Three-Year Follow-up of 2-Dose Versus 3-Dose HPV Vaccine

Protocol V503-010 (NCT01984697) was an international, 3-year safety and immunogenicity study, in which we compared administration of 2-dose regimens of 9vHPV vaccine, separated by either 6 or 12 months, in girls and boys with administration of a 3-dose regimen in a control group of young women (Fig 1). Participants were enrolled from 52 sites in 15 countries (Canada, Chile, Colombia, Czech Republic, Denmark, Israel, Malaysia, Norway, South Africa, South Korea, Spain, Taiwan, Thailand, Turkey, and the United States).

Participants were healthy girls and boys aged 9 to 14 years who were not sexually active before enrollment or young women aged 16 to 26 years with ≤4 lifetime sexual partners and no history of abnormal Papanicolaou test results or other cervical abnormalities, who agreed to use effective contraception through month 7.

The enrollment of girls and boys was stratified into 3 age strata (9–10, 11–12, and 13–14 years) of similar size.

The study was conducted in accordance with principles of Good Clinical Practice and was approved by the appropriate institutional review boards and regulatory agencies; all participants (for minors, the parent or legal guardian and participant) provided written informed consent.

The vaccine was administered as 0.5 mL intramuscular injections at each vaccination visit as a 2-dose regimen administered at 0 and 6 months ([0, 6] regimen), as a 2-dose regimen administered at 0 and 12 months ([0, 12] regimen), or as a 3-dose regimen administered at 0, 2, and 6 months ([0, 2, 6] regimen). Girls were randomly assigned 2:1:2 within each age stratum to the (0, 6) or (0, 12) 2-dose regimen or the 3-dose regimen. Boys were randomly assigned 2:1 within each age stratum to the (0, 6) or (0, 12) 2-dose regimen. Young women were assigned to the 3-dose regimen. The last dose of the assigned regimen was the month 12 vaccination for the (0, 12) regimen and the month 6 vaccination for all other regimens (Fig 1). Participants who received a 2-dose regimen received an additional vaccine dose at month 36 for exploratory immunogenicity analyses (to be reported separately).

HPV types 6, 11, 16, 18, 31, 33, 45, 52, and 58 antibody responses were assessed in serum samples collected prevaccination through month 36 by using a competitive Luminex immunoassay (cLIA).²¹  The original version of the cLIA was used through 2015 for the testing of all samples up to month 13. A new version of the cLIA was used starting in 2016 for the testing of all samples from month 24 onward. The newer version of the assay was bridged to the earlier version to ensure comparable antibody measurements between the 2 versions. Supportive analyses were conducted by using the immunoglobulin G-Luminex immunoassay (IgG-LIA).²²  The results of these assays are reported in milli-Merck units (mMU) per mL. Although the same designation is used for the unit of measurement in both assays, cLIA mMU per mL and IgG-LIA mMU per mL are different units of measurement and cannot be directly compared.

Serious adverse events (SAEs) regardless of causality were reportable from day 1 through 6 months after the last dose of the vaccination regimen, as previously described.¹⁸  Deaths and SAEs judged by the investigator to be related to the 9vHPV vaccine were reported throughout the study.

Immunogenicity was assessed in the per-protocol immunogenicity (PPI) population, consisting of participants who received all planned vaccinations within acceptable day ranges, provided serology samples within 21 to 49 days post last dose of the assigned regimen, were seronegative at day 1 for the HPV type being analyzed, and had no other protocol violations that could interfere with immunogenicity assessments. Safety was assessed in all participants who received at least 1 study vaccination and had follow-up data.

Geometric mean titers (GMTs) and associated 95% confidence intervals (CIs) were computed for all immunogenicity assessment time points between day 1 and month 36. GMTs at months 7, 12, 24, and 36 in girls and boys who received the (0, 6) regimen were compared with GMTs in girls and young women who received 3 doses. The summary measures used for the comparison were the point estimates and 95% CI estimates of the ratio of the GMTs. The month 24 and month 36 visits represent 1.5 and 2.5 years post last dose of both the (0, 6) and (0, 2, 6) dose regimens (Fig 1); therefore, immunogenicity results at months 24 and 36 can be directly compared for these dose regimens. However, the month 24 and 36 visits represent 1 and 2 years post last dose, respectively, of the (0, 12) regimen (Fig 1). Because of the different time intervals between the last dose and the month 24 and month 36 visits, the immunogenicity results at months 24 and 36 in the (0, 12) regimen cannot be directly compared with the corresponding results in the (0, 6) and (0, 2, 6) regimens.

The study was conducted between December 16, 2013, and July 24, 2017. Of 1518 enrolled participants (Fig 2), 159 discontinued the study, most commonly because of loss to follow-up or withdrawn consent. Participant baseline characteristics have been described previously.¹⁸  Briefly, the mean age of the girls and boys 9 to 14 years of age was 11.5 years, and the mean age of the young women 16 to 26 years of age was 21 years. The study was diverse with respect to race and region of residence.

Antibody responses at 1 month after the last dose of the assigned regimen were reported previously.¹⁸  Formal statistical testing revealed noninferiority of GMTs at 1 month after the last dose in the 2-dose cohorts versus young women receiving 3 doses (the noninferiority criterion required that the lower bound of the 95% CI of the GMT ratio [2-dose to 3-dose] be >0.67).¹⁸ 

Antibody responses (GMTs) over time through month 36 are presented in Fig 3. Throughout the study, anti-HPV cLIA GMTs in girls and boys who received the 2-dose or 3-dose regimen were generally higher or similar compared with those in young women who received the 3-dose regimen (Fig 3). GMTs were the highest at 1 month after the last dose, then generally declined until levels plateaued at ∼12 (for the [0, 12] regimen) to 18 months (for all other regimens) after the last dose of the regimen (Fig 3), which corresponded to the month 24 study visit. Analyses to compare GMTs in girls and boys who received the (0, 6) regimen with young women who received 3 doses were conducted at months 7, 12, 24, and 36. GMT ratios (girls and boys to young women) were the highest at month 7 and tended to decrease between months 7 and 24 and then, remained stable between months 24 and 36 (Table 1). Across HPV types, the GMT ratios between girls and boys receiving the (0, 6) regimen and young women receiving the 3-dose regimen ranged from 1.13 to 2.06 at month 24 and 1.05 to 1.93 at month 36. The lower bound of the 95% CI for the GMT ratios remained >0.67 for all HPV types (ranging from 0.97 to 1.79 and 0.90 to 1.66 at month 24 and month 36, respectively; Table 1); thus, the noninferiority criterion that was demonstrated at month 7 (as reported previously¹⁸ ) persisted at month 24 and month 36. Similar results were obtained when considering separately girls and boys who received the (0, 6) regimen (Supplemental Table 3).

Although GMTs at study visits beyond 1 month post last dose are not directly comparable between girls and boys who received the (0, 12) regimen and young women who received the 3-dose regimen, the graphical representation of GMT trends at similar time intervals post last dose of the regimen can be compared (Fig 3). For all HPV types, the GMT trends at 12 to 24 months post last dose of the regimen were higher in (0, 12) regimen recipients compared with the corresponding trends in young women who received 3 doses (Fig 3). Similar trends were observed when considering separately girls and boys who received the (0, 12) regimen (Supplemental Table 4).

We performed exploratory analyses comparing girls who received a 2- or 3-dose regimen to glean insight into the effect of 2- vs 3-dose regimens in participants of the same age group, ie, eliminating the effect of age difference. Generally, 2-dose (0, 12) regimen recipients had higher or similar GMT trends, compared with girls who received 3 doses from 1 to 24 months post last dose of the regimen. This observation applies to all HPV types except HPV-45 (Fig 4). Generally, for all HPV types, 2-dose (0, 6) regimen recipients had lower or similar GMT trends compared with girls who received 3 doses from 1 to 30 months post last dose of the regimen (Fig 4).

Analyses to compare GMTs in girls who received the (0, 6) regimen with girls who received 3 doses were conducted at months 7, 12, 24, and 36. GMT ratios (girls [0, 6] to girls [0, 2, 6]) were the highest at month 7, tended to decrease between months 7 and 24 and, then, remained stable between months 24 and 36 (Table 2). Across HPV types, the GMT ratios between girls receiving the (0, 6) regimen and girls receiving the 3-dose regimen ranged from 0.43 to 1.09 at month 24 and 0.46 to 1.11 at month 36. Similar results were obtained when comparing boys who received the (0, 6) regimen with girls who received 3 doses (Supplemental Table 5).

Administration of a 2-dose regimen in girls and boys resulted in high seroconversion in all vaccine HPV types at 1 month after the last dose of the regimen on the basis of cLIA (range: >99%–100%; Supplemental Table 6 and previously reported¹⁸ ) that persisted through 2 years ([0, 12] regimen) to 2.5 years ([0, 6] regimen) after the last dose of the regimen (range: >81%–100%; Supplemental Table 6). The seropositivity based on IgG-LIA for vaccine HPV types in 2-dose regimen recipients at 2 years ([0, 12] regimen) to 2.5 years ([0, 6] regimen) after the last dose of the regimen ranged from >91% to 100% (Supplemental Table 7).

In age-stratified analyses of immunogenicity in 2-dose regimen recipients, for each vaccination group, the HPV type-specific GMT and seropositivity trends over time that were observed in the overall vaccination group were also observed consistently across the 3 age strata within the vaccination group (Supplemental Tables 8 and 9). Across all age strata, GMTs for the 2-dose recipients were generally similar to or higher than GMTs in young women who received 3 doses (Table 1; Supplemental Tables 8 and 9).

The study was not designed for the assessment of antibody response to 1-dose administration of the 9vHPV vaccine. However, serum collection in (0, 6) regimen recipients at month 1 and month 6 (predose 2 vaccination) and in (0, 12) regimen recipients at month 12 (predose 2 vaccination), allowed for a short-term analysis of immunogenicity response by cLIA and IgG-LIA after 1-dose administration of the 9vHPV vaccine (Fig 2; Supplemental Tables 10 and 11). Generally, the 1-dose regimen of the 9vHPV vaccine did not result in 100% seropositivity at 1 month postvaccination, and the GMTs at 1 month after 1 dose were lower than GMTs after 2 or 3 doses (12–49-fold lower than the [0, 6] regimen; Table 1; Supplemental Table 10). The antibody levels declined from 1 to 6 months post dose 1 on the basis of cLIA, as shown by the percent seropositivity and GMTs in girls who received the (0, 6) regimen (Supplemental Table 10). The results at 12 months post dose 1 in participants receiving the (0, 12) regimen were consistent with this trend of declining antibody titers after a 1-dose regimen of the 9vHPV vaccine. Only HPV-58 exhibited antibody titers that did not decline rapidly from 1 to 12 months post dose 1. Supportive analyses performed using the IgG-LIA also suggested declines in antibody titers and seropositivity rates over time for some HPV types after a single 9vHPV vaccine dose (Supplemental Table 11).

A total of 34 participants experienced SAEs during the study (Supplemental Table 12). One participant in the (0, 6) group experienced abdominal pain 15 days after the month 36 dose and fully recovered 11 days later; this SAE was considered vaccine related. One girl in the (0, 2, 6) group experienced autoimmune encephalitis and status epilepticus 353 days after dose 3 and died of cardiac arrest 114 days later; none of these SAEs were considered vaccine related. Narratives for these 2 events are provided in the Supplemental Information. Thirty-two participants experienced SAEs that were not considered vaccine related and resolved (2 participants resolved with sequelae). One participant (a 9-year-old girl) discontinued from study vaccination because of a nonserious vaccine-related adverse event of transient urticaria 1 day after the first dose of the vaccine, which fully resolved.

Antibody responses in girls and boys (aged 9–14 years) who received 2 doses of 9vHPV vaccine 6 or 12 months apart were maintained over 2 to 2.5 years after administration of the second dose. GMTs were generally similar to or higher than responses in young women who received 3 doses throughout this time period, indicating that the noninferiority profile of the 2-dose regimens in girls and boys versus a 3-dose schedule in young women¹⁸  persisted through 2 to 2.5 years. The GMT ratios (2-dose [0, 6] to 3-dose) ranged from 1.05 to 1.93 at month 36; the noninferiority criterion persisted throughout the study for all HPV types, including those with the lowest GMT ratios (HPV-45 and HPV-52). Although antibody levels after the (0, 12) regimen were not directly comparable with the other dosing regimens, the GMTs during the 2 years after the last vaccine dose were generally higher than those in young women receiving 3 doses during the 2 years after the last dose. On the basis of this immunogenicity bridging, the efficacy of 2-dose schedules of 9vHPV vaccine in girls and boys 9 to 14 years of age was inferred.

In exploratory analyses, GMTs were generally similar or lower in girls who received 2 doses 6 months apart versus girls who received 3 doses. Lower GMTs have been reported in previous studies of 2-dose regimens of the quadrivalent and bivalent HPV vaccines.^12,14  In contrast, GMTs were generally higher in girls who received 2 doses 12 months apart versus girls who received 3 doses. Results from studies of 3-dose regimens of the qHPV vaccine have indicated that a longer interval between the first 2 doses resulted in higher antibody responses.^23–25  The relevant objective of this study was to compare 2-dose regimens of the 9vHPV vaccine in girls and boys with a 3-dose regimen in young women (the population and dose regimen used to demonstrate 9vHPV vaccine efficacy) to infer effectiveness of 2-dose regimens. Therefore, the clinical significance of these differences in the immunogenicity of 2-dose and 3-dose regimens in girls is unknown. On the basis of these results, the ACIP has recommended a 2-dose schedule in individuals 9 to 14 years of age, with the 2 doses administered 6 to 12 months apart.²⁰ 

The durable immune responses observed after 2 doses of the 9vHPV vaccine are consistent with findings from clinical studies of 2-dose regimens of bivalent HPV and qHPV vaccines.^{12,13,15,26,27}  In particular, antibody responses at 1 month after the last dose were noninferior in girls receiving 2 qHPV vaccine doses (6 months apart) compared with young women receiving 3 doses.¹²  Long-term follow-up data from the qHPV vaccine study revealed a decline in GMTs during the 5 years after vaccination; GMT ratios (2- vs 3-dose) tended to decrease between months 7 and 24 and then remained stable between months 24 and 36 (which is consistent with the results reported here with the 9vHPV vaccine).²⁶  After month 36, the trend in decline was similar between girls receiving the 2-dose and 3-dose regimens.²⁶  Ten years after vaccination, anti-HPV antibody responses in girls receiving either a 2-dose or 3-dose regimen were noninferior to those observed in young women receiving 3 doses.¹⁵  With this result, together with the high (>95%) efficacy and long-term effectiveness of the qHPV vaccine in young women,^28,29  it is suggested 2 doses of the qHPV vaccine would provide similar efficacy in the target population. The similar trajectories of antibody responses up to month 36 between the qHPV and 9vHPV vaccines suggest longer-term immunogenicity may also be similar, in the absence of longer-term follow-up data with the 9vHPV vaccine.

Our findings of durable antibody responses to 9vHPV vaccine in girls and boys for at least 2 to 2.5 years after the last dose of the regimen support the current ACIP and WHO recommendations for the use of 2-dose HPV vaccine schedules in young adolescents (aged 9–14 years), which have been adopted in many countries.^17,19,20,27  For individuals from the ages of 9 to 14 years, the WHO recommends 2 doses 5 to 13 months apart,¹⁷  and the ACIP recommends 2 doses separated by 6 to 12 months.²⁰  The use of fewer doses may have advantages in terms of cost and improving adherence to HPV vaccination regimens.³⁰ 

An exploratory analysis of antibody responses at 1, 6, and 12 months after a single vaccine dose indicated that not all participants seroconverted to the vaccine HPV types after vaccination, and declines in antibody titers and seropositivity rates were observed over time. Post hoc analyses of observational studies of clinical trial participants who were randomly assigned to receive 2 or 3 doses but did not complete the vaccination series provided the initial suggestion of effectiveness of single-dose HPV vaccination.^31,32  In 2 ongoing randomized trials, researchers are evaluating the short-term efficacy (2–4 years) of a 1-dose schedule to prevent infection with vaccine HPV types: the ESCUDDO (Estudio de Comparacion de Una y Dos Dosis de Vacunas Contra el Virus de Papiloma Humano) trial in Costa Rica (NCT03180034) and the KENSHE (KENya Single-dose HPV-vaccine Efficacy) trial in Kenya (NCT03675256). Considering the lower immunogenicity, decrease in GMT over time, and incomplete seroconversion observed after a single dose of the 9vHPV vaccine, it will be important to thoroughly assess the long-term effectiveness of single-dose HPV vaccination in these trials.

In the current study, we included a comprehensive evaluation of several dosing regimens in both girls and boys; however, there are several limitations. The duration of follow-up was limited to 2.5 years after the last dose of the vaccination regimen. However, a durable immune response through at least 10 years has been demonstrated after 2 doses of the qHPV vaccine.¹⁵  Together with the similarity of the immunogenicity profiles of the qHPV and 9vHPV vaccines,¹  these results provide reassurance that a 2-dose regimen of the 9vHPV vaccine can provide durable protection. In addition, this noninferiority immunobridging study was not designed to assess the efficacy of 2- versus 3-dose regimens against HPV infection and disease, given the low potential for young adolescents to be exposed to HPV. The effectiveness of the 2-dose regimens was inferred on the basis of the demonstration of noninferior immunogenicity in young women from ages 16 to 26 years who received a 3-dose regimen. Two-dose regimens were not evaluated in participants aged >14 years. This could be an area for future research because, in some studies, researchers have suggested that a 2-dose regimen of HPV vaccine may be suitable in individuals aged >14 years.³³ 

In this study, we assessed regimens with 2 doses administered either 6 or 12 months apart. Intervals longer than 6 months may have logistic advantages in some settings (eg, in school-based immunization programs or during periods of vaccine shortages), and preliminary evidence suggests longer intervals between 2 doses of HPV vaccine may still provide protective antibody levels.³⁴  Studies in which researchers rigorously validate regimens with longer than 12 months between doses remain to be conducted.

We thank the study participants, site personnel, and investigators; Michael Ritter and Sonali Rawat from Merck & Co, Inc, Kenilworth, NJ for their expert assistance in the scientific conduct of the study; Thomas Group from Merck Sharp & Dohme Corp for facilitating the immunologic testing of serum samples; and Xiaofei Hu from Merck Sharp & Dohme Corp for providing statistical expertise. Medical writing support, under the direction of the authors, was provided by Erin Bekes, PhD, of CMC AFFINITY, a division of McCann Health Medical Communications, Inc, funded by Merck Sharp & Dohme Corp, a subsidiary of Merck & Co, Inc, in accordance with Good Publication Practice guidelines.

9vHPV

9-valent human papillomavirus

ACIP

Advisory Committee on Immunization Practices

CI

confidence interval

cLIA

competitive Luminex immunoassay

GMT

geometric mean titer

HPV

human papillomavirus

IgG-LIA

immunoglobulin G-Luminex immunoassay

mMU

milli-Merck units

PPI

per-protocol immunogenicity

qHPV

quadrivalent human papillomavirus

SAE

serious adverse event

WHO

World Health Organization