Reply from the author

Abstract

To the Editor: Ponticelli and colleagues comment on the fact that we included in our meta-analysis concerning CsA-induced nephrotoxicity in autoimmune diseases (Kidney Int 54:536–545, 1998), studies in which CsA was used at higher doses than 5 mg/kg/day, and that only a few patients given low dose CsA had a persistent increase of serum creatinine of 10 to 33%. To be more specific, only 5 of the 18 included studies used doses of CsA of more than 5 mg/kg/day. The ‘few’ patients they refer to actually consist of 40% of the total population of CsA treated patients (337 patients out of 852). These patients had rheumatoid arthritis and were treated with low-dose CsA (that is, <5 mg/kg/day), and this important subgroup of patients exhibited a rise in serum creatinine of 10 to 30%. Indeed, patients with psoriasis or Crohn's disease treated with CsA had a less pronounced increase in serum creatinine levels. Ponticelli et al suggest that the rise in serum creatinine in rheumatoid arthritis patients simply reflects an increase in muscle mass due to restored physical activity. Three studies in rheumatoid arthritis patients [refs. 47, 50, 60 in our article) in which low dose CsA was used, included a control group treated with azathioprine or D-penicillamine (the other 4 studies were placebo-controlled). In these three studies, CsA was therapeutically as effective as azathioprine or D-penicillamine, hence leading to the same degree of restored physical activity. Yet, the CsA treated groups exhibited a rise of serum creatinine between 22 and 30%, whereas the azathioprine or D-penicillamine treated control groups had no rise at all. Suggesting that this ‘slight’ increase in serum creatinine is due to the restored physical activity is therefore rather simplistic. As already mentioned in our article, this higher increase in serum creatinine in patients with rheumatoid arthritis is probably due to the potentiation of the nephrotoxic effects of CsA by NSAIDs, which are commonly used by rheumatoid arthritis patients, or to the presence of pre-existent subclinical nephropathy in these patients. The analysis of morphological renal lesions found by CsA-treated patients clearly underlines the potential of inducing irreversible lesions by this drug. As described in our paper, in the two studies [refs. 71 and 72] which included pre- and post-treatment biopsies of the same patients, a de novo interstitial fibrosis and tubular atrophy was found in respectively 40 and 65% of the CsA-treated patients already after one year of treatment with low dose CsA. We agree that these renal lesions of interstitial fibrosis and tubular atrophy are not necessarily followed by a progressive loss of renal function after a treatment of one year with CsA. However, the results of long-term studies [refs. 69, 70, 85], that is, an increase of interstitial fibrosis as a function of time, clearly indicate the progressive nature of CsA-induced nephrotoxicity. Ponticelli and colleagues state that CsA nephropathy is reversible after drug withdrawal. As we discussed, the functional impairment was reversible after withdrawal of CsA, complete in six studies, but only partial in seven studies. However, follow-up studies after withdrawal clearly show that a subgroup of patients still has increased levels of serum creatinine 20 to 24 months after cessation of the drug [refs. 57 and 64]. Moreover, the morphological changes induced by CsA are not reversible and progress in a function of time. We agree that short-term treatment of CsA according to the dosage guidelines has an acceptable risk of progressive renal dysfunction, but we strongly fear that long-term therapy of more than one year will lead to a progression in renal dysfunction and morphological lesions, even with doses of CsA below 5 mg/kg/day. The necessity of CsA treatment in each patient with an autoimmune disease should therefore be carefully evaluated against the risk of nephrotoxicity.