Efficacy, effectiveness and immunogenicity of reduced HPV vaccination schedules
Published by: The Public Health Agency of Canada
Issue: CCDR Volume 50–6, June 2024: Cancer Vaccines
Date published: June 2024
ISSN: 1481-8531
Submit a manuscript
About CCDR
Browse
Volume 50-6, June 2024: Cancer Vaccines
Overview
Efficacy, effectiveness and immunogenicity of reduced HPV vaccination schedules: A review of available evidence
Joshua Montroy1, Marina I Salvadori1, Nicole Forbes1, Vinita Dubey2, Sarah Almasri1, Anna Jirovec1, Cathy Yan1, Katarina Gusic1, Adrienne Stevens1, Kelsey Young1, Matthew Tunis1
Affiliations
1 Centre for Immunization Programs, Public Health Agency of Canada, Ottawa, ON
2 Toronto Public Health and University of Toronto Dalla Lana School of Public Health, Toronto, ON
Correspondence
Suggested citation
Montroy J, Salvadori MI, Forbes N, Dubey V, Almasri S, Jirovec A, Yan C, Gusic K, Stevens A, Young K, Tunis M. Efficacy, effectiveness and immunogenicity of reduced HPV vaccination schedules: A review of available evidence. Can Commun Dis Rep 2024;50(6):166–78. https://doi.org/10.14745/ccdr.v50i06a01
Keywords: HPV, vaccination, dose-reduction, dosing schedule, effectiveness, cancer, evidence review
Abstract
Background: Current National Advisory Committee on Immunization (NACI) guidance recommends human papillomavirus (HPV) vaccines be administered as a two or three-dose schedule. Recently, several large clinical trials have reported the clinical benefit of a single HPV vaccine dose. As a result, the World Health Organization released updated guidance on HPV vaccines in 2022, recommending a two-dose schedule for individuals aged 9–20 years, and acknowledging the use of an alternative off-label single dose schedule.
Objective: The objective of this overview is to provide a detailed account of the available evidence comparing HPV vaccination schedules, which was considered by NACI when updating recommendations on HPV vaccines.
Methods: To identify relevant evidence, existing systematic reviews were leveraged where possible. Individual studies were critically appraised, and the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology was used to assess the certainty of evidence.
Results: Available evidence suggests that a one, two, or three-dose HPV vaccine schedule may provide similar protection from HPV infection. While antibody levels against HPV vaccine types were statistically significantly lower with a single dose schedule compared to two or three doses, titres were sustained for up to 16 years. The clinical significance of lower antibody titres is unknown, as there is no established immunologic correlate of protection.
Conclusion: While the available evidence on single-dose HPV vaccination schedules shows a one-dose schedule is highly effective, continued follow-up of single-dose cohorts will be critical to understanding the relative duration of protection for reduced dose schedules and informing future NACI guidance on HPV vaccines.
Introduction
Human papillomavirus (HPV) infections are the causative agent of several cancers, including virtually all cervical cancers, other anogenital cancers, as well as head and neck cancers and anogenital warts (AGW) Footnote 1Footnote 2. HPV vaccines were first authorized in 2006 and have been shown to be highly effective Footnote 3Footnote 4. In Canada, a two or three-dose schedule is recommended for healthy individuals aged 9–14 years, and a three-dose schedule is recommended for healthy individuals aged 15 years and over, and for immunocompromised individuals Footnote 5. Recently, the World Health Organization (WHO) released an updated position paper on HPV vaccination schedules, detailing that while a two-dose schedule is recommended for those over 9 years of age, an alternative off-label, single-dose schedule can be used in those aged 9–20 years Footnote 6. Several other jurisdictions, such as the United Kingdom, have since updated their HPV vaccination recommendations to include a single-dose schedule Footnote 7Footnote 8Footnote 9. This updated guidance was based on several factors, including emerging evidence indicating that a single dose of HPV vaccine provides similar levels of protection from HPV infection as multi-dose schedules Footnote 10.
Canadian provinces and territories have asked that the National Advisory Committee on Immunization (NACI) review the currently available evidence and potentially provide updated guidance on reduced HPV immunization schedules. The Public Health Agency of Canada (PHAC) has prepared this overview to review the available clinical evidence on reduced HPV vaccination schedules (with a focus on single-dose schedules), with an objective to help inform NACI evidence-informed recommendations and decision-making for vaccine programs in Canada.
Methods
Table 1 outlines eligibility criteria for studies included in this analysis. To identify relevant studies, an update of a 2022 systematic review Footnote 10 performed by Cochrane Response in collaboration with the Strategic Advisory Group of Experts on Immunization (SAGE) (which itself was a modified update of a previous Cochrane Response review Footnote 11) was performed. The updated literature search allowed for identification of any additional studies published since 2022 or any available updated data from included studies (e.g., both recent publications and proceedings from international conferences).
Criteria | Eligibility (one vs. two/three doses) |
Eligibility (two vs. three doses) |
---|---|---|
Population |
Individuals ≥9 years of age |
|
Intervention |
One dose of GARDASIL®9 or CERVARIX®. Considering limitations to evidence (e.g., limited follow-up time) on GARDASIL®9, indirect evidence from studies using GARDASIL®4 was also considered. |
Two doses of GARDASIL®9 or CERVARIX®. Considering limitations to evidence (e.g., limited follow-up time) on GARDASIL®9, indirect evidence from studies using GARDASIL®4 was also considered. |
Comparator |
Two or three doses of GARDASIL®9 or CERVARIX®(with the interval between the first and last dose in the series being at least six months). Considering limitations to evidence (e.g., limited follow-up time) on GARDASIL®9, indirect evidence from studies using GARDASIL®4 was also considered. Note: While not directly comparing the clinical benefit of HPV vaccines by the number of doses, studies evaluating the immunogenicity or vaccine efficacy/effectiveness of a one-dose HPV vaccine schedule compared to no HPV vaccine were also included. |
Three doses of GARDASIL®9 or CERVARIX®. Considering limitations to evidence (e.g., limited follow-up time) on GARDASIL®9, indirect evidence from studies using GARDASIL®4 was also considered. |
Outcomes |
Outcomes rated as critical for decision-making (deemed equally critical):
Outcomes rated as important for decision-making (deemed equally important):
|
|
Study design |
Randomized controlled trials, non-randomized trials, and observational studies. Observational studies assessed to be at a serious or critical risk of bias were excluded. |
|
For analyses comparing a single dose to zero, two, or three doses, the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology Footnote 12 was used to assess available evidence considered by NACI during guidance development. Following critical appraisal of individual studies, summary tables with ratings of the certainty of evidence using the GRADE methodology were prepared. For analyses comparing a two-dose to a three-dose HPV vaccine schedule, a methodology informed by A MeaSurement Tool to Assess systematic Reviews (AMSTAR 2) Footnote 13 was used to assess available evidence considered by NACI during guidance development. Detailed information regarding the methodology used in the update of this review can be found elsewhere.
Results
Efficacy/effectiveness against HPV infection
A GRADE assessment of the available randomized controlled trial (RCT) evidence concluded that a one-dose HPV vaccine schedule resulted in a large reduction in persistent infection compared to no vaccine (high certainty of evidence; Table 2). Currently, the KENya Single-dose HPV-vaccine Efficacy (KEN SHE) trial represents the sole RCT evidence demonstrating the efficacy of a single-dose schedule Footnote 14. This trial randomized women aged 15–20 years (n=2,275) to one dose of either GARDASIL®9, CERVARIX®, or meningococcal vaccine. After three years of follow-up, vaccine effectiveness (VE) against persistent HPV16/18 infection was 97.5% (95% CI: 90.0%–99.4%) and 98.8% (95% CI: 91.3%–99.8%) for GARDASIL®9 and CERVARIX®, respectively. Similar results were seen in the non-RCT evidence, with a single dose probably resulting in reductions in persistent Footnote 15Footnote 16, incident Footnote 16Footnote 17, and prevalent Footnote 17Footnote 18 HPV infections compared to no vaccination (moderate certainty of evidence; Table 2, Figure 1).
Number of studies | Study design | Number of events/number of participants | Effect | Certainty of evidence | Comments | ||
---|---|---|---|---|---|---|---|
Zero doses | One dose | Relative effect (95% CI) |
Absolute effect (95% CI) |
||||
Persistent HPV infection with vaccine types (follow-up ranging from 3–10 years) | |||||||
1 Footnote 14 | RCTTable 2 footnote a | 72/757 (9.5%) |
3/1,518 (0.2%) |
RR 0.02 (0.01–0.07) |
94 fewer per 1,000 (94 fewer to 88 fewer) | High | A single dose of HPV vaccine results in a large reduction in persistent HPV infections compared to no vaccine |
2 Footnote 15Footnote 16 | Post-hoc RCT analysis | A small number of events in the intervention groups across studies (n=292–2,135); high VE was estimated in each studyTable 2 footnote b | ModerateTable 2 footnote c | A single dose of HPV vaccine probably results in a large reduction in persistent HPV infections compared to no vaccine | |||
Prevalent HPV infection with vaccine types (follow-up ranging from 6–11 years) | |||||||
2 Footnote 17Footnote 18 | 1 post-hoc RCT analysis, 1 observational study | A small number of events in the intervention groups across studies (n=87–221); large reductions in infection prevalence associated with a single dose in each studyTable 2 footnote d | ModerateTable 2 footnote c | A single dose of HPV vaccine probably results in reduction in prevalent HPV infections compared to no vaccine | |||
Incident HPV infection with vaccine types (follow-up ranging from 10–11 years) | |||||||
2 Footnote 16Footnote 17 | Post-hoc RCT analysis | Number of events dissimilar between studies (n=112–2,858); however, similar reductions in risk compared to unvaccinated were observedTable 2 footnote e | ModerateTable 2 footnote c | A single dose of HPV vaccine probably results in reduction in incident HPV infections compared to no vaccine | |||
Antibody titres (follow-up ranging from 4–10 years) | |||||||
3 Footnote 18Footnote 19Footnote 20 | Observational | Varying number of participants in each study (n=30–324), with differing lengths of follow-up and magnitudes of effect across studies; however, the direction of effect was consistent across studiesTable 2 footnote f | High | A single dose of HPV vaccine results in an increased immune response compared to no vaccine | |||
Anogenital warts (follow-up of approximately 2.5 years) | |||||||
1 Footnote 21 | Observational | 523/52,779 (1.0%) |
69/9,898 (0.7%) |
aHRTable 2 footnote g 0.32 (0.20–0.52) |
7 fewer per 1,000 (8 fewer to 5 fewer) | ModerateTable 2 footnote h | A single dose of HPV vaccine probably reduces the risk of anogenital warts compared to no vaccine |
Juvenile-onset recurrent respiratory papillomatosis (JoRPP) | |||||||
N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Figure 1: Risk ratios and 95% CI for persistent, prevalent, and incident HPV vaccine-type infections, one dose compared to no dosesFigure 1 footnote aFigure 1 footnote bFigure 1 footnote cFigure 1 footnote dFigure 1 footnote eFigure 1 footnote fFigure 1 footnote gFigure 1 footnote hFigure 1 footnote iFigure 1 footnote j
Figure 1 - Text description
The studies are stratified in a table by both the type of infection (persistent, prevalent, or incident infection) and the study design (i.e., RCT or non-RCT).
Forest plot depicts risk ratio (RR) and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the risk of HPV infection as a graphic depicting the risk ratio in a forest plot and a graphic depicting the risk of bias for each study.
Study name | Events/total 1 dose |
Events/total 0 doses |
Risk ratio | Lower limit | Upper limit |
---|---|---|---|---|---|
Persistent infections, RCT evidence | |||||
Barnabas et al.Figure 1 footnote aFigure 1 footnote b | 3/1,518 | 72/757 | 0.02 | 0.01 | 0.07 |
Persistent infections, non-RCT evidence | |||||
Basu et al.Figure 1 footnote cFigure 1 footnote d | 2/2,135 | 35/1,265 | 0.03 | 0.01 | 0.14 |
Kreimer et al.Figure 1 footnote eFigure 1 footnote f | 1/292 | 17 249 | 0.05 | 0.01 | 0.37 |
Incident infections, non-RCT evidence | |||||
Kreimer et al.Figure 1 footnote g | 2/112 | 69/1,783 | 0.46 | 0.11 | 1.86 |
Basu et al.Figure 1 footnote h | 154/2,858 | 192/1,479 | 0.42 | 0.34 | 0.51 |
Prevalent infections, non-RCT evidence | |||||
Kreimer et al.Figure 1 footnote i | 2/112 | 178/1,783 | 0.18 | 0.04 | 0.71 |
Batmunkh et al.Figure 1 footnote j | 1/87 | 41/266 | 0.07 | 0.01 | 0.53 |
For Barnabas et al., risk of bias was assessed as moderate due to missing outcome data. For Basu et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results. For Kreimer et al. (2015), risk of bias was assessed as moderate due to confounding and missing data. For Kreimer et al. (2020), risk of bias was assessed as moderate due to confounding and the selection of the reported results. For Batmunkh et al., risk of bias was assessed as moderate due to confounding, selection of participants into the study, and the selection of the reported results.
A GRADE assessment of the available evidence concluded that, compared to two or three doses, there may be little to no difference in persistent, incident, or prevalent HPV infection risk with a one-dose HPV vaccine schedule (low certainty of evidence, Table 3 and Table 4, Figure 2 and Figure 3). Two RCTs evaluating the effectiveness of a two and/or three-dose HPV vaccine schedule have conducted post-hoc analyses to also estimate the VE of a one-dose schedule, with both studies reporting similar VE across all dosing schedules, up to 10 Footnote 16 or 11 years Footnote 17 (low certainty of evidence; Table 3 and Table 4, Figure 2 and Figure 3).
Number of studies | Study design | Number of events/number of participants | Effect | Certainty of evidence | Comments | ||
---|---|---|---|---|---|---|---|
Two doses | One dose | Relative effect (95% CI) |
Absolute effect (95% CI) |
||||
Persistent HPV infection with vaccine types (follow-up ranging from 4–10 years) | |||||||
2 Footnote 15Footnote 16 | Post-hoc RCT analysis | A small number of events in both the intervention (n=292–2,135) and control arms (n=611–1,452) across studies; high VE estimated for both arms in each studyTable 3 footnote a | LowTable 3 footnote bTable 3 footnote c | A single dose of HPV vaccine may result in little to no difference in persistent HPV infections compared to two doses | |||
Prevalent HPV infection with vaccine types (follow-up of 11 years) | |||||||
1 Footnote 17 | Post-hoc RCT analysis | 1/62 (1.6%) |
2/112 (1.8%) |
RR 1.11 (0.10–11.97) |
2 more per 1,000 (15 fewer to 177 more) | LowTable 3 footnote dTable 3 footnote e | A single dose of HPV vaccine may result in little to no difference in prevalent HPV infections compared to two doses |
Incident HPV infection with vaccine types (follow-up ranging from 10–11 years) | |||||||
2 Footnote 16Footnote 17 | Post-hoc RCT analysis | Number of events dissimilar between studies (n [one dose]=112–2,858; n [two doses]=62–2,166), as the baseline risk of events varies across studies; however, VE estimates for each group are similar across studiesTable 3 footnote f | LowTable 3 footnote dTable 3 footnote e | A single dose of HPV vaccine may result in little to no difference in incident HPV infections compared to two doses | |||
Antibody titres (follow-up ranging from 2–16 years) | |||||||
1 Footnote 22 | RCT | 310 | 310 | Ratio of GMTs ranging from 0.11 (0.09–0.14) to 0.21 (0.16–0.26) | N/A | High | A single dose of HPV vaccine results in a decreased immune response compared to two doses |
2 Footnote 21Footnote 23 | Post-hoc RCT analysis | Dissimilar number of participants between intervention and control arms, across all studies; however, consistent magnitude and direction of effect across studies | High | A single dose of HPV vaccine results in a decreased immune response compared to two doses | |||
Histological and cytological abnormalities (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 1/1,128 (0.9%) |
4/1,511 (2.6%) |
RR 2.99 (0.33–26.80) |
2 more per 1,000 (1 fewer to 23 more) | LowTable 3 footnote bTable 3 footnote e | A single dose of HPV vaccine may result in little to no difference in cervical abnormalities compared to two doses |
CIN2+ (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 0/1,128 (0%) |
0/1,511 (0%) |
Not estimable | Not estimable | LowTable 3 footnote bTable 3 footnote g | A single dose of HPV vaccine may result in little to no difference in CIN2+ compared to two doses |
HPV-associated cancer (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 0/1,128 (0%) |
0/1,511 (0%) |
Not estimable | Not estimable | Very lowTable 3 footnote bTable 3 footnote hTable 3 footnote i | Data insufficient to determine association |
Anogenital warts (follow-up of approximately 2.5 years) | |||||||
1 Footnote 21 | Observational | 42/8,046 (0.5%) |
69/9,898 (0.7%) |
aHRTable 3 footnote j 0.74 (0.35–1.60) |
2 more per 1,000 (5 fewer to 4 more) | LowTable 3 footnote eTable 3 footnote k | A single dose of HPV vaccine may result in little to no difference in the risk of anogenital warts compared to two doses |
Juvenile-onset recurrent respiratory papillomatosis (JoRPP) | |||||||
N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Number of studies | Study design | Number of events/number of participants | Effect | Certainty of evidence | Comments | ||
---|---|---|---|---|---|---|---|
Three doses | One dose | Relative effect (95% CI) |
Absolute effect (95% CI) |
||||
Persistent HPV infection with vaccine types (follow-up ranging from 4–10 years) | |||||||
2 Footnote 15Footnote 16 | Post-hoc RCT analysis | A small number of events in both the intervention (n=292–2,135) and control (n=1,460–11,104) arms across studies; high VE estimated for both arms in each studyTable 4 footnote a | LowTable 4 footnote bTable 4 footnote c | A single dose of HPV vaccine may result in little to no difference in persistent HPV infections compared to three doses | |||
Prevalent HPV infection with vaccine types (follow-up of 11 years) | |||||||
1 Footnote 17 | Post-hoc RCT analysis | 27/1,365 (2.0%) |
2/112 (1.8%) |
RR 0.90 (0.22–3.75) |
2 fewer per 1,000 (15 fewer to 54 more) | LowTable 4 footnote dTable 4 footnote e | A single dose of HPV vaccine may result in little to no difference in prevalent HPV infections compared to three doses |
Incident HPV infection with vaccine types (follow-up ranging from 10–11 years) | |||||||
2 Footnote 16Footnote 17 | Post-hoc RCT analysis | Number of events dissimilar between studies (n [one dose]=112–2,858; n [two doses]=1,365–2,019), as the baseline risk of events varies across studiesTable 4 footnote f | LowTable 4 footnote dTable 4 footnote e | A single dose of HPV vaccine may result in little to no difference in incident HPV infections compared to three doses | |||
Antibody titres (follow-up ranging from 2–16 years) | |||||||
1 Footnote 22 | RCT | 310 | 310 | Ratio of GMTs ranging from 0.06 (0.05–0.07) to 0.19 (0.15–0.24) | N/A | High | A single dose of HPV vaccine results in a decreased immune response compared to three doses |
2 Footnote 21Footnote 23 | Post-hoc RCT analyses | Dissimilar number of participants between intervention and control arms, across all studies; however, consistent magnitude and direction of effect across studies | High | A single dose of HPV vaccine results in a decreased immune response compared to three doses | |||
Histological and cytological abnormalities (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 1/1,037 (0.9%) |
4/1,511 (2.6%) |
RR 2.75 (0.31–24.53) | 2 more per 1,000 (1 fewer to 23 more) | LowTable 4 footnote bTable 4 footnote e | A single dose of HPV vaccine may result in little to no difference in cervical abnormalities compared to three doses |
CIN2+ (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 0/1,037 (0%) |
0/1,511 (0%) |
Not estimable | Not estimable | LowTable 4 footnote bTable 4 footnote g | A single dose of HPV vaccine may result in little to no difference in CIN2+ compared to three doses |
HPV-associated cancer (follow-up of 10 years) | |||||||
1 Footnote 16 | Post-hoc RCT analysis | 0/1,037 (0%) |
0/1,511 (0%) |
Not estimable | Not estimable | Very lowTable 4 footnote bTable 4 footnote hTable 4 footnote i | Data insufficient to determine association |
Anogenital warts (follow-up of approximately 2.5 years) | |||||||
1 Footnote 21 | Observational | 91/57,287 (0.2%) |
69/9,898 (0.7%) |
aHRTable 4 footnote j 0.63 (0.37–1.09) |
3 more per 1,000 (1 fewer to 4 more) | LowTable 4 footnote eTable 4 footnote k | A single dose of HPV vaccine may result in little to no difference in the risk of anogenital warts compared to three doses |
Juvenile-onset recurrent respiratory papillomatosis (JoRPP) | |||||||
N/A | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
Figure 2: Risk ratios and 95% CI for persistent, prevalent and incident HPV vaccine-type infections, one dose compared to either two or three dosesFigure 2 footnote aFigure 2 footnote bFigure 2 footnote cFigure 2 footnote dFigure 2 footnote e
Figure 2 - Text description
The studies are stratified in a table by both the type of infection (persistent, prevalent, or incident infection) and the comparison (i.e., 1 vs. 2 doses or 1 vs. 3 doses).
Forest plot depicts risk ratio (RR) and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the risk of HPV infection as a graphic depicting the risk ratio in a forest plot and a graphic depicting the risk of bias for each study.
Study name | Events/total 1 dose |
Events/total 2 or 3 doses |
Risk ratio | Lower limit | Upper limit |
---|---|---|---|---|---|
Persistent infections, 1 vs. 2 doses | |||||
Basu et al.Figure 2 footnote aFigure 2 footnote b | 2/2,135 | 1/1,452 | 1.36 | 0.12 | 14.99 |
Kreimer et al.Figure 2 footnote cFigure 2 footnote d | 1/292 | 3/611 | 0.70 | 0.07 | 6.68 |
Persistent infections, 1 vs. 3 doses | |||||
Basu et al.Figure 2 footnote aFigure 2 footnote b | 2/2,135 | 1/1,460 | 1.37 | 0.12 | 15.07 |
Kreimer et al.Figure 2 footnote cFigure 2 footnote d | 1/292 | 84/11,104 | 0.45 | 0.06 | 3.24 |
Incident infections, 1 vs. 2 doses | |||||
Kreimer et al.Figure 2 footnote e | 2/112 | 1/62 | 1.11 | 0.10 | 11.97 |
Basu et al.Figure 2 footnote b | 154/2,858 | 107/2,166 | 1.09 | 0.86 | 1.39 |
Incident infections, 1 vs. 3 doses | |||||
Kreimer et al.Figure 2 footnote e | 2/112 | 8/1,365 | 3.05 | 0.65 | 14.18 |
Basu et al.Figure 2 footnote b | 154/2,858 | 110/2,019 | 0.99 | 0.78 | 1.25 |
Prevalent infections, 1 vs. 2 doses | |||||
Kreimer et al.Figure 2 footnote e | 2/112 | 1/62 | 1.11 | 0.10 | 11.97 |
Prevalent infections, 1 vs. 3 doses | |||||
Kreimer et al.Figure 2 footnote e | 2/112 | 27/1,365 | 0.90 | 0.22 | 3.75 |
For Basu et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results. For Kreimer et al. (2015), risk of bias was assessed as moderate due to confounding and missing data. For Kreimer et al. (2020), risk of bias was assessed as moderate due to confounding and the selection of the reported results.
Figure 3: Risk ratios and 95% CI for persistent, prevalent and incident HPV infections, two doses compared to three dosesFigure 3 footnote aFigure 3 footnote bFigure 3 footnote cFigure 3 footnote dFigure 3 footnote e
Figure 3 - Text description
The studies are stratified in a table by the type of infection (persistent, prevalent, or incident infection).
Forest plot depicts risk ratio (RR) and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the risk of HPV infection as a graphic depicting the risk ratio in a forest plot and a graphic depicting the risk of bias for each study.
Study name | Events/total 2 doses |
Events/total 3 doses |
Risk ratio | Lower limit | Upper limit |
---|---|---|---|---|---|
Persistent infections | |||||
Basu et al.Figure 3 footnote aFigure 3 footnote b | 1/1,452 | 1/1,460 | 1.01 | 0.06 | 16.06 |
Kreimer et al.Figure 3 footnote cFigure 3 footnote d | 3/611 | 84/11,104 | 0.65 | 0.21 | 2.05 |
Incident infections | |||||
Basu et al.Figure 3 footnote b | 107/2,166 | 110/2,019 | 0.91 | 0.70 | 1.17 |
Kreimer et al.Figure 3 footnote e | 1/62 | 8/1,365 | 2.75 | 0.35 | 21.66 |
Prevalent infections | |||||
Kreimer et al.Figure 3 footnote e | 1/62 | 27/1,365 | 0.82 | 0.11 | 5.90 |
For Basu et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results. For Kreimer et al. (2015), risk of bias was assessed as moderate due to confounding and missing data. For Kreimer et al. (2020), risk of bias was assessed as moderate due to confounding and the selection of the reported results.
The Costa Rica Vaccine Trial (CVT) was originally designed to test the efficacy of a three-dose schedule of CERVARIX®in females aged 18–25 years (compared to control hepatitis A vaccine); however, approximately 20% of participants did not complete their three-dose schedule, primarily due to pregnancy or colposcopy referral, thus creating cohorts who received a one or two-dose schedule. After 11 years of follow-up, VE against prevalent HPV16/18 infection was similar among recipients of either one-dose (82.1%; 95% CI: 40.2%–97.0%), two-dose (83.8%; 95% CI: 19.5%–99.2%) or three-dose (80.2%; 95% CI: 70.7%–87.0%) schedules Footnote 17.
Similar results were also seen in the International Agency for Research on Cancer (IARC) study from India, which was originally designed to compare two and three doses of GARDASIL®in females aged 10–18 years. However, numerous participants did not complete their full vaccine schedule, as recruitment of girls into HPV trials was suspended by the Indian government in 2010. Vaccine effectiveness against persistent HPV16/18 infection was similar among women who received one (95.4%; 95% CI: 85%–99.1%), two (93.1%; 95% CI: 77.3%–99.8%) or three (93.3%; 95% CI: 77.5%–99.7%) doses after 10 years of follow-up Footnote 16.
Efficacy/effectiveness against cervical precancerous lesions
Among included studies, only the IARC trial Footnote 16 reported data on the effect of different HPV vaccine schedules on cervical precancers and HPV-related cancers. After 10 years of follow-up, 16/4,626 (0.3%) of unvaccinated women reported cervical intraepithelial neoplasia (CIN) grade 1, compared to 4/1,511 (0.3%), 1/1,128 (0.1%) and 1/1,037 (0.1%) in the one, two and three-dose groups, respectively. There were no cases of CIN2 or greater in any of the vaccine groups, regardless of the number of doses received, while 5/4,626 women (0.1%) in the unvaccinated group experienced CIN2 or greater. Additionally, there were no cases of HPV-related cancers in any of the groups.
A GRADE assessment of the available evidence concluded that there may be little to no difference in the risks of cervical abnormalities or CIN2+ between one and either two or three-dose schedules (low certainty of evidence; Table 3 and Table 4).
Efficacy/effectiveness against anogenital warts
There is currently no clinical trial evidence comparing the effect of a single dose to either two or three doses on the risk of AGW. However, an observational study from the United States (n=64,517) compared the risk of AGW in female participants who received one dose to those who received no, two, or three doses of GARDASIL®Footnote 21. Propensity score-weighted incidence rates were 761.9 (95% CI: 685.5–849.1), 256.6 (95% CI: 161.8–432.3), 194.2 (95% CI: 108.0–386.4), and 161.8 (95% CI: 124.4–214.6) per 100,000 person-years in the unvaccinated, one, two and three-dose groups, respectively. Propensity score-weighted hazard ratios (HRs) demonstrated no statistically significant difference between the groups, with HRs of 0.74 (95% CI: 0.35–1.60) and 0.63 (95% CI: 0.37–1.09) for two and three doses (compared to one), respectively (no direct comparison of the two and three-dose groups).
A GRADE assessment of the available evidence concluded that a single dose of HPV vaccine probably reduces the risk of AGW compared to no vaccine (moderate certainty of evidence; Table 2), and that there may be little to no difference in risk, compared to a two or three-dose schedule (low certainty of evidence; Table 3 and Table 4).
Antibody titres
A GRADE assessment of the available evidence concluded that a single dose of HPV vaccine results in an increased immune response compared to no vaccine Footnote 18Footnote 19Footnote 20 (high certainty of evidence; Table 2, Figure 4), and a decreased immune response compared to two or three doses (high certainty of evidence; Table 3 and Table 4, Figure 5).
Figure 4: Ratio of geometric mean titres and 95% CI comparing one dose to zero dosesFigure 4 footnote a
Figure 4 - Text description
The studies are presented in a table below.
Forest plot depicts the ratio of GMTs and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the GMTs after HPV vaccination as a graphic depicting the ratio of GMTs in a forest plot and a graphic depicting the risk of bias for each study.
Study name | HPV type | Time point (years)Figure 4 footnote a |
Ratio of GMTs | Lower limit | Upper limit |
---|---|---|---|---|---|
Safaeian et al. | 16 | 4 | 9.36 | 6.40 | 13.68 |
Safaeian et al. | 18 | 4 | 4.79 | 3.37 | 6.80 |
Batmunkh et al. | 16 | 6 | 5.73 | 3.03 | 10.84 |
Batmunkh et al. | 18 | 6 | 2.28 | 1.51 | 3.44 |
Joshi et al. | 16 | 10 | 2.05 | 1.34 | 3.15 |
Joshi et al. | 18 | 10 | 3.49 | 2.80 | 4.35 |
For Safaeian et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results. For Batmunkh et al., risk of bias was assessed as moderate due to confounding, selection of participants into the study, and selection of the reported results. For Joshi et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results.
Figure 5: Ratio of geometric mean titres and 95% CI comparing one dose to either two or three dosesFigure 5 footnote a
Figure 5 - Text description
The studies are stratified in a table by both the study design (i.e., RCT or non-RCT) and the comparison (i.e., 1 vs. 2 doses or 1 vs. 3 doses).
Forest plot depicts the ratio of GMTs and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the GMTs after HPV vaccination as a graphic depicting the ratio of GMTs in a forest plot and a graphic depicting the risk of bias for each study.
Study name | Vaccine | HPV type | Time point (years)Figure 5 footnote a |
Ratio of GMTs | Lower limit | Upper limit |
---|---|---|---|---|---|---|
1 vs. 2 doses, RCT evidence | ||||||
Watson-Jones et al. | Cervarix | 16 | 2 | 0.14 | 0.12 | 0.17 |
Watson-Jones et al. | Gardasil9 | 16 | 2 | 0.11 | 0.09 | 0.14 |
Watson-Jones et al. | Cervarix | 18 | 2 | 0.20 | 0.17 | 0.24 |
Watson-Jones et al. | Gardasil9 | 18 | 2 | 0.21 | 0.16 | 0.26 |
1 vs. 3 doses, RCT evidence | ||||||
Watson-Jones et al. | Cervarix | 16 | 2 | 0.06 | 0.05 | 0.07 |
Watson-Jones et al. | Gardasil9 | 16 | 2 | 0.12 | 0.10 | 0.15 |
Watson-Jones et al. | Cervarix | 18 | 2 | 0.09 | 0.08 | 0.11 |
Watson-Jones et al. | Gardasil9 | 18 | 2 | 0.19 | 0.15 | 0.24 |
1 vs. 2 doses, non-RCT evidence | ||||||
Joshi et al. | Gardasil | 16 | 10 | 0.29 | 0.24 | 0.34 |
Joshi et al. | Gardasil | 18 | 10 | 0.39 | 0.31 | 0.48 |
Romero et al. | Cervarix | 16 | 16 | 0.54 | 0.42 | 0.71 |
Romero et al. | Cervarix | 18 | 16 | 0.57 | 0.43 | 0.75 |
1 vs. 3 doses, RCT evidence | ||||||
Joshi et al. | Gardasil | 16 | 10 | 0.28 | 0.23 | 0.34 |
Joshi et al. | Gardasil | 18 | 10 | 0.32 | 0.25 | 0.40 |
Romero et al. | Cervarix | 16 | 16 | 0.36 | 0.29 | 0.44 |
Romero et al. | Cervarix | 18 | 16 | 0.50 | 0.41 | 0.62 |
For Watson-Jones et al., risk of bias was assessed as low. For Joshi et al., risk of bias was assessed as moderate due to confounding, and selection of the reported results. For Romero et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results.
The Dose Reduction Immunobridging and Safety study (DoRIS) from Tanzania randomized females aged 9–14 years (n=930) to receive one, two, or three doses of CERVARIX®or GARDASIL®9 Footnote 22. Antibody titres were statistically significantly lower for one-dose recipients compared to two or three-dose recipients for both vaccines (Figure 5). However, while lower titres were observed for the one-dose schedule, the antibody response was sustained through year two (end of study). In individuals who received two doses of GARDASIL®9, antibody titres were non-inferior compared to those who received three doses; however, they were significantly lower (and non-inferiority was not met) for those receiving two doses of CERVARIX®(Figure 6).
Figure 6: Ratio of geometric mean titres and 95% CI comparing two doses to three dosesFigure 6 footnote aFigure 6 footnote bFigure 6 footnote cFigure 6 footnote dFigure 6 footnote e
Figure 6 - Text description
The studies are stratified in a table by the study design (i.e., RCT or non-RCT).
Forest plot depicts the ratio of GMTs and 95% CI of individual studies.
This figure consists of a table summarizing the study characteristics for studies reporting on the GMTs after HPV vaccination as a graphic depicting the ratio of GMTs in a forest plot and a graphic depicting the risk of bias for each study.
Study name | Vaccine | HPV type | Time point (years)Figure 6 footnote a |
Ratio of GMTs | Lower limit | Upper limit |
---|---|---|---|---|---|---|
RCTs | ||||||
Watson-Jones et al. | Gardasil9 | 16 | 2 | 1.06 | 0.85 | 1.32 |
Watson-Jones et al. | Gardasil9 | 18 | 2 | 0.91 | 0.71 | 1.16 |
Watson-Jones et al. | Cervarix | 16 | 2 | 0.40 | 0.32 | 0.48 |
Watson-Jones et al. | Cervarix | 18 | 2 | 0.47 | 0.37 | 0.59 |
Bornstein et al.Figure 6 footnote b | Gardasil9 | 16 | 3 | 0.75 | 0.61 | 0.91 |
Bornstein et al.Figure 6 footnote b | Gardasil9 | 18 | 3 | 0.69 | 0.58 | 0.81 |
Bornstein et al.Figure 6 footnote c | Gardasil9 | 16 | 3 | 0.85 | 0.69 | 1.04 |
Bornstein et al.Figure 6 footnote c | Gardasil9 | 18 | 3 | 0.77 | 0.65 | 0.91 |
Bornstein et al.Figure 6 footnote d | Gardasil9 | 16 | 3 | 1.83 | 1.42 | 2.35 |
Bornstein et al.Figure 6 footnote d | Gardasil9 | 18 | 3 | 1.24 | 1.00 | 1.52 |
Bornstein et al.Figure 6 footnote e | Gardasil9 | 16 | 3 | 2.05 | 1.59 | 2.64 |
Bornstein et al.Figure 6 footnote e | Gardasil9 | 18 | 3 | 1.46 | 1.18 | 1.80 |
Leung et al. | Gardasil | 16 | 3 | 0.80 | 0.68 | 0.95 |
Leung et al. | Gardasil | 18 | 3 | 0.60 | 0.49 | 0.73 |
Romanowski et al. | Cervarix | 16 | 5 | 0.53 | 0.42 | 0.65 |
Romanowski et al. | Cervarix | 18 | 5 | 0.75 | 0.59 | 0.95 |
Ogilvie et al. | Gardasil | 16 | 10 | 1.21 | 0.75 | 1.95 |
Ogilvie et al. | Gardasil | 18 | 10 | 0.72 | 0.37 | 1.39 |
Post-hoc RCT analyses | ||||||
Joshi et al. | Gardasil | 16 | 10 | 0.98 | 0.80 | 1.20 |
Joshi et al. | Gardasil | 18 | 10 | 0.82 | 0.63 | 1.05 |
Romero et al. | Cervarix | 16 | 16 | 0.61 | 0.48 | 0.77 |
Romero et al. | Cervarix | 18 | 16 | 0.89 | 0.70 | 1.13 |
For Watson-Jones et al., risk of bias was assessed as low. For Bornstein et al., risk of bias was assessed as low. For Leung et al., risk of bias was assessed as low. For Romanowski et al., risk of bias was assessed as low. For Oglivie et al., risk of bias was assessed as low. For Joshi et al., risk of bias was assessed as moderate due to confounding, and selection of the reported results. For Romero et al., risk of bias was assessed as moderate due to confounding and the selection of the reported results.
Two post-hoc analyses of the CVT and IARC trials (data up to 16 Footnote 23 and 10 Footnote 21, respectively) have produced similar results to the DoRIS study, with a single-dose schedule producing inferior but sustained antibody titres (high certainty of evidence; Table 3 and Table 4, Figure 5).
Several RCTs provide data comparing the antibody titres of a two-dose versus three-dose schedule (Figure 6). The long-term follow-up of a Canadian RCT in girls aged 9–13 years receiving GARDASIL®demonstrated a non-inferior antibody response with two doses for HPV6, HPV11 and HPV16, ten years following vaccination (non-inferiority not met for HPV18) Footnote 24. Another RCT of girls aged 9–14 years receiving GARDASIL®demonstrated a non-inferior immune response for HPV16 and HPV18, three years following vaccination Footnote 25. In a multinational RCT using GARDASIL®9, girls aged 9–14 years were randomized to receive two (six or 12 months apart) or three doses (six months apart), while boys aged 9–14 years were randomized to receive two doses six or 12 months apart. While the individuals receiving two doses six months apart had generally lower/similar antibody levels compared to those receiving three doses, those receiving two doses 12 months apart generally had higher/similar antibody levels compared to those given three doses within six months, three years after vaccination Footnote 26, suggesting the interval between doses may be more important than the number of doses. Lastly, in an RCT of females aged 9–25 years receiving CERVARIX®, HPV16 and HPV18 antibody titres appeared slightly higher after a three-dose schedule compared to a two-dose schedule, regardless of age strata (9–14 and 15–25 years), five years following vaccination. However, no test of non-inferiority was performed Footnote 27.
Discussion
The effectiveness/efficacy and immunogenicity of various HPV vaccine schedules were reviewed. Available evidence suggests that single-dose VE against HPV infection may be similar to that of two or three doses. Antibody titres, however, indicate a lower immune response with a single dose compared to two or three doses. Currently, there is no established correlate of protection for HPV, and therefore the clinical relevance of this decreased immune response is unknown. The interpretation of results from other clinical outcomes, such as the risks of CIN and abnormal cytology, remain challenging due to limitations of the included studies. In addition to the currently available evidence outlined above, which includes follow-up for up to 16 years post-immunization depending on the study and clinical protection outcome, longer follow-up data are expected in the coming years from multiple key studies. As trials continue to accrue data, follow-up will remain important as trial participants reach the age of increased baseline risk of cervical abnormalities and associated cancers, as data for these outcomes is currently limited. Two additional RCTs from Costa Rica are underway and are expected to produce estimates of single dose VE in females 12–16 years and 18–30 years of age by 2025 and 2026, respectively Footnote 28Footnote 29.
Limitations
There are several limitations to the current data. Data are predominately limited to female adolescents and young women, with a primary focus on cervical HPV infection and cervical cancer precursors. However, several additional cancers are attributable to HPV infections (i.e., other anogenital, and head/neck cancers) Footnote 2, for which there are currently no data. While there is no clinical trial data on VE of a single vaccine dose in males, several retrospective observational studies include both biological sexes. However, only two studies report results stratified by sex, with neither study reporting a difference in HPV infection risk between dosing schedules in males Footnote 30Footnote 31. Neither study was eligible for inclusion, however, as both were considered at serious risk of bias. It is possible that different antibody levels or immunologic factors are required for protection in the female versus male genital tract, and for protection against warts and head/neck/anal cancer. Future research on the effect of single dose HPV vaccination and other HPV-related cancers, including trials where clinical outcomes are assessed among male populations, will be important for public health decision-making. Data are also currently lacking on the effect of a one-dose schedule in immunocompromised individuals. Only one observational study that provided data for this group was identified, with no difference in the incidence of abnormal cervical cytology observed between dosing schedules in HIV-positive females. This study was, however, considered at serious risk of bias and therefore not eligible for inclusion Footnote 32.
Conclusion
Current clinical data on reduced HPV dosing schedules are promising. Longer-term follow-up of clinical trial participants, as well as monitoring real-world outcomes in countries where the change to single-dose schedules have already taken place, can help better understand the duration of protection against HPV infection conferred from reduced dosing schedules. In addition, when considering population-level programmatic changes, several additional factors will likely require consideration, including impacts of a program change on acceptability and uptake of the HPV vaccine, as well as on health inequities and access to the vaccine.
Authors' statement
- JM — Conceptualization, methodology, formal analysis, data curation, writing–original draft
- NF — Conceptualization, methodology, formal analysis, data curation, writing–review & editing
- MS — Conceptualization, methodology, data curation, writing–review & editing, supervision
- VD — Conceptualization, methodology, writing–review & editing, supervision
- SA — Conceptualization, methodology, formal analysis, data curation, writing–review & editing
- AJ — Conceptualization, methodology, formal analysis, data curation, writing–review & editing
- CY — Conceptualization, data curation, writing–review & editing
- KG — Conceptualization, data curation, writing–review & editing
- KY — Conceptualization, methodology, writing–review & editing, supervision
- MT — Conceptualization, methodology, writing–review & editing, supervision
Competing interests
None.
Acknowledgements
We would like to acknowledge the contributions of the Public Health Agency of Canada Library Services and the NACI HPV Working Group.
Funding
This work was supported by the Public Health Agency of Canada.
Page details
- Date modified: