Distribution of cervical lesions in high-risk HPV (hr-HPV) positive women with ASC-US: a retrospective single-center study in China

Background To investigate distributions of cervical lesions and factors associated with the severity of the cervical lesions in high-risk HPV (hr-HPV) positive women with atypical squamous cells of undetermined significance (ASC-US) cytology. Methods Clinical information of 250,000 women who underwent HPV and cytological test was collected from January 2012 to January 2019. The association between the severity of the cervical lesions and hr-HPV genotypes, hr-HPV viral load, and ages, were analyzed in hr-HPV-positive/ASC-US women. Results 3459 hr-HPV-positive/ASC-US women were enrolled in this study. Overall, 43.51% of women with ASC-US had normal histological results, 34.35% had high-grade squamous intraepithelial lesion (HSIL), and 1.30% had cervical cancer. The rate of HSIL or worse (HSIL+) in women with single HPV16 infection (63.09%) was the highest, followed by HPV33 (57.50%), HPV51 (36.11%), HPV58 (36.11%), HPV52 (28.28%), HPV18 (26.37%), HPV66 (19.35%), HPV39 (18.92%), HPV53 (15.00%), and HPV56 (8.51%). Detection rate of HSIL+ in low, intermediate and high viral-load groups were 15.87% (n = 30), 34.91% (n = 74) and 40.68% (n = 214) (Cochran-Armitage Trend test χ2 = 35.03, p < 0.0001). Compared with the 51–60-year-old group (21.65%), the women in ≤ 30 (40.52%), 31–40 (39.67%), and 41–50 (34.22%) year-old groups had significantly higher risk of HSIL+. The women in ≤ 51–60 (2.68%) and > 60 (3.41%) year-old groups were at increased risk for cervical cancer, compared with the ≤ 30-year-old group (0.61%). Conclusions ASC-US women with HPV 16/18/33/51/52/58 single infection and multiple infections, as well as high HPV viral loads, have high risk of HSIL+.

Thinprep cytologic test (TCT) is a commonly used method for cytology test [11]. However, only 7-10% of ASC-US women was diagnosed as cervical intraepithelial neoplasia (CIN) 3, and there was very low probability of invasive carcinoma for ASC-US women [12].
There are three common HPV tests in China. The Cobas 4800 HPV test performs DNA extraction, Polymerase Chain Reaction amplification, and real time detection by an automated sample preparation, to detect 14 hr-HPV genotypes. The HPV GenoArray test simultaneous identification of 21 individual HPV genotypes and efficient detection of single or multiple HPV infection, which preforms DNA amplification and HybriBio's proprietary flow through hybridization technique. Hybrid capture 2 test (HC2) is one of the most frequently applied test to detect the presence of any 13 hr-HPV types, which can provide a quantitation of the viral DNA load.
This study aimed to investigate the distribution of hr-HPV genotypes, the association between the severity of the cervical lesions and hr-HPV viral loads, and agestratified prevalence of high-grade squamous intraepithelial lesion (HSIL) and cervical cancer, by analyzing of clinical information of HPV-positive/ASC-US women in China, which would help us find an appropriate triage of women with ASC-US cytology in the future.

Study population
All procedures in studies involving human participants were performed in accordance with ethical standards and approved by the Ethics Committee of Shandong University [2018 (054)].And the requirement for written informed consent was waived by the Ethics Committee of Shandong University.
This retrospective study included women who accepted TCT test, HPV test and colposcopy at Qilu Hospital of Shandong University from January 2012 to January 2019.
Exclusion criteria: (1) women with history of treatments to cervical lesions, such as cervical surgery, laser, freezing and medication, et al.; (2) women with malignant tumors such as ovarian cancer and endometrial cancer; (3) women with autoimmune diseases or receiving immunotherapy; (4) women with pregnancy; (5) women who had received the HPV vaccine; (7) women with smears of insufficient quality (with an absence of endocervical cells).

Cytology test and HPV test
No vaginal washing, intravaginal medication or physical therapy for 3 days before the samples were collected. No sexual behavior within 24 h and women were in nonmenstrual period when the samples were collected. Specimen collection, specimen preparations, and results of test are performed, according to the instructions of the manufacturers, respectively.
ThinPrep (Hologic Inc, Marlborough, Massachusetts, USA) preparations was used and cytological diagnosis was performed by experienced cytopathologists, according to the Bethesda system 2011 [13].
Women could choose one of three HPV tests: (

Colposcopy and guided cervical biopsies
All women, enrolled in this study, underwent visual inspection with acetic acid and insulin inspection with Lugo's iodine (VILI) and colposcopy. If any of suspicious lesion were found among the above-mentioned examinations, multi-point biopsy was taken directed by colposcopy at the suspected lesion; otherwise, a random four-quadrant cervical biopsy was obtained at the squamocolumnar junction. Women who were not satisfied with colposcopy should underwent Endocervical Curettage, if necessary. The histologic diagnosis was based on the consensus diagnosis of two experienced gynecologic pathologists. All gynecologic pathologists were blinded to results of TCT, HPV test and colposcopic examination. The results of biopsy reports include: Normal; low-grade squamous intraepithelial lesion (LSIL) (including CIN 1), highgrade squamous intraepithelial lesion (HSIL) (including CIN 2-3), and cervical cancer. HSIL or worse (HSIL+) includes HSIL and cervical cancer. The gold standard for this study was the histologic diagnosis. The funding sources acknowledged the design of the study but did not influence the process of our study, including data collection, analysis, and interpretation and the reporting of results.

Statistical analysis
Statistical analyses were performed using the SAS 9.4 (SAS Institute, Cary, NC) for windows. The categorical variables were expressed by a percentage (%). Cochran-Armitage Trend test was used to analyze the detection rate of HSIL+ in different viral load groups. The binary Logistic regression model was used to analyze the detection rate of HSIL+ and cervical cancer in different age groups. Results were considered statistically significant at p values < 0.05.

Characteristics of the study population
Clinical information of 250,000 women who underwent HPV and TCT test at Qilu Hospital of Shandong University from January 2012 to January 2019 was retrospectively collected and analyzed (Fig. 1). Among them, 7001 (2.80%) women were diagnosed as ASC-US by TCT test, including 3459 (49.41%) women with hr-HPV infection and 3542 (50.59%) women without hr-HPV infection. Finally, a total of 3,459 ASC-US women with hr-HPV infection were enrolled in this study finally.
A total of 1973 women chose the HPV typing test, identified the specific types of hr-HPV they were infected with, among whom 376 were 12 hr-HPV positive (from Cobas 4800 HPV test), 1,043 women had single HPV infection, and 554 women had multiple HPV infections. 1245 women were diagnosed with hr-HPV positive by HC2 test. The remaining hr-HPV positive 241 women who lost their results of HPV test, were classified as ANY group in this study. (We did not know the specific genotypes, nor did we know whether it is a single infection or multiple infections. We only knew that they were The average age of women was 39.86 ± 9.94 years (range: 25-80 years). The median number of pregnancies was 3 (range: 0-14) and the median number of births was 1 (range: 0-8).This study divided the age into 5 age groups ≤ 30 (n = 654, 18.91%), 31-40 (n = 1258, 36.37%), 41-50 (n = 1011, 29.23%), 51-60 (n = 448, 12.95%), and > 60 (n = 88, 2.54%).

Distribution of histological results among different genotypes in ASC-US women
The common hr-HPV genotype was shown in Table 1 and Fig. 2, in women with ASC-US.
Among women with the multiple infections, the HSIL+ detection rates in women with double hr-HPV infections, triple hr-HPV infections, and quadruple or more hr-HPV infections were 39.77%, 41.49%, and 45.00%, respectively ( Table 2).
A total of 45 (1.30%) of the women in this study were diagnosed with cervical cancer, including 24 women with According to the viral load of hr-HPV, women were divided into 3 groups: low viral load group (1 ≤ RLU/ CO < 10, n = 189), medium viral load group (10 ≤ RLU/ CO < 100, n = 212), and high viral load group (RLU/ CO ≥ 100, n = 526). The pathological distribution of these three viral-load groups was shown in Table 3. The percentage of women with normal pathology in the low-viral-load group (65.61%) was highest, followed by intermediate-viral-load group (41.98%) and high-viralload group (35.17%). However, detection rate of HSIL+ in low-viral-load group was lowest, followed by intermediate-viral-load group (34.91%) and high-viral-load group (40.68%), and there was is statistically significant between detection rate of HSIL+ and hr-HPV viral load (χ 2 = 35.03, p < 0.0001).

Distribution of histological results among age groups in ASC-US women
Age-stratified prevalence of HSIL+ and cervical cancer in hr-HPV-positive/ASC-US women is shown in Fig. 3 Table 4).

Discussion
Currently, the HPV vaccine is gradually being used in the developing countries, but it has not fully covered all the regions. The effect of the HPV vaccination will reveal after several decades. In addition, the cervical cancer screening is still an essential method to prevent and treat cervical cancer for older women. Therefore, it's significant to adopt the cervical cancer screening method to diagnose and prevent the cervical cancer. ASC-US is an ambiguous term and an exclusionary diagnosis that suggests a risk of disease rather than a definitive diagnosis of abnormal lesions. For women with ASC-US, if the diagnosis is not clear or not timely, the optimal treatment will be delayed; on the other hand, the excessive diagnosis and treatment will not only bring physical and psychological burden to the women, but also can cause adverse pregnancy outcomes [14,15]. In addition, it will increase the family and society economic burden and will not allocate medical resources reasonably.
This retrospective cross-sectional single-center study in China, lasting for 7 years, collected clinical data of 250,000 women who underwent HPV and TCT test at Qilu Hospital of Shandong University. The frequency of ASC-US was 2.80% (n = 7001) in cervical screenings of this center, which was suggested the center strictly controlled the frequency of occurrence of ASC-US and reduced the phenomenon of ascus as the "Garbage Dump", according previous research [13]. Some studies have reported that the prevalence of CIN2+ among women with ASC-US is between 5% -39% [16][17][18]. The detection rate of HSIL+ in hr-HPV-positive/ASC-US women in this study was 35.65%, which was higher than some studies. The possible reasons are as follows: (1) The target population of this study was hr-HPV-positive/ ASC-US women, while the target population of other studies were ASC-US women (including hr-HPV positive and negative women). The reported prevalence of hr-HPV among women with ASC-US in most studies was 23% to 74% [19]. (2) This study was a retrospective study with the possibility of incomplete data collection.
In this cross-sectional study, the rates of HSIL+ ranged from 8.51% of HPV 56 to 63.09% of HPV16 in ASC-US women with single hr-HPV infection. The HSIL+ detection rates in women with HPV multiple infections was more than 39.77%. Previous research reported HPV-positive/ ASC-US women would have similar the 2-year cumulative risk of HSIL+ and be clinically equivalent with women with LSIL Pap results. We look forward to relevant longitudinal researches on relationship between HPV genotypes and risk of cervical lesions.
Systematic review [20,21] and a large randomized trial [11] consistently showed that, compared with repeat cytology, the accuracy of HC2 to detect underlying HSIL+ was higher in triage of women with ASC-US cytology. Previous studies suggested that the test may optimized by using a cutoff higher than this 1.0 pg/ mL [22,23]. The rates of HSIL+ in higher viral road of women with ASC-US cytology were higher (Cochran-Armitage Trend test χ 2 = 35.03, p < 0.0001) in this study.
The highest detection rate of HSIL+ (40.52%) was in the age group of women aged ≤ 30 years, and the lowest detection rate of HSIL+ (21.65%) was in the 51-60-yearold group in this study. A large longitudinal cohort study lasting 7 years within the Guanacaste population showed older women had a similar or slightly decreased risk of HSIL+ and especially CIN 3 compared with younger women [24]. In addition, a systematic review included in 103 studies (including more than 12,400,000 women) found that HSIL prevalence in Asia peaked at a relatively younger age (25 to 40 years) [25]. The results in these studies are in coincidence with our findings. When calculating the cervical cancer detection rate, this study showed that the value of the cervical cancer detection rate increased with age, and older women had higher cervical cancer detection rates than younger women. The reasons are as follows: (1) Due to hormone levels, the use of intrauterine device, more sexual behavior and inflammation in younger women, the cells of those with benign morphological changes mixed with cells of true precancerous lesions, and it may interfere doctors' diagnosis.
(2) HPV prevalence rate among younger women are higher, and the rate of HPV prevalence is highly age-related and decreases with age [26]. According to previous study, the HPV prevalence rate could be as high as 70% for women of aged < 25 years [27]. (3) The previous research found that CIN1 and a subset CIN2 lesions are clinical manifestations of the result of productive hr-HPV infectio [28], which can regularly regress within 1-2 years spontane-ously and have a low risk to progress to invasive carcinom [29]. (4) Older women could clear newly HPV infections (including HPV16 infections) as quickly as younger women [24]. So, the lowest detection rate of HSIL+ was in the group aged 51-60 years old. (5) Lower estrogen levels, thinning of the epithelium, decreasing lactobacilli and some genital tract infections combined diabetes etc., create favorable conditions for HPV infection in older women, especially for > 60 years old women. (6) Another subset of CIN2 lesions and CIN3 lesions are clinical manifestations of the result of transforming hr-HPV infection, which is characterized by a dysregulated expression of E6 and E7 viral oncogenes [28]. It would take the long time of 20-30 years for the progression to invasive carcinoma from precancerous lesion in most patients [10]. (7) What's more, persistent hr-HPV infection is a risk factor for cervical cancer and the risk of persistence increases with the increasing age of the woman. Therefore, older ASC-US women have a higher cervical cancer detection rate.
However, this study showed there was no statistically significant increase in the rate detection of HSIL in the group of > 60 years by 36.49% (29.55% vs 21.65%), compared with the 51-60 years group, which may be related to the relatively limited sample we studied.
A 2013 Cochrane Meta-analysis studied triage ASC-US by repeat cytology versus HPV testing and found that similar pooled CIN2+ and CIN3+ specificities that likely will translate to similar overtreatment rates [12]. So, one way to solve the problem of overtreatment is to consider alternatives new method to triage women with ASC-US. Molecular markers may help identify cells with abnormal cell morphology. P16 or dual p16 and Ki-67 immunostaining on cytological preparations provides a promising method to triage HPV-positive women [30,31]. Biomarkers based on DNA methylation includes methylation markers representing various combinations of genes, such as SRY-box 1 (SOX1), PAX1, NK6 homeobox 1 (NKX6-1) [32], junctional adhesion molecule 3 (JAM3), EPB41L3, TERT, C13ORF18 [33], and CADM1/ MAL [34].
Combined with knowledge about the biology of CIN and results of this study, here are a few suggestions to triage women with ASC-US cytology. Firstly, we should ensure the quality of TCT test and strictly control the occurrence frequency of ASC-US. It has been suggested that good management requires that the frequency of ASC be maintained at < 5% of all cervical screenings [11], and ASC:SIL ratio be maintained at < 1.5. Secondly, ASC-US women with obvious inflammation should accept anti-inflammatory treatment, and repeat cytology test 4-6 months after inflammation disappearing, especially young women. If the cytological results of women ≥ ASC-US, then they are recommended to refer to colposcopy. In addition, we should strictly follow up on young women who have been diagnosed as HSIL by histological examination, rather than implementing excessive treatment. Finally, it is more and more important for us to find a new triage test.

Conclusions
ASC-US women with HPV 16/18/33/51/52/58 single infection and multiple infections, as well as high HPV viral loads, have high risk of HSIL+. In addition, the rate of HSIL+ was higher and the rate of cervical cancer was lower in younger hr-HPV-positive/ ASC-US women, in contrast with older women.

Limitations
This study was a retrospective single-center study with the possibility of incomplete data collection and limited data sources. What's more, women enrolled in this study were tested by there HPV tests and different HPV tests varied in sensitivity and specificity. Relatively large prospective studies were expected to confirm the findings of this study in the future.