A multi-center, randomized, controlled trial of parenteral nutrition titrated to resting energy expenditure in children undergoing hematopoietic stem cell transplantation ("PNTREE"): rationale and design

Abstract

Background: Children undergoing hematopoietic stem cell transplantation (HSCT) frequently require prolonged courses of parenteral nutrition (PN) as a consequence of gastrointestinal dysfunction related to preparative chemotherapy and radiation. PN has been associated with shorter engraftment time and decreased mortality during HSCT, however, it is also linked with complications, including infections, liver disease, and metabolic disturbances. Some of these complications may be a result of providing PN in excess of nutrient requirements. We previously described significant reductions in resting energy expenditure (REE), as measured by indirect calorimetry, over the course of HSCT. We also documented a decline in mid-arm muscle area, suggesting depletion of muscle mass, while triceps skinfold, a marker of fat stores, was unchanged. These results suggested the need for further study of energy expenditure, body composition and nutritional intake in this group of high risk patients.

Design and hypothesis: We hypothesize that changes in body composition affect REE during HSCT, and that standard nutritional support may lead to overfeeding. We are performing a randomized controlled trial of parenteral nutrition among children undergoing allogeneic HSCT. Subjects are randomized to receive PN designed to provide 100% of measured REE, or standard PN, i.e., 140% of estimated energy expenditure. The primary outcome variable is change in percent body fat. Secondary outcomes include glycemic control and frequency of infections, changes in REE and body composition.

Conclusion: This study will provide unique and comprehensive nutritional data and its results will guide nutritional therapy for children undergoing HSCT and possibly other catabolic patients.

Fig. 1

Study schema.

Fig. 2

Power of proposed trial to detect difference between two parenteral nutrition strategies with respect to 30-d change in percentage body fat (BF) in children undergoing stem cell transplantation. Using 20 subjects per arm (leftmost curve) ensures 80% power to detect differential change of 6% BF between standard nutritional support and matching of energy intake to expenditure (•). Comparison between subgroups (15/arm, center; 10/arm, rightmost) carries less power. Calculation assumes 5% Type I error (2-sided); no adjustment for covariates; variance and correlation parameters as specified in text.