Polymorphism of Interleukin-23 Receptor Gene But Not of NOD2/CARD15 Is Associated with Graft-versus-Host Disease after Hematopoietic Stem Cell Transplantation in Children

Article Outline

• Abstract
• Introduction
• Patients and Methods
  • Patients
  • Single Nucleotide Polymorphism (SNP) Analysis
• Results
  • Frequency of Polymorphisms
  • Frequency of GVHD
  • IL-23R Polymorphism and aGVHD
  • NOD2/CARD15 Polymorphisms and aGVHD
  • Polymorphisms of NOD2/CARD15 or IL23R and cGVHD
  • Polymorphisms of NOD2/CARD15 or IL23R and Survival
  • Results in Children up to 16 Years
• Discussion
• Acknowledgments
• References
• Copyright

Abstract

Graft-versus-host disease (GVHD) is a major cause of morbidity and mortality after hematopoietic stem cell transplantation (HSCT). The selection of a suitable donor is the most critical issue in preventing severe GVHD. Recent data suggest that the risk of GVHD does not only depend on human leukocyte antigens (HLA) but also on polymorphisms of genes that influence immune responses. We analyzed the 1142 G>A single-nucleotide polymorphism (SNP) in the interleukin-23 receptor gene (IL23R) and 3 SNPs in the NOD2/CARD15 gene in a cohort of 231 children who underwent allogeneic stem cell transplantation and/or their respective donors. No association was observed between any of the NOD2/CARD15 polymorphisms and GVHD in either donor or recipient. Likewise, the IL23R polymorphism in the recipient was not significantly associated with GVHD. We found a significantly reduced incidence of acute GVHD (aGVHD) grade II-IV in patients who were transplanted from a donor with the IL23R polymorphism (5.0% versus 33.3%; P=.009). There was no case of aGVHD grade III-IV if this polymorphism occurred in the donor. These findings could be particularly relevant for children with inborn metabolic or immunologic disorders who do not benefit from a graft-versus-tumor effect, and therefore, selection of a donor with the IL23R polymorphism might be beneficial.

Key Words: Hematopoietic stem cell transplantation, Graft-versus-host disease, Children, IL-23R, NOD2/CARD15

Back to Article Outline

Introduction 

Hematopoietic stem cell transplantation (HSCT) is a therapeutic option for many patients with progressive malignant hematologic malignancies as well as for patients with a number of inherited metabolic and immunologic disorders. One of the main complications of HSCT is the development of graft-versus-host disease (GVHD) caused by immune competent T cells in the graft. If this immune reaction is directed against leukemic cells, it is called a graft-versus-leukemia (GVL) effect, and can contribute to the therapeutic effect of HSCT.

The risk of GVHD is influenced by the constellation of major and minor histocompatibility antigens (HLA) between the patient and the donor. However, recent data suggest that the risk of GVHD also depends on polymorphisms of genes that influence immune responses [1]. Examples of such genes are NOD2/CARD15 and the interleukin-23 receptor (IL23R) gene. The NOD2/CARD15 gene encodes a protein that functions as a sensor for muramyl dipeptide, which is a cell wall component in bacteria. Among other functions it takes part in the regulation of inflammatory response through intracellular pathways involving nuclear factor kB and IkB kinase [2]. IL23R is expressed in T cells and upon stimulation induces proliferation and interferon (IFN)-γ production 3, 4.

A number of studies analyzed the association between NOD2/CARD15 polymorphisms and outcome of HSCT 5, 6, 7, 8, 9, 10, 11, 12, 13. It was initially found that the presence of at least 1 of 3 polymorphisms of the gene in either donor or recipient was associated with a higher incidence of acute GVHD (aGVHD) and treatment-related mortality (TRM). The higher TRM was not only because of aGVHD, but also because of acute or chronic respiratory failure [5]. Following studies showed that the impact on TRM is particularly pronounced in T cell-depleted transplantation [9], and that it is more prominent in HLA-identical sibling transplantation compared to unrelated donor transplantation [6]. A very strong association exists between the recipient NOD2/CARD15 variants and the development of bronchiolitis obliterans after HSCT [8].

The only available study on IL23R found a strong correlation between the polymorphism 1142 G>A in the donor and a decreased risk of aGVHD [14].

The main clinical potential of these observations is to include the NOD2/CARD15 and IL23R polymorphisms in the donor selection process. All of the above studies analyzed adult patients. Therefore, it is not clear whether the same consequences apply to pediatric patients. Children differ from adult patients in terms of underlying disease, pretreatment, comorbidity, vulnerability of organs, and the maturation stage of their immune system.

We present here the first study on the relevance of NOD2/CARD15 or IL23R polymorphisms for GVHD in pediatric patients.

Back to Article Outline

Patients and Methods 

Patients 

The study population consisted of 231 children or adolescents and their respective donors who underwent stem cell transplantation at the University Children's Hospital Jena between 1982 and 2007.

The criteria for enrolment in this study were availability of cryopreserved pretransplantation blood or bone marrow (BM) cells and the informed consent from the patients and/or their parents. The clinical characteristics of the patients are described in Table 1.

Table 1. Characteristics of Patients and Donors

		Number 231	Percent 100.0
Patient characteristics
Sex	male	149	64.5
	female	82	35.5
Median age (range)		11 (0-29)
GVHD Prophylaxis	MTX+ CsA	148	64.1
	CsA	46	19.9
	MTX	16	6.9
	no prophylaxis	21	9.1
Underlying disease	AML	53	22.9
	ALL/NHL	100	43.3
	MDS	26	11.3
	CML	18	7.8
	SAA	14	6.1
	malignant histiocytosis	1	0.4
	Wiscott-Aldrich Syndrome	1	0.4
	severe combined immunodeficiency	2	0.9
	infantile septic granoulomatosis	2	0.9
	osteopetrosis	3	1.3
	storage diseases	8	3.5
	congenital amegacaryocytosis	1	0.4
	Blackfan-Diamond-Anemia	2	0.9
Donor characteristics
Sex	male	137	59.3
	female	94	40.7
Tissue compatibility	MFD	101	43.7
	MUD	91	39.4
	MMD	25	10.8
	haploidentical	14	6.1
Cell type	bone marrow	167	72.3
	peripheral stem cells	64	27.7

MTX indicates methotrexate; CsA, cyclosporine A; AML, acute myelogenous leukemia; ALL, acute lymphoblastic leukemia; NHL, Non-Hodgkin lymphoma; MDS, myelodysplastic syndrome; CML, chronic myelogenous leukemia; SAA, severe aplastic anemia; MFD, matched family donor; MUD, matched unrelated donor; MMD, mismatched donor; GVHD, graft-versus-host disease.

Allogeneic BM (n=167) or peripheral blood stem cells (PBSCs; n=64) were used for transplantation. The donor was HLA-matched unrelated in 39% (n=91) of transplants and HLA-identical related in 44% (n=101) of transplants. Fourteen (6%) patients where transplanted from haploidentical related donors and 25 (11%) patients from HLA-mismatched donors.

Conditioning regimen was myeloablative (MA) in all cases. Two forms of GVHD prophylaxis predominated: cyclosporine A (CsA) and methotrexate (MTX) in 148 (64%) transplants and cyclosporine A (CsA) alone in 46 (20%) transplants. MTX alone was administered in 17 (7%) transplants. In 20 (9%) transplants, no GVHD prophylaxis was given. These were high-risk patients with leukemia in whom a strong GVL effect was considered indispensable for successful treatment.

aGVHD and chronic GVHD (cGVHD) was diagnosed and graded according to standard criteria 15, 16.

Single Nucleotide Polymorphism (SNP) Analysis 

Samples were obtained from blood or BM aspirates. Mononuclear cells were purified by Ficoll-Hypaque (Sigma, St. Louis, MO) density gradients (density >1.077 g/mL) and cryopreserved in liquid nitrogen. Genomic DNA was isolated using the High Pure PCR Template Preparation Kit (Roche, Mannheim, Germany) according to manufacturer's instructions. The genotype was determined using ready to use TaqMan^® SNP Genotyping Assays and the ABI PrismTM 7700 Sequence Detector (Applied Biosystems, Foster City, CA) according to manufacturers instructions. The Assay IDs were: IL-23R-1142 G>A (SNP ID: rs11209026): C_1272298_10; NOD2/CARD15 SNP8 (SNP ID: rs2066844): C_11717468_20; and NOD2/CARD15 SNP12 (SNP ID: rs2066845): C_11717466_20. Details about these assays are available at www.appliedbiosystems.com.

No ready-to-use assay was available for NOD2/CARD15 SNP13 (SNP ID: rs2066847). Therefore, a Custom TaqMan^® SNP Genotyping Assay was provided by Applied Biosystems:

In 4 randomly selected samples, the genotype of IL23R was also analyzed by direct sequencing. In each sample, the TaqMan result was verified.

We used primer IL23R 1 and IL23R 2 with the following sequences:

Polymerase chain reaction (PCR) was performed using 500 ng DNA and 1.25 units of Taq DNA polymerase (Roche), in a 50-μL reaction, which contained 10 mmol/L Tris-HCl (pH 8.8), 200 μmol/L each dATP, dGTP, dCTP, dTTP, and 0.5 μmol/L each primer. Samples were subjected to an initial 5 min denaturation at 94°C, followed by 40 cycles of 1-min denaturation at 94°C, 1-min annealing at 55°C, and 1-min extension at 72°C in a Peltier Thermal Cycler, model PTC-200 (MG Research, Watertown, MA). A final 10-min extension was performed at the end of the PCR amplification. Aliquots of the PCR (10 μL) were analyzed on a 3% Nusieve GTG agarose gel (FMC, Rockland, ME) with ethidium bromide (0.05 μg/mL) staining. To purify the PCR product for sequencing, amplified DNA was run on a 2% Nusieve agarose gel (FMC) and the ethidium bromide-stained band was located and excised. Agarose was removed using QIAquick Gel Extraction Kit (Qiagen, Chatsworth, CA) according to the manufacturer's instructions. The same PCR primers were also used for sequencing of the amplicons. Sequencing reactions using DyeNamic ET Terminators were performed according to the cycle sequencing protocol (Pharmacia Life Sciences, Freiburg, Germany). The products were ethanol precipitated and analyzed on a Prism 310 Genetic Analyzer of ABI. The sequences were aligned manually.

Back to Article Outline

Results 

Frequency of Polymorphisms 

Two hundred thirty-one donor/patient pairs were analyzed. Cell samples from the patient were available in 217 cases. Samples from the donor were available in 181 cases. Cells from donor and patient were available in 168 pairs. The frequency of the polymorphisms in patients and donors were similar to previously published data 5, 6, 7, 8, 9, 10, 11, 12, 13, 14.

The IL-23R-1142 G>A allele was found in 17 (7.8%) patients and 20 (11%) donors. None of the individuals was homozygous for this SNP. In 5 cases, the polymorphism was found in donor and patient.

The frequencies of the NOD2/CARD15 polymorphisms were: SNP8: 16 (7.4%) patients and 16 (8.8%) donors, 1 patient was homozygous; SNP12: 5 (2.3%) patients and 8 (4.4%) donors, none homozygous; SNP13: 21 (9.6%) patients and 12 (6.6%) donors, 2 patients were homozygous.

At least 1 of the NOD2/CARD15 polymorphisms was found in 41 (18.9%) patients and 33 (19.6%) donors. In 12 cases, at least 1 polymorphism was found in the donor and in the patient.

Frequency of GVHD 

aGVHD was observed in 107 patients (26%). The involved organs were skin (n=82), gut (n=50), and liver (n=33). cGVHD was observed in 51 patients (22%). The most frequently involved organs were skin (n=33), oral mucosa (n=23), liver (n=21), sicca syndrome (n=15), lung (n=9), and gut (n=10).

IL-23R Polymorphism and aGVHD 

Significantly lower levels of aGVHD were associated with the presence of the IL23R-1142 G>A allele in the donor (Spearman correlation P=.024; Figure 1). Only 1 of 20 patients (5%) who were transplanted from a donor with this polymorphism developed aGVHD higher than grade I (Table 2), whereas 53 of 161 patients (33.3%) who had a donor without this SNP developed aGVHD higher than grade I (Fisher's Exact Test, P=.009).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1 

  The IL23R-1142 G>A allele in the donor is associated with less severe and less frequent GVHD (Spearman correlation P=.024). Donor IL23R-1142 G>A positive, n=20; donor IL23R-1142 G>A negative, n=161.

Table 2. Acute GVHD and Polymorphisms of IL23R and NOD2/CARD15

GVHD indicates graft-versus-host disease; SNP, single nucleotide polymorphism.

The association between aGVHD and the IL23R-1142 G>A allele in the donor was not restricted to a specific tissue but equally pronounced for skin, gut, and liver.

To test whether this correlation was because of other differences between those patient groups, we compared the main patient and donor characteristics of transplants with an IL-23R-1142 G>A-positive donor and transplants from a donor without this polymorphism. As shown in Table 3, no significant differences were observed.

Table 3. Comparison of Transplants with IL23R-1142 G>A in the Donor Versus Wild-Type of IL23R in the Donor

		IL23R-1142 G>A in the Donor				P-Value
		Yes		No (wt)
		Number	Percent	Number	Percent
		20	100	161	100
Patient characteristics
Sex	male	12	60	102	63	.89
	female	8	40	59	37
Median age (range)		12 (3-19)		10 (0-26)		.1
GVHD Prophylaxis	MTX+ CsA	11	55	104	65	.32
	CsA	8	40	38	24
	no prophylaxis	1	5	19	12
Underlying disease	AML	3	15	37	23	.8
	ALL/NHL	11	55	68	42
	MDS	4	20	22	14
	CML	1	5	11	7
	SAA	0	0	10	6
	other	1	5	13	8
Donor characteristics
Sex	male	11	55	98	61	.64
	female	9	45	63	39
Tissue compatibility	MFD	5	25	56	35	.86
	MUD	11	55	74	46
	MMD	3	15	20	12
	haploidentical	1	5	11	7
Cell type	bone marrow	13	65	105	65	1.0
	peripheral stem cells	7	35	56	35

MTX indicates methotrexate; CsA, cyclosporine A; AML, acute myelogenous leukemia; ALL, acute lymphoblastic leukemia; NHL, Non-Hodgkin lymphoma; MDS, myelodysplastic syndrome; CML, chronic myelogenous leukemia; SAA, severe aplastic anemia; MFD, matched family donor; MUD, matched unrelated donor; MMD, mismatched donor; GVHD, graft-versus-host disease.

The frequency or severity of aGVHD was not associated with the occurrence of the IL23R-1142 G>A allele in the patient. The frequency of aGVHD grade II-IV was 18% (3 of 17) if the polymorphism was present versus 29% (58 of 200) if it was not.

In 5 of the donor-patient pairs the IL23R-1142 G>A allele was present in both individuals and no aGVHD was observed.

NOD2/CARD15 Polymorphisms and aGVHD 

The association of each of the 3 NOD2/CARD15 polymorphisms with aGVHD was analyzed in 2 ways. First, the Spearman correlation coefficient was calculated, and second, the Fisher's Exact Test was calculated for aGVHD grade 0 or I versus aGVHD grade II to IV (Table 2).

We did not observe a trend for an association of any of the 3 polymorphisms with the occurrence of aGVHD (Table 2).

The result was the same in a combined analysis of the 3 polymorphisms. At least 1 of the 3 polymorphisms was observed in 41 (18.9%) patients and 33 (19.6%) donors, but neither group differed significantly in terms of frequency or severity of aGVHD.

Polymorphisms of NOD2/CARD15 or IL23R and cGVHD 

We first analyzed the association of the 4 polymorphisms with overall cGVHD, independent of the involved organs. The genotypes of donors and patients were analyzed separately. Neither the IL23R polymorphism nor any of the 3 NOD2/CARD15 polymorphisms were significantly associated with the frequency of cGVHD. The result was the same in a combined analysis of the 3 NOD2/CARD15 polymorphisms, that is, all donors who had at least 1 of the polymorphisms and all patients, respectively.

Because the rate of cGVHD in peripheral cells versus marrow as a donor source is higher, these analyses were repeated with the type of transplant as a stratification variable. The results remained insignificant.

In a second step we performed the same analyses but separately for each of the most frequently involved organs (skin, oral mucosa, liver, saliva glands, lung, and gut). No significant associations were observed, but the subgroups are too small for reliable conclusions.

Polymorphisms of NOD2/CARD15 or IL23R and Survival 

The influence of the IL23R polymorphism and each of the 3 NOD2/CARD15 polymorphisms on overall survival (OS) was analyzed by Kaplan-Maier statistics. No significant association between survival and any of the polymorphism in either donor or patient was found (Figure 2).

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2 

  Polymorphisms of IL23R and NOD2/CARD15 are not associated with OS after stem cell transplantation in children. (A) Patient IL23R-1142 G>A positive (n=17) versus patient IL23R-1142 G>A negative (n=200). (B) Donor IL23R-1142 G>A positive (n=20) versus donor IL23R-1142 G>A negative (n=161). (C) Patient positive for any of the 3 SNPs in NOD2/CARD15 (n=41) versus patient negative for all 3 SNPs (n=176). (D) Donor positive for any of the 3 SNPs in NOD2/CARD15 (n=33) versus patient negative for all 3 SNPs (n=148).

Results in Children up to 16 Years 

Our main goal was to analyze the relevance of NOD2/CARD15 and IL23 in pediatric patients. Initially, we included all patients who were transplanted in our children's hospital. However, some of the patients were young adults, the oldest being 29 years at the time of transplantation. Therefore, all analyzes were repeated with a restriction to patients aged 16 years or younger (n=199).

The correlation between the presence of the IL23R-1142 G>A allele in the donor and lower levels of aGVHD was also significant in this subgroup (Spearman correlation coefficient P=.024). In patients with aGVHD grade 0 or I, this polymorphism was found in 32% (43/135) of the donors. In patients with aGVHD grade II to IV it was found in 0% (0/17) of the donors (Fisher's Exact Test, P=.003).

All other calculations that were performed in the total group of patients remained insignificant in this subgroup.

Back to Article Outline

Discussion 

We here report the first data on the clinical relevance of NOD2/CARD15 and IL23R polymorphisms for GVHD in children. Our results show that the polymorphism IL23R-1142 G>A in the donor is associated with a significantly lower frequency and severity of aGVHD in the patient (Figure 1). This is consistent with the only available study on IL23R-1142 G>A and aGVHD in adult patients [14].

The main limitation of our analysis is the strong heterogeneity of the study population (Table 1). However, the main characteristics of patients and donors did not significantly differ between transplants with wild-type versus IL23R-1142 G>A-positive donors (Table 3). Therefore, it is unlikely that this is a false positive finding.

The IL-23R polymorphism was not associated with superior OS. The reason could be that patients with less GVHD also experience less GVL effect, and hence, have a higher risk of relapse. In our cohort of leukemia patients the number of relapses was not higher after transplantation from a donor with the IL-23R polymorphism. A detailed analysis of this hypothesis will require prospective studies with larger numbers of patients and less heterogeneity of diseases.

Patients with inborn metabolic or immunologic disorders represent a growing proportion of pediatric stem cell transplantations. These patients do not need a GVL effect. Therefore, they would probably benefit from a donor with the IL23R polymorphism.

A number of studies have analyzed the relevance of NOD2/CARD15 polymorphisms for GVHD in adult patients 5, 6, 7, 8, 9, 10, 11, 12, 13. Initial reports suggested that the presence of any of the 3 polymorphisms (SNP8, SNP12, and SNP13) in donor or patient was associated with more severe aGVHD and TRM [5]. Most of the following studies confirmed the higher incidence of TRM but produced controversial results concerning the risk of aGVHD 6, 7, 8, 9, 10, 11, 12. Sairafi et al. [13] observed neither an association with aGVHD nor with TRM in a large single center study. Hildebrandt et al. [8] demonstrated a strong association between the recipient NOD2/CARD15 variants and the development of bronchiolitis obliterans after HSCT in adult patients.

In our pediatric patient group, we found no correlation of the NOD2/CARD15 polymorphisms in either donor or patient with aGVHD, cGVHD, or OS. The result was the same when only TRM was analyzed. It is possible that we failed to observe such an association because of the heterogeneity in the patient population. However, because we did not even see a trend for more severe aGVHD, cGVHD or a higher mortality in association with SNPs of NOD2/CARD15, it is unlikely that the genotype of NOD2/CARD15 is of major relevance for the outcome of HSCT in children.

Back to Article Outline

Acknowledgments 

Financial disclosure: The authors have nothing to disclose.

Back to Article Outline

References 

1. Dickinson AM, Middleton PG, Rocha V, Gluckman E, Holler E Eurobank members. Genetic polymorphisms predicting the outcome of bone marrow transplants. Br J Haematol. 2004;127:479–490
   • View In Article
   • MEDLINE
   • CrossRef
2. Eckmann L, Karin M. NOD2 and Crohn's disease: loss or gain of function?. Immunity. 2005;22:661–667
   • View In Article
   • MEDLINE
   • CrossRef
3. Parham C, Chirica M, Timans J, et al. A receptor for the heterodimeric cytokine IL-23 is composed of IL-12Rbeta1 and a novel cytokine receptor subunit, IL-23R. J Immunol. 2002;168:5699–5708
   • View In Article
   • MEDLINE
4. van den Eijnden S, Goriely S, De Wit D, Willems F, Goldman M. IL-23 up-regulates IL-10 and induces IL-17 synthesis by polyclonally activated naive T cells in human. Eur J Immunol. 2005;35:469–475
   • View In Article
   • MEDLINE
   • CrossRef
5. Holler E, Rogler G, Herfarth H, et al. Both donor and recipient NOD2/CARD15 mutations associate with transplant-related mortality and GvHD following allogeneic stem cell transplantation. Blood. 2004;104:889–894
   • View In Article
   • MEDLINE
   • CrossRef
6. Holler E, Rogler G, Brenmoehl J, et al. Prognostic significance of NOD2/CARD15 variants in HLA-identical sibling hematopoietic stem cell transplantation: effect on long-term outcome is confirmed in 2 independent cohorts and may be modulated by the type of gastrointestinal decontamination. Blood. 2006;107:4189–4193
   • View In Article
   • MEDLINE
   • CrossRef
7. Holler E, Rogler G, Brenmoehl J, et al. The role of genetic variants of NOD2/CARD15, a receptor of the innate immune system, in GvHD and complications following related and unrelated donor haematopoietic stem cell transplantation. Int J Immunogenet. 2008;35:381–384
   • View In Article
   • CrossRef
8. Hildebrandt GC, Granell M, Urbano-Ispizua A, et al. Recipient NOD2/CARD15 variants: a novel independent risk factor for the development of bronchiolitis obliterans after allogeneic stem cell transplantation. Biol Blood Marrow Transplant. 2008;14:67–74
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (347 KB)
   • CrossRef
9. Granell M, Urbano-Ispizua A, Aróstegui JI. Effect of NOD2/CARD15 variants in T-cell depleted allogeneic stem cell transplantation. Haematologica. 2006;91:1372–1376
   • View In Article
10. Elmaagacli AH, Koldehoff M, Hindahl H, et al. Mutations in innate immune system NOD2/CARD 15 and TLR-4 (Thr399Ile) genes influence the risk for severe acute graft-versus-host disease in patients who underwent an allogeneic transplantation. Transplantation. 2006;81:247–254
    • View In Article
    • MEDLINE
    • CrossRef
11. Madrigal A, Shaw BE. Immunogenetic factors in donors and patients that affect the outcome of hematopoietic stem cell transplantation. Blood Cells Mol Dis. 2008;40:40–43
    • View In Article
    • CrossRef
12. Mayor NP, Shaw BE, Hughes DA, et al. Single nucleotide polymorphisms in the NOD2/CARD15 gene are associated with an increased risk of relapse and death for patients with acute leukemia after hematopoietic stem-cell transplantation with unrelated donors. J Clin Oncol. 2007;25:4262–4269
    • View In Article
    • CrossRef
13. Sairafi D, Uzunel M, Remberger M, Ringdén O, Mattsson J. No impact of NOD2/CARD15 on outcome after SCT. Bone Marrow Transplant. 2008;41:961–964
    • View In Article
    • CrossRef
14. Elmaagacli AH, Koldehoff M, Landt O, Beelen DW. Relation of an interleukin-23 receptor gene polymorphism to graft-versus-host disease after hematopoietic-cell transplantation. Bone Marrow Transplant. 2008;41:821–826
    • View In Article
    • CrossRef
15. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus conference on acute GVHD grading. Bone Marrow Transplant. 1995;15:825–828
    • View In Article
    • MEDLINE
16. Sullivan KM, Agura E, Anasetti C, et al. Chronic graft-versus-host disease and other late complications of bone marrow transplantation. Semin Hematol. 1991;28:250–259
    • View In Article
    • MEDLINE