Older Jehovah’s Witnesses with Acute Myeloid Leukemia: Hypomethylating Agents and Venetoclax as a Transfusion-Sparing Approach

Abstract

The treatment of acute myeloid leukemia (AML) in Jehovah’s Witness (JW) patients poses unique challenges due to their refusal of blood transfusions. This case series reports the outcomes of four older JW patients with AML treated with azacitidine (Aza) and venetoclax (Ven), including two with hyperleukocytosis and FLT3-ITD mutations. Three patients achieved initial remission; one of these patients subsequently received gilteritinib in combination with Ven and Aza, also achieving remission. All but one therapy cycle was administered in an outpatient setting, and hematologic recovery occurred in all patients without bleeding, ischemic events, or fungal infections. Three patients experienced disease relapse at 179, 301, and 392 days post-diagnosis, while one patient remains alive 706 days post-diagnosis. This report is among the first to demonstrate that Ven and Aza can safely achieve remissions, some of which were durable, in older JW patients with AML, even those with proliferative features like hyperleukocytosis and FLT3-ITD mutations. Our central finding is that Ven and Aza represent safe and effective transfusion-sparing therapeutic options in this population, with triplet therapy incorporating gilteritinib also proving feasible with dose modifications. These findings underscore the clinical relevance of such approaches, suggesting that transfusion refusal should not preclude treatment initiation, offering meaningful clinical outcomes and potentially enhancing quality of life in this population.

1. Introduction

Standard acute myeloid leukemia (AML) treatment requires blood transfusions, posing a significant challenge for Jehovah’s Witnesses (JWs), who decline blood products. Intensive induction chemotherapy, such as the “7 + 3” regimen, typically causes prolonged marrow aplasia, necessitating transfusion support to manage severe cytopenias [1]. Historical attempts using dose-reduced intensive chemotherapy without transfusions in JW patients resulted in markedly inferior outcomes and high mortality, often due to complications from severe anemia or bleeding. This has led to hesitation among hematologists to treat such patients [2]. Venetoclax (Ven) combined with hypomethylating agents, such as azacitidine (Aza), offers a promising alternative, established as a less-intensive standard of care for older or unfit patients [3]. However, this combination is also myelosuppressive and typically requires transfusion support during induction. This study reports the outcomes of four consecutive older JW patients with AML treated with this regimen at our institution, three of whom were specifically referred due to hesitation about treatment in view of the inability to provide transfusional support. The preliminary findings from this work were presented as an abstract at the 2024 Annual Meeting of the American Society of Hematology [4].

2. Materials and Methods

This retrospective case series included four consecutive adult JW patients with newly diagnosed AML treated at our NCI-designated comprehensive cancer center. Inclusion required documented refusal of all blood products and initiation of treatment with an Aza- and Ven-based regimen. This study adhered to institutional guidelines and the ethical principles of the Declaration of Helsinki. Patient data were fully de-identified prior to analysis, and patients’ documented refusal of blood transfusions was respected.

Clinical data, including demographics, baseline AML characteristics (cytogenetics, Next-Generation Sequencing [NGS]), and treatment details, were retrieved from electronic health records. Treatment involved Aza (typically 75 mg/m² subcutaneously or intravenously for 5–7 days/cycle) and Ven (100–400 mg daily for 14–28 days/cycle, adjusted for tolerance and drug interactions). Gilteritinib (starting at 80 mg daily) was used for FLT3-ITD mutations. Transient hydroxyurea was used once for hyperleukocytosis. Supportive care included darbepoetin alfa, intravenous iron, and antimicrobials, including antifungal prophylaxis (fluconazole or posaconazole) as needed. Treatment was primarily administered in an outpatient setting.

AML diagnosis followed World Health Organization criteria [5]. Response was assessed by bone marrow aspirate and biopsy (morphology, cytogenetics, NGS) after the first cycle of therapy (or after the second cycle in one case, at the treating physician’s discretion) and as clinically indicated according to European Leukemia Net recommendations [6]. Adverse events were monitored clinically and via laboratory results.

Descriptive statistics summarized patient characteristics and outcomes. Continuous variables (like age and counts) are reported with ranges or medians where appropriate. Survival and remission durations were calculated from the date of diagnosis. No formal comparative statistical analyses were performed due to the small sample size.

3. Case 1

A 62-year-old woman with a medical history of diabetes, hypertension, and coronary artery disease presented with a white blood cell count (WBC) of 5.8 × 10⁹/L, hemoglobin (Hb) of 8.2 g/dL, and platelet count (Plt) of 174 × 10⁹/L. A bone marrow biopsy confirmed AML with 26% blasts. Cytogenetic analysis was normal, while NGS identified FLT3-ITD, NPM1, CEBPA, and TET2 mutations. She began treatment with Aza for 7 days and Ven at 200 mg daily for 28 days, co-administered with fluconazole. Her nadir blood Hb and Plt counts were 9.4 g/dL and 108 × 10⁹/L, respectively. Post-cycle 1 bone marrow evaluation showed complete remission with persistent FLT3-ITD and NPM1 mutations. Fluconazole was discontinued from cycle 2 onward, allowing for an increase in Ven dosage to 400 mg for 21 days per cycle. By the end of cycle 4, progressive cytopenias were observed (WBC 1.3 × 10⁹/L, Plt 37 × 10⁹/L, Hb 8.2 g/dL). A repeat marrow biopsy demonstrated AML relapse with 42% blasts. NGS revealed the re-emergence of monoallelic CEBPA and TET2 mutations, in addition to persistent FLT3-ITD and NPM1. Gilteritinib (80 mg daily) was initiated as salvage therapy. Hematologic recovery followed, and a bone marrow biopsy performed 2 months later confirmed complete remission. However, after 6 months on gilteritinib, the WBC count rose to 24.7 × 10⁹/L, the Plt count dropped to 33 × 10⁹/L, and the Hb count was 8.9 g/dL. A repeat marrow assessment showed AML relapse with 92% blasts and NPM1 and TET2 mutations on NGS. She was referred to hospice, gilteritinib was discontinued, and she died 392 days after initial diagnosis.

4. Case 2

A 66-year-old woman presented with a WBC count of 122 × 10⁹/L, a Hb count of 7.8 g/dL, and a Plt count of 130 × 10⁹/L. She was diagnosed with AML with 68% blasts on a bone marrow biopsy. Cytogenetic analysis was normal, and NGS identified mutations in DNMT3A, FLT3-ITD, and NPM1. Induction therapy consisted of Aza for 7 days and Ven at 400 mg for 21 days, along with a short course of hydroxyurea. Her Hb reached a nadir of 5.6 g/dL, and her platelet count peaked at 207 × 10⁹/L. A post-cycle 1 marrow biopsy revealed persistent disease with 50% blasts. From cycle 2, Ven was reduced to 14 days, and gilteritinib (80 mg daily) was initiated. During this cycle, the nadir Hb count was 5.5 g/dL, and the Plt count was 154 × 10⁹/L. She required a brief hospitalization for febrile neutropenia. A subsequent marrow biopsy showed a blast reduction to 10%, and her Hb count later improved to 12.2 g/dL. Bone marrow evaluations after cycles 4 and 8 demonstrated complete remission with normalization of cytogenetics and persistence of DNMT3A on NGS. Cycles 2–8 were delayed by a median of 2 weeks (range 0–3 weeks) due to neutropenia. A prolonged delay of 10 weeks occurred before cycle 9, at which point both Aza and Ven were held, and she continued on single-agent gilteritinib. Aza was reintroduced 19 weeks later, followed by re-initiation of Ven (100 mg for 7 days) 7 weeks after that, which was well tolerated. However, per patient preference, she received cycle 12 with Aza and gilteritinib alone. Shortly thereafter, 21 months into gilteritinib therapy, she noted significant symptomatic neuropathy affecting her feet and occasionally her hands. Concern arose that this was related to gilteritinib, so it was discontinued; her symptoms resolved, and Ven was reintroduced with cycle 13 of Aza. She was alive 706 days post-diagnosis.

5. Case 3

A 78-year-old woman with a prior history of stage I breast cancer, previously treated with radiation therapy completed 5 years earlier, presented with a WBC count of 2.2 × 10⁹/L, a Hb count of 8.7 g/dL, and a Plt count of 95 × 10⁹/L. A bone marrow biopsy established a diagnosis of therapy-related AML with 26% blasts. Cytogenetic analysis revealed deletions in chromosomes 5q and 12p. NGS identified pathogenic mutations in DNMT3A, ASXL1, EZH2, PTPN11, and RUNX1. She was started on Aza for 7 days in combination with Ven (400 mg daily for 14 days). Her nadir Hb count dropped to 5.7 g/dL, and her Plt count dropped to 10 × 10⁹/L. After the first treatment cycle, another bone marrow biopsy demonstrated complete remission, with normalization of cytogenetics and clearance of molecular abnormalities on NGS. Her Hb rose to 13 g/dL following two cycles. However, after completing 8 cycles, she developed worsening pancytopenia. A repeat marrow evaluation confirmed relapsed AML with 30% blasts and newly identified complex cytogenetics. NGS at relapse detected two DNMT3A variants. Comfort-focused care was pursued. She experienced relapse 301 days after her initial diagnosis.

6. Case 4

A 73-year-old woman presented with a WBC count of 51.6 × 10⁹/L, a Hb count of 7.9 g/dL, and a Plt count of 86 × 10⁹/L and was diagnosed with AML with 80% marrow blasts. Cytogenetic analysis showed a complex karyotype, including a rearrangement involving the FUS/16p11 locus. Initial therapy consisted of Aza administered for 5 days and Ven (100 mg for 14 days), with concurrent posaconazole prophylaxis during the first cycle. Her blood counts improved, with a nadir Hb count of 8 g/dL and a Plt count of 279 × 10⁹/L. In cycle 2, Aza was given for 7 days and Ven to 400 mg for 14 days, this time without antifungal prophylaxis. After two cycles, she achieved complete remission with incomplete hematologic recovery (CRi), with persistent cytogenetic abnormalities still detected on bone marrow biopsy. Due to neutropenia, cycles 3 and 4 were delayed by 3 and 4 weeks, respectively. With subsequent worsening pancytopenia (Plt 19 × 10⁹/L, Hb 6 g/dL, WBC 0.9 × 10⁹/L), a bone marrow biopsy performed after cycle 4 confirmed relapsed or persistent AML with 35–45% blasts and the continued presence of cytogenetic abnormalities. Following relapse 179 days after diagnosis, the patient pursued comfort care and died shortly thereafter.

All patients received darbepoetin, some with intravenous iron. All but one therapy cycle was administered in an outpatient setting. All patients achieved hematological recovery without bleeding, ischemic events, or invasive fungal infections.

7. Discussion

The management of AML in patients who are JWs presents a unique clinical and ethical dilemma, stemming from their religiously motivated refusal of allogeneic blood product transfusions [7]. Standard intensive induction chemotherapy regimens like “7 + 3” (cytarabine plus an anthracycline) invariably induce profound and prolonged myelosuppression, necessitating red blood cell and platelet transfusions as critical supportive care to prevent life-threatening complications from severe anemia and thrombocytopenia. Historically, the inability to provide this essential support has led to significant hesitation among hematologists to offer potentially curative or life-extending treatments to JW patients with AML [2].

Historical attempts to use dose-reduced intensive chemotherapy without transfusion support in JW patients resulted in markedly inferior outcomes [8]. In their study of nine JW patients under 60 years old treated with reduced-dose chemotherapy cycles aimed at preventing sustained myelosuppression, only three (33%) achieved complete remission. Furthermore, relapse was common among responders, and no durable remissions were observed. Compared to a matched control group receiving full-dose chemotherapy with transfusion support, the JW cohort faced a staggering 12.1-fold higher hazard ratio for death. Critically, the primary driver of early non-leukemic mortality in the JW group was not uncontrolled infection but rather the direct consequences of severe cytopenias. Progressive anemia led to four out of five early deaths (with terminal hemoglobin levels reported between 1.8 g/dL and 3.3 g/dL), and uncontrollable bleeding with a platelet count below 10 × 10⁹/L accounted for one early death [8]. Other reports have corroborated the increased morbidity and mortality associated with treating acute leukemia in JW patients without transfusion support [7]. The failure of dose-reduced intensive chemotherapy thus stemmed not only from potentially compromised anti-leukemic efficacy but, more critically, from the inability to manage the resultant severe anemia and thrombocytopenia without transfusions, even at lower chemotherapy doses. This experience underscores that an optimal transfusion-sparing approach must fundamentally differ from traditional intensive chemotherapy, offering not just efficacy but also a more manageable hematologic toxicity profile.

In this context, the combination of the BCL-2 inhibitor Ven with hypomethylating agents such as Aza has emerged as a promising alternative. This regimen has become a standard of care for older (typically ≥75 years) or unfit patients with newly diagnosed AML based on the results of the phase 3 VIALE-A trial [3]. In the VIALE-A trial, Grade ≥3 thrombocytopenia and anemia occurred in 46% and 28% of patients, respectively [9]. Specific data on median nadir counts and transfusion requirements in the VIALE-A Ven + Aza arm were not readily available, though transfusion independence was a secondary endpoint, implying that transfusions were frequently required [9], potentially due to underlying leukemia and/or the treatment itself.

It is critical to recognize that the success of Ven + Aza in the VIALE-A trial was established in a population that had access to and utilized transfusions. Its application in JW patients, therefore, holds paramount importance and remains largely underexplored. This series details the experience of treating four consecutive older JW patients with AML (ages 62–78 years) using Ven + Aza-based regimens without any blood product support. Significantly, three of these patients were referred to our institution specifically because their refusal of transfusions precluded treatment initiation elsewhere.

In terms of efficacy, the regimen demonstrated meaningful anti-leukemic activity in this challenging population. All four patients eventually achieved complete remission (CR or CRi) following initial treatment. As anticipated with Ven + Aza therapy, patients experienced significant cytopenias during treatment. Nadir hemoglobin levels reached profoundly low values of 5.5 g/dL to 8.0 g/dL in Cases 2, 3, and 4. Nadir platelet counts were also critically low in some cases, reaching 10 × 10⁹/L in Case 3. Case 1 initially maintained higher nadirs (Hb 9.4 g/dL, Plt 108 × 10⁹/L) and developed severe cytopenias only upon relapse. All four patients tolerated initial therapy well enough to achieve hematologic recovery without experiencing major bleeding, ischemic events, or invasive fungal infections. Furthermore, nearly all therapy cycles were administered in an outpatient setting, demonstrating the potential for reduced hospitalization burden.

Prior to this report, the published experience with Ven + Aza combinations in JW AML patients was limited to only three cases. Bock and Pollyea described two younger patients (ages 33 and 48) with high-risk AML who achieved durable remissions without transfusion support using venetoclax combined with a 50% dose reduction in azacitidine (administered for 5 days instead of 7) [10]. More recently, Page and colleagues reported the successful treatment of a 76-year-old JW AML patient using standard-dose azacitidine (7 days) but with a significantly shortened duration of venetoclax (10 days per cycle), alongside supportive granulocyte colony-stimulating factor, erythropoietin, and iron [11].

Our current series contributes significantly to this nascent literature by expanding the experience to a cohort composed entirely of older adults (ages 62–78), aligning with the typical patient population for whom Ven + Aza is indicated, similar to the case by Page et al. [11] but distinct from the younger patients reported by Bock et al. [10]. Crucially, our series includes patients presenting with highly proliferative disease features (WBC 122 × 10⁹/L in Case 2, 51.6 × 10⁹/L in Case 4) and unfavorable molecular markers like FLT3-ITD mutations (Cases 1 and 2), characteristics associated with rapid disease progression and challenging cytopenias [12].

Regarding the treatment regimens employed, our patients generally received standard doses of azacitidine (mostly 7 days/cycle, though Case 4 started with 5 days) but required substantial and individualized adjustments to the venetoclax duration (ranging from 14 to 21 days initially, with further reductions needed in Case 2) and experienced cycle delays due to cytopenias. This contrasts with the proactive 50% reduction in azacitidine dose used by Bock et al. [10] and the consistent, markedly shortened (10-day) venetoclax schedule employed by Page et al. [11]. This empirical need for shorter durations aligns with emerging evidence suggesting that reduced venetoclax exposure (e.g., 7 or 14 days per cycle) may offer comparable efficacy with potentially less myelosuppression [13,14]. Such strategies are particularly relevant in JW patients, for whom minimizing therapy-related cytopenias is critical due to the inability to use transfusion support. The dosing variability across the limited number of reported JW cases underscores the current lack of a standardized approach and highlights the empirical, toxicity-driven modifications clinicians must make when managing these patients without the safety net of transfusions. Table 1 compares baseline characteristics and outcomes for JW AML patients treated with Ven + Aza regimens in this series, the VIALE-A trial, and previously reported cases in the literature.

Table 1. Comparison of patient characteristics and outcomes in Jehovah’s Witness acute myeloid leukemia patients treated with venetoclax + azacitidine regimens.

Characteristic	Current Series (n = 4)	VIALE-A (Ven + Aza Arm, n = 286)	Bock et al., 2020 [10] (n = 2)	Page et al., 2024 [11] (n = 1)
Median Age (Range)	70 (62–78)	76 (49–91)	41 (33, 48)	76
Sex (% Female)	100%	40%	50%	0%
Acute Myeloid Leukemia Type	75% de novo, 25% secondary	75% de novo, 25% secondary	50% de novo, 50% secondary	100% de novo
Baseline White Blood Cell (Median/Range, ×109/L)	28.7 (2.2–122)	Not reported	2.4, 3.1	1.2
Baseline Hemoglobin (Median/Range, g/dL)	8.1 (7.8–8.7)	Not reported	11.0, 12.7	10.5
Baseline Platelets (Median/Range, ×109/L)	113 (86–174)	Not reported	130, 177	77
Prognostic Risk per ELN 2022 [6] (Adverse %)	50%	36% *	100%	100%
% FLT3	50% FLT3-ITD	14% FLT3 ITD or TKD	0%	0%
Azacitidine Dose/Schedule	75 mg/m2 × 7 d (n = 3), × 5 d (n = 1)	75 mg/m2 × 7 d	37.5 mg/m2 × 5 d	75 mg/m2 × 7 d
Venetoclax Dose/Duration	100–400 mg × 14–28 d initially (reduced dose due to strong CYP3A4 interaction)	400 mg × 28 d (C1), variable later	400 mg × 28 d	50 mg × 10 d (reduced dose due to strong CYP3A4 interaction)
Use of FLT3 Inhibitor	Yes (n = 2, gilteritinib)	Not applicable	No	No
CR + CRi Rate (%)	100%	66.8%	100%	100%
Median Overall Survival (Months)	Variable (deaths at ~6, ~10, ~13 mo; ongoing at 23 mo)	14.7	Not reached at time of report	Not reached at time of report

* Per ELN 2017 classification [15]. Abbreviations: CR = complete remission; CRi = complete remission with incomplete hematologic recovery; ELN = European Leukemia Net.

Table 1. Comparison of patient characteristics and outcomes in Jehovah’s Witness acute myeloid leukemia patients treated with venetoclax + azacitidine regimens.

Characteristic	Current Series (n = 4)	VIALE-A (Ven + Aza Arm, n = 286)	Bock et al., 2020 [10] (n = 2)	Page et al., 2024 [11] (n = 1)
Median Age (Range)	70 (62–78)	76 (49–91)	41 (33, 48)	76
Sex (% Female)	100%	40%	50%	0%
Acute Myeloid Leukemia Type	75% de novo, 25% secondary	75% de novo, 25% secondary	50% de novo, 50% secondary	100% de novo
Baseline White Blood Cell (Median/Range, ×109/L)	28.7 (2.2–122)	Not reported	2.4, 3.1	1.2
Baseline Hemoglobin (Median/Range, g/dL)	8.1 (7.8–8.7)	Not reported	11.0, 12.7	10.5
Baseline Platelets (Median/Range, ×109/L)	113 (86–174)	Not reported	130, 177	77
Prognostic Risk per ELN 2022 [6] (Adverse %)	50%	36% *	100%	100%
% FLT3	50% FLT3-ITD	14% FLT3 ITD or TKD	0%	0%
Azacitidine Dose/Schedule	75 mg/m2 × 7 d (n = 3), × 5 d (n = 1)	75 mg/m2 × 7 d	37.5 mg/m2 × 5 d	75 mg/m2 × 7 d
Venetoclax Dose/Duration	100–400 mg × 14–28 d initially (reduced dose due to strong CYP3A4 interaction)	400 mg × 28 d (C1), variable later	400 mg × 28 d	50 mg × 10 d (reduced dose due to strong CYP3A4 interaction)
Use of FLT3 Inhibitor	Yes (n = 2, gilteritinib)	Not applicable	No	No
CR + CRi Rate (%)	100%	66.8%	100%	100%
Median Overall Survival (Months)	Variable (deaths at ~6, ~10, ~13 mo; ongoing at 23 mo)	14.7	Not reached at time of report	Not reached at time of report

* Per ELN 2017 classification [15]. Abbreviations: CR = complete remission; CRi = complete remission with incomplete hematologic recovery; ELN = European Leukemia Net.

Our report is the first to show that Ven and Aza can safely achieve meaningful remissions, with one patient remaining in remission beyond 700 days, in older JW patients, even those with proliferative features including hyperleukocytosis and FLT3-ITD mutations. Triplet therapy with gilteritinib also proved feasible with dose modifications. Relapse remains a major concern, as observed in three of the four patients in this series, highlighting the challenges in managing AML in older adults, particularly those with genetic features often associated with higher relapse rates. These features included FLT3-ITD mutations (present in Cases 1 and 2) and complex or unfavorable cytogenetics (observed in Case 3 and Case 4), which generally confer an intermediate to adverse prognosis with an increased risk of relapse [6]. Managing relapsed disease without transfusion support is exceedingly difficult, as more intensive salvage chemotherapy options are generally precluded. While targeted agents like gilteritinib may offer a salvage pathway for patients with specific mutations, their use adds complexity and potential toxicity, as seen in Case 2. Although our series is too small for definitive conclusions on factors affecting relapse, the outcomes underscore the persistent challenge of achieving long-term disease control in this high-risk JW AML population, despite initial responses to Ven + Aza-based therapy.

This study is inherently limited by its small sample size of four patients, its retrospective design, and its single-institution setting. The possibility of patient selection bias cannot be excluded, as individuals willing to pursue AML therapy despite refusing transfusions might possess different characteristics or baseline health compared to those who decline active treatment altogether. Robust evaluation of the efficacy and toxicity of Ven + Aza-based regimens in JW AML patients requires larger, ideally prospective, multicenter studies.

Despite these limitations, this case series, combined with the few preceding reports, provides accumulating evidence that Ven + Aza-based therapy represents a viable and potentially effective treatment strategy for selected older JW patients with AML. These findings support the assertion that a patient’s refusal of blood transfusions should not serve as an absolute contraindication to offering treatment with modern therapeutic strategies that yield meaningful clinical responses and the potential for extended remission durations in this population. Transfusion refusal should not preclude treatment initiation.