Validation of the FL/IL6 score, a new dynamic cytokine-based risk-model in adults with acute myeloid leukemia, a prospective multicentre filo study.
Bibliographic record
Abstract
Abstract Background: In a recent monocentric study in adults with acute myeloid leukemia (AML), it was shown that combining the kinetic profile of plasma level concentrations of Fms-like Tyrosine Kinase 3 Ligand (FL) with that of Interleukine-6 (IL6) level evaluated at day (D) 22 during a frontline intensive induction was strongly predictive of response and overall survival (OS). Indeed, three FL kinetic profiles can be delineated during induction (FL risk-model): i) sustained increase of FL between D 1 and D22 (FLI group), ii) increase from D1 to D15, then decrease at D22 (FLD group) and iii) stagnation of low levels (FLL group). With longer follow-up, it was observed that FLI and FLD groups shared similar outcomes while the FLL group kept a very bad prognostic. Thereafter, D22 IL6 level allowed to predict survival within the combined FLI/FLD sub-group. Thus, a new FL/IL6 risk-model was proposed, distinguishing 3 sub-groups: FLI/FLD with IL-6 <15.5 pg/mL (favorable, F), FLI/FLD with IL-6 ≥ 15.5 pg/mL (intermediate, I) and FLL (adverse, A). To confirm these results, we have conducted a large prospective multicentre study. Preliminary results, focusing on the FL risk-model impact, were presented at the ASH 2023 meeting (Blood (2023) 142: 975). We now report on the impact of the FL/IL6 score in the same cohort of AML patients (pts) with a longer follow-up. Methods: The non-interventional multicentric prospective FLAMVAL study (clinicaltrials.gov: NCT04641910) was conducted on behalf of the French Innovative Leukemia Organization (FILO) and was funded by Astellas. The main objective was to validate the prognostic impact on 2y OS of the new FL/IL6 risk-model. All adults (≥ 18 years old, yo) with a non-previously treated non-M3 AML and receiving a standard intensive induction chemotherapy were eligible. FL and IL6 plasma levels were evaluated at D1, D8, D15 and D22 using a cytokine multiplex assay from Milliplex (MILLIPLEX®). The number of pts to be included was estimated at 201. All pts gave informed consent. Results: Between June 2021 and August 2022, 226 pts from 16 FILO centres were included and 203 with full available dosages were eventually considered for analyses. Median age was 59 yo [IQR:48-67], 52% were men, median white blood count (WBC) was 4.15 G/L [IQR: 2-11.5]. Of the 48 pts with FLT3 mutations (FLT3-ITD n=40, FLT3-TKD n=7, both n=1), 41 received a FLT3 inhibitor during induction. Also 106 pts (52%) received allo-HSCT. According to ELN-2017 risk-classification, 55 (27%), 87 (43%) and 61 (30%) pts had F-, I- and A-risk, respectively. Considering the whole cohort, the complete remission (CR)/CR with incomplete hematologic recovery (CRi) rate (according to ELN-2017 risk-classification) was 76%, 2y OS and EFS were 70% (95%CI: 63; 73) and 45% (95%CI: 38-52), respectively. Median OS was not reached, median EFS was 24 months (95%CI: 21; 24). According to the new FL/IL6 risk-model, 44 (22%), 81 (40%) and 74 (36%) pts had F-, I- and A-risk, respectively, while 4 pts (2%) could not be classified. The CR/CRi rates were F: 89%, I: 89% and A: 66%, p<0.001. 2-y OS were F: 84% (70-93), I: 69% (58-79) and A: 61%, p=0.04. 2y EFS were F: 55% (39-70), I: 52% (47-64) and A: 30% (20-42), p=0.003. Multivariate analyses (MA) were performed for response, OS and LFS, including the following factors: FL/IL6 risk-model, ELN-2017 risk-classification, age (< vs ≥ 60 yo), WBC (< vs ≥ 20 G/L), sex and FLT3 status. FL/IL6 F-risk (vs A-risk, HR: 0.32; 95%CI: 0.18-0.46, p<0.001) and ELN2017 F-risk (vs I-risk, HR: 0.23; 95%CI: 0.09-0.36, p=0.001; and vs A-risk, HR: 0.25; 95%CI: 0.10-0.39, p=0.001) were significantly associated with better CR/CRi rate after induction. Both FL/IL6 I-risk (vs F-risk, HR: 2.82, 95%CI: 1.20-6.63, p=0.01) and A-risk (vs F-risk, HR: 2.99; 95%CI: 1.30-6.88, p=0.01) as well as ELN-2017 I-risk (vs F-risk, HR: 3.85; 95%CI: 1.46-10.19, p=0.007) and A-risk (vs F-risk, HR: 8.31; 95%CI: 3.15-21.88, p<0.001) were significantly associated with lower OS. Also, FL/IL6 A-risk (vs F-risk, HR: 2.35, 95%CI: 1.37-4.02, p=0.002) and ELN-2017 I-risk (vs F-risk, HR: 2.19; 95%CI: 1.25-3.85, p=0.006) and A-risk (vs F-risk, HR: 2.65; 95%CI: 1.51-4.63, p=0.001) were significantly associated with lower EFS. Age, WBC, sex, and FLT3 status have no impact on response, OS and EFS. Conclusion: The new FL/IL6 cytokine-based risk-model predicts response and survivals in intensively treated adult AML pts, deserving further investigation in AML studies.
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 | 0.003 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".