Long‐Term Proton Pump Inhibitors in Lung Transplant Recipients: Near‐Universal Use Without Clear Impact on Clinical Outcomes Over a 20‐Year Observational Study

ABSTRACT

Background

Gastroesophageal reflux disease (GERD) is a known risk factor for chronic lung allograft dysfunction (CLAD), the main barrier to long‐term survival in lung transplant recipients (LTR). While proton pump inhibitors (PPI) are effective for symptomatic treatment of GERD, data are sparse on the prevalence of long‐term PPI use and association with clinical outcomes.

Methods

Single‐center retrospective cohort of adult first‐time LTR between 1999 to 2018 who were alive and CLAD‐free at 1‐year post‐transplant. PPI exposure was defined as ≥8 continuous weeks in the first year, started within 6 months post‐transplant.

Results

1308 LTR were included, with 1170 (89.4%) exposed to PPI. PPI use increased over time and remained above 90% from 2007 onwards. In multivariable analyses, LTR on PPI had similar time to CLAD and time to death/retransplant compared to those not on PPI. Similarly, no differences in outcomes were observed between LTR on once‐daily and twice‐daily dosing, or after accounting for the number of reflux episodes on 24 h pH‐impedance monitoring. Positive bronchoalveolar lavage cultures and acute rejection over the first‐year post‐transplant was also comparable.

Conclusions

Long‐term use of PPI became near‐universal without observable effect on clinical outcomes. While further research is needed, these findings question the routine use of PPI in LTR.

Abbreviations

BAL

bronchoalveolar lavage

CMV

cytomegalovirus

CF

cystic fibrosis

COPD

chronic obstructive pulmonary disease

CLAD

chronic lung allograft dysfunction

FEV1

forced expiratory volume in 1 s

GERD

gastroesophageal reflux disease

ILD

interstitial lung disease

IQR

interquartile range

ISHLT

International Society for Heart and Lung Transplantation

LTR

lung transplant recipients

PFT

pulmonary function test

PH

pulmonary hypertension

PPI

proton pump inhibitor

1. Introduction

Gastroesophageal reflux disease (GERD) is prevalent after lung transplantation. [1] The relationship between GERD and poor outcomes in lung transplant recipients (LTR) remains debated. Some studies have shown associations between GERD with acute rejection and chronic lung allograft dysfunction (CLAD), but these have been inconsistent. [2, 3] GERD is thought to drive the microaspiration of gastric contents into the lungs, contributing to graft injury and inflammation. This can eventually lead to progressive and irreversible fibrosis manifesting as CLAD, which remains the leading cause of morbidity and mortality in LTR after the first year. [4]

Proton pump inhibitors (PPI) are among the most prescribed medications for the treatment of GERD. [5] By blocking the hydrogen‐potassium ATPase pump on parietal cells in the stomach, they effectively reduce acid secretion and alleviate the typical symptoms of reflux and heartburn. However, it is unlikely that PPI modify the underlying risk of microaspiration, which is believed to be primarily a mechanical issue related to the laxity of the lower esophageal sphincter as opposed to acidity. PPI also do not reduce the frequency or volume of reflux. [6] This is supported by a study [3] which found that LTR on PPI had fewer acidic reflux episodes compared to those off PPI, but there were no differences in non‐acidic reflux episodes or bronchoalveolar lavage markers of microaspiration, both of which have been linked to CLAD. [7, 8]

In idiopathic pulmonary fibrosis (IPF), PPI have been associated with clinical benefits in observational studies, including stabilization of lung function and reduction in mortality, which led to guidelines recommending PPI for all IPF patients irrespective of GERD. [9] In LTR, there is a paucity of research on PPI use and clinical outcomes, with only a small single‐center study showing a protective effect of PPI against acute and chronic rejection. [10] As such, the latest International Society for Heart and Lung Transplantation (ISHLT) guidelines on CLAD did not make any specific recommendations on the use of PPI. [4] Despite this lack of evidence, a recent survey of 44 lung transplant centers in Europe found that 84% of LTR were treated universally with PPI for prophylaxis against CLAD, often in the absence of GERD. [11]

There has been emerging concern over the potential overuse of PPI in the general population, particularly around adverse events and safety of long‐term PPI use. Guidelines have defined this as at least 8 continuous weeks, which is a clinically relevant time frame as this is generally beyond most acute indications for PPI such as peptic ulcer disease treatment or stress ulcer prophylaxis for ICU patients. [12] A meta‐analysis observed a 1.5‐fold increased risk of community‐acquired pneumonia among patients exposed to PPI therapy. [13] Several pathophysiologic mechanisms have been proposed, including microaspiration of altered gut microbiome in the setting of elevated stomach pH, although a causal relationship remains controversial due to potential biases. This association has not been specifically tested in LTR, who are notably immunosuppressed and at higher baseline risk of pneumonia, as well as more likely to have GERD and microaspiration. [1]

Our objective in this study was to evaluate the prevalence of long‐term PPI use in LTR over two decades at our center and association with clinically‐relevant outcomes, namely CLAD, death/retransplant, positive bronchoalveolar lavage (BAL) bacterial cultures, and acute rejection.

2. Materials and Methods

2.1. Study Design

This was a single‐center retrospective cohort study of all adult first‐time lung recipients transplanted at Toronto General Hospital between January 1, 1999 to December 31, 2018, who were alive and CLAD‐free at 1‐year post‐transplant, with at least 4 total pulmonary function tests (PFT) and adequate medication data. This study was approved by the University Health Network Research Ethics Board (Protocol number 15‐9531). Informed consent was obtained as required.

2.2. Sample Size Calculation

Sample size calculation was performed with time‐to‐CLAD as the primary outcome, with an accrual time of 20 years and additional follow‐up time of 5 years. We assumed a modest hazard ratio of 0.9, a two‐sided α of 0.05, a power of 0.8, a hazard of 0.1 per year in the unexposed group, and a ratio of 0.1 unexposed to exposed. This yielded a total required sample size of 1007 patients.

2.3. Medication Exposure

Exposure to medication was retrospectively reviewed from electronic health records. Additional data was obtained from the Toronto Lung Transplant Program Database. Patients who did not have at least 8 continuous weeks of prescriptions for corticosteroids, calcineurin inhibitors, and cell cycle inhibitors within the first‐year post‐transplant were excluded as having inadequate medication data. Long‐term PPI was defined as any use of oral dexlansoprazole, esomeprazole, lansoprazole, omeprazole, pantoprazole, or rabeprazole, at a frequency of once daily or more, for a continuous duration of at least 8 weeks in the first post‐transplant year, started within 6 months post‐transplant.

2.4. Clinical Protocols and Definitions

Clinical standards of care for LTR at our center was previously described. [14] CLAD was defined as a persistent forced expiratory volume in 1 s (FEV1) decline of greater than or equal to 20% from baseline, after exclusion of other causes, with the onset defined as the date of the first FEV1 decline, in accordance with ISHLT 2019 consensus. [4] Patients who did not develop CLAD were censored on the date of the last PFT up to July 12, 2024, or when they had a decline in FEV1 greater than or equal to 20% from baseline deemed not related to CLAD. CLAD status and date of onset were individually adjudicated by a lung transplant specialist. Death/retransplant was censored on July 12, 2024. A list of bacteria considered pathogenic in BAL is provided in Supplemental Materials Appendix A. A‐score was defined as the average acute rejection grade, obtained as the sum of all A grades divided by all evaluable transbronchial biopsies. [15] Post‐transplant pH‐impedance testing was routinely performed in our center starting in 2003.

2.5. Statistical Analysis

For baseline patient characteristics, Mann‐Whitney and chi‐square test were used to compare continuous and discrete variables, respectively. Kaplan‐Meier analyses and Cox proportional hazards models were used to assess the association between PPI exposure with time to CLAD and time to death/retransplantation, landmarked from 1‐year post‐transplant. Covariates including age, sex, transplant type (bilateral vs single vs heart lung), primary disease, cytomegalovirus (CMV) serology mismatch (vs all others), transplant year (as a continuous variable), and the number of reflux episodes on 24 h pH‐impedance monitoring were included in multivariable models as indicated. Sensitivity analysis using propensity score overlap weighting was performed for time to CLAD and time to death/retransplantation, with propensity scores generated from the same covariates. The proportion of positive BAL bacterial cultures and the A‐score over the first‐year post‐transplant were analyzed using a two‐part model. First, multivariable logistic regression was used to model the association of PPI exposure with the probability of any non‐zero value, followed by multivariable linear regression to compare the mean differences between groups among patients with non‐zero values. A p‐value of less than 0.05 was considered statistically significant. All statistical analyses were performed using R version 4.5.2.

3. Results

3.1. Baseline Patient Characteristics

1308 patients met inclusion criteria (Figure 1). Median age at transplant was 56 years and 56% were male (Table 1). 81.9% of patients underwent bilateral lung transplantation. Restrictive lung disease was the most common indication for lung transplantation at 44.3%. 1170 (89.4%) were exposed to long‐term PPI, with a median (IQR) of 360 (11) cumulative days on PPI in the first post‐transplant year. 138 (10.6%) were not exposed to long‐term PPI, with a median (IQR) of 0 (0) cumulative days on PPI in the first post‐transplant year. Of patients on long‐term PPI therapy, 597 (51.0%) were on once daily dosing and 573 (49.0%) were on twice daily dosing. Compared to those not on long‐term PPI, more patients on long‐term PPI were older (median age 57 vs 52) and transplanted after 2008 (73.1% vs 29.7%). Patients who were on long‐term PPI were comparable to those not on long‐term PPI with regards to sex, transplant type (bilateral vs single vs heart/lung), primary disease, CMV serology mismatch, and FEV1 at 6 months and at 1 year.

FIGURE 1.

Cohort flow diagram. Cohort flow diagram showing inclusion and exclusion criteria, resulting in the final analysis cohort of 1308 patients, of whom 138 had no long‐term PPI exposure and 1170 had long‐term PPI exposure.

TABLE 1.

Patient baseline characteristics.

Variable	Overall	No long‐term PPI exposure	Long‐term PPI exposure	p‐value
	(n = 1308)	(n = 138)	(n = 1170)
Age, median (SD)	56 (14)	52 (14)	57 (14)	0.004
Gender, female, n (%)	575 (44.0)	51 (37.0)	524 (44.8)	0.08
Transplant type				0.70
Bilateral, n (%)	1071 (81.9)	110 (79.7)	961 (82.1)
Single, n (%)	217 (16.6)	25 (18.1)	192 (16.4)
Heart/lung, n (%)	20 (1.5)	3 (2.2)	17 (1.5)
Primary disease				0.55
COPD/Obstructive, n (%)	348 (26.6)	39 (28.3)	309 (26.4)
CF/Suppurative, n (%)	267 (20.4)	30 (21.7)	237 (20.3)
PH/Pulmonary Vascular, n (%)	39 (3.0)	6 (4.3)	33 (2.8)
ILD/Restrictive, n (%)	579 (44.3)	53 (38.4)	526 (45.0)
Other, n (%)	75 (5.7)	10 (7.2)	65 (5.6)
CMV serology mismatch, n (%)	257 (19.6)	29 (21.0)	228 (19.5)	0.67
Transplanted after 2008, n (%)	896 (68.5)	41 (29.7)	855 (73.1)	<0.001
Percent predicted FEV1 at 6 months, median (SD)	61.9 (19.7)	65.9 (19.9)	61.8 (19.7)	0.21
Percent predicted FEV1 at 1 year, median (SD)	62.4 (19.6)	64.2 (20.8)	62.3 (19.6)	0.49

3.2. Prevalence of PPI Use over Time

The prevalence of PPI use rose from around 20% in 1999 to over 90% in 2007 and remained above 90% in every subsequent year up to and including 2018 (Figure 2A). Among patients on PPI, the proportion on twice‐daily dosing was generally higher in the first decade (1999–2008) compared to the second decade (2009–2018) (Figure 2B).

FIGURE 2.

Absolute and relative number of patients exposed to long‐term PPI from 1999 to 2018. Panel A shows the absolute number of patients exposed to long‐term PPI each year, separated by once‐daily and twice‐daily dosing, over the study period from 1999 to 2018. Panel B shows the percentage of patients exposed to long‐term PPI among all patients in each year (black line), as well as the percentage of patients on twice‐daily PPI among all patients on PPI each year (orange line) over the study period from 1999 to 2018.

3.3. Association of PPI Use with CLAD and Death/Retransplant

687 (52.5%) patients developed CLAD, with a median follow‐up time after transplant of 4.9 years. 826 (63.1%) patients died or were retransplanted, with a median follow‐up time after transplant of 7.2 years. Figure 2 demonstrates Kaplan‐Meier curves for the cumulative incidence of CLAD and death/retransplantation by long‐term PPI exposure, which were not statistically different. As a sensitivity analysis, this model was restricted to patients in earlier decade (1999–2008), with similar results (Supplemental Materials Appendix B). There were also no significant differences in the time to CLAD or death/retransplant between patients on once‐daily vs twice‐daily PPI (Supplemental Materials Appendix C), Figure 3.

FIGURE 3.

Cumulative incidence of CLAD and death/retransplant by long‐term PPI exposure. Kaplan Meier curves for the cumulative incidence of CLAD (Panel A) and death/retransplant (Panel B), separated by PPI exposure, with patients not exposed to long‐term PPI represented in red and patients exposed to long‐term PPI represented in teal.

In unadjusted Cox models, long‐term PPI exposure was not significantly associated with risk of CLAD (hazard ratio [HR] 0.876; 95% confidence interval [CI] 0.700‐1.098; p = 0.251) or risk of death/retransplant (HR 1.049; 95% CI 0.855‐1.287; p = 0.647). In multivariable Cox models adjusted for age, sex, transplant type (bilateral vs single vs heart lung), primary disease, cytomegalovirus (CMV) serology mismatch, and transplant year, long‐term PPI exposure was not significantly associated with risk of CLAD (HR 0.957; 95% CI 0.745‐1.229; p = 0.730) or risk of death/retransplant (HR 0.977; 95% CI 0.778‐1.228; p = 0.844) (Table 2, Model 1). We alternatively modeled PPI exposure using the cumulative number of days on PPI in the first‐year post‐transplant as a continuous variable. Cumulative exposure to PPI was not significantly associated with CLAD (multivariable HR per 30‐day exposure to PPI of 1.00; 95% CI 0.98‐1.02; p = 0.89) or death/retransplant (multivariable HR per 30‐day exposure to PPI of 0.99; 95% CI 0.97‐1.01; p = 0.32). (Table 2, Model 2).

TABLE 2.

PPI exposure and association with CLAD and death/retransplant in unadjusted and adjusted Cox models.

	CLAD		Death/retransplantation
Variable	HR (95% CI)	p‐value	HR (95% CI)	p‐value
Model 1: Long‐term PPI exposure (minimum 8 weeks continuous) in the first‐year post‐transplant, yes vs. no
Univariable (unadjusted) models
Long‐term PPI exposure: Yes	0.876 (0.700–1.098)	0.251	1.049 (0.855–1.287)	0.647
Multivariable (adjusted) models
Long‐term PPI exposure: Yes	0.957 (0.745–1.229)	0.730	0.977 (0.778–1.228)	0.844
Age	0.995 (0.988–1.004)	0.284	1.014 (1.006–1.021)	<0.001
Sex: Male	0.985 (0.843–1.151)	0.852	0.981 (0.851–1.131)	0.791
Transplant type: Single	1.490 (1.203–1.844)	<0.001	1.772 (1.478–2.126)	<0.001
Transplant type: Heart/lung	0.759 (0.386–1.492)	0.424	1.372 (0.722–2.609)	0.335
Primary disease: CF/Suppurative	0.807 (0.603–1.081)	0.151	1.081 (0.825–1.414)	0.573
Primary disease: PH/Pulmonary vascular	0.897 (0.582–1.383)	0.623	0.585 (0.357–0.961)	0.034
Primary disease: ILD/Restrictive	0.780 (0.646–0.942)	0.010	0.973 (0.822–1.150)	0.744
Primary disease: Other	0.826 (0.562–1.213)	0.329	0.622 (0.414–0.933)	0.022
CMV serology mismatch: Yes	1.349 (1.123–1.619)	0.001	1.349 (1.141–1.596)	<0.001
Transplant year	0.990 (0.973–1.006)	0.227	1.010 (0.994–1.027)	0.221
Model 2: Cumulative PPI exposure in the first‐year post‐transplant
Univariable (unadjusted) models
Cumulative PPI exposure: Per 30 days	0.990 (0.971–1.010)	0.326	1.000 (0.983–1.018)	1.000
Multivariable (adjusted) models
Cumulative PPI exposure: Per 30 days	0.998 (0.977–1.021)	0.887	0.990 (0.971–1.010)	0.316
Age	0.996 (0.988–1.004)	0.282	1.013 (1.006–1.021)	<0.001
Sex: Male	0.987 (0.845–1.153)	0.867	0.978 (0.848–1.127)	0.755
Transplant type: Single	1.490 (1.204–1.845)	<0.001	1.771 (1.476–2.124)	<0.001
Transplant type: Heart/lung	0.760 (0.387–1.494)	0.426	1.362 (0.716–2.592)	0.346
Primary disease: CF/Suppurative	0.807 (0.603–1.080)	0.149	1.082 (0.827–1.416)	0.565
Primary disease: PH/Pulmonary vascular	0.897 (0.582–1.384)	0.623	0.579 (0.352–0.951)	0.031
Primary disease: ILD/Restrictive	0.780 (0.646–0.942)	0.010	0.971 (0.821–1.148)	0.727
Primary disease: Other	0.824 (0.561–1.210)	0.323	0.627 (0.418–0.942)	0.025
CMV serology mismatch: Yes	1.350 (1.124–1.621)	0.001	1.346 (1.138–1.591)	<0.001
Transplant year	0.989 (0.973–1.006)	0.205	1.013 (0.997–1.030)	0.117

Reference group for transplant type is bilateral. Reference group for primary disease is COPD/Obstructive.

3.4. Sensitivity Analyses

Results were materially unchanged when year of transplant was modelled as a cubic spline with 3 degrees of freedom to account for stepwise and nonlinear improvements over time or when death/retransplant was added as a competing risk for CLAD in cause‐specific Cox models. As an additional sensitivity analysis, propensity score overlap weighting yielded similar results for CLAD (HR 0.917; 95% CI 0.698‐1.204; p = 0.533) and death/retransplant (HR 0.927; 95% CI 0.728‐1.181; p = 0.541). Mean differences in covariates between groups were acceptable after weighing. (Supplemental Materials Appendix D).

3.5. Patients Who Underwent Objective Reflux Testing

In 674 patients who underwent post‐transplant 24 h pH‐impedance testing results within 1‐year post‐transplant, 631 (93.6%) were exposed to long‐term PPI and 43 (6.4%) were not. Among patients on PPI, only 100 (16%) had objective reflux as defined by ≥ 40 reflux episodes in 24 h. Multivariable models accounting for the total reflux episodes in 24 h showed that PPI exposure was not significantly associated with risk of CLAD (HR 0.892; 95% CI, 0.578‐1.377; p = 0.607) or death/retransplant (HR 0.930; 95% CI, 0.628‐1.379; p = 0.719) (Supplemental Materials Appendix E).

3.6. Association of PPI Use with Positive BAL Bacterial Cultures and Acute Rejection over the First Post‐Transplant Year

Compared to patients unexposed to PPI, patients exposed to PPI were just as likely to have at least one positive BAL culture over the first post‐transplant year (odds ratio [OR] 1.41; 95% CI, 0.95, 2.11; p = 0.09). Among patients with at least one positive respiratory culture, the mean proportion was 0.30 in those exposed to PPI, compared to 0.37 in those unexposed to PPI (p = 0.32) (Supplemental Materials Appendix F).

97 (7.4%) patients did not have any evaluable transbronchial biopsies. Among patients with at least one evaluable transbronchial biopsy, the likelihood of having at least one episode of acute rejection over the first post‐transplant year was similar in patients exposed to PPI, compared to those who were not (OR 1.38; 95% CI, 0.89, 2.13; p = 0.14). Among patients with at least one episode of acute rejection, the mean A‐score was 0.52 in those exposed to PPI, compared to 0.54 in those unexposed to PPI (p = 0.13) (Supplemental Materials Appendix H).

4. Discussion

Our study found no evidence of an association between long‐term PPI therapy in the first‐year post‐transplant with subsequent risk of CLAD or death/retransplant, nor with positive BAL cultures or acute rejection within the first‐year post‐transplant. There were several notable differences between groups, specifically the patients on long‐term PPI tended to be older and transplanted in the later era. Large registry data have shown that older age is associated with worse outcomes, while later transplant era has been associated with better outcomes. [16] Nonetheless, there remained no significant associations in any of our measured outcomes after adjustment for these potential confounders in multivariable regression models or propensity score overlap weighting.

Given that the group unexposed to long‐term PPI was small and mostly from the first era (1999 to 2008), we restricted a sensitivity analysis to patients in this era where the group sizes were more comparable, with similar results. We also performed additional analyses comparing patients who were on once daily vs twice daily PPI to determine if there was any dose‐response effect. This demonstrated no difference with higher dose PPI, despite the groups being nearly equal in size. However, this interpretation may be challenging as it is unclear why patients were prescribed twice‐daily PPI. Although it is likely that some of these patients had acute indications such as upper gastrointestinal bleeding from peptic ulcer disease, many patients may have had worse GERD, either by clinical symptoms or objective testing, to have prompted the higher dose.

We further explored this possibility in a subset of patients who had available post‐transplant 24 h pH/impedance monitoring. We used the number of reflux episodes as a continuous variable, rather than a cutoff of 40 or 80 episodes to define GERD proposed in the updated Lyon consensus 2.0 for diagnosis of GERD [17], as the thresholds defining abnormal reflux have not been validated in LTR. As with the previous analyses, we were unable to observe any clear effect from long‐term PPI use after adjusting for the number of reflux episodes. Moreover, the number of reflux episodes itself was not associated with CLAD or death/retransplant in this cohort. This highlights that the relationship between GERD with poor outcomes after lung transplant remains controversial.

Our results are discordant with a previous observational study which found a protective effect of PPI on acute and chronic rejection. [10] However, this study had a considerably smaller sample size of 188 patients and shorter follow‐up period compared to ours. Moreover, a composite outcome of acute and chronic rejection was used, but when CLAD was assessed as a separate outcome, no association with PPI use was found, which is congruent with our findings. To the best of our knowledge, there are no other observational studies, nor any published or ongoing randomized trials addressing this clinical question. Therefore, we believe that our study is likely to represent the strongest evidence to date.

In addition to finding that long‐term PPI therapy conferred no clear benefit in our study, it is important to note that we did not observe any signal for harm in the outcomes studied, reinforcing the current paradigm that PPI are safe in LTR when deemed necessary. Reassuringly, PPI use was not associated with positive BAL cultures, although we could not incorporate any clinical and radiological criteria to define pneumonia, which would have been ideal. Future research should assess other potential adverse effects linked to long‐term PPI use such as Clostridium difficile infection and osteoporosis, which were not captured in our study.

A key finding of our study was that over two decades, the proportion of LTR on long‐term PPI at our center increased dramatically to near‐universal levels above 90%. It is also notable that roughly half were on twice‐daily dosing, which is higher than the typical dose required for GERD. The use of routine PPI appears to be similarly widespread at other lung transplant centers, evidenced by a recent multicenter survey from Europe reporting that 84% of patients were on prophylactic PPI. [11]

Given the common use of long‐term PPI at many lung transplant centers worldwide, we believe that our results have potential implications for clinical practice. [11] Lung transplant recipients who do not have GERD or another appropriate indication for long‐term PPI should be considered for a trial of deprescribing. While further research is required to establish how to safely and effectively deprescribe PPI in LTR, this may significantly reduce adverse effects, healthcare costs, and pill burden. On the other hand, LTR with documented GERD and declining lung function should perhaps be treated with PPI strictly for the purpose of relieving any GERD symptoms. Importantly, PPI should not substitute or delay early fundoplication, which has been associated with improved allograft outcomes and supported by the latest ISHLT consensus report on CLAD. [4]

Our study should be interpreted in the context of several important limitations. First, this was an observational, non‐randomized study of an intervention. Although we measured and adjusted for a multitude of confounders in our analysis, it remains vulnerable to bias from residual and unmeasured confounding between the treated and control group, especially as we could not determine the clinical indication or rationale behind PPI use. For this reason, our study falls short of establishing causality on its own. Second, we used data documented in electronic health records to ascertain long‐term PPI exposure. Despite our use of data on immunosuppressants to validate this approach, it is possible that PPI data was inaccurately captured to some extent. We were also unable to assess medication adherence, which could have influenced the true exposure to PPI and its effects on outcomes. Third, a substantial proportion of patients did not have 24 h pH‐impedance monitoring data which limited the ability to adjust for reflux episodes in our primary models, and hence the results from this analysis should be considered exploratory.

In summary, our retrospective cohort study demonstrated that long‐term PPI therapy in the first year after lung transplant increased to near‐universal levels from 1999 to 2018. We found no evidence that PPI was associated with benefit or harm in clinical outcomes including subsequent risk of CLAD or death/retransplant beyond 1‐year post‐transplant, or positive BAL cultures and acute rejection within 1‐year post‐transplant. The relationship between PPI, GERD, and CLAD is highly complex, with further research ideally in the form of randomized trials needed to better inform clinical practice. It is important to note that there are many established indications for long‐term PPI which our study was unable to fully examine, such as stress ulcer prophylaxis in the early post‐transplant period, and that our findings should not discourage the appropriate use of PPI. Instead, we hope that our results encourage the field to re‐evaluate the widespread use of universal long‐term PPI for CLAD prevention. In the appropriate clinical context, LTR on long‐term PPI, unless indicated otherwise, could be considered for a trial of deprescribing.

Author Contributions

Chen Yang Kevin Zhang: Conceptualization, Methodology, Formal Analysis, Investigation, Data Curation, Writing – original draft, Writing – review & Editing, Rayoun Ramendra: Data Curation, Meghan Aversa: Data Curation, Rasheed Ghany: Software, Data Curation, Jussi Tikkanen: Data Curation, Ella Huszti: Methodology, Software, Jonathan Yeung: Supervision, Shaf Keshavjee: Supervision, Lianne G. Singer: Supervision, Tereza Martinu: Conceptualization, Resources, Writing – Review and Editing, Supervision and Project Administration

Funding

The authors received funding from the University Health Network Foundation.

Conflicts of Interest

The authors declare no conflicts of interest.

Disclosure

The authors of this manuscript have the following conflicts of interest to disclose: Lianne G Singer: Sanofi (advisory board), Renovion (DSMB), Zambon (DSMB). Tereza Martinu: Sanofi (advisory board, clinical trial), APCBio Innovations (research material), Trove Therapeutics (collaboration).

Supporting information

Supporting Information: ctr70706‐sup‐0001‐SuppMat.docx

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.