Icenticaftor, Novel Therapy for COPD: This Glass Is Half Full

Icenticaftor is a cystic fibrosis (CF) transmembrane regulator (CFTR) modulator that has been tested in patients with chronic obstructive pulmonary disease (COPD). Although a large phase II clinical trial failed to achieve its primary outcome, improvement in FEV₁ at 12 weeks, the study provides substantial evidence of response for other outcomes. Further, the available data suggest that a subset of participants in the trial accounted for the majority of the potential benefit. This suggests that development of a companion diagnostic to a priori identify potential responders will be key to efficient development and implementation of this therapy. This strategy may have general applicability in addressing the daunting hurdles that have stymied development of novel treatments for COPD.

In this issue of the Journal, Martinez and colleagues (pp. 417–427) report the results of a large (N = 974) phase II clinical trial of icenticaftor (1), a CFTR activator, in patients with COPD (2). The study failed to meet its primary goal. As with most large studies with multiple endpoints, it produced a rich set of secondary data, in this case suggesting promising future paths for the program. Importantly, the study also highlights challenges and opportunities in developing novel treatments for COPD, which are urgently needed but difficult to develop. The icenticaftor study suggests a way forward.

The study had a strong rationale. Mutations in the CFTR gene cause CF. CFTR protein modulates mucus hydration largely by regulating the secretion of anions. CFTR activity is itself regulated, most importantly by cAMP acting through protein kinase A. Modulators of CFTR processing and activity have dramatically benefitted CF treatment. In patients with CF who have the G551D mutation, improved FEV₁ over a period of several months of ivacaftor treatment helped lead to its approval by the U.S. Food and Drug Administration (3). The mechanism of FEV₁ improvement by ivacaftor is believed to be indirect, due to reduced burden of secretions and inflammation. Long-term, CFTR modulators appear to favorably alter the natural history of CF. Crucially, beneficial effects of the increasing number of CFTR modulators depend on the specific mutation affecting CFTR (4).

Altered CFTR functions are also well described in COPD (5, 6). First, CFTR mutations are relatively common, suggesting that many patients with COPD will be heterozygous for an abnormal gene. Additionally, inflammation or other oxidant stressors can lead to acquired CFTR defects. Finally, because CFTR activity can be modulated, even with normal CFTR function, pharmacologic modulators might be beneficial in COPD even if CFTR itself is normal.

Secretions are a common and often troubling feature of COPD (7). They likely contribute to ventilation inhomogeneity, leading to impaired gas exchange and compromised airflow. They are also probably a factor in an altered lung microbiome having the potential to cause progressive airway disease. No available treatment has been meaningfully demonstrated to benefit secretions in COPD. Together, this body of evidence provides a very compelling rationale to explore a CFTR modulator in COPD.

The primary outcome of the study by Martinez and colleagues was to identify a dose of icenticaftor that significantly improved airflow (measured as FEV₁) at 12 weeks. None of the tested doses achieved this goal. However, some positive findings were cautiously but optimistically presented. At 24 weeks, a dose response was observed for trough FEV₁, with an improvement of 0.035 L (significance not adjusted for multiple comparisons). Similar improvements were observed with the 300-mg dose for several cough and symptom scores, rescue medication use, and fibrinogen.

A pessimistic, glass-half-empty view might be that the study failed on its primary outcome and secondary descriptive analyses were modest at best; a 35-ml improvement is not impressive. The responder analysis demonstrated that 28% of participants treated with 300 mg exhibited the nominal minimal clinically important difference for trough FEV₁ of 100 ml; there were also more responders for the Evaluating Respiratory Symptoms in COPD cough and sputum score (odds ratio, 1.52) and total score (1.45) compared with placebo. However, an increase in “responders” from 40% to 50% would be difficult to recognize clinically, particularly for symptoms that are variably present.

A much more optimistic view leads directly to approaches to address the hurdles that have stymied development of novel COPD treatments. Like many COPD studies, the study by Martinez and colleagues used FEV₁ as the primary outcome and enrolled a broad population of participants. FEV₁ is a default measure in COPD trials because it is relatively easy to measure and is well-recognized by regulators. Additionally, icenticaftor’s effect on FEV₁ was observed in a prior 4-week proof-of-concept study (8). However, as with ivacaftor, icenticaftor likely improves FEV₁ indirectly, e.g., by reducing mucus-induced ventilation inhomogeneity, associated inflammation, or both.

More “direct” outcomes of the effect of icenticaftor might be seen in terms of symptoms or objective measures of cough or sputum. Although several symptom scores were included, the study was not optimized to maximize potential benefits. Eligible participants were current or ex-smokers >40 years of age with chronic bronchitis (based on 3 months of symptoms in the prior 2 years), a COPD Assessment Test score >10, evidence of airflow limitation, and long-acting muscarinic antagonist/long-acting β2-agonist/inhaled corticosteroid treatment for 3 months before enrollment. These criteria assured that all participants would have an impaired FEV₁ that could potentially improve, which was key for the primary outcome. However, the enrollment strategy did not assure that all participants would have cough and sputum. Selection for meaningful symptoms would have improved the power to assess this outcome.

Secondary analyses identified no responder population using the available baseline clinical and demographic data, yet they provide some evidence that responders existed. The strength of this evidence could be improved by further analysis. Were individuals who showed an FEV₁ response also responders for cough and sputum measures? Were the several included cough and sputum measures concordant? Concordance of response, particularly for different domains, would increase confidence that the overall study actually contained a responder population.

Identifying responders is a key strategy seldom exploited in COPD studies. In contrast, drug-development programs in other therapeutic areas, particularly oncology, routinely include the development of a companion diagnostic to identify potential responders (9–11). Ideally, the diagnostic and therapeutic programs are integrated, and the required science, which can be substantial for novel mechanisms and/or drug classes, is developed in parallel. The data already available from this study should provide some insight into whether a responding population is present. However, such a post hoc identifier will require validation in a new therapeutic trial.

As noted, the rationale for testing CFTR modulators in COPD is strong. This same evidence should also guide the development of companion diagnostics to identify responders. Did participants have impaired CFTR function, either genetic or acquired? Did they have abnormal retention of secretions leading to inhomogeneity of ventilation? There is a stronger rationale that these parameters could identify potential responders versus the “usual-suspect” demographic variables. Equally importantly, improving the responder population from 28% to 60%, for example, would dramatically reduce the sample size (and cost) of a required phase III study.

COPD is extraordinarily heterogeneous (12, 13). Even individuals with shared symptoms, e.g., cough and sputum production, probably comprise multiple subtypes with differing mechanisms. These subtypes will likely respond differently to medications, as observed for CF mutations and CFTR modulators. Despite the failure of icenticaftor to attain the primary outcome of the study, there is strong evidence of benefit and that a subset of patients will experience most of the benefit. Without a doubt, icenticaftor is only one of many emerging treatments targeting novel mechanisms in COPD. The icenticaftor study, despite being “negative,” shows promise. It also demonstrates that the integration of a companion diagnostic program to identify the responding population should be an essential part of drug development in COPD. An integrated strategy that addresses COPD heterogeneity has the potential to help address the daunting challenge of developing new treatments for COPD.