The Composite Health Risk Assessment Model (CHARM) Predicts Risks of Toxicities, Functional and Cognitive Decline Among Survivors of Allogeneic Hematopoietic Cell Transplantation (allo-HCT): A Prospective BMT-CTN Study 1704
Bibliographic record
Abstract
Introduction: The BMT CTN 1704 study developed and validated the CHARM that stratified risk for non-relapse mortality and overall mortality among older adults, performing better than the HCT-CI alone (Artz A et al, Blood. 2023;142:109) and equally to two machine learning models. CHARM assigns a total score for 7 health variables: increasing age, higher HCT-CI scores, lower albumin, higher C-reactive protein, higher percent of weight loss over the preceding year, lower patient-reported performance status scores, and lower cognitive score per Montreal cognitive assessment (MoCA). A CHARM calculator is available at:https://cibmtr.org/CIBMTR/OffNav/DevSandbox/CHARM-Risk-NRM-Calculator No multi-institutional prospective data exist on functional trajectories and morbidity after HCT in older patients. We now report on the association of CHARM to trajectories of secondary morbidity outcomes among allo-HCT survivors of this large, prospective study. Patients and Methods: Allo-HCT candidates, aged ≥60 years (yrs), were enrolled (n=1226) from 49 centers in the US between 2019 - 2021. The primary analysis includes 1105 patients proceeding to allo-HCT on study and secondary endpoints were assessed at day (D) 100, 180 and 365 except MoCA and organ toxicity were restricted to D100 and frailty had inadequate data for D100. A sequential multiple imputation strategy was implemented to impute endpoints for survivors at each time point with missing data. Analyses on multiply imputed datasets were conducted and the results combined using Rubin's rule. Associations between CHARM scores and secondary outcomes were analyzed using a multivariable Cox, Fine-Gray, Generalized Estimating Equations, and logistic regression model for survival, competing risks, continuous, and binary outcomes, respectively, with latter two focused on surviving patients. Models were adjusted for other variables including conditioning intensity, graft-versus-host disease (GVHD) prophylaxis regimen, disease-risk index, donor-recipient gender match, ethnicity, baseline value of the dependent variable, and visit timepoints. Results: Higher CHARM scores were associated with development of serious organ toxicities by D100 (OR: 2.05, [1.52-2.78], p<0.0001) and ≥2 worsening score on the MoCA (odds ratio (OR) 1.55 [1.16-2.1], p=0.003). Among survivors at all timepoints, higher CHARM scores were associated with greater disability by instrumental activities of daily living (IADL) (Slope -0.640 [-0.433-0.846], p<0.001) and worsening Patient-Reported Outcomes Measurement Reporting System (PROMIS) physical function (Slope -0.981 [-0.057 - -1.904], p=0.037), depression (Slope 0.763 [0.042-1.484], p=0.038) and in a lesser magnitude anxiety (Slope 0.659, p=0.076). Among survivors at D180 and D365, higher CHARM scores were associated with worse frailty (Slope 0.193 [0.081-0.305], p<0.001). CHARM scores were not associated with development of acute GVHD grades 2-4 or 3-4 but were associated with post-GVHD increased mortality (HR: 1.61, [1.25-2.08], p=0.0002). Higher CHARM scores are associated with a lower incidence of chronic GVHD (HR: 0.83, p=0.026), likely due to the effect of CHARM on the competing risk of death leaving fewer patients at risk for chronic GVHD. Conclusions: The novel primary CHARM, originally developed to predict risks of NRM, also predicts worse frailty, disability, cognitive decline, and serious organ toxicities; outcomes that are critically important to older recipients of allo-HCT. CHARM therefore informs risks of transplant morbidity, separate from risks of developing acute GVHD. Results further support adopting CHARM in practice to counsel patients, expedite HCT referrals for lower risk CHARM, and design trials for high CHARM score patients.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 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) | 0.000 | 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 teacher head, 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".