The overall prevalence of chronic kidney disease (CKD) in the adult population in the United States is 16.8%.1 CKD is associated with multiple complications, and renal osteodystrophy (ROD) is significant among them.2
Secondary hyperparathyroidism (SHPT) is an accompaniment of ROD, resulting from disturbances in the regulation of parathyroid hormone (PTH), calcium (Ca), phosphorus (P), and vitamin D. Levels of vitamin D begin to decrease as early as CKD stage 2, and simultaneously PTH levels begin to increase.3 As CKD progresses, serum levels of phosphate increase, whereas serum levels of calcium decrease,4 which serves as a stimulus to PTH secretion. Hyperphosphatemia seems to be an important risk factor for the development of SHPT; however, SHPT can occur during stage 3 of kidney failure, before the development of hyperphosphatemia.4,5 Impaired calcitriol synthesis is one of the major factors contributing to the development of secondary hyperparathyroidism in patients with CKD.4 ROD is associated with high morbidity and mortality,6-8 as patients with end-stage kidney disease (ESKD) are at increased risk of bone loss and hip fracture.9,10 The incidence of fracture in ESKD is reported to be 1% per year for the hip and about 2.6% for any other fracture,11,12 although the incidence of hip fracture in the general population is only 0.07%–0.22%.13 Duration of kidney replacement therapy, history of kidney transplant, exposure to glucocorticoids, and both very high and low levels of parathyroid hormone (PTH), in addition to general risk factors like older age, female sex, low body weight, postmenopausal status, osteoporosis history, and family history of osteoporosis or history of previous fracture, are among the significant risk factors.14 The quality of life in patients with ROD is primarily affected by musculoskeletal problems such as bone pain, muscle weakness, growth retardation, and skeletal deformity.15
The overall prevalence of chronic kidney disease (CKD) in the adult population in the United States is 16.8%.1 CKD is associated with multiple complications, and renal osteodystrophy (ROD) is significant among them.2 Secondary hyperparathyroidism (SHPT) is an accompaniment of ROD, resulting from disturbances in the regulation of parathyroid hormone (PTH), calcium (Ca), phosphorus (P), and vitamin D. Levels of vitamin D begin to decrease as early as CKD stage 2, and simultaneously PTH levels begin to increase.3 As CKD progresses, serum levels of phosphate increase, whereas serum levels of calcium decrease,4 which serves as a stimulus to PTH secretion. Hyperphosphatemia seems to be an important risk factor for the development of SHPT; however, SHPT can occur during stage 3 of kidney failure, before the development of hyperphosphatemia.4,5 Impaired calcitriol synthesis is one of the major factors contributing to the development of secondary hyperparathyroidism in patients with CKD.4 ROD is associated with high morbidity and mortality,6-8 as patients with end-stage kidney disease (ESKD) are at increased risk of bone loss and hip fracture.9,10 The incidence of fracture in ESKD is reported to be 1% per year for the hip and about 2.6% for any other fracture,11,12 although the incidence of hip fracture in the general population is only 0.07%–0.22%.13 Duration of kidney replacement therapy, history of kidney transplant, exposure to glucocorticoids, and both very high and low levels of parathyroid hormone (PTH), in addition to general risk factors like older age, female sex, low body weight, postmenopausal status, osteoporosis history, and family history of osteoporosis or history of previous fracture, are among the significant risk factors.14 The quality of life in patients with ROD is primarily affected by musculoskeletal problems such as bone pain, muscle weakness, growth retardation, and skeletal deformity.15
ROD among CKD patients can be the result of either increased bone turnover, decreased bone turnover, or a mixture of both. ROD is broadly classified into: osteitis fibrosa (OF), osteomalacia (OM), adynamic bone disease (ABD), and mixed osteodystrophy (MOD).16 Although patients with a mild to moderate degree of CKD rarely experience symptoms, these skeletal changes develop years before symptoms arise.17 As therapy for ROD varies, it is essential to establish the underlying diagnosis before assigning appropriate treatment. At present, bone biopsy is the gold standard for diagnosis of ROD; however, because of its invasive nature and overall complexity, clinicians rely on serum PTH levels for the diagnosis and treatment of ROD among CKD patients.
The purpose of our systematic review was to study the prevalence of and evaluate an association between serum biochemical markers such as Ca, P, PTH, alkaline phosphatase (ALP), bone specific alkaline phosphatase (bsALP), osteocalcin (OC) or vitamin D, and bone histology in order to predict a diagnosis of ROD.
- Top of page
Using the PubMed database, we undertook a systematic review of the literature for studies describing the distribution of ROD among CKD patients on dialysis and CKD patients not on dialysis [search 1: kidney failure, chronic (MeSH) AND renal osteodystrophy (MeSH); search 2: renal osteodystrohy (MeSH) and biochemical markers; search 3: kidney failure, chronic (MeSH) AND biochemical markers AND bone histomorphometry; search 4: dialysis (MeSH) AND biochemical markers AND bone histology].
Of all the retrieved results, we selected articles published between 1985 and 2007 that provided information on bone histology in combination with serum biochemical markers of patients with CKD. Of the 42 articles we reviewed, 3 were in Spanish, 1 was in Japanese, and the rest were in English. Of these 42 articles, we selected 13 studies that had at least 3 of these 7 serum biochemical markers: PTH, Ca, P, ALP, bsALP, OC, and vitamin D. Because serum biochemical markers were expressed either as mean ± standard deviation (SD) or standard error (SE), we converted SE to SD by using the formula SE = SD/√n. A weighted average of the mean and SD were calculated. To calculate the mean, we multiplied the number of patients in each category by the means of the serum biochemical markers and divided by the total number of patients in the study. To calculate the weighted standard deviation, the second power of each standard deviation was added, and then the square root of the sum was taken.
We divided patients into those with high-turnover (HTO) bone disease and those with low-turnover (LTO) bone disease, according to bone histology. All patients with mixed bone disease, mild, moderate, or severe hyperparathyroidism, and OF cystica were included in the highturnover group, whereas patients with ABD and OM were included in the low-turnover group. If there were several categories in each histological type such as mild, moderate, and severe hyperparathyroidism, we calculated the weighted average for serum biochemical markers in each category.
- Top of page
We reviewed 42 articles, of which 26 included dialysis patients and 16 included CKD patients not on dialysis.
TableI summarizes the overall histological distribution of ROD among dialysis patients. Of the patients receiving 1,701 bone biopsies, 41% had hyperparathyroidism, 5% OM, and 33% ABD. The rest were classified as normal bone or mixed bone disease. There was variability among studies in reporting bone histology. Some studies did not report the histological types of all biopsies, and 1 study reported 17 patients with OM among 33 with SHPT.
Table I. Overall histological distribution of ROD among dialysis patients
|Studies (n)||Biopsies (n)||ABD||OM||Normal||MOD||SHPT|
|26||1701||560 (33%)||91 (5%)||89 (5%)||230 (13.5%)||698 (41%)|
TableII summarizes the histological distribution of ROD among dialysis patients in individual studies. The presence of ABD significantly varied in each study, ranging from 4% to 63%. TableIII summarizes the histological distribution of ROD among CKD patients not on dialysis. Of these patients who received 1,316 bone biopsies, 34% had hyperparathyroidism, 19% OM, and 8% ABD. TableIV summarizes the histological distribution of ROD among CKD patients not on dialysis in the individual studies. The presence of ABD significantly varied in each study, ranging from 5.1% to 27%.
Table II. Histological distribution of ROD among dialysis patients in individual studies
|Chazan et al., 199141||57||3 (5.3%)||16 (28.1%)||—||15 (26.3%)||23 (40.3%)|
|DeVita et al., 199220||20||N/A||N/A||Aluminum associated: 2; osteoporosis: 1||5 (1 associated osteomalacia)||11 HPT; 1 OM ± HPT|
|Hercz et al., 199342||259||49.4%||4.2%||Mild 17.7%||6.2%||22.5%|
|Goodman et al., 199443||14 PD||N/A||N/A||3 mild||N/A||11 (78.6%)|
|Torres et al., 199544||HD: 49; CAPD: 32||HD 32% CAPD 48%||N/A||N/A||N/A||HD: 32%; CAPD: 14%|
|Ureña et al., 199633||42||10 (23.8%)||N/A||N/A||N/A||32 (76.2%)|
|Fletcher et al., 199721||73||3 (4%)||1||8||4||57|
|Coen et al., 199845||41||9 (22.0%)||N/A||N/A||9 (22.0%)||23 (56.0%)|
|Forerol et al., 199846||70||15||2 (2.9%)||20 mild||2 (2.9%)||31 (44.3%)|
|Olaizola et al., 199847||20||8 (40%)||N/A||5 mild (25%)||—||7 (35%)|
|Haese et al., 199948||100||35 (35%)||10 (10%)||13 (13%)||21 (21%)||21 (21%)|
|Gerakis et al., 200049||62||14||2 (3.2%)||—||6 (9.7%)||40 (64.5%)|
|Sanchez et al., 200050||57||36||N/A||16 mild||N/A||5 (8.8%)|
|Jarava et al., 200051||73||21 (28%)||2||N/A||5|| |
|Pecovnik Balon & Bren, 200040||30||5 (16%)||N/A||N/A||15||OF 10|
|Changsirikulchai et al., 200052||56||23 (41.1%)||2 (3.6%)||3 (5.4%)||11 (19.6%)||16 (28.6%)|
|Ballanti et al., 200153||37||5 (13.5%)||7 (18.9%)||4 mild (10.8%)||5 (13.5%)||16 (43.3%)|
|Marie et al., 200154||51||LBT: 28||N/A||N/A||N/A||HTO or normal: 23|
|Coen et al., 200255||39||8 (20.5%)||2 (5.1%)||N/A||17 (43.6%)||12 (30.8%)|
|Chu et al., 200356||14||4 (28.6%)||1 (7.1%)||N/A||3 (21.4%)||6 (42.9%)|
|Ng et al., 200457||153||49%||10%||12% mild||8%||21%|
|Gal-Moscovici & Popovtzer, 200527||96||26%||13%||—||21%||40%|
|Coen et al., 200558||104||LTO: 14; adynamic: 7||N/A||1 normal||29||60|
|Coen et al., 200659||38||3 (7.9%)||3 (7.9%)||N/A||11 (28.9%)||21 (55.3%)|
|Van Eps et al., 200760||17||N/A||LTO 4||3 (17.7%)||N/A||HTO: 10|
|Barreto et al., 200761||97||57 (58%)||1 (1%)||3 (3%)||24 (25%)||12 (12%)|
Table III. Histological distribution of ROD among CKD patients not on dialysis
|Studies (n)||Biopsies (n)||ABD||OM||Normal||MOD||SHPT|
|16||1316||112 (8%)||255 (19%)||269 (20%)||244 (18%)||453 (34%)|
Table IV. Histological distribution of ROD among CKD patients not on dialysis in individual studies
|Mora et al., 198362||327||N/A||OM with OF: 112||Neither OM or OF: 40||N/A||OF alone: 175|
|Lindenau et al., 198863||CKD: mild 53, moderate 40, advanced 44||N/A||CKD: mild 12, moderate 3, advanced 4||CKD: mild 5, moderate 9, advanced 5||CKD: mild 24, moderate 21, advanced 29||CKD: mild 12, moderate 7, advanced 6|
|Sellares et al., 199164||53||N/A||7 (13.2%)||1 (1.9%)||7 (13.2%)||Severe: 12; mild: 26|
|Hutchison et al., 199365||30||8 (27%)||2 (7%)||1 (3%)||4 (13%)||OF: 15|
|Sato, 199466||62||N/A||4 (6.5%)||N/A||N/A||OF: 1; mild: 57|
|Hamdy et al., 199517||176||9 (5.1%)||With HPT: 24; OM alone: 1||44 (25%)||N/A||98 (55.7%)|
|Coen et al., 199667||76||9 (11.8%)||7 (9.2%)||10 (13.2%)||MOD: mild 26, advanced 22||2 HP (2.6%)|
|Lafage et al., 199968||16||4 (25%)||2 (12.5%)||7 (43.7%)||N/A||3 OF|
|Ballanti et al., 200153||27||6 (22.2%)||3 (11.1%)||N/A||MOD: advanced 9, mild 7||2 (7.4%)|
|Coen et al., 200269||79||9 (11.4%)||8 (10.1%)||10 (12.6%)||50 (63.3%)||2 (2.5%)|
|Duothat et al., 200370||52||5 (9.6%)||10 (19.2%)||N/A||10 (19.2%)||Severe: 13; mild: 14|
|Spazovski et al., 200371||84||23%||12%||38%||18%||9% mild|
|Bervoets et al., 200372||84||19 (22.6%)||10 (11.9%)||32 (38.1%)||15 (17.9%)||8 (9.5%)|
|Lobao et al., 200473||40 (low densitometry)||52.5%||42.5%||N/A||5%||N/A|
|Lehmann et al., 200574||Stages 3/4: 35||Stages 3/4: 3 (8.6%)||Stages 3/4: 2 (5.7%)||Stages 3/4: 7 normal, 4 mild||Stages 3/4: 3 (8.6%)||Stages 3/4: OF 16 (45.7%)|
TableV presents serum biochemical markers in dialysis patients with ROD (LTO versus HTO). There were a total of 221 patients in the low-turnover and 303 in the high-turnover group. When compared with the LTO group, patients with HTO bone disease had significantly higher levels of ALP, bsALP, OC, and iPTH. There was no significant difference between 2 groups in terms of Ca and P. TableVI presents biochemical markers in CKD patients not on dialysis with ROD (LTO versus HTO). There were a total of 78 patients in the LTO group and 29 patients in the HTO group. When compared with the LTO bone disease group, patients with HTO bone disease had significantly higher bsALP, OC, and iPTH. The level of Ca, P, and ALP were not significantly different between the 2 groups.
Table V. Biochemical markers in dialysis patients with ROD (low turnover versus high turnover)
|References||Patients (n)||Ca (mg/dL)||P (mg/dL)||AP (IU/L)||bsALP||OC (_g/L)||Vit D||iPTH|
| Low turnover|| || || || || || || || |
| Hercz et al., 199342||128||9.56||5.51||73.53||N/A||N/A||N/A||77|
| Urena et al., 199632||10||9.9 ± 1.32||6 ± 1.93||167 ± 75||10.8 ± 4.2||N/A||N/A||128 ± 149|
| Fletcher et al., 199721||3||9.72 ± 0.72||6.1 ± 0.93||161 ± 53||9.9 ± 3.3||N/A||N/A||140 ± 187|
| Coen et al., 199845||9||N/A||N/A||79.31 ± 12.88||10.18 ± 1.37||43.1 ± 54.49||N/A||70.97 ± |
| Gerakis et al., 200049||14||10.3||4.1||75 (54–140)||N/A||N/A||N/A||15 (4–37)|
| Jarava et al., 200051||21||10 ± 0.7||5.4 ± 1.1||123 ± 61||20 ± 10||30 ± 29||N/A||47 ± 43|
| Monier et al., 200154||28||9.3 ± 1.05||6.1 ± 2.11||N/A||19.8 ± 18.25||21.5 ± 2.39||N/A||280|
| Coen et al., 200255||8||10.32 ± 0.63||5.24 ± 0.75||91.52 ± 35.52||12.64 ± 6.99||N/A||9 ± 4.52||101.16 ± |
|Total (n)||221|| || || || || || || |
|Mean ± SD|| ||9.66 ± 2.06||5.5 ± 3.28||86.23 ± 99.49||16.5 ± 22.6||28 ± 66.7||9 ± 4.52||105.28 ± |
|High turnover|| || || || || || || || |
| Hercz et al., 199342||99||9.89 ± 23.87||6.03 ± 0.29||146.5||N/A||N/A||N/A||40.25|
| Ureña et al., 199632||32||10.32 ± 0.92||8.56 ± 1.21||249 ± 161||66.9 ± 63.5||N/A||N/A||753 ± 670|
| Fletcher et al., 199721||29||10 ± 1||6.4 ± 0.75||361 ± 190||43.9 ± 31.5||N/A||N/A||780 ± 455|
| Coen et al., 199845||23||N/A||N/A||729.07 ± 783.5||178.77 ± 268.39||172.09 ± 98||N/A||917.42 ± |
| Gerakis et al., 200049||40||10.2||5.7||183 (95–430)||N/A||N/A||N/A||714|
| Jarava et al., 200051||45||10.01||6.52||334||62.53||N/A||N/A||749.17|
| Monier et al., 200154||23||9 ± .95||7.1 ± 2.39||N/A||35.1 ± 21.05||34.5 ± 24.45||N/A||530|
| Coen et al., 200255||12||10.55 ± 1.27||5.93 ± 1.73||362.2 ± 247.25||91.78 ± 96.64||N/A||13.4 ± 18.85||776.33 ± |
|Total (n)||303|| || || || || || || |
|Mean ± SD|| ||9.97 ± 23.94||6.12 ± 3.28||269 ± 858||71.68 ± 294.68||92.7 ± 101||13.4 ± 18.85||529 ± |
Table VI. Biochemical markers in CKD patients not on dialysis with ROD (low turnover versus high turnover)
|References||Patients (n)||Ca (mg/ dL)||P (mg/dL)||ALP (IU/L)||bsALP||OC (±g/L)||Vit D||iPTH|
|Low turnover|| || || || || || || || |
| Hutchison et al., 199365||10||N/A||4.31 ± 0.34||87 ± 20.1||36.6 ± 3.6||N/A||2.4 ± 0.8||21.6 ± 1.6|
| Coen et al., 199667||16||8.72 ± 1.52||3.94 ± 1.5||120 ± 67.52||N/A||N/A||N/A||130.5 ± 160.24|
| Coen et al., 200659||6||8.76 ± 0.4||4.9 ± 0.56||81.5 ± 28.4||N/A||45.7 ± 23.2||14.3 ± 3.5||97.8 ± 56.4|
| Coen et al., 200269||17||N/A||N/A||123.7 ± 67.47||N/A||N/A||21.78 ± 15.81||3.42 ± 126.61|
| Spasovski et al., 200371||29||8.24||6.86||74.51||23.82||31.75||N/A||213.5|
|Total (n)||78|| || || || || || || |
|Mean ± SD|| ||8.94 ± 1.57||4.83 ± 1.92||104.05 ± 118.14||27.07 ± 11.3||34.17 ± 23.2||22.88 ± 16.38||128.38 ± 169.98|
|High turnover|| || || || || || || || |
| Hutchison et al., 199365||15||N/A||5.7 ± 3.87||123 ± 42.56||62.3 ± 56.7||N/A||4.5 ± 4.72||335 ± 10.37|
| Coen et al., 199667||2||9.15 (9.1–9.2)||7.3 (6.8–7.9)||151.5 (78–225)||N/A||N/A||N/A||897 (355–1440)|
| Coen et al., 200659||3||9.24 ± 0.2||4.7 ± 0.43||117 ± 38.8||N/A||78.8 ± 46||13.6 ± 2||238 ± 33.2|
| Coen et al., 200269||2||N/A||N/A||151.5 (78–225)||N/A||N/A||23 (18–28)||797.5 (155–1440)|
| Spasovski et al., 200371||7||8.9 (6.7–11.4)||6.3 (4.7–8.7)||91 (50–223)||42 (20–126)||76 (11–183)||N/A||201 (38–1790)|
|Total (n)||29|| || || || || || || |
|Mean ± SD|| ||9.02 ± 0.2||5.86 ± 3.89||118.65 ± 57.59||55.86 ± 56.7||76.84 ± 46||7.7 ± 5.12||363.31 ± 34.78|
Figure1 shows the comparison of histological variations in ROD between dialysis patients and CKD patients not on dialysis. ABD is more common in dialysis patients than in CKD patients not on dialysis. Figures2 and 3 show a comparison of all serum biochemical markers in dialysis and CKD patients not on dialysis with LTO and HTO bone disease, respectively.
Figure 3. Comparison of serum biochemical markers in CKD patients not on dialysis with low- and high-turnover bone disease.
Download figure to PowerPoint
- Top of page
ROD comprises a wide spectrum of manifestations that includes a high-turnover state such as hyperparathyroid bone disease and a low-turnover state such as OM and ABD.18 ROD occurs early in the course of CKD and worsens as kidney function declines. Bone disease is common among patients with CKD stage 5, and by the time dialysis is initiated, almost all patients are affected.19
Our goal was to evaluate the prevalence of ROD and establish an association between serum biochemical markers associated with the rate of bone turnover such as PTH, ALP, bsALP, or OC, with underlying bone histology, in order to establish an accurate diagnosis of ROD, which is essential to assign treatment. We found an overall high prevalence of adynamic bone disease—33% and 8% among CKD patients on dialysis and CKD patients not on dialysis, respectively. However, individual studies have found a prevalence of adynamic disease as high as 58% and 52% among dialysis patients and CKD patients not on dialysis, respectively. We found hyperparathyroid bone disease was the most common type of ROD in both CKD patients on dialysis and those not on dialysis. In the dialysis group, PTH, ALP, bsALP, and OC were significantly higher in those with HTO bone disease than in those with LTO bone disease. Similarly, in CKD patients not on dialysis, PTH, bsALP, and OC were significantly higher in those with HTO bone disease than in those with LTO bone disease.
Various methods such as serum biochemical markers, imaging studies, and histopathological studies are currently used to diagnose ROD. Ca, P, PTH, ALP, and bsALP are among the most commonly used serum biochemical markers. Similar to previous studies, we found no significant association between serum Ca or P with rate of bone turnover in ROD.20,21
PTH is most commonly used as a serum biochemical marker to determine the underlying bone disease in CKD patients. Biologically active PTH circulates mostly as a 1–84 amino-acid peptide.22 PTH fragments containing carboxy-terminal parts of the molecule of varying length are also present in circulation. First-generation immunoradiometric assays of so-called intact PTH not only measured full-length PTH (1–84) but also recognized large PTH fragments lacking the amino terminus.22-24 The Nichols assay for the estimation of intact parathyroid hormone (i-PTH) overestimates parathyroid gland function by recognizing both the whole PTH-1-84 molecule (identified as a cyclase-activating PTH-CAP) and N-truncated fragments of PTH-7-84 (identified as a cyclase-inactive PTH-CIP).24 Currently, a third-generation assay, which measures 1-84 PTH, and whole PTH and provides 7-84 PTH,25 manufactured by Scantibodies (San Clemente, Calif.), and the ratio of 1-84/7-84 has shown to be associated with rate of bone turnover.26
Serum levels of PTH can predict the presence and severity of SHPT without correlating with the underlying bone disease.24,27 Levels of iPTH in dialysis patients more than 4 times normal and less than 2 times normal are associated with a greater frequency of HTO and LTO bone disease, respectively.28 Although PTH is a good indicator of bone metabolism, the sensitivity and specificity to diagnose HTO bone disease with levels <500 ng/mL and ABD disease with levels <100 ng/mL are inadequate. 28 Bone biopsy studies among dialysis patients revealed that bone remodeling and response to PTH varies among various racial groups.29,30 In a study of 76 ESKD patients, the majority of African American patients with LTO bone disease had higher serum PTH levels than those of whites with LTO bone disease.29 In our systematic review, although individual patients had variations in the correlation of PTH with underlying bone turnover, at an aggregate level there was a good correlation between the level of PTH and bone turnover among both dialysis and non-dialysis patients.
ALP is a marker of osteoblast-mediated bone formation that does provide useful information in conjunction with PTH measurement.31 The combination of low serum bsALP (≤7 ng/mL) and low serum PTH is very suggestive of LTO bone disease.32 Similarly, an elevated bsALP (>200 ng/mL) alone or in combination with increased serum PTH (>200 ng/mL) has been shown to be highly sensitive and specific for HTO bone disease.33 Although bsALP may provide some advantage, some experts believe that total serum ALP plus PTH levels are adequate to evaluate bone formation.18 In our review, we found good correlation between serum ALP and bone turnover among dialysis patients; however, such correlation was not present among CKD patients not on dialysis.
Bone-specific ALP, a more specific indicator of osteoblastic activity, was not reported in all articles used in our review. However, based on our data, there was a good correlation between bsALP and bone turnover. In dialysis patients with HTO bone disease, the levels of bsALP were almost 4 times higher than those in LTO bone disease. Similarly, in CKD patients not on dialysis with HTO bone disease, the levels of bsALP were almost double than those in LTO bone disease.
The serum level of OC—an indicator of bone formation—refl ects the rate of OC synthesis by osteoblasts. Among biomarkers, which refl ect bone formation, the OC assay is preferred because of its high discriminant power and is better characterized in terms of clinical application.34 In our review, OC was reported in a limited number of articles; however, using the aggregate data, there was a good correlation between level of OC and bone turnover. Among dialysis patients with HTO bone disease, OC levels were almost 3 times greater than those in LTO bone disease. Similarly, in CKD patients not on dialysis with HTO bone disease, the levels of OC were almost twice those of patients with LTO bone disease.
Additionally, as an indicator of osteoblastic activity, other biomarkers such as tartrate-resistant acid phosphatase, serum C-terminal telopeptide of collagen type I, and pyridinolin crosslinks as indicators of the resorption process can be specifically measured to estimate bone turnover and are important for noninvasive diagnosis of ROD.35,36
Radiographic examination of bone can provide important information regarding the presence of hyperparathyroidism such as osteopenia, subperiosteal resorption, and cysts. However, radiological finding are less sensitive and do not determine renal osteodystrophy. Among patients with advance bone disease, plain films may reveal subperiosteal resorption in severe OF or looser zones in severe OM.18 The importance of bone mineral density (BMD) measurement is unclear in patients with ROD; however, a lower BMD has shown to predict fracture risk in dialysis patients.37,38 In CKD patients, distal radius is the preferred site for BMD measurement, as BMD of the spine may be misleading because of aortic calcifications.
The prevalence of osteopenia and/or osteoporosis also increases with a decreasing glomerular filtration rate.39 In a study of patients with CKD, the highest levels of BMD in the lumbar spine, hip, and distal forearm were found in those with a glomerular filtration rate between 70 and 110 mL/min/1.73 m2 (stages 1 and 2 CKD), whereas those with a glomerular filtration rate between 6 and 26 mL/min/1.73 m2 (stage 4 CKD) had the lowest BMD levels. The abnormalities in bone metabolism that might be responsible for the decreased BMD were not characterized in these studies. The presence of osteoporosis is a strong predictor of increased risk for fractures in the general population. Although the importance of BMD and risk reduction in fractures by treatment of osteoporosis has not been established among CKD patients, serial measurement of BMD using dual energy x-ray absorptiometry can identify patients with a continuous decline in BMD and increased risk of fractures.
A combination of serum biochemical markers can predict the underlying rate of bone turnover with more accuracy. A study of 30 chronic hemodialysis patients in whom a bone biopsy was performed in conjunction with assessment of biochemical markers showed that if only PTH was taken into consideration, 36.6% of patients were correctly classified according to their diagnosis. However, if both PTH and bone densitometry were taken into consideration, 46.6% were classified correctly. Considering PTH and radiological changes in clavicular and metacarpal bones such as periosteal, endosteal, and intracortical resorption, 60% of patients were classified correctly.40
In conclusion, we found that on a collective basis serum levels of PTH, AP, bsALP, and osteocalcin are high in highturnover bone disease and low in low-turnover bone disease. Use of a combination of 3 or more markers may determine underlying renal osteodystrophy with more accuracy than individual biochemical markers. However, a bone biopsy should be encouraged in younger patients with ESKD, with PTH levels between 200 and 500 pg/dL, where other biochemical markers are not well defined.
Similar to any systematic review, lack of individual patient data, variability in histological classification within studies, and lack of use or reporting of all biochemical markers were limitations of our review.