Primary prevention: human papillomavirus vaccinationVaccine Development

Human papillomaviruses (HPV) are small non-enveloped viruses with an 8 kb circular genome that encodes eight genes including two encapsulating structural proteins, L1 and L2.1 Mucosal viruses are transmitted mainly through skin-to-skin contact, and HPV is the most common sexually transmitted infection worldwide.2 Persistent infections with oncogenic types can lead to pre-invasive lesions and cancer, and HPV-related cancers include cancer of the uterine cervix, anus, vulva, vagina, and oral cavity.3 More than 40 HPV types that affect mucosal epithelium have been described, and 12 are defined as oncogenic (types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59) and some others as possibly oncogenic.4

More than 30 years ago, animal studies showed that there could be protection against papillomavirus lesions using virions through the generation of neutralizing antibodies5 and, later, the HPV vaccines were developed based on recombinant virus-like particles (VLP) based on the L1 major capsid protein. The type-specific HPV VLP resembles the L1 major capsid protein but has no viral DNA and consequently does not produce viral infection and does not have oncogenic potential. Natural infection antibodies primarily to the L1 major capsid protein develop slowly and are only detected in a portion of infected hosts. Conversely, HPV vaccination induces high and long-standing titers of genotype-specific antibodies.6

Two companies initially developed commercial HPV vaccines (Table 1). One developed a bivalent vaccine (Cervarix) composed of HPV-16 and HPV-18 VLP, and the other developed a quadrivalent vaccine (Gardasil) with HPV-16, HPV-18, HPV-6, and HPV-11 VLP.7 Both vaccines also differ in the producer cells for the VLP and their adjuvants. The VLP are produced in baculovirus-infected insect cells for the bivalent vaccine and the quadrivalent in Saccharomyces cerevisiae yeast. The bivalent vaccine has an AS04 adjuvant, which consists of monophosphoryl lipid A (a toll-like receptor 4 agonist) and aluminum hydroxide, while the quadrivalent vaccine contains aluminum hydroxyphosphate sulfate.8 9 Later, a nonavalent vaccine (Gardasil 9) was developed, like the quadrivalent but using L1 VLPs of five additional oncogenic HPV types 31, 33, 45, 52, and 58, and using 500 µg aluminum hydroxyphosphate sulfate instead of 250 µg. Recently, a new bivalent vaccine (Cecolin) was developed containing HPV-16 and HPV-18 VLP, and it was licensed in China in 2020. The L1 proteins are produced in Escherichia coli, and the vaccine has an aluminum adjuvant.10

Table 1

Characteristics of commercially available vaccines

Immunogenicity

HPV vaccines can induce T-cell and B-cell responses. Still, the clinical effect is almost exclusively due to the induction of neutralizing antibodies, mostly immunoglobulin G, that can reach sites of cervicovaginal infection through transudation and exudation to join the virions and prevent initial infection.6 HPV vaccines can generate robust antibody responses,11 with a seroconversion rate of virtually 100% for all types after the three-dose regimens in clinical trials, and geometric mean antibody titers are comparable across ethnic groups and regions. Titers rise after each dose and decline over time,12 with some studies confirming the persistence of antibody levels up to 11 or 12 years after vaccination.13

As the initial trials used different antibody assays, it is difficult to compare immunogenicity results from quadrivalent trials using competitive Luminex immunoassay (cLIA) and bivalent trials using ELISA.14 Head-to-head immunogenicity trials were later developed with the same assay, showing that the bivalent vaccine produced higher titers,15 and it is thought to be due to the AS04 adjuvant. However, clinical practice implications are unclear, given that both vaccines produce antibodies substantially higher than natural infection, and no minimum level of protective antibody is established.

Randomized Controlled Trials

The initial randomized controlled trials (RCTs) were conducted in young women aged 15–26 years.14 16–18 Although the target recipients for HPV vaccination programs are younger women (Table 2), efficacy trials were not feasible in that age group because it would take too long to achieve enough outcomes, and cervical cancer is not a realistic outcome to be evaluated in the context of these trials as all women were followed and treated for any pre-invasive lesion. Given this consideration, outcomes included a combination of cervical intra-epithelial neoplasia (CIN) and cervical adenocarcinoma in situ caused by the vaccine-targeted types, persistent infections of HPV vaccine-targeted types, antibody titers, and the incidence of anogenital warts as surrogate outcomes. HPV vaccines were licensed in young adolescents based on bridging immunogenicity trials and safety data. Later, RCTs were conducted in women aged >26 years19 and men.20

Table 2

Human papillomavirus (HPV) vaccine schedule and dosing

The RCTs were conducted by the manufacturers with three doses administered through intra-muscular injection, and the only trial non-funded by industry was conducted through the association between the US National Cancer Institute and the Costa Rican government.17 There were differences among the trials, like the inclusion criteria, follow-up for disease outcomes, HPV assays, and some outcomes.19 The Gardasil 9 trial was later designed to compare the nonavalent vaccine with the quadrivalent one,21 22 and the primary outcome was high-grade cervical, vulvar, or vaginal intra-epithelial lesions attributable to HPV types 31, 33, 45, 52, or 58. The Cecolin trial design is comparable to those for Gardasil and Cervarix.10

Women in these trials were included independently of HPV infection status, but the primary efficacy analyses included only women who were negative for cervicovaginal HPV DNA and antibodies, and who received three vaccine doses. It was demonstrated from the beginning that vaccines are not therapeutic and do not prevent the progression of infection present at the time of vaccination.23 24

The first RCT in women aged >25 years for the bivalent vaccine was the VIVIANE trial, which included more than 5000 women,25 and for the quadrivalent vaccine, the FUTURE III trial that included more than 3500 women.26 Outcomes for these trials were different with vaccine type 6-month persistent infection or vaccine type-attributable CIN 1+ (and anogenital lesions for the quadrivalent vaccine trial). The women were randomized without regard to HPV status.

More than 25 clinical trials using HPV vaccines, including more than 70 000 patients, have been published and most were assessed in a previous systematic review with meta-analysis.19 For patients negative for high-risk HPV and aged between 15 and 26 years, high-quality evidence shows that the HPV vaccine (including bivalent and quadrivalent) reduces the risk of CIN 2+ associated with HPV 16/18 (risk ratio (RR) 0.01, 95% CI 0.00 to 0.05) (3 RCTs and 23 676 patients) and CIN 3+ associated with HPV 16/18 (RR 0.01, 95% CI 0.00 to 0.10) (2 RCTs and 20 214 patients). Moderate quality evidence shows that the HPV vaccine reduces the risk of adenocarcinoma in situ associated with HPV 16/18 (RR 0.10, 95% CI 0.01 to 0.82) (2 RCTs and 20 214 patients).

Regarding any CIN 3+ irrespective of HPV type, evidence shows a reduction in risk for vaccinated patients that varied according to HPV vaccine type. High-quality evidence shows that the risk of CIN 3+ was reduced for the bivalent vaccine (RR 0.08, 95% CI 0.03 to 0.23) (2 RCTs and 11 423 patients), and moderate-quality evidence shows that the risk of CIN 3+ was reduced for the quadrivalent vaccine (RR 0.54, 95% CI 0.36 to 0.82) (1 RCT and 9296 patients). These data should be analyzed with caution as it is highly probable that the outcomes are predominantly represented by the effect of reduction in HPV 16/18 cases, which are the most common causal agents. For adenocarcinoma in situ, irrespective of HPV type, moderate-quality evidence shows a reduction in risk for both bivalent and quadrivalent vaccines (RR 0.10, 95% CI 0.01 to 0.76) (2 RCTs and 20 214 patients).

The effects of HPV vaccine on cervical lesions in adolescent girls and women unselected for HPV DNA status at baseline varied among age groups. For patients aged 15–26 years, high-quality evidence shows that vaccine reduces the risk of CIN2+ associated with HPV 16/18 (RR 0.46, 95% CI 0.37 to 0.57) (3 RCTs and 34 852 patients) and any CIN 2+ irrespective of HPV type (RR 0.70, 95% CI 0.58 to 0.85) (4 RCTs and 35 779 patients). For patients aged 24–45 years, moderate-quality evidence shows that there is no difference in the vaccine group for the risk of CIN 2+ associated with HPV 16/18 (RR 0.74, 95% CI 0.52 to 1.05) (2 RCTs and 9200 patients). For patients aged 15–26 years, high-quality evidence shows that the risk of CIN 3+ was reduced for the bivalent vaccine (RR 0.55, 95% CI 0.43 to 0.71) (2 RCTs and 18 329 patients) and moderate-quality evidence shows that the risk of CIN 3+ was reduced for the quadrivalent vaccine (RR 0.81, 95% CI 0.69 to 0.96) (1 RCT and 17 160 patients) and adenocarcinoma in situ (RR 0.32, 95% CI 0.15 to 0.67) (2 RCTs and 34 562 patients).

These results highlight why the target population for vaccine programs must be young women before sexual onset and concordantly without previous exposure to HPV.

Safety and Adverse Effects

There are enough data on the safety of HPV vaccination from both pre-licensure trials and post-licensure evaluation.27 All pre-licensure trials found the expected adverse effects, mainly fever and injection site reactions. High-quality evidence based on RCTs shows no difference in the rate of adverse events in vaccine groups with variable follow-up times from 6 months to 7 years (RR 0.98, 95% CI 0.92 to 1.05) (23 RCTs and 71 597 patients).19 Additional reviews of Vaccine Adverse Event Reporting System data for all three licensed HPV vaccines have not identified new concerns.28

One, Two, or Three Doses?

The post hoc analysis of the Costa Rica trial suggested similar high efficacy in those women randomized to receive three doses who did not complete the scheme and those who did.29 Later trials of two dose schedules were developed, evaluating the non-inferiority of two doses in young adolescents aged 9–14 years compared with three doses as in the original efficacy trials.30

Studies on Gardasil, Cervarix, Gardasil 9, and Cecolin have been conducted, finding non-inferiority with two doses administered at intervals of 6 or 12 months.31–34 In general, non-inferiority criteria were met even with some geometric mean antibody titers higher among those aged 9–14 years who received two doses compared with controls in women aged 16–26 years who received the three doses.31 Most outcomes were assessed at 1 month follow-up after the last dose, and doubts remain regarding reduced dose schedules for older groups and whether just one dose is sufficient for obtaining high and long-term immunity. Post hoc evaluation of RCTs comparing the use of one, two, or three doses35 has shown that the efficacy is similar, and the antibody titers are high even after more than 10 years of follow-up.36

Recently, a RCT assessing single-dose HPV vaccination in young African women was published.37 The investigators used nonavalent and bivalent vaccines and were compared with meningococcal vaccination in women aged 15–20 years. The primary outcome was incident persistent vaccine-type HPV infection at 18 months. For incident persistent infections of HPV 16/18, the efficacy of the nonavalent vaccine was 97.5% (95% CI 81.7% to 99.7%; p=0.0001), the efficacy of the bivalent vaccine was 97.5% (95% CI 81.6% to 99.7%; p=0.0001), and the efficacy of the nonavalent vaccine for HPV 16/18/31/33/45/52/58 was 88.9% (95% CI 68.5% to 96.1%; p=0.0001), showing the promising outcomes for single-dose vaccine program development.

Cancer as an Outcome

As cervical cancer cannot be assessed in clinical trials, given ethical limitations, population-based studies have recently been reported assessing the long-term outcomes of cohorts of vaccinated and non-vaccinated women. An observational study conducted in the UK38 using data from more than 13 million years of follow-up of women aged 20–30 years showed that the relative reduction in cervical cancer rates by age at vaccine offer was 34% (95% CI 25% to 41%) for women aged 16–18 years, 62% (95% CI 52% to 71%) for those aged 14–16 years, and 87% (95% CI 72% to 94%) for those aged 12–13 years compared with the reference unvaccinated cohort.

In Sweden, using health registries to follow a population of 1 672 983 girls and women aged 10–30 years from 2006 through 2017, an observational retrospective study39 found that the cumulative incidence of cervical cancer was 47 cases per 100 000 persons among women who had been vaccinated and 94 cases per 100 000 persons among those who had not been vaccinated and, after the adjustment for all covariates, the incidence rate ratio was 0.12 (95% CI 0.00 to 0.34) among women who had been vaccinated before the age of 17 years and 0.47 (95% CI 0.27 to 0.75) among women who had been vaccinated at the age of 17–30 years.

Even when RCTs will never be able to show cervical cancer reduction as an outcome, the information provided by observational population studies gives us encouraging data for the expected outcomes of vaccination programs.

Vaccination after Intra-epithelial Lesion Excisional Treatment

As presented before, since the initial RCTs, it was evident that HPV vaccines do not have a therapeutic effect and do not prevent the progression of infection present at the time of vaccination. However, using the HPV vaccine after surgical treatment of patients with CIN remains controversial in our field. A non-randomized intervention study40 reported a possible effect on recurrence rate after Loop Electrosurgical Excision Procedure (LEEP) treatment for CIN 2, CIN 3, and microinvasive 1A1 cervical cancer. In total, 536 patients were initially considered for the study and 248 patients received quadrivalent vaccination. Disease recurrence, defined as CIN 2+, was observed in 6.4% of cases compared with 1.2% in the vaccinated cohort (p=0.0112). However, this study has several limitations that compromise confidence in the results. The selection of the intervention arm was not randomized; there was a high rate of positive margins in both arms (16% and 19%), the inclusion of microinvasive disease and, most importantly, the high rate of patients lost during follow-up or with short follow-up time (29% in the vaccinated cohort and 36% in the non-vaccinated cohort). Other studies have replicated similar results but with similar limitations, such as a high rate of positive margins and differences in recruiting times.41 42

Most systematic reviews that assessed the possible impact of vaccination after surgical treatment are, unfortunately, limited by the inclusion of patients who received vaccination before treatment, like the populations from initial vaccine RCTs and the combination of patients from RCTs with observational studies, which limits the interpretation of the information and represents relevant bias.43 44

So far, two RCTs have been published assessing the use of the HPV vaccine after conization, including patients with CIN 1, CIN 2, and CIN 3 diagnoses and using the quadrivalent vaccine.45 46 They were included in a recent meta-analysis of 420 patients.47 A reduction in the risk of CIN recurrence was reported (OR 0.29, 95% CI 0.16 to 0.53); however, the evidence is limited by the number of included patients, the quality of the trials, and the inclusion of CIN 1 lesions.

Currently, two larger RCTs known as the NOVEL trial (NCT03979014) and the HOPE 9 (NCT03848039) are in progress. The NOVEL trial will assess if three doses of 9-valent vaccine after surgical treatment for high-grade CIN reduce the incidence of HPV infection during follow-up; 1000 patients are expected to be recruited . It is possible that CIN recurrence outcomes will be evaluated. The HOPE 9 trial will assess the efficacy of 9-valent HPV vaccination in preventing recurrence of CIN 2+ in participants treated for high-grade CIN with LEEP. Three doses will be administered: the first dose before surgery, the second dose on the day of surgical treatment (2 months later), and the third dose 6 months after the first dose at the first follow-up visit. The total sample size is estimated to be 1220 patients.

Other trials are evaluating different vaccine interventions, including the evaluation of single-dose schemes in children and adolescents (NCT03832049, NCT03728881) and the use of an 11-valent vaccine in adults aged 18–45 years (NCT05262010).

Strategies to Improve Vaccine Programs

Some developed countries have achieved a relatively high uptake of HPV vaccination through the inclusion of the vaccine in national vaccination programs, but the uptake of HPV vaccination has been sub-optimal in many developed countries, as it is in most low to middle income countries.48

In a survey, parents whose children had not been vaccinated and reported it was not likely that they would initiate vaccination next year were interrogated. They gave reasons like safety concerns (the most common reason), lack of necessity, knowledge, and recommendation.49 This highlights a lack of understanding of HPV vaccination evidence and benefits on the part of the parent or caregiver, and the critical role of the healthcare provider in consistently educating parents or caregivers about vaccination as part of the WHO 90-70-90 strategy to eradicate cervical cancer worldwide as some studies have identified an association between direct provider recommendation and vaccine receipt.50 Community-based interventions to improve HPV vaccine coverage include patient reminders, inclusion as part of multiple vaccine administration, school-based vaccination programs, physician interventions, and social media strategies, with some promising results.51

Secondary Prevention of Cervical Cancer

Cervical screening programs aim to detect cervical cancer precursor lesions to provide appropriate follow-up and treatment. Globally, population-based (organized), non-population-based (unorganized) screening and even opportunistic screening programs have been implemented. Therefore, the coverage and success rate of screening varies between countries.52 53 Since the implementation of high coverage screening programs with conventional cytology, there has been a decrease in cervical cancer incidence and mortality, preventing between 50% and 70% of the expected cases.53–55 Additionally, some countries have included liquid-based cytology to increase the sensitivity of detection of CIN 2+,56 57 to reduce cervical cancer incidence and mortality.53 However, cytology has some significant limitations as a screening test, such as the moderate sensitivity for detecting high-grade CIN and the low reproducibility due to the subjectivity of the test.58

Low-cost alternatives to cytology screening called visual inspection with acetic acid (VIA) and visual inspection with Lugol’s iodine (VILI) have been tested in low- and middle-income countries. Both have shown promising results in underserved communities because they are based on a strategy called ‘see and treat’ to reduce the number of visits while increasing the coverage of the screening.59

Several clinical studies60–63 have determined that the implementation of HPV-based screening improves the sensitivity of detecting high-grade pre-invasive lesions and concluded that a negative result for HPV infection offers greater safety in predicting the risk of developing CIN 3 within 3 years.64 Currently, several countries have implemented primary screening with HPV testing52 58 65 and several international associations recommend it,64 66 67 recognizing that HPV infections in young women are usually transient.64 68

VIA and VILI

Visual methods (VIA and VILI) are based on the concept that either pre-invasive or invasive cervical lesions can be seen with the naked eye, without magnification, with the aid of a light source. The advantages of visual methods are that they can be taught to health workers, midwives, nurses, and paramedical workers in hands-on courses that can take 4–10 days and with the help of manuals specially created for teaching; they do not require a laboratory infrastructure; acetic acid and Lugol’s iodine are widely available and are inexpensive; and immediate results can be interpreted, allowing ‘see and treat’ strategies, providing proper treatment in a single visit, usually with cryotherapy, once an invasive lesion has been ruled out.69

Visual inspection techniques have been evaluated mainly in a low resource setting in developing countries. A positive result to VIA from International Agency for Research on Cancer criteria was defined as the presence of well-defined acetowhite areas in the transformation zone (near the squamocolumnar junction or external os) or the presence of an acetowhite growth about 1 min after direct application of a 3–5% diluted solution of acetic acid using a cotton swab or a spray. A positive result of VILI was based on the appearance of a definite yellow area in the transformation zone, close to the squamocolumnar junction or the external os, or on a cervical growth, after application of a 5% solution of Lugol’s iodine.68 70 71

In a RCT conducted in India, a single round of VIA screening was associated with a significant 25% reduction in cancer incidence and a 35% reduction in cancer mortality due to cervical cancer. A stratified analysis by age indicated that there was a significant 38% reduction in incidence and 66% reduction in mortality in the 30–39 year age group and a significant 45% reduction in mortality in the 40–49 year age group, but there was no significant reduction in incidence and mortality in the 50–59 year age group.72 According to a meta-analysis, the sensitivity and specificity of VIA were as high as 92%.73 The sensitivity of VILI varied from 44% to 92% and specificity from 75% to 85% in cross-sectional studies in India, Africa, and Latin America.74–76

Validated HPV Tests for Screening

While cervical cancer screening with HPV testing has demonstrated better risk estimation on the development of CIN 3,64 the use of validated tests is essential to ensure the effectiveness of the screening program.77 Poljak et al78 reported that 254 different commercial HPV tests and at least 425 variants of those tests were available on the global market in 2020, representing a 31% and 235% increase in the number of tests and variants, respectively, from those reported in 2015. However, 82% of the tests lack peer-reviewed publications about analytical and/or clinical evaluations, and more than 90% do not fulfill the consensus requirements to guarantee their use in a clinical setting.

The following requirements must be fulfilled to assure the quality of primary cervical screening: (1) clinical sensitivity for CIN 2+ not less than 90% compared with the reference standard in women at least 30 years of age; (2) clinical specificity for CIN 2+ not less than 98% compared with the reference standard in women at least 30 years of age; (3) intra-laboratory reproducibility and inter-laboratory agreement with a lower confidence limit of not less than 87%, determined by evaluation of at least 500 samples.79 Validation of HPV tests should be performed according to the Meijer guidelines79 or the VALGENT protocol (Validation of HPV genotyping tests),80 and should demonstrate non-inferiority sensitivity and specificity to detect CIN 2+ compared with two reference standards (Hybrid Capture 2 HPV DNA test (HC2) or GP5+/6+PCR EIA).79 80 Since GP5+/GP6+PCR EIA is no longer used and HC2 has decreased its use in screening programs, COBAS 4800 has been considered an alternative standard for comparison.81

In 2020, of the approved and validated HPV tests for primary cervical screening, seven were approved by the FDA, 13 by the Meijer guidelines, and 22 by the VALGENT protocol.78 According to Arbyn et al,81 11 high-risk HPV DNA assays fulfill all requirements for use in cervical cancer screening using clinician-collected specimens (Table 3). Other tests show partial validated status. The mRNA test (APTIMA HPV assay, Hologic, USA) met the Meijer criteria for a high-risk HPV DNA test, but longitudinal relative performance indicators are imprecise, heterogeneous, and based on few studies.81 According to recent data, the mRNA test shows similar cross-sectional and longitudinal sensitivity and slightly higher cross-sectional specificity for CIN 2+ and CIN 3+ compared with fully validated high-risk HPV DNA tests on clinically collected samples.82

Table 3

Fully validated HPV tests for cervical cancer screening for CIN2+ detection81 82

Triage of HPV-Positive Women

High-risk HPV tests are characterized by high sensitivity, thus they also detect transient infections that may be clinically insignificant. It is important to incorporate triage testing to compensate for the low specificity of HPV tests and to decrease unnecessary follow-up of positive women for transient HPV infections, maximizing the benefits of primary screening.83 There are several screening strategies with different characteristics, such as reflex cytology, HPV genotyping, p16 immunostaining, and methylation.53 66 68 77 84

Reflex cytology is recommended for all HPV-positive primary screening results due to the high-grade evidence of suitability, improving pre-invasive lesion detection rates by up to 30%.66 77 84 HPV genotyping is a useful strategy for risk stratification according to genotype.77 84 In women with HPV16 persistent infection, the 7-year cumulative risk of CIN 3+ is 21.5%, followed by 20.9% and 11.0% for genotypes 33 and 18, respectively. Although HPV33 shows a high cumulative risk, it is less frequent than types 16 or 18.85

p16ink4a is a cyclin-dependent kinase inhibitor markedly overexpressed in pre-invasive and invasive cervical lesions. p16 immunostaining alone or in dual staining with Ki67 could be a useful option for the triage of HPV-positive women.68 84 p16/Ki67 dual stain showed better risk stratification for CIN 3+ than cytologic triage, decreasing the number of colposcopy referrals.84 86

DNA methylation of HPV and host genes increases with lesion severity, thus being a promising triage strategy for high-risk HPV-positive women to predict CIN 2+.87 Approximately 10 human genes show high methylation in pre-invasive lesions and hypermethylation in invasive lesions, including CADM1, EPB41L3, FAM19A4, MAL, miR-124, PAX1, and SOX1 88 and the L1/L2 HPV genes of genotypes 16, 18, 31, 33, and 45.84 88 Methylation tests are at an early stage, but show great potential as an accurate molecular classifier. Recent studies on screening populations and colposcopy using different human gene targets showed sensitivity ranging between 69% and 83%, with specificity (for CIN 2+) ranging between 66% and 76%.87 88 Optimal sensitivity rates for CIN 2+ are achieved when several methylation markers are combined, therefore the development of multiplex polymerase chain reaction (PCR) methylation tests is an opportunity to improve the clinical sensitivity and decrease the detection time.89

Primary Screening in Vaccinated Population

The population impact of the HPV vaccine has been demonstrated, both in the reduction of pre-invasive lesions and cervical cancer.88 In addition, the prevalence of vaccine-preventable high-risk genotypes has decreased significantly,90 91 affecting the positive predictive value (PPV) of HPV tests based on genotypes 16 and 18.88 The PPV of cytology as primary screening in a group of non-vaccinated women was 21.3% compared with 17.4% in vaccinated women, being lower in the fully vaccinated group, and a prevalence of CIN 2+ of 1.6% and 1.9%, respectively. When stratified by age, PPV was 11.9% in women who received the vaccine before age 21 years and 30.7% in those who received the vaccine at age 21 years or older.92

According to a mathematical model that evaluates the health and economic outcomes of primary screening in vaccinated women with a complete schedule of either vaccine, the optimal strategy for adolescents who received all three doses of bivalent or tetravalent vaccine would be screening every 5 years starting at age 25 or 30 years with cytology or an HPV test. For those who received the complete doses of the nonavalent vaccine, the best option would be to extend screening to 10 years starting at age 30 or 35 years, with an HPV test.93

Regardless of the shift in screening programs for vaccinated populations, it is essential to establish new guidelines due to the significant negative impact of false positive results in managing these patients.92 According to the European Society of Gynaecological Oncology-European Federation of Colposcopy (ESGO-EFC) guidelines, primary cervical screening should be performed by group: (1) cervical screening in birth cohorts not offered vaccination; (2) cervical screening in unvaccinated women within vaccinated cohorts; (3) women known to have been vaccinated; (4) women in a partially vaccinated cohort.84

Self-Sampling for HPV Testing

One of the most important challenges in achieving 70% of cervical screening coverage is to include women with inadequate access to preventive care, cultural and personal barriers, language differences with the healthcare provider, low educational level, and discomfort with pelvic examination.94 The implementation of screening programs with vaginal self-sampling could be a safe and accessible option to reach women who do not participate in organized screening programs.95 96 As reported by Serrano et al,96 only 48 of 139 (35%) countries and territories have implemented primary HPV-based screening and, of these, 17 have introduced self-sampling in their guidelines for under-screened women or as a primary screening option.

Two meta-analyses evaluated the accuracy of self-samples for HPV detection as primary screening with respect to clinician samples. Both studies agree that signal amplification tests done with self-samples are up to 15% less sensitive and less specific compared with clinician samples, and even with self-samples done with PCR-based tests.95 97 Despite these differences, self-sampling is a cost-effective program that can increase primary screening coverage in low- and middle-income countries where the incidence of cervical cancer is high.95 96 98 The response to self-sampling varies according to the invitation scenario, with community campaigns and the door-to-door strategy showing the highest relative participation.95

Mailing self-sample kits is an effective strategy for under-screened women.95 The Prospective Evaluation of Self-Testing to Increase Screening (PRESTIS) trial (NCT03898167), a hybrid type 2 effectiveness-implementation pragmatic RCT, is ongoing. Its primary outcome is to evaluate the effectiveness of mailing self-sample kits, and the secondary outcome is the follow-up of patients with a positive screening test to provide meaningful data to inform the equitable delivery of screening to achieve the global goal of eliminating cervical cancer.94

Another option for self-sample collection is the detection of HPV in urine samples, which is a more accessible and acceptable method. A systematic review and meta-analysis suggest that urine testing with an appropriate HPV test may be an alternative for primary cervical screening. The pooled sensitivity and specificity for urine and clinician samples was 78% (95% CI 0.56 to 0.95); 87% (95% CI 0.83 to 0.89) and 97% (95% CI 0.93 to 1.00); 85% (95% CI 0.79 to 0.91), respectively, highlighting that signal amplification and mRNA tests were less sensitive. Although post-statistical analyses of the sampling device or urine collection time are not factors that significantly affect test performance, first-void urine and Colli-Pee showed higher detection accuracy than other conditions.99 The Validation of Human Papillomavirus Assays and Collection Devices for Self-samples and Urine Samples (VALHUDES) trial (NCT03064087) evaluates the HPV test accuracy for both urine and vaginal self-samples and the standardization of the sampling device for primary cervical screening.100 101