Real-world evidence of diagnostic testing and treatment patterns in U.S. breast cancer patients with implications for treatment biomarkers from RNA-sequencing data

1. A computer-implemented method of diagnosing HER2 status for a patient, the method comprising:

obtaining human epidermal growth factor receptor 2 (HER2) status for a specimen from analysis of immunohistochemistry (IHC) image from a first sample of the patient by applying the IHC image to a trained IHC classification model;

obtaining HER2 status for a specimen from analysis of fluorescence in-situ hybridization (FISH) image from a second sample of the patient by applying the FISH image to a trained FISH classification model;

identifying discordant HER2 status result between the HER2 status obtained from the IHC image and the HER2 status obtained from the FISH image; and

in response to the identification of discordant HER2 status, diagnosing HER2 status based on at least gene expression data from a third sample of the patient, wherein the trained IHC classification model and the trained FISH classification model are machine learning models.

2. The computer-implemented method of claim 1 , further comprising generating a HER2 discordance status report indicating biologic pathways in gene expression data.

3. The computer-implemented method of claim 1 , further comprising generating a HER2 discordance status report including an indication of a model of molecular subtype gene expression used to identify the discordant HER2.

4. The computer-implemented method of claim 3 , wherein the model of molecular subtype gene expression comprises a linear gene expression model.

5. The computer-implemented method of claim 3 , wherein the model of molecular subtype gene expression comprises a pathway gene expression model.

6. The computer-implemented method of claim 5 , further comprising generating the HER2 discordance status report including a listing of pathways identified by the pathway gene expression model.

7. The computer-implemented method of claim 3 , wherein the model of molecular subtype gene expression comprises a multiple gene linear regression gene expression model.

8. The computer-implemented method of claim 7 , further comprising generating the HER2 discordance status report including a listing of genes identified by the multiple gene linear regression gene expression model.

9. The computer-implemented method of claim 1 , further comprising generating a HER2 discordance status report indicating a molecular subtype determined from the gene expression data.

10. The computer-implemented method of claim 9 , wherein the molecular subtype is a HR+subtype, a HR+/HER2+subtype, a HR−/HER2+subtype, or a HER2-subtype.

11. The computer-implemented method of claim 9 , wherein the molecular subtype is a ER+, ER−, PR+, or PR−.

12. The computer-implemented method of claim 9 , wherein the molecular subtype is a triple negative subtype.

13. The computer-implemented method of claim 1 , further comprising adjusting a therapeutic treatment protocol based on the patterns in gene expression data.

14. The computer-implemented method of claim 1 , wherein diagnosing HER2 status based on at least gene expression data from a third sample of the patient comprises:

for each of a plurality of biological pathways in the gene expression data, determining a pathway score;

preparing a summary score for the plurality of biological pathways, based upon the pathway score for each biological pathway; and

comparing the summary score to one or more enrichment scores each associated with a pre-determined molecular subtype,

wherein diagnosing HER2 status based on at least gene expression data from a third sample of the patient comprises determining a molecular subtype of the gene expression data as corresponding to the HER2 status, based on the comparison of the summary score and the one or more enrichment scores.

15. The computer-implemented method of claim 1 , wherein the trained IHC classification model is trained with histopathology slide image data to classify HER2 status; and wherein the trained FISH classification model is trained with histopathology slide image data to classify HER2 status.

16. The computer-implemented method of claim 15 , wherein the trained IHC classification model and the trained FISH classification model are convolutional neural networks.

17. The computer-implemented method of claim 15 , further comprising, in response identifying discordant HER2 status result between the HER2 status from the IHC image and the HER2 status from the FISH image and diagnosing HER2 status based on the at least gene expression data from the third sample of the patient, providing the discordant HER2 status results and the diagnosed HER2 status based on the at least gene expression data to a hybrid classification model.

18. The computer-implemented method of claim 17 , wherein the hybrid classification model comprises a convolutional neural network.

19. The computer-implemented method of claim 1 , wherein the first sample, the second sample, and the third sample are from a single biopsy block.

20. The computer-implemented method of claim 19 , wherein the first sample, the second sample, and the third sample are different slices from the biopsy block.

21. A system having a memory and a processor, the memory storing instructions, that when executed, cause the processor to:

obtain human epidermal growth factor receptor 2 (HER2) status for a specimen from analysis of immunohistochemistry (IHC) image from a first sample of the patient using a trained IHC classification model;

obtain HER2 status for a specimen from analysis of fluorescence in-situ hybridization (FISH) image from a second sample of the patient using a trained FISH classification model;

identify discordant HER2 status result between the HER2 status obtained from the IHC image and the HER2 status obtained from the FISH image; and

in response to the identification of discordant HER2 status, diagnose HER2 status based on at least gene expression data from a third sample of the patient,

wherein the trained IHC classification model and the trained FISH classification model are machine learning models.

22. The system of claim 21 , the memory storing further instructions, that when executed, cause the processor to: generate a HER2 discordance status report indicating biologic pathways in gene expression data.

23. The system of claim 21 , the memory storing further instructions, that when executed, cause the processor to: generate a HER2 discordance status report including an indication of a model of molecular subtype gene expression used to identify the discordant HER2.

24. The system of claim 21 , wherein model of molecular subtype gene expression comprises a linear gene expression model.

25. The system of claim 21 , wherein the model of molecular subtype gene expression comprises a pathway gene expression model.

26. The system of claim 21 , the memory storing further instructions, that when executed, cause the processor to: generate the HER2 discordance status report including a listing of pathways identified by the pathway gene expression model.

27. The system of claim 21 , wherein the model of molecular subtype gene expression comprises a multiple gene linear regression gene expression model.

28. The system of claim 21 , the memory storing further instructions, that when executed, cause the processor to: generate the HER2 discordance status report including a listing of genes identified by the multiple gene linear regression gene expression model.

29. The system of claim 21 , the memory storing further instructions, that when executed, cause the processor to: generate a HER2 discordance status report indicating a molecular subtype determined from the gene expression data.