Results
The search in the different databases, retrieved 1214 citations relating to our subject. After removing the duplicates 930 were screened on titles or abstracts. We considered 39 studies possibly eligible and were retrieved for full text reading. Six studies were useful for our aim and these were included in this review (19, 22-26) (figure 1). Table 1 presents an overview of the selected studies. The characteristics of the studies are presented in Table S1 and the quality assessments are presented in supplementary S2.
Due to the heterogeneity we didn’t perform a statistical pooling of the data of these five studies together. We conducted two different meta-analyses: one including the studies of Joura et al. Giannella et al. and Hildesheim et al, which are comparable in design and method, and one including the studies of Kang et al. and Ghelardi et al which are relatively comparable as well.
Giannella et al. described two different cases of a woman with a CIN lesion treated with LEEP. The first case involved a 33-year-old woman with a CIN 3, HPV 16 positive. At her own request she was given the quadrivalent HPV vaccine (HPV 6-11-16-18) after excision. At 6 months’ follow-up, the PAP smear was normal, but the HPV 16 infection had persisted. At 18 months, CIN 3 was diagnosed with colposcopy, HPV 16 and 39 positive. The subsequent LEEP showed already an invasive carcinoma. The second case described a 35-year-old woman. At the first evaluation 6 months after LEEP, atypical squamous cells of undetermined significance were found while the HPV test was negative. The quadrivalent HPV vaccine was then administered. Twenty-four months after the conization, a CIN 3 was found with HPV type 33 positive.
The studies by Joura et al, Garland et al, and Hildesheim et al, all concern a sub analysis of a large community-based clinical trial for the evaluation of the effectivity and efficacy of HPV-vaccination in young women, irrespective of HPV state. The methods of these studies are similar. Joura et al. retrospectively analyzed data from a phase III study on the quadrivalent HPV vaccine (HPV type 6,11,16,18). Garland et al. and Hildesheim et al. analyzed data from phase III studies testing the efficacy of the bivalent vaccine (HPV type 16,18). In all three studies, routine follow-up and treatment were provided after vaccination (with HPV vaccine or hepatitis A/placebo vaccine as control, randomly assigned 1:1), if necessary, in accordance with local medical practice (table 2). The total recurrence of CIN2+ was 1.49% (12/806) in the HPV-vaccinated group versus 3.51% (37/1052) in the non-HPV vaccinated group. The combined RR was favorable for vaccination; i.e. 0.41 (95% CI 0.21-0.78; P<0,01) Figure 2.
Kang et al. retrospectively reviewed data from 737 women who had a LEEP for CIN lesions. Three hundred and sixty women had been vaccinated with the quadrivalent HPV vaccine; the other women had not been vaccinated at the time of treatment. Patient characteristics were similar between these two groups. The CIN lesion had recurred in 9 women in the vaccinated group (2.5%) versus 27 women (7.2%) in the non-vaccinated group, regardless of causal HPV-type. Regarding the vaccine- related HPV types (16 and 18) the CIN lesion had recurred in 5 women in the vaccinated group (2.5%) versus 18 women (8.5%) in the non-vaccinated group. Vaccination was an independent risk factor for recurrence (HR = 2.840; 95% CI, 1.335-6.042; P<0,01).
The latest, and only prospective, study published to date is the SPRERANZA project from Ghelardi et al. This prospective, non-randomized case-control study evaluated the post-treatment vaccination effect. Women aged 18-45 who were scheduled for cervical surgery for CIN2+ to ≤ cervical cancer IA1, were counseled on HPV diseases and HPV vaccination in a 90-minutes session. After counseling, they were invited to participate in the trial, with the option to be vaccinated or not. The vaccination costs could not be reimbursed, but a discount was offered. Eventually, 536 women signed informed consent for participation, of which 248 women received the quadrivalent HPV vaccination. Baseline characteristics were similar between the group of vaccinated women and the group of non-vaccinated women. The median follow-up period was 36 months. Data of only 344 women were analyzed: almost 200 women had been lost to follow-up or had a follow-up period shorter than 6 months.
At 6 months, there was no statistically significant difference in HPV clearance between the groups (81.4% in the vaccinated group versus 84.9% in the non-vaccinated group). Clinical recurrence was found in 6.4% of the non-vaccinated group versus 1.2% of the vaccinated group (p=0.0112). The risk of CIN2+ after cervical surgical was significant reduced by 81.2% (95%CI, 34.3-95.7). In the vaccinated group, none of the CIN2+ lesions was related to HPV vaccine types (6-11-16-18).
The studies by Kang et al. and Ghelardi et al. are the only two known studies in which vaccination was given adjuvant to cervical surgery as a treatment for CIN2+. The recurrence rate of CIN2+ in these studies combined was favorable for HPV vaccination in addition to a LEEP: 2.07% (11/532) in vaccinated women versus 6.92% (38/549) in non-vaccinated women. Resulting in a significant positive vaccination effect in addition to a LEEP (RR 0.30 95%CI 0.16-0.58 P<0,01). (figure 3)