1. ¹Therapeutic Hematology, University Hospital, Basel, Switzerland
2. ²Diagnostic Hematology, University Hospital Basel, Switzerland
3. ³Novartis Pharma Schweiz AG, Berne, Switzerland

Correspondence: D. Heim, Division of Hematology, University Hospital Basel, 4031 Basel, Switzerland. Fax:+41 61 265 4450

Received 28 September 2002; Accepted 27 November 2002.

TO THE EDITOR

Chronic myeloid leukemia (CML) is caused by the bcr/abl tyrosine kinase, the product of the Philadelphia chromosome (Ph). Imatinib mesylate (Glivec^®, Novartis, Switzerland, formerly STI571) selectively inhibits the bcr/abl tyrosine kinase and has shown significant activity in CML patients.^1,2,3

Imatinib is generally well tolerated and nonhematologic NCI grade 3 or 4 toxicities requiring discontinuation of the drug rarely occur. Hematological toxicity however is more frequent. Depending on the disease stage, up to 70% of the patients experience an NCI grade 3 or 4 neutropenia or thrombocytopenia during Imatinib therapy.⁴ The significance of the myelotoxicity was unknown when the first phase II studies were initiated in 1999. According to the phase II study protocols from Novartis, Imatinib had to be withheld for safety reasons in the event of a grade 3 or 4 thrombocytopenia or neutropenia until the absolute neutrophil count (ANC) was raised again to >1.5 10⁹/l or platelets to >50 10⁹/l. This procedure often led to prolonged breaks of the Imatinib therapy. Based on the experience from those trials, the manufacturer now recommends a stepwise dose reduction before discontinuation of the Imatinib therapy. Both procedures have their drawbacks with regard to leukemia treatment. Therapy breaks of 3–4 weeks and often on repetition were the result of these study requirements. Long-term suboptimal drug dosage may be the consequence of such drug reduction scheme. In order to avoid those therapy breaks in neutropenic patients, we investigated the use of recombinant human granulocyte-colony-stimulating factor (Filgrastim, Roche, Switzerland) in six patients with grade 3 or 4 neutropenia, who were treated with Imatinib for late chronic or accelerated phase CML.

Between October 1999 and June 2001, 39 patients with late chronic phase CML and accelerated phase CML were enrolled at our Institution in the Imatinib phase II studies and the expanded access protocols from Novartis. Nine patients (23% of the patients treated at our institution) receiving 400 or 600 mg Imatinib per day developed at least once grade 3 or 4 neutropenia.

We treated three patients with accelerated phase CML who developed a grade 3 or 4 neutropenia according to the study guidelines. Two patients progressed into blast crisis shortly (2 weeks and 1 months, respectively) after interruption of the Imatinib therapy. The third patient had a therapy break of 4 weeks. He was restarted on a reduced Imatinib dose of 400 mg per day. After 2 years on the reduced Imatinib dose he has a complete hematologic response (CHR), but no cytogenetic response. No further grade 3 or 4 hematological toxicity occurred.

Six patients were treated with one to several courses of G-CSF for their neutropenia. Details of the patient characteristics and history are summarized in Table 1. Median age was 42 years (range 20–75 years). Their ANC at study entry ranged from 1.4 to 46.2 10⁹/l (median 3.95 10⁹/l). The three patients enrolled in the accelerated phase study received 600 mg Imatinib per day, the three patients with late chronic phase CML started with 400 mg Imatinib per day. Time to neutropenia ranged from 12 to 41 days after start of the study drug (median 28 days). Five patients had one or more breaks in the Imatinib therapy for neutropenia before G-CSF was used to stimulate myelopoiesis ( Table 1). The median time to recover from neutropenia without G-CSF support and without Imatinib therapy was 28 days (range 28–42 days). With G-CSF support and continued Imatinib therapy, the neutrophil count reached 1.5 10⁹/l after a median duration of 6 days (range 1–7 days). The G-CSF dose used was 0.5 MIU/kg body weight three times per week with the exception of one patient who received 0.5 MIU/kg body weight per day because of profound neutropenia and concurrent sepsis. The G-CSF dose was reduced after resolution of neutropenia to NCI grade 1 level (>1.5 10⁹/l) and maintained at the lowest dose level that kept the neutrophils at or above NCI grade 1 level.

Table 1 - Patient and treatment characteristics.

Full table

Five of six patients are alive after a median follow-up of 15.5 months (range 11–28 months) after start of G-CSF. All five patients are continuously treated with Imatinib. Two patients achieved a major cytogenetic response. Three patients had a CHR as their best response. Three of the patients still receive G-CSF simultaneously with Imatinib.

A 20-year-old male patient ( Table 1, patient no. 1) was enrolled in the STI trial 114 having relapsed after allogeneic hematopoietic stem cell transplantation and progressed to accelerated phase CML. After 4 weeks on Imatinib he achieved a CHR. He was treated with a donor lymphocyte infusion (DLI) from the original matched unrelated donor. He achieved a complete cytogenetic response and a complete hematological chimerism could be demonstrated. Since the bcr/abl transcript was still detectable with the nested PCR method, Imatinib therapy was continued. This patient had three breaks in the Imatinib therapy because of recurring grade 3 and 4 neutropenias. After the third break we started G-CSF until the ANC reached normal values. G-CSF was stopped again and Imatinib was reintroduced at a dose of 300 mg/day. At 3 days after the restart of the Imatinib therapy, the dose had to be reduced further to 200 mg/day for grade 3 neutropenia. We then started him again on G-CSF and increased the Imatinib dose simultaneously to 400 mg/day. This patient is currently on 500 mg Imatinib per day. With 30 MIU of G-CSF every 12 days the ANC stays over 1.5 10⁹/l.

One patient ( Table 1, patient no. 2) showed a minor cytogenetic response (46% Ph-positive metaphases) 3 months after start of the Imatinib therapy and reached a partial cytogenetic response (4% Ph positive) after 6 months. From the 2nd to the 6th month on Imatinib she received G-CSF for grade 4 neutropenia. She was then taken off G-CSF and for the remaining observation period she had ANCs between 0.5 and 1.0 10⁹/l on 400 mg Imatinib.

Two patients (patients nos. 3 and 4) had one break of the Imatinib therapy before the growth factor was given for recurring neutropenia after restart of the study drug. Both patients did not achieve a cytogenetic response. Attempts to stop G-CSF failed because of recurrence of neutropenia.

Patient no. 5 ( Table 1) with accelerated phase CML had a therapy break 5 weeks after the start of the Imatinib therapy at a dose of 600 mg per day. After the toxicity resolved therapy was resumed with 400 mg per day. The ANC again dropped to less than 1.0 10⁹/l and G-CSF was added. The Imatinib/G-CSF combination therapy was carried on for 6 months. We then stopped G-CSF and reduced the Imatinib dose to 300 mg alternating with 400 mg per day. With this regimen the ANC is stable at 1.0 10⁹/l without G-CSF.

Patient no. 6 ( Table 1) with accelerated phase CML had four breaks in the Imatinib therapy, two for grade 4 thrombopenia and two for grade 4 neutropenia. After the fourth break of the Imatinib therapy, the CML evolved into blast crisis. We started Imatinib 600 mg per day and G-CSF simultaneously when the ANC was still in the normal range. This patient died 2 months after the diagnosis of CML blast crisis.

No serious adverse events owing to the use of G-CSF did occur in any of the six treated patients.

These data show that Imatinib induced neutropenia in CML can be overcome by concomitant stimulation of myelopoiesis with G-CSF. This allows for continued Imatinib therapy.

The mechanisms by which CML affects normal hemopoiesis is only partly understood so far. Displacement through the expanding clone, microenvironmental factors or inhibitory cytokines are discussed as factors responsible for the suppression of normal hematopoiesis.⁵ Additional factors to displacement alone are required to explain the fact that none of the six patients regained adequate output of normal neutrophils after tumorload reduction. Even in remission, as evidenced by achieving a major cytogenetic response and/or a CHR, neutropenia was observed.

The early clinical trials with Imatinib showed that myelotoxicity was not only dose dependent, but primarily associated with disease stage. CML patients with advanced disease had significantly more frequently grade 3 or 4 myelotoxicity than patients with chronic phase CML.⁴ The percentage of peripheral blood or marrow blasts has been identified as the only significant variable for the development of grade 3 or 4 neutropenia for patients on Imatinib therapy.⁶ The presence of significant cytopenia after 6 months of Imatinib treatment is a bad prognostic factor. The development of grade 3 or 4 cytopenia in patients with chronic phase CML treated with Imatinib is associated with a significantly higher rate of progression to accelerated phase or blastic phase CML.⁷ We postulate that the treatment interruption at least might have been a contributing factor.

These observations motivate for continued Imatinib therapy especially in those patients at high risk for disease progression. A dose reduction of Imatinib or even a therapy break in these situations should and can be avoided. G-CSF is a safe and effective drug to stimulate myelopoiesis and allows for continued Imatinib therapy in CML patients at risk for disease progression.