Physical function and quality of life in patients with chronic graft-versus-host-disease: A summary of preclinical and clinical studies and a call for exercise intervention trials in patients

Abstract

Allogeneic Hematopoietic Stem Cell Transplant, to reconstitute hematopoietic and immune status of patients undergoing myeloablative therapy for hematologic disorders, has been of great benefit in minimizing or eradicating disease and extending survival. Patients who undergo allogeneic hematopoietic stem cell transplant (allo-HSCT) are subject to many comorbidities among which the most significant, affecting quality of life (QoL) and survival, are acute (aGVHD) and chronic Graft Versus Host Disease (cGVHD), resulting from donor lymphocytes reacting to and damaging host tissues. Physical activity and exercise have clearly been shown, in both children and adults, to enhance fitness, improve symptomatology and QoL, reduce disease progression and extend survival for many diseases including malignancies. In some cases, vigorous exercise has been shown to be equal to or more effective than pharmacologic therapy. This review addresses how cGVHD affects patients’ physical function and physical domain of QoL, and the potential benefits of exercise interventions along with recommendations for relevant research and evaluation targeted at incorporating this strategy as soon as possible after allo-HSCT and ideally, as soon as possible upon diagnosis of the condition leading to allo-HSCT.

Introduction

Allogeneic hematopoietic cell transplantation (allo-HSCT) is a potentially curative therapy for several hematologic diseases. Graft-versus-host disease (GVHD) is a major complication affecting an ever increasing number of long term survivors (1–4). GVHD arises when donor T-cells act against the cells of the immunocompromised host recipient because of histocompatibility antigen disparity (2). The chronic form of GVHD (cGVHD) may be lethal and survivors have to face the challenge of severe morbidity, which considerably impairs their quality of life (QoL), particularly those domains related to physical function (see below). Moreover, prolonged cGVHD is a risk factor for other complications (5) and a prognostic factor for overall survival and relapse-free survival (6).

Systemic corticosteroid therapy is the standard first-line treatment for cGVHD, though under 40–50% respond favorably and secondary and tertiary treatments have been associated with a high rate of failure (7). In view of these poor treatment outcomes and the toxic effects of treatment, a main research objective is to identify new treatment strategies that will help preserve or even improve the QoL of these patients. Such strategies should specifically target the physical domain of QoL, thus minimizing impacts on daily living activities. It is widely known that an active lifestyle has numerous beneficial effects on the course of most chronic diseases ultimately improving patient well-being (8). It is therefore no surprise that physical exercise is often recommended as a therapeutic tool for numerous conditions, though its effects on patients with cGVHD are largely unknown. Work in this field requires valid reliable tools to accurately define the physical condition of a patient and the patient's response to a given intervention in terms of effects on daily living activities. This brief review provides an overview of our current understanding of this debilitating disease, how it affects the patients’ functional capacity and physical domain of QoL and, and what can we can learn from pre-clinical exercise intervention studies. Recommendations are also given for future exercise studies in patients targeting mainly to improve their physical function and physical function-related domains of QoL.

Summary of diagnosis criteria for cGVHD

The National Institutes of Health (NIH) Consensus Working Group for Diagnosis and Staging of cGVHD considered the clinical and pathological characteristics of GVHD (9–14) to define two main categories each with two subcategories (Figure 1): acute (aGVHD), including classic aGVHD, and late-onset aGVHD; and cGVHD, comprising classic cGVHD and overlap syndrome (9). Overlap syndrome is characterized by a poor prognosis, functional impairment, and higher symptom burden and mortality (15). The NIH also consider the necessary clinical manifestations for a diagnosis of cGVHD (Table 1) distinguishing this form from aGVHD when there is at least one diagnostic clinical sign of cGVHD or at least one distinctive manifestation confirmed by biopsy or other relevant tests, along with the exclusion of other possible diagnoses (9,16).

Figure 1.

GVHD classification following National Institutes of Health Consensus Working Group for Diagnosis and Staging of cGVHD. Abbreviations: aGVHD, acute graft-versus-host disease; cGVHD, chronic graft-versus-host disease.

Table 1.

Diagnostic manifestations of cGVHD (9).

ORGAN	FEATURES
Skin	Poikiloderma Lichen-planus like features Sclerotic features Morphea-like features Lichen-sclerosis like features
Mouth	Lichen-type features Hyperkeratotic plaques Restriction of mouth opening from sclerosis
Genital tract	Lichen-planus-like features Vaginal scarring or stenosis
Gastrointestinal tract	Esophageal web Strictures or stenosis (upper-mid third of oesophagus)
Lungs	Bronchiolitis obliterans diagnosed with lung biopsy
Muscles, fascia, joints	Fasciitis Joint stiffness or contractures secondary to sclerosis

Summary of the pathobiology of cGVHD

The events leading to the development of cGVHD have been defined in large measure using murine models that show three disease mechanisms: (i) the production of auto-antibodies (typically anti-DNA); (ii) the development of fibrosis; and (iii) abnormal thymus function (17–19). Since none of these events per se faithfully reproduce the mechanisms of human cGVHD (17,18), two other new models have appeared: one recapitulating a transition from aGVHD to a scleroderma-like form of cGVHD with salivary gland involvement and serum antibodies (20), and another reproducing a systemic disease with multi-organ involvement, in which lung and liver fibrosis are associated with CD4 T-cell and B-cell infiltration (21).

Different theories have tried to explain the pathophysiology of cGVHD, primarily involving the role of T or B-cells, thymus dysfunction or fibrotic changes. Briefly, cGVHD is linked to a high percentage of memory effector CD4⁺ CCR7⁺/CD62L^low T-cells from the donor as well as CD8⁺/CCR7⁻/CD45RA⁺ T cells, with CD4⁺ and CD8⁺ cell infiltrates detected in oral lichenoid lesions (22), and CD8⁺ infiltrates detected in the skin, intestine and oral mucosa (23,24). The expression of CD134 (OX40) on the surface of CD8⁺ and CD4⁺ T-cells has been associated with cGVHD onset, and marks an early T-cell activation by inflammatory cytokines (25). A decline in regulatory T-cells (T_regs) leads to a loss of peripheral tolerance, autoimmunity and cGVHD (19,26). In fact, patients with cGVHD have lower numbers of CD4⁺ T_regs compared to patients without cGVHD (27,28).

The role played by B lymphocytes in cGVHD has been well established (29–33), with patients showing high levels of B-cell activating factor (BAFF) (31,33–38). There could be an increased BAFF/B-cell ratio resulting in an immune pathology (31,33,36). The development of cGVHD is also related to an increase in B-cells expressing high levels of the toll-like receptor 9 (38). In addition, given that many memory B-cells express auto-reactive antibodies, the inflammation present during GVHD activates these cells to produce allo/auto-antibodies (21,39–41). Further, an immune system recovery process poor in B-cells leads to a state of immunodeficiency increasing the risk of infection (21,29,33,37). Finally, the decrease produced in naïve and transition B-cells increases the auto-reactivity of B-cells with antigenic experience (expressing CD27⁺) (33). This subgroup of B-cells circulates in patients with cGVHD who present BAFF receptors (33).

The damage caused by conditioning regimens in the thymus, previous aGVHD, atrophy and/or the prophylactic regimen, deregulates central tolerance mechanisms during the immune recovery that leads to cGVHD (42). Regarding fibrotic changes, several complement factors play a role in cGVHD. Complement factor 3 is deposited at the dermis-epidermis junction and complement factor 5 modifies liver fibrosis (43). Chemokines C-C motif ligand-3 and 2 are chemotactic for mononuclear phagocytes and are essential pro-fibrotic mediators (44). T-CD4⁺ lymphocytes are stimulated by class II antigenic determinants expressed on fibroblasts, keratinocytes and other cells, and once activated, these cells produce cytokines [IL4, transforming growth factor (TGF)β] that induce collagen production by the fibroblasts, thereby producing fibrotic lesions in the skin, liver, exocrine glands and thymus which are characteristic of cGVHD (45). Specifically, TGFβ is involved in autoimmune diseases and cGVHD (46).

Clinical manifestations of cGVHD: Implications for patients’ physical function and QoL

The clinical manifestations of cGVHD resemble those observed in an autoimmune disease such as systemic lupus erythematosus, lichen planus and scleroderma (2,9,16,40,47,48), affecting a single organ or many body zones (9). Initial signs usually appear in the oral mucosa with other organs involved such as the skin, nails, eyes, muscles, lungs, tendons, gastrointestinal tract, liver, joints, nerves, kidneys, serous membranes, genitals, heart and immune system (see Figure 2). Thus, this disease has a negative impact on the chain of interactive events involved in physical exercise capacity, mainly blood oxygenation (which depends on pulmonary function), supply of oxygenated blood to the working muscles (which depends on the cardiac pump), and muscle function.

Figure 2.

Clinical manifestations of chronic graft versus host disease (cGVHD) and how they affect patients’ daily physical function and quality of life (QoL).

Quality of life is a multidimensional concept describing patients' perceptions of the impacts of their disease and its treatment on their physical, psychological, social, financial and spiritual well-being (49–51). Chronic GVHD has a negative impact on QoL (1,15,48,50,52–68), though this is not detectable in all patients (69). According to the NIH criteria for grading cGVHD severity, patients with mild cGVHD have higher QoL than those with moderate cGVHD, and both of these categories have higher QoL than patients with severe cGVHD (63). Patients with overlap subtype of cGVHD report worse QoL and higher symptom burden than patients with classic cGVHD (15). As for age, middle-aged (41–59 years) patients report lower QoL compared to younger (18–40 years) or older patients (≥60 years) (70). Importantly, patients affected by cGVHD have significant impairment in QoL compared to age- and sex-matched US population normative data (65).

Although there is some controversy on how does tissue-specific affectation impact QoL, the presence of sclerotic, skin, joints-fascia, and/or lung involvement is thought to have the greatest deteriorating effect on the QoL of moderate-severely affected cGVHD patients (53). Other important determinants of patient-reported QoL might be gastrointestinal involvement and elevated bilirubin (71), or skin and nutrition symptoms in patients with overlap syndrome (15).

More specifically, the physical domain of QoL warrants special attention since both allo-HSCT and GVHD, acute or chronic, are known to impair physical functioning (1,50,5,61,72–76), limiting a patient's ability to carry out daily living activities (1,59,61). Long periods of bed rest or use of a wheel chair lead to muscle atrophy and cachexia, reducing functional capacity (1,77), and the adverse effects of immunosuppressive therapy (i.e., osteoporosis, osteonecrosis, diabetes mellitus, hypertension, problems in a growing child) further limit a patient's physical capacity, increasing patient morbidity (1,50,59,61,63). The clinical severity of cGVHD is associated with the magnitude of functional deterioration (15,53,61,65), especially in older patients, in whom the decline is more marked compared to their younger peers (70). Besides aging (65), other independent predictors of physical impairment in cGVHD are intensive systemic immunosuppression, reduced capacity for ambulation and greater cGVHD symptom bother (61). Other important contributors to the decline in patients’ physical functioning are upper gastrointestinal and overall liver involvement (71). Patients with overlap syndrome are thought to suffer more marked functional impairment compared to those with classic cGVHD (15). Accordingly, patient-reported QoL and physical functioning are considered important end points of clinical trials (64). These endpoints need to be measured using reliable, safe, practical, valid methods with the objectives of: i) quantifying the effects produced by cGVHD on physical fitness; ii) monitoring changes produced over time in these endpoints in response to an intervention targeted at modifying lifestyle habits; and iv) establishing a common vocabulary for the clinicians managing and caring for these patients. Currently, several assessment tools are available for the different aspects of the physical QoL domain of the patient with cGVHD (see Table 2 for a detailed description of these tests). Many of these tests are recommended by the NIH Consensus Conference on cGVHD (12,78,79). The reader is referred to Table 3 for a detailed analysis of the studies that have used these tests on patients with cGVHD.

Table 2.

A. Functional capacity tests
Assessment scale	Description	Measured variable
2MWT	The patient is instructed to walk as fast as possible a 50-foot course (25-feet each direction) with 180 degree turns at each end, and the total distance covered in 2 minutes is recorded (80).	Functional status. Functional capacity.
HGS test	Grip strength using a portable electronic dynamometer three times in the dominant hand (average of 3 measurements) (82).	Functional status. Functional capacity.
PFTs	The LFS is calculated from theFEV1and DLCO corrected for hemoglobin. The FEV1 and DLCO are converted to a numeric score as follows: >80=1; 70% to 79%=2; 60% to 69%=3; 50% to 59%=4; 40% to 49%=5; and <40%=6. The LFS=FEV1 score + DLCO score, with a possible range of 2 to 12, with higher numbers indicating worse dysfunction. The LFS (0 to 3) is derived as follows: 0 = FEV1>80% or LFS 2; 1 = FEV1 60% to 79% or LFS 3 to 5; 2 = FEV1 40% to 59% or LFS 6 to 9; and 3 = FEV1<39% or LFS 10 to 12 (87). Portable spirometry serves torecord of FEV1 after the patient is instructed to take a deep breath and exhale forcefully and rapidly (3 measures), although formal pulmonary function testing is preferable if available.	Respiratory capacity. Lung capacity.
ROM	Using a standard goniometer in the supine position, participants' mean bilateral ROM measurements are converted to the percentage of normal ROM, using the American Academy of Orthopaedic Surgeons' definition of normal ROM for each joint: measurements that exceeded the maximum value are assigned a score of 100 and joints with fixed contractures are assigned a score of 0, and an aggregate score (125) is calculated representing the patient's average degree of impairment in upper and lower body ROM.	Joint/fascia response.
P-ROM scale	It is a series of images that captures ROM separately for shoulders, elbows, wrists/fingers, and ankles (83), with lower scores indicating more limited ROM. The P-ROM total score is the sum of scores in all 4 joints, with a maximum possible score of 25.	Joint/fascia response. ROM.

B. Patient’s self-reports of functional capacity and physical domain of quality of life (QoL).
Assessment scale	Description	Measured variable
HAP	It is a 94-item self-reported assessment of energy expenditure and physical fitness. Responders are asked to indicate, whether they are “still doing this activity”, “have stopped doing this activity”, or “never did this activity”. Two scores are calculated: the MAS, which is the number of the most difficult task the respondent is “still doing”, and the AAS, calculated by counting how many activities with lower values than the MAS the respondent has “stopped doing” and subtracting this from the MAS.	Daily activity. Physical functioning
SF-36	This tool is a 36-item self-report questionnaire that examines following domains: physical functioning, role limitations as a result of physical functioning, bodily pain, general health, vitality, social functioning, role emotional functioning, and mental health. These domains can be aggregated into 2 summary scales: the PCS and the MCS, both of them are derived by using a standard algorithm to aggregate scores across the eight domains (population mean of 50; standard deviation of 10). SF-36 higher scores indicate better functioning; lower scores on the PCS indicate limitations in physical functioning and role participation, a high degree of bodily pain, and an unfavorable perception of general health (126).	QoL. Functional health and well-being
FACT-BMT	It is a 37-item self-report questionnaire that measures the effect of cancer therapy on domains including physical, functional, social/family, and emotional well-being, and bone marrow transplantation concerns. Individual domain scores can be summarized to give a total FACT-BMT score (including all subscales) or a FACT-TOI (physical well-being + functional well-being + BMT subscale) (91).	QoL
L-cGVHD-SC	It is 30-item, 7-subscale symptom scale self-administered patient that evaluates adverse effects of cGVHD on skin, vitality (energy), lung, nutritional status, psychological functioning, muscles and joints, eyes, and mouth. A summary score is created by taking the mean of all items and linearly transforming that value to a 0–100 scale (79). We can find a muscle/ joint subscale which is a summary of 4 items, i.e., joint and muscle aches, limited joint movement, muscle cramps, and weak muscles, with each item rated as follows: 0 → not at all; 1 → slightly; 2 → moderately; 3 → quite a bit; 4 →extremely.	cGVHD symptom

C. National Institutes of Health specifics scales.
Assessment scale	Description	Measured variable
NIH joint/fascia score	It uses a 0–3-point scale. 0: no symptoms; 1: mild tightness of arms or legs, normal or mild decreased ROM, and not affecting ADL; 2: tightness of arms or legs or joint contractures, erythema thought due to fasciitis, moderate decrease in ROM, and mild-to-moderate limitation of ADL; 3: contracture with significant decrease of ROM and significant limitation of ADL (unable to tie shoes, button shirts, dress self, etc.)	Tightness. ROM. ADL.
NIH lung score	It uses a 0–3-point scale. 0: no symptoms; 1: shortness of breath with stairs; 2: shortness of breath on flat ground; and 3: shortness of breath at rest or requiring oxygen.	Pulmonary function.

Abbreviations:ADL, Activities of Daily Living; DLCO, Diffusing Capacity of Carbon Monoxide; FEV₁, Forced Expiratory Volume in 1 second; HGS, Hand Grip Strength; LFS, Lung Function Score; PFTs, Pulmonary Function Tests; P-ROM, Photographic-Range of Motion scale; ROM, Range of Motion; 2MWT, 2-Minute Walk Test.

Abbreviations: AAS, Adjusted Activity Score; cGVHD, chronic Graft-versus-Host disease;FACT-BMT, Functional Assessment of Cancer Therapy-Bone Marrow Transplant; FACT-TOI,Functional Assessment of Cancer Therapy–Trial Outcome Index; HAP, Human Activity Profile; L-cGVHD-SC, Lee cGVHD Symptom Scale; MAS, Maximum Activity Score;MCS, Mental Component Score; PCS, Physical Component Score; SF-36, Medical Outcomes Study 36-Item Short-Form Health Survey.

Abbreviations: ADL, Activities of Daily Living; NIH, National Institutes of Health; ROM, Range of Motion.

Table 3.

Summary of studies reporting results on functional capacity tests or self-reports of functional capacity and physical domain of quality of life (QoL) in patients with chronic graft versus host disease (cGVHD).

Reference	N, age (average) and gender (%)	cGVHD onset (%)	NIH cGVHD global severity (%)	cGVHD NIH subtype (%)	Source of transplant (%)	Functional capacity tests	Self-reports of functional capacity or physical domain of QoL	Main results
(15)	N=427 With classic Median age in years (range): 49 (17–69) 99% adults M=64; F=36 With overlap Median age in years (range): 51 (2–79) 96% adults M=57; F=43	-	None: 0.47 Mild: 10 Moderate: 57.2 Severe: 32.3 With classic None: 3 Mild: 9 Moderate: 63 Severe: 25 With overlap Mild: 10 Moderate: 56 Severe: 34	Classic: 18 Overlap: 82	Bone marrow: 7.54 PBSC: 87.6 UC: 4.94 With classic Bone marrow: 5 PBSC: 90 UC: 5 With overlap Bone marrow: 8 PBSC: 87 UC: 5	2-MWT HGS test PFTs (and portable spirometry)	HAP SF-36 FACT-BMT L-cGVHD-SC	-Patients with overlap syndrome had lower values in 2-MWT, FEV1, SF-36 social functioning scores, HAP, and higher symptom burden than patients with classic cGVHD. -The overlap cGVHD was associated with worse overall survival (HR 2.1, 95% CI 1.1–4.7; p=0.03) and higher non-relapse mortality (HR 2.8, 95% CI 1.2–8.3; p=0.02) than classic cGVHD.
(50)	With cGVHD N=117 Median age in years (range): 44 (21–72) M=55.6; F=44.4 Without cGVHD N=59 Median age in years (range): 44 (18–67) M=45.8; F=54.2	Unknown: ? De novo ? Progressive: 20,5 Quiescent: ?	Mild: 28.3 Moderate: 42.7 Severe: 29	Classic: 100	Allo-HSCT: 100	-	HAP SF-36 FACT-BMT L-cGVHD-SC	-The HAP MAS and AAS correlated inversely with cGVHD severity (r=−0.25 for MAS and −0.24 for AAS). -Lung manifestations of cGVHD correlated with AAS (r=0.17), but not with MAS. -HAP scores correlated with subscales from other instruments measuring physical domains, especially the PCS SF-36. -There were high internal consistency of HAP and good correlation of MAS and AAS with cGVHD severity. -The HAP had higher sensitivity to change of cGVHD activity compared to the SF-36 and the FACT-BMT. -Steroid myopathy correlated with both HAP scores.
(53)	N=189 Average age in years (range): 48 (18–70). M=52; F=48	Unknown: 0.5 De novo 35.5 Progressive: 42 Quiescent: 22	Mild: 1 Moderate: 33 Severe: 66	Classic: 88 Overlap: 12	Bone marrow: 18.5 PBSC: 81 UC: 0.5	2-MWT HGS test PFTs ROM	HAP SF-36 FACT-BMT L-cGVHD-SC	-NIH global severity scores were associated (p≤0.001 to p≤0.05) with all measures, with the exception of SF-36 MCS. -Joints/fascia, skin, and lung scores showed highest number of significant associations with almost all outcome measures, having a big impact on function and QoL. -Joints/fascia and skin scores, but not lung, were associated with intensity of immunosuppression (both p<0.0001), therapeutic intent at the time of evaluation (p<0.0001 for both), and clinician 7-point global assessment of change (p<0.0001 for both). -The L-cGVHD-SC was correlated with gastrointestinal score and a NIH Global Score of “severe”. -NIH lung score of 3 versus <3 was the most strongly negatively associated with survival, with a 3-year estimated survival of 35% versus 82% respectively, p<0.0001.
(61)	N=100 Median age in years (range): 46 (20–66) M=52; F=48	Unknown: 0 De novo 40 Progressive: 43 Quiescent: 17	Mild: 5 Moderate: 45 Severe: 50	-	Bone marrow: 18 PBSC: 80 Unspecified: 2	2-MWT HGS test ROM	SF-36 L-cGVHD-SC	-The mean PCS and seven of eight SF-36 subscale scores were significantly lower (mean=36.8±10.7) than the US population norm of 50 (p<0.001). -The highest decrements were in physical function (mean=38.8±10.9) and physical role function (mean=37.8±11.8). -Significant independent predictors of impaired performance were intensive systemic immunosuppression, reduced capacity for ambulation, and greater cGVHD symptom bother (p<0.05). -Symptom bother had a direct effect on functional performance, as well as an indirect effect partially mediated by functional capacity (Sobel test, p=0.004).
(63)	N=264 Median age in years (range): 36.1±11.1 M: 63.3; F: 36.7	De novo 62.1 Progressive: 3.4 Quiescent: 34.5	Mild: 28.4 Moderate: 52.3 Severe: 19.3	Classic: 69.3 Overlap: 30.7	Allo-HSCT:100	-	SF-36	-Patients with mild cGVHD had better QoL than in those with moderate cGVHD (p<0.05). -Patients with mild or moderate cGVHD had better QoL than in those with severe cGVHD (p<0.01). -Chronic GVHD severity had the greatest significant negative impact on patients’ QoL (PCS: OR=1.4; 95% CI: 1.1–1.8; p=0.004; MCS: OR=1.6; 95% CI, 1.1–2.1; p=0.005), whereas being female was associated with a negative impact on the MCS (OR=1.6; 95% CI: 1.0–2.4; p=0.039).
(64)	N=336 Median age in years (range): 52 (19–79) M: 60; F: 40	-	Change in cGVHD severity was examined by comparing severity at each visit to severity at the previous visit.	Classic: 77 Overlap: 23	Allo-HSCT: 100	-	SF-36 FACT-BMT FACT-G	-Change in NIH-assessed global cGVHD severity over a 6-month average timeframe was not associated with change in QoL. -Changes in clinician-assessed cGVHD were associated with changes in the FACT-G and FACT-BMT scores. -Change in patient-reported cGVHD severity was correlated with change in both the SF-36 and FACT-BMT. -Patients demonstrated clinically relevant improvement or worsening in QoL compared to that at the prior visit in approximately 40% of follow-up visits. -The SF-36 and FACT-BMT perform well in capturing thepatients’ perspective, and fairly well in capturing physicians’ assessment. Neither correlates well with changes in NIH global severity scores.
(65)	N=298 → 87% completed all or part of the SF-36 and FACT-BMT Median age in years (range): 53 (20–79). M=58; F=42	-	Mild: 10 Moderate: 59 Severe: 31	Classic: 56 Overlap: 44	Bone marrow: 11 PBSC: 89	-	SF-36 FACT-BMT	-Chronic GVHD severity was independently associated with QoL, adjusting for age. -Significant differences were found most often between average QoL scores in moderate and severe cGVHD patients, with fewer significant differences between mild and moderate GVHD. -QoL composite and subscale scores were higher on average for patients with moderate GVHD compared with severe cGVHD. Differences in QoL between mild and severe cGVHD were of greater magnitude than between moderate and severe. -Mean scores for cGVHD patients were significantly lower for physical functioning, role-physical, bodily pain, general health, vitality, social functioning, and PCS than population normative data.
(70)	N=522 Adolescent and young adult (AYA) 18–40 years =22% M=49; F=51 Middle-aged: 41–59 years =53% M=57; F=43 Older: ≥60years =25% M=63; F=37	-	Moderate: 58 Severe: 42 AYA Moderate: 55 Severe: 45 Middle-aged Moderate: 59 Severe: 41 Older Moderate: 58 Severe: 42	-	AYA Bone marrow: 19 PBSC: 77 UC: 4 Middle-aged Bone marrow: 5 PBSC: 92 UC: 4 Older Bone marrow: 2 PBSC: 94 UC: 4	2-MWT	HAP SF-36 FACT-BMT L-cGVHD-SC	-Older patients had lower psychological symptom burden and lower functional status compared to AYA and middle-aged patients. -HAP scores were significantly lower for older patients compared to AYA and middle-aged patients (p<0.001). -FACT-BMT scores were higher for older patients compared to middle-aged patients, and similar to AYA patients. -Older and AYA had comparable FACT-BMT scores, which were higher than middle-aged patients scores. -Older patients reported lower SF-36 PFS scores compared to AYA and middle-aged patients, but less bodily pain (p=0.007).
(71)	N=567 Median age in years (range): 51 (2–79). 98% adults. M: 57; F: 43	-	Mild: 9 Moderate: 52 Severe: 39	-	Bone marrow: 7 PBSC: 89 UC: 4	2-MWT HGS test	HAP SF-36 FACT-G L-cGVHD-SC	-Lower gastrointestinal involvement (HR 1.67, p=0.05) and elevated bilirubin (HR 2.46, p=0.001) were associated with overall survival; both were also associated with non-relapse mortality. In multivariable analysis using all visits, gastrointestinal score greater than zero (HR 1.69, p=0.02) and elevated bilirubin (HR 3.73, p<0.001) were associated with overall survival; results were similar for non-relapse mortality. -Any esophageal involvement and gastrointestinal score greater than zero were associated with both symptoms and QoL. Elevated bilirubin was associated with QoL. -Upper gastrointestinal involvement was associated with an estimated of 31.7 feet less (p<0.001), and overall liver involvement was associated with an estimated of 19.4 feet less (p=0.001) achieved in the 2-MWT. Upper gastrointestinal was associated with HGS test (p=0.04).
(81)	N=584 Median age in years (range): 51.5 (2–79) 98% adults M=58; F=42	-	None: 1 Mild: 8 Moderate: 52 Severe: 39	-	Bone marrow: 7 PBSC: 88 UC: 5	2-MWT HGS test	HAP SF-36 FACT-BMT L-cGVHD-SC	-The 2-MWT was significantly associated with domains of the L-cGVHD-SC (overall, skin, lung, energy), SF-36 domain and summary scores, FACT summary and domain scores, and HAP scores (all p<0.001). -The 2-MWT and HGS both had significant association with global cGVHD severity. -Impaired performance in the 2-MWT was associated with higher odds of gastrointestinal, liver, and lung involvement. -Two-MWT was significantly associated with overall survival, non-relapse mortality, and failure-free survival, while no association was found for HGS test.
(84)	N=567** Present: N=164 Median age in years (range): 52 (42–58) M=57; F=43 Absent: N=403 Median age in years (range): 51 (42–60) M=58; F=42	-	Present: Mild: 2 Moderate: 50 Severe: 48 Absent: Mild: 12 Moderate: 53 Severe: 35	-	Present: Bone marrow: 6 PBSC: 89 UC: 5 Absent: Bone marrow: 7 PBSC: 89 UC: 4	2-MWT HGS test ROM (NIH joint/fascia scale) P-ROM	HAP SF-36 FACT-G L-cGVHD-SC	-There was a 29% incidence of joint/fascia manifestations in patients with cGVHD. -P-ROM scale was the most sensitive to perceived joint worsening among all scales. -The Lee muscle/joint symptom subscale was useful for capturing changes in joint-specific symptoms -The L-cGVHD-SC was useful for capturing changes in overall symptoms. -The FACT-G was sensitive to worsening but not to improvement, while the converse was true of the SF-36 PCS. -Joint improvement was evident by 3 months after the onset of joint/fascia manifestations, and these patients experienced worsening in ROM within 6 months if joint/fascia manifestations developed >3 months after diagnosis of cGVHD.
(85)	N=9 Median age in years (range): 47 (13–53) M=67; F=33	-	Mild: 11 Moderate: 22 Severe: 67	Classic: 67 Overlap: 33	Bone marrow: 55 PBSC: 45	2-MWT HGS test ROM	-	Results of this study support the need of training investigators interested in participating in cGVHD clinical trials and the need for simplifying current tools to evaluate the cutaneous involvement measurements.
(86)	N=283 Median age in years (range): 51 (2–79). M=59; F=41 Incident cases: 53%. Prevalent cases: 47%	-	Mild: 13 Moderate: 59 Severe: 28	Classic: 17 Overlap: 83	Bone marrow: 6 PBSC: 89 UC: 5	PFTs (and NIH symptom-based lung score)	HAP SF-36 FACT-BMT L-cGVHD-SC	-Complete or parcial responses at 6 months after enrollment correlated with changes in symptom burden among incident cases, but not among prevalent cases. -Response at 6 months was not associated with changes in QoL or survival outcomes in either group. -Modifications of the current response algorithm are needed to improve correlations with clinical benefits.
(87)	N=496 97% adults. 3% child. M: 58; F: 42	-	<mild: 1 Mild: 8 Moderate: 53 Severe: 38	-	Bone marrow: 7 PBSC: 88 UC: 5	PFTs (and NIH symptom-based lung score)	SF-36 FACT-BMT L-cGVHD-SC (Lee lung symptom score)	-Pulmonary dysfunction was present at enrollment in 34% of patients based on PFTs and 25% based on symptoms. -The NIH symptom-based lung score was associated with nonrelapse mortality (p=0.02), overall survival (p=0.02), patient-reported symptoms (p<0.001) and functional status (p<0.001). -Worsening of NIH symptom-based lung score over time was associated with higher nonrelapse mortality and lower survival. -NIH symptom-based lung score was correlated with the Lee lung symptom score, Lee summary symptom score, SF36-PCS, and FACT-BMT trial outcome index (all p<0.001).

All patients included were evaluated according to the National Institute of Health cGVHD staging form, with ≥70% of the patients in each study cohort having cGVHD and being predominantly adults.

Abbreviations: AAS, adjusted activity score;AYA, Adolescent and young adult;allo-HSCT, allogeneic hematopoietic cell transplantation; CI, confidence interval; F, female;FACT-BMT, Functional Assessment of Cancer Therapy of Bone Marrow Transplant; FACT-G,Functional Assessment of Cancer Therapy-General;FEV₁, forced expiratory volume in 1 second; HAP, Human Activities Profile; HGS, hand grip strength test;HR, hazard ratio; L-cGVHD-SC, Lee cGVHD Symptom Scale;M, male; MAS, maximum activity score;MCS, mental component score; NIH, National Institutes of Health;OR, odd ratio; PBSC, peripheral blood stem cells; PCS, physical component score; PFTs, pulmonary function tests; P-ROM, Photographic-ROM;ROM, range of motion; SF-36, Short Form 36; UC, umbilical cord; 2-MWT, 2-minute walk test. Definitions: Progressive onset of chronic GVHD is that following unresolved acute GVHD, quiescent onset of cGVHD is that occurring after complete resolution of acute GVHD, and de novo onset of cGVHD is that appearing in patients who have not had previous acute GVHD.

^**Characteristics of the patients grouped by presence or absence of joint/fascia manifestations at the time of enrollment.

Functional capacity tests

The 2-minute walk test (2MWT) (80) has been used to assess a dimension of the therapeutic response in patients with cGVHD (12). In patients with cGVHD, an abnormal 2MWT result has been associated with a higher symptom burden, impaired QoL, functional disability, increased mortality (81) and NIH global severity score (53). To assess arm strength as an indicator of muscle mass and the nutritional state of the patient, the hand grip strength (HGS) test has proved useful (82). However, despite significant correlation detected between NIH global severity and HGS test scores (81), NIH do not support its use (53). To assess respiratory capacity, the NIH lung scoring system has two parts: a clinical lung symptom score based on symptoms and the lung function score (LFS) which is used when pulmonary function tests are available (9). As a useful measure of joint response, the NIH consensus group recommend active-assisted upper and lower body range of motion (ROM) (12). NIH global severity scores show significant correlation with ROM (53). However, the need for a trained clinician to conduct ROM measurements is a major limitation of this technique (12). As an objective and simply alternative, the photographic-ROM (P-ROM) scale was developed (83,84) to assess joint/fascia manifestations in patients with cGVHD. In addition, the NIH joint/fascia scale was designed to evaluate the severity of GVHD manifestations in joints and fascia for baseline or cross-sectional use (9), though this test could also be used to assess the response to a treatment intervention (84).

The reduced functional capacity of patients with cGVHD has been confirmed by measurements made in the tests 2MWT (15,53,61,70,71,81,84,85), HGS (15,53,61,71,81,85,86), forced expiratory volume (FEV₁) (15,53,86,87), ROM (53,61,84,85) and P-ROM (84). Such functional decline is more marked in patients ≥60 years (70) or with overlap syndrome (as indicated by a longer 2MWT in both cases) (15). Patients with moderate-to-severe cGVHD requiring moderate-to-high levels of immunosuppression experience significant functional limitations (61), although the findings of some studies suggest no detrimental effects on QoL of immunosuppressive therapy (63,88).

Patient’s self-reports of functional capacity and physical domain of QoL

The Human Activity Profile (HAP) is a self-reported assessment tool in which activities are rated according to the energy expenditure required to perform a specific task (12,50,89). This questionnaire has proved more sensitive to changes in cGVHD status and to the presence of toxicity than other self-reported assessment methods [such as the Medical Outcomes Study 36-Item Short-Form Health Survey (SF-36), and the Functional Assessment of Cancer Therapy-Bone Marrow Transplant (FACT-BMT)] (50). Hence, it may be of use to measure the impacts of severity of cGVHD and side effects of immunosuppression on physical functioning, and the response to cGVHD therapy in clinical trials (50). Moreover, HAP shows a high internal consistency with cGVHD severity (50). The SF-36, v.2, assesses patient-reported functional health, functioning and well-being (90). The use of a norm-based scoring system helps interpret SF-36 scores such that any score below 50 is below the US population mean, and each point represents one-tenth of a standard deviation, with this scale allowing to assess some dimensions of the response to cGVHD treatment (12,90). The FACT-BMT v4.0 questionnaire includes a 10-item Bone Marrow Transplant Subscale (BMTS) (91). No significant differences between SF-36 and FACT-BMT in discriminating cGVHD severity have been detected, particularly in the physical domains (65). However, the FACT-General core survey has been recommended over SF-36 and the FACT-BMTS subscale since the core survey substantially reduces respondent burden without compromising multi-dimensional QoL comprehension or the tool's sensitivity to change (66). The Lee cGVHD Symptom Scale (L-cGVHD-SC) evaluates several symptoms of cGVHD (79) and may be effective to monitor these dimensions of the response to cGVHD therapy (12).

Chronic GVHD-affected individuals show modified SF-36 physical component scores (PCS) (61,66) and those with severe cGVHD show a reduced SF-36 mental component score (66). In patients older than 60 years compared to those younger than this age, HAP scores are lower, though QoL (FACT-BMT) is improved over that observed in patients aged 41 to 59 years (70). Further, patients with acute/chronic GVHD overlap syndrome return worse HAP and social functioning scores (SF-36) than patients with classic cGVHD (15). NIH global severity scores have been correlated with scores obtained in L-cGVHD-SC, FACT-BMT, HAP and SF-36 PCS, with the exception of the mental component of the SF-36 score (53). Those with moderate to severe cGVHD show similar SF-36 PCS scores to those reported for patients with immune-mediated disorders, yet these scores are lower than those observed in patients with common chronic medical conditions (65).

Rationale for exercise Interventions in patients with cGVHD: what can we learn from research in other chronic conditions

Given the serious consequences of cGVHD and the secondary effects of conditioning regimens, allo-HSCT, and drug treatments, new approaches to the care of patients with cGVHD are urgently needed. There is strong scientific evidence to support the therapeutic benefits of non-pharmacological interventions such as regular physical exercise [see (8) for a review of the main therapeutic and preventive benefits of exercises and some of the molecular mechanisms involved)]. This lifestyle intervention is an effective coadjuvant treatment against several chronic diseases, especially those in which an inflammatory state is produced (e.g., cancer). Its benefits include an improved cardiorespiratory capacity, measured as peak oxygen uptake, VO₂peak. The latter is one of the variables that best responds to physical exercise, and is also a strong prognostic factor of mortality attributable to chronic disease. Further benefits include improved functional and muscular capacity, diminished systemic inflammation and improved immune function, reducing some of the side-effects of cancer treatment such as fatigue (8). Despite such benefits, however, at present there is no scientific evidence of the effects of exercise in patients with cGVHD. So far, physical exercise interventions have targeted patients subjected to allo-HSCT though some programs have included patients with GVHD (92). Unfortunately, none of the studies assessing the effects of exercise on allo-HSCT recipients has analyzed separately the subset of patients suffering this condition, and in fact the authors did not even make a distinction between aGVHD and cGVHD [see (92) for a review]. Keeping these limitations in mind, among the positive exercise impacts on QoL are improved endurance/aerobic capacity, muscular strength, functional capacity, perceptions of fatigue, physical emotional and social well-being, and reductions in perceived pain scores and subdued anxiety, depression and aggressive or hostile behavior (92). Although patient cohorts and experimental designs in these studies are very heterogeneous, it seems likely that both aerobic- and resistance based exercise appears to positively impact QoL in GVHD patients.

Preclinical exercise intervention studies in cGVHD

Owing to the lack of exercise interventional research in patients, our research group decided to conduct preliminary, preclinical studies. Thus, we examined the effects of exercise in an experimental scleroderma [B10.D2 (H-2^d) → BALB/C (H-2^d)] mouse model of cGVHD. In a first study, we showed that the functional capacity (determined as maximal running velocity attained during a gradual treadmill test until exhaustion) of mice with cGVHD and receiving no drug treatment inevitably declined 12 weeks after allo-HSCT (with donor bone marrow cells and splenocytes), which was preceded by myeloablative irradiation. Yet compared to the control group, such decline was significantly less marked in those mice that had followed a moderate-intensity aerobic treadmill running program (5 sessions/week, Monday to Friday) that started shortly (2 days) after allo-HSCT and was prolonged until ~11 weeks post-transplant. The training program mimicked the widely accepted public health recommendations for physical activity in humans, that is, accumulating 30+ min of such type of moderate-intensity exercise (e.g., brisk walking, jogging) in most (if not all) days of the week (93). The cGVHD has catabolic effects at the muscle tissue level and also leads to homeostatic and hormonal alterations (92). Despite such disease effects, the trained mice showed higher activity of the musculoskeletal enzyme citrate synthase, a key marker of oxidative capacity and mitochondrial biogenesis. Thus, we propose that a shift towards a more ‘oxidative’ muscle phenotype could have positive repercussions on health and resistance to certain diseases (93). In fact, classic studies in rats by Koch and Britton’s group showed that improved oxidative pathways in mitochondria may be a common factor linking physical fitness and decreased disease risk (94,95). Further, exercise training did not negatively affect the kinetics of immune reconstitution in the context of severely debilitated and immunocompromised mice receiving lethal irradiation and transplantation (93). This finding suggests that the exercise would not negatively affect the graft-versus-tumor or the graft-versus-infection effect of the transplant. Further experiments using models adapted for these two situations (i.e., tumor or infectious agents) will elucidate these clinically important questions. Although the beneficial effects of moderate-intensity exercise on the immune function of healthy individuals are well documented (96), there is more controversy regarding its effects in immunocompromised people (97) and our results in mice with cGVHD are in line with previous findings showing no benefit but also no harm on immune reconstitution in patients with hematological cancer (98,99). Also noted in this first study, were reduced IL6 levels (93). Thus, our preliminary findings suggest that exercise might attenuate the severe physical decline (and thus benefit the physical domain of QoL) in debilitated and immunocompromised patients, such as those suffering cGVHD.

In a second study (100), we examined the effects of the same type of exercise training in the same cGVHD model, which received a standard immunosuppressive treatment for this disease, i.e., cyclosporine-A. Immunosuppressant drugs induce severe side effects at the multisystem level including muscle tissue deterioration (101,102), with some human studies suggesting a link between cyclosporine-A and myopathies associated with heart or liver transplant or with the treatment against Graves disease (103). In animal models, chronic administration of cyclosporine-A reduces physical capacity, and the magnitude of the decline is linked to the extent of mitochondrial alteration (104). Importantly, the results of our second study suggested beneficial effects on survival, the clinical course of cGVHD and on physical capacity in the exercise group, compared with the control group (100). Further, the exercise intervention had a favorable effect on different immune cell compartment kinetics, led to a less aggressive inflammatory profile, and offset the toxic effects of the immunosuppressive treatment (100). Specifically, the scleroderma cGVHD mouse model that we used is dependent on CD4⁺ cells, which activate Th2 lymphocytes secreting IL2, IL5, IL9, IL10, IL13 and TNFα. Growing levels of these cytokines are linked to the disease progression (105,106). Our exercise program led to lower levels of TNFα and IL4 after 12 weeks post-transplant compared to the control group (100). Further research is needed to determine the mechanisms explaining such exercise effects on cytokine profile. On the other hand, patients with cGVHD have reduced levels of B- and CD4⁺ T-helper cells (29) which, together with the altered lymphopoiesis and the drug effects, result in long-lasting immunodeficiency (107). Thus, the finding that the treatment program did not negatively affect immune reconstitution, and in fact tended to favor it, at least partly (i.e., we found higher blood levels of B220 and CD4⁺ cells in the exercise group at 21 days post-allo-HSCT), has potential clinical relevance (100). In this regard, Ko et al. (108) showed an improvement in the immune function of healthy mice treated with cyclosporine-A that followed an 8-week moderate-intensity treadmill training compared to those receiving the drug but not performing the training, suggesting a beneficial immunomodulatory effect of exercise that is yet to be corroborated in humans receiving the dame drug.

In a final study, physical exercise promoted the induction of myocardial autophagy in trained mice surviving 12 weeks following allo-HSCT, suggesting this could be one of the biological mechanisms mediating the beneficial effects of exercise in cGVHD in the later stages of life (109). Indeed, autophagy, a catabolic route of degradation and recycling of cellular components in all tissues, is an important cardioprotective mechanism that helps to ‘declutter’ the cell and restore its functionality, including in pathological and aging conditions, with suppression of genes involved in autophagy leading to altered cardiac function (110–112). Several studies also showed improved myocardial autophagy in trained rodents (113–116). In addition, our data also showed higher levels of antioxidant enzymes in the hearts of exercised mice (109). Thus, both training-enhanced autophagy and myocardial antioxidant capacity might confer heart-muscle protection against radiation toxicity. In this regard, De Lisio et al. showed that exercise training enhances mouse skeletal-muscle response of antioxidant and mitochondrial enzymes to radiation (117).

From preclinical studies to clinical trials: Suggested feasibility and potential limitations of exercise interventions in patients with cGVHD

It must be kept in mind that exercise interventions targeting to attenuate the physical decline of patients with cGVHD and the impact of such decline in their daily living and QoL are likely to face numerous limitations, especially at end-stages of the disease and/or in the more severely affected individuals. Figure 2 shows the clinical aspects of the disease that are more likely to affect the exercise capacity and adherence to the program. Notably, not only the manifestations of the disease per se but also the side effects of the treatment and prolonged bed rest can further deteriorate the patients’ physical condition and physical domain of QoL. The negative impact of the disease at multisystem level (cardiorespiratory, immune and gastrointestinal systems) can decrease the capacity of the patient to improve aerobic capacity and muscle anabolism in response to training interventions. Many patients have low walking capacity, or even low capacity for standing from a chair, which limits the range of exercises that can be applied. Nonetheless, our group has shown the feasibility, safety and benefits of individualized exercise interventions (often performed in the hospital setting) usually combining low-moderate intensity aerobic and resistance (i.e., weight lifting) exercises in other debilitated, fragile populations, including nursing home resident nonagenarians (118), hospitalized octogenarians (119) (most of whom, like the nonagenarians, had very poor walking capacity and serious difficulties in coping independently with common activities of daily living). Implementation of specific types of weight lifting training (i.e., leg press at low loads or simply standing from a chair several times a day) is feasible, safe and effective even in the ‘oldest old’ people (90+ years) (118) or in anorexic adolescents who often suffer metabolic myopathy due to protein malnutrition (120). Further, special ‘in-room’ exercise interventions (e.g., including light weightlifting while lying in bed) are effective and feasible in isolated, highly immunocompromised children during inpatient hospitalization for pediatric allo-HSCT (98). Finally, other special and simple interventions such as specific inspiratory load training to improve the function of breathing musculature with special, small size breathing devices (e.g., ‘Power breathe’) are easily applicable even in patients’ homes, as we have shown in children with cystic fibrosis (121). In the above mentioned fragile populations, besides the almost unanimous benefits in aerobic capacity or muscle strength, in some patient populations the exercise interventions also produced improvements in: (i) the individuals’ ability to cope independently with daily living (e.g., decreased risk of falls in nonagenarians (118) or improved performance in functional tests in children who had received allo/haploidentical HSCT) (122); or (ii) self-reported QoL, e.g., improved self-report of comfort and resilience in pediatric patients who had undergone HSCT (122).

On the other hand, based on our own experience, previous research on pediatric/adult hematological cancer (123), as well as on specific recommendations by leading institutions such as the American College of Sports Medicine (124) which state that adult cancer survivors should avoid inactivity as soon as possible after diagnosis and in fact adhere to the 2008 Physical Activity Guidelines for Americans (i.e., adults should undertake ≥ 150 min/week of moderate physical activity such as brisk walking); we recommend that patients with cGVHD also adhere to these guidelines as soon as possible after diagnosis of the disease. In fact, an active lifestyle should be ideally adopted upon diagnosis of the disease that resulted in allo-HSCT.

Future Directions

Our results reveal the potential therapeutic value of regular physical exercise in allo-HSCT recipients who suffer the devastating effects of cGVHD. Given the foreseeable increase in the number of patients that will receive a transplant from an unrelated donor in the years to come, the incidence of GVHD will also rise and this will be accompanied by increased healthcare costs. To further our understanding of cGVHD, research efforts need to focus on determining the optimal mode, intensity and volume of exercise that will lead to an improved clinical course and QoL of these patients. It will also be important to determine if exercise has a direct impact on the pathobiology of GVHD or if it can serve as an adjuvant for ATG (anti-thymocyte globulin) therapy or other pharmacological interventions aimed at dampening the severity and/or lowering the incidence of GVHD. Moreover, the timing of when the exercise intervention is delivered may also influence GVHD incidence and severity, be it before (‘prehabilitation’), during or after hospitalization for allo-HSCT. To guarantee the validity, reliability and reproducibility of the data emerging from future work, variables of interest need to be measured using adequate tools by interdisciplinary teams of experts who care for patients with cGVHD.