Abstract 1708P
Background
In the COVID-19 era, a simplified risk assessment algorithm for CIN might optimize resource utilization. Since Gr4N toxicity is frequently measured and reported in clinical trial results, Gr4N frequency might predict adverse clinical outcomes for patients who develop CIN.
Methods
We performed a meta-analysis combining data from published CIN clinical trials and the clinical trial datasets from the novel CIN-preventive agent plinabulin. Literature search terms included Grade 4 neutropenia, severe neutropenia, chemotherapy-induced neutropenia, febrile neutropenia (FN), infection, absolute neutrophil count (ANC) nadir, and duration of severe neutropenia (DSN). The search engines were NCBI, PubMed, and Google Scholar databases. To evaluate the validity of the dataset, we first evaluated correlations between the historically accepted CIN endpoints FN rate, DSN, and ANC nadir. Next, we correlated Gr4N frequency with FN rate, DSN, ANC nadir, hospitalization rate, and infection rate.
Results
Correlations of exponential equations between FN rate, DSN, and ANC nadir were in strong and statistically significant (p<0.0001) agreement with each other. Gr4N frequency had high exponential equation fit values with each: DSN (R=0.647; n=4864; p<0.0001), FN rate (R=0.44; n=4311; p<0.0001), ANC nadir (R=-0.771; n=2623; p<0.0001), hospitalization rate (R=0.686; n=850; p<0.0001), and infection rate (R=0.429; n=2042; p<0.0001). The criteria we used to satisfy the Gr4N threshold level to depict low risk for CIN outcomes was a) DSN of <1 day (considered to be not clinically significant), and b) FN risk <10% (as defined by NCCN). The 65% Gr4N threshold met these criteria, as summarized below. For all CIN variables correlated with Gr4N, the curve stayed fairly flat between Gr4N of 0 to 65%, with an exponential rise after Gr4N≥65%. Table: 1708P
| Gr4N <65% Mean (95%CI), N | Gr4N ≥65% Mean (95%CI), N | P-value | |
| DSN | 0.99 (0.97, 1.013), 2659 | 2.14 (2.084, 2.19), 2205 | <0.0001 |
| FN rate | 4.50 (4.32, 4.68), 2659 | 11.78 (11.30, 12.26), 1652 | <0.0001 |
| ANC nadir | 1.11 (1.082, 1.15), 1586 | 0.33 (0.31, 0.35), 1037 | <0.0001 |
| Hospitalization rate | 4.17 (3.70, 4.63), 435 | 15.93 (15.16, 16.70), 415 | <0.0001 |
| Infection rate | 4.38 (4.10, 4.66), 806 | 11.41 (10.39, 12.43), 1236 | <0.0001 |
Conclusions
Gr4N is a valid binary predictor of CIN outcomes and a 65% Gr4N threshold depicts low vs. high CIN outcome risk.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
BeyondSpring Pharmaceuticals Inc.
Funding
BeyondSpring Pharmaceuticals Inc.
Disclosure
R. Mohanlal: Financial Interests, Personal and Institutional, Stocks/Shares: Beyond Spring Pharmaceuticals; Financial Interests, Personal and Institutional, Full or part-time Employment: Beyond Spring Pharmaceuticals; Financial Interests, Personal and Institutional, Leadership Role: Beyond Spring Pharmaceuticals. L. Huang: Financial Interests, Personal and Institutional, Ownership Interest: Beyond Spring Pharmaceuticals; Financial Interests, Personal and Institutional, Stocks/Shares: Beyond Spring Pharmaceuticals; Financial Interests, Personal and Institutional, Funding: Beyond Spring Pharmaceuticals; Financial Interests, Personal and Institutional, Royalties: Beyond Spring Pharmaceuticals. D. Blayney: Financial Interests, Institutional, Funding: Beyond Spring Pharmaceuticals. All other authors have declared no conflicts of interest.