Search

Background: Cell-free DNA (cfDNA) is emerging as a potential biomarker for the detection of ovarian cancer (OC). Recently, we reported a method based upon cfDNA whole-genome sequencing data including the nucleosome distribution (nucleosome footprinting NF), terminal signature sequence (motif), DNA fragmentation (fragment), and copy number variation (CNV).In the present study, we explored whether multiomics early screening technology in cfDNA can be applied for early screening of ovarian cancer. Methods: Fifty-nine patients with OC and 100 healthy controls were included in this prospective study. Cell-free DNA was extracted from plasma and analyzed by low-pass whole-genome sequencing. Genomic features were obtained for all samples of the cohort, including copy number variation (CNV), 5'-end motifs, fragmentation profiles, and nucleosome footprinting (NF). An integrated scoring system termed the OC score was developed based on the performance of these four features. Results: All four features showed diagnostic potential for OC. Based on the unique genome features of cfDNA, the OC score has high accuracy in distinguishing OC patients from healthy controls (AUC 97.7%; sensitivity 94.7%; specificity 98.0%) as a new comprehensive diagnostic method for OC. The OC score showed a gradual trend from healthy controls to OC patients with different stages, especially for early OC monitoring of concern, which achieved a satisfactory sensitivity (85.7%) at a high specificity. Conclusions: This is the first study evaluating the potential of cell-free DNA for the diagnosis of primary OC using multidimensional early screening technology. We present a promising method to increase the accuracy of prediction in patients with OC. [ABSTRACT FROM AUTHOR]

The liquid biopsy is being integrated into cancer diagnostics and surveillance. However, critical questions still remain, such as how to precisely evaluate cancer mutation burden and interpret the corresponding clinical implications. Herein, we evaluated the role of peripheral blood cell‐free DNA (cfDNA) in characterizing the dynamic mutation alterations of 48 cancer driver genes from cervical cancer patients. We performed targeted deep sequencing on 93 plasma cfDNA from 57 cervical cancer patients and from this developed an algorithm, allele fraction deviation (AFD), to monitor in an unbiased manner the dynamic changes of genomic aberrations. Differing treatments, including chemotherapy (n = 22), radiotherapy (n = 14) and surgery (n = 15), led to a significant decrease in AFD values (Wilcoxon, p = 0.029). The decrease of cfDNA AFD values was accompanied by shrinkage in the size of the tumor in most patients. However, in a subgroup of patients where cfDNA AFD values did not reflect a reduction in tumor size, there was a detection of progressive disease (metastasis). Furthermore, a low AFD value at diagnosis followed a later increase of AFD value also successfully predicted relapse. These results show that plasma cfDNA, together with targeted deep sequencing, may help predict treatment response and disease development in cervical cancer. What's new? Liquid biopsy is increasingly being integrated into cancer diagnostics and surveillance. However, critical questions still remain, such as how to precisely evaluate cancer mutation burden and infer clinical implications. Here, the authors develop a new algorithm that evaluates in an unbiased manner the mutation deviation of every single site in the target regions covered by a cancer panel, including the sites that usually fail to pass the single nucleotide variations (SNV)‐calling criteria of most algorithms. By analyzing 93 cervical cancer samples, they demonstrate that this algorithm is able to efficiently monitor patient response to treatment and prognosticate tumor progression. [ABSTRACT FROM AUTHOR]