Predictive Value of HPV Testing in Self-collected and Clinician-Collected Samples Compared with Cytology in Detecting High-grade Cervical Lesions

Abstract

Background: Self-sampling has become an attractive proposition now that human papillomavirus (HPV) primary testing is being incorporated into cervical cancer screening programs worldwide. We compared predictive values of HPV testing based on self-and physician-collected samples, and cytology, in detecting high-grade cervical intraepithelial neoplasia (CIN).Methods: The Cervical And Self-Sample In Screening (CASSIS) study enrolled 1,217 women ages 16-70 years prior to scheduled colposcopies. Vaginal specimens were self-collected using the validated HerSwab device. Cervical specimens were collected by gynecologists. Specimens were tested for presence of high-risk HPV (hrHPV) by the Cobas 4800 HPV test. We estimated positive predictive values (PPV) and negative predictive values (NPV) and 95% confidence intervals (CI) for a subset of women (n ¼ 700) who underwent cervical biopsy and cytology at the actual CASSIS visit.Results: hrHPV was detected in 329 women (47%) with HerSwab and in 327 (46.7%) with physician sampling. Respective values for HPV16/18 were 119 (17%) and 121 (17.3%). On histology, 134 women had CIN1, 49 had CIN2, 48 had CIN3, 5 had CIN2/CIN3, and 3 had cancers. PPVs for CIN2 þ of any hrHPV were 28% (95% CI, 23.2-33.1) and 29.7% (95% CI, 24.8-34.9) for HerSwab and physician samples, respectively. Corresponding values for HPV16/18 were 43.7% (95% CI, 34.6-53.1) and 43.8% (95% CI, 34.8-53.1). PPV of cytology (ASC-USþ) was 26.6% (95% CI, 21.6-32.0). Corresponding NPVs (same order as PPVs) were 96.4% (95% CI, 93.9-98.1), 97.8% (95% CI, 95.6-99), 90.9% (95% CI, 88.2-93.1), 91% (95% CI, 88.4-93.2), and 94.7% (95% CI, 91.8-96.8).Conclusions: Our results confirm that HPV self-sampling has comparable performance with a physician-collected sample in detecting cervical lesions.Impact: HPV self-sampling has the potential to increase coverage in cervical cancer screening.

Dual staining for p16/Ki‐67 to detect high‐grade cervical lesions: Results from the Screening Triage Ascertaining Intraepithelial Neoplasia by Immunostain Testing study

“…The Screening Triage Ascertaining Intraepithelial Neoplasia by Immunostain Testing (STAIN‐IT) study was conducted within the CASSIS (Cervical And Self‐Sample In Screening) study; full details of the latter study design, procedures and clinical specimens have been previously described 9,10 . Briefly, 1217 women aged 16 to 70 years were enrolled between June 2015 and April 2016 at three McGill University affiliated hospitals in Montreal.…”

Self-Sampling as a Plausible Alternative to Screen Cervical Cancer Precursor Lesions in a Population with Low Adherence to Screening: A Systematic Review

“…The weakest point of these assessments is the noticeable fluctuation in the average number of women assessed, as these analyses are conducted in numerically unimportant groups, especially in underdeveloped areas. As a result, the data listed in Table 4 clearly present this potential limitation, with some exceptions, namely the works of El Zein et al [8] and Stanczuk et al [17] in developed countries, and Lorenzi et al [18] and Allende et al [19] in developing countries. It is important to emphasize, however, that in the work of Poli et al [16], who studied about 5,000 women in poor urban areas of India, a potential risk of overtreatment was noted using molecular methods associated with self-sampling.…”

“…(7) Was the chronogram sufficient to expect an association between exposure, if any? (8) For exposures that may vary in amount or level, did the study assess different exposure levels in relation to outcome (e.g., exposure categories or exposure measured as continuous variable)? (9) Were exposure measures (independent variables) clearly defined, valid, reliable, and consistently implemented across all study participants?…”

“…(14) Were the main potential confounders measured and statistically adjusted for their impact on the relationship between exposure and outcome? [3] moderate to high Possati-Resende [10] moderate to high Haguenoer [1] high Allende [19] moderate to high Maza [11] moderate to high Esber [9] moderate Gottschlich [23] moderate Laskow [13] moderate Mbatha [14] moderate El-Zein [8] moderate Poli [16] low to moderate Vahabi [7] moderate to high Stanczuk [17] moderate Winer [4] moderate Chatzistamatiou [12] moderate B Requirements for quality assessment of articles included in the CASP checklist systematic review for randomized controlled trials comprised: (1) Did the study address a clearly focused issue? (2) Has the designation of patients for treatment been randomized?…”

“…The methodological premise of self-sampling is simple, since the material is sent to a reference laboratory that will evaluate the presence of HPV. If the test is negative, the risk of developing cervical lesions in the following 5-10 years is low (there is a very high negative predictive value documented throughout various works) [8]. On the other hand, women with positive results should subsequently perform an assessment by the visual acetic acid indicator (VIA) or have a Pap smear, according to available resources [2,3].…”