Intensive chemotherapy is considered the treatment of choice for newly diagnosed patients with acute myeloid leukemia (AML). However, not all patients are eligible for this treatment because of physical limitations, comorbidities, or advanced age.1 Based on the results of two international, multicenter clinical trials (VIALE-A and VIALE-C),2,3 in 2020 the U.S. Food and Drug Administration (FDA)4 approved treatment with venetoclax (VEN) in combination with azacitidine (VEN/AZA) or low-dose cytarabine (VEN/LDAC) in over 75-year-old patients with newly diagnosed AML or in patients ineligible for intensive chemotherapy as induction therapy. In these trials, median overall survival (OS) was 14.7 months for the VEN/AZA combination versus 8.4 months for the VEN/LDAC combination.2,3

In the daily practice, hematologists often must choose between VEN/AZA and VEN/LDAC when treating AML patients who cannot receive intensive chemotherapy.

Based on the results of the VIALE-A2 and VIALE-C3 trials, the medical community has generally assumed that the VEN/AZA combination is superior to the VEN/LDAC combination in the treatment of these patients. However, assuming the superiority of one regimen over the other based solely on the results of these two clinical trials has several major limitations. The first and perhaps the most important limitation is that these trials aimed not to determine whether VEN/AZA is superior to VEN/LDAC. The second most important limitation is that the VIALE-C trial included a higher proportion of patients with secondary AML (41 %)3 compared to the VIALE-A trial (25 %);2 a factor often associated with worse OS curves.

Recent research has focused on evaluating the effect of VEN in various combinations (fludarabine, cladribine, and others) and with other indications such as myelodysplastic syndrome, relapsed or refractory AML or even first-line intensive chemotherapy,5-10 while the option of LDAC has lagged behind as a therapeutic option.

To our knowledge, there has been no formal comparison to categorically demonstrate the superiority of VEN/AZA, especially in patients with intermediate and favorable cytogenetic prognosis. This is particularly important in Latin American countries with low-income populations. It has already been pointed out that VEN/AZA is unlikely to be accessible to the entire population because of its current cost.11

For these reasons, we conducted a systematic review of real-world studies to describe OS, complete remission rate (CR), and composite complete remission rate (CCR) to assess response to the VEN/AZA or VEN/LDAC combinations in patients with newly diagnosed AML ineligible for intensive chemotherapy.

The search strategy, study selection, and data synthesis were based on the 2020 version of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.12 The protocol was submitted to PROSPERO (https://www.crd.york.ac.uk/prospero/) and was assigned the registration number CRD4202022367290. With the original search criteria, we did not find enough clinical trials with the VEN/LDAC combination to perform a meta-analysis. Therefore, we amended the protocol, updating it with the same registration number (CRD4202022367290). Therefore, the present study corresponds to a systematic review of the current literature, including only the analysis of retrospective studies with real-world experience to describe the main outcomes (OS, CR, CCR) associated with VEN/AZA and VEN/LDAC regimens.

Historically, for many years, before the introduction of hypomethylating agents, LDAC was the only medication for AML patients ineligible for chemotherapy; the benefits of LDAC were very limited in terms of OS. Subsequently, hypomethylating agents and VEN improved the OS curves.

In this systematic literature review of retrospective real-world studies, the lack of studies using the VEN/LDAC combination reflects the worldwide trend to favor the VEN/AZA combination over VEN/LDAC in the treatment of patients with AML who are not eligible for intensive therapy.15-25

In this review, we found only one study reporting the results of ten patients treated with VEN/LDAC, with a CCR rate of 90 % and a median OS not yet reached.25 This study had better results than the VIALE-C study,3 which had a CCR rate of 48 %, a CR rate of 18.8 %, and a median OS of 8.4 months. Perhaps these differences can be explained by the small sample size, short follow-up time, and the younger age of the patients in the study by Gómez-De León et al.25 We do not know the distribution of risk mutations in this study, or the distribution of treatment response based on cytogenetic risk, which may also have influenced the discrepancy in results.25 However, it is noteworthy that in this study, patients with VEN/LDAC had better CCR rates and OS than their VEN/AZA-treated counterparts (Table 1).

VIALE-C is the only Phase III clinical trial with a large number of patients providing information on the VEN/LDAC combination.3 However, it should be noted that VIALE-C included a patient population with unfavorable AML, particularly with regard to the incidence of secondary AML (41 %). In addition, prior exposure to hypomethylating agents was allowed, an exclusion criterion for the VIALE-A trial2; this is relevant because prior exposure to hypomethylating agents is associated with poor prognosis.28

As previously mentioned, the VIALE-A trial2 reported a median OS of 14.7 months, a CR rate of 36.7 %, and a CCR rate of 66.4 %. Compared with the results of the 11 retrospective VEN/AZA studies included in the present review, the median OS was 10.7 months, with a CR rate of 37.5 % and CCR of 63.3 %, thereby demonstrating similar results. In this context, it is worth noting that three studies17,18,25 reported higher CCR rates than VIALE-A.2 The first study was by Cherry et al.,17 which enrolled 143 patients on the VEN/AZA combination with a CCR rate of 71.3 %. The second study by Mirgh et al.18 included 16 patients and reported a CCR rate of 75 %, and the third study by Gomez-De Leon et al.25 enrolled 18 patients and reported a CCR rate of 72.2 %. The latter is noteworthy because the effect of clinical trials usually provides advantages in survival and response rates over real-world studies. A possible explanation for this could be differences between studies in the proportion of patients with favorable cytogenetic risk.

It is important to highlight that two of the reports with VEN/AZA included in the present review, one by Jensen et al.20 and the other by Garciaz et al.,22 reported OS curves of 9.1 and 9.4 months, respectively, which is similar to the OS curve of 8.4 months reported by the VIALE-C trial,3 but these reported OS curves are considerably shorter than the OS of 14.7 months reported in VIALE-A.2 This highlights the need to control for multiple variables in the study population to establish additional appropriate parameters and comparisons between these two regimens.

Qin Y et al.29 performed a meta-analysis on combinations of VEN with hypomethylating agents or LDAC as induction therapy for AML patients ineligible for intensive chemotherapy. They included only four clinical trials (three Phase Ib/II using VEN/hypomethylating agents and VIALE-C) and found pooled rates for CR of 40 % (95 % CI: 0.26–0.55) and CCR 64 % (95 % CI: 0.49–0.77) with a median OS of 11.7 (95 % CI: 10.15–14.18) months. It should be noted that the aim of this meta-analysis29 was to evaluate the efficacy and safety of VEN combinations and not to determine which VEN combination is more appropriate for a particular type of AML patient. Therefore, it was not possible to determine which combination (VEN/AZA or VEN/LDAC) showed greater benefit in isolation.

Even though, in this review, the most common cytogenetic risk category for the VEN/AZA combination was adverse (53.3 %) compared to VIALE-A and VIALE-C2,3 where the predominant category was intermediate, we highlight that some studies showed higher response rates compared to VIALE-A and even in the study by Cherry et al., a higher OS (16.1 months) was reported.17 We believe that this may be due to the proportion of included patients with favorable cytogenetic risk, which may have influenced the survival curves. We do not have cytogenetic risk data from real-world studies for the VEN/LDAC combination, so we can not analyze them.

It is noteworthy that in the VIALE-A trial,2 patients with favorable cytogenetic risk were excluded, resulting in reported distribution frequencies of 64 % and 36 % for the intermediate and adverse risk categories, respectively. The absence of favorable cytogenetic risk cases in the VIALE-A trial2 may explain why some real-world reports observed better response rates.

Regarding OS by cytogenetic risk, the VIALE-A trial reported a median OS for intermediate risk of 20.8 months, while for the adverse risk category the OS was 7.6 months.2 Other authors have pointed out that cytogenetic risk in conjunction with AML status (de novo versus secondary), age, and Eastern Cooperative Oncology Group (ECOG) scale functional status are independent baseline factors for OS.3

In the present work, we highlight that in the study by Cherry et al.,17 it was reported that the presence of RUNX1 mutations was associated with a better prognosis in patients receiving the VEN/AZA combination compared with intensive chemotherapy. Similarly, the VIALE-A and VIALE-C trials reported that mutations with better prognosis (IDH1/IDH2) showed favorable outcomes with VEN/AZA or VEN/LDAC.2,3 Other authors have reported high response rates, durable remissions, and OS of up to 24.5 months for the VEN/AZA combination in patients with IDH1/IDH2 mutations.30 On the other hand, patients included in both VIALE-A and VIALE-C, with poor prognostic mutations such as TP53 and FLT3, showed no beneficial effect with any combination of VEN.2,3

It should also be noted that a subgroup analysis in the VIALE-C trial3 showed that the patients who benefited most were those with a NMP1 mutation, with a CCR rate of 78 % and a median not reached OS. Interestingly, this mutation did not benefit from VEN/AZA as reported in the VIALE-A trial,2 although as we have mentioned, these are not formal direct comparisons. Based on these studies, we reason that the population that would benefit most from a formal comparison between AZA and LDAC in combination with VEN might be the favorable and intermediate-risk groups, including cases with the IDH1–2 and NPM1 mutations.

Hematologic toxicity, mainly neutropenia, was the most common adverse event in retrospective studies of the VEN/AZA or VEN/LDAC combinations, which is consistent with reports from the VIALE-A and VIALE-C studies.2,3 In VIALE-A, the main adverse events were nausea of any grade (44 %), febrile neutropenia (42 %), thrombocytopenia (45 %), and Grade 3 or higher neutropenia (42 %).2 In contrast, with VIALE-C, the main adverse events were nausea of any grade (42 %), febrile neutropenia (32 %), thrombocytopenia (45 %), and Grade 3 or higher (46 %) neutropenia.3 When comparing the safety profile, only the rate of febrile neutropenia is higher in the VEN/AZA group (42 %) than in the VEN/LDAC group (32 %). Currently, dose adjustment and the use of granulocyte colony-stimulating factor favor adequate management of neutropenia induced by VEN combinations.31 The study by Winters et al.15 was the only one that reported anemia as the main adverse event, followed by neutropenia. This may be due to the use of granulocyte colony growth factor in this patient cohort, which shortens the duration of febrile neutropenia induced by VEN.

Given the favorable response rates when VEN is combined with other drugs in patients with AML who are ineligible for intensive chemotherapy, VEN is currently considered the first-line treatment, regardless of their mutational status, except for the presence of a TP53 mutation.32

To our knowledge, there are no direct comparisons of the efficacy of VEN/AZA and VEN/LDAC in newly diagnosed AML patients. However, one expert opinion publication33 states that the only scenario in which VEN/LDAC could be chosen instead of VEN/AZA is in patients with prior exposure to hypomethylating agents.

In this era of new pharmacological combinations, drug triplets including VEN34-36 are a therapeutic modality currently being investigated for the treatment of AML to overcome resistance in patients with poor prognosis, especially those with TP53 mutations. In this scenario, cytarabine could be considered a viable combination option due to its known accessibility, low cost, efficacy, and toxicity profile.

The economic factor is another point to consider when deciding whether to combine VEN/AZA or VEN/LDAC. According to an economic analysis study, VEN/LDAC is less expensive than VEN/AZA ($75,833 versus $129,025 per year, respectively,37 and based on current costs, it has been indicated that the use of VEN/AZA is not affordable for all patients who need it, even in countries with developed economies.11 Accordingly, LDAC is significantly less expensive than AZA in Mexico ($300 versus $6500 per cycle).25 This may explain why in the publication by Gomez-De Leon et al.,25 VEN/LDAC was the most common combination in patients without insurance coverage.

From the results of this review, it appears that the VEN/LDAC combination remains a second option to VEN/AZA for patients ineligible for intensive chemotherapy, and possibly an alternative treatment for refractory/relapsed AML in unfit patients, specifically after receiving hypomethylating agents.

However, it is important to consider that this inferiority to VEN/AZA was assumed based on the results of non-comparable studies designed for other purposes. Therefore, these studies have a less-than-optimal methodology for establishing the superiority of one treatment over another, which is an aspect that could be relevant to some subgroups of patients, such as patients with intermediate or favorable cytogenetic prognosis.

On the other hand, VEN/LDAC is useful for patients who have already been treated with hypomethylating agents or when drug cost is a critical factor to consider.

Comparative clinical trials between VEN/LDAC and VEN/AZA should be performed to properly classify each combination in the treatment of AML patients who are ineligible for intensive chemotherapy.