Spectrum of renal bone disease in end-stage renal failure patients not yet on dialysis

Transcription

1 Nephrol Dial Transplant (2003) 18: DOI: /ndt/gfg116 Original Article Spectrum of renal bone disease in end-stage renal failure patients not yet on dialysis Goce B. Spasovski 1, An R. J. Bervoets 2, Geert J. S. Behets 2, Ninoslav Ivanovski 1, Aleksander Sikole 1, Geert Dams 2, Marie-M. Couttenye 2, Marc E. De Broe 2 and Patrick C. D Haese 2 1 Department of Nephrology, Clinical Center Skopje, University of Skopje, Macedonia and 2 Department of Nephrology-Hypertension, University of Antwerp, Belgium Abstract Background. During the last few years the spectrum of renal osteodystrophy (ROD) in dialysis patients has been studied thoroughly and the prevalence of the various types of ROD has changed considerably. Whereas until a decade ago most patients presented with secondary hyperparathyroidism (HPTH), adynamic bone (ABD) has become the most common lesion within the dialysis population over the last few years. Much less is known about the spectrum of ROD in end-stage renal failure (ESRF) patients not yet on dialysis. Methods. Transiliac bone biopsies were taken in an unselected group of 84 ESRF patients (44 male, age 54"12 years) before enrolment in a dialysis programme. All patients were recruited within a time period of 10 months from various centres (ns18) in Macedonia. Calcium carbonate was the only prescribed medication in patients followed up by the outpatient clinic. Results. HPTH was found in only 9% of the patients, whilst ABD appeared to be the most frequent renal bone disease as it was observed in 23% of the cases next to normal bone (38%). A relatively high number of patients (ns10; 12%) fulfilled the criteria of osteomalacia (OM). Mixed osteodystrophy (MX) was diagnosed in 18% of the subjects. There was no significant difference between groups in age, creatinine, or serum and bone strontium and aluminium levels. Patient characteristics associated with ABD included male gender and diabetes, whilst OM was associated with older age ()58 years). Conclusions. In an unselected population of ESRF patients already, 62% of them have an abnormal bone histology. ABD is the most prevalent type of ROD in this population. In the absence of aluminium or strontium accumulation the relatively high prevalence Correspondence and offprint requests to: Patrick C. D Haese, PhD, Department of Nephrology, University Hospital Antwerp, Wilrijkstraat 10, B-2650 Edegem, Belgium. dhaese@uia.ua.ac.be of a low bone turnover as expressed by either normal bone or ABD and OM is striking. Keywords: adynamic bone disease; ESRF; hyperparathyroidism; osteomalacia; predialysis; renal osteodystrophy Introduction The abnormalities in bone histology in patients with chronic renal failure (CRF), known as renal osteodystrophy (ROD), can be observed early in the course of the disease. Patients with mild to moderate degrees of CRF rarely experience symptoms, but recent studies have shown that skeletal changes may occur years before the symptoms arise [1]. Despite the recent advances in the understanding of pathogenic mechanisms and the introduction of new therapeutic regimens, ROD continues to be a diagnostic and therapeutic challenge for the practicing nephrologist [2]. Over the last decades the spectrum of ROD in dialysis patients has been studied thoroughly and the prevalence of the various types of renal bone disease changed over the years [3 5]. The gold standard for diagnosing these different types of ROD remains the histomorphometric analysis of a bone biopsy after double tetracycline labelling. The earliest reports have shown hyperparathyroid bone disease (HPTH) [6] to be the predominant form of ROD. Osteomalacia (OM) developed as a consequence of aluminium-intoxication [7], vitamin D deficiency and the presence of metabolic acidosis in CRF patients not yet on dialysis [8]. Recently, OM has also been associated with strontium accumulation [9]. In recent years, adynamic bone disease (ABD) has become the most prevalent bone lesion within the dialysis population. This type of ROD was first found in association with high bone aluminium (B-Al) levels [10]. Recent reports, however, have shown that the prevalence of ABD has increased up to 60 and 36% for # 2003 European Renal Association European Dialysis and Transplant Association

2 1160 G. B. Spasovski et al. continuous ambulatory peritoneal dialysis (CAPD) and haemodialysis (HD) patients, respectively, regardless of the absence of a distinct bone surface aluminium [3,5,11]. Studies on ROD in pre-dialysis patients are scarce. In the earliest reports the prevalence of ROD in pre-dialysis patients has varied between 40 and 100% [12 15] with predominant HPTH [4,16]. Two Spanish studies reported an increased prevalence of ABD in end-stage renal failure (ESRF) patients not yet in dialysis of 32 and 48%, respectively [17,18]. A recent study in 76 pre-dialysis patients showed a somewhat different pattern. Here the mild mixed (HPTHqOM) type of ROD (MX) was most prevalent in 36%, advanced MX was diagnosed in 29%, and ABD and OM in 12 and 9% of the patients, respectively [19]. It should be noted that in the latter study the degree of renal insufficiency widely differed from patient-topatient. In a recent Asian study in 56 pre-dialysis patients, 91% of the cases presented with ROD [20]. Mild and severe osteitis fibrosa (OF) was found in 36 and 9% and ABD and OM in 24 and 10% of the patients, respectively. A similar distribution pattern was found in the study of Ballanti et al. [21] in 27 predialysis patients with various degrees of renal failure. Here, the prevalence of ABD was 22% whilst OM was diagnosed in 11%, HPTH in 8% and MX in 59% of the patients. The differences in the spectrum of ROD reported by the various study groups can, at least in part, be explained by differences in criteria for patient recruitment (selected vs unselected populations), cohort size, genetic and dietary factors, referral rates and use of phosphate binding agents. In the present study, data on the spectrum of ROD are presented from a pre-dialysis population treated with a single phosphate binder. The purpose of the study was to: (i) describe the distribution of the different types of ROD in an unselected population of patients not yet on dialysis recruited from a single geographic area within a time period of only 10 months, (ii) establish risk factors that might influence the development of the various types of ROD. Subjects and methods Patients Bone biopsies for this prospective cross-sectional study were taken at the Department of Nephrology, Clinical Centre Skopje, Macedonia, from September 1998 to June During that period each hospitalized patient evaluated for enrolment on a maintenance dialysis programme was asked to participate in the study. Ninety per cent (93 out of 103) of the patients agreed to participate in the study and underwent a bone biopsy. Nine subjects were excluded for further evaluation because of an inadequate bone biopsy specimen or serum sample. Hence, complete data were obtained from 84 subjects. There were 44 male and 40 female patients with a mean"sd age of 54.2"12.1 years. Patients were recruited in various medical centres (ns18) in Macedonia. Thirty per cent (25 out of 84) of the patients were seen by the nephrologist for the first time during their hospitalization for bone biopsy in the frame of the present study. Fifty-six per cent (47 out of 84) of the patients were referred to a nephrologist within the last 6 months prior to entering dialysis ( late referral ). The remaining patients were followed by the outpatient clinic. Thirty per cent of the patients had been scheduled for urgent dialysis treatment and had undergone three to nine dialysis sessions before bone biopsy. This short treatment period (1 3 weeks) was considered irrelevant with respect to the evolutionupattern of ROD. In all patients serum samples were taken before the first dialysis session. Calcium carbonate (3 guday) was the only phosphate binder used and was taken by 59 patients (70%). None of the patients received either vitamin D analogues or erythropoietin before commencing dialysis. All patients had a creatinine clearance of -5 mlumin. The causes of renal failure were chronic glomerulonephritis in 13 patients (15%), interstitial kidney disease in 16 (19%), nephrosclerosis in 23 (27%) and a group of 14 patients (17%) with undefined diagnosis. Ten patients (12%) were affected by insulindependent diabetes mellitus. There were three patients with polycystic kidney disease, amyloidosis was noticed in two patients and multicystic disease, obstructive nephropathy and uric nephropathy were diagnosed in one case each. The protocol was approved by the Ethical Committee of the Medical Faculty, University of Skopje, Macedonia. Bone biopsy Transiliac bone biopsies were obtained using a Bordier- Meunier needle with an internal diameter of 5 mm, after double tetracycline labelling with a 14-day interval according to a standardized protocol [6]. The biopsies were performed 3 7 days after the second labelling session. Each specimen was transversally divided in two pieces. The largest part was used for histological examination and the second part was weighed directly after sampling and used for bulk analysis by means of electrothermal atomic absorption [22,23]. Histomorphometric analysis was performed on 5 mm Goldner stained undecalcified, methylmetacrylate embedded bone sections. For the detection and localization of aluminium the sections were stained with Aluminon 1. Unstained sections (7 mm) were used for the evaluation of tetracycline labels by fluorescence microscopy. Histological sections were analysed and quantified using a Leica DMRB microscope equipped with a colour CCD camera and a KS 400 image analysis system. Calibration of image pixel size was performed before each measurement cycle using a calibration grid. Bone area, osteoid area, osteoid perimeter, eroded perimeter and quiescent perimeter were measured by manually tracing the mineralized and osteoid area and marking erosion lacunae on the computer screen after which the system calculated the areas and perimeters. Double labelled perimeter and total perimeter were measured in a similar way on unstained sections. Interlabel distance was measured by tracing the labels, after which the system measured the distances between the labels at regular intervals, perpendicular to the labels. Per section, consecutive fields were analysed. Out of these primary measurements, the following derived parameters were calculated according to standardized procedures [24]. Bone area [BAR (%)]: the area of trabecular bone including both mineralized bone and osteoid, expressed as a percentage of

3 Spectrum of ROD in ESRF the total tissue area. Osteoid area [OAR (%)]: the measured area of osteoid expressed as a percent of total BAR. Osteoid width [OWI (mm)]: the mean width of surface osteoid seams, calculated by dividing the measured osteoid area by the length of the osteoid seams. Osteoid perimeter [OPM (%)]: trabecular bone perimeter occupied by osteoid as a percentage of the total bone perimeter. Eroded perimeter [EPM (%)]: the percentage of trabecular bone perimeter characterized by the presence of scalloped bone resorptive lacunae. Double labelled perimeter [DLPM (%)]: percentage of total endosteal perimeter exhibiting a double fluorescent tetracycline label. Mineral apposition rate [MAR (mmuday)]: the rate by which osteoid is mineralized, calculated as the average distance between midpoints of two consecutive tetracycline labels, divided by the time interval between the labelling periods. Adjusted apposition rate [AJAR (mmuday)]: mineral apposition rate averaged over the entire osteoid perimeter. The latter concept is important because in a steady state and in the absence of OM the adjusted apposition rate is the best estimate available from a biopsy of the mean rate of osteoid apposition. Mineralization lag time [MLT (days)]: mean interval between deposition and mineralization of any infinitesimal volume of matrix, averaged over the entire life span of the osteoid seam. Bone formation rate [BFR (mm 2 umm 2 uday)]: area of bone formed per unit of time, calculated as the product of mineral apposition rate and mineralizing perimeter; the latter is calculated as the sum of doubly labelled plus half of singly labelled perimeter per bone perimeter. Bone histological data as well as dynamic parameters are reported using standardized nomenclature and definitions that have repeatedly been used in previous studies of our and other groups [5,6,24 26]. Quantitative measurements are expressed in two dimensions. Thus, osteoid area [two-dimensional (2D)] corresponds to osteoid volume [three-dimensional (3D)], osteoid perimeter to osteoid surface and osteoid width to osteoid thickness. For conversion of 2D to 3D we refer to Parfitt [25]. Based on the values obtained in normal controls, classification of the various types of ROD was carried out according to the following criteria. Normal histology: osteoid area -12%, no fibrosis, BFR mm 2 umm 2 uday. Mild hyperparathyroidism (HPTH): osteoid area -12%, no fibrosis, BFR )613 mm 2 umm 2 uday. OF: osteoid area -12%, with fibrosis, BFR) 613 mm 2 umm 2 uday. OM: osteoid area )12%, no fibrosis, BFR -97 mm 2 umm 2 uday. ABD: osteoid area -12%, no fibrosis, BFR -97 mm 2 umm 2 uday. Mixed lesion: osteoid area )12%, with fibrosis. The biopsies were examined and classified by three different observers without knowledge of the biochemical and clinical findings. The bone aluminium (B-Al) and bone strontium (B-Sr) concentrations were determined by electrothermal atomic absorption spectrometry with Zeeman background correction (Perkin- Elmer Zeeman 3030; graphite furnace HGA 600) [22,23]. We also determined the calcium concentration in bone (B-Ca) to correct for differences in bone density. Calcium was determined with atomic absorption spectrometry with flame atomization (Perkin-Elmer Model 3110). B-Al and B-Sr levels of dialysis patients are considered not to be associated with aluminium or strontium-related bone disease when -14 mgug wet weight and -60 mgug wet weight, respectively [9,26]. Biochemical parameters Blood samples for the determination of biochemical parameters were obtained prior to the start of the first dialysis session. After immediate centrifugation 3 ml serum samples were stored and shipped to the laboratory in dry test tubes at 808C for the determination of a number of relevant biochemical parameters of bone turnover and aluminium and strontium. Intact parathormone (ipth) and osteocalcin were measured with an immuno radiometric assay (IRMA) (Biosource S.A, Europe). The Bio-Rad Crosslinks kit was used in combination with an in-house developed method for the HPLC measurement of pyridinolines (PYD) and deoxypyridinolines (DPYD) in urine (Bio-Rad Laboratories GmbH, Munchen, Germany) [27]. Calcium (Se-Ca), phosphorus (Se-P) and total alkaline phosphatase (TAP) in serum were measured by standard automated techniques. Aluminium (Se-Al) and strontium (Se-Sr) in serum were measured with the same technique as that used for bone analysis [22,23]. The laboratory threshold values for these parameters in dialysis patients are -30 mgul and -100 mgul, respectively [26,28]. Bone alkaline phosphatase (BAP) was quantified following electrophoretic separation on an agarose gel (ISOPAL-kit) as described previously by Van Hoof et al. [29]. Histomorphometric and (bio)chemical analyses in bone and serum were performed at the Department of Nephrology of the University of Antwerp, Belgium. Statistical analysis Comparison between groups of the various parameters under study in serum and bone was performed by means of Kruskal Wallis, followed by the Mann Whitney rank sum test for further comparison between two separate groups. A comparison of the percentage of patients belonging to categorical variables: sex, diabetes, late vs early referrals and calcium carbonate intake in the various ROD groups was performed by x 2 analysis, followed by Fisher s exact test for comparison between two separate groups. A P value was considered to be significant at a two-tailed level. Results are expressed as median (range) for serum biochemistry and bone histomorphometry. To establish the risk factors that might be associated with the development of a various types of ROD a stepwise logistic regression model was used. Non-significant covariates were eliminated by backward selection. Covariates (binominal or continuous) were considered predictive for a particular type of ROD when the estimated coefficient was significantly different from 0 based on the Wald statistic (P-0.05) and expressed in the form of odds ratios. Age, gender of the patients, diabetes, calcium carbonate intake, lateuearly referral ratio and duration of the follow-up by the nephrologist prior to entering dialysis were included in all models. Statistical analysis was performed using the SPSS 10.0 for Windows statistical software. Results 1161 The distribution of the various types of ROD in our study population, as assessed by histologic and histomorphometric examination of 84 bone biopsies is presented in Figure 1. HPTH was found to have the lowest prevalence, covering only 9% of the cases (HPTH patients all had the mild form; i.e. none of them presented with OF) whilst ABD appeared to be the most frequent renal bone disorder as it was observed in 23% of the study patients. Thirty eight (38) per cent of the patients presented with a normal

4 1162 G. B. Spasovski et al. Fig. 1. Distribution (%) of the various types of ROD of the study population. bone histology. A relatively high number of patients (ns10; 12%) fulfilled the criteria of OM. Bone histomorphometry and histodynamic parameters for the various types of ROD are presented in Table 1. Data in Table 2 represent the median (range) values of the various biochemical markers of bone turnover and a series of other relevant serum and bone parameters for each type of ROD. Here significant differences between groups were noted for ipth, OC, TAP, BAP, bone and serum calcium and ferritin (Se- Fer). Serum creatinine (Se-Cr) and phosphorous levels were not significantly different between the groups. Serum calcium significantly differed (Ps0.049) between groups whilst pair-wise comparison of the various groups indicated serum calcium levels to be significantly higher in the ABD (P-0.05) and significantly lower in the OM group (P-0.05) groups as compared with those of all other groups. Clinical data corresponding with the different bone lesions are presented in Table 3. From a review of case histories, the duration of renal failure before the start Table 1. Bone histomorphometry and histodynamic data of the various ROD groups Parameters Normal values a Normal bone ABD HPTH OM MX Kruskal Wallis BAR (10 60) 17 (10 40) 25 (10 40) 20 (10 30) 23 (10 50) NS OAR (2 14) 3 (1 8) 8 (5 12) 16 (10 32) 18 (8 33) P OWI 7.1 (4.9 12) 5.1 ( ) 9 ( ) 12.2 ( ) 15 (8 31.8) P OPM (17 86) 35 (8 81) 62 (53 94) 82 (68 90) 75 (67 93) P EPM (1 10) 2 (0.1 10) 1 (0.1 10) 3 (1 6) 6 (1 14) NS DLPM (1 15) 0.1 (0 6) 5 (4 20) 0 (0 2) 4 (1 18) P MAR ( ) 0.00 ( ) 1.16 ( ) 0.00 ( ) 0.96 ( ) P AJAR 0.25 ( ) 0.00 ( ) 0.36 ( ) 0.00 ( ) 0.35 ( ) P BFR ( ) 0.0 ( ) 758 ( ) 0.0 ( ) 625 ( ) P MLT (8 5000) 5000 ( ) 24 (10 38) 5000 ( ) 76 (24 414) P OMT (5 5000) 5000 ( ) 7.6 (4.7 11) 5000 ( ) 16 (9.8 28) P BAR, bone area (%); OAR, osteoid area (%); OWI, osteoid width (mm); OPM, osteoid perimeter (%); EPM, eroded perimeter (%); DLPM, double labelled perimeter (%); MAR, mineral apposition rate (mmuday); AJAR, adjusted apposition rate (mmuday); BFR, bone formation rate (mm 2 umm 2 uday); MLT, mineralization lag time (days); OMT, osteoid maturation time (days). All values are expressed as median (range). NS, not significant. a Subjects with normal renal function. Table 2. Biochemical data of the various ROD groups Parameters Normal Normal bone ABD HPTH OM MX Kruskal Wallis values a Se-Cr, mgudl (6 20) 9 (5 19) 10 (5 15) 10 (6 17) 9 (6 16) NS Se-Fer, nguml (25 597) b 205 ( ) 196( ) 493( ) 183 (40 677) P-0.05 IPTH, ngul (17 611) 128 (2 377) c 201 ( ) 376 (55 549) 309 ( ) P-0.01 OC, Uul (7 117) 29 (0 244) 76(11 183) d 37 (28 64) 41 (11 183) P-0.01 TAP, Uul (18 106) e 59 (5 116) f 91 (50 223) 104 (40 126) 116 (23 178) P-0.01 BAP, Uul (7 56) g 19 (1 42) h 42 (20 126) 33 (17 90) 46 (16 115) P-0.01 PYD, nmolul ( ) 131 (12 467) 160 (58 536) 147 (70 363) 163 (48 374) NS DPYD, nmolul (14 279) 23 (2 63) 35 (16 112) 35 (17 74) 42 (12 67) NS Se-Sr, mgul -30 (-100 i ) 38 (11 160) 51 (17 294) 51 (21 150) 47 (14 198) 39 (24 176) NS Se-Al, mgul -2 (-30 i ) 9 (2 23) 9 (4 150) 14 (6 24) 5 (1 18) 9 (3 16) NS Se-Ca, mgudl (3 12) 9.0 ( ) j 8.9 ( ) 6.8 (4.9 11) k 7.8 (4.8 10) P-0.05 Se-P, mgudl ( ) 6.9 ( ) 6.3 ( ) 6.8 (4.8 12) 7.4 ( ) NS B-Sr, mgug -30 (-60 i ) 41 (13 94) 38 (18 80) 48 (17 174) 46 (14 106) 43 (18 132) NS B-Al, mgug -2 (-14 i ) 3 (1 11) 3 (1 27) 3 (2 12) 3 (2 7) 3 (1 39) NS B-Ca, mgug 133 (77 197) 139 (97 196) 151 ( ) 121 (92 148) l 128 (36 262) Ps0.054 Tot. Prot., gul (45 84) 62 (34 82) 66 (56 74) 73 (51 79) 73 (51 82) NS a Subjects with normal renal function. b P-0.05 vs ABD, P-0.01 vs OM. c P-0.01 vs MX. d P-0.05 vs ABD and OM. e P-0.01 vs HPTH and MX. f Ps0.01 vs HPTH and MX. g P-0.01 vs HPTH and MX. h P-0.01 vs HPTH, OM and MX. i ESRF patients. j P-0.01 vs all other groups. k P-0.01 vs all other groups. l P-0.05 vs ABD, P-0.01 vs N. All values are expressed as median (range). NS, not significant.

5 Spectrum of ROD in ESRF 1163 Table 3. Clinical data of the various study groups (median and range) Lesion n Sex a (MuF) Age (years) medianurange Diabetes, n CaCO 3 intake, n (%) Follow-up (months) medianurange Referral time earlyulate ABD 19 13u6 55 (26 72) 6 b 14 (74%) 5 (1 23) 5u14 OM 10 3u7 63 (50 73) c 0 9 (90%) 16 (1 31) 6u4 HPTH 8 7u1 b 48.5 (39 57) 0 6 (75%) 12 (1 79) 4u4 MX 15 5u (33 70) 1 7 (50%) 2.5 (1 27) 6u9 Normal 32 16u16 54 (37 76) 3 23 (74%) 11.5 (1 148) 16u16 a P-0.05 between the groups (Pearson x 2 ). b P-0.05 (within group) Fisher s Exact Test. c P-0.05 vs HPTH and N (Mann Whitney). of dialysis was checked for. Thirty per cent of the patients referred the date of hospitalization for bone biopsy, taken in the frame of the present study, as their first contact with the nephrologist. Another 26% of the patients were followed by the nephrologist up to maximally 6 months before enrolment in the study. Hence, 56% of the patients were considered late referrals. There was a tendency towards a higher incidence of late referral patients in the ABD group (14 vs 5), which, however, was not significant. The groups were not significantly different with respect to calcium carbonate intake. Patients with OM were significantly older as compared with those presenting HPTH. The maleufemale sex ratio was significantly different between the various groups with a significantly higher proportion of males in the HPTH as compared with the other ROD groups. The aetiology of renal failure could not be associated with a particular type of ROD with the exception of diabetic nephropathy. Indeed, diabetes mellitus appeared to predispose to the development of ABD as six out of the 10 diabetic patients of the study population presented with this type of ROD (P-0.01 vs the prevalence of diabetes in other groups). In the assessment of risk factors for the various types of ROD logistic regression analysis identified three significant independent covariates in the development of ABD; i.e. patients gender, diabetes and late referral (Table 4). Although the maleufemale ratio did not differ significantly from the other ROD groups as assessed by Fisher s exact test, risk analysis revealed male patients to be at a higher risk for the development of ABD (P-0.05). Patients with diabetic nephropathy also turned out to be at a higher risk for ABD (Ps0.052), which is also reflected by the significantly higher prevalence of insulin dependent diabetes mellitus in the ABD group compared with the other forms of ROD (P-0.01). Patients who were referred to the nephrologist within the last 6 months prior to hospitalization were more likely to have ABD at the borderline of significance (Ps0.076). Logistic regression analysis identified two significant independent covariates as potential risk factors for the development of OM: age )58 years (this cut-off was shown to have the highest likelihood ratio in the finding of OM); and serum calcium -7.1 mgudl. Thus, OM was much more likely to develop in patients older than 58 years with a low serum calcium (P-0.01), as compared with the other ROD groups. Although the maleufemale gender ratio was significantly higher in patients with HPTH in comparison with the other ROD groups, logistic regression analysis revealed that male gender did not significantly (Ps 0.075) predispose to an increased risk for the development of HPTH, which perhaps, can be ascribed to the limited number of patients presenting this type of ROD. Discussion The histopathology of ROD in patients with ESRF not yet in dialysis is much less known than in dialysis patients. The present study provides a description of the current spectrum of renal bone disease in an unselected group of pre-dialysis patients recruited within a period of 10 months, from a well-defined geographic area (Macedonia) and in the absence of aluminium anduor strontium overload. The majority of patients (38%) presented with a normal bone histology, a finding that considerably differs from the ROD spectra reported previously in non-dialysed renal failure patients [4,17,18,20]. In two recent Spanish [17,18] studies and that of Hutchison et al. [4] none of the patients was reported to have normal bone histology, whilst in a very recent Asian study 8.6% of the patients were diagnosed with this Table 4. Clinicalubiochemical data identified as an independent risk factor for development of various types of ROD Parameter ABD OM HPTH MuF ORs6.3, P-0.05 ORs7.0, P Diabetes ORs6.2, Ps0.052 Lateuearly referral ORs4.1, Ps0.076 Age )58 years ORs6.0, P-0.05 Serum calcium -7.1 ORs6.3, P-0.05 OR, Odds ratio.

6 1164 G. B. Spasovski et al. histological picture [20]. Hamdy et al. found 25% (44 out of 176) of their pre-dialysis patients to present a normal bone histology [1]. Here it should be noted that patients of the latter study still had a relatively high creatinine clearance (mean 35.9 mlumin) [1]. The rather conflicting literature data may to a certain extent be explained by the non-conformity in the histological standards used for classification of the various types of ROD (some authors do not include normal bone for differential diagnosis) and the selective nature of patient recruitment by the various research groups. In the absence of any exposure to aluminium, ABD was found to be the second most prevalent (23%) histological picture next to normal bone. The observed prevalence of ABD is highly comparable with that noted in two other recent studies [20,21]. Interestingly, in agreement with our data, in the study of Shin et al. [20], ABD could also be associated with a relatively high number of late referrals (65.5%). A high prevalence (48 and 32%, respectively) of ABD was also reported by Torres et al. [17,23] and Hernandez et al. [18,24] in two Spanish pre-dialysis populations. To which extent subtle aluminium exposure might have contributed to the development of ABD in the latter populations is not clear. In agreement with two previous reports [18,20], we found a relatively high prevalence of OM (12%, 10 out of 84 patients). This relatively high number of patients with OM was noticed in the absence of any increased exposure to either aluminium or strontium as reflected by the patients low serum and bone levels of these elements. As Macedonia is a rather sunny country vitamin D deficiency is not readily expected. However, determination of both 25-OH vit D 3 and 1a,25-(OH) 2 vit D 3 in a separate group of 22 ESRF patients not yet in dialysis which were treated under the same clinical conditions and lived in the same region, showed both compounds to be in the low-normal range (25-OH vit D 3 : normal range nguml; median value study population, 37.5 nguml, five out of 22 patients -22 nguml and 1a,25-(OH) 2 vit D 3, normal range, pguml; median value study population, 53 pguml, seven out of 22 patients -36 pguml) pointing to a relative vitamin D deficiency in at least a subset of patients. In this context it is worth mentioning that in a recent study carried out in Algeria, being a sunny country also, osteomalacic lesions, diagnosed by the presence of Looser s zones, have been associated with decreased levels of 25-(OH) vit D 3, [not 1a,25-(OH) 2 vit D 3 ] [30]. In support of this statement are the decreased serum calcium levels in the osteomalacic patients of the present study. The older age of the OM patients, probably with less sun exposure and hence lower vitamin D uptake to a certain extent fits with the proposed hypothesis also. To which extent other factors, e.g. acidosis might also have played a role in the development of OM is not clear. As reported previously by others [31], could the diet and nutritional status also contribute to the incidence of OM? In the present study the diet of all patients was quite similar and mainly consisted of dairy products and vegetables. Furthermore, the total protein levels in serum, that to a certain extent reflects the nutritional status, did not differ significantly between groups. In view of this, the role of this factor in the development of OM population seems to be limited in the present study population. The relatively low prevalence of HPTH (9%) is striking and in general is lower than that reported in previous studies in pre-dialysis patients [4]. Differences in dietary habits, sun exposure or genetic predisposition, treatment with calcium carbonate (70% of the patients), are potential explanatory factors. Using logistic regression analysis a series of the clinical data were evaluated as possible risk factors for the development of a particular type of ROD. To the best of our knowledge, this is the first study in which such an approach was used in pre-dialysis patients. Male patients tended to be at a higher risk for development of HPTH (odds ratio 7.0). Literature data obtained in dialysis patients showed opposite results as they indicated females to be at a higher risk for the development of HPTH whilst male gender could be associated with a higher incidence of ABD. Oestrogen deficiency and cessation of ovarian function in older amenorrhoeic women leading to a higher susceptibility to increased bone turnover could be an attractive explanation for these findings. The higher, although not statistically significant, maleufemale ratio noticed in the ABD group of the present study, to a certain extent agrees with the above statement and previous literature data. A tendency towards a longer follow-up by the nephrologist (i.e. period between the first time the diagnosis of CRF was made and bone biopsy) in the HPTH vs the ABD group (mean follow-up periods of 27 and 8 months, respectively) was noted. Moreover the increased prevalence of late referral patients in the ABD group (14u19) was identified as an independent risk factor for the development of this type of bone disease. This observation might contribute to a better insight in the natural course of pre-dialysis ROD and might point towards an evolution from ABD towards HPTH with increasing referral time. Our finding of an association between diabetic nephropathy and ABD confirms and extends observations made in previous reports [18,20]. In the present study as much as 60% of the diabetic patients presented with ABD, a percentage being considerably higher than that reported in previous Asian and Spanish studies reporting values of 44.4 and 23%, respectively. Older age ()58 years) was found to be a risk factor for the development of OM. Reduced phosphorus intake, less sunlight exposure and hence altered vitamin D activityumetabolism might to a certain extent explain this observation, which is further supported by the low serum calcium levels in patients with OM. Our data demonstrate a relatively high prevalence of low bone turnover as expressed by normal bone, ABD and OM in the absence of an increased exposure to aluminium or strontium. The only medication in

7 Spectrum of ROD in ESRF connection with ROD was the use of calcium carbonate for phosphate binding (70% of the patients). As the percentages of patients taking calcium carbonate did not differ between groups, the possible speculation as should the high number of patients with low bone turnover be due to a relative suppression of PTH secretion secondary to the use of calcium-containing phosphate binders can not be withheld. In the absence of these various factors promoting the development of low bone turnover (i.e. aluminium, strontium and calcium carbonate) our data may again contribute to a better insight into the natural course of ROD in CRF. With respect to possible associations between various types of ROD and biochemical parameters the increased serum calcium concentrations in ABD patients and decreased levels in OM patients as compared to the other ROD groups are of particular interest. Measurement of the serum calcium level thus could be helpful for the biochemical differentiation between these two types of low turnover bone disease. The differences in serum calcium levels were also reflected by the bone calcium concentration that was significantly lower in patients of the OM group compared with those having ABD. Patients with OM also showed significantly higher levels of ferritin. Although interferences of iron with bone cell function and physicochemical interactions of the element with bone mineralization have been reported [32], further studies and additional clinical data are required to demonstrate an etiological role for iron in the development of OM. In conclusion, in an unselected group of pre-dialysis patients recruited within a period of 10 months, from a well-defined geographic area (Macedonia) and in the absence of aluminium or strontium overload, 62% of the cases already presented with an abnormal bone histology. ABD is the most prevalent type of ROD in this population. In the absence of aluminium or strontium accumulation the relatively high prevalence of OM (12%) is striking. Patient characteristics associated with ABD included male gender, late referral and diabetes, whilst OM was associated with high patient age ()58 years) and serum calcium levels -7.1 mgudl. Additional work is needed to determine whether even in sunny countries like Macedonia a deficient vitamin D status could at least in part explain the relatively high prevalence of OM and low prevalence of HPTH. Acknowledgement. The authors are grateful to Erik Snelders for expert desk editing. Conflict of interest statement. None declared. References 1. Hamdy NA, Kanis JA, Beneton MN et al. Effect of alfacalcidol on natural course of renal bone disease in mild to moderate renal failure. Br Med J 1995; 310: Malluche H, Faugere MC. Renal bone disease 1990: an unmet challenge for the nephrologist. Kidney Int 1990; 38: Sherrard DJ, Hercz G, Pei Y et al. The spectrum of bone disease in end-stage renal failure an evolving disorder. Kidney Int 1993; 43: Hutchison AJ, Whitehouse RW, Boulton HF et al. Correlation of bone histology with parathyroid hormone, vitamin D3, and radiology in end-stage renal disease. Kidney Int 1993; 44: Couttenye MM, D Haese PC, Van Hoof VO et al. Low serum levels of alkaline phosphatase of bone origin: a good marker of adynamic bone disease in haemodialysis patients. Nephrol Dial Transplant 1996; 11: Salusky IB, Coburn JW, Brill J et al. Bone disease in pediatric patients undergoing dialysis with CAPD or CCPD. Kidney Int 1988; 33: Ward MK, Feest TG, Ellis HA et al. Osteomalacic dialysis osteodystrophy: evidence for a water-borne aetiological agent, probably aluminium. Lancet 1978; 1: Coen G, Manni M, Addari O et al. Metabolic acidosis and osteodystrophic bone disease in predialysis chronic renal failure: effect of calcitriol treatment. Miner Electrolyte Metab 1995; 21: D Haese PC, Schrooten I, Goodman WG et al. Increased bone strontium levels in hemodialysis patients with osteomalacia. Kidney Int 2000; 57: Andress DL, Maloney NA, Endres DB, Sherrard DJ. Aluminum-associated bone disease in chronic renal failure: high prevalence in a long-term dialysis population. J Bone Miner Res 1986; 1: Pei Y, Hercz G, Greenwood C et al. Renal osteodystrophy in diabetic patients. Kidney Int 1993; 44: Ingham JP, Stewart JH, Posen S. Quantitative skeletal histology in untreated end-stage renal failure. Br Med J 1973; 2: Eastwood JB. Quantitative bone histology in 38 patients with advanced renal failure. J Clin Pathol 1982; 35: Cundy T, Hand DJ, Oliver DO et al. Who gets renal bone disease before beginning dialysis? Br Med J 1985; 290: Mora Palma FJ, Ellis HA, Cook DB et al. Osteomalacia in patients with chronic renal failure before dialysis or transplantation. QJMed1983; 52: Dahl E, Nordal KP, Attramadal A et al. Renal osteodystrophy in predialysis patients without stainable bone aluminum. A cross-sectional bone-histomorphometric study. Acta Med Scand 1988; 224: Torres A, Lorenzo V, Hernandez D et al. Bone disease in predialysis, hemodialysis, and CAPD patients: evidence of a better bone response to PTH. Kidney Int 1995; 47: Hernandez D, Concepcion MT, Lorenzo V et al. Adynamic bone disease with negative aluminum staining in predialysis patients: prevalence and evolution after maintenance dialysis. Nephrol Dial Transplant 1994; 9: Coen G, Mazzaferro S, Ballanti P et al. Renal bone disease in 76 patients with varying degrees of predialysis chronic renal failure: a cross-sectional study. Nephrol Dial Transplant 1996; 11: Shin SK, Kim DH, Kim HS et al. Renal osteodystrophy in pre-dialysis patients: ethnic difference? Perit Dial Int 1999; (Suppl 2): S402 S Ballanti P, Coen G, Mazzaferro S et al. Histomorphometric assessment of bone turnover in uraemic patients: comparison between activation frequency and bone formation rate. Histopathology 2001; 38: D Haese PC, Van de Vyver FL, de Wolff FA, De Broe ME. Measurement of aluminum in serum, blood, urine, and tissues of chronic hemodialyzed patients by use of electrothermal atomic absorption spectrometry. Clin Chem 1985; 31: D Haese PC, Van Landeghem GF, Lamberts LV et al. Measurement of strontium in serum, urine, bone, and soft tissues by Zeeman atomic absorption spectrometry. Clin Chem 1997; 43: Parfitt AM, Drezner MK, Glorieux FH et al. Bone histomorphometry: standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 1987; 2:

8 1166 G. B. Spasovski et al. 25. Parfitt AM. Bone histomorphometry: proposed system for standardization of nomenclature, symbols, and units. Calcif Tissue Int 1988; 42: D Haese PC, Couttenye MM, Goodman WG et al. Use of the low-dose desferrioxamine test to diagnose and differentiate between patients with aluminium-related bone disease, increased risk for aluminium toxicity, or aluminium overload. Nephrol Dial Transplant 1995; 10: Van Hoof VO, Couttenye MM, Van Brussel KA et al. Serum (deoxy) pyridiniline crosslinks in dialysis. Bone 1998; 23 [Suppl]: S Schrooten I, Elseviers MM, Lamberts LV et al. Increased serum strontium levels in dialysis patients: an epidemiological survey. Kidney Int 1999; 56: Van Hoof VO, Lepoutre LG, Hoylaerts MF et al. Improved agarose electrophoretic method for separating alkaline phosphatase isoenzymes in serum. Clin Chem 1988; 34: Ghazali A, Fardellone P, Pruna A et al. Is low plasma 25-(OH) vitamin D a major risk factor for hyperparathyroidism and Looser s zones independent of calcitriol? Kidney Int 1999; 55: Bernardino Diaz Lopéz J, Jorgetti V, Caorsi H et al. Epidemiology of renal osteodystrophy in Iberoamerica. Nephrol Dial Transplant 1998; 13 [Suppl 3]: Phelps KR, Vigorita VJ, Bansal M et al. Histochemical demonstration of iron but not aluminum in a case of dialysisassociated osteomalacia. Am J Med 1988; 84: Received for publication: Accepted in revised form: