Blood biomarkers for the prediction of outcome after cardiac arrest: an international prospective observational study within the Targeted Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest (TTM2) trial
Bibliographic record
Abstract
BACKGROUND: Prognostication of recovery in patients who are unconscious following cardiac arrest can be guided by concentrations of brain injury biomarkers in the blood. The optimal biomarker and cutoff concentrations for the prediction of outcome remain unknown. In this study, we aimed to evaluate which biomarker of brain injury is most accurate for predicting functional outcome after cardiac arrest, and to evaluate cutoff levels for the prediction of good and poor outcome. METHODS: This study was a prospective, international, observational biomarker study within the international Targeted Hypothermia versus Normothermia after Out-of-Hospital Cardiac Arrest (TTM2) trial including adults aged 18 years or older with a presumed cardiac cause or unknown cause of arrest. Patients were recruited from 24 European hospitals. Serum samples were collected at 0, 24, 48, and 72 h after admission to intensive care units. Concentrations of neuron-specific enolase, S100, neurofilament light, and glial fibrillary acidic protein were analysed with Elecsys electrochemiluminescence immunoassays. The primary outcome was 6-month good (modified Rankin Scale 0-3) or poor (modified Rankin Scale 4-6) functional outcome. Prognostic accuracy was evaluated by the area under the receiver operating characteristic curve (AUROC). The biomarker with the highest AUROC at each timepoint was compared with that of the second highest marker using DeLong's test. As pre-specified, to account for multiple comparisons using Bonferroni correction, a p value of less than 0·0125 was considered statistically significant. FINDINGS: Between April, 2018, and January, 2020, 113 (12%) of 932 eligible patients were excluded due to death, missed sampling, or missing outcome data. 661 (81%) of 819 included patients were male and 158 (19%) were female, the mean age was 64 years (SD 13), and 418 (51%) had a poor outcome. In patients who were unconscious, neurofilament light predicted functional outcome with AUROCs at 0, 24, 48, and 72 h of 0·77 (95% CI 0·73-0·80), 0·92 (0·90-0·94), 0·93 (0·91-0·95), and 0·93 (0·91-0·95), respectively. Glial fibrillary acidic protein achieved an AUROC of 0·74 (95% CI 0·70-0·77) at 0 h, 0·87 (0·84-0·90) at 24 h, 0·87 (0·84-0·90) at 48 h, and 0·87 (0·84-0·91) at 72 h. Neuron-specific enolase predicted functional outcome with an AUROC of 0·61 (95% CI 0·56-0·65) at 0 h, 0·78 (0·75-0·82) at 24 h, 0·85 (0·81-0·88) at 48 h, and 0·86 (0·82-0·89) at 72 h. S100 achieved an AUROC of 0·74 (95% CI 0·71-0·78) at 0 h, 0·84 (0·81-0·87) at 24 h, 0·79 (0·75-0·82) at 48 h, and 0·78 (0·74-0·82) at 72 h. Neurofilament light had a statistically significantly higher AUROC than the second highest marker, glial fibrillary acidic protein, at 24, 48, and 72 h (p<0·0001), but not at 0 h (p=0·27). INTERPRETATION: Neurofilament light is a highly accurate predictor of long-term outcome after cardiac arrest and superior to other relevant biomarkers evaluated in this study. FUNDING: The Swedish Research Council (Vetenskapsrådet), the Swedish Heart-Lung Foundation, the Stig and Ragna Gorthon Foundation, the Knutsson Foundation, the Laerdal Foundation, the Hans-Gabriel and Alice Trolle-Wachtmeister Foundation for Medical Research, the Bundy Academy at Lund University, Regional Research Support in Skåne, the Swedish Government, and Roche Diagnostics International.
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 | 0.004 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| 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".