Tipifarnib

Complete Remission Following Clofarabine Treatment in Refractory Juvenile Myelomonocytic Leukemia

Summary: Juvenile myelomonocytic leukemia (JMML) is the most common myeloproliferative/myelodysplastic disorder seen in chil- dren. The treatment of choice, allogeneic stem cell transplantation, provides the only known cure for the disease, but relapse after transplant is common. The authors describe a 5-year-old boy diag- nosed at age 34 months with JMML that evolved to acute myeloid leukemia. Initial treatment consisted of fludarabine and cis-retinoic acid therapy, followed by a matched sibling bone marrow transplant. After a relapse, he received a second transplant from the same donor, using peripheral blood stem cells, followed by repeated donor leukocyte infusions. After the second relapse, he received the farnesyltransferase inhibitor R115777 (tipifarnib, Zarnestra), but the leukemia persisted. When bone marrow blasts numbered 60% of the mononuclear cells, he received single-agent clofarabine induction (52 mg/m2/d) for 5 days. After three courses, he attained a remission marrow with 5% blasts and disappearance of the 5q- and 9q- cytogenetic abnormalities.

Key Words: juvenile myelomonocytic leukemia, purine agonist therapy, new drug therapy

Juvenile myelomonocytic leukemia (JMML) is the most common form of pediatric myelodysplastic syndrome of childhood, with more than 50 cases diagnosed each year in the United States. Presentation typically occurs before age 4 years, and boys are more commonly affected. Diagnostic crite- ria include a peripheral blood monocytosis greater than 1,000/mm3, less than 20% blasts and/or promonocytes in the bone marrow or peripheral blood, absence of the Bcr-Abl fusion gene, and two or more of the minor criteria (increased fetal hemoglobin for age, immature granulocytes in peripheral blood, white blood count .10,000/mm3, clonal chromosomal abnormality, and hypersensitivity to GM-CSF by myeloid progenitors).1 Dyspoiesis is often present. To date, three dis- tinct genes have been found to be mutated. These encode the GTPase activating protein neurofibromin, the Ras GTPase, and the tyrosine phosphatase SHP-2.2–4 These mutations occur exclusively.

JMML is difficult to treat, with allogeneic stem cell transplant being the only known curative therapy. In some cases, the disease is rapidly fatal; in others, long-term survival (eg, 4 years) with or without therapy has been observed. About 20% of patients develop frank leukemia. Poorer outcome is associated with platelet count less than 40,000/mm3, fetal hemoglobin less than 10%, and the presence of a cytogenetic abnormality at presentation.

CASE REPORT

A 5.5-year-old boy was initially diagnosed with JMML 33 months before referral. At diagnosis, he presented with fever, white blood count 38,900/mm3, monocyte count 7,391/mm3, platelet count 42,000/mm3, hepatosplenomegaly, and lymphadenopathy. At di- agnosis, his fetal hemoglobin was elevated at 84%, and culture of bone marrow progenitors cells showed hypersensitivity to GM-CSF. Bone marrow blasts were 9%, and cytogenetics was normal. He was treated with fludarabine, cytosine arabinoside, and cis-retinoic acid, according to Children’s Oncology Group AAML-0122. After the second cycle, he underwent a matched allo-sib bone marrow trans- plant with cytosine arabinoside (3 g/m2/dose for six doses), etoposide (500 mg/m2/dose for two doses), cyclophosphamide (45 mg/kg for two doses), and total body irradiation (1,200 cGy with lung shielding after 900 cGy). Graft-versus-host disease (GvHD) prophylaxis consisted of antithymocyte globulin, cyclosporine, and prednisone. He developed grade I acute GvHD of the gut. Nine months after the transplant, bone marrow analysis showed 4% blasts and the appear- ance of a 5q- clone. He received four courses of donor lymphocyte infusions (1 3 107 CD3+ cells/kg, then 2 3 107/kg, 5 3 107/kg, and 1 3 108/kg at 3- to 4-week intervals) without induction of GvHD. Twelve months after his initial bone marrow transplant, his marrow blast count had risen to 27%. Immunophenotype analysis revealed myelomonocytic blasts (CD13+, CD33+, CD117+, and CD64+) with aberrant CD7 expression, consistent with recurrent myelomonocytic leukemia. Variable nucleotide tandem repeat (VNTR) studies showed 50% recipient cells in the bone marrow. He underwent an allo-sib peripheral stem cell transplant with etoposide (250 mg/m2/dose for three doses), fludarabine (30 mg/m2/dose for four doses), melphalan (90 mg/m2/dose for two doses), and localized radiation (500 cGy) to Waldeyer’s ring for bulky lymph nodes. Following engraftment by day +22, the 5q- clone disappeared. The patient received pre-emptive monthly donor lymphocyte infusions for 10 months after trans- plantation, and no overt GvHD developed. Fifteen months after the second transplant, a complete blood count showed 1% peripheral blasts and a 5q- clone in the bone marrow. VNTR studies showed more than 95% donor cells in the marrow. Despite additional donor lymphocyte infusions with IL-2, the blast count continued to rise, and 2 months after the second recurrence, a bone marrow examination showed 42% blasts and chromosomal abnormalities: the 5q- seen at first relapse was again detected, but now these cells had additional chromosomal abnormalities, including 9q-. The patient then began R115777 (tipifarnib, Zarnestra) 300 mg/m2/dose twice daily. Peripheral blasts persisted, and marrow blasts rose to 71% despite a 1-month trial of R115777. Immunophenotyping of the bone marrow showed a pattern similar to the original diagnostic study and consistent with myelomonocytic lineage: CD13+, CD33+, CD117+, CD64+, and HLA-DR positive with a CD7+ subpopulation. The patient was subsequently referred to the University of Texas M.D. Anderson Cancer Center for clofarabine therapy.

A bone marrow examination prior to the start of clofarabine showed 60% blasts, again with myeloid phenotype. Morphology was M0, acute myeloid leukemia, undifferentiated. The patient received three courses of clofarabine 52 mg/m2/d for 5 days. Courses were given approximately 3 weeks apart. A bone marrow aspirate performed 21 days after the first course showed 30% blasts. Twenty-one days after the second course, 7% blasts were noted in the bone marrow, and 21 days after that, there were 5% blasts (Table 1). After the second course, VNTR studies showed 5% recipient cells. Complete cytogenetic remission was obtained after the third course. Clofarabine was well tolerated, with grade II nausea and vomiting. During the infusion, the patient experienced paroxysmal bone pain, which was relieved and prevented by administration of hydrocodone. After the first course, there was a 24-hour hospitalization for fever and neutropenia. After the second course, there was a positive direct antigen for respiratory syncytial virus obtained upon onset of symp- toms of an uncomplicated upper respiratory tract infection. The patient has undergone a third allogeneic stem cell transplant, this time using a 6/6 HLA-matched unrelated donor with peripheral blood stem cell. The preparative regimen consisted of fludarabine (30 mg/m2/d 3 4) and melphalan (90 mg/m2/d 3 2). Despite GvHD prophylaxis consisting of antithymocyte globulin and cyclosporine, he developed grade IV disease (skin, liver, and gut). He developed progressive neurologic dysfunction and died at posttransplantation day 121 with hemorrhagic leukemic infiltrates composed of myelomonocytic cells in the brain. The infiltrate had similar histomorphologic features to the leukemic cells found in the bone marrow diagnostic for JMML. There was no evidence of leukemia in the bone marrow.

DISCUSSION

Clofarabine (Cl-F-ara-A, 2-chloro-2#-fluorodeoxy-9-b- D-arabinofuranosyladenine) is a second-generation purine analog that has shown efficacy in both adult and pediatric phase 1 trials for refractory and relapsed acute lymphoblastic and acute myeloid leukemia.5,6 In a recently reported phase 1 trial in children, five patients achieved complete remission and three achieved partial remission, for an overall response rate of 32%.6 A multisite phase 2 trial for refractory and relapsed pediatric acute lymphoblastic and acute myeloid leukemias is ongoing. An overall response rate of 34% has been found in heavily pretreated children and adolescents with multiply relapsed or refractory acute myeloid leukemia.7 Seven of these responders (n = 29) received a subsequent allogeneic bone marrow transplant; three remain in remission. This is the first patient with JMML that evolved to acute myeloid leukemia treated with clofarabine.

As a next-generation purine agonist, clofarabine may be considered a hybrid of nucleoside and deoxyadenosine analogs. Because it retains the 2-halogenated aglycone of fludarabine and cladribine, it is resistant to deamination by adenosine deaminase. For cytotoxic activity, fludarabine must undergo phosphorylation by deoxycitidine kinase. Clofarabine inhibits DNA polymerases and ribonucleotide reductase. When combined with cytosine arabinoside, it potentiates in- hibition of ribonucleotide reductase. In addition, clofarabine can cause disruption of mitochondrial function with the release of cytochrome c, binding of the apoptotic protease activating factor-1 (APAF-1), and activation of the caspase-9 and -3 apoptotic pathway.
Chemotherapy is often effective in controlling JMML disease activity and is used prior to an allo-sib transplant. Approximately 30% of patients remain alive at 6 years.8 When disease recurs after transplantation, survival is poor. Immu- notherapy in the form of interferon-alpha and/or donor lymphocyte infusions may induce disease stability.9 This case report suggests that clofarabine may have activity against refractory myeloid leukemia associated with JMML and can be well tolerated. Previous treatment with fludarabine does not preclude a response to clofarabine, as confirmed in this patient. He had received fludarabine both as an induction and as part of a myeloablative regimen for the first bone marrow transplant. Although chemically related to fludarabine, clofarabine has a unique property, as described above, of affecting mitochon- drial function.10 In this patient, an isolated central nervous system (CNS) leukemic relapse occurred 121 days after transplantation. Like fludarabine and cladribine, levels of clofarabine do not persist in the CNS.11 An isolated CNS relapse of JMML has been reported once before.12

The cytogenetic abnormalities might suggest a second- ary leukemia following etoposide used in the first transplant. However, the latency period and the absence of 11q23 would be atypical.13,14 A short latency suggests that disease arose prior to high-dose chemotherapy and transplantation.15 At diagnosis, no chromosomal abnormalities were detected, but the leukemic cells detected after transplantation showed 5q-. When leukemia was detected after the second transplant, the detection again of 5q- provides strong evidence that this is recurrent disease. Some additional cytogenetic abnormalities were then detected, but only in cells with 5q-, implying clonal evolution with disease progression and relapse rather than a new leukemic population. However, the possibility that the 5q- clone represents a new, distinct neoplasm cannot be entirely excluded.

In this child who had refractory leukemia after two allogeneic bone marrow transplants, multiple donor leukocyte infusions, intensive cytotoxic chemotherapy, and farnesyl- transferase inhibitor therapy, three courses of clofarabine were well tolerated and induced a hematologic and cytogenetic re- mission. Further evaluation of clofarabine in JMML with or without leukemic transformation is warranted, although addi- tional therapy directed toward the CNS would be prudent.