# KEYNOTE-091 (PEARLS)

Source: https://onco.cc/trials/keynote-091/  
OnCo record `keynote-091` (Trial). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Adjuvant pembrolizumab after lung cancer surgery delayed recurrence for everyone, but, unexpectedly, not for the group with the most PD-L1 on their tumour.

## Summary

KEYNOTE-091, run as PEARLS by the EORTC Lung Cancer Group and the European Thoracic Oncology Platform, recruited 1,177 of 1,955 screened patients from 196 centres in 29 countries between January 2016 and May 2020. Eligible patients had completely resected stage IB (4 cm or more), II or IIIA disease of any histology and any PD-L1 level; adjuvant chemotherapy was to be considered for stage IB and strongly recommended for stage II and IIIA. Randomisation was 1:1, triple-blind, stratified by stage, previous adjuvant chemotherapy, PD-L1 expression and region.

At the second interim analysis, disease-free survival was significantly longer with pembrolizumab in the overall population, but the co-primary endpoint in the PD-L1 tumour proportion score 50 percent or greater subgroup (168 and 165 patients) was not met. That inversion, the opposite of what PD-L1 predicts in metastatic disease, is the trial's lasting puzzle and has not been explained.

The FDA approved adjuvant pembrolizumab on 26 January 2023 for stage IB (T2a 4 cm or more), II or IIIA disease after resection and platinum-based chemotherapy, without a PD-L1 requirement. In England NICE TA1037 (5 February 2025) recommends it within its marketing authorisation for resected disease at high risk of recurrence after platinum chemotherapy, subject to the commercial arrangement.

## Fields

- Kind: Trial
- Status: mixed
- Last checked: 2026-09-25
- Also known as: PEARLS; EORTC-1416-LCG; ETOP 8-15
- Registry id: NCT02504372
- Phase: 3
- Setting: Completely resected stage IB (4 cm or more), II or IIIA non-small-cell lung cancer of any histology and any PD-L1 level, after optional adjuvant chemotherapy: pembrolizumab 200 mg every three weeks for up to 18 cycles versus placebo, with dual primary endpoints of disease-free survival in the whole population and in the PD-L1 tumour proportion score 50 percent or greater population
- Sponsor: Merck Sharp & Dohme with the European Organisation for Research and Treatment of Cancer and the European Thoracic Oncology Platform
- Enrolled: 1177
- Result: Disease-free survival significantly improved in the overall population; the co-primary endpoint in the PD-L1 50 percent or greater population was not met.
- Replication: IMpower010 showed the same direction for adjuvant atezolizumab, but there the benefit was concentrated in the PD-L1-positive stage II to IIIA subgroup, the opposite pattern.

## Sources

- ClinicalTrials.gov NCT02504372: https://clinicaltrials.gov/study/NCT02504372
- PEARLS/KEYNOTE-091 interim analysis (Lancet Oncology 2022): https://doi.org/10.1016/S1470-2045(22)00518-6
- NICE TA1037: pembrolizumab for adjuvant treatment of resected non-small-cell lung cancer: https://www.nice.org.uk/guidance/ta1037

## Connected records

- cancers: [Lung cancer (all types)](https://onco.cc/cancers/lung-cancer/), [Resectable stage I to III non-small-cell lung cancer](https://onco.cc/cancers/resectable-nsclc/)
- technologies: [Immune checkpoint inhibitors](https://onco.cc/technologies/checkpoint-inhibitor/)
- targets: [PD-1](https://onco.cc/targets/pd1/)
- drugs: [Pembrolizumab](https://onco.cc/drugs/pembrolizumab/)
- companies: [Merck & Co. (MSD)](https://onco.cc/companies/merck/)
- terms: [Neoadjuvant / adjuvant / perioperative](https://onco.cc/terms/neoadjuvant-adjuvant/)
- trials: [ADAURA](https://onco.cc/trials/adaura/), [CheckMate 816](https://onco.cc/trials/checkmate-816/), [KEYNOTE-671](https://onco.cc/trials/keynote-671/), [Study to Assess Safety and Efficacy of Atezolizumab (MPDL3280A) Compared to Best Supportive Care Following Chemotherapy in Patients With Lung Cancer [](https://onco.cc/trials/nct02486718/)

---
JSON: https://onco.cc/api/v1/entities/keynote-091.json