背景

宫颈癌是全球女性最常见的恶性肿瘤之一，宫颈癌的发生与高危型人乳头瘤病毒（HR-HPV）的持续感染相关。虽然绝大多数感染可被宿主免疫系统清除，但持续存在的感染可进一步导致宫颈上皮内瘤变（CIN）和宫颈癌的发生。以HPV DNA或mRNA为靶标的分子学检测方法，与传统细胞学相比，可更有效的检测高级别宫颈上皮内瘤变（CIN2+）。由于不同HPV型别的致癌风险存在差异，且高病毒负荷通常与更高的癌变风险相关，因此HPV基因分型检测与病毒负荷定量分析，在宫颈癌筛查的分流管理中具有重要临床意义。

当前HPV病毒分型和病毒负荷测量手段主要包括杂交捕获法（Hybrid capture）、实时荧光定量PCR（Real-time quantitative fluorescence PCR, qPCR）等。杂交捕获法以病毒全长基因组为靶标，属于半定量检测技术。基于qPCR的检测方法在分型检测的基础上，较杂交捕获法定量更为准确。然而，该方法依赖标准曲线进行定量分析，多数方法针对更保守的HPV L1区段进行检测。Aptima和Onclarity检测针对HPV E6/E7的mRNA和DNA进行检测，针对mRNA的检测在区分一过性感染与持续性感染方面具有潜在优势。数字PCR（digital PCR, dPCR）分为芯片式（cdPCR）或微流控式（mdPCR）和微滴数字PCR（Droplet digital PCR, ddPCR），ddPCR法通过微滴化和泊松分布计算实现绝对定量，无需依赖标准曲线。

本研究从分子检测层面讨论HPV DNA和mRNA水平在宫颈癌及癌前病变中的分布，分别探究基于HPV DNA、mRNA的检测方法在不同癌前病变阶段检出的病毒负荷变化情况及临床检测效能。为理解病毒负荷与病变进展的相关性提供理论支持，并比较不同基因表达产物的检测方法用于宫颈癌筛查及分流表现差异，从而为优化宫颈癌筛查策略提供科学依据。

方法

本研究采用回顾性分析方法，通过ddPCR检测法对山西省襄垣县、成都市妇女儿童中心医院及四川省肿瘤医院3个研究中心的402例宫颈脱落细胞学样本进行HPV E6/E7 DNA绝对定量检测。所有样本经标准化核酸提取流程处理后，同步进行qPCR与ddPCR检测，旨在分析两种方法在检测一致性和定量相关性上的差异。将HPV型别按致癌风险分为高、中、低三层，探究病毒负荷在不同宫颈病变等级中的分布特征。

在HPV E6/E7 mRNA与DNA检测效能比较研究中，纳入1607名女性受试者（902例来自宫颈癌筛查人群，705例为门诊就诊患者），使用Aptima和Onclarity平台分别对宫颈细胞样本进行检测。研究重点评估两种方法针对高级别宫颈病变（CIN2+、CIN3+）的检测效能，包括灵敏度（Se）、特异度（Sp）、阳性预测值（PPV）和曲线下面积（AUC），同时分析HPV 16、HPV 18/45等高危型别在病变进展过程中（从正常组织到宫颈癌）的Ct值和S/CO值变化趋势，以揭示病毒负荷动态变化对检测方法性能的潜在影响。统计分析使用了R软件进行（V4.3.1）。P小于0.05（双侧）认为有统计学差异。

结果

1.ddPCR定量检测研究中共纳入402例样本，平均年龄49.04岁。细胞学等级分布为：286例NILM，53例ASC-US，22例LSIL，24例HSIL，11例ASC-H，1例AGC，1例SCC，以及1例细胞学结果不满意和3例未查。组织病理诊断方面，126例样本具有可参考的病理学结果，其中18例为正常或宫颈黏膜慢性炎症，21例为CIN1，54例为CIN2，29例为CIN3，4例宫颈癌。

E6/E7 mRNA与DNA的效能对比研究中纳入1607例样本，平均年龄48.88岁。细胞学等级分布为：772例NILM，159例ASC-US，71例ASC-H，8例AGC，64例LSIL，24例HSIL，11例ASC-H，322例SCC，5例AIS和15例ADC。组织病理诊断方面，893例正常或宫颈粘膜慢性炎症，67例CIN1，48例CIN2，85例CIN3，474例SCC，40例ADC。

2.检测阳性率：在ddPCR定量检测研究中，ddPCR法与qPCR法在14 HR-HPV中的阳性检出率整体相似，HPV 45在ddPCR检测中的检出率较qPCR更高（P = 0.041）。在E6/E7 mRNA与DNA的效能对比研究中，Aptima检测任意型别阳性的阳性率在SCC、HSIL中高于Onclarity检测任意型别阳性的阳性率（P＜0.001，P = 0.021）。

3.分型检测一致性：qPCR和ddPCR检测总体一致性较好（Kappa = 0.872），且在多数HPV型别检测中表现出较好的一致性（Kappa值均大于0.800）。

4.临床检测效能：在E6/E7 mRNA与DNA的效能对比研究中，两种方法在高级别宫颈病变（CIN2+、CIN3+）检测中表现出可比的临床性能。在筛查人群中，Aptima与Onclarity检测对CIN2+和CIN3+病变的灵敏度均为100%，AUC值在CIN2+中为0.918与0.920；在CIN3+中为0.912与0.913。

5.病毒负荷在不同细胞学等级和组织病理级别的分布情况：高风险组与中风险组，ddPCR检测病毒载量标化值随细胞学等级升高而升高，随病理等级的升高而升高；qPCR的Ct值变化与ddPCR趋势不完全对应。低风险组，ddPCR检测病毒载量标化值随细胞学等级升高而升高，病理等级变化时病毒载量变化趋势不明显，qPCR 检测值在不同等级变化趋势也均不明显。在E6/E7 mRNA与DNA的效能对比研究中，HPV16、18/45以及14种高危型HPV的mRNA水平随着正常组织到癌组织逐渐增加。HPV16、45的DNA水平随着正常组织到癌组织逐渐增加，HPV18的DNA水平在癌组织中急剧上升。

结论

本研究比较了HPV E6/E7 DNA和mRNA在HPV检测中的效能，在HPV E6/E7 DNA的两种检测中，ddPCR与qPCR检测在型别间具有良好一致性。通过对病毒定量的研究，还观察到风险分层中的高、中风险HPV（HPV16、18、45；HPV31、33、52、58）的ddPCR病毒负荷随细胞学及病理等级的严重程度增加而上升。在HPV E6/E7 DNA与mRNA的对比中，发现两种基因表达产物的方法在高级别病变的检测中具有相似的效能。此外，本研究还进一步揭示了HPV DNA和mRNA水平在宫颈病变发展中的变化趋势。本研究的结果对优化宫颈癌筛查策略及多种HPV检测技术的应用提供了科学依据。

Background

Cervical cancer is one of the most common malignancies among women globally. Its development is associated with persistent infection by high-risk human papillomavirus (HR-HPV). Although the vast majority of HPV infections can be cleared by the host's immune system, persistent infections can lead to the development of cervical intraepithelial neoplasia (CIN) and cervical cancer. Molecular detection methods targeting HPV DNA or mRNA can more efficiently detect high-grade cervical intraepithelial neoplasia (CIN2+) compared to traditional cytology. Given the varying carcinogenic risks among different HPV types and the correlation between high viral load and increased cancer risk, precise HPV genotyping and viral load quantification are of significant clinical importance in the triage management of cervical cancer screening.

Current methods for HPV genotyping and viral load measurement mainly include hybrid capture and real-time quantitative fluorescence PCR (qPCR). Hybrid capture targets the full-length viral genome and is a semi-quantitative detection technique. qPCR-based methods offer more accurate quantification than hybrid capture in addition to genotyping. However, this method relies on standard curves for quantitative analysis, and most target the more conserved HPV L1 region. Aptima and Onclarity tests target HPV E6/E7 mRNA and DNA. Detection of mRNA has potential advantages in distinguishing transient from persistent infections. Digital PCR (dPCR) includes chip-based (cdPCR), microfluidic-based (mdPCR), and droplet digital PCR (ddPCR). The ddPCR method achieves absolute quantification through droplet generation and Poisson distribution calculations, without relying on standard curves.

This study discusses the distribution of HPV DNA and mRNA levels in cervical cancer and precancerous lesions at the molecular detection level, and explores the changes in viral load detected by HPV DNA- and mRNA-based detection methods at different precancerous lesion stages and their clinical detection efficacy. It provides theoretical support for understanding the correlation between viral load and lesion progression, and compares the performance differences of detection methods for different gene expression products in cervical cancer screening and triage, to provide a scientific basis for optimizing cervical cancer screening strategies.

Methods

This study adopted a retrospective analysis approach and used the droplet digital PCR (ddPCR) technique to perform absolute quantification of HPV E6/E7 DNA in 402 cervical exfoliated cytology samples from three research centers: Xiangyuan County in Shanxi Province, Chengdu Women's and Children's Central Hospital, and Sichuan Cancer Hospital. After standardized nucleic acid extraction, all samples were simultaneously tested using quantitative PCR (qPCR) and ddPCR techniques to analyze the differences in detection consistency and quantitative correlation between the two methods. HPV genotypes are classified into high, intermediate, and low oncogenic risk tiers, and explore the distribution characteristics of viral load across different cervical lesion grades.

In the comparative study on the detection efficacy of HPV E6/E7 mRNA and DNA, 1,607 female subjects were enrolled (902 from cervical cancer screening populations and 705 outpatient clinic patients), and cervical cell samples were tested using the Aptima and Onclarity platforms, respectively. The study focused on evaluating the detection efficacy of the two methods for high-grade cervical lesions (CIN2+, CIN3+), including sensitivity (Se), specificity (Sp), positive predictive value (PPV), and area under the curve (AUC). Meanwhile, the changing trends of Ct values and signal-to-cutoff (S/CO) values of high-risk types such as HPV16 and HPV18/45 during lesion progression (from normal tissue to cervical cancer) were analyzed to reveal the potential impact of dynamic viral load changes on the performance of detection methods. Statistical analysis was performed using R software (V4.3.1), and statistical significance was set at P < 0.05 (two-sided).

Results

A total of 402 samples were included in the ddPCR quantitative detection study, with an average age of 49.04. The cytology grade distribution was as follows: 286 with NILM, 53 with ASC-US, 22 with LSIL, 24 with HSIL, 11 with ASC-H, 1 with AGC, 1 with SCC, 1 with unsatisfactory result, and 3 unexamined cases. In terms of histopathological diagnosis, 126 samples had referable pathological results, 18 cases with normal or chronic inflammation of the cervical mucosa, 21 with CIN1, 54 with CIN2, 29 with CIN3, and 4 with cervical cancer. In the study comparing the efficacy of E6/E7 mRNA and DNA, 1,607 samples were included, with an average age of 48.88. The cytology grade distribution was: 772 with NILM, 159 with ASC-US, 71 with ASC-H, 8 with AGC, 64 with LSIL, 24 with HSIL, 11 with ASC-H, 322 with SCC, 5 with AIS, and 15 with ADC. In terms of histopathological diagnosis, 893 cases with normal or chronic inflammation of the cervical mucosa, 67 with CIN1, 48 with CIN2, 85 with CIN3, 474 with SCC, and 40 with ADC.

Detection positive rates: In the ddPCR quantitative detection study, the overall positive detection rates of the ddPCR and qPCR methods for 14 HR - HPV types were similar. The detection rate of HPV45 by ddPCR was higher than that by qPCR (P = 0.041). In the study comparing the efficacy of E6/E7 mRNA and DNA, the positive rate of Aptima for detecting any type was higher than that of Onclarity in squamous cell carcinoma and HSIL lesions (P＜0.001, P = 0.021).

Genotyping detection consistency: The overall consistency between qPCR and ddPCR was good (Kappa = 0.872), and most HPV type detections showed good consistency (Kappa values were all greater than 0.800).

Clinical detection efficacy: In the study comparing the efficacy of E6/E7 mRNA and DNA, the two methods showed comparable clinical performance in detecting high-grade cervical lesions (CIN2+ and CIN3+). In the screening population, the sensitivities of Aptima and Onclarity for detecting CIN2+ and CIN3+ lesions were both 100%. The AUC values were 0.918 and 0.920 for CIN2+ and 0.912 and 0.913 for CIN3+.

Distribution of viral load in different cytology and histopathology grades: In the high-risk and medium-risk groups, the viral load detected by ddPCR generally increased with the increase in cytology grade and histopathology grade. The change in Ct values of qPCR did not fully correspond to the trend of ddPCR. In the low-risk group, the viral load detected by ddPCR increased with the increase in cytology grade, but the trend of change in viral load was not obvious with the change in histopathology grade, and the trend of qPCR detection values in different grades was also not obvious. In the study comparing the efficacy of E6/E7 mRNA and DNA, the mRNA levels of HPV 16, 18/45, and 14 high-risk HPV types gradually increased from normal tissue to cancer tissue. The DNA levels of HPV 16 and 45 gradually increased from normal tissue to cancer tissue, and the DNA level of HPV 18 increased sharply in cancer tissue.

Conclusion
This study compared the performance of HPV E6/E7 DNA and mRNA in HPV detection. In the two HPV E6/E7 DNA detection methods (ddPCR and qPCR), good consistency was observed across genotypes. Through viral load, it was also found that the ddPCR viral loads of high- and medium-risk HPVs (HPV 16, 18, 45; HPV 31, 33,52, 58) in risk stratification increased with the severity of cytological and pathological grades. In the comparison between HPV E6/E7 DNA and mRNA, the two methods targeting different gene expression products showed similar performance in detecting high-grade lesions. Additionally, this study further revealed the changing trends of HPV DNA and mRNA levels during the progression of cervical lesions. The results of this study provide a scientific basis for optimizing cervical cancer screening strategies
and the application of multiple HPV detection technologies