© 2007 American Thoracic Society doi: 10.1164/rccm.200702-226ED
Biomarkers for Cystic FibrosisAre We Progressing?National Heart and Lung Institute, Imperial College, London, United Kingdom Cystic fibrosis (CF) research has reached an interesting stage. On the basis of improvements in conventional clinical care, both quality and quantity of life have markedly improved. Large clinics are reporting declines in lung function of no more than 1% per year (1), and the predicted median survival of a child born today with CF is around 50 years (2). Although a great deal of understanding of the underlying pathophysiology has been unraveled since the cloning of the CF gene 17 years ago, new treatments based on the molecular defect in the CF transmembrane conductance regulator (CFTR) protein have yet to reach the patient. Given the average of 20 years to bring a product to market, this rate of progress is unsurprising and predictable. Encouragingly, large programs of small molecule screening and gene therapy were initiated several years ago. Both have identified lead compounds, and are nearing phase II studies. Two obvious questions arise. First, in light of noted improvements in CF outcome, do we need new treatments based on the molecular defect? The gain in survival that conventional management has achieved comes at a high cost—namely, a huge burden of treatment for patients and their families, and significant interruptions to daily life. New treatments with the potential to prevent rather than treat lung disease could improve both quantity and quality of life. Second, and common to either small molecule or gene-based approaches, is the much more difficult question, "How will I know when my treatment has been clinically successful?" Clearly, from the perspective of an uninitiated observer, the answer is straightforward: "When an increase in life expectancy and/or quality of life as measured by treatment burden, time in hospital, or similar has been reached." Sadly, we can't both have our cake and eat it. Power calculations show that unfeasibly large cohorts of patients need to be studied over protracted periods to show statistically significant benefits in the above parameters. The field has, therefore, turned to a surrogate of these clinical symptoms, namely FEV1. Importantly, this measurement of lung function reflects both acute and chronic changes in CF, and it may be crucial to distinguish these when assessing treatment effects. Symptomatic treatments, such as mucolytics, are predominantly aimed at the downstream effects of CFTR dysfunction, when infection predominates. Acute changes in FEV1 caused by factors such as mucus rearrangement have been clearly demonstrated and provide a short-term, fairly accessible endpoint (3). In contrast, novel therapies aimed at basic cellular defects in CF aim to alter lung disease at more upstream stages. Here, FEV1 may reflect more primary links to CFTR dysfunction, such as altered mucociliary clearance and inflammation. Given the chronicity of this pathophysiology, novel treatments are unlikely to provide a "quick fix"; they will need to be given over protracted time courses. FEV1 may reflect this with slower changes in rate of progression rather than improvements. Statistical calculations suggest that the use of FEV1 in this setting may be in reach for large multicenter trials (4). In contrast, the typical, academically led studies, which underpin so many of the earlier phase trials, will still be woefully underpowered. This has led to an extensive search for new outcome measures that would be more "sensitive" than FEV1 (i.e., able to distinguish a change in chronic pathophysiology with lower patient numbers and/or shorter duration of treatment). An obvious caveat is that it may be better, at least theoretically, to seek a surrogate for lung disease, rather than a surrogate for a surrogate (FEV1). The study by Mayer-Hamblett and colleagues in this issue (pp. 822–828) exemplifies this quest (5). Using a large cohort of patients, collected from four separate studies, the study shows that inflammatory markers in sputum correlate with FEV1, when studied cross-sectionally. Specifically, neutrophil count and neutrophil elastase account for the majority of the correlation, in keeping with the domination of this cell type in the endobronchial inflammation seen in CF. Such findings are encouraging, and are complementary to those reported in smaller studies that suggested similar links between inflammatory markers and FEV1 (6). Clearly, however, the link may not be causal. A further issue is correlating sputum sampling at a single time point in the evolution of lung disease with FEV1, which reflects the lifetime history of the problem. This issue can, at least in part, be ameliorated by longitudinal measurements. Additional potential surrogates to follow in CF have been suggested by other groups. In general, such surrogates fall into the categories of inflammation, infection, imaging, and lung physiology. However, in our view, there are two as yet unmet and crucial challenges to the field. First, do any of these young pretenders provide a more sensitive marker of lung disease for a novel therapy aimed at the basic defect? Second, how does the optimal one(s) correlate with assays of CFTR molecular function? The U.K. CF Gene Therapy Consortium (7) has grappled with these two challenges over a number of years. However, we have been unable to produce a shortcut to a laborious, expensive, and time-consuming program. To address the first question of identifying highly sensitive assays, we have designed a large, multidose, double-blind, placebo-controlled trial administering gene therapy over a 1-year period, assuming that this may be the minimal period over which changes in chronic lung pathophysiology have a reasonable chance to become apparent. Before treatment, a larger group of some 200 patients with CF, including children, will be longitudinally assessed over a 12- to 18-month period; approximately 100 of these subjects will then move forward into the gene therapy trial. We have assembled a large group of CFTR "clinical function" assays that will be assessed at multiple time points through both stages of this program. To address the second question, we will look for correlations between these clinical function parameters and a basket of assays related to CFTR "molecular function," including bioelectric measurements, consequent changes in airway surface liquid height, and bacterial adherence. Many groups have shown alterations in some of these markers using either small molecule or gene-based therapies (8, 9), and some of them are likely to be more sensitive than clinical assays given the high degree of variability in clinical status demonstrated by patients. However, it is unclear whether the magnitude of such changes will in any way predict alterations in assays of CFTR clinical function. Our trial will, therefore, incorporate bronchoscopic assessment of CFTR molecular function from the lower airways, and may allow us to gain a handle on how, and if, CFTR molecular function relates to more relevant clinical assays. If a relationship is established, smaller, easier, and cheaper trials using sensitive molecular function endpoints may be able to provide signposts to the clinical relevance of a potential new therapy. Both the Cystic Fibrosis Trust, through the U.K. CF Gene Therapy Consortium, and the Cystic Fibrosis Foundation, through the Therapeutic Development Network, have put into place ways in which assay development can be catalyzed. Patient databases are increasingly being unified, and one of these, Port CF, has recently been accepted by a large proportion of those in the CF field. This may, in time, allow a more universal combination of data on novel assays, likely of benefit to patients and researchers. There is still a considerable amount of hard work needed over the next few years, but CF biomarkers are progressing well. FOOTNOTES Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
Related articles in AJRCCM:
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
