Association of hospital admission and forced vital capacity endpoints with survival in patients with idiopathic pulmonary fibrosis: analysis of a pooled cohort from three clinical trials

Abstract

Background: Mortality is an impractical primary endpoint for clinical trials in patients with idiopathic pulmonary fibrosis who have mild-to-moderate physiological impairment because event rates are low. Change in forced vital capacity (FVC) is widely accepted as a surrogate for mortality and is the most common primary endpoint in clinical trials for this disorder. Use of hospital admission as a predictor for mortality, independent of FVC decline, has not been well defined. We aimed to ascertain the independent and combined association of hospital admission and at least a 10% decrease in FVC with all-cause mortality.

Methods: We did a pooled cohort study of 517 patients with idiopathic pulmonary fibrosis from three IPFnet multicentre randomised controlled trials. We compared the incidence of non-elective hospital admission and a 10% or greater reduction in FVC across strata of baseline physiological impairment. We used Cox proportional-hazards models to assess the risk of all-cause mortality associated with these surrogate events, occurring up to a predefined landmark timepoint. The three studies are registered at ClinicalTrials.gov, numbers NCT00650091, NCT00517933, and NCT00957242.

Findings: Seven patients died before the landmark timepoint. Of the 510 patients remaining, 38 (7%) were admitted to hospital up to the predefined timepoint and 58 (11%) had a categorical decrease in FVC of at least 10%. Most patients admitted to hospital did not have a 10% or greater decrease in FVC (30 vs eight). Both surrogate events were associated with subsequent time to death from any cause (hazard ratio [HR] for admission 4·05, 95% CI 1·36-12·11 vs HR for 10% or greater decline in FVC 4·68, 1·83-11·99). When causes of hospital admission were considered, only respiratory events were associated with mortality (5·97, 1·81-19·74).

Interpretation: Hospital admission might be an appropriate component of a clinically meaningful composite endpoint that improves the feasibility of clinical trials in idiopathic pulmonary fibrosis. Further studies are needed to refine the most appropriate definition of hospital admission for future trials.

Funding: US National Heart, Lung, and Blood Institute (NHLBI), and The Cowlin Family Fund at the Chicago Community Trust.

Conflict of interest statement

DECLARATION OF INTERESTS

HRC reports personal fees from Bayer, Biogen, Boehringer Ingelheim, FibroGen, Genentech, Gilead, Genoa, Inter Mune, Pfizer, Promedior, and Moerae Matrix, outside the submitted work. KKB reports grants from NIH/NHLBI during the conduct of the study, grants from NIH/NHLBI outside the submitted work, grants and personal fees from Actelion, Amgen and Gilead outside the submitted work, personal fees from Almiral, Altitude Pharma, Astra Zeneca, Bayer, Biogen/Stromedix, Boehringer Ingelheim, Celgene, Centocor, Fibrogen, Galecto, Glaxo Smith Kline, MedImmune, Novartis, Pfizer, Promedior, Roche/Genentech, Sanofi/Genzyme, and Veracyte outside the submitted work, and non-financial relationships with Bristol-Myers Squibb, GeNO, Genoa, Mesoblast, and Moerae outside the submitted work. KRF reports grants from NIH/NHLBI during the conduct of the study, board membership and personal fees from Boehringer Ingelheim outside the submitted work, personal fees from Fibrogen, Genentech, Gilead, Ikaria, Immune Works, Med Immune, Novartis, Takeda, Vertex, Veracyte, Roche, the Pulmonary Fibrosis Foundation, and UpToDate outside the submitted work, and institutional fees from Immune Works, Inter Mune and Bristol-Myers Squibb outside the submitted work. TEK reports personal fees from Inter Mune, Immune Works, Boehringer Ingelheim, Glaxo Smith Kline, Daiichi Sankyo, and UpToDate during the conduct of the study but outside the submitted work. SMP reports grants from Bristol-Myers Squibb and Gilead, outside the submitted work. GR reports grants from NIH/NHLBI during the conduct of the study, consultancies for Biogen, Boehringer Ingelheim, Centocor, Fibrogen, Gilead, Inter Mune, Actelion, Promedior, and Veracyte outside the submitted work, and data safety monitoring board membership for MedImmune outside the submitted work. KJA reports grants from NIH/NHLBI during the conduct of the study. FJM reports grants from NIH/NHLBI during the conduct of the study, steering committee membership for studies funded by Bayer, Centocor, Gilead and Promediorduring the conduct of the study but outside the submitted work, personal fees from Ikaria, Genentech, Nycomed/Takeda, Pfizer, Vertex, MedScape, Stromedix/Biogen, Axon Communications, Johnson & Johnson, Genzyme, the National Association for Continuing Education, Boehringer Ingelheim, Veracyte, the American Thoracic Society, Inova Health System, Spectrum Health System, and the University of Texas-Southwestern during the conduct of the study but outside the submitted work, and personal fees from Forest, Janssen, Glaxo Smith Kline, NycoMed/Takeda, Genentech, Actelion, Amgen, Astra Zeneca, Carden Jennings, CSA Medical, Ikaria/Bellerophon, Merck, Pearl, Roche, Sudler& Hennessey, CME Incite, the Center for Healthcare Education, Miller Medical, Paradigm, Peer Voice, Projects in Knowledge, Bayer, Forest, Grey Healthcare, Merion, Informa, Annenberg, UpToDate, MedScape, the American College of Chest Physicians, Inova Health System, St. John’s Hospital, St. Mary’s Hospital, the University of Illinois-Chicago, the University of Virginia, and Wayne State University, outside the submitted work. All other authors declare that they have no competing interests.

Copyright © 2015 Elsevier Ltd. All rights reserved.

Figures

Figure 1. Study Design

Patients from the IPFnet clinical trials formed the study population. Patients from the ACE-warfarin and PANTHER-prednisone/azathioprine/N-acetylcysteine arms were excluded to form the sample for overall time-to-event analyses. Seven additional patients were excluded for the remaining analyses, for which patients needed to be alive at the trial-specific landmark points.

Figure 1. Study Design

Patients from the IPFnet clinical trials formed the study population. Patients from the ACE-warfarin and PANTHER-prednisone/azathioprine/N-acetylcysteine arms were excluded to form the sample for overall time-to-event analyses. Seven additional patients were excluded for the remaining analyses, for which patients needed to be alive at the trial-specific landmark points.

Figure 2. Rates of hospitalization, disease progression and death

Plots show time to (A) disease progression as defined by ≥10% relative decrease in FVC, (B) first non-elective hospitalization from any cause, and (C) death from any cause. All are stratified by GAP index staging system: stage I, stage II, or stage III. Trial enrollment serves as time zero.

Figure 3. Impact of Hospitalizations and Decreases in FVC on All-Cause Mortality

Kaplan-Meier plots show time to death from any cause, stratified by (A) early disease progression as defined by ≥10% relative decrease in FVC, and (B) early hospitalization from any cause. The trial-specific landmark point serves as time zero. Number of subjects at risk is shown below the horizontal axis.

Figure 4. Impact of Composite Non-Fatal Endpoints on All-Cause Mortality

Kaplan-Meier plots show time to death from any cause, stratified by (A) composite of early non-elective, all-cause hospitalization or disease progression, and (B) composite of early respiratory hospitalization or disease progression. The trial-specific landmark point serves as time zero. Number of subjects at risk is shown below the horizontal axis.