Calcific Uremic Arteriolopathy Revisited

Calciphylaxis was first described in the early 1960s by dermatologist Hans Selye et al.¹ as local cutaneous calcification leading to inflammation and skin sclerosis, which is provoked by a mechanical or chemical “challenger” after sensitization with a “systemic calcifying factor.” In the experiments by Selye et al.,¹ rodents were “sensitized” with active vitamin D or parathyroid hormone (PTH) and noted to develop severe calcium deposition in the skin after injections with albumin or iron dextran. Soon after, calciphylaxis was reported in human patients with renal failure and notably, high calcium × phosphorus products.^2,3 In the rodents in the experiments by Selye et al.,¹ cutaneous calcium deposition caused skin lesions, whereas human patients with calciphylaxis develop cutaneous arteriolar medial calcification, which leads to thrombosis and necrotic skin ulcerations.⁴ Hence, more recently, the term calcific uremic arteriolopathy (CUA) has been used to more accurately reflect this histopathology and the affected patients—those with chronic renal failure requiring dialysis. Nonetheless, CUA has also been reported in patients with normal renal function and normal serum PTH, calcium, and phosphorus levels.⁵

CUA is a rare but devastating disease. The incidence of CUA among patients on dialysis is generally reported to be ≤5%⁶ and has been on the rise over the past 10 years.⁷ Although the low incidence of CUA may seem somewhat inconsequential in the dialysis population with its many comorbidities, the morbidity and mortality associated with CUA are substantial. Indeed, the 1-year survival after a diagnosis of CUA is <50%.⁸ If ulcers are present, sepsis is a leading cause of death, and mortality may be as high as 80%.^8,9 Moreover, CUA may be difficult to treat and is a source of considerable pain and suffering for patients.

Our understanding of CUA pathophysiology is incomplete. Disturbances of calcification inhibitors have been implicated as the instigating pathophysiologic abnormality.^4,10 However, although arteriolar vascular calcification plays a major role in the pathology of CUA and patients who are diagnosed with CUA are reported to have biochemical mineral metabolism abnormalities, it is not considered to be a feature of CKD mineral bone disorder. In fact, 70%–80% of patients on prevalent dialysis have coronary artery calcification, and many have abnormal markers of mineral metabolism¹¹; however, only a very small percentage of patients on dialysis goes on to develop CUA, and those who do develop it may have normal calcium, phosphorus, and PTH serum levels.¹⁰ Clinical factors associated with CUA are numerous and sometimes contradictory. Obesity; diabetes mellitus; women; white race; time on dialysis; abnormalities of the vitamin D-PTH axis (including elevated PTH levels); calcium × phosphorus product and alkaline phosphatase levels; liver disease; hypoalbuminemia; and medications, such as corticosteroids, warfarin, vitamin D, and calcium salts, have all been associated with CUA among patients on dialysis.^5,6,8,9 Association does not prove causation, and the time at which these factors are evaluated in relation to the diagnosis of CUA may affect their interpretation and complicate study comparison.⁶ It has even been suggested that CUA be classified into subgroups according to anatomic distribution—proximal versus distal lesions, which may develop in distinct patient populations and respond differently to therapies.¹⁰ Therefore, to further our understanding of this disorder, longitudinal data collection of parameters that are related to CUA before its diagnosis is imperative.^6,10

In this issue of the Journal of the American Society of Nephrology, Nigwekar et al.¹² report the results of a large case-control study that compared the baseline characteristics at hemodialysis initiation among 1030 patients who developed CUA with those of patients initiating hemodialysis who did not develop CUA. Over a 4-year period, patients from chronic hemodialysis units affiliated with Fresenius Medical Care North America (FMCNA) with newly diagnosed CUA were identified (patients) and matched to controls in a 1:2 ratio by age (±5 years), sex, and race. The mean incidence rate of CUA in the entire cohort of patients from chronic hemodialysis units affiliated with FMCNA during the 4-year study period was 3.49/1000 patient-years (95% confidence interval [95% CI], 3.30 to 3.72). CUA incidence was much higher among warfarin users (6.24/1000 patient-years; 95% CI, 5.87 to 6.67) compared with those who did not use warfarin (3.41/1000 patient-years; 95% CI, 3.27 to 3.75). Similar to prior studies, mortality after CUA diagnosis was high—45% died within 12 months of diagnosis. More women and individuals of white race developed CUA. In this analysis, baseline diabetes, higher body mass index, higher albumin–corrected serum calcium, higher PTH, nutritional vitamin D use, cinacalcet use, and warfarin use were associated with increased odds of CUA development. Notably, higher baseline hemoglobin and use of erythropoietin-stimulating agents were associated with reduced odds of CUA development.

Leveraging the FMCNA database and the large number of patients newly diagnosed with CUA (n=1030), Nigwekar et al.¹² were able to investigate the anatomic distribution of CUA lesions (central versus peripheral). Of those with known location of a CUA lesion (n=710), central distribution (n=582) was more common than peripheral distribution. Diabetes, higher body mass index, higher albumin–corrected serum calcium, warfarin use, and insulin use were associated with central CUA development. Among those with diabetes, receiving three or greater than three compared with no insulin injections had an increased risk of future CUA development, suggesting a dose-response relationship.

This is the largest epidemiologic study to date that assesses the risk factors associated with CUA diagnosis, incidence rate of CUA, and mortality in a patient population on hemodialysis. It is novel in that it assesses CUA risk factors that are present at dialysis initiation, significantly ahead of CUA diagnosis; indeed, the median duration between hemodialysis initiation and CUA development was 925 days (interquartile range, 273–2185 days). The large number of patients with CUA also allowed for closer evaluation of anatomic distribution of CUA lesions and linked skin trauma in the form of insulin injections with the development of CUA in a dose-response relationship.

This study moves the nephrology field closer to a more definitive list of CUA risk factors. It confirms the findings of other studies that link race, sex, diabetes, obesity, and warfarin to risk of CUA. It also corroborates the known incidence rate of CUA among patients on hemodialysis and its poor prognosis. However, the question remains: what factor or combination of factors instigates CUA? The low incidence rate of CUA conveys that not all patients on dialysis at risk of CUA will actually develop it—in fact, the majority will not. Identifying and understanding the changes that take place between dialysis initiation and development of CUA would go a long way in elucidating modifiable risk factors and may even hint at causation. Therefore, the lack of repeated measures and time–averaged biochemical analysis is a significant limitation of this study. However, rare diseases, such as CUA, pose a significant research challenge, which is further complicated by the limitations of database research. For example, the fact that there is no specific International Classification of Diseases-9 code (or International Classification of Diseases-10 code) for calciphylaxis or CUA has hindered use of other large databases. Fortunately, the FMCNA has tracked new CUA diagnoses via physician questionnaire. However, data collection was not detailed enough to reliably delineate the time of CUA diagnosis. Furthermore, the development of CUA may begin well in advance of skin lesions, further complicating the identification of CUA-instigating factors.

Nigwekar et al.¹² report strong results confirming nonmodifiable risk factors and identify other novel and modifiable risk factors for CUA in a large nationally representative cohort of patients on hemodialysis. This investigation should serve as a jumping off point for clinician and scientist alike to find ways to better understand, prevent, and treat this calamitous disorder.

Disclosures

None.