Image Analysis for Breast Cancer Prognosis

Abstract

Heterogeneity for biomarkers in a tissue sample can be calculated. A heterogeneity score can be combined with an immunohistochemistry combination score to provide breast cancer recurrence prognosis. Heterogeneity can be based on percent positivity determinations for a plurality of biomarkers according to how many cells in the sample stain positive. An immunohistochemistry combination score can be calculated. An imaging tool can support a digital pathologist workflow that includes designating fields of view in an image of the tissue sample. Based on the fields of view, a heterogeneity metric can be calculated and combined with an immunohistochemistry combination score to generate a breast cancer recurrence prognosis score.

1 - 15 . (canceled) 
     
     
         16 . An image processing system comprising:
 a slide image processing tool operable to receive a plurality of slide images depicting protein expression for respective biomarkers in a breast cancer sample from a subject;   wherein the slide image processing tool is operable to further receive fields of view within the slide images;   wherein the slide image processing tool is operable to calculate an immunohistochemistry combination score based on the slide images and fields of view within the slide images;   wherein the slide image processing tool is operable to calculate one or more heterogeneity scores based on the slide images and selections of fields of view within the slide images; and   a prognosis tool operable to accept the immunohistochemistry combination score and the one or more heterogeneity scores as input and output an indication of whether cancer is likely to recur in the subject.   
     
     
         17 - 33 . (canceled) 
     
     
         34 . A computer-implemented method comprising:
 for ER, receiving a plurality of digital fields of view in an image depicting a breast cancer sample detectably labeled with an antibody for ER;   for PR, receiving a plurality of digital fields of view in an image depicting a breast cancer sample detectably labeled with an antibody for ER;   for Ki-67, receiving a plurality of digital fields of view in an image depicting a breast cancer sample detectably labeled with an antibody for ER;   for HER2, receiving a plurality of digital fields of view in an image depicting a breast cancer sample detectably labeled with an antibody for ER;   based on the received plurality of digital fields of view for ER, calculating an H-score for ER;   based on the received plurality of digital fields of view for PR, calculating a percent positivity for PR;   based on the received plurality of digital fields of view for Ki-67, calculating a percent positivity for Ki-67;   based on the received plurality of digital fields of view for HER2, calculating a binned score for HER2; and   combining the H-score for ER, the percent positivity for PR, the percent positivity for Ki-67, and the binned score for HER2 into an immunohistochemistry combination score.   
     
     
         35 . The method of  claim 34 , wherein the combining the H-score for ER, the percent positivity for PR, the percent positivity for Ki-67, and the binned score for HER2 into an immunohistochemistry combination score comprises converting the binned score for HER2 into a binary score for HER2. 
     
     
         36 . The method of  claim 34 , wherein the calculating of the binned score for HER2 comprises: performing analysis on the digital fields of view for HER2 collectively. 
     
     
         37 . The method of  claim 34 , wherein the calculating of the binned score for HER2 comprises: for a plurality of nuclei appearing in the digital fields of view for HER2, determining whether cell membrane surrounding the nuclei are completely stained. 
     
     
         38 . The method of  claim 34 , wherein the calculating of the binned score for HER2 comprises: for a plurality of nuclei appearing in the digital fields of view for HER2, determining intensity of staining of cell membrane surrounding respective nuclei. 
     
     
         39 - 52 . (canceled) 
     
     
         53 . The method of  claim 34 , further comprising computing one or more heterogeneity scores. 
     
     
         54 . The method of  claim 53 , wherein a first heterogeneity score is computed for ER protein expression; and wherein a second heterogeneity score is computed for PR protein expression. 
     
     
         55 . The method of  claim 54 , further comprising at least a third heterogeneity score. 
     
     
         56 . The method of  claim 54 , further comprising computing a heterogeneity combination score derived from the first and second computer heterogeneity scores. 
     
     
         57 . The method of  claim 53 , further comprising computing a breast cancer recurrence prognosis score by combining the one or more heterogeneity scores with the immunohistochemistry combination score. 
     
     
         58 . The method of  claim 56 , further comprising computing a breast cancer recurrence prognosis score by combining the heterogeneity combination score with the immunohistochemistry combination score. 
     
     
         59 . The method of  claim 34 , wherein the breast cancer sample is derived from a human subject diagnosed with breast cancer. 
     
     
         60 . The method of  claim 59 , further comprising determining whether the human subject diagnosed with breast cancer is at a high risk or at a low risk of breast cancer recurrence based on the immunohistochemistry combination score. 
     
     
         61 . The method of claim  43 , wherein the breast cancer sample is derived from a human subject diagnosed with breast cancer, and wherein the method further comprises determining whether the human subject diagnosed with breast cancer is at a high risk or at a low risk of breast cancer recurrence based on the breast cancer recurrence prognosis score. 
     
     
         62 . The method of  claim 57 , wherein the breast cancer sample is derived from a human subject diagnosed with breast cancer, and wherein the method further comprises determining whether the human subject diagnosed with breast cancer is at a high risk or at a low risk of breast cancer recurrence based on the breast cancer recurrence prognosis score. 
     
     
         63 . The method of  claim 62 , further comprising administering to the human patient an appropriate for breast cancer treatment regimen if the breast cancer recurrence prognosis score is greater than the predetermined threshold value. 
     
     
         64 . The method of  claim 34 , wherein the IHC combinations score is an IHC4 score. 
     
     
         65 . The method of  claim 56 , wherein the first and second heterogeneity scores are normalized prior to computing the heterogeneity combination score.