Paricalcitol for reduction of albuminuria in diabetes.

Albeit based on robust methods, the study by Dick de Zeeuw and colleagues (Nov 6, p 2010) raises several concerns. First, little attention is given to the fact that the primary endpoint— ie, the eff ect of the vitamin D receptor activator (VDRA) on urinary albuminto-creatinine ratio (UACR)—is actually negative. The positive results on secondary endpoints should thus be used only to generate new hypotheses, not to support the effi cacy of the intervention. Besides, any reduction in UACR must also be interpreted in the context of its biological variation in patients with diabetes, which is estimated at 61%. Second, the mean concentration of vitamin D in the patients included (40 nmol/L) is far below the current recommendations and refl ects a severe defi ciency. Subgroup analysis according to the 25-hydroxyvitamin D concentration would be of interest. Third, the cost-eff ectiveness of a strategy based on paricalcitol in this indication has to be balanced with the use of the much cheaper native vitamin D for which there are physiological reasons to expect some effi cacy. Finally, the long-term safety of this strategy is questionable. Paricalcitol logically decreased concentrations of parathyroid hormone (from 90·7 to 40 μg/L in the 2 μg group). These supraphysiological doses of selective VDRA could increase the risk of low bone turnover (with the inherent increased risk of vascular calcifi cations). Data on calciuria would be valuable in this regard. Risk of renal stones or nephrocalcinosis in the long term should also be discussed.