Graphical abstract
Overview of the study. BOS: bronchiolitis obliterans syndrome; ITT: intention-to-treat; FEV1: forced expiratory volume in 1 s.
Abstract
Background
Chronic lung allograft dysfunction (CLAD) is a major obstacle to improving outcomes after lung transplantation. Bronchiolitis obliterans syndrome (BOS), characterised by progressive decline in forced expiratory volume in 1 s (FEV1) due to fibrotic scarring of the small airways, accounts for most CLAD cases. Pirfenidone, an antifibrotic agent used for idiopathic pulmonary fibrosis, was assessed for treating progressive BOS.
Methods
An investigator-initiated, international, multicentre, randomised, double-blind, placebo-controlled phase II trial was conducted in nine European lung transplant centres. Adults with bilateral lung transplants and progressive BOS were randomised (1:1) to receive pirfenidone 2403 mg·day−1 or placebo with standard care for 26 weeks. The primary end-point was change in FEV1 from baseline to week 26, with imputation for missing values.
Findings
From 1 May 2015 to 1 December 2019, 477 patients were screened, and 90 were randomised to pirfenidone (n=48) or placebo (n=42). Both groups showed continued decline in FEV1 from baseline to week 26, with no significant difference in intention to treat (ITT), ITT with imputation, or per-protocol analyses. Secondary end-points (graft loss, death, re-transplantation) were similar between groups. Treatment-related serious adverse events were distributed equally.
Interpretation
Pirfenidone did not show superiority over placebo and standard care in this exploratory trial. It cannot be recommended for treating BOS. Further research is needed to explore other treatments for CLAD to improve long-term outcomes after lung transplantation.
Shareable abstract
No difference was found when using pirfenidone versus placebo to ameliorate lung function decline in progressive BOS in lung transplant recipients https://bit.ly/4oeSDDc
Introduction
Chronic lung allograft dysfunction (CLAD) continues to be a major obstacle to long-term survival and good health after lung transplantation. CLAD is defined by a persisting decline in forced expiratory volume in 1 s (FEV1) once the FEV1 has fallen by ≥20% of the baseline value and both acute rejection and infection is excluded. CLAD can present as different phenotypes [1], with the obstructive phenotype (bronchiolitis obliterans syndrome (BOS)) being the most common, affecting 50–60% of lung transplant recipients within 5 years after transplantation. CLAD remains the leading cause of death beyond the first year [2–4]. No reliably effective treatment options for BOS are available [5]. Azithromycin has demonstrated improvement in lung function in up to 30% of patients with declining FEV1 and may prevent the development of BOS, although the phenotype of those who respond has not been well described [6, 7]. Transbronchial biopsies of patients with BOS and autopsies from deceased recipients with BOS reveal obliterative bronchiolitis due to excessive fibrotic scarring even in the early stages of BOS, which may suggest a potential beneficial role of antifibrotic agents [8, 9]. Following this rationale, pirfenidone, an antifibrotic drug registered for the treatment of idiopathic pulmonary fibrosis (IPF), was tested against placebo and standard care to see whether it would reduce the fibrotic process in the small airways by ameliorating the decline of FEV1 in progressive BOS.
Pirfenidone is an orally active, small-molecule drug that inhibits the synthesis of transforming growth factor-β, a cytokine that controls many cell functions including proliferation and differentiation, and plays a key role in the development of fibrosis [10]. In addition, it inhibits the synthesis of tumour necrosis factor-α, a cytokine that is known to play an active role in inflammation [11]. Several studies with pirfenidone have demonstrated an antifibrotic effect and inhibition of obliterative airway disease in animals as well as in human transplant models [12–14]. A few clinical case reports and single-centre series have demonstrated promising results with pirfenidone treatment, stabilising or reducing the FEV1 decline in CLAD patients for both BOS and restrictive allograft syndrome phenotypes, as well as BOS after haematopoietic cell transplantation [15–17].
We aimed to investigate the role of pirfenidone in patients with progressive BOS, in a double-blinded, randomised controlled trial. The primary end-point was change in FEV1 over 6 months in bilateral lung transplant recipients with progressive BOS, despite treatment with azithromycin next to standard of care (SOC) (shown in the supplementary material, appendix A) for ≥4 weeks.
Methods
Study design
The study was a phase II, investigator-initiated European multicentre, randomised, double-blind placebo-controlled parallel trial of pirfenidone treatment in double lung transplant recipients with BOS stages 1–3. Nine lung transplant centres across Europe (in Belgium, Denmark, Germany, Norway, Sweden, the Netherlands and the UK) participated in the study.
Participants
The study population comprised patients aged >18 years who, after a bilateral lung transplant, developed BOS stage 1–3. According to the standard of care, patients were included irrespective of pre-transplant disease, cause of BOS or type of induction treatment and/or immunosuppressive treatment, which was given at the discretion of the different centres (supplementary material, appendix A). Patients were ≥6 months past lung transplantation and had post-transplant baseline values of FEV1 (the mean of the two highest values measured ≥3 weeks apart according to International Society for Heart and Lung Transplantation criteria) [1]. All patients had received azithromycin therapy for ≥4 weeks prior to study start, receiving a minimum dose of 250 mg·day−1 for ≥3 days per week.
The study participants had documented progressive BOS, as demonstrated by at least three FEV1 measurements in the past 6 months, each ≥3 weeks apart with a total decline of ≥200 mL in FEV1 in the past 6 months and a mean decline of ≥50 mL between the last two measurements.
Exclusion criteria comprised lung re-transplantation, combined organ transplantation (including heart and lung transplantation), single lung transplantation, treatment within 4 weeks prior to inclusion with montelukast, extracorporeal photopheresis (ECP), total lymphoid irradiation (TLI), inhaled corticosteroids or bronchodilators. A complete list of inclusion and exclusion criteria is provided in the supplementary material (appendix B).
End-point adjudication was performed by the steering committee, who reviewed all inclusion and exclusion criteria as well as CLAD staging for all enrolled patients. End-point adjudication was undertaken blind to randomised treatments. The study was approved by the Danish National Board of Health (H-3-2014-136/ approved 9 January 2014) and the Danish Data Protection Agency (RH-2015-29; approved 30 April 2015) as well as by all local ethics committees for each participating centre. The trial was conducted in accordance with the Declaration of Helsinki. Both oral and written informed consent was obtained from all eligible participants. The study was registered with ClinicalTrials.gov (identifier NCT02262299) and EudraCT (identifier 2008-004771-22).
Randomisation
Randomisation was performed by a computerised central interactive web response system. The automated system assigned the study medication to each patient. All randomisation codes were generated by a statistician independent of the conduct of the trial, keeping the study personnel blinded.
Eligible subjects were randomised to pirfenidone or placebo in a 1:1 ratio. Pirfenidone and placebo were both supplied in capsules that were visually indistinguishable by study nurse or centre study pharmacy according to local practice. Pirfenidone and placebo packaging and labelling were identical. Patient, study nurse and investigator were blinded to the type of treatment, throughout the study. Data analysis was performed by a statistician independent of the trial conduct.
Procedures
Screening assessment included mandatory chest radiography, high-resolution computed tomography scan of the thorax (to exclude non-BOS pathology), spirometry with measurement of FEV1 over forced vital capacity (FVC) according to European Respiratory Society [18] criteria to demonstrate typical obstruction and ECG to exclude QTc >500 ms. Bronchoscopy with bronchoalveolar lavage (BAL) was undertaken according to standard international criteria with two aliquots of 50 mL [19]. BAL was cultured for bacteria, viruses and fungi, and a cytospin preparation or immunological BAL for differential counts was performed to exclude infection. Transbronchial biopsies were performed to exclude acute cellular rejection and blood samples (haematology and chemistry) including liver function tests as well as drug monitoring for the immunosuppressive treatment to exclude other treatable causes of FEV1 decline. Renal insufficiency (creatinine clearance <30 mL·min−1 calculated using the Chronic Kidney Disease Epidemiology Collaboration formula) was also an exclusion criterion.
Upon enrolment, patients were randomised to placebo or pirfenidone tablets of 267 mg and received a starting dose of one capsule orally three times a day for the first week (days 1–7) (801 mg·day−1) and gradually increased per week to a maximum dose of three capsules, three times per day (2403 mg·day−1). Dosing regimen was adapted by the local investigator upon patient tolerance. Adjustments in other medication was done according to the discretion of the treating physician and not registered in the study. Discontinuation of the study drug for a total of 14 days during the study was allowed without being excluded from the study. Per protocol, by 180 days, subjects should have taken (7×3+7×6+166×9) 1557 tablets. A minimum therapeutic dose of equivalent to 540 tablets had to be taken throughout the 6-month study period to complete the study. Study visits to collect outcomes were scheduled at 1, 2, 3, 4 and 6 months, with an additional blood sample at 5 months to monitor possible drug toxicity. At the last visit participants underwent a final assessment. The planned outcome measures for each visit are presented in the supplementary material (appendix C).
Outcomes
The primary outcome was absolute change in FEV1 (L) from baseline to 6 months (FEV1 at 6 months – FEV1 at baseline). Secondary outcomes included relative percentage change and categorical thresholds, as well as mortality, as follows. 1) Categorical percentage change in FEV1 (mL) from baseline to 6 months; 2) change of FVC (L); 3) change in total lung capacity (TLC) (L); 4) change in FEV1/FVC ratio; 5) number of patients with treatment failure (defined as one or more of the following: re-transplantation, death due to respiratory causes, initiation of rescue therapy for worsening of BOS, defined as a decline of ≥600 mL in three consecutive months. Rescue therapy was defined as initiation of ECP, TLI, montelukast or other treatments considered as rescue therapy by the investigator); 6) worsening in BOS stage as defined by Meyer et al. [20] in 2014 from baseline to 6 months; 7) change in percentage predicted diffusing capacity of the lung for carbon monoxide (DLCO); 8) change in functional level as assessed by the 6-min walk distance (6MWD); 9) hospital admission for any reason; 10) change in EuroQol five-dimension (EQ-5D) scale. Previous studies have suggested diversity within BOS phenotypes based on differing rates of FEV1 decline [21], suggesting that effects of treatment may vary within the heterogenous population [21, 22]. It was therefore decided to do a post hoc analysis of slow and fast decliners, as defined by change in FEV1 during the inclusion period above or below the median rate of 155 mL per month.
Statistical analysis
Power calculation
Based on retrospective data from the involved centres, a decline in FEV1 from BOS onset over 6 months was expected to be 450 mL in the placebo group. With a treatment effect of 50%, the expected decline was estimated to be 225 mL. With a β-error of 12% (power 88%), the study would be able to detect this difference with a sample size of 36 patients in each group using a two-sided t-test with a significance level of 0.05. An estimated 80 patients were planned to be recruited in the trial allowing for a dropout rate of 10%. An additional 10 patients were recruited due to higher-than-expected dropout rate.
Analysis plan
Descriptive statistics for each variable were planned with the use of parametric statistics or nonparametric approaches as appropriate in the intention-to-treat analysis (ITT). Missing data were handled by the establishment of imputation rules before the onset of the study. The imputations for missing data due to treatment failure or all-cause death was imputed as the worst of either 50% of baseline FEV1 in mL, or the last observed FEV1 was carried forward. For patients with missing 6-month data, without treatment failures or death, the worst of the last observed FEV1 carried forward prior to the 6-month visit or the next observed FEV1 after the 6-month visit was imputed (maximum 9 months after randomisation). The per-protocol evaluation was defined as a minimal exposure to study drug, which was at least three tablets a day for 180 days (i.e. 540 tablets by the 6-month visit), the availability of FEV1 at 6±1 months at the 6-month study visit and the definition of protocol violations to exclude patients from the per-protocol analysis (supplementary material, appendix E).
Primary statistical analysis
Change in FEV1 from baseline was summarised with descriptive statistics and the two-sided t-test was used to compare the change in FEV1 from baseline to the 6-month follow-up between the randomised groups. Linear mixed models were used for evaluating the decline in FEV1 for each patient through the repeated measurements of FEV1 over the 6-month study period, to better account for within patient correlation and at random missing values. Additional information regarding the statistical plan is presented in the supplementary material (appendix D), and planned statistics for secondary outcomes are described in detail in the protocol.
SAS/ACCESS 9.4 Interface to ADABAS (SAS Institute, Cary, NC, USA) was used for analysis.
The study was performed in accordance with the guidelines for good clinical practice. The statistical analysis plan was drafted by an independent statistician together with the study steering committee. An independent data monitoring committee reviewed the safety data, which was prepared by an independent statistician. The final analysis was carried out by the independent statistician.
Role of the funding source
InterMune funded the study through an unrestricted research grant and provided the study drug and placebo. During the study period (in August 2014) InterMune was purchased by Roche, who then was responsible for the study funding commitments for the later part of the study.
The funder of the study had no role in study design, data collection, data analyses, data interpretation or writing of the report.
Results
From 1 May 2015 to 1 December 2019, 477 screening visits at nine European sites were performed, resulting in a total of 90 participants being included and randomly assigned into the trial; 48 (53%) in the pirfenidone treatment group and 42 (47%) to receive placebo in the control group (figure 1, table 1). The median (interquartile range (IQR)) age of participants was 54 (45–60) years and 39 (43%) were female. Baseline demographics and clinical characteristics were balanced between groups.
FIGURE 1.
Consolidated Standards of Reporting Trials diagram of trial profile. FEV1: forced expiratory volume in 1 s; ITT: intention to treat. #: more than one reason possible; ¶: insufficient FEV1 decline.
TABLE 1.
Baseline characteristics
| Pirfenidone | Control | |
|---|---|---|
| Participants | 48 | 42 |
| Sex | ||
| Female | 22 (46) | 20 (41) |
| Male | 26 (54) | 22 (60) |
| Age years | 51.0 (45.0–60.0) | 57.5 (48.0–61.0) |
| Primary diagnosis | ||
| COPD/emphysema | 12 (25) | 20 (48) |
| Idiopathic pulmonary fibrosis | 13 (27) | 9 (21) |
| Cystic fibrosis | 10 (21) | 7 (17) |
| Primary pulmonary hypertension | 2 (4) | 0 (0) |
| Other (supplementary material, appendix G) | 11 (23) | 6 (15) |
| Baseline FEV1 L | 1.4 (1.1–1.8) | 1.6 (1.1–2.0) |
| FEV1 at inclusion % predicted | 46.5 (32.9–60.2) | 52.2 (40.1–68.5) |
| BOS stage at inclusion | ||
| Stage 1 (FEV1 >65–80%) | 9 (19) | 9 (21) |
| Stage 2 (FEV1 >50–65%) | 19 (40) | 22 (52) |
| Stage 3 (FEV1 ≤50%) | 19 (40) | 11(26) |
| FVC % predicted | 76.6 (60.9–91.6) | 81.7 (63.1–97.2) |
| FEV1/FVC ratio | 0.52 (0.44–0.62) | 0.56 (0.47–0.65) |
| TLC % predicted | 92.1 (76.4–100.5) | 97. 5 (82.8–105.4) |
| DLCO % predicted | 56.7 (47.2–66.2) | 65.8 (51.1–77.1) |
| 6MWD m | 456 (394–514) | 463 (396–569) |
Data are presented as n, n (%) or median (interquartile range). FEV1: forced expiratory volume in 1 s; BOS: bronchiolitis obliterans syndrome; FVC: forced vital capacity; TLC: total lung capacity; DLCO: diffusing capacity of the lung for carbon monoxide; 6MWD: 6-min walk distance.
Two study participants (pirfenidone n=1, control n=1) were inadvertently enrolled and randomised despite violation of the inclusion or exclusion criteria. The participant in the treatment group was never given the study drug, while the other was receiving simultaneous ECP treatment. Both were included in the ITT analysis. Detailed information about the subjects who were removed due to protocol violations are presented in the supplementary material (appendix E).
11 (12%) other participants who were inadvertently enrolled all continued in the study (equally distributed in the two groups), on study drug after review and acceptance by the steering committee (supplementary material, appendix E). All of them were included in the ITT analysis, but not in the per-protocol analysis.
17 (35%) participants in the treatment group and 14 (33%) in the control group discontinued the study prematurely (p=1.000); reasons for premature discontinuation were rescue treatment, side-effects, death or re-transplantation, protocol deviation, or other, as demonstrated in figure 1. Time to study drug discontinuation and time to study discontinuation are shown in the supplementary material (appendix F).
Four patients died during the study period, all due to progressive BOS and respiratory failure: three in the treatment group and one in the placebo group. Time to death after inclusion was 69, 143 and 192 days, for the three patients in the pirfenidone group and 172 days for the patient in the control group. Two patients underwent re-transplantation (one per group) at 155 and 113 days after randomisation.
In the 59 participants who completed the 6-month follow-up visit, a mean±sd 1463±148 tablets (94% of intended max dose) were taken by the 31 patients in the treatment group during the study period and 1515±92 tablets (97% of intended max dose) were taken by the 28 patients in the placebo group. Seven (23%) patients in the treatment group and two (7%) in the control group took <1431 tablets (92% of intended max dose), and only one (3%) participant took <1080 tablets (69% of intended max dose) in the treatment group. However, all took more than the minimum of 540 tablets (34% of intended max dose) defined in the protocol as the lower limit.
Primary end-point
Out of 48 subjects in the pirfenidone group, 17 (35%) did not complete a FEV1 measurement at 6 months. One patient was accidentally randomised without having a FEV1 measurement upon study entry, resulting in 30 (63%) subjects in the observed ITT analysis. In the control group, 28 (67%) subjects completed a FEV1 measurement at 6 months and were included in the observed ITT analysis. The median (IQR) decline in FEV1 over the 6-month period was 430 (820–250) mL in the pirfenidone group and 440 (780–280) mL in the placebo group, which was in concordance with an assumed mean decline of 450 mL as seen in the statistical plan. After 6 months, the mean FEV1 in the treatment group (observed ITT) was 120 mL higher than the placebo group, although not statistically significant (p=0.2186). Adding imputed values for missing data according to the statistical plan did not change this, with the mean FEV1 being 90 mL higher in the treatment group (p=0.3026). In the per-protocol analysis, which was performed with 27 patients in each group, the difference after 6 months of treatment remained at 130 mL, but was not significant (p=0.1741) (figure 2).
FIGURE 2.
Change in forced expiratory volume in 1 s (FEV1) by randomised group from 1 month until end of study/results of mixed linear model with a) nonimputed intention-to-treat (ITT), b) imputed ITT and c) per-protocol data.
The results of the repeated FEV1 measures in the mixed linear model showed that the average reduction in FEV1 increased with each follow-up visit (p<0.0001), indicating a continued decline in lung function over time, but this did not differ between randomised groups (p=0.5149). The results are similar for the imputed ITT (p=0.1735) and observed per-protocol (p=0.2053) analysis, as seen in figure 2.
We conducted a pre-specified post hoc analysis for median rate of change in FEV1 from baseline to end of study according to initial BOS staging at study entry. The median (IQR) rate of FEV1 change in BOS stage 1 was 200 (150–250) mL per month, 160 (130–190) mL per month in BOS stage 2, and 110 (90–120) mL per month in BOS stage 3. No significant differences were observed between treatment and placebo groups for any of the three BOS stages. The same was true for imputed data and change in rate of decline (data not shown).
A post hoc subgroup analysis of slow versus fast decliners was also performed. The rate of change in FEV1 up to 6 months before inclusion was used to define slow and fast decliners by dichotomising the population at the median decline of 155 mL per month. The fast decliners (40%) did show reduced FEV1 decline in the study period when treated with pirfenidone versus placebo, and the difference was statistically significant in the observed ITT (pirfenidone (n=11) versus placebo (n=12)) (p=0.0336) and the observed per-protocol (n=10 versus n=11) (p=0.0225) groups analysis. However, it was not statistically significant in the ITT (n=22 versus n=23) group when employing single-value imputation to account for subjects missing from the 6-month end-point due to death, treatment failure, etc. (p=0.1755), as well as when adjusting for baseline FEV1 between groups among ITT and per-protocol analyses.
Secondary end-points
We did not observe any categorical changes in percentage FEV1 from baseline to 6 months of treatment in randomised subjects using observed ITT analysis (p=0.1395).
FVC decline was comparable between the two groups. FVC declined 220 mL over the 6-month period in the treated group and 250 mL in the placebo group (p=0.7874) (figure 3a). The repeated-measures model showed a significant reduction in FVC by each visit, as was observed for FEV1 (p=0.0010), without statistically significant difference between groups (p=0.9052). The FEV1/FVC ratio did not decrease significantly over the study period (p=0.2503) and remained similar between both groups (p=0.0784) (figure 3b). Furthermore, TLC did not change significantly from baseline to 6 months between groups (20 mL, 95% CI −260–310 mL; p=0.8659) (figure 3c). Two (2.4%) patients in the control group were reported to develop the mixed phenotype compared to none in the treatment group. DLCO remained almost stable over the 6-months study period with a median (IQR) decline of only 1.21% (−8.32–2.57%) in the treatment group and 1.59% (−13.1–1.20%) in the placebo group, without significant difference between groups (2.95%, 95% CI −2.60–8.50%; p=0.2901) (figure 3d).
FIGURE 3.
a) Change in forced vital capacity (FVC) by randomised group; b) change in forced expiratory volume in 1 s (FEV1)/FVC ratio by randomised group; c) change in total lung capacity (TLC) by randomised group; d) change in diffusing capacity of the lung for carbon monoxide (DLCO) by randomised group.
There was a median decline in 6MWD from baseline to 6 months of 0 m (IQR −30–16 m) in the treatment group and 12.5 m (IQR −62–14 m) in the placebo group, with a mean difference between groups of −2.6 m (95% CI −56–51 m; p=0.9225). We did not see a significant change in BOS stage between study groups (p=0.2400).
Treatment failure with the need for rescue therapy, death due to respiratory causes or re-transplantation were all secondary end-points and are summarised in table 2. All deaths in the study were related to progressive BOS by developing respiratory failure.
TABLE 2.
Treatment failure, initiation of rescue therapy, death due to respiratory causes, and re-transplantation in randomised patients in the intention-to-treat population
| Pirfenidone | Placebo | |
|---|---|---|
| Treatment failure | 10 | 10 |
| Initiation of rescue therapy | 6 | 8 |
| Death due to respiratory causes | 3 | 1 |
| Re-transplantation | 1 | 1 |
Data are presented as n. p=0.5278.
We did not find any significant difference between groups in the EQ-5D questionnaire: mobility (p=0.7625), self-care (p=0.0936), usual activities (p=0.7058), anxiety/depression (p=0.3892) and visual analogue scale (p=0.7300).
Each secondary end-point had >10% missing values in both randomised groups. However, since the primary end-point was not statistically significant in the observed ITT, single imputed ITT or per-protocol analysis, and the secondary end-points were not statistically significant in the observed ITT analysis, imputation was not performed for secondary end-points.
There was a total of 220 adverse events in 45 (93.8%) patients in the pirfenidone group and 152 events in 37 (88.1%) controls, but without significant difference between groups (p=0.4654) (data not shown). The most frequent adverse events were dyspnoea (p=1.000), nausea (p=0.112), diarrhoea (p=0.564) and fatigue (p=0.564). When reducing number of adverse events to events that could have a possible or probable relationship to the study medication, only nausea (p=0.0264) was statistically significantly different between groups, and most common in the active treatment group.
Dyspnoea was the single most frequent serious adverse event, which was significantly different between groups with 18.8% in the treatment group versus 2.4% in the control group (p=0.0175).
Discussion
This is the first international multicentre randomised placebo-controlled trial in lung transplant recipients’ patients with chronic lung allograft dysfunction of the BOS phenotype after lung transplantation. 6 months of treatment with pirfenidone did not show any significant benefit compared to placebo regarding decline in FEV1 in patients with progressive BOS. There was no difference in any of the secondary end-points as well as number of treatment failures, deaths and re-transplantations.
The number of re-transplants combined with the number of deaths adds up to a total of seven out of 90, equivalent to 8% of the study population with graft loss over a period of 26 weeks, depicting the severity of progressive BOS, and is comparable to previous reports. Finlen-Copeland et al. [23] had a 2-year mortality >30% after the initial diagnosis of BOS in a retrospective cohort. A BOS diagnosis early after transplantation and a higher BOS-stage were both risk factors for death. In two randomised trials, Corris et al. [6] reported no deaths within a 12-week study period, while Ruttens et al. [24], studying the effect of montelukast, had an overall mortality rate of 6.5% after 12 months, increasing to 20% and 30% after 2 years in the treatment and control groups, respectively [25, 26].
80% of the patients in the pirfenidone group received full treatment dose, and no one received fewer than the minimum 540 tablets. In the CAPACITY studies, an effect was documented after 24 weeks of pirfenidone treatment, and was consistent throughout the 72 weeks of the study [27]. Therefore, insufficient treatment should not play a role in our study.
The number of adverse events was almost equal in both treatment groups, apart from nausea, which was significantly more frequent in the pirfenidone group. We would expect more gastrointestinal symptoms in this group, as has been reported previously with pirfenidone in IPF [28], but our findings might reflect the general burden of symptoms and side-effects caused by the standard-of-care lung transplant medication [29].
The number and character of serious adverse events were similar in both groups, although only dyspnoea was significantly more common in the pirfenidone treatment group as opposed to the control group. Dyspnoea is a known and common (>10%) side-effect related to pirfenidone treatment. Moreover, dyspnoea is a common and an expected symptom in this BOS population with declining lung function, especially at a higher BOS stage, as was seen in the treatment group at baseline. This higher initial BOS stage could also explain the increased number of deaths due to respiratory failure in the treatment group compared to the control group.
Overall, there was no major difference in adverse events or side-effects between groups, which suggests that there was only limited risk of inadvertently un-blinding the study arm to the patient or treating physician.
We aimed to include lung transplant recipients with BOS who were progressive (i.e. with declining FEV1) despite treatment with azithromycin on top of standard care. This trajectory is well known in any lung transplant programme [5, 21, 26]. As described, there was a significant decline in FEV1 between study visits in both groups and there was no change in FEV1/FVC ratio, nor in TLC over time.
Reflecting the complexity of this population and the challenges for inclusion within the time frame of the protocol, we found it crucial to implement adjustments to the inclusion criteria. One of these was to extend the inclusion criteria to contain BOS stage 3, since a substantial number of patients continued to decline in FEV1, leading to a higher BOS stage within the time of workup and screening, and consequently were missed for inclusion in the study.
With the introduction of the updated CLAD criteria in 2019, we conducted a post hoc analysis of CLAD stages at inclusion, which showed that there was a larger proportion of CLAD stage 3 (23%) and 4 (17%) in the treatment group as opposed to the placebo group (17% and 7%, respectively), although this was not statistically significant (p=0.3440).
The histopathological correlate of BOS is constrictive bronchiolitis, with obliterative bronchiolitis lesions as a pathognomonic finding. A substantial part of the small airways have already been obliterated when BOS is diagnosed by a ≥20% decline in FEV1 [30]. Therefore, we would expect the potential effect of altering the fibrotic process at a less severe grade to be more efficacious in stabilising the rate of decline in lung function at a higher plateau, which suggests that earlier management of BOS may mitigate progression to more severe grades [31]. Although a larger proportion of BOS stage 3 patients were included in the treatment group, since patients were not stratified according to BOS group, a post hoc analysis of primary outcome did not find any significant difference for mean change in FEV1 from baseline to end of study by BOS stage; nor was there a change in rate of decline (data not shown), although numbers are too small to draw firm conclusions. These findings concur with the STOP-CLAD trial, where stratifying by CLAD stage did not show any treatment effect, although enrolment was not met [32].
Among fast decliners, which declined with a rate similar to that which has been shown in other studies [24], we did see a signal of possible treatment effect of pirfenidone when analysing per protocol and ITT. However, when adjusting for difference in baseline FEV1, this effect was no longer significant. Similarly, when including imputed data for missing values due to death or withdrawal from the study, this was no longer significant.
Even though characterisation of CLAD has improved with the differentiation into additional phenotypes (BOS, restrictive allograft syndrome, mixed, undefined) the definitions continue to rely on clinical findings [1]. Within the clinical syndrome BOS, heterogenicity is still seen, e.g. in different profiles of lung function trajectories and difference in response to treatment. This likely reflects a similar heterogenicity in underlying pathogenesis. Improved understanding of the pathogenesis of BOS could aid identification of pharmacological targets in certain subgroups of CLAD patients.
Our study has important limitations. Inclusion took longer than expected and additional amendments were necessary to achieve sufficient patients included. Although the study did achieve the inclusions based on the power calculation, the effect parameter was estimated to be a 50% reduction of FEV1 decline, equivalent to a 225 mL difference between cases and controls. This amount of reduction is clinically relevant, but might be optimistic, which could introduce the risk of underpowering the study. We only observed a difference of 130 mL between groups. Furthermore, there was a substantial dropout rate, although not different from previous randomised BOS trials [6, 24]. Inclusion was not matched on BOS grade, which may be a confounding factor that can affect the treatment effect, as previously mentioned. In previous studies on pirfenidone treatment, patients with a FVC >50% predicted were included, and the treatment significantly reduced the decline in lung function. However, in our study, we included patients with a FEV1 <50% of their best value, with a higher proportion in the treatment group. This difference in patient selection may have contributed to the neutral treatment outcomes observed for pirfenidone. Additionally, even though this was a randomised controlled trial, other factors that can affect the progression of BOS were not measured or matched for in this study, and could have been distributed unevenly in the two groups. Strengths of the study include the relatively large study population and well-characterised cohort of progressive BOS with only limited risk of accidental unblinding towards study personnel.
Conclusion
Although previous observational case reports and experimental data have been promising, our randomised controlled trial did not show a statistically significant benefit of pirfenidone versus placebo with standard of care treatment on lung function decline in patients with BOS stage 1–3 at the time of treatment start. Patients had additional side-effects in the treatment group versus placebo group. Despite the limitations of the study, the neutral outcome combined with the increased incidence of side-effects suggests that the use of pirfenidone is currently not justified in this population. Therefore, our findings support the exploration of other treatments that can prevent or stop this condition in order to further improve long-term results after lung transplantation.
Further observations can be concluded from this study, including the recognition of the complexities in undertaking a randomised clinical trial in this population of patients despite the evident need. The value of collaborating centres forming trial networks to enable effective trial designs in lung transplantation is supported.
Acknowledgements
All project leaders, research nurses, colleagues, staff of the transplant teams who have participated and made the study possible. The study DSMB was unpaid, academic and professionally chaired by Christian Benden (Boston Children's Hospital and Harvard Medical School, Boston MA, USA), and included Elisabeth Bendstrup (Aarhus University Hospital, Århus, Denmark) and Frederica Meloni (University of Padova and AOPD, Padova, Italy). Without the visions of now retired senior consultant Martin Iversen (Copenhagen, Denmark), this study would never have been conceptualised. Finally, Cornelia G. Crone (NOH, Copenhagen, Denmark) for critical reading and feedback.
Footnotes
This clinical trial was prospectively registered with ClinicalTrials.gov as NCT02262299.
This article has an editorial commentary: https://doi.org/10.1183/13993003.02412-2025
Ethics statement: The study was approved by the Danish National Board of Health (H-3-2014-136, approved 9 January 2014) and the Danish Data Protection Agency (RH-2015-29, approved 30 April 2015), as well as by all local ethics committees for each participating centre. The trial was conducted in accordance with the Declaration of Helsinki. Both oral and written informed consent was obtained from all eligible participants.
Author contribution: The study was conceptualised by M. Perch, P. Corris, G.M. Verleden, E. Verschuuren, R. Vos and J. Gottlieb. The protocol was drafted by P. Corris, G.M. Verleden, E. Verschuuren and J. Gottlieb. The steering committee included M. Perch, P. Corris, G.M. Verleden, E. Verschuuren and J. Gottlieb, who contributed to the statistical analysis plan, discussed with the statistician, and verified the underlying data. M. Perch managed the study, and was the sponsor as well as PI, participated in processing, and wrote the original draft of the manuscript. All authors were involved in patient inclusion, critical revisions and approving the final draft for submission. All authors had access to the aggregated data and accept responsibility for the decision to submit for publication.
Conflict of interest: M. Perch reports support for the present study from Intermune/Roche, grants from PulmonX and Therakos, payment or honoraria for lectures, presentations, manuscript writing or educational events from Zambon, participation on a data safety monitoring board or advisory board with Chiesi, PulmonX and Takeda, and leadership roles with European Respiratory Society, Group 8.02 Transplantation, Associate Medical Director for the International Thoracic Transplant registry, and treasurer of the European Society for Heart and Lung Transplantation. P. Corris reports participation on a data safety monitoring board or advisory board with MSD and Aerovate. J. Magnusson reports grants from Boehringer Ingelheim, Vicore Pharma, Takeda Pharma and AstraZeneca, payment or honoraria for lectures, presentations, manuscript writing or educational events from Mallinckrodt, Boehringer Ingelheim, GSK and AstraZeneca, and consulting fees from Boehringer Ingelheim, Vicore, Takeda and AstraZeneca. R. Vos reports support for the present study from Research Foundation-Flanders (FWO), and consulting fees from Zambon. N. Kneidinger reports payment or honoraria for lectures, presentations, manuscript writing or educational events from Roche. I. Leuckfeld reports support for the present study from Rigshospitalet, Copenhagen, Denmark, and payment or honoraria for lectures, presentations, manuscript writing or educational events from Boehringer Ingelheim. J. Gottlieb received reports grants from Deutsche Forschungsgemeinschaft, Zambon and German Center of Lung Research, consulting fees from Atheneum, payment or honoraria for lectures, presentations, manuscript writing or educational events from Novartis, AstraZeneca, CSL Behring, Takeda, Theravance, Pierre Fabre, Springer Healthcare, European Research Network and Moderna, expert testimony for Therevance, Pierre Fabre, Springer Healthcare, European Research Network and Moderna, participation on the data monitoring board for the ScanCLAD study, stocks in Pfizer, and receipt of equipment, materials, drugs, medical writing, gifts or other services from Pfizer. The remaining authors have no potential conflicts of interest to disclose.
Support statement: This work was supported by an unrestricted institutional grant from InterMune and Roche. Funding information for this article has been deposited with the Open Funder Registry.
Supplementary material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material: appendix A
ERJ-02154-2024.Appendix_A
Supplementary material: appendix B
ERJ-02154-2024.Appendix_B
Supplementary material: appendix C
ERJ-02154-2024.Appendix_C
Supplementary material: appendix D
ERJ-02154-2024.Appendix_D
Supplementary material: appendix E
ERJ-02154-2024.Appendix_E
Supplementary material: appendix F
ERJ-02154-2024.Appendix_F
Supplementary material: appendix G
ERJ-02154-2024.Appendix_G
Data availability
Data can be shared according to GDPR. Figures with pulmonary function tests over time and related to CLAD were constructed in the study. Study protocol, statistical analysis plan and informed consent will all be made available.
References
- 1.Verleden GM, Glanville AR, Lease ED, et al. Chronic lung allograft dysfunction: definition, diagnostic criteria, and approaches to treatment – a consensus report from the Pulmonary Council of the ISHLT. J Heart Lung Transplant 2019; 38: 493–503. doi: 10.1016/j.healun.2019.03.009 [DOI] [PubMed] [Google Scholar]
- 2.Chambers DC, Perch M, Zuckermann A, et al. The International Thoracic Organ Transplant Registry of the International Society for Heart and Lung Transplantation: thirty-eighth adult lung transplantation report – 2021; focus on recipient characteristics. J Heart Lung Transplant 2021; 40: 1060–1072. doi: 10.1016/j.healun.2021.07.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Burton CM, Carlsen J, Mortensen J, et al. Long-term survival after lung transplantation depends on development and severity of bronchiolitis obliterans syndrome. J Heart Lung Transplant 2007; 26: 681–686. doi: 10.1016/j.healun.2007.04.004 [DOI] [PubMed] [Google Scholar]
- 4.Verleden GM, Raghu G, Meyer KC, et al. A new classification system for chronic lung allograft dysfunction. J Heart Lung Transplant 2014; 33: 127–133. doi: 10.1016/j.healun.2013.10.022 [DOI] [PubMed] [Google Scholar]
- 5.Bedair B, Hachem RR. Management of chronic rejection after lung transplantation. J Thorac Dis 2021; 13: 6645–6653. doi: 10.21037/jtd-2021-19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Corris PA, Ryan VA, Small T, et al. A randomised controlled trial of azithromycin therapy in bronchiolitis obliterans syndrome (BOS) post lung transplantation. Thorax 2015; 70: 442–450. doi: 10.1136/thoraxjnl-2015-207080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Vos R, Vanaudenaerde B, Verleden S, et al. A randomised controlled trial of azithromycin to prevent chronic rejection after lung transplantation. Eur Respir J 2011; 37: 164–172. doi: 10.1183/09031936.00068310 [DOI] [PubMed] [Google Scholar]
- 8.Vanstapel A, Verleden SE, Verbeken EK, et al. Beyond bronchiolitis obliterans: in-depth histopathologic characterization of bronchiolitis obliterans syndrome after lung transplantation. J Clin Med 2021; 11: 111. doi: 10.3390/jcm11010111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Müller C, Rosmark O, Åhrman E, et al. Protein signatures of remodeled airways in transplanted lungs with bronchiolitis obliterans syndrome obtained using laser-capture microdissection. Am J Pathol 2021; 191: 1398–1411 doi: 10.1016/j.ajpath.2021.05.014 [DOI] [PubMed] [Google Scholar]
- 10.Liu H, Drew P, Gaugler AC, et al. Pirfenidone inhibits lung allograft fibrosis through l-arginine-arginase pathway. Am J Transplant 2005; 5: 1256–1263. doi: 10.1111/j.1600-6143.2005.00876.x [DOI] [PubMed] [Google Scholar]
- 11.Visner GA, Liu F, Bizargity P, et al. Pirfenidone inhibits T-cell activation, proliferation, cytokine and chemokine production, and host alloresponses. Transplantation 2009; 88: 330–338. doi: 10.1097/TP.0b013e3181ae3392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.McKane BW, Fernandez F, Narayanan K, et al. Pirfenidone inhibits obliterative airway disease in a murine heterotopic tracheal transplant model. Transplantation 2004; 77: 664–669. doi: 10.1097/01.TP.0000113162.48048.AA [DOI] [PubMed] [Google Scholar]
- 13.Zhou H, Latham CW, Zander DS, et al. Pirfenidone inhibits obliterative airway disease in mouse tracheal allografts. J Heart Lung Transplant 2005; 24: 1577–1585. doi: 10.1016/j.healun.2004.11.002 [DOI] [PubMed] [Google Scholar]
- 14.Bizargity P, Liu K, Wang L, et al. Inhibitory effects of pirfenidone on dendritic cells and lung allograft rejection. Transplantation 2012; 94: 114–122. doi: 10.1097/TP.0b013e3182584879 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ihle F, von Wulffen W, Neurohr C. Pirfenidone: a potential therapy for progressive lung allograft dysfunction? J Heart Lung Transplant 2013; 32: 574–575. doi: 10.1016/j.healun.2013.02.004 [DOI] [PubMed] [Google Scholar]
- 16.Bennett D, Fossi A, Valentini ML, et al. Pirfenidone in chronic lung allograft dysfunction after lung transplantation: a single-centre experience. Eur Respir J 2017; 50: Suppl. 61, OA1990. doi: 10.1183/1393003.congress-2017.OA1990 [DOI] [Google Scholar]
- 17.Matthaiou EI, Sharifi H, O'Donnell C, et al. The safety and tolerability of pirfenidone for bronchiolitis obliterans syndrome after hematopoietic cell transplant (STOP-BOS) trial. Bone Marrow Transplant 2022; 57: 1319–1326. doi: 10.1038/s41409-022-01716-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Stanojevic S, Kaminsky DA, Miller MR, et al. ERS/ATS technical standard on interpretive strategies for routine lung function tests. Eur Respir J 2022; 60: 2101499. doi: 10.1183/13993003.01499-2021 [DOI] [PubMed] [Google Scholar]
- 19.Martinu T, Koutsokera A, Benden C, et al. International Society for Heart and Lung Transplantation consensus statement for the standardization of bronchoalveolar lavage in lung transplantation. J Heart Lung Transplant 2020; 39: 1171–1190. doi: 10.1016/j.healun.2020.07.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Meyer KC, Raghu G, Verleden GM, et al. An international ISHLT/ATS/ERS clinical practice guideline: diagnosis and management of bronchiolitis obliterans syndrome. Eur Respir J 2014; 44: 1479–1503. doi: 10.1183/09031936.00107514 [DOI] [PubMed] [Google Scholar]
- 21.Todd JL, Finlen Copeland CA, Neely M, et al. FVC and FEV1 decline in chronic lung allograft dysfunction (CLAD) Phenotypes. J Heart Lung Transplant 2016; 35: S223–S224. doi: 10.1016/j.healun.2016.01.633 [DOI] [Google Scholar]
- 22.Greer M, Dierich M, De Wall C, et al. Phenotyping established chronic lung allograft dysfunction predicts extracorporeal photopheresis response in lung transplant patients. Am J Transplant 2013; 13: 911–918. doi: 10.1111/ajt.12155 [DOI] [PubMed] [Google Scholar]
- 23.Finlen-Copeland CA, Snyder LD, Zaas DW, et al. Survival after bronchiolitis obliterans syndrome among bilateral lung transplant recipients. Am J Respir Crit Care Med 2010; 182: 784–789. doi: 10.1164/rccm.201002-0211OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Ruttens D, Verleden SE, Demeyer H, et al. Montelukast for bronchiolitis obliterans syndrome after lung transplantation: a randomized controlled trial. PLoS One 2018; 13: e0193564. doi: 10.1371/journal.pone.0193564 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hofstetter E, Boerner G. Survival after lung transplantation in Europe – CLAD as major cause of death. J Heart Lung Transplant 2022; 41: Suppl., S286. doi: 10.1016/j.healun.2022.01.703 [DOI] [Google Scholar]
- 26.Todd JL, Neely ML, Finlen Copeland CA, et al. Prognostic significance of early pulmonary function changes after the onset of chronic lung allograft dysfunction. J Heart Lung Transplant 2019; 38: 184–193. doi: 10.1016/j.healun.2018.10.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Noble PW, Albera C, Bradford WZ, et al. Pirfenidone in patients with idiopathic pulmonary fibrosis (CAPACITY): two randomised trials. Lancet 2011; 377: 1760–1769. doi: 10.1016/S0140-6736(11)60405-4 [DOI] [PubMed] [Google Scholar]
- 28.Lancaster LH, de Andrade JA, Zibrak JD, et al. Pirfenidone safety and adverse event management in idiopathic pulmonary fibrosis. Eur Respir Rev 2017; 26: 170057. doi: 10.1183/16000617.0057-2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Khor YH, Goh NS, Wong AW, et al. Impact of concomitant medication burden on tolerability of disease-targeted therapy and survival in interstitial lung disease. Ann Am Thorac Soc 2022; 19: 962–970. doi: 10.1513/ANNALSATS.202108-980OC [DOI] [PubMed] [Google Scholar]
- 30.Verleden SE, Vasilescu DM, Willems S, et al. The site and nature of airway obstruction after lung transplantation. Am J Respir Crit Care Med 2014; 189: 292–300. doi: 10.1164/rccm.201310-1894OC [DOI] [PubMed] [Google Scholar]
- 31.Ross DJ, Lewis MI, Kramer M, et al. FK 506 ‘rescue’ immunosuppression for obliterative bronchiolitis after lung transplantation. Chest 1997; 112: 1175–1179. doi: 10.1378/chest.112.5.1175 [DOI] [PubMed] [Google Scholar]
- 32.Combs MP, Belloli EA, Gargurevich N, et al. Results from randomized trial of pirfenidone in patients with chronic rejection (STOP-CLAD study). J Heart Lung Transplant 2024; 43: 1468–1477. doi: 10.1016/j.healun.2024.05.013 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material: appendix A
ERJ-02154-2024.Appendix_A
Supplementary material: appendix B
ERJ-02154-2024.Appendix_B
Supplementary material: appendix C
ERJ-02154-2024.Appendix_C
Supplementary material: appendix D
ERJ-02154-2024.Appendix_D
Supplementary material: appendix E
ERJ-02154-2024.Appendix_E
Supplementary material: appendix F
ERJ-02154-2024.Appendix_F
Supplementary material: appendix G
ERJ-02154-2024.Appendix_G
Data Availability Statement
Data can be shared according to GDPR. Figures with pulmonary function tests over time and related to CLAD were constructed in the study. Study protocol, statistical analysis plan and informed consent will all be made available.