Non-inferior antibody levels for HPV16/18 after extended two-dose schedules compared with a six-month interval: findings of a systematic review and meta-analysis

ABSTRACT

Protection after human papillomavirus (HPV) vaccination can be maximized by optimizing vaccination schedules. We systematically reviewed immunogenicity and effectiveness of HPV vaccines administered 6 months apart compared with longer intervals. Seroconversion to vaccine-type HPV was non-inferior for 12- compared with 6-month intervals, but inconclusive for comparison of 36–96 months with 6 months. A 12-month interval showed non-inferior (margin 0.5) vaccine-type HPV antibody responses compared with a 6-month interval. Compared to 6 months, an interval of 36–96 months resulted in non-inferior antibody responses for HPV6 and high-risk types HPV16 and 18, but did not lead to a non-inferior antibody response for HPV11 (GMR 0.63, 95% CI:0.41–0.97). Data on the effectiveness of extended two-dose schedules were limited. Our findings indicate that HPV immunization programs could adopt a 12-month interval instead of 6 months for increased flexibility without compromising immunogenicity. Further evaluation to confirm the immunogenicity and effectiveness of intervals beyond 12 months is warranted.

KEYWORDS:

• Antibody levels
• extended intervals
• human papillomavirus vaccine
• two-dose schedule
• immunization schedule
• non-inferiority

Introduction

The human papillomavirus (HPV) is the most common sexually transmitted infection.^1,2 It is estimated that almost all cervical cancer cases are attributable to high-risk HPV types.³ Benign anogenital warts are caused by low-risk HPV types and can significantly reduce quality of life.^4,5 Immunization using the bivalent, quadrivalent or nonavalent HPV vaccines is effective in protecting against HPV infections and precursor lesions of anal and cervical cancer.⁶ Immunization is a vital component of the global strategy for elimination of cervical cancer as a public health problem by 2030, and one of the aims is to fully vaccinate 90% of girls worldwide against HPV by age 15.⁷ Providing as much flexibility as possible in the vaccination schedule, assuming equivalent effectiveness, could help in realizing this global target, to ensure that vaccine schedules meet local public health realities.

Current recommendations for administering two-doses of HPV vaccines to immunocompetent individuals are: 5–13 months apart for the bivalent vaccine if recipient 9–14 years,⁸ 6 or 12 months apart for the quadrivalent vaccine if recipient 9–13 years⁹ and 6–12 months apart for the nonavalent vaccine if recipient 9–14 years,¹⁰ for older age groups three-doses are recommended. School-based immunization is widely used for delivery of HPV vaccines to the primary target population who are early adolescents.¹¹ A majority of school-based immunization programs deliver HPV vaccines on a 0,6-month schedule.^11,12 Therefore at least two immunization clinics must be held six months apart in each participating school within an already congested nine-month academic calendar. This causes logistical challenges^13,14 that limit opportunities for completing HPV vaccination, which have recently been worsened while trying to deliver the program during the COVID-19 pandemic.

In addition to programmatic benefits, there might be immunological benefits from using schedules with an interval longer than 6 months. Indications from Bergman et al are that extension of the two-dose interval up to 12 months, may generate higher antibody levels compared to an interval of two to six months.¹⁵ It is not clear whether this trend continues when the two-dose interval extends beyond 12 months. Secor et al did not find higher antibody levels for extended interval schedules longer than 12 months relative to standard intervals. However, the comparisons made by Secor et al were based on pooled data for two and three dose schedules and are not definitive for two dose schedules.¹⁶ Further neither study investigated the relative effectiveness of extended and standard interval two dose schedules against HPV-infected and HPV-related disease outcomes. Comparisons of this kind, could help in delimiting optimal two-dose intervals. This systematic review evaluates HPV vaccine immunogenicity and effectiveness for two doses administered six months apart compared with extended intervals.

Methods

Protocol and registration

The protocol for this systematic review was registered in the International Prospective Register of Systematic Reviews (PROSPERO), under protocol number: CRD42019141959.

Literature search

Systematic literature searches were updated till July 10, 2020. Indexed databases searched were: MEDLINE (Ovid), EMBASE (Ovid), Cochrane Central Register of Clinical Trials, PUBMED, and CINAHL. The original search strategy was developed in Medline (Supplemental Appendix 1) and modified for use with the other platforms, based on their specific syntax and thesaurus. Unindexed databases were searched as follows: the term “HPV vaccine” was used to search GSK study register, MERCK clinical trials, EU Clinical Trials Register and Drugs@FDA, whilst the term “HPV vaccine schedule” was used to search ClinicalTrials.gov and International Clinical Trials Registry Platform.

Study selection

A two-stage selection process was used and consisted of title/abstract screening followed by full-text screening. The screening process was performed in duplicate (ACF and RD). Studies were included if they fulfilled all of the following criteria: study participants were 9–26 years inclusive at first HPV vaccine dose; administered any combination of approved prophylactic HPV vaccines in a two-dose schedule to females and/or males; and study presented data on immunogenicity, effectiveness or efficacy of two doses of HPV vaccine administered at 6 months (± 1 month) as well as two doses administered more than 7 months apart. Studies were excluded if they: included populations with morbidities only; administered a vaccine formulation that is different from the licensed formulation; are reported in a language other than English.

Data extraction and quality assessment

ACF and RD independently extracted data from selected studies into standardized data collection forms, after which ACF checked both sets of extracted data for accuracy. Extracted data included: publication details, general study characteristics, intervention and comparator (HPV vaccine and schedule) and outcomes of interests namely seroconversion, geometric mean concentration or titers (GMC/T), B-cell and T-cell frequencies and rates of HPV infection or HPV-related diseases. The potential for bias and confounding in each study was assessed using the Cochrane Risk of Bias tool version 2 for randomized studies¹⁷ and the Cochrane Risk of Bias in Non-Randomized Studies of Intervention (ROBINS-I) tool for observational studies.¹⁸

Analyses and certainty assessment

To validly compare effects across studies, within-study relative measures for the immunogenicity and effectiveness of the extended interval (0,>7 month) versus the standard (0,6 month) interval were obtained. Using the relative within-study effect measures allowed controlling for inter-study methodological variations such as differences in definition of seroconversion or assays of antibody response across studies.¹⁹ Immunogenicity was assessed by seroconversion difference as well as antibody response using the geometric mean ratio (GMR) comparing the GMC or GMT for the extended versus the standard interval. For randomized control trials (RCTs), measures for the per protocol populations were used to avoid misclassification of vaccination schedules. Non-inferiority was tested by constructing two-sided 95% confidence intervals (CI) and comparing them to pre-specified non-inferiority criteria. 95% CI for the seroconversion estimates were calculated using Wilsons exact approach without continuity correction. These non-inferiority criteria were previously reported²⁰ and specify that the lower limit of the 95% CI be > 0.5 for GMR and > −10% for the difference in seroconversion. Where non-inferiority was demonstrated based on ratios or proportions, superiority was assessed to determine if the extended interval was better than 6 months. Superiority required the lower limit of the 95% CI to exceed 1 for ratios and 0 for differences in proportions. Incidence rate difference (IRD) was calculated using incidence count and person years data extracted from study reports. The unavailability of non-inferiority margins for IRD and the intricacies involved in defining meaningful margins for this measure, discouraged such analysis here. The certainty of the evidence available to assess each study objective was determined using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE).²¹

Results

Study selection

The systematic search yielded 1451 unique records, of which 11 publications reporting on four different studies satisfied the selection criteria after full-text screening (Figure 1). Table 1 describes the characteristics of included studies.

Table 1. Characteristics of included studies

						Outcomes Reported
						Immunogenicity			Efficacy/Effectiveness
Study/Trial ID	Study Location	Study Participants (sex, age in years)	Intervention (Vaccine & Schedule)	Sample Size (Total Vaccinated)	Publication (Author_Year)	Seroconversion	GMC/T	Cellular Immunogenicity	AGW	Incident: HPV/Cancer/Pre-cancer
NCT01381575 & EUCTR2011-000757-22-IT	Canada, Taiwan, Thailand, Germany, Italy	F, 9–14 F, 9–14	2 v @ 0,6 m 2 v @ 0,12 m	550	GSK 2011*^+	●	●	●	º	º
				415	Huang et al 2017	●	●	●	º	º
				965	Puthanakit et al 2016	●	●	●	º	º
NCT01984697	Canada, Chile, Columbia, Czech Republic, Denmark, Israel, Maleysia, Norway, South Korea, South Africa, Spain, Taiwan, Thailand, Turkey, United States	F, 9–14 M, 9–14 FM, 9–14	9 v @ 0,6 m 9 v @ 0,6 m 9 v @ 0,12 m	301 301 301 903	Merck 2013*	●	●	º	º	º
					Iversen et al 2016	●	●	º	º	º
					Joohee et al 2016	●	●	º	º	º
					Yan et al 2016 Khurana 2016	● º	● ●	º º	º º	º º
NCT0256795 & NCT03431246	Canada	FM, 9–14 F, 9–14	9 v @ 0,6 m 4 v/9 v @ 36–96 m	173 31 204	Gilca et al 2019	●	●	º	º	º
Lamb et al 2017	Sweden	F, ≤16 F, 17–19	4 v @ 4–7 m 4 v @ 8 + m 4 v @ 4–7 m 4 v @ 8 + m	8095 1894 9989 2965 615 3580	Lamb et al 2017	º	º	º	●	º

● = available data, º = no data, 2 v = bivalent HPV vaccine, F = female, M = male, FM = female and male, 4 v = quadrivalent HPV vaccine, 9 v = nonavalent HPV vaccine, 4 v/9 v = mixed regimen of 4 v then 9 v HPV vaccine, GMC/T = geometric mean concentration/titers from enzyme linked immunosorbent assay, competitive luminex immunoassay or pseudoviron based neutralization assay, AGW = ano-genital warts, * = source from which data were extracted when multiple publications available, ⁺ = two identical trial records available but data extracted from NCT0138157

Figure 1. Screening and selection

* The eleven articles are from 4 distinct studies. ⁺ Data from all three immunogenicity studies were used in meta-analysis.

Seroconversion

At 1 month post last dose, non-inferior seroconversion to vaccine type HPV was observed for the 12-month interval versus the 6-month interval when using the bivalent and nonavalent HPV vaccines (Figure 2a). For the bivalent vaccine, non-inferiority of a 12-month versus 6-month interval for seroconversion was shown to 1 year post last dose. Non-inferior seroconversion could not be concluded when the mixed quadrivalent/nonavalent regimen was administered at an interval of 36 to 96 months compared with 6 months (Figure 2b).

Figure 2. (A) Difference in seroconversion to HPV vaccine types after two doses of HPV vaccine administered at 12 versus 6 months apart, at 1 month after last HPV vaccine dose. (B) Difference in seroconversion to HPV vaccine types after two doses of HPV vaccine were administered at 36 to 96 versus 6 months apart, at 1 m after last HPV vaccine dose

The line at −10% is the pre-specified non-inferiority margin for the difference of the 12-month interval group versus 6-month interval group. The line at 0% represents no difference.The line at −10% is the pre-specified non-inferiority margin for the difference of the 36 to 96-month interval group versus the 6-month interval group. The line at 0% represents no difference.

Antibody response (Supplemental Appendix 2)

The administration of the bivalent HPV vaccine with a 12-month interval was non-inferior compared to administration using a 6-month interval for HPV18 (GMR 1.11, 95% CI: 1.00 to 1.23) at 1 month post last dose (Figure 3a). GMR comparing the 12-month to the 6-month interval for each individual vaccine type HPV showed substantial heterogeneity for the peak antibody response to HPV16 and 18 at 1 month post last dose (I² = 76%), precluding pooling of these data. For the bivalent vaccine, the 12-month interval had antibody responses superior to the 6-month interval for both HPV16 (GMR 1.22, 95% CI: 1.09 to 1.37 at 6 months later and 1.27 95% CI: 1.13 to 1.42 at 12 months later) and HPV18 (GMR 1.22, 95% CI: 1.08 to 1.38 at 6 months later and 1.35, 95% CI: 1.19 to 1.55 at 12 months later). At 1 month post last dose, a two-dose interval of 36 to 96 months between two doses was non-inferior to 6 months for HPV6, 16 and 18, but not HPV11 (GMR 0.63, 95% CI: 0.41 to 0.97, Figure 3b).

Figure 3. (A) Ratios for antibody levels of HPV vaccine-types for two-dose schedules administered at intervals of 12 months versus 6 months at 1 month post last HPV vaccine dose. (B) Ratios for antibody levels of HPV vaccine-types for two-dose schedules administered at intervals of 36–96 months versus 6 months at 1 month post last HPV vaccine dose

The line at 0.5 is the pre-specified non-inferiority margin for the ratio of the 12-month interval group versus 6-month interval group. The line at 1 represents no difference.The line at 0.5 is the pre-specified non-inferiority margin for the ratio of the 36 to 96-month interval group versus 6-month interval group. The line at 1 represents no difference.

Cellular immune response

Memory B-cell and T-cell immune responses were measured in a subset of participants in a study evaluating the bivalent vaccine (NCT01381575). No CD8+ T cell response to HPV16/18 was detectable for any of the evaluated schedules. The memory B-cell and CD4+ T-cell response were of similar magnitude for the 0, 12 m and 0, 6 m schedules, at 36 m post-dose-one. This study however was not powered to detect differences in cellular immune responses after two doses were administered 12 or 6 months apart.²²

Efficacy and effectiveness

There were no studies reporting efficacy data for comparing the extended and standard intervals, while a single study provided data for comparing effectiveness. Lamb et al.²³ provided incidence data that was used to calculate anogenital wart IRDs for the 8 + m interval schedule relative to the 4–7 m interval schedules. The IRD among females younger than 17 years was 256.21 (95% CI −12.11– 524.53) per 100,000 person-years and among those aged 17–19 years IRD was 448.62 (95% CI – 49.41– 946.64) per 100,000 person-years.

Quality and certainty assessments

Risk of bias assessment indicated a low risk of bias for RCTs evaluating the bivalent and nonavalent vaccines, while the observational studies were assessed to have an overall moderate risk of bias (Supplemental Appendix 3). In applying the GRADE certainty assessment, effect estimates for the bivalent and nonavalent studies were deemed to have high certainty due to them being RCTs with low risk of bias and over 100 participants per study group. The effect estimates based on the observational studies however were assessed to be of low certainty due to their moderate risk of bias and wide confidence intervals (Supplemental Appendix 4). Supplemental Appendix 2 provides an overview of calculated effect estimates and associated certainty for all outcomes assessed in this study.

Discussion

We systematically reviewed the non-inferior immunogenicity and effectiveness of two doses of HPV vaccine when administered at an extended interval of 7 or more months apart, versus a standard interval of 6 months. Our findings extend upon the trend reported by Bergman et al.¹⁵ where antibody levels increase as dose intervals increase from 2 to 6 months and from 6 to 12 months. Consistent with this trend, we found that for the bivalent and nonavalent vaccines, increasing the two-dose interval from 6 to 12 months resulted in superior antibody response to all HPV types, except for the bivalent vaccine for HPV18, although an interval of 12 months was non-inferior to an interval of 6 months. Although superiority was shown when an interval of 12 months was compared to 6 months for HPV6, 11 and 16, it is unclear how this finding would translate into effectiveness given the lack of a correlate of protection.²⁴ The non-inferiority of a 12-month interval to a 6-month interval persisted to 1 year post last dose for the antibody response that the bivalent vaccine induces to HPV16 and 18. This is the maximum period for which follow-up data are available.

Although not included in this review, which used the defined minimum required interval of five months as comparator, in line with recommendations of WHO/EMA, several previous studies included shorter intervals. These studies found lower antibody levels when comparing a one or two month interval compared with an interval of six months.^25,26

Our findings of non-inferior immunogenicity for the two-dose schedule administered over a year are comparable to an evaluation of extended three-dose schedules, were extended schedules up to 12 and months showed non-inferior antibody schedules one month after the third dose.²⁷ At 29/30 months after the first dose also a schedule extended up to 24 months showed non-inferior antibody responses. Evaluation of this extended dosing schedule (0,12,24 months) at 20 months after the first dose also resulted in non-inferior antibody responses. In addition, for all evaluated schedules the authors compared the antibody responses before administration of the third dose. Interestingly, the antibody titers were comparable or higher after two-doses administered at 0,12 compared to a six month interval for HPV6/11/16/18 before administration of the third dose at 24 and 12 months respectively.²⁸

We did not find a trend of higher antibody levels with increasing interval length to continue when the interval was 36–96 months. Rather, increasing the dose interval from 6 months to between 36 and 96 months resulted in non-inferior antibody response to HPV6 and high-risk HPV types 16 and 18, but not HPV11. It is notable that the single study providing data to compare the 36–96 months interval to the standard interval was not powered to assess superiority of the extended interval. The HPV11 comparison is the only antibody response that failed to demonstrate non-inferiority between extended and standard two-dose intervals. In so doing, it failed to indicate adequate protection by the immunobridging principle, since previous studies showed that vaccine-type antibody responses elicited by the standard (6 months) two-dose schedule in girls were non-inferior to antibody responses elicited by three doses in adolescents and women.^29–31 The clinical significance of the HPV11 antibody response failing to achieve non-inferiority is unclear since there is no known minimum antibody level for protection against HPV infection.²⁴ Studies have shown that memory B-cells may trigger a protective anamnestic response, even if plasma antibody levels are low or undetectable.^32,33

Only one study reported on the effectiveness of two-dose schedules of six or more months using AGW as the outcome, no studies were found that evaluated the efficacy or effectiveness against HPV infections, precursor lesions or ultimately HPV-associated cancers. Bergman et al.¹⁵ also highlighted a lack of available data for comparing HPV vaccine effectiveness when two-dose schedules with extended and standard intervals are used.

Booster vaccine doses mediate their effect through activation of new naïve B cells (not stimulated by the priming dose) and re-activation of memory B cells (elicited by the priming dose). The interval between prime and boost vaccine doses can mediate the interaction between the vaccine antigen and both naïve and memory B cells via the presence of antigen-specific antibody at time of boost.^34–36 Extending the interval between doses will likely result in lower antibody levels at the time of booster dose administration, which may allow greater opportunity for stimulation of B cell responses by antigen at this time.³⁶ This may explain the generally superior/higher antibody responses observed for a 12 month interval versus a 6 month interval. The duration of time over which the antibody response remains protective is another important consideration for HPV vaccines. The primary target age for HPV administration is 9–14 years,² but peak acquisition of HPV extends up to age 25 in many populations.³⁷ Evaluation of the long-term effectiveness of both two-dose schedules with 6-month interval as well as those with extended dose intervals provided by HPV vaccines is therefore an area for further studies, as it is ideal that HPV vaccines remain protective against infection throughout the years of peak HPV exposure.

The higher antibody responses for a 12-month interval suggests that it could be used, instead of 6 months, for school-based HPV immunization programs. With the need to improve the resilience of school-based immunization to disruptions, prolonged intervals could be explored as an option to reduce annual HPV vaccine needs during periods of HPV vaccine shortage. When using a 0,6 month schedule in school-based immunization programs logistical challenges^13,14 limit opportunities for completing HPV vaccination since missed school clinics are not easily rescheduled and timing restrictions make it difficult for such programs to adapt to disruptions. Unanticipated school closures can complicate, if not altogether eliminate opportunities for, completion of HPV vaccine schedules as illustrated by the current COVID-19 pandemic.

The 12-month interval could help to streamline series completion. COVID-19 school closures eliminated opportunities to complete the 0,6 month schedule for students in the single target grade for the 2019/2020 school year. Adoption of a 0,12 month schedule allows for two doses to be administered across two consecutive school years. Simultaneous administration of the single HPV vaccine dose to students initiating HPV vaccination in the upcoming school year as well as to those whose schedules were incomplete for the current school year could allow smooth completion of regimen. Further during the academic year, the removal of the 6-month wait between doses that is imposed by a 0,6 month schedule, gives greater flexibility for arranging additional clinics to catch up on missed appointments. In settings where health facility-based immunization is done, a single visit in a year offers convenience and should still be more advantageous than two visits. In implementing a 0,12 month schedule, it will be important to protect against continued extension of the two-dose interval, since data to rigorously evaluate the impact of such longer intervals on vaccine performance is lacking. Considering that HPV vaccines are most effective when administered prior to HPV exposure,³⁸ each jurisdiction should utilize knowledge of its median age of onset of sexual contact, when selecting the consecutive target grades for implementing a 0,12 month schedule for its school-based HPV vaccination program.

Currently, disruptions in the HPV vaccine supply chain is a looming concern that could prevail until 2024.³⁹ This global shortage has led to suggestions for the two-dose interval to be extended to 3–5 years to significantly alleviate current shortfall in HPV vaccine supplies.⁴⁰ The non-inferiority in HPV16 and 18 antibody response observed for the interval of 36–96 months relative to 6 months is encouraging as it suggests that delaying the second dose for up to 96 months could be as beneficial for cervical cancer prevention as delivery of two doses within six months. Consequently, it supports recommendations that routine HPV immunization programs focus on and prioritize the delivery of at least one dose to all eligible children during the COVID-19 pandemic, with the second dose to be delivered at an opportune time.⁴¹ Further well powered studies with comparison to adult women on a 3-dose schedule are however needed to confirm the immunogenicity of the extended interval of 36–96 months. Notably, such long intervals present increased risks for initiation of sexual contact before series completion and will require prudent planning and implementation of strategies to minimize this risk.

This systematic review is the most comprehensive synthesis to evaluate the performance of two-dose intervals of 6 months to intervals longer than 6 months. It included a range of study designs, in contrast to Bergman et al.¹⁵ that included only RCTs in assessing intervals of 6 months with longer intervals. Other strengths of this review are the use of time post last dose for true time-matched comparison of schedules and the comparison of relative measures between extended and standard intervals within studies, to minimize the influence of between-study methodological variations. This study was however limited by the paucity of available data on the effect of intervals longer than 12 months on vaccine efficacy and effectiveness. There is a clear need for high quality empirical evidence to assess the impact of extended interval for two-dose schedules on the performance of HPV vaccines in preventing HPV infections and related outcomes.

To conclude, we found that a 12-month interval is non-inferior to a 6-month interval for two-dose HPV vaccine schedules with regard to immunogenicity for all HPV types. Consequently, HPV immunization programs could safely adopt a 12-month interval for more flexibility. An interval extending between 36 and 96 months demonstrated non-inferiority to a 6-month interval with regard to immunogenicity for HPV6, 16 and 18 but not HPV11. This is a promising finding that needs to be further explored in light of recommendations promulgating the use of 3–5 year intervals for HPV vaccine administration during the HPV vaccine shortage. Further studies are needed to confirm the durability of the protection provided by two-dose vaccination schedules with a 12-month interval and the impact of extended intervals on the effectiveness of HPV vaccines. Altogether these are valuable insights for policy and decision makers involved in HPV immunization programs.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors would like to thank Kathryn Hornby, UBC reference librarian for critical review of the search strategy.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website at https://doi.org/10.1080/21645515.2021.1926182.

Additional information

Funding

RD has received salary support from the Canadian Immunization Research Network and the Michael Smith Foundation for Health Research; Canadian Immunization Research Network; Michael Smith Foundation for Health Research Scholar Award.