Clinical Performance of Self-collected Specimen HPV-DNA vs Clinician- collected Specimen HPV-mRNA to Detect High-risk HPV and High-grade Cervical Lesions and Cancer

Objective: Self- collected specimens to detect high-risk (hr) HPV and high-grade cervical lesions (CIN2+) has been introduced aiming to increase cervical cancer screening coverage. The performance of self- collected specimen compared to clinician collected specimen is one major concern. This study aimed to compare self-sampling HPV-DNA and clinician-sampling HPV-mRNA to detect hr-HPV and high-grade cervical lesions. Methods: Women with abnormal cervical cytology and/ or positive hr-HPV who attended the colposcopy clinics in 10 tertiary hospitals in Bangkok were enrolled. Self-collected specimens were evaluated for HPV DNA using Cobas® 4800 HPV test prior to the clinician-collected specimens which were tested for HPV mRNA with APTIMA® HPV Assay. Subsequent colposcopy with biopsy was performed. The detection rates of hr-HPV from both HPV tests and their performance to detect high-grade lesions pathology were compared. Results: Data from 497 women’s specimens were analyzed. Both samplings had 86.8% concordance rate in detecting hr-HPV (Kappa 0.670; 95% confidence interval [CI] 0.599-0.746, P value < 0.001). The sensitivity (95% CI) of self-collected specimen HPV DNA and clinician- collected specimen HPV-mRNA to detect high-grade lesions were 91.8% (85.4%-96.0%) and 90.2% (83.6%-94.9%) respectively. The corresponding negative predictive values (95% CI) were 91.9% (85.6%-96.0%) and 91.7% (86.0%-95.7%) respectively. Conclusion: HPV DNA testing from self-collected specimen to detect HR-HPV demonstrates high concordance with HPV mRNA testing from clinician-collected specimen. The sensitivity and negative predictive value of both tests to detect high-grade lesions are comparable.


Introduction
Cervical cancer is the fourth common cancer in women worldwide [1].Up to 604000 new cases and 342000 deaths were reported in 2020 and most cases occurred in low-and middle-income countries [1,2].In Thailand, cervical cancer is the third most common women cancer with ASIR and mortality rate of 16.4 and 7.4 per 100,000 women-years respectively [1].
Primary prevention with HPV vaccine is the ideal method to reduce the incidence of cervical cancer.However, the vaccination coverages are still low in some countries [3].Hence, secondary prevention with cervical cancer screening is still the main key to reduce its incidence [4].Cytologic testing, which was the main Clinical Performance of Self-collected Specimen HPV-DNA vs Clinician-collected Specimen HPV-mRNA to Detect High-risk HPV and High-grade Cervical Lesions and Cancer method of cervical cancer screening for several decades had been recently replaced by primary HPV testing due to its higher sensitivity to detect cervical cancer.
Not only the high sensitivity of the screening tool, a high coverage of target population to have screening is also crucial for cervical cancer reduction.Despite a wide availability of the screening services by health care providers, one major problem is still encountered.The women, themselves, may not adhere to the screening recommendation.Previous studies in many countries [5,6] as well as in Thailand [7][8][9] had revealed many reasons for the non-adherence.This included fear, embarrassment, no time, distance to the service, negligence, or cultural background.
With many reasons of women not to undergo screening  3 Chiang Mai University, Thailand. 4King Chulalongkorn Memorial Hospital, Chulalongkorn University, Thailand.*For Correspondence: saranya.jibi@gmail.comby the physician or in health service unit, self-collected HPV testing might be an alternative option to overcome these barriers and to increase the population to participate with the cancer screening program.In 2020 WHO announced that using self-sampling HPV testing can help to reach a global target of 70% coverage of screening by 2030 [10].
One major concern of self-collected specimen HPV testing is the efficacy to detect high-risk HPV (hr-HPV) and pre-invasive cervical lesions comparing to health-care worker collection.Some previous studies had shown a good correlation of self-and clinician-collected specimen to detect hr-HPV [11,12].However, those studies focused on the HPV DNA testing.
From the evidence that detection of high-risk HPV E6 and E7 mRNA might be specific than detection of HPV DNA.Some authors reported similar sensitivity and slightly higher specificity of HPV messenger RNA (HPV mRNA) and HPV DNA testing to detect highgrade cervical intraepithelial neoplasia or more severe lesions [13].In 2021, WHO has suggested HPV mRNA detection using samples taken by the health-care provider as a primary screening test [14].
However, data comparing the performance between self-collected specimen HPV-DNA vs clinician-collected specimen HPV-mRNA are still limited.This study aimed to compare the performance of self-collected specimen HPV-DNA vs clinician-collected specimen HPV-mRNA to detect hr-HPV and high-grade cervical lesions and cancer (High-grade squamous intraepithelial lesion or worse: CIN 2+).

Study population and study design
This study was approved by the Central Research Ethics Committee of Thailand (COA-CREC082/2021).A multicenter prospective trial collaboration of 10 institutes in Thailand which was conducted under the support of the Thai Gynecologic Cancer Society (TGCS).This study was one among the large project along with a few other studies with other objectives.
The study population included women aged ≥ 18 years who attended colposcopy clinic due to abnormal cervical cytology and/or abnormal HPV testing during October 2021-May 2022.The exclusion criteria were women who were pregnant, diagnosed with cervical cancer, previous treatment with hysterectomy, radiation therapy or chemotherapy, had active bleeding or used vaginal douching within 48 hours before collecting the specimen.The process of study conduct is shown in Figure 1 Specimen collection Self-collected specimen and HPV DNA testing The purpose of the study was explained to each woman who met inclusion criteria.After obtaining written informed consent, the process of self-collected specimen was described to the participants by a research assistant along with manufacturer's illustrated instructions and a video demonstration.The self-collected specimen procedure was performed in a separate room using self-collected specimen tools from the Aptima multitest swab specimen collection Kit (Hologic, Inc., San Diego, CA, USA).

Clinician-collected specimen and HPV mRNA testing
The clinician who performed colposcopy collected the cervical sampling using ThinPrep specimen collection kit for mRNA testing (Aptima® HPV assay ®, Hologic, Inc., USA).

Colposcopy and cervical tissue collection
After the clinician specimen collection, colposcopy, and cervical biopsy with or without endocervical sampling were performed.Tissue obtained were sent to pathologic laboratory of each institution for histopathological analysis.

Data collection and Statistical analysis
Data collected were: baseline characteristics, history of cervical cancer screening, the results of HPV tests, and subsequent histopathology.Statistical analyses using IBM SPSS Statistics for Windows, Version 28.0 (IBM Corporation, Armonk, NY, USA) were carried out.A p-value of < 0.05 was considered as statistically significant.Cohen's Kappa coefficients was calculated to evaluate the agreement between two sampling collection method.The scale to express the strength of the agreement were as follow: 0.00-0.20 (low), 0.21-0.40(fair), 0.41-0.60(moderate), 0.61-0.80(substantial) and ≥ 0.81 (almost perfect).

Results
Overall, 500 participants were accessed for eligibility.Except one whose age was under 18 years old, a total of 499 participants were enrolled in the study.Five of them were further excluded: 2 had missing of socio-economic characteristic feature, 1 had hysterectomy and 2 had history of cervical cancer.A total of 494 participants had vaginal self-collected specimen and were included for the analysis.
The mean age of 497 women was 39.28 + 11.36 years.335 of them (67.4%) were married.353 (71.0%) have education bachelor's degree or higher and 326 women (65.6%) have family income over 24,000 THB per month (672 USD) (Table 1).424 women (85.8%) are in pre-menopause status.Only 6 (1.2%) of them never Figure1.Enrollment of 500 Women who had Abnormal Cervical Cancer Screening Test and were Scheduled to Undergo Colposcopy had sexual activity.The most used method was oral contraceptive pills (162 women or 32.8%) among 330 women (66.4%) ever use any methods of contraception.Of note, 436 women (87.7%) had never received HPV vaccination, whereas 72% were still sexually active (Table 2) Among 70.0% who had ever had cervical cytology testing, 22.7% had history of abnormal cervical cytology.On the other hand, out of 21.7% who ever had HPV testing 15.3% had HR-HPV detected.We found 149 (30.0%) and 389 women (78.3%) had never had cervical cancer screening or particularly HPV assays respectively.Among 91 (18.3%) women who reported history of abnormal cervical lesions, 61 women (12.3%) were diagnosed CIN1/LSIL/HPV effect (Table 3).
The indication for colposcopy included abnormal cytology and/or positive HVP test.Abnormal cervical cytology in 142 women were: ASC-US/ASC-H/AGC.60 women (12.1%) were presented with hr-HPV test without cervical cytology.

Discussion
There are five HPV assays which have been approved by US-FDA as HPV screening: Hybrid CaptureII, Cervista, Cobas, Aptima and BD Onclarity [15].Technology to detect high-risk HPV includes DNA-RNA hybridization, signal amplification and PCR based which detected hr-HPV DNA.To date, Aptima is only HPV assays to detected viral mRNA (E6/E7) which is transcriptional active virus which seems more specific to high grade cervical lesion [15].The systematic review and meta-analysis demonstrated slightly higher specificity (1.03 [95%CI: 1.02-1.04]) of mRNA HPV testing over HPV DNA testing for detection of CIN2+ with similar sensitivity (0.98 [95%CI: 0.95-1.01])[16].Cumulative detection rates of high-grade cervical lesion after negative mRNA or negative DNA screening were also comparable [16].Currently, the mRNA HPV test for cervical cancer screening was approved and recommended by WHO in 2021.
Previous studies reported that vaginal self-collected specimen technique could increase participation of women in cervical cancer screening [17,10,18].Regarding the clinical performance of self-collected specimen, the results comparing to the clinician-collected specimen were still controversial.The accuracy of self-collected specimen HPV testing was validated against clinician-collected specimens across different settings [19][20][21][22][23][24][25].The difference might lie on multiple factors such as attitude of the women, collection technique, performance of the test, and laboratory logistics of the specimen.The meta-analysis from 36 studies enrolled 154,556 women to identified accuracy of HPV test on selfcollected versus clinician-collected sampling for cervical cancer prevention revealed pooled sensitivity of HPV testing on self-collected specimen was lower than clinicalcollected specimen for high grade cervical lesion (CIN2+) [19].Interestingly, some PCR-based assays showed similar sensitivity on both techniques of sample collection [19].This study which was a multicenter prospective clinical trial demonstrated comparable clinical performance of self-collected specimen HPV-DNA and clinician-collected specimen HPV-mRNA to detect high grade cervical lesion in abnormal cervical cytology and/or HPV testing patients who attended colposcopy clinic in Bangkok, Thailand.We demonstrated similarly high prevalence of hr-HPV detection in self-collected specimen HPV-DNA and clinician-collected specimen HPV-mRNA (74.7% and 70.4% respectively).These rates were comparable with the results from one systematic review on the performance of HPV mRNA [16] However, our prevalence of hr-HPV in self-collected specimen and clinician-collected specimen were higher than 46.6% and 48.0%respectively from Tiiti's study [26].The difference may partly lie on the characteristic features of women included in the study.Our population were women who attended colposcopy clinic with abnormal cervical cytology and/or positive HPV test whereas their study included general women who sought for service in their gynecology department [26].
Our study demonstrated high agreement (86.8%) of self-collected specimen HPV-DNA versus cliniciancollected specimen HPV-mRNA with kappa of 0.670 (95%CI: 0.599-0.746).The concordance rate between self-collected specimen HPV-DNA and cliniciancollected specimen HPV-mRNA in this study is substantial agreement which is higher than the studies of Phoolcharoen et al. which reported 74.5% (k=0.46)concordance rate between vaginal self-and cliniciancollected HPV-DNA [21] and to the study of Aranda Flores et al. which found 78.2% concordance rate (k=0.34)for HPV-DNA and 92.5% (k=0.40) for HPV-mRNA [19,24].Moreover, the clinical performance of clinician-collected specimen HPV-mRNA to detect CIN2+ in our study which showed high sensitivity (90.2[95%CI: 83.6-94.9]and high negative predictive value (91.7[95%CI: 86.0-95.7]were consistent with a prior systematic review demonstrated high relative sensitivity for CIN2+ for clinician-collected specimen HPV mRNA (98[95%CI: 0.95-1.01])[16].Our study revealed slightly high sensitivity for self-collected specimen-HPV DNA (91.8[95%CI:85.4-96.0])compared with prior systematic review which relative sensitivity of self-collected specimen-HPV DNA (0.88[95%CI: 0.85-0.91])for CIN2+ [19].Since our study focused on only the women with abnormal cervical screening as the indications for colposcopy, the rates of hr-HPV may be higher than normal women.Nevertheless, findings from this group of women highlight the difference or similarity of HPV detection rates between the 2 tests.
The strength of our study which could be noted is that it is a prospective multicenter study, the number of participants was large enough to assess a clinical association between different techniques of cervical specimens' collection (self-vs clinician-) and HPV assays (HPV-DNA vs HPV-mRNA).Our findings support that self-sampling can serve as an alternative screening test to a physician sampling.Furthermore, either HPV-DNA test or -mRNA test could be used in clinical practice for cervical screening.
In conclusion, HPV-DNA testing from self-collected specimen demonstrates high concordance with HPV-mRNA testing from clinician-collected specimen to detect high-risk HPV.The sensitivity and negative predictive value of both tests to detect high grade lesions are also comparable.To increase coverage of cervical cancer screening in Thailand, the self-collected specimen might be another assay for women who are inconvenience to seek the healthcare providers.

Table 3 .
History of Cervical Screening and Abnormal Findings (n=497)

Table 5 .
Agreement of Self-Collected Specimen HPV DNA and clinician-collected specimen HPV mRNA (N=494)

Table 6 .
Histological Analysis according to HPV Results

Table 7 .
The Clinical Performance of HPV Assays to Detected High Grade Cervical Lesion