Abstract 804
Background
Randomised controlled trials (RCTs) in ABC have reported significantly longer progression free survival (PFS) and overall survival (OS) with the SERD, fulvestrant (500mg) in first (versus an AI) and second line settings. In contrast, none of the four RCTs comparing AIs versus Tam (SERM) individually reported a difference in OS. Of note in these trials between 6.8% - 55% of tumours were HR unknown or negative. The present meta-analysis intends therefore to restrict the comparison to HR+ tumours.
Methods
Anonymised, individual patient level data were obtained for three RCTs (EORTC, 0027 & 0030): for the remaining Femara RCT, odds ratios (ORs) with confidence intervals were obtained from the Clinical Study Report &/or publications. Details of the studies, AI used & patient numbers are shown (Table).
Results
ORs for clinical benefit rate (CBR), duration of CB (DoCB), PFS & OS are shown for each individual study & for combined trials grouped 1-4 and 1-3 (with & without Femara). P-values for CBR with ORs>1 favour AIs; P-values for DoCB, PFS & OS with ORs <1 favour AIs.Table: 47P
| Treatment | Study | Total no. of pts | No of HR+ pts | CBR | DoCB | PFS | OS |
|---|---|---|---|---|---|---|---|
| Tam vs | (n) | (n) | OR | OR | OR | OR | |
| Exe (1) | EORTC | 371 | 346 | 1.85 | 1.02 | 0.86 | 1.06 |
| Ana (2) | 0027 | 665 | 298 | 1.12 | 0.73 | 0.84 | 1.77 |
| Ana (3) | 0030 | 353 | 317 | 1.71 | 0.99 | 0.77 | 1.0 |
| Let (4) | Femara | 907 | 599 | 1.62 | 0.81 | 0.70 | 1.06 |
| Subtotal (1-4) | 2296 | 1560 | 1.56 | 0.88 | 0.82 | 1.1 | |
| p-value | <0.001 | 0.08 | 0.007 | 0.18 | |||
| Subtotal (1-3) | 1389 | 961 | 1.51 | 0.93 | 1.2 | ||
| p-value | 0.008 | 0.46 | 0.26 |
Conclusions
AIs produce significantly increased CBR than Tam. AIs result in a significantly longer PFS than Tam. The meta-analysis did not show improved OS for AIs compared to Tam. This finding is different from the SERD, fulvestrant (500mg), where significant improvement in CBR & PFS total have been associated with significant improvement in OS.
Editorial acknowledgement
Clinical trial identification
Legal entity responsible for the study
University of Nottingham.
Funding
University of Nottingham.
Disclosure
J. Robertson: Consulting/non-CME fees: Novartis, AstraZeneca, Cullinan Oncology; Contracted research: Novartis, AstraZeneca - all payments to author\'s institution; Stock: Oncimmune, Carrick Therapeutics; Expert testimony: AstraZeneca. J. Lichfield: Employee, Share options: AstraZeneca. All other authors have declared no conflicts of interest.