Plasma sodium

The sodium concentration of commercially available dialysis solutions varies between 132 mmol/L and 134 mmol/L. Because of the shallow concentration gradient between dialysate and blood, diffusion of sodium is usually lower than its convective transport. Only in situations where there is a marked difference between plasma and dialysate-sodium concentrations does the diffusion become clinically important.

Researchers have tested methods for removal of sodium in PD. They included:

		decreasing the dialysate-sodium concentration;
		increasing the dialysate volume;
		increasing the number of exchanges;
		increasing the use of hypertonic solutions; and
		using colloid osmotic agents.

When used for dwells of at least eight hours, a 7.5% icodextrin-based dialysis solution can be expected to increase sodium removal more than will a 3.86% glucose solution. This assumption is based on the observation that icodextrin does not induce Na+ sieving. Furthermore, net ultrafiltration is higher with 7.5% icodextrin when the dwell time exceeds eight hours, leading to a higher convective sodium loss. However, comparisons between the two solutions with regard to net sodium removal have not been published. Furthermore, it can be concluded that the sodium concentration in dialysis solutions used for APD should be lower than in those employed for CAPD. This case is especially true for anuric patients and for patients using large quantities of high glucose solutions. Although the number of studies on this subject is rather small, a 3.86% glucose-based solution with a sodium concentration between 125 mmol/L and 130 mmol/L is likely to be beneficial for most APD patients without residual renal function.

This has been shown both retrospectively and within the DIANA study, where serum sodium levels decreased from 138.1 ± 0.7 mmol/L to an average concentration in the follow-up visits of 135.9 ± 0.8 mmol/L (p < 0.05). At 12 months, the serum sodium concentration increased to a non-significant difference from baseline. Serum osmolality increased, but did not differ significantly from glucose users at any visit.

BP control and fluid balance

Cardiovascular disease is the leading cause of mortality and morbidity in end-stage renal disease (ESRD) patients undergoing dialysis, and this is most likely related to long-standing hypertension along with abnormal lipid profiles, ventricular hypertrophy and dysfunction as well as a high level of sympathetic activation. Initiation of peritoneal dialysis (PD) often results in excellent blood pressure (BP) control, possibly better than with haemodialysis (HD). However the improvement is rarely sustained and many patients become progressively hypertensive with time.

Better fluid volume control and greater clearance of vasoconstrictor factors compared with HD have been considered responsible for the initial improvement whereas late decline of peritoneal ultrafiltration and fluid retention have been implicated as potential causes of long-term deterioration in BP control.

High peritoneal membrane transport characteristics are an adverse prognostic factor in CAPD, and one possible explanation is that poor ultrafiltration in these patients leads to chronic fluid overload. It is suggested that fluid retention causes dilutional hypoalbuminaemia in CAPD, and thus the link between serum albumin concentrations and survival in CAPD may be due partly to complications of fluid overload.

Use of icodextrin for the daytime dwell in APD results in improved fluid balance and blood pressure control compared with 2.27% glucose. Multifrequency bioimpedance detected clinically important changes in fluid content and a significant short-term improvement in control of hypertension in patients on APD beginning use of icodextrin for the daytime dwell.

Glucose stability/near-patient glucose testing

Poor glycaemic control is a risk factor for the development of end-stage renal disease, and poor glycaemic control can impact on survival on dialysis. In addition, renal disease makes glycaemic control more problematic with a higher incidence of hypoglycaemia. The associated co-morbidity so often seen in diabetic ESRD patients, such as retinopathy, autonomic neuropathy and macrovascular disease, makes avoidance of hypoglycaemia important.

Patients with diabetes on dialysis face particular problems because of the unpredictable effect that renal failure and dialysis per se have on insulin and glucose metabolism. There are several factors and mechanisms specific to ESRD that alter the relationship between insulin requirements and blood glucose concentration, and the magnitude of their effect is proportional to the degree of loss of renal function.

Most fundamentally there is reduced insulin catabolism: 30 % of insulin catabolism takes place in the kidney so that a reduction of functional renal mass will prolong insulin half-life and reduce insulin requirements. In addition, in renal failure reduced calorie intake and weight loss may be important elements tending to reduce insulin requirements in insulin-treated patients with diabetes. Conversely, there is often a reduction in physical activity and increasing insulin resistance is often present.

Furthermore, PD is associated with significant glucose absorption and this leads to a positive energy balance of approximately 500 kcal/day. Consequently, in patients with diabetes this leads to hyperglycaemia and the need for increased insulin requirements to counteract this. Common PD treatment regimens include CAPD taking the form of four or five daily exchanges of glucose-containing fluid, with peritoneal “dwell-times” of four to six hours. This raises insulin requirements throughout the 24 hours of dialysis.

Switching patients to and from HD and PD (as may be needed typically as short HD sojourns to allow recovery from PD-technique failures: herniae, peritonitis and catheter blockage) will have the greatest, and most immediate, impact on insulin requirements Changing from CAPD to APD results in a need for adjustment of the insulin regimen (more at night to limit nocturnal hyperglycaemia, and less by day to limit daytime hypoglycaemia) when the glucose load is concentrated overnight rather than throughout the 24 hours.

Another approach to the loss of ultrafiltration due to premature loss of the glucose-mediated osmotic gradient is simply to use even richer glucose solutions, as very hypertonic solutions (2.27 % w/v and 3.86 % w/v), though these cause yet further serious peritoneal structural and functional changes. Icodextrin is only licensed for once-daily administration because its metabolites accumulate rapidly. In the serum of control patients there is no maltose or isomaltose detectable. Patients with dialysis-dependent renal failure had elevated levels of isomaltose (23.6 +/- 8.3 mg/L) but low levels of maltose (< 3.0 mg/L). Treatment with icodextrin resulted in elevated plasma levels of maltose (range: 500 to 1600 mg/L), while levels of isomaltose declined to 9.8 +/- 5.2 mg/L (P < 0.0001 vs baseline levels). The reasons for these changes have not been elucidated.

The use of icodextrin rather than very high concentration glucose solutions may itself significantly reduce insulin requirements and improve insulin resistance. But in addition, there is a unique problem using icodextrin. The polymer is absorbed and its metabolites can have an effect on reagent sticks glucose measurements, giving rise to erroneously high blood sugar readings.

Many glucose test strips are based on the enzyme glucose oxidase catalysing the conversion of glucose in the presence of oxygen to hydrogen peroxidase and gluconic acid. Hydrogen peroxidase may then be detected by a peroxidase dye indicator whereby the dye is oxidised and changes colour. Hydrogen peroxidase can also be detected electrochemically at a positive charged electrode and the current flow measured is proportional to glucose levels. Reducing sugars, such as galactose (such as seen in galactosaemia) and maltose, are also readily oxidised and hence will cause overestimation of glucose results. Some glucose test strips are based on glucose dehydrogenase catalysing the conversion of glucose to gluconic acid and reduced NAD (nicotinamideadenine dinucleotide), the latter being a measure of glucose concentration.

It has been reported that icodextrin may cause erroneously high glucose results in some glucose test strips. On reviewing these studies the Accutrend Sensor (Boehringer Mannheim), One touch II (Lifescan) and the Glucocard Memory (Menarini Diagnostics) cause positive interference with icodextrin and hence overestimation of glucose readings. The systems which use a glucose dehydrogenase enzyme (Accutrend Sensor, Boehringer Mannheim) appeared to cause the most interference. Glucotouch (Lifescan, Johnson & Johnson) and One Touch profile (Lifescan) showed no interference with icodextrin. When the individual metabolites of icodextrin were looked at it was mainly maltose and maltotriose which caused the positive interference. It seems that overestimation of glucose decreases as the molecular size of the saccharide increases.

We have completed a detailed examination of the effect of icodextrin dialysis on a broad range of glucose test strips (see table on next page)