Abstract
bicarbonate-buffered solutions, to better correct acid-base balance.
icodextrin solutions, to increase ultrafiltration in water overloaded patients.
amino-acid solutions, to improve nutritional status in malnourished patients and to reduce glucose load.
Our paper will discuss this last topic.
Malnutrition in Peritoneal Dialysis Patients
Malnutrition has been reported to be an important indicator of mortality risk in continuous ambulatory peritoneal dialysis (CAPD), and it affects, more or less heavily, 8% - 33% of patients (1). The causes of malnutrition in CAPD patients are the same as in other uremic subjects: that is, hormonal imbalances, impaired metabolism of amino acids, and low protein intake. In addition, CAPD involves unavoidable daily absorption of glucose (100 - 200 g) and losses of amino acids (2 - 4 g) and proteins (5 - 15 g) into the dialysate (2,3). Protein losses may increase, even double, during peritonitis episodes and over the succeeding 2 - 3 weeks.
Glucose absorption brings an extra charge of 400 -800 kcal per day that, in some patients, may lead to obesity and dyslipidemia. Both the absorbed glucose and the increased abdominal pressure (due to the presence of the fluid in the peritoneal cavity) can lessen appetite, further decreasing protein intake in patients who already tend to avoid meat because of uremia.
Abnormalities in Amino-Acid Concentrations
Altered concentrations of free amino acids in plasma and within cells are just one evidence of the complex imbalance of nitrogen metabolism in uremia.
In a normal subject, plasma concentration of total free amino acids averages 2500 – 3000 μmol/L with an essential-to-nonessential ratio of 0.6. Renal clearances are lower than 1 mL/min for most amino acids, but for some—histidine, taurine, glycine—non negligible values of 4 – 7 mL/min are reached, and the renal clearances of “no threshold” amino acids approximate inulin clearance.
Twenty-two amino acids take part in the protein synthesis. In addition, some amino acids are precursors of non protein hormones or neurotransmitters. For instance, tyrosine gives rise to thyroid hormones and catecholamines, tryptophan to serotonin, and glutamine to gamma amino butyric acid (GABA). Other amino acids—ornithine, citrulline, arginine—participate in the urea cycle in the liver. Nitric oxide is synthesized from arginine, and so on. These metabolic pathways may also be disturbed in uremic patients.
In CAPD, peritoneal clearances average 3 – 4 mL/ min for all amino acids, with only minor differences depending on electrical charge (4). Abnormalities in free amino acid concentrations in CAPD are partly uremia-related and partly peculiar to CAPD (Table 1). The most frequently reported alterations in plasma are the reduced levels of many essential amino acids, including valine, and of some nonessential amino acids such as histidine, serine, tyrosine, and taurine. On the other hand, levels of some nonessential amino acids, such as citrulline, are increased, as are levels of the “no threshold” amino acids (5).
Amino Acid Abnormalities in Continuous Ambulatory Peritoneal Dialysis Patients
Concentrations within muscle cells do not reflect plasma concentrations. Lindholm and colleagues (6,7) have repeatedly found increased concentrations of both essential and nonessential amino acids in muscle cells, with the exception only of taurine and perhaps valine. Concentrations in erythrocytes are intermediate between plasma and muscle. Altogether, the intracellular profile is less altered in CAPD than in uremic patients treated with other renal replacement therapies, while plasma alterations are more pronounced. The latter observation could be a consequence of the hyperinsulinemia present in CAPD.
Metabolism of amino acids occurs mainly in liver and in muscle. Branched-chain amino acids (valine, leucine, isoleucine) are primarily metabolized in muscle. Degradation proceeds through a reversible transamination followed by an irreversible oxidative decarboxylation, and may be more or less complete depending on the level of acidosis and on the energetic request of the muscle. Branched-chain amino acids and their keto-analogues have shown anabolic properties in both in vivo and in vitro studies (8,9).
Sulfur–amino-acid pathway proceeds from methionine, an essential amino acid, to taurine, which represents an end-product. In CAPD, levels of methionine and taurine are reduced, while levels of the intermediate products (homocysteine, cysteine, cysteinesulfinic acid) are increased, maybe as a consequence of metabolic stops at these levels. Taurine could have an important role in regulating calcium ion fluxes across the membranes. Low taurine has been related to fatigue and cardiomyopathy, and homocysteine represents a risk factor for cardiovascular diseases (10). Supplementation with vitamins B12, B6, and folic acid, which are cofactors to the enzymes involved, reduces homocysteine concentration, but does not normalize it (10,11).
Serine is normally synthesized in the kidney from glycine, therefore depletion of serine progresses as renal function declines. Serine depletion has been reported in neurological and psychotic syndromes, and, in experimental animals, in vitamin B6 deficiency (11,12). Tyrosine, too, is largely produced in kidney from phenylalanine. The reaction is catalyzed by phenylalanine-hydroxilase, which is inhibited in uremia (13).
The peritoneum is used as a dialytic membrane in CAPD. Peritoneal administration is also a physiological way to supply amino acids. It is not surprising, then, that researchers have sought to use this form of feeding in malnourished CAPD patients.
Clinical Experience with Amino-Acid Solutions
The use of amino acids in dialysis fluid has followed two paths:
The addition of amino acids to glucose-based solutions
The use of amino acids as the osmotic agent.
In 1968, Gjessing (14) added 25 mL of a 5% amino-acid solution for intravenous infusion to each litre of fluid for intermittent peritoneal dialysis (IPD). The aim was to compensate for non-urea nitrogen losses into the dialysate. Gjessing showed that this simple procedure substantially reduced the fall in serum protein and the loss of amino acids during treatment.
Thereafter, experiments were conducted in several countries on using solutions that contain amino acids and glucose together. However, problems with sterilization and stability limited the wider use of such solutions. More recently, amino acids have been added to glycerol-based solutions (15) to obtain the advantage of lowering the concentrations of both substances without losing ultrafiltration capacity.
Oreopoulos and co-workers (16) were the first to use amino acids instead of glucose as an osmotic agent in patients on CAPD. They found that the amino-acid solution was well tolerated and effective. This solution had more clinical and commercial success. In the years since, several groups have evaluated the nutritional benefits of replacing glucose with amino acids in various formulations and schedules (Table 2).
Clinical Studies with Amino-Acid Solutions in Adult Continuous Ambulatory Peritoneal Dialysis Patients
Result was considered “improved” when total protein, albumin or transferrin, nitrogen balance, or total body nitrogen was statistically increased. Variation in plasma concentrations of amino acids was not, by itself, considered improvement of nutritional status.
Effects on Dialysis
The capacity of amino-acid solutions to remove waste products and water was investigated, after preliminary experiments in animals, in several single-cycle studies in humans by Oreopoulos (16) and his group (17,18) at the University of Toronto, and by other researchers afterward (19-33, Table 2).
Those studies showed that amino acids
have a suitable ultrafiltration capacity;
are efficient in removing urea, creatinine, and potassium, the 1.1% solution being as effective as a 1.36% glucose solution with respect to these parameters;
are absorbed from the dialysate by 70% - 90% on average, with a twofold to threefold increase in plasma levels, peaking at 1 hour and returning to near basal values at the end of a 6-hour dwell; and
neither damage peritoneal membrane functions, nor reach dangerous concentrations in plasma.
Effects on Nutrition
Most of the studies listed in involve a small number of patients, last for short periods ranging from some weeks to a few months, use different criteria for enrolling patients, and use different schedules of treatment, so that results are neither statistically validated nor comparable.
Only two studies were randomized. One, from Misra and co-workers (30), examined the effect of the 1.1% solution on a nutritionally unselected group of 18 patients. The researchers observed favorable results only in malnourished patients. The other study was a multicenter study coordinated by Jones (31), in which 54 patients were enrolled to a 3-month amino-acid dialysis with one or two daily exchanges. Nutritional benefit was manifested in several ways and was more pronounced in hypoalbuminemic patients.
From the literature and from our own experience, we are now substantially convinced that amino-acid solutions may carry nutritional benefit, but only when the following conditions are met:
The amino-acid mixture must have a “renal composition.”
A simultaneous caloric load must be delivered.
Acidosis must be fully corrected.
Patients should be selected on the basis of malnutrition arising from low protein intake.
Renal Composition
By “renal composition,” we mean a formulation of amino acids aimed at supplying amino acids lacking in plasma, on the hypothesis that the correction, where possible, of plasma abnormalities could improve protein synthesis. Therefore, the essential amino acids, plus serine, tyrosine, and histidine, will be present at higher concentrations, and nonessential amino acids, such as glycine and alanine, at lower concentrations. In the light of the Lindholm group's (6) observation that concentrations of amino acids in muscle, the main site of protein synthesis of the body, do not correlate to plasma concentrations, this hypothesis may appear quite simple. But renal formulations have, in fact, attained better results. Starting from the composition of Travasol (Baxter Healthcare Corporation, Deerfield, Illinois, U.S.A.), in which alanine and glycine represented more than 50% of the amino-acid content (molar concentration), the proportion of essential amino acids has been progressively increased and that of nonessential amino acids reduced (Table 3).
Evolution in the Composition of Amino-Acid Solutions Used by Authors Over Time
Acidosis
Amino-acid solutions carry some degree of risk of acidosis. Acidosis may worsen renal bone disease and stimulate protein catabolism and oxidation of amino acids, blunting the beneficial effects of the amino-acid supply (34). That is why, in parallel with a more balanced amino-acid composition, the buffer power of the solutions had to be increased over time. The most recent formulation uses up to 40 mEq/L lactate (Table 2). Still, the increase in lactate may be not sufficient, and alkali administration by mouth should be started or increased (33).
Caloric Load
Caloric intake greatly influences nitrogen retention in healthy subjects as well as in uremic ones. Bergström (35) demonstrated that energy intake directly correlates with nitrogen balance in CAPD patients, and Delarue (36), studying the effects of amino-acid dialysis solution on leucine metabolism, showed that ingestion of a caloric meal inhibited protein breakdown and reinforced the positive effect on protein balance. We therefore believe that lack of a caloric supplement was one important reason for the lack of benefit reported by some authors who evaluated amino-acid solutions used overnight.
Side Effects
Increase in plasma urea levels is one expected consequence of an increased nitrogen load and of the oxidization of some of the amino acids supplied (Table 4). From the Delarue study (35), which evaluated the acute effects of a 1.1% solution (Nutrineal: Baxter Healthcare) on leucine metabolism, it appears that only 16% of the absorbed amino acids are oxidized. Another 43% are used for protein synthesis, and the remaining 41% enter the amino-acid pool.
Side Effects of Amino-Acid Solutions
In our study with one amino-acid bag per day (24), urea levels rose by 28% on average, from 140 to 180 mg/dL in the second month. Thereafter, they slowly decreased. In another of our studies (37), which used two bags per day, plasma urea increased by 46%, reaching 240 mg/dL in one patient. This increased level of azotemia may cause loss of appetite, nausea, and vomiting, as some authors have reported in the past, mainly with higher concentrations or multiple exchanges. Patients treated with one 1.1% amino-acid exchange usually tolerate it well. Two exchanges should be delivered only to patients with very low protein intake; in these cases, a proportional increase in dialysis dose should be considered. Assuming that 80% of the administered amino acids are absorbed, each bag will increase protein intake by nearly 18 g, which must be considered in calculating the individual requirement.
Among untoward side effects, most authors mention a decrease in pH, bicarbonate, or total CO—that is, a worsening of acidosis. This effect is not completely blunted even with 40 mEq/L lactate. Increasing the alkali supply by mouth is therefore advisable (33).
With respect to peritoneal function, neither ultrafiltration capacity, nor urea clearance, nor creatinine clearance has been reported to change on dialysis with amino-acid solution. Increased permeability to proteins, possibly induced by increased PGE2 generation, is reported by Steinhauer (38) with more concentrated solutions. Douma and co-workers (39) reported an increased permeability to low molecular weight solutes, but not to proteins, with the 1.1% amino-acid solution (Nutrineal). It seems likely that, if amino-acid solutions have an effect on peritoneal transport kinetics, the effect should not impair amino-acid dialysis over time.
Conclusions
Clearly, amino-acid supplementation cannot improve malnutrition arising from causes other than a low protein intake (for example, under-dialysis, intercurrent illness, gastrointestinal troubles, recurrent peritonitis), at least not until those causes are removed or corrected.
The only dialysis solution now, to our knowledge, commercially available (Nutrineal) meets many of the requirements arising from clinical experience. Nevertheless it can be optimized. For instance, Brulez and co-workers (40) observed that absorption of methionine from dialysate induces an increase in the plasma level of homocysteine. They suggested that the methionine content of the fluid be lowered. Better correction of acidosis has been obtained by Jones and co-workers (41) with a reduction in methionine, lysine, and arginine, and the addition of asparagine and glutamine.
New solutions for CAPD, containing amino acids together with other osmotic substances (Table 5) will soon be available for routine use. These solutions will increase the complexity of PD treatment, but will render it more suitable to the needs of individual patients. The following schedules should become advisable:
Amino-Acid-Containing Solutions for Continuous Ambulatory Peritoneal Dialysis (Already Tested in Humans)
glucose/amino acids/glucose/icodextrin
glycerol/amino acids/glycerol/icodextrin in diabetics)
The increase in azotemia that represents a limit to the number of exchanges with amino-acid solution and to the amino-acid dose in general, could be minimized with solutions containing amino acids and keto-analogues. This combination is already available as a nutritional integrator by mouth. Keto-analogues of branched-chain amino acids have shown an anabolic effect on protein metabolism (8,9).
Bicarbonate will probably resolve the problem of acidosis in CAPD as it did in hemodialysis (HD); but, because bicarbonate precipitates Ca and Mg as carbonate, dialysis bags subdivided into two or three chambers have been devised (42). Such bags will also allow the association of bicarbonate and amino acids.