www.kisupplements.
org
Summary and comparison of 2017 updated and 2009
KDIGO CKD-MBD recommendations
2017 revised KDIGO CKD-MBD
recommendations 2009 KDIGO CKD-MBD recommendations Brief rationale for updating
3.2.1. In patients with CKD G3a–G5D with 3.2.2. In patients with CKD G3a–G5D with evidence Multiple new prospective studies have
evidence of CKD-MBD and/or risk factors for of CKD-MBD, we suggest that BMD testing not be documented that lower DXA BMD predicts
osteoporosis, we suggest BMD testing to performed routinely, because BMD does not incident fractures in patients with CKD G3a–
assess fracture risk if results will impact predict fracture risk as it does in the general G5D. The order of these first 2
treatment decisions (2B). population, and BMD does not predict the type of recommendations was changed because a
renal osteodystrophy (2B). DXA BMD result might impact the decision to
perform a bone biopsy.
3.2.2. In patients with CKD G3a–G5D, it is 3.2.1. In patients with CKD G3a–G5D, it is The primary motivation for this revision was
reasonable to perform a bone biopsy if reasonable to perform a bone biopsy in various the growing experience with osteoporosis
knowledge of the type of renal osteodystrophy settings including, but not limited to: unexplained medications in patients with CKD, low BMD,
will impact treatment decisions (Not Graded). fractures, persistent bone pain, unexplained and a high risk of fracture. The inability to
hypercalcemia, unexplained hypophosphatemia, perform a bone biopsy may not justify
possible aluminum toxicity, and prior to therapy withholding antiresorptive therapy from
with bisphosphonates in patients with CKD-MBD patients at high risk of fracture.
(Not Graded).
4.1.1. In patients with CKD G3a–G5D, This new recommendation was provided in
treatments of CKD-MBD should be based on order to emphasize the complexity and
serial assessments of phosphate, calcium, and interaction of CKD-MBD laboratory parameters.
PTH levels, considered together (Not Graded).
4.1.2. In patients with CKD G3a–G5D, we 4.1.1. In patients with CKD G3a–G5, we suggest There is an absence of data supporting that
suggest lowering elevated phosphate levels maintaining serum phosphate in the normal efforts to maintain phosphate in the normal
toward the normal range (2C). range (2C). In patients with CKD G5D, we suggest range are of benefit to CKD G3a–G4 patients,
lowering elevated phosphate levels toward the including some safety concerns. Treatment
normal range (2C). should aim at overt hyperphosphatemia.
4.1.3. In adult patients with CKD G3a–G5D, we 4.1.2. In patients with CKD G3a–G5D, we suggest Mild and asymptomatic hypocalcemia (e.g., in
suggest avoiding hypercalcemia (2C). maintaining serum calcium in the normal range (2D). the context of calcimimetic treatment) can be
In children with CKD G3a–G5D, we suggest tolerated in order to avoid inappropriate
maintaining serum calcium in the age- calcium loading in adults.
appropriate normal range (2C).
4.1.4. In patients with CKD G5D, we suggest 4.1.3. In patients with CKD G5D, we suggest using Additional studies of better quality are
using a dialysate calcium concentration a dialysate calcium concentration between 1.25 available; however, these do not allow for
between 1.25 and 1.50 mmol/l (2.5 and 3.0 and 1.50 mmol/l (2.5 and 3.0 mEq/l) (2D). discrimination of benefits and harms between
mEq/l) (2C). calcium dialysate concentrations of 1.25 and
1.50 mmol/l (2.5 and 3.0 mEq/l). Hence, the
wording is unchanged, but the evidence grade
is upgraded from 2D to 2C.
4.1.5. In patients with CKD G3a–G5D, decisions 4.1.4. In patients with CKD G3a–G5 (2D) and G5D Emphasizes the perception that early
about phosphate-lowering treatment should (2B), we suggest using phosphate-binding agents “preventive” phosphate-lowering treatment is
be based on progressively or persistently in the treatment of hyperphosphatemia. It is currently not supported by data (see
elevated serum phosphate (Not Graded). reasonable that the choice of phosphate binder Recommendation 4.1.2).
takes into account CKD stage, presence of other The broader term “phosphate-lowering”
components of CKD-MBD, concomitant therapies, treatment is used instead of phosphate
and side effect profile (Not Graded). binding agents since all possible approaches
(i.e., binders, diet, dialysis) can be effective.
(Continued on next page)
Kidney International Supplements (2017) 7, 1–59 19
summary and comparison of 2017 updated and 2009 KDIGO CKD-MBD recommendations www.kisupplements.org
2017 revised KDIGO CKD-MBD
recommendations 2009 KDIGO CKD-MBD recommendations Brief rationale for updating
4.1.6. In adult patients with CKD G3a–G5D 4.1.5. In patients with CKD G3a–G5D and New evidence from 3 RCTs supports a more
receiving phosphate-lowering treatment, we hyperphosphatemia, we recommend restricting general recommendation to restrict calcium-
suggest restricting the dose of calcium-based the dose of calcium-based phosphate binders based phosphate binders in
phosphate binder (2B). In children with CKD and/or the dose of calcitriol or vitamin D analog in hyperphosphatemic patients across all
G3a–G5D, it is reasonable to base the choice of the presence of persistent or recurrent severities of CKD.
phosphate-lowering treatment on serum hypercalcemia (1B).
calcium levels (Not Graded).
In patients with CKD G3a–G5D and
hyperphosphatemia, we suggest restricting the
dose of calcium-based phosphate binders in the
presence of arterial calcification (2C) and/or
adynamic bone disease (2C) and/or if serum PTH
levels are persistently low (2C).
4.1.8. In patients with CKD G3a–G5D, we 4.1.7. In patients with CKD G3a–G5D, we suggest New data on phosphate sources were deemed
suggest limiting dietary phosphate intake in limiting dietary phosphate intake in the treatment to be included as an additional qualifier to the
the treatment of hyperphosphatemia alone or of hyperphosphatemia alone or in combination previous recommendation.
in combination with other treatments (2D). It is with other treatments (2D).
reasonable to consider phosphate source (e.g.,
animal, vegetable, additives) in making dietary
recommendations (Not Graded).
4.2.1. In patients with CKD G3a–G5 not on 4.2.1. In patients with CKD G3a–G5 not on dialysis, The Work Group felt that modest increases in
dialysis, the optimal PTH level is not known. the optimal PTH level is not known. However, we PTH may represent an appropriate adaptive
However, we suggest that patients with levels suggest that patients with levels of intact PTH response to declining kidney function and has
of intact PTH progressively rising or above the upper normal limit of the assay are first revised this statement to include “persistently”
persistently above the upper normal limit for evaluated for hyperphosphatemia, hypocalcemia, above the upper normal PTH level as well as
the assay be evaluated for modifiable factors, and vitamin D deficiency (2C). “progressively rising” PTH levels, rather than
including hyperphosphatemia, hypocalcemia, “above the upper normal limit.” That is,
high phosphate intake, and vitamin D It is reasonable to correct these abnormalities with treatment should not be based on a single
deficiency (2C). any or all of the following: reducing dietary elevated value.
phosphate intake and administering phosphate
binders, calcium supplements, and/or native
vitamin D (Not Graded).
4.2.2. In adult patients with CKD G3a–G5 not on 4.2.2. In patients with CKD G3a–G5 not on dialysis, Recent RCTs of vitamin D analogs failed to
dialysis, we suggest that calcitriol and vitamin D in whom serum PTH is progressively rising and demonstrate improvements in clinically
analogs not be routinely used. (2C) It is remains persistently above the upper limit of relevant outcomes but demonstrated
reasonable to reserve the use of calcitriol and normal for the assay despite correction of increased risk of hypercalcemia.
vitamin D analogs for patients with CKD G4–G5 modifiable factors, we suggest treatment with
with severe and progressive calcitriol or vitamin D analogs (2C).
hyperparathyroidism (Not Graded).
In children, calcitriol and vitamin D analogs
may be considered to maintain serum calcium
levels in the age-appropriate normal range
(Not Graded).
4.2.4. In patients with CKD G5D requiring PTH- 4.2.4. In patients with CKD G5D and elevated or This recommendation originally had not been
lowering therapy, we suggest calcimimetics, rising PTH, we suggest calcitriol, or vitamin D suggested for updating by the KDIGO
calcitriol, or vitamin D analogs, or a analogs, or calcimimetics, or a combination of Controversies Conference in 2013. However,
combination of calcimimetics with calcitriol or calcimimetics and calcitriol or vitamin D analogs due to a subsequent series of secondary and
vitamin D analogs (2B). be used to lower PTH (2B). post hoc publications of the EVOLVE trial, the
It is reasonable that the initial drug selection for Work Group decided to reevaluate
the treatment of elevated PTH be based on Recommendation 4.2.4 as well. Although
serum calcium and phosphate levels and other EVOLVE did not meet its primary endpoint, the
aspects of CKD-MBD (Not Graded). majority of the Work Group members were
It is reasonable that calcium or non-calcium-based reluctant to exclude potential benefits of
phosphate binder dosage be adjusted so that calcimimetics for G5D patients based on
treatments to control PTH do not compromise subsequent prespecified analyses. The Work
levels of phosphate and calcium (Not Graded). Group, however, decided not to prioritize any
We recommend that, in patients with hyper- PTH-lowering treatment at this time because
calcemia, calcitriol or another vitamin D sterol calcimimetics, calcitriol, or vitamin D analogs
be reduced or stopped (1B). are all acceptable first-line options in G5D
patients.
20 Kidney International Supplements (2017) 7, 1–59
www.kisupplements.org summary and comparison of 2017 updated and 2009 KDIGO CKD-MBD recommendations
2017 revised KDIGO CKD-MBD
recommendations 2009 KDIGO CKD-MBD recommendations Brief rationale for updating
We suggest that, in patients with hyper-
phosphatemia, calcitriol or another vitamin D
sterol be reduced or stopped (2D).
We suggest that, in patients with hypocalcemia,
calcimimetics be reduced or stopped depend-
ing on severity, concomitant medications, and
clinical signs and symptoms (2D).
We suggest that, if the intact PTH levels fall
below 2 times the upper limit of normal for the
assay, calcitriol, vitamin D analogs, and/or cal-
cimimetics be reduced or stopped (2C).
4.3.3. In patients with CKD G3a–G5D with 4.3.3. In patients with CKD G3a–G3b with Recommendation 3.2.2 now addresses the
biochemical abnormalities of CKD-MBD and biochemical abnormalities of CKD-MBD and low indications for a bone biopsy prior to
low BMD and/or fragility fractures, we suggest BMD and/or fragility fractures, we suggest that antiresorptive and other osteoporosis
that treatment choices take into account the treatment choices take into account the therapies. Therefore, 2009 Recommendation
magnitude and reversibility of the biochemical magnitude and reversibility of the biochemical 4.3.4 has been removed and 2017
abnormalities and the progression of CKD, abnormalities and the progression of CKD, with Recommendation 4.3.3 is broadened from CKD
with consideration of a bone biopsy (2D). consideration of a bone biopsy (2D). G3a–G3b to CKD G3a–G5D.
4.3.4. In patients with CKD G4–G5D having
biochemical abnormalities of CKD-MBD, and low
BMD and/or fragility fractures, we suggest
additional investigation with bone biopsy prior to
therapy with antiresorptive agents (2C).
5.5. In patients with G1T–G5T with risk factors 5.5. In patients with an estimated glomerular 2009 Recommendations 5.5 and 5.7 were
for osteoporosis, we suggest that BMD testing filtration rate greater than approximately 30 ml/ combined to yield 2017 Recommendation 5.5.
be used to assess fracture risk if results will min/1.73 m2, we suggest measuring BMD in the
alter therapy (2C). first 3 months after kidney transplant if they
receive corticosteroids, or have risk factors for
osteoporosis as in the general population (2D).
5.7. In patients with CKD G4T–G5T, we suggest
that BMD testing not be performed routinely,
because BMD does not predict fracture risk as it
does in the general population and BMD does
not predict the type of kidney transplant bone
disease (2B).
5.6. In patients in the first 12 months after 5.6. In patients in the first 12 months after kidney The second bullet is revised, consistent with
kidney transplant with an estimated transplant with an estimated glomerular filtration the new bone biopsy recommendation (i.e.,
glomerular filtration rate greater than rate greater than approximately 30 ml/min/1.73 2017 Recommendation 3.2.2).
approximately 30 ml/min/1.73 m2 and low m2 and low BMD, we suggest that treatment with
BMD, we suggest that treatment with vitamin vitamin D, calcitriol/alfacalcidol, or
D, calcitriol/alfacalcidol, and/or antiresorptive bisphosphonates be considered (2D).
agents be considered (2D). We suggest that treatment choices be influ-
We suggest that treatment choices be enced by the presence of CKD-MBD, as indi-
influenced by the presence of CKD-MBD, as cated by abnormal levels of calcium, phosphate,
indicated by abnormal levels of calcium, PTH, alkaline phosphatases, and 25(OH)D (2C).
phosphate, PTH, alkaline phosphatases, and It is reasonable to consider a bone biopsy to
25(OH)D (2C). guide treatment, specifically before the use of
It is reasonable to consider a bone biopsy to bisphosphonates due to the high incidence of
guide treatment (Not Graded). adynamic bone disease (Not Graded).
There are insufficient data to guide treatment There are insufficient data to guide treatment after
after the first 12 months. the first 12 months.
25(OH)D, 25-hydroxyvitamin D; BMD, bone mineral density; CKD, chronic kidney disease; CKD-MBD, chronic kidney disease–mineral bone disorder; DXA, dual-energy x-ray
absorptiometry; PTH, parathyroid hormone; RCT, randomized controlled trial.
Changes to above summarized recommendations resulted in renumbering of several adjacent guideline statements. Specifically, 2009 Recommendation 4.1.6 now becomes
2017 Recommendation 4.1.7; 2009 Recommendation 4.1.8 now becomes 2017 Recommendation 4.1.9; 2009 Recommendation 4.3.5 now becomes 2017 Recommendation
4.3.4; and 2009 Recommendation 5.8 now becomes 2017 Recommendation 5.7.
Kidney International Supplements (2017) 7, 1–59 21
chapter 3.2 www.kisupplements.org
Chapter 3.2: Diagnosis of CKD-MBD: bone
3.2.1: In patients with CKD G3a–G5D with evidence of without CKD in the Health, Aging and Body Composition
CKD-MBD and/or risk factors for osteoporosis, we Study, a prospective study of community-living individuals, 70
suggest BMD testing to assess fracture risk if results to 79 years of age at enrollment.13 A total of 587 (21%) of the
will impact treatment decisions (2B). 2754 participants had CKD, and among those, 83% and 13%
had CKD G3a and G3b, respectively. In adjusted analyses, the
Rationale fracture HR for each SD lower femoral neck BMD was 2.14
It is well established that patients with CKD G3a–G5D have (95% CI: 1.80–2.55) in participants without CKD, and 2.69
increased fracture rates compared with the general popula- (95% CI: 1.96–3.69) in those with CKD. Similar results were
tion,2–4 and moreover, incident hip fractures are associated observed for total hip BMD. When limited to hip fractures, the
with substantial morbidity and mortality.5–9 At the time of adjusted femoral neck BMD HRs were 5.82 (95% CI: 3.27–
the 2009 KDIGO CKD-MBD guideline, publications 10.35) among those with CKD and 3.08 (95% CI: 2.29–4.14)
addressing the ability of dual-energy X-ray absorptiometry among those without CKD. Interaction terms demonstrated
(DXA) measures of bone mineral density (BMD) to estimate that the association of BMD with fracture did not differ in those
fracture risk in CKD were limited to cross-sectional studies with versus without CKD. However, the association of femoral
comparing BMD in CKD patients with and without a prev- neck BMD with fracture was significantly less pronounced (test
alent fracture. The results were variable across studies and for interaction, P ¼ 0.04) among those with PTH > 65 pg/ml
across skeletal sites. In light of the lack of evidence that DXA (6.9 pmol/l; HR: 1.56, 95% CI: 0.90–2.70) compared with those
BMD predicted fractures in CKD patients as it does in the with a PTH # 65 pg/ml (6.9 pmol/l; HR: 2.41, 95% CI: 2.04–
general population, and the inability of DXA to indicate the 2.85) in all participants combined. This is noteworthy in light
histological type of bone disease, the 2009 Guideline recom- of the similar pattern observed in dialysis patients, as described
mended that BMD testing not be performed routinely in above.10
patients with CKD G3a to G5D with CKD-MBD. Further- West et al. reported the results of a prospective cohort
more, the lack of clinical trials in patients with low BMD and study of 131 predialysis participants, mean age 62 years, fol-
CKD also limited the enthusiasm for measuring BMD in the lowed up over a 2-year interval.12 At baseline, the proportions
first place. with CKD G3a to G3b, G4, and G5 were 34%, 40%, and 26%,
The current evidence-based review identified 4 prospective respectively. DXA BMD was measured in the total hip, lum-
cohort studies of DXA BMD and incident fractures in adults bar spine, and ultradistal and one-third radius at baseline and
with CKD G3a to G5D (Supplementary Tables S7–S12). 2 years. Low BMD at all sites, and a greater annualized per-
These studies demonstrated that DXA BMD predicted frac- centage decrease in BMD predicted fracture. For example, in
tures across the spectrum from CKD G3a to G5D multivariate models, each SD lower total hip BMD was
(Supplementary Tables S7–S12).10–13 In the earliest study, associated with an odds ratio (OR) of fracture of 1.75 (95%
DXA BMD was measured annually in 485 hemodialysis (HD) CI: 1.30–2.20). The ROC AUC ranged from 0.62 in the spine
patients (mean age: 60 years) in a single center in Japan.10 In to 0.74 in the ultradistal radius in adjusted models.
adjusted Cox proportional analyses, lower baseline femoral Most recently, Naylor, et al.11 assessed the ability of the
neck and total hip BMD predicted a greater risk of fracture; Fracture Risk Assessment Tool (FRAX) to predict a major
for example, the hazard ratio (HR) was 0.65 (95% confidence osteoporotic fracture in 2107 adults $ 40 years of age in the
interval [CI]: 0.47–0.90) for each standard deviation (SD) Canadian Multicenter Osteoporosis Study, including 320 with
higher femoral neck BMD. In receiver operating characteristic an eGFR # 60 ml/min/1.73 m2. Of these, 72% and 24% had
(ROC) analyses stratified according to parathyroid hormone CKD G3a and G3b, respectively. FRAX with BMD, FRAX
(PTH) below or above the median value of 204 pg/ml (21.6 without BMD, and the femoral neck T-score all predicted
pmol/l), the area under the curve (AUC) for femoral neck fractures (AUC: 0.65 to 0.71); the AUC was highest for femoral
BMD was 0.717 in the lower stratum and 0.512 in the higher neck T-score with inclusion of fall history. Importantly, the
stratum. Of note, higher serum bone-specific alkaline phos- AUCs did not differ between those with and without CKD.
phate levels also predicted incident fractures. There is growing evidence that DXA BMD predicts frac-
In the second study, Yenchek et al. assessed whether DXA tures in healthy children and adolescents, and those with
total hip and femoral neck BMD were associated with incident chronic disease.14,15 However, no studies have examined the
nonspine fragility fractures in participants with estimated associations among DXA BMD and fractures in children and
glomerular filtration rate (eGFR) < 60 ml/min/1.73 m2 and adolescents with CKD. In light of the lack of evidence that the
22 Kidney International Supplements (2017) 7, 1–59
www.kisupplements.org chapter 3.2
ability of DXA BMD to predict fracture in children with CKD Research recommendations
is different than in adults, no specific recommendations are RCTs are needed to determine whether interventions based
provided for children. However, it should be noted that on DXA BMD are associated with lower fracture rates, and
children and adolescents with CKD frequently exhibit sub- whether the effects vary based on clinical variables such as
stantial growth failure. Given that DXA measures of areal the baseline PTH level, underlying cause of kidney disease,
BMD (g/cm2) underestimate volumetric BMD (g/cm3) in and CKD GFR category.
children with short stature,16 DXA results should be adjusted Prospective studies are needed to determine whether
for bone size, consistent with the 2013 International Society alternative imaging techniques, such as quantitative CT,
of Clinical Densitometry (ISCD) Pediatric Official Posi- improve fracture prediction in CKD.
tions.17 Prediction equations to adjust DXA results for height Prospective studies are needed in children and adolescents
Z-score are now available,16 and the impact on DXA BMD to determine whether DXA predicts fractures in children
Z-scores in children with CKD is substantial.18 Finally, a single- and to determine whether the ISCD recommendations to
center study in 171 children with CKD G2 to G5D reported measure whole-body and spine BMD in children are the
that lower cortical volumetric BMD in the tibia, as measured appropriate sites in the context of CKD.17 Hip and radius
by peripheral quantitative computed tomography (CT), pre- BMD pediatric reference data are now available and predict
dicted fractures over a 1-year interval (Supplementary incident fractures in healthy children and adolescents.27,28
Tables S7–S12).19 The HR per unit lower cortical BMD
Z-score was 1.75 (95% CI: 1.15–2.67; P < 0.01). 3.2.2: In patients with CKD G3a–G5D, it is reasonable to
The evidence-based review also evaluated clinical trials of perform a bone biopsy if knowledge of the type of
the effects of osteoporosis medications on BMD in CKD G3a renal osteodystrophy will impact treatment decisions
to G5D (Supplementary Tables S1–S6). Prior analyses of large (Not Graded).
randomized clinical trials (RCTs) evaluating medications for
the treatment of postmenopausal osteoporosis (risedronate, Rationale
alendronate, teriparatide, and raloxifene) were described in Renal osteodystrophy is defined as abnormal bone histology
the 2009 Guideline. These trials specifically excluded patients and is 1 component of the bone abnormalities of CKD-
with an elevated serum creatinine, hyperparathyroidism, or MBD.29 Bone biopsy is the gold standard for the diagnosis
abnormal alkaline phosphate levels (i.e., CKD-MBD).20–23 and classification for renal osteodystrophy. As detailed in the
However, post hoc analyses found that these drugs had 2009 KDIGO CKD-MBD Guideline,30 DXA BMD does not
similar efficacy on improving BMD and reducing fracture distinguish among types of renal osteodystrophy, and the
incidence in individuals with moderately reduced eGFR, diagnostic utility of biochemical markers is limited by poor
compared with those with mildly decreased or normal eGFR. sensitivity and specificity. Differences in PTH assays (e.g.,
Three new trials were identified. The denosumab study was intact vs. whole PTH) and reference ranges have contributed
also a post hoc analysis of an RCT in women with post- to differences across studies. Unfortunately, cross-sectional
menopausal osteoporosis and normal PTH levels.24 The studies have provided conflicting information on the use of
analysis demonstrated efficacy of denosumab in decreasing biomarkers to predict underlying bone histology. This is not
fracture risk and increasing BMD in 2817 women with CKD surprising given the short half-lives of most of the circulating
G3a to G3b and 73 with CKD G4. Here, the risk of hypo- biomarkers, and the long (3–6 months) bone remodeling
calcemia associated with denosumab in advanced CKD re- (turnover) cycle.
quires mentioning. The remaining 2 new trials on KDIGO recently led an international consortium to
alendronate25 and raloxifene26 were small studies (<60 par- conduct a cross-sectional retrospective diagnostic study of
ticipants) that did not exclude patients with evidence of CKD- biomarkers (all run in a single laboratory) and bone biopsies
MBD. These studies did not show consistent beneficial effects in 492 dialysis patients.31 The objective was to determine the
on DXA BMD. Generally, a major limitation is the lack of predictive value of PTH (determined by both intact PTH
data on fracture prevention by such therapeutic interventions [iPTH] and whole PTH assays), bone-specific alkaline phos-
in advanced CKD (especially in CKD G5–G5D). phatase (bALP), and amino-terminal propeptide of type 1
In summary, the aforementioned 4 prospective studies procollagen (P1NP) as markers of bone turnover. Although
evaluating BMD testing in adults with CKD represent a iPTH, whole PTH, and bALP levels were associated with bone
substantial advance since the original guideline from 2009. turnover, no biomarker singly or in combination was suffi-
Despite the fact that they were conducted across a spectrum ciently robust to diagnose low, normal, and high bone turn-
of CKD severity, the finding that hip BMD predicted fractures over in an individual patient. The conclusion was in support
was consistent across studies, and 2 studies demonstrated of the 2009 KDIGO Guideline to use trends in PTH rather
associations comparable to those seen in the absence of than absolute “target” values when making decisions as to
CKD.11,13 Based on these insights, if a low or declining BMD whether to start or stop treatments to lower PTH. Table 1
will lead to additional interventions to reduce falls or use provides the sensitivity, specificity, and positive and negative
osteoporosis medications, then BMD assessment is predictive value of PTH in helping clinicians determine
reasonable. therapies, demonstrating the challenges clinicians face. Thus,
Kidney International Supplements (2017) 7, 1–59 23
chapter 3.2 www.kisupplements.org
Table 1 | Utility of KDOQI and KDIGO PTH thresholds for diagnostic decision making
KDOQI* KDIGOD
Sens Spec PPV NPV Sens Spec PPV NPV
Differentiating low-turnover from non–low-turnover bone disease, 69% 61% 72% 58% 66% 65% 73% 57%
or “When do I stop therapy?”
Differentiating high-turnover from non–high-turnover bone disease, 58% 78% 35% 90% 37% 86% 35% 87%
or “When do I start therapy?”
iPTH, intact parathyroid hormone; KDIGO, Kidney Disease: Improving Global Outcomes; KDOQI, Kidney Disease Outcomes Quality Initiative; NPV, negative predictive value;
PPV, positive predictive value; PTH, parathyroid hormone; Sens, sensitivity; Spec, specificity.
*Using serum iPTH < 150 pg/ml (16 pmol/l) for lower and > 300 pg/ml (32 pmol/l) for upper threshold.
þ
Using serum iPTH < 130 pg/ml (14 pmol/l) for lower and > 585 pg/ml (62 pmol/l) for upper threshold (2X and 9X of upper limit of normal for assay).
Reproduced with permission from Sprague SM, Bellorin-Font E, Jorgetti V, et al. Diagnostic accuracy of bone turnover markers and bone histology in patients with CKD treated
by dialysis. Am J Kidney Dis. 2016;67:559–566.
the Work Group encourages the continued use of trends in antiresorptive treatment (denosumab) has proven to be
PTH to guide therapy, and when trends in PTH are incon- effective in CKD G3a to G3b and G4, as discussed in
sistent, a bone biopsy should be considered. Recommendation 3.2.1. The growing experience with osteo-
A bone biopsy should also be considered in patients with porosis medications in patients with CKD increases the
unexplained fractures, refractory hypercalcemia, suspicion of comfort of treating patients with low BMD and a high risk of
osteomalacia, an atypical response to standard therapies for fracture with antiresorptive therapy, although definitive trials
elevated PTH, or progressive decreases in BMD despite are lacking. Furthermore, additional data clearly support that
standard therapy. The goal of a bone biopsy would be to: (i) the incidence of fracture is markedly increased in patients
rule out atypical or unexpected bone pathology; (ii) deter- with CKD, and thus the inability to perform a bone biopsy
mine whether the patient has high- or low-turnover disease, may not justify withholding antiresorptive therapy to patients
which may alter the dose of medications to treat renal at high risk of fracture. Thus, the Work Group voted to
osteodystrophy (e.g., initiate or discontinue calcimimetics, remove the requirement of bone biopsy prior to the use of
calcitriol, or vitamin D analogs); or (iii) identify a minerali- antiresorptive therapy for osteoporosis because the use of
zation defect that would alter treatment (e.g., stop intake of these drugs must be individualized in patients with CKD.
aluminum, or aggressively treat hypophosphatemia or However, it is still prudent that these drugs be used with
vitamin D deficiency). caution and that the underlying renal osteodystrophy be
The 2009 Guideline recommended a bone biopsy prior to addressed first. With regard to efficacy, one may speculate that
antiresorptive therapy in patients with CKD G4 to G5D and antiresorptive therapies confer less benefit in the absence of
evidence of biochemical abnormalities of CKD-MBD, low activated osteoclasts, as is the case in adynamic bone disease.
BMD, and/or fragility fractures. The rationale was that low Moreover, additional side effects such as acute kidney injury
BMD may be due to CKD-MBD (e.g., high PTH) and that may also merit consideration in CKD G3a to G5.
lowering PTH is a safer and more appropriate therapy than an In summary, bone biopsy is the gold standard for the
antiresorptive. In addition, there was concern that assessment of renal osteodystrophy and should be considered
bisphosphonates would induce low-turnover bone disease. in patients in whom the etiology of clinical symptoms and
This was based on a single cross-sectional study in 13 patients biochemical abnormalities is in question, and the results may
with CKD G2 to G4 that were referred for bone biopsy after a lead to changes in therapy. With this statement, the Work
variable duration of bisphosphonate therapy.32 To date, Group is well aware that experience concerning performance
studies in patients with CKD have not definitively demon- and evaluation of bone biopsies is limited in many centers.34
strated that bisphosphonates cause adynamic bone disease. With this in mind, in addition to the growing evidence that
Furthermore, the concerns in patients with CKD are only antiresorptive therapies are effective in patients with CKD
theoretical, as it is well established that antiresorptive medi- G3a to G3b and G4, and the lack of robust evidence that these
cations suppress bone formation rates, even in the absence of medications induce adynamic bone disease, the guideline no
kidney disease. For example, in an RCT of zoledronic acid for longer suggests that a bone biopsy be performed prior to
the treatment of postmenopausal osteoporosis, bALP levels initiation of these medications.
were 59% lower in the zoledronic acid group compared with
the placebo group at 12 months.33 Research recommendation
Despite these limitations, in weighing the risk-benefit ratio Prospective studies of circulating biomarkers are needed to
of bisphosphonate treatment, the 2009 KDIGO Guideline determine whether they can predict changes in bone
suggested a biopsy prior to therapy. Since 2009, an additional histology.
24 Kidney International Supplements (2017) 7, 1–59
