- Design
- PRISMA 2020 systematic review and meta-analysis of phase 3 randomised trials, fixed-effect primary analysis with RoB 2 and GRADE
- Population
- Six phase 3 trials, 4,137 patients randomised across parent reports, with advanced or metastatic oesophageal squamous cell carcinoma
- Primary outcome
- Overall survival with first-line PD-1 inhibitor-based therapy versus chemotherapy alone
- Effect
- Pooled HR 0.68 (95% CI 0.63-0.74), I-squared 0%; larger effect in PD-L1-high disease but not formally poolable
Six phase 3 trials have tested first-line PD-1 blockade added to chemotherapy in advanced oesophageal squamous cell carcinoma - KEYNOTE-590 (squamous subgroup), CheckMate-648 (the nivolumab plus chemotherapy arm), ESCORT-1st, JUPITER-06, ORIENT-15 and RATIONALE-306. This PRISMA-guided synthesis pools them, with RoB 2 risk-of-bias assessment and GRADE.
Overall survival improved substantially: pooled hazard ratio 0.68 (95% CI 0.63-0.74), with I-squared of 0%, meaning the six trials agreed closely. Grade 3 or worse treatment-related adverse events were common in both arms; immune-related toxicity was more frequent with PD-1 inhibitors and needs active monitoring rather than a warning.
The unresolved question is biomarker selection, and the authors are careful about it. Effect appeared larger in PD-L1-high disease, but the trials measured PD-L1 with different assays and reported subgroups differently, so it could not be pooled formally. The magnitude of benefit in PD-L1-low or negative disease therefore remains uncertain - which is precisely the population where funders and patients most need a number.
So the practical position: chemoimmunotherapy is a contemporary first-line standard in advanced oesophageal squamous carcinoma, and the survival evidence for the population as a whole is strong and consistent. Where PD-L1 is low or negative and the drug is self-funded, as it usually is in Indian practice, the honest answer to a patient asking what it will do for them specifically is that the trials do not say. Assay heterogeneity means a combined positive score from one laboratory is not interchangeable with another's, so do not treat a threshold as a hard eligibility line.
- Offer first-line chemoimmunotherapy in advanced oesophageal squamous cell carcinoma where it is available
- Record which PD-L1 assay and scoring system was used - they are not interchangeable across trials
- Discuss cost honestly where PD-L1 is low or negative; the benefit there is not quantified
- Set up immune-related toxicity monitoring at the start: thyroid, liver, colitis, pneumonitis
- Chemotherapy backbones differed between trials, so local practice may not match any single one
The statistics, in plain English
A pooled hazard ratio of 0.68 means roughly a third lower rate of death at any moment, and the narrow interval (0.63-0.74) reflects over 4,000 randomised patients. An I-squared of 0% says the trials agreed - reassuring, though with only six studies the statistic is unreliable, which the authors state. Note two of the six contributed only part of their population: KEYNOTE-590's squamous subgroup and CheckMate-648's nivolumab arm. Subgroup data that cannot be harmonised is a real limitation, not a technical footnote: it means the biomarker question was not answered rather than answered negatively.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for oncology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free