Clinical outcomes and peritoneal histology in patients starting peritoneal dialysis are related to diabetic status and serum albumin levels

http://www.kidney-international.org & 2008 International Society of Nephrology

Clinical outcomes and peritoneal histology in patients starting peritoneal dialysis are related to diabetic status and serum albumin levels J-C Contreras-Vela´zquez1, V Soto2, Y Jaramillo-Rodrı´guez1, L-I Samaniego-Rı´os1, V Quin˜ones-Pe´rez1, M A´vila2, D Amato2 and R Paniagua2 1

Departments of Surgery, Pathology and Internal Medicine, Hospital General de Zona No. 18, Torreo´n, Coahuila, Me´xico and 2Unidad de Investigacio´n Me´dica en Enfermedades Nefrolo´gicas, Hospital de Especialidades, Centro Me´dico Nacional Siglo XXI, Instituto Mexicano del Seguro Social, Me´xico, D.F., Me´xico

Peritoneal morphological changes seem to be related to dialysis solutions bioincompatibility and to infections, but the uremic milieu per se may also contribute to peritoneal changes. The influence of diabetes and diabetes-associated comorbidities on peritoneal histological changes in the pre-dialysis stage have been insufficiently studied. The aim of this study is to analyze the effect of diabetes and serum albumin levels on peritoneal histology and certain clinical variables such as peritoneal permeability, technique failure, and general mortality in patients starting peritoneal dialysis (PD) treatment. Eighteen PD patients without diabetes (uremic non-diabetic group, U-ND) and 65 with diabetes (uremic diabetic group, U-D) were studied prospectively. Clinical and biochemical variables were registered, and a parietal peritoneum biopsy was obtained at the time of the peritoneal catheter placement. Peritoneal histology was evaluated by light microscopy and immunohistochemistry. A control group of 15 non-uremic, non-diabetic (NU-ND) patients who underwent non-complicated elective abdominal surgery was also studied and used as control. The proportion of patients with peritoneal morphological changes as evaluated by light microscopy was higher in the two groups of uremic patients than in the control. The U-D group had higher mesothelial loss (40.9 vs 29.4%), higher mesothelial basement membrane thickening (45.5 vs 23.5%), higher proportion of vascular wall thickening/sclerosis (39.7 vs 11.1%), and higher proportion of inflammatory infiltrate (45.4 vs 23.6%) than the U-ND group. Uremic patients had lower density of mesothelial cells and higher density of inflammatory cells than the control, as evaluated by immunohistochemistry. These changes were even more striking in the U-D group than in the U-ND group. On the other hand, inflammatory infiltration to the peritoneum,

Correspondence: R Paniagua, Unidad de Investigacio´n Me´dica en Enfermedades Nefrolo´gicas, Hospital de Especialidades, Centro Me´dico Nacional Siglo XXI, Av Cuauhtemoc 330, Col Doctores Me´xico, DF CP 06725, Mexico. E-mail: [email protected] or [email protected] S34

mesothelial cell loss, and mesothelial basement membrane thickening were associated with higher technique failure and mortality. However, when the serum albumin level was introduced into the model, the aforementioned associations became nonsignificant. In conclusion, uremia and diabetes were associated with important peritoneal histological changes before starting PD treatment. Diabetes associated with uremia was more strongly related to the peritoneal changes than uremia per se. Hypoalbuminemia and peritoneal inflammatory infiltrate were markedly associated with technique failure and mortality in patients starting PD treatment. Kidney International (2008) 73, S34–S41; doi:10.1038/sj.ki.5002599 KEYWORDS: peritoneal dialysis; peritoneal histology; diabetes; serum albumin; mortality; inflammation

Preservation of peritoneal integrity is probably the most important challenge for patients with chronic kidney disease treated by peritoneal dialysis (PD). Increased peritoneal permeability with ultrafiltration failure and sclerosing peritonitis are two of the leading causes of treatment failure in the long term.1–3 Chronic exposures to dialysis solutions and infections have been related to morphological and functional changes of the peritoneal membrane.4,5 However, the uremic milieu per se may also play a significant role in these changes. Peritoneal biopsies from non-dialyzed patients with chronic kidney disease and patients on hemodialysis have shown structural changes, such as mesothelial cell loss, neovascularization, and fibrosis that cannot be attributed to PD solutions.6,7 Experimental diabetes has also been reported to induce peritoneal anatomical changes mainly in the capillary vessels.8 This same study has also shown that diabetes may induce functional changes, as neoangiogenesis was associated with higher peritoneal permeability in the diabetic rats.8 In patients starting PD, solute transport and vascular endothelial growth factor serum levels are correlated.9 The association of non-vascular histological changes with Kidney International (2008) 73, S34–S41

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

peritoneal function has not been demonstrated, probably because in most of the studies, peritoneal biopsies were obtained after the peritoneal function was lost or when the patients required catheter removal for other reasons.6,10 Whether these changes represent a risk for technique failure or mortality is not known. The aim of this study was to analyze peritoneal anatomical changes in incident PD patients, their relationship to peritoneal permeability, and their role in technique failure and mortality. RESULTS

There were 18 uremic non-diabetic (U-ND) patients, 65 uremic diabetic (U-D) patients, and 15 non-uremic, nondiabetic (NU-ND) controls. The main clinical and demographical data are shown in Table 1. The most significant difference is that the U-D group had a higher proportion of malnutrition, and the U-ND group had a higher proportion of patients starting dialysis in urgent manner due to late referral. Table 2 shows the main biochemical results. The U-D group had higher glucose levels, and lower albumin and creatinine levels as compared to the U-ND group. Table 3 Table 1 | General characteristics

n Gender (M/F) Age (years) Weight (kg) Height (m) Body mass index (kg/m2) SGA (% with moderate to severe malnutrition) Late referral (%)

NU-ND controls

U-ND patients

15 2/13 36.572.9 78.374.8 1.6170.02 30.0571.41

18 7/11 54.874.5** 70.272.9 1.6570.02 26.0871.28

U-D patients

0

28.5

53.5*

39.0

24z

65 34/31 60.971.2** 68.472.0* 1.6270.01 25.970.59**

F, female; M, male; NU-ND, non-uremic, non-diabetic; SGA, subjective global assessment; U-D, uremic, diabetic; U-ND, uremic, non-diabetic. *Po0.05 vs NU-ND; **Po0.01 vs NU-ND. z Po0.05 vs U-ND.

shows the histological results, as analyzed by light microscopy. Both groups of patients had important morphological changes as compared to the control, with mesothelial loss and mesothelial basement membrane thickening. Loss of mesothelial cells and thickening of mesothelial basement membrane as well as vascular wall thickening and sclerosis and inflammatory infiltration were more striking in the U-D group than in the U-ND group. Figure 1 shows illustrative images for comparison. The immunohistochemical study corroborated theses results. Markers of peritoneal surface integrity such as epithelial membrane antigen, cytokeratin, carcinoembryonic antigen, and mesothelial cells were significantly lower in the U-D group than in the U-ND group (Table 4 and Figure 2). The immunostaining density for inflammatory cell markers was consistent with the light microscopy results. Density for common leukocyte antigen, T and B cells, and macrophages were higher in the U-D group. Figure 3 shows illustrative images for comparison between groups. There was lack of integrity of the vascular Table 3 | Light histology data

n Mesothelium (% with loss) Mesothelial basement membrane (% with focal loss, fragmentation, or thickening) Neoangiogenesis (%) Blood vessels (% with thickening or scleroses) Inflammatory infiltration Acute Chronic

NU-ND controls

U-ND patients

U-D patients

15 14.5

18 29.4*

65 40.9**,z

14.3

23.5*

45.5**,z

0 10.0

11.1 11.1

5.9 39.7**,z

14.3 14.4

17.5 23.6*

19.7 45.4*,zz

NU-ND, non-uremic, non-diabetic; U-D, uremic, diabetic; U-ND, uremic, non-diabetic. *Po0.05 vs NU-ND; **Po0.01 vs NU-ND. z Po0.05 vs U-ND; zzPo0.01 vs U-ND.

Table 2 | Biochemical data

n Serum glucose (mg per 100 ml) Serum urea (mg per 100 ml ) Serum creatinine (mg per 100 ml) Serum albumin (g per 100 ml) Serum total protein (g per 100 ml) D/P creatinine Hemoglobin (g per 100 ml) Hematocrit (%) Leukocytes (cells  103 per mm3) Neutrophils (cells  103 per mm3) Lymphocytes (cells  103 per mm3)

NU-ND controls

U-ND patients

U-D patients

15 98.574.2 23.272.7 0.8570.05 4.5670.18 6.5570.23

18 94.373.4 184.8728.3** 11.5371.54** 4.3770.30** 6.0270.23 0.6770.03 8.970.5** 27.171.7** 7.9370.88 5.5570.84 1.5970.19*

65 161.4713.4** 161.0712.1** 7.6070.48**,z 3.6570.14**,z 6.0270.13 0.6670.01 9.370.2** 28.170.8** 8.8770.48 6.4870.47* 1.5770.10*

13.170.3 39.471.0 7.9470.92 5.2170.84 2.0670.21

D/P, dialysate/plasma; NU-ND, non-uremic, non-diabetic; U-D, uremic, diabetic; U-ND, uremic, non-diabetic. *Po0.05 vs NU-ND; **Po0.01 vs NU-ND. z Po0.05 vs U-ND.

Kidney International (2008) 73, S34–S41

S35

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

a

b

c

d

e

f

g

h

i

Figure 1 | Parietal peritoneal biopsies (original magnification  200) stained with (a–c) hematoxylin and eosin (H&E) and (d–f) Masson’s trichrome. Peritoneum from an NU-ND patient (control group) appears without histological alterations. (a, d) Mesothelial cells appear as monolayer and no stroma alterations. Peritoneum from a U-ND patient. Mesothelium is preserved but hyperplasic and reactive. (b, e) Blood vessels are increased in number. (c, f) Peritoneum from a U-D patient where intermittent mesothelium, vascular sclerosis, and perivascular inflammatory cells are observed. The last images (original magnification  400) in (g–i), stained with H&E, are from U-D patients in which perivascular inflammatory cells are observed.

Table 4 | Immunohistochemistry data

n Epithelial membrane antigen Cytokeratin Carcinoembryonic antigen Mesothelium Common leukocyte antigen (CD45) B cell (anti-CD20) B cell (anti-CD79a) T cell (anti-CD3) T cell (anti-CD45RO) Macrophages (anti-CD68) Myeloperoxidase Endothelium (CD31) Endothelium (Ulex europeaus lectin) VCAM ICAM

NU-ND controls

U-ND patients

U-D patients

15 170727 160725 340758 135723 107749 119754 112769 107760 83764 150788 66728 388773 286734 5107113 281730

18 125727 96718* 247745 154734 128755 77736 67727 119753 85739 78733 71735 204747* 169728* 225743** 267743

65 106712* 108711 207720* 124715z 187730z 141730z 122722z 134724 118722z 143726z 56715 206718* 220729z 263725** 295720

ICAM, intracellular cellular adhesion molecule; NU-ND, non-uremic, non-diabetic; U-D, uremic, diabetic; U-ND, uremic, non-diabetic; VCAM, vascular cellular adhesion molecule. Data are expressed as pixels per field. *Po0.05 vs NU-ND; **Po0.01 vs NU-ND. z Po 0.05 vs U-ND; zzPo0.01 vs U-ND.

endothelium in both groups of patients, which was more noticeable in the U-D group than in the U-ND group. Vascular cellular adhesion molecule density was significantly decreased in both groups of patients. Significant correlations among light microscopy and immunohistochemistry variables were found with correlation S36

coefficients higher than 0.3 (Po0.01) for all pairs of variables. Peritoneal equilibrium test results were similar to those previously reported in a Mexican population,11 with no significant differences for the dialysate/plasma ratio between U-D and U-ND groups (Table 2). There were also no Kidney International (2008) 73, S34–S41

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

significant correlations between the dialysate/plasma ratio and any light microscopy or immunohistochemistry variables. There were 10 dropouts in the U-ND group and 46 in the U-D group. Deaths were significantly (Po0.05) higher in the U-D group (n ¼ 38) than in the U-ND group (n ¼ 5). However, in the Cox model, diabetes was not an independent risk factor for technique failure or mortality. Several models were considered for the analysis of factors associated with technique failure or mortality, including only the most representative variables. For instance, to analyze the effect of inflammation, only inflammatory infiltrate as evaluated by light microscopy was considered, and to evaluate cellular changes, only light microscopy results were included. This was done because of the high colinearity found among light microscopy and immunohistochemistry variables. For general mortality, the most representative model is shown in Table 5; inflammatory infiltration was the most significant independent factor for mortality (Figure 4). Patients with chronic inflammatory infiltration or the combination of acute and chronic inflammatory infiltration had the worst outcomes, and acute inflammatory infiltration does not appear to have negative effect on survival. However, when serum albumin was included in the model, it became the only statistically significant variable. The same procedure was used to analyze risk factors associated with technique failure (Table 6). Inflammatory infiltrate, assessed by light microscopy, and leukocytes and B cell immunohistochemical markers were significantly and independently associated with technique failure. When serum albumin was included in the model, it was the dominant risk factor. It should be noted that residual renal function was negligible and protein losses in the PD effluent did not happen.

a

b

c

d

e

f

Figure 2 | Immunostaining HBME-1 (mesothelium antibody) of mesothelial surface layer of the parietal peritoneum (original magnification  1000). (a, b) Cells from mesothelium of control group are organized in a continuous row without any alterations. Mesothelium from the U-ND group has reactive changes and the nuclei are irregular. (c, d) Microvilli are preserved although there is shortening. Mesothelium from the U-D group display discohesive cellular surface and massive infiltration of granulocytes and macrophages. (e, f) Positive immunostained cells have reactive changes with vacuolization of cytoplasm.

DISCUSSION

The results show that the peritoneal membrane of patients starting in a PD program has important histological changes, such as mesothelial loss, mesothelial basement membrane thickening, vascular wall sclerosis, and hyperplasia, as well as inflammatory cell infiltrates. These changes are more striking in the diabetic patient. These anatomical changes are not related to peritoneal permeability, but they are associated with hypoalbuminemia and represent a risk factor for technique failure and mortality. Previous studies have demonstrated similar changes such as mesothelial layer loss, mesothelial basement membrane thickening, vascular wall sclerosis, and hyperplasia.6–8 These structural changes may be due to an unspecific effect of the uremic milieu, but the precise pathophysiology is unknown. Diabetes increases the magnitude of these changes, as has been demonstrated experimentally.8 The frequency and intensity of inflammatory infiltrates have not been previously reported. The presence of these infiltrates seems not to be incidental or artifactual, as there is internal consistency among several variables obtained through light microscopy and immunohistochemistry. It is Kidney International (2008) 73, S34–S41

a

b

c

Figure 3 | Parietal peritoneal biopsies immunostained with CD45 (common leukocytary antigen) (original magnification  200). Peritoneum from a control patient (a) without inflammatory infiltration and peritoneum from a U-ND patient (b) with a few inflammatory cells in a field. (c) In the U-D patients, many perivascular inflammatory cells are observed. S37

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

Table 5 | Factors associated with patient mortality Variable Model 1 Mesothelial loss (+) Mesothelial basement membrane (thickening, fragmentation, or loss) () Vascular damage (fibrosis, necrosis, obstruction) (+) Inflammatory infiltration () No Acute Chronic

B

Significance

Exp (B)

95% CI

0.72 0.45

0.17 0.34

2.06 0.64

0.73–5.76 0.25–1.60

0.6

1.83

0.92–3.62

0.54 2.41 1.3

0.08 0.01 0.01 0.01 0.01

0.21 0.09 0.27

0.07–0.63 0.02–0.47 0.10–0.73

0.91 0.92

0.11 0.07

2.49 0.4

0.81–7.55 0.14–1.08

1.26

0.55–2.84

0.42 0.82 0.42 0.55

0.07–0.63 0.02–0.47 0.09–1.82 0.36–0.85

Model 2 Mesothelial loss (+) Mesothelial basement membrane (thickening, fragmentation, or loss) () Vascular damage (fibrosis, necrosis, obstruction) (+) Inflammatory infiltration () No Acute Chronic Serum albumin (g)

0.87 0.19 0.88 0.59

0.59 0.58 0.27 0.85 0.24 0.007

Model 3 Epithelial membrane antigen (Px) Mesothelium (Px) Common leukocyte antigen (Px+) B cell (anti-CD79a) (Px+) T cell (anti-CD3) (Px)

0.002 0.002 0.002 0.002 0.005

0.37 0.48 0.03 0.26 0.01

1 1 1 1 0.99

1.00–1.01 1.00–1.01 1.01–1.02 0.99–1.01 0.98–0.99

Model 4 Epithelial membrane antigen (Px) Mesothelium (Px) Common leukocyte antigen (Px+) B cell (anti-CD79a) (Px+) T cell (anti-CD3) (Px) Serum albumin (g)

0.003 0.004 0.001 0.002 0.002 0.703

0.37 0.09 0.58 0.45 0.27 0.003

1 1 1 1 1 0.49

0.99–1.00 0.99–1.00 0.99–1.00 0.99–1.00 0.99–1.00 0.31–0.79

0.23

CI, 95 % confidence interval, B, exponent from regression analysis; Exp (B), change in dependent variable per unit change of independent variable; Px, pixels with negative immunostaining; Px þ , pixels with positive immunostaining.

1.0

Survival

0.8 2

0.6

Inflammatory infiltration

0.4 1

4 Acute and chronic

3

3 Chronic only

0.2

2 Acute only 4

0.0 0

12

24

1 None 36

Months

Figure 4 | Graph shows the role of inflammatory infiltration on patient survival. Chronic inflammatory infiltration or the combination of chronic and acute inflammatory infiltration exerts adverse outcomes in patients starting in a PD program. The graph represents the model 1 in Table 5.

important to emphasize that only incident patients before starting PD were included in this study, and that a significant proportion of them had a delayed referral to the PD program. Therefore, the impact of the uremic milieu, the extracellular S38

fluid volume expansion, and comorbidity on peritoneal histological changes may have been enhanced. The cause of the inflammatory infiltrate is not evident. Bacterial translocation from the intestinal lumen may be implied. The aforementioned hypothesis was put forward to explain the low level of chronic inflammation observed in PD patients, but it has not been demonstrated.12 Increased intestinal epithelial permeability reported in uremic patients13,14 supports it, as well as motility disorders, edema, and intestinal anatomical changes present in uremic patients, which may favor an exaggerated proliferation of intestinal bacterial flora.15,16 Such deleterious effects may be further aggravated in diabetic patients, since intestinal motility disorders associated with diabetic neuropathy may enhance bacterial proliferation.17,18 Mesothelial loss, fibrosis, and neoangiogenesis are the dominant peritoneal morphological changes in PD patients and they are part of a process known as mesothelial– mesenchymal transdifferentiation. This process begins when mesothelial cells lose their polarity, detach, and migrate to deeper tissue layers, transforming their mesothelial phenotype to myofibroblasts, which in turn produce extracellular Kidney International (2008) 73, S34–S41

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

Table 6 | Factors associated with technique failure Variable Model 1 Mesothelial loss (+) Mesothelial basement membrane (thickening, fragmentation, or loss) () Neoangiogenesis (+) Vascular damage (fibrosis, necrosis, obstruction) (+) Inflammatory infiltration () No Acute Chronic

B

Significance

Exp (B)

95% CI

0.78 0.54

0.14 0.26

2.17 0.58

0.78–6.03 0.23–1.48

0.53 0.66

1.7 1.93

0.39–7.49 0.96–3.88

1.94 2.83 1.67

0.48 0.07 0.04 0.01 0.01 0.02

0.14 0.06 0.19

0.03–0.65 0.01–0.44 0.05–0.75

0.9 0.93

0.11 0.07

2.47 0.4

0.81–7.55 0.15–1.08

0.07 0.22

0.93 1.25

0.12–7.27 0.55–2.85

0.45 0.88 0.44 0.55

0.04–4.68 0.06–13.09 0.05–3.92 0.36–0.85

Model 2 Mesothelial loss (+) Mesothelial basement membrane (thickening, fragmentation, or loss) () Neoangiogenesis (+) Vascular damage (fibrosis, necrosis, obstruction) (+) Inflammatory infiltration () No Acute Chronic Serum albumin (g)

0.81 0.13 0.82 0.59

0.95 0.6 0.75 0.5 0.92 0.46 0.01

Model 3 Epithelial membrane antigen (Px) Mesothelium (Px) Common leukocyte antigen (Px+) B cell (anti-CD79a) (Px+) T cell (anti-CD3) (Px)

0.002 0.001 0.002 0.002 0.004

0.44 0.52 0.13 0.21 0.03

1 1 1 1 0.99

0.99–1.01 0.99–1.01 0.99–1.01 0.99–1.01 0.98–0.99

Model 4 Epithelial membrane antigen (Px) Mesothelium (Px) Common leukocyte antigen (Px+) B cell (anti-CD79a) (Px+) T cell (anti-CD3) (Px) Serum albumin (g)

0.002 0.004 0.004 0.002 0.002 0.615

0.39 0.06 0.72 0.34 0.22 0.005

1 0.99 1 1 0.99 0.54

0.99–1.01 0.99–1.01 0.99–1.01 0.99–1.01 0.99–1.01 0.35–0.83

CI, 95 % confidence interval, B, exponent from regression analysis; Exp (B), change in dependent variable per unit change of independent variable; Px, pixels with negative immunostaining; Px þ , pixels with positive immunostaining.

matrix and induce angiogenesis.19,20 Transforming growth factor-b seems to play a central role in this process. Our present data support the previous assumption. Transforming growth factor-b production increases stimulated by inflammatory processes are caused by dialysis solutions bioincompatibility, glycation advanced products, and infections.21,22 Our patients may be in a very early stage of the mesothelial–mesenchymal transdifferentiation process, when the negative effects of the uremic milieu, fluid retention, and diabetes complication reach a peak just before PD initiation. The association of peritoneal morphological changes with clinical outcomes has not been properly studied. In experimental models, histological changes induced by local inflammation and late glycation products are related to higher peritoneal membrane permeability secondary to vascular proliferation and vasodilatation.8,23,24 In patients on PD, this morpho-functional association has been only partially studied in a stage when permeability disorders or ultrafiltration failure are evident.11,19,20 In this study, an association between morphological changes and peritoneal permeability as evaluated by conventional peritoneal equiliKidney International (2008) 73, S34–S41

brium test (performed 1 month after catheter implantation) could not be demonstrated. In previous studies, a relationship between peritoneal transport type and biochemical variables associated with peritoneal architectural changes had not been found.9,25 The effect of local inflammation on clinical outcomes is not surprising. Inflammation, assessed by markers such as C-reactive protein, interleukins, and circulating leukocytes has been considered as a risk factor for mortality.26–29 However, histological evidence of local peritoneal inflammation has not been previously related to clinically relevant outcomes. In this study, inflammatory infiltrate was an independent risk factor for mortality in several Cox models. This finding is relevant because it discloses the important role played by the patients’ clinical conditions at the initiation of PD treatment on clinical outcomes. Diabetes and age did not reach statistical significance as risk factors for mortality. Inflammation was independently associated with mortality and was also related to delayed referral to the PD program, more severe uremic syndrome, and higher extracellular fluid volume expansion.30–32 Serum albumin was the more relevant S39

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

risk factor for mortality. Albumin is widely accepted as a marker of nutritional status, inflammation, and functional status of the peritoneal membrane.33–35 In this study, there is a high colinearity of peritoneal histological damage and serum albumin levels. Because serum albumin is a marker of more generalized damage, it annuls the effect of peritoneal histological lesion when both variables are included in the statistical models. The relationship between local inflammation and technique failure is not clear. Loss of the mesothelial cell layer and inflammation can predispose to increased peritoneal adhesion. We hypothesize that delayed initiation of PD treatment can make the patients more susceptible to catheter entrapment by the omentum. Diabetes increases the risk for inflammation, and this can explain the high rate of early technique failure reported in the Mexican PD patients.36 External validity cannot be established in this small-group single-center study, but the general characteristics of these patients are not different from other PD populations, except for the high prevalence of diabetic patients. These findings are in agreement with the previous body of knowledge regarding the deleterious effects of uremic milieu and diabetes on peritoneal histology as well as with the intestinal translocation of bacteria or bacterial products as a source of inflammation in chronic kidney disease patients. These findings warrant further research to be confirmed. MATERIALS AND METHODS Patients with chronic kidney disease before starting PD treatment from a single dialysis center in a general hospital were eligible. The procedures were in accordance with the ethical standards of the Institutional Committee on Human Experimentation and with the Declaration of Helsinki Principles (as revised in 1983). After approval of the protocol by the Institutional Review Board, signed informed consent forms were obtained from each participating patient. Clinical and demographic data of the patients were registered and a venous blood sample after overnight fast was obtained for biochemical analysis. A 2  2  0.5 cm fragment of parietal peritoneum was obtained at the time of peritoneal catheter surgical placement. Tissue samples were fixed in 10% formalin, included in paraffin, and cut in 0.2–0.3 mm sections. Two pathologists analyzed the slides blindly and independently, evaluating at least five fields from each section. Images were recorded and analyzed using an Olympus BX-51 microscope, a COOL-Snap Pro Plus camera, and the image analyzer software Image Pro Plus v5.1w. Histological sections were stained with hematoxylin and eosin, periodic acid Schiff, Masson’s trichrome, reticulum, and Alcian blue. Mesothelial morphological changes were analyzed using a semiquantitative scale (1 ¼ normal, 2 ¼ focal loss, 3 ¼ degenerated-lost). The mesothelial basement membrane was also evaluated semiquantitatively (1 ¼ normal; 2 ¼ fragmented, focal loss, or total loss). Angiogenesis and vascular wall damage (thickening, fibrosis, necrosis, and lumen obstruction) were evaluated in blood vessels. Inflammatory infiltrate was classified as normal-absent, acute, and chronic according to the characteristics of the predominant inflammatory cells. Other sections underwent de-paraffination for immunohistochemical studies by a conventional avidin–biotin complex technique. Antibodies used for this study were mouse anti-human epithelial S40

membrane antigen, mouse anti-human cytokeratin (low molecular weight), rabbit anti-human carcinoembryonic antigen (polyclonal), mouse anti-human carcinoembryonic antigen (monoclonal), mouse anti-human mesothelial cell (HBME-1), mouse anti-human CD45 (common leukocytary antigen), mouse anti-human CD20 (B cell), mouse anti-human CD79a (B cell), mouse anti-human CD3 (T cell), mouse anti-human CD45RO (UCHL-1, T cell), mouse anti-human CD68 (macrophages), rabbit anti-human CD myeloperoxidase, mouse anti-human CD31 (endothelial cell), rabbit anti-plant Ulex europeaus lectin (Dako, Deutschland Gmbh, Hamburg, Germany); mouse anti-human vascular cellular adhesion molecule, and mouse anti-human intracellular adhesion molecule (Serotec, Kidlington, Oxford, UK). Each variable was evaluated quantitatively according to dye intensity and size of the observed field. A peritoneal equilibrium test was carried out 1 month after catheter placement.37 All the patients were free from infections and catheter-related complications when the peritoneal equilibrium test was done. All the biochemical tests were performed with the standard techniques. All the data are shown as means7s.e.m. for continuous variables and as proportions for nominal variables. Differences between groups were analyzed by Student’s t-test or non-parametrical tests as necessary. Pearson’s or Spearman’s correlations were used as necessary. Survival was analyzed by a Cox model. All statistical tests were carried out with the SPSSw v.10.0 software. DISCLOSURE

RP is supported by Consejo Nacional de Ciencia y Tecnologı´a, Me´xico (CONACYT) Grant U42396-M, 2002 and grants from Baxter since 2000. DA is supported by CONACYT Grant Salud-2003-C01-055. All the other authors declared no competing interests.

ACKNOWLEDGMENTS

We thank Histology Technicians Benito Cha´vez-Renterı´a and Armando Medina-Cruz from the Instituto Nacional de Cardiologı´a Ignacio Cha´vez and Ms Susan Drier for her assistance in preparing this paper. REFERENCES 1.

Davies SJ, Phillips L, Griffiths AM et al. What really happens to people in long-term peritoneal dialysis? Kidney Int 1998; 54: 2207–2217. 2. Selgas R, Bajo MA, Del Peso G et al. Preserving the peritoneal dialysis membrane in long-term peritoneal dialysis patients. Semin Dial 1995; 8: 326–332. 3. Krediet RT, Lindholm B, Rippe B. Pathophysiology of peritoneal membrane failure. Perit Dial Int 2000; 20(Suppl 4): S22–S42. 4. Topley N. Membrane longevity in peritoneal dialysis: impact of infection and bio-incompatible solutions. Adv Ren Replace Ther 1998; 5: 179–184. 5. Topley N, Alobaidi Hm, Davies M et al. The effect of dialysate on peritoneal phagocyte oxidative metabolism. Kidney Int 1988; 34: 404–411. 6. Williams JD, Craig KJ, Topley N et al. Peritoneal dialysis: changes to the structure of the peritoneal membrane and potential for biocompatible solutions. Kidney Int 2003; 63(Suppl 84): S158–S161. 7. Williams JD, Craig KJ, Topley N et al. Peritoneal Biopsy Study Group. Morphologic changes in the peritoneal membrane of patients with renal disease. J Am Soc Nephrol 2002; 13: 470–479. 8. Stoenoiu MS, De Vriese AS, Brouet A et al. Experimental diabetes induces functional and structural changes in the peritoneum. Kidney Int 2002; 62: 668–678. 9. van Esch S, Zweers MM, Jansen MA et al. Determinants of peritoneal solute transport rates in newly started non-diabetic peritoneal dialysis patients. Perit Dial Int 2004; 24: 554–561. 10. Bertoli SV, Buzzi L, Ciurlino D et al. Morpho-functional study of peritoneum in peritoneal dialysis patients. J Nephrol 2003; 16: 373–378. Kidney International (2008) 73, S34–S41

J-C Contreras-Vela´zquez et al.: Diabetes, peritoneal histology, and clinical outcomes in PD

11.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22. 23.

24.

Paniagua R, Amato D, Correa-Rotter R et al. Correlation between the peritoneal equilibration test and the dialysis adequacy and transport test for peritoneal transport type characterization. Perit Dial Int 2000; 20: 53–59. Goncalves S, Pecoits-Filho R, Perreto S et al. Associations between renal function, volume status and endotoxaemia in chronic kidney disease patients. Nephrol Dial Transplant 2006; 21: 2788–2794. Magnusson M, Magnusson KE, Sundqvist T et al. Impaired intestinal barrier function measured by differently sized polyethylene glycols in patients with chronic renal failure. Gut 1991; 32: 754–759. Zuckerman MJ, Buchwald D, Watts MT. Intestinal permeability to 51Cr-EDTA in patients on chronic ambulatory peritoneal dialysis. Nephron 1994; 68: 408–409. el-Lakany S, Eagon PK, Gavaler JS et al. Gastrointestinal function, morphology, and immune status in uremia. Nutrition 1990; 6: 461–468. Strid H, Simren M, Stotzer PO et al. Patients with chronic renal failure have abnormal small intestinal motility and a high prevalence of small intestinal bacterial overgrowth. Digestion 2003; 67: 129–137. Sarno S, Erasmas LP, Haslbeck M et al. Orocaecal transit, bacterial overgrowth and hydrogen production in diabetes mellitus. Ital J Gastroenterol 1993; 25: 490–496. Cuoco L, Montalto M, Jorizzo RA et al. Eradication of small intestinal bacterial overgrowth and oro-cecal transit in diabetics. Hepatogastroenterology 2002; 49: 1582–1586. Ya´n˜ez-Mo´, Lara-Pezzi E, Selgas R et al. Peritoneal dialysis induces an epithelial–mesenchymal transition of mesothelial cells. N Eng J Med 2003; 348: 403–413. Aroeira LS, Aguilera A, Selgas R et al. Mesenchymal conversion of mesothelial cells as a mechanism responsible for high solute transport rate in peritoneal dialysis: role of vascular endothelial growth factor. Am J Kidney Dis 2005; 46: 938–948. Offner FA, Feichtinger H, Stadlmann S et al. Transforming growth factor-beta synthesis by human peritoneal mesothelial cells. Induction by interleukin-1. Am J Pathol 1996; 148: 1679–1688. Border WA, Noble NA. Transforming growth factor b in tissue fibrosis. N Engl J Med 1994; 331: 1286–1292. Flessner MF, Choi J, Vanpelt H et al. Correlating structure with solute and water transport in a chronic model of peritoneal inflammation. Am J Physiol Renal Physiol 2006; 290: F232–F240. Zareie M, De Vriese AS, Hekking LH et al. Immunopathological changes in a uraemic rat model for peritoneal dialysis. Nephrol Dial Transplant 2005; 20: 1350–1361.

Kidney International (2008) 73, S34–S41

25.

26.

27.

28.

29.

30.

31.

32.

33. 34.

35.

36.

37.

Rodrigues AS, Almeida M, Fonseca I et al. Peritoneal fast transport in incident peritoneal dialysis patients is not consistently associated with systemic inflammation. Nephrol Dial Transplant 2006; 21: 763–769. Paniagua R, Frı´as Y, Ventura MJ et al. C-reactive protein and antiChlamydia pneumoniae antibodies as risk factors of cardiovascular death in incident patients on peritoneal dialysis. Perit Dial Int 2003; 23: 132–137. Chung SH, Heimburger O, Stenvinkel P et al. Association between residual renal function, inflammation and patient survival in new peritoneal dialysis patients. Nephrol Dial Transplant 2003; 18: 590–597. Qureshi AR, Alvestrand A, Divino-Filho JC et al. Inflammation, malnutrition, and cardiac disease as predictors of mortality in hemodialysis patients. J Am Soc Nephrol 2002; 13(Suppl 1): S28–S36. Johnson DW, Wiggins KJ, Armstrong KA et al. Elevated white cell count at commencement of peritoneal dialysis predicts overall and cardiac mortality. Kidney Int 2005; 67: 738–743. Kazmi WH, Gilbertson DT, Obrador GT et al. Effect of comorbidity on the increased mortality associated with early initiation of dialysis. Am J Kidney Dis 2005; 46: 887–896. Khan SS, Xue JL, Kazmi WH et al. Does predialysis nephrology care influence patient survival after initiation of dialysis? Kidney Int 2005; 67: 1038–1046. Chen W, Guo LJ, Wang T. Extracellular water/intracellular water is a strong predictor of patient survival in incident peritoneal dialysis patients. Blood Purif 2007; 25: 260–266. Jansen MA, Termorshuizen F, Korevaar JC et al. Predictors of survival in anuric peritoneal dialysis patients. Kidney Int 2005; 68: 1199–1205. Paniagua R, Amato D, Vonesh E et al. Effects of increased peritoneal clearances on mortality rates in peritoneal dialysis: ADEMEX, a prospective, randomized, controlled trial. J Am Soc Nephrol 2002; 13: 1307–1320. Mallamaci F, Tripepi G, Cutrupi S et al. Prognostic value of combined use of biomarkers of inflammation, endothelial dysfunction, and myocardiopathy in patients with ESRD. Kidney Int 2005; 67: 2330–2337. Lean˜os-Miranda A, Gardun˜o-Espinoza J, Martı´nez-Garcı´a MC et al. Factores asociados a la falla del catiter de dia´lisis peritoneal en insuficiencia renal cro´nica. Rev Invest Clin 1997; 49: 189–195. Twardowski ZJ. The fast peritoneal equilibration test. Semin Dial 1990; 3: 141–142.

S41