This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
Neo-AEGIS was a randomized phase 3 trial evaluating treatment approaches for adenocarcinoma of the oesophagus and oesophago-gastric junction.
| Feature | Neo-AEGIS |
|---|---|
| NCT ID | NCT01726452 |
| Title | NEOadjuvant Trial in Adenocarcinoma of the oEsophagus and oesophagoGastric Junction International Study (Neo-AEGIS) |
| Phase | Phase 3 |
| Status | Completed |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
2. Clinical Question
Population
Patients with adenocarcinoma of the oesophagus, oesophago-gastric junction, oesophageal tumours, or junctional tumours.
Interventions
The registry lists chemotherapy agents, radiation therapy, and perioperative treatment components including epirubicin, cisplatin or oxaliplatin, fluorouracil/capecitabine, paclitaxel, carboplatin, docetaxel, leucovorin, and radiation of 41.4 Gy in 23 fractions.
Comparator
The registry describes a randomized two-arm comparison; specific comparator assignment details are not reported in the registry information used here.
Primary question
The primary statistical question was whether randomized treatment strategies differed with respect to overall survival.
3. Trial Design
Patients were allocated randomly.
Parallel-group phase 3 design with two arms.
No masking was used.
Cancer Trials Ireland; sponsor type: network.
4. Endpoints
| Endpoint | Definition | Time frame |
|---|---|---|
| Overall survival | Calculated from date of randomisation with an event registered on the date of death from any cause. Patients lost to follow-up or without recorded death at database freeze are censored at last follow-up. | At end of trial — up to 3 years in follow up |
5. Planned Analysis
The registry identifies overall survival as the primary endpoint but does not contain posted statistical analyses or reported efficacy estimates.
Overall survival endpoints are commonly analysed using methods such as Kaplan–Meier estimation to describe survival over time, log-rank testing for comparisons between randomized groups, and Cox proportional-hazards models to estimate hazard ratios. These methods account for censoring of patients whose event status is not observed during follow-up.
The ClinicalTrials.gov record does not report posted formal statistical analyses or outcome estimates for the primary endpoint.
6. Statistical Methodology
Overall survival analysis
Overall survival is a time-to-event endpoint. The statistical analysis focuses on the distribution of time from randomisation until death while appropriately handling patients whose survival status is not observed beyond their last follow-up.
The Kaplan–Meier estimator describes the probability of remaining event-free over time while incorporating censored observations.
Hazard ratios
A Cox proportional-hazards model is commonly used in randomized survival studies to estimate the relative event rate between treatment groups. A hazard ratio compares instantaneous event rates under the model and is not the same as an absolute risk difference.
7. Statistical Methods Explained
Why is overall survival a time-to-event endpoint?
Overall survival measures the time until death and therefore requires methods that account for different follow-up durations among patients.
Why are patients censored in survival analysis?
Censoring allows patients with incomplete observed follow-up to contribute information up to the point where their outcome status was last known.
What does a hazard ratio represent?
A hazard ratio summarizes the relative instantaneous event rate between groups under a statistical model. It does not represent the probability that an individual patient will experience an event.
Why does randomization matter?
Randomization creates treatment groups intended to be comparable at baseline, allowing differences in outcomes to be interpreted within the randomized comparison.
8. Limitations
- No posted results: The registry record does not report statistical analyses or outcome estimates.
- Endpoint scope: Only the registered primary endpoint information is available in the record.
- Follow-up interpretation: Survival analyses depend on completeness of follow-up and appropriate handling of censoring.
- Generalizability: Trial eligibility criteria and study setting determine which patient populations are represented.
9. Why This Trial Matters Statistically
Neo-AEGIS illustrates several important principles in clinical trial statistics: randomized treatment comparison, phase 3 study design, parallel-group allocation, and analysis of overall survival as a censored time-to-event endpoint.
| Concept | Application |
|---|---|
| Randomization | Two-arm randomized comparison |
| Time-to-event analysis | Overall survival from randomisation |
| Censoring | Handling patients without observed death at database freeze |
| Clinical interpretation | Separating treatment effects from statistical uncertainty |