The Effect of Diabetes on Pathologic Complete Response Among Patients With Esophageal Cancer

Diabetes is a common comorbidity in the U.S. and is associated with adverse outcomes in a variety of disease processes. Other cancer specialties have shown an association of diabetes with poor oncologic outcomes. We hypothesized that pathologic complete response (pCR) would be less likely among diabetic patients with esophageal cancer who underwent neoadjuvant chemoradiation therapy followed by esophagectomy resulting in worse overall survival (OS). We performed a retrospective chart review at 2 high-volume academic hospitals of all patients with esophageal cancer who received neoadjuvant chemoradiation therapy followed by esophagectomy from 2010-2019. Patients were excluded if they had histology other than squamous cell carcinoma or adenocarcinoma, did not receive multi-agent chemotherapy or received a radiation dose <39.6 Gy. The primary outcome of interest was pCR and secondary outcome was OS. Multivariable logistic regression was used to assess the likelihood of pCR and Cox hazard analysis was used to assess OS. In total, 244 patients met inclusion criteria: 190 (77.9%) were non-diabetic and 54 (22.1%) were diabetic. Diabetic and non-diabetic patients were similar in age, sex, institution where they received treatment, ASA class, comorbidities, histologic sub-type, clinical T and N stage, chemotherapy regimen and radiation dose. Diabetic patients were more likely to have a higher body mass index (29.1 vs 25.9, p < 0.001) and hypertension (87.0% vs 47.9%, p < 0.001). On univariable analysis, diabetes was the only factor associated with decreased likelihood of pCR (p = 0.04). Multivariable analysis showed diabetes was again the only factor associated with a decreased likelihood of pCR (OR 0.32, p = 0.03). Cox survival analysis showed that older age (HR 1.03, p = 0.02) and overall posttreatment pathologic stage 2 (HR 2.16, p = 0.03), stage 3 (HR 3.25, p < 0.001), and stage 4 (HR 5.75, p < 0.001) compared to pCR were associated with worse OS, however diabetes alone had no effect (HR 1.01, p = 0.98). This multi-institutional study shows that diabetes adversely affects pCR in patients receiving neoadjuvant treatment for esophageal cancer. Almost a quarter of patients with esophageal cancer have diabetes suggesting implications for management of these patients. Future studies are warranted to determine the optimal neoadjuvant treatment strategy for esophageal cancer patients with diabetes.