A risk stratification model for toxicities in phase 1 immuno-oncology (P1-IO) trials.
Bibliographic record
Abstract
2648 Background: Despite an exponential increase in the number of potential targets in the immuno-oncology (IO) field, lack of risk models to predict toxicities remains a challenge in early drug development. We proposed a risk stratification strategy and investigated whether different IO classes may be associated with incremental risk of toxicities. Methods: A systematic search for IO studies from 01/2014 to 10/2020 was conducted. Among the 12053 abstracts screened, 254 reporting phase I-IO trials were selected. Type of IO treatment (IO monotherapy [mono] vs IO combination [comb]), therapeutic class (e.g. IO, molecularly targeted agents [MTA]), and dose escalation method (rule-based, model-based and model-assisted) were collected. A risk scoring model was developed after expert consensus: treatment-related deaths (1:yes, 0:no), incidence of G3/G4 treatment related adverse events (TRAE) or treatment emergent adverse events (TEAE) (0 if 1-29%,1 if ≥30-49%, 2 if ≥50%), incidence of ≥G2 cytokine release syndrome (1:yes, 0:no), incidence of ≥G2 encephalopathy (1:yes 0:no) and incidence of dose-limiting toxicity (DLT) (0:no, 1:lab, 2:clinical). Risk categories were defined by summing all points (0 = low, 1-2 = intermediate, 3+ = high) and were correlated with type of IO treatment, therapeutic class and dose escalation method. Results: Of 254 P1-IO trials reviewed, 228 (90%) were scorable, 26 were not (25 due to lack of AE data). Up to 10/26 (38%) of non-scorable studies were cell therapies. Among the 228 scorable studies, 120 (53%) scored 0, 65 (28%) scored 1-2, 43 (19%) scored 3+; 24 (11%) and 125 (55%) did not provide no. of pts with G3/G4 TRAEs or TEAEs respectively. A significant association was observed between risk categories and therapeutic class (p<0.001) (see table). Additionally, IO-MTA and IO-IO were both associated with an increased risk of toxicity as compared to IO-mono (OR=3.91 (95%CI 1.7-9.2), p=0.002) and (OR=2.79 (95%CI 1.2-6.5), p=0.02) respectively. There was no association between dose escalation method and risk of toxicity (p=0.95), but 182 (92%) used rule-based methods. Conclusions: Our results suggest that different IO classes are associated with different risks of toxicity. This risk classification strategy may guide future clinical trial design. Additionally, standards for reporting toxicities in P1-IO trials are urgently needed.[Table: see text]
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 | 0.041 | 0.072 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.009 |
| Bibliometrics | 0.006 | 0.003 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 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".