3456 - Predicting Pathologic Complete Response following Clinical Complete Response in Esophageal Carcinoma

Purpose/Objective(s): Esophagectomy has long been considered the standard treatment for esophageal cancer, with pathologic complete response (pCR) serving as a favorable prognostic indicator. However, surgery is associated with high morbidity and even possible mortality. If clinical complete response (cCR) can predict pCR, this could potentially serve as a tool to determine whether surgery is necessary. This study aims to assess the predictiveness of cCR for pCR in esophageal carcinoma, stratified by histology, notably adenocarcinoma or squamous cell carcinoma (SCC).

Materials/Methods: This retrospective study included patients diagnosed with adenocarcinoma or SCC who underwent esophagectomy following concurrent chemoradiation (CRT) between 2015 and 2022. Descriptive statistics and graph were used for data summarization. Chi-square or t-tests were applied for group comparisons, as appropriate. Two-by-two tables were used to calculate the positive predictive value (PPV) and negative predictive value (NPV) of cCR, which was defined as complete response on PET/CT or endoscopy.

Results: Of the 105 eligible patients, 79 (75.2%) were male and 78 (74.3%) had adenocarcinoma. The average age during CRT was 63.9 years. The most common chemotherapy regimen was carboplatin and paclitaxel (82.9%). The most common radiation regimen was 50.4 Gy given in 28 fractions (79%). PET/CT was used in 97.6% patients to assess cCR. Endoscopic biopsy was used in only 47% of cases after PET/CT had confirmed PET-CR. The PPV of a cCR was 38% for adenocarcinoma (p=0.092), 100% for SCC (p<0.001), and 57% for the entire cohort (p<0.001, Table 1). The p-value is meant to represent the association between cCR and pCR. The mean interval between CRT and cCR was 37.3 days and between CRT and pCR was 66.1 days, with no significant difference by histology. Survival time did not differ significantly between patients with and without cCR (58.4 months vs. 52.1 months, p = 0.331) or between those with and without pCR (60 months vs. 51.5 months, p=0.19).

Conclusion: The data suggests that cCR may be a strong predictor for pCR in the overall cohort and SCC, but it has more limited predictive value for adenocarcinoma. The 100% PPV from SCC may be partially attributed to the smaller sample size. A larger sample size may yield more robust results, helping to refine the decision-making process regarding the need for esophagectomy.