This study showed that AAC was associated with all-cause and CVD mortality in HD patients, as had been reported [1] [2] [3]. The major causes of death were CVD, infection, and malnutrition. The combined mortality of infection and malnutrition was higher than that of CVD in the most severe AAC group. The HD patients with more severe AAC were older than those with milder AAC, which may influence an increase in death from infection and malnutrition. These findings highlight the importance of paying more attention to the occurrence of fatal infection and the management of nutrition in HD patients with severe AAC.
In this study, CVD was the most common cause of death in HD patients with all AAC grades. A previous study showed that CVD mortality accounted for 37.7% of total death in patients with severe AAC, and those rates by AAC grade 0, 1, 2, and 3 were 5.3%, 12.7%, 18.9%, and 24.4%, respectively [2]. However, in this study, there was no clear difference in CVD mortality among AAC grades. The low prevalence of diabetes mellitus, a well-known poor prognostic factor, may have influenced no increase in CVD mortality in HD patients with severe AAC [9]. In addition, ischemic stroke mortality tended to occur more frequently in patients with severe AAC. This result was compatible with a previous report that the incidence of cerebral infarction was associated with calcification of the thoracic aorta [10]. Moreover, a recent report showed that CVD risk profiles in both sexes that were differentially associated with calcification at multiple vascular sites [11]. Although gender association was not clear in this study, this report supports the importance of assessing AAC, which is a potential risk factor for CVD.
The HD patients with severe AAC in this study were older than patients on maintenance dialysis in Japan and were more likely to die from infection and malnutrition [5]. Elderly HD patients may have malnutrition, inflammation, and atherosclerosis (MIA) syndrome that is associated with an elevated mortality in dialysis patients[12]. Malnutrition and inflammation were significantly associated with AAC progression, which was evaluated by using lateral lumbar radiography [13]. Moreover, Okamoto et al. reported that poor nutritional status was an independent risk factor for the progression of aortic calcification in HD patients [14].
Further, it has been reported that elderly HD patients often died of infectious diseases, including pneumonia and sepsis [15]. In this study, the mean serum albumin level tended to be lower in more severe AAC groups. Therefore, the prevalence of MIA syndrome may be higher in patients with higher AAC grades, and those patients may have contributed to the increase in mortality due to infection and malnutrition. On the other hand, the association between AAC grade 3 and all-cause mortality remained significant after adjustment of covariates, suggesting the existence of other factors that we did not adjust in a multivariate analysis.
There are various methods for the evaluation of vascular calcification related to mortality. It has been reported that AAC evaluated by coronary Computed Tomography (CT) and 3D-CT can also predict CV events [16] [17]. In addition, there were several reports that the calcification score on abdominal radiograph can predict CV events [18] [19]. It is necessary to consider a more appropriate evaluation method, including not only predictive hit ratio but also cost and invasiveness. Although Bohn et al. evaluated that AAC did not provide additional information on the prediction of mortality, we found the causes of death in patients on HD much differ among different AAC severities [20]. Thus, we emphasize that the assessment of AAC is useful in predicting all causes of death and should pay attention to especially infection and nutrition.
The study has several strengths. The design of the study was simple and the AAC assessment was easy to perform on a routine chest radiograph. The long observation period of this study allowed us to detect the clear difference in all-cause mortality and CVD mortality despite the survival rate of HD patients in Japan is much longer than that in Western countries [21]. A recent randomized controlled trial showed an intensive control of hyperphosphatemia resulted in reduction of AAC [22]. The strong association between AAC and all-cause mortality in this study supports the necessity of an appropriate management of hyperphosphatemia in patients on HD. On the other hand, there are several limitations in this study. Firstly, the AAC grading based on a plain chest radiograph may be biased by the investigators. However, this method has several advantages such as simplicity and cost, and thus is available in clinical practice. Secondly, we could not exclude the effect of changes in AAC and laboratory data during the study period on all-cause mortality and causes of death, since our analyses were based on baseline data. Thirdly, this study was conducted on patients from two HD clinics in Japan that may limit the generalizability of the study results. Although the baseline characteristics of the patients in this study appeared to be close to the nation-wide Japanese dialysis cohort, our patients were slightly younger, had a longer dialysis vintage, and used more phosphate-binders than those of the JSDT population [5].