Abstract
Background and Objective
Chronic graft versus host disease (GVHD)‐associated myositis targeting skeletal muscle is a relatively rare but potentially debilitating complication following allogeneic hematopoietic stem cell transplantation (HSCT). We reviewed the clinicopathological features of GVHD‐associated myositis among patients receiving allogeneic HSCT to elucidate the cellular pathogenesis.
Methods
We retrospectively reviewed clinical data and muscle biopsy results from 17 consecutive patients diagnosed with GVHD‐associated myositis at our institution between 1995 and 2019. Immunostaining findings of GVHD‐associated myositis were compared to those of patients with anti‐tRNA‐synthetase antibody‐associated myopathy (ASM) (n = 13) and dermatomyositis (DM) (n = 12).
Results
The majority of patients with GVHD‐associated myositis showed subacute or chronic progression of mild to moderate limb weakness together with elevated serum creatine kinase. These patients also exhibited mild C‐reactive protein elevation but were negative for myositis‐related autoantibodies. Programmed death‐1 (PD‐1)‐positive cells were observed in muscle interstitium adjacent to myofibers expressing human leukocyte antigen (HLA)‐DR. The interstitium was also HLA‐DR‐positive, similar to biopsy samples from ASM patients but not DM patients. The proportions of HLA‐DR‐positive muscle fibers and PD‐1‐positive interstitial cells were significantly higher in GVHD and ASM samples than DM samples. The PD‐1‐positive cells were mostly CD‐8‐positive lymphocytes.
Discussion
GVHD‐associated myositis is characterized by HLA‐DR‐positive myofibers and infiltration of PD‐1‐positive lymphocytes. These features distinguish GVHD‐associated myositis from DM but not from ASM.
Introduction
Graft versus host disease (GVHD) is a potentially serious complication of organ transplantation in which immune cells from the donor organ induce an immune response in the host. 1 While GVHD may occur after many different allogeneic organ transplantation procedures, it is particularly severe following hematopoietic stem cell transplantation (HSCT) and blood transfusions because donor immune tissues are directly implanted in the host. 1 , 2 The ensuing GVHD is classified as either acute or chronic depending on the clinical features rather than the temporal relationship with transplantation. 1
In GVHD following HSCT, there may be varying degrees of damage to multiple organs, including the skin and viscera, while damage to muscle and nervous tissues is relatively rare. 3 , 4 For instance, among Japanese allogeneic stem cell transplantation recipients, the reported 5‐year cumulative incidence of myositis caused by chronic GVHD was only 0.54%. 5 GVHD of the muscles, termed GVHD‐associated myositis, is difficult to distinguish from other myositis based on clinicopathological features, 6 and there is currently insufficient knowledge of the unique pathological characteristics of GVHD‐associated myositis for timely diagnosis and treatment.
Major histocompatibility complex (MHC) class II antigen, human leukocyte antigen (HLA)‐DR, and MHC class I antigen HLA‐ABC are used as histopathological markers of myositis. 7 , 8 While HLA‐ABC is a high‐sensitivity marker in biopsied muscle specimens from patients with inflammatory myopathies, specificity is low because HLA‐ABC is also expressed in noninflammatory myopathies and neurogenic diseases. 8 By contrast, HLA‐DR has low sensitivity but has greater specificity as a marker for all major inflammatory myopathies, especially inclusion body myositis. 9 The diagnostic utility of myofiber HLA‐DR expression in anti‐tRNA‐synthetase (ARS) antibody‐associated myopathy (ASM) has also been reported. 10 However, except for a case report on the expression of non‐necrotic fibers in perifascicular regions, 11 the detailed profile of myofiber HLA‐DR expression in GVHD‐associated myositis is largely unknown. 5 , 6 , 12
Programmed death‐1 (PD‐1) is expressed on activated CD4+ and CD8+ T cells, B cells, monocytes, dendritic cells, and natural killer cells, and interacts with its cognate ligands programmed death‐ligand 1 (PD‐L1) and programmed death‐ligand 2 (PD‐L2). 13 As binding of PD‐1 to PD‐L1 provides inhibitory signals that regulate T‐cell activation, PD‐L1 expressed on non‐hematopoietic cells limits effector T‐cell responses, thereby protecting tissues from immune‐mediated damage. 14 Several studies have reported that the PD‐L1 pathway is activated by interferon‐gamma (IFN‐γ) and induces apoptosis of infiltrating alloreactive donor T cells in lung and liver tissues of GVHD patients. 15 , 16 , 17 , 18 In addition, IFN‐γ may also induce the expression of HLA‐DR molecules in muscle fibers of patients with inflammatory myositis of unknown cause without skin rash. 19 However, the functions of PD‐1 and IFN‐γ and their associations with HLA‐DR have yet to be elucidated in GVHD‐associated myositis. Here, we investigated the clinical and histopathological features of GVHD‐associated myositis cases following allogeneic HSCT to elucidate the molecular and cellular pathogenesis.
Methods
Patients and clinical data
We retrospectively analyzed the medical records and muscle biopsy results from 17 consecutive patients diagnosed with GVHD‐associated myositis following allogeneic HSCT at our hospital between 1995 and 2019. Clinicodemographic items gathered for analysis included sex, age, reason for allogeneic stem cell transplantation (underlying disease), delay between transplantation and emergence of GVHD‐associated myositis, initial symptoms, and blood chemistry parameters such as serum creatinine kinase (CK), together with electromyogram (EMG) and muscle imaging results. A muscle biopsy sample was taken from the biceps or quadriceps of all patients. Four of these patients (No. 11, 13, 15, 17) with skin symptoms also received skin biopsy. All patients received clinical and neurological assessments, routine blood tests, and urinalysis. Neurological assessments were performed by at least two neurologists for each patient. The strength of each target muscle was graded using the Medical Research Council scale for muscle strength. Electromyography was performed with concentric needles in the deltoid, biceps brachii, first dorsal interosseous, thenar, lumbar paraspinal, quadriceps, and tibialis anterior muscles. The myopathic changes analyzed from EMG recordings included motor units with low amplitude, short duration, polyphasic waveforms, and early recruitment at muscle contraction.
Immunohistochemistry
Muscle biopsy was performed in all patients as described previously. 19 , 20 , 21 , 22 , 23 All specimens were obtained by open biopsy, snap‐frozen in isopentane, chilled on dry ice, and preserved at −80°C until further analyses. Specimens were cut by routine methods into 10‐μm cryostat sections for hematoxylin/eosin (H&E), modified Gomori's trichrome, nicotinamide adenine dinucleotide (NADH) tetrazolium reductase, adenosine triphosphatase (ATPase), and alkaline phosphatase (ALP) staining. Sections except seven patients were also immunostained with mouse monoclonal immunoglobulin G antibodies against HLA‐ABC (1:4000, BD Pharmingen, San Diego, CA, G46‐2.6), HLA‐DR (1:300, BD Pharmingen, L243), CD4 (1:150, Dako, Glostrup, Denmark, 4B12), CD8 (1:150, Dako, C8/144B), CD68 (1:300, Dako, KP1), membrane attack complex (MAC) (1:100, Abcam, Cambridge, UK, C5b‐9 [aE11]), myxovirus resistance protein (MxA) (1:1000, Merck, Kenilworth, NJ, M143), and PD‐1 (1:100, Abcam, NAT105) or with a rabbit monoclonal antibody against PD‐L1 (1:10,000, Cell Signaling Technology, Danvers, MA, E1L3N). Immunolabeling was visualized using the Vectastain ABC Kit (Vector Laboratories, Burlingame, CA). Briefly, cryostat sections were fixed in cold acetone, air dried, blocked with the appropriate serum for the secondary antibody source, immunostained, and then treated with Vectastain ABC Kit reagents according to the manufacturer's instructions. Antibody binding was visualized by 3,3′‐diaminobenzidine tetrahydrochloride (DAB). Each muscle biopsy specimen was coded and analyzed independently by two neurologists trained in muscle pathology (TK and SN) blinded to diagnosis and clinical status using an Olympus microscope.
To evaluate the disease specificity of findings from GVHD‐associated myositis patients, we also conducted immunostaining of specimens from 13 patients with ASM (male, 4; female, 9; mean age, 54.8 ± 8.32 years) and 12 patients with dermatomyositis (DM; male, 5; female, 7; mean age, 61.7 ± 9.18 years) obtained at our hospital between October 2016 and December 2019. All patients with DM were positive for antibodies against transcriptional intermediary factor 1‐γ (TIF1‐γ) or melanoma differentiation‐associated gene 5 (MDA5). Because the pathology of dermatomyositis has been reported to differ depending on the subtype of autoantibody, 24 we excluded a case of anti‐Mi‐2 antibody who was the only patient we diagnosed during the same period.
We counted the total number of myofibers and HLA‐DR‐positive myofibers in four 100× fields of view randomly selected from prespecified areas of interest for each specimen and calculated the proportion (%) of HLA‐DR‐positive myofibers. The number of PD‐1‐positive cells in the muscle bundle and the muscle bundle area were also measured in four 100× fields of view for each specimen and the PD‐1‐positive cell density calculated in number of cells/mm2. At least two specialists in muscle pathology similar to those described above conducted imaging and immunostaining measurements using ImageJ (https://imagej.nih.gov) with the Cell Counter plugin. Additionally, CD68 and HLA‐DR immunohistochemistry and ALP staining were also performed on the serial sections from ASM and GVHD‐associated myopathy. To confirm the association with macrophage infiltration, a case of immune‐mediated necrotizing myopathy (IMNM) positive for anti‐SRP antibodies was similarly examined for CD68, HLA‐DR, and ALP, as IMNM is pathologically characterized by intense macrophage infiltration. 25
We divided the GVHD cases into two groups, those who were treated with and without immunosuppressants or steroids, and compared the density of PD‐1 and the positivity rate of HLA‐DR between them.
Dual immunofluorescence histochemistry
Double immunofluorescence staining was performed after fixation and blocking as above. Briefly, sections were incubated overnight at 4°C in a mixture of the following primary antibodies (from the same suppliers and with the same dilutions as above or as indicated): PD‐1, CD3 (1:1000, Abcam, Cambridge, UK, SP‐7), CD4, CD8, PD‐L1, and HLA‐DR. Sections were then washed, incubated with a mixture of secondary antibodies for 2 h at room temperature, washed again, wet‐mounted, and stored at 4°C.
Statistical analysis
Statistical analysis was performed using R software (www.Rproject.org). In cases of non‐normal distribution, quantitative variables of two groups were compared using Mann–Whitney U‐test. After testing for normality with the Shapiro–Wilk test, Kruskal–Wallis with Steel‐Dwass post hoc tests were performed using JMP® 9 (SAS Institute Inc., Cary, NC, USA) to compare two or more groups that were not normally distributed. We conducted measurement at a confidence level of 95%. All p‐values were two‐tailed and statistical significance is defined as p < 0.05.
Standard protocol approvals, registrations, and patient consents
This study was conducted in accordance with the tenets of the Declaration of Helsinki, the Ethics Guidelines for Human Genome/Gene Analysis Research, and the Ethical Guidelines for Medical and Health Research Involving Human Subjects endorsed by the Japanese government. The study protocol was approved by the Ethics Review Committee of Nagoya University Graduate School of Medicine (No. 2016‐0157). All participants were informed of the purpose of the study and provided written informed consent. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
Results
Clinical features
The clinical characteristics of GVHD‐associated myositis are summarized in Tables 1 and 2. In all cases except one, GVHD was diagnosed from lesions in other organs before muscle biopsy was performed, including skin lesions observed in eleven patients, pulmonary lesions in two patients, and hepatic lesions in one patient. Among patients with skin lesions, seven developed skin sclerosis and two erythema or hyperpigmentation, while there were no detailed descriptions for the remaining two patients. In all seven cases, skin sclerosis was widely distributed on the limbs or trunk. No patients had myositis‐specific findings such as Gottron's papule, mechanic's hands, or heliotrope rush. Skin biopsy was performed in four patients (No. 11, 13, 15, and 17), but the histopathological findings were nonspecific.
Table 1.
Characteristics of patients with GVHD‐associated myositis.
| Patient no. | Patient gender | Age at transplantation | Diagnosis | Donor | Donor source | GVHD prophylaxis | Acute GVHD | Chronic GVHD |
|---|---|---|---|---|---|---|---|---|
| 1 | F | 24 | ALL | Sister | BMT | – | – | – |
| 2 | M | 44 | CTCL | Unrelated | BMT | CyA | – | Skin, mouth, lung |
| 3 | M | 35 | CML |
Unrelated |
BMT | PSL, CyA | Skin, meningitis | Skin, lung, cystitis |
| 4 | F | 20 | AML | Sister | BMT | CyA | – | Mouth, GI |
| 5 | F | 43 | AML | Sister | BMT | CyA | – | Mouth |
| 6 | F | 29 | ML | Unknown | BMT | PSL, TAC | – | Skin |
| 7 | F | 31 | AML | Unrelated | BMT | – | – | Skin |
| 8 | M | 30 | AML | Brother | PBSCT | – | – | Skin |
| 9 | F | 21 | AML | Unrelated | CBT | TAC, L‐PAM | Skin | – |
| 10 | F | 47 | AML | Sister | PBSCT | PSL | – | Skin |
| 11 | F | 46 | ML | Sister | PBCST | – | – | Skin |
| 12 | M | 44 | MDS | Unrelated | BMT | TAC | Skin | – |
| 13 | M | 62 | MDS | Sister | BMT | PSL | – | Skin, lung, mouth |
| 14 | M | 55 | MM | Unrelated | BMT | PSL, TAC | Skin | Liver |
| 15 | M | 58 | AML | Sister | PBSCT | PSL, TAC | Skin, mouth, GI | Skin |
| 16 | M | 68 | AML/MLD | Unrelated | BMT | PSL, MMF | GI | – |
| 17 | M | 37 | Ph+ALL | Unrelated | BMT | – | – | Skin |
ALL, acute lymphocytic leukemia; AML, acute myelogenous leukemia; AML/MLD, acute myeloid leukemia with multilineage dysplasia; BMT, bone marrow transplantation; CBT, cord blood transplantation; CML, chronic myelogenous leukemia; CTCL, cutaneous T‐cell lymphoma; CyA, cyclosporine A; F, female; GI, gastrointestinal symptoms; GVHD, graft versus host disease; L‐PAM, melphalan; M, male; MDS, myelodysplastic syndromes; ML, malignant lymphoma; MM, multiple myeloma; MMF, mycophenolate mofetil; PBSCT, peripheral blood stem cell transplant; Ph+ALL, philadelphia chromosome‐positive acute lymphoblastic leukemia; PSL, prednisolone; TAC, tacrolimus.
Table 2.
Muscle‐related findings of GVHD‐associated myositis.
| No. | Age at onset of myositis | Time from transplantation (months) | Myalgia | Muscle weakness | Severity a | CK (U/I) | CRP (mg/dL) | Myositis‐specific autoantibody | |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 25 | 13 | − | Neck and proximal upper limbs | Symmetrical | Mild | 1812 | 0.2 | ND |
| 2 | 45 | 16 | + | Proximal four limbs | Asymmetrical | Moderate | 7230 | 1.51 | ND |
| 3 | 37 | 26 | + | Proximal and distal four limbs | Asymmetrical | Mild | 1257 | 5.8 | ND |
| 4 | 20 | 7 | + | Neck and proximal four limbs | Asymmetrical | Severe | 1780 | 7.1 | Jo‐1 (−) |
| 5 | 45 | 26 | − | Proximal four limbs | Symmetrical | Severe | 1902 | 0 | ND |
| 6 | 31 | 15 | − | Proximal lower limbs | Asymmetrical | Moderate | 29 | 1.07 | Jo‐1 (−) |
| 7 | 33 | 32 | − | Proximal lower limbs | Asymmetrical | Moderate | 57 | 0.14 | Jo‐1 (−) |
| 8 | 31 | 14 | − | Proximal and distal lower limbs | Symmetrical | Moderate | 63 | ND | ND |
| 9 | 21 | 2 | + | − | − | Severe | 8640 | ND | Jo‐1 (−) |
| 10 | 48 | 11 | − | Proximal lower limbs | Symmetrical | Severe | 20 | 0.04 | Jo‐1 (−) |
| 11 | 49 | 32 | − | Neck and proximal upper limbs | Symmetrical | Moderate | 2812 | 1.17 | Jo‐1 (−) |
| 12 | 44 | 10 | + | − | − | Moderate | 1747 | 8.4 | Jo‐1 (−) |
| 13 | 64 | 25 | − | Proximal lower limbs | Symmetrical | Moderate | 71 | 0.11 | Jo‐1 (−), MDA5 (−), Mi‐2 (−), TIF‐γ (−) |
| 14 | 55 | 11 | + | − | − | Mild | 1883 | 0 | ND |
| 15 | 60 | 20 | + | Proximal lower limbs | Symmetrical | Moderate | 23 | 0.12 | ARS (−), MDA5 (−), Mi‐2 (−), TIF1‐γ (−), SRP (−), Ku (−) |
| 16 | 71 | 23 | + | − | − | Mild | 572 | 8.39 | ARS (−) |
| 17 | 39 | 25 | + | − | − | Mild | 4537 | 10.9 | ARS (−) |
ARS, aminoacyl tRNA synthetase; CK, creatinine kinase; CRP, c reactive protein; GVHD, graft versus host disease; MDA5, melanoma differentiation‐associated gene 5; ND, not determined; SRP, signal recognition particle; TIF1‐γ, transcriptional intermediary factor1‐γ.
Mild, able to look after own affairs without assistance; Moderate, require some help, but able to walk unassisted; Severe, unable to walk unassisted.
The majority of patients with GVHD‐associated myositis exhibited subacute or chronic progression of mild to moderate limb weakness (Table 2). The mean age at the time of GVHD‐related myositis onset was 42.2 ± 7.6 years and the mean period from allogeneic HSCT until myositis onset was 18.1 ± 4.5 months. All patients had myalgia or muscle weakness as the initial symptom. Mean serum CK was 2026 ± 2553 IU/L (range, 20–8640 IU/L) and was above the normal range in 11 of 17 patients. Antibodies to antigens specific for myositis, such as ARS, MDA5, Mi‐2, TIF1‐γ, signal recognition particle (SRP), or Ku, were not detected in any patient examined. The clinical course was benign in all patients, and myositis was not a direct cause of death in any case. However, one (patient No. 11) required artificial ventilation due to respiratory muscle weakness.
Histopathology
The pathological findings for each case are shown in Table 3. Relatively few necrotic fibers and infiltrating cells were observed in 11 cases (Fig. 1A), while in the remaining six cases, a large number of infiltrating cells and well‐defined regions of connective tissue proliferation were observed in some muscle bundles (Fig. 1B). Modified Gomori's trichrome staining showed similar results to H&E stains (Fig. 1C). Samples from all patients demonstrated widely distributed HLA‐ABC expression (Fig. 1D) and the majority of samples also expressed HLA‐DR, although HLA‐DR‐positive myofibers were less numerous than HLA‐ABC‐positive fibers (Fig. 1E). HLA‐DR‐positive myofibers were arranged in multifocal or perifascicular patterns. The distribution of CD68‐positive cells on serial sections was nearly consistent with HLA‐DR, suggesting that the majority of HLA‐DR staining in the increased interstitium was probably due to macrophages (Fig. 1F).
Table 3.
Pathological findings of GVHD‐associated myositis.
| No. | Necrosis and regeneration | HLA‐ABC a | HLA‐DR a | PD‐1 b | PD‐L1 b | MxA b | MAC b | CD4 b | CD8 b | CD68 b |
|---|---|---|---|---|---|---|---|---|---|---|
| 1 | + | +++ | + | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 2 | + | +++ | − | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 3 | + | ++ | + | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 4 | + | ++ | ++ | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 5 | + | ++ | ++ | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 6 | + | +++ | ++ | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 7 | + | ++ | ++ | + | − | − | − | + | + | + |
| 8 | + | ++ | ++ | + | − | − | − | − | − | + |
| 9 | + | ++ | ++ | N/A | N/A | N/A | N/A | N/A | N/A | N/A |
| 10 | − | +++ | ++ | + | − | − | − | − | + | + |
| 11 | + | +++ | ++ | ++ | + | − c | − | ++ | +++ | +++ |
| 12 | + | +++ | ++ | + | + | − c | − | +++ | +++ | +++ |
| 13 | + | +++ | ++ | + | − | − | − | + | + | + |
| 14 | + | +++ | ++ | + | + | − | − | + | + | + |
| 15 | + | +++ | ++ | + | − | − | − | + | ++ | ++ |
| 16 | + | +++ | ++ | ++ | + | − | − | + | + | ++ |
| 17 | + | ++ | ++ | + | + | − c | − | + | + | + |
GVHD, graft versus host disease; HLA, human leukocyte antigen; MxA, myxovirus resistance protein; MAC, membrane attack complex; N/A, not available; PD‐1, programmed death‐1; PD‐L1, programmed death‐ligand 1.
The number of plus signs indicates approximately the degree of positive myofibers. (+, single or scattered; ++, continuous but not diffuse; +++, diffuse).
The number of plus signs indicates the approximate degree of cell infiltration into endomysium. (+, <30/mm2; ++, 30 << 100/mm2; +++, >100/mm2).
Myofibers were not stained, but only interstitium was weakly stained.
Figure 1.
Histopathological analysis of muscle biopsy samples from patients with GVHD‐associated myositis. (A) In most samples, a few scattered muscle fibers exhibited necrosis and regeneration without immune cell infiltration (H&E staining). (B) Some muscle bundles exhibited cellular infiltration with stromal hyperplasia (H&E staining). (C) Modified Gomori's trichrome staining showed similar results to HE stains, with no evidence of abnormal structures in the cytoplasm. (D–F) Serial sections showing the immunoreactivity of HLA and CD68. Fibers in most samples diffusely expressed HLA‐ABC (D) and most samples also expressed HLA‐DR (E), albeit at lower levels. The distribution of CD68‐positive cells on serial sections was nearly consistent with HLA‐DR (F). (G–I) Serial sections showing HLA‐DR‐positive myofibers and adjacent HLA‐DR‐positive interstitium (G) containing PD‐1‐positive cells (H) and PD‐L1‐positive cells (I). (J–L) Serial sections showing the distribution of lymphocytes and macrophages. CD8 staining (J) and CD4 staining (K) for leukocytes, and CD68 staining for macrophages. (L). (Bar: 100 μm). GVHD, graft versus host disease; H&E, hematoxylin and eosin; HLA, human leukocyte antigen; PD‐1, programmed death‐1; PD‐L1, programmed death‐ligand 1.
Cells positive for PD‐1 were also observed in all patient samples. These cells infiltrated within the endomysium in all but one case and were colocalized with the HLA‐DR‐positive interstitium (Fig. 1G,H). In addition, PD‐L1 immunoreactivity was detected in PD‐1‐ and HLA‐DR‐positive interstitium (Fig. 1I). Infiltration of CD4‐ and CD8‐positive lymphocytes was also observed in all patients (Fig. 1J,K). The distributions of those lymphocytes were wider than the distributions of PD‐1‐positive cells and HLA‐DR‐positive interstitium, and did not overlap with that of CD68‐positive macrophages (Fig. 1L).
PD‐1‐related pathology in GVHD‐associated myositis
To determine if these PD‐1‐positive cells were CD4‐ or CD8‐positive lymphocytes, we performed double immunofluorescent staining. All PD‐1‐positive cells were CD3‐positive (Fig. 2A–C), while some were CD4‐positive (Fig. 2D–F) and most were CD8‐positive (Fig. 2G–I). We also confirmed that PD‐1‐positive cells surrounded PD‐L1‐positive myofibers (Fig. 2J–L), and that HLA‐DR‐positive interstitium was always adjacent to PD‐L1‐positive myofibers (Fig. 2M–O).
Figure 2.
Fluorescence immunohistochemical staining of muscle samples from GVHD‐associated myositis patients confirming immune cell infiltration and PD‐1/PD‐L1 interactions. (A–C) All PD‐1‐positive cells were CD3‐positive. (D–F) Only some of PD‐1‐positive cells were CD4‐positive. (G–I) Most PD‐1‐positive cells were CD8‐positive. (J–L) Infiltration of PD‐1‐positive cells surround a myofiber expressing PD‐L1. (M–O) A myofiber expressing PD‐L1 and HLA‐DR is inundated by proliferation of HLA‐DR‐positive interstitial cells. GVHD, graft versus host disease; HLA, human leukocyte antigen; PD‐1, programmed death‐1; PD‐L1, programmed death‐ligand 1.
Comparison with other types of myositis
Next, we compared the histopathological findings of GVHD‐associated myositis to ASM and DM. Similar to GVHD‐associated myositis, the muscle specimens from ASM patients also demonstrated PD‐1‐positive cells colocalized with HLA‐DR‐positive interstitium (Fig. 3A–D). By contrast, PD‐1‐positive cells and HLA‐DR‐positive interstitium were less frequently observed in DM, though perifascicular myofibers in DM were often positive for HLA‐DR without adjacent interstitium proliferation or PD‐1‐positive cells (Fig. 3E,F). The proportion HLA‐DR‐positive myofibers was higher in GVHD and ASM than DM (difference between groups 8.3 (GVHD and DM), −0.97 (GVHD and ASM), −7.1 (ASM and DM); 95% CI 7.5 to 41.9, −20.6 to 20.4, −38.4 to 0.0; p = 0.0065, 0.94, 0.035, respectively), and the number of PD‐1‐positive cells infiltrating into muscle bundles was also significantly larger in GVHD and ASM than DM (difference between groups 9.6 (GVHD and DM), 2.6 (GVHD and ASM), −8.4 (ASM and DM); 95% CI 2.3 to 44.3, −8.0 to 22.4, −12.9 to −0.076; p = 0.0011, 0.64, 0.0089, respectively) (Fig. 3G,H). In ASM, increased ALP activity overlaps with CD68‐positive macrophages and HLA‐DR in the perimysium (Fig. 3I–K). In GVHD myositis, although CD68 immunoreactivity overlaps with HLA‐DR mainly in the endomysium (Fig. 3M,N), ALP staining was not observed but nonspecific staining of regenerating fibers (Fig. 3L). Similar staining pattern was also found in INMM (Fig. 3O–Q).
Figure 3.
Differences in histopathological features among GVHD‐associated myositis, anti‐tRNA‐synthetase antibody‐associated myopathy, and dermatomyositis. (A–F) Serial sections of HLA‐DR‐ and PD‐1‐immunostained biopsied muscle from a representative patient with GVHD‐associated myositis (A, B), anti‐tRNA‐synthetase antibody‐associated myopathy (ASM) (C, D), and dermatomyositis (DM) (E, F). HLA‐DR‐positive myofibers were observed in all cases, but the proliferating interstitium were seen in GVHD‐associated myositis (A) and ASM (C), but not in DM (E). PD‐1‐positive cells were present within the HLA‐DR‐positive interstitium of GVHD‐associated myositis (B) and ASM (D) samples. Infiltration of PD‐1‐positive cells into muscle bundles was rarely observed in DM (F). (G, H) The proportion of HLA‐DR‐positive myofibers (G) and the number per unit area (mm2) of PD‐1‐positive cells (H) in muscle bundles of each disorder. (I–Q) ALP staining and immunohistochemistry of CD68 and HLA‐DR on serial sections. Increased ALP activity overlap with CD68 immunoreactivity and HLA‐DR in the perimysium of ASM (I–K). In GVHD and immune‐mediated necrotizing myopathy (IMNM), although the distribution of CD68‐positive cells was consistent with the expression of HLA‐DR, ALP staining was scarcely found (L–Q). *p < 0.05 by Steel‐Dwass test. Bar = 100 μm in A–F and I–Q. ALP, alkaline phosphatase; ASM, anti‐tRNA‐synthetase antibody‐associated myopathy; DM, dermatomyositis; GVHD, graft versus host disease; HLA, human leukocyte antigen; IMNM, immune‐mediated necrotizing myopathy; PD‐1, programmed death‐1.
Between the GVHD patients with and without immunosuppressive drugs or steroids, there were no significant differences in either the density of PD‐1 (17.0 vs. 27.0 cells/mm2, p = 0.591) or the positivity rate of HLA‐DR (33.5% vs. 25.4%, p = 0.364).
Discussion
This retrospective review of GVHD‐associated myositis patients revealed that most cases show subacute or chronic progression, and are often accompanied by other organ lesions. All patients with GVHD‐associated myositis had hematological disorders and underwent HSCT. Most patients exhibited mild to moderate muscle weakness, but a few demonstrated only myalgia. Moreover, the clinical course was benign and responded well to immunotherapy, in agreement with previous studies. 5 , 12
Histopathology further revealed infiltration of PD‐1‐positive immune cells in the muscle interstitium adjacent to PD‐L1‐positive myofibrils, as well as high HLA‐DR expression in both muscle and interstitium. The association between local PD‐1 and HLA‐DR expression was also observed in muscles from ASM patients but rarely in DM patients. HLA‐DR expression is shown more characteristic in ASM than in DM, 10 being consistent with our findings here. A key observation of the present study is the expression of HLA‐DR in myofibers and interstitium adjacent to myofibers in GVHD‐associated myositis and ASM but not in DM. We also confirmed that PD‐1‐positive cells frequently infiltrated into the muscle interstitium accompanied by HLA‐DR expression.
PD‐1 is a known target of immune checkpoint inhibitors, which activate the antitumor responses of CD8‐positive T cells. Administration of immune checkpoint inhibitors after allogeneic HSCT is associated with the risk of severe or even lethal GVHD. 26 In our GVHD‐associated myositis cohort, the majority of PD‐1‐positive cells were also CD8‐positive and colocalized with PD‐L1 and HLA‐DR staining. Given that PD‐L1 induced by IFN‐γ, 27 which also upregulates HLA‐DR, 19 , 28 has a protective effect on GVHD of lung and liver, 15 , 16 , 17 our results suggest that activation of the PD‐1/PD‐L1 pathway suppresses immune cell activity and ensuing tissue damage in GVHD‐associated myositis.
GVHD‐associated myositis is considered a histologically nonspecific form of myositis related to infiltration of CD8‐positive lymphocytes. 12 , 29 It has also been reported that GVHD‐associated myositis has pathological features similar to those of DM, including perifascicular atrophy. 30 , 31 However, no patient in our study exhibited histopathological signs considered specific for DM, such as perifascicular atrophy 32 or capillary MAC deposition. 33 Sarcoplasmic MxA expression, which is reported to be a histopathological marker for DM, 34 was also negative in all cases. Skin manifestations were the most common complicating symptom of GVHD in our study, but the findings of skin biopsies were also not suggestive of DM. Instead, our investigations of muscle HLA‐DR and PD‐1 expression patterns collectively indicate that GVHD‐associated myositis is more closely related pathologically to ASM.
Increased ALP activity in the perimysium has been reported in ASM and DM, especially in Mi2 antibody‐positive cases. 10 , 35 In the present study, ALP staining colocalizes with CD68 and HLA‐DR in ASM. In GVHD and IMNM, however, ALP activity was not observed despite the overlap of CD68 immunoreactivity and HLA‐DR. These disease‐associated staining patterns may be due to, at least partially, the spatial distribution of macrophage infiltration, which is dominant in the perimysium in ASM but mainly in the endomysium in GVHD and IMNM.
This study has several limitations. First, the retrospective study design limits follow‐up investigation and may introduce selection bias. Second, the diagnosis was biopsy‐based and performed in a tertiary care institution, which may compound selection bias. Third, the small sample size and ethnic homogeneity limit the generalizability of these findings. Fourth, information on autoantibodies was not available for all the patients. Fifth, the muscle weakness reported in patients with GVHD may also stem from other factors, such as drug side effects and complications of different autoimmune diseases such as myasthenia gravis. Finally, it is possible that steroids or immunosuppressive drugs altered the histologic findings. Although there were no substantial differences in either the density of PD‐1 or the positivity rate of HLA‐DR between the GVHD patients with and without immunosuppressive drugs or steroids, the sample size might be too small to evaluate the precise role of therapies on the pathological findings. Although it is difficult to completely eliminate potential cofactors, the efforts of several clinicians and pathologists ruled out conditions other than myositis.
In summary, these clinical and histopathological examinations reveal leukocyte and macrophage infiltration into the muscle interstitium of GVHD‐associated myositis patients. These infiltrating cells express PD‐1 and are associated with elevated HLA‐DR and PD‐L1 in adjacent myofibers. A similar expression pattern was found in ASM but not DM, suggesting that these findings can aid in differential diagnosis between GVHD‐associated myositis and DM.
Author Contributions
Conception and design of the study: TK, AM, HN, and MK. Acquisition and analysis of data: all authors. Drafting the manuscript and figures: TK, SN, and MK.
Conflict of Interest
We have no conflict of interest to disclose.
Acknowledgment
This work was funded by JSPS KAKENHI Grant Numbers JP20H00527 and JP23H00420.
Funding Statement
This work was funded by Japan Society for the Promotion of Science grants JP20H00527 and JP23H00420.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1. Jagasia MH, Greinix HT, Arora M, et al. National Institutes of Health consensus development project on criteria for clinical trials in chronic graft‐versus‐host disease: I. The 2014 diagnosis and staging working group report. Biol Blood Marrow Transplant. 2015;21(3):389‐401.e1. doi: 10.1016/j.bbmt.2014.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ferrara JL, Levine JE, Reddy P, Holler E. Graft‐versus‐host disease. Lancet. 2009;373(9674):1550‐1561. doi: 10.1016/S0140-6736(09)60237-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Shulman HM, Kleiner D, Lee SJ, et al. Histopathologic diagnosis of chronic graft‐versus‐host disease: National Institutes of Health consensus development project on criteria for clinical trials in chronic graft‐versus‐host disease: II. Pathology working group report. Biol Blood Marrow Transplant. 2006;12(1):31‐47. doi: 10.1016/j.bbmt.2005.10.023 [DOI] [PubMed] [Google Scholar]
- 4. Grauer O, Wolff D, Bertz H, et al. Neurological manifestations of chronic graft‐versus‐host disease after allogeneic haematopoietic stem cell transplantation: report from the consensus conference on clinical practice in chronic graft‐versus‐host disease. Brain. 2010;133(10):2852‐2865. doi: 10.1093/brain/awq245 [DOI] [PubMed] [Google Scholar]
- 5. Oda K, Nakaseko C, Ozawa S, et al. Fasciitis and myositis: an analysis of muscle‐related complications caused by chronic GVHD after allo‐SCT. Bone Marrow Transplant. 2009;43(2):159‐167. doi: 10.1038/bmt.2008.297 [DOI] [PubMed] [Google Scholar]
- 6. Stevens AM, Sullivan KM, Nelson JL. Polymyositis as a manifestation of chronic graft‐versus‐host disease. Rheumatol (Oxf). 2003;42(1):34‐39. doi: 10.1093/rheumatology/keg025 [DOI] [PubMed] [Google Scholar]
- 7. Cresswell P. Assembly, transport, and function of MHC class II molecules. Annu Rev Immunol. 1994;12:259‐293. doi: 10.1146/annurev.iy.12.040194.001355 [DOI] [PubMed] [Google Scholar]
- 8. Benveniste O, Goebel HH, Stenzel W. Biomarkers in inflammatory myopathies‐an expanded definition. Front Neurol. 2019;10:554. doi: 10.3389/fneur.2019.00554 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Rodríguez Cruz PM, Luo YB, Miller J, Junckerstorff RC, Mastaglia FL, Fabian V. An analysis of the sensitivity and specificity of MHC‐I and MHC‐II immunohistochemical staining in muscle biopsies for the diagnosis of inflammatory myopathies. Neuromuscul Disord. 2014;24(12):1025‐1035. doi: 10.1016/j.nmd.2014.06.436 [DOI] [PubMed] [Google Scholar]
- 10. Tanboon J, Inoue M, Hirakawa S, et al. Muscle pathology of antisynthetase syndrome according to antibody subtypes. Brain Pathol. 2023;33(4):e13155. doi: 10.1111/bpa.13155 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Shiota T, Eura N, Hasegawa A, Kiriyama T, Sugie K. Pathological features of inflammatory myopathy as a manifestation of chronic graft‐versus‐host disease after allogeneic bone marrow transplantation. Neuropathology. 2022;42(4):309‐314. doi: 10.1111/neup.12816 [DOI] [PubMed] [Google Scholar]
- 12. Saw JL, Sidiqi MH, Mauermann ML, Alkhateeb H, Naddaf E. Immune‐mediated neuromuscular complications of graft‐versus‐host disease. Muscle Nerve. 2021;63(6):852‐860. doi: 10.1002/mus.27214 [DOI] [PubMed] [Google Scholar]
- 13. Francisco LM, Sage PT, Sharpe AH. The PD‐1 pathway in tolerance and autoimmunity. Immunol Rev. 2010;236:219‐242. doi: 10.1111/j.1600-065X.2010.00923.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Keir ME, Butte MJ, Freeman GJ, Sharpe AH. PD‐1 and its ligands in tolerance and immunity. Annu Rev Immunol. 2008;26:677‐704. doi: 10.1146/annurev.immunol.26.021607.090331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Wang H, Yang YG. The complex and central role of interferon‐γ in graft‐versus‐host disease and graft‐versus‐tumor activity. Immunol Rev. 2014;258(1):30‐44. doi: 10.1111/imr.12151 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Yi T, Chen Y, Wang L, et al. Reciprocal differentiation and tissue‐specific pathogenesis of Th1, Th2, and Th17 cells in graft‐versus‐host disease. Blood. 2009;114(14):3101‐3112. doi: 10.1182/blood-2009-05-219402 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Yang HR, Chou HS, Gu X, et al. Mechanistic insights into immunomodulation by hepatic stellate cells in mice: a critical role of interferon‐gamma signaling. Hepatology. 2009;50(6):1981‐1991. doi: 10.1002/hep.23202 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Bolinger B, Engeler D, Krebs P, et al. IFN‐gamma‐receptor signaling ameliorates transplant vasculopathy through attenuation of CD8+ T‐cell‐mediated injury of vascular endothelial cells. Eur J Immunol. 2010;40(3):733‐743. doi: 10.1002/eji.200939706 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Inukai A, Kuru S, Liang Y, et al. Expression of HLA‐DR and its enhancing molecules in muscle fibers in polymyositis. Muscle Nerve. 2000;23(3):385‐392. doi: [DOI] [PubMed] [Google Scholar]
- 20. Nakanishi H, Koike H, Matsuo K, et al. Demographic features of Japanese patients with sporadic inclusion body myositis: a single‐center referral experience. Intern Med. 2013;52(3):333‐337. doi: 10.2169/internalmedicine.52.8910 [DOI] [PubMed] [Google Scholar]
- 21. Noda T, Iijima M, Noda S, et al. Gene expression profile of inflammatory myopathy with malignancy is similar to that of dermatomyositis rather than polymyositis. Intern Med. 2016;55(18):2571‐2580. doi: 10.2169/internalmedicine.55.6706 [DOI] [PubMed] [Google Scholar]
- 22. Noda S, Koike H, Maeshima S, et al. Transforming growth factor‐β signaling is upregulated in sporadic inclusion body myositis. Muscle Nerve. 2017;55(5):741‐747. doi: 10.1002/mus.25405 [DOI] [PubMed] [Google Scholar]
- 23. Maeshima S, Koike H, Noda S, et al. Clinicopathological features of sarcoidosis manifesting as generalized chronic myopathy. J Neurol. 2015;262(4):1035‐1045. doi: 10.1007/s00415-015-7680-0 [DOI] [PubMed] [Google Scholar]
- 24. Tanboon J, Inoue M, Saito Y, et al. Dermatomyositis: muscle pathology according to antibody subtypes. Neurology. 2022;98(7):e739‐e749. doi: 10.1212/WNL.0000000000013176 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Preuße C, Goebel HH, Held J, et al. Immune‐mediated necrotizing myopathy is characterized by a specific Th1‐M1 polarized immune profile. Am J Pathol. 2012;181(6):2161‐2171. doi: 10.1016/j.ajpath.2012.08.033 [DOI] [PubMed] [Google Scholar]
- 26. Ortega Sanchez G, Stenner F, Dirnhofer S, et al. Toxicity associated with PD‐1 blockade after allogeneic haematopoietic cell transplantation. Swiss Med Wkly. 2019;149:w20150. doi: 10.4414/smw.2019.20150 [DOI] [PubMed] [Google Scholar]
- 27. Wiendl H, Mitsdoerffer M, Schneider D, et al. Human muscle cells express a B7‐related molecule, B7‐H1, with strong negative immune regulatory potential: a novel mechanism of counterbalancing the immune attack in idiopathic inflammatory myopathies. FASEB J. 2003;17(13):1892‐1894. doi: 10.1096/fj.03-0039fje [DOI] [PubMed] [Google Scholar]
- 28. Ding M, Huang T, Zhu R, et al. Immunological behavior analysis of muscle cells under IFN‐γ stimulation in vitro and in vivo. Anat Rec (Hoboken). 2018;301(9):1551‐1563. doi: 10.1002/ar.23834 [DOI] [PubMed] [Google Scholar]
- 29. Shulman HM, Cardona DM, Greenson JK, et al. NIH consensus development project on criteria for clinical trials in chronic graft‐versus‐host disease: II. The 2014 pathology working group report. Biol Blood Marrow Transplant. 2015;21(4):589‐603. doi: 10.1016/j.bbmt.2014.12.031 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Allen JA, Greenberg SA, Amato AA. Dermatomyositis‐like muscle pathology in patients with chronic graft‐versus‐host disease. Muscle Nerve. 2009;40(4):643‐647. doi: 10.1002/mus.21353 [DOI] [PubMed] [Google Scholar]
- 31. Ollivier I, Wolkenstein P, Gherardi R, et al. Dermatomyositis‐like graft‐versus‐host disease. Br J Dermatol. 1998;138(3):558‐559. doi: 10.1046/j.1365-2133.1998.02155.x [DOI] [PubMed] [Google Scholar]
- 32. Dalakas MC, Hohlfeld R. Polymyositis and dermatomyositis. Lancet. 2003;362(9388):971‐982. doi: 10.1016/S0140-6736(03)14368-1 [DOI] [PubMed] [Google Scholar]
- 33. Jain A, Sharma MC, Sarkar C, et al. Detection of the membrane attack complex as a diagnostic tool in dermatomyositis. Acta Neurol Scand. 2011;123(2):122‐129. doi: 10.1111/j.1600-0404.2010.01353.x [DOI] [PubMed] [Google Scholar]
- 34. Uruha A, Nishikawa A, Tsuburaya RS, et al. Sarcoplasmic MxA expression: a valuable marker of dermatomyositis. Neurology. 2017;88(5):493‐500. doi: 10.1212/WNL.0000000000003568 [DOI] [PubMed] [Google Scholar]
- 35. Pestronk A. Acquired immune and inflammatory myopathies: pathologic classification. Curr Opin Rheumatol. 2011;23(6):595‐604. doi: 10.1097/BOR.0b013e32834bab42 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.