Table 2.

Studies that evaluated HPV vaccine effectiveness by number of doses: analyses and main findings

Endpoint/vaccine/authors	Study population age (years) at vaccination and outcome		Buffer^a (months)	Sensitivity analyses by age group/ buffer/dose interval^b	Comparison with unvaccinated		Formal comparison between doses

					Effect (95% CI)	Comments

Vaccine-type HPV infection

Quadrivalent vaccine

Chandler 2018	≤26^c	14–26	0	No/No/No	No significant effectiveness for ≥1 dose		1 vs 3 doses OR = 0.99 (0.33–2.96) 2 vs 3 doses OR = 0.60 (0.17–2.12)

Widdice 2019	16.2 wave 1 (mean) 15.1 wave 2 (mean)	13–26	0	Yes/No/No	No significant effectiveness for ≥1 dose	• Similar results for the analysis restricted to men vaccinated at ≥15 years and men vaccinated before sexual initiation, and men vaccinated after sexual initiation	Number of doses (0,1,2,3) not associated with ≥1 vaccine-type HPV or HPV16 and/or 18

Sonawane 2019	≤26^c	18–26	0	No/No/No	Difference in predicted probability 3: aPD = −4.3 (−4.6, −4.0) 2: aPD = −1.7 (−2.4, −0.1) 1: aPD = −5.0 (−5.6, −4.5)		1 vs 3 doses p-value = 0.70 2 vs 3 doses p-value = 0.40 1 vs 2 doses p-value = 0.12

Markowitz 2020	≤29^c	20–29	1	Yes/No/No	Overall results 3: aPR = 0.17 (0.11–0.26) 2: aPR = 0.15 (0.05–0.47) 1: aPR = 0.25 (0.10–0.62) Results for those with first dose at age ≤18 years 3: aPR = 0.08 (0.04–0.15) 2: aPR = 0.07 (0.01–0.47) 1: aPR = 0.08 (0.01–0.54)	• Similar results for unadjusted analyses and controlling for race/ethnicity and age at screening	3 vs 1 dose PR = 1.06 (0.14–8.09) 3 vs 2 doses PR = 1.17 (0.15–8.96) 2 vs 1 dose PR = 0.90 (0.06–14.36)

Batmunkh 2020	11–17	16–26	0	No/No/No	1: aPR = 0.08 (0.01–0.56)	• Adjusted for income and employment status	No

Bivalent vaccine

Kavanagh 2014	15–17	20–21	0	Yes/No/No	3: aOR = 0.43 (0.34–0.55) 2: aOR = 0.68 (0.42–1.12) 1: aOR = 0.95 (0.51–1.76)	• Differences by number of doses still observed when stratified by age at vaccination	No

Cuschieri 2016	15–17	20–21	0	No/No/No	3: aOR = 0.27 (0.20–0.36) 2: aOR = 0.45 (0.29–0.69) 1: aOR = 0.52 (0.31–0.83)		No

Kavanagh 2017	12->18	20–21	0	Yes/No/No	3: aOR = 0.40 (0.33–0.48) 2: aOR = 0.75 (0.57–0.99) 1: aOR = 0.89 (0.63–1.25)	• When stratified by age at first dose, 3-dose effectivness was highest in the youngest group and lower with age, but all were significant (range: 28.9%–89.1%)	No

Hoes 2021	12–13	14–17	0	No/No/No	2: aHR = 0.16 (0.035–0.73)	• Study conducted when routine 2-dose vaccination program recommended	No

Anogenital warts

Quadrivalent vaccine

Herweijer 2014	10–19	10–24	3	Yes/Yes/No	3: aIRR = 0.20 (0.17–0.23) 2: aIRR = 0.32 (0.26–0.40) 1: aIRR = 0.54 (0.43–0.68)	• Similar results for age groups 10–16 and 17–19 • Similar results for buffers of 0–12 months, except effectiveness for 1 dose was not significant among those vaccinated at 17–19 using buffers of 0 and 1 month(s)	3 vs 1 dose aIRR = 0.37 (0.28–0.48) 3 vs 2 doses aIRR = 0.63 (0.48–0.82) 2 vs 1 dose aIRR = 0.59 (0.43–0.81) • With buffer periods >4 months, no significant difference between 3 and 2 doses

Blomberg 2015	12–27	12–27	1	Yes/No/Yes	1: IRR = 0.51 (0.46–0.56)		3 vs 2 doses IRR = 0.46 (0.39–0.54) 2 vs 1 doses IRR = 0.44 (0.37–0.51) • With dose interval >4 months, no significant difference for 3 vs 2 doses • Similar results when stratified by age at vaccination

Dominiak-Felden 2015	10–21	16–23	1	Yes/No/No	3: aIRR = 0.12 (0.07– 0.21) 2: aIRR = 0.34 (0.14–0.83) 1: aIRR = 0.63 (0.35–1.16)	• 3 dose effectiveness estimates were higher for those vaccinated at age <15 and 15–17 years than ≥18 years • 3 dose effectiveness estimates higher with buffers >1 year	No

Perkins 2017	9–25	9–25	0	No/Yes/Yes	3: aIRR = 0.53 (0.46–0.60)		3 vs 1 doses aIRR = 0.82 (0.71–0.95) 3 vs 2 doses aIRR = 0.89 (0.78–1.03) • With 1 yr buffer period, no change in findings (data not shown) • Similar results with interval >5 months for 2 doses

Navarro-Illana 2017	14	14–19	0	No/No/No	3: aRR = 0.24 (0.15–0.34) 2: aRR = 0.36 (0.14–0.68) 1: aRR = 0.39 (0.13–0.80)		No

Lamb 2017	10–19	10–27	0	Yes/No/Yes	No analyses of 3, 2 or 1 doses compared to 0		• Higher effectiveness of 3 vs 2 doses, when 1^st and 2^nd doses administered 0–3 or >8 months apart but not 4–7 months • Similar results stratified by age at vaccination

Hariri 2018	16–17 (mean)	11–28	6 from last dose 12 from first dose	No/Yes/Yes	6 month buffer from last dose 3: aHR = 0.23 (0.17–0.31) 2^d: aHR = 0.32 (0.17–0.59) 1: aHR = 0.81 (0.60–1.08) 12 month buffer from first dose 3: aHR = 0.20 (0.15–0.27) 2^d: aHR = 0.24 (0.13– 0.44) 1: aHR = 0.32 (0.20–0.52)		6 month buffer from last dose 3 vs 1 dose aHR = 0.29 (0.20–0.42) 3 vs 2^d doses aHR = 0.74 (0.38–1.43) 2^d vs 1 dose aHR = 0.39 (0.20–0.76) 12 month buffer from first dose 3 vs 1 dose aHR = 0.63 (0.37–1.09) 2 vs 1 dose aHR = 0.74 (0.35–1.60)

Zeybek 2018	9–26	10–31	3 from last dose	Yes/No/Yes	Results for 15–19 yr-olds 3: aRR = 0.58 (0.49–0.70) 2: aRR = 0.67 (0.51–0.89) 1: aRR = 0.65 (0.49–0.85)	• No significant effect in older or younger age groups • Similar results with 2-dose interval <6 or ≥6 months	• No significant differences for 3 vs 1, 3 vs 2, or 2 vs 1 doses (only p-values reported)

Willows 2018	9–26	10–33	0	Yes/No/No	Results for those vaccinated at age 9–18 years 3: aHR = 0.4 (0.3–0.7) 2: aHR = 1.4 (0.6–3.3) 1: aHR = 0.6 (0.2–1.8)	• No significant effect in those vaccinated at older ages	No

Baandrup 2021	12–30	12–30	1	Yes/No/No	Results for first dose at age 12–14 years 3: aIRR = 0.16 (0.15–0.18) 2: aIRR = 0.22 (0.18–0.26) 1: aIRR = 0.29 (0.22–0.38)	• Results presented for 4 age at vaccination groups; in oldest age, ≥19 years, significant effect only with 3 doses • Significant but lower effectiveness for 3, 2 and 1 doses in 15–16 and 17–18 age groups than 12–14 years	Results for first dose at age 12–14 years 3 vs 1 dose: aIRR = 0.56 (0.43–0.73) 2 vs 1 dose: aIRR = 0.76 (0.56–1.03)

Cervical abnormalities ^e

Quadrivalent vaccine

Gertig 2013	12–19	12–21	0	No/No/No	Outcome summarized: CIN2+ 3: aHR = 0.61 (0.48–0.78) 2: aHR = 1.02 (0.68–1.53) 1: aHR = 1.47 (0.97–2.23) Outcome summarized: CIN3/AIS 3: aHR = 0.53 (0.36–0 .77) 2: aHR = 0.87 (0.46–1.67) 1: aHR = 1.40 (0.75–2.61)	• Similar results for CIN2 as an outcome	No

Crowe 2014	12–26	11–31	0	Yes/Yes/No	Outcome summarized: high grade histological lesions 3: aOR = 0.54 (0.43–0.67) 2: aOR = 0.79 (0.64–0.98) 1: aOR = 0.95 (0.77–1.16)	• Buffer periods from 1 to 12 months - no consistent impact on estimates • Similar results among those vaccinated at ages 15–18 and 19–22 years	No

Brotherton 2015	12–26	12–30	0	Yes/Yes/Yes	Results for those vaccinated prior to screening Outcome summarized: CIN2+ 3: aHR = 0.71 (0.64–0.80) 2: aHR = 1.21 (1.02–1.44) 1: aHR = 1.19 (0.99–1.43) Outcome summarized: CIN3/AIS 3: aHR = 0.69 (0.58– 0.81) 2: aHR = 1.17 (0.92–1.48) 1: aHR = 1.41 (1.12–1.77)	• Similar results for those vaccinated after screening, stratified by age at vaccination • With longer buffer periods, some effectiveness for 2 and 1 doses • No difference by interval between 2 doses	No

Hofstetter 2016	11–20	11–27	1	Yes/No/Yes	Outcome summarized: any abnormal cytology 3: aHR = 0.58 (0.48–0.69) 2: aHR = 0.81 (0.66–0.99) 1: aHR = 1.05 (0.88–1.26)	• Similar results when stratified by age at vaccination, although 2 doses not always significant • Highest effectiveness, although only significant for 3 doses, for those vaccinated at ages 11–14 years	No

Kim 2016	10–15	18–21	0	No/No/No	Outcome summarized: high grade cytology 3: aOR = 0.48 (0.28–0.81) 2: aOR = 0.17 (0.02–1.20) 1: aOR = 0.45 (0.11–1.83)		No

Silverberg 2018	14–26	18–34	6	Yes/No/No	Outcome summarized: CIN2+/AIS 3: aRR = 0.78 (0.66–0.91) 2: aRR = 1.02 (0.82–1.28) 1: aRR = 0.89 (0.73–1.09) Outcome summarized: CIN3+/AIS 3: aRR = 0.64 (0.48–0.84) 2: aRR = 0.97 (0.67–1.41) 1: aRR = 0.90 (0.65–1.24)	• Highest 3 dose effectiveness among those vaccinated at youngest ages	No

Dehlendorff 2018	13–29	13–30	0	Yes/No/Yes	Outcome summarized: CIN2+/AIS (age ≤16 years) 3: aIRR = 0.23 (0.11–0.49) 2: aIRR = 0.44 (0.10–2.03) 1: aIRR = 0.23 (0.01–5.24)	• Similar results for age at vaccination 1719 years	2 vs 3 doses aIRR = 1.60 (1.05–2.24) • No significant difference between 2 and 3 doses when interval between dose 1 and 2 >5 months and age at vaccination <20 years

Brotherton 2019	≤13–22	15–22	0	Yes/No/Yes	Outcome summarized: CIN2+ 3: aHR = 0.59 (0.54–0.65) 2: aHR = 0.61 (0.52–0.72) 1: aHR = 0.65 (0.52–0.81) Outcome summarized: CIN3+/AIS 3: aHR = 0.43 (0.35–0.53) 2: aHR = 0.42 (0.27–0.64) 1: aHR = 0.66 (0.41–1.06)	• Similar results for time-varying dose status, CIN3+, buffers of 0 and 12 months, and alternate status based on timing between 1 and 2 doses	Outcome summarized: CIN2+ 3 vs 1 doses aHR = 0.91 (0.74–1.13) 3 vs 2 doses aHR = 0.97 (0.83–1.14) 2 vs 1 doses aHR = 0.94 (0.73–1.21) Outcome summarized: CIN3+/AIS 3 vs 1 doses aHR = 0.66 (0.41–1.05) 3 vs 2 doses aHR = 1.04 (0.68–1.57) 2 vs 1 doses aHR = 0.64 (0.35–1.16)

Verdoodt 2020	12–16	17–25	0 (6 months for compari son betwee n doses)	Yes/No/No	Outcome summarized: CIN2+/AIS 3: aIRR = 0.43 (0.36–0.51) 2: aIRR = 0.49 (0.32–0.76) 1: aIRR = 0.34 (0.13–0.87) Outcome summarized: CIN3+/AIS 3: aIRR = 0.37 (0.30–0.45) 2: aIRR = 0.38 (0.22–0.66) 1: aIRR = 0.38 (0.14–0.98)	• Similar results by age at vaccination, but only significant for <23 year-olds	Outcome summarized: CIN2+/AIS 3 vs 1 dose aIRR = 0.99 (0.64–1.53) 2 vs 1 doses aIRR = 1.00 (0.61–1.64) Outcome summarized: CIN3+/AIS 3 vs 1 dose aIRR = 0.95 (0.60–1.51) 2 vs 1 doses aIRR = 0.89 (0.53–1.52)

Johnson Gargano 2020	12–26	18–39	24	Yes/Yes/No	Outcome summarized: CIN2+/AIS 3: aOR = 0.26 (0.20–0.35) 2: aOR = 0.45 (0.30–0.69) 1: aOR = 0.53 (0.37–0.76)	• aORs were slightly higher using 1 and 12 month buffer periods and lower using a 36 month buffer period, but all showed significant effectiveness • Effectiveness was higher in earlier birth cohort and lower in later birth cohort	3 vs 1 dose aOR = 0.61 (0.38–0.99) 3 vs 2 dose: aOR = 0.64 (0.39–1.05) 2 vs 1 dose aOR = 0.96 (0.55–1.68)

Rodriguez 2020	9–26	9–31	12	Yes/No/Yes	Outcome summarized: CIN2/3 Age at first dose 15–19 years 3: aHR = 0.66 (0.55–0.80) 2: aHR = 0.72 (0.54–0.95) 1: aHR = 0.64 (0.47–0.88)	• Study underpowered for <15 year-olds • No vaccine effectiveness against high-grade cytology or against CIN2/3 for those who received first dose at age >20 years	No

Innes 2020	14–21	20–24	0	Yes/No/No	Outcome summarized: high-grade histology (≥1 dose <18 years) 3: IRR = 0.66 (0.60–0.72) 2: IRR = 0.81 (0.63–1.03) 1: IRR = 1.10 (0.85–1.45)	• No significant effectiveness against high-grade histology for ≥1 dose among women vaccinated at ≥18 years	No

Bivalent vaccine

Pollock 2014	15->18	20–21	0	No/No/No	Outcome summarized: CIN3 3: aOR = 0.45 (0.35–0.58) 2: aOR = 0.77 (0.49–1.21) 1: aOR = 1.42 (0.89–2.28)		No

Cameron 2017	14->18	20–21	0	No/No/No	Outcome summarized: CIN3 Significant effect only with 3 doses	• Vaccinated in each deprivation category, compared with unvaccinated in most deprived	No

Palmer 2019	12–>18	20	0	Yes/No/No	Outcome summarized: CIN3+ 2: aOR = 0.77 (0.48–1.24) 1: aOR = 1.19 (0.70–2.05)	• Effect of 3-dose vaccination larger with younger age at vaccination ranging from 0.14 to 0.85	No

Acuti Martellucci 2021 ^f	14–>30	17–32	1,6,12	Yes/Yes/No	Outcome summarized: Any abnormal cytology, youngest birth cohort (1990–1993), 1-month buffer duration 3: aOR = 0.44 (0.14–1.43) 2: aOR = 0.65 (0.20–2.16) 1: aOR = 0.43 (0.17–1.05)	• Sensitivity analyses also for vaccine type (both bivalent and quadrivalent used), high and low grade cytology, and buffer duration	No

Abbreviations: IRR, incidence rate ratio; aIRR, adjusted incidence rate ratio; RR, relative risk; aRR, adjusted RR; aOR, adjusted odds ratio; HR, hazard ratio; CI, confidence intervals; CIN3, cervical intraepithelial neoplasia grade 3; CIN2+, CIN grade 2 or worse; AIS, adenocarcinoma in situ; Significant, 95% CI does not include 1.

Buffer period is the lag time between vaccination and counting of outcomes. This column shows buffer period in main analysis.

Interval between doses for 2-dose vaccine recipients.

Not explicitly stated in paper.

Data presented for 2 doses are those with an interval ≥6 months between doses.

Some articles presented several outcomes for cervical cytological or histological abnormalities. In this table, we summarize results for the outcome most proximal to cervical cancer.

Either bivalent or quadrivalent vaccine.