br diagnoses at each cancer program
diagnoses at each cancer program per year: <5, 5-9, 10-19, and ³20. The papers analyzed by Markar et al used 5-10 ER annually as low volume and 20 ER and above annually as high volume (6). Pathologic and Perioperative Outcomes
N1, N2, N3, Nx), M-stage (M0, M1, Mx), overall TNM stage (1, 2, 3, unknown, AJCC 7th edition), grade (well differentiated, moderately differentiated, poorly differentiated, undifferentiated, unknown), tumor size (<3 cm, 3-5 cm, ³5 cm), pathologic margins (negative, positive, unknown), length of stay, readmission, 30-day mortality, 90-day mortality, and overall survival in 5, 10, and 15 years. Patients deceased within 30 and 90 days from the date of operation were included in the full cohort. Pathologic variables in the definitive CR group were based upon clinical staging.
All analyses were performed using SAS, version 9.4 (SAS Institute, Inc. Cary, North Carolina). Baseline patient characteristics and post-operative outcomes were compared in each treatment modality. After removing unknown values, a chi-square test was used for categorical variables, and ANOVA was used for continuous variables. The Kaplan-Meier method was used to generate survival curves and they were compared using the log-rank test. After excluding T-stage, N-stage, and M-stage due to multicollinearity with overall AJCC Stage category, a multivariable Cox proportional hazard model using stepwise selection was applied to estimate hazard ratios (HR) of predictors of survival time. Following univariate analysis for multiple independent variables, those with significant trend P<0.05 were selected for multivariable models. Margin status was also excluded from the model as we considered this to be a performance measure for facility type. Survival analyses were adjusted for treatment category, facility type, annual facility volume, age, sex, race, education, living location, insurance status, income, year of diagnosis, Charlson-Deyo score, tumor size, distance of tumor from incisors, tumor differentiation, and overall AJCC stage. All comparisons were 2-sided and statistical significance was defined as a p<0.05. Bonferroni adjustment was made for multiple comparisons. Multivariable logistic regression models were evaluated for discrimination using the cumulative C index (0.62).
In unadjusted analysis, patients in the neoadjuvant therapy group were more likely to have tumors located 30 cm or more from the incisors and less likely to have positive margins (all p<0.05) whereas patients in the definitive CR and adjuvant therapy groups were more likely to have tumors that were T3/T4, 4μ8C positive, and AJCC stage 3 (all p<0.05). Patients in the neoadjuvant therapy group had a significantly shorter median length of hospital stay (11 vs. 14 days; p<0.0001) and were less likely to be readmitted (6.2% vs. 11.2%; p<0.0001) compared to patients in the surgery alone groups. Additionally, patients in the neoadjuvant therapy group had a significantly lower 30-day (4.8% vs. 8.4%; p<0.0001) and 90-day (11.2% vs. 16.6%; p<0.0001) mortality compared to the esophagectomy alone group, but not the adjuvant therapy group (0.7% and 3.7%, respectively) (Table 1).
In a subgroup analysis stratifying patient into each treatment group, annual surgical volume remained a significant predictor of mortality on multivariable analysis in all treatment groups except adjuvant therapy (p=0.3219) (Figure 4A). In the neoadjuvant therapy group, compared to patients treated at facilities performing greater than 20 esophagectomies, patients treated at facilities performing 10-19 esophagectomies (HR 1.553, 95% CI 1.077-2.252), 5-9 esophagectomies (HR 1.406, 95% CI 1.002-1.996), and <5 esophagectomies (HR 1.569, 95% CI 1.168-2.147) per year were associated with
In cilia retrospective study, using the NCDB, we evaluated current trends in the treatment of ESCC and the impact of different treatment modalities on long-term survival outcomes. Eighty percent of patients did not undergo surgical therapy and were treated with definitive CR despite all patients having AJCC Stage 3 or less disease. Neoadjuvant therapy was utilized in only 9% of patients and was associated with a significant improvement in 5- and 10-year overall survival (33.6% and 22.1%, respectively) adjusting for patient, facility, and tumor related characteristics compared to patients receiving definitive CR (20.4% and 11.4%, respectively), esophagectomy alone (28.3 and 17.6%, respectively), and esophagectomy followed by adjuvant therapy (30.3% and 19.2%, respectively).
EC is associated with a poor prognosis with 5-year relative survival rates of 40% for local disease and 22% for regional disease (7). Although some studies suggest that multimodality treatment with preoperative chemotherapy, radiotherapy, or combined CR may improve overall survival in patients with EC compared to surgery alone (8, 9), others have demonstrated equivocal results (10-13). In a recent meta-analysis of 27 randomized controlled trials, Pasquali et al. (14) demonstrated that survival was significantly improved in patients with ESCC receiving neoadjuvant CR (HR 0.82, 95% CI 0.72-0.93) followed by surgery or neoadjuvant chemotherapy followed by surgery (HR 0.89, 95% CI