Minimal residual disease profiling predicts pathological complete response in esophageal squamous cell carcinoma

Abstract

Accurate presurgical prediction of pathological complete response (pCR) can guide treatment decisions, potentially avoiding unnecessary surgeries and improving the quality of life for cancer patients. We developed a minimal residual disease (MRD) profiling approach with enhanced sensitivity and specificity for detecting minimal tumor DNA from cell-free DNA (cfDNA). The approach was validated in two independent esophageal squamous cell carcinoma (ESCC) cohorts. In a cohort undergoing neoadjuvant, surgical, and adjuvant therapy (NAT cohort), presurgical MRD status precisely predicted pCR. All MRD-negative cases (10/10) were confirmed as pCR by pathological evaluation on the resected tissues. In contrast, MRD-positive cases included all the 27 non-pCR cases and only one pCR case (10/10 vs 1/28, P < 0.0001, Fisher's exact test). In a definitive radiotherapy cohort (dRT cohort), post-dRT MRD status was closely correlated with patient prognosis. All MRD-negative patients (25/25) remained progression-free during the follow-up period, while 23 of the 26 MRD-positive patients experienced disease progression (25/25 vs 3/26, P < 0.0001, Fisher's exact test; progression-free survival, P < 0.0001, log-rank test). The MRD profiling approach effectively predicted the ESCC patients who would achieve pCR with surgery and those likely to remain progression-free without surgery. This suggests that the cancer cells in these MRD-negative patients have been effectively eliminated and they could be suitable candidates for a watch-and-wait strategy, potentially avoiding unnecessary surgery.

Keywords: Adjuvant therapy; Definitive radiotherapy; Esophageal squamous cell carcinoma; Minimal residual disease; Neoadjuvant therapy; Pathological complete response.

Fig. 1

Clinical characteristics and ctDNA fractions of the neoadjuvant therapy cohort. (A) Study overview. Patients diagnosed with locally advanced ESCC were randomized into two groups: the socazolimab + TP arm (n = 20) and the placebo + TP arm (n = 18). Following neoadjuvant treatment, all participants underwent video-assisted thoracoscopy esophagectomy. In the socazolimab + TP arm, patients with pathological residues after surgery were administered adjuvant therapy. Tumor biopsy tissue samples were collected before neoadjuvant treatment by esophagogastroscopy. Blood was collected at three distinct phases: pretreatment (baseline, T0); before surgery (Tb); and one month after surgery (Tp). TP denotes the combination of nab-paclitaxel and cisplatin. (B) Patient overview, tumor characteristics, and MRD profiling outcomes. “Stage” refers to the clinical stage of patients upon enrollment. “T stage,” “N stage,” and “M stage” denote the pathological stages of the excised tumor. Percentage of residual tumor cells, pathological evaluation of the proportion of residual tumor cells in resected tumor tissue. NA, not available. (C) ctDNA fractions at T0 grouped by pathological responses. (D) ctDNA fractions of all the cases at T0, Tb, and Tp. (E) ctDNA fractions at Tb grouped by pathological response. (F) Distribution of MRD-positive and MRD-negative cases at Tp categorized on adjuvant therapy and relapse. G and H. Kaplan-Meier survival analysis depicting progression-free survival (PFS) probability based on adjuvant therapy in (G) MRD-positive patients and (H) MRD-negative patients at Tp. P value is not provided in (H) as none of the patients exhibited progression. NS, no significance; ** P < 0.01; ***P < 0.001

Fig. 2

The MRD status and prognosis of definitive radiotherapy cohort. (A) Study design. Fifty-one patients underwent either RT alone or RT-based concurrent comprehensive treatment. Tumor biopsy samples were collected prior to treatment. Blood samples were collected after the dRT (post-dRT). RT stands for radiotherapy. (B) Overview of patient characteristics and MRD outcomes. Stage, clinical stage of patients at enrollment. (C) ctDNA fractions at post-dRT classified based on prognosis. (D) Kaplan-Meier survival analysis shows the probability of progression-free survival (PFS) as determined by MRD status of post-dRT. ****P < 0.0001