Mutation profiling, tumour burden assessment, outcome prediction and disease monitoring by circulating tumour DNA in peripheral T-cell lymphoma

In this prospective study, we aimed to evaluate the utility of circulating tumour DNA (ctDNA) in peripheral T-cell lymphomas (PTCLs). Plasma cell-free DNA (cfDNA) was obtained, and the mutational profile was assessed in 47 patients with newly diagnosed mature T- and NK-cell lymphoma. To validate the mutations detected in cfDNA, paired tumour tissue samples were available for 36 patients. Targeted next-generation sequencing was performed. A total of 279 somatic mutations involving 149 genes were identified in the 47 cfDNA samples. The overall sensitivity of plasma cfDNA in discovering biopsy-confirmed mutations was 73.9% with a specificity of 99.6%. When we considered only mutations with variant allele frequency >5% in the tumour biopsy, the sensitivity increased to 81.9%. Pretreatment ctDNA concentration and the number of mutations were highly correlated with tumour burden indicators including lactate dehydrogenase, Ann Arbor stage and International Prognostic Index score. Patients with high ctDNA level (>1.9 log ng/mL) had significantly lower overall response rates, inferior 1-year progression-free survival and overall survival rates than those with low level. Longitudinal analysis of ctDNA showed a strong agreement between ctDNA dynamics and the radiographic response. In conclusion, our study demonstrates that ctDNA may serve as a promising tool for mutational profiling, tumour burden assessment, outcome prediction and disease monitoring in PTCLs.

Automated summary

In this study, the utility of circulating tumor DNA (ctDNA) in peripheral T-cell lymphomas (PTCLs) was evaluated. Plasma cell-free DNA (cfDNA) was obtained and mutational profiling was conducted in 47 patients with newly diagnosed PTCLs. The results showed that ctDNA had high sensitivity and specificity for detecting mutations confirmed by tumor biopsy. Furthermore, ctDNA concentration and number of mutations were strongly correlated with tumor burden indicators and could predict patient outcomes. Longitudinal analysis of ctDNA also showed agreement with radiographic response. Overall, this study suggests that ctDNA may be a promising tool for mutational profiling, tumor burden assessment, outcome prediction, and disease monitoring in PTCLs.