Abstract 2120
Background
In patients (pts) with NSCLC, the efficacy of atezo (anti-PDL1) correlates with PD-L1 expression on tumor cells (TC) and tumor-infiltrating immune cells (IC). We further examined the association between tumor mutation load (ML) and efficacy of atezo in pts with NSCLC.
Methods
Pretreatment tumor specimens from 454 2L+ NSCLC pts treated on three Ph 2 trials of atezo monotherapy (POPLAR, a trial comparing atezo vs docetaxel (doc); BIRCH and FIR, single-arm studies in PD-L1–selected pts) were available for targeted genetic sequencing using the FM1 panel of 315 cancer-related genes. ML was quantified for each sample and efficacy was assessed in groups defined by 75th (high), 50th (median) and 25th (low) percentile of the study-specific ML. TIL infiltration was assessed by H&E staining. Teff gene expression was assessed with iChip. Atezo efficacy was examined at the following data cutoffs: POPLAR, May 8, 2015; BIRCH, May 28, 2015; FIR, Jan 7, 2015.
Results
Across all samples, the median ML was 9.9/MB (range 0-444.1, 25th-75th percentile 6.3-16.6). OS, PFS and ORR were improved in pts with increased ML treated with atezo in both unselected pts (POPLAR) and PD-L1–selected pts (BIRCH, FIR; see Table). ML appeared to predict atezo efficacy independently of PD-L1 status. ML was not associated with efficacy in pts treated with doc in POPLAR. Associations of ML with PD-L1 expression on TC and IC, TIL infiltration and Teff gene expression will be presented.
Conclusions
We demonstrated for the first time that increased tumor ML assessed by the FM1 targeted sequencing panel is associated with improved outcomes with atezo in 2L+ NSCLC. The association between ML and atezo efficacy was seen in both unselected and PD-L1-selected NSCLC pts. ML did not appear to be prognostic in pts treated with doc. Therefore, in addition to PD-L1, ML by FM1 may be an independent predictor of improved responsiveness to atezo in 2L+ NSCLC.
Clinical efficacy of atezo monotherapy in ML subgroups
| Unselected 2L+ NSCLC pts (atezo- vs doc- treated pts, POPLAR) | ||||
|---|---|---|---|---|
| ML cutoff | BEP n = 87 | Low n = 66 | Median n = 46 | High n = 23 |
| OS (atezo vs doc), HRa (95% CI) | 0.58 (0.31, 1.07) | 0.55 (0.27, 1.10) | 0.51 (0.22, 1.17) | 0.22 (0.05, 1.03) |
| PFS (atezo vs doc), HRa (95% CI) | 0.88 (0.55, 1.42) | 0.69 (0.40, 1.21) | 0.49 (0.25, 0.97) | 0.33 (0.12, 0.96) |
| ORR (atezo vs doc, per RECIST v1.1) | 15% vs 15% | 16% vs 12% | 21% vs 11% | 30% vs 8% |
| PD-L1–selected 2L+ NSCLC pts (atezo-only treated pts, BIRCH + FIR) | ||||
| ML cutoff | BEP n = 367 | Low n = 367 | Median n = 367 | High n = 367 |
| OS (above vs below the cutoff), HRb (95% CI) | NA | 1.07 (0.72, 1.59) | 0.91 (0.64, 1.29) | 0.66 (0.42, 1.03) |
| PFS (above vs below the cutoff), HRb (95% CI) | NA | 0.82 (0.63, 1.08) | 0.66 (0.52, 0.85) | 0.49 (0.35, 0.67) |
| ORR (above vs below the cutoff; per RECIST v1.1) | 18% | 19% vs 10% | 22% vs 13% | 26% vs 15% |
BEP, biomarker-evaluable population; Doc, docetaxel; HR, hazard ratio; ML, mutational load. aHR comparing efficacy-evaluable pts treated with atezo vs doc at or above the ML cutoff. bHR comparing efficacy-evaluable pts with tumors at or above the ML cutoff vs patients below the ML cutoff.
Clinical trial identification
N/A
Legal entity responsible for the study
F. Hoffmann-La Roche Ltd
Funding
F. Hoffmann-La Roche Ltd
Disclosure
M. Kowanetz, A. Sandler, G. Hampton, L. Amler, P.S. Hegde, D.S. Shames, C. Cummings, V. Leveque, S. Flynn, S. Mocci, G. Shankar, D. Waterkamp: Genentech employee. W. Zou: Roche employee, Roche research funding. N. Rizvi: Consulting: Merck, AZ, Roche, Novartis, Lilly Co-founder and shareholder: Gritstone Oncology. A.I. Spira: Roche funded clinical trial. G.M. Frampton: Foundation Medicine employee and Shareholder. R. Funke: Genentech employee with stocks. M. Ballinger: Genentech employee and stock. M. Hellmann: Consultancy: Genentech, Merck, AZ, BMS, Neon Research support: Genentech, BMS.