You are on page 1of 8

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/47698662

Lateral lumbar Xray assessment of abdominal


aortic calcification in Australian haemodialysis
patients

Article in Nephrology November 2010


Impact Factor: 2.08 DOI: 10.1111/j.1440-1797.2010.01420.x Source: PubMed

CITATIONS READS

8 317

9 authors, including:

Kenneth Lau Matthew J Damasiewicz


Stanford University Monash University (Australia)
47 PUBLICATIONS 1,149 CITATIONS 22 PUBLICATIONS 79 CITATIONS

SEE PROFILE SEE PROFILE

Peter G Kerr
Monash University (Australia) and Monash
244 PUBLICATIONS 6,287 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: Peter G Kerr
letting you access and read them immediately. Retrieved on: 29 June 2016
Nephrology 16 (2011) 389395

Original Article

Lateral lumbar X-ray assessment of abdominal aortic calcication in


Australian haemodialysis patients nep_1420 389..394

NIGEL D TOUSSAINT, EUGENIE PEDAGOGOS, KENNETH K LAU,2 STEFAN HEINZE,4 GAVIN J BECKER,3
1,3 3

JENNIFER BEAVIS,3 KEVAN R POLKINGHORNE,1 MATTHEW J DAMASIEWICZ1 and PETER G KERR1


Departments of 1Nephrology and 2Radiology, Monash Medical Centre, Clayton, and Departments of 3Nephrology and 4Radiology, The Royal Melbourne
Hospital, Parkville, Victoria, Australia

KEY WORDS: ABSTRACT:


cardiovascular disease, chronic kidney disease,
haemodialysis, lateral lumbar X-ray, mineral Aim: Vascular calcification is prevalent in patients with chronic kidney
metabolism, vascular calcication. disease. Abdominal aortic calcification (AAC) can be detected by X-ray,
although AAC is less well documented in anatomical distribution and sever-
Correspondence: ity compared with coronary calcification. Using simple radiological imaging
Dr Nigel D Toussaint, Department of
we aimed to assess AAC and determine associations in prevalent Australian
Nephrology, Monash Medical Centre, 246
haemodialysis (HD) patients.
Clayton Road, Clayton, Vic. 3168, Australia.
Email: nigel.toussaint@monash.edu Methods: Lateral lumbar X-ray of the abdominal aorta was used to deter-
mine AAC, which is related to the severity of calcific deposits at lumbar
Accepted for publication 25 October 2010. vertebral segments L1 to L4. Two radiologists determined AAC scores, by
Accepted manuscript online 3 November 2010. semi-quantitative measurement using a validated 24-point scale, on HD
patients from seven satellite dialysis centres. Regression analysis was used
doi:10.1111/j.1440-1797.2010.01420.x
to determine associations between AAC and patient characteristics.
Results: Lateral lumbar X-ray was obtained in 132 patients. Median age of
SUMMARY AT A GLANCE patients was 69 years (range 2990), 60% were male, 36% diabetic, median
The main nding of this study is that duration of HD 38 months (range 6230). Calcification (AAC score 1) was
abdominal aortic calcication detected by present in 94.4% with mean AAC score 11.0 1 6.4 (median 12). Independent
lateral lumbar X-ray is highly prevalent in predictors for the presence and severity of calcification were age (P = 0.03),
Australian HD patients and is associated duration of dialysis (P = 0.04) and a history of cardiovascular disease
with CV disease, increasing age and (P = 0.009). There was no significant association between AAC and the pres-
duration of HD. This semi-quantitative
ence of diabetes or time-averaged serum markers of mineral metabolism,
inexpensive method is useful in detecting
vascular calcication and may assist
lipid status and C-reactive protein.
cardiovascular risk stratication in Conclusions: AAC detected by lateral lumbar X-ray is highly prevalent in our
Australian HD patients. cohort of Australian HD patients and is associated with cardiovascular
disease, increasing age and duration of HD. This semi-quantitative method
of determining vascular calcification is widely available and inexpensive and
may assist cardiovascular risk stratification.

Cardiovascular (CV) disease is highly prevalent in patients general population, with advancing age and duration of
with chronic kidney disease (CKD) and a significant cause of dialysis being consistent risk factors.710
mortality in patients with end-stage kidney disease on Experimental and clinical studies support a strong link
dialysis.13 The latest Australian and New Zealand Dialysis between abnormalities of vascular calcification and bone and
and Transplant (ANZDATA) Registry report indicated that mineral metabolism in CKD11 and the clinical entity Chronic
34% of the 15 dialysis deaths per 100 patient years at risk Kidney Disease Mineral and Bone Disorder (CKD-MBD)
were due to CV causes.4 Mineral and bone disorders in CKD was recently defined to better recognize this relationship.
are major risk factors for CV morbidity and mortality and This term encompasses clinical, biochemical and imaging
these disturbances are associated with increased vascular abnormalities of vascular calcification and renal bone disease
calcification.5,6 Many studies have reported increased vascu- and their intimate associations with the outcomes of CV
lar calcification in dialysis patients compared with the disease and mortality.5 Assessment of vascular calcification as

2011 The Authors


Nephrology 2011 Asian Pacic Society of Nephrology 389
ND Toussaint et al.

a marker of CKD-MBD may be beneficial in clinical practice 2010. Analysis of the data presented comprises baseline information
given the high prevalence and functional significance of this for an implementation project to improve the management of CKD-
complication, especially in the dialysis population. Detection MBD in HD patients. Medical charts of HD patients were reviewed to
of vascular calcification may assist risk stratification and determine a brief clinical history and document relevant patient
demographics. Data collected from the ANZDATA Registry were also
guide treatment changes to reduce CV complications.
used to obtain further medical information about individual
Vascular calcification scores derived by computed tomog-
patients. A history of CV disease was defined as either the presence
raphy (CT) have been shown to add incremental prognostic of documented coronary artery disease, peripheral vascular disease
information to conventional risk factors for the prediction of or cerebrovascular disease as reported by the latest ANZDATA
CV events and mortality in both the general population12,13 report.4 The observational study was approved by the local ethics
and those with CKD.14 Recent international clinical guide- committees.
lines for the management of CKD-MBD produced by KDIGO
(Kidney Disease: Improving Global Outcomes) recommend
lateral abdominal radiographs be used as an alternative to CT
Lateral abdominal radiographs
to detect vascular calcification and suggest that those patients Lateral lumbar X-ray was performed in a standing position using
with the presence of vascular calcification be considered at standard radiographic equipment (Fig. 1). A minimum of 8 cm of
highest CV risk.15 Although CT, with reproducible measure- tissues anterior to the lumbar spine, which would include abdominal
ments, constitutes the gold standard for quantification of aorta, had to be visible: the focus-film distance was 100 cm. Other
vascular calcification, lateral lumbar X-ray images can detect parameters were: 94 kV, 33200 mAs (depending on the body
aortic calcification with good sensitivity and specificity.16 habitus) and the estimated dose of radiation was approximately
15 mGy.
These simple abdominal radiographs are also inexpensive
Abdominal aortic calcification was assessed using a previously
and involve less radiation exposure than CT.
validated 24-point scale.22,23 All subjects were assessed indepen-
The presence of aortic calcification is a risk marker for CV dently by two investigators (KK and SH) blinded to patient demo-
disease and is associated with coronary artery disease, stroke graphics and serum markers. For the 24-point score, calcified
and heart failure in the general population.17,18 Aortic calci- deposits along the anterior and posterior longitudinal walls of the
fication is linked to aortic stiffness, which leads to left ven- abdominal aorta adjacent to each lumbar vertebra from L1 to L4
tricular hypertrophy, myocardial fibrosis and subsequent were assessed using the midpoint of the intervertebral space above
heart failure and sudden death. In the CKD population and below the vertebrae as the boundaries. Calcifications were
reported risk factors for aortic calcification include calcium graded as follows: 0, no aortic calcific deposits; 1, small scattered
phosphate product, age, duration of dialysis, blood pressure, calcific deposits less than one-third of the corresponding length of
smoking and diabetes.19,20 Few studies have examined aortic the vertebral level; 2, medium quantity of calcific deposits about
one-third or more, but less than two-thirds of the corresponding
calcification with simple radiographs in CKD patients,
vertebral length; 3, severe quantity of calcifications of more than
although a Japanese study of 515 dialysis patients assessed
two-thirds or more of the corresponding vertebral lengths. The
the predictive validity of detecting abdominal aortic calcifi- scores, obtained separately for the anterior and posterior walls, result
cation (AAC) using lateral lumbar X-ray and reported an in a range from 0 to 6 for each vertebral level and 0 to 24 for the total
increase in CV mortality more than twofold in patients with score.
AAC compared with those without.21 The aim of the present Intrarater reliability was determined by one radiologist (KK)
study was to assess the prevalence of AAC in a cohort of re-scoring 20 images (blinded by previous AAC scores) and inter-
Australian dialysis patients using lateral abdominal radiogra-
phy and determine patient characteristics significantly asso-
ciated with this complication.
a b

METHODS

Study subjects
Subjects examined in this study were haemodialysis (HD) patients
associated with the Departments of Nephrology at Monash Medical
Centre, Clayton (MMC), and The Royal Melbourne Hospital,
Parkville (RMH), Victoria, Australia. Inclusion criteria were duration
of HD at least 3 months and age greater than 18 years. Patients from
seven different satellite dialysis units associated with MMC and RMH
were included after the introduction of a protocol to recommend HD
patients undertake a lateral lumbar radiograph for CV risk stratifi-
cation according to the KDIGO CKD-MBD guidelines. Of the 335 HD Fig. 1 Lateral lumbar spine X-rays of two patients showing (a) scattered
patients asked to perform an X-ray at MMC and RMH, 132 (39.4%) abdominal aortic calcications (AAC score 13) and (b) diffuse abdominal aortic
completed this investigation between December 2009 and May calcications (AAC score 24).

2011 The Authors


390 Nephrology 2011 Asian Pacic Society of Nephrology
Aortic calcication on lumbar X-ray

rater reliability was determined by double reading of images from a Table 1 Demographic characteristics of patient studied (n = 132)
subgroup of 20 patients assessed by both radiologists.
Median (range) or frequency

Age (years) 69 (2990)


Laboratory values Gender (% male) 59.7
DM (%) 36.3
Biochemical data were obtained using local routine laboratory
CAD (%) 48.4
methods. Serum markers measured were those addressing mineral
PVD (%) 35.9
metabolism, including calcium (corrected), phosphate, intact par-
CVD (%) 15.9
athyroid hormone (PTH) and 25-hydroxy vitamin D, as well as
Smoking (%) 57.0
haemoglobin, albumin, C-reactive protein and lipid profile. Serum
Vintage of HD (months) 38 (6230)
was taken prior to HD during the mid-week session in all patients
Dialysis hours per session (hours) 4 (35)
and was collected over four consecutive months and time-averaged. Vascular access (%)
Total serum calcium was adjusted for albumin levels using the con- AVF 84.9
version factor; corrected calcium = calcium + 0.02 mmol/L (40 Graft 7.9
albumin).24 Catheter 4.8

Approximately 63.4% had a history of cardiovascular disease (either CAD,


Statistical analysis PVD or CVD). Current or former smoker. AVF, arteriovenous stula; CAD,
coronary artery disease; CVD, cerebrovascular disease; DM, diabetes mellitus;
Results are expressed as mean 1 SD, median (and range) or fre-
HD, haemodialysis; PVD, peripheral vascular disease.
quency (as percentage). Intraclass correlation coefficients were
determined to assess intrarater and interrater reliability of AAC
scores. Univariate and multivariate logistic regression was performed
Table 2 Serum markers for cohort (n = 132)
to determine significant associations between AAC severity and
other variables, adjusted for potential confounders. Using the Mean 1 SD or median (range) Normal range
median AAC score of 12, patients with an AAC score of 12 or higher
Haemoglobin (g/L) 115.6 1 13.3 120160
were compared with patients with an AAC score below 12 (refer-
Albumin (g/L) 35.2 1 3.4 3545
ence category, odds ratio (OR) of 1.0). All variables with P-value less
Calcium (mmol/L) 2.23 1 0.19 2.202.60
than 0.10 in the univariate analysis were included in the multivari-
Phosphate (mmol/L) 1.59 1 0.47 0.801.50
ate regression model. Stepwise backward elimination was used, PTH (pmol/L) 26 (0.1317.6) 1.17.7
beginning with the variable with the highest P-value. Inspection of 25(OH) D (nmol/L) 43 (12132) 75250
the change in the adjusted R2 and performance of a likelihood ratio CRP (mg/L) 5.1 (1180.9) 05
test were both used to confirm that deleted factors did not contribute Cholesterol (mmol/L) 3.6 1 1.0 <5.5
to the model. ORs and their 95% confidence intervals were Triglycerides (mmol/L) 1.7 1 1.0 <2.0
reported. A P-value of <0.05 was considered to be statistically sig-
nificant. Intercooled Stata 10.0 (StataCorp, College Station, Texas, Time-averaged for four levels over four consecutive months. Time-averaged
for two levels over four consecutive months. 25(OH) D, 25-hydroxy vitamin D;
USA) was used for all statistical analysis.
CRP, C-reactive protein; PTH, parathyroid hormone.

RESULTS
Of 132 patients who underwent lateral lumbar X-rays, 78 cent of subjects had AAC score 1. No significant gender
were from MMC and 54 from RMH. Demographics and differences were observed. A frequency distribution histo-
clinical characteristics of subjects are presented in Table 1. gram of AAC determined from the lateral lumbar X-rays
Patients were predominantly male (59.7%) with median age using the 24-point measurement scale is shown in Figure 2.
69 years. Thirty-six per cent were diabetic with median There were vertebral fractures noted on X-ray in nine (7%)
dialysis vintage 38 months and 63.4% had CV disease. Labo- patients. The intraclass correlation coefficients assessing
ratory values, including serum markers of CKD-MBD and intrarater and interrater reliability were 0.94 and 0.91
inflammation, are displayed in Table 2. Time-averaged mean respectively.
serum phosphate for the cohort was 1.59 1 0.47 mmol/L and
65.5% of patients had at least one serum phosphate level Factors associated with AAC scores
>1.60 mmol/L over the four consecutive months of testing.
The percentage of HD patients with 25-hydroxy vitamin D On univariate logistic regression AAC was significantly asso-
deficiency or insufficiency was 81.5%, with a median serum ciated with age (P = 0.002) (Fig. 3). There was no significant
25-hydroxy vitamin D level of 43 nmol/L. relationship between AAC and serum markers of mineral
metabolism (phosphate, calcium, PTH, 25-hydroxy vitamin
D), lipids, C-reactive protein or presence of diabetes
Abdominal aortic calcication
(Table 3). There was a positive association between AAC and
The mean AAC score assessed by lateral lumbar radiograph dialysis vintage (P = 0.004). There was also a significant rela-
was 11.0 1 6.4 with a median score of 12. Ninety-four per tionship between AAC and known CV disease (P = 0.003).

2011 The Authors


Nephrology 2011 Asian Pacic Society of Nephrology 391
ND Toussaint et al.

Table 3 Univariate logistic regression analysis between AAC and patient


8

characteristics

Odds ratio 95% CI P-value

Age 1.06 1.021.09 0.002


6

Diabetes 1.07 0.532.19 0.84


Dialysis vintage 1.01 1.001.02 0.004
Per cent

Cardiovascular disease 3.27 1.507.11 0.003


4

Smoking 1.29 0.612.74 0.50


Calcium 0.83 0.145.01 0.83
Phosphate 0.60 0.132.77 0.51
2

PTH 0.99 0.971.00 0.05


25(OH) D 1.00 0.991.02 0.64
CRP 0.99 0.981.01 0.88
Cholesterol 0.75 0.481.00 0.05
0

0 5 10 15 20 25 Triglycerides 0.64 0.411.03 0.06


AAC
Statistically signicant values P < 0.05 in bold, all P < 0.10 in italics. Median
Fig. 2 Frequency distribution of abdominal aortic calcication (AAC) using the AAC used for logistic regression: patients with score 12 were compared with
24-point score from lateral abdominal radiographs. patients with score <12 (reference category, odds ratio of 1.0). 25(OH) D,
25-hydroxy vitamin D; AAC, abdominal aortic calcication; CRP, C-reactive
protein; PTH, parathyroid hormone.
25

Table 4 Multivariable logistic regression analysis showing independent pre-


20

dictors of AAC

Odds ratio 95% CI P-value


15
AAC score

Age (per 1 year increase) 1.042 1.0041.082 0.03


Dialysis vintage (per 1 year increase) 1.012 1.0011.023 0.04
10

Cardiovascular disease (positive 3.237 1.3437.801 0.009


history)

Median AAC used for logistic regression: patients with score 12 were com-
5

pared with patients with score <12 (reference category, odds ratio of 1.0).
PTH, cholesterol and triglycerides were also initially included in the multivari-
0

ate model but were not statistically signicant and excluded after stepwise
20 40 60 80 100
Age (years)
backward elimination. AAC, abdominal aortic calcication; PTH, parathyroid
hormone.
Fig. 3 Abdominal aortic calcication (AAC) scores in individual patients in
relation to their age.
bution and severity of aortic calcification in this population,
in contrast to coronary calcification, has been less well docu-
Multiple logistic regression analysis was used to investigate mented in the literature. Studies report between 40% and
independent predictors of the severity of aortic calcification 95% of patients having aortic calcification depending on the
(divided by median AAC score). Age, duration of dialysis, CV age of patients, presence of comorbidities such as diabetes
disease, PTH, cholesterol and triglycerides were all included and method of detection of vascular calcification.1921,2530 We
initially in the model before stepwise backward elimination report that 94% of subjects on HD in our study had the
(Table 4). Independent predictors for the presence and presence of aortic calcification detected by lateral abdominal
severity of AAC included in the final model were age (OR radiography.
1.042/year (1.004, 1.082), P = 0.03), duration of dialysis (OR Aortic calcification is associated with all-cause and CV
1.012/year (1.001, 1.023), P = 0.04) and the presence of CV mortality among dialysis patients.21,31 The use of CT for mea-
disease (OR 3.237 (1.347, 7.801), P = 0.009). surement of aortic calcification is a highly reliable clinical
research tool, being more sensitive and providing quantita-
tive calcification measurement. However, CT scans are
DISCUSSION
expensive and deliver a substantial dose of radiation. Given
We report the first Australian study to determine the pres- the prognostic significance of vascular calcification, it has
ence and extent of aortic calcification using lateral lumbar been suggested that simple assessments might be substituted
X-ray in patients on dialysis. Vascular calcification is highly for more sophisticated radiological techniques. A CV calcifi-
prevalent in patients on dialysis, although the specific distri- cation index has been recently developed comprising demo-

2011 The Authors


392 Nephrology 2011 Asian Pacic Society of Nephrology
Aortic calcication on lumbar X-ray

graphic information, time on dialysis and simple imaging In the CKD population, age and vintage of dialysis are
procedures that are widely available such as lateral lumbar associated with the presence of aortic calcification in a
X-ray and echocardiogram.32 Few studies, however, have number of previous studies,21,3739 although the association
systematically examined the use of plain X-rays to assess with dialysis duration is less clear with calcification in the
vascular calcifications,33 but these simple radiological images thoracic aorta.37 In our study, we assessed for associations
may be a good initial investigation and can potentially dis- with AAC and found increasing age, longer dialysis vintage
tinguish between intimal and medial calcification.2,34 and a history of CV disease to be significant independent
Aortic calcification detected on lateral lumbar X-ray has factors. This finding is consistent with the CORD study,
been shown to be predictive of heart failure, ischaemic heart which reported the same three independent predictive
disease and stroke as well as overall CV disease incidence and factors for the presence and severity of ACC. In the study by
mortality in the general population and in patients with Bellasi et al. assessing the criterion validity of lateral lumbar
CKD.18,35,36 The scoring system for measuring the AAC score X-ray for determining AAC, 140 prevalent HD patients, with
is a validated method developed from the general popula- a mean age of 55 years and dialysis vintage of 2.7 years, were
tion. The 24-point scale for quantification of aortic calcifica- examined and the mean AAC was 4.4.16 The mean AAC
tion from lateral lumbar X-ray was described by Kauppila score in our study (11.0) was greater and may be explained
et al.22 in a subgroup of participants of the Framingham heart by patients in our study generally having both increased age
study. We demonstrate in a cohort of Australian HD patients and duration of dialysis.
the usefulness of this scoring system showing the high preva- There were no significant associations between AAC and
lence and severity of AAC in this population. Our study serum markers of mineral metabolism including calcium,
examines a representative sample of the Australian HD phosphate, PTH and 25-hydroxy vitamin D levels in our
population as the patient demographics of our cohort are cohort. This lack of association with the biochemical markers
similar to those of the overall Australian HD population of CKD-MBD, despite time-averaged serum levels, may
reported by ANZDATA.4 relate to small patient numbers in our study although there
One European cross-sectional study, the CORD (Calcifica- is inconsistency in the literature regarding this relationship.
tion Outcome in Renal Disease) study, recently reported on Some observational studies have reported a positive correla-
933 dialysis patients showing AAC present on lateral lumbar tion between phosphate, PTH and calcium and vascular cal-
X-rays in 81% of subjects.26 This is the largest radiological cification;8,9,19,20,33 however, many others have not reported
documentation of arterial calcification in dialysis patients to this association10,25,26,37 and reasons for these inconsistencies
date. The higher prevalence of AAC in our study compared are unclear. Increased C-reactive protein20 and the presence
with the CORD study may reflect an older population (mean of diabetes21 have also been associated with aortic calcifica-
age 66.9 vs 61.4 years) and greater percentage of diabetics tion in patients on dialysis but there was no significant asso-
(36.3 vs 22.9%) in our cohort, two factors reported to be ciation with these factors in our study perhaps also related to
associated with increased vascular calcification (although we small sample size.
note that in our study there was no significant association Despite recommendations for determination of vascular
between diabetes and AAC). The CORD study also excluded calcification in dialysis patients as part of clinical practice,15
patients with significant comorbidities and with estimated there is uncertainty as to whether altering the progression of
life expectancy of less than 6 months. Despite the smaller vascular calcification will impact patient outcomes and there
patient numbers compared with the CORD study, the impor- is no clear evidence-based protocol or algorithm for thera-
tance of our study is that the inclusion criteria include all HD peutic strategies after yielding a positive AAC test. The pres-
patients having undertaken an X-ray, not restricted by ence of vascular calcification may be regarded as a
comorbidities or expected lifespan. Therefore, our cohort and complementary component incorporated into the decision
the prevalence of AAC may again be more representative of making of how to individualize treatment of CKD-MBD.
the general HD population in Australia and not a selected Detection of AAC may potentially be useful to improve iden-
study sample. tification of dialysis patients who may benefit from more
There is variable prevalence of AAC on lateral lumbar aggressive treatment of risk factors such as stricter control of
X-ray in the general population depending on the cohort abnormal mineral metabolism and avoidance of exces-
examined. A study of post-menopausal women aged sive exogenous calcium intake or better management of
55 years and over reported 59% had detectable AAC23 and traditional CV risk factors such as glycaemic control and
the Framingham heart study participants (mean age dyslipidaemia.
54 years) had a prevalence of 37% in men and 27% in The reliability of the AAC scoring on lateral lumbar X-ray
women (although at follow up 25 years later the prevalence was assessed in our study and shown to be excellent both
in this cohort was 86% in both genders).22 We are not aware within and between investigators reporting measurement of
of any study looking at the general Australian population to vascular calcification from the images (intraclass correlation
assess the proportion of those with the presence (or degree) coefficients 0.94 and 0.91 respectively). Reliability of scoring
of aortic calcification. AAC from X-ray has also been examined in previous studies

2011 The Authors


Nephrology 2011 Asian Pacic Society of Nephrology 393
ND Toussaint et al.

including the CORD study (n = 933) where double reading of We found that AAC detected by lateral lumbar X-ray is
64 radiographs revealed a correlation coefficient of 0.9.26 highly prevalent in Australian HD patients and is associated
There is an inverse relationship between bone mineral with increasing age, duration of HD and a history of CV
density and CV disease, including vascular calcification, in disease. The semi-quantitative method of determining AAC
patients with CKD.40 Reduced bone mineral density and on plain X-ray is more widely available, has less radiation
increased fracture rates are common in the dialysis popula- and is less expensive than CT and could form part of CV risk
tion41 and a recent Spanish study of 193 prevalent HD stratification in patients on dialysis.
patients reported that vertebral fractures resulted in greater
mortality with a relative risk of 4.8.42 In the latter study,
fractures were associated with increased vascular calcifica- ACKNOWLEDGEMENTS
tion. The presence of a previous fracture is also known to be
The authors would like to thank Brian Livingston
an important risk factor for subsequent fracture. In dialysis
(ANZDATA) for his assistance with obtaining patient infor-
patients Danese et al. reported that a history of any fracture
mation from registry data. NT is the recipient of a National
resulted in a hazard ratio of 8.33 for hip and 7.32 for verte-
Health and Medical Research Council (NHMRC) National
bral fracture and symptomatic vertebral fracture was associ-
Institute of Clinical Studies (NICS) Fellowship. Although this
ated with a more than sevenfold increased risk of subsequent
Fellowship is supported by NHMRC the views expressed
fracture.43 Lateral lumbar X-ray may therefore be helpful to
herein are those of the authors and are not necessarily those
evaluate for the presence of fractures and be a useful indi-
of the NHMRC.
cator of future fracture risk in the dialysis population. In our
study 7% of patients were incidentally noted to have verte-
bral fractures on the lateral abdominal radiograph images. REFERENCES
This prevalence is lower than other reported studies. A
1. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of
single-centre study of Japanese HD patients reported the
cardiovascular disease in chronic renal disease. Am. J. Kidney Dis.
prevalence of fractures was 15% in men and 30% in
1998; 32: S11219.
women44 and in another study the prevalence of vertebral 2. Blacher J, Guerin AP, Pannier B, Marchais SJ, London GM.
fractures was 32% in diabetic and 13% in non-diabetic dialy- Arterial calcifications, arterial stiffness, and cardiovascular risk in
sis patients.45 The reason for a lower prevalence in our study end-stage renal disease. Hypertension 2001; 38: 93842.
is likely because the radiologists were not asked to specifi- 3. London GM. Cardiovascular calcifications in uremic patients:
cally assess for fractures and therefore only moderate verte- Clinical impact on cardiovascular function. J. Am. Soc. Nephrol.
2003; 14: S3059.
bral fractures may have been reported, but those of a mild
4. McDonald S, Excell L, Livingston B. ANZDATA Registry Report 2008.
degree (less than 25% of the vertebral height) may not have
Adelaide: Australian and New Zealand Dialysis and Transplant
been documented. Registry, 2009.
There were a few limitations in our study. First, the study 5. Moe SM, Drueke T, Lameire N, Eknoyan G. Chronic kidney
is cross-sectional and therefore does not directly show how disease-mineral-bone disorder: A new paradigm. Adv. Chronic
detection of AAC on lateral lumbar radiographs predicts inci- Kidney Dis. 2007; 14: 312.
dent CV events in the dialysis population. Also, only 40% of 6. Toussaint ND, Kerr PG. Vascular calcification and arterial stiffness
in chronic kidney disease: Implications and management.
the HD population had undergone lateral lumber X-ray for
Nephrology (Carlton) 2007; 12: 500509.
analysis of the data in this study and therefore with potential
7. Braun J, Oldendorf M, Moshage W, Heidler R, Zeitler E, Luft FC.
selection bias, findings may not necessarily be representative Electron beam computed tomography in the evaluation of cardiac
of all HD patients associated with both institutions. It is calcification in chronic dialysis patients. Am. J. Kidney Dis. 1996;
unclear why there was a low uptake of patients undertaking 27: 394401.
the X-ray over a 6 month period despite the investigation 8. Goodman WG, Goldin J, Kuizon BD et al. Coronary-artery
being recommended to all HD patients by physicians, but we calcification in young adults with end-stage renal disease who are
undergoing dialysis. N. Engl. J. Med. 2000; 342: 147883.
speculate that patients may have either refused or not under-
9. Raggi P, Boulay A, Chasan-Taber S et al. Cardiac calcification in
stood the task, mislaid their requests, or may still be planning
adult hemodialysis patients. A link between end-stage renal
on undertaking the X-ray but not yet performed this. disease and cardiovascular disease? J. Am. Coll. Cardiol. 2002; 39:
However, we believe that included patients were not signifi- 695701.
cantly different from the excluded patients with respect to 10. Mehrotra R, Budoff M, Christenson P et al. Determinants of
demographical and clinical characteristics as shown by simi- coronary artery calcification in diabetics with and without
larities when comparing our study patients to the overall nephropathy. Kidney Int. 2004; 66: 202231.
11. Moe SM. Vascular calcification and renal osteodystrophy
dialysis population at both centres (median age 69 years,
relationship in chronic kidney disease. Eur. J. Clin. Invest. 2006; 36
36% diabetes, 61% male and median duration of dialysis
(Suppl 2): 5162.
38 months). 12. Shaw LJ, Raggi P, Schisterman E, Berman DS, Callister TQ.
In conclusion, aortic calcification is known to be predictive Prognostic value of cardiac risk factors and coronary artery calcium
of CV events and is easily identified by non-invasive testing. screening for all-cause mortality. Radiology 2003; 228: 82633.

2011 The Authors


394 Nephrology 2011 Asian Pacic Society of Nephrology
Aortic calcication on lumbar X-ray

13. LaMonte MJ, FitzGerald SJ, Church TS et al. Coronary artery 29. Fujiu A, Ogawa T, Matsuda N, Ando Y, Nitta K. Aortic arch
calcium score and coronary heart disease events in a large cohort calcification and arterial stiffness are independent factors for
of asymptomatic men and women. Am. J. Epidemiol. 2005; 162: diastolic left ventricular dysfunction in chronic hemodialysis
4219. patients. Circ. J. 2008; 72: 176872.
14. Matsuoka M, Iseki K, Tamashiro M et al. Impact of high coronary 30. London GM, Marchais SJ, Gurin AP, Boutouyrie P, Mtivier F,
artery calcification score (CACS) on survival in patients on chronic de Vernejoul MC. Association of bone activity, calcium load, aortic
hemodialysis. Clin. Exp. Nephrol. 2004; 8: 548. stiffness, and calcifications in ESRD. J. Am. Soc. Nephrol. 2008; 19:
15. Kidney Disease: Improving Global Outcomes (KDIGO) CKD-MBD 182735.
Work Group. KDIGO clinical practice guideline for the diagnosis, 31. Spiegel DM, Raggi P, Smits G et al. Factors associated with
evaluation, prevention, and treatment of Chronic Kidney mortality in patients new to dialysis. Nephrol. Dial. Transplant.
Disease-Mineral and Bone Disorder (CKD-MBD). Kidney Int. 2009; 2007; 22: 356872.
76 (Suppl 113): S1130. 32. Muntner P, Ferromosca E, Bellasi A et al. Development of
16. Bellasi A, Ferramosca E, Muntner P et al. Correlation of simple cardiovascular calcification index using simple imaging tools in
imaging tests and coronary artery calcium measured by computed hemodialysis patients. Nephrol. Dial. Transplant. 2007; 22: 50814.
tomography in hemodialysis patients. Kidney Int. 2006; 70: 33. Goldsmith DJ, Covic A, Sambrook PA et al. Vascular calcification in
16238. long-term haemodialysis patients in a single unit: A retrospective
17. Schousboe JT, Taylor BC, Kiel DP et al. Abdominal aortic analysis. Nephron 1997; 77: 3743.
calcification detected on lateral spine images from a bone 34. London GM, Guerin AP, Marchais SJ et al. Arterial media
densitometer predicts incident myocardial infarction or stroke in calcification in end stage renal disease: Impact on all-cause and
older women. J. Bone Miner. Res. 2008; 23: 40916. cardiovascular mortality. Nephrol. Dial. Transplant. 2003; 18:
18. Walsh CR, Cupples LA, Levy D et al. Abdominal aortic calcific 173140.
deposits are associated with increased risk for congestive heart 35. Wilson PW, Kauppila LI, ODonnell CJ et al. Abdominal aortic
failure: The Framingham heart study. Am. Heart J. 2002; 144: calcific deposits are an important predictor of vascular morbidity
7339. and mortality. Circulation 2001; 103: 152934.
19. Shigematsu T, Kono T, Satoh K et al. Phosphate overload 36. van der Meer IM, Bots ML, Hofman A, del Sol AI,
accelerates vascular calcium deposition in end-stage renal disease van der Kuip DA, Witteman JC. Predictive value of noninvasive
patients. Nephrol. Dial. Transplant. 2003; 18: 869. measures of atherosclerosis for incident myocardial infarction: The
20. Yamada K, Fujimoto S, Nishiura R et al. Risk factors of the Rotterdam Study. Circulation 2004; 109: 108994.
progression of abdominal aortic calcification in patients on 37. Moe SM, ONeill KD, Fineberg N et al. Assessment of vascular
chronic haemodialysis. Nephrol. Dial. Transplant. 2007; 22: calcification in ESRD patients sing srpial CT. Nephrol. Dial.
20327. Transplant. 2003; 18: 11528.
21. Okuno S, Ishimura E, Kitakani K et al. Presence of abdominal 38. Taniwaki H, Ishimura E, Tabata T et al. Aortic calcification in
aortic calcification is significantly associated with all-cause and hemodialysis patients with diabetes mellitus. Nephrol. Dial.
cardiovascular mortality in maintenance hemodialysis patients. Transplant. 2005; 20: 24728.
Am. J. Kidney Dis. 2007; 49: 41725. 39. Nguyen PTH, Coche E, Goffin E et al. Prevalence and determinants
22. Kauppila LI, Polak JF, Cupples LA, Hannan MT, Kiel DP, of coronary artery calcification assessed by chest CT in renal
Wilson PW. New indices to classify location, severity and transplant recipients. Am. J. Nephrol. 2007; 27: 32935.
progression of calcific lesions in the abdominal aorta: A 25-year 40. London GM, Marty C, Marchais SJ, Guerin AP, Metivier F,
follow-up study. Atherosclerosis 1997; 132: 24550. de Vernejoul MC. Arterial calcifications and bone
23. Schousboe JT, Wilson KE, Hangartner TN. Detection of aortic histomorphometry in end-stage renal disease. J. Am. Soc. Nephrol.
calcification during vertebral fracture assessment (VFA) compared 2004; 15: 194351.
to digital radiography. PLoS ONE 2007; 2: e715. 41. Jadoul M, Albert JM, Akiba T et al. Incidence and risk factors for
24. Varley H, Gowenlock AH, Bell M. Calcium, magnesium, hip or other bone fractures among hemodialysis patients in the
phosphorus and phosphates. In: Varley H, Gowenlock AH, Bell M Dialysis Outcomes and Practice Patterns Study. Kidney Int. 2006;
(eds). Practical Clinical Biochemistry, 5th edn. London: Heinemann, 70: 135866.
1980; 85077. 42. Rodriguez-Garcia M, Gomez-Alonso C, Naves-Diaz M,
25. Toussaint ND, Lau KK, Strauss BJ, Polkinghorne KR, Kerr PG. Diaz-Lopez JB, Diaz-Corte C, Cannata-Anda JB; Asturias Study
Relationship between vascular calcification, arterial stiffness and Group. Vascular calcifications, vertebral fractures and mortality in
bone mineral density in a cross-sectional study of prevalent haemodialysis patients. Nephrol. Dial. Transplant. 2009; 24: 23946.
Australian haemodialysis patients. Nephrology (Carlton) 2009; 14: 43. Danese MD, Kim J, Doan QV, Dylan M, Griffiths R, Chertow GM.
10512. PTH and the risks for hip, vertebral, and pelvic fractures among
26. Honkanen E, Kauppila L, Wikstrom B et al. Abdominal aortic patients on dialysis. Am. J. Kidney Dis. 2006; 47: 14956.
calcification in dialysis patients: Results of the CORD study. 44. Yamaguchi T, Kanno E, Tsubota J, Shiomi T, Nakai M, Hattori S.
Nephrol. Dial. Transplant. 2008; 23: 400915. Retrospective study on the usefulness of radius and lumbar bone
27. Ogawa T, Ishida H, Matsuda N et al. Simple evaluation of aortic density in the separation of hemodialysis patients with fractures
arch calcification by chest radiography in hemodialysis patients. from those without fractures. Bone 1996; 19: 54955.
Hemodial. Int. 2009; 13: 3016. 45. Inaba M, Okuno S, Kumeda Y, Yamakawa T, Ishimura E,
28. Ogawa T, Ishida H, Akamatsu M et al. Progression of aortic arch Nishizawa Y. Increased incidence of vertebral fracture in older
calcification and all-cause and cardiovascular mortality in chronic female hemodialyzed patients with type 2 diabetes mellitus. Calcif.
hemodialysis patients. Int. Urol. Nephrol. 2010; 42: 18794. Tissue Int. 2005; 76: 25660.

2011 The Authors


Nephrology 2011 Asian Pacic Society of Nephrology 395