Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Sep 1.
Published in final edited form as: Br J Haematol. 2021 Jun 7;194(5):927–931. doi: 10.1111/bjh.17573

Isocitrate dehydrogenase inhibitor-driven differentiation may resemble secondary graft failure in post-allogeneic haematopoietic cell transplantation relapsed acute myeloid leukaemia

Adrianne Rasche 1,2, Emily F Mason 2,3, Stephen A Strickland 2,4, Michael Byrne 2,4, P Brent Ferrell 2,4
PMCID: PMC8936480  NIHMSID: NIHMS1732566  PMID: 34096047

Targeted isocitrate dehydrogenase (IDH) inhibitor (IDHi) therapy is frequently used in patients with IDH-mutated acute myeloid leukaemia (AML). Here, we describe two patients with IDH-mutated AML who relapsed after allogeneic haematopoietic cell transplant (HCT) and were treated with IDHi therapy. Both patients showed evidence of differentiation, a known effect of IDHi therapy, which manifested as a significant but transient drop in donor chimaerism.

Patient 1

A 58-year-old female with a history of breast carcinoma, treated with chemoradiotherapy, developed pancytopenia approximately two years after diagnosis. Bone marrow aspiration and core biopsy (BMBx) showed AML. Cytogenetic testing showed an abnormal karyotype: 46,XX,t(3;5)(q5;q34) [20], and single-gene molecular testing was positive for an fms-like tyrosine kinase 3 (FLT3) gene internal tandem duplication (ITD) mutation (mutant allelic ratio not available). Next-generation sequencing (NGS) was not performed. She was induced with 7 + 3, achieved a complete remission (CR), and received three cycles of high-dose cytarabine (HiDAC) consolidation plus a FLT3 inhibitor. Single-gene testing performed after consolidation chemotherapy was negative for a FLT3-ITD mutation. Following myeloablative conditioning chemotherapy, she underwent allogeneic HCT with a 9/10 A antigen mismatched unrelated donor. She achieved full donor chimaerism at day +30, single-gene molecular testing performed on the day +30 BMBx was negative for a FLT3-ITD mutation, and she began post-HCT maintenance with a FLT3-inhibitor at day +81.

The patient’s course was remarkable for mild chronic graft-versus-host disease (GVHD) of the skin and upper gastrointestinal tract. Immunosuppression was discontinued at day +623. She relapsed with FLT3-ITD-negative disease at day +651. A BMBx revealed 20–25% blasts (Fig 1AC). Cytogenetics showed a mixed abnormal female (recipient) and normal male (donor) karyotype: 46,XX,t(3;5)(21:q31)[8]/46, XY[12], consistent with the bone marrow engraftment (BME) studies reporting 39% donor DNA (Fig 2A). NGS identified an IDH1 R132H mutation with a variant allele frequency (VAF) of 17%. The patient enrolled in a clinical study of azacitidine in combination with the IDH1i olutasidenib (NCT02719574). After one cycle, BMBx revealed a moderately hypercellular marrow with trilineage haematopoiesis, left-shifted myeloid maturation, multilineage dysplasia and no increase in blasts (Fig 1DF). BME studies showed a further reduction in donor DNA to 8%, which was reflected in her karyotype (Fig 2A). Approximately two weeks later, the white blood cell (WBC) count increased, eventually peaking at 26 400/μl, with a neutrophil predominance. Sorted peripheral blood (PB) chimaerism testing after cycle 2 showed a CD33+ (myeloid) donor chimaerism of 3%. Following three cycles of AZA/IDH1i, the WBC count normalized, and repeat BMBx revealed a normocellular marrow with maturing trilineage haematopoiesis and no increase in blasts (Fig 1G, H). Cytogenetic analysis showed a normal male (donor) karyotype, and both BME and sorted CD33+ PB chimaerism studies reported 100% donor DNA. The patient completed six cycles without interruption before receiving her first donor lymphocyte infusion (DLI) at day +840, with three subsequent DLI doses at days +911, +979, and +1137. Now >2 years out from her relapse, she remains in CR on single-agent ivosidenib maintenance.

Fig 1.

Fig 1.

Morphologic and genetic evidence of leukaemic cell differentiation with isocitrate dehydrogenase (IDH) inhibitor therapy. (A–H) Patient #1: prior to initiation of IDH inhibitor therapy (A–C), the bone marrow showed 20–25% blasts, which by aspirate morphology (A) were intermediate in size with oval to irregular nuclei and scant cytoplasm (➞). The core biopsy (B) was hypercellular for age and showed increased blasts with irregular nuclei and scant cytoplasm (➡), which were positive for CD117 (C). After one cycle of azacitidine + IDH1i (D–F), the aspirate smear (D) showed relative myeloid hyperplasia with left-shifted myeloid maturation, with frequent myelocytes and metamyelocytes (⊳) but with no increase in blasts. The core biopsy (E) was hypercellular for age and showed left-shifted myeloid elements with abundant eosinophilic cytoplasm (►), consistent with maturation past the blast stage. A CD117 stain (F) showed no increase in blasts. A bone marrow biopsy performed after three cycles of therapy (G, H) showed trilineage haematopoiesis with a full range of maturation and no increase in blasts. (I, J) Patient #2: the bone marrow aspirate smear at original diagnosis (I) showed 84% blasts, which were small to intermediate in size with scant cytoplasm and occasional Auer rods. After four weeks of IDH2i (J), blasts showed monocytic features, with larger size, folded nuclei and more abundant cytoplasm with occasional cytoplasmic vacuolization. All aspirate and H&E images: 1000×. CD117 immunostain images: 400×.

Fig 2.

Fig 2.

Laboratory and genetic data during treatment. (A) Patient #1: at pre-treatment, the patient had recurrence of 20–25% leukaemic blasts, return of her original cytogenetic abnormality in 8/20 metaphases and was 39% donor by BME studies. Following one cycle of IDHi therapy (C2D1), blasts were no longer increased; however, her original cytogenetic abnormality was present in an increased number of metaphases (17/20), and BME donor chimaerism decreased to 8%. At C3D1, the WBC count increased to 26 K/μl, with 82% mature neutrophils, and sorted PB chimaerism showed 3% donor in CD33+ cells. By C4D1, the white blood cell count had normalized, her BMBx showed no increase in blasts with return of a normal male (donor) karyotype and BME showed 100% donor. End of treatment assessments confirmed ongoing remission. (B) Patient #2: at pre-treatment, the patient had immunophenotypic, cytogenetic and molecular evidence of low-level disease recurrence. Following one cycle of IDHi therapy (C2D1), blasts had increased to 15–20%, trisomy 11 was present in 29% of cells by FISH, BME dropped to 85% and sorted PB chimaerism for CD33+ cells showed 87% donor. By C4D1, the blast percentage had decreased to 11.5%; however, the percentage of cells with trisomy 11 by FISH increased to 35%, BME further decreased to 78% donor and sorted PB CD33+ chimaerism nadired at 53% donor. At C7D1, BMBx showed measurable residual disease by flow cytometry; trisomy 11 was seen in 4% of cells by FISH, and BME increased to 89%. BMBx, bone marrow biopsy; BME, bone marrow engraftment; CG, cytogenetics; EOT, end of treatment; FISH, fluorescence in situ hybridization; IP, immunophenotype; NGS, next-generation sequencing; PB CD33+, peripheral blood sorted chimaerism for CD33+ cells; WBC, white blood cell.

Patient 2

A 63-year-old female presented with a WBC of 144 500/μl and 90% circulating myeloblasts. BMBx showed AML (Fig 1I), and cytogenetic testing showed an abnormal karyotype with trisomy 11 (47,XX,+11[20]). NGS identified a mutation in IDH2 (p.Arg172Lys; 95% VAF). She achieved CR after 7 + 3 induction and received three cycles of HiDAC consolidation. Minimal residual disease testing was not performed after induction or consolidation. Following reduced intensity conditioning chemotherapy, she underwent allogeneic HCT with a 10/10-matched unrelated donor.

The patient had no acute or chronic GVHD and immunosuppression was discontinued on day +433. On day +1108, PB chimaerism fell to 97%, and a BMBx was performed, which showed <5% blasts; however, blasts had an abnormal immunophenotype identical to the diagnostic BMBx. Cytogenetics revealed a mixed abnormal female (recipient) and normal male (donor) karyotype: 47,XX,+11[3]/46,XY[17], and SNaPshot molecular testing showed the low-level presence of the IDH2 R172K mutation (Fig 2B). These findings were consistent low-level disease recurrence, and single-agent enasidenib was initiated. Repeat BMBx performed after four weeks of IDH2i therapy showed a normocellular marrow with decreased maturing myeloid elements and 15–20% blasts (Fig 1J). Although blasts were increased in the bone marrow, they showed morphologic evidence of monocytic differentiation (Fig 1J), which was not prominent at diagnosis (Fig 1I) and was favoured to represent a differentiation response to IDH2i therapy. Trisomy 11 was present in 8/20 metaphases by karyotype and in 29% of cells by fluorescence in-situ hybridization (FISH). NGS identified the IDH2 R172K mutation (13% VAF). Sorted CD33+ PB chimaerism studies showed a drop in donor chimaerism to 87%. Repeat BMBx after 12 weeks of therapy showed a decrease in blasts to 10% but a rise in the percentage of cells with trisomy 11 by FISH to 35%, and further decline in sorted PB chimaerism studies for CD33+ cells, which nadired at 53% donor DNA. Six months after relapse, the patient is stable on single-agent IDH2i therapy, with measurable residual disease detected by flow cytometry, and sorted PB chimaerism of 87% in CD33+ cells.

Discussion

We present two patients with post-HCT relapsed AML who were successfully treated with IDHi therapy. In patient 1, BME studies showed a sharp decline in donor chimaerism following one cycle of IDH1i, raising concern that graft function may not recover. However, with a growing understanding of IDHi-mediated differentiation, and data indicating that time to maximal response can be prolonged and variable,1,2 the treatment was continued. She achieved a CR with normalization of her WBC count and recovery of full donor chimaerism. Patient 2 also showed a decline in donor chimaerism with IDH2i treatment and continues therapy with improving BME studies.

These cases illustrate the transient changes related to leukaemic cell differentiation in the PB and bone marrow associated with IDHi use after allogeneic HCT. Declining donor chimaerism should be interpreted with caution and should not prompt an interruption or change in therapy without clear indicators of treatment failure. The patients presented here showed evidence of treatment response, with declining blasts over time. In addition, both patients showed expansion of a cytogenetically abnormal, disease-related myeloid clone without an increasing blast population. Thus, we hypothesize that the disease-related cytogenetic abnormalities were present within maturing myeloid elements, which would be consistent with a differentiation process.

The commercially available IDH1 inhibitor ivosidenib and IDH2 inhibitor enasidenib are being evaluated in the post-HCT maintenance setting (NCT03564821 and NCT03515512 respectively). Early data from the phase 1 enasidenib maintenance study indicate that the drug is well tolerated, with relapse occurring in 2/16 patients.3 In the relapsed/refractory setting, modest monotherapy response rates, need for extended dosing to maximum response and efficacy of post-HCT venetoclax-based therapies, particularly in IDH1/2-mutated myeloid diseases, are barriers to the widespread use of these agents.4

Leukaemic blast differentiation is commonly seen with IDHi therapy and may be associated with ‘differentiation syndrome’ in approximately 10% of patients.5,6 In a recent study of bone marrow morphology in patients receiving IDHi therapy in combination with intensive induction chemotherapy, 30% of patients showed morphologic or immunophenotypic evidence of a differentiation response.7 Experience with these agents in the post-transplant setting and their impact on traditional BME monitoring is limited. Recognizing these effects has important implications for management and consolidation. These cases, of two consecutive patients treated at our institution, illustrate the successful use of IDHi therapy on donor/recipient chimaerism and caution against the use of applying traditional parameters in evaluating treatment response. Our experience suggests that patients with evidence of differentiation, particularly with reduction or disappearance of leukaemic blasts, should be treated to maximum response regardless of donor chimaera, as recovery of donor haematopoiesis is possible.

Conflicts of interest

PBF receives research funding from Forma Therapeutics, Incyte and Astex Pharmaceuticals. MB receives research funding from Karyopharm Therapeutics.

References

  • 1.Stein EM, DiNardo CD, Pollyea DA, Fathi AT, Roboz GJ, Altman JK, et al. Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood. 2017;130:722–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.DiNardo CD, Stein EM, de Botton S, Roboz GJ, Altman JK, Mims AS, et al. Durable remissions with ivosidenib in IDH1-mutated relapsed or refractory AML. N Engl J Med. 2018;378:2386–98. [DOI] [PubMed] [Google Scholar]
  • 3.Fathi AT, Li S, Soiffer RJ, Levis M, Mims AS, Devine SM, et al. A phase I study of the IDH2 inhibitor enasidenib as maintenance therapy for IDH2-mutant myeloid neoplasms following hematopoietic cell transplantation. 2020 Virtual ASH Annual Meeting and Exposition; 2020. [Google Scholar]
  • 4.Byrne M, Danielson N, Sengsayadeth S, Rasche A, Culos K, Gatwood K, et al. The use of venetoclax-based salvage therapy for post-hematopoietic cell transplantation relapse of acute myeloid leukemia. Am J Hematol. 2020;95:1006–14. [DOI] [PubMed] [Google Scholar]
  • 5.Birendra KC, DiNardo CD. Evidence for clinical differentiation and differentiation syndrome in patients with acute myeloid leukemia and IDH1 mutations treated with the targeted mutant IDH1 Inhibitor, AG-120. Clin Lymphoma Myeloma Leuk. 2016;16:460–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Amatangelo MD, Quek L, Shih A, Stein EM, Roshal M, David MD, et al. Enasidenib induces acute myeloid leukemia cell differentiation to promote clinical response. Blood. 2017;130:732–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Mason EF, Pozdnyakova O, Roshal M, Fathi AT, Stein EM, Ferrell PB, et al. A novel differentiation response with combination IDH inhibitor and intensive induction therapy for AML. Blood Adv. 2021;5: 2279–83. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES