Abstract 58P
Background
To compare the relative survival benefits & toxicities among treatments for non-driver mutated metastatic NSCLC (mNSCLC).
Methods
RCTs since inception, to June 1st 2021, were systematically searched from databases, trial registries & annual meeting abstracts. All trials comparing survival & toxicity between at least two of the following treatments were included: chemotherapy (CT) (single/multi-agent); PD1, PDL1 or CTLA4 immune checkpoint inhibitors (ICI) (single/multi-agent) (with/without CT), &; radiotherapy followed by either chemotherapy (RT+CT) or immune checkpoint inhibitors. The primary outcomes were risk of death at 1 year, risk of progression at 6 months & 1 year, & overall grade 3 or higher (G3+) toxicities. A Bayesian network meta-analysis using a random-effects model & empirical Markov Chain Monte Carlo simulation of multiple interventions was used. Results were expressed as risk ratios (RR)[95% Credible Interval (CrI)] & ranked using SUCRA scores. All CT regimens were merged into a single category & ICIs were divided into two categories (PD1 or PDL1; CTLA-4). Local & global consistency and sensitivity analyses were performed on the network (after splitting treatments). All treatments satisfied the transitivity assumption. Risk-of-bias (RoB) & confidence were assessed with Cochrane RoB & CiNeMA tools.
Results
30 RCTs (n = 14904; 36 possible comparisons; 12 direct comparisons) comparing 9 treatments were included. The combination of RT+PD1/PDL1 was ranked highest & demonstrated the lowest risk of death [RR = 0.47 (0.30–0.72)][SUCRA = 95%], progression [6 m RR = 0.39 (0.23–0.63)[SUCRA = 96%]; 1 yr RR = 0.74 (0.59–0.95)][SUCRA = 87%], & G3+ toxicities [RR = 0.74 (0.57–0.96)][SUCRA = 81%]. Results are shown below [RR (95% CrI)]. All sensitivity analyses favored RT+PD1/PDL1 over other treatments.
Table: 58PResults
| Death-1yr | Progression-6m | Progression-1yr | G3+Toxicity | |
|---|---|---|---|---|
| CT (ref) | ||||
| CT + RT | 0.56 (0.3–0.92) | 0.47 (0.26–0.77) | 0.77 (0.55–1) | 0.77 (0.55–0.99) |
| CTLA4 | 0.93 (0.69–1.21) | 1.07 (0.84–1.35) | 0.97 (0.88–1.05) | 0.85 (0.58–1.16) |
| CTLA4 + CT | 0.85 (0.60–1.2) | 0.89 (0.65–1.23) | 1 (0.93–1.08) | 0.82 (0.68–0.96) |
| PD1/PDL1 | 0.84 (0.79–0.9) | 0.91 (0.85–0.97) | 0.9 (0.89–0.92) | 0.9 (0.87–0.93) |
| PD1/PDL1 + CT | 0.84 (0.77–0.92) | 0.7 (0.63–0.78) | 0.83 (0.8–0.86) | 0.85 (0.81–0.9) |
| PD1/PDL1 + CTLA4 | 0.86 (0.75–0.99) | 0.9 (0.78–1.04) | 0.86 (0.82–0.9) | 0.87 (0.8–0.94) |
| PD1/PDL1 + CTLA4 + CT | 0.7 (0.54–0.9) | 0.7 (0.57, 0.97) | 0.8 (0.77–0.91) | 0.8 (0.73–0.96) |
| PD1/PDL1 + RT | 0.47 (0.3–0.72) | 0.39 (0.23–0.63) | 0.74 (0.59–0.95) | 0.74 (0.57–0.96) |
Conclusions
The results suggest that combining RT with PD1/PDL1 inhibitors could improve outcomes over other treatments in non-driver mutated metastatic NSCLC.
Legal entity responsible for the study
K. S. Chufal.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.