key: cord-0727267-2l5m0kx3
authors: Razia, Deepika; Mittal, Sumeet K.; Bansal, Sandhya; Ravichandran, Ranjithkumar; Smith, Michael A.; Walia, Rajat; Bremner, Ross M.; Mohanakumar, Thalachallour; Tokman, Sofya
title: Lung Transplant Candidates With Pretransplant Gastroesophageal Reflux and Antibodies to Lung Self-antigens Have Shorter CLAD-free Survival After Transplant
date: 2022-02-11
journal: Transplant Direct
DOI: 10.1097/txd.0000000000001294
sha: 74a64e486e07458fc507f9d3dd7fd3c07aba049e
doc_id: 727267
cord_uid: 2l5m0kx3

Pre–lung transplant (LTx) gastroesophageal reflux (GER) and circulating antibodies against the lung self-antigens (SAbs) collagen V and K-alpha-1 tubulin may predispose recipients to chronic lung allograft dysfunction (CLAD). We aimed to study the association of pre-LTx GER or pre-LTx SAbs with CLAD. METHODS. In this retrospective analysis of patients who underwent LTx between 2015 and 2019, pre-LTx GER and SAbs were dichotomously defined as present or absent. The study group comprised recipients with either GER‚ SAbs, or both, and the control group comprised recipients without GER or SAbs. Endpoints included CLAD and survival. RESULTS. Ninety-five LTx recipients were divided into a study group (n = 71; 75%) and a control group (n = 24; 25%). Pretransplant GER was associated with pre-LTx SAbs (odds ratio [95% confidence intervals], 5.022 [1.419-17.770]; P = 0.012). In addition, the study group (either GER‚ SAbs, or both) had a higher risk of CLAD (hazard ratio [95% confidence intervals], 8.787 [1.694-45.567]; P = 0.010) and lower CLAD-free survival after LTx than the control group (P = 0.007); however, overall survival was similar between the 2 groups (P = 0.618). CONCLUSIONS. GER was associated with elevated SAbs in LTx candidates, and either GER, SAbs, or both were associated with CLAD in LTx recipients. This association suggests that GER may cause an immune response to normally sequestered lung-associated self-antigens that drives ongoing lung injury.

Lung transplantation (LTx) can be a life-extending option for patients with end-stage lung disease; however, longterm survival after LTx is shorter than long-term survival after other solid organ transplants, and mortality is driven by chronic lung allograft dysfunction (CLAD). 1 A number of risk factors for the development of CLAD have been identified and include primary graft dysfunction (PGD), 2 acute cellular rejection (ACR), development of donorspecific antibodies (DSAs), and recurrent infections. 1 Gastroesophageal reflux (GER) and the resultant aspiration may also be a risk factor for CLAD, 3, 4 and pre-LTx GER may be associated with early allograft injury and higher 1-y mortality. 5, 6 Furthermore, elevated titers of pre-LTx antibodies to the lung self-antigens (SAbs) collagen V (Col-V) and K-alpha-1 tubulin (Kα1T) have been identified in LTx recipients (LTxRs) with CLAD 7,8 and may have a dual role as both a biomarker and a propagator of lung injury. In this retrospective cohort study, we hypothesized that the presence of GER in LTx candidates may drive lung injury through a resultant expression of SAbs and that both GER and SAbs before transplant are a risk factor for the development of CLAD after transplant.

All included patients consented and were prospectively enrolled for research as part of a National Institutes of Healthapproved study (NIH HL056643) in accordance with the principles of the Declaration of Helsinki. After institutional review board approval (PHXB-16-0027-10-18 dated March 7, 2016), we retrospectively retrieved pre-LTx SAb assay results for samples collected between 2015 and 2019. The esophageal disease center database was also queried to extract pre-LTx 24-h GER testing data at the time of LTx evaluation for these patients. From these data, patients with unavailable 24-h GER testing or 24-h GER testing done on acid suppression therapy and patients declined after LTx candidacy evaluation or remaining on the United Network for Organ Sharing waitlist as of March 31, 2020, were excluded. Serum samples obtained closest to the time of LTx listing were selected and analyzed for SAbs (Col-V and Kα1T) by ELISA as previously described. 8 Thus, samples serially obtained throughout the evaluation process, particularly among patients whose LTx listing was deferred, were discarded. This allowed for relative uniformity in the time frame between sample acquisition and LTx. Patients transplanted between January 2015 and June 2019 were included in the analysis. Patient selection for LTx was standardized and based on International Society for Heart and Lung Transplantation criteria. 9

The presence of SAbs was dichotomously defined as present (anti-Col-V or anti-Kα1T titers ≥106 ng/mL and ≥116 ng/ mL, respectively) or absent. The presence of GER was also dichotomously defined as present (DeMeester score ≥14.72) or absent. Uni-and multivariate analyses were performed to ascertain predictors of SAbs in LTx candidates. LTxRs with either GER‚ SAbs, or both were assigned to the study group, and recipients without GER or SAbs were assigned to the control group. Electronic charts of LTxRs were reviewed, and the study and control groups were compared for post-LTx outcomes and survival. The study group was further divided into 3 subgroups for stratified analysis of outcome: subgroup 1 = GER(+) SAbs(+); subgroup 2 = GER(-) SAbs(+); and subgroup 3 = GER(+) SAbs(-).

The primary endpoint of the study was time from LTx to first detection of CLAD. International Society for Heart and Lung Transplantation criteria were used to define CLAD. 1, 10 Restrictive allograft syndrome and bronchiolitis obliterans syndrome were not separately defined for the study. The secondary endpoint was time from LTx to all-cause mortality and overall survival. All outcomes were adjudicated by transplant physicians blinded to GER results and SAb titers.

Clinical covariates of LTxRs were recorded throughout the study and included graft ischemia time, use of cardiopulmonary bypass or extracorporeal membrane oxygenation (ECMO), post-LTx length of hospital stay, baseline allograft function, pretransplant respiratory infections, PGD grade 3 at 72 h post-LTx, ACR (≥A2 or ≥B1R), and de novo development of DSAs (>1000 mean fluorescence intensity) to mismatched donor HLAs any time after LTx. Baseline lung allograft function was defined as the mean of the 2 best forced expiratory volumes in 1 s and forced vital capacity measurements during the first 6 mo after LTx. We did not differentiate between respiratory colonization and infection. Thus, pre-LTx respiratory infections were defined as having a respiratory specimen (sputum, tracheal aspirate, or bronchoalveolar lavage) with a positive bacterial culture or a nasopharyngeal swab or nasal wash with a positive viral polymerase chain reaction test within 6 mo before serum sample collection.

Ambulatory esophageal pH monitoring was performed using a catheter-based dual, proximal and distal esophageal, electrode probe (Digitrapper 400pH; Medtronic, Minneapolis, MN), and pH score was calculated using standard variables. 11 Briefly, the catheter-based pH probe was passed transnasally and positioned 5 cm above the upper border of the lower esophageal sphincter (defined on high-resolution manometry). Testing was done off acid suppression therapy (7 d for proton pump inhibitors and 3 d for H2 receptor blockers).

DSAs to HLA SAbs Kα1T and Col-V were detected via ELISA assay, as previously described. 8 In brief, ELISA plates were coated overnight at 4 °C with either recombinant Kα1T (1 μg/mL) or Col-V (1 μg/mL; Sigma-Aldrich, St. Louis, MI) in PBS and blocked for 2 h with 1% BSA. Samples from pre-LTx patients and healthy volunteers were diluted 1:1000 for Col-V and 1:1250 for Kα1T and loaded. Color was developed using tetramethylbenzidine substrate, and SAbs were detected using horseradish-peroxidase conjugated antihuman IgG (1:10,000) and read at 450 nm. SAbs (Kα1T, Col-V) were considered positive in readings 2 standard deviations above the mean of healthy controls (116 ng/mL for K-α1T and 106 ng/mL for Col-V). Antibody concentrations were calculated using standard curves of known concentrations of anti-Kα1T (Santa Cruz Biotechnology, Dallas, TX) or anti-Col-V (Abcam, Cambridge, United Kingdom). DSAs were detected using single antigen beads (One Lambda, ThermoFisher Scientific, Waltham, MA) using a Luminex platform.

Induction therapy included high-dose corticosteroids (methylprednisolone) before perfusion of each lung allograft and antilymphocyte therapy with basiliximab, rituximab, or antithymocyte globulin, with basiliximab as the induction agent of choice. The initial maintenance immunosuppressive regime was uniform across the study period and consisted of a corticosteroid (prednisone), an antiproliferative agent (mycophenolate mofetil or mycophenolic acid), and a calcineurin inhibitor (tacrolimus or cyclosporine).

All analyses were performed using IBM SPSS Statistics for Windows, version 23.0 (IBM Corp, Armonk, NY)‚ and R package 4.1.0 (R Foundation for Statistical Computing, Vienna, Austria). Continuous variables were expressed as median (interquartile range [IQR]), and categorical variables were expressed as frequencies (percentage). Differences in continuous variables between study and control groups were assessed using the nonparametric Kruskal-Wallis test. The chisquared or Fisher exact test was used to compare categorical variables between groups.

Univariate analysis was used to determine the predictors of SAbs in LTx candidates. Variables with P < 0.2 were included in a multivariate regression model. Included covariates were demographic characteristics of the LTx candidate, lung allocation score (LAS), underlying lung disease, pre-LTx hospitalization for respiratory infections, and pre-LTx GER. In addition, covariates relevant to the outcome (SAbs) in the literature, that is, type of transplantation, 12 were also included. Next, a univariate Cox regression was conducted with proportional hazard assumption for the endpoint of CLAD in LTxRs. For the multivariate Cox proportional hazard analysis, covariates based on statistical significance (P < 0.2 in univariate analysis) and variables known to be clinically important, 2,7 that is, PGD, DSA, and ACR, were purposefully selected. Significance was evaluated at the 0.05 alpha level. Confounding was defined as present if the change in the parameter estimate on the model adjusted for PGD, DSA, and ACR was >15% than the unadjusted model. Finally, stratified analysis was conducted to assess the relationship of CLAD to subgroups 1, 2, and 3 compared with the control group. The survival time was calculated from LTx to first detection of CLAD or death. Not meeting any of the 2 endpoints by the last clinical follow-up defined CLAD-free survival; time from LTx to death defined overall survival. Cumulative CLAD-free and overall survival was calculated using the Kaplan-Meier method. The log-rank test was used to compare survival rates between groups. A P value for the subgroups' survival analysis was adjusted using the Bonferroni method. Statistical significance was set at P ≤ 0.050.

In total, serum samples from 166 patients were analyzed for the presence of SAbs. Patients with inadequate GER testing (n = 27) and those on the United Network for Organ Sharing waitlist at the end of the study period (n = 44) were excluded; 95 LTxRs formed the study cohort ( Figure 1 ). The median (IQR) patient age, body mass index, and LAS at the time of LTx were 65 y (56-71), 26 kg/m 2 (22) (23) (24) (25) (26) (27) (28) (29) (30) , and 38 (34-45), respectively; 50.5% (n = 48) were men, and the most common indication for LTx was idiopathic pulmonary fibrosis (IPF). A bilateral LTx was performed in 96.8% (n = 92) of patients, and 4 had a redo-LTx.

Lung transplant recipients were assigned to the study group (n = 71, 74.7%) or the control group (n = 24, 25.3%). Baseline characteristics, including age, gender, body mass index, LAS, and underlying lung disease‚ were comparable between the study and control groups ( Table 1 ). Post-LTx clinical outcomes including graft ischemic time, use of cardiopulmonary bypass, ECMO rescue for severe PGD, length of postoperative hospital stay, and baseline allograft function were also comparable between the study and control groups. The prevalence of PGD and DSA among the study group was significantly higher than the prevalence in the control group (64.8% versus 37.5%, P = 0.019‚ and 70.4% versus 37.5%, P = 0.004, respectively), whereas the prevalence of ACR was similar between the study and control groups (32.4% versus 20.8%, P = 0.283). Notably, the prevalence of CLAD was significantly higher in the study group than in the control group (42% versus 13%, P = 0.008; Table 1 ).

In LTx candidates, underlying lung disease, pretransplant GER, and history of pretransplant hospitalization for respiratory infections were associated with SAbs on univariate Table 3 ).

We suspected a confounding association of PGD, DSA, and ACR between the study group and CLAD. The covariates were sequentially entered one at a time for adjustment, and the percent change in parameter estimate as compared with the unadjusted model was noted. PGD and ACR did not retain a confounding association; however, DSA was significantly related to CLAD and possibly made an important contribution in the presence of GER and SAbs ( Table 4) .

Irrespective of confounding, the probability of CLAD-free survival was significantly lower in the study group than in the control group (57.7% versus 87.5%, P = 0.007; Figure 2 ). The median (IQR) CLAD-free survival in the study group and the control group was 22 and 35 (22-40) mo, respectively.

Next, a synergistic association between GER and SAbs was investigated for a possible dose response relationship with CLAD. The study group was further divided into 3 subgroups. Subgroup 1 (n = 39; 41%) included GER(+) SAbs(+) recipients‚ subgroup 2 (n = 18; 19%) included GER(-) SAbs(+) recipients‚ and subgroup 3 (n = 14; 15%) included GER(+) SAbs(-) recipients. CLAD-free survival in subgroups was compared with the control group. Stratified analysis showed the risk of CLAD was significantly higher among patients in (Table 5) ; however, a synergistic association between GER and SAbs with CLAD was not identified. The probability of CLAD-free survival of subgroups 1, 2, and 3 was significantly lower than that of the control group (61.5%, 55.6%, 50.0%, and 87.5%, respectively; P = 0.044; Figure 3 ), with significant intergroup differences compared with the control group (P = 0.023, 0.013, and 0.007, respectively). The median (IQR) CLAD-free survival of subgroup 1, 2, and 3 was 24 (14, 36), 21 (13, 39) , and 22 (13, 29) mo, respectively.

Finally, the median (IQR) post-LTx follow-up time of the study cohort was 45 mo (33-50). The overall mortality in the cohort was 23% (n = 22). Mortality on index admission was 3.2% (n = 3); 1 patient had severe PGD requiring ECMO support‚ 1 had a posttransplant cerebrovascular accident and subsequent multiorgan system failure‚ and 1 had recurrent respiratory infections, delirium, multiorgan dysfunction, and resultant failure to thrive. At the end of follow-up, the deaths of 2 patients (both in the control group) were attributed to SARS-CoV-2 infection ( Table 2 ). The probability of overall survival was comparable between the study group and the control group (78.9% versus 70.8%, P = 0.618; Figure 4 ). The median (IQR) survival of the study group and the control group was also comparable (39 [25-48] and 41 [30-47] mo, respectively; P = 0.532); however, the similarity of overall survival between groups may be related to short follow-up time.

Aspiration is the inhalation of oropharyngeal or gastric contents into the larynx and lower respiratory tract. 13 Asymptomatic microaspiration refers to aspiration of small volumes of oropharyngeal or gastric secretions into the lungs. GER and the resultant microaspiration are associated with lung disease and can contribute to recurrent exacerbations and lung disease progression. 4, 14 In addition, data from LTx literature suggest that chronic microaspiration is associated with CLAD. 3 In fact, several studies have suggested that early fundoplication improves survival and decreases CLAD in LTxRs, presumably through reducing the frequency of microaspiration events. 15, 16 In addition to direct airway and parenchymal damage, GER without aspiration may also produce functional changes in the respiratory tract. 17 Animal and human studies have shown that GER without aspiration (ie, GER affecting the distal esophagus alone) may increase airway resistance and promote airway inflammation by releasing proinflammatory mediators. 18, 19 Our data suggest that GER can also trigger a pneumotoxic SAbs-mediated immune response.

Col-V and Kα1T are sequestered, yet immunogenic, selfantigens restricted to the lungs. 20 SAbs have been detected in both LTx candidates and LTxRs. In a sentinel study by Tiriveedhi et al, 8 pre-and posttransplant serum samples were analyzed from 317 LTxRs with underlying chronic obstructive pulmonary disease (n = 161), IPF (n = 50), cystic fibrosis (n = 55), or other lung diseases (n = 51) who underwent LTx between 2000 and 2011. This study demonstrated that 18% of LTxRs with chronic obstructive pulmonary disease (29 of 161; P = 0.033), 34% with IPF (17 of 50; P = 0.0006), 29% with cystic fibrosis (16 of 55; P = 0.0023), and 19.6% with other lung diseases (10 of 51; P = 0.044) had preexisting SAbs. Furthermore, patients with pre-LTx SAbs had a significantly higher post-LTx incidence of PGD (88% versus 54%, P < 0.05), DSA (70% versus 45%, P < 0.01), and bronchiolitis obliterans syndrome (90% versus 38%, P < 0.001) than patients without pre-LTx SAbs.

We have demonstrated that GER in LTx candidates may result in lung injury and lead to immune exposure to Col-V Step-wise analysis to study confounding association of chronic lung allograft dysfunction with primary graft dysfunction, donor-specific antibodies, and acute cellular rejection and Kα1T antigens, with resultant development of SAbs. This immune response is a potential marker of GER-induced lung injury in LTx candidates. GER and resultant microaspiration episodes may render the pulmonary microenvironment proinflammatory 21, 22 and prone to antigen-antibody reactions, complement cascade activation, and release of proinflammatory cytokines; 7 however, in addition to GER, other factors such as infections and environmental exposures may trigger SAbs. The current study also suggests that pre-LTx GER and SAbs may drive lung allograft injury after LTx leading to CLAD. CLAD is an important cause of mortality in LTxRs and is likely the end result of a variety of immune, infectious, and inflammatory injuries to the allograft. Our study suggests that causes of these injuries include pretransplant GER and SAbs, in addition to other well-known risk factors such as PGD, ACR, and DSA. Importantly, although DSA confounded the relationship between GER, SAbs, and CLAD (Table 4) , an independent relationship between GER, SAbs, and CLAD was still identified. This suggests a mechanistic pathway between pre-LTx GER and SAbs, with SAbs being both a biomarker and propagator of lung injury. In addition, a review of the literature suggests that there may be a mechanistic relationship between SAbs and DSA posttransplant with SAbs driving lung injury leading to immunologic exposure of HLA antigens and the resultant development of DSA. 8, 20 The absence of synergy between GER and SAbs with CLAD may be due to the multifactorial, and potentially transient, nature of SAb elevation and CLAD, as both develop as a result of a variety of injuries. Our study has limitations. This analysis is a single-center, retrospective cohort study, with a relatively small sample size, which makes establishing causation difficult; however, this study does illustrate a correlation between pre-LTx GER and SAbs with adverse post-LTx outcomes that warrants further evaluation in larger, multicenter trials. The results presented demonstrate a potentially important, clinically relevant finding that pre-LTx GER, in conjunction with SAbs to Col-V and Kα1T, predisposes recipients to CLAD after LTx. Although including SAbs in the pretransplant evaluation is feasible, additional validation is warranted before these serological markers impact clinical decision making.

In conclusion, Pre-LTx GER was associated with the presence of SAbs. Furthermore, pre-LTx GER and SAbs were associated with CLAD after LTx; however, the effect was not synergistic, likely due to the multifactorial nature of CLAD.

Chronic lung allograft dysfunction: definition, diagnostic criteria, and approaches to treatment-A consensus report from the Pulmonary Council of the ISHLT

Impact of immediate primary lung allograft dysfunction on bronchiolitis obliterans syndrome

Bile acid aspiration and the development of bronchiolitis obliterans after lung transplantation

Gastro-oesophageal reflux and aspiration in patients with advanced lung disease

Pre-transplant impedance measures of reflux are associated with early allograft injury after lung transplantation

Pretransplant gastroesophageal reflux compromises early outcomes after lung transplantation

Antibodies to self-antigens predispose to primary lung allograft dysfunction and chronic rejection

Pre-transplant antibodies to Kα1 tubulin and collagen-V in lung transplantation: clinical correlations

A consensus document for the selection of lung transplant candidates: 2014-an update from the Pulmonary Transplantation Council of the International Society for Heart and Lung Transplantation

Report of the ISHLT Working Group on primary lung graft dysfunction, part I: definition and grading-A

Consensus Group statement of the International Society for Heart and Lung Transplantation

Utilization of wireless pH monitoring technologies: a summary of the proceedings from the esophageal diagnostic working group

Prevalence of antibodies to lung self-antigens (Kα1 tubulin and collagen V) and donor specific antibodies to HLA in lung transplant recipients and implications for lung transplant outcomes: single center experience

Aspiration pneumonitis and aspiration pneumonia

Extraesophageal manifestations of GERD

Maxwell Chamberlain Memorial Paper. Early fundoplication prevents chronic allograft dysfunction in patients with gastroesophageal reflux disease

Improved lung allograft function after fundoplication in patients with gastroesophageal reflux disease undergoing lung transplantation

Gastro-oesophageal reflux disease and non-asthma lung disease

Possible mechanisms of influence of esophageal acid on airway hyperresponsiveness

Comparison of airway responses following tracheal or esophageal acidification in the cat

An important role for autoimmunity in the immunopathogenesis of chronic allograft rejection

Aspiration, localized pulmonary inflammation, and predictors of early-onset bronchiolitis obliterans syndrome after lung transplantation

Exposure of airway epithelium to bile acids associated with gastroesophageal reflux symptoms: a relation to transforming growth factor-beta1 production and fibroblast proliferation

The authors thank Kristine Nally for her editorial assistance and Marco Marchionni for illustrations.