Purpose: Bexarotene is a retinoic X receptor agonist that has been shown in vitro to inhibit growth and induce differentiation of myeloid leukemic cell lines. We therefore conducted a phase I dose escalation study to assess the maximum tolerated dose, toxicities, and activity of bexarotene in patients with non-M3 acute myeloid leukemia (AML).

Experimental Design: We enrolled patients with active non-M3 AML who had either relapsed or refractory disease or were not eligible for standard cytotoxic chemotherapy. Cohorts of three to six patients received escalating doses of daily oral bexarotene ranging from 100 to 400 mg/m2 until evidence of disease progression or unacceptable adverse events occurred.

Results: Twenty-seven patients, with median age of 69 years (range, 51-82 years), were treated. Twenty-four (89%) patients had undergone prior chemotherapy. At the highest dose level tested (400 mg/m2), three of six patients had to reduce their dose of bexarotene due to grade 3 adverse events. The maximum tolerable dose of bexarotene was determined to be 300 mg/m2. Clinical activity was manifested by 4 (15%) patients with reduction in bone marrow blasts to ≤5%, 11 (41%) patients with improved platelet counts, and 7 (26%) patients with improved neutrophil counts. Three patients with relapsed AML survived >1 year while taking bexarotene. Leukemic blast differentiation was suggested by the presence of the leukemic cytogenetic abnormality in mature circulating granulocytes and the occurrence of differentiation syndrome.

Conclusions: The recommended dose of bexarotene for future studies is 300 mg/m2/d. Bexarotene is well tolerated in patients with non-M3 AML and has evidence of antileukemic activity.

Retinoid receptors control cellular maturation and proliferation, providing a rationale for exploring the activity of retinoid agonists as cancer therapeutic agents. All-trans retinoic-acid (ATRA) is a well-studied and effective treatment for acute promyelocytic leukemia (APL; refs. 13). APL is characterized by the presence of the t(15;17), which fuses the retinoic acid receptor (RAR) α gene to part of the promyelocytic leukemia gene, creating an aberrant retinoic acid receptor fusion protein that leads to dysregulated myeloid differentiation (4). ATRA is able to overcome this maturation arrest and reestablish myeloid differentiation (4, 5). ATRA has become part of standard first-line therapy for APL and has lead to significant improvements in survival (6). Unfortunately, APL is an uncommon subtype of acute myeloid leukemia (AML) and ATRA therapy has no clinical effect on the more common non-M3 AML (7).

Bexarotene is another retinoid agonist that is commonly used in the treatment of T-cell lymphomas (8, 9). In contrast to ATRA, which has strong affinity for RAR receptors and only weak binding to retinoic X receptor (RXR) receptors, bexarotene is biased toward binding to RXRs (10, 11). In vitro, bexarotene inhibits growth and induces differentiation of HL-60 leukemia cell lines as well as leukemic cells from patients with AML (12). Based on this preclinical rationale, we initiated a phase I dose escalation study to establish a safe treatment dose and to explore the toxicity and preliminary efficacy of bexarotene in non-M3 AML.

Patients. All patients were enrolled and treated at the Abramson Cancer Center at the University of Pennsylvania Medical Center as part of an institutional review board–approved clinical trial. Written informed consent was obtained from all patients. Eligible patients were those with active biopsy-proven non-M3 AML, age ≥18 years, Eastern Cooperative Oncology Group performance status of 0 to 2, prior treatment with chemotherapy or not eligible for conventional chemotherapy, and recovered from any toxicities of prior therapy. Major exclusion criteria included history of prior pancreatitis, active alcohol abuse, WBC count >10,000/μL, serum triglycerides levels >1,000 mL/dL, other concurrent AML therapies, history of myeloablative allogeneic stem cell transplant, uncontrolled active infection, known active central nervous system involvement with AML, use of blood growth factors within the past week, or significant organ dysfunction (total bilirubin, transaminase greater than three times the upper limit of normal or creatinine >4 mg/dL).

Study design. This dose escalation trial followed the standard three-plus-three cohort escalation design. Daily dose levels evaluated ranged from 100 to 300 mg/m2 in 50 mg/m2 increments and then 400 mg/m2. Dose-limiting toxicity (DLT) was defined as grade ≥4 (National Cancer Institute Cancer Therapy Evaluation Program 3.0) toxicity occurring within 28 days of drug initiation. Patients with DLT that had resolved or had intolerable grade <4 adverse events would be allowed to reinitiate bexarotene at the next lower dose level. Dose escalations were not permitted for individual patients. New patient cohorts were opened after 28 days of follow-up had passed for all patients in the previous dose cohort and if no DLTs had occurred in the initial three patients or only one DLT had occurred in six patients in the prior dose cohort. The maximum tolerated dose was defined as the dose level below, which more than one of three or two of six patients experienced DLT.

Treatment plan. Bexarotene capsules were administered orally daily with food. Patients continued treatment as long as they were on study. Patients were also concurrently initiated on either 48 mg/d fenofibrate or 10 mg/d atorvastatin as prophylactic treatment for hypertriglyceridemia. Patients were not allowed to take other concurrent antileukemic therapies or erythoid or myeloid growth factors while on study. Exceptions to this rule were made for hydroxyurea for the first 21 days of study enrollment for initial control of peripheral blood leukemic blasts in patients with WBC >10,000/μL and for myeloid growth factors for up to 3 weeks in the case of infection. Patients with DLT events or who were unable to tolerate bexarotene for grade 3 event would withhold drug for up to 3 weeks to resolve the adverse event. They were then eligible to reinitiate bexarotene at the previously tested next lower dose level. Patients continued on treatment until they experienced progressive disease as manifested by WBC >20,000/μL with >75% blasts or unacceptable toxicity.

Patient monitoring. Patients were evaluated at a minimum of every 2 weeks with full history, physical evaluation, and laboratory studies. Triglycerides and thyroid function studies were done at 2 weeks and then monthly thereafter. Bone marrow biopsy was done at 2 weeks, 2 months, and then every 3 months thereafter.

Response assessment. Patient response to therapy was based on changes in their peripheral blood counts and bone marrow blast percentage. Bone marrow response was defined as a decrease in bone marrow core blast percentage of >50% or an absolute blast percentage of ≤5%. Patients with neutrophil count improvement had to have had a pre-bexarotene absolute neutrophil count (ANC) <1,500/mm3 with at least a 100% improvement or absolute increase of >500/mm3, whichever was greater, after initiation of therapy. Platelet response was defined as achievement of transfusion independence (baseline platelet count of <20,000/mm3) or an absolute increase of 50,000/mm3. The patient bone marrow and hematologic response data were also evaluated using the modified International Working Group response criteria for myelodysplasia in an effort to provide a standardized method for response (13).

Flow separation and fluorescence in situ hybridization analysis. Peripheral blood granulocytes were purified by cell sorting for CD16 and CD45 as described previously (14). Granulocytic purity was confirmed by cytospin and microscopic examination. Fluorescence in situ hybridization analysis was done on the purified granulocytes using probes for the given patient's known leukemic cytogenetic abnormality.

Statistical analysis. All survival distributions and rates were calculated according to the Kaplan-Meier method. Response variables were examined using the χ2 test. All P values reported are two tailed. Statistical analysis was done with the Stata software (Stata version 9, 2005).

Patient characteristics. As shown in Table 1, 27 adult patients with non-M3 AML were treated on study. This patient population had a median age of 69 years with poor prognostic factors (48% with prior myelodysplastic syndrome, 44% with primary refractory, and 26% with complete response lasting <12 months). Eighty-nine percent had had prior induction chemotherapy with 59% having had ≥2 induction attempts. Nineteen percent had undergone prior autologous stem cell transplantation. At the time of enrollment, the majority had significant cytopenias and five patients had WBC >10,000/μL and were on hydroxyurea.

Table 1.

Patient characteristics

CharacteristicsNo. patients (N = 27)Percentage of patients
Sex   
    Male 19 70 
    Female 30 
Age (y)   
    Median 69  
    Range 51-82  
Prognostic factors   
    Prior myelodysplastic syndrome 13 48 
    Unfavorable cytogenetics 14 52 
Prior treatment   
    Primary refractory 12 44 
    Complete response <12 mo 26 
    Chemotherapy naive 11 
    ≥2 induction attempts 16 59 
    Prior autologous bone marrow transplant 19 
French-American-British AML subtype   
    M0 15 
    M1 22 
    M2 30 
    M4 22 
    M5 
    NOS 
Initial blood counts   
    ANC <1,000/mm3 20 74 
    Packed RBC transfusion dependent 26 96 
    Platelet transfusion dependent 18 67 
On hydroxyurea 19 
CharacteristicsNo. patients (N = 27)Percentage of patients
Sex   
    Male 19 70 
    Female 30 
Age (y)   
    Median 69  
    Range 51-82  
Prognostic factors   
    Prior myelodysplastic syndrome 13 48 
    Unfavorable cytogenetics 14 52 
Prior treatment   
    Primary refractory 12 44 
    Complete response <12 mo 26 
    Chemotherapy naive 11 
    ≥2 induction attempts 16 59 
    Prior autologous bone marrow transplant 19 
French-American-British AML subtype   
    M0 15 
    M1 22 
    M2 30 
    M4 22 
    M5 
    NOS 
Initial blood counts   
    ANC <1,000/mm3 20 74 
    Packed RBC transfusion dependent 26 96 
    Platelet transfusion dependent 18 67 
On hydroxyurea 19 

NOTE: Unfavorable cytogenetics are defined as >2 cytogenetic abnormalities, inv(3), t(3,3), t(6,9), and -7.

Dose escalation and maximum tolerated dose determination. A total of 27 patients were treated on six dose levels between 100 and 400 mg/m2 (Table 2). The median number of days on therapy was 39 with a range of 4 to >527 days. Bexarotene was well tolerated overall with only one DLT consisting of altered mental status being recorded (Table 3). In retrospect, it was likely that the single reported DLT was due to concurrent sepsis rather than study drug. At the highest tested dose level (400 mg/m2), three of six patients had their dose of bexarotene reduced due to grade 3 rash, anorexia, or fatigue that was felt by the patients to be intolerable. Although an official maximum tolerated dose as determined by the occurrence of DLTs was not identified, a dose of 300 mg/m2 was established as the suggested patient tolerable dose. Due to the patients' preexisting active AML, data pertaining to leukemia-related side effects, including neutropenia, leukopenia, anemia, thrombocytopenia, petechiae, and weight loss, were not used in the assessment of DLT. Overall bexarotene was well tolerated with dose-dependent and reversible toxicities.

Table 2.

Dose escalation scheme

Treatment dose level (mg/m2)No. patients treatedNo. days on study per patientRespective reason for discontinuing treatment
100 17, 355, 52 PD, infection, PD 
150 14, 64, 14, 112, 396, 33 DLT, choice, central nervous system AML, DS, PD, DS 
200 30, 4, 50 Death, sepsis, PD 
250 29, 25, 158 Choice, PD, PD 
300 10, 4, 517+, 17, 49, 53 PD, SAE, AL, death, PD, PD, choice 
400 137, 39, 10, 18, 78, 188 PD, choice, PD, choice, SAE, death 
Treatment dose level (mg/m2)No. patients treatedNo. days on study per patientRespective reason for discontinuing treatment
100 17, 355, 52 PD, infection, PD 
150 14, 64, 14, 112, 396, 33 DLT, choice, central nervous system AML, DS, PD, DS 
200 30, 4, 50 Death, sepsis, PD 
250 29, 25, 158 Choice, PD, PD 
300 10, 4, 517+, 17, 49, 53 PD, SAE, AL, death, PD, PD, choice 
400 137, 39, 10, 18, 78, 188 PD, choice, PD, choice, SAE, death 

Abbreviations: PD, progressive disease; DS, differentiation syndrome; AL, still alive and on study; choice, patient choice; SAE, serious adverse event.

Table 3.

Summary of grade ≥3 treatment-related nonhematologic adverse events

Toxic eventNo. patients experiencing toxicity dose cohort and toxicity grade
100 mg/m2
150 mg/m2
200 mg/m2
250 mg/m2
300 mg/m2
400 mg/m2
343434343434
Mental status change 
Differentiation syndrome 
Rash 
Muscle/bone pain 
Dyspnea 
Elevated amylase 
Fatigue 
Chest pain 
Anorexia 
Edema 
Syncope 
No. patients       
Toxic eventNo. patients experiencing toxicity dose cohort and toxicity grade
100 mg/m2
150 mg/m2
200 mg/m2
250 mg/m2
300 mg/m2
400 mg/m2
343434343434
Mental status change 
Differentiation syndrome 
Rash 
Muscle/bone pain 
Dyspnea 
Elevated amylase 
Fatigue 
Chest pain 
Anorexia 
Edema 
Syncope 
No. patients       

NOTE: Leukemia-related side effects (neutropenia, leukopenia, anemia, thromobocytopenia, petechiae, and weight loss) were not collected. Due to early therapeutic intervention, it was not possible to grade hypertriglyceridemia and hypothyroidism.

The most common significant side effects of bexarotene were hypertriglyceridemia and hypothyroidism. These adverse events were not graded due to prophylactic and early therapeutic intervention. Hypertriglyceridemia occurred in 19 (70%) patients despite the prophylactic use of antihyperlipidemic agents and appeared to be more severe in patients treated with higher bexarotene doses. Aggressive use of antihyperlipidemic agents controlled patient triglyceride levels in all patients. Hypothyroidism was detected in 11 (41%) of patients and was easily treated with low doses of thyroid replacement hormone.

Response and treatment outcome. No patients achieved a complete response with blood count recovery. However, 5 (19%) patients achieved significant reduction in bone marrow blasts with 4 (15%) patients reducing their bone marrow blast percentage to ≤5% (Fig. 1; Table 4A). By modified International Working Group criteria, 5 (19%) of patients achieved a marrow complete response. Eleven (41%) patients had improvements in their platelet counts with 5 transfusion-dependent patients achieving transfusion independence (Table 4B). Three patients with improved platelet counts who were not transfusion dependent were on hydroxyurea at the time of study enrollment. Therefore, some of their platelet count improvement may have been due to discontinuation of hydroxyurea. By modified International Working Group criteria, 4 (15%) of 26 patients who were thrombocytopenic at the time of study initiation had a platelet response. None of the patients with an International Working Group criteria platelet response were on hydroxyurea at the time of study initiation. Seven (26%) patients, who had ANC <1,500/mm3 at study enrollment, had improvement in their neutrophil counts (Table 4C). By modified International Working Group criteria, 2 (10%) of 20 patients who were neutropenic at the time of study initiation had a neutrophil response. Median overall survival is 3.4 months with a range of 0.1 to >18.8 months. Three patients survived >1 year on study, with 2 having significant improvement in their blood counts (Fig. 2). Twenty-three patients have died due to progressive disease (13), infection (8), stroke (1), or bleeding (1). No deaths noted on study have been felt to be due to adverse events related to bexarotene but universally reflect issues related to the patients' underlying leukemia.

Fig. 1.

Bone marrow aspirate from patient 9 with H&E stain at ×40 magnification. A, pre-bexarotene with leukemic blasts. B, post-bexarotene with evidence of myeloid maturation.

Fig. 1.

Bone marrow aspirate from patient 9 with H&E stain at ×40 magnification. A, pre-bexarotene with leukemic blasts. B, post-bexarotene with evidence of myeloid maturation.

Close modal
Table 4.

Response to bexarotene

A. Bone marrow response
Patient no.Pre-bexarotene
Post-bexarotene
Cellularity% BlastsCellularity% Blasts
15 10 50 <5 
40 21 80 
15 40 90 95 20 
18 50 35 20 
26 25 25 20 <5 
B. Platelet response
 
  
Patient no.
 
Pre-bexarotene platelet count
 
Best post-bexarotene platelet count
 
61 124 
Transfusion dependent 40 
Transfusion dependent 60 
13 40* 208 
14 42* 292 
15 20 73 
16 70* 159 
18 Transfusion dependent 91 
20 44 222 
22 Transfusion dependent 42 
27 Transfusion dependent 48 
C. Neutrophil response
 
  
Patient no.
 
Pre-bexarotene ANC/mm3
 
Best post-bexarotene ANC/mm3
 
364 3,540 
570 2,759 
1,037 23,328 
14 208 1,416 
15 189 26,207 
18 28 1,200 
23 1,242 11,899 
A. Bone marrow response
Patient no.Pre-bexarotene
Post-bexarotene
Cellularity% BlastsCellularity% Blasts
15 10 50 <5 
40 21 80 
15 40 90 95 20 
18 50 35 20 
26 25 25 20 <5 
B. Platelet response
 
  
Patient no.
 
Pre-bexarotene platelet count
 
Best post-bexarotene platelet count
 
61 124 
Transfusion dependent 40 
Transfusion dependent 60 
13 40* 208 
14 42* 292 
15 20 73 
16 70* 159 
18 Transfusion dependent 91 
20 44 222 
22 Transfusion dependent 42 
27 Transfusion dependent 48 
C. Neutrophil response
 
  
Patient no.
 
Pre-bexarotene ANC/mm3
 
Best post-bexarotene ANC/mm3
 
364 3,540 
570 2,759 
1,037 23,328 
14 208 1,416 
15 189 26,207 
18 28 1,200 
23 1,242 11,899 

NOTE: Platelet counts are shown in thousands per mm3. For designation of neutrophil response, a patient had to have had pre-bexarotene ANC <1,500/mm3 with at least a 100% improvement or absolute increase of >500/mm3, whichever was greater. Patients were not allowed to have taken granulocyte colony-stimulating factor at the time of ANC assessment.

*

Patients 13, 14, and 16 were initially on hydroxyurea at the time of their pre-bexarotene platelet count.

Fig. 2.

Graph of peripheral blood neutrophil, leukemic blasts, and platelet counts over time in patients 8 (bottom) and 18 (top). Patient 8 was removed from the study for progressive disease on day +396. Patient 18 remains on study on day 565+.

Fig. 2.

Graph of peripheral blood neutrophil, leukemic blasts, and platelet counts over time in patients 8 (bottom) and 18 (top). Patient 8 was removed from the study for progressive disease on day +396. Patient 18 remains on study on day 565+.

Close modal

Evidence of myeloid differentiation. Three patients with improving ANC were studied for evidence of myeloid differentiation. Peripheral blood granulocytes were purified and subjected to fluorescence in situ hybridization analysis for the patient's known cytogenetic abnormality. The majority (92%, 100%, and 96%) of the peripheral blood granulocytes in the 3 patients studied (patients 8, 9, and 15) contained the patients known leukemic cytogenetic abnormality [5q-, t(8,21), and 5q-] suggesting a leukemic origin with myeloid differentiation.

Two patients (patients 7 and 9) developed a syndrome similar to the differentiation syndrome induced by ATRA or arsenic trioxide (ATO) in APL therapy. This was characterized by respiratory distress, dry cough, weight gain/edema, and pericardial and/or pleural effusion in the setting of a rapidly rising neutrophil count. No pulmonary infiltrates were noted by computed tomographic scans. Symptoms in both patients resolved within 48 hours after discontinuation of bexarotene and initiation of dexamethasone. Radiographic findings resolved within 2 to 4 weeks.

This phase I study has defined a clinically active and tolerable dose of bexarotene for patients with non-M3 AML. Importantly, we identified no severe DLTs in this population. In addition, we identified laboratory and clinical evidence of differentiation of non-M3 myeloid leukemic cells in response to retinoid X receptor activation as well as evidence of clinical response in chemotherapy refractory non-M3 AML. The findings establish retinoid X receptor agonists as a novel class of therapeutic agents for the treatment of non-M3 AML.

This trial has identified 300 mg/m2 as the maximum patient tolerable dose for bexarotene in patients with AML. Our study was originally done in a phase I format to identify AML-specific side effects of bexarotene that might reduce the standard treatment dose. In previous phase I and II studies in patient with cutaneous T-cell lymphoma and solid tumors, bexarotene was tested up to 600 mg/m2 and found to be safe but difficult to tolerate (9, 15). In addition, in a trial of cutaneous T-cell lymphoma, there was no difference in efficacy between 300 mg/m2 and higher doses (9). Based on those trials, the standard treatment dose of bexarotene is 300 mg/m2. Our trial confirms that although the 400 mg/m2 dose level is safe, the long-term patient tolerability is poor. When treating patients with relapsed or refractory AML or older patients ineligible for standard chemotherapy, palliative therapies must be well tolerated over time to ensure patient compliance and maximized quality of life. Given the differences in long-term drug tolerability between 300 and 400 mg/m2, we believe that 300 mg/m2 should be the dose used in future clinical trials for AML.

The adverse events associated with bexarotene use in AML patients closely mirror those reported previously in cutaneous T-cell lymphoma and lung cancer patients. Hypertriglyceridemia and hypothyroidism were frequent, but with close observation and early intervention, easily manageable (9). Events that affected quality of life including bone pain, nausea, anorexia, edema, and rash occurred but were low grade, dose dependent, and reversible. Before this trial, there was a concern that bexarotene adverse events may not be well tolerated in AML patients or that previously unreported adverse events might occur. In our AML patient population, bexarotene did not appear to worsen the clinical status of our patients.

We observed in two patients a differentiation syndrome similar to that induced by ATRA or ATO in APL, which suggests that bexarotene can induce leukemic differentiation in non-M3 AML (14, 16). Although our patients had a syndrome similar to that seen with ATRA- or ATO-induced differentiation syndrome, no pulmonary infiltrates were noted. It is possible that pulmonary infiltrates may occur in future patients treated with bexarotene or that bexarotene-induced differentiation syndrome is in some way different from that caused by ATRA or ATO. In M3 AML patients treated with ATRA or ATO, differentiation syndrome typically occurs as the WBC is rising due to leukemic blast differentiation induced by ATRA or ATO. The syndrome is felt to be due to cytokine release induced by the maturing myeloid cells. In patients with non-M3 AML treated with ATRA on previous clinical trials, neither clinical activity nor differentiation syndrome was noted, suggesting that the two activities are linked. Given that differentiation syndrome and similar pulmonary syndromes have not been reported previously in nonleukemia patients treated with bexarotene, it is likely that this reaction is not a direct side effect of bexarotene but rather that it is a therapeutic consequence of the leukemic differentiating activity induced by bexarotene (7). In our two patients, differentiation syndrome was rapidly resolved with cessation of bexarotene and institution of steroids. Neither patient was restarted on bexarotene within our clinical trial due to this significant adverse event. After achieving a bone marrow blasts reduction to <5% and clearance of peripheral blood blasts with bexarotene, one patient did well clinically for 1 month after drug discontinuation before presenting with explosive recurrent disease and expired without ever restarting bexarotene (16). The second patient restarted bexarotene off study at half his previous dose without recurrent differentiation syndrome but ultimately developed progressive disease a few weeks later. It remains to be seen if patients can be rechallenged with full-dose bexarotene after initially developing differentiation syndrome. With close monitoring for differentiation syndrome, bexarotene appears to be a safe and well-tolerated therapy in patients with non-M3 AML.

Our data suggest that bexarotene induces the differentiation of leukemic blasts into functioning neutrophils. In three responding patients tested, essentially all peripheral blood neutrophils contained the respective patient's known leukemic cytogenetic abnormality, suggesting a leukemic origin to these cells. Unfortunately, only three patients' peripheral blood were tested by fluorescence in situ hybridization and no patient had peripheral blood fluorescence in situ hybridization done pre-study, limiting our ability to quantitate any changes induced by bexarotene. Two patients experienced a pulmonary syndrome reminiscent of ATRA or differentiation syndrome seen in M3 AML patients. The finding of differentiation syndrome in our patients further suggests that the mechanism of bexarotene may be from differentiation of the leukemic blasts. Given that bexarotene has only low affinity for RAR and that ATRA has no effect on non-M3 AML, one must postulate a different mechanism than pure RAR agonist activity. RAR and RXR receptors are known to heterodimerize and interact with one another. It is possible that these RAR/RXR heterodimers have downstream activities that affect neutrophil differentiation. Whereas leukemic blast differentiation in M3 AML may occur through the RAR pathway, RXR activity may be needed to achieve similar effects in non-M3 AML. In addition, given that bexarotene has low-level affinity for RAR receptors as well as high affinity for RXR receptors, costimulation of both RAR and RXR receptors may be involved in the differentiation of non-M3 AML. Further laboratory studies will be needed to identify the mechanism of action of bexarotene.

Significant evidence of clinical activity was seen as manifested by improved blood counts and reduction in bone marrow blasts. Several patients showed increases in their neutrophil counts or platelet counts. In patients with active AML, these effects are important as pancytopenia with resulting infections and transfusion dependency is common. Unlike current cytotoxic therapies for AML, bexarotene appears to have relatively minimal bone marrow suppressive side effects and leads to improved blood counts in patients while on therapy. Currently, many patients with AML are treated with supportive care only due to the potential risks of worsening pancytopenia from chemotherapy. Bexarotene may provide a new therapeutic option that would not carry the risks of conventional chemotherapy. The positive effects of this therapy can be long-lasting in some patients, freeing patients from platelet transfusions and significantly improving their overall quality of life. Patients with AML frequently do not die directly from progressive AML but rather indirectly from infection and bleeding complications. Although most hematologic responses were short lived in our phase I study, often due to progressive AML, it is hoped that with proper dosing of bexarotene that the responses may be more robust and longer lived. In our study, three patients with relapsed or refractory AML have lived >1 year while on bexarotene alone, suggesting the possibility for improvement in survival in a subset of patients.

Although we have identified evidence of clinical activity of bexarotene monotherapy in non-M3 AML patients, the overall activity was low and transient in the majority of patients. As this was a phase I dose-finding study, it is hoped that, with proper dose administration and patient selection in future trials, the overall effect of bexarotene monotherapy will be greater. The ultimate use of bexarotene may, however, be in conjunction with other antileukemic agents. Other retinoic acids such as ATRA have been shown to improve survival in M3 AML patients when used in conjunction with conventional chemotherapy and may improve patient outcome as well when used with nonchemotherapeutic agents such as ATO. Further clinical investigations using bexarotene with other agents are ongoing.

We have shown that bexarotene has significant clinical activity in non-M3 AML and identified a safe and well-tolerated dose in this patient population. There is a significant need for novel well-tolerated therapies in patients with refractory non-M3 AML and in older patients not otherwise eligible for induction chemotherapy. Given that bexarotene has minimal bone marrow suppressive side effects and works in a manner different from cytotoxic chemotherapy, it is a promising agent for these patients and should be further investigated.

C. Andreadis is a member of the Speakers' Bureau of Eisai; A. Kemner is employed by GlaxoSmithKline, LLC.

Grant support: Ligand Pharmaceuticals, Inc. and Eisai, Inc. (trial support) and American Society of Clinical Oncology Career Development Award (D.E. Tsai). M. Carroll is a clinical scholar of the Leukemia and Lymphoma Society of America. A. Bagg is supported in part by a Specialized Center of Research grant from the Leukemia and Lymphoma Society of America.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Prior Presentation of Data in Part: This is the only complete and final report of our phase I study of bexarotene in acute myeloid leukemia.

Preliminary data were presented at the American Society of Clinical Oncology Annual Meeting 2006, American Society of Hematology Annual Meeting 2006, and American Society of Clinical Oncology Annual Meeting 2007.

1
Garattini E, Terao M. Atypical retinoids: an expanding series of anti-leukemia and anti-cancer agents endowed with selective apoptotic activity.
J Chemother
2004
;
16
:
70
–3.
2
Tallman MS, Andersen JW, Schiffer CA, et al. All-trans retinoic acid in acute promyelocytic leukemia: long term outcome and prognostic factor analysis from the North American Intergroup protocol.
Blood
2002
;
100
:
4298
–302.
3
Tallman MS. Acute promyelocytic leukemia as a paradigm for targeted therapy.
Semin Hematol
2004
;
41
:
27
–32.
4
Melnick A, Licht JD. Deconstructing a disease: RARα, its fusion partners, and their roles in the pathogenesis of acute promyelocytic leukemia.
Blood
1999
;
93
:
3167
–215.
5
Warrell RP, Frankel S, Miller WH, et al. Differentiation therapy of acute promyelocytic leukemia with tretinoin.
N Engl J Med
1991
;
324
:
1385
–93.
6
Lengfelder E, Saussele S, Weisser A, et al. Treatment concepts of acute promyelocytic leukemia.
Crit Rev Oncol Hematol
2005
;
56
:
261
–74.
7
Burnett AK, Milligan D, Prentice AG, et al. A comparison of low-dose cytarabine and hydroxyurea with or without all-trans retinoic acid for acute myeloid leukemia and high-risk myelodysplastic syndrome in patients not considered fit for intensive treatment.
Cancer
2007
;
109
:
1114
–24.
8
Tsai DE, Aqui NA, Vogl DT, et al. Successful treatment of T-cell post-transplant lymphoproliferative disorder with the retinoid analog bexarotene.
Am J Transplant
2005
;
5
:
2070
–3.
9
Duvic M, Hymes K, Hearld P, et al. Bexarotene is effective and safe for treatment of refractory advanced stage cutaneous T cell lymphoma: multinational phase II-III trial results.
J Clin Oncol
2001
;
19
:
2456
–71.
10
Boehm M, Zhang M, Badea B, et al. Synthesis and structure-activity relationship of novel retinoid X receptor-selective retinoids.
J Med Chem
1994
;
37
:
2930
–41.
11
Boehm M, Zhang L, Zhi L, et al. Design and synthesis of potent retinoid X receptor selective ligands that induce apoptosis in leukemia cells.
J Med Chem
1995
;
38
:
3146
–55.
12
Kizaki M, Dawson MI, Heyman R, et al. Effects of novel retinoid X receptor-selective ligands on myeloid leukemic differentiation and proliferation in vitro.
Blood
1996
;
87
:
1977
–84.
13
3Cheson BD, Greenberg PL, Bennett JM, et al. Clinical application and proposal for modification of the International Working Group response criteria in myelodysplasia.
Blood
2006
;
108
:
419
–25.
14
Tsai DE, Luger SM, Kemner A, et al. Evidence of myeloid differentiation in non-M3 acute myeloid leukemia treated with the retinoid X receptor agonist bexarotene.
Cancer Biol Ther
2007
;
6
:
18
–21.
15
Rizvi N, Marshall JL, Dahut W, et al. A phase I study of LGD1069 in adults with advanced cancer.
Clin Cancer Res
1999
;
5
:
1658
–64.
16
DiNardo CD, Ky B, Vogl DT, et al. Differentiation syndrome in Non-M3 acute myeloid leukemia treated with the retinoid X receptor agonist bexarotene. Med Oncol 2008. In press.

Supplementary data