Estrid Høgdall

Validating reference-based algorithms to determine cell-type heterogeneity in ovarian cancer DNA methylation studies

AbstractInformation about cell composition in tissue samples is crucial for biomarker discovery and prognosis. Specifically, cancer tissue samples present challenges in deconvolution studies due to mutations and genetic rearrangements. Here, we optimized a robust, DNA methylation-based protocol, to be used for deconvolution of ovarian cancer samples. We compared several state-of-the-art methods (HEpiDISH, MethylCIBERSORT and ARIC) and validated the proposed protocol in an in-silico mixture and in an external dataset containing samples from ovarian cancer patients and controls. The deconvolution protocol we eventually implemented is based on MethylCIBERSORT. Comparing deconvolution methods, we paid close attention to the role of a reference panel. We postulate that a possibly high number of samples (in our case: 247) should be used when building a reference panel to ensure robustness and to compensate for biological and technical variation between samples. Subsequently, we tested the performance of the validated protocol in our own study cohort, consisting of 72 patients with malignant and benign ovarian disease as well as in five external cohorts. In conclusion, we refined and validated a reference-based algorithm to determine cell type composition of ovarian cancer tissue samples to be used in cancer biology studies in larger cohorts.

The IGF–PAPP-A–Stanniocalcin Axis in Serum and Ascites Associates with Prognosis in Patients with Ovarian Cancer

Pregnancy-associated plasma protein-A (PAPP-A) and PAPP-A2 modulate insulin-like growth factor (IGF) action and are inhibited by the stanniocalcins (STC1 and STC2). We previously demonstrated increased PAPP-A and IGF activity in ascites from women with ovarian carcinomas. In this prospective, longitudinal study of 107 women with ovarian cancer and ascites accumulation, we determined corresponding serum and ascites levels of IGF-1, IGF-2, PAPP-A, PAPP-A2, STC1, and STC2 and assessed their relationship with mortality. As compared to serum, we found highly increased ascites levels of PAPP-A (51-fold) and PAPP-A2 (4-fold). Elevated levels were also observed for IGF-1 (12%), STC1 (90%) and STC2 (68%). In contrast, IGF-2 was reduced by 29% in ascites. Patients were followed for a median of 38.4 months (range: 45 days to 8.9 years), during which 73 patients (68.2%) died. Overall survival was longer for patients with high serum IGF-1 (hazard ratio (HR) per doubling in protein concentration: 0.60, 95% CI: 0.40–0.90). However, patients with high ascites levels of IGF-1 showed a poorer prognosis (HR: 2.00 (1.26–3.27)). High serum and ascites IGF-2 levels were associated with increased risk of mortality (HR: 2.01 (1.22–3.30) and HR: 1.78 (1.24–2.54), respectively). Similarly, serum PAPP-A2 was associated with mortality (HR: 1.26 (1.08–1.48)). Our findings demonstrate the presence and activity of the IGF system in the local tumor ecosystem, which is likely a characteristic feature of malignant disease and plays a role in its peritoneal dissemination. The potential clinical implications are supported by our finding that serum levels of the proteins are associated with patient prognosis.

Identification of stably expressed microRNAs in plasma from high-grade serous ovarian carcinoma and benign tumor patients

Abstract Background Ovarian cancer is a lethal gynecological cancer and no reliable minimally invasive early diagnosis tools exist. High grade serous ovarian carcinoma (HGSOC) is often diagnosed at advanced stages, resulting in poorer outcome than those diagnosed in early stage. Circulating microRNAs have been investigated for their biomarker potential. However, due to lack of standardization methods for microRNA detection, there is no consensus, which microRNAs should be used as stable endogenous controls. We aimed to identify microRNAs that are stably expressed in plasma of HGSOC and benign ovarian tumor patients. Methods and results We isolated RNA from plasma samples of 60 HGSOC and 48 benign patients. RT-qPCR was accomplished with a custom panel covering 40 microRNAs and 8 controls. Stability analysis was performed using five algorithms: Normfinder, geNorm, Delta-Ct, BestKeeper and RefFinder using an R-package; RefSeeker developed by our study group [1]. Among 41 analyzed RNAs, 13 were present in all samples and eligible for stability analysis. Differences between stability rankings were observed across algorithms. In HGSOC samples, hsa-miR-126-3p and hsa-miR-23a-3p were identified as the two most stable miRNAs. In benign samples, hsa-miR-191-5p and hsa-miR-27a-3p were most stable. In the combined HGSOC and benign group, hsa-miR-23a-3p and hsa-miR-27a-3p were identified by both the RefFinder and Normfinder analysis as the most stable miRNAs. Conclusions Consensus regarding normalization approaches in microRNA studies is needed. The choice of endogenous microRNAs used for normalization depends on the histological content of the cohort. Furthermore, normalization also depends on the algorithms used for stability analysis.

Human papillomavirus prevalence in first, second and third cervical cell samples from women HPV-vaccinated as girls, Denmark, 2017 to 2024: data from the Trial23 cohort study

BACKGROUND Danish women vaccinated with the 4-valent human papillomavirus (HPV) vaccine (HPV types: 6/11/16/18) at age 14 in 2008 reached screening age in 2017, allowing assessment of long-term effects on prevalence, persistence and incidence of HPV infections. AIM To examine the HPV status of cervical samples over time among women vaccinated as girls. METHODS Between February 2017 and February 2024, residual material from cytology-analysed samples collected through the ‘Trial23’ study, part of the national screening programme, was tested for HPV16/18 and non-vaccine high-risk (HR) HPV types. Prevalence in first, second and third samples, and persistence and incidence between samples were calculated. RESULTS Over 7 years, 8,659 women provided at least one sample, 5,835 at least two and 2,461 at least three. In 7,800 vaccinated women, HPV16/18 prevalence was 0.4% (95% confidence interval (CI): 0.2–0.5), 0.3% (95% CI: 0.1–0.4) and 0.2% (95% CI: 0.0–0.4) in three consecutive samples. Prevalence of non-vaccine HR HPV was 32% (95% CI: 31–33), 28% (95% CI: 27–29) and 31% (95% CI: 29–33). Persistence of HPV16/18 and non-vaccine HPV among vaccinated women was 40% and 53%. In adjusted analyses comparing vaccinated vs unvaccinated women, incidence was significantly lower for HPV16/18 (adjusted relative risk (aRR) < 0.10) while incidence of non-vaccine HR HPV types was higher (aRR: 1.66; 95% CI: 1.12–2.45). No significant difference was observed for persistence. CONCLUSION Our study provides real-world evidence of stable protection against HPV16/18 infections in women vaccinated as girls. Less intensive screening seems reasonable until women vaccinated with the 9-valent vaccine reach screening age, when screening should be reconsidered.

Are we ready for translational research based on material and data from the Danish CancerBiobank and can we gain new knowledge from biobank registration?

Bio‐and GenomeBank, Denmark (RBGB) is a nationwide infra‐structure. Danish CancerBiobank (DCB) is a biobank in RBGB. The aim is to describe the degree of biological material collected and stored in DCB for patients diagnosed with primary ovarian cancer registered in The Danish Gynecologic Cancer Database (DGCD). Furthermore, to investigate the concordance between predicted organ of disease registered in RBGB at time of sampling (presumed diagnosis) with final diagnosis for patient. Data extraction from DGCD and DCB. Biological materials are present for 1.347 (62%) of 2.172 patients with primary ovarian cancer (OC). The median age of OC patients were 68 years (range: 18–90 years). Median age of patients with biological material in DCB was 67 years and for patients without biological material in DCB 69 years (p ≤ 0.0001). The histological subtypes for the 1347 OC patients with biological material were 911 (68%) serous adenocarcinoma, 97 (7%) endometrioid adenocarcinoma, 80 (6%) mucinous adenocarcinoma, 58 (4%) clear cell carcinoma, and for 201 (15%) no information were registered. For 327 patients (24%), the presumed diagnosis was hematological with a final diagnosis of OC. Using clinical data and biological material including pre‐analytical data regarding the biological material the possibility for translational research is optimal. Furthermore, information registered through daily working procedures may propose the need for additional biomarkers to aid clinicians to stratify patients to treatment in correct fast‐track packages.

Noncoding RNA (ncRNA) Profile Association with Patient Outcome in Epithelial Ovarian Cancer Cases

AbstractOvarian cancer (OC) is the second most frequent type of gynecological cancers worldwide. In the past decades, the development of novel diagnostic and prognostic biomarkers available for OC has been limited, reflecting by the lack of specificity of such markers or very costly management. Microarray expression profiling has shown very effective results in exploring new molecular markers for patients with OC. Nonetheless, most screenings are focused on mutations or expression of molecules that are translated into proteins, corresponding to only 2% of the total human genome. In order to account for the vast majority of transcripts, in the present exploratory study, we assessed the expression levels of a comprehensive panel of noncoding RNA in different subtypes of OC. We further evaluated their association with patient overall survival (OS) and aggressive forms of the disease, such as tumor type, stage, and chemotherapy resistance. By microarray profiling in a total of 197 epithelial OC patients (162 serous carcinomas, 15 endometrioid carcinomas, 11 mucinous carcinomas, and 9 clear cell carcinomas), we found two candidates, SNORA68 and SNORD74, which associated with OS and poor clinicopathological features. The overexpression of those two targets combined was correlated with shorter OS and progression-free survival. That association was further observed to correlate with a more aggressive form of the disease. Overall, the results indicate that a panel comprised of SNORA68 and SNORD74 may be clinically relevant, where patients could be offered a more individualized, targeted follow-up, given its further validation on future prospective clinical studies.

DNA Methylation in Ovarian Tumors—a Comparison Between Fresh Tissue and FFPE Samples

AbstractAmong women, ovarian cancer (OC) is one of the most severe forms of malignancy, accounting for a low 5-year survival rate, of approximately 52%. Early symptoms are unspecific and hence hard to detect. The origin of OC and its subtypes are still unclear, underlying the need for efficient diagnostic biomarkers. In that regard, epigenetics studies are emerging in cancer diagnostics, with encouraging outcomes. Among them, DNA methylation profiling has shown that the origins of the cancer epigenome are associated with molecular factors that are crucial to carcinogenesis, such as regulation of oncogenes and tumor suppressors. Furthermore, those events have been detected in abnormal cell morphology before neoplastic formation, indicating its potential crucial use in the OC diagnostics in the future. Nonetheless, studies are limited, and whether methylation analysis can be performed optimally in formalin-fixed paraffin-embedded (FFPE) preparations of OC cases is still elusive. In the present report, we investigated the performance of DNA methylation analysis in FFPE samples, compared to their matched fresh frozen tissue in a small cohort of OC samples. We found that the overall DNA methylation profile in FFPE tissue showed high concordance to that found in fresh frozen tissue, and accounting for the small cohort size, the differentially methylated sites found primarily in frozen tissue, compared to benign samples, were also reproducible in FFPE. Overall, by using samples from our current clinical setting of tissue preservation, these preliminary observations might provide insights into the clinical use of FFPE tissues in methylation studies without critically compromising the outcome.

The Diagnostic Value of Circulating Cell-Free HPV DNA in Plasma from Cervical Cancer Patients

Circulating cell-free HPV DNA (ccfHPV DNA) may serve as a marker for cervical cancer. In this study, we used digital droplet PCR (ddPCR) to detect and quantify ccfHPV DNA in plasma from patients with HPV16- or HPV18-associated cervical cancer. Blood samples from 60 patients diagnosed with cervical cancer (FIGO IA1-IVA) at Aarhus or Odense University Hospital (June 2018 to March 2020) were collected prior to treatment, and patients were subdivided into an early stage (n = 30) and a late-stage subgroup (n = 30) according to disease stage. Furthermore, blood samples from eight women with HPV16- or 18-associated premalignant conditions (CIN3), and 15 healthy controls were collected. ddPCR was used to analyze plasma from all participants. ccfHPV DNA was detected in 19 late-stage patients (63.33%), 3 early stage patients (10.00%), and none of the CIN3 patients or controls. Quantitative evaluation showed significant correlations between ccfHPV DNA level and stage, tumor score, and tumor size. Thus, our results indicate that ccfHPV DNA may not be a useful marker for early detection of cervical cancer. However, for patients with advanced stage cervical cancer, ccfHPV DNA level represents a promising tool to establish tumor burden, making it useful for establishing treatment response and monitoring the disease.

Comparative Performance of Methylation Array and Bisulfite Sequencing in Ovarian Tissue Samples and Cervical Swabs

Introduction DNA methylation has emerged as a promising tool for the early detection of ovarian cancer. Consequently, accurate and cost-effective methods for detecting DNA methylation are essential. Although the Infinium Methylation Array provides broad coverage, its high cost limits clinical utility. Bisulfite Sequencing (BS) represents a potential alternative for biomarker validation and diagnostic assay development, provided it can reliably reproduce array-based methylation profiles. This study aims to assess the concordance between BS and Infinium Methylation Array data in ovarian cancer tissues and cervical swabs. Methods DNA from 55 ovarian cancer tissues and 25 cervical swabs underwent bisulfite conversion and was analyzed using the Infinium Methylation Array and a custom BS panel. We compared the results, focusing on overall methylation levels, Spearman correlation between beta values, and Bland-Altman analysis. We also assessed whether sample clustering patterns by diagnosis were consistent across methods. Results Methylation profiles generated by bisulfite sequencing were consistent with those obtained using the Infinium Methylation Array. We observed strong sample-wise correlation between platforms, particularly in ovarian tissue samples. Agreement was slightly lower in cervical swabs, likely due to reduced DNA quality. Diagnostic clustering patterns were broadly preserved across both methods. Conclusion Our results show that BS can reliably replicate results from the Infinium Methylation Array and presents a cost-effective option for analyzing larger sample sets. Moreover, our work may serve as a best-practice guide, as it highlights key challenges associated with working with sequencing library preparation.

No Bacterial Biomass Detected in Tissue From Patients With Ovarian Cancer and Serous Tubal Intraepithelial Carcinomas Using 16S rDNA Sequencing

The ovarian oncobiome is subject to increasing scientific focus, but a potential link between bacterial dysbiosis and ovarian carcinogenesis remains controversial. Our primary aim was to characterize the bacterial microbiota in epithelial ovarian cancer samples. Secondarily, we aimed to compare results from the bacterial microbiota in formalin‐fixed, paraffin‐embedded ovarian tissue samples from 194 patients with epithelial ovarian cancer, fallopian tube tissue samples from 16 patients with serous tubal intraepithelial carcinomas and in benign fallopian tube tissue samples from 25 patients. Bacterial abundance was investigated by means of 16S rDNA Next Generation Sequencing. The 16S rDNA sequencing run produced a very low number of bacterial reads. Only seven samples displayed bacterial reads above 5000. Validation of detected bacterial reads by qPCR was negative and indicative of results being background amplification. Our results indicate no amount of bacterial biomass in the ovarian cancer, benign fallopian tube and in the samples with serous tubal intraepithelial carcinomas. The findings underline the importance of validating results from bacterial microbiota studies with additional technical assays before any conclusion may be drawn.