Myositis, polyserositis with a large pericardial effusion and constrictive pericarditis as manifestations of chronic graft-versus-host disease after non-myeloablative peripheral stem cell transplantation and subsequent donor lymphocyte infusion


The clinical features of chronic graft-versus-host disease (cGVHD) following a non-myeloablative peripheral blood stem cell (PBSC) transplant may differ from those that occur after a conventional allograft. We describe a man with Hodgkin's disease refractory to chemotherapy and radiotherapy who was transplanted from an HLA-identical brother, who developed cGVHD characterised, in particular, by polymyositis, polyserositis with a large pericardial effusion and constrictive pericarditis, 1 month after donor lymphocyte infusion for relapsed disease. Constrictive pericarditis has not been previously reported after a conventional allograft, and none of these features have been reported after a non-myeloablative transplant. The course of cGVHD necessitated potent immunosuppression leading to the presumed loss of graft-versus-lymphoma (GVL) effect. Bone Marrow Transplantation (2001) 27, 231–233.


The current approach to the diagnosis and management of chronic graft-versus-host disease (cGVHD) is based on extensive experience of conventional allografting. It is unknown whether clinical manifestations and response to treatment may be different, if cGVDH occurs in the setting of a non-myeloablative transplant. We describe a patient who developed myositis, polyserositis with a large pericardial effusion and constrictive pericarditis following a non-myeloablative stem cell transplant and subsequent donor lymphocyte infusion. cGVHD progressed despite treatment with steroids and cyclosporin A.

Case report

A 36-year-old man with Hodgkin's disease (nodular sclerosis) who was in second complete remission after BEAM chemotherapy and autologous stem cell rescue presented with back pain. Computed tomography (CT) demonstrated retroperitoneal relapse. Physical examination and routine pre-transplant investigations (blood count, biochemistry studies, echocardiogram) were normal, apart from lung function tests which showed slightly reduced transfer factor (67% predicted). HIV, HbsAg tests and viral serology (hepatitis C, VZV, CMV) were negative in both the patient and the donor.

He underwent a non-myeloablative PBSC transplant from his HLA-identical brother. Conditioning included fludarabine 25 mg/m2 (day −6 to day −2) and cyclophosphamide 1 g/m2 (day −3 and −2 ). GVHD prophylaxis consisted of methotrexate 10 mg/m2 (days +1, +3, +6 and +11) and cyclosporin A 5 mg/kg from day −1, reduced to 3 mg/kg from day +4 and adjusted to maintain the level between 150 and 300 μg/l. The dose of nucleated cells infused on day 0 was 6.44 × 108/kg and CD34+ cells 4.86 × 106/kg. Chimerism was evaluated by analysis of minisatellite variable number of tandem repeats on whole blood. The patient had 63% donor-derived cells on day 16 and converted to full donor chimaerism by week 12. The dose of cyclosporin A was reduced from week 13.

Six months after the transplant he reported recurrence of the back pain. CT and positron emission tomography (PET) findings were consistent with relapse in the left para-aortic region. Cyclosporin A was discontinued and he was treated with radiotherapy to the site (35 Gy in 20 fractions). However, the back pain did not improve and the patient received a donor lymphocyte infusion (1 × 107 CD3+ cells/kg).

Clinical response to the procedure was difficult to assess, since 3 weeks later he started to complain of severe myalgia predominantly affecting the muscles of the lower limbs, which coincided with the development of cGVHD of the skin and the liver (both biopsy proven). There was no muscle weakness or CK rise. He was commenced on 40 mg of prednisolone with some symptomatic improvement.

At the end of the following month the patient presented with worsening dyspnoea, peripheral oedema and signs of cardiac tamponade. Chest X-ray showed cardiomegaly and a small right- sided pleural effusion. Echocardiogram confirmed the presence of a large pericardial effusion which was drained. The fluid was sterile and contained no malignant cells.

The next month was remarkable for the development of keratoconjunctivitis sicca confirmed with the dry Schirmer's test, Raynaud's phenomenon, widespread arthralgia with global restriction of large joints in a capsular pattern, worsening myalgia and proximal muscle weakness. Antiınuclear antibodies were positive (titre 1/80) with a nucleolar pattern, having been known to be negative before the transplant. An antibody to unidentified extractable nuclear antigen was detected, but anti-Scl-70, anti-centromere and anti-PM/Scl antibodies were negative. CK remained normal and EMG showed no evidence of a myopathic process. However, MRI appearance of the lower limbs muscles was consistent with myositis which was subsequently confirmed on muscle biopsy (Figure 1).

Figure 1

(a) Coronal T1 weighted magnetic resonance image of the lower thighs. Muscle and femoral marrow return a normal signal and a muscle bulk is normal. (b) Coronal short-tau inversion recovery (STIR) sequence at the same anatomical level. The normal high signal from subcutaneous fat and marrow has been nulled. There is diffuse high signal oedema within the muscles, indicating an active myositis.

Three weeks later the night sweats recurred and the steroids were withdrawn, which resulted in severe generalised muscles weakness. CK rose to 492 IU/l and the EMG at this time demonstrated classic myopathic features. The patient made a dramatic clinical response to intravenous methylprednisolone.

He was subsequently re-commenced on cyclosporin A in addition to oral steroids. Despite aggressive immunosuppression, his clinical course was characterised by significant residual muscle weakness, extension of cutaneous involvement and, in particular, worsening dyspnoea, peripheral oedema and markedly deranged liver function tests (cholestatic pattern). Echocardiogram was normal, but CT demonstrated new pericardial thickening (Figure 2) and cardiac catheterisation confirmed the diagnosis of constrictive pericarditis. The CT showed no evidence of disease relapse. In view of progression of cGVHD, the patient was treated with anti-lymphocyte globulin (ALG). He developed hypotension and oliguria after the first infusion of horse ALG and required a short admission to the intensive care unit. He was later treated with rabbit ALG which he tolerated well. However, the night sweats persisted. A PET scan at that time was consistent with relapse of Hodgkin's disease in the neck and multiple visceral sites.

Figure 2

Axial CT scan with intravenous contrast medium at the level of the right and left ventricles. Note anterior pericardial thickening (arrows).


Non-myeloablative PBSC transplant combined with donor lymphocyte infusion (DLI) is a novel method of immunotherapy for refractory haematological malignancies.1 Recently published studies have demonstrated promising results, including those in patients with Hodgkin's disease.2

We describe a patient with relapsed Hodgkin's disease who developed extensive cGVHD following a non-myeloablative PBSC transplant and subsequent donor lymphocyte infusion from an HLA- identical donor. His presentation was dominated by polymyositis, polyserositis with a large pericardial effusion and constrictive pericarditis.

Polymyositis has been reported in 23 recipients of conventional allogeneic bone marrow transplantation.3 The clinical presentation is indistinguishable from that of idiopathic polymyositis. Unlike our case, 10 out of 11 patients achieved a complete remission following treatment with cyclosporin A and prednisolone although one patient had mild residual weakness. CK ranged from 742 IU/l to 6000 IU/l with a mean of 3519 IU/l. Interestingly, in our case, raised CK was only detected once, being moderately elevated at 492 IU/l despite profound generalised weakness. It is also of note that the CK rise and EMG changes lagged behind the clinical, histological and radiological findings.

Polyserositis with pleural and pericardial effusions has been reported after a conventional allograft.4 However, as far as we are aware, constrictive pericarditis and the pattern of organ involvement by cGVHD seen in our patient have not been documented previously.

We suggest that transient mixed chimaerism may play an important role in the pathogenesis of cGVHD in the setting of a non-myeloablative transplant. Animal studies and clinical trials have demonstrated that the presence of host haemopoietic cells in the mixed chimeras dictates the severity of graft-versus-host reactions.5 The potency of this effect has recently been proven by a successful HLA-mismatched BMT which exerts a graft-versus-lymphoma effect in the presence of stable mixed chimerism with grade II acute GVHD amenable to treatment.6 One can also speculate that the co-existence of donor and recipient haemopoiesis provides the environment for interaction between the donor T cell and recipient antigen-presenting cells leading to the generation of additional allo-reactive T cell clones.

DLI is an established rescue procedure for relapsed disease after a conventional allograft, with GVHD being the major complication.7 Administration of DLI in multiple escalating doses appears to reduce the incidence and severity of GVHD.8 In the non-myeloablative transplant, this can be achieved by giving the DLI during the phase of mixed donor chimaerism.910 Thus, we propose that administration of the DLI as a single dose after the establishment of full donor chimaerism together with cessation of inhibitory effect of host haemopoietic cells may explain the severity of cGVHD in our patient.

In summary, we report an unusual presentation of GVHD after a non-myelablative PBSC transplant and DLI from an HLA-identical sibling donor for progressive Hodgkin's disease, which necessitated the use of potent immunosuppression resulting in loss of graft-versus-lymphoma effect.

Figure 3

Skeletal muscle biopsy showing variation in fibre size and an interstitial infiltrate of small round lymphoid cells. The latter are CD3 positive by immunoperoxidase staining. H&E ×640.


  1. 1

    AM Carella, R Champlin, S Slavin et al. Mini-allografts: ongoing trials in humans Bone Marrow Transplant 2000 25: 345–350

    Article  Google Scholar 

  2. 2

    P McSweeney, D Niederwieser, J Shizuru et al. Outpatient allografting with minimally myelosuppressive, immunosuppresive conditioning of low-dose TBI and postgrafting cyclosporin and mycophenolate mofetil Blood 1999 94:: (Suppl. 1) 393a (Abstr.)

    Google Scholar 

  3. 3

    M Parker, H Openshaw, S Forman . Myositis associated with graft-versus-host disease Curr Opin Rheumatol 1997 9: 513–519

    Article  Google Scholar 

  4. 4

    A Toren, A Nagler . Massive pericardial effusion complicating the course of chronic GVHD Bone Marrow Transplant 1997 9: 805–807

    Google Scholar 

  5. 5

    PA McSweeney, R Storb . Mixed chimerism: preclinical studies and clinical applications Biol Blood Marrow Transplant 1999 5: 192–203

    Article  Google Scholar 

  6. 6

    M Sykes, F Preffer, S McAfee et al. Mixed lymphohaemopoetic chimerism and graft-versus-lymphoma effects after non-myeloablative therapy and HLA-mismatched bone marrow transplantation Lancet 1999 353: 1755–1759

    Article  Google Scholar 

  7. 7

    RH Collins, O Shpilberg, W Drobyski et al. Donor leukocyte infusion in 140 patients with relapsed malignancy after allogeneic bone marrow transplantation J Clin Oncol 1997 15: 433–444

    Article  Google Scholar 

  8. 8

    F Dazzi, RM Szydlo, C Craddock et al. Adoptive immunotherapy for patients who relapse after allografting for chronic myeloid leukaemia: a comparison of two regimens of donor lymphocyte infusion Blood 1998 92: (Suppl. 1) 653 (Abstr.)

    Google Scholar 

  9. 9

    TR Spitzer, SL McAfee, R Sackstein et al. Induction of mixed chimerism and potent anti-tumour responses following nonmyeloablative conditioning therapy and HLA-matched and mismatched donor bone marrow transplantation for refractory haematological malignancies Blood 1998 92: (Suppl. 1) 519a (Abstr.)

    Google Scholar 

  10. 10

    M Sykes, SL Saidman, F Preffer et al. Mixed lymphohaemopoetic chimerism in adult recipients following nonmyeloablative therapy and HLA-matched or mismatched bone marrow transplantation, and the effect of donor leukocyte infusion Blood 1998 92: (Suppl. 1) 520a (Abstr.)

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to J Cavenagh.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Silberstein, L., Davies, A., Kelsey, S. et al. Myositis, polyserositis with a large pericardial effusion and constrictive pericarditis as manifestations of chronic graft-versus-host disease after non-myeloablative peripheral stem cell transplantation and subsequent donor lymphocyte infusion. Bone Marrow Transplant 27, 231–233 (2001).

Download citation


  • polyserositis
  • pericardial effusion
  • constrictive pericarditis
  • myositis
  • graft-versus-host disease

Further reading