Predictors of CIN Study Series 5.0:Stability Study in Vaginal Self-samples

Background Cervical cancer is caused by persistent infection with high risk human papillomavirus types (Hr-HPV) but can be prevented if detected at an early stage. Prevention through routine cervical screening is traditionally done with liquid-based cytology sampling where the cells collected by clinicians are examined to identify abnormal features. Another method is HPV testing performed on the same liquid-based samples as for cytology. HPV DNA testing is more sensitive, but less specific, than cytology for cervical screening and both tests are broadly acceptable to women (Forrest, McCaffery et al. 2004; Waller, McCaffery et al. 2006; Szarewski, Cadman et al. 2009). Although clinician-taken samples are the gold standard, self-sampling may be a useful alternative. HPV DNA testing using self-collected (SC) samples has been shown to have acceptable sensitivity and specificity for HPV testing compared to samples taken by clinicians (Petignat, Faltin et al. 2007; Szarewski, Cadman et al. 2007; Arbyn, Verdoodt et al. 2014). A high level of concordance between self and clinician sample for HPV DNA detection of 0.87 (95% Confidence Intervals (CI), 0.82-0.91) was shown in Petignat's systematic review and meta-analysis (Petignat, Faltin et al. 2007). HPV DNA testing is well established but newer HPV testing technologies are available and emerging.

HPV DNA testing has most commonly been performed on cervical samples collected in a liquid medium (generally Specimen Transport Medium (STM) or PreservCyt). However, this is not ideal for home SC samples because there are often restrictions on postage of biological samples in liquid, risks (although low) pertaining to the transport medium itself and issues around refrigeration of the liquid. The removal of the necessity for a liquid transport medium (and especially any subsequent refrigeration) would greatly improve the accessibility of SC sampling and reduce costs. This would also increase the potential for cervical screening in those areas where there may be logistical difficulties such as lack of regular, frequent postal collections or refrigerated storage facilities.

There have been several studies which have investigated dry sample collection for HPV DNA detection (Shah, Daniel et al. 2001; Krech, Castriciano et al. 2009; Feng, Cherne et al. 2010; Cerigo, Coutlee et al. 2012; Darlin, Borgfeldt et al. 2013; Eperon, Vassilakos et al. 2013). However, the differing methodologies and requirements for refrigeration have restricted the generalisability of the results. They have also not addressed the question of how long the sample on the dry collection device would remain viable, especially under higher temperatures.

Rationale and Risks/Benefits Evidence increasingly suggests that, at least in temperate climates, HPV DNA testing from a SC sample using a dry swab provides similar sensitivity to a wet swab (where the SC sample is placed in liquid specimen transport medium) (Wolfrum, Koutsky et al. 2012). Dry transport could make HPV DNA testing from SC samples more accessible and widely available. In the UK for example, postal packaging is simplified when no liquid is involved. HPV primary screening is being explored in clinical trials and even being introduced into some screening programmes. In the Netherlands for example SC HPV DNA testing within the primary screening programme is planned for introduction for non-attenders in 2016 (Rijksinstituut voor Volksgezondheid en Milieu 2014).

The rationale of this stability study is to ascertain whether SC samples kept under dry conditions will still be useful if left for a period of time in warm conditions. This would be of particular use in the many low and middle income countries where cervical cancer incidence and mortality are highest. These countries have frequently found it very difficult to set up national cervical screening programmes due to lack of infrastructure, access to efficient refrigeration and a postal service.

Bart's Health National Health Service (NHS) Trust serves a diverse population from a range of socio-economic and ethnic backgrounds in inner London. The colposcopy clinic (which was at Bart's Hospital but relocates to Royal London Hospital in spring 2015) has approximately 1050 new women attending and 1900 women attending for follow-up per year. The participation in this project will have no implications for the ongoing care of any woman who agrees to participate. In previous studies and in other United Kingdom (UK) based screening programmes in the UK, such as for Chlamydia trachomatis, the use of swabs by women themselves has been shown to be achievable and safe.

Histology and cytology results from within three months prior to the visit and up to six months post-visit will be collected. There may be a difference in results depending on the grade of disease and this will be confirmed or excluded with this information.

Trial design This study is a small preparatory project aiming to enrol 60 women from a colposcopy service and comparing their SC samples taken with two dry flocked swabs and the CE marked device HerSwabTM (Eve Medical, Toronto, Canada). Each woman will take all three samples. The swabs will all be refrigerated immediately and then will be either frozen immediately at -20ºC (SR0) to preserve the sample or incubated at 25ºC for either one (SA1) or two weeks (SA2). HerSwab samples will be included in the incubation part of the study (HA1 and HA2). The order of sampling will be randomised to remove any bias associated with testing order and roughly equal numbers of first, second and third swabs will be allocated to each swab management regimen. The study will precede and inform a multi-phase study with the overall aim of comparing SC HPV testing using different devices, under different conditions and using different HPV testing assays.

Feasibility In the period 1st January 2013 to 31st December 2013, 1052 new patients attended the Bart's colposcopy unit. In the same time period 1930 attended for follow up colposcopy or treatment, This would mean that over a 12 month period 2982 women attended colposcopy. There are therefore sufficient numbers of women who can potentially be recruited to the study. The service will still receive these referrals when it relocates to the Royal London Hospital.

Randomisation Procedures Randomisations will be carried out using an in-house computer application. The Centre for Cancer Prevention will generate a randomised list linking participant numbers (PNOs) with a pack number and a code identifying the order of sampling. The order will be randomised 1:1:1 to dry flocked swab (for freezing on arrival in the laboratory) (SR0), dry flocked swab (for incubation at 25ºC for one (SA1) or two weeks (SA2) on arrival in the laboratory), or HerSwab (for incubation at 25ºC for one (HA1) or two weeks (HA2) on arrival in the laboratory) (please refer to Study Scheme Diagram). PNOs will be assigned to numbered packs containing three self-sample kits, labelled to show the order in which the samples should be taken. 60 women consenting to take part will be given the next pack in sequential order as far as possible as they attend the clinic and consent to the study. All samples will be immediately refrigerated at 2-8ºC in clinic until transfer to the laboratory.

Schedule of intervention for each visit Assessment Visit 1 Three self-samples taken x

Study devices Women will be asked to self-collect three vaginal samples using two dry flocked swabs and the CE marked device HerSwabTM (Eve Medical, Toronto, Canada). HerSwab utilises a flocked swab contained within a plastic applicator, with the intention of making self-sampling easier for women.

Procedure for Collecting Data Cervical or vaginal cytology and histology data will be collected from the referral cytology result, which led to their attending the colposcopy service, up to six months post appointment date. All study data entry and management will be undertaken at the Centre for Cancer Prevention.

All HPV, cytology and histology results will be entered onto, or imported into, the study's ORACLE database by the Centre for Cancer Prevention (CCP) study team. All databases are protected and only accessible to named staff.

Consent forms will be securely stored at CCP.

Laboratory Assessments Extraction and detection of Human DNA by a quantitative polymerase chain reaction (qPCR) assay from all samples.

Extraction and detection of HPV DNA (primarily HPV 16) by a qPCR assay from all samples.

Validated HPV test from all samples yielding CT or RLU values.

End of Study Definition The end of study would be defined as when the final histology and cytology data were collected for the last participant recruited. This would be approximately eight months after the date of visit of the last recruit. Eight months would allow time for the data to be entered onto the hospital database.