Genetic Determinants of Response to Guadecitabine (SGI-110) in AML
Bibliographic record
Abstract
Abstract Background: Guadecitabine is a second generation hypomethylating drug with improved pharmacokinetics and pharmacodynamics compared to decitabine and demonstrated clinical activity in both treatment naïve (tn) and relapsed-refractory (rr) AML. Previous studies reported similar response rates to guadecitabine in different cytogenetic subsets but it remains unknown whether this extends to genetic changes. Methods: Pre-treatment blood or bone marrow derived DNA was available for analysis from a total of 220 patients with AML enrolled on guadecitabine phase I-II trials (121 rrAML and 99 tnAML). We included only patients treated at therapeutic doses. Mutations in a panel of 54 genes was studied by the TruSight Myeloid Sequencing Panel (Illumina) and deep sequencingon the Illumina HiSeq platform. FLT3 mutations were separately determined by PCR analysis. The sequence data was analyzed for mutations using the TruSeq Amplicon Application in the Illumina BaseSpace Suite. Putative mutation calls were further filtered by sequencing read quality, minimum variant allelic fraction, and presence in the dbSNP and COSMIC databases. Results: In aggregate, responses to guadecitabine in rrAML were 15 CR (12%), 12 PR/CRi/CRp and 94 non responders (NR) and in tnAML, responses were 34 CR (34%), 21 PR/CRi/CRp and 44 NR. Overall, a median of 1 (range 0 - 5) mutation was present in each patient, with no significant differences between tn and rr AML. The most frequently mutated genes were ASXL1 (16.8%), TET2 (14.1%), IDH2 (10.9%), NPM1 (10%), RUNX1 (9.5%), DNMT3A (9.1%), NRAS (9.1%), FLT3-ITD (8.6%), U2AF1 (8.2%), IDH1 (6.8%), TP53 (5%), and KRAS (4.5%). The distribution of mutations was as expected for a group of patients with rrAML and elderly tnAML. We used Fisher's exact tests to compare mutation frequencies between patients who achieved CR and those who did not achieve CR. When we evaluated rrAML and tnAML separately, none of the genes showed a significantly different mutation rate between response subgroups. We then examined the population as a whole (N=220) and found that mutations in NRAS were significantly lower in patients who achieved CR (0/49) compared to those who did not (20/171, p=0.009). NRAS and KRAS mutations were inversely correlated, and when we considered the two genes together, mutations were present in 1/49 CR patients compared to 28/171 non-CR patients (p=0.007). Overall, CR rate was 3.4% in patients with RAS mutations compared to 25.1% in patients without such mutations. There was a similar significant trend for IDH2 mutations to be lower in CR patients (1/48) compared to non-CR patients (23/172, p=0.02) but this was not seen for IDH1. None of the mutations in other epigenetic regulators (DNMT3A, ASXL1, EZH2, TET2, U2AF1 or WT1) were significantly different between CR and non-CR patients individually or when we considered mutations in any of 8 epigenetic regulators (mutated in 22/49 CR patients vs. 93/171 non-CR patients, p=0.26). RAS mutations were higher in rrAML (22/121, 18.2%) than in tnAML (7/99, 7.1%, p=0.017) which may explain the lower CR rate in this group. Patients with PR/CRi/CRp were genetically similar to NR. Conclusions: In patients with AML treated with guadecitabine, RAS pathway mutations and IDH2 mutations are associated with a lower likelihood of achieving a CR. Disclosures Kropf: Celgene: Consultancy; Takeda: Consultancy. Roboz:Cellectis: Research Funding; Agios, Amgen, Amphivena, Astex, AstraZeneca, Boehringer Ingelheim, Celator, Celgene, Genoptix, Janssen, Juno, MEI Pharma, MedImmune, Novartis, Onconova, Pfizer, Roche/Genentech, Sunesis, Teva: Consultancy. Yee:Novartis Canada: Membership on an entity's Board of Directors or advisory committees, Research Funding. Taverna:Astex Pharmaceuticals: Employment. Naim:Astex Pharmaceuticals: Employment. Azab:Astex Pharmaceuticals, Inc.: Employment. Issa:Teva Pharmaceutical Industries: Consultancy; Astex Pharmaceuticals: Consultancy.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".