Duke University Medical Center The Role of Surfactant Protein A & D in Lung Allograft Rejection Matthew Schechter, BSc, C. Ashley Finlen Copeland, MSW, Francine Kelly, BS, Scott M. Palmer, MD, MHS. Division of Pulmonary and Critical Care Medicine - Duke University Medical Center, Durham, NC ABSTRACT HYPOTHESES Figure 1: Recipient T11208C genotype (SP-D 3’ UTR) affects Figure 4: SP-D levels in transplanted lung influenced by the recipient’s Background Lung transplantation is characterized by high rates of rejection, especially compared to other types of solid organ transplants. Our lab has progression to BOS > 5 yrs. post transplant genotype developed the hypothesis that the lung’s innate immune system, unique among transplanted organs, influences the outcome of lung transplantation. The Donor single nucleotide polymorphisms (SNPs) within the SP-A and SP-D pulmonary collectins, surfactant protein (SP)-A and SP-D, regulate the innate host defense of the lung and may provide a link between the innate and acquired immune systems. Although these proteins have been implicated in numerous lung diseases, their role in lung transplantation remains unstudied. genes explain the interindividual variation in the development of PGD p = .0622 p = .0322 p = .0705 p = .0731 We hypothesized that single nucleotide polymorphisms (SNPs) in the SP-A and SP-D genes influence the development of early and late graft dysfunction. SNPs within the recipient SP-A and SP-D genes explain the interindividual 900 900 Mean SP-D Concentration Therefore, we determined the effect of SP-A and SP-D polymorphisms in the donor on the development of primary graft dysfunction (PGD) and the effect of Mean SP-D Concentration variation in the development of BOS 800 ANOVA: p = .0813 these polymorphisms in the recipient on the development of bronchiolitis obliterans syndrome (BOS), the major cause of late death after lung 800 ANOVA: transplantation. We also theorized that SP-D levels are associated with allograft function, and therefore followed the trends of SP-D levels in a group of SP-D levels in the bronchioalveolar lavage fluid are associated with chronic 700 700 p = .0931 patients that eventually developed BOS and a matched BOS-free cohort. allograft function. 600 600 (ng/mL) (ng/mL) Methods A cohort of lung transplant recipients was established using recipients and donors from Duke University Medical Center. Patients were screened 500 500 for four SP-D polymorphisms and three polymorphisms from each of the SP-A genes (SP-A1 and SP-A2). For the SP-D polymorphisms, commercially 400 400 available TaqMan assays were used. Given the homology between the two SP-A genes, gene-specific primers were first used to amplify the region 300 300 containing the polymorphism of interest, followed by allelic determination using custom-designed TaqMan assays. We examined the association between donor SP-A & SP-D genotypes and the presence or absence of severe (Grade 3) PGD using a chi-square test and compared the PaO2/FiO2 ratio 24 hours post-transplant, based on donor genotype, using an ANOVA. A Kaplan-Meier log rank test was used to determine the effect of recipient SP-A and SP-D RESULTS 200 100 0 200 100 polymorphisms on the onset of BOS and post-transplant graft survival. A commercially-available ELISA for SP-D was performed on bronchoalveolar lavage 0 TT TC CC TT TC CC fluid (BALF) collected from transplant recipients during surveillance bronchoscopies. We examined the relationship between recipient SP-D genotypes and TABLE 1: SP-A1, SP-A2 & SP-D POLYMORPHISMS ANALYZED IN THE LUNG mean SP-D levels over the course of the transplant using ANOVA & Student’s t-tests. TRANSPLANATION POPULATION rs1923537 (3' UTR) rs721917 (Met11Thr) Results A total of 407 lung transplant recipients and 212 donors were genotyped for the 10 surfactant SNPs. All SNPs had a no-call rate <5%. Relevant Minor 900 data was entered into database in order to establish well-defined clinical phenotypes of the early and late graft function in all transplant patients. 900 Mean SP-D Concentration Mean SP-D Concentration Homozygosity for the T allele within the 3’ UTR of SP-D was associated with decreased rates of both severe PGD and late (> 5yrs) BOS. The genotype was Absolute Genetic Amino Acid Nucleotide allele 1.00 800 p = .0357 800 p = .0369 Survival Distribution Function also associated with increased SP-D levels in the BALF. Lung transplant recipients with a Thr/Thr genotype at αα11 had significantly decreased SP-D levels Gene position RS number region substitution substitution frequency 700 700 as well as increased rates of severe PGD In a cross-sectional analysis of BALF, we found that BAL SP-D levels were significantly decreased in BOS SP-A1 chr10: rs1059047 Exon 1 T/C αα 19: Val → 6.5% 600 600 (ng/mL) (ng/mL) patients, with higher BOS grade corresponding to lower of SP-D levels. 74656728 Ala 0.75 500 500 Conclusions These studies indicate that polymorphic forms of SP-D might modulate the immune system’s response to lung transplantation and support our SP-A1 chr10: rs1136450 Exon 1 G/C αα 50: Val → 22.7% 400 400 overall hypothesis for a role of pulmonary innate immunity in modulating lung allograft outcomes. A decrease in SP-D, due to either genetic or environmental 74656820 Leu 300 300 factors, is associated with chronic graft dysfunction. Furthermore, the 3’ UTR variant associated with increased SP-D levels seems to be protective against 0.50 severe PGD and late progression to BOS. The pulmonary collectins provide a potentially novel target for therapies designed to help prevent rejection and SP-A1 chr10: rs4253527 Exon 4 C/T αα 219: Arg → 2.2% 200 200 improve outcomes following lung transplantation. 74658868 Trp 100 100 SP-A2 chr10: rs1059046 Exon 1 A/C αα 9: Asn → 17.4% 0.25 0 0 TT C* T* CC 75314238 Thr rs1923537 (3' UTR) rs721917 (Met11Thr) INTRODUCTION SP-A2 chr10: 75314238 rs17886395 Exon 2 G/C αα 91: Ala → Pro 10.9% 0.00 Lung transplant is a viable option for patients with end-stage lung disease. Although immediate post-transplant survival has improved to over 80% at 1 year, the 5-year survival rate remains poor at 50%, well behind other solid organ transplants. 1 SP-A2 chr10: 75312069 rs1965708 Exon 4 C/A αα 223: Gln → Lys 11.4% 0 500 1000 1500 2000 bos23freedays 2500 3000 3500 4000 4500 RESULTS/CONCLUSIONS SP-D chr10: rs1923537 3’ UTR T/C n/a 35.4% STRATA: DUTR=CC Censored DUTR=CC Primary Graft Dysfunction (PGD) is an important early complication of lung allograft transplantation. Severe (Grade 2 & 3) 75537958 (T11208C) DUTR=TC DUTR=TT Censored DUTR=TC Censored DUTR=TT Of all the polymorphisms studied, only two SP-D SNPs (3’UTR & Met11Thr) PGD occurs in ~12% of transplants, and as many as half of all lung transplant recipients may experience a mild form of this SP-D chr10: rs721917 Exon 2 T/C αα 11: Met → 41.7% seemed to influence either early or late lung allograft function disease. PGD has a 30-day mortality rate of over 60%.2 PGD is believed to be due to ischemia/reperfusion injury, whereby 75549076 Thr activation of neutrophils leads to an accumulation of reactive oxygen species that causes acute lung injury (similar to ARDS).3 SP-D chr10: rs2243639 Exon 5 C/T αα 160: Ala → 43.8% The genotype associated with increased SP-D levels in the BAL (TT 3’UTR) was 75692660 Thr The long-term survival of lung transplants is limited by bronchiolitis obliterans (BO), characterized by fibrotic scarring and also associated with decreased rates of severe PGD & progression to BOS > 5 destruction of the bronchioles by a dense, lymphocytic infiltration. BO manifests clinically as progressive airflow obstruction SP-D chr10: rs3088308 Exon 8 A/T αα 270: Thr → 6.2% Figure 2: Recipient SP-D genotypes are associated with the years post-transplant 75688804 Ser termed bronchiolitis obliterans syndrome (BOS). Nearly half of all lung transplant patients develop BOS within 5 years of development of severe (Grade 3) PGD 24 hours post-transplant transplantation. BO is thought to the result of chronic allograft injury, whether caused by alloimmune responses (acute TABLE 2: BOS GENETIC ASSOCIATION COHORT NO (%) p = .212 The genotype associated with decreased SP-D levels (Thr/Thr at αα 11) is rejection, lymphocytic bronchiolitis) or alloimmune-independent processes (infection, gastroesophageal reflux, PGD).4 (N= 356) 100 PGD 0-2 300 associated with increased rates of severe PGD PGD Grade @ 24 hrs Male 206 (57.9%) PGD 3 These results suggest that SP-D levels in the BALF of lung transplant patients PaO2/FiO2 ratio 75 Any incidence of PGD is associated with poorer overall transplant function and increased long-term complications, including Female 150 (42.1%) 200 development of BOS.2,3 Median Age at Transplant 56 (42, 62) 50 p = .028 may be influenced by the recipient’s genotype. Whether the differences in SP-D Although the exact mechanism of neither PGD nor BOS/BO is completely understood, increasing evidence indicates that the Native Disease levels directly contribute to the development of either PGD or late BOS requires pulmonary innate immune system contributes to the development and progression of both conditions. Obstructive 157 (44.1%) 25 100 further investigation. Pulmonary Vascular 5 (1.4%) Surfactant protein (SP)-A and SP-D, members of the collectin family of soluble pattern recognition receptors, regulate the Cystic Restrictive 69 (19.4%) 125 (35.1%) 0 0 This study represents the largest analysis of donor and recipient genotypes upon lung transplantation outcomes; however further validation of these results in TC TT CC pulmonary innate immune response & are known to affect both acute and chronic inflammation. 5 TT TC CC Bilateral Transplant 330 (92.7%) SP-D 3' UTR SP-A & SP-D are synthesized and secreted by type II pneumocytes. These cells are damaged by chronic inflammation, and Single Transplant 26 (7.3%) SP-D 3' UTR independent multicenter cohorts is needed. p = .151 evidence suggests that over time, recipient-derived cells repopulate the damaged lung epithelium, meaning that both donor Patients with Acute Rejection (AR) 282 (79.2%) 100 p = .052 Given 350 Median Number of Episodes/Patient 2 (1, 3) PGD 0-2 the diverse functions of SP-A & SP-D in pulmonary immunity, future PGD Grade @ 24 hrs. and recipient genotypes may influence protein function. 6,7 PGD 3 300 Median Cumulative Rejection Grade/Patient 3 (2, 6) studies should examine manipulation of these innate molecules in the prevention PaO2/FiO2 Ratio 75 250 p = .15 Patients with CMV Pneumonia 82 (23.0%) 200 or treatment of PGD and BOS 50 Patients with Nissen within 6 months of transplant 94 (26.4%) METHODS Median Number of PFT’s per Patient 28 (19, 36) 25 150 100 BOS Genetic Association (GA) Cohort: Median Number of Bronchoscopies/Patient Prior to BOS 8 (5,12) 50 Genomic DNA was isolated from 356 patients from a total of 424 lung transplant recipients who met the study inclusion criteria: adult Caucasian patients undergoing first lung transplant at Duke University Medical Center between January 1, 1998 and June 25, 2008 who were BOS eligible (e.g. survived longer than 180 days and had pulmonary function data from which to assess BOS). Patients undergoing heart/lung Time from Transplant to BOS (Q1, 66th percentile) 5.82 years (3.15, 8.79) 0 0 TT TC CC TT TC CC FUNDING: transplantation were excluded from the study. BOS was defined according to standard ISHLT criteria based on serial Forced Expiratory Volume (FEV1) measurements.8 Censor date was July 1, 2009 to ensure every patient had at least one year of follow-up. Demographic and post-transplant characteristics of the BOS genetic association cohort are summarized in Table 2. Overall Survival (Q1, 60th percentile) 8.47 years (4.20, 9.02) SP-D Met11Thr SP-D Met11Thr SCCOR grant: NIH / NHLBI 1P50-HL084917-011, Project 3 (to Palmer) PGD GA Cohort: K24 HL091140 (Palmer) Genomic DNA was isolated from 263 cadaveric donors transplanted at Duke University Medical Center between November 1, 1998 and January 1, 2005. We only included the 188 Caucasian donors in our study. Stead Scholarship, 2007-2008 (to Schechter) PGD-T24 was defined according to the ISHLT guidelines9 using the PaO2 and FiO2 values from the arterial blood gas closest to 24 hrs. post-transplant. PGD grades 0 and 1 were combined into a single grade (0/1) for analysis given the lack of chest X-rays 24 hours post-transplant. Demographic and post-transplant characteristics of the PGD genetic association cohort are summarized in table 3. TABLE 3: PGD GENETIC NO. (%) TABLE 4: SP-D CROSS- NO. (%) Determination of SP-A & SP-D Polymorphisms: The surfactant protein SNPs studied are outlined in Table 1. SNPs were chosen for inclusion based upon the following criteria: common (generally >5% MAF) non-synonomous SNPs with a probable functional effect. ASSOCIATION COHORT (N=188) SECTIONAL COHORT (N=38) Figure 3: SP-D levels decrease with increasing BOS grade REFERENCES: Probability of functional effect was determined by a review of the literature to identify SNPs that had been previously associated with other pulmonary diseases and/or were located within a functionally-important region (1) Zhang P, Summer WR, Bagby GJ et al. Innate immunity and pulmonary host defense. Immunol Rev 2000; of the protein. For the SP-D SNPs, commercially available TaqMan assays (Applied Biosystems, Foster City, CA) were used for genotyping. Given the homology between the two SP-A genes (SFTPA1 & SFTPA2), Male 91 (48.4%) Male 25 (66%) 183:310-51 gene-specific primers were designed to amplify a 600-800bp region containing the SNP of interest, followed by a custom-designed TaqMan assay (Applied Biosystems, Foster City, CA) in order to determine the * genotype of each SNP. The TaqMan assay results were performed at least in duplicate, with 100% reproducibility of results and a no-call rate of <5% for each SP-S SNP. Female 97 (51.6%) Female 13 (34%) ** (2) Christie JD, Sager JS, Kimmel SE et al. Impact of primary graft failure on outcomes following lung SP-D Concentration (ng/mL) SP-D Cross-sectional Analysis Cohort: Median Age at Transplant 54 (40, 61) Median Age at Transplant 56 (38, 61) 2000 ‡ 2000 transplantation. Chest. 2005; 127:161-5. From October 1, 2007 to October 29, 2008, we obtained bronchoalveolar lavage fluid (BALF) from 112 patients who had received a lung transplant at Duke University Medical Center (underwent transplant from 1998- 2008). Of those patients, 58 patients met the following inclusion criteria: patient received a cadaveric transplant, survived l> 6 months post-transplant and had no concurrent rejection or infection at the time of BAL Native Disease Native Disease (3) Carter YM, Gelman AE, Kreisel D. Pathogenesis, management, and consequences of primary graft p = .0067 SP-D Concentration sample. BOS grades (0-p to 3) were assigned based upon the 2001 updated classification for BOS. 8 112 Duke lung transplant recipients (underwent transplant from 1998-2008) were screened for possible inclusion. Obstructive 98 (52.1%) Obstructive 15 (39.5%) dysfunction. Sem Thorac Cardiovas Surg 2008; 20:165-72. 21 patients had BOS 0-p at the time of BAL collection, and were therefore excluded from the study, leaving 38 patients who compromise the analysis cohort (Table 3). Bronchoalveolar lavage fluid (BALF) was * p = .0102 collected at the time of bronchoscopy, processed and stored at -80ºC until analysis. SP-D levels in the BAL fluid using a commercially-available ELISA (Biovendor, Candler, NC) according to the manufacturers Pulmonary Vascular 13 (6.9%) Pulmonary Vascular 4 (10.5%) 1500 1500 (4) Sato M, Keshavjee S. Bronchiolitis obliterans syndrome: alloimmune-dependent and -independent injury instructions. ** p = .0469 Cystic 32 (17.0%) Cystic 9 (23.7%) with aberrant tissue remodeling. Sem Thorac Cardiovas Surg 2008; 20:173-82. (ng/mL) SP-D Longitudinal Analysis Cohort: In order to follow SP-D levels during the post-transplant period, we identified 23 patients who eventually developed BOS for whom we had at least four BALF samples. For 14 patients, we had at least two samples Restrictive 45 (23.9%) Restrictive 10 (26.3%) ‡ p = .0568 (5) Pastva AM, Wright JR, Williams KL. Immunomodulary roles of surfactant proteins A and D: Implications in from before the development of BOS (pre-BOS) and two samples from after the development of BOS (post-BOS). Only one pre-BOS sample was available for three of these patients, while two patients had no pre- 1000 1000 lung disease. Proc Am Thorac Soc 2008; 4:252-8. BOS samples. The remaining three patients had multiple pre-BOS samples, but only one post-BOS BALF sample. The average number of samples per patients was 5.39. We then identified a group 20 patients who Bilateral Transplant 163 (86.7%) Bilateral Transplant 37 (97%) remained free of BOS that matched the BOS patients in terms of average number of BALF samples available (5.10 samples per patients; p = .55 vs BOS patient population) and average length of follow-up (1292.8 Single Transplant 25 (13.3%) Single Transplant 1 (3%) (6) Kleeberger W, Versmold A, Rothamel T et al. Increased chimerism of bronchial & alveolar epithelium in days vs. 1463.0 days for BOS patients; p = .33). BOS grades (0-p to 3) were assigned based upon the 2001 updated classification for BOS.6 Bronchoalveolar lavage fluid (BALF) was collected at the time of bronchoscopy, processed and stored at -80ºC until analysis. SP-D levels in the BAL fluid using a commercially-available ELISA (Biovendor, Candler, NC) according to the manufacturers instructions. human lung allografts undergoing chronic injury. Amer J Path 2003; 162:1487-94. PGD-T24 grade Average time to BOS Onset 1494 days 500 500 Analysis: PGD 0/1 58 (30.9%) (7) Spencer H, Rampling D, Aurora P et al. Transbronchial biopsies provide longitudinal evidence for epithelial In the BOS genetic association cohort, the association of each SNP with survival, time to BOS1 and time to BOS 2/3 using a Kaplan-Meier log rank test. For the PGD cohort, the PaO2/FiO2 ratio at 24h was considered as a continuous variable, and the relationship between both donor and recipient genotype was assessed using an ANOVA. Results are expressed as mean ± standard error (SEM). Comparisons between two genotype PGD 2 52 (27.6%) BOS Grade at Bronch Date chimerism in children following sex mismatched lung transplantation. Thorax 2005; 60:60-62. groups utilized a Student’s t-test. The presence or absence of severe (Grade 3) PGD at 24h was also considered as an ordinal value, whereby a likelihood ratio was used. In both cohorts, each SNP was considered in PGD 3 46 (24.5%) No BOS 18 (47%) 0 0 (8) Christie JD, Carby M, Bag R et al. Report of the ISHLT Working Group on Primary Lung Graft Dysfunction a dose-response and recessive inheritance model. A p-value < .05 was considered statistically significant. Unknown 32 (17.0%) BOS 1 10 (26%) No BOS BOS 1 BOS 2 BOS 3 No BOS/BOS 1 BOS 2/3 part II: definition. J Heart Lung Transplant 2005; 24:1454-9. For the cross-sectional analysis of SP-D levels, the relationship between BOS grade and SP-D concentration was assessed using an ANOVA. Comparisons between two groups, as well as the comparison between BOS 2 6 (16%) severe (Grade 2/3) BOS and No BOS/BOS 1, were completed using a Student’s t-test. All graphs show average SP-D levels by BOS Grade ± SEM. Association between recipient genotype and average SP-D levels Median PaO2/FiO2 ratio 270.0 (9) Estenne M, Maurer JR, Boehler A et al. Bronchiolitis obliterans syndrome 2001: an update of the diagnostic over the post-transplant period were assessed by ANOVA. The means of two genotypes were compared using a Student’s t-test. All graphs show average SP-D levels by genotype ± SEM. Each SNP was considered BOS 3 4 (11%) BOS Status BOS Status in a dose-response and recessive inheritance model. A p-value < .05 was considered statistically significant. (24 hours post-transplant) (187.5, 345.2) criteria. J Heart Lung Transplant 2002; 21:297-310.