Evaluation of the immunogenicity of the quadrivalent HPV vaccine using 2 versus 3 doses at month 21: An epidemiological surveillance mechanism for alternate vaccination schemes

Abstract

The cost of HPV vaccines and the need for 3 doses remains a barrier for their inclusion in routine vaccination schedules for girls in low and middle income countries. In a non-inferiority study, we aimed to compare the immunogenicity of a standard 3 doses and a 2 doses schedule. We enrolled 450 participants in an open-label non-randomized clinical trial to evaluate the immunogenicity induced at different ages by the licensed HPV6/11/16/18 quadrivalent vaccine in a 2 doses schedule (0–6 months, n = 150 girls aged 9–10 y) and 3 doses schedule (0, 2, and 6 months; n = 150 girls aged 9–10 y and n=150 women aged 18 to 24 years). To assess the antibody response, blood samples were obtained at Month 7 and 21 after the first vaccination from participants in all study groups. cLIA testing was performed at Merck Research Laboratories. Antibody levels were expressed as milli-Merck units (mMU) per ml. Primary outcome was non-inferiority (95% CI, lower bound >0.5) of the geometric mean titers (GMT) ratios for HPV6, HPV11, HPV16 and HPV18 antibodies 7 and 21 months after the first dose among girls receiving 2 doses compared with young women and girls receiving 3 doses. All vaccinees were seropositive for both HPV16 and HPV18 antibodies at month 7. At month 21, 98.5 and 56.6% of women 18–24 y old were seropositive for HPV16 and 18, respectively. For girls in the three doses group, seropositivity rates were 99.3 and 86.3% for HPV16 and 18, respectively. For girls in the two doses group rates were 99.3 and 70.2% for HPV16 and 18, respectively. The two doses schedule was non-inferior compared to the 3 doses schedule in same-age girls and to the group of adult women after 21 months of the first vaccine dose. Our results are in agreement with similar trials evaluating the immune response of a 2 doses schedule of both HPV vaccines, supporting the recent WHO recommendation as well as the Mexican policy to incorporate the 2 doses schedule for girls aged 9–11 y.

Keywords:

• alternative vaccination scheme
• HPV quadrivalent vaccine
• Mexico
• non-inferiority

Introduction

It is estimated that HPV genotypes 16 and 18, which are found in bivalent and quadrivalent HPV vaccines, contribute to at least 70% of invasive cervical cancer cases in Latin America.¹ These vaccines are highly effective,² providing high levels of immunogenicity³ with an acceptable safety profile.⁴ The maximum cost-benefit ratio has been estimated to occur when the vaccine is administered before exposure to HPV,⁵ and universal HPV vaccination schemes should therefore focus on adolescent girls who have not yet become sexually active.⁶ These vaccines have been incorporated into public vaccination policies with 3 doses schemes administered over a period of 6 months.⁷

High effectiveness of the tetravalent vaccine in preventing genital warts has been observed in Sweden⁸ and Denmark,⁹ and a herd immunity effect has been documented in Australia.¹⁰ In the United States, 4 y after the introduction of a vaccination program, 56% reductions in HPV prevalence have been observed in women between 14 and 19 y of age despite the low coverage rates achieved for the second and third doses.¹¹ Reduction in the incidence of high grade cytological abnormality has also been observed in Australia¹² and in Denmark.¹³

Studies of antibody responses and duration following HPV vaccination have shown a maximum peak in antibody titers 7 months after beginning the vaccination scheme.^14,15 Subsequently, a gradual decline in antibody levels is observed, and by the 24^th month these levels stabilize and remain constant until at least the 60^th month.¹⁶ To date, the minimum level of antibodies correlated with clinical protection is not yet known,¹⁷ but a correlation has been established between the presence of antibodies and a protective effect. However, it has been observed that in women between 10 and 17 years of age, the geometric mean titers (GMTs) of IgG are at least twice as high as those in women aged from 18 to 25 y.¹⁸ The immune response of 9–11 y old girls after 2 doses of quadrivalent HPV (qHPV) vaccine is similar to or greater than that the one obtained after 3 doses in women 16–26 y of age, the age group for which the efficacy of the vaccine has been demonstrated.¹⁹ In subjects vaccinated under the traditional scheme, the application of an extra (fourth) dose of the vaccine 60 months after the first vaccination resulted in a rapid and vigorous memory immune response against all 4 HPV genotypes included in the qHPV vaccine, which exceeded the levels of GMTs observed one month after the third dose.²⁰ This response was similar to the response observed with the Hepatitis B vaccine.²¹

In 2008 Mexico implemented an extended vaccination schedule of 0–6–60 months in girls younger than 16 y.²² There is now evidence that alternative vaccination schemes are not inferior to the traditional scheme in terms of immunogenicity.²³ In April 2014 the World Health Organization Strategic Advisory Group of Experts in Immunization recommended a 2 doses schedule (0–6 Months) for girls under 15 y of age, taking into consideration evidence from non-inferiority immunogenicity trials.²⁴ Currently, the European Union, in addition to some other countries worldwide approved the administration of 2 doses of HPV vaccine.²⁵ Furthermore, since April 2014, the regulatory authority in Mexico has approved the use of 2 doses of the HPV vaccine in girls aged 9–14 y.²⁶

In this article we present the immunogenicity achieved using a qHPV vaccine at Month 21 in 3 comparison groups of Mexican women. These results are useful for monitoring the effect of the vaccine in terms of the immune response generated by a 2 doses schedule compared to the standard scheme.

Results

Four hundred fifty participants were enrolled in this trial. All 150 girls assigned to the standard schedule received complete doses. Of the 150 girls assigned to the 2 doses schedule, 98% (148/150) received the complete vaccine series. One hundred forty three (95.3%) women aged 18–24 y received the qHPV vaccine on the standard 3 doses schedule (Fig. 1). Regardless of their group, participants received the vaccine within the 6-month interval between the first (month 0) and last dose (month 6).

Figure 1. CONSORT diagram. Figure legend. Flowchart of participants disposition throughout the study.

The time range between the last dose of the vaccine and the blood sample collection for the first serological evaluation (Month 7) ranged from 0.7–1.6 months for all girls (independently of the vaccine schedule group), and 0.7–1.7 months for the adult women. All of the groups complied with the predefined second serologic assessment between Months 19 and 23 after the first dose.

All of the vaccinees were seropositive for both HPV16 and HPV18 antibodies at Month 7 (Table 1). However, only 56.6% of the women aged 18–24 y were seropositive for HPV18 at Month 21.

Table 1. Seropositivity prevalence

Month 7
Antigen	Women (18–24 years)
	Two dose (M, 0–6) n=145	Three dose (M, 0–2–6) n=150	Three dose (M, 0–2–6) n = 141
	Seropositivity % (n)	Seropositivity % (n)	Seropositivity % (n)
HPV16	100	100	100
HPV18	100	100	100
HPV6	97.2 (141)	98.7 (148)	97.1 (137)
HPV11	99.3 (144)	100	100
Month 21
Antigen	Women (18–24 years)
	Two dose (M, 0–6) n=141	Three dose (M, 0–2–6) n=146	Three dose (M, 0–2–6) n = 136
	Seropositivity % (n)	Seropositivity % (n)	Seropositivity % (n)
HPV16	99.3 (140)	99.3 (145)	98.5 (134)
HPV18	70.2 (99)	86.3 (126)	56.6 (77)
HPV6	95.7 (135)	95.9 (140)	89 (121)
HPV11	97.2 (137)	100	97.1 (132)

Serostatus cutoff values: 20, 16, 20, and 24 mMU/ml for HPV6, 11, 16, and 18, respectively, according to previous validation studies. (Opalka D. 2003; Dias D. 2005)

The distribution of serum antibody titers for HPV6/11/16/18 of girls with 2 doses regime and adult women with 3 doses according to the quartiles categories of antibody titers defined in the reference group of girls receiving the 3 doses regime is shown in Table 2. We observed the 2 doses group tended to have a greater distribution of HPV16 antibodies across the top quartiles values defined by the reference group distribution: 19.1% were in the lowest quartile and 28.4% were in the highest quartile. In contrast, a large proportion of girls (29.8%) in the 2 doses group were seronegative and 24.8% in the lowest quartile of anti-HPV18 of the reference group. Regarding women aged 18–24 y we should emphasize the large proportion (43.4%) of this group who were seronegative for HPV18 at month 21, while only 5.1% of this group had comparable levels to those in the top quartile.

Table 2. Distribution in quartiles of titers of HPV16/18/6/11 antibodies, as measured in milli-Merck units per mL (mMU/mL), for the cohort of 3 study groups at Months 7 and 21 according to protocol cohort analysis

Month 7						Month 21
Quartiles of titers for specific HPV type in girls 9–10 y with 3-dose of HPV vaccine (M, 0–2–6) n = 150		2-dose group (M, 0–6) 9–10 y n = 145	3-dose group (M, 0–2–6) 18–24 y n = 141			Quartiles of titers for specific HPV type in girls 9–10 y with 3-dose of HPV vaccine (M, 0–2–6) n = 146		2-dose group (M, 0–6) 9–10 y n = 141		3-dose group (M, 0–2–6) 18–24 y n = 136
Titers	Quartile	%	p-value	%	p-value	Titers	Quartile	%	p-value	%	p-value
mMU/mL	%					mMU/mL	%
Antigen HPV16
<20 (seronegative)	0	0		0		<20 (seronegative)	0.7	1.5		2
20–3877	25	33.8	0.391	65.2	0.000*	20–143	26	19.9	0.563	27.9	0.852
3878–7959	25	23.4	0.898	23.4	0.899	144–289	24	18.5	0.599	22.8	0.909
7960–20109	25	22.8	0.806	7.1	0.213	290–956	24.7	32.6	0.435	33.1	0.410
20110–79836	25	20	0.651	4.3	0.262	957–8140	24.7	28.4	0.716	14.7	0.380
Antigen HPV18
<24 (seronegative)	0	0		0		<24 (seronegative)	13.7	29.8	0.169	43.4	0.017*
24–468	25.3	42.1	0.090	61.7	0.000*	24–48	22.6	24.8	0.831	22.1	0.962
469–1317	24.7	31.7	0.483	29.1	0.662	49–133	20.6	18.5	0.836	21.3	0.947
1318–2368	25.3	17.2	0.448	7.1	0.213	134–260	21.9	17	0.649	8.1	0.308
2369–17127	24.7	9	0.229	2.1	0.371	261–2160	21.2	9.9	0.358	5.1	0.322
Antigen HPV6
<20 (seronegative)	1.3	2.8		2.8		<20 (seronegative)	4	4.3	0.986	11	0.618
20–83	25.3	31.0	0.566	15.6	0.380	20–37	24	19.1	0.651	21.3	0.798
84–241	24	20.0	0.700	18.4	0.597	38–59	24	19.1	0.651	26.5	0.809
242–1807	24.7	26.9	0.827	51.1	0.008	60–166	24	30.5	0.523	22.1	0.856
1808–8786	24.7	19.3	0.605	12.1	0.289	67–1166	24	27	0.769	19.1	0.647
Antigen HPV11
<16	0	0.7		0		<16 (seronegative)	0	2.8		2.9
(seronegative)
16–839	25.3	42.8	0.077	61	0.000*	16–70	26	26.2	0.984	39.7	0.172
840–1663	24.7	22.1	0.800	28.4	0.714	71–151	24	29.1	0.617	33.8	0.338
1664–2766	25.3	13.1	0.288	3.5	0.277	152–277	25.3	17.7	0.480	15.4	0.380
2767–21005	24.7	21.4	0.748	7.1	0.225	278–1367	24.7	24.1	0.953	8.1	0.235

Note. Quartiles distribution were compared using girls 9–10 y with 3-dose of HPV vaccine as reference. Statistical tests were 2 sided with a significance threshold of 0.05. Numbers with * correspond to P values <0.05.

The GMTs for both HPV-type antibodies at Months 7 and 21 for all groups are shown in Table 3. The GMTs for HPV6, 11, 16 and 18 antibodies at 21 months after the first vaccine dose in the 2 doses group of girls were higher and statistically non-inferior to those in the group of adult women. Non-inferiority was demonstrated because the lower limit of the 95% CI of the GMT ratios for HPV6, HPV11, HPV16 and HPV18 for girls (2 doses) compared with young women (3 doses) as well as girls (3 doses) 21 months after the first dose was above the predefined limit of 0.5 (Table 3). When comparing the 2 doses vs. the 3 doses group among girls, antibodies were higher in the 3 dose group, but the levels for those receiving 2 doses were statistically non-inferior (Table 3).

Table 3. Comparisons of the geometric mean titers (GMTs) of HPV16, 18, 6 and 11 antibodies, as measured in milli-Merck units per mL (mMU/mL), for the cohort of 3 study groups at Months 7 and 21 according to protocol cohort analysis

					Month 7
	Girls (9–10 years)				Women (18–24 years)		GMT ratio (95% CI)
	Two dose (M, 0–6)		Three dose (M, 0–2–6)		Three dose (M, 0–2–6)
Antigen	n	GMT (95% CI), mMU/mL	n	GMT (95% CI) mMU/mL	n	GMT (95% CI) mMU/mL	Girls 2 dose/ Women 3 dose	Girls 2 dose/ Girls three dose
HPV-16	145	5136.7 (4035.8 – 6538.0)	150	6538.9 (5220.0 – 8191.2)	141	2408.8 (2003.5 – 2896.1)	2.13 (1.58 – 2.89)	0.79 (0.57–1.10)
HPV-18	145	605.0 (503.2 – 727.4)	150	1086.5 (890.6 – 1325.5)	141	343.7 (291.9 – 404.7)	1.76 (1.38–2.25)	0.56 (0.43*–0.73)
HPV-6	141	306.2 (228.3–410.7)	148	420.9 (316.7 – 559.4)	137	387.3 (305.4 – 491.2)	0.79 (0.54 – 1.15)	0.15 (0.48* – 1.09)
HPV-11	144	968.3 (800.1 – 1171.9)	150	1409.3 (1173.7 –1692.0)	141	629.9 (540.4 – 734.2)	1.54 (1.20 – 1.96)	0.69 (0.53 – 0.89)
					Month 21
	Girls (9–10 years)				Women (18–24 years)
	Two dose (M, 0–6)		Three dose (M, 0–2–6)		Three dose (M, 0–2–6)		GMT ratio (95% CI)
Antigen	n	GMT (95% CI), mMU/mL	n	GMT (95% CI) mMU/mL	n	GMT (95% CI) mMU/mL	Girls 2 dose/ Women 3 dose	Girls 2 dose/ Girls 3 dose
HPV-16	140	412.8 (338.1 – 504.1)	145	354.6 (288.9 – 435.3)	134	276.2 (226.0 – 337.7)	1.49 (1.12 – 1.98)	1.16 (0.88 – 1.55)
HPV-18	99	93.8 (76.4 – 115.3)	126	126.2 (105.3 – 151.3)	77	73.9 (61.3 – 89.1)	1.27 (0.96 – 1.67)	0.74 (0.57 – 0.98)
HPV-6	135	95.6 (81.0 – 112.8)	140	79.3 (67.4 – 93.2)	121	74.4 (63.3 – 87.3)	1.29 (1.02 – 1.62)	1.21 (0.96 – 1.52)
HPV-11	137	130.2 (109.3 – 155.0)	146	150.1 (128.3 – 175.5)	132	85.8 (73.9 – 99.6)	1.52 (1.21 – 1.91)	0.87 (0.69 – 1.10)

Primary outcome was non-inferiority (95% CI, lower bound >0.5) of the geometric mean titer (GMT) ratios for HPV16, HPV18. HPV6 and HPV11 for girls (2 doses) compared with young women (3 doses) as well as girls (3 doses) 7 and 21 months after the first dose. The seronegative participants at 7 and 21 months were excluded from GMT´s estimation.

The kinetics of antibodies against HPV6/11/16/18 in the 2 doses group of girls aged 9–10 y followed a similar pattern to those observed in both other groups; after a peak response one month after the third dose (month 7), a decline in antibody titers was observed at Month 21 (Fig. 2). However, the levels remained above those reported in women with natural infection with the exception of HPV 11, whose titers dropped to the level of natural infection.²⁷

Figure 2. Kinetics of antibody responses to HPV-16/18/6/11 from quadrivalent vaccine. *GMTs, geometric mean titers; milli-Merck units per mL (mMU/mL); CI, confidence interval; M 0,2,6, HPV-16/18/6/11 vaccine at standard schedule at 0,2,6 months; M 0–6–60, HPV-16/18/6/11 vaccine at extended schedule at 0–6–60 months. Plateau (dotted line), GMT values for Canadian women ages 16–26 y who were vaccinated with 3 doses of the quadrivalent vaccine.²⁸

Safety was evaluated in all 450 of the participants who received at least one dose and who attended the subsequent post-vaccination appointment. Among the solicited reports of adverse events following any vaccination, we had very infrequent reports in all 3 comparison groups (Table 4). Pain at the injection site was the most common local symptom across groups. No participants withdrew from the study because of an adverse event, and there were no serious or fatal events.

Table 4. Symptoms reported during the next vaccination period by schedule of vaccine administration and age (total vaccinated cohort)

	Standard schedule (M, 0–2–6)		Alternative schedule
Adverse events	Girls 9–10 y 3 dose (N=150 )	Women 18–24 y 3 dose (N=150 )	Girls 9–10 y 2 dose (N=150 )
Solicited local symptoms
Pain
Postdose M,0	3%	15%	8%
Postdose M,2	5%	7%	—
Postdose M,6	5%	5%	5%
Redness			0%
Postdose M,0	0%	0%	2%
Postdose M,2	3%	2%	—
Postdose M,6	1%	2%	0%
Solicited general symptoms
Headache
Postdose M,0	0%	0%	1%
Postdose M,2	1%	0%	—
Postdose M,6	1%	0%	1%
Fever
Postdose M,0	0%	0%	1%
Postdose M,2	0%	0%	—
Postdose M,6	0%	0%	1%
Dizziness
Postdose M,0	0%	1%	0%
Postdose M,2	0%	0%	0%
Postdose M,6	0%	0%	0%

Discussion

Our findings are consistent with previous studies documenting that in terms of immunogenicity 2 doses of the HPV6/11/16/18 vaccine in girls 9 to 10 years-old are not inferior to 3 doses among girls at the same age group as well as women 18 to 24 y.

In 2008 the Mexican government adopted an extended HPV vaccination scheme (0–6–60 month) focusing on 9 to 11 years-old girls. This schedule was adopted trying to achieve maximum antibody levels just before their sexual activity begins, under the hypothesis that the delay in application of the third dose would lead to an increase in antibody levels. This was a controversial decision, since in the absence of efficacy studies for a 2 doses HPV vaccination schedule it assumed plausibility of protection from HPV infection as a function of the immune response at an early age.

The principal basis for alternative 2 doses vaccination schedules is that the magnitude of the immune response is inversely proportional to the age at which the vaccine is applied,⁴ and the application of 2 doses of HPV vaccines at a minimum 6-month interval appears to provide a level of protection similar to 3 doses, during the initial years after vaccination.²⁸ The serum levels of HPV antibodies observed after vaccination last longer the younger the age at vaccination as long as there is a 6 to 12 month interval between the first and second doses.²³ The principal public health implications of this policy are to lower costs and flexibility in the interval between doses, which could improve operational efficiency and allow increased coverage in low- and middle-income countries, as well as eliminate the observed low adherence to the third dose.²⁹

In relation to persistence of immune response to the HPV vaccine, a number of studies have reported that the magnitude of response could increase when the interval between the first and second dose is between 6 and 12 months.⁴ In fact, a study on immunogenicity after the application of the HPV6/11/16/18 vaccine in a cohort of adolescents which included Mexican young men and women recently documented persistence of a high level of antibody levels for HPV types 6, 11, 16 and 18.³⁰ The HPV6, 11, 16 and 18 antibodies levels we observed in the present study are consistent with the non-inferiority of 2 doses in young girls as compared with 3 doses in adult women, in accordance to a similar previous study.²⁸ We recognize that the efficacy of HPV vaccine depends not only on the amount of antibodies induced by vaccines, but it is also relying on some other factors like the quality of these antibodies, or the cellular immunity such as immune memory.³¹ However, the evaluation of those immune factors is out of the scope for the present analysis. We strongly recommend further evaluations in this area.

We should emphasize that a large proportion of women aged 18–24 y (43.4%) were seronegative for HPV 18 at month 21, while among the girls with 3 doses and 2 doses, only 13.7 or 29.8% of them were seronegative, respectively. This reduction of seropositivity for HPV18 by cLIA is consistent with findings in aprevious study in Colombia, which reported a 45% of seropositivity for HPV18 at month 72, nevertheless a sustained absence of disease due to HPV18, as well as due to the other HPV vaccine types was observed.³² However, when the total IgG assay is used, the proportion of women who are seropositive for HPV18 at month 72 is substantially higher (81.5% overall). The high seronegativity rate for HPV18 deserves further evaluation to better understand the long term protection of less than 3 doses schedule. In addition, it should be kept in mind that the cLIA assay is a competitive assay that measures only the subset of neutralizing antibodies that compete with the specific monoclonal antibody used in the assay for VLP surface binding. Thus, the assay can underrepresent the potentially protective antibody response induced by the vaccine.³³

Alternative two doses HPV vaccination schedules are emerging as a new global strategy, including in high-income countries. Between the months of December, 2013 and February, 2014 the European Medicines Agency, which is responsible for scientific evaluation and supervision of medicines, approved the use of 2 doses for the bivalent and tetravalent vaccine for girls under 15-years-old. After this, Great Britain,³⁴ the Provinces of Quebec and British Columbia in Canada³⁵ Switzerland,³⁶ South Africa³⁷ and Austria³⁷ adopted policies with 2 doses for HPV vaccination in girls under 15-years-old. A two-dose system has also been approved in some other countries like Bangladesh, Pakistan, the Philippines, Ghana, Nigeria, South Africa, El Salvador, Guyana, Haiti, Panama, Brazil, Colombia, Chile, Guatemala among others.³⁸

In Mexico, the medicine regulation agency (COFEPRIS) approved a 2 doses (0–6 months) schedule for the bivalent and HPV6/11/16/18 vaccines in 11-year-old girls on April 28th, 2014.²⁶ Similarly, Chile is currently preparing for implementation of a 2 doses scheme (0–12 months) and 52 countries have begun processes of approval of 2 doses schemes within their regulating agencies. In this context, October 24th, 2014 the World Health Organization made public its recommendation of 2 vaccine doses for girls 14-years-old or under, derived from the recommendations of their Strategic Advisory Group of Experts (SAGE).³⁹

Although the HPV vaccines are highly efficacious and immunogenic, to date there is no immune correlate of protection against cervical disease. Duration of protection of a 2 doses scheme requires further evaluation. To add to our knowledge and understanding of the protection against HPV infection induced by HPV vaccination, future studies addressing effectiveness of alternative schedules should evaluate virological endpoints such as 6-month persistent HPV 16/18. These virogical endpoints have been recommend by IARC as surrogate of clinical end-points in HPV trials.^40,41

We recognize some limitations of our trial. The lack of randomization among our population may not guarantee a homogenous distribution of socio-demographic or other relevant population characteristics, and we did not collect individual socio-demographic data to confirm potential differences. However, girls are from schools located in the urban area of Cuernavaca, Mexico, and they are both public schools that receive students from lower to middle-class socioeconomic backgrounds living close to those schools.

Vaccination should be one element in efforts to strengthen cervical cancer prevention and control efforts, in addition to early detection and treatment with guaranteed quality control. An epidemiological surveillance system should be incorporated using a centralized laboratory that quantifies HPV16 and 18 antibodies in the vaccinated population to evaluate the immune response to 2 doses and to estimate the possible protective effect, as this study among Mexican women did. We currently face the situation that most anti-HPV vaccine immunogenicity studies have used assays developed by the companies that develop the vaccines, which are not easily available to research groups, unless they participate in studies funded by the pharmaceutical industry. This means that there is a need to develop, validate and standardize serological studies that are easily available for the benefit of epidemiological surveillance systems. Given this, we should implement mechanisms for monitoring antibodies in alternative HPV vaccination schedules as a central part of an epidemiological surveillance system that could provide alerts of a decrease in the potential impact on cervical cancer prevention.

Material and Methods

This study is an open-label, nonrandomized clinical trial (NCT01717118) to evaluate the immunogenicity induced by the HPV6/11/16/18 vaccine administered using an alternative (0–6 months) and the standard (0–2–6 months) vaccination schedules.

Our objective was to conduct a preliminary non-inferiority analysis among 9 to 10 y girls having received 2 doses of qHPV vaccine compared with girls aged 9 to 10 y and women aged 18 to 24 y 21 months after the administration of a standard 3 doses scheme.

The trial was conducted in accordance with the International Conference on Harmonization of Good Clinical Practice Guidelines. The protocol was approved by the Institutional Review Board of the National Institute of Public Health (# 833), and all of the subjects provided written, informed consent; for underage participants, parents provided the written consent.

Study population

The study was conducted in the metropolitan area of Cuernavaca in Mexico. The study population of girls consisted of female students who were recruited at 18 public primary schools. In the state of Cuernavaca there are 97 primary and public schools that offer the morning shift. From those, 18 were selected by logistical convenience; they were the primary schools that were not already recruited by the study that made simultaneously in the same area. Each school has an average of 200 girls distributed along 6 y of primary education, and approximately 30 girls (from 18 to 42 per class) in the fifth year, those are the girls aged 9 and 10 y This age section is currently the target of the national vaccination program. All the girl's parents of 5^th grade were invited to an informative session where they were told about the study and were invited to participate. This was supported by the school Direction. The attendance was of 87% of parents (470/540). From these attending parents only 75% of them (352) gave their consent for the inclusion of their daughter's in the study; the participation rate across schools varied from 45 to 82%. The intervention made inside the schools was allocated by clusters, the first one was the intervention of a 2-dose vaccination schedule (7 schools) and the second one (11 schools) was the control cluster. From these groups we formally included 150 participants to reach the pre-defined total sample size per arm. A third group was created with adult women aged 18–24 y recruited at a center for clinical studies of the Instituto Nacional de Salud Pública named Centro Médico Cuauhtémoc (Fig. 1). Most of the recruited adult women were those who responded to the promotion of the study made in universities of the metropolitan area of Cuernavaca, where the invitation was made to adult women. Some additional women attended on their own following recommendation of other participants, and because they were teachers or related to the study girls. We formally recruited the first 150 women that accepted and complied with all the inclusion criteria. We did not document refusal rates. Exclusion criteria included fever at the time of vaccination, previous HPV vaccination, and history of allergy to vaccine components, history of thrombocytopenia, administration of any other vaccine during the 15 d prior to study enrollment, immunosuppression, diarrhea, vomiting, and blood disease. Likewise, the presence of diarrhea or vomiting at the time of administering the HPV vaccine was considered for exclusion.

Antibody assays

We used Gardasil®, a quadrivalent human papillomavirus (Types 6, 11, 16, and 18) vaccine (Merck, Whitehouse Station, NJ, USA) provided by the Mexican Ministry of Health. To assess antibody response, blood samples were obtained at 7 and 21 months after the first vaccination from the participants in all study groups. Merck cLIA testing was performed at Merck Research Laboratories as previously described,⁴² and antibody levels were expressed as milli-Merck units (mMU) per ml. The PsV NAb assay was performed as previously described.⁴³ Briefly, HPV16 and HPV18 PsV containing a reporter plasmid encoding red fluorescent protein (RFP) were prepared and titrated in 293TT cells. The sera were serially diluted, mixed with 100 infectious units of the respective HPV PsV, and inoculated onto 293TT cells in microtiter plates. The cultures were read by fluorescence microscopy after 4 to 6 d The endpoint was the highest dilution of serum that completely blocked the expression of RFP (100% neutralization). Seven and 21 month samples were tested in duplicate in the same assay run, and GMTs were calculated. Subjects were considered to be neutralizing antibodies seropositive for the respective HPV type if the GMT was 20, 16, 20, and 24 mMU/ml for HPV6, 11, 16, and 18, respectively, according to previous validation studies.⁴⁴ The testing laboratories were blinded to the dosing regimens.

Safety

The study staff recorded safety profile assessments of local symptoms (pain and redness at the injection site) and general symptoms (fever, headache, fatigue, and gastrointestinal symptoms that included nausea, vomiting, diarrhea and/or abdominal pain, arthralgia, myalgia, and rash) at the next scheduled appointment after the administration of each vaccine. The staff also inquired about and recorded any serious adverse events at each contact and among women withdrawing from the study.

Statistical analysis

The immunogenicity analysis was conducted on study participants who complied with protocol procedures and had data available for at least one vaccine antibody result. For the immunogenicity objectives, immunogenicity data (GMTs) with a 95% confidence interval (CI) were computed for HPV6/11/16/18 antibodies at Months 7 and 21 for the 3 comparison groups.

To compare the distributions of antibody response across the study groups, the serum antibody titers for the HPV6/11/16/18 antigens of the girls in the 3 doses group were categorized into quartiles and used as a reference, because we considered that the optimal immune response expected by HPV vaccine may be observed in this particular group, as has been documented in previous studies. Then, the HPV antibody distributions of the other 2 groups were compared to the reference group distribution.

Primary outcome was non-inferiority (95% CI, lower bound >0.5) of geometric mean titer (GMT) ratios for HPV6, HPV11, HPV16 and HPV18 for girls (2 doses) compared with women aged 18–24 y (3 doses) as well as girls (3 doses) at months 7 and 21 after the first dose. Our non-inferiority definition was based on criteria previously used in immunogenicity clinical trials of the qHPV vaccine¹⁹ that included results from Mexican women.⁴ The seronegative participants at 7 and 21 months were excluded from the GMT's estimation.

The natural infection GMTs that we used to examine the kinetic immune response correspond to the GMTs previously reported in subjects who had cleared a natural infection.⁴⁵ To evaluate the kinetics of the immune response across groups, we used the GMT obtained from serologic measurements at Months 7 and 21 after the first vaccine was administered. The plateau curve was adapted from the GMTs achieved until Month 36 in a non-inferiority immunogenicity trial of the qHPV vaccine in girls aged 16–26 y from Canada who received the vaccine on the standard schedule (M, 0–2–6). ²⁸

Statistical analyses were performed with STATA Version 12.0 (Stata Corp. College Station, TX, USA).

Disclosure of Potential Conflicts of Interest

Stanley M has acted as consultant for SPMSD Lyon France, MSD Merck and GSK Biologicals Wavre Belgium. Muñoz N receives honoraria from Merck MSD for attending meetings of the HPV Global Advisory Board. JS and EL have received funding through their institutions for research projects from Qiagen, GSK, Merck, Roche, Becton Dickinson, Gen-Probe, DICIPA, Silanes, and Arbor Vita. None of the other authors have potential conflict of interest to report.

Funding

Funding was provided by the Ministry of Health (Mexico) and the National Institute of Public Health (Mexico). Immunogenicity testing was provided by Merck, which had no role in the study design and conduct, data collection, management, analysis, or interpretation or in the preparation, review, or approval of the manuscript.