AML (Part 2): Treatment, APL, Relapse and Special Populations

Continued from: AML (Part 1): Biology, Diagnosis, Classification and Risk

Treatment: Induction (Fit Patients)

• "7+3": cytarabine 100 to 200 mg/m² CIVI x 7d + daunorubicin 60 to 90 mg/m² IV x 3d (or idarubicin 12 mg/m² x 3d). Day 14 marrow: if blasts >5% and marrow not hypocellular, consider re-induction (7+3 or HiDAC). CR 70 to 80% in <60 yo; 40 to 60% in fit >60 yo.
• Daunorubicin dose: 90 mg/m² superior to 45 mg/m² in pts <60 yo (ECOG-E1900, Fernandez NEJM 2009: median OS 23.7 vs 15.7 mo, HR 0.72; benefit in favorable and intermediate risk, and age <50; update also showed benefit with DNMT3A/NPM1 and age 50 to 60 with FLT3-ITD/NPM1). 60 mg/m² is current standard in younger fit pts and older pts who tolerate anthracycline (UK NCRI AML17 and HOVON showed 90 = 60). ALFA-9801: daunorubicin 80 = idarubicin 12; anthracycline intensification and rIL-2 maintenance did not improve outcomes. Meta-analysis (Teuffel BJH 2013): idarubicin reduced remission failure (RR 0.81) mainly when DNR:IDA dose ratio <5; high-dose daunorubicin reduced remission failure and overall mortality; high-dose DNR (90 x 3 or 50 x 5) and IDA (12 x 3) achieve 5-yr survival 40 to 50%.
• CPX-351 (Vyxeos): liposomal cytarabine + daunorubicin in a fixed 5:1 molar ratio (synergistic cytoreduction). FDA 2017 (adults) and Mar 2021 (pediatric ≥1 yo) for therapy-related AML or AML with myelodysplasia-related changes (AML-MRC, the FDA label term). Lancet JE et al, JCO 2018 (pts 60 to 75, approved regardless of age): median OS 9.56 vs 5.95 mo, HR 0.69, vs 7+3; prolonged cytopenias. 5-yr OS update Lancet JE, Lancet Haematol 2021.
• FLT3+ induction:
  • Midostaurin 50 mg PO BID days 8 to 21 of induction/consolidation (RATIFY/CALGB 10603, Stone NEJM 2017; FLT3-ITD or TKD; median OS 74.7 vs 25.6 mo, HR 0.78; CR 58.9% vs 53.5%).
  • Quizartinib 35.4 mg PO daily (induction days 8 to 21, consolidation days 6 to 19) for FLT3-ITD only; maintenance starts 26.5 mg daily, up to 53 mg if QTcF permits (QuANTUM-First, Lancet 2023; FDA Jul 2023).
  • Pearl: quizartinib is FLT3-ITD only, NOT TKD; midostaurin covers both.
• CD33+ (CBF and selected non-adverse-risk): add gemtuzumab ozogamicin (GO, Mylotarg) 3 mg/m² (up to one 4.5 mg vial) days 1, 4, 7 of induction (ALFA-0701, Castaigne Lancet 2012; long-term Lambert Haematologica 2019: EFS/OS benefit confined to favorable and intermediate cytogenetics, biggest in CBF AML; meta-analysis Hills Lancet Oncol 2014). Risk = VOD/SOS, especially within 3 mo of allo-HSCT.
• Cladribine: cladribine + 7+3 improved survival in pts >50, high initial WBC, or unfavorable cytogenetics. Clofarabine: benefit limited to intermediate-risk subsets, with added grade 3/4 toxicity and more deaths in remission.
• Hyperleukocytosis (WBC >50 to 100k): hydroxyurea, leukapheresis (avoid in APL, DIC), prompt induction. PRBC transfusion BEFORE WBC reduction can worsen leukostasis.
• TLS prophylaxis: allopurinol or rasburicase (esp. monocytic, high WBC, APL); aggressive IVF (urine alkalinization no longer recommended).
• Diffuse alveolar hemorrhage (DAH): can occur during induction (dyspnea, fever, cough; radiologic/bronchoscopic diagnosis), especially with monocytic (M4/M5) leukemic pulmonary infiltration killed by chemotherapy. Gingival infiltration (~5% at presentation) is most common in M4/M5.

Treatment: Unfit / Older Adults

• Venetoclax + azacitidine (or decitabine): VIALE-A (DiNardo NEJM 2020): median OS 14.7 vs 9.6 mo, HR 0.66 vs aza alone; CR/CRi 66% (vs 28%). Standard for unfit AML. Mechanism: venetoclax is a BH3-mimetic that targets antiapoptotic BCL-2, displacing BIM to activate BAX/BAK, causing mitochondrial outer-membrane permeabilization and caspase-mediated apoptosis. Benefit especially in IDH1, FLT3, NPM1 but not TP53. Resistance via MCL-1 and BCL-xL (myelomonocytic AML expresses MCL-1 and is more resistant). TLS risk on cycle 1 ramp-up (QD ramp 100 to 200 to 400 mg); moderate CYP3A inhibitors require ≥50% dose reduction, strong inhibitors (posaconazole, voriconazole) ≥75% (100 mg; 70 mg with posaconazole); day-21 marrow for cytopenias.
• Venetoclax + LDAC: VIALE-C, alternative for those not aza candidates; less robust OS benefit.
• Oral decitabine-cedazuridine (Inqovi) + venetoclax (FDA May 2026): all-oral option for newly dx AML ≥75 yo or unfit for intensive induction (ASTX727-07, CR 41.6%).
• Ivosidenib + aza (IDH1+): AGILE (Montesinos NEJM 2022): median OS 24 vs 7.9 mo. Approved May 2022 for IDH1-mut newly dx unfit AML.
• Glasdegib + LDAC: hedgehog (Smoothened) inhibitor; OS benefit (BRIGHT AML 1003, pts ≥75, cardiac disease, ECOG 2, or Cr >1.3) but limited use given ven-aza dominance.
• Azacitidine monotherapy: for those who cannot tolerate a venetoclax combination.
• Oral azacitidine (CC-486, Onureg): maintenance for AML in CR/CRi after induction not proceeding to transplant. QUAZAR AML-001 (Wei NEJM 2020): median OS 24.7 vs 14.8 mo, HR 0.69. 300 mg PO days 1 to 14 of a 28-day cycle. FDA Sept 2020.

Targeted Therapies: Comprehensive Reference

• FLT3 inhibitors (Type 1 = midostaurin, gilteritinib, active on both TKD and ITD; Type 2 = sorafenib, quizartinib, ITD only):
  • Midostaurin (Rydapt, Apr 2017): newly dx FLT3-ITD/TKD AML w/ 7+3 + consolidation (RATIFY).
  • Gilteritinib (Xospata, Nov 2018): R/R FLT3-mut AML monotherapy (ADMIRAL, Perl NEJM 2019: OS 9.3 vs 5.6 mo vs salvage chemo, HR 0.64; EFS 2.8 vs 0.7 mo); also studied as post-allo maintenance (MORPHO, Levis JCO 2024: RFS not significantly improved overall (HR 0.68, P=0.052); RFS benefit only in MRD+ pts (HR 0.52), no OS benefit shown). AEs: anemia, thrombocytopenia, hypokalemia, constipation.
  • Quizartinib (Vanflyta, Jul 2023): newly dx FLT3-ITD AML w/ 7+3 + consolidation + 3-yr maintenance (QuANTUM-First); REMS for QTc.
  • Sorafenib: added to 7+3 improved EFS in a phase 2 trial (off-label; not FDA-approved for AML), with increased toxicity.
• IDH inhibitors (mutant IDH produces the oncometabolite 2-hydroxyglutarate, which inhibits alpha-KG-dependent dioxygenases and impairs differentiation; IDH1 in 6 to 10%, IDH2 in 10 to 13%):
  • Ivosidenib (Tibsovo, IDH1): R/R IDH1-mut AML (Jul 2018, ORR ~42%); newly dx unfit IDH1-mut AML monotherapy (May 2019, frail ORR ~55%); newly dx IDH1-mut AML + azacitidine (May 2022, AGILE, CR ~45%).
  • Olutasidenib (Rezlidhia, Dec 2022): R/R IDH1-mut AML monotherapy (Cortes Blood 2024, CR+CRh 35%).
  • Enasidenib (Idhifa, Aug 2017, IDH2): R/R IDH2-mut AML monotherapy (ORR ~40%, CR 19%).
  • Differentiation syndrome in 11 to 14%; treat immediately with steroids. Can occur up to 6 months out; start corticosteroids immediately and follow label interruption rules: withhold olutasidenib when DS is suspected (boxed warning); interrupt ivosidenib/enasidenib for severe DS persisting despite steroids.
• Menin inhibitors (disrupt menin-KMT2A interaction to reverse aberrant HOXA9/MEIS1 expression in KMT2A-r and NPM1-mut leukemias; class effects: differentiation syndrome, QTc prolongation; resistance via MEN1 binding-site mutations M327I, G331R/D):
  • Revumenib (Revuforj, Nov 15, 2024): R/R acute leukemia w/ KMT2A translocation, ages ≥1 yr (AUGMENT-101, Issa JCO 2024, CR+CRh 22.8% in published efficacy cohort (21% in FDA label)); label expanded 2025 to R/R NPM1-mut AML (CR+CRh 23%).
  • Ziftomenib (Komzifti, Nov 13, 2025): R/R NPM1-mut AML in adults (KOMET-001, Wang JCO 2025, CR+CRh 22% (21.4% in FDA label); 600 mg PO daily).
• BCL2 inhibitor: Venetoclax (Venclexta, Oct 2020 full approval): newly dx unfit AML + aza/dec/LDAC (VIALE-A, VIALE-C).
• CD33 ADC: Gemtuzumab ozogamicin (Mylotarg, re-approved Sept 2017): newly dx CD33+ AML; R/R CD33+ AML; pediatric newly dx CD33+ AML ≥1 mo (Jun 2020); R/R CD33+ AML ≥2 y.
• Hedgehog inhibitor: Glasdegib (Daurismo, Nov 2018): newly dx AML ≥75 or unfit + LDAC (BRIGHT 1003); largely supplanted by ven-aza.
• Oral HMA maintenance: CC-486 / oral azacitidine (Onureg, Sept 2020): post-induction maintenance in AML CR/CRi not proceeding to allo-HSCT (QUAZAR AML-001).
• Liposomal 7+3: CPX-351 (Vyxeos, Aug 2017 adults; Mar 2021 pediatric): t-AML or AML-MR.
• Differentiation syndrome (DS): complication of FLT3, IDH, and menin inhibitors. Sx: dyspnea, pulmonary infiltrates, fever, edema, weight gain, pleural/pericardial effusions, hypotension, AKI. Treat: dexamethasone 10 mg IV q12h; hold inhibitor for severe DS; supportive care. Can be fatal if untreated.
• FLT3 inhibitor pearls: QTc monitoring for all (esp. quizartinib, REMS). Strong CYP3A/azole interactions: reduce dose for quizartinib; monitor toxicity or consider alternatives for midostaurin and gilteritinib. Resistance via FLT3 F691L (gatekeeper) and D835 (activation-loop) mutations; D835 drives type II inhibitor resistance.

Consolidation & Allo-HSCT

• Timing: follow-up marrow 21 to 28 days after induction. Hypoplasia = cellularity <20% with residual blasts <5%; if seen, repeat marrow in 7 to 14 days. Even excellent responders relapse without consolidation.
• HiDAC consolidation: cytarabine 3 g/m² q12h days 1, 3, 5 x 3 to 4 cycles for favorable risk (CALGB 8525, Mayer 1994: benefit in CBF AML). 1.5 g/m² for older pts.
• Cerebellar toxicity from HiDAC: dose/schedule-related, worse with renal dysfunction and older age (no safe threshold; occurs at standard 3 g/m²); monitor pre/post, hold for nystagmus, dysmetria, dysarthria; do not rechallenge. Older age and renal dysfunction increase risk (neurotoxicity 55% vs 7% with renal failure); consider consolidative autologous transplant if HiDAC cannot continue.
• Allo-HSCT in CR1 by ELN risk:
  • Favorable: generally no transplant in CR1 if MRD clears (HiDAC/chemo); persistent or rising MRD can justify transplant.
  • Intermediate: consider transplant (data favors benefit).
  • Adverse: YES transplant in CR1.
• Transplant evidence: EORTC/GIMEMA (Suciu Blood 2003): DFS favored allo (43.4% vs 18.4%) for adverse cytogenetics. Meta-analysis: allo improved 5-yr OS from 45% to 52% (intermediate risk) and 20% to 31% (poor risk); no benefit in good-risk cytogenetics.
• TP53-mutated AML: very poor prognosis; transplant outcomes poor; consider clinical trials.
• Conditioning: MAC (Bu/Cy or TBI/Cy) for younger fit; RIC (Flu/Bu, Flu/Mel) for older/comorbid. BMT-CTN 0901 (Scott JCO 2017): RIC increased relapse and reduced RFS/OS vs MAC; for young, high-risk, or MRD+ patients, prefer MAC. IV busulfan + cyclophosphamide gives less non-relapse mortality and better leukemia-free survival (Copelan Blood 2013).
• Donor selection priority: MSD > MUD (10/10) > haplo w/ PTCy approx MUD > UCB. PTCy now standard in haplo and increasingly in MUD (BMT-CTN 1703: PTCy/Tac/MMF superior to Tac/MTX for GVHD-free relapse-free survival).
• MRD-guided decisions: MRD+ at end of induction to consider transplant; NPM1 MRD by qPCR (UK NCRI AML17/19), CBF AML by RUNX1::RUNX1T1 or CBFB::MYH11 transcripts, flow MRD in others. MRD+ pre-transplant markedly increases relapse.

Maintenance Therapy

• Oral azacitidine (Onureg): for fit pts in CR1/CRi who do not proceed to transplant. 300 mg PO days 1 to 14 of a 28-day cycle. QUAZAR AML-001: OS 24.7 vs 14.8 mo.
• Sorafenib maintenance: SORMAIN (Burchert JCO 2020), RFS benefit post-transplant in FLT3-ITD AML (primary endpoint, HR 0.39; OS trend only, HR 0.52, not significant) (off-label; not FDA-approved for this indication).
• Quizartinib maintenance: part of QuANTUM-First post-consolidation in FLT3-ITD; FDA-approved in this setting.
• Midostaurin maintenance: included in RATIFY for 12 mo post-consolidation but no clear OS benefit from the maintenance phase alone.

Acute Promyelocytic Leukemia (APL): t(15;17) PML::RARA

• Hallmark: t(15;17)(q24.1;q21.2), fusing RARA (17q) to PML (15q) to make PML::RARA, which silences promyelocyte-differentiation genes causing promyelocytic arrest. Faggot cells, hypergranular promyelocytes (M3) or microgranular variant (M3v, high WBC); Auer rods; MPO strongly +; CD34-, HLA-DR-, CD117+, CD13+, CD33+.
• Coagulopathy: DIC + hyperfibrinolysis to bleeding (intracranial hemorrhage = early death). Replace aggressively: cryoprecipitate (fibrinogen >150), platelets >30 to 50k, FFP for prolonged PT/PTT. Avoid heparin. Avoid routine antifibrinolytics (e.g., tranexamic acid); rely on ATRA and aggressive blood-product support.
• CRITICAL: start ATRA at suspicion: do NOT wait for cytogenetic confirmation (continue empiric ATRA while RT-PCR is pending and discontinue only when molecular testing excludes APL, not on a negative FISH alone (cryptic PML::RARA is FISH-negative)). ATRA 45 mg/m²/d divided BID.
• Risk stratification (Sanz): low = WBC ≤10k AND platelets >40k; intermediate = WBC ≤10k AND platelets ≤40k; high = WBC >10k.
• Low/intermediate risk induction: ATRA + ATO (APL0406/Lo-Coco NEJM 2013: 100% CR, 2-yr EFS 97% vs 86%, better OS, less hematologic toxicity vs ATRA + chemo). New standard; maintenance often omitted.
• High-risk induction: ATRA + ATO + idarubicin or gemtuzumab ozogamicin (for cytoreduction), or ATRA + anthracycline-based chemo +/- cytarabine. Example regimens: ATRA + daunorubicin + cytarabine; ATRA + idarubicin (days 2,4,6,8); ATRA + ATO + GO.
• Differentiation syndrome (APL DS, formerly retinoic acid syndrome): up to 25% of ATRA/ATO-treated patients; onset median ~12 d (range day 0 to 46), bimodal with peaks in weeks 1 and 3 of induction. Fever, rising WBC, dyspnea, hypoxia, pleural/pericardial effusions, weight gain, hypotension, AKI; CXR diffuse infiltrates. Treat: dexamethasone 10 mg IV q12h at first suspicion, continued until resolution (minimum 3 days; NCCN 3 to 5 days), then taper over ~2 weeks; hold ATRA/ATO only if severe. Prophylactic corticosteroids (prednisone 0.5 mg/kg or dexamethasone) for high-risk disease, WBC >10k (NCCN; ELN: consider at WBC >5 to 10k or rising WBC; benefit uncertain).
• ATRA toxicity: pseudotumor cerebri (increased CSF production; treat with acetazolamide, topiramate, therapeutic LP, diuretics, rarely shunt; hold ATRA and may rechallenge).
• ATO toxicity: QTc prolongation (blocks potassium currents; correct K, Mg; ECG monitoring), hepatotoxicity, neuropathy. ATO binds the PML moiety of PML::RARA, causing degradation, differentiation, and apoptosis.
• MRD monitoring: PML::RARA RT-qPCR q3mo for 2 yr post-completion. Molecular relapse (confirm by repeat marrow in 2 weeks) to preemptive ATO +/- ATRA to reinduction.
• Relapsed APL: ATO if not previously exposed (or if >6 mo since ATO). If CR2 and PCR-negative to autologous HSCT; if PCR-positive to allogeneic HSCT; ATO x6 cycles if not a transplant candidate. Single-agent GO is an option.

Refractory / Relapsed AML

• Salvage chemo: MEC (mitoxantrone/etoposide/cytarabine), FLAG-IDA, HiDAC. 30 to 40% CR; bridge to transplant. Late relapse in pts <60 may respond to repeating the initial successful regimen.
• Targeted salvage:
  • FLT3-mut: gilteritinib (ADMIRAL, Perl NEJM 2019, OS 9.3 vs 5.6 mo, HR 0.64).
  • IDH1-mut: ivosidenib (Tibsovo) or olutasidenib (Rezlidhia, CR+CRh 35%).
  • IDH2-mut: enasidenib (Idhifa, Stein Blood 2017, ORR 40%, CR 19%).
  • KMT2A-translocated acute leukemia (AML or ALL): revumenib (AUGMENT-101, CR+CRh 21%; FDA Nov 2024, age ≥1 yr).
  • NPM1-mut: revumenib (label expansion 2025, CR+CRh 23%) or ziftomenib (KOMET-001, CR+CRh 21.4%; FDA Nov 2025, adults).
• Allo-HSCT: the only curative option for most R/R AML; consider trials.
• Special consideration: test for actionable mutations BEFORE starting non-targeted salvage; molecular profiling at relapse is essential.
• Post-transplant relapse: DLI +/- azacitidine; re-induction; consider a second transplant in select cases.

Supportive Care Pearls

• Transfusion thresholds: platelets <10k (or <20k w/ fever/bleeding); Hb <7 to 8. Use gamma-irradiated products to prevent transfusion-associated GVHD in immunocompromised patients.
• Antifungal prophylaxis: posaconazole during induction (Cornely NEJM 2007, superior to fluconazole/itraconazole for mold prevention). Adjust when starting venetoclax (CYP3A4: voriconazole/posaconazole require major ven dose reduction: 100 mg with voriconazole, 70 mg with posaconazole).
• Antibacterial: levofloxacin during neutropenia (institutional).
• PJP/HSV prophylaxis: PJP prophylaxis for T-cell-suppressive therapy, prolonged steroids, or transplant (not neutropenia alone); HSV prophylaxis for HSV-seropositive patients during induction or allo-HSCT.
• CMV monitoring: in transplant recipients; letermovir prophylaxis (Marty NEJM 2017) for CMV+ allo recipients, first 100 days routinely; extend to 200 days in patients at risk for late CMV.

Special Populations

• Older adults (≥75 or comorbid): ven + aza standard; assess fitness (geriatric assessment, ECOG, comorbidity index).
• Pediatric AML: different protocols (AAML1031); DS-associated AML (M7) has excellent outcomes with reduced-intensity chemo.
• AML in pregnancy: first trimester carries the highest fetal death/spontaneous abortion rate (~37.5% vs 9.7% second, 0% third; Chang & Patel 2015); CR rates 100%/81%/67% across trimesters. First trimester: treat without delay; discuss elective termination given high fetal chemo risk (cytarabine is an antimetabolite), then prompt standard induction. Do not defer induction to a later trimester; second/third: standard 7+3 induction generally acceptable (daunorubicin preferred over idarubicin).
• CNS AML: LP at diagnosis if CNS symptoms; if positive, IT chemo (MTX, AraC) 2x/wk until clear then weekly x4 to 6 wk. Screening LP before first consolidation for monocytic differentiation, MPAL, WBC >40,000, extramedullary disease, or high-risk APL. Do not give IT chemo or HiDAC concurrently with cranial radiation. Routine CNS prophylaxis is NOT standard for AML (unlike ALL).
• Myeloid sarcoma (chloroma/granulocytic sarcoma): extramedullary AML, most commonly skin (leukemia cutis); treat as AML (intensive chemo if fit, HMA + venetoclax or targeted therapy if not); treat all myeloid sarcoma (including isolated) with systemic AML-directed therapy; use radiation/surgery for local control or residual disease, not as sole curative treatment. ~52% carry a targetable mutation (NPM1 > IDH1/2 > FLT3 > MLL).
• Down syndrome myeloid proliferations: transient abnormal myelopoiesis (DS-TAM) resolves spontaneously over the first months of life (observe; treat severe cases with low-dose cytarabine); GATA1 mutations occur in all cases; up to 20 to 25% progress to ML-DS.
• BPDCN: CD123+ neoplasm; tagraxofusp (SL-401, IL-3-diphtheria toxin fusion; preferred) with risk of capillary leak syndrome (replace albumin if <3.5); alternatives are ALL-type (hyper-CVAD) or AML-type induction; bridge to allo-HSCT.

Prognostic Pearls: Quick Reference

• Best prognosis: APL (>90% cure w/ ATRA+ATO); CBF AML w/ HiDAC consolidation; mutated NPM1 without FLT3-ITD (and without adverse-risk cytogenetics).
• Worst prognosis: TP53-mutated (median OS <12 mo); complex/monosomal karyotype; t-AML, -7/-5, secondary AML from MDS.
• Improved by allo-HSCT in CR1: adverse-risk and most intermediate-risk AML.
• NPM1 + DNMT3A + FLT3-ITD: "triple mutant," worse than NPM1 alone.
• FLT3-ITD high allelic ratio: historically adverse; ELN 2022 dropped it from risk stratification given FLT3 inhibitors.

AML high yield

• Induction: 7+3 gives 70 to 80% CR in <60 yo and 40 to 60% in fit >60 yo; young/fit patients get daunorubicin 60 to 90 mg/m² (E1900: 90 > 45; 90 = 60 in AML17/HOVON).
• CPX-351 for pts 60 to 75 with t-AML, secondary AML, or AML-MR.
• Midostaurin days 8 to 21 for FLT3 mutation (RATIFY).
• Gemtuzumab days 1,4,7 (or day 1) in favorable/intermediate risk, especially inv(16)/t(8;21).
• HMA + venetoclax: VIALE-A, venetoclax + azacitidine vs aza alone: median OS 14.7 vs 9.6 mo, CR+CRi 66% vs 28%.
• APL: ATRA + arsenic for low/intermediate risk; add GO for high risk.
• Targeted inhibitors have activity in both frontline and R/R settings; MRD-guided strategies are emerging.