Utilization of Guideline-Concordant Therapy for Clinical T2N0M0 Esophageal Adenocarcinoma

R. C. Jacobs¹, A. B. Chang², D. J. Vitello², J. E. Williams³, K. C. Lung², D. M. Avella Patino², A. Bharat², S. S.. Kim², D. D. Odell³, D. J. Bentrem^2
1Northwestern University Feinberg School of Medicine, Boston, Massachusetts ²Northwestern University, Chicago, Illinois ³University of Michigan, Ann Arbor, Michigan

Purpose: To 1) describe guideline-concordant utilization of treatment modalities for clinical T2N0M0 (cT2N0M0) esophageal adenocarcinoma stratified by tumor risk (high-risk: tumor size ≥3cm or high-grade histology; low-risk: tumor size < 3cm or low-grade histology), 2) evaluate predictors of guideline-concordant treatment utilization, and 3) analyze survival outcomes by treatment modality.

Methods: Patients diagnosed with cT2N0M0 esophageal adenocarcinoma were identified within the National Cancer Database from 2004-2020. Treatment modalities included esophagectomy alone, neoadjuvant chemoradiation with esophagectomy, perioperative chemotherapy with esophagectomy, and definitive chemoradiation. Multivariable Poisson regression with robust variance was used to identify predictors of guideline concordance (low-risk tumors: esophagectomy alone; high-risk tumors: neoadjuvant chemoradiation with esophagectomy, perioperative chemotherapy with esophagectomy, or definitive chemoradiation for patients who refuse or cannot tolerate surgery). A multivariable Cox proportional hazard model was estimated to evaluate the association of survival with the interaction of treatment modality and tumor risk category. Propensity score matching was done to establish cohorts of patients receiving guideline concordant versus guideline non-concordant care. A Kaplan-Meier curve with log-rank test was estimated to evaluate survival in propensity score-matched cohorts of patients receiving guideline concordant versus guideline non-concordant care.

Results: There were 3,823 patients included in analysis, with 997/3,823 (26.1%) of patients with low-risk tumors. No significant difference in treatment modalities offered among patients between low tumor risk and high tumor risk was observed (p=0.86). For patients with low-risk tumors, 216/997 (21.7%) underwent guideline concordant care (esophagectomy alone), with the rest undergoing guideline non-concordant care (202/997 (20.3%) underwent neoadjuvant chemoradiation with esophagectomy, 22/997 (2.2%) underwent perioperative chemotherapy with esophagectomy, and 218/997 (21.9%) underwent definitive chemoradiation). For patients with high-risk tumors, 883/2,826 (31.3%) underwent guideline concordant care (622/2,826 (22.0%) underwent neoadjuvant chemoradiation with esophagectomy, 80/2,826 (2.8%) underwent perioperative chemotherapy with esophagectomy, and 181/2,826 (6.4%) underwent definitive chemoradiation who refuse or cannot tolerate surgery). Higher hospital-level annual esophagectomy volume was a predictor of receipt of guideline concordant therapy (≥8 per year compared to < 1 esophagectomy per year: aRR 2.07 (1.49-2.88)), and older age was a predictor of receipt of guideline non-concordant therapy (age >75 years old compared to < 60 years old: aRR 0.54 (0.45-0.66). In propensity score-matched multivariable Cox proportional hazard modeling, patients who underwent guideline concordant care were associated with having a lower risk of death compared to those who underwent guideline non-concordant care (aHR 0.70 (0.57-0.86)).

Conclusion: Most patients with cT2N0M0 esophageal adenocarcinoma did not receive guideline concordant care. Predictors of guideline concordance include high surgical volume and age less than 60 years. Multimodality therapy is overutilized in patients with low-risk tumors. Further work investigating discordance in treatment utilization and tumor risk is necessary to optimize treatment outcomes for patients with cT2N0M0 esophageal adenocarcinoma.

Identify the source of the funding for this research project: National Cancer Institute training grant number: 5R38CA245095