Abstract:
This chapter addresses peritoneal dialysis solutions, peritoneal dialysis adequacy, and prescriptions and residual kidney function.
Keywords
biocompatible, glucose degradation products, glucose polymer solutions, PD adequacy, PD solutions, peritoneal membrane transport, residual kidney function, solutions
Outline
Peritoneal Dialysis Solutions, 480
Constituents of Peritoneal Dialysis Solutions, 480
Dialysate Buffer, 482
Dialysate Calcium, 482
Glucose-Based Solutions, 483
Non–Glucose-Based Peritoneal Dialysis Solutions, 485
Future Development in Peritoneal Dialysis Solutions, 492
Conclusions, 492
Dialysis Adequacy and Prescription, 493
Defining Dialysis Adequacy, 493
Measuring Biochemical Indices of Dialysis Adequacy, 493
Normalization Factor for Urea and Creatinine Clearance, 495
Estimation of Residual Kidney Function, 495
Frequency of Monitoring of Biochemical Indices of Dialysis Adequacy, 495
Peritoneal Equilibration Test, 495
Conclusions, 504
Residual Kidney Function, 504
Importance of Residual Kidney Function in Peritoneal Dialysis, 504
Decline of Residual Kidney Function, 506
Conclusion, 508
The use of a novel portable or wearable equilibrium peritoneal dialysis (PD) technique was first described by Moncrief et al. in 1976 and named as continuous ambulatory peritoneal dialysis (CAPD) 2 years later. In the last 40 years, significant advances have been made in the understanding of peritoneal membrane function, residual kidney function (RKF), and dialysis adequacy and in the development of PD technology, solutions, and prescription. A significant growth in PD use has been reported worldwide, with some countries, such as Hong Kong and Thailand, adopting a “PD first policy.” A more thorough understanding of PD solutions, adequacy, and prescription is therefore essential for successful delivery of PD programs worldwide.
This chapter covers three major issues in relation to PD:
- 1.
PD solutions. This includes conventional standard glucose PD solutions, more biocompatible PD solutions, and solutions using alternative osmotic agents as well as other novel solutions that are under development. Various other constituents of PD solutions are also reviewed and their importance discussed in light of more recent studies.
- 2.
PD adequacy and prescription. This section defines standard indices of dialysis adequacy, peritoneal membrane function and their measurements. Prescription strategies used to increase PD dose are reviewed. Specific discussions will be made on ultrafiltration failure (UFF) and management of volume overload.
- 3.
Residual kidney function. This section reviews the importance of RKF in its contribution to the overall clearance and clinical outcomes in PD patients. Potential strategies in retarding loss or decline of RKF are discussed.
Peritoneal Dialysis Solutions
Constituents of Peritoneal Dialysis Solutions
Conventional commercially available PD solutions contain sodium (132 to 135 mmol/L), calcium (1.25 to 1.75 mmol/L), magnesium (0.5 mmol/L), chloride (95 to 103.5 mmol/L), and lactate (35 to 40 mmol/L) and varying concentrations of glucose/dextrose ranging from 1.36%/1.5% to 2.27%/2.5% and 3.86%/4.25%. This results in an overall osmolality of 344 to 347, 395 to 398, and 483 to 486 mOsmol/L, respectively ( Table 31.1 ). The concentrations of various components of PD solutions were designed to facilitate ultrafiltration and removal of water-soluble uremic toxins while maintaining electrolyte and acid–base balance of PD patients. The various glucose concentrations of PD solutions provide an osmotic gradient of different degrees that allows ultrafiltration to take place across the peritoneum. However, standard glucose-based PD solutions are acidic in pH (5.0 to 5.8) to prevent dextrose caramelization during the sterilization procedure. As a result of their high lactate, high glucose concentration, high osmolality, and high levels of glucose degradation products (GDPs), long-standing use of these standard glucose PD solutions has been associated with progressive peritoneal membrane injury, neovascularization, peritoneal sclerosis, and fibrosis. Furthermore, the low pH, high osmolarity, and glucose content of standard glucose PD solutions inhibit phagocytic functions of peritoneal leukocytes and impair host immune defense mechanisms. Instillation of these solutions may also be associated with inflow pain in some patients.
| Dianeal Standard Glucose PD Solutions | Physioneal | Icodextrin or Extraneal or 7.5% Solution | Amino acid or Nutrineal or 1.1% Solution | StaySafe Standard Glucose PD Solutions | StaySafe Balance | Bica Vera | |
|---|---|---|---|---|---|---|---|
| Sodium (mmol/L) | 132 | 132 | 133 | 132 | 134 | 134 | 134 |
| Calcium (mmol/L) Low Standard Ultralow | 1.25 (PD-2) 1.75 (PD-4) 1.00 (PD-1) | 1.25 1.75 — | — 1.75 — | 1.25 — — | 1.25 1.75 — | 1.25 1.75 — | — 1.75 — |
| Magnesium (mmol/L) | 0.25 | 0.25 | 0.25 | 0.25 | 0.5 | 0.5 | 0.5 |
| Chloride (mmol/L) | 96 | 95 (Physioneal-40) 101 (Physioneal-35) | 96 | 105 | 103.5 | 100.5 for 1.25 mmol Ca 101.5 for 1.75 mmol Ca | 104.5 |
| Lactate (mmol/L) | 40 | 15 (Physioneal-40) 10 (Physioneal-35) | 40 | 40 | 35 | 2 | 0 |
| Bicarbonate (mmol/L) | — | 25 | 0 | 0 | 0 | 35 | 35 |
| Glucose (mmol/L) | 1.36%, 76 2.27%, 126 3.86%, 214 | 1.36%, 75.5 2.27%, 126 3.86%, 214 | 0 | 0 | 1.5%, 83.2 2.3%, 126.1 4.25%, 235.8 | 1.5%, 83.2 2.3%, 126.1 4.25%, 235.8 | 1.5%, 83.25 |
| Icodextrin (g/L) | 0 | 0 | 75 | 0 | 0 | 0 | 0 |
| Amino acids (mmol/L) | 0 | 0 | 0 | 87 a | 0 | 0 | 0 |
| Overall pH | 5.5 | 7.4 | 5.2 | 6.6 | 7 | 7 | 7.4 |
| Approximate osmolality | PD-2 1.36%, 346 mOsm 2.27%, 396 mOsm 3.86%, 485 mOsm PD-4 1.36%, 345 mOsm 2.27%, 395 mOsm 3.86%, 483 mOsm PD-1 1.36%, 344 mOsm 2.27%, 394 mOsm 3.86%, 484 mOsm | 1.36%, 344 mOsm 2.27%, 395 mOsm 3.86%, 483 mOsm | 284 mOsm | 365 mOsm | 1.25 mmol/L Ca solution 1.5%, 356 mOsm 2.3%, 399 mOsm 4.25%, 509 mOsm 1.75 mmol/L Ca solution 1.5%, 358 mOsm 2.3%, 401 mOsm 4.25%, 511 mOsm | 1.25 mmol/L Ca solution 1.5%, 356 mOsm 2.3%, 399 mOsm 4.25%, 509 mOsm 1.75 mmol/L Ca solution 1.5%, 358 mOsm 2.3%, 401 mOsm 4.25%, 511 mOsm | 358 mOsm |
a The 1.1% solutions consist of a combination of amino acids, including histidine, valine, isoleucine, alanine, leucine, arginine, lysine, glycine, methionine, proline, phenylalanine, serine, threonine, tyrosine, and tryptophan.
Dialysate Buffer
The buffer present in standard PD solutions is lactate, with concentrations varying between 35 to 40 mmol (see Table 31.1 ). In patients with normal liver function, lactate is rapidly converted to bicarbonate such that 1 mM lactate absorbed generates 1 mM bicarbonate. The rapid metabolism of lactate to bicarbonate maintains the high dialysate to plasma lactate concentration gradient necessary for continued absorption without accumulation of lactate in the circulation. Absorption may be reduced with hypertonic glucose PD solution, because increased ultrafiltration may dilute the concentration of lactate in the peritoneal cavity, thus decreasing the concentration gradient for diffusion. Lactate in standard PD solution is normally present in a racemic mixture of approximately equimolar concentration of d -lactate and l -lactate isomers. l -Lactate was suggested to be more rapidly absorbed than d -lactate.
Bicarbonate is regarded as the most physiological and biocompatible buffering system. However, calcium and magnesium precipitate in the presence of bicarbonate and with an alkaline pH. Thus magnesium and calcium must be omitted from bicarbonate-containing dialysate. Patients doing PD with these calcium-free and magnesium-free solutions may develop deficits of calcium and magnesium. A dual-chamber dialysate bag in which one chamber contains the bicarbonate buffer of 34 mmol/L and the other contains a solution with calcium and magnesium has therefore been designed. The two solutions are mixed together only before instillation into patients’ abdomen, thus preventing calcium and magnesium carbonate precipitation. The neutral pH bicarbonate solution has been found to be well tolerated and effective in ameliorating metabolic acidosis and is currently used in daily clinical practice in some parts of the world. Treatment of metabolic acidosis is well known to be associated with downregulation of muscle ubiquitin–proteasome complex and inhibition of muscle degradation. More details are discussed in the Biocompatible Peritoneal Dialysis Solutions section. Bicarbonate-containing solutions (e.g., 25 mmol/L) may also be available in a mixture form with lactate (15 mmol/L), because solutions with pure supraphysiological concentration of bicarbonate have been associated with more abdominal pain than mixed bicarbonate and lactate solution.
Dialysate Calcium
Patients treated with PD may be in negative, positive, or neutral calcium balance, depending on various factors, including dialysate calcium concentration, dietary calcium intake, dose of calcium-based phosphorus binders used, and use of vitamin D analogs. Hou et al. found that a high dialysate calcium concentration of 1.75 mM may lead to positive calcium balance and increased plasma calcium, whereas a low dialysate calcium concentration of 0.75 mM may lead to negative calcium balance and decreased plasma calcium. A dialysate calcium concentration of 1.25 mM maintained neutral calcium balance with minimal perturbation in plasma calcium ( Fig. 31.1 ). However, plasma calcium does not reflect body calcium balance because 99% of calcium resides in the bone. The addition of vitamin D analogs and calcium-based binders may increase extracellular calcium and increase the risk for positive calcium balance even when 1.25 mM calcium dialysate is used.
In fact, it is somewhat of a misnomer that current commercially available high-calcium (1.75 mM) PD solution is termed standard calcium solution, whereas 1.25 mM calcium concentration solution is termed low-calcium solution but in fact is the more physiological calcium concentration (see Table 31.1 ). As nicely demonstrated in two calcium balance studies, an elemental calcium intake of 1.5 to 2 g induced positive calcium balance in patients with chronic kidney disease (CKD). There are concerns that high-calcium dialysate may induce low parathyroid hormone and has been associated with a higher risk for cardiovascular death in hemodialysis patients. On the other hand, low-calcium dialysate may stimulate an increase of parathyroid hormone and worsen secondary hyperparathyroidism. Although an earlier study suggested that low-calcium (1.25 mM) dialysate use may increase parathyroid hormone in the short term, addition of calcium-based phosphorus binders and vitamin D analogs appeared to inhibit parathyroid hormone increase and stabilize its level over time. Furthermore, use of low-calcium (1.25 mM) dialysate may be associated with a lower incidence of hypercalcemia in PD patients. A previous small randomized controlled trial (RCT) found no difference in bone histomorphometry for PD patients randomly assigned to receive 1.25 mM or 1.75 mM calcium solution for 1 year. There were some suggestions that PD patients using 1.25 mM calcium dialysate may have less progression in arterial stiffness than those receiving 1.75 mM calcium dialysate.
In a 2016 prospective RCT, 425 hemodialysis patients with intact parathyroid hormone ≤ 300 pg/mL were randomly assigned to receive either 1.25 mM or 1.75 mM calcium dialysate; those receiving 1.75 mM calcium dialysate had significantly more progression in coronary artery calcium score over 24 months than those using 1.25 mM calcium dialysate. Subgroup analysis indicated that progression of coronary artery calcification was found mainly in the subgroup with poor phosphorus control, namely serum phosphorus level ≥ 4.7 mg/dL. Furthermore, the 1.25-mM low-calcium dialysate group had a significant decrease in the prevalence of histologically diagnosed low bone turnover, from 85% to 41.8%, whereas the prevalence of low bone turnover did not change over 24 months in the 1.75-mM high-calcium dialysate group. Although a similar study is not available in PD patients, both the Kidney Disease Improving Global Outcomes 2017 CKD-mineral bone disease (MBD) guideline and the International Society of Peritoneal Dialysis (ISPD) Adult Cardiovascular and Metabolic guidelines suggested a 1.25-mM calcium-containing PD solution be used to avoid positive calcium balance or hypercalcemia.
Glucose-Based Solutions
One study found that initial ultrafiltration rate across the peritoneum is directly proportional to the initial glucose osmotic gradient. Generally, 1.36%/1.5% glucose/dextrose solution generates a 100 to 200 mL ultrafiltration, 2.27%/2.5% glucose/dextrose solution generates a 200 to 400 mL ultrafiltration, and 3.86%/4.25% glucose/dextrose solution generates more than 400 mL ultrafiltration. A 4-hour dwell of 2.5% PD dextrose solution has been used as the standard for peritoneal membrane equilibration test (PET). A net ultrafiltration of more than 200 mL from a standard 4-hour dwell of 2.27%/2.5% glucose/dextrose or more than 400 mL from a standard 4-hour dwell of 3.86%/4.25% glucose/dextrose solution is regarded as sufficient ultrafiltration. Values less than this indicate relative UFF. Symptoms of UFF may not manifest overtly until RKF has declined significantly or is completely lost.
Other than serving as an osmotic agent to facilitate ultrafiltration, glucose in the PD solution provides an important source of energy. The amount of glucose absorbed varies proportional to the concentration of glucose in the dialysate. The amount of glucose absorbed per liter of dialysate (y) can be predicted using the equation of Grodstein et al. as y = II.3 × x – 10.9, r = 0.96, in which x denotes the concentration of glucose from the dialysate. With an average peritoneal solute transport rate (PSTR), it is estimated that almost two-thirds of the PD fluid glucose is absorbed during a 4-hour dwell and more than 85% in an 8-hour dwell. This translates to an obligatory absorption of 43 g and 73 g of glucose with an 8-hour dwell of 2.5% and 4.25% solutions, respectively. Patients receiving four PD exchanges with three exchanges of 1.5% dextrose and one exchange of 2.5% were estimated to have absorbed 88 g glucose in a day. Typically glucose absorbed from PD fluids may amount to 100 to 300 g/day and may account for approximately 12% to 34% of total energy input. Thus it may lead to weight gain in some patients, although this has not been reproduced in other studies. Hypertonic glucose solutions may also add satiety and reduce appetite in PD patients.
Glucose-based PD solutions have various other local and systemic adverse effects ( Fig. 31.2 ).
Local Effects
Glucose exerts direct cytotoxic effects on peritoneal mesothelial cells and causes diabetiform changes in the postcapillary venules, leading to changes in peritoneal membrane structure and function. The standard heat sterilization of glucose-based PD solutions accelerates the generation of GDPs. The formation of GDPs during heat sterilization may be reduced by lowering the pH of the solution. GDPs are locally cytotoxic and glycated local proteins form advanced glycation end-products (AGEs) and contribute to the long-term bioincompatibility of PD solutions. GDP and AGEs exert damaging effects on the peritoneal membrane by causing mesothelial cell loss, inflammation, submesothelial fibrosis, calcification, vasculopathy, and diabetiform neoangiogenesis.
Data from the Peritoneal Biopsy Registry indicated a thickening of the submesothelial compact zone and development of a diabetiform occlusive vasculopathy of small arterioles and postcapillary venules with neovascularization with increasing time on PD. Peritoneal biopsy specimens taken from PD patients with low ultrafiltration capacity demonstrated that AGE accumulation in the peritoneal membrane was positively correlated with the development of severe interstitial fibrosis and microvascular sclerosis. All these negative effects on the peritoneum decrease the effectiveness of the peritoneum as a dialysis membrane, resulting in an increased PSTR with time on dialysis and intensifying the need for using hypertonic glucose solutions. This may eventually lead to peritoneal membrane failure (PMF), increased risk for fluid overload, and protein energy wasting (PEW). Several studies reported that a high PSTR was associated with worse patient survival, and some studies also reported a trend toward worse PD technique survival with high PSTR. A more contemporary analysis based on a nationally representative cohort of PD patients in the United States, of whom 87% were treated with automated peritoneal dialysis (APD), reinforced a similar finding; that is, the dialysate to plasma (D/P) creatinine ratio was associated with survival outcome and linearly associated with mortality. For every 0.1 unit increase in the D/P creatinine ratio, the adjusted mortality risk increased by 7% (95% confidence interval [CI], 1.02 to 1.13), the adjusted hospitalization risk increased by 5% (95% CI, 1.03 to 1.06). On the other hand, ultrafiltration volume was inversely related with hospitalization rate but not with all-cause mortality.
Systemic Effects of Glucose-Based Peritoneal Dialysis Solutions
The cumulative, systemic PD glucose absorption through the peritoneum may aggravate various metabolic disturbances, including insulin resistance, hyperglycemia, accumulation of atherogenic visceral fat, weight gain, and dyslipidemia, and worsen glycemic control in PD patients with diabetes.
Insulin resistance has been found to be associated with a higher cardiovascular mortality and worse clinical outcomes in dialysis patients. New-onset hyperglycemia has been reported in end-stage renal disease (ESRD) patients sometime after commencing PD treatment. Intraperitoneal glucose absorption gives rise to higher plasma glucose levels as well as a more extended period of hyperinsulinemia compared with an equivalent dose of oral glucose. Hyperglycemia not only stimulates hyperinsulinemia but also increases mitochondrial superoxide generation, which activates four key pathways—namely, the polyol pathway, hexosamine pathway, protein kinase C pathway, and AGE pathway—causing hyperglycemic damage. Activation of these pathways induces the expression of endothelin 1, vascular epithelial growth factor, transforming growth factor beta, fibronectin, collagen, and leptin genes and at the same time downregulates endothelial nitric oxide synthase and upregulates transcription factor nuclear factor kappa B. Together with an increased oxidative stress, this results in a heightened inflammatory response, insulin resistance, and peritoneal membrane damage as well as a higher risk for accelerated atherosclerosis.
Derangements in glucose and insulin metabolism are evident in early CKD. With progression of CKD, peritubular insulin uptake increases, which compensates for the decline in the metabolism of filtered insulin. As kidney function further deteriorates, insulin clearance also decreases. In addition, insulin resistance and tissue insensitivity to insulin increase, resulting in suboptimal insulin secretion in response to a glucose load or hyperglycemia. Usually around 60% of the glucose in PD solutions may be reabsorbed during the dwell, thus causing disturbed carbohydrate metabolism even in nondiabetic PD patients, further aggravating insulin resistance. Hyperinsulinemia also contributes to hypertriglyceridemia. Insulin enhances hepatic triglyceride synthesis and indirectly reduces the metabolism of very-low-density lipoprotein cholesterol.
Use of glucose-based PD solutions has been associated with more weight gain and fat mass increase than non–glucose-based PD solutions. Increased truncal fat mass and visceral adiposity over time using glucose-based PD solutions has been positively associated with serum leptin and inflammation and negatively associated with serum adiponectin, , providing indirect evidence of an important link between truncal fat mass and cardiometabolic risk markers. Increased abdominal adiposity has been implicated to contribute significantly to a higher cardiovascular risk in PD patients. An in vitro study found that leptin secretion is directly stimulated in adipocytes being exposed to glucose-based PD solutions via activation of the hexosamine pathway. The degree of leptin secretion correlated directly with the amount of glucose being exposed. Similar stimulation on leptin secretion was not noted with non–glucose-based solutions. In keeping with these experimental data, higher leptin levels were reported in PD patients than in hemodialysis patients.
Several studies have examined whether long-standing exposure to glucose-based PD solutions may be associated with worse clinical outcomes. So far, most available studies were retrospective and results were inconsistent. A retrospective analysis from Taiwan suggested that a higher initial glucose load within the first 6 months of initiation of PD was associated with higher prevalent diabetes mellitus, lower serum albumin, and lower RKF and a higher risk for technique failure but not mortality. Another extended follow-up study using time-dependent covariate analysis reported that long-term glucose exposure was associated with a greater risk for mortality and technique failure. In keeping with these observations, a retrospective analysis from China found that use of PD solutions of higher glucose concentration over the initial 6 months of PD treatment was associated with a greater risk for all-cause and cardiovascular disease mortality. Metabolic syndrome, which consists of a constellation of clinical features including central obesity, hypertension, atherogenic dyslipidemia, hyperglycemia, and insulin resistance, has been found to be associated with adverse cardiovascular outcomes in nondiabetic PD patients.
Hyperglycemia and hyperinsulinemia induced by hypertonic glucose PD solutions may also be associated with short-term adverse hemodynamic changes. A 4.25% dextrose solution has been found to increase blood pressure and cardiac output acutely in the absence of any acute changes in left ventricular diameter, whereas non–glucose-based PD solution was not associated with similar hemodynamic disturbance.
Non–Glucose-Based Peritoneal Dialysis Solutions
Glucose Polymer Solutions
Icodextrin is a starch-derived, branched, water-soluble glucose polymer with an average molecular weight between 13,000 and 19,000 Da. The current commercially available formulation is a 7.5% icodextrin solution with a sodium concentration of 133 mmol/L and a lactate concentration of 40 mmol/L and is isoosmotic (284 mOsmol/L) (see Table 31.1 ). Icodextrin is not significantly metabolized in the peritoneum. Instead it is slowly absorbed into the bloodstream via the lymph vessels, with around 40% being absorbed after a 12-hour period, and is metabolized into oligosaccharides and maltose by circulating α-amylase. The maltose cannot be metabolized in the circulation of humans because maltase is not in the circulation but is present in the kidney and intracellularly in the body. Nevertheless, there is no evidence to date that maltose accumulates within patients treated with icodextrin. Because icodextrin stays in the peritoneal cavity for a considerable period and very little gets reabsorbed, icodextrin is a superior osmotic agent and has better ultrafiltration capacity compared with conventional glucose, especially with longer dwell hours. Icodextrin was approved for use by the Food and Drug Administration of the United States in 2002 to increase ultrafiltration in PD patients.
Effects on ultrafiltration, volume status, and technique survival
Over the years, considerable clinical data demonstrated the efficacy of icodextrin in increasing ultrafiltration in PD patients compared with standard glucose solutions. When used as once-daily overnight long dwell in CAPD or long day dwell in continuous cycler PD (CCPD), icodextrin can achieve a net ultrafiltration equivalent to or even more than that of 2.27% or 3.86% glucose solution, depending on the peritoneal membrane transport characteristics and length of dwell. Generally the ultrafiltration, relative to glucose, is greater in patients with high PSTR. In patients with high or high-average peritoneal membrane transport, the ultrafiltration volume with icodextrin is significantly larger than that with 2.27% or 3.86% PD solution. As a result of better ultrafiltration, more sodium is also removed by convection using icodextrin compared with 2.27% glucose PD solution. Thus it has been used as a salvage therapy in patients with clinically inadequate ultrafiltration or PMF. There is a suggestion that icodextrin may extend the life of PD treatment.
Two earlier short-term RCTs both demonstrated better overall ultrafiltration with icodextrin compared with conventional glucose solution, resulting in better volume status as assessed by bioimpedance. Konings et al. also suggested a decrease in left ventricular mass index with icodextrin. Davies et al. reported that an icodextrin group had weight loss, whereas a standard glucose solution group had weight gain. This may be explained by more fluid removal or less fat mass gain with icodextrin. Another prospective RCT found that icodextrin improved technique survival rate in PD patients with background diabetic nephropathy, though a more rapid decline in RKF was found over a 24-month follow-up period.
Several meta-analyses have reviewed the use of icodextrin versus standard glucose PD solutions in relation to peritoneal ultrafiltration and small solute clearance ( Table 31.2 ). All studies showed significantly better daily peritoneal ultrafiltration with icodextrin than standard glucose PD solutions. The systemic review by Cho et al. additionally found a significantly lower incidence of uncontrolled fluid overload with icodextrin compared with standard glucose PD solutions without compromising RKF. Icodextrin may be a useful salvage therapy in PD patients with refractory fluid overload or UFF and may prolong technique survival. However, so far there are no prospective trials examining the effects of icodextrin on patients’ survival as the primary endpoint. Based on available evidence, the ISPD Adult Cardiovascular and Metabolic Guidelines 2015 recommended that once-daily icodextrin be considered as an alternative to hypertonic glucose PD solutions for long dwells in those experiencing difficulties to maintain euvolemia as a result of insufficient peritoneal ultrafiltration, taking into account the PSTR. The level of recommendation for the statement was given a grading of 1B. Similarly, the European Best Practice Working Group (EBPG) recommended icodextrin be used as the long dwell in high transporter patients with a net peritoneal ultrafiltration <400 mL during a PET with a 3.86% glucose solution.
| Author, Year | No. of Studies (N) | Key Results |
|---|---|---|
| He et al., 2011 | Total nine trials: Two trials ( n = 131), peritoneal creatinine clearance Five trials ( n = 528), peritoneal creatinine clearance rate Four trials ( n = 508), peritoneal urea clearance rate | Significant ↑ in peritoneal efficiency ratio with icodextrin vs. control (MD 6.84, 95% CI, 4.43–9.25) Significant ↑ in peritoneal creatinine clearance rate with icodextrin vs. control (MD 0.51, 95% CI, 0.35–0.67) Significant ↑ in peritoneal urea clearance rate with icodextrin vs. control (MD 0.43, 95% CI, 0.26–0.61) |
| Cho et al., 2013 and Cho et al., 2014 (same findings) | Total 11 trials: Four trials ( n = 102), peritoneal ultrafiltration Two trials ( n = 100), fluid overload episodes Four trials ( n = 114), RKF Three trials ( n = 69), urine volume ( n = 69) | Significant ↑ in peritoneal ultrafiltration with icodextrin vs. control (MD 448.54, 95% CI, 289.28–607.80 mL/d) Significant ↓ in uncontrolled fluid overload episodes with icodextrin vs. control (RR 0.30, 95% CI 0.15–0.59) Did not compromise RKF (standardized MD 0.12, 95% CI, −0.26 to 0.49) or urine output (three trials; 69 patients; MD −88.88, 95% CI, −356.88 to 179.12 mL/d) with icodextrin use for up to 2 years |
Effects on metabolic profile
Use of icodextrin as the long-dwell solution also minimizes glucose exposure and incurs less metabolic derangement compared with standard glucose PD solutions. Changes in triglycerides level have been suggested to correlate with the amount of peritoneal glucose absorption. Several prospective RCTs found that icodextrin, by minimizing glucose absorption, may improve glucose metabolism, improve insulin sensitivity, and reduce dyslipidemia compared with standard glucose PD solutions. An earlier prospective open-label multicenter study from Japan found that switching PD patients from standard glucose PD solutions to icodextrin reduced triglycerides and mean total and low-density lipoprotein cholesterol levels. Among those with glycated hemoglobin ≥6.5%, glycemic control also improved with icodextrin. A randomized study from Mexico found that diabetic PD patients using icodextrin as the long dwell had reduced insulin requirement, lower fasting glucose, lower glycated hemoglobin, lower serum triglycerides, and fewer adverse events compared with a standard glucose group. IMPENDIA and EDEN trials evaluated the clinical effects of a glucose-sparing PD regimen, namely icodextrin and amino acid–based PD solutions, in a multicenter study of 251 diabetic PD patients. The study reported an average of 0.5% improvement in glycated hemoglobin (as the primary study outcome) in the intervention group versus standard glucose group at 6 months. The mean glycated hemoglobin reduced from 7.7% to 7.2%. Secondary outcomes included a significant reduction in very-low-density lipoprotein cholesterol, serum triglyceride, and apo-B levels in the intervention group compared with control group. However, a significantly higher incidence of treatment-related adverse events, serious adverse events, hypoglycemia, and volume overload as well as study withdrawal were reported in the intervention group. Icodextrin may also have favorable effects on plasma adipokines.
Another prospective open-label RCT from Japan failed to find a significant benefit of icodextrin on glycemic control or lipid profile in diabetic PD patients, but both glycemic control and lipid profile were secondary study outcomes. A more recent multicenter open-label short-term (90 days) RCT of 60 nondiabetic automated PD patients found that icodextrin use as the long dwell significantly reduced insulin resistance as denoted by homeostatic model assessment (HOMA) index. There are some suggestions that using twice-daily exchanges of icodextrin in place of standard glucose PD solution may reduce peritoneal glucose exposure by up to 50%, thus lowering systemic glucose absorption by 60%. Longer-term studies are needed to evaluate the safety of this approach because icodextrin metabolite levels will be higher than a single daily exchange.
Effects on residual kidney function
Although there were some earlier suggestions that increased ultrafiltration with icodextrin may jeopardize RKF, a systematic review by Cho et al. did not find any significant difference between icodextrin and standard glucose PD solution in relation to decline in RKF and urine volume. However, RKF and urine volume were not primary outcomes in these studies. In a prospective multicenter RCT from Korea examining changes in residual glomerular filtration rate (GFR) and urine volume as the primary outcomes, there were some suggestions that icodextrin may be associated with slower residual urine volume loss compared with standard glucose PD solutions over a period of 12 months in both the intention-to-treat analysis and per-protocol analysis.
Potential adverse effects
Use of icodextrin may be associated with potential adverse events such as sterile peritonitis or skin rash as a result of allergy to starch. Sterile peritonitis with icodextrin has been described and was related to contamination of icodextrin by peptidoglycan, which is a constituent of bacterial cell walls. Clinically, patients with sterile or chemical peritonitis may remain well despite having cloudy effluent. The differential cell count of PD fluid is associated with leukocytosis with predominantly eosinophilia but not neutrophils. PD effluent usually clears up rapidly on withdrawal of icodextrin. According to the systematic review by Cho et al., the incidence of chemical peritonitis was not increased with icodextrin compared with standard glucose PD solutions. The reported incidence of skin rash with icodextrin use is around 10% but may be up to 18.9%. It is usually mild and localized to the palm of hands. Occasionally, a more severe form of exfoliating dermatitis may occur, requiring withdrawal of icodextrin.
The presence of icodextrin and its metabolites in plasma may interfere with some laboratory analytical methods on glucose measurements. For example, certain glucometers that use glucose dehydrogenase–pyrroloquinoline quinone will overestimate blood glucose in patients using icodextrin.
Biocompatible Peritoneal Dialysis Solutions
Neutral pH, low-GDP solutions
Epithelial-to-mesenchymal transition (EMT) of peritoneal mesothelial cells is a hallmark feature in the peritoneum of PD patients and was suggested in earlier studies to play an essential role in the initiation of peritoneal fibrosis, leading to peritoneal membrane function decline and failure. In vitro experiments found that mesothelial cells culturing in neutral pH, low-GDP solutions had less EMT than in standard glucose solutions. Effluent mesothelial cells grown ex vivo from patients treated with neutral pH, low-GDP solutions also had a trend toward maintaining more of an epitheloid phenotype with less induction of proinflammatory cytokine production, whereas mesothelial cells from patients treated with lactate-buffered standard glucose PD solution invariably transdifferentiated into a nonepithelial phenotype by 24 months. This led to the development and clinical application of more “biocompatible” PD solutions to minimize adverse effects to the peritoneum. However, a later experimental study used inducible genetic fate mapping and demonstrated type I collagen–producing submesothelial fibroblasts as specific progenitors of α-smooth muscle actin–positive myofibroblasts that accumulated progressively in various models of high-glucose solutions or transforming growth factor β 1 –induced peritoneal fibrosis. Notably, only submesothelial fibroblasts but not mesothelial cells expressed α-smooth muscle actin after induction of peritoneal fibrosis in mice. Furthermore, pharmacological inhibition of platelet derived growth factor receptor, which is expressed by submesothelial fibroblasts but not mesothelial cells, attenuated peritoneal fibrosis. These data provide important novel evidence that submesothelial fibroblasts may be the key cell source of myofibroblasts driving peritoneal injury and fibrosis.
For the glucose-based PD solutions to be more biocompatible clinically, a dual-chamber bag system was designed to allow heat sterilization and storage to occur at a lower pH in a separate bag to minimize GDP generation. Some of the low-GDP solutions used bicarbonate instead of lactate as the buffering system to minimize exposure to lactate as well. Thus mixing the contents of the two chambers just before use would produce a more physiological and neutral pH of around 7.0 (see Table 31.1 ). Earlier preclinical studies found that use of neutral pH, low-GDP solutions was associated with better preserved peritoneal membrane morphology and function and better host immune defense. Several subsequent observational studies also suggested similar findings, that use of neutral pH, low-GDP solutions may be associated with more favorable peritoneal membrane morphology, less systemic inflammation, and better patient survival outcomes. Numerous RCTs were conducted in the last 10 years on neutral pH, low-GDP solutions; these are summarized in Table 31.3 .
| Reference | n | I/P | Follow-up Duration | Results |
|---|---|---|---|---|
| Feriani et al., 1998 | 69 | I | 24 weeks | No significant difference was found in decline in residual renal clearances between biocompatible, 2.7 ± 2.8–2.8 ± 4.4 mL/min, and standard glucose group, 3.6 ± 4.1–3.0 ± 3.3 mL/min. |
| Szeto et al., 2007 | 50 | 1 | 12 months | No significant difference was found in the rate of decline of RKF between neutral pH, low-GDP solutions (–1.19 ± 2.23 mL/min/1.73 m 2 ) and control group (–1.02 ± 3.27 mL/min/1.73 m 2 ). |
| Choi et al., 2008 | 104 | 1 | 12 months | No significant difference was found in the change in renal clearances or urinary volume between neutral-pH, low-GDP solution and standard glucose group. |
| Fan et al., 2008 | 93 | I | 12 months | No significant difference was found in the change in RKF between biocompatible and standard glucose groups. |
| Kim et al., 2009 | 91 | 1 | 12 months | No significant difference was found in urine volume in the two groups at any time point. Using a mixed-models method and adjusting for baseline differences in kidney function, the residual GFR at 12 months was significantly higher in the neutral pH, low-GDP solution group ( P = 0.048). |
| Haag-Weber et al., 2010 | 69 | I | 18 months | The rate of decline in RKF was significantly slower with biocompatible PD solution Gambrosol Trio (–1.5%/month [95% CI, –3.07% to 0.03%]) vs. standard glucose solution (–4.3%/month [95% CI, –6.8% to –2.06%, P = 0.04]); also, there was a slower decline in 24-hour urine volume (12 vs. 38 mL/month, P = 0.02) |
| Johnson et al., 2012 | 185 | I | 2 y | No significant difference was found in the slope of decline in RKF in either year 1 (–0.22 vs. –0.28 mL/min/1.73 m 2 ) or year 2 (–0.09 vs. –0.10 mL/min/1.73 m 2 ). A significantly longer time to anuria was reported in individuals assigned to neutral pH, low-GDP solution ( P = 0.009). |
| Lai et al., 2012 | 125 | I | Mean, 2.3 years | No significant difference was found in RKF in patients at end of follow-up (2.30 ± 0.36 vs. 1.69 ± 0.28 mL/min/1.73 m 2 ); urine volume was higher in neutral pH, low-GDP solution group (745.7 ± 107.57 vs. 475.1 ± 77.69 mL/d). |
| Lui et al., 2012 | 150 | I | 12 months | No significant difference was found in the rate of decline in RKF between the group using 1.5%/1.1%/7.5% vs. standard solutions (–0.76 ± 1.77 vs. –0.91 ± 1.92 mL/min/1.73 m 2 per year; P = 0.60); patients using the biocompatible PD fluids had better preservation of daily urine volume (959 ± 515 vs. 798 ± 615 mL/d; P = 0.02). |
| Park et al., 2012 | 146 | I | 12 months | No significant difference was found in RKF (2.9 ± 2.3 vs. 2.9 ± 3.1 mL/min) or urine volume (625 ± 488 vs. 644 ± 575 mL) in neutral pH, low-GDP solutions vs. standard solutions group. |
| Cho et al., 2013 | 60 | I | 12 months | No significant difference was found in RKF (2.4 ± 1.7 vs. 2.2 ± 2.1 mL/min) or urine volume (714 ± 537 vs. 682 ± 460 mL) in neutral pH, low-GDP solution vs. standard glucose solution group. |
Effects on residual kidney function and urine volume
The Euro Balance Trial was the first RCT that examined the effects of neutral pH, low-GDP solutions on the peritoneal membrane and RKF in a crossover design. The study found that use of neutral pH, low-GDP solutions is accompanied by a significant improvement in peritoneal membrane integrity as reflected by higher levels of peritoneal cancer antigen 125 (CA125) and procollagen peptide and significantly decreased circulating AGE levels compared with standard glucose solution over a 12-week time frame. Renal urea and creatinine clearance (CrCl) and urine volume also increased after exposure to neutral pH, low-GDP solutions.
To date, the balANZ trial represents the largest RCT with the longest follow-up (2 years) that examined the effect of neutral pH, low-GDP solution on RKF and various other patient outcomes. It was an investigator-initiated, multicenter, multicountry, open-label, parallel-design RCT including 185 incident PD patients. The primary outcome of the study was slope of RKF decline, which did not have a significant difference between the neutral pH, low-GDP solution group and the control group (standard glucose solution) up to 2 years. However, the biocompatible group had significantly longer times to anuria ( P = 0.009), longer time to first peritonitis episode ( P = 0.01), and lower rates of peritonitis (0.30 vs. 0.49 episodes per year, P = 0.01) compared with the standard glucose solution group. Nevertheless, these were secondary outcomes. No significant difference was reported between the two groups in relation to technique survival and patient survival. In keeping with the findings from the balANZ trial, several other investigators found that neutral pH, low-GDP solutions attenuated decline of RKF in incident PD patients in an RCT setting. However, some other controlled trials did not reproduce similar results, which may be related to the rather short study durations of up to 12 months and relatively small sample sizes.
The findings from these RCTs were summarized in the systematic review by Cho et al. Essentially, the use of neutral pH, low-GDP solutions appeared to produce better preservation of RKF and greater urine volumes. In the Cochrane review, the benefit of low-GDP solutions in preserving RKF was found most significantly between 12 to 23 months ( P = 0.0005) but also from 24 months and beyond, though less significant ( P = 0.04). Similarly, low-GDP solutions were beneficial in preserving urine volume between 12 to 23 months ( P = 0.0005) and from 24 months and beyond ( P = 0.04) 80 ( Fig. 31.3 ). In addition, in 2016 the Trio study involving 67 incident PD patients also reported significantly slower decline in RKF in the lactate-buffered low GDP solution (Gambrosol Trio) group compared with the standard glucose PD solution (see Table 31.3 ). Based on these data, the 2015 ISPD Adult Cardiovascular and Metabolic Guidelines recommended that neutral pH, low-GDP solutions be considered for better preservation of RKF if used for 12 months or more (grading 2B).
Effects on peritonitis risk
In the balANZ trial, the time to first peritonitis episode was longer and overall peritonitis rates were lower with neutral pH, low-GDP solution compared with standard glucose solutions. A separate analysis on the same trial suggested a broad reduction in peritonitis caused by gram-positive, gram-negative, and specifically non- Pseudomonas gram-negative organisms as well as less severe peritonitis in patients using biocompatible solutions. However, similar findings were not reported in other trials. The Cochrane review did not find any difference in peritonitis rates between neutral pH, low-GDP solutions and standard glucose PD solutions.
Effects on peritoneal solute transport and ultrafiltration volume
The effects of biocompatible solutions on ultrafiltration volume were inconclusive. Several studies suggested lower ultrafiltration volume with neutral pH, low-GDP solutions compared with standard glucose solutions. In a separate analysis of the balANZ trial, the mean D/P creatinine ratio at 4 hours remained stable in the neutral pH, low-GDP solutions group but increased significantly in the control group over 2 years. Ultrafiltration volume was lower in the neutral pH, low-GDP solution group compared with the control group up to 6 months. However, over 2 years, ultrafiltration volume increased significantly in the neutral pH, low-GDP solution group but remained stable in the control group. The differential effects of neutral pH, low-GDP solutions on peritoneal solute transport and ultrafiltration over time remain poorly understood.
Another study suggested that neutral pH, low-GDP solution may be associated with a higher ultrafiltration volume. Overall the recent Cochrane and Canadian systematic reviews suggested a trend toward lower ultrafiltration volume with biocompatible PD solutions compared with standard glucose solutions but not reaching statistical significance. No significant difference was found in the peritoneal creatinine clearance.
Other clinical outcomes
Earlier studies suggested a potential benefit in reducing inflow pain with more biocompatible PD solutions. However, a recent Cochrane review suggested that the trend toward lower incidence of inflow pain with neutral pH, low-GDP solutions did not reach statistical significance. There is no convincing evidence to indicate that use of neutral pH, low-GDP solutions improve patients’ survival and technique survival. So far, no studies have been adequately powered to examine the impact of neutral pH, low-GDP solutions on clinical hard outcomes such as hospitalizations or cardiovascular events.
Amino Acid Peritoneal Dialysis Solutions
Amino acid PD solution was launched back in the 1990s with an aim to supplement and replace nitrogen losses in PD patients. PD patients may lose up to 3 to 4 g/day of amino acids and 4 to 15 g/day of proteins even in the stable condition. The amount of peritoneal protein and amino acids loss may increase further with peritonitis. The only commercially available amino acid PD solution is 1.1% solution, which contains 87 mmol/L of amino acids, the majority (61%) of which is essential amino acids (see Table 31.1 ). A 1.1% amino acid solution exerts a similar osmotic force to 1.36% glucose solutions, thus providing an ultrafiltration volume comparable to that achieved with 1.36% glucose solutions. The peak plasma amino acid concentration is usually achieved around an hour. Usually approximately 72% to 82% of amino acids are absorbed in a single daily dwell and this may amount up to 18 g/day, thus providing a good source of protein supplement without adding phosphorus load.
The 1.1% amino acid solution has been found to have a very small vasodilatory effect on peritoneal blood flow and may increase small solute transport. Earlier balance studies suggested that the nitrogen absorbed from a single daily dwell of 1.1% amino acid solution is sufficient to offset the daily losses of amino acids and protein from the peritoneum. Compared with using glucose solution only, combined amino acids and glucose PD solutions have been found to improve protein kinetics and whole-body protein synthesis in PD. Nevertheless, this was not accompanied by a parallel increase in serum albumin, suggesting different pathways involved in albumin and skeletal muscle protein synthesis.
Several RCTs have been conducted to evaluate the safety of 1.1% amino acid solution and its efficacy in treating malnutrition, or what is now called protein-energy wasting , in PD patients. An earlier 12-week multicenter study reported an increase in circulating insulin growth factor I in PD patients randomly assigned to 1.1% solution compared with controls, suggesting an increase in protein synthesis. A subgroup analysis indicated an increase in plasma prealbumin and transferrin levels, yet no change in midarm muscle circumference levels among those with plasma albumin <35 g/L. Another RCT using 1.1% amino acid solution was conducted in Chinese malnourished PD patients. In the 3-year study, participants who were randomly assigned to use 1.1% solution had an improvement in serum albumin, triglyceride, and total cholesterol level as well as a reported increase in dietary protein intake and an increase in normalized protein equivalent of nitrogen appearance (nPNA). Some anthropometric markers improved, although composite nutrition scores did not differ between treatment and control groups. Interestingly, the nutritional benefits appeared more prominent in women, whose lean body mass and body mass index were well maintained with combined amino acids and glucose PD solution but not with glucose solutions. The study was not powered to detect a difference in survival outcomes because only 60 participants were included. No significant difference was noted in inflammatory profile, mortality rates, and duration of hospitalization. Another 6-month crossover study failed to find any positive benefit on lipid metabolism with 1.1% amino acid solution in PD patients. In relation to peritoneal membrane function, preliminary data suggested that 1.1% amino acid solution may be more biocompatible to the peritoneal membrane and mesothelial cells than standard glucose solutions, as reflected by both in vitro and in vivo response of peritoneal CA125.
In summary, the overall clinical value of 1.1% amino acid solution in PD patients has remained equivocal and uncertain. Generally, 1.1% amino acids PD solution is safe. Potential adverse effects include increased nausea and anorexia. Some patients may develop mild metabolic acidosis. This may be ameliorated by adjusting to use a bicarbonate-based buffer in the remaining PD exchanges. 1.1% amino acids solution may be reserved as a glucose-sparing solution in PD subjects and also for use in subjects at risk or exhibit features of PEW syndrome. Adequately powered randomized studies are needed to better define the outcome benefits of this solution.
Low-Sodium Peritoneal Dialysis Solutions
Sodium removal is largely achieved by convection in PD through ultrafiltration. The concept of using low-sodium dialysate is to enhance sodium removal by diffusion, given the larger concentration gradient between plasma and dialysate sodium. However, low-sodium dialysate is currently not in use in clinical practice. In a 2009 nonrandomized controlled trial by Davies et al., two novel solutions designed from predictions using the three-pore model were compared. In one group, sodium concentration was 115 mmol/L and glucose concentration was increased to 2.0% to compensate for reduced osmolality. In the other group, sodium concentration was 102 mmol/L and glucose concentration was unchanged (2.5%). Both solutions were substituted for one 3- to 5-hour exchange per day and no change was made to the rest of the PD regimen. The results indicated that use of the compensated low-sodium dialysate increased the diffusive component of sodium removal while maintaining ultrafiltration. This in turn translated to an improvement in blood pressure, thirst, and fluid status. However, uncompensated low-sodium dialysate had no effect at all on blood pressure, thirst, and fluid status, suggesting that these benefits cannot be achieved just by manipulating the dialysate sodium level alone.
In a prospective multicenter double-blind RCT conducted to prove noninferiority of total weekly urea clearance (Kt/V) with low-sodium versus standard-sodium PD solutions, 108 patients were randomly assigned to low-sodium dialysate (sodium concentration 125 mmol/L) versus standard sodium dialysate (sodium concentration 134 mmol/L) for a 12-week treatment. The noninferiority of low-sodium PD solution for total Kt/V could not be confirmed. Although no difference was found in PD Kt/V, renal Kt/V was different between the two groups, being higher in standard-sodium than in low-sodium group. Low-sodium group showed higher mean daily dialysate sodium removal than high-sodium group. Blood pressure also decreased more with the low-sodium group, resulting in less antihypertensive medication use. Freida et al. created their own compensated “bimodal solution” using a combination of icodextrin and glucose solution. The combination solution had a sodium concentration of 121 mmol/L, and a 15-hour dwell was associated with enhanced net ultrafiltration (mean 990 mL) and sodium removal (mean 158 mmol) compared with 7.5% icodextrin (mean net ultrafiltration 462 mL, mean net sodium removal 49 mmol) or 3.86% glucose-based PD solution (mean net ultrafiltration –85 mL, mean net sodium removal 16 mmol). Compared with 7.5% icodextrin, the combination solution was also associated with significantly higher urea and creatinine clearances, by 41% and 26%, respectively.
Another nonrandomized trial from the same group of investigators using the bimodal solution combining colloids and crystalloids solution found that over a 4-month period, net ultrafiltration and peritoneal sodium removal using bimodal solution as the long dwell increased more than twofold compared with baseline. The estimated change (95% CI) from baseline in net daytime ultrafiltration was 150% (106% to 193%) for the bimodal solution versus 18% (–7% to 43%) for icodextrin and was highly significant. The estimated change from baseline in peritoneal sodium removal was 147% (112% to 183%) for the bimodal solution versus 23% (–2% to 48%) for icodextrin ( P < 0.001). The estimated change from baseline in ultrafiltration efficiency was also significantly higher with bimodal solution than icodextrin (71% vs. –5%). These data suggest that a bimodal solution based on a mixture of glucose (2.6%) and icodextrin (6.8%) significantly improved both ultrafiltration and peritoneal sodium removal and warrants further evaluation.
Future Development in Peritoneal Dialysis Solutions
Hyperbranched polyglycerol (0.5 to 3 kDa) is a nontoxic, nonimmunogenic, water-soluble polyether polymer and has been found to be an efficacious and biocompatible osmotic agent in a rodent model of PD. Preclinical data suggested that hyperbranched polyglycerol functioned as colloids and induced osmosis mainly through capillary small pores and provided superior fluid and waste removal compared with Physioneal solution. More recent animal data indicated that hyperbranched polyglycerol solution may preserve peritoneal membrane function and structure better than standard glucose solution in a rat model of chronic PD. Pyruvate has been suggested as the other alternative buffer. However, all these alternative solutions require further testing and evaluation before clinical application.
Conclusions
The standard glucose-based PD solutions have adverse effects both locally to the peritoneal membrane and systemically to the metabolic profile of PD patients. There is a need to move to a glucose-sparing regimen for PD patients. Preclinical and observational studies suggested that biocompatible neutral pH, low-GDP solutions may improve peritoneal membrane morphology. However, this was not confirmed in the RCT setting. Data from RCTs indicated that the biocompatible neutral pH, low-GDP solutions preserved RKF and urine volume better than standard glucose solutions without finding any increase in adverse events and thus should be adopted more readily clinically in patients with preserved RKF. The latest evidence from a systematic review of RCTs suggested that glucose polymer solution significantly increased peritoneal ultrafiltration and reduced episodes of uncontrolled volume overload compared with standard glucose solutions. None of these solutions produced a difference in the peritonitis risk compared with standard glucose solutions.
Dialysis Adequacy and Prescription
Defining Dialysis Adequacy
Starting in the 1990s, the term peritoneal dialysis adequacy was used to denote small solute clearance, namely Kt/V normalized to total body water and CrCl normalized to body surface area. Total weekly Kt/V and CrCl are each composed of two components, namely clearance from RKF and clearance from PD. The dialysis component is modifiable. Thus the term dialysis prescription was introduced to denote the prescription of PD therapy to optimize dialysis adequacy to reach adequacy targets.
The United States Centers for Medicare and Medicaid Services ESRD Dialysis Incentive Program included Kt/V as a dialysis adequacy comprehensive clinical measure. However, there is increasing recognition that the term dialysis adequacy should encompass more than small solute clearance, Kt/V. Adequacy of dialysis should reflect measures that maximize the sum of survival, cardiovascular outcomes, extracellular volume status, CKD-MBD control, and patient-centered outcomes such as quality of life, appetite, nutrition status, and sleep, in addition to biochemical indices of dialysis adequacy.
The American Society of Nephrology Dialysis Advisory Group recently proposed a multidimensional measure that moves beyond small solute clearance and that aims to quantify dialysis adequacy. This multidimensional measure of optimal dialysis encompasses not only small solute removal, ultrafiltration rate, and extracellular volume overload but also left ventricular geometry, higher weight range middle molecule and phosphate removal, blood pressure variability, serum potassium control, anemia management, and, most importantly, patient-reported outcomes ( Fig. 31.4 ). There are evolutionary changes coming from the concept of adequacy of dialysis to optimal dialysis in the Standardised Outcomes in Nephrology (SONG)–PD Initiative, which accounts more for patient-centered outcomes such as self-reported quality-of-life measures ( Fig. 31.5 ). Adequacy of dialysis should not just include numerical indices of biochemical or clinical parameters but also, more importantly, how patients themselves feel with the therapy received.
Measuring Biochemical Indices of Dialysis Adequacy
The measurement of biochemical indices of dialysis adequacy requires the simultaneous collection of all PD effluent drained out and all urine output passed within the same 24-hour period together with a blood sample for serum urea and creatinine collected around the same time frame. In practical terms the blood sample is usually collected at the time when patients return both the 24-hour PD effluent and simultaneous 24-hour urine to the dialysis center. The PD component of Kt/V and CrCl is calculated by measuring urea and creatinine quantity in a 24-hour collection of PD fluid, respectively ( Table 31.4 ). These values are divided by serum urea and creatinine levels to give the PD Kt/V and CrCl, respectively. The PD and residual renal component of urea and CrCl are then added to give a total weekly Kt/V and CrCl, respectively. Both Kt/V and CrCl values are conventionally expressed as weekly rather than daily to allow comparisons with hemodialysis.
| Indices of Dialysis Adequacy | Formulas |
|---|---|
| Total weekly Kt/V | = (Daily PD + renal urea clearance) × 7 and normalized by V |
| Daily PD Kt/V | = [(24-hour PD volume in L) × (24-hour PD fluid urea concentration in mmol/L) / Plasma urea concentration in mmol/L] and normalized by V |
| Daily renal Kt/V | = [(24-hour urine volume in L) × (24-hour urine urea concentration in mmol/L) / Plasma urea concentration in mmol/L] and normalized by V |
| Total weekly CrCl | = (Daily PD + renal creatinine clearance) × 7 and normalized by BSA |
| Daily PD CrCl | = [(24-hour PD volume in L) × (24-hour PD fluid creatinine concentration in umol/L) / Plasma creatinine concentration in μmol/L] and normalized by BSA |
| Daily renal CrCl | = [(24-hour urine volume in L) × (average of 24-hour urine urea and creatinine concentration in μmol/L) / Plasma creatinine concentration in μmol/L] and normalized by BSA |
| nPNA (g/d) equations ∗ Bergstrom formula I Bergstrom formula II Randerson I Randerson II Teehan Blumenkrantz I Blumenkrantz II | = 20.1 + (7.5 × UNA in g/d) = 15.1 + (6.95 × UNA in g/d) + (dialysate + urine protein in g/24 h) = 10.76 × (UNA/1.44 + 1.46) and UNA is in g/d = 10.76 × (UNA + 1.46) and UNA is in mg/min = 6.25 × (UNA + 1.81 + 0.031 ∗ BW) and UNA is in g/d = 34.6 + 5.86 × UNA in g/d = 22.5 + 6.16 × UNA in g/d |
| Residual GFR (mL/min/1.73m 2 ) | = Average of (24-h urine urea clearance + creatinine clearance in mL/min) |
| Normalization factor | Formulas |
| V (L) | = 2.447 + (0.3362 × BW in kg) + (0.1074 × BH in cm) – (0.09516 × age in years) for male patients |
| V (L) | = –2.097 + (0.2466 × BW in kg) + (0.1069 × BH in cm) for female patients |
| BSA | = 0.007184 × BW in kg 0.425 × BH in cm 0.725 |
∗ These equations make the assumption that the patient is in steady state, where urea nitrogen output equals urea generation. The Randerson equation also assumes that the average daily protein loss in the dialysate is 7.3 g/day. In PD patients with substantial protein losses in dialysate or urine, these losses must be added to the equation in calculating nPNA.
Notably, when 24-hour PD fluid and urine collections are repeated on separate occasions in the same patient receiving the same PD prescription, significant intraindividual variations occur, especially in the renal component of clearance. These variations may be accounted for by inaccuracies in the 24-hour urine and dialysate collection because they are cumbersome to do or by the timing of blood sampling. There could also be genuine day-to-day variation in the amount of dietary protein intake and fluid intake; urine volume, peritoneal ultrafiltration volume, and degree of equilibration of PD fluids; timing of PD exchanges; and concentrations of PD glucose solutions used. In CAPD either the patient is asked to bring all the dialysate drained out over the 24-hour period to clinic and the total 24-hour dialysate volume is measured in the laboratory, or the patient can measure the volumes of all dialysate drained out over a 24-hour period, mix up all the dialysates drained out well, and bring back a representative volume to the dialysis center. In automated PD the dialysate volumes involved are even larger. APD patients are usually trained to record or measure total cycler effluent volumes at home using the machine reading and bring back a representative aliquot of the dialysate to clinic for measurement of urea and creatinine concentrations.
In CAPD the timing of blood sampling for urea and creatinine may not be as critical because generally serum urea and creatinine do not fluctuate significantly during the day. In automated PD, however, there may be a 10% or greater variation in serum urea and creatinine from a trough value after completion of PD cycles in the morning to a peak value before the patient resumes PD cycles in the evening. This may be a problem especially if patients have no PD fluids dwell during daytime. Thus in patients receiving APD with no PD fluids during daytime, serum samples should be taken approximately halfway through the noncycling period, which means mostly in the early afternoon.
The target total weekly Kt/V values achieved in PD are usually around half to two-thirds of those in hemodialysis. This difference may be explained by the fact that, in contrast to hemodialysis, PD provides a continuous removal of small solute clearance and thus is greater than intermittent hemodialysis that delivers a similar quantity of clearance. Furthermore, the continuous nature of PD avoids the day-to-day fluctuations in small solute clearance as found in intermittent hemodialysis. The concept of a “peak concentration hypothesis” put forward by Keshaviah et al. suggested that peak levels rather than mean levels of small solutes determined uremic toxicity. Theoretically, continuous modalities may be better, because peak levels of uremic waste solutes should be lower for a given clearance than as in the case of intermittent modalities.
Normalization Factor for Urea and Creatinine Clearance
Total weekly Kt/V is normalized to total body water, V, which can be estimated using anthropometric formulas such as Watson or Hume, which are based on age, gender, body weight, and height information. Estimates of V using the Watson formula are on the average slightly lower compared with the gold standard methods for estimating total body water such as deuterium oxide dilution. However, the discrepancy varies substantially in patients with extreme body weight, such as obesity. By far, the Watson formula remains the most commonly adopted equation used to calculate V. Total weekly CrCl is normalized to body surface area (BSA), which is usually estimated by the du Bois formula. In general, edema-free body weight should be used in the formulas to calculate V and BSA (see Table 31.4 ). If patients have significant weight loss as a result of PEW, the ideal body weight rather than actual body weight has been suggested to be used in the formula. This is because wasted patients would have a misleadingly high normalized clearance value as a result of significant weight loss, whereas obese patients would have a misleadingly low value because of overweight. Ideal body weights provided by the National Health and Nutrition Evaluation Survey tables give the median body weight of North Americans of the same age, sex, height, and frame as the patient and are regularly updated. However, it is uncertain whether these values are also directly applicable to non-white populations.
Estimation of Residual Kidney Function
The renal component of urea and CrCl is calculated in the same way as with a 24-hour urine collection, except that in the case of creatinine clearance, an average of residual renal urea and CrCl is typically used (see Table 31.4 ). This is because unmodified urine CrCl substantially overestimates the true GFR because of tubular secretion of creatinine, whereas using renal Kt/V underestimates GFR. Both creatinine and urea are freely filtered by the glomeruli. As part of the biochemical indices of dialysis adequacy, residual GFR is estimated using the average of 24-hour urine urea and CrCl and is normalized to BSA (see Table 31.4 ). It is important to take into account both the peritoneal membrane transport characteristics and RKF of each patient when doing PD prescription and interpreting biochemical indices of dialysis adequacy.
Frequency of Monitoring of Biochemical Indices of Dialysis Adequacy
For all CAPD or APD patients, total weekly Kt/V and RKF should be measured around 4 weeks after initiation of PD therapy. Thereafter, they should ideally be monitored at least once every 6 months to allow for prescription modification if necessary.
Peritoneal Equilibration Test
PET is a simple clinical method that assesses the diffusive transport capacity of urea, creatinine, and other solutes across the semipermeable membrane, as well as ultrafiltration, in different patients.
Patients who are high transporters equilibrate very quickly and have excellent diffusive transport capacity. However, they tend to have a low ultrafiltration volume using standard glucose PD solutions because the osmotic gradient for glucose dissipates relatively quickly ( Fig. 31.6 ). These patients may do better with short dwell times as in APD. However, long hours’ day dwell in APD using standard glucose solution may be a problem in high transporters because PD fluids will be largely reabsorbed. Glucose polymer solution or icodextrin would be preferred as the long day dwell solution for high transporters receiving APD. In contrast, low transporters ultrafiltrate well but equilibrate slowly. Thus longer dwell times may be more effective in removing small solutes in low transporters. Generally, Kt/V is much less affected by peritoneal transport characteristics than CrCl in CAPD because more than 90% of Kt/V and equilibration occurs with the long dwell hours of CAPD, regardless of peritoneal transport characteristics. This differs from creatinine clearance, which may have two to three times difference between low and high transporters even after a 4- to 6-hour dwell (see Fig. 31.6 ). In APD, dwell time is usually shorter than CAPD except for the long day dwell. Thus peritoneal membrane transport characteristics may be considered in the prescription of PD modality and regimen. In patients with complete loss of RKF, there could be limitations in achieving optimal clearance targets, depending on the peritoneal transport characteristics.
Importance of Dialysis Adequacy and Defining Numerical Targets for Biochemical Indices of Dialysis Adequacy
The CANUSA (Canada-USA) study was a prospective multicenter cohort study of 680 incident CAPD patients from Canada and United States that examined the importance of adequacy. Over a 2-year period of follow-up, a significant positive association was identified between the total weekly Kt/V and CrCl and clinical outcomes including overall survival. For every 0.1 unit higher total weekly Kt/V, the relative risk for mortality was lower by 6%, and for every 5 L higher total weekly CrCl, the relative risk for mortality was lower by 7%. Subsequent to the CANUSA study, Maiorca et al. also reported an association between weekly Kt/V >1.96 and better survival in another prevalent cohort of CAPD patients followed over 3 years. These two important studies formed an important basis for the recommended total weekly Kt/V target of at least 2.0 and a CrCl target of at least 60 L/week/1.73 m 2 for those receiving CAPD with high and high-average transporters and a CrCl target at least 50 L/week/1.73 m 2 for those with low and low-average transporters by the National Kidney Foundation Dialysis Outcome Quality Initiative (K/DOQI) Guidelines in 1997 and 2000. If there is discordance in achieving these targets, Kt/V was recommended to be the immediate determinant of dialysis adequacy because it directly reflects protein metabolism and is less affected by extreme variations in RKF. However, a cause for the discrepancy should be looked for and the patient should be followed closely for signs of underdialysis. One potential explanation for the discrepancy could be patients being underweight, which may result in a small V and overall a higher Kt/V.
Subsequent to these studies, two large prospective RCTs found no significant difference in the mortality risk for PD patients by increasing peritoneal small solute clearance. In the ADEquacy of PD in MEXico (ADEMEX) study, increasing weekly peritoneal Kt/V from 1.62 to 2.13 (weekly CrCl from 46.1 to 56.9 L/week/1.73 m 2 ) had no significant effect on mortality risk (relative risk 1.00, 95% CI, 0.8 to 1.24). In another randomized trial from Hong Kong, 320 PD patients with baseline renal Kt/V <1.0 were randomly assigned to three groups of different Kt/V targets, namely 1.5 to 1.7, 1.7 to 2.0, and >2.0. Patients in the group with the lowest Kt/V target 1.5 to 1.7 had more clinical problems and were more likely to be withdrawn from the study by their physicians based on clinical ground. The overall 2-year survival was 84.9% with no significant difference reported between the group with a Kt/V target of 1.7 to 2.0 and the group with a Kt/V target >2.0. Another retrospective analysis from an administrative database of anuric PD patients in the United States demonstrated higher mortality rates in patients with a weekly Kt/V <1.7.
With the exception of the study from Hong Kong, all the other studies failed to find a relationship between peritoneal small solute clearance and survival in PD patients. The ADEMEX RCT and the Hong Kong PD RCT formed the key basis for a subsequent revision of the 2006 K/DOQI, 2005 EBPG, and 2006 ISPD guidelines, all recommending a minimum weekly Kt/V target of 1.7 in CAPD patients. There are no trial data to support benefit of further increasing dialysis to reach a total weekly Kt/V beyond 2.0 or a CrCl of more than 60 L/week/1.73 m 2 . Furthermore, there are trial data to suggest that a weekly total Kt/V <1.7 was associated with more clinical problems and greater need for erythropoietin. In the Netherlands Cooperative Study on the Adequacy of Dialysis (NECOSAD) study, peritoneal Kt/V <1.5 and CrCl <40 L/week/1.73 m 2 were associated with higher mortality. The 2005 EBPG guideline, 2006 ISPD guideline update, and K/DOQI 2006 guideline suggested a minimum weekly peritoneal Kt/V of 1.7 for anuric PD patients.
In APD a cycler is used to perform multiple overnight exchanges. A minimum weekly Kt/V target of ≥1.7 per week is recommended and is in line with the 2006 K/DOQI guideline. In APD patients who are anuric or have minimal urine volume, an additional manual day exchange is suggested to increase PD clearance, because the recommended Kt/V targets for APD patients are based on studies in CAPD patients who underwent continuous PD over a 24-hour period. In the 2006 ISPD guideline, a separate target for CrCl was not suggested for CAPD. In APD, because of more variable relations between Kt/V and CrCl, an additional target of 45 L/week/1.73 m 2 for CrCl was recommended.
Novel concepts in defining dialysis adequacy or optimal dialysis
The minimum target Kt/V defines the minimum but not necessarily the optimal dose of dialysis that should be performed. Optimal dialysis prescription should be personalized, based on an assessment of RKF, peritoneal membrane function, status of clinical and biochemical parameters, acid–base status, uremic symptoms, volume status, nutrition profile, control of mineral metabolism, and patients’ self-perceived well-being on dialysis (see Fig. 31.4 ). In some situations, dialysis dose may need to be increased despite achieving the minimum weekly Kt/V target of 1.7. Examples of these situations are persistent acidosis, persistent uremic symptoms such as nausea and anorexia, clinical manifestations of PEW, and hyperphosphatemia despite dietary restriction and use of phosphorus binders.
Ultrafiltration and Volume Control as a Target for Dialysis Adequacy
Ultrafiltration is an important parameter for assessing adequacy of dialysis, and ultrafiltration has been found to be associated with survival in anuric APD patients; ultrafiltration volume less than 750 mL/day was associated with a higher mortality. However, a numerical target for daily ultrafiltration volume was not formulated or recommended by different guideline bodies. It is because the overall volume status depends also on the residual urine volume as well as salt and fluid intake of patients and there may be substantial intraindividual variation. Both the 2006 ISPD guidelines and the recent 2015 ISPD Cardiovascular and Metabolic guidelines emphasized the importance of maintaining euvolemia in PD patients as one of the treatment goals in PD and recommended that attention be paid to both the urine volumes and PD ultrafiltration volumes.
The negative findings from the ADEMEX study that increasing small solute clearance did not improve survival of PD patients suggested the need to look for factors other than small solute clearance that may be important in determining clinical outcomes of PD patients. A reanalysis of data from the CANUSA study indicated that every 250 mL of urine output was associated with a 36% lower mortality risk, suggesting the importance of urine volume in determining the clinical outcome of PD patients. Sodium and fluid removal are associated with survival in PD patients. The lower the sodium and fluid removal from PD, the greater the mortality risk ( Fig. 31.7 ). Extracellular volume overload is a highly prevalent complication in PD patients. Using bioimpedance spectroscopy, the estimated prevalence of extracellular volume overload was at least more than 50% and was even higher when patients completely lost their RKF. Extracellular volume overload is associated with a higher risk for mortality in PD patients. Apart from loss of RKF, UFF with low drain volume is another important cause of volume overload (see Ultrafiltration Failure). Other non–membrane-related causes of volume overload include dietary nonadherence with excess salt and fluid intake, nonadherence to dialysis prescription, and inappropriate choice of PD solution strengths that may affect the PD outflow. Mechanical complications such as dialysate leaks, hernias, catheter malposition, and loculations in the peritoneal cavity may also cause low drain volume. In fact, retroperitoneal dialysate leaks have been suggested to be an important cause of acute UFF and can be diagnosed using computed tomography peritoneogram. Hyperglycemia may also contribute to low drain volume by decreasing the glucose osmotic gradient–driven ultrafiltration.
