key: cord-0859146-3qd3ilzt
authors: Wang, Runzhi; Lee, Kristen; Gaydos, Charlotte A.; Anderson, Jean; Keller, Jean; Coleman, Jenell
title: Performance and Acceptability of Self-Collected HPV Testing Among Women Living with HIV
date: 2020-07-30
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2020.07.047
sha: e9b19aa292d7655703732b631c014da6d41cc478
doc_id: 859146
cord_uid: 3qd3ilzt

Abstract Objective To assess the validity, reliability, and acceptability of self-collected HPV tests in U.S. women living with HIV (WLHIV). Methods WLHIV ≥ age 30 years underwent self-collected (clinic and home) and clinician-collected HPV tests. Sensitivity and specificity analyses were performed using clinician-collected HPV test as the comparator. Unweighted kappa statistic evaluated validity and reliability of self-collected HPV test, and level of agreement between clinician-collected mRNA test and an DNA test that was used for routine care. A 13-question survey assessed acceptability. Results Among 70 participants, median age was 50, 75% had an undetectable HIV RNA, and 11% had a CD4 count <200. Nearly 63% had at least one positive HPV test. The sensitivity and specificity for the self-collected HPV test were 84.6% (95% CI: 65.1%, 95.6%) and 62.9% (95% CI: 44.9%, 78.5%), respectively; κ = 0.5 (95% CI: 0.2, 0.7). The agreement between the two self-collected tests was good (κ = 0.8, 95% CI: 0.5, 1,0). There was good agreement between clinician-collected mRNA tests and DNA tests (κ = 0.8, 95% CI: 0.7, 1.0). Self-collection was highly acceptable. Conclusions Among WLHIV, self-collected HPV tests had good sensitivity and moderate specificity compared to clinician-collected HPV tests. The reliability between self-collected testing locations was good. Self-collected HPV testing had high acceptability.

High-risk human papillomavirus (HPV) causes almost all cervical cancers (Walboomers et al. 1999) . The American Cancer Society estimates 13,170 new cervical cancer cases in the United States (U.S.) in 2019 and 4,250 deaths (Siegel et al. 2019) . Most HPV infections are asymptomatic and can be cleared, but if the infection persists and is left untreated, over time, it can lead to precancerous changes that may develop into cancer.

One of the risk factors for persistence and progression of HPV-related disease is immunodeficiency (IARC Working Group on the Evaluation of Carcinogenic Risks to Humans 2012). Women living with HIV/AIDS (WLHIV) have a 4-fold higher risk of developing cervical cancer than the general population but are under-screened (Engels et al. 2006; Chaturvedi et al. 2009; Oster et al. 2009; Peprah et al. 2018 ). In addition, infection with multiple HPV types is seen more often among WLHIV compared to women without HIV (Levi et al. 2002; Moscicki et al. 2004; Menon et al. 2016) , and faster HPV-related disease progression has been observed (Denslow et al. 2014 ).

J o u r n a l P r e -p r o o f HPV testing is a notable advancement in cervical cancer screening, given its ability to offer greater reassurance of low cancer risk compared to cytology alone. A randomized controlled trial (Ronco et al. 2010 ) comparing conventional liquid-based cytology vs.

HPV-based testing (either alone or combined with cytology) showed that HPV-based screening was more effective in preventing cervical cancer by detecting persistent highgrade lesions earlier. Additionally, HPV testing has a high sensitivity for detecting precancerous lesions (Bulkmans et al. 2007; Naucler et al. 2007; Clad et al. 2011) . A negative HPV test provides greater reassurance of low cervical precancer risk compared with a negative Pap test (Cuzick et al. 2006; Gyllensten et al. 2012; Leinonen et al. 2012; Ogilvie et al. 2012; Malila et al. 2013; Ronco et al. 2014; Wright et al. 2015) .

Furthermore, compared to cytology, HPV testing is less subjective, more reproducible, and needs less training and expertise for users (Cuzick et al. 2006) . The American Society for Colposcopy and Cervical Pathology (ASCCP) and the United States Public Service Task Force (USPSTF) support primary HPV screening as one of the first-line cervical cancer prevention strategies among women older than 25 years or 30 years, respectively (Huh et al. 2015; U.S. Preventive Services Task Force et al. 2018) . Studies have shown that HPV self-collection increases cervical cancer screening participation in healthy women who do not routinely attend traditional cervical cancer screening programs, and the acceptability has been favorable (De Alba et al. 2008; Racey et al. 2013; Nelson et al. 2015; Winer et al. 2016) . Therefore, self-collected samples potentially can increase the uptake of cervical cancer screening in WLHIV who are under-screened by offering screening at primary care sites, HIV specialty clinics, or J o u r n a l P r e -p r o o f non-clinical sites (e.g., home) that do not routinely perform pelvic examinations; however, there are very few studies to support this assertion. Therefore, this study aimed to assess the validity, reliability, and acceptability of HPV self-collected tests in a clinical setting and at home compared to conventional clinician-collected HPV tests in U.S. WLHIV.

WLHIV age 30 years who were undergoing a Pap test with HPV DNA testing as part of routine care were eligible for inclusion. Women who were pregnant, had a hysterectomy, and had genital tract cancer were excluded. Recruitment took place at a large academic multidisciplinary clinic in Baltimore, Maryland. Institutional Review Board approval was obtained.

Participants were instructed on how to perform an unsupervised vaginal HPV mRNA swab (Aptima ® , Hologic, San Diego, CA) collection. At the time of the clinic visit, participants were instructed to insert a cytobrush into the vagina as far as possible, turn the brush five full rotations, and then place the brush into a vial of transport media. Next, participants underwent a pelvic examination, during which clinicians collected a research cervical HPV mRNA swab, and a routine cervical sample for HPV and Pap cotest. The clinical laboratory used an assay that detected HPV DNA (Qiagen ® , Hilden, Germany), although both types of HPV tests are commercially available. Participants J o u r n a l P r e -p r o o f were given a home collection kit that contained instructions on self-collection, one collection brush, storage containers, and a pre-addressed, postage-paid return cardboard envelope. Two weeks after the clinic visit, participants were reminded by a text message or phone call to self-collect at home and to mail the swab back. Women were given a $20 gift card for their participation. Women with a positive result from any HPV test were phoned and advised to discuss the results with their clinicians for further

Adolescents with HIV) were used to identify women who required colposcopy.

Electronic medical records were abstracted to collect age, race, most recent HIV RNA concentration, CD4 T-cell count, antiretroviral therapy use, and smoking status. Due to the large proportion of women of Black race in the HIV clinic, race was classified to Black vs non-Black. In addition, we classified the most recent HIV RNA concentration into two groups: detectable (≥ 20 copies/mL) and undetectable (< 20 copies/mL).

The clinician-collected (CC) cervical sample was tested for HPV DNA using a hybrid capture II (HC2) DNA nucleic acid hybridization assay that detected HPV DNA Types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59 and 68. Positive samples had a relative light unit  1pg/mL. Individual genotyping results were not available. Equivocal HPV DNA results were reclassified as positive (Knoepp et al. 2007 ). For research purposes, the self-collected (SC) and CC samples were tested for HPV using Aptima ® transcription-mediated amplification RNA test that detected E6/E7 messenger RNA of J o u r n a l P r e -p r o o f the same HPV types as HC2 plus type 66. Positive samples had signal-to-cutoff values of ≥0.5. Invalid HPV mRNA were not reclassified as there are no data available to inform the reclassification.

A 13-question survey was administered after self-collection at the clinic to assess the acceptability of self-collection. These questions assessed if patients: 1) understood the instructions; 2) felt comfortable; 3) felt relaxed; 4) felt in control; 5) felt they were taking care of their health; 6) felt convenient; 7) felt anxious; 8) felt embarrassed; 9) felt it was painful; 10) worried that they might do the test wrong; 11) would use the swab again; 12) would recommend the swab to family and friends; and 13) would use the swab in a clinic. A 5-point Likert scale from strongly agree to strongly disagree was used. The survey also queried, "How easy or hard was it for you to collect the vaginal specimen using the brush?" with responses ranging from Very Easy, Easy, Ok, to Hard, or Very Hard.

To evaluate the ability of SC HPV mRNA tests to correctly identify HPV infection, sensitivity and specificity analyses with 95% confidence intervals (CI) were performed over 2×2 contingency tables using the CC HPV mRNA test as the comparator.

Unweighted Cohen's kappa statistic was used to evaluate the level of agreement between SC and CC HPV mRNA tests. McNemar test with Fisher's exact was used to assess the significance of discordance. To examine the test-retest performance of the J o u r n a l P r e -p r o o f SC HPV mRNA tests, first (clinic) and repeated (home) test results were compared using the unweighted kappa statistic. The same method was used to compare the two CC HPV tests (mRNA test and DNA test). Kappa of 0.81-1.00 was interpreted as very good, 0.61-0.80 as good, 0.41-0.60 as moderate, 0.21-0.40 as fair, and <0.20 as poor (Altman 1991) . Based on a one-sided significance level of 0.05 and 80% power, 45 participants were needed to detect a Kappa value > 0.4, given that the expected Kappa was 0.7. Participants with missing data or invalid test results were excluded from individual analyses. Statistics were performed using STATA version 14.1 (Stata Corp, College Station, Texas, USA).

Seventy-five women enrolled and four were excluded based on age or incomplete data.

We also excluded one participant who had invalid results on CC HPV mRNA tests leaving 70 evaluable participants. All of 70 participants also underwent SC HPV mRNA tests at the clinic with 9 invalid samples (13%). Fifty-five participants performed home collection with 7 invalid samples (13%). Any HPV test positivity prevalence was 63% (95% CI: 51%, 74%).

The median age of participants was 50 (interquartile range: 41-56), and over 92% of participants were of Black race. Over half were current smokers (51%), the majority were taking antiretroviral therapy (94%), and 76% had an undetectable HIV RNA J o u r n a l P r e -p r o o f concentration. About 11% of the participants' most recent CD4 T-cell count was <200 (Table 1) .

Using the CC HPV mRNA test as the comparator, the sensitivity for the SC HPV mRNA test was 84.6% (95% CI: 65.1%, 95.6%) and specificity was 62.9% (95% CI: 44.9%, 78.5%) ( Table 2 ). There was a difference in the proportion of positive vs. negative results between these two HPV mRNA tests with marginal significance under the McNemar test (P= 0.049). The agreement between the CC HPV mRNA test and the SC HPV mRNA test was moderate, with a kappa value of 0.5 (95% CI: 0.2, 0.7). In addition, the reliability of the SC HPV mRNA tests (clinic vs. home) was good, with a kappa value of 0.8 (95% CI: 0.5, 1.0). We performed a sensitivity analysis that included invalid test results as negative or positive. The reliability of SC HPV tests (clinic vs. home) showed a good level of agreement if invalid results were included as negative results (= 0.7, 95% CI: 0.5, 0.9), and a moderate level of agreement if invalid results were included as positive results (= 0.5, 95% CI: 0.3, 0.7). The agreement between CC HPV mRNA tests and CC HPV DNA tests was good (=0.8, 95% CI: 0.7, 1.0).

Sixty-one participants (87%) completed the survey. About 79% of women thought selfcollection was easy. No participants described self-collection as hard or very hard.

Almost all the women (97%) agreed or strongly agreed that they understood the selfsampling instructions. Less than a third (27%) of participants felt anxious about self-J o u r n a l P r e -p r o o f sampling and felt it was painful (24%). Overall, 79.0% agreed that they would use the swab again. An overwhelming majority (90%) agreed that self-sampling was convenient and comfortable. Additionally, 86% of participants would recommend the swab to family and friends. (Figure 1) 

Our study found 63% of WLHIV with HPV infection, which is similar to other studies among WLHIV (Kojic et al. 2011 ) and much higher than women without HIV (2.5%-4.2%) (Sargent et al. 2008) . HIV induced immunodeficiency is thought to impede HPV clearance, resulting in the persistence of HPV infection, along with an increased risk of cervical cancer and pre-cancer (Denny et al. 2012) . Thus, developing a widely acceptable and accessible screening strategy and guideline to optimize cervical cancer prevention among WLHIV is crucial.

Compared to the clinician-collected HPV test, we found that self-collected tests had good sensitivity, moderate specificity, and moderate agreement. This finding is in agreement with one previous study that showed moderate concordance of self-collected vaginal samples compared with clinician-collected cervical samples (Cho et al. 2019) . In our study, the test-retest performance of the self-collected strategy was good and the overall performance of self-collected tests was acceptable, although it was still inferior to tests using clinician-collected cervical samples. Findings from several studies also revealed similar concerns. A large population-based cervical cancer screening study in China showed self-collected HPV testing sensitivity and specificity for detecting precancerous lesions was 86.2% and 80.7%, respectively; whereas clinician-collected J o u r n a l P r e -p r o o f HPV testing sensitivity and specificity was 97% and 82.7% (Zhao et al. 2012) . Results from a meta-analysis of 36 studies examining self-collected versus clinician-collected samples ) also reported lower pooled sensitivity and specificity of HPV testing on self-collected than clinician-collected samples.

Our study had a higher than expected rate of invalid testing in the self-collected samples compared to clinician-collected samples. However, the level of agreement between the first (clinic) and repeated (home) test results of self-collection was good, despite the invalid test results. These invalid test results may have been due to insufficient sample material (Engesaeter et al. 2016) . It is possible that participants did not adhere to the instructions to insert the brush into the vagina deep enough or did not rotate the brush enough times, which could have translated to lower sample yield.

Additionally, a cytobrush was used for collection, which may have been uncomfortable to use and may have led to hesitancy with sample collection. A higher number of invalid tests were not reflected in other self-collection studies that have been done on an average-risk population (Saville et al. 2018; Smith et al. 2018) . However, these studies used HPV DNA tests and different swabs. Also, compared to self-collected samples, clinician-collected samples likely had greater yield since clinicians are experienced with genital sample collection and performed it under speculum examination. Lower sample material could be a drawback with self-collected methods, but it is potentially associated with types of swabs or assays. Future studies should investigate the impact of different collection devices and techniques on assay test performance similar to studies that have been performed for other female genital tract analyses (Dezzutti et al. 2011 ).

Another important finding of our study is the high acceptance of self-collected HPV testing among WLHIV. Women had positive attitudes toward this testing strategy and thought testing was easy. Also, they felt they were taking care of themselves and would use it again. This indicates that the self-collected HPV testing strategy may be able to increase cervical cancer screening, if the purpose of HPV testing is made clear to patients. For example, one of the largest randomized clinical trials in the U.S. of home HPV testing reported only 12% of participants returned the home HPV test kit, and there was no difference in detection of precancerous lesions compared to usual care (Winer et al. 2019) . The poor kit return in this trial might be explained by the patients' lack of knowledge regarding the superior performance of HPV testing, the requirement to still attend the clinic for usual care, or the awareness that the results could not replace usual care screening methods (Moss et al. 2019 ). Our study builds on this literature by describing HPV self-collection with a specific focus on an underserved urban population with a higher risk of cervical cancer. Moreover, unlike a large-scale study testing its hypothesis under ideal highly controlled conditions, our study examines the "real-world" experience of HPV self-collection among WLHIV. An earlier study from our clinic found that WLHIV were less likely to come to gynecology visits as compared to primary HIV care visits and one of several reasons was fear or discomfort associated with the pelvic examination (Tello et al. 2010) . Given that data are insufficient to recommend pelvic examinations in the absence of symptoms (U.S. Preventive Services Task Force et al.

) and the trend of making pelvic examinations contingent on medical history or symptoms (ACOG 2018), the rationale for HPV self-collection at home or at non-gyn J o u r n a l P r e -p r o o f clinic sites is strengthened. HPV self-collection could result in more WLHIV being screened for cervical cancer, which may justify its slightly inferior performance compared to clinician-collected sampling methods. Self-collected HPV testing could be a particularly effective strategy to reach underserved women, especially those living in Health Professional Shortage Areas such as rural areas or those with barriers to attending medical appointments like lack of transportation or work/child care responsibilities.

Our study has some limitations. First, we had a relatively small study population, which limited the certainty of interpreting our results. However, our study was adequately powered and provided important evidence for future study of primary HPV cervical cancer screening strategies in WLHIV. Second, this study was limited by a higher than expected number of invalid results, although they did not appear to greatly impact the reliability of the tests. Last, our study population was predominately Black WLHIV enrolled at one urban site, which may not represent all WLHIV. However, Black patients accounted for 42% of the new HIV diagnoses in the U.S. (CDC 2019) and our population is a reflection of the nation's racial disparity.

Our examination of SC highlights the specific needs of this population of U.S. WLHIV and the opportunity to improve HPV screening in high resource settings. Compared to the clinician-collected HPV test, the self-collected HPV test had good sensitivity and moderate specificity among WLHIV. Because of the possible inadequate HPV sample J o u r n a l P r e -p r o o f material resulting from self-collection, future studies should focus on how to improve self-collection techniques (including urine-based testing) since it is highly accepted by women and may improve cervical cancer screening in WLHIV who are at higher risk of cervical cancer. 

Committee Opinion No. 754: The Utility of and Indications for Routine Pelvic Examination

Some common problems in medical research

Accuracy of human papillomavirus testing on self-collected versus clinician-collected samples: a meta-analysis

Human papillomavirus DNA testing for the detection of cervical intraepithelial neoplasia grade 3 and cancer: 5-year follow-up of a randomised controlled implementation trial. The Lancet

Preliminary); vol. 30

Risk of Human Papillomavirus-Associated Cancers Among Persons With AIDS. JNCI: Journal of the National Cancer Institute

Comparison of urine, selfcollected vaginal swab, and cervical swab samples for detecting human papillomavirus (HPV) with Roche Cobas HPV

Performance of the Aptima high-risk human papillomavirus mRNA assay in a referral population in comparison with Hybrid Capture 2 and cytology

Overview of the European and North American studies on HPV testing in primary cervical cancer screening

Selfsampling for human papillomavirus in a community setting: feasibility in Hispanic women

Human papillomavirus, human immunodeficiency virus and immunosuppression

Incidence and progression of cervical lesions in women with HIV: a systematic global review

Performance of swabs, lavage, and diluents to quantify biomarkers of female genital tract soluble mucosal mediators

Trends in cancer risk among people with AIDS in the United States

Quality assurance of human papillomavirus (HPV) testing in the implementation of HPV primary screening in Norway: an inter-laboratory reproducibility study

Primary high-risk HPV screening for cervical cancer in post-menopausal women

Use of primary high-risk human papillomavirus testing for cervical cancer screening: interim clinical guidance

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Biological agents

Resolution of equivocal results with the Hybrid Capture II high-risk HPV DNA test: a cytologic/histologic review of 191 cases

Human papillomavirus infection and cytologic abnormalities of the anus and cervix among HIV-infected women in the study to understand the natural history of HIV/AIDS in the era of effective therapy (the SUN study)

Detection rates of precancerous and cancerous cervical lesions within one screening round of primary human papillomavirus DNA testing: prospective randomised trial in Finland

High Prevalence of Human Papillomavirus (HPV) Infections and High Frequency of Multiple HPV Genotypes in Human Immunodeficiency Virus-Infected Women in Brazil

The HPV test has similar sensitivity but more overdiagnosis than the Pap test--a randomised health services study on cervical cancer screening in Finland

Epidemiology of HPV Genotypes among HIV Positive Women in Kenya: A Systematic Review and Meta-Analysis

Persistence of Human Papillomavirus Infection in HIV-Infected and -Uninfected Adolescent Girls: Risk Factors and Differences, by Phylogenetic Type

Home Screening for Human Papillomavirus Falls Short in Initial Application, Remains Promising

Human Papillomavirus and Papanicolaou Tests to Screen for Cervical Cancer

Human Papillomavirus Infection in Women Who Submit Self-collected Vaginal Swabs After Internet Recruitment

Primary cervical cancer screening with HPV testing compared with liquid-based cytology: results of round 1 of a randomised controlled trial --the HPV FOCAL Study

Prevalence of Cervical Cancer Screening of HIV-Infected Women in the United States

Guidelines for the prevention and treatment of opportunistic infections in adults and adolescents with HIV: recommendations from the Centers for Disease Control and Prevention, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America

Utilization of Pap testing among women living with HIV enrolled in primary care in Baltimore, Maryland: A 10-year longitudinal study

Self-collected HPV testing improves participation in cervical cancer screening: a systematic review and meta-analysis

Efficacy of HPV-based screening for prevention of invasive cervical cancer: follow-up of four European randomised controlled trials

Efficacy of human papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial neoplasia: a randomised controlled trial. The Lancet Oncology

Prevalence of type-specific HPV infection by age and grade of cervical cytology: data from the ARTISTIC trial

Self-collection for underscreened women in a National Cervical Screening Program: pilot study

Cancer statistics, 2019. CA A Cancer J Clin

Mailed Human Papillomavirus Self-Collection With Papanicolaou Test J o u r n a l P r e -p r o o f Referral for Infrequently Screened Women in the United States: Sexually Transmitted Diseases

Barriers to recommended gynecologic care in an urban United States HIV clinic

Screening for Gynecologic Conditions With Pelvic Examination: US Preventive Services Task Force Recommendation Statement

Human papillomavirus is a necessary cause of invasive cervical cancer worldwide

Assessing Acceptability of Self-Sampling Kits, Prevalence, and Risk Factors for Human Papillomavirus Infection in American Indian Women

Effect of Mailed Human Papillomavirus Test Kits vs Usual Care Reminders on Cervical Cancer Screening Uptake, Precancer Detection, and Treatment: A Randomized Clinical Trial

NPV (95% CI)

We would like to acknowledge the patients who participated in the project and Perry Barnes, Nicole Quinn, and Rosemary Ramroop for their assistance with the lab assay and recruitment.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.