Pharmacology-Guided venetoclax dose reduction with azacitidine in older or unfit acute myeloid leukemia: A North African experience from a Moroccan single-center cohort and exploratory hospital cost-effectiveness analysis

Abstract

Introduction

Azacitidine-venetoclax has improved outcomes in older or intensive chemotherapy-ineligible acute myeloid leukemia (AML) patients. However, real-world experience and pharmacoeconomic evidence remain scarce in resource-limited settings, particularly in North Africa, where drug pricing and supportive care constraints significantly shape clinical practice.

Methods

A retrospective single-center study was conducted at a tertiary military teaching hospital in Morocco (December 2019–December 2025). Fifteen older or unfit AML patients receiving azacitidine-based regimens were included. Venetoclax was dose-reduced to 100 mg/day under concomitant voriconazole prophylaxis based on CYP3A4 inhibition principles. Kaplan-Meier survival analysis was performed. An exploratory pharmacoeconomic analysis was conducted from a hospital perspective, incorporating drug acquisition costs, infection-related hospitalization expenses, and quality-adjusted life-years (QALYs) derived from the literature.

Results

Complete remission was achieved in 78% of first-line azacitidine-venetoclax recipients (n = 9) and 50% with azacitidine alone (n = 4). Infectious complications occurred in 73.3% of patients. Median overall survival was not reached; estimated survival probabilities were 76% at 12 months and 57% at 24 months. The incremental cost-effectiveness ratio (ICER) of azacitidine-venetoclax versus azacitidine alone was approximately 183,000 MAD per QALY gained (≈2× Morocco's GDP per capita) after incorporating infection-related costs. Venetoclax acquisition price was the primary cost-effectiveness driver on sensitivity analysis.

Conclusions

Pharmacology-guided azacitidine-venetoclax regimens are clinically feasible in resource-limited settings, with encouraging survival outcomes. Infection management, antifungal prophylaxis strategies, and targeted drug pricing negotiations are key determinants of cost-effectiveness. Prospective multicenter studies are warranted.

Keywords

Acute myeloid leukemia venetoclax azacitidine pharmacoeconomics cost-effectiveness Morocco older patients

Introduction

Acute myeloid leukemia (AML) is a malignant hematologic disorder characterized by clonal proliferation of myeloid blasts leading to bone marrow failure, with hemorrhagic and infectious complications.¹ In older patients, AML is frequently associated with comorbidities, reduced tolerance to intensive chemotherapy (IC), and adverse biological features, making therapeutic decision-making challenging.² At tertiary care centers in Morocco, management is adapted to patient characteristics and local resources, with a substantial role for lower-intensity approaches when IC is not feasible.³

Azacitidine plus venetoclax (Aza-Ven) has demonstrated survival benefits in pivotal trials,¹ but venetoclax exposure is markedly increased by strong CYP3A inhibitors, requiring substantial dose reductions to mitigate toxicity.^4,5 Strong CYP3A inhibitors such as posaconazole and voriconazole necessitate at least a 75% reduction of the venetoclax maintenance dose from 400 mg to 100 mg daily to avoid excessive drug exposure and associated toxicity.^4,5 In the Moroccan context, access to comprehensive cytogenetic and molecular profiling remains heterogeneous, and budgetary constraints at hospital level require careful consideration of both clinical effectiveness and economic feasibility when implementing innovative therapies.

We aimed to describe the real-world implementation of a pharmacology-guided, dose-reduced venetoclax regimen with azacitidine using voriconazole prophylaxis in older or intensive chemotherapy-ineligible AML patients treated at a Moroccan tertiary center between December 2019 and December 2025, to report clinical outcomes including overall survival, and to provide exploratory hospital-perspective cost-effectiveness estimates with explicit modeling of infection-related resource use.

Methods

Study design, setting, and period

This was a single-center retrospective descriptive study conducted in the Clinical Hematology Department of a tertiary military teaching hospital in Morocco, from December 2019 to December 2025.

Study population

We included AML patients diagnosed according to WHO 2022 criteria,^6,7 aged ≥65 years and/or deemed ineligible for intensive chemotherapy by the treating hematologist, with exploitable medical records. Ineligibility for IC was determined by physician assessment based on frailty, comorbidities (cardiac, renal, hepatic dysfunction), poor performance status (ECOG ≥2), or anticipated intolerance to anthracycline-based induction. We excluded patients who were eligible for IC, patients with acute leukemia other than AML, and incomplete or non-exploitable medical records.

Baseline assessment and collected variables

Baseline evaluation included history and physical examination, complete blood count, bone marrow aspirate, immunophenotyping, cytogenetic and molecular testing when available. Routine biochemistry, coagulation tests, blood group/phenotyping, viral serologies (HBV/HCV/HIV), imaging, and cardiac assessment (ECG, echocardiography with LVEF) were collected as per departmental practice. Because molecular testing was not systematically available, proportions for molecular variables are reported only among patients with documented results.

Treatment approach (departmental practice)

For patients eligible for IC, the reference induction regimen was 7 + 3 (cytarabine for 7 days plus an anthracycline for 3 days).⁸ In older or comorbid patients ineligible for IC, a lower-intensity strategy was favored using azacitidine alone or Aza-Ven, depending on drug access and clinical feasibility.

In the Aza-Ven regimen used at this center, azacitidine was administered at the standard dose of 75 mg/m²/day for 7 days. Venetoclax was escalated (100 mg day 1, 200 mg day 2, 300 mg day 3) and then administered at 100 mg/day with concomitant voriconazole 200 mg twice daily for 28 days in cycle 1, with subsequent cycles adapted (21 to 7 days of venetoclax based on AML profile, response and tolerability).^9,10 This practice is consistent with the pharmacologic principle that strong CYP3A inhibitors warrant at least a 75% venetoclax dose reduction after ramp-up; at steady dosing the 100 mg daily maintenance dose represents a 75% reduction from the commonly used 400 mg daily target dose in the absence of strong inhibitors.^4,5 The local rationale for using voriconazole was dual: primary antifungal prophylaxis and pharmacology-guided venetoclax dose reduction to reduce costs while maintaining expected therapeutic effectiveness based on pharmacokinetic principles.

Outcomes and definitions

Clinical complications were collected and categorized as infectious, metabolic, and digestive events as reported in the medical records. Clinical outcomes included complete remission (CR), early death (death during induction before response assessment), early relapse (<12 months from documented CR), failure (no response after first cycle), and status at last follow-up as recorded. Response assessment was performed after the first treatment cycle using bone marrow aspiration between days 28 and 42, in addition to peripheral blood count assessment. Given the retrospective nature of the study, outcome definitions reflect chart documentation; CR as reported may not strictly distinguish between morphologic CR and CR with incomplete count recovery (CRi). Future analyses should explicitly standardize response criteria (e.g., ELN 2022 definitions) and time-to-event endpoints.

Data handling and statistical analysis

Data were extracted using a standardized collection form and analyzed descriptively using Microsoft Excel, with means, medians, ranges, and frequencies as appropriate. All data were anonymized, and no nominative identifiers were included in the analysis dataset. Missing data were handled using a complete-case approach. Variables with missing information were not imputed and were excluded from specific analyses when unavailable. The number of patients included in each analysis is specified where applicable.

Overall survival was analyzed using the Kaplan-Meier method, with survival time defined as the interval between treatment initiation and death from any cause or last follow-up (censored observation). Median survival and 95% confidence intervals were estimated using the Kaplan-Meier estimator. Survival probabilities at 12 and 24 months were calculated from the survival function. Survival analyses were performed using R statistical software version 4.3.0 (R Foundation for Statistical Computing, Vienna, Austria).¹¹

Exploratory pharmacoeconomic analysis

We conducted an exploratory cost and cost-effectiveness analysis from the hospital perspective comparing azacitidine alone versus azacitidine plus venetoclax in older or intensive chemotherapy-ineligible AML patients.^12,13 Costs were expressed in Moroccan dirhams (MAD) using 2024–2025 values and were discounted at 3% annually. The time horizon of the base-case model corresponded to six treatment cycles, with extrapolation of survival beyond this period based on the observed distribution of clinical outcomes (remission, relapse, failure, death) in the cohort.

Confidentiality statement: Medication costs were calculated using publicly available Moroccan retail/list prices (prix public de vente, PPV) from the national AMMPS drug database, including for venetoclax, azacitidine, and voriconazole. Hospital procurement prices may differ from listed PPV values; however, since publicly verified PPV figures were available for all three agents, these were used consistently across all strategies to ensure transparency and reproducibility of the pharmacoeconomic model.^14,15

Drug costing assumptions: Azacitidine cycle cost was calculated for a mean body surface area of 1.7 m² and dosing of 75 mg/m²/day for 7 days, corresponding to a total required dose of 892.5 mg per cycle. In our department, cost estimation was based on a local reconstitution and administration practice using refrigerated preparation conditions within the validated stability window described in the product information, allowing residual volume from a partially used vial on one treatment day to be administered to the same patient on the following day within the accepted storage interval. Under this institutional practice, a full 7-day cycle could be delivered using 9 vials of 100 mg (900 mg total supplied), with only minimal residual wastage at cycle completion; this negligible remainder was not separately costed in the exploratory model. Venetoclax costs were calculated using the departmental regimen, with a total of 3100 mg during cycle 1 ramp-up (100 mg day 1 + 200 mg day 2 + 300 mg day 3 + 100 mg/day × 25 days) and subsequent cycles modeled at 21 days of 100 mg/day (2100 mg per cycle). Unit costs were derived from the publicly available Moroccan prix public de vente (PPV) listed in the national AMMPS drug database: venetoclax 100 mg film-coated tablets at 560 MAD per tablet (VENCLYXTO, box of 14 tablets; PPV 7841 MAD); azacitidine 100 mg powder for injection at 1919 MAD per vial; and voriconazole 200 mg tablets at 69 MAD per tablet (generic, box of 30 tablets; PPV 2076 MAD). All PPV data were retrieved from the AMMPS national drug database in 2025.¹⁵ Oral voriconazole prophylaxis was modeled as 200 mg twice daily for 28 days per cycle (11,200 mg per cycle), with unit costs derived from PPV for 200 mg tablets (boxes of 30 tablets) reported in the same national lists.

This assumption reflects local hospital practice and may not be generalizable to centers where same-patient next-day use of reconstituted azacitidine is not operationally feasible.

Infection-related costs: In addition to drug acquisition, we incorporated other direct medical costs, including treatment administration, hospitalizations for infectious complications, supportive care, and end-of-life care. Given that infectious events were the most frequent complications in our cohort (11/15 patients, 73.3%) and that all such episodes required hospitalization and intravenous antimicrobial therapy, infection-related costs were modeled explicitly for each strategy. For the base-case analysis, we assumed that patients receiving azacitidine alone experienced on average 0.7 infectious episodes per patient, each requiring a mean hospital stay of 6 days, whereas patients treated with azacitidine–venetoclax experienced on average 1.1 infectious episodes per patient, each with a mean hospital stay of 7 days; these assumptions were informed by the observed infection burden in the cohort (11/15 patients, 73.3%) and from the clinical impression of more intensive myelosuppression in the venetoclax-based group; they were not directly observed per-patient rates from a larger independent institutional dataset and should be interpreted accordingly.

Model structure and effectiveness: A simple decision-analytic model with two health states (progression-free and post-progression) was used. Time spent in each health state was approximated from chart-documented treatment duration, occurrence of relapse or failure, and survival status at last follow-up, acknowledging the uncertainty related to retrospective data and heterogeneous follow-up times. Literature-derived utility weights for AML health states were applied to each health state to estimate quality-adjusted life years (QALYs),¹⁶ and both costs and QALYs were discounted at 3% per year. One-way sensitivity analyses varied key medication cost inputs, notably venetoclax, by ±50% to explore the impact of drug price uncertainty on incremental cost-effectiveness ratios (ICERs).

Scenario analysis: full-dose versus pharmacology-guided venetoclax dosing with azole prophylaxis

To explore the economic implications of our pharmacology-guided dosing strategy, we conducted a scenario analysis comparing two alternative implementations of azacitidine–venetoclax over six cycles. Scenario A represented a hypothetical full-dose regimen without strong CYP3A inhibition, with venetoclax given at 400 mg/day for 28 days per cycle and no systematic voriconazole prophylaxis. Scenario B represented our local practice, combining azacitidine with venetoclax dose reduction under azole prophylaxis, modeled as the departmental regimen (3100 mg total in cycle 1 followed by 21 days of 100 mg/day in cycles 2–6) plus oral voriconazole 200 mg twice daily for 28 days per cycle.

Using publicly available PPV data from the AMMPS national database, venetoclax at full dose (Scenario A: 400 mg/day × 28 days × 6 cycles = 672 tablets of 100 mg) would cost approximately 376,368 MAD over six cycles, versus approximately 76,170 MAD with pharmacology-guided dose reduction (Scenario B: 136 tablets of 100 mg over six cycles). This represents a venetoclax acquisition saving of approximately 300,000 MAD, which substantially exceeds the additional voriconazole prophylaxis cost of approximately 23,251 MAD over the same period. The net drug acquisition saving of pharmacology-guided Scenario B versus full-dose Scenario A is therefore approximately 277,000 MAD over six cycles — a greater than 50% reduction in total venetoclax-related expenditure. These findings confirm that pharmacology-guided venetoclax dose reduction under azole prophylaxis represents a strongly cost-advantaged strategy in the Moroccan context, provided clinical and pharmacokinetic assumptions underlying the dose adjustment are met.

In a context where both venetoclax and azole antifungals are relatively costly, the net budget impact of choosing a reduced-dose venetoclax plus voriconazole strategy versus a full-dose venetoclax regimen without azole prophylaxis therefore depends critically on their relative prices, reinforcing the need for coordinated price negotiations for both agents.

Primary antifungal prophylaxis cost-effectiveness

In addition to the comparison between azacitidine alone and azacitidine–venetoclax, we explored the potential economic impact of systematic primary antifungal prophylaxis within the azacitidine–venetoclax strategy. In our departmental protocol, oral azole prophylaxis is already implemented (voriconazole 200 mg twice daily for 28 days per cycle), and was therefore included in the base-case costs. Using our assumptions, this prophylaxis represents an additional drug cost of approximately 23,000 MAD per patient over six cycles, whereas each infectious episode in the azacitidine–venetoclax group was costed at 18,900 MAD (7 days of hospitalization and intensive intravenous antibiotics). Under these parameters, primary prophylaxis would need to avert more than one severe infectious episode per patient to become cost-saving purely from a hospital-cost perspective, which is not compatible with the observed mean infection rate of 1.1 episodes per patient in our cohort. These exploratory estimates suggest that, under the current cost assumptions and the modeled infection rates, primary antifungal prophylaxis would need to avert more than one severe infectious episode per patient to become cost-saving from a hospital-budget perspective alone. However, this purely cost-driven inference must be interpreted with extreme caution given the very limited sample size (n = 15) and the fact that the only early infection-related death in this cohort occurred in the azacitidine–venetoclax group, highlighting the life-threatening consequences of invasive fungal disease in this population. Accordingly, the decision to implement primary antifungal prophylaxis should be guided primarily by clinical considerations — notably the prevention of invasive fungal infections and associated mortality — rather than by hospital cost estimates derived from a small retrospective series. Any net impact on QALYs and overall cost-effectiveness warrants evaluation in larger, prospective datasets.

Ethical approval

This retrospective analysis used routinely collected clinical data. All data were anonymized prior to analysis and no nominative identifiers were extracted. According to institutional policy for retrospective observational studies using anonymized data, formal ethics committee approval was not required. The study complied with the ethical principles of the Declaration of Helsinki and applicable local regulations.

Results

Baseline demographic and clinical characteristics

Fifteen older or unfit AML patients were managed between December 2019 and December 2025, including 9 men (60%) and 6 women (40%), with a mean age of 67 years (range 58–76). The most frequent clinical presentation was an anemic syndrome (100%), followed by tumor syndrome (7/15, 46.7%), infectious syndrome (5/15, 33.3%), and infiltrative manifestations (2/15, 13.3%). On complete blood count, anemia was present in all patients (15/15, 100%), with frequent thrombocytopenia (13/15, 86.7%), neutropenia (10/15, 66.7%), leukopenia (7/15, 46.7%), and circulating blasts (13/15, 86.7%) as reported in the files. Reported comorbidities and medical history included diabetes, smoking, heart disease, prior myelodysplastic syndrome, myeloproliferative neoplasms, solid cancer, and others with interindividual heterogeneity.

Hematologic and biological features

Bone marrow morphology classification (FAB) included AML2 (2/9, 22.2%), AML4 (3/9, 33.3%), and unclassifiable cases (4/9, 44.4%) as documented among patients with available FAB classification (n = 9). Immunophenotyping confirmed myeloid lineage in all tested patients (n = 9), with frequent expression of CD13 (8/9, 88.9%), CD33 (7/9, 77.8%), and MPO (9/9, 100%), and variable expression of HLA-DR (6/9, 66.7%), CD34 (5/9, 55.6%), and CD117 (4/9, 44.4%). Cytogenetics were available for all 15 patients and were normal in 7/15 (46.7%) of cases. Among the abnormal karyotypes, complex karyotype (≥3 abnormalities) was the most frequent (3/15, 20.0%), followed by other diverse abnormalities in 2/15 (13.3%). Specific recurrent abnormalities included t(8;21) in 1/15 (6.7%), inv(16) in 1/15 (6.7%), and isolated trisomy 8 (+8) in 1/15 (6.7%). Molecular testing was heterogeneous and incomplete; one IDH2 mutation (PCR) and one ASXL1 abnormality (NGS) were reported, while several tested patients had no detected abnormality. Systematic HIV/HBV/HCV serologies were negative in the series. Baseline characteristics are summarized in Table 1.

Table 1.

Baseline characteristics (n = 15).

Variable	Result
Total patients	n = 15
Sex	Men: 9 (60%); Women: 6 (40%); M/F ratio: 1.5
Age, years	Mean: 67 (range: 58–76)
Clinical presentation	Anemic syndrome: 15/15 (100%); Tumor: 7/15 (46.7%); Infectious: 5/15 (33.3%); Infiltrative: 2/15 (13.3%)
CBC abnormalities	Anemia: 15/15 (100%); Thrombocytopenia: 13/15 (86.7%); Neutropenia: 10/15 (66.7%); Leukopenia: 7/15 (46.7%); Circulating blasts: 13/15 (86.7%)
Karyotype (n = 15)	Normal: 7/15 (46.7%); Complex (≥3): 3/15 (20.0%); t(8;21): 1/15 (6.7%); inv(16): 1/15 (6.7%); +8: 1/15 (6.7%); Other: 2/15 (13.3%)
Immunophenotype (n = 9)	CD13: 8/9 (88.9%); CD33: 7/9 (77.8%); MPO: 9/9 (100%); HLA-DR: 6/9 (66.7%); CD34: 5/9 (55.6%); CD117: 4/9 (44.4%)
FAB (n = 9)	AML2: 2/9 (22.2%); AML4: 3/9 (33.3%); Unclassifiable: 4/9 (44.4%)
Molecular (selected)	IDH2 mutation: 1 case; ASXL1 abnormality: 1 case
Serologies	HIV/HBV/HCV: All negative

Treatment, complications, and outcomes

Initial treatment consisted of azacitidine alone in 4 patients (26.7%) and azacitidine-venetoclax in 9 patients (60%), while two patients (13.3%) had previously received intensive 7 + 3 induction elsewhere and were later managed at relapse with azacitidine-venetoclax.

First-line outcomes (n = 13): Among the 13 patients who received azacitidine-based therapy as first-line treatment (4 azacitidine alone; 9 azacitidine-venetoclax), complete remission was documented in 9 patients (69%), one early death (8%) occurred during induction due to septic shock, two early relapses (<12 months, 15%) were reported, and one failure (8%) was reported. By treatment group: (1) Azacitidine alone (n = 4): Complete remission in 2 patients (50%), early relapse in 1 patient (25%), failure in 1 patient (25%). (2) Aza-Ven first-line (n = 9): Complete remission in 7 patients (78%), early death in 1 patient (11%), early relapse in 1 patient (11%). (3) Aza-Ven at relapse after prior IC (n = 2): These patients are reported separately as salvage cases.

Complications: Infectious complications predominated (11/15, 73.3%), followed by metabolic (3/15, 20.0%) and digestive (2/15, 13.3%) complications. All patients who developed infectious complications required hospitalization and antimicrobial therapy; one case of invasive aspergillosis was documented in the early death case.

Of note, the only early infection-related death (septic shock with invasive aspergillosis) occurred in the azacitidine–venetoclax subgroup (1/9 first-line patients, 11%), underscoring the importance of antifungal prophylaxis and vigilant supportive care in this setting.

Survival analysis

Kaplan-Meier survival analysis was performed for the entire cohort (n = 15). The median overall survival was not reached, with more than 50% of patients alive beyond 40 months of follow-up. At 12 months, the estimated survival probability was 76.0%, and at 24 months, the estimated survival probability was 57.0%. The Kaplan-Meier curve is shown in Figure 1.

Figure 1.

Kaplan-Meier overall survival curve for the entire cohort (n = 15). The Kaplan-Meier curve shows overall survival for all 15 patients treated with azacitidine-based regimens (azacitidine alone or azacitidine-venetoclax) between December 2019 and December 2025. The median overall survival was not reached, with more than 50% of patients alive beyond 40 months of follow-up. The estimated survival probabilities were 76.0% at 12 months and 57.0% at 24 months. Censored observations (patients alive at last follow-up) are indicated by vertical tick marks. The shaded area represents the 95% confidence interval.

Second-line therapy: Four patients received second-line therapy after failure or early relapse, including salvage strategies with azacitidine-venetoclax (n = 2), one case proceeding to allogeneic hematopoietic stem cell transplantation after azacitidine-venetoclax, and one case treated with azacitidine-ruxolitinib with hematologic improvement without morphologic CR sustained over 24 months.

Status at last follow-up: At last follow-up, 11 patients (73.3%) were alive and 4 (26.7%) had died. The four deaths included one early death from septic shock with invasive aspergillosis during first-line induction in the azacitidine–venetoclax subgroup, two deaths after early relapse despite second-line therapy (one post-transplant), and one death after primary failure and subsequent treatment. Treatment outcomes are summarized in Table 2.

Table 2.

Treatment, complications, and outcomes.

Parameter	Result
Initial treatment	Azacitidine alone: 4 (26.7%); Aza-Ven: 9 (60%); Prior IC → Aza-Ven at relapse: 2 (13.3%)
First-line outcomes (n = 13)	CR: 9 (69%); Early death: 1 (8%); Early relapse: 2 (15%); Failure: 1 (8%)
Azacitidine alone (n = 4)	CR: 2 (50%); Early relapse: 1 (25%); Failure: 1 (25%)
Aza-Ven first-line (n = 9)	CR: 7 (78%); Early death: 1 (11%); Early relapse: 1 (11%)
Complications (n = 15)	Infectious: 11/15 (73.3%); Metabolic: 3/15 (20.0%); Digestive: 2/15 (13.3%)
Second-line therapy	4 patients: Aza-Ven salvage (n = 2), Allo-HSCT (n = 1), Aza-Ruxolitinib (n = 1)
Survival (Kaplan-Meier)	Median OS: Not reached (>40 months); 12-month survival: 76.0%; 24-month survival: 57.0%
Status at last follow-up	Alive: 11 (73.3%); Dead: 4 (26.7%)
Deaths (n = 4)	Early death (septic shock + aspergillosis): 1; Post-relapse deaths: 2; Post-failure death: 1

Exploratory pharmacoeconomic results

Using publicly available Moroccan prices as proxies and modeling six treatment cycles under the assumptions described above, the initial base-case analysis without explicit breakdown of infection-related costs estimated the mean total direct medical cost per patient at approximately 112,626 MAD for azacitidine alone and 212,206 MAD for azacitidine-venetoclax. Corresponding mean QALYs were 0.9 and 1.5, respectively, yielding an incremental cost of 99,580 MAD and an ICER of approximately 165,966 MAD per QALY gained for azacitidine-venetoclax versus azacitidine alone.

When infection-related costs (hospitalization days and intravenous antibiotic therapy for infectious complications) were explicitly incorporated, the mean infection-related cost per patient was estimated at 10,500 MAD for azacitidine alone and 20,790 MAD for azacitidine-venetoclax. Adding these costs to the previously modeled drug and other direct medical costs increased the total mean cost per patient to 123,126 MAD for azacitidine alone and 232,996 MAD for azacitidine-venetoclax. This resulted in a revised incremental cost of 109,870 MAD, while the incremental QALY gain remained at 0.6, yielding an updated ICER of approximately 183,000 MAD per QALY gained. Compared with the initial base-case ICER (approximately 165,966 MAD/QALY), this refined estimate suggests that explicitly accounting for infectious complications modestly worsens the cost-effectiveness profile of the azacitidine-venetoclax strategy, reflecting the higher burden of infection-related resource use in this group. Results are summarized in Table 3.

Table 3.

Exploratory cost-effectiveness results with explicit infection-related costs.

Strategy	Mean cost (MAD)	Mean QALYs	Incremental cost (MAD)	Incremental QALYs	ICER (MAD/QALY)
Azacitidine alone	123,126	0.9	—	—	—
Aza + venetoclax, pharmacology-guided dose reduction (100 mg/day) + voriconazole	232,837	1.5	109,711	0.6	∼183,000
Aza + venetoclax, full dose (400 mg/day), no voriconazole (hypothetical comparator)	509,784	1.5	386,658	0.6	∼644,000

Base-case using publicly available Moroccan PPV data (AMMPS national drug database, 2025); unit costs cited from official sources for all three agents.

Note: The full-dose venetoclax scenario (Scenario A, 400 mg/day) is presented as a hypothetical comparator reflecting standard venetoclax dosing in the absence of strong CYP3A inhibition; it was not implemented in this cohort. Infection-related costs are included in all strategies using identical assumptions. Unit costs sourced from AMMPS national drug database (April 2026): azacitidine 1919 MAD/vial; venetoclax 560 MAD/100 mg tablet (VENCLYXTO, box of 14); voriconazole 69 MAD/200 mg tablet (generic, box of 30). ICER: incremental cost-effectiveness ratio; QALYs: quality-adjusted life-years; MAD: Moroccan dirhams.

To further contextualize the economic benefit of pharmacology-guided dose reduction, a hypothetical full-dose venetoclax scenario (400 mg/day without azole prophylaxis) was added as a third comparator in Table 3, using the same PPV-based unit costs. At full dose, the estimated total cost per patient over six cycles would reach 509,784 MAD, yielding an ICER of approximately 644,000 MAD/QALY — more than seven times Morocco's GDP per capita — compared with ∼183,000 MAD/QALY for the pharmacology-guided dose-reduced strategy.

Sensitivity analysis: In one-way sensitivity analysis focusing on venetoclax acquisition cost, a 50% reduction in venetoclax price decreased the ICER to approximately 119,000 MAD/QALY, whereas a 50% increase in venetoclax cost raised the ICER to around 246,000 MAD/QALY. These ranges, together with the explicit inclusion of infection-related costs, underscore the central role of both drug pricing and toxicity-related resource use in shaping the cost-effectiveness profile of azacitidine-venetoclax combinations in this setting.

Discussion

This single-center real-world series describes predominantly marrow failure at presentation (including universal anemia and frequent thrombocytopenia) and a high burden of infectious morbidity among older or IC-ineligible AML patients. Lower-intensity approaches predominated (azacitidine alone in 26.7% and azacitidine-venetoclax in 60%), reflecting tolerance constraints, comorbidities, and drug availability in the Moroccan setting.

Pharmacology-Guided venetoclax dosing

Our departmental regimen relied on pharmacology-guided venetoclax dose reduction under voriconazole prophylaxis. This approach aligns with the general recommendation to reduce venetoclax by at least 75% when strong CYP3A inhibitors are necessary, to mitigate excessive exposure and toxicity risk.^4,5 The 100 mg daily maintenance dose used in our protocol represents this 75% reduction from the standard 400 mg dose used without inhibitors. This strategy served a dual purpose: antifungal prophylaxis in a high-risk population and cost reduction while maintaining expected therapeutic effectiveness based on pharmacokinetic principles.

The scenario analysis comparing full-dose venetoclax (400 mg/day without voriconazole, Scenario A) versus our dose-reduced regimen with azole prophylaxis (Scenario B) highlighted that the pharmacology-guided strategy substantially reduced venetoclax milligram use (approximately 4.9-fold reduction over six cycles) but added approximately 23,000 MAD in voriconazole acquisition costs. The net budget impact therefore depends critically on the relative prices of venetoclax and azole antifungals. Our findings reinforce that, to fully leverage the economic benefit of venetoclax dose reduction strategies under strong CYP3A inhibition, pricing policies must address both the targeted agent and the mandatory antifungal prophylaxis through coordinated price negotiations or patient access schemes.

Furthermore, our analysis of primary antifungal prophylaxis within the azacitidine–venetoclax strategy indicated that, under current cost assumptions, prophylaxis would need to prevent more than one severe infectious episode per patient to become cost-saving from a hospital perspective. Given the observed infection rate of 1.1 episodes per patient in our cohort, primary prophylaxis is unlikely to reduce total hospital costs and should instead be justified primarily on clinical grounds—namely, prevention of life-threatening fungal disease and associated mortality—with potential QALY gains requiring evaluation in larger prospective cohorts.

Clinical outcomes and survival

Despite encouraging remission reporting (69% CR rate among first-line patients, 78% in the azacitidine-venetoclax subgroup), infectious complications were frequent (73.3%), emphasizing that supportive care and anti-infective prevention pathways are central to safely delivering venetoclax-based regimens in this population. The single early death (8% of first-line patients) was associated with septic shock and invasive aspergillosis, highlighting the vulnerability of this patient population.

The observed CR rate of 78% with Aza-Ven in our first-line subgroup is numerically higher than the CR rate reported in the pivotal VIALE-A trial (36.7%), and appears consistent with some real-world series that have described higher response rates in routine practice.^1,17,18 However, this apparent difference must be interpreted with extreme caution given our very small sample size (n = 9 in the first-line azacitidine-venetoclax group), retrospective single-center design, potential differences in baseline risk profiles, and lack of standardized ELN response criteria and time-to-event endpoints. Similarly, the 50% CR rate observed with azacitidine alone in our limited sample (n = 4) is higher than the 17.9% reported in VIALE-A, but the small denominator in our cohort precludes any robust comparative conclusion and should be viewed as purely descriptive.

Kaplan-Meier survival analysis of the entire cohort showed that median overall survival was not reached, with more than 50% of patients alive beyond 40 months of follow-up, and estimated survival probabilities of 76.0% at 12 months and 57.0% at 24 months. While the VIALE-A trial reported a median OS of 14.7 months for the azacitidine-venetoclax arm versus 9.6 months for azacitidine alone,¹ direct comparison with our mixed cohort (including both azacitidine alone and azacitidine-venetoclax patients, as well as salvage cases) is not appropriate given the small sample size, heterogeneous treatment groups, and retrospective design. Nevertheless, the proportion of patients alive at last follow-up (73.3%) and the observed survival curve suggest that pharmacology-guided dose-reduced venetoclax regimens can achieve clinically meaningful survival outcomes in routine practice, although prospective studies with larger cohorts and standardized follow-up are needed to confirm these findings.

Cost-Effectiveness considerations

From an exploratory hospital-perspective pharmacoeconomic standpoint, azacitidine-venetoclax was associated with higher estimated direct medical costs and higher estimated QALYs compared with azacitidine alone. In the initial base-case analysis using drug and other direct medical costs but without explicit breakdown of infection-related resource use, the ICER was approximately 165,966 MAD per QALY gained. When hospital days and intravenous antibiotic therapy for infectious complications were explicitly incorporated, the total mean cost per patient increased to 123,126 MAD with azacitidine alone and 232,996 MAD with azacitidine-venetoclax, resulting in a revised incremental cost of 109,870 MAD and an updated ICER of around 183,000 MAD per QALY gained. This value corresponds to roughly 1.5–2.0 times Morocco's 2024 GDP per capita (approximately 90,000 MAD), and can be viewed against historically proposed willingness-to-pay benchmarks of 1–3× GDP per capita for middle-income countries, while recognizing that these thresholds are increasingly debated and should be interpreted as flexible reference points rather than rigid decision rules.¹⁹

The refined analysis highlights that, beyond drug acquisition prices, infection-related resource use is a non-trivial driver of overall costs, particularly in this frail elderly AML population receiving venetoclax-based regimens. Because medication costs were approximated using public PPV/list prices rather than confidential hospital acquisition costs, utilities were derived from international literature rather than local preference data, and infection-related resource use was modeled using average assumptions instead of detailed patient-level micro-costing, the resulting ICERs should be interpreted as exploratory and hypothesis-generating for Moroccan decision-makers rather than definitive reimbursement evidence.

The addition of the full-dose venetoclax comparator to Table 3 further underscores the economic rationale for pharmacology-guided dose reduction. At full dose (400 mg/day without azole prophylaxis), the estimated ICER would reach approximately 644,000 MAD/QALY — more than seven times Morocco's GDP per capita — rendering this approach economically untenable by any standard willingness-to-pay threshold applicable in Morocco. By contrast, the pharmacology-guided dose-reduced strategy with voriconazole achieves an ICER of approximately 183,000 MAD/QALY (−2× GDP per capita) at a net drug acquisition saving of approximately 277,000 MAD over six cycles. This demonstrates that the economic benefit of pharmacology-guided dosing is not marginal but transformative in a resource-limited setting, and reinforces the need for coordinated price negotiations for both venetoclax and azole antifungals.

Nevertheless, the sensitivity of the ICER to venetoclax price and to infection-related costs suggests that targeted price negotiations, patient access schemes, and optimized supportive care and antimicrobial stewardship could meaningfully improve the value-for-money profile of azacitidine-venetoclax in routine practice.

Limitations

This study has several important limitations. Its retrospective, single-center design and small sample size (n = 15) limit statistical power and preclude robust comparative inference between treatment strategies. Incomplete molecular profiling constrained formal ELN risk stratification, and the variable follow-up duration limits the precision of survival estimates. The exploratory pharmacoeconomic analysis relies on several assumptions: drug costs were approximated using public list prices rather than actual hospital acquisition costs; azacitidine cycle costing also reflected a local reconstitution workflow designed to minimize vial wastage, which may not be replicable across all centers; health-state utilities were derived from international literature, and infection-related resource use was modeled using average assumptions rather than patient-level micro-costing. Accordingly, the resulting ICERs should be interpreted as illustrative and hypothesis-generating rather than definitive. These findings should therefore be viewed strictly as descriptive, supporting the need for larger prospective multicenter studies with standardized response criteria, explicit time-to-event endpoints, and detailed local cost data.

Clinical and policy implications

These findings reinforce the need for several concrete actions at clinical and policy levels:

Optimized supportive care protocols: Systematic anti-infective prevention strategies (including careful consideration of breakthrough fungal infection risk), improved access to comprehensive cytogenetic and molecular diagnostics for risk stratification and measurable residual disease monitoring, and organizational optimization for febrile neutropenia pathways and standardized follow-up in older AML patients.¹⁷

Infection management and antimicrobial stewardship: From a hospital management perspective, the refined costing approach confirms that infection-related resource use is a non-negligible driver of total costs, particularly in patients receiving azacitidine-venetoclax. In this elderly, highly immunocompromised population, each additional febrile neutropenia episode can translate into several days of hospitalization and high-cost broad-spectrum antibiotics, thereby amplifying the budget impact of venetoclax-based regimens beyond drug acquisition alone. This underscores the importance of robust supportive care pathways, early detection and streamlined management of febrile neutropenia, and evidence-based antimicrobial stewardship to contain costs while maintaining patient safety.

Sustainable financing strategies: The exploratory ICER of approximately 183,000 MAD/QALY (roughly 1.5–2.0× GDP per capita) suggests that azacitidine-venetoclax regimens fall into an economic zone requiring policy dialogue. Targeted price negotiations for both venetoclax and concomitant azole antifungals, patient access schemes, risk-sharing agreements, and integration into national cancer control programs could improve affordability and equitable access.

Prospective data collection: To move beyond exploratory estimates, Moroccan centers should implement prospective registries capturing standardized response criteria (ELN 2022), time-to-event endpoints with adequate sample size, detailed resource use (including infection episodes, supportive care, and end-of-life costs), and patient-reported outcomes to generate robust local health-technology assessment evidence. The feasibility of pharmacology-guided dose reduction with azole antifungals offers a potential strategy to improve access in resource-limited settings, but requires careful attention to infection risk and hematologic toxicity monitoring.

Conclusions

In this Moroccan cohort of 15 older or intensive chemotherapy-ineligible AML patients, pharmacology-guided venetoclax dosing (100 mg daily with concomitant voriconazole) plus azacitidine was frequently used and associated with encouraging complete remission rates (78% in the first-line azacitidine-venetoclax subgroup) and prolonged survival (median not reached at 40 months; 76% alive at 12 months, 57% at 24 months), albeit at the cost of a high burden of infectious complications (73.3%). While the exploratory cost-effectiveness analysis, which explicitly incorporated infection-related resource use, suggested potential quality-adjusted survival gains at increased incremental cost (base-case ICER approximately 183,000 MAD/QALY, approximately 1.5–2.0× GDP per capita), these findings remain highly uncertain and should primarily be viewed as generating hypotheses for future research and policy dialogue.

The scenario analysis comparing dose-reduced venetoclax with voriconazole prophylaxis versus full-dose venetoclax without azole highlighted the critical importance of coordinated pricing strategies for both the targeted agent and mandatory concomitant prophylaxis. Furthermore, our analysis of primary antifungal prophylaxis indicated that, under current cost structures, prophylaxis is unlikely to be cost-saving from a hospital budget perspective and should instead be justified primarily on clinical grounds, with potential QALY gains requiring validation in larger prospective studies.

Overall, our results highlight both the clinical feasibility of implementing azacitidine-venetoclax regimens with pharmacology-guided venetoclax dose reduction in a North African tertiary care setting, and the parallel need for optimized supportive care pathways, strengthened anti-infective prevention, broader access to comprehensive diagnostic work-up, and context-appropriate strategies to secure sustainable financing of innovative AML therapies in Morocco. These exploratory findings, derived from a small single-center cohort, support the need for prospective, multicenter studies to confirm efficacy, refine treatment strategies, and further improve survival and quality of life in elderly AML patients in resource-limited settings.

Footnotes

Acknowledgements

The authors thank the medical and nursing staff of the Clinical Hematology Department for their contribution to patient care and data availability. No writing assistance or third-party editorial support was received in the preparation of this manuscript

ORCID iD

Baztami Youssef

Author contributions

Baztami Youssef: Conceptualization, Methodology, Pharmacoeconomic modeling, Formal analysis, Data curation, Writing – original draft, Writing – review & editing.

Jebrane Ilham: Conceptualization, Clinical supervision, Resources (patient cohort), Validation, Writing – review & editing.

Aznag Mohamed Amine: Formal analysis (survival analysis, Kaplan-Meier), Software (R statistical computing), Visualization, Writing – review & editing.

Chakara Rim: Investigation (clinical data collection), Data curation, Writing – review & editing.

Siham Ahchouch: Methodology (decision-analytic model, cost-effectiveness analysis), Validation, Writing – review & editing.

Abderrahim Raissi: Supervision, Project administration, Resources, Writing – review & editing.

All authors have read and approved the final version of the manuscript and consent to its submission to the Journal of Oncology Pharmacy Practice.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data sharing statement

The data that support the findings of this study are not publicly available due to institutional confidentiality policies and patient privacy regulations. Anonymized, aggregated data may be made available from the corresponding author upon reasonable request and subject to applicable ethical and institutional approvals.

References

DiNardo

Jonas

Pullarkat

, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med 2020; 383: 617–629.

Döhner

Wei

Appelbaum

, et al. Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood 2022; 140: 1345–1377.

Heuser

Ofran

Boissel

, et al. Acute myeloid leukaemia in adult patients: ESMO clinical practice guidelines for diagnosis, treatment and follow-up. Ann Oncol 2020; 31: 697–712.

Agarwal

DiNardo

Potluri

, et al. Management of venetoclax-posaconazole interaction in acute myeloid leukemia patients: evaluation of dose adjustments. Clin Ther 2017; 39: 359–367.

AbbVie Inc. Venclexta (venetoclax) [prescribing information]. North Chicago, IL: AbbVie Inc., 2020.

Arber

Orazi

Hasserjian

, et al. The 2016 revision to the world health organization classification of myeloid neoplasms and acute leukemia. Blood 2016; 127: 2391–2405.

Khoury

Solary

Abla

, et al. The 5th edition of the world health organization classification of haematolymphoid tumours: myeloid and histiocytic/dendritic neoplasms. Leukemia 2022; 36: 1703–1719.

Tallman

Wang

Altman

, et al. Acute myeloid leukemia, version 3.2019, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw 2019; 17: 721–749.

Wei

Döhner

Pocock

, et al. Oral azacitidine maintenance therapy for acute myeloid leukemia in first remission. N Engl J Med 2020; 383: 2526–2537.

10.

Pollyea

Pratz

Letai

, et al. Venetoclax with azacitidine or decitabine in patients with newly diagnosed acute myeloid leukemia: long term follow-up from a phase 1b study. Am J Hematol 2021; 96: 208–217.

11.

R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing, 2023. https://www.R-project.org (2023, accessed 28 February 2026).

12.

Drummond

Sculpher

Claxton

, et al. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford: Oxford University Press, 2015.

13.

Husereau

Drummond

Augustovski

, et al. Consolidated health economic evaluation reporting standards 2022 (CHEERS 2022) statement: updated reporting guidance for health economic evaluations. BMC Med 2022; 20: 23.

14.

Ministère de la Santé et de la Protection Sociale, Royaume du Maroc. Liste des médicaments remboursables et tarifs de référence. https://www.sante.gov.ma (2024, accessed 28 February 2026).

15.

Agence Marocaine des Médicaments et des Produits de Santé (AMMPS). Base de données nationale des médicaments — Prix publics de vente. Disponible sur : https://ammps.sante.gov.ma/basesdedonnes/listes-medicaments [accessed: avril 2026].

16.

Beusterien

Davies

Leach

, et al. Population preference values for treatment outcomes in chronic lymphocytic leukaemia: a cross-sectional utility study. Health Qual Life Outcomes 2010; 8: 50.

17.

Chua

Roberts

Reynolds

, et al. Chemotherapy and venetoclax in elderly acute myeloid leukemia trial (CAVEAT): a phase ib dose-escalation study of venetoclax combined with modified intensive chemotherapy. J Clin Oncol 2020; 38: 3506–3517.

18.

Stein

DiNardo

Fathi

, et al. Molecular remission and response patterns in patients with mutant-IDH2 acute myeloid leukemia treated with enasidenib. Blood 2019; 133: 676–687.

19.

Woods

Revill

Sculpher

, et al. Country-level cost-effectiveness thresholds: initial estimates and the need for further research. Value Health 2016; 19: 929–935.