Clinical Outcomes Of Patients With Pathologic Residual Nodal Disease After Neo-Adjuvant Chemoradiation For Esophageal Carcinoma

Abstract

To analyze the effect of neoadjuvant chemoradiation (NA-CRT) on regional lymph nodes according to radiation field coverage and its influence on clinical outcomes. Between 2013 to 2019, 201 patients with Stage II-IVa esophageal cancer underwent NA-CRT followed by radical esophagectomy at our institution, out of which 34 patients harbored pathologic residual nodal disease. All patients received radiotherapy [median dose/fractionation: 50 Gy (Range: 39.6-50.4) / 25 Fx (Range: 22-28)] via IMRT or 3DCRT technique and concurrent weekly chemotherapy [Median cycles: 5 (Range: 1-6)]. Target design was similar to Chemoradiotherapy for Esophageal cancer followed by Surgery Study (CROSS). CT datasets with associated dose distributions for each patient were reviewed to determine whether the residual nodal disease was situated in-field or out-of-field (located outside any delineated clinical target volume). In addition, demographics (age; sex), clinical (clinical T, N, M; histology) and treatment-related (radiotherapy dose, technique and duration; chemotherapy; surgical details; ypT stage) were collected. Pathologic nodal characteristics were also collected [nodal positivity; location of positive nodes (in-field/out-of-field); extranodal extension (ENE), total nodes dissected]. After univariable analysis, to compensate for small sample size, Principal Component Analysis (PCA) (Kaiser-Meyer-Olkin test, Bartlett’s test and Correlation Matrix Determinant, set at >0.6, <0.0001 and > 0.001, respectively) was performed to select variables for Multivariable Cox Regression Modelling. The primary endpoints were overall survival (OS) and recurrence-free survival (RFS) (p < 0.05 was significant). The median follow up, OS and RFS of the entire cohort was 19 months, 24 months (95% CI: 9.4-38.6) and 21 months (95% CI: 9.6-32.4), respectively. The median number of total dissected and positive nodes were 16 (Range: 6-40) and 2 (Range: 1-8), respectively. 82.3% of patients harbored out-of-field positive nodes and the crude incidence of out-of-field and in-field positive nodes was 13.9% (50/360 nodes) and 11.4% (28/245 nodes), respectively. None of the variables significantly influenced OS or RFS on univariable analysis. PCA identified 9 candidate variables which influenced OS and RFS. On multivariable modelling, the variables influencing OS and RFS were presence of ENE (HR: 7.2; p = 0.004 and HR: 9.2; p = 0.009, respectively) and pathologic residual disease at primary site (HR: 4.8; p = 0.023 and HR: 9.5; p = 0.14, respectively). In addition, presence of out-of-field nodes (HR: 40.8; p = 0.002) also influenced RFS. In patients with residual nodal disease after NACRT, both OS and RFS is negatively influenced by ENE & residual primary disease. The presence of out-of-field nodes negatively influences RFS.