A Phase I Trial of Bexarotene, a Retinoid X Receptor Agonist, in non-M3 Acute Myeloid Leukemia: Evidence of Myeloid Differentiation and Clinical Activity.

In vitro, bexarotene has been shown to inhibit the proliferation of non-M3 AML cell lines and induce differentiation of leukemic blasts. This phase I study was designed to evaluate the safety of escalating doses of bexarotene in patients with non-M3 AML. Bexarotene was administered orally daily until disease progression occurred. Prophylactic antihyperlipidemic agents were used in all patients. Six dose levels ranging from 100 to 400mg/m² are planned. Dose escalation occurred in cohorts of 3–6 patients based on dose-limiting toxicity. Twenty patients have been enrolled in 5 dose cohorts (100–300mg/m²) with enrollment demographics: 13M/7F, median age 69 (range 51–82), 11 prior MDS, 8 primary refractory, median number of induction attempts 2, no prior induction chemotherapy 2, prior autologous stem cell transplant 4, 19 blood transfusion dependent, 13 platelet transfusion dependent, and 16 neutropenic. Consistent with reported toxicity, 2 patients developed hypothyroidism, 7 patients developed grade 2–4 hypertriglyceridemia and 1 patient developed grade 2 pancreatitis. Two patients developed a syndrome reminiscent of retinoic acid syndrome, consisting of dyspnea/hypoxia, pleural/pericardial effusions, weight gain/edema and dry cough in the setting of a rapidly rising neutrophil count. This syndrome resolved within 48 hours of stopping bexarotene and initiating steroids. No CR’s were noted, however significant evidence of drug activity were seen. Six patients showed evidence of neutrophil response (pretreatment median ANC 286/μL, range 28–1,037/μL, posttreatment ANC 3,150/μL, range 1,100–27,207/μL). Flow sorted peripheral blood neutrophils were collected from three of these patients and examined by FISH. Between 92–100% of purified neutrophils contained the patient’s leukemic cytogenetic abnormality suggesting differentiation of the leukemic blasts. Bone marrow blasts decreased to <6% in 3 patients. Nine patients with platelet counts <100,000/μL had increases in their platelet counts >30,000/μL (peak range 40–292x10³/μL). Four of these patients with platelet counts <20,000/μL had improvement in their counts to 40–91,000/μL and became transfusion independent. One patient, with sustained clinical response, had 2 independent leukemic clones (7- and 7q- respectively) and no normal metaphases on bone marrow cytogenetic analysis prior to treatment. After 2 months of treatment, cytogenetic analysis showed 93% normal metaphases and no evidence of the original clones. However, 7% of metaphases consisted of 2 new leukemic clones with multiple cytogenetic abnormalities derived from his original clones. This is suggestive of partial eradication of the original leukemic clones and reestablishment of normal hematopoesis but with clonal evolution of the leukemic clones. Daily oral bexarotene is well tolerated at the dose levels studied to date. Early evidence for clinical activity has been seen as exemplified by improvement in platelet counts, increased neutrophil counts, cytogenetic effects, and decreased bone marrow blast counts. We postulate that bexarotene may work by inducing leukemic blast differentiation in non-M3 AML and may represent a novel non-cytotoxic treatment modality. Patient enrollment is ongoing.