Intradialytic parenteral nutrition: where do we go from here?

SINCE THE BEGINNING of the eighties, malnutrition seemed to be a determinant of survival in maintenance hemodialysis.1, 2 In a recent European study, 24% of the 7,123 patients presented with a body mass index lower than 20, 62% a lean body mass lower than 90% of expected value, 20% with a serum albumin level under 35 g/L, and 36% with a serum transthyretin (prealbumin) less than 300 mg/L.3 In malnourished hemodialysis patients, it is estimated that the yearly mortality rate is close to 30%.4, 5, 6 Several methods of nutritional supplementation were developed to compensate for insufficient food intake and to reach the recommended targets of 35 kcal and 1.2 g protein/kg/day.7, 8

Intradialytic parenteral nutrition (IDPN), an easy-to-perform and usually well-tolerated treatment, is widely used in hemodialysis units: in the early nineties, 7.5% of patients received IDPN in large North American series.5 However, IDPN development occurred in the absence of clear demonstration of its usefulness. IDPN is a cyclic parenteral nutrition given 3 times weekly through the venous line of the dialysis circuit. Energy supply is usually given as a glucose-lipid mixture. Because no clinical benefit of specific amino acid solutions has been shown so far, standard amino acid solutions are most often used. IDPN can bring up to 800 to 1,200 kcal and 30 to 60 g of protein per dialysis session, ie, approximately, in a 60-kg patient, 5 to 8 kcal and 0.2 to 0.4 g protein/kg/day. Thus, it must be underlined that IDPN allows reaching optimal MHD requirements only if the patient’s spontaneous feeding is equal to or above 25 kcal and 0.8 g protein/kg/day. Therefore, when spontaneous food intake is below these levels, a greater daily nutritional support, most often through an enteral route, is required to reach the recommended energy and protein intakes.

IDPN must be assessed in terms of metabolic, nutritional, and outcome benefits. Protein and energy homeostasis was investigated during hemodialysis in 7 patients with and without IDPN.9 Patients were studied 2 hours before, during, and 2 hours after a hemodialysis session, using a primed constant infusion of L-(1-13C) leucine and L-(ring-2H5) phenylalanine. IPDN resulted in a positive energy balance concomitant with a change from a catabolic state to a positive protein balance, both on the whole-body level and on the muscle level, eg, in the forearm. Altogether, 30 studies addressed the nutritional effects of IDPN. Most of these reports come from cohort studies. Five nonrandomized trials compared patients treated using IDPN with control patients.4, 5, 10, 11, 12 Although these studies were heterogeneous with regard to the nutritional supply, number of patients, length of treatment, and nutritional assessment, it can be noted that all of these reports showed an improvement of the measured nutritional outcomes. Only 3 randomized controlled trials have been published.13, 14, 15 Two of them only lasted for a few days and tested the feasibility of IDPN. One randomized controlled trial addressed the nutritional effects of a 3-month IDPN regimen given in 12 patients compared with 14 controls and showed an improvement of body weight, arm muscle circumference, triceps skin fold, plasma albumin and prealbumin, interdialytic creatinine production, and skin tests of delayed hypersensitivity.15

Together, these data argue for metabolic and nutritional efficacy of IDPN. However, a key point for clarifying the indications of IDPN is to evaluate its effects on morbidity and mortality and to better characterize the patients who could best benefit from this treatment. Three retrospective studies suggested that IDPN improved survival in severely malnourished hemodialysis patients: in a series of 72 depleted patients treated with a 6-month IDPN, Foulks et al reported that, when nutritional status was improved, the 9-month survival was increased and the hospitalization rate reduced 16; in an 81-patient series including 50 patients treated using a 9-month IDPN and 31 controls, Capelli et al observed that the 2-year survival was correlated to IDPN administration 4; Chertow et al analyzed the outcome of 24,196 patients treated in the National Medical Care hemodialysis centers and observed that among the 1,679 patients receiving IDPN, patients with serum albumin ≤34 g/L showed an improved 1-year survival.5

Many investigators underline the need for controlled studies addressing the effect of IDPN on patient outcome. Such studies require large multicenter series of patients and long-term treatment and follow-up. Moreover, the multiplicity of causes of hospitalization and death, not directly linked to malnutrition, makes it difficult to show the proper effect of refeeding. The goal of the current French Intradialytic Nutrition Evaluation study is to evaluate, in a prospective randomized controlled design, the effects of IDPN on mortality, hospitalization rate, nutritional status, dialysis efficacy, and cost of treatment. Thirty-eight hemodialysis centers within the French Study Group of Nutrition in Dialysis participated in this study. Among 3,583 patients screened, 390 were eligible according to the following criteria: adult patients less than 82 years old, hemodialysis treatment duration of more than 6 months, and 2 of the following malnutrition criteria: body mass index less than or equal to 20 kg/m2, dry body weight loss within 6 months greater than or equal to 10%, serum albumin ≤35 g/L, prealbumin ≤300 mg/L. One hundred and eighty-six patients were included (male-female ratio = 0.9, 68.3 ± 10.4 years old; diabetes, 24%; hemodialysis vintage, 98 ± 99 months) from January 1, 2001, to December 31, 2002. Baseline data showed severe malnutrition: body mass index, 22.7 ± (SD) 4.2; predialysis serum albumin, 31.6 ± 4.0 g/L; prealbumin, 240.5 ± 50.3 mg/L; creatinine, 648 ± 177 μmol/L; normalized protein catabolic rate, 1.09 ± 0.35 g/kg/day; weekly hemodialysis time, 13.0 ± 2.4 h; Kt/V, 1.69 ± 0.35; Karnofsky score, 66 ± 16. Patients were randomized into 2 groups: IDPN group, receiving IDPN thrice weekly, 13.8 kcal (glucose-lipid: 50:50) and 0.62 g amino acids/kg per dialysis session, ie, 5.9 kcal/kg/day and 0.27 g amino acid/kg/d during 1 year, and the control group, without IDPN. For ethical reasons, both groups were given oral supplementation during the same period. After a mean follow-up of 15 months, overall hospitalization and mortality rates were 4 days/mo and 24.4%, respectively. These preliminary data showed adequate randomization and power to answer the question of the efficacy of IDPN on patient outcome. Furthermore, this study may help to identify criteria for selecting patients in whom outcome improvement will be observed. Final results, using intention-to-treat analysis, will be available in early 2005.