Cardiovascular (CV) disease is an important global issue, especially in dialysis patients. It is the leading cause of deaths in end-stage renal disease (ESRD) patients which accounting for up to 40% of deaths among these patients during the first 3 years of dialysis. In addition, CV mortality in ESRD patients is about 10 to 20 times higher compared with general population. The impact of cardiovascular disease on patients starting dialysis has recently received increasing attention.[2,4]
The atherosclerosis-related vascular complication is the most common cause of CV disease. Once atherosclerosis develops, the atheromatous plaque or fibrous fatty plaque forms with calcium deposition will cause narrowing the blood vessel lumen and weakening the media which cause coronary artery disease, stroke and peripheral artery disease.[4,5] The CV events associated with atherosclerosis are more often fatal in patients with chronic kidney disease than in individuals without chronic kidney disease. In addition to CV mortality, non-CV mortality was also increased to the same extent as mortality from CV disease.[2,7] Accurate risk assessment for future CV events can help clinician to guide clinical management and the accurate tool to predict the long-term outcomes is still lacking.
Abdominal aorta calcification (AAC) is a good clinical tool to evaluate the atherosclerosis occurring in abdominal aorta and is a marker of subclinical atherosclerotic disease. It correlates with asymptomatic coronary artery disease and is also highly predictive of subsequent CV outcomes in general population.[10,11] There were several studies showed that AAC could predict CV outcomes in patients under hemodialysis[12,13] but the data, especially the long-term data, for patients under peritoneal dialysis (PD) was still limited.[14,15] In addition, previous studies of AAC were mostly focused on CV outcomes and total mortality instead of non-CV mortality. The role of AAC in prediction of mortality from non-CV diseases in dialysis patients has not yet been reported. In this prospective cohort study, we aimed to analyze the long-term predictive values of AAC to CV and non-CV mortality in PD patients.
We prospectively enrolled 123 patients who undergoing PD for more than 3 months since February 2009. Pregnant women, patients with clinical signs of active infection, and those with a prior renal transplant were excluded. The baseline characteristics, medical history and medication usage were carefully recorded, and biochemical parameters were tested during initial evaluation. All patients provided written informed consent. This study was approved by the Institutional Review Board of National Taiwan University Hospital (approval numbers NTUH-REC No. 200808062R and NTUH-REC No. 201007080R).
These patients were prospectively followed in our peritoneal dialysis center after enrollment. The causes of mortality from CV and non-CV diseases were carefully recorded during follow-up. CV mortality was defined as mortality due to acute coronary syndrome, peripheral artery disease, sudden cardiac death (SCD), life threatening arrhythmia (ventricular tachycardia/ventricular fibrillation), progressive heart failure, ischemic and hemorrhagic stroke. SCD was defined as cardiac arrest occurring suddenly and within 1 h of witnessed symptom onset. The patients suffered from other causes of death were categorized as the non-CV mortality. The patients who received renal transplantation were censored in this study.
2.3 Outcomes predictor
AAC ratio is the predictor of interest in this study. AAC was measured by the standard 64-multiple detector computed tomography (MDCT) scan (LightSpeed VCT, GE Healthcare, Milwaukee, WI) in all subjects. The calcified area was calculated based on an attenuation range of >150 Hounsfield units using image analysis software (ImageJ, version 1.45, National Institutes of Health, Bethesda, MD). The percentages of the area of the whole aorta affected by aortic calcification were calculated from the images of 4 consecutive slices just above the iliac bifurcation level.[17,18] The images were reviewed independently by 2 radiologists who were blinded to the patients’ clinical characteristics.
The baseline demographic data, biochemistry analysis and echocardiogram including HbA1c, serum creatinine, PD KT/V, lipid profiles, serum electrolytes, C-reactive protein (CRP) and left ventricular ejection fraction (LVEF) were also analyzed as covariates in this study.
Transthoracic echocardiography (iE33 xMATRIX Echocardiography System, Philips, Amsterdam, Netherlands) was performed in all patients and LVEF was measured by M-mode measurements or long-axis area-length method in accordance with the recommendations of the American Society of Echocardiography.
2.6 Statistical analysis
Data were expressed as mean ± standard deviation for normally distributed data which were tested by Kolmogorov-Smirnov test. Comparisons of data between groups were made by the independent t test. Differences between proportions were calculated by the Chi-square test or Fisher's exact test. Cox regression analysis was used to explore the associations between variables and outcomes. Significant determinants in univariable Cox regression analysis (P < .05), including age, LVEF, AAC ratio, were then tested in multivariable Cox regression analysis with stepwise subset selection to identify independent predictors of outcomes. The discrimination abilities of AAC to CV, non-CV and total mortality were assessed using the time-dependent receiver operating characteristic (ROC) curve analysis. The optimal cutoff point of the AAC ratio with the highest area under curve (AUC) to predict CV mortality was obtained by multiple comparisons of Cox regression probability. The Kaplan-Meier survival curves according to the cutoff were plotted and log-rank tests were used for comparison. The predicted probability of an event in the Cox model was obtained using the ‘phreg’ procedure in SAS version 9.4 (SAS Institute, Cary, NC). The optimal cutoff point of a marker in the survival analysis was determined using R version 3.6.1 (R Development Core Team) and the “survminer” package (Version 0.4.6 updated on Sep 3, 2019). Other analyses were done using SPSS version 25 for Windows (SPSS Inc., IL, USA). The significance level of the statistical analysis was set at .05.
3.1 Patients (Table 1)
After median 6.8 (interquartile range, 3.6–9.2) years follow-up, there were 18 CV mortality, 24 non-CV mortality and 42 total mortality. The cumulative CV mortality, non-CV mortality and total mortality were 20.4%, 26.6%, 41.6%, respectively. The causes of CV mortality including 7 life threatening arrhythmia, 3 cardiogenic shock, 7 sudden death and 1 hemorrhagic stroke. The causes of non-CV mortality were mainly sepsis, especially peritonitis, except 1 patient with lung adenocarcinoma.
In CV mortality analysis, the AAC ratios were significantly higher, 32.6 ± 18.6% in patients died due to CV causes compared with others without CV mortality (including non-CV mortality and survived patients). The patients died from CV causes were also significantly older compared with others without CV mortality. The sex, underling systemic disease, duration of PD before enrollments, medication use at enrollments, biochemical data and left ventricular systolic function were well balanced between 2 groups.
3.2 The predictors of CV mortality (Tables 2 and 4)
In univariable Cox regression analysis, older age, lower LVEF, and higher AAC ratio were significant predictors of CV mortality. Every increased 1% of AAC ratio increased 6% risk of CV mortality, (hazard ratio, 1.060; 95% CI, 1.033–1.089). These 3 predictors were then tested in multivariable Cox regression analysis and only AAC ratio remained in the model. Then we adjusted the AAC ratio with age, sex, baseline medication use, DM, HTN and LVEF in 5 different models. The AAC ratio remained to be a significant predictor of CV mortality after adjustment.
3.3 The predictors of non-CV mortality (Table 3)
In univariable Cox regression analysis, older age and longer duration of PD before enrollment were significant predictors of non-CV mortality. Both age and duration of PD were remained in the model after multivariable Cox regression analysis. The AAC was not associated with non-CV mortality (hazard ratio, 1.015; 95% CI, 0.992–1.039).
3.4 The predictive power of AAC to CV and non-CV mortality (Fig. 1 and Table 5)
The time-dependent ROC curve analysis showed that AAC ratio could predict CV mortality and total mortality, AUC:0.787 and AUC:0.685, respectively. However, AAC ratio could not predict the non-CV mortality, AUC:0.537.
3.5 The optimal cutoff value for AAC ratio to predict CV mortality (Fig. 2)
We determined the optimal cutoff value for AAC ratio to predict CV mortality using ROC curve analysis. The best cutoff value for AAC ratio was 39% to predict CV mortality (hazard ratio, 8.01; 95% CI, 3.14–20.44). The KM survival curve analysis showed that PD patients with AAC ratio >39% had higher risks of CV mortality compared with other patients with AAC ratio ≤39% (Fig. 2).
Cardiovascular disease is the leading cause of morbidity and mortality in ESRD patients. In daily practice, predicting the clinical outcomes of PD patients remains to be a challenge. Atherosclerosis-related vascular calcification is commonly observed in dialysis patients and has been highly associated with CV outcomes.[5,13,20,21] There are many risk factors contribute to atherosclerosis and vascular calcification such as age, smoking, hypertension, dyslipidemia, diabetes mellitus, uremia, mineral metabolism particularly hyperphosphatasemia, chronic inflammation, fetuin-A and osteoprotegerin.[22–30] In the advanced stage of atherosclerosis, the calcium deposit in the vascular wall is frequently observed and can be used for outcomes prediction.[5,13,20,21]
The vascular calcification can be measured from many different sites including coronary arteries,[32,33] heart valves, carotid arteries, thoracic aorta, abdominal aorta and peripheral arteries. Among all these sites, AAC is one of the best choice due to it can be easily measured with quantitative results and can be used for long-term serial follow-up.[12,28] AAC tends to progress in ESRD patients during dialysis and it is associated with congestive heart failure, arterial stiffness,[40,41] effects of renal denervation, autonomic dysfunction and worse heart rate variability.[43,44] Furthermore, AAC has been reported to predict CV events and mortality in ESRD patients.[13,21,45]
There are 2 main commonly employed methods to quantify AAC including plain abdominal X ray with Kauppila score and CT. The advantage of plain X ray including lower cost and radiation dose but it is difficult to quantify AAC. In contrast, CT can easily quantify AAC with objective and reproducible results. In addition, the abdominal CT is commonly performed to assess other pathologies therefore incidentally available for many patients. Furthermore, CT can clearly recognize the patterns of calcification. In dialysis patients, it is characterized by mineral deposition in the tunica media, in contrast to non-dialysis populations, where calcification predominately deposits in the atheromatous plaque. Previous study supported that CT appeared to be more sensitive than plain X-rays at detecting vascular calcifications in hemodialysis patients. Tsushima et al developed a method to measure the percentage of calcified volume against whole vascular volume using CT[17,18] and Mori et al also described and validated a new volume-rendering approach to quantify aortic calcification using commercially available software. Currently, CT remains the reference standard in the measurement of AAC. In this study, we used CT to evaluate the ratio of AAC and confirmed its strong predictive value of CV mortality. Even after adjustments with traditional risk factors, AAC remained to be the excellent predictor of CV mortality.
Compared with CV mortality, the studies about the association between non-CV mortality and AAC are limited. At the best of our knowledge, the present study is the first one demonstrating that AAC could not predict the non-CV mortality. Mäkelä et al recently reported that AAC can predict total mortality and CV events in PD patients. Unfortunately, they didn’t report the association between non-CV cause of deaths and AAC. In this study, we also showed that AAC could predict total mortality but the predictive power was mainly contributed from CV mortality. In the present study, instead of AAC, age and duration of PD before enrollment were significant predictors of non-CV mortality. The possible explanation of our finding was that the main causes of non-CV mortality were sepsis, especially peritonitis, and cancer which were not strongly associated with atherosclerosis.
Interestingly, the incidence of non-CV mortality was even higher than CV mortality in PD patients in our study. The data from the European Renal Association-European Dialysis and Transplant Association Registry also supported our finding which showed that the age-adjusted CV mortality in dialysis patients was 8.8 times higher, whereas the non-CV mortality was also increased 8.1 times compared with general population. This emphasizes that in addition to the reduction of CV risk factors to improve CV outcomes, the prevention of non-CV mortality deserves equal attention. In the future, research for the reduction of mortality in dialysis patients should focus on both CV and non-CV causes of death.
There are several limitations to this study. First, this is a small cohort study and all the patients in this study were Taiwanese. The statistic power was limited due to small case number. Further larger clinical studies with different ethnicities enrollment are needed to confirm our findings. Second, the time gap between chronic kidney disease to initiation of PD might also influence the severity of AAC and outcomes which was lacked in current study. Third, our study group is only limited to PD patients, and further studies are needed to elucidate whether the same association between AAC and CV, non-CV mortality also exists in hemodialysis patients or even general population. The risks of non-CV mortality might be different between hemodialysis and PD patients.
AAC has excellent prognostic value of long-term CV mortality in PD patients. The predictive power is high even after adjust with multiple clinical variables. However, AAC is not associated with non-CV mortality. Since the non-CV mortality contributed a large portion of total cause of mortality, further studies should focus on the risk assessment and management of non-CV cause of deaths in PD patients.
J.W.H. conceived and designed the experiments. C.H.T., I.J.T. and Y.H.L. analyzed the data. C.H.T and I.J.T. wrote the paper. L.Y.L. and J.W.H. made scientific comments on the manuscript.
. Szeto CC, Wong TY, Chow KM, et al. Are peritoneal dialysis
patients with and without residual renal function equivalent for survival study? Insight from a retrospective review of the cause of death. Nephrol Dial Transplant 2003;18:97782.
. de Jager DJ, Grootendorst DC, Jager KJ, et al. Cardiovascular and noncardiovascular mortality among patients starting dialysis. JAMA 2009;302:17829.
. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis 1998;32: (Suppl 3): S112119.
. Powers WJ, Rabinstein AA, Ackerson T, et al. Guidelines for the early management of patients with acute ischemic stroke: 2019 update to the 2018 guidelines for the early management of acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2019;50:e344418.
. Campean V, Neureiter D, Varga I, et al. Atherosclerosis and vascular calcification in chronic renal failure. Kidney Blood Press Res 2005;28:2809.
. Drueke TB, Massy ZA. Atherosclerosis in CKD: differences from the general population. Nat Rev Nephrol 2010;6:72335.
. Jager KJ, Lindholm B, Goldsmith D, et al. Cardiovascular and non-cardiovascular mortality
in dialysis patients: Where is the link? Kidney Int Suppl 2011;1:213.
. O’Connor SD, Graffy PM, Zea R, et al. Does nonenhanced CT-based quantification of abdominal aortic calcification outperform the Framingham risk score in predicting cardiovascular events in asymptomatic adults? Radiology 2019;290:10815.
. An C, Lee HJ, Lee HS, et al. CT-based abdominal aortic calcification score as a surrogate marker for predicting the presence of asymptomatic coronary artery disease. Eur Radiol 2014;24:24918.
. Wilson PW, Kauppila LI, O’Donnell CJ, et al. Abdominal aortic calcific deposits are an important predictor of vascular morbidity and mortality. Circulation 2001;103:152934.
. Bastos Goncalves F, Voute MT, Hoeks SE, et al. Calcification of the abdominal aorta as an independent predictor of cardiovascular events: a meta-analysis. Heart 2012;98:98894.
. Inoue H, Shimizu S, Watanabe K, et al. Impact of trajectories of abdominal aortic calcification over 2 years on subsequent mortality: a 10-year longitudinal study. Nephrol Dial Transplant 2018;33:67683.
. Okuno S, Ishimura E, Kitatani K, et al. Presence of abdominal aortic calcification is significantly associated with all-cause and cardiovascular mortality
in maintenance hemodialysis patients. Am J Kidney Dis 2007;49:41725.
. Mäkelä S, Asola M, Hadimeri H, et al. Abdominal aortic calcifications predict survival in peritoneal dialysis
patients. Perit Dial Int 2018;38:36673.
. Yoon HE, Park BG, Hwang HS, et al. The prognostic value of abdominal aortic calcification in peritoneal dialysis
patients. Int J Med Sci 2013;10:61723.
. Ramesh S, Zalucky A, Hemmelgarn BR, et al. Incidence of sudden cardiac death in adults with end-stage renal disease: a systematic review and meta-analysis. BMC Nephrol 2016;17:78.
. Miwa Y, Tsushima M, Arima H, et al. Pulse pressure is an independent predictor for the progression of aortic wall calcification in patients with controlled hyperlipidemia. Hypertension 2004;43:53640.
. Tsushima M, Koh H, Suzuki M, et al. Noninvasive quantitative evaluation of early atherosclerosis and the effect of monatepil, a new antihypertensive agent. An interim report. Am J Hypertens 1994;7(Pt 2):154S60S.
. Mitchell C, Rahko PS, Blauwet LA, et al. Guidelines for performing a comprehensive transthoracic echocardiographic examination in adults: recommendations from the American Society of Echocardiography. J Am Soc Echocardiogr 2018;32:164.
. Bhan I, Thadhani R. Vascular calcification and ESRD: a hard target. Clin J Am Soc Nephrol 2009;4: (Suppl 1): S102105.
. Verbeke F, Van Biesen W, Honkanen E, et al. Prognostic value of aortic stiffness and calcification for cardiovascular events and mortality in dialysis patients: outcome of the calcification outcome in renal disease (CORD) study. Clin J Am Soc Nephrol 2011;6:1539.
. Paloian NJ, Giachelli CM. A current understanding of vascular calcification in CKD. Am J Physiol Renal Physiol 2014;307:F891900.
. Speer MY, Giachelli CM. Regulation of cardiovascular calcification. Cardiovasc Pathol 2004;13:6370.
. Taniwaki H, Ishimura E, Tabata T, et al. Aortic calcification in haemodialysis patients with diabetes mellitus. Nephrol Dial Transplant 2005;20:24728.
. Dellegrottaglie S, Sanz J, Rajagopalan S. Vascular calcification in patients with chronic kidney disease. Blood Purif 2006;24:5662.
. Wang AY, Woo J, Lam CW, et al. Associations of serum fetuin - a with malnutrition, inflammation, atherosclerosis and valvular calcification syndrome and outcome in peritoneal dialysis
patients. Nephrol Dial Transplant 2005;20:167685.
. Libby P. Inflammation in atherosclerosis. Nature 2002;420:86874.
. Golledge J. Abdominal aortic calcification: clinical significance, mechanisms and therapies. Curr Pharm Des 2014;20:58348.
. Pugliese G, Iacobini C, Blasetti Fantauzzi C, et al. The dark and bright side of atherosclerotic calcification. Atherosclerosis 2015;238:22030.
. Chen NX, Moe SM. Uremic vascular calcification. J Investig Med 2006;54:3804.
. Berliner JA, Navab M, Fogelman AM, et al. Atherosclerosis: basic mechanisms. Oxidation, inflammation, and genetics. Circulation 1995;91:248896.
. Agatston AS, Janowitz WR, Hildner FJ, et al. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol 1990;15:82732.
. Alexopoulos N, Raggi P. Calcification in atherosclerosis. Nat Rev Cardiol 2009;6:6818.
. Otto CM, Burwash IG, Legget ME, et al. Prospective study of asymptomatic valvular aortic stenosis. Clinical, echocardiographic, and exercise predictors of outcome. Circulation 1997;95:226270.
. de Weert TT, Cakir H, Rozie S, et al. Intracranial internal carotid artery calcifications: association with vascular risk factors and ischemic cerebrovascular disease. AJNR Am J Neuroradiol 2009;30:17784.
. Budoff MJ, Nasir K, Katz R, et al. Thoracic aortic calcification and coronary heart disease events: the multi-ethnic study of atherosclerosis (MESA). Atherosclerosis 2011;215:196202.
. Rocha-Singh KJ, Zeller T, Jaff MR. Peripheral arterial calcification: prevalence, mechanism, detection, and clinical implications. Catheter Cardiovasc Interv 2014;83:E212220.
. Goodman WG, Goldin J, Kuizon BD, et al. Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis. N Engl J Med 2000;342:147883.
. Walsh CR, Cupples LA, Levy D, et al. Abdominal aortic calcific deposits are associated with increased risk for congestive heart failure: the Framingham Heart Study. Am Heart J 2002;144:7339.
. McEniery CM, McDonnell BJ, So A, et al. Aortic calcification is associated with aortic stiffness and isolated systolic hypertension in healthy individuals. Hypertension 2009;53:52431.
. Guerin AP, London GM, Marchais SJ, et al. Arterial stiffening and vascular calcifications in end-stage renal disease. Nephrol Dial Transplant 2000;15:101421.
. Courand PY, Pereira H, Del Giudice C, et al. Abdominal aortic calcifications influences the systemic and renal hemodynamic response to renal denervation in the DENERHTN (renal denervation for hypertension) trial. J Am Heart Assoc 2017;6:
. Tsai CH, Lin C, Ho YH, et al. The association between heart rhythm complexity and the severity of abdominal aorta calcification
in peritoneal dialysis
patients. J Am Heart Assoc 2017;6:e007062.
. Chesterton LJ, Sigrist MK, Bennett T, et al. Reduced baroreflex sensitivity is associated with increased vascular calcification and arterial stiffness. Nephrol Dial Transplant 2005;20:11407.
. Huang JW, Lien YC, Yang CY, et al. Osteoprotegerin, inflammation and dyslipidemia are associated with abdominal aortic calcification in non-diabetic patients on peritoneal dialysis
. Nutr Metab Cardiovasc Dis 2014;24:23642.
. Kauppila LI, Polak JF, Cupples LA, et al. New indices to classify location, severity and progression of calcific lesions in the abdominal aorta: a 25-year follow-up study. Atherosclerosis 1997;132:24550.
. Davies MR, Hruska KA. Pathophysiological mechanisms of vascular calcification in end-stage renal disease. Kidney Int 2001;60:4729.
. NasrAllah MM, Nassef A, Elshaboni TH, et al. Comparing different calcification scores to detect outcomes in chronic kidney disease patients with vascular calcification. Int J Cardiol 2016;220:8849.
. Mori S, Takaya T, Kinugasa M, et al. Three-dimensional quantification and visualization of aortic calcification by multidetector-row computed tomography: a simple approach using a volume-rendering method. Atherosclerosis 2015;239:6228.
. Karohl C, D’Marco Gascon L, Raggi P. Noninvasive imaging for assessment of calcification in chronic kidney disease. Nat Rev Nephrol 2011;7:56777.