Further steps to improve efficacy and safety in acute myeloid leukemia

Long-term follow-up of VIALE-A

Venetoclax 400 mg QD in combination with azacitidine 75 mg/m2 on days 1–7 has been approved for the treatment of patients with newly diagnosed acute myeloid leukemia (AML) who are ineligible for intensive chemotherapy based on the phase III VIALE-A trial that met its primary endpoint of overall survival (OS) at the time of the interim analysis conducted in March 2020 [1]. The study included patients aged ≥ 75 years or 18-74 years with comorbidities ineligible for standard induction regimens. While the experimental arm (n = 286) received the combination, the control arm was treated with placebo plus azacitidine (n = 145). At ASH 2022, the results of the 100 % OS analysis were reported to determine the long-term survival benefit of venetoclax plus azacitidine [2].

After a median follow-up of approximately 43 months, no new safety signals had emerged for venetoclax plus azacitidine or azacitidine monotherapy compared to the previous analysis. Grade ≥ 3 adverse events (AEs) were observed in 98.6 % vs. 96.5 %. The overall AE rates were similar across the arms, while hematologic AEs occurred more frequently with the combination.

Venetoclax plus azacitidine continued to show significant OS improvement over azacitidine monotherapy, with a 42 % mortality reduction (median OS, 14.7 vs. 9.6 months; HR, 0.58; p < 0.001). All subgroups derived continued OS benefit. In patients with IDH1/2 mutations, median OS was 19.9 vs. 6.2 months, which translated into a 69 % risk reduction (HR, 0.314; p < 0.001). Approximately 68 % of patients treated with venetoclax plus azacitidine achieved complete remission (CR) or CR with incomplete hematologic recovery (CRi) as best response. The median duration of CR/CRi was 18.2 vs. 10.7 months. Within this group, median OS was longer in patients who had obtained MRD < 10-3 compared to those with MRD ≥ 10-3 (34.2 vs. 18.7 months). Median duration of CR for patients on the combination treatment (22.1 months) was approximately 5 months longer than the respective result observed with venetoclax plus aza–citidine at the time of the primary analysis (17.5 months). Both analyses yielded median duration of CR of approximately 13 months in the control arms.

The 7 + 7 scheme

The VIALE-A regimen which contains continuous venetoclax exposure confers considerable toxicity including high rates of grade 3/4 febrile neutropenia and post-remission grade 4 cytopenia [1]. Therefore, Willekens et al. assessed the efficacy and safety of a reduced venetoclax schedule as a potential means to improve tolerability [3]. This multicentric retrospective study included a frail first-line AML population in whom venetoclax had been restricted to 7 days per cycle and was administered concurrently with azacitidine (7 + 7 scheme). The patients’ median age was comparable to the that of the VIALE-A study population, although 37 % had ECOG performance status 2-4, and 29.3 % showed comorbidities that had been defined as exclusion criteria in VIALE-A. Moreover, the group was enriched for patients with therapy-related AML (32 %). In the cohort aged < 75 years, prior or concomitant neoplasia was mainly responsible for the ineligibility for intensive treatment (47 %). Poor cytogenetic risk was present in 32.9 %. Overall, nearly 70 % of patients fell into the “adverse risk” category according to ELN 2022.

A total of 82 patients were included in the analysis. After a median follow-up of 4.8 months, 39 (48 %) were still on treatment. Discontinuation was due to failure/relapse in 63 % and toxicity in CR/CRi or morphologic leukemia-free state (MLFS) in 30 %. The median number of cycles was 4 (1–28). Increasing proportions of patients achieved CR/CRi with every cycle, which resulted in a global response of 68.3 % that was similar to the results noted with conventional 28-day venetoclax exposure. CR/CRi rates ranged from 43.7 % to 100 % according to the genetic set-up, with a normal karyotype being associated with higher responses while the presence of TP53 correlated with the lowest CR/CRi rate. Estimated OS and event-free survival (EFS) findings were encouraging and most favorable in the cohort with NPM1 and/or IDH2 mutations (Table 1).

Supplemental dose reductions remained frequent in the post-remission phase (60.7 % of CR/CRi patients). The cycles were shortened to 5 days in almost all patients. Moreover, the venetoclax dose was reduced to 200 mg/d in half of them, and the interval between cycles was expanded to 35 days in approximately 30 %. Seventy percent of patients with further dose reductions were still on therapy at the time of the analysis, and their estimated median OS was 25.8 months.

Early toxicity in cycle 1 was similar to the toxicity observed with the conventional venetoclax schedule. Febrile neutropenia occurred in 48.8 %, and 88.3 % of patients required platelet or erythrocyte transfusions. The median delay before cycle 2 was 13 days. Patients with CR/CRi had a median 36-day delay to neutrophil recovery > 1 G/L and a 31-day delay to platelet recovery > 100 G/L. As the authors emphasized, longer follow-up is necessary for the confirmation of the OS and EFS results as well as the assessment of toxicity after cycle 2.

Table 1 Overall survival and event-free survival with reduced venetoclax exposure in the total population and subgroups

BGB-11417 plus azacitidine

Another approach to optimize Bcl-2-inhibitor–based AML treatment is the use of improved agents such as the potent and selective Bcl-2 inhibitor BGB-11417. In an AML xenograft model, BGB-11417 has shown greater tumor reduction than venetoclax at the same dose level, both alone and when combined with azacitidine [4]. The phase IB/II dose-finding and dose-escalation study BGB-11417-103 is assessing BGB-11417 plus azacitidine in treatment-naïve AML patients unfit for intensive chemotherapy and patients with relapsed/refractory AML who had no prior Bcl-2-inhibitor or azacitidine exposure. As of the data cutoff, 57 patients were dosed in four cohorts (40 mg, 80 mg or 160 mg for 10 days or 160 mg for 28 days in 28-day cycles). The treatment-naïve and relapsed/refractory groups comprised 31 and 26 individuals, respectively.

According to the update presented at ASH 2022, BGB-11417 plus azacitidine was generally well tolerated [5]. Dose-limiting toxicities occurred only in the 80 mg cohort, and no clinical tumor lysis syndrome was observed. Laboratory tumor lysis syndrome emerged in one patient treated with 160 mg for 10 days who fully recovered. The analysis showed no dose-dependent toxicities, and the maxi–mum tolerated dose was not reached. Neutropenia represented the most common treatment-emergent grade ≥ 3 AE (65 %) but was manageable with dose modifications. Grade ≥ 3 thrombocytopenia and anemia were observed in 51 % and 35 %, respectively. Nausea and constipation prevailed among the most common non-hematologic treatment-emergent AEs but were mostly grade 1 and 2.

BGB-11417 plus azacitidine proved active in both treatment-naïve and relapsed/refractory settings at the four dose levels tested. The combination gave rise to CR and CR with partial hematologic recovery (CRh) in 65 % and 50 % of treatment-naïve and relapsed/refractory patients, respectively. CR/CRh in the treatment-naïve group was mostly obtained by the end of cycle 1. The 80-mg cohort experienced the longest treatment duration (median, 7 cycles) and the highest CR rates (Figure 1). Almost half of 27 MRD-evaluable patients achieved MRD negativity. Efficacy analyses of molecular subgroups, safety expansion, and evaluation of higher doses of BGB-11417 are ongoing, and the inclusion of patients with AML who failed hypomethylating agents is planned.

Figure 1: BGB-11417 plus azacitidine: response rates across the four dose levels

Figure 1: BGB-11417 plus azacitidine: response rates across the four dose levels

Magrolimab on top of AV

Although the administration of Bcl-2 inhibition plus hypomethylating agents is undoubtedly effective in newly diagnosed AML patients ineligible for intensive chemotherapy, the survival rates beyond 2 years are low [1]. Jia et al. showed that combined blockade of the CD47-SIRPa axis by the first-in-class anti-CD47 antibody magrolimab plus azacitidine/venetoclax increases phagocytosis in AML cell lines regardless of TP53 mutation status in AML preclinical models, thus facilitating macrophage-mediated anti-leukemia efficacy [6]. In the clinical setting, phase IB results have indicated efficacy and tolerability of magrolimab plus azacitidine in AML patients unfit for intensive chemotherapy [7].

Magrolimab plus azacitidine and venetoclax is being assessed in a phase I/II trial containing three phase II cohorts: patients with de novo and secondary AML (frontline cohort), those with relapsed/refractory venetoclax-naïve disease, and those with relapsed/refractory disease after prior venetoclax. The recommended phase II dose of magrolimab has been established at 1 mg/kg on days 1 and 4 followed by 15 mg/kg on day 8 and 30 mg/kg once weekly from day 11 of cycle 1 until the end of cycle 2. From cycles 3 onward, magrolimab 30 mg/kg was administered every 2 weeks. At ASH 2022, Daver et al. reported efficacy and safety findings for the frontline and relapsed/refractory AML groups [8].

Favorable results in frontline de novo AML

In the cohort treated in the frontline setting (n = 43), the disease was classified as adverse according to the ELN 2017 risk stratification in almost all of the patients, and high-risk cytogenetics were present in the majority of cases. Nevertheless, 72 % responded by achieving CR or CRi, with CR/CRi rates being similar across TP53-mutated and TP53-wildtype patients (Table 2). MRD negativity resulted in 67 %. In patients with de novo AML, median duration of response and median OS had not been reached in either TP53 cohort at the time of the analysis. Twelve-month OS rates were 83 % and 53 % for the TP53-wildtype and TP53-mutant groups, respectively. The cohort with secondary AML fared worse than the de novo cohort; here, median OS was 9.6 and 7.6 months for the TP53-wildtype and TP53-mutant cohorts, respectively.

A propensity matched analysis demonstrated improved OS with the triple combination compared to venetoclax plus hypomethylating treatment in TP53– mutated patients (p = 0.02), although the median follow-up and numbers remained small. In the TP53 wildtype population, the sample size was insufficient for a historical comparison. TP53-mutated patients who went on to receive stem cell transplant (SCT) experienced higher recurrence-free survival compared to those who did not (16.3 vs. 4.2 months); the same applied to OS (16.6 vs. 9.8 months).

In patients with relapsed/refractory AML (n = 36), the activity of magrolimab plus azacitidine and venetoclax was modest. Only 11 % and 44 % developed CR/CRi in the venetoclax-exposed and venetoclax-naïve groups, respectively. Recurrence-free survival and OS were low in both cohorts.

Overall, the treatment with magrolimab plus venetoclax and azacitidine was safe in this population with very high risk. The analysis revealed no unexpected AEs, and none of the patients discontinued therapy due to treatment-related events. Infusion reactions occurred in 8 patients (10 %), with three of them experiencing grade 3 events, although further reactions were effectively mitigated based on dexamethasone premedication. Eighteen individuals (23 %) had grade ≥ 3 anemia; in the frontline cohort, hemoglobin levels decreased by a median of 1.2 g/dl (0–3.9 g/dl) after the first magrolimab infusion. The authors recommended careful monitoring of hemoglobin levels, especially between days 1 and 10 of cycle 1. The randomized phase III ENHANCE-2 (NCT04778397) and ENHANCE-3 (NCT05079230) studies are further assessing magrolimab in frontline AML.

Table 2 Responses to treatment with magrolimab, azacitidine and venetoclax in the frontline setting

Olutasidenib: relapsed/refractory IHD1-mutated disease

IDH1 mutations are found in 7–14 % of patients with AML [9]. The potent, selective, oral, small-molecule IDH1 inhibitor olutasidenib has demonstrated a manageable tolerability profile and clinical activity in the completed phase I portion of the open-label, multicenter, phase I/II 2102-HEM-101 study [10]. Results from the pivotal phase II cohort of patients with relapsed/refractory AML who received olutasidenib 150 mg BID were reported at ASH 2022 [11]. The efficacy-evaluable and safety sets included 147 and 153 patients, respectively. They had received a median of 2 prior therapies, and 12 % had undergone SCT. Refractory disease was present in 35 %.

The CR/CRh rate, which constituted the primary endpoint, was 35 %, with 32 % of patients having developed CR. Median duration of CR/CRh and CR was 25.9 and 28.1 months, respectively. Notably, response rates were similar for 12 patients who had received prior venetoclax; here, the CR/CRh and CR rates amounted to 33 % and 25 %, respectively. In the total population of 153 individuals, median OS was 11.6 months. The OS findings differed considerably according to the depth of response (Figure 2). In the group with CR/CRh, median OS had not been reached at the time of the analysis, and the 18-month OS rate was 78 %. Across all response groups, 34 % of patients who were platelet- and/or red blood cell-transfusion–dependent at baseline reached transfusion independence at 56 days. Those who achieved CR/CRh had comparatively higher rates of transfusion independence than other responders. Sixteen patients in the total population (11 %) were able to proceed to SCT following olutasidenib treatment.

With regard to safety, the most common AEs included nausea, differentiation syndrome, and leukocytosis. Differentiation syndrome occurred in 14 % of patients, with grade ≥ 3 events in 9 % and one fatal case. However, most cases resolved with treatment interruption, dexamethasone therapy and/or supportive measures. Likewise, hepatic AEs that occurred in 25 % (grade 3, 12 %) were manageable with dose modifications and concomitant medications. QTc prolongation occurred in 8 % (grade 3, 1 %). In their summary, the authors pointed out that olutasidenib induced durable remissions and transfusion indepen–dence with a well-characterized and manageable safety profile. The observed activity was deemed to represent a therapeutic advance in this poor-prognosis patient population.

Figure 2: Overall survival according to the depth of response to olutasidenib therapy

Figure 2: Overall survival according to the depth of response to olutasidenib therapy

MRD-guided pre-emptive FLT3-targeted therapy

FLT3-mutated AML confers a high risk of relapse [12]. Sequential MRD monitoring using quantitative PCR (qPCR) has the potential to enable pre-emptive intervention as it can identify impending hematologic relapse that inevitably follows untreated molecular relapse and is associated with poor outcomes [13, 14]. According to unpublished data from the UK NCRI AML19 study, most patients with FLT3-positive AML have co-mutations suitable for molecular qPCR MRD assessment, with NPM1 mutations occurring most frequently in 55 %.

Given the scarcity of data on whether and how to treat molecular relapse/persistence, Othman et al. hypothesized that treatment with a FLT3 inhibitor might be effective in patients who had FLT3 mutation at diagnosis and experience molecular failure. The scientists retrospectively identified 51 AML patients from 27 UK hospitals who had FLT3-ITD or FLT3-TKD mutation at diagnosis, underwent MRD monitoring of NPM1 or a fusion gene based on qPCR, and experienced molecular failure that was treated with single-agent FLT3 inhibitor therapy (i.e., gilteritinib, sorafenib and quizartinib in 65 %, 19 % and 16 %, respectively). Types of failure included molecular relapse, molecular persistence > 2 copies/100 ABL and molecular progression. Among the 51 patients assessed, most had molecular relapse (49 %). In 51 %, prior midostaurin therapy had been administered. Thirty-one percent had received allogeneic SCT.

FLT3 inhibitor monotherapy for molecular failure gave rise to high rates of molecular response and excellent OS [15]. In the total group, 63 % of patients achieved CR or molecular response. The CR rate was as high as 93 % in the group with prior allogeneic SCT. At 2 years, 79 % of all patients were alive, 69 % were event-free, and 58 % had molecular EFS. After prior midostaurin, the molecular response rate was lower (54 %) than in non-pretreated patients (74 %). As many as 19 patients could be bridged to allogeneic SCT, and six were administered donor lymphocyte infusion (DLI) with FLT3 inhibition. Responses deepened after SCT or DLI, with the overall CR rate increasing from 48 % to 96 % in these 25 patients.

The treatment was well tolerated. In the first 4 cycles, no neutropenia < 0.5 x 109/L or thrombocytopenia < 25 x 109/L was found in 25 patients for whom data were available. No transfusions were administered in this group. Eight patients underwent 12 hospital admissions, with a median length of stay of 4.5 days.

At the time the patients were treated, highly sensitive FLT3 MRD assays had not been available. Data from 35 individuals showed that next-generation sequencing using the get ITD assay allows for improved detection of FLT3-ITD and thus for better patient selection [16]. Overall, the authors noted that in the presence of an appropriate marker, sequential molecular MRD monitoring can be utilized to guide pre-emptive therapy.


  1. DiNardo CD et al., Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med 2020; 383(7): 617-629
  2. Pratz KW et al., Long-term follow-up of the phase 3 VIALE-A clinical trial of venetoclax plus azacitidine for patients with treatment-naïve acute myeloid leukemia ineligible for intensive chemotherapy. ASH 2022, abstract 219
  3. Willekens C et al., Reduced venetoclax exposition to seven days of azacitidine is efficient in treatment-naïve patients with acute myeloid leukemia. ASH 2022, abstract 222
  4. Data on file. BGB-11417 Investigator Brochure
  5. Shortt J et al., Preliminary safety and efficacy of BGB-11417, a novel Bcl-2 inhibitor, in combination with azacitidine in patients with acute myeloid leukemia. ASH 2022, abstract 1443
  6. Jia Y et al., Combined blockade of CD47-sirpa interaction by 5F9 (magrolimab) and azacitidine/venetoclax therapy facilitates macrophage-mediated anti-leukemia efficacy in AML pre-clinical models. Blood 2021; 138 (Suppl 1): 510
  7. Sallman DA et al., The first-in-class anti-CD47 antibody magrolimab combined with azacitidine is well-tolerated and effective in AML patients: phase 1b results. ASH 2020, abstract 330
  8. Daver NG et al., Phase I/II study of azacitidine, venetoclax and magrolimab for newly diagnosed and relapsed/refractory AML. ASH 2022, abstract 61
  9. Medeiros BC et al., Isocitrate dehydrogenase mutations in myeloid malignancies. Leukemia 2017; 31(2): 272-281
  10. Watts JM et al., Olutasidenib alone or with azacitidine in IDH1-mutated acute myeloid leukaemia and myelodysplastic syndrome: phase 1 results of a phase 1/2 trial. Lancet Haematol 2022 Nov 9; S2352-3026(22)00292-7
  11. Cortes J et al., Olutasidenib (FT-2102) induces durable complete remissions in patients with relapsed/refractory mIDH1 acute myeloid leukemia. Results from a planned interim analysis of a phase 2 pivotal clinical trial. ASH 2022, abstract 2757
  12. Stone RM et al., Midostaurin plus chemotherapy for acute myeloid leukemia with a FLT3 mutation. N Engl J Med 2017; 377(5): 454-464
  13. Perl AE et al., Gilteritinib or chemotherapy for relapsed or refractory FLT3-mutated AML. N Engl J Med 2019; 381(18): 1728-1740
  14. Ivey A et al., Assessment of minimal residual disease in standard-risk AML. N Engl J Med 2016; 374(5): 422-433
  15. Othman J et al., High molecular response rate and overall survival with FLT3 inhibitors as MRD-guided salvage treatment for molecular failure in AML. ASH 2022, abstract 829
  16. Blätte TJ et al., getITD for FLT3-ITD-based MRD monitoring in AML. Leukemia 2019; 33(10): 2535-2539

© 2023 Springer-Verlag GmbH, Impressum

More posts