Role of nutrition support during induction chemoradiation therapy in esophageal cancer.

Abstract

Preoperative chemoradiation therapy (CRT) potentially benefits a subgroup of patients with esophageal cancer. The ability to administer aggressive CRT may depend on the initial nutritional status and the ability to sustain nutrition during therapy. Parenteral nutrition support during CRT may lead to complications that limit its usefulness and negate any potential benefit. Data were analyzed to evaluate the role of parenteral nutrition support (PNS) in patients receiving CRT. Forty-five consecutive patients with locoregional esophageal cancer, enrolled in a phase I/II trial of induction CRT, were analyzed. On the basis of the nutrition support received, two groups were defined as follows: group I (with PNS, n = 30) and group II (without PNS, n = 15). Results were compared in terms of chemotherapy (CT) dose tolerated, morbidity of CRT, response rates, and surgical outcome in groups with and without PNS. The two groups were comparable for demographic data, stage and site of disease, and performance status. There was no significant difference between the groups in the nutritional parameters (weight and serum albumin) before and after CRT. Group I patients received significantly more (% of total calculated dose) CT compared with group II (5-fluorouracil [5-FU], 86.4% vs 68.8%, p = .02; cisplatin [CDDP], 90.8% vs 78.2%, p = .05; and interferon alpha-2b [IFN-alpha], 95.4% vs 79.8%, p = .05, in groups I and II, respectively). Major (grade III/IV) adverse effects of CT were hematologic (group I, 93.3% vs group II, 86.6%, p = .59) and gastrointestinal (group I, 56.67% vs group II, 33.3%, p = .2). Postsurgical staging revealed complete response in 10 (22%) and a major response in 23 (51%) patients, although the response rates were similar in the two groups (group I, 76.6% vs group II, 66.6%, p = .8). Surgical morbidity (51.8% vs 61.5%, p = .73), mortality (7.4% vs 7.6%, p = 1.00), and hospital stay (22.5 vs 19.6 days, p = .63) were also similar in the two groups. PNS can be provided to these patients without an increased risk of CRT or resection-related morbidity. Although early and prolonged PNS facilitates administration of complete CRT doses, no benefit is derived from the administration of more CRT in the present regimen. The utility of PNS in this setting is unclear and, until further clarified, should not be applied routinely to this cohort of patients.