The clinical significance of 5% change in vital capacity in patients with idiopathic pulmonary fibrosis: extended analysis of the pirfenidone trial

Abstract

Background: Our phase III clinical trial of pirfenidone for patients with idiopathic pulmonary fibrosis (IPF) revealed the efficacy in reducing the decline of vital capacity (VC) and increasing the progression-free survival (PFS) time by pirfenidone. Recently, marginal decline in forced VC (FVC) has been reported to be associated with poor outcome in IPF. We sought to evaluate the efficacy of pirfenidone from the aspects of 5% change in VC.

Methods: Improvement ratings based on 5% change in absolute VC, i.e., "improved (VC ≥ 5% increase)", "stable (VC < 5% change)", and "worsened (VC ≥ 5% decrease)" at month 3, 6, 9 and 12 were compared between high-dose pirfenidone (1800 mg/day; n = 108) and placebo (n = 104) groups, and (high-dose and low-dose (1200 mg/day; n = 55)) pirfenidone (n = 163) and placebo groups. PFS times with defining the disease progression as death or a ≥ 5% decline in VC were also compared between high-dose pirfenidone and placebo groups, and low-dose pirfenidone and placebo groups. Furthermore, considering "worsened" and "non-worsened (improved and stable)" of the ratings at months 3 and 12 as "positive" and "negative", respectively, and the positive and negative predictive values of the ratings were calculated in each group.

Results: In the comparison of the improvement ratings, the statistically significant differences were clearly revealed at months 3, 6, 9, and 12 between pirfenidone and placebo groups. Risk reductions by pirfenidone to placebo were approximately 35% over the study period. In the comparison of the PFS times, statistically significant difference was also observed between pirfenidone and placebo groups. The positive/negative predictive values in placebo and pirfenidone groups were 86.1%/50.8% and 87.1%/71.7%, respectively. Further, the baseline characteristics of patients worsened at month 3 had generally severe impairment, and their clinical outcomes including mortality were also significantly worsened after 1 year.

Conclusions: The efficacy of pirfenidone in Japanese phase III trial was supported by the rating of 5% decline in VC, and the VC changes at month 3 may be used as a prognostic factor of IPF.

Trial registration: This clinical trial was registered with the Japan Pharmaceutical Information Center (JAPIC) on September 13th, 2005 (REGISTRATION NUMBER: JAPICCTI-050121).

Figure 1

Categorical analysis based on 5% changes in VC at months 3, 6, 9, and 12. Improvement ratings based on 5% changes in VC were defined as "improved (VC 5% increase)", "stable (VC < 5% change)", and "worsened (VC 5% decrease)", using VC values measured at months 3, 6, 9, and 12. a) high-dose vs. placebo groups, b) pirfenidone-treated (high + low-dose) vs. placebo groups. The results are shown by the frequencies of improved (white areas), stable (gray areas), and deteriorated (black areas). P-values by Wilcoxon's test are indicated at the right.

Figure 2

Kaplan-Meier plot of Progression-Free Survival (PFS) times in groups of IPF patients. a) The disease progression was defined by a ≥ 5% decline in VC from baseline or death. b) The disease progression was defined by a ≥ 5% decline in VC from baseline on two consecutive occasions or death. Solid line: high-dose; broken line: low-dose; bold broken line: placebo. The distribution of PFS times were compared with log-rank test.

Figure 3

K-M plot of PFS times with origin at month 3 in groups of patients with and without 5% decline in VC at month 3. Solid line with closed circle: No decline in VC at month 3; broken line with plus: a ≥ decline in VC at month 3. P-value was 0.8835 with log-rank test.