Background: Data for this analysis is provided by pooled data from 2 randomized phase III trials (046 and 048) comparing Ixa + Cape versus cape alone in women with MBC refractory to or pretreated with anthracyclines (A) and taxanes (T). Pre-specified sub-group analyses (SGA) were undertaken in MBC patients with PPF including triple negative phenotype (TN), symptomatic KPS 70-80 (KPS), and post adjuvant rapid relapse (PARR, recurrence ≤ 12 months after adjuvant A and T). For each SGA, a clinically meaningful PFS increase was seen with Ixa + Cape (Table 1), consistent with the overall population (OP). A significant overall survival (OS) improvement was not observed in the OP. In the SGA, the OS results were as follows: TN (10.3 vs 9.0 months, p=0.1802c), PARR (15.1 vs 12.5 months, p=0.2081c) and KPS (12.3 vs 9.5 months, p=0.0015c). Across SGA, treatment related adverse events (AEs) were more common for Ixa + Cape, but similar to the OP. A Q-TWiST analysis was undertaken for each SGA to evaluate the trade off between toxicity and PFS.Methods: The area under the OS curves for each SGA was partitioned into 3 health states: 1) Toxicity (TOX) - time the subject spent prior to progression with grade 3/4 treatment toxicity, 2) TWiST - time spent without grade 3/4 AEs prior to progression, 3) Relapse (REL) - time from progression to death or end of follow-up. Utility weights (UW) between 0 and 1 were assigned to the TOX and REL health states with the base case analysis using 0.5 for TOX and REL. The TWiST UW was assumed to equal 1.0 as the most favorable health state achievable. The utility weighted sum of the mean health state durations was derived and treatment comparisons were made. Sensitivity analyses were performed varying the UW of the TOX and REL states between 0 and 1 and calculating the differences in the weighted sum for all combinations.Results: For each SGA, all UW combinations for REL and TOX were associated with greater observed quality adjusted survival (QAS) durations for Ixa + Cape compared to Cape. The base case showed improvements in mean QAS in favor of Ixa + Cape (Table 1), with differences ranging from 4.9 to 9.6 weeks and the PARR group deriving the largest benefit.

Median PFS and Mean QAS in Pooled PPF MBC Sub-Groups*

  Median PFSa, months Mean QASb, weeksa 
Patient Sub-Groups Ixa + Cape Cape HR (95% CI) P value c Ixa + Cape Cape P value c 
TN (N = 191) 4.2 (N = 208) 1.7 0.63 (0.52 -0.77) <0.0001 (N = 213) 35.3 (N = 230) 30.4 0.0031 
KPS (N = 268) 4.6 (N = 257) 3.1 0.76 (0.64 - 0.90) 0.0021 (N = 314) 39.8 (N = 292) 33.4 <0.0001 
PARR (N = 123) 5.6 (N = 111) 2.8 0.58 (0.45 - 0.76) < 0.0001 (N = 149) 45.6 (N = 144) 36.0 0.0007 
  Median PFSa, months Mean QASb, weeksa 
Patient Sub-Groups Ixa + Cape Cape HR (95% CI) P value c Ixa + Cape Cape P value c 
TN (N = 191) 4.2 (N = 208) 1.7 0.63 (0.52 -0.77) <0.0001 (N = 213) 35.3 (N = 230) 30.4 0.0031 
KPS (N = 268) 4.6 (N = 257) 3.1 0.76 (0.64 - 0.90) 0.0021 (N = 314) 39.8 (N = 292) 33.4 <0.0001 
PARR (N = 123) 5.6 (N = 111) 2.8 0.58 (0.45 - 0.76) < 0.0001 (N = 149) 45.6 (N = 144) 36.0 0.0007 

*Pooled data from trials 046 and 048 a PFS computed on all randomized 046 patients and 048 patients randomized to measurable disease stratum. b QAS computed on all randomized patients from both studies c Not adjusted for multiple comparisons

Conclusions: This Q-TWIST analysis supports the positive benefit:risk of Ixa + Cap in MBC patients with poor prognostic features, previously treated with an A and T.

Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5051.