Abstract
“
The widespread application of the Papanicolaou smear for the prevention of cervical cancer has led to a substantial decrease in the prevalence of the disease. However, the majority of cervical cancer cases (60%) are still occurring in women who experience an absence or deficiency of screening [1]. Therefore, an important immediate gain in cervical cancer prevention could be attained by increasing screening participation among women who are currently either unscreened or screened infrequently, regardless of the test used [1–7].
High-risk human papillomavirus (HPV; hr-HPV) types are associated with over 99% of invasive cervical cancers and, nowadays, detection of hr-HPV DNA is considered the primary screening method [2]. Several methods considered to be clinically validated for use in primary screening have been applied for the detection of hr-HPV DNA, involving either nonamplification, hybridization-based or PCR-based assays 2,3]. Comparative studies have demonstrated that molecular techniques for hr-HPV DNA detection exhibit higher diagnostic sensitivity and reproducibility compared with cervical cytology for detecting cervical intraepithelial neoplasia (CIN) grade 2 (CIN2) or grade 3 (CIN3), which are the high-grade lesion precursors of invasive cervical cancer [3,4]. In addition, a pooled analysis of four randomized controlled trials, SWEDESCREEN, POBASCAM, ARTISTIC and NTCC demonstrated that a lower CIN3 incidence was recorded when HPV DNA testing was applied instead of cytology [2]. Furthermore, in this analysis HPV-based cervical screening versus conventional cytology provided 60–70% greater protection against invasive cervical carcinomas, prompting the authors to recommend the implementation of HPV-based cervical screening with triage from the age of 30 years at intervals of 5 or more years [2]. This age restriction could be anticipated, since hr-HPV testing has higher sensitivity compared with cytology, but exhibits lower specificity. The specificity can be increased by screening women over 30 or 35 years of age and by adjusting the threshold for HPV positivity (e.g., 2 pg/ml instead of 1 pg/ml on the Hybrid Capture 2 assay [Qiagen, MD, USA]) [3]. Furthermore, it is the high-negative predictive value of HPV testing that allows the safe increase of screening intervals to at least 3–5 years for women who test negative. Still, the enhanced sensitivity of HPV testing and the potential for application of high-throughput automated methods, makes it an attractive alternative to cytology. Further to its role in cervical cancer screening, hr-HPV testing is valuable in the triage of women with borderline cytology, aiming to reduce those who are referred to colposcopy, in addition to monitoring women following treatment of high-grade CIN, as it is better at predicting recurrent disease than cytology and colposcopy [4].
At present, all the above applies mainly to cervical sampling, which is the standard choice in hr-HPV DNA detection. However, it requires a gynecological examination to enable the physician-obtained cervical sampling. It should be noted that, as recently highlighted in a European population, the most common reason claimed by women for not attending screening was that they felt “uncomfortable with vaginal examination.” This evidence suggests that self-sampling, which bypasses this internal examination, should be further explored as a strategy to increase the coverage of cervical screening and follow-up programs [5]. In that respect, self-collected urine, vaginal, rectal and pharyngeal samples provide a different, cost-effective option and have been extensively investigated for screening outside the traditional clinical setting [5–13]. The majority of those studies aimed to: explore the specificity and sensitivity of self- compared with clinician-collected samples; to investigate their potential to increase population coverage, including those unwilling to provide samples in traditional settings; and to identify the ideal collection method, kit and molecular assay for HPV detection.
It has recently been shown that, when appropriate instructions are given, self-sampling of specimens leads to similar results in the subsequent molecular HPV detection, compared with those collected by healthcare professionals. In particular, a recent meta-analysis clearly suggests that self-sampling molecular HPV testing is indicated for screening women who are at risk for cervical lesion development, as it increases population coverage [7]. Furthermore, hr-HPV testing on self-samples appeared to be at least equally, if not more, sensitive for CIN2 or worse compared with cytology on clinician-obtained cervical samples; although, often less specific [8]. Overall, according to our experience and several other reports, self-sampling, irrespective of the collection method used, has been shown to be an appropriate alternative to physician-collected cervical sampling, particularly in underserved women [8–13]. Vaginal samples, which represented the self-collected option mostly applied in the above studies, are usually acquired by the women after they have been instructed to insert the swab or brush into the vagina, to rotate it three times in both directions and to place it in one of the used self-sampling kits.
As far as the performance of different sampling sites and collection strategies, it has been previously reported, in a nonpregnant population, that hr-HPV detection using self-collected vaginal samples yields highly concordant results with those using physician-collected cervical samples [8–14]. In some cases, somewhat lower sensitivity has often been observed [10–14]; however, recent studies applying PCR methods have shown similar sensitivity to that of physician-collected cervical samples [8,15]. Furthermore, promising results are currently been reported regarding HPV detection in the antenatal clinic [16]. Overall, the existing evidence currently suggests that cervical cancer screening conducted by hr-HPV detection on vaginal self-samples exhibits a somewhat lower or comparable sensitivity with clinician samples, but superior sensitivity compared with cytology on clinician samples.
It is evident that the self-sampling method, storage conditions, sample preparation, DNA extraction and the molecular method applied for the vaginal HPV DNA detection may all have an important impact on HPV DNA detection, as different sensitivities were reported with different assays. In particular, Belinson et al. reported that self-collected vaginal specimens showed lower sensitivity and specificity than physician-collected endocervical specimen analyzed with the nonamplification, hybridization-based method; however, using more sensitive PCR-based assays could improve the sensitivity to comparable levels [17,18]. In another recent review among different HPV DNA test assays used, Hybrid Capture 2, Aptima (Gen-Probe Inc., CA, USA), Cervista (Third Wave Technologies, Hologic, Inc., MI, USA) and certain PCRs were less sensitive on self-samples; however, it should be kept in mind that the variations observed in clinical performance probably reflect the use of different combinations of collection devices and HPV tests [8].
“Further standardization and optimization of vaginal human papillomavirus self-sampling, and detection techniques in the future might enable its effective incorporation in cervical cancer screening and premalignant disease triage, and follow-up.”
Given the need to increase the proportion of women participating in hr-HPV screening programs, considerable efforts are focused on the incorporation of noninvasive self-sampling methods not only in vaginal but in urine samples as well, as reported by several groups previously [11,12,19,20]. Urine samples are acceptable in clinical practice for the screening of chlamydial and gonorrheal infections. However, hr-HPV DNA detection in urine exhibits considerable disadvantages, the most important being that the sample is not collected at the initial disease site and only comprises spontaneously exfoliated cells. Recent publications have suggested that detecting HPV DNA in urine may be achievable and could become a useful tool; however, it requires further improvement and standardization of the corresponding technology before recommending it for implementation [19,20]. Owing to the fact that urine samples can be obtained easily via a noninvasive self-sampling method, potentially optimized tests of urine for HPV DNA detection could become valuable approaches for further investigations in the cervical cancer screening field. Last but not least, self-sampling collection methods for HPV detection have also been applied in male populations, and it was concluded that they exhibit optimal acceptability and feasibility for use in large epidemiologic studies or surveillance efforts [21].
In the recent era of financial restrictions, there are, in almost every public health system, limited resources, often making it hard to cover the high costs of cytology-based screening and of appropriate numbers of qualified healthcare professionals. Self-sampling has the potential to overcome some of these limitations, since it can use community-based organizations, street outreach settings and the internet. Self-sampling organized through the internet can reduce the resources allocated for visits to qualified healthcare professionals. Furthermore, self-sampling is a key tool in order to test women for HPV infection, particularly in developing countries or remote areas, as it would be possible to send these women kits containing brushes, medium collection supplies and illustrative instructions [22]. Such self-collected vaginal samples could subsequently be tested in central laboratories in a costly manner. In fact, a US study already found that the lifetime cost of in-home self-collection for the detection of hr-HPV followed by in-clinic cytology triage was slightly lower than clinician-based screening [23].
As noted above, there is one main concern with molecular HPV testing, applying to all sampling sites, which relates to its lower specificity as it cannot separate transient from persistent infections, with only the latter being associated with increased risk of high-grade CIN and cancer. Women who test hr-HPV positive on their self-sample are usually triaged to cytology, which, as mentioned, is quite expensive [24]. It has been reported that the reduction in specificity associated with self-sampling and every hr-HPV testing could be tackled by the application of appropriate molecular triage methods, which can be applied in the same self-sampled specimens [24].
In conclusion, self-sampled vaginal HPV DNA detection is a woman-friendly acceptable additional tool that could be implemented to increase participation to cervical cancer prevention programs. This has been already suggested for nonresponders and special populations; however, we believe that it could very well be applied on a large scale, owing to its cost–effectiveness. Further standardization and optimization of vaginal HPV self-sampling, and detection techniques in the future might enable its effective incorporation in cervical cancer screening and premalignant disease triage, and follow-up.
Financial & competing interests disclosure
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.