Treatment of FLT3-ITD-Positive Acute Myeloid Leukemia Relapsing after Allogeneic Stem Cell Transplantation with Sorafenib

Abstract

Patients with acute myeloid leukemia (AML) and internal tandem duplication of FMS-like tyrosine kinase receptor-3 gene (FLT3-ITD) mutation have poor prognoses and are often treated with allogeneic hematopoietic stem cell transplantation (HSCT). Sorafenib, an inhibitor of multiple kinases including FLT3, has shown promising activity in FLT3-ITD-positive AML. We treated 16 patients with FLT3-ITD-positive AML who relapsed after HSCT with sorafenib alone (n = 8) or in combination with cytotoxic chemotherapy (n = 8). The number of circulating blasts decreased in 80% of cases, but none of the patients achieved complete remission (CR); 3 achieved partial remission. Two patients were bridged to a second transplantation but both relapsed within 3 months of the transplantation. Median overall survival (OS) was 83 days, with none surviving more than a year. Sorafenib is not effective in the treatment of FLT3-ITD-positive AML relapsing after HSCT. Preventive strategies after HSCT may be more suitable for these high-risk patients.

INTRODUCTION

FMS-like tyrosine kinase receptor-3 (FLT3) is a transmembrane protein important in proliferation and survival of hematopoietic stem cells (HSC) upon activation [1]. Internal tandem duplication of the juxtamembrane domain of the FLT3 gene (FLT3-ITD) leads to kinase activity and activation of downstream signaling pathways including the MAPK pathway [2]. FLT3-ITD has been reported in approximately a one-quarter of patients with acute myeloid leukemia (AML) and is associated with higher relapse rates and shorter survival [3–5]. As the mutation portends an increased risk of disease relapse following chemotherapy alone, allogeneic hematopoietic stem cell transplantation (HSCT) is frequently proposed if a donor is identified. Although a proportion of patients will be cured following this approach, a significant number will suffer disease recurrence following the transplantation.

Sorafenib is an oral, small-molecule, multikinase inhibitor that may restrain proliferation of leukemia cells by inhibition of the MAPK pathway through raf-1 induction of apoptosis through mcl-1 [6,7], in addition to directly targeting mutant FLT3 [8]. It was found to be active in patients with FLT3-ITD-positive AML in phase I trials [9,10]. Sorafenib has also been successfully used to treat relapsed FLT3-ITD-positive AML following allogeneic HSCT [11–16]. Here, we reviewed our experience with this drug in patients with FLT3-ITD-positive AML who relapsed after allogeneic HSCT.

MATERIALS AND METHODS

We identified all patients who received sorafenib for at least 7 days, either alone or with chemotherapy, to treat FLT3-ITD-positive AML relapse after allogeneic HSCT in our institution. The retrospective chart review protocol was approved by the institutional review board (IRB). Demographic and transplant-related information was collected, as well as relapse-specific data, sorafenib dose, and duration of treatment. Treatment response was defined according to International Working Group criteria [17]. Overall survival (OS) was defined as time from sorafenib initiation to death. Actuarial survival curves were estimated according to the Kaplan-Meier method, and the significance of differences between the curves was estimated by the log-rank test.

RESULTS

Sixteen patients were treated (Table 1). Four patients had a second transplantation before sorafenib therapy, whereas 12 received sorafenib after the first transplantation. Only 3 patients (19%) were in complete remission (CR) at the time of the first transplantation. The preparative regimen was of reduced intensity (n = 7) or myeloablative (n = 9). Three patients received CD34-selected stem cells; hence, neither received graft-versus-host disease (GVHD) prophylaxis nor developed GVHD. Of the remaining 13 patients, 4 developed acute GVHD (aGVHD) of the skin before disease relapse. All 13 were receiving tacrolimus for prophylaxis or treatment of GVHD in addition to mycophenolate mofetil (n = 3) or systemic steroids (n = 1) at the time of disease recurrence. The median remission duration following transplantation was 3 months (range: 1–7 months).

Table 1.

Patient Characteristics and Response to Therapy (N = 16)

Characteristics	n (%)
Median age at diagnosis*	34 (20–63)
Antecedent hematologic history	4 (25%)
Cytogenetic risk based on AML diagnosis
High	13 (81%)
Intermediate	2 (13%)
Unknown	1 (6%)
Sorafenib use before transplantation	6 (38%)
Complete remission at transplantation	3 (19%)
Donor type
Matched related	5 (31%)
Mismatched related	4 (25%)
Matched unrelated	5 (31%)
Unrelated cord blood	2 (13%)
Remission duration following transplantation (months)*	3 (1–7)
Posttransplantation salvage therapy before sorafenib	7 (44%)
Number of salvage regimens before sorafenib*	0 (0–5)
Sorafenib therapy
Alone	8 (50%)
In combination with chemotherapy	8 (50%)
Duration of sorafenib treatment in days (range)*
Alone	39 (10–100)
In combination with chemotherapy	7 (7–32)
WBC count before sorafenib (103/μL)*	22.6 (0.6–119)
Peripheral blast percentage before sorafenib*	65% (0%–80%)
Median percentage decrease in circulating peripheral blasts (n = 12)*	50% (0%–88%)
Bone marrow blast percentage before sorafenib*	58.5% (12%–88%)
Median absolute decrease in bone marrow blast percentage*	0% (0–46)
CR following sorafenib	None
PR following sorafenib	3 (19%)
New or worsening GVHD following sorafenib	1 (6%)
Bridged to second transplantation	2 (13%)
Time from bridged second transplantation to relapse (days)*	(53 and 106)

WBC indicates white blood cell; CR, complete remission; PR, partial remission; GVHD, graft-versus-host disease.

^*Median (range).

Sorafenib Treatment

Six patients (38%) had received sorafenib before HSCT, either as part of the induction therapy or as a salvage regimen. In 9 patients (56%), sorafenib with or without chemotherapy was the first salvage therapy following allogeneic HSCT. The drug was given either alone or in combination with other cytotoxic therapy in 8 (50%) and 8 (50%) patients, respectively.

Sorafenib was used as a single agent orally twice daily at 400 mg (n = 6), or 600 mg (n = 2), on a 3-week cycle, either 5 days on therapy and 2 days off weekly, or 14 days on therapy and 7 days off therapy. When combined with chemotherapy, sorafenib was given as 400 mg daily (n = 4) or 400 mg twice daily (n = 4). Combined therapy included cytarabine and idarubicin (n = 7), or azacitidine (n = 1). Median duration of single-agent sorafenib treatment was 39 days, whereas median duration of sorafenib administration with chemotherapy was 7 days. Of 8 patients who received sorafenib alone, 4 developed grade ≥2 adverse events that included a grade 2 increase in alanine aminotransferase, grade 3 fatigue, grade 3 diarrhea, and grade 3 hyperbilirubinemia. None of the patients experienced worsening or development of GVHD after sorafenib treatment.

Response to Sorafenib

Nine patients had pre- and postsorafenib bone marrow aspirations performed: 3 patients (19%) achieved a partial remission. The responders included 2 patients who received sorafenib alone, and 1 patient received sorafenib in combination with other chemotherapy agents. A bone marrow examination was not performed in the remaining 7 patients because of disease progression (apparent from peripheral blast counts), inadequate bone marrow samples, or patient death. Peripheral blast data was available in the majority (75%) of patients (unless pancytopenic).

The median decrease in bone marrow blast percentage was 0%, whereas median absolute reduction in peripheral blood blast percentage was 50%. A reduction in the number of circulating blasts was evident in 80% of the cases, similarly distributed in the single agent versus combination therapy subgroups. Given the low response rate, no dose or schedule appeared superior. Of the 6 patients who had received sorafenib before transplantation, 1 achieved a partial remission; 3 had at least a 50% reduction in peripheral blood blasts, while 2 were refractory. Median OS was 83 days (Figure 1). There was no difference in survival between patients who received sorafenib alone or in combination with chemotherapy.

Figure 1.

(A) OS of all patients treated with sorafenib. (B) OS of patients treated with single-agent sorafenib and those treated in combination with cytotoxic therapy (P = .45).

Two patients were “bridged” to a second transplantation. One was a 22-year-old male transplanted after failing induction chemotherapy who achieved a brief remission after the first transplantation (2 months). Sorafenib and azacitidine induced a reduction in marrow blasts, and a second transplantation was performed, after which he relapsed in 3 months. The second patient was a 61-year-old female who relapsed 14 days after the first transplantation. She then received single-agent sorafenib, which led to a partial remission, followed by a second allogeneic HSCT. Her disease relapsed 1 month after this transplantation.

DISCUSSION

Here, we reported a series of 16 FLT3-ITD-positive AML patients treated with sorafenib after failing allogeneic HSCT. Our experience is in contrast to that reported in the literature, which includes complete molecular responses. Although this could be attributed to the particularly advanced disease in our cases—only 3 patients underwent transplantation in CR, and the majority relapsed within 100 days of HSCT—it is striking that median survival here was <3 months. Another potential reason for failing to respond to sorafenib is the dose schedule used here. It is possible that alternative schedules, such as continuous dosing (single agent), or a different combination with other agents may induce a sustainable inhibition of FLT3, resulting in improved clinical response. In addition, Pratz and collaborators [18] recently showed that the FLT3 ligand can interfere with the action of kinase inhibitors, a mechanism that conceivably could be operative here. Nevertheless, response to sorafenib was dismal when compared with the CR rate of 44%—previously reported from our institution—achieved in AML/myelodysplastic syndrome (MDS) patients with relapsed disease after HSCT who were treated with chemotherapy [19].

It has been hypothesized that the antileukemia effect of FLT3 inhibitors could be improved by disruption of stroma–leukemia interaction through CXCR4 inhibition. FLT3-ITD appears to activate CXCR4 signaling. Zeng et al. [20] combined sorafenib with the CXCR4 inhibitor AMD3465, demonstrating complete blockage of prosurvival signaling pathways in AML cell lines. Furthermore, AMD3465 enhanced sorafenib-induced apoptosis in samples from primary AML patients with FLT3 mutation. This concept is under investigation in the transplantation-and chemotherapy-only scenarios [21].

Others have postulated that sorafenib could increase the prevalence and severity of GVHD after T cell replete HSCT based on preliminary findings from a mice study [22]. Three patients were reported elsewhere to either develop or experience worsening of GVHD after sorafenib initiation [11,12,15]. However, none of our patients experienced worsening GVHD with sorafenib, indicating that larger number of patients will have to be treated before conclusions can be drawn.

Leukemia relapse after HSCT remains a major cause of treatment failure [23]. Current treatment options are limited and include donor lymphocyte infusion, chemotherapy, and second HSCT. Treatment choice is usually based on institutional experience and disease characteristics, but only a small minority of patients will benefit from current salvage therapies. Longer remission duration after HSCT is probably the single most important prognostic factor [24]. Prevention of relapse after HSCT in patients with high-risk disease may be a better strategy, ideally using FLT3 inhibitors for patients harboring this mutation. Modification of preparative regimens with the inclusion of more effective antileukemic agents [25], or the use of “sequential conditioning” with cytoreductive chemotherapy followed by reduced-intensity conditioning (RIC) [26,27], may also reduce relapse rates. Furthermore, preemptive treatment of patients with rising quantitative FLT3 polymerase chain reaction levels, before hematologic relapse, is a possible approach to be investigated. In a broader sense in AML, we are currently investigating maintenance of remission therapy with low-dose azacitidine [28]. Interestingly, this drug was demonstrated by others to increase the expression of C/EBPδ, a transcription factor that is frequently suppressed in AML and exhibits growth inhibitory properties in FLT3-ITD-positive AML cell lines [29].

Our results would argue against using sorafenib as described here to treat FLT3-ITD AML relapse after allogeneic transplantation. Innovative approaches are urgently needed to prevent and treat recurrence in this setting.