Is Idiopathic Pulmonary Fibrosis Now Treatable?

Idiopathic pulmonary fibrosis (IPF) is a disease that almost inevitably progresses, and some have suggested that no treatments exist that modify its course (1, 2). But this view requires qualification. Although improvements in physiologic outcome measures in response to treatment have been disappointing, it is also fair to say that there have been no prospective studies of antiinflammatory drugs involving sufficiently large numbers of patients on which to make a clear statement of efficacy; specifically, there has never been a study of the efficacy of corticosteriods in a randomized controlled trial or a controlled clinical trial (3, 4). By the same token, however, there is no evidence that antiinflammatory drugs have no effect, particularly on disease stabilization. After all, coarse reticular change on CT never improves (5), and the best response in many cases is the prevention of disease worsening (i.e., disease stability). Nonetheless, based on the perception of lack of efficacy of antiinflammatory agents, there has been increasing momentum to explore agents that might modulate the most obvious and prevalent component of the histopathology-fibrogenesis (6).

In the last year, we have seen the results of the largest randomized double-blind placebo-controlled study of any drug in idiopathic pulmonary fibrosis, interferon-? (7). These results were somewhat disappointing. The primary endpoint of progression-free survival time (defined as improvement in time to mortality, worsening of predicted forced vital capacity = 10%, or an increase in A-a gradient = 5 mm Hg) was not met, although it could be argued that at least one of the endpoints was insufficiently reproducible to provide reliable measures (8). Nonetheless, there were a number of exploratory endpoints that have encouraged the launch of a second study of interferon-?, recruitment for which is underway. Two other studies of antifibrogenic drugs, anti-tumor necrosis factor-a (Etanercept) and an anti-endothelin dual receptor antagonist (Bosentan) have completed enrollment, and results should be available during the forthcoming months. A study of antioxidants (NAC) has also been completed.

With this background, it is encouraging to read the report in this issue of the American Journal of Respiratory and Critical Care Medicine on the efficacy of pirfenidone in IPF. Azuma and coworkers (pp. 1040-1047) report the results of a double-blind, placebo-controlled, multicenter trial of pirfenidone in Japanese patients with IPF (9). Although the primary endpoint, the difference in the change in lowest oxygen saturation by pulse oximetry between baseline and 9 months during a 6-minute exercise test, was not achieved, there was a trend in favor of pirfenidone at this time point. However, in a prespecified subset of patients, the lowest Sp^sub O^sub 2^^ at 6 and 9 months was better in the pirfenidone group. Of the other secondary endpoints, the number of individuals with improved or stable vital capacity and total lung capacity was greater in the pirfenidone group than the placebo group. Of importance, the Data Safety Monitoring Board advised the cessation of the study before completion because five patients in the placebo group had acute exacerbations of disease during the course of the study, by contrast with none in the pirfenidone group.

Does this mean that we now really have an effective treatment for IPF? There is no doubt that this study provides significant encouragement, but there are certain caveats. A shortcoming in the study, the use of a 6-minute exercise test that has not been validated in this disease, is a potential problem; the authors acknowledge this. Although this is a logical patient-friendly functional assessment of outcome, we have no idea how robust this measurement is, nor whether it can be reproduced in other centers and how it correlates to more established indices of change. This needs to be explored more formally. Furthermore, the mean changes in lowest Sp^sub O^sub 2^^ are quite small (< 2% change) in the subgroup of individuals who completed the 6-minute exercise test at the 6-month time point, the subset with the most significant difference between placebo and pirfenidone. With these residual uncertainties, I suggest that we cannot recommend this drug as first line therapy at the present time, even though a previous uncontrolled open label study by one of the authors hinted at a pirfenidone benefit (10).

The problem of endpoint raises a broader issue; what is the most appropriate measure of change to be used in clinical trials? At the time of the inception of this study, the area was murky. Since then, there have been three studies that have shown the robustness of change in FVC and DL^sub CO^ as a measure of likely subsequent outcome in IPF (11-13). Together with one study that showed that desaturation on 6-minute walk testing was a strong predictor of mortality (14), these data provide quite compelling support for the use of physiologic indices. What we really need, however, is a good serologic index that is repeatable, reliable, and accurate and that is a surrogate marker predicting progression. To date none has emerged but, of the candidates, KL-6 shows some promise (15).

The study of Azuma and colleagues highlights another important issue. This is now the second large study conducted in a proper prospective fashion that has evaluated IPF and been published. In the context of one other study under review and two others that will be analyzed hopefully during this coming year, we have finally achieved a significant goal-moving away from the uncontrolled anecdote that has historically "informed" therapeutic decisions on IPF. Further armed by the study of Azuma and coworkers, we need to move forward in cementing a collaborative network of international clinicians to put into place an iterative approach to clinical trials of novel therapy. IPF has a mortality greater than some cancers, and we need to follow the model of cancer trials; comparison of current "best therapy" with novel drugs. These studies need to be large-hence the collaborative-and powered to get answers. A first step would be a larger, more international, study of pirfenidone.

Conflict of Interest Statement: R.M.d.B. has received consultancy and Advisory Board fees from Intermune, Actelion, and Centocor and lecture fees from Intermune, Actelion, GlaxoSmithKline, and AstraZeneca.