Screening the key pathogenic mutations in Chinese patients with colorectal cancers using liquid and tissue biopsy.

Abstract

e15637 Background: Colorectal cancer (CRC) is the third most common cancer worldwide. Over the past decades, significant progress has been made in the clinical treatment of CRC. Targeted therapy is an optional approach that has successfully prolonged overall survival for CRC patients. It is necessary to determine the personalized targets for guiding the usage of specific drugs, which requires biopsy to obtain pathological samples for analysis. Both tissue biopsy and liquid biopsy are commonly used in clinical diagnosis, while each has its own advantages and limitations. The mutation detection rate of tissue biopsy is usually higher, and more genes can be detected, but most tissue biopsies require surgery to obtain samples and may be biased due to the intra-tumor heterogeneity. Liquid biopsy is less harmful to patients, but the detection sensitivity is much lower than tissue biopsy. Methods: In this study, 353 paraffin tissue section samples were used as tissue biopsy samples and 112 blood samples were used as liquid biopsy samples. All samples are from Chinese patients and are collected from December 2021 to August 2022. Whole-genome sequencing was performed, and several processes have been conducted to eliminate noise, duplicates and reads with low depth, therefore the gene mutations detected in the samples were considerably reliable. We selected APC, TP53, KRAS, PIK3CA and BRAF as the key pathogenic genes for CRC, analyzed the mutation detection results in tissue and blood samples. The detection difference between two types of biopsies was assessed using Pearson's Chi-squared test. Results: Among all samples, 21 (5.95%) tissue biopsy samples and 37 (33.04%) liquid biopsy samples did not contain any key pathogenic genes. For all biopsy samples, the detection rates for APC in tissue biopsy samples (n = 236, 66.86%) and liquid biopsy samples (n = 49, 43.75%) demonstrate significant difference (p = 2.02e-05). The differential detection rates were also observed in TP53 (n = 223 vs. n = 51, 63.17% vs. 45.54%, p = 0.0014) and KRAS (n = 183 vs. n = 39, 51.84% vs. 34.82%, p = 0.0024). There is no significant difference in detection rates of PIK3CA (n = 77 vs. n = 15, 21.81% vs. 13.39%, p = 0.0699) and BRAF. Only 25 (7.08%) tissue biopsy samples and 8 (7.14%) liquid biopsy samples were found to detect BRAF (p = 1). Conclusions: Generally, there is no doubt that the overall detection sensitivity of tissue biopsy is higher than that of liquid biopsy for CRC, and our analysis provided the statistical support. But for some relatively rare disease-causing genes, there is not much difference between the two biopsy methods. For patients, tissue biopsies can be more painful than liquid biopsies, so how to improve the detection of liquid biopsy in CRC is still a subject that requires efforts.