Nephrogenic Systemic Fibrosis and Diabetes Mellitus FREE

Over the past decade, specialists in all areas of medicine have become increasingly aware of and interested in nephrogenic systemic fibrosis (NSF), yet much of this devastating disease's pathogeny remains a mystery. Many risk factors have been implicated, including liver disease,¹ erythropoietin,² proinflammatory and hypercoagulable conditions, dialysis, and gadolinium exposure.³ Diabetes mellitus (DM) and renal failure are both associated with fibrosing disorders. While a strong association exists between kidney disease and NSF, diabetes has never before been proposed as a risk factor. The most common cause of renal failure in the United States is DM.^4- 5 Seven percent of the population has diabetes, and every third patient with DM develops nephropathy, accounting for about 45% of end-stage renal disease (ESRD).⁴ Herein, we review 295 cases of NSF to determine if diabetes could be a risk factor for NSF development.

To examine the prevalence of DM among patients with NSF, a PubMed literature search was performed for articles published between 1997 and September 2011 containing the terms nephrogenic systemic fibrosis and nephrogenic fibrosing dermopathy. Resulting articles from any country in any clinical setting that provided clinical descriptions and data for each patient with histologic evidence of NSF were analyzed. Articles containing known previously reported cases were excluded. A total of 81 articles describing 295 patients were included in our final analysis (eReferences). Each patient's age, sex, gadolinium exposure, dialysis history, cause of renal failure, comorbidities, and diagnosis of DM were tabulated. Prevalence rates of diabetes and other causes of renal disease were calculated and compared with current data from the United States Renal Data System (USRDS)⁵ using a χ² goodness-of-fit test.

Table 1 summarizes demographic data. All patients had some degree of renal failure. Most had been undergoing hemodialysis and/or peritoneal dialysis, although 12 patients (4.1%) were reported to never have had dialysis. Most had chronic renal disease or ESRD, although 26 patients (8.8%) had acute kidney injury. Fifty-seven patients were reported to have diabetes (19.3%). All but 1 were undergoing dialysis (or dialysis information was not provided). The mean (SD) age of the diabetic group was 57.5 (12.7) years; 49% were men. The mean (SD) age of the nondiabetic group was 47.6 (16.4) years. Of all 295 patients, 54% had exposure to gadolinium; 41% had unknown or unrecorded gadolinium exposure histories; and 5% had never been exposed to gadolinium. Of the diabetic patients, 45.6% had exposure to gadolinium, 50.9% had unknown or unrecorded exposure histories, and 3.5% had no exposure. Of the 12 patients who never had dialysis, all had either been exposed to gadolinium or had unknown or unrecorded exposure histories.

Table 2 lists the 229 cases of ESRD where a primary cause was identified. The prevalence of diabetes among these patients with NSF (18.8%) was much lower than the expected USRDS rate of DM in all patients with ESRD (45%).⁵ Patients with NSF also had a higher rate of glomerulonephritis than expected, even when only “true ” glomerulonephritides were analyzed (excluding focal segmental glomerulosclerosis, membranous and human immunodeficiency virus nephropathies, nephrotic syndrome, and minimal change disease). Results of a χ² goodness-of-fit test ( χ²  =  424.69) (P   <  1.0   ×  10 ⁻⁴) indicate statistical significance in this difference in ESRD cause distribution among patients with NSF patients compared with expected USRDS rates.

Besides renal disease, the most prevalent comorbidities among all patients are listed in Table 3.

Prevalence rates of renal failure and gadolinium exposure among patients with NSF in this analysis are consistent with those found in other studies.³ Our data indicate an underrepresentation of diabetes mellitus among patients with NSF. To our knowledge, this is the first report of a possible protective quality of diabetes against NSF. Glycosylated tissue may have a low affinity for gadolinium. Pharmacologic treatments for DM (insulin or oral hypoglycemics) may confer protection. Diabetic patients often have comorbid conditions that may contribute to or confound this finding. Though rarely described in NSF reports, variables such as cause, type, duration, and glycemic control of diabetes may offer clues. Our data also indicate an underrepresentation of hypertension and an overrepresentation of glomerulonephritis, urologic disease, and cystic kidney disease. Whether this reflects infectious, autoimmune, other specific inflammatory changes, or an unidentified confounder is beyond the scope of this analysis. Further research may elucidate the reason for this relationship between diabetes and NSF and aid in developing new treatments.