NDT Advance Access originally published online on November 7, 2006
Nephrology Dialysis Transplantation 2007 22(2):484-490; doi:10.1093/ndt/gfl621
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Measurement of vascular calcification using CT fistulograms
1Department of Nephrology and 2Department of Radiology, Monash Medical Centre, Clayton, Victoria, Australia
Correspondence and offprint requests to: Dr Nigel D. Toussaint, Department of Nephrology, Monash Medical Centre, 246 Clayton Road, Clayton, Victoria 3168, Australia. Email: Nigel.Toussaint{at}med.monash.edu.au
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Vascular calcification (VC), precipitated by calcium and phosphate imbalance, is a major contributor to cardiovascular disease (CVD) in chronic kidney disease (CKD). Electron-beam computed tomography (EBCT) quantitatively assesses coronary artery calcification (CAC), with VC scores predictive of atherosclerosis and cardiac events in the general and CKD population. EBCT is not readily available but spiral CT can also provide quantitative assessment of the extent of VC. CT fistulograms can be used as initial investigation for arterio-venous fistula (AVF) problems in haemodialysis (HD). The images obtained include thoracic aorta, brachio-cephalic, subclavian and common carotid arteries which allow assessment of the extent of VC in these vessels. No study to date has combined the CT fistulogram with concurrent determination of VC.
Methods. We hypothesize that a single investigation for AVF management may also provide information on VC. We retrospectively analysed CT fistulograms on 28 HD patients determining VC scores (in Hounsfield units) in AVF, subclavian and carotid arteries and aorta. We correlated these scores with patient demographics, serum markers of mineral metabolism (time averaged for the period 6 months prior to CT) and calcium-based phosphate binders.
Results. Patients (60.7% male) had a median age of 59 years and 46.4% were diabetic. The mean duration of dialysis was 17.5 months. CT fistulograms showed predominantly aortic (75% of patients) and subclavian (75%) calcifications, with only 21.4% having carotid VC and minimal VC at the level of AVF. Median VC scores were 619.8 (0–1481.4) for aorta and 521.7 (0–1139.6) for subclavian (scores of >400 indicate severe atherosclerotic disease), but there was no significant correlation with serum markers or duration of HD. Increasing age correlated significantly with greater VC in aortic (R = 0.53, P = 0.003) and subclavian (R = 0.40, P = 0.03) vessels, as well as with the number of VC sites involved. CAC was present in most patients (89.3%) but CAC scores were not able to be determined because of cardiac movement.
Conclusion. Concurrent determination of the degree of calcification in certain vessels may be possible from CT studies assessing AVF structure. VC scores provided by CT fistulograms could contribute to HD patient CVD risk assessment but studies with larger patient numbers are required to determine their relevance.
Keywords: arterio-venous fistula; cardiovascular risk; computed tomography; end-stage renal disease; haemodialysis; vascular calcification
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The leading cause of mortality in patients with chronic kidney disease (CKD) is cardiovascular disease (CVD) and patients on dialysis have a 3–30-fold increase in mortality, depending on the age group examined, compared with the general population [1]. This excess in mortality is not explained by the presence of traditional cardiovascular risk factors such as diabetes, smoking, hypertension and elevated cholesterol levels [2]. Addressing these risk factors in the general population, with aggressive preventative measures to reduce the prevalence and severity of CVD, is crucial. However, a phenomenon of reverse epidemiology has been noted in the CKD population, and in relation to certain risk factors, like hypertension and hypercholesterolaemia, there may be a reduction in the relative risk of death [3].
Vascular calcification (VC) is highly prevalent in dialysis patients and has profound effects on cardiovascular function. Calcium and phosphate imbalance in CKD contribute to VC and CVD, and the calcium x phosphate product (Ca x P) is an independent risk factor for VC and mortality in dialysis [4]. Arterial calcification and arterial stiffness are also independent predictors of all-cause and CVD mortality in CKD [5]. VC is therefore becoming increasingly appreciated as a unique and important CVD risk factor for patients on dialysis.
Studies using electron-beam computed tomography (EBCT) have accurately and quantitatively assessed coronary artery calcification (CAC), with total calcification scores proven to be strongly predictive of coronary artery atherosclerosis and of major future adverse cardiac events in the general population [6–8]. In patients on dialysis, calcification scores are also markedly increased, especially at a younger age, and progress more rapidly, although the prognostic significance of EBCT in this population remains to be determined [9–11]. EBCT is not yet available in Australia, although helical or multi-slice spiral computed tomography (CT) through the aorta and other large arteries is also representative of the extent and severity of VC [12,13]. The determination of VC may have prognostic implications for dialysis patients.
Multi-slice CT scans to assess the anatomy of arterio-venous fistulae (AVF) are useful non-invasive investigational tools for haemodialysis (HD) access difficulties. Recent studies have demonstrated the benefits of CT fistulography, with superior 3D angiographic images, and CT can be considered as an alternative imaging method to digital subtraction angiography (DSA), currently the gold standard investigation for anatomic delineation of stenoses in AVF [14,15] (Figure 1). As part of the imaging process, details of the thoracic aorta, brachiocephalic, bilateral subclavian and common carotid arteries can be obtained, but no study to date has combined this investigation with the concurrent determination of VC in these vessels. We hypothesize that a single investigation for AVF management may also provide information on VC. We report on the associations between calcification scores based on Hounsfield units, determined by CT fistulograms, and patient demographic and mineral metabolism data in a cohort of patients who had undergone investigation for HD access problems.
|
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We retrospectively analysed CT fistulograms in 28 HD patients determining VC scores in AVF, subclavian and carotid arteries, and thoracic aorta. Patients were all undergoing HD at Monash Medical Centre, Clayton, Australia, or satellite dialysis units associated with this institution. All the patients had difficulties with HD access within the month preceding the CT fistulogram. The cohort selected were those patients who had consecutively undergone CT investigation of their AVF between December 2005 and March 2006.
Serum markers analysed were those addressing mineral metabolism, including calcium (corrected), phosphate, Ca x P, intact parathyroid hormone (PTH) and alkaline phosphatase (ALP), as well as haemoglobin, albumin and C-reactive protein (CRP). The laboratory markers were all time-averaged for the 6 months preceding the CT fistulogram and a mean level was determined. (Patients may have undergone between two to six serum levels of the various parameters over that 6-month period). Serum was analysed using Synchron LX 20 Pro autoanalyzer (Beckman Coulter Inc., Fullerton, CA, USA). Total serum calcium was adjusted for albumin levels using the conversion factor; corrected calcium = calcium + 0.02 mmol/l x (40 – albumin) [16]. Intact PTH levels were measured by immunometric assay (Immunolite 1000, Diagnostic Products Corperation, Los Angeles, CA, USA) with a normal range of PTH between 1.1–7.7 pmol/l).
Medical charts were reviewed for clinical history and medications, and supplemented with information obtained from dialysis nurses at the HD centre where the patient attended. Patients were considered to have a history of coronary artery disease if there was previous abnormal cardiac investigation or a history of myocardial infarction or angina. Hypertension was defined as having a documented history of high blood pressure and taking, or having taken, blood pressure lowering agents. Medications were recorded, including calcium-based phosphate binders and cholesterol lowering agents. No patients in our study were taking warfarin.
Contrast enhanced CT scans (CT fistulograms) were performed on patients for assessing AVF dysfunction using GE medical systems Lightspeed 16® multi-slice spiral CT scanner. The scanning parameters were 120 kVp and 40 mAs. The image acquisition was obtained at 0.75 mm thickness. The images were reconstructed back to 2 mm thickness for viewing on the workstation. All patients had fistulae in the upper limbs. The upper limb where there was a fistula was laid by the side of the patient during CT examination, while the contralateral arm was elevated above the head to reduce artefact. The contralateral arm was also used as the site of intravenous contrast administration to reduce artefact. Ultravist 300 of 100 ml was used as the intravenous contrast medium. The scanning range was from the level of mid neck to the level just below the fistula in the upper limb.
Apart from assessing the anatomical information of the AVF, the sites and corresponding Hounsfield units of any VC in the AVF, aorta, subclavian and common carotid arteries were noted by a single radiologist who was blinded to the patient demographic and serum results. VC scoring was performed by the CT software simulating the Agatston method as previously described [17]. Scores were calculated by multiplying the area of each calcified lesion in the various vessels by a weighting factor corresponding to the peak pixel intensity for each lesion to yield lesion-specific VC scores. The results were recorded and tabulated.
Results are expressed as mean ± SD or median (and range). Pearson's correlation was used in order to investigate the relationship between variables. A P-value < 0.05 was considered statistically significant. Stata 8.0 (StataCorp, Texas, USA) was used for all statistical analysis.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patient demographics are shown in Table 1. Patients were predominantly male (60.7%) with a median age of 59 years (range 22–77). Of all the patients, 46% were diabetic and the median duration of dialysis was 8 months (range 1–136). Three patients (10.7%) were undertaking nocturnal home HD (8 h nightly for 7 nights each fortnight). Most patients had a history of hypertension (78.6%), with 67.9% taking anti-hypertensive medication (Table 2). Calcium-containing phosphate binders were also prescribed to the majority of patients (78.6%). Table 3 shows laboratory values including the mean serum markers of mineral metabolism.
|
|
|
Calcification scores for the major vessels analysed are shown in Table 4. CT fistulograms showed predominantly aortic (75% of patients) and subclavian (75%) calcifications (Figure 2), with only 21.4% having carotid VC and minimal VC at the level of AVF. CAC was in fact present in most patients (89.3%) but CAC scores could not be accurately determined because of the natural cardiac movement being present during the image acquisition. The image artefacts from the cardiac movement could be rectified if the ECG gating was applied during scanning. This was not performed since the CT scanning was primarily performed for the AVF assessment.
|
|
Median VC scores were 619.8 (0–1481.4) for aorta and 521.7 (0–1139.6) for subclavian. Calcification scores of >400 indicate severe atherosclerotic disease. There was no correlation between the VC at aorta and subclavian sites (R = 0.18, P = 0.37) (Figure 3). Increasing age correlated significantly with greater VC in aorta (R = 0.53, P = 0.003) and subclavian (R = 0.40, P = 0.03) (Figure 4), as well as the number of VC sites involved. Table 5 reveals the relationship between serum markers and calcification, with Pearson's correlation and P-values demonstrating no statistically significant associations between VC scores and serum markers of mineral metabolism. There was no significant correlation between VC and gender (male gender vs aortic VC R = 0.27, P = 0.16), duration of dialysis or the use of calcium-based phosphate binders. Conventional CVD risk factors of hypertension, diabetes and hypercholesterolaemia (determined by the prescription of cholesterol-lowering statin therapy) also did not correlate with VC in any vessel.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We present a cross-sectional analysis of a cohort of HD patients with observational data comparing calcification scores and potential risk factors for the development of VC. VC may be a manifestation of a repair response to vessel injury in the presence of calcium-phosphate derangement. This hypothesis is attractive since uraemia causes vessel injury, most notably at the endothelial cell level. VC in patients on dialysis, noted radiologically for decades, has been of increasing interest within the renal literature. Of the two co-localizing forms, medial calcification is related to excess calcium and phosphate levels, and has a direct adverse effect on vascular distensibility contributing to arterial stiffness. Intimal calcification is observed as a part of advanced atherosclerosis, is inflammatory, and is associated with cholesterol deposition. Whether VC is medial or intimal can usually not be determined by spiral CT imaging, but it is the medial calcification that predominates in the renal failure population, and is becoming increasingly appreciated as an important risk factor for CVD.
It has been suggested that screening methods to assess the degree of VC could potentially be worthwhile in general clinical practice given the high prevalence and functional significance of these structural changes [18]. Detection using imaging modalities may allow accurate risk stratification and changes in management to address the extent of VC. EBCT is a highly sensitive method to determine VC with very low exposure to ionized radiation, however, cost and availability limit its use. Modern 8- and 16-slice spiral CT scanners have been shown to be effective in evaluating calcification in coronary arteries and the aorta in dialysis patients, and have also been applied to peripheral arteries including the superficial femoral artery [12,13].
CT fistulography accurately depicts stenoses in dysfunctional AVF, with the additional benefit of having the entire arterial and venous aspects of the AVF being demonstrated in a single less invasive examination. Concurrent determination of the degree of calcification in certain vessels is also possible during the CT fistulogram because of advanced multi-slice CT technology, fine slice image data acquisitions and superior 3D reconstruction capabilities. The calcification score based on Hounsfield units can be easily obtained using the workstation software by circling the region of interest giving better overall measurement of the calcification load.
We have demonstrated the use of CT fistulograms to calculate VC scores in vessels, predominantly the subclavian arteries and thoracic aorta. Previous studies of HD patients have shown 72–88% prevalence of VC in coronary arteries, 61–80% in the aorta and 73% in the superficial femoral artery [9–13,19]. Our study is comparable, with 75% of patients having VC at the aortic and 75% at subclavian levels. Unfortunately, due to the CT technique and the retrospective nature of this study, we were unable to determine CAC scores in patients, but 89% had some degree of VC at the coronary artery levels.
Similar to previous studies with spiral CT and VC, we found no association between calcification scores and abnormalities of mineral metabolism such as hypercalcaemia, hyperphosphataemia, raised Ca x P and hyperparathyroidism. Moe et al. [13] demonstrated the use of spiral CT to determine CAC scores and aortic VC on 33 dialysis patients and 38 renal transplant patients showing a significant correlation with increasing age and duration of dialysis for CAC and only increasing age associated with aortic calcification. Sigrist et al. [12] reported on 134 subjects (60 HD, 28 peritoneal dialysis and 46 with CKD) having multi-slice spiral CT scans through the superficial femoral artery, and revealed that more heavily calcified patients were significantly older and predominantly males. In our study, we discovered that the only significant association with greater VC determined by CT fistulograms was increasing age. This risk factor has also prevailed across multiple studies of VC quantification including EBCT and ultrasound [10,11,20].
One limitation to our study was that the population chosen were those who had developed AVF problems requiring further investigation. Whether these patients are more likely to have greater VC, with greater risk of stenotic and thrombotic complications, is not known as we have not compared our group with HD patients with well functioning AVF. AVF problems are more likely to occur in patients who are older, diabetic and with systemic or vascular pathologies [21]. In our cohort, mean serum albumin was relatively low (median 31.4 g/l) and CRP levels quite high (median 22.1 mg/l), perhaps representing that this small population of HD patients has an increased inflammatory status, and therefore may have greater VC than those without AVF problems. There was no significant correlation however, between VC in any vessel and CRP or albumin.
A further limitation is in the measurement of VC. Measuring Hounsfield units only determines the density of the VC, and may not truly reflect the amount of calcification load in arteries. In order to do this, measurement of the area and volume of all calcification is required with complicated calculations, and this becomes very labour intensive. As the calcification needs to be manually delineated, any scoring of VC has a degree of operator dependability.
The current gold standard for measuring VC is ECG-gated EBCT or multi-slice CT. Given that our study retrospectively analysed CT fistulograms, these CT investigations were primarily performed for AVF assessment and were therefore not ECG-gated. Not having comparative imaging modalities, such as EBCT, in our study, is certainly a weakness for determination of the accuracy of diagnostic CT fistulograms to provide potential VC information, and prospective studies comparing CT fistulography with gold standard would be useful. Also, using 100 ml of contrast media may be a disadvantage with CT fistulograms (vs about 50 ml used with DSA for AVF investigation), especially in patients with residual renal function given the potential for nephrotoxicity, although this has not formerly been studied.
Overt CVD is very common at initiation of dialysis, with macrovascular complications caused by atherosclerosis of increased prevalence in CKD. Traditional CVD risk factors are only partly responsible for the incidence of ischaemic heart disease, cardiac failure and sudden death. Perhaps of more importance is arteriosclerosis and arterial stiffening, contributing to left ventricular hypertrophy, predictive of patient survival on dialysis. These structural changes are closely linked to the extensive and progressive medial VC in the CKD population. Like the majority of reported studies assessing VC by various imaging modalities, our study revealed no significant correlations with VC and conventional risk factors, such as smoking, hypercholesterolaemia and diabetes, in any vessel.
Hypertension has previously been reported in one study to be associated with VC [11]. Unfortunately, we were not able to assess the degree of hypertension for each individual patient or calculate mean systolic or diastolic pressures because of the retrospective nature of this study, although there was no correlation between VC and a history of hypertension.
Non-conventional risk factors in CKD patients, such as abnormalities in mineral metabolism, have instead been implicated as a primary cause for VC. Previous studies have linked the presence of VC with hypercalcaemia, hyperphosphataemia and an elevated Ca x P [9,10]. Although the absence of significant correlations with mineral metabolism markers, as in our study, is not a new finding [12,13], the development and progression of VC is likely a multifactorial process and longitudinal studies are required to further clarify the importance of these factors in relation to the degree of calcification.
Although our study did not reveal a significant correlation, the use of calcium-based phosphate binders has previously been reported to be a predisposition for the development of VC [10]. However, we did not have information about the cumulative dosage of phosphate binder administration. We were also not able to correlate calcification findings with the prescription of vitamin D medications, although there have been few published reports to date showing an association with vitamin D and VC [22,23], or with serum vitamin D levels as these were not routinely measured at our institution prior to the study.
The differentiation between intimal and medial calcification, and between arteriosclerosis and atherosclerosis is often not clear-cut, as these conditions are likely inter-related. At present, CT determination to distinguish these disorders and their distribution is limited. In our study, there was less VC in the carotid arteries, compared with aorta and subclavian arteries, perhaps because the carotids are less atherogenic, but no study to date has used spiral CT to look at carotid artery calcification. Ultrasound-based methods have previously been utilized to study VC in the carotid artery [24]. There was also no significant correlation in our study with VC at the level of the carotids, probably because of the small percentage of patients having calcification in this area (21.4%).
Our study presents a new concept with potential dual benefit from the use of CT fistulography which has not previously been published. Unfortunately, our sample size is small, and we were not able to assess certain variables such as lipid status, AVF function or cardiac function. However, determination of VC using spiral CT of AVF is described with similar VC scores to those previously reported in the dialysis population. Also, VC correlated with age similar to previous CT studies. VC scores provided by CT fistulograms may contribute to HD patients’ CVD risk assessment but studies with larger patient numbers are required to determine their relevance. A prospective study evaluating CT fistulograms in comparison with ECG-gated EBCT or multi-slice CT in a larger HD population would be valuable.
Conflict of interest statement. None declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Foley RN, Parfrey PS, Sarnak MJ. (1998) Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 32:S112–S119.[Web of Science][Medline]
- Zoccali C. (2000) Cardiovascular risk in uremic patients—is it fully explained by classical risk factors? Nephrol Dial Transplant 15:454–457.
[Free Full Text] - Kalantar-Zadeh K, Block G, Humphries M, et al. (2003) Reverse epidemiology of cardiovascular risk factors in maintenance dialysis patients. Kidney Int 63:793–808.[CrossRef][Web of Science][Medline]
- Block GA, Hulbert-Shearon TE, Levin NW, et al. (1998) Association of serum phosphorus and calcium x phosphate product with mortality risk in chronic hemodialysis patients: a national study. Am J Kidney Dis 31:607–617.[Web of Science][Medline]
- London G, Guerin A, Marchais S, et al. (2003) Arterial media calcification in end-stage renal disease: impact on all-cause and cardiovascular mortality. Nephrol Dial Transplant 18:1731–1740.
[Abstract/Free Full Text] - Wong ND, Hsu JC, Detrano RC, et al. (2000) Coronary artery calcium evaluation by electron beam computed tomography and its relation to new cardiovascular events. Am J Cardiol 86:495–498.[CrossRef][Web of Science][Medline]
- Rumberger JA, Brundage BH, Rader DJ, et al. (1999) Electron beam computed tomographic coronary calcium scanning: a review and guidelines for use in asymptomatic persons. Mayo Clin Proc 74:243–252.[Abstract]
- Kondos GT, Hoff JA, Sevrukov A, et al. (2003) Electron-beam tomography coronary artery calcium and cardiac events: a 37-month follow-up of 5635 initially asymptomatic low- to intermediate-risk adults. Circulation 107:2571–2576.
- Raggi P, Boulay A, Chasan-Taber S, et al. (2002) Cardiac calcification in adult hemodialysis patients. A link between end-stage renal disease and cardiovascular disease? J Am Coll Cardiol 39:695–701.
[Abstract/Free Full Text] - Goodman WG, Goldin J, Kuizon BD, et al. (2000) Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med 342:1478–1483.
[Abstract/Free Full Text] - Braun J, Oldendorf M, Moshage W, et al. (1996) Electron beam computed tomography in the evaluation of cardiac calcification in chronic dialysis patients. Am J Kidney Dis 27:394–401.[Web of Science][Medline]
- Sigrist M, Bungay P, Taal MW, et al. (2006) Vascular calcification and cardiovascular function in chronic kidney disease. Nephrol Dial Transplant 21:707–714.
[Abstract/Free Full Text] - Moe SM, O’Neill KD, Fineberg N, et al. (2003) Assessment of vascular calcification in ESRD patients using spiral CT. Nephrol Dial Transplant 18:1152–1158.
[Abstract/Free Full Text] - Ye C, Mao Z, Rong S, et al. (2006) Multislice computed tomographic angiography in evaluating dysfunction of the vascular access in hemodialysis patients. Nephron Clin Pract 104:c94–c100.[CrossRef][Web of Science][Medline]
- Polkinghorne KR, Yii MK, Yik T, et al. (2006) Multi-slice CT fistulography: a preliminary comparison to digital subtraction angiography. Nephrology (Carlton) 11:Suppl 2, A30.[CrossRef]
- Varley H, Gowenlock AH, Bell M. (1980) Calcium, magnesium, phosphorus and phosphates. In Varley H, Gowenlock AH, Bell M (Eds.). Practical clinical biochemistry 5th edn (Heinemann, London, UK) pp. 850–877.
- Agatston AS, Janowitz WR, Hildner FJ, et al. (1990) Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 15:827–832.[Abstract]
- McIntyre CW. (2006) Is it practical to screen dialysis patients for vascular calcification? Nephrol Dial Transplant 21:251–254.
[Free Full Text] - Chertow GM, Burke SK, Raggi P. (2002) Sevelamer attenuates the progression of coronary and aortic calcification in hemodialysis patients. Kidney Int 62:245–252.[CrossRef][Web of Science][Medline]
- Blacher J, Guerin AP, Pannier B, et al. (2001) Arterial calcifications, arterial stiffness, and cardiovascular risk in end-stage renal disease. Hypertension 38:938–942.
[Abstract/Free Full Text] - Giorcelli G, Tricerri A, Vacha G. (1998) 20 years’ experience with "difficult" vascular access. Minerva Urol Nefrol 50:29–33.[Medline]
- Milliner DS, Zinsmeister AR, Lieberman E, et al. (1990) Soft tissue calcification in pediatric patients with end-stage renal disease. Kidney Int 38:931–936.[Web of Science][Medline]
- Goldsmith DJ, Covic A, Sambrook PA, et al. (1997) Vascular calcification in long-term haemodialysis patients in a single unit: a retrospective analysis. Nephron 77:37–43.[Web of Science][Medline]
- Guerin A, London G, Marchais S, et al. (2000) Arterial stiffening and vascular calcification in end-stage renal disease. Nephrol Dial Transplant 15:1014–1021.
[Abstract/Free Full Text]
Accepted in revised form: 26. 9.06
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. Atkinson Age-related medial elastocalcinosis in arteries: mechanisms, animal models, and physiological consequences J Appl Physiol, November 1, 2008; 105(5): 1643 - 1651. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||