Since patients’ prognosis depends on the lesions identified by kidney biopsy (KB), we aimed to evaluate predictors of non-diabetic kidney disease (NDKD) in diabetic subjects and to assess their kidney outcome as compared to diabetic nephropathy (DN).
180 adults diagnosed by KB with DN (n = 120) or NDKD (n = 60), over a 10 year time-span, were retrospectively included and followed for a mean of 48.1 (95% CI 43.1-53.1) months. Patients with superimposed specific lesions over DN and with steroid-induced diabetes were excluded. The primary endpoint was renal replacement therapy (RRT) initiation. Only subjects who were alive at the end of follow-up (73 with DN and 38 with NDKD) entered the kidney survival analysis.
Membranous nephropathy (9%) was the most common NDKD. Predictors for NDKD were shorter duration of diabetes (OR 0.88; 95% CI 0.81-0.96, p = 0.004), absence of diabetic retinopathy (OR 0.08; 95% CI 0.01-0.44, p = 0.003), and nephrotic syndrome at presentation (OR 3.55; 95% CI 1.39-9.04, p = 0.008). Subjects with NDKD needed RRT later as those with DN [82 (95% CI 67-97.1) vs. 45 (95% CI 34-56.5) months, p = 0.001]. In an adjusted Cox model, biopsy diagnosed DN independently predicted RRT (OR 4.43; 95% CI 1.54-12.7, p = 0.006). Other predictors were lower eGFR, higher proteinuria, and absence of renin-angiotensin inhibitor therapy.
As one-third of the investigated subjects had NDKD, and NDKD was associated with a better kidney survival, independently predicted by the type of glomerular lesion, KB appears the most reliable tool to guide therapy and to assess outcome in patients with diabetic kidney disease.