Antiacid therapy does not improve disease outcomes in IPF—results from three randomized controlled trials
Running header: Antiacid therapy in IPF
Michael Kreuter MD,1,2 Wim Wuyts MD,3 Elisabetta Renzoni MD,4 Dirk Koschel MD,5 Toby M Maher MD,4 Martin Kolb MD,6 Derek Weycker PhD,7 Paolo Spagnolo MD,8 Klaus-Uwe Kirchgaessler MD,9 Felix JF Herth MD,1,2 Ulrich Costabel MD10
1 Pneumology and Respiratory Critical Care Medicine, Center for Interstitial and Rare Lung Diseases, Thoraxklinik, University of Heidelberg, Heidelberg, Germany; 2 Translational Lung Research Center Heidelberg (TLRCH), German Center for Lung Research (DZL), Heidelberg, Germany; 3 Unit for Interstitial Lung Diseases, Department of Respiratory Medicine, University Hospitals, Leuven, Belgium; 4National Institute for Health Biomedical Research Unit, Royal Brompton Hospital and National Heart and Lung Institute Imperial College, London, United Kingdom; 5 Department of Pulmonary Diseases, Fachkrankenhaus Coswig, Centre for Pulmonary Diseases and Thoracic Surgery, Coswig, Germany; 6 Firestone Institute for Respiratory Health, Department of Medicine, Pathology & Molecular Medicine, McMaster University, Hamilton, Ontario, Canada; 7 Policy Analysis Inc. (PAI), MINERVA Health Economics Network, Ltd., Brookline, MA, USA; 8 Medical University Clinic, Kanton Hospital Baselland and University of Basel, Liestal, Switzerland; 9F. Hoffmann-La Roche Ltd, Basel, Switzerland; 10Interstitial and Rare Lung Disease Unit, Ruhrlandklinik, University Hospital, University of Duisburg- Essen, Essen, Germany.
Corresponding author:
Prof. Dr. Michael Kreuter
Department of Pneumology and Respiratory Critical Care Medicine, Center for Interstitial and Rare Lung Diseases, Thoraxklinik, University of Heidelberg and Translational Lung Research Center Heidelberg (TLRCH), German Center for Lung Research (DZL), Heidelberg, Germany
Röntgenstr. 1, D-69126
Heidelberg, Germany
Phone: +49 6221 396-1201
Email:
Word count: 2863 (3000 maximum)
Funding: This study was funded by F. Hoffmann‐La Roche, Ltd.
Key words: idiopathic pulmonary fibrosis, gastroesophageal reflux disease, antiacid therapy, progression-free survival
ABSTRACT
Background: Gastroesophageal reflux disease is a potential risk factor for the development and progression of idiopathic pulmonary fibrosis (IPF). The effect of antiacid therapy(AAT) on disease progressionremains uncertain.
Methods: Patients with IPF from the placebo groups of 3 trials of pirfenidone (CAPACITY 004,CAPACITY 006, and ASCEND) were included in this post-hoc analysis. AAT use at baseline was analyzed for pulmonary function, exercise tolerance, survival, hospitalizations, and adverse events (AEs) over 52 weeks by bivariate and multivariate analyses. Disease progression was defined as decrease in forced vital capacity (FVC) ≥10% predicted, decrease ≥50 m in the 6-minute walk distance (6MWD), or death.
Findings: Of 624 patients, 47% received AAT. There was no significant difference at 52 weeks in disease progression (AAT vs non-AAT: 37·8% vs 40·5%; P=0·4844),all-cause and IPF-related mortality (6·9% vs 6·6%; P=0·8947 and 3·8% vs 5·1%; P=0·4251), absolute FVC decline ≥10% (16·8% vs 19·2%; P=0·4411), or mean change in FVC (% predicted: -4·9% vs -5·5%, P=0·3355; liters: -0·2 L vs -0·2 L, P=0·4238). There was a non-significant higher rate of hospitalization in the AAT group (22·3% vs 16·2%; P=0·0522). When stratified by baseline FVC (<70% or ≥70%), disease progression, mortality, FVC, 6MWD, and hospitalization were not different between groups. AEs were also similar;however, overall infections and pulmonary infections were significantly higher in the <70% FVC AAT group (74·3% vs 61·6%; P=0·0174 and 13·9% vs 6·1%; P=0·0214).
Interpretation: AAT did not improve outcomes in patients with IPF and may bepotentiallyassociated with an increased risk of infection in those with more-advanced disease.
Funding: F. Hoffmann‐La Roche.
INTRODUCTION
Idiopathic pulmonary fibrosis (IPF) is a chronic, irreversible, and progressive lung disease of unknown etiology with a median survival from the time of diagnosis of 2–3 years.1 It is characterized by a progressive decline in lung function, worsening dyspnea, and diminished exercise tolerance. Two therapies, pirfenidone and nintedanib, are approved for the treatment of IPF; both have demonstrated significant slowing of disease progression compared with placebo.2–4
Gastroesophageal reflux disease (GERD) is prevalent in 10%–20% of the Western population, with symptoms of heartburn, dyspepsia, regurgitation, and chest pain.5 Diagnosis of GERD can be established bya combination of symptoms, endoscopy testing, ambulatory reflux monitoring, and response to antiacid therapy (AAT). Recommended treatments include lifestyle interventions, such as weight loss, head of bed elevation, tobacco and alcohol cessation, avoidance of late-night meals, and cessation of foods that can potentially aggravate reflux symptoms, and the use of AAT with histamine H2 receptor antagonists (H2 blockers) or proton pump inhibitors (PPIs).5
The incidence of GERD in patients with IPF is higher than in the general population, and it has been reported to range between 8% and 87%6–9; variations in incidence may depend on the method of diagnosis used, such as patient-reported symptoms, physician reports, or pH probe testing, and the variability between sites in collecting and reporting information. The increased incidence of GERD in patients with IPF may be due to shared risk factors for these conditions, including age and smoking.10 In addition, the increased recoil of the fibrotic lung may dilate the lower esophageal sphincter and potentially increase reflux. GERD may also play an important role in the development and progression of IPF, including acute exacerbations.11,12 In fact, GERD is a risk factor for microaspiration, which may cause repeated lung injury and worsening of IPF. AAT may decrease the risk for acidic microaspiration–associated lung injury or damage.13–15
Current IPF treatment guidelines give a conditional recommendation for the use of AAT in patients with IPF, albeit with very low confidence in estimates of effect16becausedata on AAT impacting outcomes in IPF are limited. However, a retrospective analysis reported that GERD-related therapy, especially AAT, was associated with less radiologic fibrosis and was an independent predictor of longer survival time in patients with IPF.11 An additional study reported on the outcome of IPF patients randomized to placebo in three National Heart, Lung, and Blood Institute IPF Clinical Research Network (IPFnet)–sponsored randomized controlled trials.17 After adjusting for sex and pulmonary function, patients who received AAT had significantly less deterioration of pulmonary function than those not being treated.
The objective of this study was to further investigate the effect of AAT on the composite endpoint of disease progression in patients randomized to placebo in three large, phase 3 trials of pirfenidone in IPF.
METHODS
Source and Study Populations
The study population included all individuals with IPF randomized to placebo in three phase 3 multinational trials (CAPACITY studies 004 and 006 and ASCEND study 016).2,3Eligibility criteria for the trials have been previously described.2,3 Briefly, inclusion criteria included: age 40–80 years; a diagnosis of IPF made within the previous 48 months; no evidence of improvement in disease severity over the previous year; a predicted forced vital capacity (FVC) ≥50%, hemoglobin-corrected predicted diffusing capacity of the lung for carbon monoxide (DLCO)≥35% (≥30% in ASCEND), either predicted FVC or predicted DLCO≤90% (both ≤90% in ASCEND), and a 6-minute walk distance (6MWD) ≥150 m. All trial participants provided written informed consent, and the ethics committee or institutional review board at each participating institution approved the protocol for each trial. The study population was stratified into two subgroups based on use (i.e., yes versus no) of AAT (either H2 blockers or PPIs) at trial baseline.
Data Collection
Data collected in the aforementioned trials included patient demographic and clinical characteristics (e.g., age, sex, and comorbidity profile), pulmonary function (e.g., FVC and DLCO), exercise tolerance (6MWD), dyspnea (University of California at San Diego Shortness of Breath Questionnaire [UCSD-SOBQ]), medication use (e.g., H2 blockers and PPIs) and indication for use, adverse events, hospitalizations, and vital status. The total score of the UCSD-SOBQ ranges from 0–120 and increases with extent of dyspnea.18
FVC, 6MWD, and UCSD-SOBQ were measured at trial baseline and periodically during the trial; DLCO was assessed after baseline only in the CAPACITY trials. Medication use was documented at trial baseline and subsequently during the trial. Vital status was assessed at pre-specified time points until the follow-up visit or entry into an extension study, whichever occurred earlier. The primary cause of death and its relation to IPF were assessed in a blinded fashion by an independent mortality assessment committee in the ASCEND trial3 and by the site investigators in the CAPACITY trials.2 Safety outcomes are reported as events that occurred during the period from baseline to 28 days after the last dose of the study drug. The CAPACITY trials’ duration was 72–120 weeks, and the ASCEND trial duration was 52 weeks.
Study Outcomes
The primary study outcome, disease progression, was defined as death due to any reason,FVC decrease (absolute) ≥10%, or 6MWD decrease ≥50 m, and was ascertained over the 1-year period from trial baseline. Functional worsening (FVC decrease ≥10% and/or 6MWD decrease ≥50 m), was considered only when observed on two consecutive occasions, at least 6 weeks apart. Progression free survival (PFS) was defined as time to the first occurrence of any one of the following: a confirmed decrease of ≥10% predicted FVC, a confirmed decrease of ≥50 m in the 6MWD, or death. Secondary outcomes included all-cause and IPF-related mortality, FVC decrease (relative) ≥10%, FVC decrease (absolute) ≥5%, FVC decrease (relative) ≥5%, all-cause hospitalization, and selected adverse events (gastrointestinal [GI] adverse effects, infections, and pulmonary infections).
Statistical Analyses
Demographic and clinical characteristics of the study population were evaluated both separately by trial and collectively, as well as stratified by baseline use of AAT. Crude (i.e., unadjusted) risks of binary study outcomes, as well as changes from baseline in FVC and 6MWD, among baseline users of AAT versus baseline non-users were compared. Statistical comparisons were undertakenusing an independent-samples ttest for continuous variables and a chi-square test for categorical variables.
AAT use was examined against study outcomesusing a shared frailty model (an extension of the Cox proportional hazards modelthat adjusts for intra-cluster [i.e., intra-trial] correlation), without and with adjustment for age, sex, smoking status, lung function, and comorbidity profile. Survival analyses were based on the Kaplan-Meier estimator and were evaluated using the log-rank test. Only observed data were employed (i.e., missing values were not imputed). Individuals were censored at the time of loss to follow-up, at the time of lung transplantation, or at the end of the 1-year follow-up period, whichever occurred first. The presence of multicollinearity, hazards assumptions, and treating death as a competing risk (where appropriate) were evaluated using published methods.19,20
Role of the funding source
This was an investigator-initiated analysis. The funder of the study oversaw study design and data collection. All authors had access to the data and interpreted them.The corresponding author had full final responsibility for the decision to submit for publication.
RESULTS
Patients
A total of 624 patients were included in the study cohort. Baseline demographics and clinical characteristics for patients in both groups were similar (Supplemental Table 1).
Of the 624 patients, 291 (46·6%) received AAT (88·0% PPIs, 8·2% H2 blockers, 3·8% PPI and H2 blockers; Table 1). Of the 624 patients, 291 (46·6%) received AAT (88·0% PPIs, 8·2% H2 blockers, 3·8% PPI and H2 blockers; Table 1). Of the 291 patients receiving AAT, 38 stopped after baseline; of those 333 patients not receiving AAT, 83 patients started after baseline.Baseline characteristics were similar between AAT users and nonusers, with the exception of a significantly higher proportion of AAT users having sleep apnea (21·3% vs 9·9%; P = <0·0001), GERD (84·9% vs 22·2%; P = <0·0001), hiatus hernia (16·5% vs 3·9%; P = <0·0001), or Barrett’s esophagus (2·1% vs 0·3%; P = 0·0371) compared with patients who did not receive AAT. Cardiovascular risk factors were also more prevalent in the AAT group, with significantly higher rates of hypertension (61·5% vs 48·4%; P = 0·001) and hypercholesterolemia (55·0% vs 40·5%; P = 0·0003) than in the no AAT group. AAT was prescribed primarily for GERD (84·2%), followed by dyspepsia (2·8%) and gastritis (2·8%; Table 2).
Efficacy Outcomes
Mean follow-up was similar between groups (342 days for AAT vs 346 days for no ATT; P = 0·5246; Table3). AAT did not result in a significant between-group difference for disease progression vs no ATT (37·8% vs 40·5%, P = 0·4844). AAT users had similar PFS compared with no AAT users (hazard ratio [HR], 0·9 [95% CI, 0·7–1·2]; P = 0·404; Figure 1A). For each component of the disease progression composite endpoint, a similar number of patients in the AAT and no AAT groups had a qualifying event, including death (4·8% vs 5·4%, respectively), absolute FVC decrease ≥10% (10·3% vs 10·2%, respectively), and a decrease of ≥50 m in the 6MWD (22·7% and 24·9%, respectively). In the AAT and no AAT groups, the rates of all-cause mortality (6·9% vs 6·6%, respectively; P = 0·8947) and IPF-related mortality (3·8% vs 5·1%, respectively; P = 0·4251) were also similar, regardless of AAT. The risk of death from IPF at 1 year was not significantly reduced with AAT compared with no AAT (HR, 0·8 [95% CI, 0·4–1·6]; P = 0·4637; Figure 1B). AAT users had similar mean (SD) changes in FVC from baseline to week 52 compared with no AAT users (% predicted: -4·9% vs -5·5%, P=0·3355; liters: -0·2 L vs -0·2 L, P=0·4238; Table 3). Absolute and relative changes in %FVC from baseline and 6MWD decreases ≥50 m after 52 weeks were similar between patients who did and did not receive AAT (Table3). There was a marginally significantly higher rate of hospitalization in the AAT group (22·3% vs 16·2%; P = 0·0522).
When patients were stratified by baseline FVC (≥70% or <70%; Table 4), no differences were observed in disease progression or mortality between the two groups. Of patients with <70% predicted FVC, disease progression rates were 43·8% for those who received AAT compared with 45·7% for those who did not receive AAT (P = 0·7971); similarly, among patients with percentage-predicted FVC ≥70%, the rates did not differ significantly (32·7% vs 36·7%, respectively; P = 0·4528). Furthermore, changes in FVC, 6MWD, and all-cause hospitalization were not significantly different between the AAT groups when the patients were stratified by baseline FVC. Similar results were obtained when baseline FVC was stratified by ≥60% or <60% (data not shown).
When assessed by bivariate analysis using a shared frailty model, AAT was not significantly associated with PFS, any of the components of the PFS composite score, all-cause hospitalization, or IPF-related mortality (Table 5). Similar results were observed in multivariate analyses using a shared frailty model.
Safety Outcomes
Patients who received AAT had similar rates of all-cause hospitalization, GI adverse effects, infections, and pulmonary infections as patients who did not receive AAT (Table3). When patients were stratified by baseline FVC, GI adverse effects were similar regardless of AAT use. However, in patients with FVC <70% predicted, infections were significantly higher with AAT use than no AAT use (74·3% vs 61·6%; P = 0·0174). Similar differences were also observed forpulmonary infection rates (13·9% vs 6·1%; P = 0·0214).
DISCUSSION
In this post-hoc analysis of patients with IPF randomized to placebo in 3 large controlled trials, AAT did notyieldclinically significant improvements in outcomes to 52 weeks. There was no relationship between AAT and PFS, mortality, or adverse events. Patients with advanced IPF (<70% FVC) who received AAT had similar rates of PFS and mortality, but had higher infection rates (both pulmonary and non-pulmonary) than patients who did not receive AAT.
AAT has been given a conditional recommendation for use in the 2015 IPF treatment guidelines, and this is unchanged from the 2011 guidelines.16 Retrospective analyses have reported that patients who received AAT had slower disease progression as assessed by decline in FVC and improved survival compared with patients who did not receive AAT.11,17 In an analysis of the placebo arms of 3 randomized controlled trials of patients with IPF, AAT use was associated with a significantly smaller decrease in FVC, although no differences were observed for all-cause mortality or all-cause hospitalization.17 Fewer acute exacerbations were also observed among patients who received AAT. Additional studies suggested AAT helped stabilize IPF.21
Our findings do not support any beneficial effect of AAT in patients with IPF, in contrast with those of previous studies. We performed multiple sensitivity analysis using stratification by FVC <70% or ≥70% and ATT use in patients with GERD only and saw similar results to our main analysis (data not shown).One explanation could be related to differences in patient characteristics. In the phase 3 CAPACITY and ASCEND trials, a minority of patients had advanced disease as assessed by functional impairment. Patients awaiting lung transplant were also excluded, and AAT may potentially benefit this patient population.22 In the IPFnet-sponsored trials, mean baseline percentage of predicted FVC was ≈ 59% in STEP-IPF, ≈ 58·5% in ACE-IPF, and ≈ 71% in PANTHERcompared with ≈ 75% in CAPACITY and ≈ 68% in ASCEND.2,3,17 While it has been suggested that GERD is more prevalent in patients with advanced IPF,our data analysis of patients stratified based on FVC <70% or ≥70% did not show a significant difference in terms of effects by AAT. Because the IPFnet-sponsored studies were not analyzed separately, it is not possible to ascertain whether STEP-IPF, which included patients with advanced IPF, contributed the most to the results. On the other hand, with regard to the retrospective analysis suggesting that AAT may confer a survival benefit, it is possible that the diagnosis of GERD in patients with IPF may have introduced a lead-time bias, thus accounting for the better prognosis associated with AAT use.11Furthermore, different follow-up times (e.g., 30 weeks in the IPFnet study and 52 weeks in our study) may have contributed to different results. Although comorbidities, such as cardiovascular disease and sleep apnea, were more prevalent among patients who received AAT and may haveinfluenced the results, disease progression and survival analysis were adjusted for comorbidities and no association was observed in multivariate models.
In this analysis, while there were no significant differences in the overall cohort, among patients with advanced IPF (e.g., FVC <70%) who received AAT there was a significantly higher incidence of infections, both pulmonary and non-pulmonary, consistent with previous studies demonstrating a higher incidence of ventilator-associated and community-acquired pneumonias among patients treated with AAT.23,24Baseline FVC of <70% was chosen because this was the mean FVC of the patient population and allowed for fair statistical evaluations in sufficiently large subgroups of patients. For patients with FVC<60% similar results were obtained (data not shown).Due to the retrospective nature and nonrandomized comparisons, results should be interpreted with caution.Moreover, we note that due to a relatively small sample size, these analyses were likely underpowered to detect meaningful differences.