Application Of NGS In HCCS Screening And The Construction Of A Clinical Prediction Model

Preface:According to WHO statistics,colorectal cancer(CRC)is the third most common malignancy after female breast and lung cancer among new cancers worldwide.About 5%-10% of them are clearly associated with germline gene mutations and are known as hereditary colorectal cancer syndrome(HCCS).How to rapidly and accurately diagnose new CRC as HCCS has become a research hotspot worldwide.In this study,we analysed and summarised the clinical characteristics and detection rates of germline pathogenic mutations in people at high risk of HCCS,while constructing predictive models to help address how to more effectively screen for HCCS in the clinical setting.Materials and methods:1.640 patients with primary diagnosis of CRC who attended the Department of Geriatric Surgery at Xiangya Hospital,Central South University between October 2018 and November 2021 were included in this study.Patients who met the high-risk factors for HCCS were screened for inclusion in the study,and their clinical data and NGS results were combined to perform germline genetic characterization.2.The clinical data of 133 patients with a final pathological diagnosis of CRC were included,and the independent risk factors affecting the detection of HCCS-related pathogenic mutations by NGS were screened using binary logistic regression using SPSS 26.0 software,and finally a column line graph(Nomogram)was established based on the independent risk factors identified using R language The predictive value of the nomogram was evaluated using ROC curves.Results:1.A total of 110 patients met the entry criteria for HCCS risk factors,and a total of 36 HCCS germline mutations were detected,including 13 in APC,19 in mismatch repair(MMR),2 in EPCAM,1 in MYTYH and 1in STK11.A large number of mutations of unknown clinical significance were also detected.It can be found that the detection rate of germline pathogenic genes in patients meeting the high risk of HCCS is significantly higher than the incidence of HCCS previously reported in the literature.2.By performing a binary logistic regression multifactorial analysis on 133 patients with pathologically confirmed CRC,we found that age at presentation(P < 0.05),ABO blood group(P < 0.05),tumour marker CA125(P < 0.05),tumour site(P < 0.05),pathological type(P < 0.05),tumour histological grade(P < 0.05),and family history of cancer(P <0.05)were statistically significant(P < 0.05)as independent risk factors for predicting whether mutations in mutant germ line genes could be detected.The columnar line plot ROC curve using R language had an AUC = 0.983 and had good predictive value.Conclusions:1.For individuals suspected of having risk factors for HCCS,NGS testing should be recommended to determine whether they carry HCCS-associated germline mutations,and NGS-based WES testing may increase the detection rate of pathogenic germline mutations in the suspected HCCS population.2.For patients with pathologically confirmed CRC,age at presentation,ABO blood group,tumour marker CA125,family history of cancer,tumour location,pathological type and tumour histological grade are independent risk factors for the detection of HCCS-associated germline mutations in individuals.The predictive model based on these independent risk factors has a high predictive value.The predictive variables in the line graph are all assigned corresponding scores,and the total score obtained after summing the scores corresponds to their predictive probability,which is the predicted probability of detecting germline mutations.