Successful treatment of refractory pure red cell aplasia associated with lymphoproliferative disorders with the anti-CD52 monoclonal antibody alemtuzumab (Campath-1H)


Howard A. Liebman, Division of Hematology, Kenneth Norris Jr Cancer Center, Rm 3466, 1441 Eastlake Ave., Los Angeles, CA 90033, USA. E-mail:


Summary. Acquired pure red cell aplasia (PRCA) is a rare, but significant, complication of lymphoproliferative disorders. It is characterized by anaemia, absence of red cell precursors in the bone marrow and normal granulopoiesis and megakaryopoiesis. We describe two patients with refractory pure red cell aplasia associated with chronic lymphocytic leukaemia (CLL) and a large granular CD8 T-lymphocytic leukaemia (LGL) respectively. Both patients had failed multiple treatment regimens for PRCA and were transfusion dependent. Both patients were subsequently treated with the anti-CD52 humanized monoclonal antibody, alemtuzumab, receiving total doses in excess of 300 mg. Response to treatment, as documented by a rapid increase in the reticulocyte count, occurred as early as the third infusion. At the time of this report, both patients remain in complete remission with normal haemoglobin levels. Alemtuzumab appears to be an effective and well-tolerated therapy for pure red blood cell aplasia associated with lymphoproliferative disorders.

Pure red cell aplasia (PRCA) is a rare cause of severe anaemia (Dessypris, 1991; Charles et al, 1996). The disorder is characterized by an absence of reticulocytes in the peripheral blood with normal platelet and neutrophil counts. Bone marrow examination shows markedly decreased or absent red blood cell precursors with normal myeloid and megakaryocytic maturation (Dessypris, 1991; Charles et al, 1996). PRCA is known to occur in association with a number of autoimmune and lymphoproliferative disorders, including lupus erythematosis (Habib et al, 2002), rheumatoid arthritis (Dessypris et al, 1984), lymphoma (Carloss & Tavassoli, 1979), Hodgkin's disease (Morgan et al, 1978), chronic lymphocytic leukaemia (CLL) (Chikkapa et al, 1986), benign and malignant thymomas (Shigeo et al, 1997; Murakawa et al, 2002) and large granular lymphocytic leukaemia (LGL) (Lacy et al, 1996). In these disorders, PRCA is believed to result from the development of autoreactive cytotoxic lymphocytes that target antigens expressed on early red cell precursors (Dessypris, 1991; Charles et al, 1996). PRCA can also result from a chronic proliferative infection with parvovirus B19 in immunosuppressed individuals (Young, 1995). Recently, PRCA has also been reported in association with the development of antierythropoietin antibodies during treatment with recombinant erythropoietin (Casadevall et al, 2002).

In patients with thymoma-associated PRCA, surgical removal of the thymoma can result in a remission in approximately one-third of patients (Shigeo et al, 1997; Murakawa et al, 2002). In patients refractory to thymectomy or in cases of PRCA not associated with thymoma, treatment has traditionally involved the use of cytotoxic or immunosuppressive medications (Clark et al, 1984; Marmont, 1991; Yamada et al, 1997). Agents such as corticosteroids, cyclophosphamide, antithymocyte globulin and cyclosporine have been used in the treatment of PRCA. Recently, two refractory cases of PRCA have been reported that were successfully treated with the anti-CD20 humanized monoclonal antibody, rituximab (Auner et al, 2002; Ghazal, 2002).

In this report, we describe two patients with refractory PRCA associated with lympoproliferative disorders who were successfully treated with the anti-CD52 humanized monoclonal antibody, alemtuzimab (Campath 1H). One patient had developed PRCA in association with CLL during treatment with fludarabine. The second patient developed PRCA in association with a benign thymoma and a monoclonal CD8 LGL. Both patients were resistant to conventional cytotoxic and immunosuppressive therapy, including a failure in one patient to respond to rituximab.

Case reports

Patient 1

A 68-year-old Hispanic man was diagnosed with CLL in October 2000 in El Salvador. He had received one cycle of chlorambucil before he arrived in the United States. He presented at our institution in April 2001 with haemoglobin 8·2 g/dl, white cell count 8·6 × 109/l with 42% neutrophils, 50% lymphocytes, 6% monocytes, 2% eosinophils and platelets 258 × 109/l. Diagnosis of CLL was reconfirmed based upon morphological findings and flow cytometry showing peripheral blood lymphocytes that were positive for CD19, dim CD20, CD43, dim SIg lambda, CD5 and negative for CD33, CD79, CD10, FMC 7, CD23, Bcl-1 with a CLL score of 4. Bone marrow aspiration and biopsy specimen showed 30–40% cellularity with 64% lymphocytes and 6·5% erythroid precursors. Therapy was changed to fludarabine 25 mg/m2/d given on four consecutive days.

After completion of the first cycle of fludarabine, the patient was followed weekly with repeated blood counts with a gradual decrease in the haemoglobin, despite recovery of his white blood cell count. He was admitted 6 weeks after treatment with complaints of fatigue, weakness, dizziness and dyspnoea on exertion. The patient was febrile, and physical examination was unremarkable except for pallor and tachycardia. There was no hepatosplenomegaly. A complete blood count showed a severe normochromic, normocytic anaemia with haemoglobin of 1·5 g/dl and normal white cell and platelet counts. A reticulocyte count was < 5 × 109/l (< 0·1% corrected). Despite a positive direct antiglobulin test (Coombs' test), there was no evidence of active haemolysis as characterized by a normal serum lactate dehydrogenase (186 U/l; normal range: 89–215U/l), normal total bilirubin (11·97 µmol/l; normal range 0·0–22·23 µmol/l), direct bilirubin (5·13 µmol/l; normal range 0·0–6·84 µmol/l) and a normal haptoglobin (1·92 mg/l; normal range 0·6–2·7 mg/l). Serological and molecular studies for infection with parvovirus B19 and cytomegalovirus were negative. Repeat bone marrow aspiration and biopsy showed an absence of red cell precursors and no residual CLL by morphological examination.

The diagnosis of PRCA was made. He was initially treated with high-dose prednisone 100 mg/d, intravenous immunoglobulin 1 g/kg/d for 3 d and blood transfusions. After 3 weeks of corticosteroid therapy, there was no improvement in the reticulocyte count. Treatment was then changed to cyclosporine 300 mg bid and cyclophosphamide 100 mg/d for 3 weeks without improvement in haemoglobin. He was then treated with rituximab 375 mg/m2, given weekly for eight infusions in combination with cyclosporine and prednisone. Prednisone was gradually tapered over this period because of poor patient tolerance. Cyclosporine was discontinued 1 week after initiation of rituximab because of elevations in hepatic enzymes. There was a gradual improvement in the patient's haemoglobin to 8·8 g/dl without blood transfusion. However, 4 weeks after the discontinuation of this regimen, his haemoglobin decreased again to 5 g/dl with a reticulocyte count of < 5 × 109/l. He again required blood transfusion every 2–3 weeks.

The patient was then started on treatment with intravenous alemtuzumab (Campath-1H). He received an initial dose of 3 mg given on d 1 and 2, 10 mg on d 3 and 30 mg on d 4. Subsequent treatment was given at 30 mg on Monday, Wednesday and Friday weekly for 4 weeks. The patient's total dose of alemtuzumab was 316 mg. The patient's reticulocyte count began to increase after the first week and, on completion of alemtuzumab therapy, his haemoglobin was 10 g/dl. Two weeks after completion of treatment, the patient's haemoglobin was 12 g/dl. Four months after completion of therapy, the patient's haemoglobin was 13 g/dl. The patient remained in remission from PRCA 9 months after completion of therapy, but with residual CLL by bone marrow flow cytometry studies.

Patient 2

The patient, a 52-year-old gentleman, was in good health until he presented to a local emergency room with complaints of severe fatigue and dyspnoea on exertion developing slowly over a 2-month period. At that time, his haemoglobin was 5·8 g/dl, and the packed cell volume was 15%. His white blood cell count was 12 × 109/l with 65% atypical lymphocytes. His reticulocyte count was < 0·1% corrected, and review of his peripheral blood smear showed a normochromic, normocytic anaemia with a prominent population of large granular lymphocytes. An extensive workup included a bone marrow aspiration and biopsy, which showed absent red blood cell precursors and increased numbers of large granular lymphocytes that marked as CD8 thymic lymphocytes. He was treated with oral cyclophosphamide 100 mg and prednisone 60 mg daily, which was discontinued after 12 weeks because the patient remained transfusion dependent without improvement in his PRCA.

The patient self-referred to our institution for a second opinion. A full evaluation was repeated including a repeat bone marrow aspiration and biopsy. The bone marrow showed no evidence of dysplasia in the megakaryocytic and myeloid cell lines with normal chromosome studies, but an absence of red blood cell precursors. These studies confirmed the diagnosis of PRCA with a population of large granular lymphocytes that marked positive by flow cytometry for CD3, CD8, CD57 and negative for CD4, CD19, CD16 and CD56. The lymphocytes were found to be monoclonal by T-lymphocyte receptor gamma-chain gene rearrangement. During that evaluation, the patient was also found to have a partially calcified mediastinal mass on a computerized tomography scan. The mediastinal mass was subsequently removed with the final pathology finding of a benign thymoma. Three weeks after thymectomy, the patient continued to be transfusion dependent with a quantitative reticulocyte count of < 5 × 109/l. He was started on treatment with cyclosporine 300 mg/d and dexamethasone 40 mg/d for 4 d every 2 weeks. After 5 weeks of therapy, the patient continued to require blood transfusions every 2–3 weeks without improvement in his reticulocyte count.

Treatment with alemtuzumab (Campath-1H) was started, and his cyclosporine was decreased to 50 mg/d. A standard dose-escalating regimen was begun, with the initial dose of 3 mg given intravenously (i.v.) on d 1, 10 mg on d 2 and 30 mg on d 3, then 30 mg i.v. twice a week, on an outpatient basis. The patient received a total dose of 133 mg of alemtuzumab. After the third infusion of alemtuzumab, his reticulocyte count increased to 219·7 × 109/l (4·5% corrected) with a marked decrease in the circulating lymphocyte count to < 0·1 × 109/l. However, 5 weeks after completion of alemtuzumab therapy, the peripheral blood showed an increased number of LGL lymphocytes and his reticulocyte count decreased to 19 × 109/l. He was given a second 5-week course of alemtuzumab at a total dose of 330 mg. He again responded with a rapid increase in his reticulocyte count and, at the time of this report, 5 months after completion of alemtuzumab treatment, he has a haemoglobin of 14·4 g/dl and a reticulocyte count of 43·5 × 109/l.


It has now been demonstrated clearly that, in a majority of patients with PRCA, populations of T-gamma lymphocytes can be found in peripheral blood and bone marrow that are capable of inhibiting the growth of erythroid progenitor burst-forming and colony-forming units in vitro (Mangan et al, 1982; Hara et al, 1990). In patients with CLL-related PRCA, increased numbers of bone marrow and peripheral blood T-gamma lymphocytes expressing higher concentrations of membrane IgG-Fc receptors have been observed, compared with patients with equivalent stages of CLL and normal erythropoiesis. In addition, the T-gamma lymphocytes decrease markedly with effective therapy for PRCA (Mangan et al, 1981, 1982; Mangan & D'Alessandro, 1985). PRCA in patients with in thymomas has also been associated with clonally expanded T lymphocytes with characteristic LGL morphology (Handgretinger et al, 1999). The suppression and elimination of the clonal T lymphocytes has resulted in haematological remissions (Handgretinger et al, 1999). A less frequent pathogenic mechanism resulting in PRCA was reported by Casadevall et al (1996), who demonstrated antierythropoietin antibodies in a patient with PRCA. The antibodies inhibited erythropoietin receptor binding, therefore inhibiting differentiation of erythroid progenitors in vitro. This mechanism could account for the reported responses of some patients with PRCA to the anti-CD20 humanized monoclonal antibody, rituximab (Auner et al, 2002; Ghazal, 2002).

Based upon its immunopathogenesis, the therapy of PRCA usually consists of immunosuppressive and/or cytotoxic medications directed primarily to the suppression of cellular immunity. However, despite well-documented responses to immunosuppressive therapy, a significant number of patients do not respond to treatment or relapse after withdrawal of treatment (Clark et al, 1984; Charles et al, 1996). Furthermore, both short- and long-term use of immunosuppressive and cytotoxic agents can be associated with significant morbidity and sometimes mortality. Therefore, more effective, less toxic therapy needs to be developed.

Recently, two reports demonstrated the successful treatment of PRCA with the humanized anti-CD20 monoclonal antibody, rituximab (Auner et al, 2002; Ghazal, 2002). Our first patient, with a diagnosis of CLL and PRCA, achieved a partial response to combination therapy with rituximab, cyclosporine and prednisone, but he relapsed 4 weeks after completing treatment having never obtained a normal haemoglobin level. In addition, our patient, unlike the patient with CLL and PRCA reported by Ghazal (2002), was in a morphological CLL remission before he received rituximab for PRCA.

Alemtuzumab is a humanized IgG monoclonal antibody specific for the CD52 antigen that presents on human lymphocytes and monocytes. Although approved for the treatment of relapsed B-lymphocyte CLL, alemtuzumab appears to be more cytotoxic to thymic (T) lymphocytes, resulting in prolonged and profound depletion of CD4 and CD8 lymphocyte subpopulations (Brett et al, 1996). It has been proposed that depletion of autoimmune T-lymphocyte subpopulations could subsequently modulate the immune system resulting in long-term remissions in autoimmune disorders (Willis et al, 2001). Willis et al (2001) treated 21 patients with autoimmune cytopenias, including four patients with PRCA, with alemtuzumab at a dose of 10 mg/d for 10 d. One of these four patients had a low-grade B non-Hodgkin's lymphoma. Two out of four patients with PRCA responded to treatment, with one patient subsequently relapsing. In this report, we treated two patients with refractory PRCA with alemtuzumab, both of whom responded. The first patient received 30 mg every other day for 4 weeks, total dose 316 mg, but the second patient received only 133 mg of alemtuzumab during his initial course of treatment. After an early relapse, he received a total of 330 mg, resulting in a more durable remission that approached 6 months at the time of writing. Response to treatment was associated with depletion of circulating T and B lymphocytes including the CD8 large granular lymphocytes in the second patient. The only side-effects in our patients were mild infusion reactions. Both patients received prophylaxis with bactrim and acyclovir, and neither had an infectious complication.

Based upon our clinical experience and the previous report of Willis et al (2001), alemtuzumab appears to be a safe and effective agent for the treatment of PRCA, although the durability of these remissions remains to be determined. Whether other concurrent or post-treatment immunosuppressive agents such as cyclosporine or mycophenolate are necessary for a sustained remission needs to be determined in future studies.