Editorial 

May 9, 2023

The Future of Clinical Trials in Idiopathic Pulmonary Fibrosis

Anna J. Podolanczuk, Luca Richeldi, Fernando J. Martinez

JAMA. 2023;329(18):1554-1555. doi:10.1001/jama.2022.23955

Idiopathic pulmonary fibrosis (IPF) currently affects about 100 000 individuals in the US, and millions globally. Patients living with IPF experience disabling symptoms and progressive loss of lung function.¹ The 2 treatments approved by the US Food and Drug Administration to treat IPF, nintedanib and pirfenidone, slow disease progression but do not stop or reverse lung fibrosis.^2,3 Both drugs are poorly tolerated by a substantial number of patients, and use is further limited by high cost and lack of perceived benefit on symptoms and quality of life. Many patients with IPF are eager to participate in clinical trials. Yet in the era of approved antifibrotic therapy, patients and their physicians must consider how potential investigational drugs will fit into the current standard of care.

In this issue of JAMA, the results from the ISABELA 1 and ISABELA 2 randomized clinical trials,⁴ one of the first phase 3 clinical trial programs in IPF to be completed since the approval of nintedanib and pirfenidone, are presented. The 2 identically designed trials randomized a total of 1306 patients to 1 of 2 doses of oral ziritaxestat (an autotaxin inhibitor that lowers levels of lysophosphatidic acid, which is a profibrotic mediator) or placebo. The trials allowed the continuation or initiation of background antifibrotic therapy.

The trials were terminated early after an interim analysis raised safety concerns and revealed a lack of efficacy in the treatment groups. Neither dose of ziritaxestat demonstrated any benefit on the rate of decline for forced vital capacity (FVC) over 52 weeks. There was no benefit on any of the reported secondary outcomes. In fact, time to first respiratory-related hospitalization, respiratory-related mortality, and first acute IPF exacerbation were all worse in the ziritaxestat groups. All-cause mortality was highest in the high-dose ziritaxestat group, and in the ISABELA 2 trial, all-cause mortality was double that of placebo. Respiratory-related deaths were the primary cause of death.

The ISABELA trials highlight the complexity of conducting IPF clinical trials in the era of approved antifibrotic therapy.⁵ Each trial had 3 groups: high-dose ziritaxestat (600 mg/d), low-dose ziritaxestat (200 mg/d), and placebo. It also had 3 standard of care treatment options: pirfenidone, nintedanib, and neither; there was approximately equal distribution of participants among these 3 options. The study was well designed and conducted. Participants were recruited from 227 sites in 26 countries. There were missed clinic visits due to the COVID-19 pandemic, but this does not seem to have impacted the results. The enrolled patients had a centrally reviewed diagnosis of IPF and mild to moderate disease. There was good target engagement. Despite early termination, the large sample size and small mean differences make it unlikely that a clinically significant beneficial effect was missed.

A close examination of analyses stratified by background standard of care therapy reveals high rates of treatment-related, treatment-emergent adverse events among patients taking both nintedanib and ziritaxestat. Dose reductions and interruptions of nintedanib were more common in those taking high-dose ziritaxestat, and dose reductions or interruptions of the study drug were more common in those taking nintedanib. Doses of pirfenidone and nintedanib were generally lower among participants taking ziritaxestat. These findings were more pronounced in the ISABELA 2 trial, which enrolled a higher proportion of Asian patients (28% vs 6%) due to the geographic distribution of study sites. It is unclear if these differences in the standard of care explain the lack of efficacy of ziritaxestat, but regional heterogeneity and pharmacogenomic interactions should be examined in more detail.

The clear failure of ziritaxestat provides important lessons for future clinical trials in IPF. Late-phase clinical trials must be based on robust and high-quality preclinical and clinical data. The ISABELA trials were designed and conducted after a phase 2a trial demonstrated a smaller decline in FVC at 12 weeks among 23 patients with IPF randomized 3:1 to ziritaxestat vs placebo (ie, 17 patients treated with the active drug and 6 treated with placebo).⁶The phase 2a trial⁶ did not allow for use of background antifibrotic therapy even though ziritaxestat increases plasma levels of nintedanib. The small sample size and short duration of the phase 2a trial⁶ were significant limitations.

Future phase 2 trials must be sufficiently powered to assess safety and heterogeneity of treatment response in the setting of background antifibrotic therapy. Using short-term (3-month) change in FVC can improve the efficiency of IPF clinical trials and may be a reasonable surrogate outcome in a phase 2 study.⁷ A recent analysis found that a 3-month decline in FVC is associated with increased mortality and short-term outcomes predicting disease progression at 1 year.⁸ However, shorter trials need to enroll more patients, and ensure high density of data to adequately demonstrate safety and evaluate surrogate outcomes, as has been done in a recent phase 2 study of a novel oral phosphodiesterase 4B inhibitor.⁹

Clinical trials in IPF must consider racial, ethnic, geographic, and sex differences in disease pathogenesis and response to therapy. Idiopathic pulmonary fibrosis is a complex disease with multiple genetic determinants and molecular pathways contributing to lung fibrosis.¹ Significant heterogeneity in clinical outcomes commonly evaluated in clinical trials (including mortality) exist among different racial and ethnic groups.¹⁰ Racial and ethnic differences in genetic drivers of disease susceptibility and progression have been demonstrated. For example, a single-nucleotide variation in the promoter region of MUC5B has been identified as the strongest known genetic risk factor for IPF among White and Hispanic individuals, but was found to be very rare in a Korean population.^11,12Regional differences in outcomes may also reflect differences in exposures, including air pollution and diet. These issues must be considered and prioritized at each stage of the drug development process to ensure participant safety and generalizability of the results. Studies should be large enough to examine subgroups based on biologically plausible mechanisms. Alternatively, precision-based approaches to identify patients most likely to respond to a therapy using genetics, biomarkers, and imaging features could enhance patient selection for clinical trials, as is currently being done in an ongoing phase 3 trial (funded by the National Institutes of Health) of N-acetylcysteine for patients with IPF and the TOLLIP TT genotype.¹³

Trials should consider incorporating composite outcomes that include clinically relevant events. Change in FVC over 1 year has been accepted by the US Food and Drug Administration as a clinically relevant efficacy measure in IPF, and both currently available antifibrotic therapies were approved based on their effect on this outcome.¹⁴Although a decline in FVC is associated with increased mortality, a reduction in the rate of FVC decline in response to treatment has little clinical meaning for patients and use of FVC as a primary outcome can be complicated by missing data. Alternative measures include the composite outcome of all-cause mortality and respiratory hospitalizations. The feasibility of using this outcome to evaluate the efficacy of antimicrobial therapy for the treatment of IPF was demonstrated in Clean-UP-IPF.¹⁵

Moving forward with these lessons will be challenging, especially as IPF therapies improve and trials need to increase in size and duration to detect smaller treatment effects while patients are continuing to receive standard of care treatment. Innovative approaches to clinical trial design, including adaptive trials and bayesian analyses, may enhance trial efficiency, and we expect that these approaches will be increasingly used. Large trials like the ISABELA trials should collect multidimensional data that will generate new insights about the underlying disease and response to therapy. Industry partners should consider making these data publicly available so that even if the investigational therapy fails, the knowledge generated will ensure the success of future studies.

There is an urgent need for better treatments for IPF. There are many promising drugs in development, stemming from an investment in research into the mechanisms driving lung fibrosis. It is our hope that innovative approaches to clinical trials will improve their efficiency and safety while accelerating the approval of effective therapies so that patients living with IPF may benefit.