Whole-genome/exome analysis of circulating tumor DNA and comparison to tumor genomics from patients with heavily pre-treated ovarian cancer: subset analysis of the PERMED-01 trial

Abstract

IntroductionThe poor prognosis of ovarian carcinoma (OvC) is due to the advanced stage at diagnosis, a high risk of relapse after first-line therapies, and the lack of efficient treatments in the recurrence setting. Circulating tumor DNA (ctDNA) analysis is a promising tool to assess treatment-resistant OvC and may avoid iterative tissue biopsies. We aimed to evaluate the genomic profile of recurrent heavily pre-treated OvC.MethodsWe performed tumor panel-based sequencing as well as low-coverage whole-genome sequencing (LC-WGS) of tumor and plasma collected in patients with ovarian cancer included in the PERMED-01 trial. Whole-exome sequencing (WES) data of plasma samples were also analyzed and compared to mutation and copy number alteration (CNA) tumor profiles. The prognostic value [progression-free survival (PFS)] of these alterations was assessed in an exploratory analysis.ResultsTumor and plasma genomic analyses were done for 24 patients with heavily pretreated OvC [67% high-grade serous carcinoma (HGSC)]. Tumor mutation burden was low (median 2.04 mutations/Mb) and the most frequent mutated gene was TP53 (94% of HGSC). Tumor CNAs were frequent with a median of 50% of genome altered fraction. Plasma LC-WGS and WES detected ctDNA in 21/24 cases (88%) with a median tumor fraction of 12.7%. We observed a low correlation between plasma and tumor CNA profiles. However, this correlation was significant in cases with the highest circulating tumor fraction. Plasma genome altered fraction and plasma mutation burden (p = 0.011 and p = 0.041, respectively, log-rank tests) were associated with PFS.ConclusionsCombination of LC-WGS and WES can detect ctDNA in most pre-treated OvCs. Some ctDNA characteristics, such as genome altered fraction and plasma mutation burden, showed prognostic value. ctDNA assessment with LC-WGS may be a promising and non-expansive tool to evaluate disease evolution in this disease with high genomic instability.Clinical Trial Registrationhttps://clinicaltrials.gov/ct2/show/NCT02342158, identifier NCT02342158.

Concordance between whole exome sequencing of circulating tumor DNA and tumor tissue

“…In Sabatier, R. et al . [ 31 ], the average ctDNA fraction was 12.7%, and all 11 samples with Pearson’s correlation coefficients between ctDNA’s and tumor tissue’s copy number alteration profiles of 0.3 or lower were those with ctDNA fraction of less than 15%. Other studies have speculated that low ctDNA fractions were behind poor data quality [ 32 , 33 ], but the exact ctDNA fractions were not measured or reported.…”

“…For tumor DNA detection, as low as 3% ctDNA fraction was reportedly enough [ 17 ]. For mutational load calculation [ 33 ] and copy number alteration analysis [ 31 ], data quality issues began to emerge in samples with low ctDNA fractions. For WES, ctDNA fraction may be an explanation for the wide variation in reported levels of concordance between ctDNA’s and tumor tissue’s mutational profiles [ 34 , 35 ].…”

“…The ability to ensure high quality of tumor variant characterizations is important, especially for WES, as it incurs cost on the patients. Our findings indicate that a recommended cutoff for ctDNA fraction for WES of 5–10% [ 17 , 30 ] may be too low [ 31 ] and that the appropriate cutoff may be cancer tissue-specific ( Fig 2A ), possibly due to the different shedding of tumor cells from various tissues into the bloodstream [ 35 ]. Although the high concordance between variants of top cancer-related genes ( Fig 4 ) in WES of ctDNA and tumor tissue may suggest that WES of ctDNA alone is enough for clinical applications, the large numbers of unique variants in WES of both samples (S3 Table in S1 File ) indicate that performing WES on both samples can uncover considerably more information.…”

Platelet RNA enables accurate detection of ovarian cancer: an intercontinental, biomarker identification study

“…The AUC of a published model with three exosomal miRNAs for OC diagnosis was 0.8337 ( Chen et al, 2022 ). A French group reported that the plasma sequencing can detect ctDNA 88% OC cases ( Sabatier et al, 2022 ). Comparing with these models, the performance of our TEPOC is higher or comparable.…”

Emerging applications of tumour-educated platelets in the detection and prognostication of ovarian cancer

“…Although detecting early-stage may improve the prognosis of ovarian cancer, there is currently a large unmet need for diagnostic biomarkers as no screening tools have been proven effective to date ( Siegel et al, 2021 ). Studies have examined the diagnostic performance of other liquid biopsy components, including CTCs (79.4% sensitivity and 92.2% specificity) ( Wang et al, 2022 ), exosomal miRNAs (model AUC = 0.8337) ( Chen et al, 2022 ), and ctDNA (sensitivity = 88%, specificity 80%) ( Passiglia et al, 2018 ; Sabatier et al, 2022 ). Major advantages of TEPs include the extensive interactions between platelets and tumour cells, high availability in the peripheral blood, lack of interference from genomic DNA given their anuclear structure, and ease of isolation.…”