 years of post vaccination follow-up. The HPV-/ AS0- adjuvanted vaccine, does not protect against the HPV positive CIN+ if viral infection already exists and persistent HPV infection has been established. The low efficacy observed in the PATRICIA trial among baseline positives already pointed to this direction. Vaccination of adult women of whom up to 0+% already have been exposed to hrhpv infection is not an effective HPV vaccination strategy. In conclusion, ten years post vaccination the AS0-adjuvanted HPV-/ vaccine shows continued efficacy against CIN+ irrespectively of HPV type. Our results also suggest that the wide crossprotective efficacy of the HPV-/ vaccine reported in clinical trials against HPV types //, is true for the associated CIN+ end-point in the long-term context. It is warranted to continue the long-term follow-up of the HPVvaccination trial cohorts to all HPV-associated invasive cancer end-points. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

13 Page of Transparency declaration The lead author ML affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained. Data sharing Full dataset and statistical code available from the lead author at llmale@uta.fi. Consent was not obtained but the presented data are anonymised and risk of identification is low. Contributors ML, GD, JPaa, EP and FS designed the study. ML and EP participated in planning the statistical analysis. TL did the statistical analyses. DA, MK, JPa, TP and MS-M conducted the study as study doctors, and TE and KN coordinated the study. KH, TE and EP were responsible for the registry linkages. CL, PN, and JD were responsible for the lab analyses. ML and JPaa wrote the first draft of the manuscript, all the co-authors contributed to the writing in important intellectual content by revisions and comments. Acknowledgements The retrieval of histopathological blocks from Finnish pathology laboratories is gratefully acknowledged. Conflicts of interest DA, ML, JP and JD have received grants from GSK group of companies and/or Merck & Co. Inc. through their employers (DA, Family Federation Finland; ML; University of Tampere; JP and PN University of Helsinki; JD and ML, Karolinska Institute) for HPV vaccination studies. GD was and FS is employee of GSK group of companies. FS has received shares and stock options from the GSK group of companies. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

14 Page of BMJ Open References. IARC International Agency for Research on Cancer. A Review of Human Carcinogens, Biological Agents. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans 00; Part B: (Lyon, France).. zur Hausen H. HPV vaccines: what remains to be done? Expert Rev Vaccines 0;0:0-.. Munoz N, Kjaer S, Sigurdsson J, et al. Impact of human papillomavirus (HPV)-/// vaccine on all HPV-associated genital diseases in young women. J Natl Cancer Inst 00;0:-.. Lehtinen M, Paavonen J, Wheeler C, et al. Overall efficacy of HPV-/ vaccine against the most stringent cervical pre-cancer end-points: end-of study report of a double blind, randomized trial. Lancet Oncol 0;:-.. Lehtinen M, Dillner J. Clinical HPV vaccine trials and beyond. Nature Rev Clin Oncol 0;0:0-0.. Joura E, Giuliano A, Iversen O-E, et al. A -valent HPV vaccine against infection and intraepithelial neoplasia in women. NEJM 0;:-.. Chang M-H, Chen C-J, Lai M-S, et al. Universal hepatitis B vaccination in Taiwan and the incidence of hepatocellular carcinoma in children. New Engl J Med;:-.. McMahon BJ, Bulkow LR, Singleton RJ, et al. Elimination of hepatocellular carcinoma and acute hepatitis B in children years after a hepatitis B newborn and catch-up immunization program. Hepatology 0;:0-.. Lehtinen M, Idänpään-Heikkilä I, Lunnas T, et al. Population-based enrolment of adolescents in a long-term follow-up trial of human papillomavirus vaccine efficacy. Int J STD AIDS 00;:-. 0. Lehtinen M, Apter D, Dubin G, et al. Enrolment of,000 adolescent women to cancer registry follow-up for long-term human papillomavirus vaccine efficacy: guarding against guessing. Int J STD AIDS 00;:-. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

15 Page of Lehtinen M, Herrero R, Mayaud P, et al. Studies to assess the long-term efficacy and effectiveness of HPV vaccination in developed and developing countries. Vaccine 00;(S):-.. Rana M, Huhtala H, Apter D, et al. Cancer registry based follow-up in the understanding of long-term protection of human papillomavirus vaccination against cervical carcinoma. Int J Cancer 0;:-.. Paavonen J, Jenkins D, Bosch X, et al. Efficacy of a prophylactic adjuvanted L VLP vaccine against infection with HPV/ in young women: an interim analysis of a phase III double-blind, randomized controlled trial. Lancet 00;: -0.. Paavonen J, Halttunen M, Hansson B-G, et al. Feasibility studies on HPV vaccination. J Clin Virol 000;:-0.. Petäjä T, Pedersen C, Poder A, et al.. Long-term persistence of humoral and mucosal immune response to HPV-/ ASO-adjuvanted vaccine in preteen/adolescents girls & young women. Int J Cancer 0;:-.. Woodhall SC, Eriksson T, Nykanen A-M, et al. Impact of HPV vaccination on quality of life. Eur J Contracept Reproduct Health Care 0;:-.. Eriksson T, Torvinen, Lehtinen MM, et al. Impact of HPV/ vaccination on quality of life. Eur J Contracept Reproduct Health Care 0;::-.. Kohdunkaulan, ulkosynnytinten ja emättimen solumuutokset. Duodecim 00; -.. Lagheden C, Eklund C, Kleppe SN, et al. Validation of a standardized extraction method for formalin-fixed paraffin-embedded tissue samples, J. Clin. Virol. 0;0:-. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

16 Page of BMJ Open Ewell M. Comparing methods for calculating confidence intervals for vaccine efficacy. Stat Med ;:-.. Chan IS. Exact tests of equivalence and efficacy with a non-zeron lower bound for comparative studies. Stat Med ;:-.. Ronco G, Dillner J, Elfstrom M, et al. Efficacy of HPV-based screening for prevention of invasive cervical cancer: follow-up of four European randomised controlled trials. Lancet 0;:-.. Apter D, Wheeler CM, Paavonen J, et al. Efficacy of the human papillomavirus and (HPV-/) AS0-adjuvanted vaccine against cervical infection and precancer in young women: final event-driven analysis of the randomized, double-blinded PATRICIA trial. Clin Vac Immunol 0;:-.. Bosch XF, Robles C, Diaz M, et al. HPV Faster: Broadening the perspectives in the prevention of HPV related cancers. Nature Rev Clin Oncol 0;:-. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

17 Page of Table. Demographic characteristics of the cohorts subjected to Finnish Cancer Registry follow-up. HPV-/ vaccinated HAV-vaccinated Unvaccinated Category (N= ) (N= ) (N= ) Age at enrolment - years - years - years Age at passive follow-up - years - years - years Response rate* 0 (.%) 00 (.%) (.%) Sexual debut (mean age). years.0 years. years No. of life-time partners 0 (.%) (.%) 0 (.%) (.%) (.%) (.%) (.%) (.%) 0 (0.%) - (.%) (.%) 0 (.%) 0 or more (.%) (.%) (.%) No. of partners in the past months 0 (.%) 0 (0.%) (.%) (.%) 0 (.%) (.%) (0.%) (.%) 0 (.%) or more 0 (.%) (.%) (.%) Use of contraception BCP (ever) 0 (.%) 0 (.%) 0 (.%) Condom use (.%) (.%) 00 (.%) no use (.%) (.%) (.%) No contraception (0.%) (0.%) (0.%) (ever) *returned questionnaires at the age of - years when the passive follow-up was started last year BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

18 Page of BMJ Open Table. Incidence rate (/ women years) of cervical intraepithelial neoplasia grade and invasive cancer (CIN+) in cluster-randomized cohorts of - to -year-old HPV-/ vaccine recipients, and unvaccinated originally - to -year old women. Passive followup was by the population-based Finnish Cancer Registry up to 0 years post vaccination. End point of Vaccine Control the follow-up N Person yrs n Rate# N Person yrs n Rate# FCR registered CIN+ diagnoses Active 0 Intermittent Passive KI re-reviewed CIN+ diagnoses Intermittent Passive 0 Active (0- years), Intermittent (-. years), Passive (.-0 years) #incidence/ women years BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

19 Page of Table. Characteristics of CIN cases identified among the recipients of the HPV-/ vaccine in the Finnish Cancer Registry based passive long-term follow-up of the clinical trial participants between. and 0 years post vaccination. Number of Age at Baseline cervical doses received Date of End-point (CIN) HPV-0 enrolment HPV DNA status in diagnosis HPV DNA status participant years HPV doses Sep 00 HPV no years HPV doses May 0 HPV no years HPV doses Mar 0 HPV no years HPV doses Apr 0 n.a. yes* years HPV doses Mar 0 HPV yes # *One HPV DNA test after the end of the PATRICIA trial. years before the CIN+ diagnosis # CIN+ diagnosis made before start of the passive follow-up BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

20 Page of BMJ Open Table. Vaccine efficacy (VE, % confidence interval) against cervical intraepithelial neoplasia grade and invasive cancer (CIN+) associated with vaccine and/or non-vaccine HPV types in women vaccinated in 00/00 with the HPV-/ vaccine between ages - to - years and in an age-aligned control cohort of originally - to -year-old women passively followed via Finnish Cancer Registry for up to 0 years post vaccination. Vaccine Control End-point (CIN+) N Person yrs n Rate# N Person yrs n Rate# VE (% CI) HPV 0 (-0,) HPV. 00 (-00,00) HPV/ (-,) HPVA 0 (-,) HPVA a 0 00 (-,00) HPVA/A b 00 (-,00) HPV// 00 (-0,00) All protected HPV types c 0 (-0,) All protected HPV types d 00 (-0,00) All non-protected HPVs e.0 00 (-,00) All detected HPV types (-,) All detected HPV types fb 00 (-,00) Total f 0 (-,) Total all (.,) A=HPV/////, A=HPV////, All vaccine protected HPV types: ////////, All non-vaccine protected HPV types: /////////////,// and 0 a (excluding co-infections with ) b (excluding co-infections with /) c HPV//////// d HPV////// (excluding co-infections with /) e HPV///////////////0/ (excluding co-infections with /), f HPV positive and HPV negative original FCR registered CIN+ diagnoses #incidence/ women years BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

21 Page 0 of Legend for the figure. Figure. Consort diagram with relevant invitation, enrolment and exclusion criteria/steps. *Due to migration not eligible to Finnish Cancer Registry follow-up. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

22 Page of HPV/ vaccinated Continued active follow-up and/or crossvaccination No Vaccinated/00-.0 invited - year-olds + 0 one excluded* Consented, eligible HPV/ vaccinated LTFU participants. Yes No Unvaccinated/00. HAV vaccinated Continued active follow-up and/or crossvaccination All excluded Yes +. BMJ Open Yes excluded 0. invited - year-olds.0 respondents Declined to participate or had already participated in an HPV vaccination trial No mjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright. Consented, eligible unvaccinated LTFU participants of the combined cohort. Unvaccinated/00. invited - year-olds 0.0 respondents Declined to participate or had already participated in an HPV vaccination trial No Yes.0 excluded

23 Page of Section/Topic Title and abstract Introduction Background and objectives CONSORT 00 checklist of information to include when reporting a randomised trial* Item No Checklist item a Identification as a randomised trial in the title b Structured summary of trial design, methods, results, and conclusions (for specific guidance see CONSORT for abstracts) a Scientific background and explanation of rationale b Specific objectives or hypotheses Reported on page No Methods Trial design a Description of trial design (such as parallel, factorial) including allocation ratio b Important changes to methods after trial commencement (such as eligibility criteria), with reasons NA Participants a Eligibility criteria for participants b Settings and locations where the data were collected Interventions The interventions for each group with sufficient details to allow replication, including how and when they were - actually administered Outcomes a Completely defined pre-specified primary and secondary outcome measures, including how and when they were assessed b Any changes to trial outcomes after the trial commenced, with reasons NA Sample size a How sample size was determined b When applicable, explanation of any interim analyses and stopping guidelines Randomisation: Sequence a Method used to generate the random allocation sequence reference generation b Type of randomisation; details of any restriction (such as blocking and block size) reference Allocation concealment mechanism Mechanism used to implement the random allocation sequence (such as sequentially numbered containers), describing any steps taken to conceal the sequence until interventions were assigned Implementation 0 Who generated the random allocation sequence, who enrolled participants, and who assigned participants to interventions Blinding a If done, who was blinded after assignment to interventions (for example, participants, care providers, those reference CONSORT 00 checklist Page on January 0 by guest. Protected by copyright. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from

24 Page of BMJ Open assessing outcomes) and how b If relevant, description of the similarity of interventions NA Statistical methods a Statistical methods used to compare groups for primary and secondary outcomes b Methods for additional analyses, such as subgroup analyses and adjusted analyses NA Results Participant flow (a a For each group, the numbers of participants who were randomly assigned, received intended treatment, and, figure diagram is strongly recommended) b were analysed for the primary outcome For each group, losses and exclusions after randomisation, together with reasons figure Recruitment a Dates defining the periods of recruitment and follow-up b Why the trial ended or was stopped NA Baseline data A table showing baseline demographic and clinical characteristics for each group table Numbers analysed For each group, number of participants (denominator) included in each analysis and whether the analysis was figure by original assigned groups Outcomes and a For each primary and secondary outcome, results for each group, and the estimated effect size and its - estimation precision (such as % confidence interval) b For binary outcomes, presentation of both absolute and relative effect sizes is recommended NA Ancillary analyses Results of any other analyses performed, including subgroup analyses and adjusted analyses, distinguishing NA pre-specified from exploratory Harms All important harms or unintended effects in each group (for specific guidance see CONSORT for harms) Discussion Limitations 0 Trial limitations, addressing sources of potential bias, imprecision, and, if relevant, multiplicity of analyses 0 Generalisability Generalisability (external validity, applicability) of the trial findings Interpretation Interpretation consistent with results, balancing benefits and harms, and considering other relevant evidence, Other information Registration Registration number and name of trial registry Protocol Where the full trial protocol can be accessed, if available, HPV-0 Funding Sources of funding and other support (such as supply of drugs), role of funders *We strongly recommend reading this statement in conjunction with the CONSORT 00 Explanation and Elaboration for important clarifications on all the items. If relevant, we also recommend reading CONSORT extensions for cluster randomised trials, non-inferiority and equivalence trials, non-pharmacological treatments, herbal interventions, and pragmatic trials. Additional extensions are forthcoming: for those and for up to date references relevant to this checklist, see CONSORT 00 checklist Page on January 0 by guest. Protected by copyright. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from

25 Ten-year follow-up of human papillomavirus vaccine efficacy against the most stringent cervical neoplasia endpoint registry-based follow-up of randomized controlled trial Journal: BMJ Open Manuscript ID bmjopen-0-0.r Article Type: Research Date Submitted by the Author: -Mar-0 Complete List of Authors: Lehtinen, Matti; Tampereen Yliopisto, School of Health Sciences; Karolinska Institutet, Department of Laboratory Medicine Lagheden, Camilla; Karolinska Institutet, Department of Laboratory Medicine Luostarinen, Tapio; Karolinska Institutet, Department of Laboratory Medicine Eriksson, Tiina; Tampereen Yliopisto, School of Health Sciences Apter, Dan; VL Medi Harjula, Katja; Tampereen Yliopisto, School of Health Sciences Kuortti, Marjo; Tampereen Yliopisto, School of Health Sciences Natunen, Kari; Tampereen Yliopisto, School of Health Sciences Palmroth, Johanna; Tampereen Yliopisto, School of Health Sciences Petäjä, Tiina; Tampereen Yliopisto, School of Health Sciences Pukkala, Eero; Cancer Society of Finland Siitari-Mattila, Mari; Tampereen Yliopisto, School of Health Sciences Struyf, Frank ; GlaxoSmithKline Biologicals SA Site de Wavre Nieminen, Pekka; Helsingin Yliopisto Paavonen, Jorma; University Central Hospital, Department of Obstetrics and Gynecology Dubin, Gary; Takeda Pharmaceuticals International Dillner, Joakim; Karolinska Institutet, Department of Medical Epidemiology and Biostatistics BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from <b>primary Subject Heading</b>: Oncology Secondary Subject Heading: Infectious diseases Keywords: HPV vaccines, long-term follow-up, vaccine efficacy, CIN+, cervical cancer, randomized controlled trial on January 0 by guest. Protected by copyright.

26 Page of BMJ Open Ten-year follow-up of human papillomavirus vaccine efficacy against the most stringent cervical neoplasia end-point registry-based follow-up of randomized controlled trial Matti Lehtinen,, Camilla Lagheden, Tapio Luostarinen, Tiina Eriksson, Dan Apter, Katja Harjula, Marjo Kuortti, Kari Natunen, Johanna Palmroth, Tiina Petäjä, Eero Pukkala, Mari Siitari-Mattila, Frank Struyf, Pekka Nieminen, Jorma Paavonen, Gary Dubin, Joakim Dillner University of Tampere, Tampere, Finland; Karolinska Institute. Stockholm, Sweden; Family Federation Finland, Helsinki, Finland; Finnish Cancer Registry, Helsinki, Finland; GSK Biologicals, Wavre, Belgium; University of Helsinki, Helsinki, Finland; Takeda Pharmaceuticals International, Switzerland Address for correspondence: Matti Lehtinen, Department of Laboratory Medicine, Karolinska Institute, Stockholm, Sweden, matti.lehtinen@uta.fi Funding GlaxoSmithKline Biologicals SA (Belgium), Academy of Finland, Finnish Cancer Organizations and the Swedish Cancer Society BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

27 Page of Objective Due to long lag time between infection/cancer diagnoses human papillomavirus (HPV) vaccination programs will deliver vaccine efficacy (VE) estimates against cancer end-points late. Cancer registry follow-up of population-based, randomized trial cohorts of vaccinated and unvaccinated women speeds up the VE estimation. Methods We report interim results with person years of Finnish Cancer Registry -based follow-up of individually and/or cluster randomized cohorts of HPV-/ vaccinated and unvaccinated adolescent women enrolled in June 00/00, and between May 00 and April 00, respectively. The cohorts comprised - to -year-old unvaccinated women (NCT0), and and - to -year-old HPV-/ vaccinated women participating the PATRICIA (NCT00) and HPV-0 (NCT00) trials, respectively. The agealigned passive follow-up started months after the clinical trials end. Results During the follow-up of. to 0 years post enrolment we identified cases of cervical intraepithelial neoplasia grade (CIN) and cases of invasive cervical cancer (ICC) in the unvaccinated cohort, and CIN cases in the HPV-/ vaccinated women. Diagnostic blocks were available for HPV typing from % of the cases. CIN+ lesions were detectable in cases. HPV was found in of 0 unvaccinated CIN+ cases, and in CIN+ cases in the HPV-/ vaccinated women. The latter were all baseline positive for cervical HPV DNA. Baseline data was not available for the unvaccinated women. Intention-to-treat VE against any CIN+ was % (% CI, ). Conclusions Ten years post vaccination the AS0-adjuvanted HPV-/ vaccine shows continued efficacy against CIN+ irrespectively of HPV type. Vaccine efficacy was not observed in baseline HPV DNA positive subjects. All the trial cohorts were commenced before July 00 BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

28 Page of BMJ Open Strengths and limitations of this study *Country-wide cancer registry follow-up of sizeable randomised cohorts of HPV vaccinated and unvaccinated women for person years (up to0 years post vaccination) provides most reliable vaccine efficacy estimates against cancerous end- points. *Retrieval of most diagnostic histopathological blocks and state-of-science identification of the causal HPV type in the lesion enable identification of HPV type-specific vaccine efficacy estimates. *The per protocol defined interim analysis has limited statistical power. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

29 Page of Introduction High-risk (hr) human papillomaviruses (HPVs) cause up to % and % of cancers in females and males. Bivalent, quadrivalent and nonavalent vaccines against HPV types /, ///, and ////////, respectively, have an acceptable safety profile and are highly efficacious against a number of infections with hrhpvs and associated precancers. - Proof of vaccine efficacy (VE) against HPV-associated cancers is, however, not easy to reach due to the long lag time between exposure to the virus and diagnosis of the associated cancer. This phenomenon is not uncommon, and has for instance hindered the determination of VE against hepatitis B virus (HBV) associated hepatocellular carcinoma, since the time lag between virus exposure and diagnosis of carcinoma is to., Proving the concept of vaccine induced protection against (one of) the major HPV-associated cancers, i.e., in situ and invasive cervical carcinoma is not only of conceptual but also of practical importance. It would guarantee the impact of primary cancer prevention via prophylactic HPV vaccination, guide targeting this prophylaxis, and provide the scientific basis for understanding how and when such proof will be available for other HPV-associated cancers such as non-cervical anogenital cancers, and head and neck cancers. Our objective is to determine VE against cervical cancer. To accomplish this we identify invasive cervical cancer (ICC) and intraepithelial neoplasia grade (CIN+) incidence in passive, population-based cancer registry follow-up of a randomized clinical HPV vaccine trial cohort (PATRICIA), and a cluster-randomized control cohort enrolled in The follow-up of originally adolescent females has previously proven to be feasible. With the observed CIN+ incidence of./ in the control cohort, which equals that of the Finnish female population of similar age./00 000, it is well powered to verify % VE against CIN+ and ICC, 0 and years post vaccination, respectively. - We report interim results on the efficacy of the bivalent HPV/ vaccine against overall and HPV type-specific CIN+ end-points. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

30 Page of BMJ Open Methods Study design and Ethics Our cluster-randomized follow-up involves separate birth cohorts of clinical trial participants,, aged - years and non-participants,0 aged - years assigned according to the start of PATRICIA trial in May 00 by ML (the principal investigator for phase III HPV vaccination trials in Finland). The former were also individually randomized. The trials, establishing the unvaccinated control cohort and their Finnish Cancer Registry (FCR) based follow-up were approved by the Finnish National Ethical Review Board (TUKIJA: /00 and /00), respectively. Patient involvement Patients (with cervical neoplasia / condyloma) were not involved in the design of this study in 000. A study on the feasibility of population-based enrolment was done in. Enrolment Starting in May 00 originally all 0 (Q/-Q/ born) - year old Finnish females resident in trial communities were sent invitations to participate to the PATRICIA trial (HPV-00, NCT00) on the immunogenicity, safety and efficacy of the AS0-adjuvanted HPV-/ vaccine against HPV/ positive CIN+ (fig. ). By June 00 a total of 0 women participated. They were randomly assigned to HPV-/ or hepatitis A-virus (HAV) vaccination in : ratio to receive three doses of the HPV-/ vaccine or the HAV vaccine at months 0, and, followed by active follow-up visits with month interval up to years., In addition, - year olds received doses of the HPV-/ vaccine by May 00 in a concomitant HPV-0 trial which ended years later. Following the end of the active follow-up in May 00, approximately 0% of the HAV-vaccine recipients chose HPV-/ cross-vaccination during HrHPV DNA positives at the last PATRICIA trial visit ( from the HPV-/ and HAV cohorts) continued active clinical follow-up (HPV-0 study protocol) for approximately. years after the end of the PATRICIA trial. They had annual HPV testing cytology until HPV DNA negative or exit colposcopy after years. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

31 Page of An unvaccinated control cohort from entire adjacent birth cohorts of - year olds was recruited inviting 0 and in June 00/ by the Finnish Population Register Centre as described (NCT0). - All PATRICIA, and earlier FUTURE trial participants were excluded from the control cohort. There were no health care interventions targeted to the control cohort. Follow-up Both the vaccinated and unvaccinated control cohorts responded at the age of - years to a questionnaire on life habits with special emphasis on sexual health, i.e., at the beginning of the passive registry-based follow-up., Opportunistic vaccination by Gardasil or Cervarix vaccines after their licensures in 00 and 00 was considered based on questionnaires in 00 and 00, and among the vaccinated PATRICIA and HPV-0 trial cohorts in 00. In addition, vital status and emigration were updated until the end of 0. Invitations to cervical screening were sent to all study participants at the age of years, also in communities which organize the screening from age 0 onwards. With over-lapping time-windows of. years we age-aligned the passive follow-up for the different birth cohorts. The study outcomes were invasive cervical cancer and cervical intraepithelial neoplasia grade (CIN+) diagnosed during the passive follow-up. According to local standard of care, women with cytological abnormalities were referred to colposcopy biopsy for histopathological diagnosis within a month period following cytology. Thus, the passive, Finnish Cancer Registry (FCR) -based follow-up was started months after the end of active clinical follow-up of the PATRICIA (and 0) trial and 0 trials. For the control cohort the passive follow-up was age-aligned by a comparable time-period. Based on age-specific incidence of CIN+ in the Finnish female population ( the enrolled cohorts of HPV-/ vaccinated women and control women exceed the numbers required for at least 0% power to identify statistically significant % vaccine efficacy against CIN+.,0 An interim analysis for CIN+ was planned to take place after years of passive follow-up, and the final analysis for ICC after 0 years of passive follow-up. BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.

32 Page of BMJ Open The FCR is population-based and receives cancer notifications from the entire country with 00 % coverage, and 0% coverage for CIN ( Individually, the passive follow-up of the different cohorts extended until 0 years post vaccination or receipt of informed consent (unvaccinated women) up to the end of 0. Following Finnish national ethical committee clearances in 00 and 00, registers of HPV-/ vaccinated and unvaccinated cohorts were established and have since been maintained at the University of Tampere.,0 Permission to link these registers with the FCR for the identification new cancer cases until 0 was obtained from the Finnish Institute for Health & Welfare in 00. For the interim analysis the age-aligned HPV-/ vaccine (N = ) and the unvaccinated control cohort (N = ) were linked using personal identifiers with the FCR to determine the incidence (per person years) of CIN and ICC (CIN+) during the overlapping.-year follow-up periods of passive follow-up. Histopathological block retrieval and re-analysis Diagnostic, formalin-fixed histopathological blocks were identified by the permission of Valvira, a department of the Finnish Ministry of Health and Social Welfare. An experienced pathologist confirmed that the retrieved archival diagnostic block contained a CIN+ lesion. All eligible blocks were sectioned according to a PCR-proof manner as described. Extraction, amplification and typing of the lesional HPV DNA was performed as previously described. Statistic analysis Vaccine efficacy (VE) was calculated on the intention-to-treat (ITT) principle including all individuals regardless of baseline HPV status receiving at least one HPV-/ vaccine dose in the arm of HPV vaccinated using statistical software SAS. software (SAS Institute, Cary, NC, USA) according to Ewell 0 and Chan. The % confidence intervals were based on exact binomial distribution of number of vaccinated cases conditional on total number of cases. 0, BMJ Open: first published as 0./bmjopen-0-0 on August 0. Downloaded from on January 0 by guest. Protected by copyright.