FLA-IDA salvage chemotherapy combined with a seven-day course of venetoclax (FLAVIDA) in patients with relapsed/refractory acute leukaemia

To the editor

Venetoclax is a potent orally available BH3 mimetic promoting apoptosis through selective inhibition of pro-survival protein B-cell lymphoma 2 (BCL2) (Bhola et al., 2016).

Single-agent venetoclax was studied in AML demonstrating some biological activity and good tolerability, though response rates were short-lived (Konopleva et al., 2016).

Only limited data is available for the efficacy of venetoclax in relapsed/refractory AML patients. Reported response rates vary between 21% and 64% in the relapse setting when venetoclax was combined with HMA or LDAC and predictors of response are yet unknown (DiNardo et al., 2018a, Aldoss et al., 2018). Even less data is available on the feasibility of combining venetoclax with standard induction or salvage chemotherapy regimens in younger, medically fit patients with acute leukaemia. DiNardo and colleagues combined venetoclax (200 mg orally days 1-21) with the commonly used salvage regimen FLAG-IDA in relapsed or refractory AML patients (DiNardo et al., 2018b). This regimen provided notable activity with an overall response rate of 73%. We report the safety and initial efficacy of a different dosing scheme of venetoclax with FLA-IDA chemotherapy based on data reported to our venetoclax registry (ClinicalTrials.gov NCT03662724). Patients aged 18 years or older with refractory or relapsed acute myeloid leukaemia as defined by the 2016 World Health Organization criteria (Arber et al., 2016), who had been treated with fludarabine, cytarabine and idarubicin combined with venetoclax and were reported to the venetoclax registry (venreg.org) were included in the analysis. Venetoclax was administered at a dose of 100 mg once daily perorally days 1-7 due to mandatory co-medication with a CYP3A4 inhibitor for fungal prophylaxis. All patients had given written informed consent to the off-label use of venetoclax, genetic analysis and use of clinical data according to the Declaration of Helsinki and institutional guidelines. The registry was approved by the local Ethics Review Committee (ethical vote No.7972_BO_K_2018) and registered at https://www.clinicaltrials.gov (NCT03662724). The control patients were selected from the in-house database of Hannover Medical School and were treated with FLA-IDA between 2000 and 2018 for relapsed or refractory AML. Eighty-one patients were identified that fulfilled the selection criteria of the FLAVIDA patients and were compared to the FLAVIDA patients included in this analysis (detailed information about the dosing scheme are provided in Supplementary materials). Molecular MRD was assessed as described previously (Thol et al., 2018).

The data cut-off for this analysis was August 15, 2019. Thirteen patients had received FLAVIDA and had safety and efficacy outcomes reported. Comparison of demographic and baseline characteristics indicate that the control cohort was representative of the patients included in the FLAVIDA analysis (Table 1, Supplementary Table S1). Most commonly observed treatment-related toxicities of any grade were haematological and infectious adverse events being reported in all patients. The most common grade 3/4 all-causality event was neutropenic fever (77%), while thrombocytopenia, anaemia, and neutropenia were reported in all patients at nadir (100%). Non-haematologic adverse events of all grades occurred in all patients with gram-negative bacteraemia (23%) being most frequent and requiring intravenous broad-spectrum anti-infective therapy (Supplementary Table S2).

Table 1. Patient demographics and baseline characteristics of FLAVIDA and FLA-IDA control patients.

Baseline characteristics	FLAVIDA (n=13)	FLA-IDA control (n=81)	P
Age (years)     Median (range)	49 (18-62)	52 (22-72)	0·39
Sex n (%)     Male     Female	8 (61·5) 5 (38·5)	45 (55·6) 36 (44·4)	0·69
Type of AML n (%)     De novo     Secondary     Missing	8 (61·5) 5 (38·5) -	61 (75·3) 19 (23·5) 1 (1·2)	0·49
FAB Subtype n (%)     M0     M1     M2     M4     M4eo     M5     M6     M7     Missing	- 4 (30·8) 3 (23·1) 3 (23·1) - 3 (23·1) - - -	11 (13·6) 11 (13·6) 12 (14·8) 12 (14·8) 1 (1·2) 16 (19·8) - 1 (1·2) 17 (21)	0·27
ELN classification 2017 n (%)     Favourable     Intermediate     Adverse     Missing	2 (15·4) 7 (53·8) 4 (30·8) -	17 (21) 35 (43·2) 18 (22·2) 11 (13·6)	0·46
Complex karyotype n (%)     Yes     No     Missing	2 (15·4) 11(84·6) -	7 (8.6) 66 (81.5) 8 (9.9)	0.41
Treatment lines before FLAV/IDA     Median (range)	1 (1-5)	1 (1-4)	0·42
WBC at start of salvage (x109/l)     Median (range)     Missing	3·0 (0·5-77) -	2·8 (0·1-117·3) 14	0·45
Haemoglobin at start salvage (g/dl)     Median (range)     Missing	9·3 (6·5-11·6) -	9·3 (6·6-14·9) 14	0·65
Platelet count at start of salvage (x109/l)     Median (range)     Missing	47 (5-202) -	51 (5-731) 14	0·58
Blasts in BM at start of salvage (%)     Median (range)     Missing	70 (12-100) 1	50 (10-100) 42	0·06
Blasts in PB at start of salvage (%)     Median (range)     Missing	7 (0-98) -	15 (10-90) 38	0·68
FLT3 status n (%)     Wildtype     Mutated     Missing	9 (69·2) 4 (30·8) -	49 (60·5) 13 (16) 19 (23·5)	0·87
NPM1 status n (%)     Wildtype     Mutated     Missing	10 (76·9) 3 (23·1) -	43 (53) 19 (23·5) 19 (23·5)	0·27
AlloHCT before FLAV/IDA n (%)     Yes     No	6 (46·2) 7 (53·8)	22 (27·2) 59 (72·8)	0·17
AlloHCT after FLAV/IDA n (%)     First alloHCT     Second alloHCT     No     Missing	7 (53·8) 2 (15·4) 4 (30·8) -	49 (60·5) 6 (7·4) 25 (30·9) 1 (1·2)	1·0
Use of G-CSF     Concomitant to chemotherapy     After chemotherapy     No use     Missing	- 8 (61·5) 5 (38·5) -	50 (61·7) 6 (7·4) 18 (22·2) 7 (8·6)	-

Dates are n (%) or median (range) unless otherwise noted. ECOG=Eastern Cooperative Oncology Group; WBC=white-cell count; BM=bone marrow; PB=peripheral blood; alloHCT=allogeneic hematopoietic cell transplantation.

The median time to neutrophil recovery (≥0·5x10⁹/l) in responding FLAVIDA-treated patients was 33 days (95% CI, 26-40) and 32 days (95% CI, 28-37) in the FLA-IDA control cohort (P=0·46) from day one of chemotherapy (Figure 1A). Recovery times for ANC ≥1·0x10⁹/l for responding FLAVIDA patients were 35 days (95% CI, 34-36), and were similar to those observed in the control cohort (39 days (95% CI, 36-42) (P=0·16) (Figure 1B)). Median times for platelet recovery (≥50x10⁹/l) were 36 days (95% CI, 27-45) in the FLAVIDA cohort vs. 39 days (95% CI, 19-59) in the control cohort (P=0·24) (Figure 1C). Recovery times for platelet count >100x10⁹/l for responding FLAVIDA patients were 40 days (95% CI, 28-51) compared to 41 days (95% CI, 28-54) in the control cohort (P=0·31) (Figure 1D). Thus, ANC and platelet recovery were overall comparable between FLAVIDA and FLA-IDA regimens suggesting that a seven-day course of venetoclax does not add significant toxicity to the FLA-IDA regimen. (Supplementary Table S3). Importantly, patients with a previous history of alloHCT had almost identical recovery times of neutrophils and platelets as non-transplanted patients.

Figure 1. Blood count recovery in responding patients after FLAVIDA and FLA-IDA regimens for neutrophils (ANC) and platelets (PLT).

Cumulative incidence of blood count recovery amongst responding patients in the FLAVIDA cohort (red line) and patients in the FLA-IDA control cohort (blue line). Panel A displays the cumulative incidence of neutrophil recovery to ANC ≥ 0·5x10⁹/l. Panel B shows the corresponding cumulative incidence of neutrophil recovery to ANC ≥ 1·0x10⁹/l. Panel C displays the cumulative incidence of platelet recovery to PLT ≥ 50x10⁹/l. Panel D displays the cumulative incidence of platelet recovery to PLT ≥ 100x10⁹/l.

The median follow-up in the FLAVIDA cohort was 9·3 months and 37·5 months in the FLA-IDA control cohort. The ORR after one cycle of FLAVIDA was 69% (n=9) (Supplementary Table S3) with a median CR duration of 7·3 months (range, 2·2-12·2 months) compared to 47% in the FLA-IDA control cohort (43% ORR in the cohort receiving a four-day schedule; 48% in the cohort with a five-day dosing scheme) and a median duration of CR/CRi of 11·3 months (range, 0·6-196·6 months). Molecular MRD negativity was achieved in 2 of 9 (22%) CR/CRi patients (Supplementary Tables S4 and S5). Nine FLAVIDA patients were consolidated with alloHCT and two received donor-lymphocyte infusions.

For FLAVIDA patients the estimated 6-month OS was 76% compared to 76% in the control cohort, the 6-month EFS was 52% compared to 46% in the control cohort and the 6-month RFS was 76% compared to 77% in the control cohort (Supplementary Figures S1 A–C).

In our study, short-term venetoclax in combination with intensive salvage chemotherapy was well tolerated and demonstrated a high CR/CRi rate of 69% including molecular remissions in 2 of 9 evaluable patients including patients with high-risk molecular features and adverse prognostic factors.

The combination of venetoclax and azacitidine reduce oxidative phosphorylation in LSCs likely through reduction of glutathione levels, which reduce electron transport chain-complex II activity and increase reactive oxygen species (Pollyea et al., 2018).

Importantly, resistance to venetoclax may be induced by upregulation of MCL1, which binds free BIM and reduces the apoptotic potential (Luedtke et al., 2017).

The major concern with adding novel drugs to existing regimens is toxicity. Therefore, safety and toxicity are important outcomes in this early investigational therapy. Here, common grade 3/4 AEs were haematologic and were managed with transfusions and anti-infective therapy. Interestingly, we found a very similar CR/CRi rate with our 7 day venetoclax schedule (69%) compared to the previously reported 21 and 14 day schedules of venetoclax (73%) (DiNardo et al., 2018b). With a median follow-up of 9·3 months in the FLAVIDA cohort we observed similar EFS, RFS or OS compared to FLAIDA treated patients. Longer follow up will show whether the improved CR rate including molecular remissions can improve long-term outcome in relapsed/refractory acute leukaemia patients.