Lung transplant recipients with telomere-mediated pulmonary fibrosis have increased risk for hematologic complications

Abstract

Idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs) are enriched for short telomere length (TL) and telomere gene rare variants. A subset of patients with nontransplant short-TL are at increased risk for bone marrow (BM) dysfunction. We hypothesized that IPF-LTRs with short-TL and/or rare variants would be at increased risk for posttransplant hematologic complications. Data were extracted from a retrospective cohort of 72 IPF-LTRs and 72 age-matched non-IPF-LTR controls. Genetic assessment was done using whole genome sequencing or targeted sequence panel. TL was measured using flow cytometry and fluorescence in-situ hybridization (FlowFISH) and TelSeq software. The majority of the IPF-LTR cohort had short-TL, and 26% of IPF-LTRs had rare variants. Compared to non-IPF controls, short-TL IPF-LTRs were more likely to have immunosuppression agents discontinued due to cytopenias (P = .0375), and BM dysfunction requiring BM biopsy was more prevalent (29% vs 4%, P = .0003). IPF-LTRs with short-TL and rare variants had increased requirements for transfusion and growth factor support. Multivariable logistic regression demonstrated that short-TL, rare variants, and lower pretransplant platelet counts were associated with BM dysfunction. Pretransplant TL measurement and genetic testing for rare telomere gene variants identified IPF-LTRs at increased risk for hematologic complications. Our findings support stratification for telomere-mediated pulmonary fibrosis in lung transplant candidates.

1. Introduction

Idiopathic pulmonary fibrosis (IPF) is the leading indication for lung transplantation in North America.¹ Earlier work demonstrated that patients with IPF are enriched for short telomere length (TL) (> 50%) and approximately 12% harbor germline mutations in telomerase and telomere-maintenance genes.^2–7 Most recently, we have shown that 24% of IPF-LTRs are enriched for rare variants in the telomere-maintenance genes, compared with 12% in nontransplanted IPF patients.⁸ We also found that IPF-LTRs had a similar risk for acute cellular rejection (ACR) events in the first posttransplant year and chronic lung allograft dysfunction (CLAD).⁸ In addition, previous studies reported an increased incidence of CLAD in IPF-LTRs with short-TL.^9,10 Cytomegalovirus (CMV) infection, a common complication in LTRs, has been previously associated with increased CLAD risk.^11–15 We and others have reported an increased susceptibility to CMV complications in short-TL patients with or without lung transplant due to impaired T cell immunity.^7,16,17 Further, we have reported an increased risk for Epstein-Barr virus (EBV)-associated posttransplant lymphoproliferative disorder (PTLD) in IPF-LTRs.¹⁸ Results on posttransplant hematologic complications in patients with telomere-mediated pulmonary fibrosis are mixed, with some studies reporting increased complications ^16,19–21 and another finding no difference compared with control patients.⁹

In the current study, we initially identified IPF-LTRs with short-TL and/or rare variants to have an increased incidence of discontinuation of select immunosuppression (IS) medications (eg, mycophenolate mofetil) due to cytopenias. Based on this, we hypothesized that stratification of IPF-LTRs by TL or rare variants in the telomere-maintenance genes would identify individuals at increased risk for posttransplant hematologic complications. We also found that cytopenias led to an increased need for hematologic support and/or hematologic evaluation requiring bone marrow biopsy (BMBx) in patients with short-TL and rare variants. Together, these findings support the idea that measurement of TL and genetic testing are useful tools in stratifying IPF-LTRs who are at increased risk for posttransplant hematologic complications.

2. Materials and Methods

2.1. Subjects

This is a single-center retrospective cohort study that was approved by the University of Pittsburgh institutional review board and all study participants signed informed consent prior to enrollment. All patients were diagnosed with IPF according to the consensus guidelines of the American Thoracic Society and European Respiratory Societies at the time of their enrollment.^22,23 Patient subjects enrolled in the University of Pittsburgh Medical Center (UPMC) lung transplant research registry between March 2015 and June 2019 were assessed for eligibility in the study. Briefly, 29% (151/529) of patients in our transplant registry were diagnosed with IPF, and 72 had peripheral blood mononuclear cells available for research FlowFISH determination for TL assessment, with 69 of 72 patients also undergoing genetic testing. We then identified 72/378 aged-matched non-IPF controls for comparison (Supplementary Fig. S1). Patient characteristics, including age, gender, family history, diagnosis at transplant, and type of transplant (single vs double), were collected via retrospective chart review. Pretransplant hematologic parameters, IS regimens, induction regimen, discontinued medications, blood product requirement, growth factor support, and BMBx results were obtained by retrospective chart review. Transfusion requirement was assessed 90 days after lung transplant to avoid immediate postoperative complications and also did not include acute events such as surgeries and gastrointestinal bleeding, among others. It is also the protocol of the UPMC lung transplant team to discontinue the cell cycle inhibitor prior to discontinuation of the antiviral or antibacterial prophylaxis for cytopenias. The standard protocol at our center for the administration of growth factor support is for an absolute neutrophil count of < 1000 cells/μL.

2.2. TL measurement

TL measurement of the entire IPF-LTR cohort was determined using FlowFISH at the Johns Hopkins Pathology Laboratories.³ Additionally, 57 IPF-LTRs underwent whole genome sequencing (WGS), 12 had targeted panel sequencing, and 3 had no genetic testing. TL was measured using TelSeq application in WGS samples.^24,25 A nomogram was established for TelSeq using 901 noninterstitial lung disease samples from the UPMC Genome Center from patients between the ages of 18 and 95 years old, as previously described.⁸

2.3. Genetic analysis

Genetic characterization was performed using WGS and the Genome Aggregation Database, as previously described.⁸ Briefly, we focused on the 7 telomere-maintenance genes that have been previously reported in patients with IPF: TERT, TERC (or TR), RTEL1, PARN, TINF2, NAF1, and DKC1.^4,26–31 Using the American College of Medical Genetics guidelines, the genes were further classified as either pathogenic, likely pathogenic, or variants of uncertain significance.³² Additionally, next-generation sequencing was performed on a subcohort of IPF-LTRs who underwent BMBx to determine somatic gene mutations in the bone marrow (BM). Genetic characterization was then compared with TL and hematologic outcomes.

2.4. Statistical Analysis

The baseline characteristics of IPF-LTRs and non-IPF-LTRs were compared using χ² test or Fisher exact tests to compare categorical variables. A multivariable logistic regression analysis was performed to determine if telomere genetics and pretransplant blood counts were predictive of hematologic outcomes in IPF-LTRs. Variables evaluated for inclusion in the multivariable logistic regression included sex, age, pretransplant white blood cell count, pretransplant hemoglobin (Hgb), pretransplant red blood cell mean corpuscular volume and pretransplant platelet count (PLT). Variables were included in the final model if they had a P value of < .20.

3. Results

3.1. The majority of IPF-LTRs have short telomeres in the presence or absence of rare gene variants in the telomere-maintenance genes

To assess posttransplant hematologic complications, we evaluated a single-center cohort of 144 LTRs with IPF vs non-IPF (n = 72 per group), as shown in Table 1, with the derivation of this convenience cohort shown in Supplementary Figure S1 consort diagram (please see methods for details). The groups were similar in age, with a male preponderance in the IPF group (P = .002), as expected. The majority of the non-IPF-LTR group had a chronic obstructive pulmonary disease/emphysema diagnosis as indication for transplant, and the majority of LTRs in each group underwent double lung transplant. We also observed a trend in the discontinuation of IS medications in IPF-LTRs compared to non-IPF controls (P = .054), with mycophenolate mofetil being the most commonly discontinued medication (Table 1). We also compared the baseline characteristics within the IPF-LTR cohort stratified by short vs long TL and found no significant differences (Supplementary Table S1).

Table 1.

Baseline characteristics of study cohort.

	IPF (n = 72)	non-IPF (n = 72)	P value
Age (y)	65 (48–72)	66 (47–71)	NS
Male, n (%)	56 (77)	38 (53)	.002
Diagnosis at transplant, n (%)
IPF
COPD/emphysema	72 (100)	50 (70)	NS
Other		22 (30)
Family history
IPF, n (%)	9 (12.5)	NA	NS
Induction, n (%)
Basiliximab	31 (44)	30 (42)	NS
Campath	41 (57)	42 (58)	NS
IS drugs, n (%)
Triple-drug*	49 (68)	59 (82)	NS
IS discontinuation#	23 (32)	13 (18)	.054
Viral ppx reduction	10 (14)	14 (19)	NS
Bacterial ppx reduction	2 (3)	7 (9)	NS
Transplant procedure, n (%)
Single	21 (29)	17 (24)	NS
Double	51 (71)	55 (76)	NS

Abbreviations: COPD, chronic obstructive pulmonary disease; IPF, idiopathic pulmonary fibrosis; IS, immunosuppression; NA, not applicable; NS, nonsignificant; ppx, prophylaxis.

^*Triple-drug immunosuppression: calcineurin inhibitor, antiproliferative (mycophenolate mofetil or azathioprine), and prednisone.

^#Discontinuation of immunosuppression drugs secondary to cytopenia. Mycophenolate mofetil was most commonly discontinued. IPF 48% (11/23) and Non-IPF 92% (12/13).

We next determined TL using clinically validated flow cytometry and FlowFISH and found that 68% (49/72) of IPF-LTRs had significantly short lymphocyte TL (Fig. 1A) and 80% (41/51) of IPF-LTRs had short granulocyte TL (≤ 10th percentile for age) (Fig. 1B). TL was measured using TelSeq software in 79% (57/72) of IPF-LTRs who underwent WGS. Of these, 58% (33/57) of IPF-LTRs had telomeres below the 10th percentile (Fig. 1C). In addition, age-adjusted mean TL (TelSeq) was significantly lower in these 57 IPF-LTRs compared with the 901 normal control group (P < .0001) (Fig. 1D). Of the 57 IPF-LTRs who underwent WGS and FlowFISH, we compared their TLs between both assays and found that these correlated (Fig. 1E).

Figure 1.

Telomere length (TL) measurement of idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs) using flow cytometry and fluorescence in-situ hybridization (FlowFISH) and TelSeq software. TL was measured using FlowFISH on the IPF-LTR cohort (72 patients). (A) Telegram showing 68% (49/72) of IPF-LTRs with short lymphocyte TL (≤ 10th percentile). (B) Telegram showing 80% (41/51) with short granulocyte TL (≤ 10th percentile). (C) A total of 79% (55/72) underwent WGS, and TL length was measured using TelSeq software. (D) IPF-LTR cohort with TL below the age-adjusted mean compared with 901 normal control nomograms. (E) Correlation of FlowFISH lymphocyte TL and TelSeq assay in IPF-LTRs (n = 57).

To genetically characterize our IPF-LTR cohort, we used WGS or a targeted sequence panel that included the telomere-associated genes. We examined the 7 telomere-maintenance genes that have previously been associated with pulmonary fibrosis: TERT, TERC (also known as TR), RTEL1, PARN, NAF1, TINF2, and DKC1. We identified rare variants in the telomere-associated genes in 26% (19/72) of our IPF-LTR cohort, consistent with our previous report of 24% incidence (Table 2). We used the American College of Medical Genetics guidelines to classify these rare variants and found that 16.67% (12/72) were either pathogenic or likely pathogenic, similar to our previous findings (Table 2). Notably, all of the 19 IPF-LTRs with rare variants had short lymphocyte TL (≤ 10th percentile). Together, these data indicate that IPF-LTRs with short TL have a high incidence of rare gene variants in the telomere-maintenance genes.

Table 2.

Rare variants in telomere-maintenance genes in idiopathic pulmonary fibrosis lung transplant recipient (IPF-LTR) cohort.

IPF-LTR No.	Age (y)	Gender	Family history	Mutant gene	Protein position	Prior disease association	gnomAD MAF*	Interpretation	Lymph TL percentile
8	5	M	Yes	PARN	F123L	No	Absent	VUS	1st-10th
9	55	M	Yes	RTEL1	A687G	No	Absent	Likely pathogenic	<1st
10	69	F	Yes	PARN	E585Dfs*4	Exome sequencing pts with IPF PMID:28099038	9.67 × 10−5	Likely pathogenic	1st-10th
13	76	M	No	RTEL1	R1264H	Ballew et al33, 2013	1.31 × 10−4	Pathogenic	1st-10th
14	57	M	No	DKC1	S422T	No	Absent	VUS	1st-10th
33	48	M	No	PARN	E58Dfs*4	Exome sequencing patients with IPF PMID:28099038	9.67 × 10−5	Likely pathogenic	1st-10th
35	59	F	No	RTEL1	R754Q	No	1.34 × 10−4	Likely pathogenic	<1st
36	56	M	No	PARN	K211R	No	Absent	VUS	1st-10th
70	61	F	No	RTEL1	R449X	No	Absent	Pathogenic	<1st
97	67	M	No	TERT	D129Tfs*221	No	Absent	Pathogenic	<1st
169	67	M	No	DKC1	M204R	No, outside of pathogenic area	Absent	VUS	1st-10th
171	72	F	Yes	NAF1	S297_D298ins	No	Absent	Likely pathogenic	<1st
174	72	M	Yes	TERC	N/A	No	Absent	VUS	<1st
237	69	M	Yes	PARN	Splicing	No	Absent	Pathogenic	<1st
301	72	M	No	RTEL1	L710R	PMID: 23453664	Absent	VUS	1st-10th
314	50	M	No	RTEL1	R974X	Cogan et al28, 201 5 (PMID: 28099038, 25607374,25047097, 23959892,23453664, and 23329068)	3.24 × 10−5	Pathogenic	<1st
317	56	M	No	TINF2	A323P	No, outside of pathogenic area	1.08 × 10−5	Benign	<1st
368	70	M	No	TERT	P632L	PMID: 28192371	Absent	Pathogenic	1st-10th
483	64	F	No	PARN	Splicing	No	Absent	Pathogenic	<1st

Abbreviations: gnomAD, Genome Aggregation Database; MAF, minor allele frequency; Lymph, lymphocyte; TL, telomere length; VUS, variant of unknown significance.

^*The gnomAD MAF is derived from data from a total of 138,632 individuals (123,136 exomes and 15,496 genomes).

3.2. IPF-LTRs with short telomeres are at increased risk for BM dysfunction and hematologic evaluation requiring BMBx, but not rejection outcomes

Next, we stratified IPF-LTRs by lymphocyte TL to assess IS medication discontinuation from 3-drug to 2-drug regimens due to cytopenia. IPF-LTRs with short TL had a significant discontinuation of the third immunosuppression agent, most often mycophenolate mofetil, compared to non-IPF control LTRs but not long telomere IPF-LTRs (Fig. 2A). Of the 17 patients with short-TL with discontinuation of IS medication, 14 (82%) underwent hematologic evaluation with BMBx, a significantly higher incidence compared to non-IPF controls (Fig. 2B). We then compared the induction regimen given between the IPF-LTRs and non-IPF-LTR controls and found no significant differences between the use of alemtuzumab and simulect (Table 1). We further assessed allograft outcomes comparing ST-LTRs who had IS medication discontinued compared with the control groups and found no difference in freedom from ACR or CLAD censored at 2 and 5 years, respectively, using Kaplan Meier analyses (Supplementary Fig. S2). We also determined the incidence of donor-specific antibodies and found no differences with 43% (3/7) of short TL IPF-LTRs vs 25% (1/4) of long TL IPF-LTRs (P = .55) or compared with the 55% (6/11; P = .63) of non-IPF controls . Only 1 patient with short TL IPF-LTR was treated for antibody-mediated rejection, and there were no patients in the control groups. Together, these data do not show differences in allograft rejection outcomes in short TL IPF-LTRs with IS medication discontinued compared with controls.

Figure 2.

Immunosuppression (IS) drug discontinuation and hematologic evaluation requiring bone marrow biopsy (BMBx) in idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs) stratified for telomere length (TL). (A) IPF-LTRs stratified by lymphocyte TL to assess IS medication discontinuation from 3 to 2 drug regimens due to cytopenias, where short-TL IPF-LTRs had a significantly increased discontinuation of the third drug (most often mycophenolate mofetil) compared with the non-IPF-LTRs. (B) In total, 17 short-TL IPF-LTRs had discontinuation of IS drug, and 82% (14/17) underwent hematologic evaluation with BMBx, a significantly increased incidence compared to non-IPF controls (P = .0003). The majority of IPF-LTRs undergoing BMBx had short lymphocyte (C) and granulocyte (D) TLs by FlowFISH. Delta TL from the age-adjusted mean showed similar lymphocyte TL (E) among IPF-LTRs but lower granulocyte TL in those undergoing BMBx (F).

The majority of IPF-LTRs who underwent BMBx demonstrated short lymphocyte and granulocyte TL by FlowFISH, as shown in Figure 2C, D. We next analyzed the ΔTL from the age-adjusted mean and found that while IPF-LTRs had reduced TL, those undergoing BMBx had a significantly lower granulocyte TL (Fig. 2E, F). The IPF-LTRs with short TL and their BMBx results are shown in Table 3, and controls (IPF-LTRs with long TL and non-IPF-LTRs) are shown in Supplementary Table S2. The median time to hematologic evaluation and BMBx for IPF-LTRs with short TL was 1230 days posttransplant and did not differ from controls (1079 days; P = .5). Within the first year of transplant, only 14% (2/14) of patients who underwent BMBx developed BM failure, with 1 patient receiving alemtuzumab induction and the other receiving simulect. In IPF-LTRs with short TL, 50% (7/14) had hypocellular BMs, and 23% (3/13) demonstrated somatic mutations in the BM. Although the patient with 13q deletion did not develop significant BM failure, the patient with partial trisomy 1q had significant pancytopenia but did not have the clinical diagnosis of myelodysplastic syndrome established.³⁴ The third patient with an IGH/MYC gene rearrangement developed a severe case of PTLD and succumbed to the disease. MYC gene rearrangement has been previously shown to be associated with poorer outcomes in PTLD.³⁵ IPF-LTRs who underwent BMBx were comparable to those who did not in terms of age and sex. IS medication discontinuation occurred at a significantly higher rate in those who underwent BMBx (81% vs 19%, P < .0001) (Table 4). Importantly, 88% (14/16) of IPF-LTRs who underwent BMBx had short lymphocyte TL by FlowFISH compared with 63% (35/56) of IPF-LTRs without BMBx (P = .058) (Table 4). Furthermore, IPF-LTRs with short TL undergoing BMBx had significantly increased likelihood of having rare variants identified in the known telomere-maintenance genes: 71% (10/14) of IPF-LTRs with short TL undergoing BMBx compared with 26% (9/35) of IPF-LTRs without BMBx (P = .003) (Table 4). Taken together, these data demonstrate an enrichment in short TL with rare variants in the telomere-maintenance genes among IPF-LTRs with BM dysfunction that underwent BMBx compared with those who did not.

Table 3.

Bone marrow biopsy (BMBx) results and genetic assessment of short TL in idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs).

IPF-LTR No.	Age (y)	Gender	Indication for BMBx	Time PostLTx	BMBx interpretation	Lymphocyte TL (percentile)	BM NGS	Telomere gene variant
652	70	F	Pancytopenia	Day 1687	- Normocellular w/ trilineage hematopoiesis	1st-10th	None	Not done
					- No morphologic evidence of involvement by PTLD
8	55	M	Anemia/TCP	Day 1424	- Normocellular w/ trilineage hematopoiesis and dyspoietic features	1st-10th	None	PARN
10	69	F	Pancytopenia	Day 166	- Normocellular w/ trilineage hematopoiesis w/ very mild plasmacytosis (3.4%)	1st-10th	None	PARN
9	55	M	Pancytopenia	Day 858	- Moderately hypocellular w/ stromal injury, decreased trilineage hematopoiesis, dyspoietic changes, increased blasts (1.3%)	<1st	None	RTEL1
36	56	M	TCP	Day 1078	- Variably cellular w/ trilineage hematopoiesis	1st-10th	None	PARN
					- Megakaryocytes present in normal numbers
237	69	M	Anemia/TCP	Day 2149	- Hypocellular w/ erythroid-predominant trilineage hematopoiesis and mild dyspoiesis	<1st	Partial trisomy 1q	PARN
33	48	M	Pancytopenia	Day 1203	- Mildly hypercellular w/ erythroid-predominant trilineage hematopoiesis and mild plasmacytosis	1st-10th	None	PARN
64	71	M	Pancytopenia	Day 1780	- hypercellular marrow, increased iron stores	1st-10th	13q Deletion	None
97	67	M	Pancytopenia	Day 1257	- Slightly hypocellular w/ erythroid-predominant trilineage hematopoiesis	<1st	None	TERT
328	58	F	Pancytopenia	Day 2365	- CD19-, CD56+ Plasma cell population (2%−3%), hypocellular marrow w/ erythroid predominance, left shifted myeloid maturation	1st-10th	IGH/MYC gene rearrangement	None
14	57	M	Pancytopenia	Day 1878	- Hypocellular w/ trilineage hematopoiesis and stromal changes	1st	one	DKC1
257	66	M	TCP	Day 1198	- Normocellular w/ trilineage hematopoiesis	<10th	None	None
70	62	F	Anemia/TCP	Day 349	- Hypocellular w/ erythroid-predominant trilineage hematopoiesis	<1st	None	RTEL1
477	72	M	Pancytopenia	Day 258	- Hypocellular w/ trilineage hematopoiesis	<10th	Unknown	None

Abbreviations: BM NGS, bone marrow next-generation sequencing; LTx, lung transplant; PTLD, posttransplant lymphoproliferative disorder; TCP, thrombocytopenia; TL, telomere length, w/, with.

Table 4.

Characteristics of idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs) who underwent hematologic evaluation.

	IPF no BMBx (n = 56)	IPF BMBx (n = 16)	P value
Median age, (y)	65 (48–72)	64 (48–71)	NS
Male, n (%)	44 (78)	12 (75)	NS
IS discontinuation *	11 (19)	13 (81)	<.0001
Short telomere length	35 (62)	14 (87)	.058
Telomere rare variant in short telomere LTRs †	9/35 (26)	10/14 (71)	.003
Transfusion support in IPF-LTRs	8 (14)	12 (75)	<.001
G-CSF support in IPF-LTRs	33 (59)	15 (94)	<.01

Abbreviations: BMBx, bone marrow biopsy; IPF- Idiopathic pulmonary fibrosis; LTR, lung transplant recipient; IS, immunosuppression; NS, nonsignificant, G-CSF, granulocyte colony-stimulating factor.

^*Discontinuation of immunosuppression secondary to cytopenia.

^†Rare variants were only checked in IPF-LTRs.

3.3. IPF-LTRs with short TL, rare variants in telomere genes, and those undergoing BMBx require more hematologic support posttransplant

We next evaluated the requirements for blood transfusion and growth factor support (granulocyte colony-stimulating factor; G-CSF) in the IPF-LTRs compared to non-IPF-LTRs, stratifying for TL in the former. Short TL IPF-LTRs were more likely to require blood product transfusions (>90 days posttransplant surgery) compared with long TL IPF-LTRs or non-IPF-LTRs (Fig. 3A). In addition, short-TL IPF-LTRs were also more likely to require G-CSF support for neutropenia compared to non-IPF-LTRs but not long-TL IPF-LTRs (Fig. 3B). These findings are consistent with prior studies.^16,19 Further, we compared posttransplant hematologic support in IPF-LTRs undergoing BMBx for BM dysfunction vs those without BMBx. As expected, IPF-LTRs undergoing BMBx had a significantly increased incidence of either transfusion or G-CSF support compared with those without BMBx (Table 4). Additionally, we found that IPF-LTRs with rare variants in the telomere-maintenance genes were more likely to require transfusion support (53%; 10/19) compared with those without rare variants (19%; 10/53) (P = .005). However, 68% (13/19) of IPF-LTRs with rare variants required G-CSF support at a similar rate compared with those without rare variants (66%; 35/53, P = 85). Together, these data show that in IPF-LTRs, the presence of short TL with or without the rare gene variants is associated with BMBx, need for blood products, and G-CSF support.

Figure 3.

Blood transfusion and granulocyte colony-stimulating factor (G-CSF) support among idiopathic pulmonary fibrosis lung transplant recipients (IPF-LTRs) vs non-IPF-LTRs. (A) Short-telomere length (TL) of IPF-LTRs demonstrated increased transfusion requirements compared with either long-TL IPF-LTRs or non-IPF-LTRs >90 days posttransplant surgery. (B) Short-TL IPF-LTRs had an increased G-CSF requirement for neutropenia when compared with either long-TL IPF-LTRs or non-IPF-LTRs.

3.4. TL and/or rare variants are associated with BM dysfunction/biopsy, and transfusion support in IPF-LTRs

We next performed multivariable logistic regression analyses to assess whether telomere genetic data (TL and rare variants), along with pretransplant complete blood count parameters (Hgb, white blood cell count, PLT count, and mean corpuscular volume), were predictive of hematologic outcomes in IPF-LTRs (Table 5). Although short-TL trended toward increased risk for BMBx (odds ratio [OR] = 4.14, P = .09) (similar to univariate regression, Table 4), the detection of a rare variant was significantly associated with BM dysfunction and biopsy (OR = 5.89, P < .01). Additionally, a lower pretransplant PLT count was found to be associated with increased risk for BMBx (OR =0.98, P < .02). Short-TL was independently associated with an increased likelihood of transfusion support (OR = 16.43, P = .01), along with a pretransplant PLT count < 200 K PLT/mcL (OR = 5.72, P = .03). Similarly, the presence of rare variants was associated with increased risk (OR = 4.07, P = .03), along with a PLT count < 200 K PLT/mcL (OR = 6.46, P = .01) and lower age. Unexpectedly, older age appeared to be a protective factor for transfusion support (OR = 0.87, P = .01). Together, these data support that TL and telomere genetics are associated with BM dysfunction with biopsy and transfusion support posttransplant in IPF-LTRs, along with pretransplant PLT count and age.

Table 5.

Multivariable logistic regression of telomere genetic data and pretransplant hematologic parameters.

Outcome: bone marrow biopsy
	OR	95% CI	P value
Short telomeres	4.14	0.81–21.11	.09
Age (y)	0.92	0.84–1.01	.07
PreTx platelet count	0.98	0.97–1.00	.02
	OR	95% CI	P value
Rare Variant	5.89	1.58–21.97	<.01
Age (y)	0.94	0.86–1.02	.14
PreTx platelet count	0.98	0.97–1.00	<.02
Outcome: transfusion support
	OR	95% CI	P value
Short telomeres	16.43	1.89–142.80	.01
Age (y)	0.87	0.79–0.97	.01
PreTx platelet count < 200 k	5.72	1.34–24.45	.03
	OR	95% CI	P value
Rare variant	4.07	1.12–1.05	.03
Age (y)	0.91	0.82–0.99	.045
PreTx platelet count < 200 k	6.46	1.61–25.97	.01
PreTx hemoglobin	0.78	0.58–1.05	.10

Abbreviations: CI, confidence interval; OR, odds ratio; Tx, therapy/treatment.

4. Discussion

Here, using 2 separate methods to measure TL—FlowFISH and TelSeq—we found that the majority of IPF-LTRs had short TL in our cohort. These findings were consistent with our previous studies and support an enrichment of telomere-mediated disease in IPF lung transplant recipients.⁸ Importantly, FlowFISH is the most widely used and clinically validated method for the detection of telomere-mediated disease and is being increasingly used to assess potential transplant candidates, with the standard cutoff to define short-TL being the 10th percentile or less when using FlowFISH based on population nomograms.³⁶ In contrast to the FlowFISH assay, which requires the isolation of peripheral blood mononuclear cells, the TelSeq assay has the advantage of only necessitating small amounts of DNA to estimate TL from WGS data based on the relative abundance of telomeric reads. However, the TelSeq assay has a reduced dynamic range compared with FlowFISH, and we recently reported that comparison of TelSeq measurements from different sequencing pipelines may be difficult, underscoring the importance of TelSeq assay validation using a population nomogram to compare age-corrected TL (Alder et al⁸). Nevertheless, this assay may be useful in stratifying IPF-LTRs for short-TL and in predicting certain transplant outcomes when properly performed.⁸ Moreover, we found a correlation between the FlowFISH and TelSeq assays in our IPF-LTR cohort.

We also tested the hypothesis that stratification by TL and telomere genetics in our IPF-LTR cohort would elucidate differences in discontinuation of IS medications and hematologic complications resulting in BMBx and hematologic support. We observed that short-TL and/or rare variants in the telomere-maintenance genes were associated with increased likelihood of BM failure, the need for hematologic support, and discontinuation of IS medications. Although others have suggested that alemtuzumab induction regimens contribute to posttransplant hematologic complications,³⁷ we found no significant difference in the incidence of posttransplant hematologic complications with the use of either alemtuzumab or basiliximab.

Our initial finding when we compared our IPF-LTR cohort to age-matched non-IPF-LTRs was a trend in discontinuation of IS medications in IPF-LTRs due to cytopenias. When we stratified for TL using FlowFISH, we found that short-TL was associated with discontinuation of IS drugs (3-drug to 2-drug), particularly in those with significant BM dysfunction who underwent BMBx. These IPF-LTRs undergoing BMBx were enriched for short TL with rare variants in the telomere-maintenance genes, along with increased requirements for transfusion support (predominantly red blood cells) and G-CSF support for leukopenia. These results are consistent with 2 prior reports in smaller lung transplant cohorts with TERT or TR rare variants. Another study that measured TL alone using a polymerase chain reaction assay in patients with pulmonary fibrosis undergoing transplant did not detect significant differences in cytopenias (but did in macrocytosis) in those with short TL < 10% vs those with TL > 10%.⁹ However, stratification by TL in the Panther IPF Trial showed impaired tolerance of IS drugs in patients with IPF with short-TL.³⁸ Collectively, these data suggest a susceptibility to IS in patients with short TL and that assessment of both TL and telomere gene rare variants is useful in identifying IPF-LTRs who are at increased risk for BM dysfunction.

Our data are consistent with an earlier study in nontransplant patients showing that patients with pulmonary fibrosis and concomitant BM failure have a very high likelihood of short-TL with a germline variant in telomerase.^39,40 Although patients with pulmonary fibrosis and overt BM dysfunction can present for lung transplantation as a manifestation of the short telomere syndrome, the majority of patients with pulmonary fibrosis have intact pretransplant hematologic parameters. It is only after the initiation of standard 3-drug IS and other transplant therapies that carry BM toxicity (eg, valganciclovir, sulfamethoxazole/-trimethoprim, and certain antibiotics) that significant BM dysfunction may ensue. Our findings suggest utility in the measurement of TL and reflex genetic testing for rare variants in the telomere-maintenance genes in identifying those at increased risk for hematologic complications and relative intolerance to standard IS regimens used in lung transplantation. Indeed, we and others have reported a similar or increased risk for ACR burden and incidence of CLAD in short-TL IPF-LTRs compared with controls.^8,9,20,41 It is likely that the relative intolerance to 3-drug IS in short-TL IPF-LTRs is due to reduced BM reserves, as is known in patients with IPF.⁴⁰ Importantly, in addition to TL and telomere gene rare variants, our findings show that stratification by pretransplant platelet count and younger age were also useful predictors of BM failure or hematologic support, whereas other pretransplant hematologic parameters were not. Although older age might be an unexpected protective factor, these data are consistent with our previous finding that short TL-LTRs are significantly younger compared with patients with nontransplanted IPF.⁸ Therefore, IPF-LTRs with short-TL and/or rare variants, along with lower normal-range PLT counts appear to be among the highest-risk patients for posttransplant BM dysfunction and those requiring hematologic support.

In IPF-LTRs with short-TL undergoing BMBx, we found hypocellularity to be the most common feature in 50% of cases. We also identified somatic mutations in 3 of 14 short-TL IPF-LTRs who underwent BMBx using next-generation sequencing. These data are consistent with recent reports demonstrating that short-TL patients develop clonal hematopoiesis.^42,43 Interestingly, although we did not observe myelodysplastic syndrome in any of our short TL IPF-LTRs, 3 patients developed and died of complications of PTLD. We recently reported an increased risk for PTLD among IPF-LTRs compared to non-IPF-LTRs. These findings are also consistent with our previous report of increased CMV complications and impaired T cell immunity among short-TL IPF-LTRs and those with short-TL who have not undergone transplantation.^7,17 Taken together, these data highlight an increased susceptibility to herpes viral infections in patients with short TL due to impaired adaptive immunity.

In summary, our findings indicate that IPF-LTRs can be stratified for potential BM dysfunction, the need for hematologic support, and the relative intolerance of standard 3-drug IS following lung transplant. These outcomes are interactive and may increase risk for both allograft outcomes (eg, discontinuation of IS) or extra-pulmonary complications. Our findings support TL measurement to stratify IPF-LTRs together with reflex genetic testing for telomere gene rare variants. Currently, the majority of lung transplant centers do not assess for telomere-mediated disease in their candidates with IPF. Together, short-TL and the identification of telomere rare variants, along with the pretransplant PLT count, may lead to personalized medicine approaches to mitigate severe BM dysfunction (eg, earlier discontinuation of cell cycle inhibitors or antiviral therapies) before profound cytopenias. However, a potential downside to discontinuing IS drugs is the development of ACR and CLAD, which we and others have found to occur similarly or at increased rates in short-TL IPF-LTRs compared with controls. Future studies are needed to determine whether TL/genetic testing and newer transplant therapies with reduced BM toxicity are as effective in preventing these posttransplant complications.

4.1. Study Limitations

We recognized several limitations in our study. Our single-center study cohort, although the largest to date to evaluate hematologic complications in short TL-LTRs, is still relatively small in terms of sample size. Nevertheless, we found significant differences BM failure and requirements for hematologic support that warrant further studies and might guide the clinical approach to patients with IPF with short TL and/or rare variants for the telomere genes.

Indeed, we further recognize variability in the practice and approach to these patients among transplant physicians, even within individual centers. Our data provides statistical evidence of risk for hematologic complications in short TL-LTRs and provides a rationale for studies to further explore BM sparing regimens to mitigate hematologic risk. We should further point out that we did not measure granulocyte TL in all of our patients with IPF, and more studies are needed to elucidate possible discordance with lymphocyte TL. Last, we did not measure TL in our patients with non-IP. Although patients with non-IPF are not enriched for short TL, a small percentage (approximately 1%) of patients with chronic obstructive pulmonary disease/emphysema may harbor short TL and telomere gene rare variants.⁴⁴

Abbreviations:

ACR

acute cellular rejection

BMBx

bone marrow biopsy

BM NGS

bone marrow next-generation sequencing

CI

confidence interval

CLAD

chronic lung allograft dysfunction

CMV

cytomegalovirus

gnomAD

Genome Aggregation Database

Hbg

hemoglobin

IPF

idiopathic pulmonary fibrosis

IPF-LTRs

IPF lung transplant recipients

IS

immunosuppression

Lymph

lymphocyte

OR

odds ratio

PLT

platelets

PTLD

posttransplant lymphoproliferative disorder

TCP

thrombocytopenia

TL

telomere length

Tx

therapy/treatment

UPMC

University of Pittsburgh Medical Center

VUS

variant of unknown significance