Bone Marrow Transplantation, (1999) 23, 881–887 1999 Stockton Press All rights reserved 0268–3369/99 $12.00 http://www.stockton-press.co.uk/bmt Extracorporeal photopheresis (ECP) in the treatment of chronic graft- versus-host disease (GVHD) FJ Child1, R Ratnavel1, P Watkins1, D Samson2, J Apperley2, J Ball3, P Taylor3 and R Russell-Jones1 1Skin Tumour Unit, St John’s Institute of Dermatology, St Thomas’ Hospital, London; 2Department of Haematology, Hammersmith Hospital, London; and 3Department of Haematology, Rotherham District General Hospital, Rotherham, UK Summary: morbidity and mortality. Acute GVHD affects 40–50% of allogeneic stem cell transplant recipients and chronic The aim of our study was to assess the efficacy of extra- GVHD affects 30–50% of those surviving the acute phase corporeal photopheresis (ECP) in chronic graft-versus- of transplantation.1,2 host disease (GVHD). Eleven patients with chronic Prevention of GVHD is important because treatment of cutaneous GVHD were studied. Four had mucosal established disease is often unsatisfactory. Measures such involvement and five had pulmonary involvement. All as pre-transplant T cell depletion and immunosuppression had failed to improve on first- and second-line therapy. with methotrexate, cyclosporin A and prednisolone are Three patients received ECP alone; the remainder con- routinely undertaken. tinued to receive steroids and/or immunosuppressive Chronic GVHD may be limited to localised skin involve- therapy. Patients received ECP twice monthly for 4 ment and liver dysfunction, or the disease may be more months and then once monthly for 3 months. They were extensive with generalised skin involvement and additional evaluated by serial skin scores, mucosal and skin pho- organ involvement. Patients with limited disease may not tography, pulmonary function tests, biochemical and require any treatment; however, those with extensive dis- haematological parameters. Nine patients showed objec- ease require treatment to decrease the chance of progression tive evidence of cutaneous improvement with a mean to end stage disease. Many patients respond to oral gluco- reduction in skin score of 48% overall. In the 10th corticoids3 and others respond to a combination of oral glu- patient, skin scores and oral involvement improved on cocorticoids and cyclosporin A.4 Refractory chronic GVHD twice monthly ECP but deteriorated when reduced to has been successfully treated with other agents including once monthly. The final patient died from renal failure azathioprine, thalidomide5 and photochemotherapy in the secondary to cyclosporin toxicity. Two out of five form of PUVA.6–10 There have also been a number of anec- patients with lung involvement showed a mild improve- dotal case reports11–13 and small studies14,15 reporting the ment in pulmonary function tests. Liver function tests success of extracorporeal photopheresis (ECP). were abnormal in five patients; they improved in one The aim of our study was to assess the safety and effi- and deteriorated in three. All patients receiving con- cacy of UVAR/UVADEX ECP (Therakos UK, Bracknell, comitant immunosuppressive/steroid therapy were able UK), in the treatment of patients with skin manifestations to reduce drug dosages by trial completion. Our results of chronic GVHD which had failed to respond to indicate that ECP can benefit patients with cutaneous standard therapies. and mucosal chronic GVHD who have failed on first- and second-line therapies. The effect on the systemic manifestations of GVHD is less consistent. Patients and methods Keywords: extracorporeal photopheresis; graft-versus- host disease; bone marrow transplantation Eleven adult patients of median age 40 years (range 18–47 years) were studied. All had received HLA-matched bone marrow transplants for haematological malignancies, from sibling (n = 8) or unrelated (n = 3) donors. They were Graft-versus-host disease (GVHD) is a frequent and major recruited nationally from Haematology and Dermatology complication of allogeneic stem cell transplantation. It is a departments within the United Kingdom. They were treated systemic disease with prominent cutaneous manifestations. at two centres: the Department of Haematology, Rotherham Ͻ It occurs either as an acute ( 100 days post- District General Hospital (patients 1–6) and St John’s Insti- Ͼ transplantation) or a chronic ( 100 days post- tute of Dermatology, St Thomas’ Hospital, London transplantation) disease and contributes significantly to (patients 7–11). Informed consent to take part in the study was obtained from all patients and approval from the local ethics committees had been granted. All patients had Correspondence: Dr F Child, Skin Tumour Unit, St John’s Institute of Dermatology, St Thomas’ Hospital, Lambeth Palace Road, London SE1 chronic cutaneous GVHD, and all had failed to respond to 7EH, UK conventional immunosuppressive protocols. All patients Received 3 September 1998; accepted 26 November 1998 had at least one of the following clinical features: Extracorporeal photopheresis in chronic GVHD FJ Child et al 882 (1) skin thickening, sclerosis, hyper- or hypo-pigmentation, pheresis bowl, saving the plasma and buffy coat layers. or ulceration; Forty millilitres of buffy coat and a volume of plasma were (2) joint contractures or fascial thickening; collected during each cycle. The red blood cells and all (3) oral ulceration; additional plasma were reinfused into the patient before (4) lichenoid skin changes. beginning the next collection cycle. Two hundred microgrammes of UVADEX were injected into the buffy The clinical data on all patients enrolled in the study are coat bag at the end of the first cycle. shown in Table 1. The median time from onset of chronic Photoinactivation of the leukocyte-enriched blood began GVHD to start of ECP therapy was 17 months (range 5– during the first cycle of buffy coat collection. Plasma and 59 months). Five patients had AML, three had ALL, two buffy coat were mixed with 200 ml of heparinized saline, had CML and one had multiple myeloma. Four patients had followed by pumping of the mixture continuously through mucosal involvement and five had pulmonary involvement. the photoactivation chamber where they were exposed to Three patients received ECP alone (patients 2, 3 and 8); UVA light; the tubing within the chamber was of uniform the remainder received additional steroids and/or immuno- thickness, allowing homogenous penetration of the UVA. suppressive therapy. Two were treated with prednisolone During the irradiation phase, UVADEX is activated by alone (patients 7 and 10), two with prednisolone and cyclo- UVA light, with concomitant pyrimidine base cross-linking sporin A (patients 6 and 11), patient 4 was taking predniso- of DNA in the leukocytes. The total photopheresis process lone and azathioprine, patient 5 was taking azathioprine took 3–4 h and treatment was performed on 2 consecu- and cyclosporin A and the other 2 patients (1 and 9) were tive days. taking a combination of prednisolone, azathioprine and cyclosporin A. Treatment protocol The following criteria were also fulfilled: haemoglobin Ͼ9 g/dl, platelets Ͼ20 × 109/l, neutrophils Ͼ1 × 109/l; body Patients were treated with ECP twice monthly for the first weight of Ͼ40 kg and adequate cardiac function to 4 months and were assessed clinically at 0, 4, 8 and 14 tolerate ECP. weeks. Wherever possible patients were kept on a constant dose of concomitant immunosuppressive therapy. After 4 ECP procedure months, ECP was continued once monthly for a further 3 months. Those who continued to improve were maintained Using the UVAR/UVADEX (liquid methoxsalen) system on monthly ECP following completion of the study. (Therakos UK) this involved collection of leukocyte- enriched blood, photoactivation and then reinfusion of the Cutaneous evaluation blood. Venous access was obtained peripherally or via a Hickman line and aliquots of whole blood were removed. Skin involvement was assessed by a member of the team Three hundred millilitres of plasma and 240 ml of buffy trained in the appropriate method of skin scoring. All mem- coat were separated and saved in six cycles of blood collec- bers involved in skin scoring had received a common train- tion, centrifugation and reinfusion. During each cycle, ing. Each individual patient was skin scored by the same whole blood was centrifuged and separated in a paediatric team member at each visit. Assessments were carried out Table 1 Clinical data on all patients enrolled in the study Patient 1 2 3 4 5 6 7 8 9 10 11 Age/sex 41/M 28/M 36/M 44/M 47/M 18/M 46/F 28/M 37/M 47/M 40/F Disease CML ALL AML AML AML ALL AML AML ALL IgA MM CML Donor sibling sibling sibling sibling sibling sibling MUD sibling MUD sibling MUD Onset GVHD 13 32 4 7 7 4 3 4 46 7 1 (months post-BMT) Disease type progressive progressive de-novo de-novo de-novo quiescent de-novo progressive de-novo progressive progressive Organs involved Skin + +++++++++ + Mucosal − −−+−−+−++ − Lung + +++−+−−−− − Liver + −++−−+−+− + Treatment Pre-ECP Prednisolone + −++−+++++ + Cyclosporin A + −−+++−++− + Azathioprine − −++++−−++ − Thalidomide + −−+−−−−−− + PUVA + +−−−−+−−− − Study completed yes yes no yes yes yes yes yes yes yes no MM = multiple myeloma. Extracorporeal photopheresis in chronic GVHD FJ Child et al 883 at 0, 4, 8 and 14 weeks, 6 months and at study completion. he had withdrawn from the trial after 4 months. Patient 5 Skin biopsies were taken from representative areas prior experienced an almost complete response. Figure 2 illus- to starting ECP to confirm the presence of predominantly trates the improvement in skin disease in patient 2 who lichenoid or sclerodermoid histological features. However, achieved a 63% reduction in skin score by study com- histology was not used to assess the effect of ECP as sclero- pletion. Six patients were able to reduce their dose of con- dermoid changes are difficult to grade histologically. comitant prednisolone (patients 1, 4, 6, 7, 9, 10); the dose Lichenoid changes histologically were represented clini- of cyclosporin A was also reduced in patient 1 and the dose cally by either erythema and scaling or papules.