T. K. Jenkins4, A. N. Lopez4, G. A. Sarosi1,2, K. Ben-David3, R. M. Thomas1,2 1University Of Florida,Department Of Surgery,Gainesville, FL, USA 2North Florida/South Georgia Veterans Health System,Department Of Surgery,Gainesville, FL, USA 3Mount Sinai Medical Center,Department Of Surgery,Miami Beach, FL, USA 4University Of Florida,College Of Medicine,Gainesville, FL, USA
Introduction: While prior research has shown that preoperative (preop) enteral access is feasible and safe in patients to support their nutrition prior to esophagectomy, controversy exists regarding its necessity, as subjective dysphagia is a poor indicator of need for enteral access. We hypothesized that patients who underwent preop enteral access prior to esophagectomy for cancer fared no better than those who had surgical enteral access performed at the time of esophagectomy.
Methods: An IRB approved retrospective database of patients undergoing esophagectomy for esophageal malignancy from 2007-2014 was established. Clinicopathologic factors were recorded including preop enteral access, weight change, nutritional labs, preop cancer stage, operative details, and perioperative complications.
Results: One hundred fifty-six patients were identified, of which 99 (63.5%) received preop chemoradiation (cXRT) prior to esophagectomy. Since preop cXRT can influence perioperative nutrition, this group comprised the study cohort. Fifty (50.5%) underwent preop enteral access [esophageal stent (1), gastrostomy (14), jejunostomy (32), nasoenteric (1), combination (2); “access group”] prior to cXRT followed by esophagectomy and feeding jejunostomy unless it was pre-existing. There was no difference in demographics, preop tumor staging, or operative details between the access and non-access groups. No difference was noted between access and non-access groups in subjective dysphagia [n=43 (86%) vs 37 (75.5%), respectively; p=0.2)] or mean preop serum albumin (gm/dl) [3.9 (range 3.1-4.5) vs 4 (range 3.3-6.4), respectively; p=0.2]. To account for potential cXRT delays, there was no difference in median time from diagnosis to surgery in the access vs non-access groups (126d vs 126d, p=0.5). Comparing weight loss 6mo preop to surgery, the access group had a mean 5.2% weight loss (range -29.4 – +6.6%) vs 4.5% reduction (range -19.4% – +68.2%) in the non-access group (p=0.8). Additionally, mean weight loss 6mo preop to 6mo postop was similar in the access vs non-access groups [-11.2% (range -44% – +5.3%) vs -15.4% (range -34.1% – -1.4%), respectively p=0.1]. Complication rates between access and non-access groups (64% vs 51%, respectively; p=0.2) were likewise similar. In patients with reported dysphagia, there was no difference in weight change 6mo preop to 6mo postop in the access vs non-access group (-11% vs -15.2%, p=0.1; respectively).
Conclusions: Despite the bias of establishing enteral access prior to preop cXRT for esophageal malignancy in candidates for esophagectomy, there was no difference in weight change, preop albumin, or complication rates in patients who had preop enteral access versus those who did not. Patients with esophageal malignancy should therefore proceed directly to appropriate neoadjuvant and surgical therapy with enteral access performed at the time of definitive resection or reserved for those with obstruction confirmed on endoscopy.