To Supplement or Not

Abstract

Protein energy wasting (PEW) is a common feature in maintenance dialysis patients (1), and its prevalence has not been improved in the past 15 years (2). Despite being a life-saving procedure, hemodialysis unfortunately has many adverse effects. Among them, it is a catabolic event in response to inflammatory cytokines, loss of nutrients into the dialysate, anorexia, and multiple septic complications, which over time induce a progressive wasting and eventually cachexia. Any attempt to improve patients’ spontaneous nutritional intake is of interest because it overcomes other catabolic conditions and reverses the wasting process (1). The optimal route of administration of supplemental nutrients has been debated for years. Although intradialytic parenteral nutrition (IDPN) support seems the gold standard and easiest answer, there are caveats that prevent full efficiency such as cost, a limited provision (<1000 kcal three times a week; i.e., 400–450 kcal/d), and nausea that reduces subsequent spontaneous intake. In addition, IDPN in some way gratifies the dialysis staff, who may then pay less attention to the decline in nutritional status of patients over time. Oral nutritional supplementation (ONS) is safe, cheaper, and can be taken during the dialysis treatment as well as at home on nondialysis days. The metabolic, nutritional efficacy, and safety of ONS are now well established (1,3–5). In a recent study, Lacson et al. supplied maintenance hemodialysis patients who had a serum albumin ≤3.5 g/dl with intradialytic oral supplements until they reached a serum albumin of 4.0 g/dl (6). They showed that survival of the supplemented patients was significantly improved by 9% (intention-to-treat analysis) and up to 34% (as treated) compared with those not receiving supplements (6). Despite the observational nature of their study, the survival benefit reported by Lacson et al. makes an important contribution by showing that intradialytic ONS may affect clinically relevant hard outcomes such as survival. ONS can be offered to patients at many occasions during the week: during the dialysis sessions three times per week, and at home generally by the way of two supplements per day. Dividing this amount over four to five doses may increase the daily tolerance. Information should be given to the patient to avoid replacing meals by these supplements and to take them 1–2 hours after meals and at bedtime to avoid decreasing appetite. Because renal supplements contain less potassium and less phosphorus, they may be preferred in cases of uncontrolled metabolic disorders. Concentrated products (2 calories per milliliter) may avoid fluid overload because they could deliver up to 500 kcal in only 250 ml per day. In this issue of CJASN, Cheu and colleagues retrospectively analyzed the Fresenius Medical Care (FMC) database from 2006 to 2008 (7). They extracted data on 470 patients who were at risk of developing PEW and analyzed their outcome whether they received ONS afterward. To minimize potential biases, the adjusted hospitalization and mortality percentages were calculated for analyses based on ONS use and analyses based on ONS indication. The intervention group was reduced to 276 patients who effectively received ONS and was also compared with a control group from the Centers for Medicare and Medicaid Services Clinical Performance Measures Project during the same time period. The authors were able to show that ONS was associated with improved hospitalization risk in both the analyses based on ONS use and the analyses based on ONS indication. Cheu et al. interpreted the results as partly favorable, and were possibly disappointed by a rather low serum albumin response over time. Indeed, after 6 months, the albumin increment was no longer different compared with the FMC control group who did not take oral supplements. This article addresses a number of interesting points. First, the diagnosis of risk of starting a wasting process was made using the serum albumin value of 3.8 g/dl as recommended by the PEW nomenclature conference (8). This value is higher than other general guidelines, but fits with the higher mortality risk observed in maintenance dialysis patients (9). It probably identifies patients who are not yet malnourished but on the verge of becoming so. This early wasting phase is likely the best moment to identify and correct insufficient intakes too frequently observed in these patients. A recent randomized study using this threshold value indeed showed that an oral supplement given for 3 months was able to restore patients’ intake back to the recommended values and improved serum albumin in patients who effectively increased their intake (5). It should be noted, however, that the 3.8 g/dl albumin threshold refers to the bromocresol green measurement technique (8) and may differ if albumin is measured differently. This information is lacking in the article by Cheu et al. The fact that serum albumin increase was not as important as expected has also been observed in other studies. This could be due to an underpowered study, a slower metabolic response of albumin compared with prealbumin, a decrease in patients’ spontaneous nutritional intake, or a reduced compliance with ONS. For the sake of study power, which was discussed in the article by Cheu et al., no sample calculation could reliably be done when patients’ files were analyzed retrospectively. One analyzes what one gets. However, mortality is reduced in patients receiving ONS, as shown by Cheu et al. in their Figure 3. The trend (P=0.09) suggests that with an increase in sample size, the study may have become significant. Thus, further trials should use this important information in order to be adequately designed. Although serum albumin remains the most robust survival predictor in maintenance dialysis, it may not be the best nutritional repletion marker. Serum prealbumin may indeed provide more information (10), particularly when serum albumin is close to the normal range (11). Results from the French Intradialytic Nutrition Evaluation (FINE) study showed that although serum albumin and prealbumin did rise soon after starting the supplementation intervention and were sustained over 2 years, serum albumin change was not predictive of survival, whereas serum prealbumin change strongly identified those patients who would survive at 2 years of follow-up (1). It would have been helpful if Cheu et al. could have examined this issue if the data were available. Finally, two other points deserve comment because they are related to patients’ food intake. Unfortunately, we do not have information on patients’ spontaneous food intake, and it makes it difficult to separately analyze the effects of the ONS from patients’ spontaneous intake. Indeed, as already mentioned, most patients have difficulty eating enough calories (more than protein, in fact), and many studies report intakes at approximately 20–25 kcal/kg per day, which is insufficient (12,13) even if they have a low energy expenditure (14). If ONS is taken any time during the day and progressively replaces meals, this will be of no benefit to the patient, and education should be provided to counteract this fact. Second, because Cheu et al. did not provide compliance records, it is possible that patients abandoned part or whole supplementation over time, which may explain the lack of improvement after 6 months of starting ONS. In the FINE study, food records and compliance logs were obtained regularly and the mean amount of ONS taken during 1 year was 1.3 U/d for two that were initially prescribed (1). Supplements by definition only fulfill partial needs (generally no more than 5 kcal and 0.4 g of protein per kg per day), and this is the most important point to remember. For patients with a spontaneous intake <0.8 g of protein/kg per day and calories <20 kcal/kg per day, neither ONS nor IDPN will bring sufficient nutrients to improve nutritional status. Thus, more aggressive intervention is indicated such as tube feeding or total parenteral nutrition for a minimum of 3–6 months. In our experience, tube feeding works very well. Although this study has some limitations (e.g., nonrandomization of the patients to the intervention, lack of information on the dietary intake and compliance of the patients), Cheu et al. should be commended for highlighting the potential effect of ONS in an integrated disease management strategy. In no way does their study contradict that of Lacson et al. (6) and it adds to the increasing number of studies confirming a beneficial effect of ONS in patients undergoing maintenance hemodialysis. Large-scale controlled prospective trials are needed to evaluate the effectiveness of oral supplements on clinically relevant outcomes. Disclosures D.F. has consulted for Abbott, Danone, and Fresenius-Kabi.