Campath 1H is a humanized monoclonal antibody (class G1) directed against CD52, an antigen present on the vast majority of all human lymphocytes but not expressed on neutrophils or platelets. The role of Campath 1H in the treatment of (chronic lymphocytic leukaemia (CLL) has been evaluated in patients who have had multiple previous therapies, and it has also proved to be very active in the first-line treatment of the disease. We describe a patient who achieved a good partial response to Campath 1H but subsequently developed severe refractory thrombocytopenia.
A 49-year-old man was diagnosed as suffering from B-cell CLL in July 1996. The white cell count at presentation was 49 × 109/l, haemoglobin 15.2 g/dl and platelets, 239 × 109/l. In addition he had previously been diagnosed as having rheumatoid arthritis (RA) for which he had received non-steroidal anti-inflammatory agents.
Treatment for CLL was initiated in July 1997 following evidence of disease progression with a rapid doubling time. First-line therapy was chlorambucil 10 mg/d for 28 d. There was little response in either the lymphadenopathy or white cell count and the treatment was changed to single-agent fludarabine, given as a 5 d course of 25 mg/m2/d intravenously, every 28 d. Due to an exacerbation of symptoms related to his RA and in the face of disappointing results from fludarabine therapy, which produced no reduction in his disease, treatment was stopped after two cycles. Disease-modifying agents for the RA were indicated and a regime of weekly oral methotrexate, oral sulphasalazine and pulsed intravenous methylprednisolone was started.
By June 1998 his RA was relatively quiescent but the white cell count had risen to 181 × 109/l with falling haemoglobin and platelet counts. Further treatment for CLL was required. Campath 1H was given on a dose-escalating regime of 3 mg intravenously on day 1, 10 mg on day 2, and 30 mg on day 3. Treatment continued thrice weekly at a dost of 30 mg intravenously, for a total of seven doses. He suffered major morbidity from serum sickness despite premedications with chlorpheniramine and paracetamol, and subsequently the subcutaneous route was used; this was better tolerated and treatment continued thrice weekly for a total of 22 doses. A break from therapy was necessitated by an episode of line-associated sepsis, which required admission for antibiotics and line removal.
Coincident with the end of treatment an abrupt fall in the platelet count was noted. Bone marrow aspirate and trephine biopsies were performed and showed only nodular residual CLL and plentiful megakaryocytes, suggestive of peripheral platelet destruction. Sulphasalazine was discontinued and due to the development of troublesome bleeding, platelet support was given. Screening for IgM and IgG antiplatelet antibodies confirmed the presence of both classes of antibody reactive to GpIIb/IIIa, GpIb, GpIa/IIb, HLA class 1 activity independent of HPA 1, 2, 3 and 5 confirming an autoimmune rather than alloimmune process.
Unfortunately the severe thrombocytopenia persisted despite further platelet transfusions and was unresponsive to intravenous immunoglobulin given daily for 5 d at a dose of 0.4 g/kg/d. The patient developed frank haematuria, and was found to be direct Coombs test positive. He became anaemic secondary to a gastrointestinal bleed and also suffered a retinal haemorrhage. Daily platelets were administered together with red cells as required, but still the platelet count remained <10 × 109/l. It was decided to perform a splenectomy after plasmapheresis and intensive platelet support. The immediate post-operative period was satisfactory although no improvement in the platelet count was seen; an isotope scan failed to reveal active residual splenic tissue. Further treatment consisted of vincristine and cyclophosphamide but these salvage measures were also unsuccessful and intra-abdominal bleeding could not be controlled and the patient died 6 d after removal of the spleen.
CLL has long been known to be associated with autoimmune phenomena. Autoimmune haemolytic anaemia (AIHA) occurs in 10–20% of patients with CLL at some stage in their disease. It is postulated that a disturbance in the function of immunoregulatory T cells may be fundamental to the development of the autoimmune phenomena (Hamblin et al, 1986).
It is now recognized that purine analogues can trigger severe AIHA and that commonly further courses of fludarabine or even alternative chemotherapeutic agents may lead to exacerbations of haemolysis (Orchard et al, 1998). All purine analogues produce a profound and long-lasting depletion in T lymphocytes as does Campath 1H, and this may be the mechanism of the autoimmune phenomenon.
The incidence of clinically significant antiplatelet autoantibodies in CLL is lower than those directed against red cells, but they are still reported as a cause of thrombocytopenia in 2% of patients. Although antiplatelet antibodies are reported in association with RA they seldom cause severe thrombocytopenia, and the production of clinically significant problems by sulphasalazine is a very rare event indeed. Montillo et al (1994) reported the development of autoimmune thrombocytopenia in a patient with CLL during treatment with fludarabine therapy. To date, Campath 1H therapy has not been associated with an increased incidence of either AIHA or ITP. Indeed, there are reports of the successful treatment of autoimmune neutropenia with Campath 1H (Killick et al, 1997).
Corticosteroids represent the conventional first-line therapy for autoimmune thrombocytopenic purpura, but a long-term complete response (platelet count >100 × 109/l) is achieved by only 20% of patients. Splenectomy is advocated as the treatment of choice for non-responders as it cures 60–80% of patients. Refractory patients are a challenge, since all subsequent therapies offer unpredictable response rates and often their side-effects are severe. Our patient failed to respond to all salvage measures as is often the case in severe refractory ITP.