This report concerns a case of long-lasting pure red cell aplasia (PRCA) with a duration of 178 days after major ABO-incompatible allogeneic peripheral blood stem cell transplantation (PBSCT). The patient needed red blood cell transfusion every week from day 54 following PBSCT. He showed no evidence of GVHD and the dose of cyclosporin A (CsA) was reduced rapidly from day 123, followed by the development of chronic GVHD around day 145. The patient no longer needed transfusions from day 167, the reticulocyte count began to increase on day 179, and antidonor isohemagglutinin titers became undetectable. Chronic GVHD induced by tapering of CsA thus appeared to be related to improvement in PRCA.
Major ABO incompatibility between donor and recipient red blood cells (RBC) occurs in approximately 10% to 15% of human leukocyte antigen (HLA)-matched allogeneic hematopoietic stem cell transplant (HSCT).1 Although time to neutrophil and platelet engraftment, incidence of GVHD, and survival are not affected by ABO incompatibility,2 pure red cell aplasia (PRCA) is seen in about 15–20% of HLA-matched major ABO-incompatible pairs.3,4,5 PRCA after major ABO-incompatible HSCT does not usually need any treatment except for transfusion of RBC, but may require other treatments if it persists. PRCA as a complication of major ABO-incompatible HSCT is sometimes difficult to treat.4 We describe the clinical course of a patient with long-lasting PRCA of 178 days duration after HLA-matched major ABO-incompatible peripheral blood stem cell transplantation (PBSCT). After rapid tapering off of cyclosporin A (CsA), the patient showed signs of chronic GVHD and an increase in the reticulocyte count, eventually resulting in improvement in PRCA.
A 27-year-old man with AML (M2) underwent allogeneic PBSCT from an HLA-matched sister in May 2001. There was major ABO incompatibility between the patient (O, CcDEe) and donor (A, CcDEe). Anti-A isoagglutinin titers of the patient before PBSCT were 1:8 (IgM) and 1:256 (IgG). PBSCT was performed using total body irradiation and cyclophosphamide as conditioning. Peripheral blood stem cells were mobilized from the donor with G-CSF and collected using a COBE Spectra (Gambro BCT, Lakewood, CO, USA). The graft was cryopreserved at −80°C until use. A total of 1.2 × 109/kg mononuclear cells including 5.1 × 106/kg CD34-positive cells was transplanted. The volume of donor RBC contaminating the graft was 9.6 ml.
GVHD prophylaxis consisted of short-term methotrexate and CsA. Granulocytes exceeded 0.5 × 109/l on day 13 after PBSCT, and platelets exceeded 50 × 109/l on day 18. Cytogenetic analysis of bone marrow (BM) on day 29 showed 46, XX but no RBC precursors. On day 49, the white blood cell (WBC) count was 3.9 × 109/l with 58% granulocytes and, the platelet count was 148 × 109/l. Although the anti-A isoagglutinin titers had decreased to 1:2 (IgM) and 1:32 (IgG) on day 49, anemia (hemoglobin (Hb) 7.3 g/dl) persisted. There were no signs of hemolysis. Serum EPO, as measured by radioimmunoassay, was 403 mIU/ml on day 49. These findings led to a diagnosis of PRCA. Absolute reticulocyte counts remained below 0.5 × 104/μl between days 4 and 175.
The patient was given rhEPO (Kirin Brewery, Tokyo, Japan) at a dose of 12000 IU s.c. three times a week between days 50 and 69 to enhance the generation of donor RBC. Because of unacceptable side-effects of general malaise and loss of appetite, however, rhEPO was discontinued without any response. An examination of BM performed on the day following cessation of rhEPO revealed no more than 2.5% of RBC precursors, so that the patient required regular transfusion of host-type RBC between days 54 and 166 (Figure 1). DNA analysis on day 60 for the detection of parvovirus B19 by polymerase chain reaction was negative. Repeated examinations of antigenemia for cytomegalovirus were also negative.
On day 123 the patient had already received 30 units of RBC after PBSCT, his serum ferritin level had increased to more than 2000 ng/ml, and there were no signs of GVHD. We decided to taper the CsA rapidly by 25% every 2 weeks in order to activate the donor lymphocytes and eradicate any remaining host plasma cells that might produce anti-A isoagglutinins. Accordingly, the dose of CsA was reduced from 200 mg/day p.o. (day 123) to 50 mg/day p.o. (day 151). The patient started to have skin eruptions with itching at around day 145, which spread over his trunk and extremities followed by liver dysfunction. Schirmer's test performed on day 161 showed reduction in tear formation, and the result of a skin biopsy was compatible with chronic GVHD. Because the chronic GVHD was deteriorating, we increased the dose of CsA again to 200 mg/day p.o. and started prednisolone (PSL) 1 mg/kg/day p.o. from day 166. This therapy resulted in improvement in the chronic GVHD. Furthermore, the patient no longer needed RBC support from day 167 and the reticulocyte count reached 5.6 × 104/μl on day 179. Type A RBC were detected in the peripheral blood on day 186 and anemia showed signs of improvement. The anti-A isoagglutinin titers decreased to 1:1 (IgM) and 1:4 (IgG) on day 186, and bone marrow aspirate obtained on day 188 contained 26% RBC precursors. His Hb level reached and then remained at more than 12.0 g/dl without additional treatment, and chronic GVHD has remained under control up to the time of writing.
Although ABO incompatibility does not appear to determine the outcome of HSCT in terms of incidence of graft rejection or GVHD, severity of GVHD, or survival rate,2 PRCA is one of the main resultant complications. PRCA occurs more frequently when the recipient's blood type is O and donor's A,6 as was the case with our patient.
Plasmapheresis combined with immunoadsorption,7 rhEPO,8 methylprednisolone (m-PSL),9 antilymphocyte globulin,10 and a combined approach11 have been reported to be effective for the treatment of prolonged PRCA after ABO-incompatible HSCT. However, it is necessary to wait for spontaneous regeneration of donor-type erythropoiesis unless these therapies are effective. First, we used rhEPO for the PRCA, but this failed. Our patient therefore needed alternative treatment to avoid further RBC transfusions because development of hemochromatosis appeared to be imminent.
The mechanism of PRCA after major ABO-incompatible HSCT is unclear, but presence of recipient isoagglutinins against ABH antigens on donor red cells may play an important role in its pathogenesis.4 Our patient needed a great number of transfusions, and the passive transfer of anti-A isoagglutinins by repeated transfusions of host-type RBC may also have contributed to the delayed regeneration of donor-type RBC.
We decided to rapidly taper off the dose of CsA in order to cause a graft-versus-host reaction, which might lower anti-A hemagglutinins by eradicating the remaining host plasma cells. Chronic GVHD developed and RBC of donor type regenerated concomitantly with a decline in anti-A isoagglutinin titers, as expected.
PSL (1 mg/kg/day) orally administered for the treatment of chronic GVHD cannot be completely excluded as an explanation for the observed Hb increase. However, we could not find any publications indicating that a lower dosage of oral PSL (0.5–1 mg/kg/day) is effective for PRCA complicating major ABO-incompatible HSCT. A few case reports9,11 mention that intravenous m-PSL (10–20 mg/kg/day) improved PRCA after major ABO-incompatible HSCT which had been resistant to a low dose of PSL. Yang et al12 recently reported a case of PRCA after major ABO-incompatible HSCT which showed improvement several weeks after the dose of PSL had been increased from 0.5 to 1.5 mg/kg/day p.o. Our patient required RBC transfusions every week from day 54, and did not start PSL until day 166, when he had his last RBC transfusion. The improvement shown thus seems to have occurred too early to be an effect of low-dose PSL therapy.
Although spontaneous regeneration of erythropoiesis is a possibility chronic GVHD itself seemed to have accelerated erythroid recovery.
After chronic GVHD had developed, the patient's anti-A isoagglutinin titers (IgM and IgG) decreased quickly and became undetectable on day 240. Mielcarek et al13 reported that antidonor isohemagglutinin titers disappeared faster in patients with acute GVHD than in those without acute GVHD in terms of HLA-matched, related, and ABO-mismatched HSCT. They suggested a graft-versus-plasma cell effect. It is also possible that chronic GVHD could cause a faster destruction of isoagglutinin-producing plasma cells. It therefore seems worthwhile to try rapid tapering of immune suppressants as treatment of PRCA complicating major ABO-incompatible HSCT. With such an approach, however, chronic GVHD must be very carefully monitored.