This study shows that AAC in MHD patients is closely related to age and hs CRP, which is an independent predictor of LVMI increase over time.
In this study, 87.5% patients had various degrees of AAC, which was higher than the reports of previous researches. Okuno et al. investigated 515 MHD patients and found that AAC was present in 291 patients (56.5%) [13]. Their data gathering time was from August 1998 to July 1999. It is possible that the innovation of photography technology, especially the introduction of digital photography technology, improves the detectable rate of AAC at this stage. Biyik et al. reported the prevalence of AAC was 71.8% in 259 MHD patients [14]. Although their technical conditions of photography were equivalent to those of this study, their participants were younger (mean age: 56 years), which could partly explain their slightly lower prevalence. More recently, Liu et al. conducted a multi-center study including 24 centers from 9 cities across mainland China [4]. More than 1110 HD patients were recruited, and they reported a similar VC prevalence to this study (80.8%). It should be noted that the dialysate calcium concentration in this study was 1.75mmol/L. Whether it plays a role in the high detectable rate of VC needs further researches to explore [15].
It has been confirmed that increased LVM is closely related to poor prognosis of MHD patients, left ventricular hypertrophy (LVH) is often considered as a surrogate end point for all-cause and cardiovascular mortality outcomes. Paoletti et al. followed up 123 MHD patients for 10 years, and found elevated LVMI is the strongest predictor of SCD [16–17]. This study once again proves patients with increased LVMI have higher 5-year mortality.
Actually, bone mineral metabolism disorder (MBD) has been identified an important risk factor of LVH in ESRD patients. Elsheikh et al. found that there was a highly-significant, negative correlation between serum vitamin D and left ventricular mass (LVM) in the patients, and that there was a highly-significant increase in the mean PTH levels of the HD patients compared with the controls [18]. Nielsen et al. reported evaluated associations between FGF23 levels, echocardiography and prognosis in patients on hemodialysis, and demonstrated that FGF23 was significant positive correlated with LVMI [19]. However, the contribution of VC, an important component of MBD, for LVM changes over time has not been fully evaluated in previous studies. Only several cross-sectional correlations were exhibited. A study of 49 MHD patients showed aortic calcification index was independently associated with LVH [6]. Hwang analyzed a larger study population (n = 341) and obtained the same results. VC was independently associated with a 2.42-fold increased risk of LVH (95% CI, 1.26–4.65). In another study, FGF 23, soluble Klotho, and pulse wave velocity were also assessed at the same time, and VC still correlated to high LVM [20]. Those patients with the more widespread VC had the highest LVMI and relative wall thickness. Our results reveal for the first time that MHD patients with elevated LVMI over time have more severe baseline AAC, and that baseline AAC is an independent predictor for LVMI deterioration. Theoretically these results are reliable. Firstly, arterial calcification can increase left ventricular work, which will result in left ventricular remodeling and hypertrophy in the long run. Secondly, many factors associated with uremia, such as FGF23, Klotho, Oxidative stress, inflammation, endothelial dysfunction, insulin resistance and so on may simultaneously participate in the occurrence and development VC and LVH [2, 21]. Some studies directly evaluated the predictive value of AAC for outcomes of MHD patients.A meta-analysis reported that Patients with chronic kidney disease with any or more advanced AAC had a higher risk of cardiovascular events (RR, 3.47; 95% CI, 2.21–5.45), fatal cardiovascular events (RR, 3.68; 95% CI, 2.32–5.84), and all-cause mortality (RR, 2.40; 95% CI, 1.95–2.97) [5]. A 6-year follow-up showed that the risk of CVD mortality was increased by a factor of 3.14 in patients in the moderate to severe AAC group (hazard ratio 3.14 (1.04–9.44), P = 0.042) [22]. A common point of these studies is that cardiac parameters were not analyzed at the same time. Impact of AAC on prognosis might be achieved at least partly by increasing LVM, which need further tests from more mechanism and intervention studies.
As mentioned, while large LVM is associated with poor prognosis, whether direct intervention for LVM can benefit MHD patients is controversial. Badve et al. assessed 25 intervention studies and found that there was no clear and consistent association between intervention-induced LVM change and mortality [23]. Based on the findings of this study, maybe attenuating VC can be followed by the improvement of left ventricular remodeling, which might be a promising approach for better outcomes. Both animal and clinical experiments suggest that calcimimetics may improve VC [24–25]. While the EVOLVE trial did not reach its primary end point [26], imbalance in subjects' age at randomization and discontinuation rates are two of the reasons that the lack of mortality benefit is in question. Subsequently, Lozano-Ortega et al. systematically evaluated four high-quality studies (two observational and two RCTs). The results indicated calcimimetic therapy might in fact reduce mortality among patients receiving maintenance dialysis [27]. In this study, the purpose of baseline data analysis is to search for possible interference targets of VC. The results show that age and hs-CRP are independently related to AAC scores. VC is an aging or premature aging manifestation, which can account for its association with age. Zhu et al. reported older age was an independent predictor for moderate to severe calcification [22]. Hs-CRP is a surrogate marker of microinflammation status. Choi et al. revealed that hs-CRP was a significant risk factor for AAC progression (Odds ratio 1.561; 95% confidence interval 1.038–2.348) [28]. Inflammation plays a critical role in the development of VC [29]. Unlike age, microinflammation status in MHD patients is not completely uncontrollable. Improving dialysis quality, controlling diabetes, and so on can help to alleviate inflammation [30], whose effect on both VC and left ventricular remodeling need more and larger research to identify.
In this study, AAC scores are used to evaluate VC. For one thing, a lateral abdominal X-ray is a convenient and economic method [31]. On the other hand, Abdominal aorta is a susceptible site for atherosclerosis and calcification, so the presence of AAC is related to the occurrence of CV disease and death in patients [32–34]. Therefore, both the 2009 Kidney Disease: Improving Global Outcomes (KDIGO) guideline and the 2017 KDIGO guideline update suggested that in patients with CKD G3a-G5D, a lateral abdominal X-ray can be used to assess the presence of vascular calcification by estimating the calcification of abdominal aorta [35].
There are some limitations in this study. The sample size is not large enough. The application intensity of renin-angiotensin system inhibitors was not included in the analysis because the patients would adjust the treatment intermittently during the follow-up.