Pure red cell aplasia in patients with chronic lymphocytic leukaemia treated with cladribine


Pure red cell aplasia (PRCA) is a rare haematological syndrome with severe anaemia, reticulocytopenia and the absence of red blood cell precursors in the bone marrow, associated with quantitatively and qualitatively normal megakaryocytic and myeloid cell lines (Chikkappa et al, 1986). It occurs in approximately 6% of chronic lymphocytic leukaemia (CLL) patients, in both B-cell type (B-CLL) and T-cell type, most often in the course of disease, but also at presentation. The purine analogues, especially fludarabine (FA) and cladribine (2-chlorodeoxyadenosine, 2-CdA), comprise a novel group of agents with high activity in low-grade lymphoid malignancies, including B-CLL (Robak et al, 2000).

The influence of purine analogues on PRCA in patients with CLL has not been defined. There are a few reports suggesting that the treatment of CLL with FA may induce PRCA in patients with this disease (Leporrier et al, 1993). On the other hand, FA may be helpful in the treatment of PRCA secondary to CLL (Jaccard et al, 1993; Ribeiro et al, 1999; Shimoni et al, 1999). The influence of 2-CdA on the PRCA in CLL patients is even less understood.

We reviewed the records of 470 patients with B-CLL treated at our institutions with 2-CdA from January 1992 to May 2000 and analysed the occurrence of PRCA, as well as the effect of 2-CdA, in patients with co-existence of these diseases. In the analysed group, 139 patients were treated with 2-CdA as a single drug, 262 in combination with prednisone, 30 in combination with cyclophosphamide and 39 in combination with cyclophosphamide and mitoxantrone. The patients received 1–12 (mean 4) courses of treatment. In 236 patients, 2-CdA was the first-line treatment and 234 patients were previously treated with chlorambucil and/or COP (cyclophosphamide, vincristine, prednisone)/CHOP (cyclophosphamide, hydroxydoxorubicin, vincristine, prednisone). Detailed analysis of the treatment results has been presented elsewhere (Robak et al, 2000). PRCA was diagnosed according to the standard criteria (Chikkappa et al, 1986). In this analysis, we have not included patients who, after treatment with 2-CdA, developed bone marrow (BM) aplasia with associated thrombocytopenia and/or granulocytopenia. Platelet count > 100 × 109/l and granulocytes > 2·5 × 109/l in peripheral blood were required. Response criteria for PRCA treatment were those established by Kwong et al (1996). PRCA was diagnosed in eight patients (Table I). In three patients, PRCA was noticed before initiation of treatment, and in five shortly after treatment with 2-CdA alone (two patients), in combination with cyclophosphamide (one patient) or cyclophosphamide and mitoxantrone (two patients). Patients previously treated with 2-CdA showed 0·9% (5/470) rate of PRCA. The frequency of PRCA seemed to be higher in the patients treated with 2-CdA in combination with cyclophosphamide or cyclophosphamide and mitoxantrone. We observed three episodes of PRCA in the group of 69 patients (4·3%) treated with this schedule, and only 2 out of 401 patients (0·5%) treated with 2-CdA alone or in combination with prednisone.

Table I.  Clinical features and response to therapy of eight patients with CLL and PRCA, treated with cladribine before or after diagnosis of PRCA.
  Blood at PRCA
before PRCA

No of
Treatment (no of
courses) of PRCA



before PRCA
(no of
and outcome
  • *

    Haematological data before 2-CdA treatment.

  • Abbreviations: CLL, chronic lymphocytic leukaemia; PRCA, pure red cell aplasia; Hb, haemoglobin; Hct, haematocrit; Ret, reticulocytes; L, lymphocytes; G, granulocytes; PLT, platelets; BM, bone marrow; E, erythroid precursors; ET, erythrocyte transfusion (number of units); HDM, high-dose methylprednisolone; P, prednisone; Chl, chlorambucil; CY, cyclophosphamide; EPO, erythropoietin; COP, CY + vincristine + P; LOP, Chl + sporine + P; CSA, cyclosporine A; 2-CdA, 2 chlorodeoxyadenosine, cladribine; CC, 2-CdA + CY; CMC, 2-CdA + mitoxantrone + CY; DAN, danazol; CR, complete response; PR, partial response; NR, no response.

1F/714·011·40·23·13·113389%1·5%  42-CdA(2)1ET(12), HDM, Chl + 
P-CR (38+)
2-CdA (1)
2ET(15), P + DAN-NR,
EPO-CR (5)
3F/695·716·20·46·92·111567%2%  3CC (1)1ET(7), PR(5 +)
4M/585·215·80·06·82·510695%0%42CMC (1)1ET(16) P-NR, CSA-CR
M/634·815·20·08·41·613385%1·0%  5CMC (2)1ET(10), P-NR, COP(2)
– PR, CY-CR (10+)
M/555·4*17·30·17·42·547092%1·5%313ET(20), HDM, P + 
(20); 2-CdA(2)-NR
7M/626·619·2 34·03·121071%2·0%  01ET-(10), 2-CdA +
P(3)-NR, P-PR (10+)
F/476·318·603·62·514146·5%0·5%48Chl + P (2),
COP (2)
1ET(15), 2-CdA +
P(2)-NR CSA-CR(12+)

In all five patients, PRCA developed after 4–8 weeks after one or two cycles of 2-CdA applied as the first-line treatment. In all five patients, there was only one incident of PRCA. The signs of anaemia passed either spontaneously (case number 3) or after immunosuppressive treatment with cyclosporine A (CSA) or corticosteroids. One patient remitted after the administration of erythropoietin (case number 2).

In three patients, 2-CdA was started after PRCA diagnosis with no improvement of erythroid parameters. In contrast, two of these patients responded to treatment with CSA. These observations are consistent with other publications, indicating that CSA has an important effect in the treatment of PRCA, both primary as well as secondary to other diseases including CLL. Our observations are not similar to some case reports in which the authors indicate that another purine analogue, FA, might be efficient in patients with CLL/PRCA (Jaccard et al, 1993; Ribeiro et al, 1999).