María-del-Carmen Prado-Uribe,1 Guadalupe Alcántara,1 and Dante Amato1
Unidad de Investigación Médica en Enfermedades Nefrológicas,1 Hospital de Especialidades, Centro MédicoNacional Siglo XXI; Hospital General de Zona 27,2 Hospital General de Zona 47,3 Hospital General de
Zona 8,4 Hospital General de Zona 25,5 Instituto Mexicano del Seguro Social, México City, México
Correspondence to: R. Paniagua, Unidad de InvestigaciónMédica en Enfermedades Nefrológicas, Hospital de Especial-idades, Centro Médico Nacional Siglo XXI, Av. Cuauhtemoc 330,Col. Doctores, México, D.F., C.P. 06725, México.
♦♦♦♦♦ Background: Icodextrin-based solutions (ICO) have clini-cal and theoretical advantages over glucose-based solutions(GLU) in fluid and metabolic management of diabetic peri-toneal dialysis (PD) patients; however, these advantageshave not yet been tested in a randomized fashion.♦♦♦♦♦ Objective: To analyze the effects of ICO on metabolic andfluid control in high and high-average transport diabeticpatients on continuous ambulatory PD (CAPD).♦♦♦♦♦ Patients and Methods: A 12-month, multicenter, open-label, randomized controlled trial was conducted to com-pare ICO (n = 30) versus GLU (n = 29) in diabetic CAPDpatients with high-average and high peritoneal transportcharacteristics. The basic daily schedule was 3 × 2 L GLU(1.5%) and either 1 × 2 L ICO (7.5%) or 1 × 2 L GLU (2.5%)for the long-dwell exchange, with substitution of 2.5% or4.25% for 1.5% GLU being allowed when clinically neces-sary. Variables related to metabolic and fluid control weremeasured each month.♦♦♦♦♦ Results: Groups were similar at baseline in all measuredvariables. More than 66% of the patients using GLU, butonly 9% using ICO, needed prescriptions of higher glucoseconcentration solutions. Ultrafiltration (UF) was higher(198 ± 101 mL/day, p < 0.05) in the ICO group than in theGLU group over time. Changes from baseline were more pro-nounced in the ICO group than in the GLU group for extra-cellular fluid volume (0.23 ± 1.38 vs –1.0 ± 1.48 L, p < 0.01)and blood pressure (systolic 1.5 ± 24.0 vs –10.4 ±30.0 mmHg, p < 0.01; diastolic 1.5 ± 13.5 vs –6.2 ±14.2 mmHg, p < 0.01). Compared to baseline, patients inthe ICO group had better metabolic control than those inthe GLU group: glucose absorption was more reduced (–17 ±44 vs –64 ± 35 g/day) as were insulin needs (3.6 ± 3.4 vs –
9.1 ± 4.7 U/day, p < 0.01), fasting serum glucose (8.3 ± 36.5vs –37 ± 25.8 mg/dL, p < 0.01), triglycerides (54.5 ± 31.9vs –54.7 ± 39.9 mg/dL, p < 0.01), and glycated hemoglobin(0.79% ± 0.79% vs –0.98% ± 0.51%, p < 0.01). Patients inthe ICO group had fewer adverse events related to fluid andglucose control than patients in the GLU group.♦♦♦♦♦ Conclusion: Icodextrin represents a significant advan-tage in the management of high transport diabetic patientson PD, improving peritoneal UF and fluid control and reduc-ing the burden of glucose overexposure, thereby facilitat-ing metabolic control.
Perit Dial Int 2009; 29:422–432 www.PDIConnect.com
Cardiovascular diseases are the main cause of morbid-ity and mortality among dialysis patients, in both
hemodialysis and peritoneal dialysis (PD) (1–3). Fluidoverload and hypertension underlie these disturbances.Fluid retention is the origin of hypertension in most di-alysis patients. Adequate control of sodium balancemakes antihypertensive drugs largely unnecessary (4–7).Control of extracellular fluid volume (ECFv), and there-fore hypertension, retards the loss of residual renal func-tion and decreases morbidity and mortality (8–10).Notwithstanding, control of ECFv is not an easy goal fordialysis patients to achieve.
One of the most important issues in PD therapy todayis how to minimize the use of glucose as osmotic agentin PD solutions in order to avoid its metabolic side ef-fects. Use of hypertonic glucose has been associated withhyperglycemia, hyperinsulinemia, and obesity (11–13).Other disadvantages include bioincompatibility, ad-vanced glycated end-product generation, peritoneal
PDI JULY 2009 – VOL. 29, NO. 4 ICODEXTRIN IN DIABETIC PD PATIENTS
patients received 3 × 2 L 1.5% glucose exchanges. Inaddition, patients in the control group received GLU, atleast 1 bag with 2.5% glucose, in the long dwell and pa-tients in the icodextrin group received ICO (7.5%) in thelong dwell. Liberal use of 2.5% or 4.25% GLU was allowedin both groups in order to reach treatment goals. Dietarysodium intake prescription was 50 mmol/day for bothgroups. To achieve the best compliance with the prescrip-tion, patients received seven different menus, with thecooking methods for the usual Mexican meals, suggestedby a dietitian. When needed, insulin was administeredsubcutaneously.
SAMPLE SIZE
Sample size calculation was based on the differencesin reduction of ECFv. A net difference of more than 1.0 Land a total standard error of 0.4 were expected. Withα = 0.05 and two tails, the power was calculated to be(1 – β) = 0.80 for a sample of 30 patients per group(Power & Precision v 2.0w; Biostat, Englewood, NJ, USA).
PATIENTS
All the studied patients signed an informed writtenconsent. The participating patients were recruited fromfour general hospitals belonging to the InstitutoMexicano del Seguro Social and located in the metro-politan area of Mexico City (Zone General Hospitals num-bers 8, 25, 27, and 47).
An initial screening was performed among the preva-lent CAPD populations of the participating hospitals.Adult patients with diabetes mellitus with high and high-average peritoneal transport status were included with-out selection by gender, residual renal function, or timeon dialysis. A simplified version of the peritoneal equili-bration test (29) was performed within a month beforerandomization. The cutoff point for each category wasas previously described in the Mexican population (30).Patients were excluded when seropositive for hepatitis Bor HIV, if they had malignancies, or were receiving im-munosuppressive medication. Patients that had had aperitonitis episode 1 month or less before being screenedwere also excluded. Twenty-three patients left the studybefore completing the scheduled follow-up for the fol-lowing reasons: patient decision, medical decision, kid-ney transplant, or address change.
CLINICAL OUTCOMES
Primary outcomes were improvement in peritoneal UF,reduction of ECFv, and metabolic control. Secondary
damage, and, in the long term, loss of ultrafiltration (UF)capacity (14). In a significant number of patients classi-fied as high transporters, glucose use as osmotic agenthas its limitations (15–17) due to rapid peritoneal glu-cose absorption into the circulation with dissipation ofthe osmotic gradient. It is common practice when fac-ing such a situation to use solutions with higher glucosecontent, resulting in increases in the adverse metaboliceffects of glucose exposure. The disadvantages of hyper-tonic glucose are even greater in diabetic patients, manyof whom are high transporters (18,19).
In the 1990s, icodextrin, a glucose polymer, began tobe used in Europe as a replacement for the traditionalglucose in PD solutions for the long dwell in both con-tinuous ambulatory PD (CAPD) and automated PD(20–22). The effectiveness of icodextrin as oncotic agenthas been demonstrated, as the gradient can be main-tained at adequate UF values for 8 – 12 hours (23). Otheradvantages of icodextrin-based solutions (ICO) are thatthey are well tolerated, lack the metabolic side effectsof glucose (23), and enhance the clearance of small- andmiddle-size molecules as a consequence of increasedconvective flow (24,25). In spite of these potential ben-efits of icodextrin, there are no randomized controlledtrials available in PD patients with both diabetes andhigh-transport peritoneal membrane characteristics.Such information would be particularly important for PDpopulations with high diabetes rates, as in Mexico andmany other countries (26–28). The objective of thepresent study was to compare the clinical efficacy of ICOto that of glucose-based solutions (GLU) in diabetic pa-tients on PD with high and high-average peritonealtransport status.
MATERIAL AND METHODS
STUDY DESIGN
A prospective, randomized, controlled, open-labelclinical trial was conducted. The study protocol was ap-proved by the Local Research Committees of all the par-ticipating hospitals and the National Commission forScientific Research of the Instituto Mexicano del SeguroSocial (registration number: 2004-3601-0004). Thestudy was also registered in the Cochrane Registry forClinical Trials (CRG040600073).
The ongoing target of the treatment was to reach con-trol of blood pressure (BP; <130/80 mmHg) and edema(no clinical edema) through an increment in peritonealUF. Patients were randomly assigned to one of the twoarms of the study. Assignment was in a 1:1 ratio througha central randomization center. After randomization, all
outcomes were hospitalizations and therapy-relatedcomplications.
CLINICAL AND BIOCHEMICAL ASSESSMENTS
Visits were scheduled every week during the firstmonth and every month for the remaining 12 months.Clinical evaluations were done in each visit. Laboratoryassessments were done at baseline, at weeks 2 and 4,and every month for hematological and glycated hemo-globin data, as well as glucose, urea, and creatinine lev-els in serum, urine, and peritoneal effluent. Serumsamples were also analyzed for total cholesterol and trig-lycerides. Biochemical analyses were performed by stan-dard methods (Synchron CX-5 analyzer; Beckman, Brea,CA, USA). Serum albumin and glycated hemoglobin lev-els were measured at baseline and at months 1 – 6, 9,and 12 by nephelometry (Array; Beckman). Body com-position analyses were carried out at baseline, week 2,and every month by whole body multiple-frequency bio-impedance spectroscopy and body fluid compartmentswere derived with the software provided by the manu-facturer (Biodynamics, Seattle, WA, USA) (31,32).
STATISTICAL ANALYSIS
Data are presented as mean and standard deviationfor continuous variables and as proportions for categori-cal variables. ANOVA was used to analyze differencesbetween groups in variables recorded repeatedly overtime. Analysis of hospitalizations and peritonitis weremade by Poisson models. Statistical analysis was madewith SPSSw v14 (SPSS Inc., Chicago, IL, USA).
RESULTS
BASELINE CHARACTERISTICS
Fifty-nine patients were randomized and received thetreatment. The accrual time was 3 months (10 October2004 to 1 January 2005): 30 patients were randomizedto the ICO group and 29 to the GLU group. Follow-up wasat least 12 months; the study was terminated on 2 Feb-ruary 2006. At baseline both groups were similar in allthe analyzed variables (Table 1), including demographicand clinical characteristics, as well as comorbidities,laboratory measurements, and dialysis parameters.
INTERVENTION EFFECTS
At baseline most patients were using at least two 2.5%glucose bags or one 4.25% glucose bag per day. During
the first days of the study, patients followed the prescrip-tion required by protocol with subsequent adjustmentsto reach their goals. The actual prescription scheduleover time is shown in Figure 1. In the control group,nearly two thirds of the patients needed more than one2.5% glucose bag or at least one 4.25% glucose bag perday to reach the treatment target. In the ICO group, only9% of the patients needed a daily bag with a glucose con-centration higher than 1.5%. Figure 2 shows changes inperitoneal UF along the study period. Ultrafiltration re-mained higher in the ICO group than in the GLU groupthroughout the study.
OUTCOMES
Changes in body weight and body fluids are shown inFigure 3. During the first days of the study, patients inthe GLU group had increments in both body weight andECFv, probably due to the change in dialysis regimen,which could seem weaker in comparison to the basalschedule, causing a temporary change in fluid control.Body weight remained stable in the ICO group through-out the study; in contrast, it increased progressively inthe GLU group. Total body water (TBW) decreased in bothgroups but the change was more pronounced and fasterin the ICO group. ECFv decreased significantly in the ICOgroup from the first week of treatment, whereas it re-mained unchanged in the control group (GLU). Reduc-tions in TBW were more pronounced than those observedin ECFv, probably due to the reduction in intracellularwater. It is known that end-stage renal disease patients,as well as those with primary malnutrition, show incre-ments in intracellular water and in intracellular concen-trations of sodium. In this condition, we attributed thedifference between TBW and ECFv to a reduction in intra-cellular edema.
Systolic and diastolic BP decreased in the ICO groupfrom soon after beginning treatment to the ninth month,after which BP differences from baseline were less pro-nounced [Figure 4(a)]. In the GLU group, systolic anddiastolic BP remained essentially unchanged at the be-ginning of the study. For a brief period from month 5 tomonth 8, diastolic BP decreased, and from month 7 tomonth 8, systolic BP decreased. After that, diastolic BPreturned to baseline and systolic BP had significant in-crements over the initial values. It is remarkable that theearly changes in BP were in accordance with those ob-served in TBW and with ECFv in the ICO group (r = 0.32,p < 0.01). However, the changes in BP in the late follow-up period were inconsistent with the body fluid data. Thisfinding suggests a mechanism of overhydration-inde-pendent hypertension due perhaps to stimulation of the
PDI JULY 2009 – VOL. 29, NO. 4 ICODEXTRIN IN DIABETIC PD PATIENTS
renin–angiotensin system. Changes in antihypertensivetherapy were consistent with the changes in BP [Figure4(b)].
Soon after the beginning of the treatment, peritonealclearances of small-size molecules increased in the ICOgroup. After month 6, both Kt/V and creatinine clear-ance returned to baseline (Figure 5). In the controlgroup, Kt/V and creatinine clearance decreased signifi-
cantly over time. In the ICO group, glomerular filtrationrate (GFR; calculated as mean urea and creatinine renalclearance) and urine volume dropped from the initialvalues of the first week. This reduction was statisticallysignificant compared to the control group, but GFR re-mained stable for the remainder of the study period [Fig-ure 5(c)]. These changes were consistent with thereduction in body fluids.
TABLE 1Baseline Clinical Characteristics and Laboratory Measurements for the Two Groups
Parameter Glucose Icodextrin p Valuec
N 29 30Age (years) 60.5±9.3 58.9±7.9 0.47Gender (female/male)a 13/16 18/12 0.30Weight (kg) 61.5±12.3 62.1±10.4 0.85Height (cm) 159.0±13.7 156.6±7.4 0.41BMI (kg/m2) 24.10±3.41 25.18±3.74 0.70Waist/hip ratio 0.97±0.20 0.95±0.06 0.60Systolic BP (mmHg) 139.8±29.4 148.9±24.1 0.20Diastolic BP (mmHg) 80.2±16.4 84.9±15.0 0.27Total body water (% body weight) 59.49±7.77 59.42±7.44 0.90Extracellular body fluid volume (% body weight) 26.95±2.99 26.86±3.02 0.70Comorbiditya
BMI = body mass index; BP = blood pressure; PD = peritoneal dialysis; D/P Cr PET 4 hours = dialysate/plasma creatinine ratio at4 hours of peritoneal equilibration test; pCrCl = peritoneal creatinine clearance; rCrCl = renal creatinine clearance; pKt/V = perito-neal Kt/V; rKt/V = renal Kt/V.a Frequency.b Median | interquartile range.c p Values are based on Student’s t-test, Wilcoxon rank-sum test, or chi-square tests according to the variable characteristics.Values are mean±SD.
At baseline, values of metabolic control parameterswere similar in both groups; two thirds of patients wereusing insulin. During the follow-up, insulin medicationwas individually reconsidered and adjusted according tofasting serum glucose values. Metabolic control was bet-ter in the ICO group, which presented lower glucose ex-posure and glucose absorption, than in the GLU group
(Figure 6). Insulin requirements decreased significantlyand progressively in the ICO group, whereas an inversepattern was observed for the control group (Figure 6).Fasting serum glucose levels decreased over time in theICO group and were statistically different from baselineafter month 6 [Figure 7(a)], whereas glucose levels didnot change in the control group. The changes in triglyc-eride levels were less consistent [Figure 7(b)] during thefirst 6 months; however, thereafter a divergent trend wasobserved, with triglyceride levels decreasing in the ICOgroup and increasing in the GLU group. This decrease intriglyceride levels in the ICO group after month 6 wasparallel to the significant decrease in fasting serum glu-
Figure 3 — Significant changes were found in body composi-tion and edema. Body weight is shown in panel A. Patientsusing glucose-based dialysis solutions for the long dwell (GLUgroup; open symbols) had significant increments over base-line. Changes in total body water (TBW; B: solid lines) and ex-tracellular fluid volume (ECFv; dashed lines) are shown inpanel B. A faster and deeper reduction in TBW was seen in pa-tients using icodextrin-based dialysis solutions for the longdwell (ICO group; closed symbols) than in the GLU group. ForECFv, ICO patients had a significant and stable reduction frombaseline values; ECFv remained unchanged from baseline val-ues in GLU patients. *p < 0.05 GLU versus ICO; +p < 0.01 GLUversus ICO.
Figure 1 — Over time, patients using glucose-based dialysissolutions for the long dwell (GLU group) needed more frequentreplacement of 1.5% glucose bags with bags with higher glu-cose concentrations: for 2.5% glucose, >50% of GLU groupversus <10% of patients using icodextrin-based dialysis solu-tions (ICO group) (p < 0.01); for 4.25% glucose, >20% GLUgroup versus none in ICO group (p < 0.01). GLU > 1 bag 2.5%glucose: open squares on dashed line; ICO ≥ 1 bag 2.5% glu-cose: closed squares on dashed line; GLU ≥ 1 bag 4.25% glu-cose: open circles on solid line; ICO ≥ 1 bag 4.25% glucose:closed circles on solid line.
Figure 2 — Ultrafiltration was higher in patients using icodex-trin-based dialysis solutions for the long dwell (ICO group;closed circles) than in patients using glucose-based dialysissolutions (GLU group; open circles), even with more frequentuse of solutions with glucose concentration higher than 1.5%in the latter group. Mean difference between groups over timewas 197 mL/day (p < 0.006).
PDI JULY 2009 – VOL. 29, NO. 4 ICODEXTRIN IN DIABETIC PD PATIENTS
cose levels observed in the same group after month 6.Total cholesterol levels were similar in the two groups atbaseline and remained statistically unchanged duringthe follow-up. Glycated hemoglobin rose in the controlgroup and dropped in the ICO group [Figure 7(c)], withthe changes being statistically significant from month 3to the end of follow-up.
The numbers of dropouts were 11 in the GLU groupand 12 in the ICO group; the remaining patients com-pleted the scheduled follow-up and were available forstatistical evaluation. Adverse events related to fluidoverload and metabolic control were more frequent inthe GLU group compared to the ICO group (Table 2).There were no skin rashes, aseptic peritonitis, or aller-gic symptoms in the studied patients. Infectious peri-tonitis rates were similar in both groups. Overall causes
Figure 4 — Systolic (solid lines) and diastolic (dashed lines)blood pressure (BP) values were significantly lower in the pa-tients using icodextrin-based dialysis solutions for the longdwell (ICO group; closed symbols) throughout the study; *p <0.05 GLU versus ICO; +p < 0.01 GLU versus ICO (A). In patientsusing glucose-based dialysis solutions for the long dwell (GLUgroup; open symbols), systolic BP rose at the end of follow-up. Changes in antihypertensive therapy resembled those seenin BP: *p < 0.05 ICO versus GLU; +p < 0.01 ICO versus GLU (B).
Figure 5 — Changes in small-size molecule peritoneal clear-ances increased from baseline after the beginning of follow-up in patients using icodextrin-based dialysis solutions for thelong dwell (ICO group; closed circles), and were higher than inpatients using glucose-based dialysis solutions (GLU group;open circles) throughout the follow-up [panel A for perito-neal Kt/V (pKt/V); panel B for peritoneal creatinine clearance(pCrCl)]. Changes in glomerular filtration rate (GFR) are shownin panel C. Patients in the ICO group had a significant and sus-tained decrease compared to patients in the GLU group. Theeffect was mainly during the first week of treatment and wasrelated to contraction in extracellular fluid volume. *p < 0.05GLU versus ICO; +p < 0.01 GLU versus ICO.
of hospitalizations and days of hospitalization did notdiffer between the groups (Table 3).
DISCUSSION
The data here presented suggest that ICO treatment issuperior to GLU treatment, allowing for better metabolic
control and improved ECFv control in diabetic patients onPD. This is probably the first randomized trial evaluatingthe effect of icodextrin versus glucose as osmotic agentfor the long-dwell exchange in high and high-averagetransport diabetic patients treated with PD.
Since its introduction, ICO has proven its utility in UFmanagement in PD patients. Icodextrin solution has
Figure 6 — Significant reductions were seen in peritoneal glu-cose exposure (A), peritoneal glucose absorption (B), and in-sulin requirements (C) in patients using icodextrin-baseddialysis solution for the long dwell (ICO; larger closed circles).GLU = patients using glucose-based dialysis solutions (smalleror open circles). +p < 0.01 GLU versus ICO; *p < 0.05 GLU ver-sus ICO.
Figure 7 — Reduced glucose load was associated with lower lev-els of fasting serum glucose (A), serum triglycerides (B), andglycated hemoglobin (Hb a1c) (C) in patients using icodextrin-based dialysis solution for the long dwell (ICO; closed circles).GLU = patients using glucose-based dialysis solutions (opencircles). *p < 0.05 GLU versus ICO; +p < 0.01 GLU versus ICO.
PDI JULY 2009 – VOL. 29, NO. 4 ICODEXTRIN IN DIABETIC PD PATIENTS
been compared to GLU of different glucose concentra-tions and with different periods of follow-up. In most ofthese studies, ICO was superior to GLU in fluid removal,even in patients with high or high-average peritonealtransport status (21,23,33). Furthermore, ICO was moreeffective than the GLU with the highest glucose concen-tration (34) in improving UF in the long dwell exchangein patients on automated PD. The data presented hereare consistent with published papers. The group usingICO exhibited higher UF volumes than the GLU group.With most of those studies, there was a by-design dif-ference related to the treatment target. ICO has beenpreviously compared to f ixed GLU prescr iptions(21,25,34). In this study, as in the aforementioned studyby Davies et al. (33), ICO was compared to the optimalPD schedule to achieve the treatment target of maintain-ing the patient without hypertension and edema. In theactual prescription for the GLU group, patients receivedone extra 2.5% glucose bag (more than 60% of the pa-tients) or one 4.25% glucose bag (a third of the patients)instead of one or more of the 1.5% glucose bags. On theother hand, less than 10% of the patients in the ICOgroup needed replacement of 1.5% glucose solutions.Ultrafiltration declined equally in both groups over time.In the GLU group, the changes may have been due to in-creased glucose exposure but not so in the ICO group(35,36). Therefore, an alternative explanation for the
ICO group is contraction in fluid compartments, as sug-gested by reduction in urine volume after initiation ofICO treatment (37).
Changes in TBW and ECFv were congruent with thosein UF. A more effective fluid removal by ICO caused alarger and faster reduction in TBW and ECFv. Similar find-ings have been reported elsewhere (33).
Blood pressure had a paradoxical behavior. Systolicand diastolic values decreased until month 6 in the ICOgroup and followed the same trend as reductions in TBWand ECFv. After that, BP reductions were less significantin spite of reductions in fluid compartments. Incrementsin natriuretic peptides with the long-term use of ICO havebeen reported (38,39), suggesting that icodextrin me-tabolites may exert an oncotic effect and increase bloodvolume. We did not measure blood volume in order tosupport this hypothesis. An alternative explanation isexcessive fluid removal beyond dry weight and a second-ary activation of vasoactive hormones, as was found inintradialytic hypertension in hemodialysis patients(40,41). Unchanged BP values and late BP increments inthe GLU group may be explained by the hemodynamiceffects of hyperglycemia, such as increased heart rateand systolic volume (42).
Changes in UF may explain variations in some dialysisadequacy-related measurements. Peritoneal creatinineclearance and peritoneal Kt/V increased significantly,while urine volume and GFR decreased. Association ofperitoneal removal of small-size molecules with convec-tive diffusion has been previously reported (24,25,43).As UF increases, more effective removal of urea and creat-inine is expected. Increments in UF and a secondary re-duction in ECFv have been proposed to explain the dropin urine volume and residual renal function in patientstreated with ICO (37). It is very likely that the initial de-crease in GFR in the ICO group may be related to an ini-tial excessive UF, as denoted by the accentuated parallel
TABLE 3Hospitalization/Days of Hospitalization as Events/
Patient-Year Due to All Causes
Glucose Icodextrin p Valuea
Admissions 1.32 1.28 0.62Hospitalization days 10.68 7.66 0.16
a Poisson.
TABLE 2More Frequent Adverse Events
Glucose (n) Icodextrin (n) Glucose (rate)a Icodextrin (rate)a p Value (Poisson)
initial decrease in TBW in the ICO group, thereby leadingto an initial dehydration in some patients. The fact thatGFR remained stable thereafter in the ICO group wouldreinforce this hypothesis and its clinical importance.
Intuitively, the improvement in metabolic control as-sociated with ICO treatment could be expected. However,controversial results have been published in noninter-ventional studies and in studies with small numbers ofpatients with respect to continuously monitored glucosevalues, fasting serum glucose levels, or other metabolicmarkers (44–47). Data from the present study showed aclear advantage of ICO in metabolic control. Peritonealglucose exposure and peritoneal glucose absorptionwere reduced in the ICO group and, in spite of reducedneeds for insulin, lower levels of fasting serum glucoseand triglycerides could be demonstrated. In addition,glycated hemoglobin increased in the GLU group anddecreased in the ICO group.
Fluid overload and metabolic control-related adverseevents were seen less frequently in the ICO group, whichis in accordance with the physical and biochemical find-ings. There were no differences in peritonitis rates be-tween the groups. Some adverse events specificallyrelated to ICO, such as skin reactions and noninfectiousperitonitis (23), did not occur. The improved quality con-trol in the manufacture of icodextrin by measuring pep-tidoglycan levels (48) may be one possible contributingfactor.
In summary, icodextrin-based solutions may representa significant advantage for the management of high andhigh-average transport diabetic patients treated withCAPD, as demonstrated in this randomized trial. Improve-ments in metabolic control, reduction of the burden ofglucose exposure, and optimization of fluid managementin these patients are important evidence of the clinicalbenefits of icodextrin for diabetic PD patients.
DISCLOSURE
The sponsors did not participate in the study design,data collection, data analysis, data interpretation, orwriting of this report. The authors did not have any kindof relationship, commercial or otherwise, with Baxter,S.A. de R.L., México. The authors are employees of theIMSS. The corresponding author had full access to all dataand had final responsibility for submitting for publication.
ACKNOWLEDGMENTS
We thank Baxter, S.A. de R.L., México (grant 2005/23-585),and Instituto Mexicano del Seguro Social (IMSS) for their fi-nancial support.
The authors also thank Ms. Susan Drier for her assistancein preparing the manuscript. Part of this paper was presentedin abstract form at the 11th Congress of the ISPD.
REFERENCES
1. Foley RN, Parfrey PS, Sarnak MJ. Epidemiology of cardio-vascular disease in chronic renal disease. J Am Soc Nephrol1998; 9(Suppl 1):S16–23.
2. Collins AJ, Hao W, Xia H, Ebben JP, Everson SE, ConstantiniEG, et al. Mortality risks of peritoneal dialysis and hemo-dialysis. Am J Kidney Dis 1999; 34:1065–74.
3. Murphy SW, Foley RN, Barrett BJ, Kent GM, Morgan J, BarréP, et al. Comparative mortality of hemodialysis and peri-toneal dialysis in Canada. Kidney Int 2000; 57:1720–6.
4. Coles GA. Have we underestimated the importance of fluidbalance for the survival of PD patients? Perit Dial Int 1997;17:321–6.
5. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,Barre PE. Impact of hypertension on cardiomyopathy,morbidity and mortality in end-stage renal disease. Kid-ney Int 1996; 49:1379–85.
6. Ritz E, Koch M. Morbidity and mortality due to hyperten-sion in patients with renal failure. Am J Kidney Dis 1993;21(5 Suppl 2):113–18.
7. Günal AI, Duman S, Özkahya M, Töz H, Asci G, Akcicek F,et al. Strict volume control normalizes hypertension inperitoneal dialysis patients. Am J Kidney Dis 2001; 37:588–93.
8. Özkahya M, Ok E, Cirit M, Aydin S, Akcicek F, Basci A, et al.Regression of left ventricular hypertrophy in haemo-dialysis patients by ultrafiltration and reduced salt intakewithout antihypertensive drugs. Nephrol Dial Transplant1998; 13:1489–93.
9. Ates K, Nergizoglu G, Keven K, Akar H, Karatan O, DumanN, et al. Effect of fluid and sodium removal on mortalityin peritoneal dialysis patients. Kidney Int 2001; 60:767–76.
10. Paniagua R, Amato D, Vonesh E, Mujais S, Horl WH. Pre-dictive value of natriuretic peptides in patients on peri-toneal dialysis: results from the ADEMEX trial. Clin J AmSoc Nephrol 2008; 3:407–15.
11. Burkart J. Metabolic consequences of peritoneal dialy-sis. Semin Dial 2004; 17:498–504.
12. Heimburger O. Obesity on PD patients: causes and man-agement. Contrib Nephrol 2003; 140:91–7.
13. Jolly S, Chatatalsingh C, Bargman J, Vas S, Chu M,Oreopoulos DG. Excessive weight gain during peritonealdialysis. Int J Artif Organs 2001; 24:197–202.
14. Serlie MJ, Struijk DG, de Blok K, Krediet RT. Differences influid and solute transport between diabetic and nondia-betic patients at the onset of CAPD. Adv Perit Dial 1997;13:29–32.
15. Wang T, Heimbürger O, Waniewski J, Bergström J,Lindholm B. Increased peritoneal permeability is associ-ated with decreased fluid and small-solute removal andhigher mortality in CAPD patients. Nephrol Dial Transplant
PDI JULY 2009 – VOL. 29, NO. 4 ICODEXTRIN IN DIABETIC PD PATIENTS
1998; 13:1242–9.16. Cueto-Manzano AM, Correa-Rotter R. Is high peritoneal
transport rate an independent risk factor for CAPD mor-tality? Kidney Int 2000; 57:314–20.
17. Brimble KS, Walker M, Margetts PJ, Kundhal KK, RabbatCG. Meta-analysis: peritoneal membrane transport, mor-tality, and technique failure in peritoneal dialysis. J AmSoc Nephrol 2006; 17:2591–8.
18. Lamb EJ, Worrall J, Buhler R, Harwood S, Cattell WR,Dawnay AB. Effect of diabetes and peritonitis on the peri-toneal equilibration test. Kidney Int 1995; 47:1760–7.
19. Lin JJ, Wadhwa NK, Suh H, Cabralda T, Patlak CS. Increasedperitoneal solute transport in diabetic peritoneal dialy-sis patients. Adv Perit Dial 1995; 11:63–6.
20. Mistry CD, Mallick NP, Gokal R. Ultrafiltration with an isos-motic solution during long peritoneal dialysis exchanges.Lancet 1987; 2:178–82.
21. Mistry CD, Gokal R, Peers E, and the MIDAS Study Group. Arandomized multicenter clinical trial comparing isosmolaricodextrin with hyperosmolar glucose solutions in CAPD.Kidney Int 1994; 46:496–503.
22. Alsop RM. History, chemical and pharmaceutical devel-opment of icodextrin. Perit Dial Int 1994; 14(Suppl 2):S5–12.
23. Frampton JE, Plosker GL. Icodextrin. A review of its use inperitoneal dialysis. Drugs 2003; 63:2079–115.
24. Krediet RT, Douma CE, van Olden RW, Ho-dac-PannekeetMM, Struijk DG. Augmenting solute clearance in perito-neal dialysis. Kidney Int 1998; 54:2218–25.
25. Wolfson M, Piraino B, Hamburger RJ, Morton AR. Icodex-trin Study Group: a randomized controlled trial to evalu-ate the efficacy and safety of icodextrin in peritonealdialysis. Am J Kidney Dis 2002; 40:1055–65.
26. Su-Hernández L, Abascal-Macías A, Méndez-Bueno FJ,Paniagua R, Amato D. Epidemiologic and demographicaspects of peritoneal dialysis in Mexico. Perit Dial Int 1996;16:362–5.
27. Amato D, Alvarez-Aguilar C, Castaneda-Limones R,Rodriguez E, Avila-Diaz M, Arreola F, et al. Prevalence ofchronic kidney disease in an urban Mexican population.Kidney Int Suppl 2005; 97:S11–17.
28. Paniagua R, Amato D, Vonesh E, Correa-Rotter R, RamosA, Moran J, et al. Effects of increased peritoneal clear-ances on mortality rates in peritoneal dialysis: ADEMEX,a prospective, randomized, controlled trial. J Am SocNephrol 2002; 13:1307–20.
29. Twardowski ZJ. The fast peritoneal equilibration test.Semin Dial 1990; 3:141–2.
30. Paniagua R, Amato D, Correa-Rotter R, Ramos A, VoneshEF, Mujais S. Correlation between the peritoneal equili-bration test and the dialysis adequacy and transport testfor peritoneal transport type characterization. Perit DialInt 2000; 20:53–9.
31. Hannan WJ, Cowen SJ, Plester CE, Fearon KC, deBeau A.Comparison of bio-impedance spectroscopy and multi-fre-quency bio-impedance analysis for the assessment of ex-
tracellular and total body water in surgical patients. ClinSci (Lond) 1995; 89:651–8.
32. Sun G, French CR, Martin GR, Younghusband B, Green RC,Xie Y, et al. Comparison of multifrequency bioelectricalimpedance analysis with dual-energy X-ray absorptiom-etry for assessment of percentage body fat in a large,healthy population. Am J Clin Nutr 2005; 81:74–8.
33. Davies SJ, Woodrow G, Donovan K, Plum J, Williams P,Johansson AC, et al. Icodextrin improves the fluid statusof peritoneal dialysis patients: results of a double-blindrandomized controlled trial. J Am Soc Nephrol 2003; 14:2338–44.
34. Finkelstein F, Healy H, Abu-Alfa A, Ahmad S, Brown F, GehrT, et al. Superiority of icodextrin compared with 4.25%dextrose for peritoneal ultrafiltration. J Am Soc Nephrol2005; 16:546–54.
35. Davies SJ, Brown EA, Frandsen NE, Rodrigues AS,Rodriguez-Carmona A, Vychytil A, et al. Longitudinal mem-brane function in functionally anuric patients treated withAPD: data from EAPOS on the effects of glucose and ico-dextrin prescription. Kidney Int 2005; 67:1609–15.
36. Cooker LA, Holmes CJ, Hoff CM. Biocompatibility of ico-dextrin. Kidney Int Suppl 2002; 81:S34–45.
37. Konings CJ, Kooman JP, Gladziwa U, van der Sande FM,Leunissen KM. A decline in residual glomerular filtrationduring the use of icodextrin may be due to underhydra-tion. Kidney Int 2005; 67:1190–1.
38. Davies SJ, Plum J, Woodrow G, Johansson AC, Williams P,Donovan K, et al. Increasing our understanding of themechanism of action of icodextrin [Abstract]. Perit DialInt 2004; 24(Suppl 2):S33.
39. Bouchi R, Babazono T, Inoue A, Tanaka M, Tanaka N, HaseM, et al. Icodextrin increases natriuretic peptides in dia-betic patients undergoing CAPD. Perit Dial Int 2006; 26:604–7.
40. Chou KJ, Lee PT, Chen CL, Chiou CW, Hsu CY, Chung HM,et al. Physiological changes during hemodialysis in pa-tients with intradialysis hypertension. Kidney Int 2006;69:1833–8.
41. Chen J, Gul A, Sarnak MJ. Management of intradialytichypertension: the ongoing challenge. Semin Dial 2006;19:141–5.
42. Selby NM, Fonseca S, Hulme L, Fluck RJ, Taal MW, McIntyreCW. Hypertonic glucose-based peritoneal dialysate is as-sociated with higher blood pressure and adverse haemo-dynamics as compared with icodextrin. Nephrol DialTransplant 2005; 20:1848–53.
43. Asghar RB, Diskin AM, Spanel P, Smith D, Davies SJ. Influ-ence of convection on the diffusive transport and sievingof water and small solutes across the peritoneal mem-brane. J Am Soc Nephrol 2005; 16:437–43.
44. Hithaishi C, Lobbedez T, Padmanabhan S, Pineda ME,Oreopoulos DG. No beneficial effect of icodextrin on bloodglucose control. Perit Dial Int 2004; 24:199–200.
45. Marshall J, Jennings P, Scott A, Fluck RJ, McIntyre CW.Glycemic control in diabetic CAPD patients assessed by
continuous glucose monitoring system (CGMS). Kidney Int2003; 64:1480–6.
46. Furuya R, Odamaki M, Kumagai H, Hishida A. Beneficialeffects of icodextrin on plasma level of adipocytokines inperitoneal dialysis patients. Dialysis & Transplantation2006; 21:494–8.
T, Nakayama M, et al. Effects of icodextrin on glycemicand lipid profiles in diabetic patients undergoing perito-neal dialysis. Am J Nephrol 2007; 27:409–15.
48. Martis L, Patel M, Giertych J, Mongoven J, Taminne M,Perrier MA, et al. Aseptic peritonitis due to peptidogly-can contamination of pharmacopoeia standard dialysissolution. Lancet 2005; 365:588–94.
