BCL-2 expression aids in the immunohistochemical prediction of the Oncotype DX breast cancer recurrence score

Abstract

Background: The development of molecular techniques to estimate the risk of breast cancer recurrence has been a significant addition to the suite of tools available to pathologists and breast oncologists. It has previously been shown that immunohistochemistry can provide a surrogate measure of tumor recurrence risk, effectively providing a less expensive and more rapid estimate of risk without the need for send-out. However, concordance between gene expression-based and immunohistochemistry-based approaches has been modest, making it difficult to determine when one approach can serve as an adequate substitute for the other. We investigated whether immunohistochemistry-based methods can be augmented to provide a useful therapeutic indicator of risk.

Methods: We studied whether the Oncotype DX breast cancer recurrence score can be predicted from routinely acquired immunohistochemistry of breast tumor histology. We examined the effects of two modifications to conventional scoring measures based on ER, PR, Ki-67, and Her2 expression. First, we tested a mathematical transformation that produces a more diagnostic-relevant representation of the staining attributes of these markers. Second, we considered the expression of BCL-2, a complex involved in regulating apoptosis, as an additional prognostic marker.

Results: We found that the mathematical transformation improved concordance rates over the conventional scoring model. By establishing a measure of prediction certainty, we discovered that the difference in concordance between methods was even greater among the most certain cases in the sample, demonstrating the utility of an accompanying measure of prediction certainty. Including BCL-2 expression in the scoring model increased the number of breast cancer cases in the cohort that were considered high certainty, effectively expanding the applicability of this technique to a greater proportion of patients.

Conclusions: Our results demonstrate an improvement in concordance between immunohistochemistry-based and gene expression-based methods to predict breast cancer recurrence risk following two simple modifications to the conventional scoring model.

Keywords: Computer-assisted diagnosis; Digital pathology; Prognostic markers; Staining.

Fig. 1

Transformation of staining attributes. a The ER and PR percent positive cells metric was transformed to a diagnostic score using logistic transformation. As indicated by the dotted lines, the curve is half maximum at the diagnostic cut-off of 10%, producing a score of 0.5. Gray points indicate the relationship between H-score and percent positive cells using simulations. b The logistic transformation is applied to the Ki-67% positive cells metric using a diagnostic cut-off of 14% to produce a diagnostic score. c Her2 score is transformed to a diagnostic score by dividing its value by 3, maintaining a linear relationship

Fig. 2

Representative examples of BCL-2 staining.We selected representative images from three cases that exhibited low (left panel), intermediate (center panel), and high (right panel) intensity staining for BCL-2. The corresponding H-scores are shown

Fig. 3

Estimation of RS from IHC. The RS for each sample is represented on the x-axis. The IHC scores generated by linear regression, following data transformation and the addition of BCL-2, are represented on the y-axis. Dotted lines represent the categorical boundaries that distinguish low, intermediate, and high risks of recurrence. The category to which each sample is assigned based on IHC score is indicated by color; blue = low, green = intermediate, red = high

Fig. 4

Concordance of IHC-based methods with RS. Concordance rates were computed on 158 cases using cross-validation for each of the following methods tested (starting with the left-most bar): Magee Score #3 with the coefficients and variables described in Klein, et al. [24]; Magee Score #3 with coefficients recomputed based on our cohort; IHC Score using ER, PR, and Ki-67% positive staining, and Her2 score; IHC Score using the logistic transformation applied to ER, PR, and Ki-67; IHC Score after inclusion of BCL-2 H-score; IHC Score using both the logistic transformation and the inclusion of BCL-2 H-score

Fig. 5

Cumulative concordance rate scales with certainty. Concordance rates were computed based only on the samples with certainty values greater than the value indicated on the x-axis. The proportion of cases used to compute the concordance value is shown on the y-axis to the left, where 1 indicates that all 158 samples were used. Standard deviation of the rates based on 10,000 iterations are represented by the shaded regions

Fig. 6

Prediction of chemotherapy effect from IHC. Chemotherapy effect was calculated from RS and plotted against IHC score for each case. Chemotherapy effect was estimated from the model presented by Paik, et al. [13] by subtracting the 10-year likelihood of recurrence at a given RS for patients treated with Tamoxifen + chemotherapy from those treated with Tamoxifen alone. The category to which each sample is assigned based on IHC score is indicated by color; blue = low, green = intermediate, red = high