Impact of HPV-16/18 AS04-adjuvanted vaccine on preventing subsequent infection and disease after excision treatment: post-hoc analysis from a randomized controlled trial

Women included in the present evaluation were the participants who underwent excision treatment at baseline and the follow-up period in a phase II/III, double-blind, randomized trial conducted in Jiangsu province (Binhai, Jintan, Lianshui and Xuzhou CDCs), China. In this trial, women aged 18–25 years were randomized (1:1) to receive HPV-16/18 AS04-adjuvanted vaccine(n = 3026) or aluminum hydroxide (n = 3025) as a placebo at months 0, 1 and 6. Enrolment in this trial started in October 2008 and follow-up lasted for 72 months (14 visits). The trial was conducted according to The Code of Ethics of the World Medical Association (Declaration of Helsinki) and the International Conference on Harmonisation Good Clinical Practice guidelines. The study protocol and informed consent were approved by the ethics committees of the Center for Disease Control and Prevention (CDC) Jiangsu Province and the Cancer Foundation of China. The trial was registered at the ClinicalTrials.​gov (number NCT00779766) and adhered to CONSORT guidelines. Before the study-specific procedures, written informed consent was obtained from each participant. The further details of the trial have been described in previous published papers [13‐15]. Briefly, women after excision treatment were followed up by collecting cervical exfoliated cell samples in gynecologic physical examination at each study visit. The cervical samples were evaluated by HPV DNA PCR testing and cytology. For all cytology diagnoses of Atypical Squamous Cells of Undetermined Significance (ASCUS), the central laboratory had additionally performed HC2 High-Risk HPV DNA Test™ (Qiagen Inc., Gaithersburg, MD) on residual PreservCyt® specimen. The subjects were referred to colposcopy if they had cytology ASCUS with HPV positive results (by HC2 HPV DNA test), or cytology low-grade squamous cell intraepithelial lesion or worse (LSIL+) independent of HPV DNA results.

In this post-hoc analysis, we intend to explore the impact of HPV vaccination on women after excision treatment by focusing on intent-to-treat participants who received at least one dose of vaccine or placebo. Three level of analysis would be conducted in focusing on HPV infection (Fig. 1: analytical cohort 1), LSIL analysis (Fig. 1: analytical cohort 2) and recurrence of precancerous lesions, respectively. In the HPV infection-level analysis, each infection instead of a woman was considered as the unit of analysis to estimate the incidence rate of HPV16/18 (vaccine-specific types), HPV31/33/45(cross-protective types) [16], 14 HR-HPV (any of the 14 high-risk HPV types) and 11 LR-HPV (any of the 11 low-risk HPV types) infection. Persistent infection was defined as at least two positive HPV DNA PCR assays for the same viral genotype with no negative DNA sample between the two positive DNA samples, over an interval of 6 months or longer. Newly acquired infection (new infection) was defined as the positive detection by PCR of an episode of infection by HPV type(s) in a subject who was negative at excision treatment visit for the considered HPV type(s). Given the high efficacy of the HPV-16/18 vaccine against new infections, we also conducted analysis restricted to newly detected infection after treatment.

Cervical exfoliated cell samples for HPV DNA testing and cytological evaluation were collected at each study visit. Cytological evaluation was performed at the Cancer Institute of the Chinese Academy of Medical Sciences (CICAMS). And the results were interpreted according to the Bethesda 2001 classification system [17]. Participants with abnormal results were managed in accordance with the pre-specified algorithms, which had been described previously [13‐15]. Biopsy and excisional treatment specimens were analyzed by a panel consisted of three expert gynecological pathologists. The data for women with CIN diagnoses were reviewed by an independent endpoint committee who had been blinded to make final case assignments.

The automatic analyzer SPF10 PCR-DEIA-LiPA25 version 1 (manufactured by Labo Biomedial Product, Rijswijk, the Netherlands based on licensed Innogenetics technology) detected HPV DNA of cervical, biopsy samples and tissue specimens including 14 high-risk HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68) and 11 low-risk HPV types (6, 11, 34, 40, 42, 43, 44, 53, 54, 70 and 74).

Comparisons of enrollment characteristic, accumulative clearance rate and persistent infection rate between the two groups were conducted by Pearson χ² or Fisher’s exact tests, as appropriate. Vaccine efficacy (VE) was calculated as 1-(incidence rate in vaccinated/incidence rate in unvaccinated) and the corresponding 95% confidence intervals (95% CIs) around vaccine efficacy were calculated by exact conditional procedure based on the observed case split between vaccine and placebo recipients, adjusted for the person time in each arm. In the squamous intraepithelial lesion (SIL) analysis, we analyzed the occurrence rate of LSIL+ and HSIL+ stratified by specific-type after excisional procedure. Results of vaccine efficacy were considered as statistically significant if the estimates and corresponding 95% CIs were above zero. Kaplan-Meier method with log-rank test was applied to compare the difference in occurrence of LSIL+ between the two groups. All statistical tests were two-sided and only p-values < 0.05 were considered statistically significant. The statistical analyses were performed using WinPEPI version 4.0 [18] or SPSS 23.0 (SPSS, Inc., Chicago, Illinois).

168 women (vaccine, n = 87; placebo, n = 81) received excisional treatment for their first cervical lesion identified at this clinical trial. One woman in each group underwent loop electrosurgical excision procedure (LEEP) at the last visit (visit 14), and thus 166 women had been followed up and were categorized based on cytological results and HPV status at enrollment. Comparison of analytical cohort stratified by study arm revealed the balance regarding to cytology results, and HPV status (HPV type) at enrollment (Table 1).

Women who underwent cervical treatment surgery were followed for a median of 50.0 months after treatment (vaccine arm: 49.5, IQR: 32.0–64.0; placebo arm: 50.0, IQR: 30.3–63.5), corresponding to a median of 9 study visits (vaccine arm: 9, IQR: 5–12; placebo arm: 8, IQR: 5–11). The median duration between enrollment and treatment was 17 months for those in the vaccine arm (IQR:4.0–30.0 months) and 17 months for women in the placebo arm (IQR:6.0–34.8 months).

In the infection-level analysis, 10 women (vaccine, n = 5; placebo, n = 5) were excluded because of no follow-up after excision treatment (Fig. 1). Finally, 158 women (vaccine, n = 82; placebo, n = 76) were included in the analysis. There was no significant difference in the distribution of baseline HPV infection between the two groups (Fisher’ exact test: 3.64, P = 0.458). The woman was included in this analysis from the day when the women received an excisional procedure (LEEP or cone) for a first cervical lesion to the day of their last follow-up visit, considering the fact that women could be infected by HPV or had abnormal cytological results for several times after excisional treatment. Among all the infections that treated women had, 71.06% were HR-HPV infection and of these 52.49% were the result of new infections. We observed a significant effect of vaccination on acquiring 14 HR-HPV infection (VE 27.0%; 95%CI 4.9, 44.0%), and a nonsignificant but positive vaccine efficacy estimate of 28.1% (95% CI -62.9,68.3%) for HPV16/18 infection after excision treatment (Table 2). Vaccine efficacy against new infections after treatment for 14 HR-HPV infection was estimated to be 32.0% (95%CI 1.8,52.8%), and was 41.2% (95%CI -162.7,86.8%) for HPV-16/18 infection, which was consistently positive in this restricted analysis but didn’t reach statistical significance.

Then, we evaluated whether the viral clearance rates or HPV persistent infection rates differed by vaccination status. Results were shown in Fig. 2. No evidence from this study showed that vaccination led to viral faster clearance or had any effect on the persistent infection rate. The accumulative clearance rates of the vaccine group and placebo group were 88.9, 81.6% for HPV16/18 infection(P = 0.345), 81.3, 80.0% for HPV31/33/45 infection(P>0.999), 63.4, 48.7% for 14 HR-HPV infection(P = 0.062), respectively. No significant difference on the persistent infection rate of HPV16/18(4.4% vs 2.6%,P>0.999), HPV31/33/45(18.8% vs 10.0%,P = 0.637),14 HR-HPV (20.7% vs 17.1%,P = 0.561) was identified between the two groups.

When LSIL+ was examined as the outcome, nonsignificant but positive vaccine efficacy was estimated at 45.5% (95%CI -15.5, 74.2%) for 14 HR-HPV infection, and 85.1% (95% CI -23.5, 98.2%) for the newly detected outcome after treatment, respectively. Similar patterns were observed for HPV-16/18 infection as well.

15 women (vaccine, n = 7; placebo, n = 8) were excluded because of no cytological results (Fig. 1). Finally, 153 women were included in the final analysis (vaccine, n = 80; placebo, n = 73). The distribution of cytological results at baseline was similar between two groups (Fisher’ exact test: 8.80, P = 0.051). In this analysis, the follow-up time was defined as the duration from the day of excisional procedure to the day of ascertainment of the subsequent disease end point (LSIL+ incidence), while for women without a subsequent disease end point, until the day of their last follow-up visit. The median follow-up time was 46.0 months among women who were treated. The median follow-up time for vaccine group and placebo group were 48.5 months, 44.0 months, respectively. The vaccine efficacy for the occurrence of LSIL+ was 55.3% (95%CI -12.1, 82.2%), irrespective of HPV DNA types. There was no significant difference between two groups (P = 0.088, Fig. 3).

In the vaccine group, one woman had been detected with vaginal intraepithelial neoplasia grade 2 (VAIN2), one with VAIN1, and one with CIN1 post-surgery. Of the two women in the placebo group, one developed CIN2 and one had CIN1 after treatment. The women (Case 4: Fig. 4) who developed CIN2 in the placebo group was HPV16, 31, 33 DNA positive, with HSIL predicted by cytology at gynecological examination at visit3 (6 month). She underwent cold knife conization treatment at visit 3(6 months) and CIN3 was diagnosed by histology; the margins of the excisional material were disease-free. After 6 months she had cytological ASCH (atypical squamous cell, but cannot exclude high-grade squamous intraepithelial lesion) results and was referred to colposcopy. CIN2 was diagnosed on punch biopsy and HPV genotyping on cervical tissue sample revealed HPV31, 33 positive. The one VAIN2 case in the vaccine group (Case 2: Fig. 4) occurred in a woman who was HPV16 DNA positive, with a cytology HSIL result at baseline (visit1). She underwent LEEP treatment at baseline (visit1) and CIN3 was diagnosed by histology; the margins of the excisional material were disease-free. At 12 months, she had ASCUS on cytology and was referred to colposcopy. VAIN2 was diagnosed on punch biopsy and HPV genotyping on cervical tissue sample revealed HPV16 positive.