Mutational Dynamics of Primary and Recurrent Esophageal Squamous Cell Carcinoma Treated with Chemoradiotherapy

Abstract

Understanding genomic alterations in recurrent esophageal squamous cell carcinoma (ESCC) following chemoradiotherapy (CRT) will lead to improved therapeutic management in patients with ESCC. Recent next generation sequencing studies have revealed the mutational landscapes of primary ESCC. However, no study has investigated the mutational profile of recurrent ESCC following CRT. The aim of this study is to clarify the mutational dynamics of primary and recurrent ESCC treated with CRT. We performed multi-region whole-exome sequencing (WES) of 11 tumors (seven primary and four recurrent specimens) and two paired-peripheral blood samples from two patients with superficial ESCC (staged as T1bN0M0) treated with definitive CRT. The radiation dose was 65.4 Gy in 35 fractions concurrent with 2 cycles of cisplatin plus 5-fluorouracil. To detect CRT-resistant subclones existing in only part of the tumor, tumor biopsy specimens were obtained by multi-region sampling during endoscopy at the time of diagnosis and relapse. DNA extracted from tumor and normal samples was sequenced using a next generation sequencer. The differences between the specific single-nucleotide variant (SNV) patterns of primary and recurrent ESCC were assessed using Fisher’s exact test. The mean numbers of mutations per sample across protein-coding exon regions in primary and recurrent ESCC were as follows: 82 (range, 77−86) and 151 (range, 141−161) in one case; 82 (range, 77−86) and 113 (range, 108−118) in another case. Notably, only 28% (range, 23−33%) of somatic mutations were consistently detectable in both primary and recurrent ESCC, indicating that recurrent ESCC has a distinct mutational profile compared to primary ESCC. Indeed, we found 140 and 99 mutations specific to recurrent ESCC in each case. To further characterize the mutational dynamics following CRT, we analyzed the patterns of SNVs in primary and recurrent ESCC. In recurrent ESCC, we observed enrichment of cytosine to adenine transversions (P < 0.05). On the other hand, cytosine to thymine transitions were the most prevalent SNV in primary ESCC, but had a decreased frequency in recurrent ESCC (P = 0.058). These dynamic SNV patterns suggest that the mutational mechanisms of recurrent ESCC differ from that of primary ESCC. Multi-region WES of paired primary and recurrent ESCC treated with CRT demonstrated that the mutational profile of recurrent ESCC is characterized by increased mutational load and distinct SNV patterns. These observations should be confirmed in a larger patient group.