Fibroneer Trials: More than Meets the Eye
Article information
The landmark studies of pirfenidone and nintedanib demonstrating effects on reduction in forced vital capacity (FVC) decline and potential benefits in terms of acute exacerbation have led to their use as standard of care anti-fibrotic therapy in the treatment of idiopathic pulmonary fibrosis (IPF) and progressive pulmonary fibrosis (PPF) [1]. Although no survival advantage was observed within the observation period of 52 weeks, recent data using real world IPF cohorts have suggested a trend to survival benefit [2]. However, drug tolerability is often limited by gastrointestinal and hepatotoxic side effects, resulting in discontinuation of therapy [3]. Hence, the pursuit to develop newer agents to target fibrotic interstitial lung disease (ILD), of which the preferential phosphodiesterase 4B (PDE4B) inhibitor (nerandomilast), in a phase 2 randomized double-blind placebo-controlled trial, showed promise in primary endpoint targets by slowing down FVC decline compared to placebo [4].
Fibroneer-IPF trial is a phase 3, double-blind trial, involving 1,177 patients with IPF randomized in a 1:1:1 ratio to receive nerandomilast at a dose of 18 mg twice daily, 9 mg twice daily, or placebo respectively, with stratification according to background anti-fibrotic therapy (nintedanib or pirfenidone) versus none [5]. Fibroneer-ILD trial is a phase 3, double-blind, placebo-controlled trial involving 1,176 patients with PPF randomized in a 1:1:1 ratio to receive the PDE4B inhibitor, nerandomilast at a dose of 18 mg twice daily, 9 mg twice daily, or placebo respectively, with stratification according to background anti-fibrotic (nintedanib) versus no anti-fibrotic therapy. In both related trials, the primary end point was the absolute change from baseline in FVC, measured in milliliters (mL), at week 52 [6]. Secondary endpoints included time to first acute exacerbation, hospitalization for respiratory cause, or death, over the duration of the trial.
We limit our discussion to patients on background anti-fibrotic therapy who received nintedanib, due to observed drug interactions between pirfenidone and nerandomilast during the study.
Both trials met the primary endpoint for efficacy. Although secondary endpoints were not met as a composite, signals for reduction in secondary endpoint events (death, acute exacerbation or death, absolute FVC decline >10% or death) were observed, more evidently in Fibroneer-ILD. In terms of drug safety, liver function abnormalities previously reported with the use of nintedanib were not significant, compared to historical nintedanib trials [3,7]. This may be more important in the Asian context, particularly, where liver function abnormalities (14.4%) were significantly more common [3,8], compared to the overall population (4.9%–5.2%) [7].
What initially seemed to be puzzling, however, was that FVC decline in patients from the placebo group on background anti-fibrotic therapy was greater than that of the placebo group not on background anti-fibrotic therapy. Interestingly, the magnitude of FVC decline on background therapy (–191.6 mL, Fibroneer-IPF; –180.9 mL, Fibroneer-ILD) was similar to that of treatment naive groups in the historical INPULSIS 1 and 2 (–223.0 mL, pooled)7 and INBUILD (–187.8 mL) trials [9]. This observation was made in both trials, and appeared to contradict our knowledge of current anti-fibrotic drug efficacy.
We would like to share our thoughts on this:
(1) Disease duration and trajectory at baseline are important considerations in data interpretation. In Inpulsis 1&2 trials, mean disease duration of patients were 1.6 and 1.7 years respectively, compared to mean disease duration in Fibroneer-IPF and Fibroneer-ILD of 3.5 and 4.2 years, respectively (18.2 months on anti-fibrotic therapy in Fibroneer-ILD). Additionally, 17% of patients in Fibroneer-IPF (19% on background anti-fibrotic therapy and 12% without background therapy) were on oxygen therapy at baseline, compared to less than 10% in pooled data of Inpulsis 1&2.
Given the natural history of IPF with median survival time of 3 years from diagnosis [10], and recent data showing similar survival patterns in PPF [11], we suggest:
• Patients on background anti-fibrotic therapy likely represented those who had more advanced disease, due to inherent (genetic), or acquired drug resistance to background anti-fibrotic therapy over time. Doctors are more likely to enroll patients already failing standard therapy. This explains FVC decline in the placebo group appearing similar to historical placebo groups, as well as less impactful effects of FVC decline with nerandomilast (Figure 1).
Change in forced vital capacity at week 52 in the placebo group in idiopathic pulmonary fibrosis (IPF) and progressive pulmonary fibrosis trials. ILD: interstitial lung disease.
• Patients not on background anti-fibrotic therapy, given the same disease duration, may have represented more indolent disease or an early progressive trajectory, generally not requiring treatment. This explains modest FVC decline in the placebo group, and more favorable drug-response effects on FVC decline with nerandomilast appearing similar to historical treatment groups (Figure 1).
• The magnitude for relative reduction of FVC decline in the Fibroneer-ILD treatment group not on background therapy (47% [9 mg twice a day, bid]; 38% [18 mg bid]) was notably inferior to the historical INBUILD trial (57%). The reason for this is less clear. However, surrogate indicators of disease trajectory, and disease duration of INBUILD subjects from the time they developed a PPF phenotype was not known.
• Although the trial did not achieve composite secondary endpoints, positive signals on risk reduction of death in Fibroneer-ILD, but not in Fibroneer-IPF trial with higher dose (18 mg bid) nerandomilast, may suggest a role for anti-inflammatory modulation of the drug. In reality, however, the distinction between IPF and non-IPF ILD, especially in the absence of histopathological evidence, may not always be so clear. Therefore, it may be important to analyse if there was a predominant subgroup of PPF patients where a survival advantage may have been observed.
(2) The early curve separation in FVC decline at 2 weeks between treatment and placebo groups in the overall population was particularly impressive, and may suggest a role for the use of nerandomilast in acute exacerbation. However, it was also observed that at 24 weeks, treatment and placebo curves appeared to parallel in both Fibroneer trials. This contrasted with FVC curves in Inpulsis 2, which showed continued curve separation. However, similar trends in treatment and placebo FVC curves noted in Inpulsis 1 and INBUILD at week 36 did not translate to reduced drug efficacy in extension trials [12,13]. As such, the potential long-term efficacy of nerandomilast should not be discredited, while we eagerly await the results of Fibroneer-On.
Notes
Authors’ Contributions
Conceptualization: all authors. Methodology: Teo FSW. Formal analysis: all authors. Data curation: Teo FSW. Project administration: all authors. Visualization: all authors. Software: Song JW. Validation: all authors. Investigation: all authors. Writing - original draft preparation: Teo FSW. Writing - review and editing: all authors. Approval of final manuscript: all authors.
Conflicts of Interest
Jin Woo Song is an associate editor of the journal, but he was not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflicts of interest relevant to this article were reported.
Funding
No funding to declare.