Evaluation of the Oncotype Dx recurrence score (RS) in hormone-positive ipsilateral breast tumor recurrences.

Abstract

e11032 Background: IBTR (ipsilateral breast tumor recurrence) is considered a marker of risk for metastasis. Most IBTRs occur in the vicinity of the index tumor while elsewhere recurrences may represent new primaries. Oncotype DX®, 21-gene expression assay, quantifies a recurrence score (RS) in node negative, ER+ breast cancer. This modality can guide treatment decisions and has been extended to select node positive patients. Concordance of RS between the primary and IBTR is presented here. Methods: Archival blocks corresponding of 11 patients were obtained but only 8 were evaluable. The blocks were submitted using a blinded identification scheme to link the primary and recurrent tumor blocks. All were ER positive breast cancers. Six were N0, one N1, and one NX. Some assay results were preexisting. RS and single gene expression profiles for ER, PR and HER2 were compared. Scatterplots and descriptive statistics were produced for this exploratory hypothesis-generating (but not hypothesis-testing) study. Results: The time interval between primary and recurrence ranged from 12 to 216 months (mean 73.5 months). Among the 8 paired samples, RS’s were similar in 5 and different in 3. In the latter, RS shifted from 2 to 29, 15 to 23 and 25 to 14. These IBTRs occurred in a different quadrant, were simultaneous with axillary recurrence, or appeared after trastuzumab therapy. Overall, the Pearson correlation coefficient was 0.68 (95% CI -0.09, 0.93). Only one case shifted from ER+ to ER-; two cases from PR+ to PR- and one from PR- to PR+. No change in HER2 status was noted. The paired differences for ER were -0.01 (95% CI -1.09, 1.07), PR -0.74 ( 95% CI -2.43, 0.96) and HER2 -0.51 (95% CI -1.20, 0.17). The Pearson correlation coefficients and 95% CI were ER 0.72 (-0.02, 0.94); PR 0.61 (-0.22, 0.91); and HER2 0.89 (0.45, 0.98) respectively. Conclusions: The RS of the IBTR was similar to primary breast cancer in 5 of 8 cases in this exploratory pilot study. Hormone receptor status changed in only one of 8 cases. The clinical significance of these observations and the utility of these results in guiding therapeutic strategies is unknown.