© 2007 American Thoracic Society doi: 10.1164/rccm.200708-1185ED
Daily Cyclophosphamide for SclerodermaAre Patients with the Most to Gain Underrepresented in this Trial?
Royal Brompton Hospital
Sotiria Hospital for Diseases of the Chest
University of Michigan In diffuse fibrosing lung disease, therapeutic studies have been inconclusive. Statistically significant treatment effects are reported in a number of placebo-controlled studies (1–3). However, the clinical benefits of these effects have been uncertain in idiopathic pulmonary fibrosis (1), sarcoidosis (2), and, until now, pulmonary fibrosis in scleroderma (SSc) (3). In particular, it is not known whether apparent benefits after a limited period of treatment translate into changes in longer term outcome. Tashkin and colleagues make a radical and very welcome departure from usual practice in this field. They had previously reported the first placebo-controlled evidence of a therapeutic effect with cyclophosphamide in patients with SSc and lung disease (3). Their follow-up data, reported in this issue of the Journal (pp. 1026–1034) (4), show clearly that the beneficial pulmonary effects of oral cyclophosphamide persist for 6 months after cessation of treatment but are largely lost at 1 year. The internal consistency of the observations is reassuring—the trends with regard to the primary endpoint (FVC) and a number of secondary endpoints were strikingly similar at all time points. There was uniformity in the lack of a treatment effect on measures of gas transfer (which are notoriously subject to confounding by pulmonary vascular disease in SSc). The findings, including those during follow-up, were robust when an alternative logistic analytic strategy was used. Overall, the follow-up data reinforce the credibility of the earlier report and establish that therapeutic benefits from short-term oral cyclophosphamide do not endure for very long. Taken together, the two reports clearly show that the treatment effect amounts largely to the prevention of progression of fibrotic disease. During the first year, group differences were ascribable to declines in FVC in the placebo group, and particularly in patients with more extensive fibrosis on computed tomography. After cessation of cyclophosphamide, the therapeutic effect was lost because disease progression occurred selectively in previously treated patients with more advanced disease. The observations amply justify the pathogenic concept of an alveolitis in SSc—that inflammation precedes and leads to fibrosis— but, paradoxically, it appears that treatment benefits across SSc apply more to predominantly fibrotic than to inflammatory disease. If so, the notion that reversible pulmonary disease must be present, to justify prolonged therapeutic intervention in SSc, is incorrect. The findings of Tashkin and colleagues are consistent with the view that, for patients with progressive pulmonary fibrosis, stability is tantamount to therapeutic success. Herein lies the problem. It is correct to question whether the average absolute effect on FVC of 2.5% is worth the price of cyclophosphamide toxicity, especially if the benefit of treatment is transient. However, it can be argued that an "average treatment effect" is a meaningless mean value when extrapolated to individual patients in clinical practice. Conventionally, the great desideratum in therapeutic studies is placebo-controlled evaluation. In reality, placebo-controlled studies are subject to major selection biases, especially when open therapy is available. In a recent placebo-controlled study of intravenous cyclophosphamide in SSc (5), patients and referring physicians preferred open therapy when disease was overtly severe or progressive (personal communication, R. K. Hoyles). In the studies of Tashkin and colleagues, the FVC at baseline exceeded 70% of predicted in approximately half the cohort and, during the follow-up period, open therapy was thought by the primary treating physician to be necessary in only 24 of 113 patients. If severe, more progressive disease is indeed seriously underrepresented, the attainment of complete stability with treatment should not be discounted, simply because an average treatment effect is small. For these reasons, the small but significant benefits of infliximab in sarcoidosis (2), and marginally significant benefits of intravenous cyclophosphamide in SSc (5), should not be dismissed too readily. How, then, can clinicians apply the findings of this study to routine practice? Crucially, the treatment effect was largely confined to patients with more advanced disease, and was therefore much larger in this subgroup. In SSc, an FVC benefit of approximately 9% was reported in open trials of oral cyclophosphamide (6). If patients with less extensive and progressive disease are selectively enrolled in a placebo-controlled study, those with aggressive disease, justifying vigorous therapeutic measures, may be overrepresented in pilot reports of open treatments. If so, it can be argued that the true "average treatment effect" of oral cyclophosphamide is likely to lie somewhere between that in the study of Tashkin and colleagues and that reported in open pilot studies: clinicians should, therefore, reasonably hope for a much larger benefit in patients with severe or overtly progressive disease. The follow-up data provided by this study are also highly informative when the balance between therapeutic efficacy and toxicity is considered. In patients with more extensive fibrosis and a higher likelihood of disease progression, early benefits from treatment are more likely to outweigh the cumulative risk of long-term complications, such as malignancy. However, the dissipation of most benefits of cyclophosphamide after 18 months strongly suggests that prolonged immunosuppression is needed to maintain stabilization. Previous studies of long-term daily oral cyclophosphamide treatment of lupus nephritis and Wegener's granulomatosis, conditions that have a high relapse rate when treated for only 1 year, document increasingly unacceptable levels of toxicity associated with longer treatment duration (6, 7). In the long-term National Institutes of Health trial of immunosuppression for lupus nephritis, daily oral cyclophospamide treatment, when compared with monthly intravenous cyclophosphamide (which markedly reduces cyclophosphamide exposure), was associated with a higher 20-year mortality (0.58% vs. 0.20%) (8). In Wegener's patients treated with daily cyclophosphamide for 2 to 3 years, with cumulative doses sometimes exceeding 100 g, severe complications included amenorrhea, cutaneous malignancy, hemorrhagic cystitis, myelodysplastic syndromes, and bladder cancer (9). Three strategies used to minimize cyclophosphamide exposure in rheumatic disease are potentially applicable to lung disease in SSc:
Given the likelihood that patient selection will favorably alter the risk/benefit ratio of the protocol used by Tashkin and coworkers, and the good prospects for developing a modified regimen of longer duration that is better tolerated, this study signals an improved outlook for patients with scleroderma lung disease. FOOTNOTES
Conflict of Interest Statement: A.U.L. received REFERENCES
Related articles in AJRCCM:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Deceased. 
2,000 for speaking at conferences organized by Actelion. W.J.M. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript.