Utility of Human Papillomavirus Testing for Cervical Cancer Screening in Korea

(1) Background: Cervical cancer is one of the most common cancers in Korean women. This study was performed to discover the utility of HPV (Human Papillomavirus) testing in screening of cervical lesions and to provide the prevalence of HPV and the genotype distribution in a single center of Korea. (2) Methods: A total of 15,141 women who underwent both HPV testing and cervical cytology were enrolled in this retrospective medical record review study. (3) Results: HPV testing showed higher sensitivity than cytology for the detection of histological high-grade squamous lesions. Furthermore, the sensitivity and specificity of HPV testing varied depending on the method used. The BD Onclarity™ HPV assay had higher sensitivity (90%) than the MyHPV CHIP™ kit (all types of HPV: 82%; high-risk HPV: 76%) for high-grade squamous lesions. A combination of MyHPV CHIP™ and cytology detected 90.9% (30/33) of histological high-grade squamous lesions. A combination of BD Onclarity™ HPV assay and cytology detected 96.55% (84/87) of histological high-grade squamous lesions. In addition, HPV prevalence and genotype distribution were different depending on the HPV testing method used. (4) Conclusion: HPV testing showed higher sensitivity than cytology, but the sensitivity and specificity of HPV testing had variation depending on the method used.


Introduction
Cervical cancer is one of the most common cancers in Korean women, and in 2019, 2856 new cases were detected and 834 deaths occurred [1]. Since the introduction of the Papanicolaou test for cervical screening, the incidence and mortality rates of cervical cancer have been decreasing year by year in Korea. From 1999 to 2014, the age-standardized incidence and mortality rates decreased from 16.3 to 9.0 per 100,000 women and 2.6 to 2.1 per 100,000 women, respectively, in Korea [2]. In 2019, the age-standardized incidence was 6.9 per 100,000 women and the age-standardized mortality rate was 1.6 per 100,000 women [1]. For this reason, the national health insurance service in Korea has provided the Papanicolaou test free of charge biennially to women aged 20 or above. In 1983 and 1984, HPV16 and HPV18 were isolated from cervical cancer, respectively [3], and various types (high-risk) of HPV (Human Papillomavirus) DNA were found in almost all cervical cancer specimens. Human papillomavirus infection has been established as the major etiology of cervical cancer [4,5]. In recent research, primary high-risk HPV screening showed a more improved sensitivity for the detection of cervical lesions than cytology alone [6][7][8][9][10]. This study was performed to discover the utility of HPV

Cytology
The cytological diagnoses were obtained from pathologic reports using electronic medical records retrospectively. The slides had Papanicolaou staining, and the Papanicolaou-stained slides had been diagnosed using an optical microscope at the department of pathology, Samsung Changwon Hospital. Three cytologists and five pathologists had been involved in the diagnostic process, and 3999 conventional smears and 11,142 liquid-based cytology tests (all, Surepath) had been performed.

Histological Diagnosis
The histological diagnoses were obtained from pathologic reports using electronic medical records retrospectively. All diagnoses had been performed in the department of pathology, Samsung Changwon Hospital. In the 15,141 women, only 195 women had undergone cervical resection in the department of gynecology, Samsung Changwon Hospital. Formalin-fixed paraffin-embedded (FFPE) blocks had been made using the cervical tissues. All tissues had been sectioned with 5-µm-thick microtome and stained by H&E (hematoxylin and eosin) stain. The stained slides had been diagnosed using an optical microscope by three pathologists.

HPV Detection by MyHPV CHIP™
The MyHPV CHIP™ kit is a PCR (Polymerase chain reaction)-based DNA microarray system. It was approved by the Ministry of Food and Drug Safety of the Republic of Korea (Korea Food and Drug Administration, KFDA) in July 2004. All assays were performed according to the manufacturer's protocol. Cervical cells for HPV testing were obtained using the soft brushes separately from Pap smears. The brushed tissues were supplied to the MyHPV CHIP™ kits and DNA was isolated. Target HPV DNA was amplified by PCR using specific primers with β-globin DNA (internal control). According to the protocol provided by the MyGene Co, Seoul, Korea, DNA isolation and PCR were performed. The PCR products were denatured by heating for 5 min at 95 • C, and then they were mixed with hybridization solution and applied to the DNA chips. The hybridizations were performed at 43 • C for 90 min, and then the products were washed with saline-sodium phosphate-EDTA and air-dried at room temperature. Hybridized HPV DNA signals were detected using a DNA chip scanner (Perkin-Elmer GmbH, Überlingen, Germany). Positive signals revealed a band of 150 base pairs on the gel electrophoresis of the HPV-PCR. The absence of these 150 base pairs on the gel electrophoresis was considered "HPV-negative". This system detected 11 types of high-risk HPV (types 16,18,31,33,35,39,45,51,52,56, and 58), 8 types of low-risk HPV (types 6, 11, 34, 40, 42, 43, 44, and 54), and other types of HPV (unspecified HPV).

HPV Detection by BD Onclarity™ HPV Assay
The cervical brush diluents in cases of conventional smear and the aliquots of the SurePath sample were used for BD Onclarity HPV testing. The cervical brush diluents did not require pre-handling before HPV testing, but each SurePath sample was vortexed briefly and a 0.5 mL aliquot was supplied to 1.7 mL of a BD HPV LBC diluent. Then, the samples (cervical brush diluent and LBC diluent) were pre-warmed for 30 min at 120 • C on the BD pre-warming heater with positive and negative controls. The pre-warmed specimens were loaded onto the BD Viper LT System after cooling. Then, all of the steps, including extraction and amplification of target DNA, were performed automatically by the BD Viper LT System. Next, the specimens with fluorescent detector probes were transferred to a three-wall, four-channel real-time PCR system for genotyping with the internal controls in each wall. The signals were detected by the integrated VIPER LT reader. The system was used for high-risk types 16,18,31,45,51, and 52 and other high-risk genotypes by group (P1: 33/58; P2: 56/59/66; P3: 35/39/68).

Statistical Analysis
All analyses were performed using R version 3.6.2., the R Foundation for Statistical Computing, (Auckland, New Zealand). The difference of HPV high-risk prevalence between MyHPV CHIP™ and the BD Onclarity™ HPV assay was calculated by Fisher's exact test. The difference between the prevalence of high-risk HPV and the prevalence of cytological squamous lesions was calculated by using the Exact McNemar's test in the package "exact2×2". The correlation between high-risk HPV infection and cytological diagnosis was calculated using Spearman's correlation. To compare the proportion of histological high-grade squamous lesions in the negative cytology/positive HPV group with the positive cytology/negative HPV group, Fisher's exact test was performed. The odds ratio of HPV infection in the normal cytology group and the positive cytology group was calculated by Fisher's exact test. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (PLR), and negative likelihood ratio (NLR) were calculated by using "epi.tests" in the package "epiR". The differences of sensitivity and specificity in MyHPV CHIP™, the BD Onclarity™ HPV assay, and cervical cytology were calculated by using Exact McNemar's test in the package "exact2×2". The ROC (receiver operating characteristic) curves drawn by using ROC in the package "Epi".
From March 2016 to June 2017, the BD Onclarity™ HPV assay was conducted for HPV detection. HPV genotypes in the BD Onclarity™ HPV assay group are shown in Table 3. The HPV tests were positive in 1124 out of 9817 women (11.45%) in the BD Onclarity™ HPV assay.

HPV Infection and Histological Diagnosis
With the MyHPV CHIP™ HPV method, 6.38% of those with negative cytology (threshold: ASCUS or worse)/negative HPV infection had histological high-grade squamous lesions (CIN2 or worse). Furthermore, 30.77% of those with negative cytology (threshold: ASCUS or worse)/positive high-risk HPV infection had high-grade squamous lesions, and 13.04% of those with positive cytology (ASCUS or worse)/negative HPV infection also had high-grade squamous lesions. There was a difference between those with negative cytology/positive high-risk HPV (30.77%, 4/13) and those with positive cytology/negative HPV (13.04%, 3/23) in the prevalence of high-grade cervical lesions, but it was not statistically significant (p = 0.225). Additionally, 67.74% of those with positive cytology (ASCUS or worse)/positive high-risk HPV had high-grade squamous lesions in their histologic diagnosis; 5.36% of those with negative cytology (threshold: LSIL or worse)/negative HPV infection had histological high-grade squamous lesions; 27.78% of those with negative cytology (LSIL or worse)/positive high-risk HPV infection and 21.43% of those with positive cytology (LSIL or worse)/negative HPV infection had high-grade squamous lesions; and 76.92% of those with positive cytology (LSIL or worse)/positive high-risk HPV infection had high-grade cervical lesions in their histologic diagnosis. From the total 33 histological high-grade squamous lesion cases, three showed negative cytology and negative HPV with the MyHPV CHIP™ method. This means that the combination of MyHPV CHIP™ and cytology detected 90.9% (30/33) of histological high-grade squamous lesions (Table 10). With the BD Onclarity™ HPV assay, 6% of those with negative cytology (threshold: ASCUS or worse)/negative HPV infection had high-grade squamous lesions (CIN2 or worse) in their histologic diagnosis; and 56.52% of those with negative cytology (ASCUS or worse)/positive high-risk HPV infection and 33.33% of those with positive cytology (ASCUS or worse)/negative HPV infection had high-grade squamous lesions. There was a difference between negative cytology/positive HPV (56.52%, 13/23) and positive cytology/negative HPV (33.33%, 6/18) in prevalence of high-grade squamous lesions, but it was not statistically significant (p = 0.209). Furthermore, 80.25% of those with positive cytology (ASCUS or worse)/positive high-risk HPV infection had high-grade squamous lesions in their histologic diagnosis; 5.45% of those with negative cytology (threshold: LSIL or worse)/negative HPV infection had histological high-grade squamous lesions; 65.79% of those with negative cytology (LSIL or worse)/positive high-risk HPV infection and 46.15% of those with positive cytology (LSIL or worse)/negative HPV infection had high-grade squamous lesions; and 80.3% of those with positive cytology (LSIL or worse)/positive high-risk HPV infection had high-grade squamous lesions in their histologic diagnosis. From the total 87 histological high-grade squamous lesion cases, 3 showed negative cytology and negative HPV with the BD Onclarity™ HPV assay. This means that the combination of BD Onclarity™ HPV assay and cytology detected 96.55% (84/87) of histological high-grade squamous lesions (Table 11). In the normal cytology group of MyHPV CHIP™, the odds ratio of HPV infection was 5.64 (95% CI 0.82-44.89; p = 0.0425) and the odds ratio of high-risk HPV was 6.38 (95% CI 0.91-51.48; p = 0.032) for histological high-grade squamous lesions. In the positive cytology group (ASCUS or worse) of MyHPV CHIP™, the odds ratio of HPV infection was 9.23 (95% CI 2.20-56.59; p = 0.00046) and the odds ratio of high-risk HPV was 10.42 (95% CI 2.85-45.76; p = 7.825 × 10 −5 ) for histological high-grade squamous lesions. In the normal cytology group of the BD Onclarity™ HPV assay, the odds ratio of HPV high-risk infection was 19.19 (95% CI 4.26-124.33; p = 4.448 × 10 −6 ) for histological high-grade squamous lesions. In the positive cytology group (ASCUS or worse) of the BD Onclarity™ HPV assay, the odds ratio of HPV high-risk infection was 7.9 (95% CI 2.33-29.96; p = 0.00019) for histological high-grade squamous lesions.
The current studies indicate that HPV testing showed higher sensitivity than cytology, but the sensitivity and specificity of HPV testing varied depending on the HPV testing method used [22][23][24]. Our study also shows similar results to previous studies. The HPV testing showed higher sensitivity than cytology (BD Onclarity™ HPV assay 90%; MyHPV CHIP™ 82%; ASCUS or worse 81%), but there was a slight difference between MyHPV CHIP™ and cytology (threshold: ASCUS or worse). In our study, the BD Onclarity™ HPV assay had higher sensitivity (90%) than the MyHPV CHIP™ kit (all types of HPV: 82%, high-risk HPV: 76%) for high-grade squamous lesions, and MyHPV CHIP™ had higher specificity (79%) than the BD Onclarity™ HPV assay (69%).
The addition of HPV testing to cytology increased sensitivity and specificity compared to cytology alone. Additionally, the addition of cytology to HPV testing increased sensitivity and specificity compared to HPV testing alone. Thus, HPV testing and cytology were complementary.
The obvious limitation in this study was a small sample size of confirmed histological diagnoses. Only about 14.5% of women of abnormal cytology or high-risk HPV infection underwent cervical resection in our hospital (19.4% in MyHPV CHIP™; 12.3% in the BD Onclarity™ HPV assay). The majority of cases (73.59%) were women identified through systemic health screening. Many of them underwent cervical resection in other gynecological hospitals and we could not access their histologically confirmed diagnoses. Besides this, we failed to exclude pregnant women and vaccinated women due to their incomplete medical information.

Conclusions
HPV testing showed higher sensitivity than cytology, and the sensitivity and specificity of HPV testing varied depending on the HPV testing method. HPV prevalence and genotype distribution were also different depending on the method. The BD Onclarity™ HPV assay had higher sensitivity than the MyHPV CHIP™ kit for high-grade squamous lesions. Furthermore, the combination of cytology and HPV testing increased sensitivity and specificity compared to one method alone; thus, cytology and HPV testing are complementary.