Open AccessPatients with high body mass index tend to have lower stage of esophageal carcinoma at diagnosis
Author(s) -
Hayashi Y.,
Correa A. M.,
Hofstetter W. L.,
Vaporciyan A. A.,
Mehran R. J.,
Rice D. C.,
Suzuki A.,
Lee J. H.,
Bhutani M. S.,
Welsh J.,
Lin S. H.,
Maru D. M.,
Swisher S. G.,
Ajani J. A.
Publication year - 2011
Publication title -
diseases of the esophagus
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.115
H-Index - 63
eISSN - 1442-2050
pISSN - 1120-8694
DOI - 10.1111/j.1442-2050.2011.01290.x
Subject(s) - medicine , body mass index , gastroenterology , cohort , stage (stratigraphy) , esophageal cancer , adenocarcinoma , reflux , cancer , disease , biology , paleontology
SUMMARY High body mass index (H‐BMI; ≥25 kg/m 2 ) is common in US adults. In a small cohort of esophageal cancer (EC) patients treated with surgery, H‐BMI and diagnosis of early stage EC appeared associated. We evaluated a much larger cohort of EC patients. From a prospectively maintained database, we analyzed 925 EC patients who had surgery with or without adjunctive therapy. Various statistical methods were used. Among 925 patients, 69% had H‐BMI, and 31% had normal body mass index (<25 kg/m 2 ; N‐BMI). H‐BMI was associated with men ( P < 0.001), Caucasians ( P = 0.064; trend), lower esophageal localization ( P < 0.001), adenocarcinoma histology ( P < 0.001), low baseline cT‐stage ( P = 0.003), low baseline overall clinical stage ( P = 0.003), coronary artery disease ( P = 0.036), and diabetes ( P < 0.001). N‐BMI was associated with weight loss ( P < 0.001), alcohol abuse ( P = 0.056; trend), ever/current smoking ( P = 0.014), and baseline cN+ ( P = 0.018). H‐BMI patients with cT1 tumors ( n = 110) had significantly higher rates of gastresophageal reflux disease symptoms ( P < 0.001), gastresophageal reflux disease history ( P < 0.001), and Barrett's esophagus history ( P < 0.001) compared with H‐BMI patients with cT2 tumors ( n = 114). Median survival of N‐BMI patients was 36.66 months compared with 53.20 months for H‐BMI patients ( P = 0.005). In multivariate analysis, older age ( P < 0.001), squamous histology ( P = 0.002), smoking ( P = 0.040), weight loss ( P = 0.002), high baseline stage ( P < 0.001), high number of ypN+ ( P = 0.005), high surgical stage ( P < 0.001), and American Society of Anesthesia scores, three out of four ( P < 0.001) were independent prognosticators for poor overall survival. We were able to perform propensity‐based analysis of surgical complications between H‐BMI and N‐BMI patients. A comparison of fully matched 376 patients (188 with H‐BMI and 188 with N‐BMI) found no significant differences in the rate of complications between the two groups. This larger data set confirms that a fraction of H‐BMI patients with antecedent history is diagnosed with early baseline EC. Upon validation of our data in an independent cohort, refinements in surveillance of symptomatic H‐BMI patients are warranted and could be implemented. Our data also suggest that H‐BMI patients do not experience higher rate of surgical complications compared with N‐BMI patients.