Metabolic, Acid-Base, and Electrolye Aspects of Peritoneal Dialysis
Rajnish Mehrotra
While peritoneal dialysis (PD) provides effective control of many of the diverse consequences of uremia, the therapy itself has unique effects on several metabolic parameters that are important for the health of patients with end-stage renal disease.
I. HYPERGLYCEMIA: With PD, ultrafiltration is induced by exerting either crystalloidal osmotic or oncotic pressure across the peritoneal barrier. This is achieved with PD solutions that contain supraphysiologic concentrations of glucose; some prescriptions also include once-daily treatment with either icodextrin or amino-acid-based dialysis solution. Each one of these substances is absorbed systemically during the course of the PD dwell leading to systemic metabolic effects. Treatment with glucose- or icodextrin-based PD solutions results in an obligatory daily absorption of 50-150 g of carbohydrates. This obligatory carbohydrate absorption is higher with greater use of more hypertonic solutions, and in individuals with a faster peritoneal solute transfer rate. Absorbed icodextrin is metabolized, not to glucose but to a variety of oligosaccharides and to the disaccharide maltose (Moberley, 2002)
In some individuals with diabetes mellitus, this obligatory absorption results in poorer glycemic control and requires significant adjustments in therapy. These may include an increase in total daily insulin dose, or initiation of insulin or other glucoselowering therapy for individuals who did not previously need such treatment. Hence, it is imperative to increase the intensity of home glucose monitoring in diabetic patients for the first few weeks after initiating PD, or whenever the prescribed tonicity of glucose-based dialysate is increased. Worse glycemic control is associated with worse outcomes in PD patients, but it is unclear whether this is association or cause and effect (Duong, 2011). There are limited data to determine how much PD increases the incidence of new-onset diabetes, but one Chinese study suggests that about 8% of nondiabetic patients become diabetic (Szeto, 2007). Therefore, blood glucose levels should also be measured every 1-3 months in nondiabetic PD patients.
Just as glucose-based dialysis solutions may worsen glycemic control, glucose-sparing regimens may improve it. These glucose-sparing regimens generally comprise substitution of one glucose-based exchange with icodextrin; the greater ultrafiltration with icodextrin during the long dwell may allow for the use of lower concentrations of glucose for the other dwells (Paniagua 2008). Substituting a second glucose-based exchange with aminoacid dialysate allows for further reduction in systemic glucose absorption. In IMPENDIA, a recently completed randomized trial, the HbA1c of individuals treated with a regimen in which two bags of glucose-based exchange were substituted with one bag each of icodextrin and amino-acid dialysate was 0.6% lower compared with individuals treated entirely with glucose-based dialysate (Li, 2013). Glucose-sparing regimens should be considered in individuals with diabetes treated with PD when there is difficulty in achieving glycemic control.
II. WEIGHT GAIN: The effects of increased body weight in PD are complex. In hemodialysis patients, increased body weight is associated with improved survival, but the evidence is conflicting in PD patients, and there is a concern that obesity may predispose to catheter problems and exit site infection (Johnson, 2012). Patients often gain weight after initiation of dialysis irrespective of modality; this generally reflects gain in fat rather than in lean body mass. This weight gain is, at least in part, a result of increased dietary energy and protein intake following the amelioration of uremic anorexia with initiation of dialysis. In patients treated with PD, some of the weight gain is attributed to the obligatory systemic carbohydrate absorption. However, large head-to-head comparisons do not support the notion that patients treated with PD are more likely to gain significant weight when compared with individuals treated with hemodialysis (Lievense, 2012). Replacing glucose with icodextrin for the long day dwell in automated PD (APD) or for the night dwell in continuous ambulatory PD (CAPD) results in smaller gains in body weight, but this could reflect differences in total body water rather than body fat. Limited evidence suggests that the sites for deposition of excess body fat differ by dialysis modality with greater gain in visceral fat in PD patients; the clinical relevance of this is unclear (Choi, 2011). Despite this uncertainty, it is prudent to limit exposure to more hypertonic glucose dialysis solutions in order to avoid excessive weight gain.
III. PERITONEAL PROTEIN LOSS: During PD, proteins in the blood— primarily albumin —move into the dialysate down their concentration gradient across the peritoneal barrier and are lost as the dialysate is drained. The daily peritoneal protein loss with PD averages 6-8 g and is substantially increased during episodes of peritonitis. As a result of this obligatory daily loss, serum albumin may decrease in patients starting treatment with PD and is often lower than in individuals undergoing hemodialysis.
This daily peritoneal protein loss is generally not modifiable, and its clinical relevance remains unclear. The evidence associating the higher daily peritoneal protein loss with all-cause mortality, cardiovascular events, or protein-energy wasting is, at best, inconsistent (Balafa, 2011). Moreover, the lower serum albumin in patients treated with PD does not seem to put these patients at any higher risk than patients undergoing hemodialysis. All these considerations suggest that PD can be safely continued in patients that are otherwise well but have modest decrements in serum albumin levels with the therapy.
IV. LIPID ABNORMALITIES: