Cyclosporin A for the treatment of patients with chronic idiopathic thrombocytopenic purpura refractory to corticosteroids or splenectomy


M. C. Kappers-Klunne, MD, Department of Haematology, University Hospital Rotterdam, PO Box 2040, 3000 CA Rotterdam, The Netherlands. E-mail:


Patients with chronic immune thrombocytopenic purpura (ITP) who are unresponsive to corticosteroids require splenectomy, but if this fails, treatment is difficult. We tried to induce durable remissions in ITP patients refractory to corticosteroids before or after splenectomy by applying strong immunosuppression with the combination of cyclosporin A (CyA 5 mg/kg/d) and prednisone (0·4 mg/kg/d). Patients were assigned to one of two groups. Group 1, 10 patients refractory to prednisone; and group 2, 10 patients refractory to at least prednisone and splenectomy. Overall response rate was 55% (50% in group 1 and 60% in group 2 patients). Nine of the 10 patients in group 1 finally had a splenectomy because of relapse, insufficient response or toxicity of CyA. Thirty percent of the patients discontinued CyA because of side-effects; hypertension, severe headache and muscle pain being the most frequent encountered. It is concluded that CyA treatment does not avoid, but only postpones, splenectomy in chronic ITP patients who are refractory to corticosteroids. However, CyA can be useful in a subgroup of patients with corticosteroid- and splenectomy-refractory ITP, but treatment toxicity is high.

Chronic idiopathic thrombocytopenic purpura (ITP) is an autoimmune disorder characterized by platelet destruction caused by an antiplatelet autoantibody resulting in platelet phagocytosis in the mononuclear phagocytic system (Karpatkin, 1997). It is classically defined by a decreased number of circulating platelets and a normal to supernormal bone marrow megakaryocyte mass, occurring in the absence of agents or systemic diseases known to induce thrombocytopenia.

Glucocorticosteroids have been accepted as the initial treatment for ITP, although only a third of the patients will have a long-term response. Splenectomy is the treatment of choice for patients failing on corticosteroids. This results in a much higher cure rate than any medical regimen, with sustained complete and partial response rates of 60% and 12% respectively (Berchtold & McMillan, 1989). However, in older patients response rates may be much lower (Guthrie et al, 1988; Cortelazzo et al, 1991).

Treatment is regarded unnecessary in patients with a platelet count of > 40 × 109/l. Severe bleeding episodes have only been recorded at platelet counts less than 30 × 109/l with an increased incidence in patients aged > 60 years (Cortelazzo et al, 1991). Thus, in patients failing to respond to corticosteroids and splenectomy with a platelet count of < 30 × 109/l, further treatment should be considered. Therapeutic options for these patients include a wide range of treatment regimens such as immunosuppression with high-dose methylprednisolone, cyclophosphamide, azathioprine or vinca-alkaloids; blockade of the mononuclear–phagocytic system with high dose gamma-globulin or IgG anti-Rh (D); and miscellaneous regimens such as danazol, ascorbic acid, staphylococcal protein-A plasmapheresis, interferon-α (McMillan, 1997) and, very recently, monoclonal antibodies (Bussel et al, 1999; Periotta et al, 1999) and even autologous stem cell transplantation (Lim et al, 1997). Treatment results generally have been disappointing.

In ITP the precise mechanism of auto-antibody formation is unknown but, more recently, T-cell dysfunction has been suggested as the most important causative factor rather than abnormal B-cell reactivity as was previously assumed (Mylvanganam et al, 1988; Ware & Howard, 1993).

Cyclosporin A (CyA) is a powerful suppressant of T-cell function (Kahan, 1989; Denton et al, 1999). CyA selectively inhibits both antigen-induced activation of CD4+ lymphocytes and the production by these cells of interleukin 2 (IL-2) and other cytokines. This action results in an indirect inhibitory effect on the growth and differentiation of B lymphocytes.

In view of its action it may be expected that CyA could induce and, especially, maintain remission in autoimmune disorders, particularly in those with mechanisms mediated by T cells. The drug has proven effectiveness in psoriasis and other autoimmune diseases (Ellis et al, 1991; Wells & Tugwell, 1993).

Corticosteroids have both immunosuppressive and non-specific anti-inflammatory properties. They inhibit IL-1 generation, thereby potentiating the action of cyclosporin on helper T cells (Kahan, 1989). The strong immunosuppressive effect of the combination of CyA and corticosteroids is potentially very effective in arresting the immune response. Therefore, we initiated a prospective study aimed at attaining a durable remission with this regimen in patients refractory to corticosteroids alone or to corticosteroids and splenectomy.

Patients and methods

Patients Patients who presented to the University Hospital Rotterdam and some affiliated hospitals were included in the study if they fulfilled the following criteria: chronic ITP defined as isolated thrombocytopenia, diffuse intravascular coagulation and pseudothrombocytopenia being excluded; absence of splenomegaly as documented by ultrasound sonography and normal bone marrow morphology with normal or increased numbers of megakaryocytes; refractoriness to conventional-dose prednisone as initial treatment or to both prednisone and splenectomy; serum creatinine < 120 μmol/l; informed consent. The protocol was approved by the medical ethical committee of our hospital.

Treatment Patients were assigned to one of two groups. Group 1 comprised chronic ITP patients or patients presenting with a relapse after a remission duration of > 6 months, who were refractory to prednisone (1·5 mg/kg/d) for 3 weeks. At this point, prednisone was tapered to zero within 2–4 weeks and oral CyA was started 6 mg/kg/d orally (in two doses) and continued for at least 4 weeks. If complete response (CR) had been attained, CyA was tapered by 50 mg/d every 2 weeks. If by 4 weeks the patient was in partial response (PR), CyA was continued until maximum platelet counts were reached, but no longer than 3 months. At the point of maximum platelet count, CyA was tapered in the same way as for patients who had reached CR. If after 4 weeks of CyA, the platelet count was or had returned to < 40 × 109/l, the patient was regarded as no response (NR) and CyA was stopped and the patient proceeded to splenectomy.

Group 2 patients were refractory to at least prednisone and splenectomy. Patients started on CyA 5 mg/kg/d (in two doses) in combination with prednisone 0·4 mg/kg/d. In the case of CR, CyA was tapered in the same way as in group 1 patients. If by 12 weeks PR was reached, CyA and prednisone were continued until maximum platelet count, but not longer than 4·5 months. At that point CyA was stopped and prednisone was gradually tapered and stopped. In the case of NR at 12 weeks, both CyA and prednisone were stopped. In all patients CyA treatment was monitored by weekly clinical and laboratory examination during the first 4 weeks, after which the interval between monitoring visits was extended, dependent on the patients tolerance of treatment.

Response criteria Complete response (CR), platelet count > 110 × 109/l for > 12 weeks. Partial response (PR), platelet count > 40 × 109/l for > 8 weeks. No response (NR), platelet count always < 40 × 109/l. Transient response (TR), platelet count > 40 × 109/l for > 2 but less than 4 weeks.

Toxicity Assessment of toxicity included monitoring of clinical features, such as blood pressure, paraesthesias, gastrointestinal complaints, gingival hypertrophy, headache and determination of whole blood or plasma CyA levels and serum creatinine. The intended therapeutic range is identical to target levels aimed at in the first 3 months after kidney transplantation. Initially, CyA concentration was determined on plasma samples using a monoclonal antibody radioimmunoassay. Later, CyA levels were determined in whole blood with the EMIT 2000-CsA immunoassay (Dade Behring, San Jose, Ca, USA). In one patient a polyclonal antibody assay was used (Abbott TDX, Hoofddorp, The Netherlands) to measure blood CyA concentration.


Treatment results

Twenty patients were included in the study, 10 patients in group 1 and 10 patients in group 2.

In group 1 (Table I) CR was reached in three cases. In patient 1, CR lasted for 2 years off treatment; a relapse was triggered by infection. In patient 3, CR lasted for 2 years off treatment until a spontaneous relapse occurred. Patient 7 reached CR but, after stopping CyA, remained dependent on low-dose prednisone (5 mg/d) to maintain safe platelet levels. Two patients in group 1 (patients 2 and 4) reached PR lasting 324 and 95 d, respectively, at which time CyA had to be stopped because of side-effects and the patients proceeded to splenectomy. Two patients (patients 5 and 8) attained TR and three cases (patients 6, 9 and 10) qualified as NR. In patient 6, CyA was started when his platelet count was 140 × 109/l because of preceding high-dose prednisone. After starting CyA, his platelet count further increased to 166 × 10 9/l, but quickly decreased to < 20 × 109/l thereafter on tapering prednisone, so he was regarded a NR. Nine patients ultimately underwent splenectomy because of relapse or insufficient response. Splenectomy resulted in durable remission in all patients in group 1 for a mean follow-up of > 2 years.

Table 1.  (A and B) Cyclosporin A (CyA) in ITP patients refractory to corticosteroids (group 1).

(A) : (B)

Age (years)
Days from
Period of
CyA (d)
CyA dose

135/M1801366CRPred 1 mg/kg/d
269/M213246–7·5PRPred 0·8 mg/kg/d
364/F1380832·5–5CRPred 0·3 mg/kg/d
418/F64956–8PRPred 1 mg/kg/d
630/M2431376NRPred 0·5 mg/kg/d
784/F211016CRPred 1 mg/kg/d
852/M39215TRPred 0·4 mg/kg/d
938/F23895NRPred 0·15 mg/kg/d
 Platelets (× 109/l)Days from start CyA to  
  MaximumPlateletsMaximumCyA levels* 
PatientStartcount> 100 × 109/lcountMeanMedianToxicity
  • CR, complete remission; NR, no response; PR, partial remission; TR, transient response; Pred, prednisone.

  • *

    Target levels, 150–300 ng/ml.

  •   †40–80 ng/ml.

  •   n.a, not applicable.

231072342347564Fatigue, nausea, gum hyperplasia → stop CyA
43673n.a.52165140Fatigue, muscle ache, burning → stop CyA
53545n.a.4274240Muscle ache, mild/transient serum creatinine
73018721101310320Hypertension, peripheral oedema → stop CyA
9436n.a.4100n.a.Head/muscle ache, hypertension → stop CyA
10n.a.8n.a.n.a.189210Gum hyperplasia

Two patients in group 2 (Table II) were in continuing CR lasting for > 4 (patient 11) and > 2 years (patient 12), respectively, and a third (patient 17) who initially reached CR converted to PR lasting for > 1 year off treatment. Three patients (patients 15, 16 and 18) reached PR. In two cases, CyA was stopped (after 91 and 31 d of CyA respectively) because of side-effects. Four cases (patients 13, 14, 19 and 20) in this group were NR.

Table II.  (A and B) Cyclosporin A (CyA) in ITP patients refractory to at least prednisone and splenectomy (group 2).



Age (years)

Time from
Period of


1160/FP, IVIG, S0·32775CRP* 0·4 mg/kg/d
1247/FP, S1·61967CRP 0·4 mg/kg/d
1329/MP, IVIG, S0·07605–7·5NRP 0·6 mg/kg/d
1436/FP,S, IVIG, vitC,
danazol, im
1551/MP, IVIG, S1·91537PRP 0·2 mg/kg/d
1650/FP, S12915PRNone
1755/FP, HD, S0·82275CRP 0·4 mg/kg/d
1834/FP, S, HD6·41245PRNone (HD immediately
preceding CyA
1966/MP, IVIG, S,
im, dap, vinc
0·7744–5NRP 0·3 mg/kg/d,
Intermittent IVIG
2049/MP, S, im, IVIG0·5563NRNone
 Platelets (× 109/l)Days from start CyA to  
  MaximumPlateletsMaximumCyA levels* 
PatientStartcount> 100 × 109/lcountMeanMedianToxicity
  • *

    P, prednisone.

  • IVIG, intravenous immunoglobulin; HD, high-dose dexamethasone; S, splenectomy; im, immuran; vinc, vincristine; dap, dapsone; vitC, vitamin C.

  • *

    Target levels: 150–300 ng/ml.

  •   †40–80 ng/ml.

  •   ‡300–800 ng/ml.

  •   n.a., not applicable.

12242548281619Severe muscle ache, paraesthesias
133n.a.n.a.n.a.6468none headache → stop CyA
162060n.a61233215head/muscle ache, ↑ serum creatinine → stop CyA
1919n.a.n.a.n.a.222225severe muscle ache, paraesthesias ache


Mean plasma/blood CyA levels always remained within the therapeutic range except in four patients (patients 7, 9, 12 and 14). Remarkably, patient 12 reached long-term CR despite suboptimal plasma CyA levels. Patient 15 attained long-term PR on mean blood CyA levels above target values. Despite optimal or sometimes even suboptimal CyA levels, the majority of the patients experienced side-effects. Hypertension, occurring in six cases, and severe muscle pain and/or headache in seven cases were the most frequently encountered side effects. Discontinuation of CyA was deemed necessary in six patients (patients 2, 4, 7, 9, 14 and 16). Reasons for stopping are shown in Tables IB and IIB.


In this prospective study we examined the feasibility of intensive immunosuppression using the combination of CyA and corticosteroids to induce remission in ITP patients refractory to corticosteroids before or after splenectomy. Based on the assumption of a T cell-mediated mechanism for aberrant B-cell activity in ITP (Mylvanganam et al, 1988), combining CyA with corticosteroids may result in more profound immunosuppression.

Reports on therapeutic measures after failure of initial corticosteroid treatment to prevent splenectomy in chronic ITP (group 1 patients) are scarce and concern only small patient numbers or incidental cases. Schultz et al (1995) described the results of CyA treatment in five children with ITP who were refractory to intravenous immunoglobulin and prednisone, with two of them also unreactive to anti-Rhesus D immune globulin. The CyA dose in this study was 5 mg/kg/d, which was increased to 10 mg/kg/d in cases of NR after 2 weeks. Responses were minimal and transient, lasting no longer than 4 weeks at the cost of intolerable toxicity, which probably, in part, has to be ascribed to the relatively high CyA levels required to obtain an increase in platelet count in these children. The toxicity observed in this study is in accordance with our own experience but the response seemed to be better in our patients.

In our study, 3 of the 10 patients in group 1 reached CR, all lasting > 1 year off treatment. The two partial responders in this group had to discontinue CyA because of side-effects. Ultimately, all but one patient in this group had a splenectomy because of relapse. Thus, CyA merely postponed splenectomy but could not prevent it.

CyA has also proven to be effective in refractory ITP (group 2 patients). The first reports concern only anecdotal cases and doses used varied widely, as did treatment results (Kelsey et al, 1985; Miescher & Beris, 1985; Velu et al, 1987; Matsumara et al, 1988). Siegel et al (1991) observed seven responses in nine patients with platelets under 50 × 109/l. Emilia et al (1996) recently reported four cases with CR and two cases with PR on CyA in five refractory patients with ITP and one patient with Evans syndrome. Most patients remained drug-dependent but, in contrast to our own observations, side-effects were moderate and transient. The combined CR and PR rate in our group 1 and 2 patients was 55%. The response rate in splenectomy-refractory patients (group 2) was 60%. The majority experienced major toxicity with hypertension, severe muscle pain and headache being the main reasons for discontinuing CyA. Overall 30% of the patients discontinued CyA because of side-effects. Only four cases (two in each group) were free of notable side-effects.

We conclude that subsequent use of CyA after failure of initial corticosteroid therapy cannot prevent, but only postpone, splenectomy in chronic ITP. However, the combination of CyA and low-dose corticosteroids can be useful in a small percentage of patients with refractory ITP, but treatment toxicity is high and the success rate appears not to be greater than any other treatment modality reported to date.


We thank the physicians who sent us clinical and laboratory data of their patients: Dr C. van der Heul, St. Elisabeth Ziekenhuis, Tilburg, Dr W. L. Blok and H. Kettner, MD, Ziekenhuis Walcheren, Vlissingen.