Abstract
Objective
The aim of this study was to evaluate the effects of a combined 25 mmol/L bicarbonate/15 mmol/L lactate-based solution (Bic/Lac), compared to a 35 mmol/L lactate solution (Lac) — the most commonly used solution for patients in southern Europe — on the venous plasma bicarbonate level in patients treated with continuous ambulatory peritoneal dialysis (CAPD).
Design
This was a randomized, parallel, controlled, open-label study, with patients studied for a period of 3 months preceded by a 1-month baseline and followed by a 1-month follow-up. Patients used the 35 mmol/L lactate solution during baseline and follow-up periods.
Setting
Four Spanish nephrology centers.
Patients
Thirty-one (20 Bic/Lac, 11 Lac) well-dialyzed (creatinine clearance > 55 L/week/1.73 m2 body surface area) CAPD patients.
Interventions
Blood samples were taken for biochemistry tests at all visits. A physical examination was completed at baseline and month 3, and a medical update was completed after 1, 2, and 3 months, and at the follow-up visit. Adverse-event monitoring and notation of prescription changes were carried out continuously.
Main Outcome Measure
Effect on venous plasma bicarbonate level.
Results
Venous plasma bicarbonate rose by 3.1 mmol/L (confidence intervals 1.6 – 4.8), from a baseline level of 23.0 mmol/L during the treatment period in those patients treated with Bic/Lac (p < 0.05 vs Lac). The number of acidotic patients (venous plasma bicarbonate < 24 mmol/L) was statistically significantly reduced at every treatment period visit in the Bic/Lac group (p < 0.05). There were no adverse findings with respect to vital signs, physical examination, or clinical symptoms, apart from one death in the control group.
Conclusions
The new Bic/Lac solution allowed better correction of acid–base status than the lactate solution.
For PD, a bicarbonate-based fluid is highly desirable, as it would provide a natural body buffer. Today, this has become possible and several alternatives are available, including bicarbonate-based and bicarbonate/lactate-based solutions (15-17). Several studies indicate that the bicarbonate/lactate-based solutions are the better option (10, 18,19).
In Europe, most patients use predominantly a solution with 40 mmol/L lactate, and this solution was used as the control in previous clinical trials. In southern Europe, however, the majority of patients use solutions with 35 mmol/L lactate. The aim of this study was to evaluate the effects of a combined 25 mmol/L bicarbonate/15 mmol/L lactate-based solution (Physioneal; Baxter SpA, Sesto Fiorentino, Italy), compared with Dianeal PD1 (with 35 mmol/L lactate, Dianeal PD1; Baxter Healthcare, Castlebar, Ireland), on venous plasma bicarbonate levels in patients treated with continuous ambulatory peritoneal dialysis (CAPD). Many patients on lower levels of buffer tend to remain acidotic, and the purpose of this study was, therefore, to investigate if acidosis could be better corrected with this new solution without causing significant alkalosis.
Subjects and Methods
Study Design
This was a randomized, prospective, controlled, open-label multicenter study comparing a new 25 mmol/L bicarbonate/15 mmol/L lactate PD (test) solution with a standard 35 mmol/L lactate-buffered PD (control) solution over a 3-month treatment period. The treatment period was preceded by a 1-month baseline period, and followed by a 1-month follow-up period during which all patients used the control solution.
Study Solutions
The test solution in this trial was presented in a two-chambered bag and contained 25 mmol/L bicarbonate and 15 mmol/L lactate (total buffer, 40 mmol/L; Physioneal). The control solution contained 35 mmol/L lactate (Dianeal PD1). The composition of the two solutions is given in Table 1. The two solutions also differed with respect to their calcium, magnesium, and chloride content. Both products were steam sterilized according to normal commercial procedures and presented in a twin-bag configuration, where the drain and new solution bags are part of the same set.
Composition of the Test Peritoneal Dialysis Solutions
Bic/Lac = bicarbonate/lactate.
Dianeal PD1 (Baxter Healthcare, Castlebar, Ireland).
Physioneal (Baxter SpA, Sesto Fiorentino, Italy).
Study Population
This study took place in four dialysis centers in Spain. A total of 33 consecutive eligible patients were recruited. Each patient gave written informed consent prior to participation, and approval for the study was given by the local ethics committee for each center. This trial conformed to the Declaration of Helsinki and the Good Clinical Practice Guidelines for Clinical Trials in the European Union (20).
Prior to entry into the study, patients had been on CAPD for at least 2 months and were being treated with a 35 mmol/L lactate dialysis solution (Dianeal PD1), using an integrated disconnect system (twin-bag), for at least 1 month. They had a normalized (to body surface area) glomerular filtration rate of less than or equal to 7 mL/minute/1.73 m2 (average of urea and creatinine clearance), and used three or four 2- to 2.5-L exchanges per day, 7 days per week, with no dry period. Their total weekly creatinine clearance (peritoneal and renal) was greater than or equal to 55 L/1.73 m2 body surface area, as determined by the PD Adequest program (version 1.4a; Baxter Healthcare). The creatinine clearance target was used rather than Kt/V urea because it was the most commonly used method in the participating centers. Patients were excluded if they had completed a course of antibiotics for peritonitis during the previous 30 days, had other serious illnesses including the need for hospitalization during the previous 30 days, were known to be HIV-positive, were pregnant or lactating, or were adding bicarbonate to the bags or taking bicarbonate orally and were judged not able to discontinue bicarbonate use for the duration of the study. The latter group was excluded because, if such use were to continue during the study, it would affect venous plasma bicarbonate levels.
Study Procedures
After giving informed consent, all patients were treated for a further month (baseline period) with the lactate-based dialysis solution and were then randomized, in a ratio of 2:1, to the test solution — the bicarbonate/lactate mix (to increase the experience on the test solution) — or to continue on the lactate-based control solution, respectively. Each center was allocated a discrete randomization block (3 patients/block; the investigators were not aware of the block size) to ensure that appropriate numbers of patients were assigned to the different treatment groups in each center. Patients continued using the test or control dialysis solution during the 3-month study period and were required not to take oral or intraperitoneal sodium bicarbonate during the study period. Dialysis prescriptions were to be kept constant.
A medical history was taken before entry and a medical update, recording all changes in medication, symptom profile, and dialysis prescription, was completed every month. Blood samples were taken after overnight oral fasting, at monthly intervals, and analyzed at the local hospitals for sodium, potassium, total calcium, phosphate, urea, creatinine, glucose, albumin, and C-reactive protein (CRP). Plasma bicarbonate was measured in venous blood samples taken anaerobically. These samples were generally analyzed within 15 minutes in a blood gas analyzer for actual bicarbonate, pCO2, total CO2, and pH. The normal range used to establish acid–base status was 24 – 32 mmol/L bicarbonate (21). Venous rather than arterial samples were taken since it was regarded as unethical to expose patients to arterial punctures as part of a clinical trial when properly taken venous samples can give adequate and reproducible results (22).
Statistical Analysis
The primary efficacy variable, the venous plasma bicarbonate concentration, was analyzed by a repeated measures analysis of covariance. Demographic data were compared using Student's t-test or Wilcoxon's rank-sum test for continuous variables, and Pearson's chi-square test or Fisher's exact test for categorical variables. The choice of test depended on the distribution and type of data. Variables evidencing a meaningful baseline difference between the treatment groups were used as covariates in a secondary efficacy analysis model. A p value of less than 0.05 was regarded as statistically significant.
There was no intention in this study to perform a statistical hypothesis test for venous plasma bicarbonate changes between the two treatment groups. The objective of this study was to evaluate the effect of the bicarbonate/lactate solution relative to Dianeal PD1 by constructing a 95% confidence interval around the difference between the two treatments’ changes from their baseline levels. The sample size in this study was calculated to keep the confidence interval width between the center and the low or high end within 2.5 mmol/L venous plasma bicarbonate. Based on this requirement and an estimated standard deviation of 3 mmol/L, this study (with a 2:1 ratio allocation of patients between the test and control groups) required 16 patients in the test group and 8 patients in the control group (a total of 24 patients). With an anticipated 20% dropout rate, it was recommended that approximately 30 patients be enrolled to ensure the required sample size of 24 patients would be met at the end of the study.
Results
Demographics
A total of 33 patients were enrolled in the study; 2 dropped out before randomization, leaving 20 who were randomly allocated to the bicarbonate/lactate-based solution and 11 to the lactate-based solution. The baseline demographics of the two groups are shown in Table 2. The two groups were similar in all respects. All patients, except one control patient who died following a cerebrovascular accident, completed the study.
Baseline Demographics of Intent-to-Treat Population a
nPNA = protein equivalent of nitrogen appearance normalized to standard body weight.
The intent-to-treat population is defined as those patients who started the treatment period and took at least one dose of study solution.
Values are means with ranges. None of the differences between the two groups were statistically significant.
Venous Plasma Bicarbonate
In the intent-to-treat population (all patients who started the treatment period and took at least one dose of study solution), with respect to mean visit-by-visit values, it was noted that in the bicarbonate/lactate group there was a rise in venous plasma bicarbonate of approximately 3 mmol/L at each on-treatment visit, from a baseline level of 23.0 mmol/L, and a subsequent return to baseline values during follow-up (Figure 1, Table 3). There were no statistically significant changes to venous plasma bicarbonate levels in the control group, remaining between 22 and 23 mmol/L. The baseline values in the two groups were not statistically significantly different. The between-treatment difference was statistically significant, whether one considered the visit-by-visit comparison or the analysis of variance (p < 0.05 at each visit, or 0.001 overall, respectively).
Selected Laboratory Results Throughout the Study (Intent-to-Treat Population a )
Bic/Lac = bicarbonate/lactate-buffered peritoneal dialysis solution; pCO2 = partial pressure CO2.
The intent-to-treat population is defined as those patients who started the treatment period and took at least one dose of study solution.
All patients used Dianeal PD1 (Baxter Healthcare, Castlebar, Ireland) at month -1, day 0, and follow-up.
p < 0.05 versus control group (Student's t-test). No other differences were statistically significant.
The value of n in the control group is reduced by 1 at month 3 and follow-up because 1 control patient dropped out. All values shown are mean ± standard deviation. All values are in mmol/L except pCO2, which is in mmHg.
Summary of venous plasma bicarbonate levels by visit (intent-to-treat group). All patients used the control solution (Dianeal PD1; Baxter Healthcare, Castlebar, Ireland) at month –1, day 0, and follow-up. The between-group differences were statistically significant at months 1, 2, and 3 (p < 0.05). The change from baseline was statistically significant at months 1, 2, and 3 in the bicarbonate/lactate (Bic/Lac) group, but at no visits in the control group (p < 0.05, Student's t-test). Analysis of variance of change from baseline also revealed an overall, statistically significant, between-group difference (p < 0.05). Values are mean ± standard deviation.
These analyses were repeated in a subgroup of patients (6 controls, 15 bicarbonate/lactate) who had no changes in calcium-containing phosphate binders [levels of which may affect plasma bicarbonate levels (23)]. The same increases in venous plasma bicarbonate were seen in this subgroup, although the level of statistical significance was reduced (but remained statistically significant) because of the smaller sample size. These changes in venous plasma bicarbonate were mirrored by similar changes in total venous plasma CO2. The pH in the bicarbonate/lactate group went up, from 7.3 to 7.4, during the treatment period. In the control group, pH fell from a baseline value of 7.4 to 7.3. The treatment period venous pH values were statistically significantly different at months 2 and 3 (p < 0.05).
During the treatment period, fewer venous plasma bicarbonate values below 24 mmol/L (lower end of the normal range) were recorded for patients in the bicarbonate/lactate group than for patients in the control group (Table 4); the between-group differences with respect to the number of values in the low/in-range categories were maximal at month 2 and were statistically significant at months 1, 2, and 3 (p < 0.05). There were no venous plasma bicarbonate levels above 32 mmol/L in either group (upper end of normal range) during any part of the study.
Bic/Lac = bicarbonate/lactate-buffered peritoneal dialysis solution.
All other values were in the range 24 – 32 mmol/L. There were no values above 32 mmol/L at any time.
The intent-to-treat population is defined as those patients who started the treatment period and took at least one dose of study solution.
All patients used Dianeal PD1 (Baxter Healthcare, Castlebar, Ireland) at month -1, day 0, and follow-up.
p < 0.05, Cochran–Mantel Haenszel test.
There were 11 control and 20 Bic/Lac patients at every visit except month 3 and follow-up, when the total was reduced to 10 because 1 control patient dropped out.
Vital Signs and Physical Examination
There were no clinically significant differences in vital signs or results of physical examination between the patient groups at any time-point, nor were there obvious changes from baseline in either group.
Other Biochemical Parameters
There were occasional fluctuations in the other laboratory parameters measured, but these appeared to be random. There was no evidence of any systematic change in any parameter (apart from blood gas parameters) over time, or of any difference between the patient groups. There were no statistically significant differences between the groups with respect to serum calcium levels, despite the fact that the calcium level was 1.75 mmol/L in the lactate-based solution and 1.25 mmol/L in the bicarbonate/lactate mix. This was also true when the patients who had changes to calcium-containing phosphate binders were removed from the analysis. Patients were permitted the use of compounds such as calcitriol to control calcium levels, and it is possible that this prevented measurable changes. A display of pertinent blood results (those that relate to acid–base balance) is given in Table 3.
As would be expected for patients receiving CAPD, a number of mostly solution-unrelated adverse events were recorded. These included peritonitis (3 patients using bicarbonate/lactate and 1 patient while using the control product), abnormal laboratory test findings, and hypertension. The peritonitis rate was not calculated, as the duration of the study and the number of patients were insufficient to allow valid calculation. There were no cases of alkalosis or acidosis reported in the bicarbonate/lactate group, but there were two cases of asymptomatic acidosis in the control group.
Discussion
In this randomized controlled study, venous plasma bicarbonate levels were studied in patients initially treated with Dianeal PD1 (containing 35 mmol/L lactate). Because of the higher buffer level in the bicarbonate/lactate solution, it was predicted that the venous plasma bicarbonate levels would rise when patients were switched. This was indeed observed: a rise in venous plasma bicarbonate of around 3 mmol/L was seen. Nolph et al., switching patients from a 35 mmol/L lactate solution to a 40 mmol/L lactate solution, saw a rise of 3.5 mmol/L in plasma bicarbonate, suggesting that bicarbonate and lactate act in much the same way (24).
The bicarbonate level (25 mmol/L) for the test formulation was selected because it corresponds to the physiologic level found in blood. The test solution is supplemented with 15 mmol/L L-lactate to help maintain acid–base balance. The pCO2 in this formulation is also at the physiologic level rather than the supranormal level required for a pure bicarbonate solution (16).
It should be noted that the lactate and the bicarbonate/lactate solutions also differed with respect to their calcium, chloride, and magnesium contents. Kellum noted that acid–base balance can be influenced by the ion-charge balance in plasma (25). However, no differences were noted between the groups with respect to calcium or phosphorus, and only venous plasma bicarbonate, venous total CO2, and venous plasma pH were statistically significantly increased in the bicarbonate/lactate group. However, since parathyroid hormone (PTH) was not measured in this study, the full effects of these differences in ionic contents cannot be judged.
The Dialysis Outcomes Quality Initiative guidelines recommend that low plasma bicarbonate levels should be avoided (26). Acidosis is known to adversely affect protein metabolism and, in fact, may be an important stimulus for net protein catabolism in chronic uremic patients (27). The prevalence of malnutrition is reported to be 34% – 60% in dialysis patients, and malnutrition is a strong predictor of increased morbidity and mortality (28). Several factors cause or contribute to malnutrition, such as low protein and energy intake due to inadequate dialysis and the catabolic effects of acidosis (29).
In addition, observations from animal and clinical studies suggest that metabolic acidosis affects several organs and other physiological functions, including depression of cardiac function and stimulation of PTH activity (30). Since metabolic acidosis has been clearly proven to have a number of detrimental effects, it is recommended that all efforts be made to correct it (30-32). Many physicians use the normal venous plasma bicarbonate range as a target, but higher levels have been advocated as it has been found that, at bicarbonate values around 30 mmol/L, protein breakdown is reduced (2,30,33). Also, Stein et al. have reported that the CAPD patient's nutritional state (increases in body weight and midarm circumference) is improved when acidosis is fully corrected using a 40 mmol/L lactate solution compared to the lower venous plasma bicarbonate levels achieved with a 35 mmol/L lactate solution (2). In the present study, statistically significantly fewer patients had venous plasma bicarbonate levels below 24 mmol/L in the bicarbonate/lactate group compared with the control group, showing that pre-existing acidosis can be corrected by the bicarbonate/lactate solution. Interestingly, this was not countered by an increase in the incidence of significant alkalosis (plasma bicarbonate levels > 32 mmol/L). It is true that acidosis may also be corrected with oral alkali supplements, such as calcium carbonate and sodium bicarbonate, but this can require large oral doses (up to 12 g/day) that may be hard to tolerate on a fluid-restricted diet (authors’ clinical experience). Also, as noted by Bushinsky, the magnitude of the effect of calcium carbonate on plasma bicarbonate levels is difficult to predict, and a strict correlation may not always be seen (34).
Overall, the new 25 mmol/L bicarbonate/15 mmol/L lactate solution allowed better correction of acid–base status than the 35 mmol/L lactate solution in most of the patients studied. There seemed to be no new adverse events associated with this solution, which appears to be safe, effective, and appropriate for use in a wide range of patients. It may be particularly indicated in patients who need a high calcium level PD solution but who cannot reach target venous plasma bicarbonate levels with this type of solution.
Footnotes
Acknowledgments
This study was supported by a grant from Baxter Healthcare Ltd.
We thank the patients who agreed to participate in this study and the nurses who played such an integral part in the performance of many of the tests and in data collection. The authors also acknowledge the data management contributions of Noëlle Alexandre at Baxter Healthcare and the statistical work performed by Metadata Solutions Ltd.