Consensus on cytokine release syndrome (CRS) monitoring practices: A Delphi approach
Bibliographic record
Abstract
Abstract Background Cytokine release syndrome (CRS) is a well-known and potentially fatal side effect of T cell redirecting therapy including therapies with chimeric antigen receptor T-cell (CAR-T) therapy and bispecific T-cell engagers (TCE) (Kanate, 2023 & Kanate, 2020). Current consensus grading and management guidelines for CRS stipulate that patients need to be monitored for at least seven days and up to four weeks after CAR-T or TCE infusion (Lee, 2019; NCCN, 2025). As more CRS incidence data is becoming available, monitoring is shifting from hospital to the outpatient or even home-based setting. Digital remote patient monitoring (RPM) strategies can support this important safety monitoring while enhancing patient convenience and reducing clinic resource utilization. The objective of this research was to establish consensus on type of digital data needed for CRS monitoring using RPM in post-discharged patients receiving CAR-T/TCE. Methods A modified Delphi method for consensus was used. We convened a steering committee of six medical experts in hematology and oncology which comprised physician investigators, a nurse leader, and industry trialists. Consensus statements were informed by various data sources: our steering committee, a review of clinical guidelines, protocols, and treatment management guidelines, and other digital technology currently available for patient-self capture of symptoms and physiological biomarkers. A total of 57 statements, reflecting current monitoring practices, perceived benefits of remote monitoring, and signs and symptoms commonly associated with onset of CRS, were developed and tested in the first round of the Delphi survey. Altogether, 203 health care professionals with experience in CAR-T and T-cell treatment, across six countries (US = 100, UK = 20, Spain =20, Sweden=20, Australia = 20 & France = 20) were surveyed. Statements that did not reach a pre-set 75% consensus were updated and retested in the second and final round. Results were analyzed by country, occupation, and years of experience in cell therapy treatments. Results A total of 203 respondents answered our survey, of which 56.2% were oncologists/hematologists, 38.4% were oncology nurses, and 5.5% were physician assistants. In the first round, 43 out of 57 statements reached consensus with 19 statements achieving >90% agreement. More than 90% of our respondents agreed that there is a benefit in remote monitoring to improve patient experience. More than 75% also agreed that heart rate/pulse rate, oxygen saturation (SpO2), and temperature were important markers to monitor (90%). Fatigue (75%), vomiting (81%), light headedness/dizziness (84%), myalgias (84%), headache (85%), chills (91%), and dyspnea (93%) were suggested as relevant to monitoring in patients at risk of CRS. The critical post-infusion monitoring period was identified as 36-48 hours post TCE dosing (85%) and 72 to 96 hours in CAR-T (85%). There was also an agreement that remote monitoring can be discontinued after two weeks post TCE infusion (77%) and four weeks post CAR-T infusion (81%). Thresholds for each vital sign and the frequency of monitoring post discharge from hospital required a second round. Further analysis of results showed that there were geographical differences in CRS monitoring practices. Conclusion The results of this study demonstrate a clear consensus on the value and benefits of RPM in monitoring for CRS, the critical window for monitoring in TCE and CAR-T, the appropriate monitoring duration post TCE infusion and CAR-T infusion, and the key symptoms and vital signs to monitor. However, disagreement remains regarding the optimal frequency of monitoring and the specific vital sign thresholds that should trigger concern for CRS. Future research should aim to address these gaps in order to contribute to the development of an effective RPM system tailored for CRS management.
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 | 0.231 | 0.161 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.007 | 0.004 |
| Science and technology studies | 0.006 | 0.006 |
| Scholarly communication | 0.005 | 0.006 |
| Open science | 0.004 | 0.019 |
| Research integrity | 0.004 | 0.005 |
| Insufficient payload (model declined to judge) | 0.005 | 0.001 |
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".