Abstract 946P
Background
Though approved for refractory RM HNSCC pts, Nivolumab (Nivo) only benefits to a minority of pts. Nivolumab efficacy depending on the activation of the patient’s immune system, we hypothesized that host genetics could predict outcome.
Methods
Germinal DNA was prospectively obtained from 90% of RM HNSCC pts of the French TopNIVO trial (NCT03226756). All pts received Nivo 3 mg/kg, Q2W. Genotyping was done using a custom panel of 165 single nucleotide polymorphisms (SNP) covering 85 immune response-related genes (Mass-Array, Agena Bioscience). Impact of genotypes on overall response rate (ORR=CR+PR) was assessed by elastic-net penalized multivariate logistic regression. A train set of 235 pts was used for building the predictive model, subsequently applied to a validation set of 75 pts. Clinical model included age, gender, performance status, type of recurrence, number of previous lines.
Results
Pts characteristics (N=310) were: 81% male, 22% >70 years, 85% PS<2, 15% PS=2, 38% had loco-regional recurrence, 28% metastatic recurrence and 33% both. ORR was 16% and was higher in women or metastatic relapse. ORR was significantly correlated with SNPs in ERCC1 (rs11615), ICOS (rs11889031), IFNL4 (rs12979860), FCGR2A (rs1801274), VEGFA (rs2010963, rs3025039), IL2RA (rs2104286), UNG (rs246079), JAK1 (rs2780890), STAT6 (rs3024971), CTLA4 (rs3087243) and IDO1 (rs3808606) genes: accuracy of the final genetic model (AUC=0.89, 95%CI 0.83-0.90) was higher than that of the clinical model (AUC=0.63, 95%CI 0.58-0.75). Performance of a risk score derived from the genetic model (including the above 12 SNPs) for predicting ORR is shown below but wasn't confirmed on the validation set. The impact of immunogenetics on OS will be presented at the meeting. Table: 946P
| Risk score (splitted by quartiles) | Train set (N=235) | Validation set (N=75) | ||
| CR+PR / Ntot | ORR % | CR+PR / Ntot | ORR % | |
| ≤ Q1 | 0 / 59 | 0 | 3 / 19 | 15.8 |
| Q1-Q2 | 3 / 59 | 5.0 | 0 / 19 | 0 |
| Q2-Q3 | 8 / 58 | 13.8 | 3 / 18 | 16.7 |
| ≥Q3 | 29 / 59 | 49.2 | 3 / 19 | 15.8 |
Conclusions
Germinal immunogenetics could improve prediction of ORR in RM HNSCC pts treated with Nivo as compared to clinical factors. External validation with more pts is needed.
Clinical trial identification
NCT03226756.
Editorial acknowledgement
Legal entity responsible for the study
Unicancer Head and Neck Group.
Funding
BMS.
Disclosure
E.B. Saada: Advisory/Consultancy, Travel/Accommodation/Expenses: BMS; Advisory/Consultancy, Travel/Accommodation/Expenses: MSD; Advisory/Consultancy, Travel/Accommodation/Expenses: AZ. A. Daste: Advisory/Consultancy, Travel/Accommodation/Expenses: BMS; Advisory/Consultancy, Travel/Accommodation/Expenses: AZ; Advisory/Consultancy: MSD; Advisory/Consultancy: Merck Serono. J. Fayette: Advisory/Consultancy, Travel/Accommodation/Expenses: BMS; Advisory/Consultancy, Travel/Accommodation/Expenses: MSD; Advisory/Consultancy, Travel/Accommodation/Expenses: AZ; Advisory/Consultancy, Travel/Accommodation/Expenses: Merck Serono. D. Cupissol: Advisory/Consultancy, Travel/Accommodation/Expenses: BMS; Advisory/Consultancy, Travel/Accommodation/Expenses: Merck Serono. J. Guigay: Advisory/Consultancy, Research grant/Funding (institution): Merck Serono; Advisory/Consultancy, Research grant/Funding (institution): BMS; Advisory/Consultancy: Innate Pharma; Advisory/Consultancy: AZ. All other authors have declared no conflicts of interest.