Telomere Length of Pulmonary Fibrosis Patients Associated with Chronic Lung Allograft Dysfunction and Post-Lung Transplantation Survival

Abstract

Background

Prior studies have shown that patients with pulmonary fibrosis with mutations in the telomerase genes have a high rate of certain complications after lung transplantation. However, few studies have investigated clinical outcomes by leukocyte telomere length.

Methods

We conducted an observational cohort study of all pulmonary fibrosis patients who underwent lung transplantation at a single center between January 1, 2007 and December 31, 2014. Leukocyte telomere length was measured from a sample of blood collected prior to lung transplantation and subjects were stratified into two groups (telomere length <10^th versus ≥10^th percentile). The primary outcome was post-lung transplant survival. Secondary outcomes included incidence of allograft dysfunction, non-pulmonary organ dysfunction and infection.

Results

Approximately one-third (32%) of subjects had a telomere length below the 10^th percentile. Telomere length <10^th percentile was independently associated with worse survival (HR 10.9, 95% CI 2.7–44.8, p=0.001). Telomere length <10^th percentile was also independently associated with a shorter time to the onset of chronic lung allograft dysfunction (CLAD) (HR 6.3, 95% CI 2.0–20.0, p=0.002). Grade 3 primary graft dysfunction occurred more frequently in the <10^th percentile group compared to the ≥10^th percentile group (28% vs 7%, p=0.034). There was no difference in the incidence of acute cellular rejection, cytopenias, infection or renal dysfunction in the two groups.

Conclusions

Telomere length <10^th percentile was associated with worse survival and shorter time to onset of CLAD, and thus represents a biomarker that may aid in the risk stratification of pulmonary fibrosis patients prior to lung transplantation.

Introduction

Since implementation of the current lung allocation score system, pulmonary fibrosis has become the leading indication for lung transplantation(1, 2). Patients with idiopathic pulmonary fibrosis (IPF) account for the largest proportion of patients awaiting, and dying while awaiting, lung transplantation. Despite rigorous pre-transplant evaluation and selection, post-lung transplant median survival is 5.7 years for all recipients but only 4.7 years for those with pulmonary fibrosis(2). One of the main limitations to survival after lung transplantation is chronic lung allograft dysfunction (CLAD). Treatment of CLAD remains a challenge since its underlying pathogenesis is not well understood. Immunosuppressive medications are used to prevent and treat rejection, but these medications are commonly associated with many side effects.

Telomeres consist of nucleotide repeats (TTAGGG) located on the end of chromosomes that serve to protect these ends during cell replication. Telomeres normally shorten with age, but excessive shortening can lead to activation of DNA damage-signaling pathways resulting in cellular senescence(3). Pathogenic rare variants in several different genes in the telomere pathway (TERT, TERC, PARN, RTEL1, NAF1) are found in patients with familial pulmonary fibrosis(4–7). Heterozygous mutations in these genes are associated with short telomere lengths and a rapidly-progressive form of pulmonary fibrosis, which is most commonly characterized as idiopathic pulmonary fibrosis (IPF)(8). Small observational studies have found that pulmonary fibrosis patients with TERT or TERC mutations have high rates of bone marrow failure, infection, renal dysfunction, and allograft dysfunction after lung transplant(9–11). However, these studies were limited by small numbers of patients and the absence of a comparison cohort.

Although pathogenic mutations are rare, short telomere lengths are relatively common in patients with pulmonary fibrosis. Age-adjusted telomere lengths <10^th percentile of normal are found in ~40% of patients with familial pulmonary fibrosis and in ~25% without a family history of lung fibrosis(12, 13). As the side effects of immunosuppression medications overlap the broad clinical spectrum of short telomere syndromes, it can be difficult to identify the clinical phenotypes that are specifically related to intrinsic patient characteristics. In this study, we sought to characterize clinical outcomes associated with age-adjusted telomere length in pulmonary fibrosis patients who underwent lung transplantation. We hypothesized that short telomere lengths would be associated with shorter post-transplant survival times and higher rates of allograft dysfunction, non-pulmonary organ dysfunction, and infection.

Methods

This is a prospective observational cohort study that included patients from the University of Texas Southwestern Medical Center (UTSW; Dallas, TX). All patients provided written informed consent and provided a sample of blood upon enrollment. Patients were recruited without regard to family history or any clinical manifestation of a short telomere syndrome. Each patient underwent lung transplantation between January 1, 2007 and December 31, 2014. Patients were excluded if they did not have a pre-transplant diagnosis of pulmonary fibrosis, were transplanted elsewhere or were enrolled after transplantation.

Clinical information was retrospectively extracted from the electronic medical record. All patients where maintained on a three drug immunosuppression regimen including a calcineurin or mTOR inhibitor (cyclosporine, tacrolimus, or sirolimus), an anti-metabolite (azathioprine or mycophenolate mofetil), and a corticosteroid (prednisone). Protocol-driven patient assessments included serial laboratory tests, pulmonary function tests and surveillance bronchoscopies.

Clinical Variable Definitions

Survival time was calculated from date of transplant to death or censor date (September 30, 2015). Cause of death was adjudicated by transplant pulmonologists (V.K., F.T.). Presence and severity of primary graft dysfunction (PGD) was determined by degree of hypoxemia and by chest radiography at 48 and 72 hours(14). Grade 3 PGD was defined as a PaO₂/FiO₂ ratio of <200 mmHg and pulmonary infiltrates on chest x-ray(15). PGD could not be assessed in 18 patients because of missing FiO₂ data (8 in the <10^th percentile group; 10 in the ≥10^th percentile group). Acute cellular rejection (ACR) was determined by histopathologic evaluation of transbronchial biopsy specimens(16). The ACR score represents the sum of the histopathologic “A” scores divided by the number of biopsies(17). Clinical rejection was defined as an acute deterioration in allograft function as evidenced by spirometric decline or worsened chest imaging; supporting histopathologic evidence was not required. CLAD was defined as a persistent decline in forced expiratory volume in 1 second (FEV₁) <80% of baseline (average of two best FEV₁ values after transplantation), not due to infection or ACR(18). Time to CLAD onset was calculated from the date of transplant. CLAD was further subdivided into bronchiolitis obliterans syndrome (BOS) and restrictive-CLAD (R-CLAD), which were defined by a forced vital capacity (FVC) of ≥80% or <80% baseline FVC, respectively, at the time of CLAD onset(19).

Infection was defined as the presence of pathogens isolated from sterile sites or presence of virus from nasal or respiratory specimens. Cytopenias were defined as leukopenia (white blood cell count <4.0 K/ul), anemia (hemoglobin <12.0 g/dl for women and <12.4 g/dl for men), thrombocytopenia (platelet count <150 K/ul) and macrocytosis (mean corpuscular volume >98 fL). Number of transfusions was tabulated, excluding those required within 30 days of transplantation. Acute renal failure was defined as an increase in serum creatinine to ≥1.5 times baseline within a 7 day period; chronic renal failure was defined as reduction in glomerular filtration rate to <60 ml/min/1.73 m² for ≥3 months. Elevated liver function tests (LFT) were defined as aspartate transaminase ≥100 U/L, alanine transaminase ≥100 U/L or alkaline phosphatase ≥150 U/L. Cirrhosis was determined by imaging or liver biopsy. Presence of malignancy was based on pathologic specimens. Venous thromboembolism (VTE) was defined as presence of either pulmonary embolism or deep venous thrombosis. Pulmonary embolism (PE) was diagnosed based on visualization by CT angiogram or by high-probability V/Q scan. Deep venous thrombosis (DVT) was diagnosed via ultrasonography. The total number of immunosuppression and antibiotic prophylaxis drugs was tabulated.

Telomere Length Measurement

Telomere length was measured using quantitative PCR from genomic DNA isolated from peripheral blood leukocytes using an Autopure LS (Qiagen, Valencia, CA)(12, 20, 21). Telomere length was represented as a logarithm-transformed relative ratio of telomere to single copy gene (T/S); age-adjusted telomere length was calculated using normal controls and represented as either an observed minus an expected (O-E) value or the percentile for a given age. Transplant physicians clinically managing patients were blinded to the leukocyte telomere length result. Sequencing of telomere-related genes was not systematically performed. Five subjects with telomere-related gene mutations (all with telomere lengths <10^th percentile) were included in other studies(8, 11).

Statistical Analysis

Patient characteristics were compared using Fisher’s exact test (categorical variables), T-test (normally distributed continuous variables), or Mann-Whitney-Wilcoxon test (non-normally distributed continuous variables). Median survival time and time to onset of CLAD were calculated from Kaplan-Meier curves; groups were compared using the log rank test. Cox proportional hazards models were used in univariable and multivariable analysis to assess the relative effects of clinically relevant covariates. Survival analysis included Grade 3 PGD, ACR score, CLAD, total infection rate and telomere length (<10^th percentile) as covariates. Time to onset of CLAD analysis included Grade 3 PGD, ACR score, pulmonary infection rate and telomere length (<10^th percentile) as covariates. In sensitivity analyses, telomere length (O-E) was treated as a continuous variable. We noted no evidence of non-proportional hazards in the Cox models. All analyses were performed using R version 3.2.2 software (www.R-project.org).

Results

Of 319 patients who underwent lung transplantation at UTSW between January 1, 2007 and December 31, 2014, 136 were consented for study (Supplementary Figure). Fifty four were excluded because of a diagnosis other than pulmonary fibrosis or because consent was obtained after transplantation. A total of 82 subjects were included and divided into two groups: 26 subjects (32%) with leukocyte telomere length <10^th percentile and 56 subjects (68%) with telomere lengths ≥10^th percentile. More patients in the telomere length <10^th percentile group had a clinical diagnosis of IPF and had a lung explant histopathologic diagnosis of usual interstitial pneumonia as compared to the ≥10^th percentile group.

There was no difference between the two groups with regard to patient demographics or severity of lung disease (Table 1). More individuals in the <10^th percentile group had a baseline macrocytosis as compared those in the ≥10^th percentile group (10 (39%) vs 7 (12%), p=0.017), but there was no difference in rate of leukopenia, anemia, or thrombocytopenia. The majority of patients in both groups underwent double-lung transplantation.

Table 1.

Pre-Transplant Clinical Characteristics

	Total N=82	<10th Percentile, N=26	≥10th Percentile, N=56	P-value
Age at Transplant, mean±SD, yr	59±9	60±6	58±10	0.51
Male, N (%)	57 (69.5)	21 (80.8)	36 (64.3)	0.197
Ethnicity, N (%)
White	71 (86.6)	26 (100)	45 (80.4)	0.15
Black	5 (6.1)	0	5 (8.9)
Hispanic	3 (3.7)	0	3 (5.4)
Asian	3 (3.7)	0	3 (5.4)
Family History of Pulmonary Fibrosis, N (%)	22 (27%)	8 (31%)	14 (25%)	0.60
Pulmonary Fibrosis Diagnosis, N (%)
IPF	50 (61)	22 (85)	28 (50)	0.003
CTD-ILD	10 (12)	3 (11)	7 (12)	1.0
Chronic HP	8 (10)	1 (4)	7 (12)	0.42
Non-UIP IIP	6 (7)	0	6 (11)	0.17
Unclassifiable fibrosis	5 (6)	0	5 (9)	0.17
Othera	3 (4)	0	3 (5)	0.55
Explant Pathologic Pattern, N (%)
UIP	63 (76.8)	26 (100)	37 (66.1)	<0.001
Non-UIP	19 (23.2)	0	19 (33.9)
Former Smoker, N (%)	42 (51.2)	14 (53.8)	28 (50.0)	0.81
Pack-Yr, median (IQR)	16 (10 – 36)	15 (10 – 48)	16 (9 – 31)	0.77
PFT Prior to Transplant, mean±SD (N)b
FVC absolute	1.90±0.75 (76)	1.93±0.58 (26)	1.88±0.83 (50)	0.46
FVC % predicted	44.6±14.9 (76)	43.5±13.4 (26)	45.2±15.7 (50)	0.81
DLCO absolute	5.8±2.5 (55)	5.4±2.4 (16)	5.9±2.5 (39)	0.46
DLCO % predicted	21.8±10.9 (55)	18.8±10.9 (16)	23.0±10.8 (39)	0.20
Bone Marrow Dysfunction, N (%)
Leukopeniac	5 (6.1)	1 (3.8)	4 (7.1)	1
Anemiad	24 (29.3)	7 (26.9)	17 (30.4)	0.80
Macrocytosise	17 (20.7)	10 (38.5)	7 (12.5)	0.017
Thrombocytopeniaf	7 (8.5)	2 (7.7)	5 (8.9)	1
Number of Lungs, N (%)
Single	12 (14.6)	3 (11.5)	9 (16.1)	0.74
Double	70 (85.4)	23 (88.5)	47 (83.9)

^aOther diagnsoses include Combined pulmonary fibrosis and emphysema (n=1), Hermansky-Pudlak syndrome (n=1) and Sarcoidosis (n=1);

^bPFT within one year of transplant;

^cLeukopenia defined as WBC <4 k/ul;

^dAnemia defined as hemoglobin level <12.0 g/dl for women and <12,4 g/dl for men;

^eMacrocytosis defined as mean corpuscular volume >98 fL;

^fThrombocytopenia defined as platelet count <150 K/ul.

Abbreviations: IPF, Idiopathic Pulmonary Fibrosis; CTD-ILD, Connective Tissue Disease associated Interstitial Lung Disease; HP, Hypersensitivity Pneumonitis; Non-UIP IIP, Non-Usual Interstitial Pneumonia Idiopathic Interstitial Pneumonia;

Survival

Patients were followed for a mean duration of five years after lung transplantation. There were 14 (54%) deaths in the telomere length <10^th percentile group compared to 10 (18%) deaths in the telomere length ≥10^th percentile group during the follow up period (p=0.0015) (Table 2, Figure 1). The overall survival time was significantly different between the two groups (log rank test p=0.019). In univariable analysis, telomere length <10^th percentile and total number of infections per year were associated with decreased survival time (HR 2.6, 95% CI 1.1–6.0, p=0.023 for telomere length) (Table 3). In multivariable analysis telomere length <10^th percentile was associated with worse survival (HR 10.9 (2.7–44.8), p=0.001), independent of other clinically relevant risk factors including Grade 3 PGD, ACR score, CLAD or infection rate. Similar results were found after excluding single lung transplant recipients (n=56, telomere length <10^th percentile HR 17.1, 95% CI 3.2–91.3, p=0.001, data not shown). Telomere length as a continuous variable was significantly associated with survival in univariable and multivariable analysis (HR 0.01 (95% CI 0–0.25) per unit change in telomere length, p=0.005, Supplemental Table 1). There was no single cause of death that was overrepresented in the group of patients with short telomere lengths. Causes of death included CLAD, infection, malignancy, cardiovascular complications and others (Supplemental Table 2).

Table 2.

Post-Transplant Clinical Characteristics

	Total N=82	<10th Percentile, N=26	≥10th Percentile, N=56	P-value
Follow up Time, mean±SD, yr	5.0±2.5	5.1±2.7	5.0±2.5	0.85
Death, N (%)	24 (29.3)	14 (53.8)	10 (17.9)	0.0015
Survival, median (95% CI)	7.1 (7.0-)	6.2 (2.3-)	-	0.019
Primary Graft Dysfunction	N=64	N=18	N=46
Grade 3 PGD, N (%)	8 (12.5)	5 (27.8)	3 (6.5)	0.034
Acute Cellular Rejection	N=82	N=26	N=56
ACR Score, median (IQR)a	0.33 (0–0.50)	0.27 (0–0.40)	0.37 (0–0.57)	0.099
Chronic Lung Allograft Dysfunction	N=82	N=26	N=56
Time to Onset of CLAD, median (95% CI)	5.3 (3.8-)	2.7(1.5-)	-	0.0054
CLAD Present, N (%)	26 (31.7)	13 (50.0)	13 (23.2)	0.022
BOS Present, N (%)	16 (19.5)	7 (26.9)	9 (16.1)	0.37
R-CLAD Present, N (%)	10 (12.2)	6 (23.1)	4 (7.1)	0.066
Bone Marrow Dysfunctionb	N=74	N=23	N=51
Leukopenia, N (%)c	20 (27.0)	7 (30.4)	13 (25.5)	0.78
Anemia, N (%)d	42 (56.8)	14 (60.9)	28 (54.9)	0.80
Macrocytosis, N (%)e	41 (55.4)	16 (69.6)	25 (49.0)	0.13
Thrombocytopenia, N (%)f	15 (20.3)	4 (17.4)	11 (21.6)	0.76
Infectious Complications	N=82	N=26	N=56
Total Infections per Yr, median (IQR)	0.96 (0.22–2.29)	1.04 (0.44–1.87)	0.90 (0.15–2.40)	0.72
Pulmonary Infections per Yr, median (IQR)	0.61 (0.03–1.84)	0.61 (0.25–1.28)	0.60 (0–1.86)	0.95
Non-Pulm. Complications	N=82	N=26	N=56
Acute Renal Failure, N (%)	54 (65.9)	16 (61.5)	38 (67.9)	0.62
Chronic Renal Failure, N (%)	50 (61.7)	18 (72.0)	32 (57.1)	0.23
RRT Required, N (%)	2 (2.4)	1 (3.8)	1 (1.8)	0.54
Elevated LFT, N (%)	37 (45.1)	17 (65.4)	20 (35.7)	0.017
Cirrhosis, N (%)	1 (1.2)	0	1 (1.8)	1

^aAcute Cellular Rejection (ACR) score calculated by adding all pathologic “A” scores and dividing by number of transbronchial biopsies;

^bCell counts assessed at 1 year post transplant,

^cLeukopenia defined as white blood cell count <4 k/ul,

^dAnemia defined as hemoglobin <12 g/dl for women and <12.4 g/dl for men;

^eMacrocytosis defined as mean corpuscular volume >98 fL,

^fThrombocytopenia defined as platelets <150 K/u.

Abbreviations: N, number; SD, standard deviation; CI, confidence interval; IQR, interquartile range; PGD, primary graft dysfunction; ACR, acute cellular rejection; CLAD, chronic lung allograft dysfunction; BOS, bronchiolitis obliterans syndrome; R-CLAD, restrictive-chronic lung allograft dysfunction; Yr. year; RRT, renal replacement therapy; LFT, liver function tests

Figure 1. Survival time post-lung transplantation in patients stratified by leukocyte telomere lengths.

Survival of pulmonary fibrosis patients as depicted in a Kaplan-Meier survival plot. Patients with leukocyte telomere lengths <10^th percentile have worse survival than patients with telomere length ≥10^th percentile (log rank, p=0.019).

Table 3.

Univariable and Multivariable Analysis of Survival

	Univariable Analysis			Mulitvariable Analysis (N=64)
	N	HR (95% CI)	P-value	HR (95% CI)	P-value
Telomere <10th percentile	82	2.62 (1.14–6.02)	0.023	10.9 (2.68–44.8)	0.001
Grade 3 PGD	64	1.6 (0.36–7.21)	0.53	1.93 (0.37–10.0)	0.43
ACR Score	82	0.23 (0.05 – 1.08)	0.063	2.28 (0.31–16.9)	0.42
CLAD	82	1.2 (0.52 – 2.75)	0.67	7.10 (1.10–46.3)	0.041
Total Infections/year	82	1.23 (1.12 – 1.36)	<0.001	1.63 (1.34–1.98)	<0.001

The reported HRs are for the presence of binary predictors (presence of telomeres <10^th percentile, Grade 3 PGD or CLAD) or for a 1-unit difference in quantitative predictors (ACR score or total infections/year).

Abbreviations: N, number; HR, hazard ratio; PGD, primary graft dysfunction; ACR, acute cellular rejection; CLAD, chronic lung allograft dysfunction

Lung Allograft Dysfunction

Patients with telomere length <10^th percentile had higher rates of allograft dysfunction compared to patients with telomere lengths ≥10^th percentile (Table 2). Data to assess primary graft dysfunction (PGD) was available for 64 of the 82 patients. For this subset of patients, there was a significant difference in the incidence of Grade 3 PGD between the two groups (28% in the telomere length <10^th percentile group vs. 7% in the ≥10^th percentile group, p=0.034). Likewise, there was a higher incidence of CLAD in the short telomere group (50% in the telomere length <10^th percentile group vs. 23% in the telomere length ≥10^th percentile group, p=0.022, Figure 2). Interestingly, there was a trend toward a higher incidence of the restrictive form of CLAD (R-CLAD) in the short telomere group, although this did not reach significance. Telomere length <10^th percentile was an independent predictor for the time to onset of CLAD after adjusting for other covariates including Grade 3 PGD, acute cellular rejection score and pulmonary infections per year, (HR 6.3, 95% CI 2.0–20.0, p=0.002, Table 4). Similar results were found after excluding single lung transplant recipients (n=52, TL <10^th percentile HR 7.3 (95% CI 2.2–24.6) p=0.001), data not shown. Telomere length as a continuous variable was significantly associated with time to development of CLAD in the univariable analysis (HR 0.20 (95% CI 0.04–0.98) per unit change in telomere length, p=0.047); after correcting for relevant covariates there was a trend toward significance (HR 0.11 (95% CI 0.01–1.09), p=0.059, Supplemental Table 3).

Figure 2. Time to development of Chronic Lung Allograft Dysfunction (CLAD) in patients stratified by leukocyte telomere lengths.

Time to the onset of CLAD as depicted in a Kaplan-Meier plot. Patients with leukocyte telomere length <10^th percentile have a shorter time to the onset of CLAD as compared to patients with telomere lengths ≥10^th percentile (log rank, p=0.005). CLAD is defined as the persistent decline in forced expiratory volume in 1 second (FEV₁) less than 80% of the individual baseline FEV₁ (the average of the two best FEV₁ values after transplantation) which cannot be attributed to concurrent infection or acute cellular rejection.

Table 4.

Univariable and Multivariable Analysis of Time to Onset of Chronic Lung Allograft Dysfunction (CLAD)

	Univariable Analysis			Multivariable Analysis (N=58)
	N	HR (95% CI)	P-value	HR (95% CI)	P-value
Telomere <10th percentile	82	2.85 (1.32 – 6.17)	0.0077	6.3 (2.0–20.0)	0.002
Grade 3 PGD	64	1.32 (0.17–10.0)	0.78	0.39 (0.05–3.4)	0.39
ACR Score	82	1.09 (0.35 – 3.41)	0.89	4.6 (0.85–24.7)	0.076
Pulmonary Infections/year	82	1.51 (1.17 – 1.95)	0.0018	1.8 (1.1–2.7)	0.011

The reported HRs are for the presence of binary predictors (presence of telomeres <10^th percentile or grade 3 PGD) or for a 1-unit difference in quantitative predictors (ACR score or total infections/year).

Abbreviations: N, number; HR, hazard ratio; PGD, primary graft dysfunction; ACR, acute cellular rejection

Other manifestations of organ dysfunction

Non-pulmonary organ dysfunction generally occurred at similar rates across the two groups. Although there was a trend that those with short telomere lengths received more packed red blood cell transfusions per year (Supplemental Table 2), there was no significant difference in the incidence of anemia, leukopenia or thrombocytopenia, at 30 days (Supplemental Table 2) or at 1 year post-transplant (Table 2). Renal dysfunction was common but did not differ between the two groups. Acute and chronic renal failure occurred in 66% and 62% of patients, respectively. Overall, only two patients required dialysis, lasting no longer than three days. Patients with short telomeres had higher rates of LFT elevations after transplantation but the incidence of liver cirrhosis was rare. There was no difference in the rate of pulmonary and non-pulmonary infections or in the type of pathogens (Supplemental Table 2).

The incident rate of venous thromboembolism (both pulmonary embolism and deep venous thrombosis) was similar between both groups (Supplemental Table 2). Post-transplant malignancy was diagnosed in 15 (18%) patients, with no difference between telomere length groups. Most common cancers were non-melanomatous skin cancers; only one patient developed a non-skin malignancy, an adenocarcinoma of unknown primary.

Medication Regimen

All patients were maintained on a triple drug immunosuppression regimen after lung transplantation (Table 5). However, patients in the <10^th percentile group were exposed to median of 5 drugs (within the three drug classes) compared to median of 3 drugs in the ≥10^th percentile group (p=0.005). There was no difference in rate of basiliximab induction or prophylactic antibiotic usage between the two groups.

Table 5.

Use of Immunosuppressive Medications After Lung Transplantation

	Total N=82	<10th Percentile, N=26	≥10th Percentile, N=56	P-value
	N=67	N=21	N=46
Basilixumab Induction, N (%)	41 (67)	13 (62)	28 (61)	1
Triple Immunosuppressive Regimen, N (%)	N=82	N=26	N=56
Immediately Post Transplant	80 (100)	25 (100)	55 (100)	1
At 1 year Post Transplant	74 (100)	23 (100)	51 (100)	1
At 3 years Post Transplant	43 (100)	13 (100)	30 (100)	1
At 5 years Post Transplant	19 (100)	6 (100)	13 (100)	1
At 7 years Post Transplant	5 (100)	1 (100)	4 (100)	1
Total Number Immunosuppressive Medications Used, median (IQR)a	5.0 (3.0–5.0)	5.0 (4.25–6.0)	3.0 (3.0–5.0)	0.005
Total Number Prophylactic Antimicrobial Drugs Used, median (IQR)	2.0 (2.0–3.0)	2.0 (2.0–3.0)	2.0 (2.0–2.0)	0.21

^aExposure to medications including Prednisone, Imuran, Cellcept/Mycophenolate, Cyclosporine, FK-506/Tacrolimus and Sirolimus.

Discussion

Here we compared clinical outcomes of patients who underwent lung transplantation for pulmonary fibrosis stratified by telomere length. While previous studies have investigated post lung transplantation outcomes for individuals with mutations in telomere-related genes(9–11), the association of recipient leukocyte telomere length to patient outcomes has not been extensively studied. We found that a leukocyte telomere length <10^th percentile of normal was associated with shorter post-transplantation survival as well as a higher incidence of Grade 3 PGD and a shorter time to the onset of CLAD. Utilization of the 10^th percentile cut-off, representing a degree of telomere shortening shared by most (>80%) telomere-related mutation carriers(6), allows for dichotomization of the full cohort for comparison of clinical outcomes.

Only one prior study has investigated the effect of telomere length on transplant outcomes(22). This study is similar in that the cohort was stratified by telomere length, and the group with shorter lengths represented ~1/3 of the total. However, this study differs in a number of ways. First, this study measured telomere length from recipient blood leukocytes, and not from explanted lung tissue, so as to generate control- and age-adjusted percentiles. Second, this study included 100%, and not ~40%, of subjects who underwent lung transplantation for pulmonary fibrosis, so as to increase the homogeneity of patient responses to lung transplantation. Third, the enrollment period (7 years) was more than double the prior study so long-term outcomes, such as survival, could be assessed. Fourth, we found higher rates of PGD and CLAD in the group of patients with telomere lengths <10^th percentile. Finally, the prior study did not show an association between recipient telomere length and survival as the current study does.

The results of this study also differ from prior studies of pulmonary fibrosis patients harboring a heterozygous pathogenic mutation in one of the telomerase genes(9–11). We found no significant difference in blood cell counts after lung transplantation; however, the overall incidence of leukopenia (27%), anemia (57%), and thrombocytopenia (20%) was common for the entire cohort. We also did not find a significant increase in renal dysfunction associated with telomere length; instead, both acute and chronic renal dysfunction was commonly found across the entire cohort and was not specifically associated with telomere length. Similarly, we did not find an association between incident malignancy or infection and leukocyte telomere length. The only short telomere-associated phenotype seen at a higher incidence in the patients with short telomere lengths was an elevation of LFTs. Elevation of the LFTs likely led to more immunosuppression medication changes for patients with shorter telomere lengths, as this group had, on average, exposure to more medications that the group with longer telomeres.

We found a significant increase in PGD and CLAD in patients with the shorter telomere lengths. PGD accounts for the majority of early morbidity and mortality after transplant(23–25) and has been linked to various donor, recipient, and surgical characteristics(26–28). Recipient telomere lengths may potentially explain the previously reported association between reperfusion injury and native fibrotic, as opposed to obstructive, lung disease(29). Like PGD, the pathogenesis of CLAD is multifactorial. CLAD results in non-reversible allograft damage after insults such as infection(30, 31), gastroesophageal reflux(32, 33), elevated donor-specific antibodies (DSAs)(34–36), and episodes of acute cellular rejection(17, 37). CLAD subtypes include an obstructive defect (bronchiolitis obliterans syndrome) and a restrictive defect (R-CLAD). Although it did not reach statistical significance, short telomere lengths were more closely associated with the restrictive subtype of CLAD. This is notable because the pulmonary pathologic manifestations of R-CLAD, including diffuse alveolar damage(38, 39), acute fibrinous organizing pneumonia(40) and pleuroparenchmal fibroelastosis(41), have been described for patients with telomere-related gene mutations(8). While type II alveolar epithelial cell senescence contributes to the development of pulmonary fibrosis(42, 43), this mechanism cannot be responsible for CLAD in these patients following lung transplantation. We could not fully assess for an association between CLAD and DSAs, as the latter were not measured for the entire cohort.

The association between short telomere lengths and CLAD does not prove causality, but suggests that the immune dysregulation and/or the injuries related to infections associated with telomere shortening may be linked to CLAD in the post-transplant period. Shortened telomere lengths in T cells are associated with the loss of expression of the CD28 costimulatory molecule and results in broad reprogramming of the CD28- T cells and an altered adaptive immunity(44) and reduced clonal expansion(45). Shortened telomere lengths may also contribute to the risk or persistence of pulmonary infections which may trigger the onset of CLAD. Short telomere length are associated with higher rates of upper respiratory viral infections in healthy subjects(46) and individuals with short telomere lengths may be less tolerant to infections once they occur(47). Further study will be needed to investigate the underlying pathogenesis of CLAD in patients with short telomere lengths.

Despite an extensive pre-transplant recipient selection process, the ability to predict complications and survival after lung transplantation is limited. Telomere length is a biomarker that could aid in risk stratification of IPF patients at the time of their diagnosis not only to predict risk of progressive disease and the need for lung transplantation(20, 48), but also to predict risk of post-transplant complications, such as PGD and CLAD. Although the overall survival of patients with telomere length <10^th percentile was significantly lower than patients with telomere length ≥10^th percentile, the short telomere length patients still had an acceptable survival time as compared to national benchmarks.

The present study has several limitations. First, this was a single center study. Overall, the percentage of patients with familial pulmonary fibrosis (27%) was higher than has been described at other centers(49) and may lead to an enrichment of patients with short-telomere lengths. Second, because we limited inclusion to patients with pulmonary fibrosis, these results cannot be extrapolated to those with other indications for lung transplant. Finally, we were unable to assess donor demographic information or donor telomere lengths with post-transplant outcomes.

Thus, telomere length represents a potential biomarker that can easily be assessed during the pre-transplant evaluation to identify those patients at increased risk for post-transplant complications such as PGD and CLAD. The results from this single site study are compelling and warrant replication in a larger, multicenter cohort. If replicable, these studies may lead the way to additional investigations probing the link between lung transplant recipient short leukocyte telomere lengths, allograft dysfunction and survival.