Roles of Cancer Histology Type and HPV Genotype in HPV ctDNA Detection at Baseline in Cervical Cancer: Implications for Tumor Burden Assessment

Human papillomavirus (HPV) is the most common sexually transmitted infection affecting the anogenital tract. HPV infects mucosal epithelial tissues [1]. It is associated with a variety of conditions, ranging from innocuous lesions to cancer. The majority of HPV infections resolve spontaneously without causing harmful conditions. The immune system is typically able to clear the virus within several years. However, HPV infections can persist for a longer duration. Such a persistent infection can increase the risk of developing cervical cancer [2].

Currently, more than 200 genotypes of HPV have been identified. More than 40 genotypes can infect the anogenital tract. Approximately 20 genotypes are commonly encountered in clinical practice. HPV infection is the etiologic agent of almost all cervical cancers. The International Agency for Research on Cancer has classified HPV genotypes into three categories based on their association with cervical and anogenital cancers: high oncogenic risk, low oncogenic risk, and unclear oncogenic risk [3]. Among high-risk genotypes, HPV16 and HPV18 are the most common causes of HPV-associated cervical cancer, accounting for approximately 70% of HPV-associated cervical cancers [4]. Most cases of squamous cell carcinoma and adenocarcinoma are caused by HPV infection. However, the genotype spectrum of adenocarcinoma is narrower than that of squamous cell carcinoma, with HPV18 being proportionally more important [5‒7]. From limited data, HPV18 is less common in low-grade squamous precancer [8] but is more common in early or rapidly progressing cervical cancer. It is associated with a high frequency of integration of HPV DNA into the human genome [7, 9].

Circulating tumor DNA (ctDNA) is a tumor-derived fraction of cell-free DNA. It is widely used as a reliable biomarker for tumor monitoring. Its applications include assessing treatment response, detecting residual disease after surgery, monitoring resistance, diagnosing cancer, profiling molecular characteristics, selecting treatment, and understanding tumor heterogeneity in different cancer types [10‒12].

HPV ctDNA can be detected in the blood of patients with HPV-associated tumors. This is because the HPV viral genome can be present in cells as an episome or integrated into human DNA [13, 14]. Studies have demonstrated the clinical value of HPV ctDNA [15, 16]. Its efficacy as a predictive and prognostic biomarker has been validated not only in cervical cancer [14, 17] but also in anal and oropharyngeal cancers [18‒20]. The advent of liquid biopsy and the detection of HPV ctDNA in plasma have provided a noninvasive diagnostic tool and a potential prognostic biomarker for cervical cancer.

Droplet digital PCR (ddPCR) is a sensitive technique for detecting ctDNA, including HPV ctDNA. It is well recognized for its precision, even when sample quantities are small [21, 22]. While the application of ddPCR in measuring HPV ctDNA has been widely acknowledged, preoperative (baseline) detection rates of HPV ctDNA in cervical cancer have exhibited a broad range, ranging from 31% to 100% [14, 23‒25]. Previous studies on HPV ctDNA have mainly concentrated on its ability to monitor residual tumors after treatment and assess recurrence [26, 27]. The main finding regarding baseline HPV ctDNA is its association with tumor burden [26]. However, observed differences in baseline ctDNA detection rates across studies raise questions about potential contributing factors beyond tumor burden. In this study, we investigated factors that might influence baseline HPV ctDNA detection, focusing on histology type and high-risk HPV genotypes (HPV16 and HPV18). Results of this study might improve clinical utility of baseline HPV ctDNA analysis in assessing tumor burden. We also explored the potential relationship between baseline HPV ctDNA status and progression-free survival (PFS) in cervical cancer. The focus of this study was on identifying factors affecting the detection of HPV ctDNA in cervical cancer through a baseline study. Results of this study might serve as a reference for future research on prospective disease surveillance studies based on HPV ctDNA.

This study included 92 patients diagnosed with HPV16- or HPV18-associated cervical cancer. These patients were histologically confirmed to have 2018 International Federation of Gynecology and Obstetrics (FIGO) stage IA2–IIIC2 cervical squamous cell carcinoma, adenocarcinoma, or adenosquamous carcinoma from 2013 to 2020. At the start of this study, some patients were staged following the FIGO 2009 classification of cervical cancer and transitioned to the FIGO 2018 classification. Only patients with either HPV16- or HPV18-associated cervical cancer confirmed by HPV test, including PCR, DNA chip and hybrid capture, from Pap cytology test were included.

We used blood plasma collected at baseline prior to treatment, including surgery and chemoradiation therapy, from 92 patients with HPV16- or HPV18-associated cervical cancer. Blood samples analyzed in this study were collected during the preoperative evaluation process prior to surgery. All samples were collected as part of a standardized preoperative procedure. Peripheral blood was collected into EDTA tubes. Plasma was immediately separated and stored at −195°C.

Blood plasma and clinical data used in this study were provided by the Biobank of Seoul National University Hospital, a member of Korea Biobank Network. This study was approved by the Institutional Review Board (IRB) of Seoul St. Mary’s Hospital (IRB approval number: KC21SIDI0441) and Seoul National University Hospital (IRB approval number: H-2106-206-1231).

DNA was isolated from 700 μL of plasma using an ARA MagNA Cell-Free DNA Isolation Kit (LAS, Gimpo, Republic of Korea). ddPCR was performed to detect circulating high-risk genotypes HPV16 and HPV18. Isolated cfDNA was assayed for HPV16 ctDNA or HPV18 ctDNA using a QX200 system (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Each reaction was performed in a total volume of 20 μL, containing 2X ddPCR^™ Supermix for Probes (no dUTPs, Bio-Rad Laboratories), HPV16 E7 or HPV18 E7 specific primers at 450 nM each, and the maximum volume of the sample template. The reaction mix was loaded into a QX200 Droplet Generator to produce nanoliter-sized droplets. These droplets were transferred to a plate in a C1000 Touch Thermal Cycler (Deep Well) (Bio-Rad Laboratories). Cycling conditions for PCR were as follows: 1 cycle of enzyme activation at 95°C for 10 min, 40 cycles of denaturation at 94°C for 30 s and annealing at 60°C for 1 min, and 1 cycle of enzyme deactivation at 98°C for 10 min, followed by cooling and storage at 4°C. The QX200 Droplet Reader (Bio-Rad Laboratories) and QuantaSoft software were used to measure the number of positive and negative droplets and estimate HPV DNA copies/μL.

HPV16 ddPCR assays were performed for HPV16-associated cervical cancer, while HPV18 ddPCR assays were performed for HPV18-associated cervical cancer. Both HPV16 and HPV18 ddPCR assays were performed for cases where cervical cancer was co-associated with both HPV16 and HPV18. Primers and probes used for HPV16 ddPCR assays were as follows: forward primer 5′-AGG ATG AAA TAG ATG GTC CA-3′, reverse primer 5′-GTA GAG TCA CAC TTG CAA CA-3′, and probe 5′-TGG ACA AGC AGA ACC GGA CAG AGC CC-3′. HPV18 ddPCR assays utilized the following primers and probes: forward primer 5′-GA AGC CAG AAT TGA GCT AGT-3′, reverse primer 5′-CAG AAA CAG CTG CTG GAA TG-3′, and probe 5′-AGT AGA AAG CTC AGC AGA CGA CCT TCG AG-3′. Based on these primer/probe designs, the amplicon size was 88 bp for HPV16 and 70 bp for HPV18. Distilled water was used as a negative control. We confirmed that HPV ctDNA was not detected in distilled water using either the HPV18 primer/probe set or the HPV16 primer/probe set. As pretests, HPV16 ddPCR assays were performed for five HPV16-associated advanced stage cases and HPV18 ddPCR assays were performed for one HPV18-associated advanced stage case. HPV ctDNA was detected in these 6 cases. A sample was considered HPV ctDNA positive if at least one single droplet containing HPV amplicons was detected.

Chi-squared test or Fisher’s exact test was used to determine statistical significance for categorical data. The Jonckheere-Terpstra test was used to analyze the trend in concentrations across the FIGO 2018 stage. PFS was defined as the time from starting treatment to tumor recurrence or death. Kaplan-Meier curves and log-rank tests were used to compare PFS. All statistical analyses were performed using R software version 4.3.2. A p value of less than 0.05 was considered statistically significant.

Among the total 92 cervical cancer cases, 78 (84.8%) were early stage (stage FIGO ≤ IB2) and 14 (15.2%) were locally advanced stage (stage FIGO IB3–3C2). The most common histology type was squamous cell carcinoma with 68 (73.9%) cases, followed by adenocarcinoma with 21 (22.8%) cases and adenosquamous cell carcinoma with 3 (3.3%) cases. The number of HPV16-associated cancers was 67 (72.8%). Given the proportional minority of HPV18, all HPV18-associated cancers, including cases in which cervical cancer was associated with both HPV16 and HPV18, were regarded as HPV18-associated cancers for statistical analysis. The number of HPV18-associated cancers was 25 (27.2%), including 5 cases where cervical cancer was associated with both HPV16 and HPV18. In addition to HPV16 and HPV18, other genotypes were identified in 11 samples. Detailed clinical and pathologic data of the 92 patients are presented in Table 1.

The median number of accepted droplets was 17,519 (range: 9,665–20,707). According to our study definition, which considered a sample to be HPV ctDNA positive if at least one single droplet containing HPV amplicons was detected, HPV ctDNA was detected at baseline in 38 (41.3%) but not detected in 54 (58.7%) of the 92 samples.

The HPV ctDNA detection rate at baseline was significantly higher in the advanced stage group (71.4%) than in the early stage group (35.9%) (p = 0.028). The concentration (copies/20 μL) of HPV ctDNA was also significantly higher in the advanced stage group (p < 0.001) (Fig. 1a, b).

In HPV16-associated cancers, HPV ctDNA positivity at baseline was 34.3%. In HPV18-associated cancers, HPV ctDNA positivity was significantly higher at 60.0% (p = 0.048) (Fig. 2a). In each histologic type of cervical cancer, either squamous cell carcinoma or adenocarcinoma (including adenosquamous carcinoma), HPV ctDNA positivity at baseline was higher in cancers associated with HPV18 than in those associated with HPV16. In a subgroup analysis, the difference was significant in the adenocarcinoma (including adenosquamous) group (p = 0.047) but not in the squamous cell carcinoma group (p = 0.550) (Fig. 2b). In 5 cases associated with both HPV16 and HPV18, HPV16 ctDNA was detected in only 1 case, while HPV18 ctDNA was detected in 4 cases. Even in the case where both genotypes HPV ctDNA were identified (case number 63, Table 2), the concentration of HPV18 ctDNA was higher at 3.3 copies/μL than that at 1.5 copies/μL for HPV16 ctDNA. In the analysis by FIGO stage, HPV ctDNA positivity at baseline was higher for HPV18-associated cancers than for HPV16-associated cancers in both early and advanced stage groups. However, in a subgroup analysis, the difference was not significant in either stage (Fig. 2c).

The detection rate of adenocarcinoma including adenosquamous carcinoma (adenocarcinoma/adenosquamous carcinoma) was higher at 54.2% than the detection rate of squamous cell carcinoma at 36.8%, although the difference was not statistically significant (p = 0.212) (Fig. 3a). The detection rate of adenocarcinoma/adenosquamous carcinoma was consistently higher than the detection rate of squamous cell carcinoma across HPV16 and -18 genotype groups, although the difference was not significant in either histology group (Fig. 3b). In the analysis by FIGO stage, adenocarcinoma/adenosquamous carcinoma had a higher detection rate than squamous cell carcinoma in both early and advanced stages, although the difference was not statistically significant in either stage (Fig. 3c).

The median follow-up duration from the initial diagnosis was 38.6 months (range: 0.8–108.8 months) in the entire cohort. Twelve (13.0%) cases showed recurrence and 1 (1.1%) case died. The median follow-up duration from the initial diagnosis was 38.1 months (range: 0.8–100.7 months) in the detected group. The median follow-up duration from the initial diagnosis was 39.0 months (range: 8.3–108.8 months) in the undetected group. Kaplan-Meier curves showed no significant difference in PFS between detected and undetected groups (p = 0.240) (Fig. 4).

This study found that HPV ctDNA detection rates at baseline in cervical cancer were influenced by histology type and HPV genotype, in addition to FIGO stage. Adenocarcinoma had a higher HPV ctDNA detection rate at baseline than squamous cell carcinoma. HPV ctDNA positivity at baseline was significantly higher for HPV18 than for HPV16.

Previous studies have shown that levels of HPV ctDNA at baseline in cervical cancer patients are correlated with FIGO stage [23, 26, 28, 29]. Our findings showed that locally advanced cervical cancers had higher HPV ctDNA detection rates at baseline than the early-stage cervical cancer, supporting the premise that HPV ctDNA at baseline could serve as a valuable tool in assessing tumor burden and disease progression. The correlation between HPV ctDNA level at baseline and FIGO stage highlights the potential usefulness of this biomarker for reflecting tumor burden. However, the variability in ctDNA detection rate at baseline across studies [14, 23‒25] suggests that factors beyond tumor stage and burden can affect HPV ctDNA dynamics. Although HPV ctDNA at baseline is a biomarker for tumor burden, its interpretation should be considered within a broader context.

This analysis focused on patients with HPV16- and HPV18-associated cancers, which account for approximately 70% of cervical cancer patients [4]. Our study found a significant difference in ctDNA detection rate at baseline between HPV genotypes 16 and 18. Notably, the HPV ctDNA positivity for HPV18 was higher, even when considering different FIGO stages. This indicates a potential HPV genotype-specific variation in baseline HPV ctDNA level and suggests that HPV genotype can play a critical role in HPV ctDNA dynamics. This may be related to different oncogenic mechanisms of different HPV genotypes that influence tumor progression and ctDNA release patterns. The most frequent oncogenic HPV genotypes, 16 and 18, have been found in approximately 50% and 20% of cervical cancers, respectively [30]. While HPV16 is the most common genotype in cervical cancer, HPV18 is generally associated with more aggressive disease and poorer survival [31‒33] and high frequency of integration of HPV DNA into the human genome [7, 9]. Its aggressive and high integration into the human genome might have contributed to our results. However, previous studies on cervical cancer have reported lower HPV ctDNA positivity for HPV18 than for HPV16 [23, 24, 28], while others found no difference in detection rate by HPV genotype [26]. These studies did not consider tumor burden or histology type, which could affect HPV ctDNA when analyzing HPV genotype. In contrast, our study adjusted for FIGO stage and histology type and found that the higher baseline HPV ctDNA detection rates in HPV18 remained consistent. Therefore, further studies are needed to investigate HPV ctDNA at baseline according to HPV genotype using a larger number of samples and adjusting for tumor stage and histology type.

Our results showed higher HPV ctDNA detection rates in adenocarcinoma than in squamous cell carcinoma, although the difference was not significant based on the p value. However, this finding of higher detection rate in adenocarcinoma/adenosquamous carcinoma was consistent in both early and advanced stages and in cervical cancers associated with either HPV16 or HPV18 genotype. Although differences in these findings were not significant, the small sample size might have influenced these results. Thus, further studies with larger samples are needed. The variation in ctDNA detection rate might be due to inherent biological differences between adenocarcinoma and squamous cell carcinoma that can affect the release of ctDNA into the bloodstream. In contrast to our results, one study has reported a higher frequency of HPV ctDNA detection at baseline in squamous cell carcinoma than in adenocarcinoma [23]. Another study has shown no significant association between baseline HPV ctDNA level and histology type [26]. Therefore, additional in-depth studies focusing on HPV ctDNA at baseline in relation to histology types are needed to gain a better understanding.

In the analysis of PFS, baseline HPV ctDNA detection did not show a significant prognostic value. However, high baseline HPV ctDNA is known to be associated with an increased risk of progression [24]. The small sample size of our study and the uneven distribution of patients into specific FIGO stages might have contributed to these different results.

Squamous cell carcinoma antigen has been used to monitor squamous cell carcinoma [34]. Carcinoembryonic antigen (CEA) and cancer antigen 125 (CA 125) can be used to monitor disease in HPV-associated cervical adenocarcinoma [35]. While squamous cell carcinoma antigen, CEA, and CA 125 can be useful in the surveillance and management of some patients, their clinical value is limited because they are expressed in only a small proportion of patients. In addition, in some cases, they lack specificity. Integrating HPV ctDNA analysis with traditional serum biomarkers could improve the accuracy of cervical cancer surveillance and management.

A limitation of our study is the lack of postoperative HPV ctDNA results. It is known that posttreatment HPV ctDNA levels are important for identifying residual tumors and monitoring recurrence; however, our study was conducted using preoperative samples only. Therefore, further studies are needed to determine whether our findings are applicable to posttreatment HPV ctDNA levels. Understanding the impact of histology type and HPV genotype could provide a valuable foundation for future research in disease monitoring and tailoring posttreatment surveillance strategies to adjust for these specific factors. However, the number of samples analyzed for results was small. Thus, further research studies with larger sample sizes and samples with posttreatment are needed. Our study focused mainly on histology type and specific HPV genotypes, HPV16 and HPV18. Other additional high-risk HPV genotypes need to be studied in the future. HPV ctDNA levels at baseline also might be influenced by other yet-undiscovered factors, including HPV ctDNA isolation methods, other HPV genotypes, and additional clinical factors not evaluated in this study. There was also a possibility of false-positive results in HPV ctDNA analysis due to the presence of nontumor HPV ctDNA such as HPV ctDNA from normal cells or other types of tumors in the bloodstream. A comparative analysis of baseline HPV ctDNA levels in blood samples against nontumor samples is needed to help us understand the specificity and relevance of the results.

In conclusion, our study highlights the complexity of using HPV ctDNA at baseline as a biomarker in tumor burden for cervical cancer. HPV18 and adenocarcinoma were associated with higher baseline HPV ctDNA detection rates. The variation in ctDNA level, which was influenced by both histology and HPV genotype, suggested a need for different approaches when using this biomarker in clinical practice. This study provides an understanding of the influence of histology type and HPV genotype on the detection of HPV ctDNA at baseline in cervical cancer patients. This understanding could also provide a valuable basis for disease monitoring in a posttreatment setting and highlights the need for a considered approach for using this biomarker in clinical practice.

The Institutional Review Board (IRB) of Seoul St. Mary’s Hospital (IRB approval No. KC21SIDI0441) and Seoul National University Hospital (IRB approval No. H-2106-206-1231) approved this study and waived the requirement to obtain written informed consent due to the retrospective nature of this study. All procedures were performed in accordance with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

The authors have no conflicts of interest to disclose.

This work was supported by a grant (No. 2020R1F1A1070341) of the National Research Foundation (NRF) funded by the Korea government (MSIT of Korea). This work was also supported by Seoul St. Mary’s Hospital. Biospecimens and data used in this study were provided by the Biobank of Seoul National University Hospital, a member of Korea Biobank Network. The funder had no role in the study design, data collection, data analysis, or reporting of this study.

The study’s conception and design was contributed by, previous versions of the manuscript were commented by, and final manuscript was and read approved by all authors. Material preparation, data collection and analysis were performed by M.L., E.L., J.K., K. L., S.L., S.H., H.K., and A.L. The first draft of the manuscript was written by M.L.

Miseon Lee and Eun Ji Lee contributed equally to this work.

The data that support the findings of this study are not publicly available due to their containing information that could compromise the privacy of research participants but are available from M.L. upon reasonable request.