Rapid transporters on maintenance peritoneal dialysis

INTRODUCTION — It is now well appreciated that peritoneal dialysis patients have different peritoneal membrane transport characteristics. These differences are best classified and determined by use of the peritoneal equilibration test (PET) [1] (see "Peritoneal equilibration test"). If the standard PET is done and all recommended measurements obtained, this test has helped characterize the relationship between dwell time, solute transport, glucose absorption, drain volume, and net solute removal. Those patients who have the highest rates of diffusive transport are classified as rapid transporters. According to PET testing in different populations, approximately 15 percent of patients will be rapid transporters at the start of peritoneal dialysis.

As a result of the high rates of diffusive transport, rapid transporters transport small solutes (such as urea, creatinine, and glucose) quickly, leading to equilibration between the dialysate small-solute concentration and that of the blood relatively early in a dwell (figure 1). These patients also rapidly absorb dialysate glucose, leading to early dissolution of the crystalloid osmotic gradient between dialysate and blood that is required to sustain ultrafiltration. Once the osmotic gradient is dissipated, the stimulus for ultrafiltration is gone, and ultrafiltration ceases. However, there is slow but continuous absorption of fluid via the peritoneal lymphatics, potentially leading to poor net ultrafiltration, low drain volume, and potential systemic volume expansion. Lower drain volumes could potentially lead to lower solute clearance and volume overload.

The rate of diffusion of any solute or the dialysate-to-plasma ratio of that solute after a timed dwell is related to the perfusion of existing capillaries and the vascularity of (capillary density) a specified peritoneal surface area. Increased rates of diffusion can be the result of increased perfusion of existing capillaries (increased effective surface area) or an increase in the number of capillaries per unit area (an anatomic increase in capillary number/unit of surface area), both which can be inherent or acquired. Acute and chronic inflammation of the peritoneal cavity related to peritonitis can increase relative perfusion rates, and, over time, the number of capillaries/unit space may increase. These changes can increase rates of diffusive transport. Some peritoneal dialysis patients are rapid transporters at the start of dialysis; others become rapid transporters over time.

If one were to design a peritoneal dialysis prescription based upon transport characteristics alone, ignoring patient convenience or lifestyle constraints, attempting to optimize ultrafiltration, drain volumes, and creatinine clearance, rapid transporters would do best with short dwell times (1.5 to 3 hours/dwell). In theory, when using only glucose-containing solutions, such patients would do best with automated therapies (automated peritoneal dialysis [APD]), such as nightly intermittent peritoneal dialysis (NIPD) or nightly cycler therapy with a last-bag fill (first morning fill) and a midday exchange when only using glucose-containing solutions. By contrast, low-average or low transporters would do best with prolonged dwells, such as those associated with continuous ambulatory peritoneal dialysis (CAPD) or continuous cycling peritoneal dialysis (CCPD) with fewer overnight exchanges (figure 1). Low-average or low transporters may need large, instilled volumes, which enhance clearance by maximizing contact surface area. In such patients, solute clearance is slow but maintained throughout the dwell, while ultrafiltration is sustained for longer dwell times because of the slow rates of transport for both small solutes and glucose. Patients who are high-average transporters would typically do well with either therapy.

Despite these theoretical concepts, most patients could do either CAPD or APD, if one is knowledgeable in peritoneal dialysis kinetics and peritoneal dialysis fluids and willing to individualize the patient's prescription. This is especially true in patients who have residual kidney function, which makes it easier to maintain euvolemia and solute removal. In an extensive observational cohort review of 42,942 patients on CAPD and 23,439 on APD in the United States who started peritoneal dialysis during the years 1996 to 2004 and were followed through September 2006, there was no effect demonstrated of modality on survival risk [2]. In this review, there was no adjustment for transport type. However, in a prior review of patients on peritoneal dialysis in Australia and New Zealand, which adjusted for transport type, there also was no demonstrable difference in risk of death for CAPD or APD [3]. By contrast, an analysis of a more contemporary cohort of patients in peritoneal dialysis in Australia and New Zealand found a lower risk of death in high transporters treated with APD compared with those on CAPD [4]. At our center, approximately 85 percent of all peritoneal dialysis patients are on APD.

The removal of larger solutes, such as B2-microglobulin, is dwell-time dependent, even among rapid transporters. Thus, despite the fact that these patients could likely meet their minimal Kt/V urea targets with NIPD alone (ie, without a daytime dwell), most have a peritoneal dialysis dwell while not on NIPD to optimize the removal of middle-molecules.

This topic discusses various aspects of physiology, pathophysiology, and outcomes among rapid transporters who are receiving maintenance peritoneal dialysis. Other aspects of management that may be relevant to the care of such patients are discussed elsewhere:

●Selection of the peritoneal dialysis modality (see "Evaluating patients for chronic peritoneal dialysis and selection of modality")

●Prescribing peritoneal dialysis (see "Prescribing peritoneal dialysis")

●Available peritoneal dialysis solutions (see "Peritoneal dialysis solutions")

●Management of hypervolemia (see "Management of hypervolemia in patients on peritoneal dialysis")

RAPID TRANSPORT AND MORTALITY — In addition to the potential for suboptimal ultrafiltration volumes and inadequate solute clearance, historical cohort studies had suggested that rapid transport status may be an independent predictor of enhanced mortality among patients undergoing continuous peritoneal dialysis [5-8]. These data were summarized in a 2006 meta-analysis of 20 observational studies, with 19 studies pooled to generate a summary mortality risk based upon transport status, as defined by the ratio of creatinine in the dialysate to plasma after a standardized four-hour dwell (dialysate to plasma ratio [D/P]) [8]. For every 0.1 increase in the D/P value, there was an increase in the relative risk of death of 1.15 (95% CI 1.07-1.23). Thus, as compared with patients with low transport status, an increased mortality risk of 22, 46, and 77 percent was noted for low-average, high-average, and high transporters, respectively. In addition, there was a trend for an increased relative risk for death-censored technique failure with every 0.1 increase in the D/P value (1.18, 95% CI 0.96-1.46). (See "Prescribing peritoneal dialysis".)

The reason for decreased observed survival among these patients is unclear, although most of the deaths were cardiovascular in nature. Postulated mechanisms include fluid overload, malnutrition, increased protein losses, and chronic inflammation [9,10]. The 2006 meta-analysis reviewed data from patients who were mainly on continuous ambulatory peritoneal dialysis (CAPD), a therapy in which rapid transporters would be predicted to do poorly with ultrafiltration and, consequently, low-drain volume and solute clearances. In a small cohort of patients who were all on the same CAPD prescription, compared with low transporters, rapid transporters were more likely to be hypertensive (0 versus 100 percent) and have left ventricular hypertrophy (LVH; 33 versus 100 percent) [11]. In follow-up of these hypertensive patients, blood pressure improved when their prescription was altered to increase ultrafiltration. When these and other observations were noted, more attention was paid to individualizing the peritoneal dialysis prescription for patients with high peritoneal membrane transport. Subsequent, more contemporary data suggest that, if there is an increased baseline risk associated with transport status, it may be more related to clinician practice and management rather than an increased risk due to transport status alone.

The following studies support the theory that the risk may be due to how patients are managed rather than transport status alone [4,12-17].

●In a single-center, observational, cohort study, peritoneal transport status predicted the risk of death in patients who started peritoneal dialysis between 1990 and 1997 [13]; in patients who started peritoneal dialysis between 1998 and 2005, transport status was not predictive of relative risk of death (figure 2) [13]. During the later vintage, patients were more likely to be treated with automated peritoneal dialysis (APD) or alternative osmotic agents, such as icodextrin. With such maneuvers, patients with high membrane transport can be managed on peritoneal dialysis without the increased observed mortality risk demonstrated in historical studies.

●The European APD outcomes study showed that risk of death at one year was not related to transport type in patients treated with APD and icodextrin [14].

●In a single-center study, among 193 incident patients who started peritoneal dialysis between 2000 and 2004 (78 percent of whom were on icodextrin or treated with APD), survival was not related to transport type (71, 69, and 67 percent for rapid, high-average and low-average transporters, respectively) [16].

●A large, historical registry report from Australia and New Zealand confirmed that peritoneal transport type was only a significant predictor of outcome in patients on CAPD, not in those on APD [17]. In a subsequent review of patients on peritoneal dialysis in Australia and New Zealand, which adjusted for transport type, there also was no demonstrable difference in risk of death for CAPD or APD [3]. In addition, in a more contemporary observational study of 628 rapid transporters from Australia and New Zealand (Australia and New Zealand Dialysis and Transplant Registry [ANZDATA]) who started peritoneal dialysis between 1999 and 2004, better survival was associated with APD compared with CAPD [4].

●In the Global Fluid Study, where data on 959 patients from 10 centers in the United Kingdom, South Korea, and Canada were evaluated, higher transport was associated with increased mortality in prevalent but not incident patients [18]. Systemic markers of inflammation, but not dialysate cytokine levels, were associated with all-cause mortality in both incident and prevalent patients.

Finally, in one review, this survival disadvantage was not observed if rapid transporters transferred to hemodialysis [19]. This suggests that early transfer to hemodialysis may be indicated among such patients. However, the reasons for and timing of transfer were not noted.

By contrast, a review of data from a United States dialysis provider demonstrated an association between D/P creatinine values with all-cause mortality (adjusted hazard ratio per 0.1 unit higher -1.07, 95% CI 1.02-1.13) and hospitalization rates (adjusted hazard rate ratio per 0.1 unit higher -1.05, 95% CI 1.03-1.06). Four-hour drain volume was inversely correlated with hospitalization rate, but not all-cause mortality, and neither value predicted technique failure. However, despite the fact that the majority of patients were on APD, only approximately 5 to 8 percent of patients were treated using icodextrin for the long dwell (which should result in a better ultrafiltration volume than a single dextrose exchange), and data on change in peritoneal equilibration test (PET) results from baseline were not known [20].

In summary, we believe that there is no demonstrable survival advantage for APD over CAPD in rapid transporters and that rapid transporters can be managed with peritoneal dialysis if they are managed closely and therapy is individualized as needed [12]. Furthermore, if there is an increase in all-cause mortality associated with rapid transport status, it is likely due to a systemic effect, which is associated with transport rate. At our unit, we do not routinely transfer rapid transporters to hemodialysis, unless there is a clinical indication such as uncontrolled volume overload, hypertension, or malnutrition.

STABILITY OF PERITONEAL MEMBRANE TRANSPORT OVER TIME — Long-term success of peritoneal dialysis is dependent upon the ability of the peritoneal membrane to maintain acceptable solute transfer rates and meet the patient's ultrafiltration needs. Clinical experience suggests that this is the case in most patients; in most studies, creatinine and glucose equilibration rates remained stable in approximately 70 percent of patients at 12 to 18 months and in over 50 percent at 24 months [21]. These clinical observations have been confirmed by histologic examination showing only minimal pathologic changes over time in most patients on maintenance peritoneal dialysis, unless there have been repeated bouts of peritonitis [22].

Some patients are inherently rapid transporters. This is either possibly due to a genetic predisposition or due to an associated comorbidity (where there are high levels of dialysate vascular endothelial growth factor [VEGF] or interleukin-6 [IL-6] [23]) or, in the absence of comorbidity, an increase in the peritoneal mesothelial cell mass (characterized by high peritoneal effluent CA-125 appearance rates, a surrogate for estimating mesothelial cell mass and health) [24]. The risk for these changes may be genetically mediated. An association has been demonstrated between IL–6 gene polymorphisms and peritoneal transport [18,25].

Occasionally, peritoneal transport changes over time. The most frequent acquired change occurs during an episode of peritonitis when there is an increased perfusion per capillary because of inflammation. However, an increase in peritoneal transport is associated with peritoneal angioneogenesis in a subgroup of long-term patients [26].

Another common change that occurs is a gradual increase in solute transport (figure 3) [21]. As noted above, this is usually associated with a decrease in net ultrafiltration rate, a change that historically has been called type 1 membrane failure [27]. The development of rapid transport characteristics is responsible for approximately 70 to 80 percent of cases with permanent loss of ultrafiltration. (See "Inadequate solute clearance in peritoneal dialysis".)

Patients who become rapid transporters after a long time on peritoneal dialysis are likely to have a different pathophysiologic process than those who are rapid transporters at the initiation of dialysis. The rise in solute transport rate has been associated with increasing duration of dialysis and the use of hypertonic dextrose exchanges [28,29]. It is thought that the glucose or the glucose degradation products in these solutions are responsible for the observed changes in the peritoneal membrane over time.

To assess the relationship among changes in solute transport and ultrafiltration over time and patient characteristics, a single-center study was performed of 574 incident peritoneal dialysis patients in which a standard peritoneal equilibration test (PET) was performed at least annually from 1990 to 2003 [29]. During the first six months, there was an increase in solute transport (ie, an increase in dialysate to plasma ratio [D/P] creatinine), but no decrease in ultrafiltration ability. Subsequently, an increase in solute transport and a decrease in ultrafiltration capacity were observed. The eventual ultrafiltration failure in a subset of 48 patients correlated in part with early exposure to higher intraperitoneal glucose concentrations, suggesting that the use of more hypertonic glucose solutions was associated with subsequent increases in transport. Denudation of the mesothelial surface has been noted histologically, similarly to what is seen with peritonitis.

An accumulation of advanced glycosylation end-products in the peritoneal vessel wall may also alter peritoneal permeability. Increased immunohistochemical staining of glycosylated products appears to correlate with increased solute transport rates among longstanding peritoneal dialysis patients [30,31]. The large amount of advanced glycosylation end-products found in conventional peritoneal dialysis fluids may be the direct result of the heat sterilization process and other processes [32-35].

In some patients with acquired increase in transport, a temporary transfer to hemodialysis with "resting" of the peritoneum has been associated with an improvement in ultrafiltration and a return to baseline peritoneal transport characteristics. Shortening the dwell time improves net ultrafiltration and thus also maximizes clearances. As a result, transfer to nightly intermittent peritoneal dialysis (NIPD) or daytime automated peritoneal dialysis (DAPD), which is the use of cycler during the day in nonambulatory patients, may help optimize net ultrafiltration per gram of glucose absorbed and increase appetite. However, it is not known whether this approach will protect the membrane from further damage.

Peritoneal mesothelial cells have a functional local renin-angiotensin system [36]. When unregulated, this can lead to angiogenesis and fibrosis. The production of intermediary cytokines such as VEGF and transforming growth factor beta (TGF-beta) can be reduced with use of angiotensin-converting enzyme (ACE) or angiotensin receptor blocker (ARB) therapies [37]. The only published retrospective, observational study correlating changes in transport over time with ACE or ARB use suggests that these agents may be protective, in that patients being administered these agents were less likely to have a change in transport over time [38]. Further data are needed, but certainly there are other indications for ACE or ARB use (preservation of residual kidney function, cardiovascular protection, hypertension control), so it may be reasonable to use these agents as first line for one of these conditions because it may also prevent peritoneal membrane changes.

CHANGES IN TRANSPORT ASSOCIATED WITH PERITONITIS — Ultrafiltration decreases during peritonitis due to a transient increase in small-solute clearances that is associated with an enhanced rate of glucose absorption, as seen in rapid transporters. This is likely due to an increase in perfusion of existing capillaries due to the local inflammatory response. In addition, peritonitis is often associated with a marked increase in dialysate protein losses. Preliminary results suggest that the use of icodextrin dialysate may help preserve ultrafiltration capabilities in this setting [39]. (See "Peritoneal dialysis solutions", section on 'Glucose polymer-containing solutions (icodextrin)' and "Management of hypervolemia in patients on peritoneal dialysis", section on 'Icodextrin dialysate'.)

These changes are usually transient, but, in some patients, they persist as a result of progression to peritoneal fibrosis and perhaps sclerosing peritonitis. Permanent injury to the peritoneal membrane is most common in long-term peritoneal dialysis patients after severe episodes of peritonitis that fail to respond rapidly to treatment. Patient death and catheter loss during an episode of peritonitis have been associated with a large drop in the plasma albumin concentration that is probably due to increased protein loss in the dialysate [40]. One possible way to minimize the development of peritonitis-induced hypoalbuminemia is the administration of intraperitoneal amino acids [41].

These data highlight the need for us to monitor transport characteristics over time and during episodes of peritonitis, not only to be able to match dwell time with transport type to maximize small-solute clearances and net ultrafiltration, but also to monitor for pathologic changes in the peritoneal membrane. Some clinicians simply anticipate ultrafiltration problems during peritonitis and empirically reduce oral fluid intake and/or increase the use of hypertonic exchanges. (See "Management of hypervolemia in patients on peritoneal dialysis" and "Microbiology and therapy of peritonitis in peritoneal dialysis", section on 'Dialysis prescription'.)

RELATIONSHIP BETWEEN TRANSPORT TYPE AND PLASMA ALBUMIN — Cross-sectional data from continuous ambulatory peritoneal dialysis (CAPD) patients have shown an inverse correlation between peritoneal transport characteristics (as estimated from the four-hour dialysate to plasma [D/P] creatinine ratio) and the plasma albumin concentration [42]. This relationship tends to be independent of dialysis dose, and hypoalbuminemia is typically accompanied by other signs of malnutrition [43]. The association of higher creatinine transfer rates with hypoalbuminemia and malnutrition is probably related to enhanced dialysate protein losses, which can be as high as 15 g/day; in comparison, the mean for all patients on CAPD is approximately 6 g/day. Another explanation is that the high-transport status is caused by chronic inflammation, which itself is associated with low serum albumin concentration.

There are probably other factors that contribute to the tendency toward malnutrition in rapid transporters. As an example, patients on long dwells often require hypertonic fluids to achieve net ultrafiltration and maintain euvolemia. It is possible that the increased glucose load may suppress appetite [44]. Another possibility is that these patients tend to be slightly volume expanded and are hypoalbuminemic in part due to dilution [45]. The role of comorbid diseases and chronic inflammation also must be considered.

SUMMARY AND RECOMMENDATIONS — Long-term success of peritoneal dialysis is dependent upon the ability of the peritoneal membrane to provide adequate solute clearances and ultrafiltration over time. Peritoneal transport characteristics can be monitored most easily by a peritoneal equilibration test (PET). We generally obtain a PET after four weeks on chronic therapy to use as a baseline value and then repeat the test when clinically indicated, such as when there is unexplained volume overload, hypertension, suboptimal Kt/V, or change in drain volumes.

If a patient is found to be a rapid transporter on baseline PET, close attention must be paid to the adequacy of ultrafiltration and nutrition. With continuous ambulatory peritoneal dialysis (CAPD), for example, the loss of residual kidney function (specifically the volume of urine) may lead to fluid retention and volume-dependent hypertension, resulting in the need for a greater use of hypertonic exchanges.

Most patients with high rates of peritoneal transport (diffusion) will have a tendency for hypoalbuminemia, and some will actually be malnourished. The clinician should consider transferring malnourished patients to short-dwell therapy, such as nightly intermittent peritoneal dialysis (NIPD) or daytime automated peritoneal dialysis (DAPD). Both NIPD and continuous cycling peritoneal dialysis (CCPD) have multiple short dwells at night. The major advantage of NIPD in rapid transporters is that the abdomen is dry during the day, thereby possibly minimizing some of the protein losses [46], although not all studies have confirmed this [47,48]. If possible, one should consider using icodextrin for the long daytime dwell. This will minimize glucose absorption, possible appetite suppression, and excessive fluid absorption, while maintaining 24-hour dialysate clearances of substances such as middle molecules. Supplemental parenteral nutrition or, when available, intraperitoneal amino acid solutions should be considered if the patient continues on CAPD. Some patients may need a trial of hemodialysis to see if the malnutrition improves.

Issues concerning the use and effectiveness of icodextrin in this setting are presented separately. (See "Inadequate solute clearance in peritoneal dialysis".)

Adequacy of dialysis — One study in the early 1990s showed that rapid transporters treated with NIPD have a lower rate of dialysate protein loss and a lesser need for hypertonic dialysate solutions [46], whereas other studies showed protein losses to be similar in all types of peritoneal dialysis therapies [47,48]. If the dialysis dose is adequate, then the reductions in protein loss and glucose absorption should permit an increase in energy intake and at least partial correction of the malnourished state. Among the benefits noted in a study of patients initially switched from CAPD to NIPD were [46]:

●An increase in weekly Kt/V from 1.60 to 1.77

●An increase in weekly creatinine clearance from 50.7 to 54.7 L/1.73 m² body surface area

●A rise in the plasma albumin concentration from 3.40 to 3.49 g/dL

●An elevation in protein catabolic rate, an estimate of protein intake, from 0.54 to 0.57 g/kg per day

●A reduction in dialysate protein losses from 6.76 to 5.39 g/day

The lack of normalization of the plasma albumin concentration in this setting may be related to persistent dialysate protein losses, volume overload if the patients volume status is not well managed on NIPD (dry days), or inadequate dialysis. If the patient is malnourished and not doing well, a higher than minimal Kt/V may be needed.

Changes in the peritoneal membrane — Finally, the clinician must address whether a change over time from an average to rapid transport is associated with other significant underlying peritoneal membrane pathology. As noted above, only minimal histologic changes are present in most patients, although peritoneal fibrosis or early sclerosing peritonitis can occur. These patients may have additional symptoms of abdominal pain, diarrhea, nausea, and vomiting. The diagnosis is usually based upon clinical parameters and indirectly via radiologic procedures. It can be confirmed by pathologic findings at exploratory laparotomy, if indicated. Affected patients may benefit from "resting" the peritoneal membrane by a temporary transfer to hemodialysis. Healing and return of baseline peritoneal membrane transport characteristics can usually be achieved by resting the membrane for six months or more.