Mangaonkar, Abhishek A.

Abstract Background: Isocitrate dehydrogenase (IDH) mutations are frequently implicated in the development of acute myeloid leukemia (AML). In a recent study (Salman MY, Blood Adv 2025), it was reported that IDH2 mutation subtypes, R140 and R172, may respond differently to intensive chemotherapy. In this study, we evaluated mid-cycle and end-of-cycle bone marrow blast responses in AML patients with IDH2 R140, IDH2 R172, and IDH1 R132 mutations, aiming to determine whether these subtypes exhibit different response dynamics. Methods: We reviewed the charts of AML patients diagnosed at Mayo Clinic, Rochester, between the years 2016 - 2023 that had available next generation sequencing (NGS) data. Institutional review board approval was obtained. All data were recorded at the time of AML diagnosis. Patients without molecular data or not treated with intensive chemotherapy (7+3 or CPX-351) were excluded from this study. Bone marrow blast percentages were assessed at mid-induction (days 14-21) and end of cycle (counts recovery). Blast clearance dynamics and complete remission rates were compared across IDH mutation subgroups and correlated with outcomes. GraphPad Prism 10.0.0 was used for statistical analysis. Results: Out of 381 AML patients, 31 had IDH2 mutations (23 with R140 and 8 with R172), and 8 had IDH1 mutations (all R132). Among patients with IDH2 mutations, those with R172 had numerically higher median mid-cycle bone marrow blast percentage compared to R140 (10% vs. 4%, p = 0.37). There was no significant difference in mid-cycle blast percentages between patients with IDH1 and IDH2 mutations (4.5% vs. 4%, p = 0.8). At mid-cycle, 68% of patients with IDH2 R140 had blast counts <5%, compared to 38% of those with IDH2 R172, suggesting a trend for faster mid-cycle clearance in the R140 group, though not statistically significant (OR 3.57, p = 0.2). In a broader comparison, 60% of IDH2-mutated patients and 50% of IDH1-mutated patients achieved blast clearance (blasts <5%) at mid-cycle, though not statistically significant (p = 0.7). At the end of the induction, patients with IDH2 R140 had significantly lower median bone marrow blast percentages compared to those with R172mutations (2% vs. 4%, p = 0.01). Similarly, patients with IDH1 mutations showed low blast counts compared to the IDH2 group (1.5% vs. 3%, p = 0.1). A higher proportion of IDH2 R140 patients achieved blast clearance compared to those with R172 mutations (100% vs. 75%, p = 0.06), while proportion of patients who achieved blast clearance were similar between IDH1 and IDH2 patients (87.5% vs 93.1%, P = 0.5). Analysis of overall complete remission rates (including both complete remission [CR] and CR with incomplete hematologic recovery [CRi]) within the IDH2-mutated group showed a higher rate in patients with the R140 mutation compared to those with R172 (87% vs. 75%), though the difference was not statistically significant (p = 0.5). When comparing IDH1- and IDH2-mutated patients, overall complete remission rates were 75% and 84%, respectively, with no significant difference observed (p = 0.6). Kaplan–Meier analysis of time to end-of-induction blast clearance revealed no statistically significant difference between IDH2 R140 and R172 patients (median time to clearance: 31 vs. 37 days, p = 0.3). Similarly, no difference was observed between IDH1 and IDH2-mutated patients (33 vs. 31 days, p = 0.9). Overall survival analyses showed no significant difference between IDH2 R140 and R172 groups (median 66.8 months vs not reached, P = 0.4). Likewise, no difference was observed between IDH1 and IDH2 patients (44.4 months vs not reached, P = 0.7), indicating similar overall survival across mutation subtypes. Conclusion: IDH2-R140 was associated with more effective blast cytoreduction following intensive induction compared to IDH2-R172. Although not statistically significant, IDH2 R172consistently showed trends toward higher blast percentages, delayed response, and lower overall complete remission rates. These findings align with the recent study reporting significantly slower blast reduction and delayed response in R172-mutated AML. Also, no significant differences were found between IDH1- and IDH2-mutated patients in terms of response kinetics or overall survival. Together, these findings highlight potential biological differences among IDH mutation subtypes that may impact treatment response and require further investigation in larger prospective cohorts.

Abstract Background: Germline predisposition states (GPS), including inherited bone marrow failure syndromes (IBMFS) are seen in 5-10% of individuals with myeloid neoplasms (MN). Given overlap with somatic driver genes, blood, saliva, and buccal swabs are suboptimal specimens for GPS testing in MN; the latter two due to contamination with blood cells(Jerez, 2024). Skin biopsy with fibroblast culture-derived DNA remains the gold standard, but the process is expensive, takes 6-8 weeks to complete, and fibroblast culture failure can occur in some patients with telomere biology disorders. Given the need to rapidly screen donors for HSCT and adjust therapy/conditioning in certain IBMFS, we developed a rapid hematologic germline testing algorithm using hair follicle-derived DNA. Methods The study was conducted after IRB approval through the Mayo Clinic GPS/IBMFS clinic. To perform PCR-based hair follicle testing, we obtained 15-25 hair follicles (bulbs) per patient. The bulbs were placed in a microcentrifuge tube and subjected to a 24-hour digestion and extraction protocol using a modified version of the QIAGEN DNA Purification from Dried Blood Spots protocol. DNA quality was assessed by Nanodrop and Qubit.Testing included either validation of known or suspected germline variants (e.g., RUNX1, DDX41, ETV6), or panel testing (e.g., GPS or IBMFS). Pathogenic variants were validated by repeat testing in a CAP/CLIA certified lab prior to medical decision making. 23 patients underwent whole exome sequencing (WES) in a CAP/CLIA lab, and 13 patients had concurrent germline testing using alternate DNA sources (8 skin biopsies, 1 saliva, 4 buccal swabs). Results Ninety-three patients (49 male, 53%) underwent hair follicle-derived germline DNA testing (median age 63 years). Indications for germline testing included the presence of a suspected variant on somatic testing with variant allele frequency (VAF) between 35%-60% ( n= 66, 71%), patient age <50 years (n=26, 28%), antecedent thrombocytopenia (n=15, 16%), presence of syndromic features (n=4, 4%), patient history of multiple malignancies (n=4, 4%), and positive family history (n=15, 16%). Time to reporting variants after receipt of hair follicle-derived DNA was <14 days. There were zero complications associated with the procedure. DNA quality as assessed by Nanodrop revealed a median nucleic acid concentration of 25.64 ng/μL. The median A260/A280 ratio was 1.86 (range 0.701-2.133) and the median A260/A230 ratio was 1.27 (range -8.193-9.223). DNA quality was insufficient for chromosomal microarray (CMA) assessments. Of 93 patients evaluated, 67 (72%) had ≥1 germline variant, with 3 having >1 germline variant for a total of 72 variants. Pathogenic/likely pathogenic variants identified included: DDX41 (30, 41.7%), TP53 (4, 5.6%), ANKRD26 (4, 5.6%), RUNX1 (4, 5.6%), GATA2 (3, 4.2%), TERT (3, 4.2%), JAK3 (2, 2.8%), SAMD9/SAMD9L (2, 2.8%), ATM (2, 2.8%), and 1 each (1.4%) in BCL2, BCOR, BPGM, CDKN2A, CEBPA, CHEK2, CSF3R, CXCR4, DNMT3A, ETV6, GATA1, MPL, MPO, POT1, PTPN11, RPL5, and SH2B3. In the 13 patients that had concurrent germline testing, results were identical in all cases, except for 1 false negative in a buccal specimen (PTPN11) and 1 case of germline mosaicism, where the TP53 variant was identified in the blood and hair follicle DNA but not in skin fibroblast DNA. After testing, 30 (44.7%) patients had a change in their management plan: 8 (27%) with DDX41-GPS were observed due to known latency of DDX41-mutant MN, 6 (20%) were referred for HSCT based on the germline variant (CEBPA, DDX41, GATA1, GATA2, RPL5, TP53), 6 (20%) were placed on an enhanced surveillance plan for cancer screening (1 ANKRD26, 3 TERT, 2 TP53), 4 (13%) patients undergoing HSCT had related donors removed from consideration as they were carriers (2 DDX41, 1 ANKRD26, and 1 SAMD9), 2 (7%) patients with neutropenia and underlying germline variants (1 ATM, 1 DDX41) had G-CSF therapy discontinued due to concerns for clonal stimulation, and 1 (3%) patient with a germline CDKN2A mutation (hematological diagnosis- CMML) was recommended to avoid radiation therapy for prostate cancer. Conclusion In summary, we validate a rapid and simple hair follicle-derived DNA based germline assessment for patients with hematological disorders and demonstrate a positive impact on clinical care in 44% of patients. We are currently optimizing DNA quality so that germline copy number alterations can be better detected using CMA.

Abstract Introduction: TERT promoter (pTERT) variants are among the most frequent genetic alterations in human cancers, where they drive oncogenesis through telomerase reactivation. However, somatic, pathogenic pTERT variants have also been seen in non-cancerous conditions such as telomere biology disorders (TBD), where they attempt to rescue the bone marrow failure phenotype (adaptive compensation). We conducted this study to assess the phenotype and outcomes of pTERT mutations in TBD patients. Methods: After IRB approval, the Mayo Clinic institutional TBD database was queried to identify all patients with somatic pTERT mutations. We have designed a customized error corrected sequencing panel that covers somatic mutations (52 genes) associated with adaptive and maladaptive outcomes in TBD patients. This covers the promoter region for TERT and can detect variants down to 0.5% variant allele fraction (VAF) with 10,000 X sequencing depth. Clinical data was retrospectively abstracted. Results: We identified 7 pTERT variants (3 hotspots) in 6 (two females (33%), mean age 61 years, 4 alive). The three hotspots were: (1) c.-57A>C (2 individuals, mean VAF 17.1%; range 11.6%-22.6%), (2) c.-124C>T (4 individuals, mean VAF 12.9%; range 1.9%-41.3%), and (3) c.-146C>T (1 individual, VAF 1.2%). One individual had both the c.-124C>T (VAF 41.3%) and the c.-146C>T somatic pTERT variants, occurring in the context of a germline TERT variant. The median VAF for all 7 pTERT variants was 6.1% (range 1.2%-41.3%). The hotspot, c.-57A>C, occurred in the context of pathogenic germline TERT (VAF 49.9%) variants with pathogenic somatic co-variants in PPM1D (avg. VAF 1.5%). The second hotspot, c.-124C>T, occurred in the context of germline TERT (avg. VAF 49.6%) and RTEL1 (VAF 47.3%) variants with pathogenic somatic co-variants including ASXL1 (VAF 1%), DNMT3A (VAF 8%), and PPM1D (VAF 4%). Five of six assessable patients with pTERT variants had shortened telomere lengths (TL), with results pending for one individual. Among those with the c.-57A>C variant, granulocyte TLs were at or below the 10th percentile, while lymphocyte TLs were below the 10th percentile in one patient and below the 1st percentile in another. For those with the c.-124C>T variant, all granulocyte TLs were below the 1st percentile. Lymphocyte TLs in this group included one at the 10th percentile and two at or below the 1st percentile. The median follow up for all patients was 2 years. Hotspot 1 (c.-57A>C) variants were associated with normocellular bone marrow biopsies with one patient having normal blood counts and the other having chronic macrocytic anemia (grade 2) and thrombocytopenia (< grade 1). In patients with hotspot 2 and 3 somatic variants, bone marrow cellularity ranged from 15–50% (1 patient had grade 2-3 hypocellularity), with normal bone marrow morphology, except for polytypic plasma cells (10–15%) identified in a patient with IgG kappa multiple myeloma (currently MRD negative on maintenance therapy). This patient also had a history of head and neck cancer and follicular lymphoma. Of these four patients, only one had normal blood counts (co-occurring c.-146C>T variant); the rest had cytopenias (n=3, the worst being grade 2 thrombocytopenia), progressive macrocytosis (n=3), or eosinophilia (n=1). Comorbidities encountered in these patients included myasthenia gravis (1), hypogonadism (1), and sarcoidosis (1). Moreover, four patients (2 each for hotspots 1 and 2) presented with idiopathic pulmonary fibrosis, while liver fibrosis was seen in 4 patients (also 2 each for hotspots 1 and 2). At last follow up, none of these patients had documented progression to myelodysplastic syndrome or acute myeloid leukemia. Conclusions: We describe the spectrum and clinical features of somatic pTERT mutations in TBD patients. pTERT mutations involved the following hotspot regions, c.-124C>T, c.-146C>T, and c.-57A>C, and co-occurred with other somatic variants (i.e., PPM1D and DNMT3A). None of the 6 patients with pTERT mutations had a severe bone marrow failure phenotype or developed myeloid neoplasms at last follow up, exemplifying a unique mechanism of adaptive somatic rescue.

Abstract Prognosis in acute myeloid leukemia (AML) patients is influenced by various factors, most importantly, patient clinical status and cytogenetic/molecular abnormalities. Within those factors, elevated lactate dehydrogenase (LDH) is considered a poor prognostic indicator in AML patients. Our study aims at exploring the different factors associated with elevated LDH and their impact on overall survival (OS). After obtaining IRB approval, we reviewed the charts of 381 AML patients diagnosed at Mayo Clinic, Rochester, between the years of 2016 and 2023. All data was recorded at the time of AML diagnosis. Patients with no cytogenetic and molecular data were excluded from this study. LDH levels were recorded if measured within 10 days of diagnosis and treatment for AML was not yet given. Both risk stratification and response evaluation were done following the 2022 European LeukemiaNet (ELN) guidelines. OS was calculated from the date of AML diagnosis to the date of last follow-up or death. We used BlueSky Statistics V 10.3.4 for statistical analysis. We identified 299 patients (78.4%) who had elevated LDH levels (normal ≤ 222 U/L) at the time of diagnosis with a median of 417 U/L (range, 223-7989). The median age for those patients was 68 years (range, 18-94), with 78 (26.1%) of them having a prior myeloid disease (MDS, MPN, or MDS/MPN) before AML diagnosis. By risk stratification, more than half of the patients had an adverse risk profile (n=171, 57.2%), while 71 (23.7%) and 57 (19.1%) had an intermediate and favorable risk, respectively. Most of those patients had an abnormal karyotype (n=204, 68.7%), with complex karyotype (n=103, 34.7%) being the most common abnormality, followed by monosomy 7 (n=50, 16.8%) and trisomy 8 (n=46, 15.5%). Next-generation sequencing (NGS) was performed on 268 (89.6%) patients with elevated LDH, and the most common mutations were TP53 (n=60, 22.3%), FLT3-ITD (n=59, 21.1%), and NPM1 (n=51, 18.5%). TP53 mutations were more common in patients with normal/low white blood cell count (WBC) in contrast with FLT3 and NPM1, which were the most common mutations in patients with high WBC. After grouping the patients according to their LDH levels (group 1: 223-444 U/L, group 2: 444-666 U/L, group 3: > 666 U/L), we found that FLT3-ITD mutations were more likely to occur at higher LDH levels i.e. groups 2 and 3 (p= 0.002). The majority of patients with elevated LDH (n=262, 87.6%) received therapy following diagnosis, with the most common treatment regimens being a combination of a hypomethylating agent and venetoclax (HMA+VEN) (n=97, 37%) and a (7+3)-based induction therapy (n=77, 29.4%). When comparing patients with elevated LDH (n=299, 78.4%) to patients who had non-elevated LDH (n=82, 21.5%), no difference in both median age at diagnosis and risk stratification was found. We noticed an increase in WBC (median, 7.7 vs. 2.1) (p<0.001) along with an increase in peripheral blasts (median, 28% vs. 7%) (p<0.001) in elevated LDH patients. Among patients with non-elevated LDH, 21% (n=17) had a complex karyotype compared to 34.7% (n=103) in the elevated LDH group (p=0.022). The molecular profile was different between the 2 groups, where the most common mutations in patients with non-elevated LDH were IDH2 (n=19, 25.3%), RUNX1 (n=17, 23%), and DNMT3A (n=16, 21.6%). All patients (n=82, 100%) in the non-elevated LDH group received therapy after diagnosis. There was no difference in terms of treatment regimens used between the 2 groups. Moreover, no statistical difference in remission rates after one line of therapy was found between those groups where non-elevated LDH patients had a 65.9% remission rate compared to the 55.3% remission rate in patients with elevated LDH (p=0.098). Patients with elevated LDH had a shorter OS than patients with non-elevated LDH (9.5 vs. 26.4 months, p=0.0008). However, median OS was similar across the different groups of elevated LDH. Using univariate analysis in our cohort (n=381), age, HSCT, risk stratification, FLT3 mutation, and LDH elevation all had an impact on OS. However, when comparing those variables using multivariate analysis, age and FLT3 mutation lost their significance (p-value of 0.25 and 0.84, respectively).Complex karyotype and FLT3 mutations were more common in patients with elevated LDH. These patients tend to have higher WBC and peripheral blasts. Elevated LDH had a negative impact on patients' OS regardless of patients' risk classification.

Abstract RAS pathway (RASp) mutations induce proliferative features, and promote transformation in chronic myelomonocytic leukemia (CMML). However, the unique clonal landscape and hierarchy of distinct RASp mutations remain unexplored. To characterize the landscape, architecture, and implications of unique RASp mutations in CMML, we evaluated a cohort of 814 patients with CMML. We identified 461 RASp mutations among 342 patients (42%). N/KRAS and CBL mutations were the most common, frequently involved the P-loop or RING domains, respectively, and frequently appeared as dominant events (63% and 65%, respectively). BRAF, NF1, and PTPN11 mutations spanned throughout the gene structure, and frequently appeared as subclonal events (75%, 64%, and 59%, respectively). CBL mutations frequently occurred in codominance with SRSF2 and multihit TET2, and were enriched for KIT mutations. PTPN11 mutations more frequently co-occurred with SETBP1 and DNMT3A mutations, and were infrequently codominant with TET2 or ASXL1. RASp mutations predicted for shorter overall survival (hazard ratio [HR], 1.55; 95% confidence interval [CI], 1.15-2.07; P = .0075) and leukemia-free survival (LFS; HR, 1.67; 95% CI, 1.26-2.20; P = .0011), influenced outcomes of myelodysplastic and TET2 mutant CMML, and cooperated with IDH2 and RUNX1 mutations to induce shorter LFS. These data set the bases for refined genomic classifications of CMML, and underscore the need to develop RAS-directed therapies for patients with CMML.