Glycemic control in diabetic patients undergoing PD can be facilitated by the constant infusion of glucose and insulin into the peritoneal cavity. Maintenance of a balanced glucose-insulin administration via the IP route can theoretically result in better glucose utilization, more physiological insulin administration and avoidance of wide fluctuations in the plasma concentration of both substances. Consequently, improvement in glycemic control and nutrition, reduction of hyperinsulinemia and elimination of multiple daily injections would be expected. The effects on lipid metabolism and their ultimate influence on clinical outcomes are more difficult to assess.
Insulin injected into the peritoneal cavity is absorbed into the portal circulation, mimicking endogenously secreted insulin. The portal circulation transports insulin to the liver where 50% is bound to receptors in a single pass and the remainder reaches the systemic circulation. The intra-hepatic concentration is regulated by the glucose and amino acid concentrations in blood. Insulin inhibits hepatic glycogenolysis, gluconeogenesis and ketogenesis and enhances glycogen and fatty acid synthesis. Insulin clearance during the first pass through the liver regulates its plasma concentration and reduces hyperinsulinemia. IP insulin use is associated with lower basal insulin levels and faster insulin release in response to acute glucose loads. Insulin absorption is faster when administered into an empty peritoneal cavity than when diluted with PD solution. Insulin kinetics have demonstrated better and more predictable absorption of IP insulin whereas SQ administration is affected by tissue degradation of insulin and regional variations in absorption due to fluctuations in tissue perfusion or sequestration.
Hyperinsulinemia has been associated with a high risk of atherosclerosis. Therefore, reducing the circulating levels of insulin should reduce atherogenic risk. However, the literature offers conflicting results. The effects of IP insulin on serum lipids have been reported to be beneficial by some and detrimental by others. The confusing interpretation of the available data may be related to methodological differences among the studies, poor understanding of the ultimate effects of specific lipid profiles on clinical outcome, lack of adjustment for co-morbid conditions and incomplete database.
Advantages and disadvantages of IP insulin administration
The table below summarizes the potential advantages and disadvantages of IP insulin administration in PD patients. Aside from the improved glycemic control, a more physiological route of absorption through the portal circulation, reduced hyperinsulinemia and avoidance of insulin injections, some interesting observations have been made related to other metabolic processes. Higher levels of plasma hydroxy-vitamin D have been observed with IP insulin as compared to SC administration in patients with comparable glycemic control.
|Better glycemic control||Higher total insulin doses||Peritonitis rates|
|More physiologic absorbtion||Higher cost||Protein losses|
|Avoidance of injections||Subcapsular steatosis and focal necrosis||Lipid profiles (?)|
|Higher vitamin D levels||Malignant omentum syndrome|
|Less hyperinsulinemia (in|
|the absence of high|
There are few disadvantages to the use of IP insulin in PD. The need for higher total insulin doses adds to the cost of therapy. Hepatic subcapsular steatosis (HSS) and focal necrosis have also been reported as necropsy findings. The clinical implications of these observations remain unclear1,2. Torun et al. evaluated the association of HSS with clinical parameters in non-diabetic CAPD patients and diabetic patients receiving IP or SQ insulin in a cross-sectional study2. HSS was only detected in five of the eight diabetic patients receiving IP insulin. None of the diabetics receiving SQ insulin and none of the non-diabetic patients developed HSS. Comparison of diabetic patients receiving IP insulin who had HSS with those without HSS showed higher daily insulin dosage, BMI, serum triglycerides and D/Pcreatinine at 4 hours, and a lower D0/D4. No differences in dialysate-glucose load was found between diabetic patients receiving IP insulin who had HSS and those who did not, but mean daily insulin dosage among those receiving IP insulin who had HSS was approximately twice as high as those without HSS. We may infer from these data that IP insulin plays a more important role in the pathogenesis of HSS than glucose levels in diabetic patients and that HSS is associated with higher daily insulin requirements, obesity, hypertriglyceridemia and high peritoneal transport in diabetics receiving IP insulin. However, this was a cross-sectional study with observations made at a single moment in the patient’s lifetime, allowing the possibility that HSS was due to increased peritoneal transport rather than the dose or route of insulin administration. High peritoneal transport rates increase insulin absorption and glucose absorption, which in turn may lead to obesity, hyperinsulinemia and hypertriglyceridemia. Thus, we are left with the important question: Is IP insulin responsible for the development of HSS and increased peritoneal transport, or are patients with increased peritoneal transport more likely to develop HSS? Until we have an answer to this question, it seems reasonable to avoid the use of IP insulin in patients with high peritoneal transport and also to monitor all patients undergoing IP insulin therapy with periodic PETs3. If there is evidence of increased peritoneal transport, as reflected in an increase in D/Pcretinine or a decrease in D0/DGlucose, strong consideration should be given to discontinue IP insulin.
A rare instance of malignant omentum syndrome has also been reported, whereby insulin is trapped in the omentum in response to foreign protein. Based on the available literature, the incidence of peritonitis in diabetic patients is not significantly different than in non-diabetics and the use of IP or SQ insulin does not seem to influence peritonitis rates. Protein losses are not affected by the addition of insulin to the dialysate. As previously discussed, the effect of IP insulin on lipid profiles and their clinical consequences remain controversial.
Determinants of insulin dose
The total dose of IP insulin required for glycemic control is significantly higher than that which the patient was administering SQ prior to initiation of PD. The reasons for this increment in insulin dose are several. Insulin doses increase by approximately 15% after initiation of PD, probably due to the increased glucose load resulting from absorption of glucose from PD solutions (90 to 150 gm/day). The total caloric load varies with the patient's specific transport rate, ultrafiltration requirements and the use of hypertonic exchanges. In addition, the total IP insulin dose increases by 100 to 200% over the total previous SC dose due to hepatic binding, adsorption to the solution bags and tubing and unabsorbed insulin discarded in the peritoneal effluent. It is estimated that 50 to 60% of the insulin is discarded in the peritoneal effluent unused. Another 15% is bound to the plastic bag and administration tubing.
The PD prescription also influences insulin dose. The uniform time intervals of CAPD throughout the day provide the opportunity to adjust the insulin dose according to the caloric load and glucose monitoring. However, simple and reliable methods for IP insulin administration in CCPD or in NIPD have also been used with remarkably good results.
Regimens for IP insulin administration: Practical considerations
- The goal is to maintain the fasting blood glucose levels near 100 mg/dl (5.5 mmol/l), with postprandial sugars of < 200 mg/dl (10 mmol/l) and HbA1C as close to normal as possible
- Blood sugar monitoring at HS, morning, mid-day and as needed may be necessary for tight glucose control
- Additional monitoring is usually required with episodes of peritonitis
- Use only regular insulin IP. Do not use NPH insulin because it precipitates and is poorly absorbed across the peritoneal membrane
- Longer needles (1.5 inches or 3.8 cm) are preferred to assure that the full dose of insulin is injected into the solution rather than being trapped in the injection port
- Mix insulin with the dialysis solution before infusion. The diluted insulin solution will result in slow and continuous diffusion. Ensure proper mixing by inverting the bag several times
- The insulin dose should never be adjusted without consultation with the physician
Guidelines for IP insulin administration with various PD modalities
- Add the previous total daily SQ insulin dose, include all NPH and regular doses
- Multiply by two for the total initial IP dose using regular insulin only. Subsequent adjustments will probably result in three times the total previous SQ dose.
- For CAPD: divide 85% of the total dose into the diurnal exchanges and the remaining 15% in the nocturnal exchange(s). Perform exchanges 0.5 hr prior to meals.
- For CCPD: place 50% of the total dose in the nocturnal cycles (equally divided among the bags) and 50% in the diurnal exchange.
- For CCPD with additional daytime exchanges (PD Plus): divide the insulin equally among all exchanges.
- If a 2.5% solution is used, increase the base insulin by approximately 10%. If a 4.25% solution is used, increase by approximately 20%.
- Adjustments to the amount of insulin added to the dialysate bags should be made according to glycemic control. CAPD patients should check their blood glucose prior to each exchange to determine the dose of IP insulin to be administered. APD patients require diligent monitoring to achieve blood glucose control.
- A sliding scale can be used based on blood glucose monitoring, e.g:
2 U for blood glucose > 20
4 U for blood glucose > 400
6 U for blood glucose > 600
2 U for blood glucose < 100
Peritonitis, other infectious processes and hypercatabolic states
During peritonitis and other infectious processes glycemic control may deteriorate due to increases in peritoneal transport rates resulting in rapid absorption of glucose from the peritoneum. Consequently there is reduced ultrafiltration requiring hypertonic solution and perpetuation of hyperglycemia. Other hypercatabolic states may require higher caloric and protein administration and in extreme cases, the use of total parenteral nutrition, dictating the need for more ultrafiltration and higher insulin doses. Regular and frequent monitoring of blood glucose; frequent and effective adjustment of IP insulin doses, supplemented with SQ or intravenous (IV) doses; and early diagnosis and treatment of the intercurrent illnesses are essential. If glycemia becomes hard to control, temporary discontinuation of PD, transfer to hemodialysis and SQ or IV sliding scale insulin are recommended.
During hospitalizations, well controlled diabetic patients are also prone to experience wide glycemic fluctuations, ranging from hyperosmolar states to frank hypoglycemia. These are most commonly precipitated by changes in dietary intake, changes in PD prescription (transfer from CAPD to APD or vice versa), inactivity, administration of drugs that affect glucose metabolism or inadvertent or inappropriate change in insulin dose. Frequent monitoring of blood sugar cannot be over emphasized.
The route of insulin administration for diabetic patients undergoing PD should be based on patient preference, patient ability to comply with therapy, peritoneal transport status, coexisting co-morbid conditions and cost of therapy.
- Khalili K, et al. Hepatic subcapsular steatosis in response to intraperitoneal insulin delivery: CT findings and prevalence. AJR Am J Roentgenol 180:1601-1604, 2003
- Torurun D, et al. Hepatic subcapsular steatosis as a complication associated with intraperitoneal insulin treatment in diabetic peritoneal dialysis patients. Perit Dial Int 25:596-600, 2005
- Thorp ML, et al. Three diabetic peritoneal dialysis patients receiving intraperitoneal insulin with dosage adjustment based on capillary glucose levels during peritoneal equilibrium tests. Am J Kidney Dis 43:927-929, 2004