Peritoneal transport testing

Transcription

1 THOROUGH CRITICAL APPRAISAL Peritoneal transport testing Vincenzo La Milia Nephrology and Dialysis Department, A. Manzoni Hospital, Lecco - Italy Ab s t r a c t Different tests can be used to provide valuable information about the function of the peritoneal membrane. The data obtained can be useful for tailoring dialysis adequacy, for the analysis of clinical problems such as ultrafiltration failure or to predict the development of more severe peritoneal membrane derangement. The most widely used peritoneal function test is the peritoneal equilibration test (PET), developed and described by Twardowski in PET is performed using a 2.27% glucose solution, and it lasts 4 hours. It measures peritoneal solute transport and ultrafiltration (UF), and it gives the possibility to categorize patients (high, high-average, low-average and low transporters). However, a PET with 3.86% glucose provides better information on UF, on the phenomenon of sodium sieving and an assessment of free water transport. Two recently developed tests (Mini- PET, Double Mini-PET) are promising tools to assess free water transport and the osmotic conductance to glucose. It is possible to integrate the 3.86%-PET with the new tests to obtain a complete PET: the Uni-PET (all these tests in 1 PET). The new insights into peritoneal function need a new standardization of the PET. It would be worth having a machine (PET machine) which performs the PET automatically to avoid possible mistakes during the performance of the manual PET and to allow an universal standardization of the test. Key words: Peritoneal equilibration test, Peritoneal transport, Ultrafiltration failure In t r o d u c t i o n The peritoneal membrane is a living tissue with variable functions influenced by several factors (endogenous and exogenous). The monitoring of the anatomic and functional characteristics of the peritoneal membrane over time is very important for tailoring dialysis adequacy, for the analysis of clinical problems such as ultrafiltration (UF) failure (UFF) (1) and to predict the development of severe peritoneal membrane damage. The analysis of the anatomic changes of peritoneal tissue can be performed only during the placement or removal of a peritoneal catheter or during other surgical procedures. Other methods can be used to provide valuable information about peritoneal structure by indirect analysis of the peritoneal membrane (2). Several tests can be used to evaluate the function of the peritoneal membrane and in particular the transport of solutes and the transport of fluids (peritoneal UF) through it. Why do we need peritoneal tests? We need peritoneal tests because the peritoneal transport is very complex: several transports occur simultaneously. Mathematical models and experimental studies on animals are used to understand the role of the individual transport components of this complex system. Several mathematical models have tried to explain the peritoneal transport. The simplest and most popular model is the membrane model, used by Pyle et al (3) and then by other authors. Another model is the pore model (4, 5), based on the concept of transport through a cylindrical pore across the membrane. Based on this model, the peritoneal barrier is an heteroporous structure with 3 types of pores: large pores (LP) with a radius of 250 Å, small pores (SP) with a radius of Å and ultrasmall pores (USP), or aquaporin-1, with a radius of 3-5 Å, which are permeable only to water. The most sophisticated model of peritoneal transport is the distributed model (6, 7) where the blood capillaries are placed within the tissue at different distances from the peritoneal surface. These models have been designed to investigate the fundamental properties of the peritoneal barrier, and because of their complexity, they are not useful in everyday clinical practice. For the clinical evaluation of the peritoneal transport, we need simple functional tests Società Italiana di Nefrologia - ISSN

2 La Milia: PET Fig. 1 - Traditional curves of peritoneal equilibration and classes of transporters assessed as the ratio between the glucose concentrations at t time and at the beginning of the PET (D t / D 0 ) and as the ratio between the dialysate creatinine concentration and plasma creatinine concentration (D/P Creat ) during the PET. The traditional values (11, 12) of D/P Creat that are used for the classification of patients on PD are reported. The peritoneal equilibration test (PET) More than 20 years ago, Twardowski proposed the peritoneal equilibration test (PET) (8) to evaluate the capacity of the peritoneal membrane to transport solutes and its ability to generate UF in peritoneal dialysis (PD) patients. Due to its simplicity, PET has had in recent years a wide distribution throughout the world, and it is now the reference test for other more complex tests that are not so easy to apply in common clinical practice (9, 10). PET is a semiquantitative assessment of the transport capacity of the peritoneal membrane determined by the speed of equilibration of the concentrations of a solute between plasma and dialysis solution. The concentration ratio between dialysate and plasma (D/P) of a given solute, after a specified time, indicates the speed of equilibration between the concentrations. A high solute D/P means that the balance between dialysate and plasma is reached quickly, and therefore the peritoneal permeability for the solute is high. Instead of the glucose D/P, the ratio between the concentration of glucose in dialysate after a certain time (t) and the concentration of glucose in the dialysate at the beginning of the test (D t ) is used. The original Twardowski PET is a test lasting 4 hours, with a solution containing glucose at 2.27% concentration, which evaluates the D/P for some small solutes, especially creatinine (D/P Creat ), and the D t of glucose. From the analysis of D/P Creat and D t during the PET, it is possible to draw curves of peritoneal membrane permeability (Fig. 1), and, based on the value of D/P Creat (D t is less commonly used) at the end of the PET, patients may be classified into 4 categories: high transporters (H), high-average transporters (HA), low-average transporters (LA) and low transporters (L) (8, 13). Usefulness of PET The great merit of Twardowski and his PET is having first shown that there is considerable variability in the transport of small solutes through the peritoneal membrane. This variability is even greater when considering UF. This simple observation led to the first and fundamental conclusion that the behavior of the peritoneal membrane is not uniform but may differ substantially (8, 14). Patient variability in peritoneal permeability to small solutes is multifactorial, influenced by age, sex, race, genetic factors, clinical factors (diabetes, cardiovascular comorbidity, inflammation), body surface area, etc (15). The intrapatient variability of peritoneal permeability to small solutes over time can be influenced by prolonged exposure of the peritoneal membrane to PD solutions, high glucose concentrations, modality of dialysis, manual (continuous ambulatory peritoneal dialysis [CAPD]) or automated peritoneal (APD) dialysis (16), peritonitis, etc (11, 17). The results of PET can be used for the prescription of the most appropriate PD modality, to achieve the greatest possible depuration and UF in a given patient (18). PET is also an indispensable tool for monitoring over time the function of the peritoneal membrane. 634

3 Prescription of dialysis modalities From the beginning, PET has been used for the prescription of the most appropriate peritoneal dialysis modality in the individual patient, as reported elsewhere (18). Initially the aim was to prescribe the dialysis modality which could obtain the greatest possible depuration. The best depuration would theoretically be obtained in high transport patients, but this high peritoneal transport of small solutes was frequently associated with increased mortality and morbidity (12). The water and sodium retention in these patients, especially in CAPD due to the rapid dissipation of peritoneal osmotic gradient, is probably one of the causes of this increased mortality. PET, in its simplicity, offered the possibility of dealing with high transporter patients in the appropriate way. In such patients, the knowledge of the rapid absorption of glucose has led to the proscription of very long dwells and to the prescription of short dwells, which have found their natural application in APD (18). More recently, the solutions with icodextrin (19), which have the ability of generating an adequate UF also in high transporter patients, have been used in daytime long dwells with a considerable increase in UF and depuration. The characteristics of peritoneal transport suggest treating high transporters with APD in combination with a daytime long dwell with icodextrin solution. A recent meta-analysis (20) of some prospective observational studies has confirmed a worse prognosis (especially in terms of survival) in high transporters than in patients with lower peritoneal transport characteristics. Usefulness for patient follow-up The periodic performance of PET in PD patients allows the monitoring of the characteristics of peritoneal membrane transport and the evaluation of how the functionality can be influenced by the time on PD, the type of solution used, intercurrent diseases of the peritoneum (peritonitis) or by other factors (13-15, 21-26). Standardization of PET PET has a variability coefficient that is less than 10% for the transport of small solutes, but can increases to 25%-50% (25) when considering UF. For this reason the standardization of the PET is necessary. Standardized PET would be highly reproducible and would allow the comparison of the results of the same patient during the clinical follow-up (27, 28) or the comparison between different populations. The standardization of PET should address the following points: (i) duration of the exchange (usually overnight) before the PET, (ii) volume of infusion and infused solution, (iii) position of patient during the infusion and the drainage, (iv) duration of the infusion and drainage, (v) methods of sampling and storage of blood and dialysate samples, and (vi) laboratory methods. Duration of the exchange (usually overnight) before the PET In the original version of the PET, it was recommended that the night before the PET, a dwell lasting 8-12 hours should be performed (8, 29). It is essential that the patient just before the PET has not had an empty peritoneal cavity. Therefore, it is advisable that the patient arrive at the dialysis center with a full peritoneal cavity and that the dwell before the test (if APD) should not last less than minutes (30-33). Some dialysis centers prefer to perform a night dwell of 8-12 hours even if the patient is on APD (26). It has been shown that the D/P Creat values are greater and those of D t are lower when, in the dwell preceding the PET, icodextrin is used instead of a solution containing 1.36% or 2.27% glucose (34-37). In patients who chronically use icodextrin for the night dwell, it is necessary that the night dwell immediately preceding the PET is performed with a solution containing glucose (1.36% or 2.27%). Infusion volume and infused solution The classic PET is performed with 2,000 ml of a solution containing a 2.27% glucose concentration. When PET was performed with a volume of 1,500 ml, the results were not different from the PET performed with 2,000 ml (38). However, we have only small studies of the infusion volume, it is therefore recommended that testing be done with a volume of 2,000 ml. All of the bags with the PD solutions (other than for APD) have a nominal volume of 2,000 ml, but when that volume is measured, the result is almost always a volume greater than this quantity (33, 39, 40). Moreover, the maneuver of flush before fill is performed with the aim of washing the line after the drainage of the previous dwell, with an unknown quantity of fresh solution that ends up in the drainage bag. The lack of quantification of the volume used for the flush before fill, which on average is 200 ml (41), can cause mistakes in the quantification of PET UF (39) and may also cause alterations in the classical parameters of PET, because the dilution of the solutes (urea, creatinine, 635

4 La Milia: PET etc.) contained in the drained dialysate lead to values of D/P for these solutes that are lower than the real values and values of D t that are higher. In both cases, there will be an underestimation of the patients belonging to the higher categories (H and HA). The overestimation of UF in PD due to the overfill of bags and to the flush before fill, could be very large when considering the 24 hours of UF in CAPD (42) and appears to be less important in APD (43). This important overestimation of UF in PD patients (especially in CAPD) could have potentially disastrous clinical consequences. It is possible that this mistake in the calculation of UF may have contributed to the water and sodium retention typical of many PD patients (42). The solution used in the classic PET contains 2.27% glucose. However, in recent years there has been considerable interest in the state of hydration of patients in PD and on peritoneal UF. This is due to the demonstration that a 20%- 30% increase in the total dialysis dose of an adequate PD (Kt/V=1.9, creatinine clearance = 60 l/week per 1.73 m 2 ) is not associated with improved survival (44) and that the removal of fluid (and sodium) is very important for the survival of patients. This was demonstrated in patients who due to their transport characteristics have low peritoneal UF, as high transporters (H) (20), or have reduced total water excretion (45), as anuric patients (46, 47). To better study the peritoneal UF, it has been proposed to replace the classic 2.27%-PET with a PET using a 3.86% solution (33). In fact the peritoneal UF is more easily quantifiable with the 3.86%-PET, because the UF achieved with the 3.86% solution is greater. So the 3.86%-PET allows a better identification of patients with UFF. A patient is defined as having UFF when after a 2.27%-PET lasting 4 hours, the UF is <100 ml, or after a 3.86%-PET lasting 4 hours the UF is <400 ml, using a 2-L bag (33). It is therefore evident that the 2.27%-PET is more prone to errors of proper evaluation of UF. Indeed, the coefficient of variation of UF is about 50% with the 2.27%-PET (25), while is only 7.8% with 3.86%-PET (26). In addition, the 3.86%-PET allows the assessment of the Na sieving coefficient during the first part of the test, expressed by the D/P of Na at 60 minutes (33). According to the three-pores theory (5), the D/P of Na at 60 minutes is the indirect expression of the free water transport of the peritoneal membrane. The free water transport is, in turn, the expression of glucose osmotic conductance of the peritoneal membrane. A high osmotic conductance to glucose (and hence a proper transport of free water and a low value of D/P of Na at 60 minutes) indicates that the peritoneal membrane function is normal. For all these reasons, it is preferable to use the 3.86%-PET. The use of a 3.86% glucose solution could result in loss of comparability with previous data obtained using the classic 2.27%-PET. Actually, the studies that have compared the 3.86%-PET with 1.36%-PET (48, 49) and 2.27%-PET (50, 51) have shown that, using the D/ P Creat, there are no differences in the classification of patients using the various types of solution for the PET. In conclusion, the 3.86%-PET is preferable, compared with the classic 2.27%-PET, for greater accuracy in determining peritoneal UF and for the opportunity to assess, albeit indirectly, the peritoneal membrane transport of free water. Finally, because of the lower coefficient of variation, the 3.86%-PET is a more reproducible test for studying the peritoneal UF in prospective studies. Position of patient during infusion and drainage In the standardization of classic PET (8), the patient drains the fluid of the night in the orthostatic position to allow the largest possible drainage; in the orthostatic position, peritoneal fluid tends to collect in the bottom of the pelvic cavity, where the extreme end of the peritoneal catheter should be positioned to obtain the best conditions for drainage. During the infusion of the solution used for PET, the patient should be supine and rotate from one side to another after every 400 ml of infusion (2 minutes each), to better mix the residual volume infused with the solution. There are no studies in the scientific literature on any need to rotate the patient on their sides to facilitate the mixing of the infused solution with the residual volume. However, it is advisable to carry out this maneuver, at least before the first drainage of dialysate. In the standardization of the classic PET, to take a dialysate sample, 200 ml of dialysate is drained into the drainage bag, of this, 10 ml is collected, in a sterile way (for the lab), and the remaining 190 ml is reinfused into the peritoneal cavity. After taking the dialysate sample, the patient is allowed to get up and walk freely; to do this, it is necessary to disconnect and reconnect many times the patient from the infusion-drainage system. To avoid problems, it is preferable that the patient remains connected to the infusion-drainage system and, therefore, remains in a lying or sitting position for the duration of the test. Of course, this requires the presence of the patient for all of the PET duration in the place where the test is performed (usually a local office of the reference dialysis center or, more rarely, at home) but avoids the possible interference of intra-abdominal pressure due to an orthostatic position and walking, on the mechanisms of peritoneal transport and peritoneal UF (52). 636

5 Duration of infusion and drainage The peritoneal cavity must be completely emptied before the PET, the drainage should be done in the upright position, lasting at least 20 minutes. The solution used for PET should be infused as quickly as possible, usually no more than 10 minutes. Indeed, the zero time of the test coincides with the end of the solution infusion. However, also during the infusion of the first milliliters of solution, the transport of water and solutes through the peritoneal membrane between the blood and the solution is starting. A longer infusion time would lead to the extension of the actual total time of the PET with the possibility that solute D/P ratios are higher and D t ratios are lower than the actual values. For the same reason, the time of drainage at the end of the test should be as quick as possible and, always to standardize PET, should not exceed 20 minutes. Timing and methods of sampling and storage of blood and dialysate samples In the standardization of the classic PET, dialysate drawings are made at time 0 (immediately after the end of solution infusion), at 120 minutes from the start of the PET and at the end of the test, after complete drainage of the peritoneal cavity. In all the cases, the volume of the sample is 10 ml and should be taken into account when calculating the peritoneal UF. In the classic PET, at time 0 and 120 minutes, a drainage of 200 ml of dialysate is performed, the sample (10 ml) is taken and the remaining quantity (190 ml) is reinfused into the peritoneal cavity. This maneuver involves a potential risk for peritonitis and should be avoided. A more secure mode may be to perform first the reinfusion, leaving an aliquot of dialysate for the sample in the bag. In the new PETs it is preferable to perform the drainage only after 60 minutes (for the quantification of free water transport) and at the end of the test. The final sample of dialysate (time 240 minutes) must be taken after complete draining of the peritoneal cavity. In the classic PET (8), the concentration of glucose in the dialysate taken at time 0, and not the concentration of glucose in the fresh solution before being infused, is used to calculate the D t. This is to avoid the influence of residual volume in the peritoneal cavity after the complete drainage of the previous exchange; nevertheless, the exchanges between blood and dialysate through the peritoneal membrane start during the first moments of infusion, and certainly a part of the residual volume measured with the classical methodology (8) is due to the transport of solutes. For this reason, it would be better to use the concentration of glucose measured in the fresh solution. We suggest leaving an aliquot of fresh peritoneal solution in the bag at the end of the infusion and then taking the sample of fresh solution. In the classical test, 2 blood samples were taken: the first at the end of the drainage for the night, immediately before instilling the solution chosen for the PET, and the second at the end of PET immediately after the drainage and the average was used for calculation of D/P (11). The test was later simplified by taking a single blood sample in mid-pet (at 120 minutes of the test) (13). In the new PETs, 2 blood samples are taken: the first at 60 minutes into the test, to evaluate the Na sieving, and the second at the end of the PET for the calculations of D/P and D t. Laboratory methods It is well-known that high concentrations of glucose interfere with certain methods of measuring the concentrations of creatinine (53), and for that reason, it becomes necessary to use a correction factor, to be calculated for each individual laboratory. This correction factor is derived from the relationship between the concentrations of creatinine, determined with the non-enzymatic method, and increasing concentrations of glucose up to the concentrations found in solutions for PD (8). The plasma concentration of creatinine (and other solutes) should be corrected for plasma water (54) before calculating the D/P, thus to avoid obtaining values of D/P of some solutes (such as urea in certain situations) higher than a unit, which is a kinetic paradox. To calculate the D/P of Na (or the reduction of dialysate sodium concentration, of Na) at 60 minutes of a PET with a 3.86% glucose solution, the use of direct ion selective electrodes (ISEs) should be avoided, while indirect ISE (which is used in most laboratories) can be used. Indirect ISE gives results comparable to that of flame photometry which is the best method for the determination of sodium in the infusion liquid and in dialysate (55). Timing of PET Both retrospective (56) and prospective (57) studies have shown that the characteristics of small solutes peritoneal transport changed significantly during the first month of peritoneal dialysis treatment, and remained stable thereafter. For this reason, some guidelines (58) suggest per- 637

6 La Milia: PET forming the first (baseline) PET after 4-8 weeks of PD. However, performing the PET earlier could uncover useful information regarding changes of peritoneal transport of small solutes during different PD modalities such as CAPD and APD (16). In the case of peritonitis, PET should be performed at least 1 month after peritonitis resolution because the inflammation causes an increase in peritoneal transport of small solutes and a marked reduction of UF (59). More controversial is the issue regarding repetition over time of PETs in the same patient. Indeed some guidelines (58), given the substantial stability of the peritoneal transport over time in most patients, recommend not repeating the PET in any scheduled manner but repeating it only if problems (such as hydro-saline retention or underdialysis) arise. Other guidelines (33) recommend performing a PET at least once a year and every time there is a clinical reason for it. Obviously, the repetition of a PET (especially the modified PET using a 3.86% glucose solution) at least once a year, allows us to anticipate, at least in some cases, the onset of clinical problems. In conclusion: (i) the first PET should be performed 4-8 weeks after the start of PD, (ii) the PET should be performed at least 1 month after any episode of peritonitis, and (iii) a PET should be performed at least once a year and as many times as there are clinical problems related to peritoneal transport. Classification o f p a t i e n t s a c c o r d i n g t o data from PET The data from a PET should be interpreted and applied clinically. The first problem is the comparability of data obtained with a PET, for classification of patients. For many years, PET data were compared with those of the original results of Twardowski et al (8, 13). It is important to underline the small number of tests and patients studied (103 PETs performed in 86 patients), the extremely variable time on PD of patients ( months), the geographical area of patients, etc. For all of these reasons, clinicians should not use the original data of Twardowski to classify their patients; however, such data are extremely useful to make comparisons between different populations. It is therefore preferable to classify patients according to the results of PETs in their own dialysis center using the mean and SD of the patients to obtain the classification. Of course this can be a problem when PD patients are very few in a small dialysis center, and in this case it is preferable to compare the data with those in the literature. Some studies have attempted to establish reference values for the parameters obtained with PET with the 3.86% glucose solution (60). In Table I the original results of Twardowski et al (8, 13) are compared with those drawn from different populations (14, 25, 26, 60-62). Despite the different populations, the values TABLE I D/P CREAT AND CLASSES OF TRANSPORTERS IN VARIOUS PATIENT POPULATIONS ON PERITONEAL DIALYSIS Twardowski (8, 13) TARGET (USA) (61) Mexico (62) UK (25) ANZA-DATA (14) The Netherlands (60) Italy (26) H >0.81 >0.79 >0.80 >0.78 >0.81 >0.82 >0.80 H-A Average SD L-A L <0.50 <0.55 <0.56 <0.52 <0.57 <0.62 <0.62 Number of patients D/PCreat = dialysate to plasma creatinine ratio; H = high transporter; H-A = high-average transporter; L = low transporter; L-A = low-average transporter; SD = standard deviation. 638

7 of D/P Creat that could have a clinical interest are those close to or above 0.80 and those close to 0.60 or lower. Of course, it is best to consider the values of peritoneal permeability (expressed by D/P Creat ) as a continuous entity and not as belonging to totally separate categories, as is done by classifying patients into classes of transporters, but these limits well represent the categories of patients at risk, because patients with values of D/P Creat at or above 0.80 are exposed to all of the risks associated with high or fast transport (low UF, tendency to sodium and water retention), while patients with values at or below 0.60 may be at risk of underdialysis in case of inappropriate dialysis prescription. Table I can be a useful tool to compare and classify patients in cases where there are of limited numbers of them. Finally, several authors (63) use the terms fast and slow transporters instead of high and low transporters because the PET measures the velocity of the equilibration of small solutes between plasma water and dialysate. These terms can be used interchangeably. Pre s c r i p t i v e an d di a g n o s t i c utilities o f modified PET w i t h 3.86% g l u c o s e s o l u t i o n The usefulness of the data obtained with a PET for determining the best dialysis mode requirement has already been shown (18). Briefly, in high or fast transporters, long dwells with glucose solutions, especially with low osmolarity (as in 1.36% solutions) must be avoided, while APD with a long daytime dwell with icodextrin is the best option. In contrast, in low or slow transporters, short dwells should be avoided because of the risk of underdialysis, while CAPD is the most suitable choice. In cases of inadequate treatment of these patients with CAPD, the volumes can be increased; otherwise, these patients should be moved to hemodialysis. The use of a modified PET with a 3.86% glucose solution is necessary for the diagnosis of loss of UF by the peritoneal membrane. The guidelines of the International Society of Peritoneal Dialysis (33) have accurately described what to do in cases of patients with hydro-saline overload and how to make a diagnosis of UFF. In summary, after excluding mechanical problems associated with the catheter, which are easily demonstrated with an abdomen X-ray, it is necessary to perform a 3.86%-PET lasting 4 hours. A peritoneal UF less than 400 ml at the end of the test is compatible with the diagnosis of UFF. According to the characteristics of peritoneal transport for small solutes, strategies have been suggested for the treatment of PD patients with UFF. However, a modified PET with a 3.86% glucose solution does not enable us to discover exactly what the mechanism is that has generated the UFF, because with this test it is not possible to discern if the patient had a peritoneal UF in the first part of the exchange with the hypertonic solution and if therefore the UF was reabsorbed for rapid dissipation of the osmotic gradient (in this case, the indication is for APD as likely to be effective), or if the patient produced little or no UF. In the latter case, there is low or no osmotic conductance to glucose, and the APD prescription would have no effect, in which case, icodextrin can be used, otherwise the patient should be moved to hemodialysis treatment. The analysis of D/P Na (or the reduction of sodium concentration in dialysate, Na) at 60 minutes indicates only, in a semiquantitative way, the functionality of the aquaporin-1 channels, but does not allow us to quantify the transport of free water through these pores or the osmotic conductance to glucose. In conclusion, a modified PET with a 3.86% glucose solution allows the classification of patients according to their capacity for peritoneal membrane transport of small solutes. Furthermore, this test allows us to make a diagnosis of UFF and provides a semiquantitative assessment of free water transport. In the case of patients with UFF, it is necessary to perform a more complete test for an accurate diagnosis of the genesis of the UFF and for the most appropriate therapeutic strategy, as explained in the next section. Qu a n t i f i c a t i o n o f f r e e w a t e r t r a n s p o r t a n d o s m o t i c c o n d u c t a n c e t o g l u c o s e of peritoneal membrane The PET has provided a huge contribution to the knowledge of the pathophysiology of the peritoneal membrane with the advantage of being very simple in its execution and interpretation. However, the PET is a test whose results are the set of a series of physiopathological mechanisms that can not be captured only with the analysis of the D/P of some solutes or with the D t of glucose. Even the changes made to the classic tests such as the use of a 3.86% glucose solution allow us to avoid some mistakes in evaluating peritoneal UF but do not explain how it actually happens. The analysis of D/P Na was the first, simple attempt to measure (although in a semiquantitative way) the UF generated through the various pores of the peritoneal barrier. The need to measure the UF in its various components, and to understand its genesis, led to the development of other tests easily performable in common clinical practice and easy to interpret. The design of the Mini-PET (64) made it possible 639

8 La Milia: PET Fig. 2 - Simulation of the intraperitoneal volume during PET with a 3.86% solution in a normal subject (line A), in a subject in whom there is an early ultrafiltration and then a partial reabsorption of ultrafiltrate (line B) and in a subject in whom there is not any ultrafiltration (line C). to quantify the transport of free water through ultrasmall transcellular pores or aquaporin-1 channels, and to separate it from other components of peritoneal UF. A precise quantification of free water transport is possible by Mini- PET assuming that the sodium diffusion is negligible during a PET with a 3.86% glucose solution lasting only 1 hour; in this situation, the UF through small pores (UFSP) is the peritoneal clearance of Na (amount of peritoneal sodium removal divided by plasma water sodium concentration), and the free water transport is the difference between the total peritoneal UF and the UFSP (64). The Double Mini-PET (65) allows the assessment of the freewater transport and the so-called osmotic conductance to glucose of the peritoneal membrane (that is the ability of the peritoneal membrane to generate UF when exposed to the osmotic stimulus given by glucose with more or less hypertonic solutions). The Double Mini-PET consists of 2 consecutive PETs, each lasting 1 hour, the first carried out with a 1.36% glucose solution and the second with a 3.86% glucose solution. Based on complex assumptions (the osmotic transient principle ) (4), the Double Mini-PET is performed with very simple methods and requires no computer processing but only the aid of a simple pocket calculator. Otherwise, quantification of osmotic conductance to glucose is usually done by a peritoneal test with a volume marker and using a complex kinetic model (66). In cases of UFF, the Double Mini-PET can provide additional guidance to the prescription of the most appropriate peritoneal dialysis modality or of the need to move the patient to hemodialysis. For example, in Figure 2, the normal curve (A) shows the increase in intraperitoneal volume during an exchange with a 3.86% glucose solution, while both the curves B and C will result in a UFF at the end of a test of 4 hours. In the case of curve B, there is some UF during the first part of the exchange, and later this UF is canceled because of the reabsorption of fluid from the peritoneum. In this case, a prescription of APD is correct and can be reinforced by a prescription of icodextrin in the long dwell. In the case of curve C, the peritoneal membrane has a reduced or absent osmotic conductance to glucose, and the glucose is unable to generate adequate UF in any part of the exchange. In this case, APD would be unable to generate UF, and the only attempt that can be made is using icodextrin (both in patients on APD and on CAPD), with the advice that patients should be moved to hemodialysis if icodextrin is insufficient to obtain an appropriate UF (of course, this is valid if the patient has no residual renal function). The UF after 1 hour, obtained by single or Double Mini-PET with a 3.86% glucose solution, measures the amount of early peritoneal UF and identify patients who may benefit from short peritoneal dwells. The extent of the osmotic conductance to glucose measures the potential of the peritoneal membrane to produce UF when subjected to different osmotic forces resulting from different concentrations of glucose. In other words it indicates the amount of UF achievable when the concentration of glucose in the PD solutions is increased. Furthermore, the Double Mini-PET allows us to quantify free water transport and the functionality of the aquaporin-1 channels. In patients with reduced or no free water transport there is a marked reduction in peritoneal UF (approximately 40%-50%) due to the osmotic strength of glucose. In these patients, the increased concentration of glucose in PD bags will not result in an increase in UF (UF can only increase through SP if their osmotic conductivity is not compromised), and so it is preferable to prescribe icodextrin which acts with a mechanism different from glucose. With the systematic implementation of this test, it would be possible to evaluate the behavior of USP and their role in the loss of peritoneal UF capacity. Moreover, one could ascertain whether or not there are isolated deficits of pores (small and ultrasmall). The analysis of a few patients with UFF (65) using the Double Mini-PET showed that the reduction in the osmotic conductance of these patients depends on the functional 640

9 damage of both small and ultrasmall pores (or aquaporin-1 channels). However, we need more clinical studies to assess whether the Double Mini-PET could be a predictive tool for early damage of the peritoneal membrane, long before the development of UFF or other problems of peritoneal transport. Finally, dwell times for the Double Mini-PET are very similar to those used in APD, so this test enables an evaluation of how the transport of small solutes and the transport of liquids occur when using the solution with the lowest osmolality (1.36% glucose) and the solution with the highest osmolality (glucose 3.86%) in dwell times that are typical of APD. The Double Mini-PET still needs to be validated in a broader survey of PD patients. Using the Mini-PET and the Double Mini-PET, it is possible to classify PD patients based on their D/P Creat after only 1 hour of testing, with good agreement with the classification obtained after 4 hours (64, 65). However, the agreement is not complete, the interpretation of D/ P Creat can be difficult and the application of a correction factor to translate the results to those of 4 hours of PET, is not possible (67). These factors limit the application of Double Mini-PET for the overall assessment of peritoneal function (64). It is necessary, then, to perform 2 PETs, the 3.86%-PET lasting 4 hours and the Double Mini-PET, to obtain all of the parameters that are today measurable in clinical practice. The combined PET Some authors (68) have combined the 3.86%-PET and the Mini-PET into only 1 test. During a 3.86%-PET, the peritoneal cavity is drained after 60 minutes to make a determination by weighing the dialysate volume at that time. After taking a dialysate sample, the drained volume is reinfused and left for another 3 hours. With this combined PET, no differences were found in UF and small solute transport in comparison with those found with a 3.86%-PET without temporary drainage. The combined PET is a good method to avoid a double PET on separate days. However, this method does not allow us to obtain the osmotic conductance to glucose, while it is possible to obtain it combining the 3.86%-PET with the Double Mini-PET. The combined PET (3.86%-PET and Double Mini-PET) or Uni-PET, is the test that allows us with only 1 test lasting 5 hours to obtain all of the functional parameters that are today measurable in clinical practice. Moreover, performing a dwell lasting 8 hours with a 1.36% glucose solution during the night preceding the Uni-PET, it is possible to estimate if a patient has a high peritoneal reabsorption, avoiding the prescription of a long dwell with this solution. In Table II, a procedure is given for the implementation of a Uni-PET as rigorously as possible. Table III shows the laboratory exams to be carried out on samples taken during the tests. Table IV shows the mathematical formulae used to calculate the parameters during the test. Future perspectives It is desirable in the near future that the Uni-PET (3.86%-PET combined with the Double Mini-PET) be carried out through the aid of a device, similar to the existing equipment for APD (a PET machine). An automated PET would reduce the costs in terms of human resources and the risk of mistakes. Importantly, it would be a standardized method. The next step should be a centralized collection (possibly on-line from the PET machine) of the data derived from the automated PET, to create a register of the peritoneal transport characteristics in PD patients. This could lead to the construction of reference graphs that could be used by any nephrologist for the evaluation of patients and for the prescription of the best treatment for each PD patient. Other tests There are other tests that do not require the use of peritoneal markers and that have the aim of evaluating the characteristics of peritoneal membrane transport. These tests are the personal dialysis capacity (PDC) test (69), the dialysis adequacy and transport test (DATT) (70) and the accelerated peritoneal examination (APEX) test (71). The PDC test lasts 24 hours and is performed by the patients at home with 5 exchanges, with different dwell times and glucose concentrations (69). The test uses a computerized mathematical model based on the 3-pore model (5) to estimate parameters such as (i) the surface area over diffusion distance (A 0 /Δ X ), which represents the effective surface area available for diffusion; (ii) the reabsorption of fluid from the peritoneal cavity; and (iii) the large pore flow (J v L). The superiority of the PDC test over the classical PET has been claimed (72, 73). However, the PDC test has a number of disadvantages such as the risk of inaccuracies (the test is performed by the patient at home) due to overfill of bags, the maneuver of flush before fill, the collection of aliquots by the patients or the transport of the bags to the center. Furthermore the PDC test requires the use of a complex computerized mathematical model, and it does not take into consideration the sodium kinetics, the free water transport and the osmotic conductance to glucose. Finally, the PDC test has been performed in a limited number of patients in comparison with the PET. 641

10 La Milia: PET TABLE II STANDARDIZATION OF THE UNI-PET: STEPS 1. Patient should have a full abdominal cavity before the beginning of the test (for continuous ambulatory peritoneal dialysis [CAPD], the overnight dwell of 8-10 hours should be performed; for automated peritoneal dialysis [APD], if an overnight dwell is not possible, the nightly program should end with infusion of dialysate that should be left in place for 1 hour at least. For other peritoneal dialysis programs, the overnight dwell should be performed as for CAPD patients). 2. Icodextrin should not be used for the overnight dwell before the PET. A 1.36% (the best option) or 2.27% glucose solution should be used also in CAPD patients who usually use icodextrin for the overnight dwell. Icodextrin can be used for the daily long dwell the day before the test in APD patients. 3. Complete drainage (lasting 20 minutes at least) of the dialysate of the overnight dwell in a bag and measurement of the volume of overnight dwell. 4. A 1.36% glucose solution should be used for the first part of the test i. Weighing of the bags and of the lines before the beginning of the test. The weighing should be repeated at the end of the test, to assess the infused volume. ii. Connection of the bag to the patient and infusion of fresh solution (10 minutes) leaving approximately 10 ml of it in the bag. At the end of the infusion, this sample should be taken for laboratory analysis. iii. Exchange lasting 60 minutes after the end of the infusion. iv. Complete drainage (20 minutes) of the abdominal cavity. Measurement of the dialysate volume; 10 ml of dialysate should be taken as a sample for laboratory analysis. 5. Blood samples. 6. A 3.86% glucose solution should be used for the second part of the test i. Weighing of the bags and of the lines before the beginning of the test. The weighing should be repeated at the end of the test, to assess the infused volume. 7. Connection of the bag to the patient and infusion of fresh solution (10 minutes) leaving approximately 10 ml of it in the bag. At the end of the infusion, this sample should be taken for laboratory analysis. i. Exchange lasting 60 minutes after the end of the infusion. ii. Complete drainage (20 minutes) of the abdominal cavity. Measurement of the dialysate volume. Reinfusion of the dialysate (10 minutes) leaving approximately 10 ml of it in the bag. At the end of the infusion, this dialysate should be taken as a sample for laboratory analysis. iii. Exchange lasting 3 hours after the end of the reinfusion. iv. Complete drainage (20 minutes) of the abdominal cavity. Measurement of the dialysate volume; 10 ml of dialysate should be taken as a sample for laboratory analysis. 8. Blood samples. 9. Disconnect the patient or begin the usual dialysis program. TABLE III UNI-PET: LABORATORY ANALYSES 1. Blood sample: plasma glucose, plasma urea, plasma creatinine, plasma sodium and plasma total proteins. 2. Sample of fresh solution: glucose and sodium. 3. Sample of dialysate after the drainage: creatinine, glucose and sodium. An enzymatic method should be used to assess the dialysate creatinine concentration. If the Jaffé method (53) is used, the values should be corrected. Flame photometry or indirect ion selective electrodes (ISE) should be used to assess the sodium concentration in blood and dialysate. The direct ISE should not be used (55). 642

11 TABLE IV UNI-PET: FORMULAE 1. D/P Creat : ratio between the dialysate creatinine concentration (mg/dl) at the end of the test (the correction factor should be used if necessary) and the creatinine concentration in plasma water (creatinine PW ) (mg/dl) (54): Creatinine PW = u creatinine P, where creatinine P is expressed in mg/dl. u = [1/(1-V lip TotProteins P )] V lip = fractional volume of plasma lipids = TotProteins P = plasma total proteins concentration in g/dl 2. D t : ratio between the dialysate glucose concentration (mg/dl) at the end of the test and the fresh solution glucose concentration (mg/dl). 3. Na sieving: it is advisable to consider the decrease in dialysate sodium concentration (ΔD Na ) after 60 minutes of the test with the 3.86% solution instead of the D/PNa after 60 minutes (34). The ΔDNa is the difference in Na concentration (mmol/l) between the fresh solution and the dialysate drained after 60 minutes. 4. Free water transport (FWT) is assessed during the second part of the test (3.86% glucose solution); FWT is the total UF (UFT) in ml of this test, minus the small pores UF in ml (UFSP) (64): FWT = UFT UFSP UFSP is assessed using the Na clearance: UFSP = [NaR 1,000]/Na p, where Na p is expressed in mmol/l NaR (mmol) is the Na removed during the second part of the test with the 3.86% solution. NaR is calculated as follows: NaR = [drained dialysate volume Na concentration in the drained dialysate] - [volume of the dialysate before infusion Na concentration in the dialysate before infusion], where dialysate volume is expressed in liters, and Na concentration in mmol/l. Na P = plasma sodium. 5. Osmotic glucose conductance (OCG) in ml/min per mm Hg (64): OCG = {(V 3.86 V 1.36 )/ [19.3 (G 3.86 G 1.36 ) t]} 1.7 V 3.86 and V 1.36 (ml) are the dialysate volumes drained at the end of the first hour of the first (3.86% glucose solution) and the second part (1.36% glucose solution) of the test; 19.3 (mm Hg/mmol/L) is the product of the absolute temperature and the constant of gases at 37 C; G 3.86 and G 1.36 are the molar glucose concentrations (mmol/l) in the dialysate before the infusion. G 3.86 and G 1.36 are calculated as follows: G = glucose /18, where glucose is expressed in mg/dl; t is the mean duration time of the exchanges with the 1.36% solution and the 3.86% solution until the temporary drainage. Each time should be the duration of the exchange (60 minutes) + 50% of the duration of the infusion and of the drainage (usually t = = 75 minutes); 1.7 is a factor of correction (65) The transport measures obtained from the PET can be applied to a 24-hour period to obtain D/P creat and D t in the DATT. DATT was used in the ADEMEX trial (44, 74), and several studies have demonstrated good correlations between DATT and PET, but it has only been validated in CAPD patients and not in APD patients. The APEX test (71) summarizes in a single number the peritoneal permeability both to glucose and urea: hence, it represents the time at which glucose and urea equilibration curves (using percentages as units) cross; the shorter the APEX time, the higher the peritoneal permeability, and, conversely, the longer the APEX time, the lower the peritoneal permeability. The APEX test has a limited use in the dialysis community and is more frequently used in children (75). 643

12 La Milia: PET TABLE V THE AVAILABLE TESTS FOR THE EVALUATION OF PERITONEAL MEMBRANE FUNCTION Test Test procedure details Measure of small solute clearance Additional measurements obtained (in comparison with standard 2.27%-PET) 2.27%-PET 4-hour test exchange D/PCreat; Dt/D0 glucose 2.27% glucose solution 3.86%-PET 4-hour test exchange D/PCreat; Dt/D0 glucose Maximal UF capacity after 4 hours 3.86% glucose solution Measures of Na sieving: D/PNa(60) D/PNa(60) D/PNa(0) D/PNa(60) / D/PNa(0) ΔDNa DATT 3 exchanges 1.5% and 1 overnight D/PCreat; Dt/D0 glucose (over 24- Dialysis adequacy measurements exchange 2.5%: 24-hour patient s hour collection of dialysate) typical dwell volume used PDC test 5 exchanges lasting 24 hours A0/Δx JvL Varying glucose concentrations JvAR 24-hour urine collection Dialysis adequacy measurements 2 blood samples performed at home Mini-PET 1-hour test exchange 3.86% glucose solution D/PCreat; Dt/D0 glucose after 1 hour Double Mini-PET Two 1-hour test exchanges D/PCreat; Dt/D0 glucose after 1 (performed consecutively) hour 1st exchange 1.36% 2nd exchange 3.86% Combined 3.86%-PET 4-hour test exchange 3.86% glucose solution Temporary drainage of dialysate after 1 hour to assess the volume by weighing and taking a dialysate sample. Then reinfusion of dialysate (left in place for another 3 hours) D/PCreat; Dt/D0 glucose Maximal UF capacity after 1 hour Measures of Na sieving (all) FWT UFSP Minimal and maximal UF capacity after 1 hour Measures of Na sieving (all) FWT UFSP OCG Maximal UF capacity after 1 and 4 hours Measures of Na sieving (all) FWT UFSP Uni-PET Two test exchanges D/PCreat; Dt/D0 glucose Maximal UF capacity after 1 and 4 hours (performed consecutively) Measures of Na sieving (all) 1st exchange 1.36% lasting 1 hour FWT 2nd exchange 3.86% lasting 4 hours: UFSP Temporary drainage of dialysate after 1 hour to OCG assess the volume by weighing and taking a dialysate sample. Then reinfusion of dialysate (left in place for another 3 hours) A0/Δx = area parameter; DATT = dialysis adequacy and transport test; D/PCreat = dialysate to plasma creatinine ratio; D/PNa(60) = ratio of dialysate to plasma sodium at 60 minutes; D/PNa(60) - D/PNa(0) = difference between dialysate to plasma sodium ratio at 60 minutes and initial dialysate to plasma sodium ratio; D/PNa(60) / D/PNa(0) = ratio of dialysate to plasma sodium ratio at 60 minutes and initial dialysate to plasma sodium ratio; Dt/D0 = dialysate glucose ratio; FWT = free water transport (aquaporin-1 transport); JvAR= absorption parameter; JvL = large pore fluid flux; OCG = osmotic conductance to glucose; PET = peritoneal equilibration test; PDC = personal dialysis capacity test; UF = ultrafiltration; UFSP = ultrafiltration through the small pores. 644