An estimation model for Oncotype DX recurrence score using routine histopathologic variables.

Authors

null

Hyunseok Kim

The Johns Hopkins University School of Medicine, Baltimore, MD

Hyunseok Kim , Christopher Umbricht , Peter B. Illei , Maria Cristina Figueroa Magalhaes , Catherine Pesce , Michele Maiko Gage , Charles Mylander , Martin Rosman , Lorraine Tafra , Kala Visvanathan , Leslie Cope , Antonio C. Wolff

Organizations

The Johns Hopkins University School of Medicine, Baltimore, MD, The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD, Cancer Institute of Sao Paulo, Sao Paulo, Brazil, NorthShore University HealthSystem, chicago, IL, Walter Reed National Military Medical Center, Bethesda, MD, Anne Arundel Medical Center, Annapolis, MD

Research Funding

No funding sources reported

Background: The gene expression profile assayOncotypeDx (ODX) is frequently used to guide adjuvant chemotherapy decisions for patients with estrogen receptor (ER)-positive, lymph node (LN)-negative breast cancer. We hypothesized that in cases where ODX is being considered, the observed recurrence score (RS) category can be accurately estimated using routinely available, less costly histopathologic variables. Methods: We retrospectively reviewed pathology reports between June 2006 and August 2012 from Johns Hopkins (JH) patients (n=301) with early stage ER-positive, LN-negative breast cancer, for whom ODX was ordered. We developed a linear regression model using routine histopathologic markers (ER and progesterone receptor [PR] expression, HER2 status, tumor grade, and Ki67) to calculate an Estimated Recurrence Score (ERS), and correlated it with the observed ODX RS assay result. This model was internally cross-validated in the JH cohort, and externally validated in a separate 326-patient cohort from three Maryland community settings. Results: In the JH cohort, 244 patients had an observed RS ≤ 25 (80%) and 57 patients had an observed RS above 25 (20%). When the ERS was < 21 (n=200), we accurately classified 95% of them (191) who were found to have a low risk(observed RS ≤ 25). Similarly, 95% of those (18/19) with an ERS > 30 fell into a high risk category with observed RS > 25. An accuracy of 93% was observed in the external validation cohort (Table). Conclusions: We developed a clinical estimator of the ODX RS using cases in which the clinician chose to order ODX RS in routine clinical practice. For more than 80% of patients in the external validation cohort, ERS was estimated to be < 21 or > 30 and in this group, the Hopkins ERS model correctly predicted the observed RS category (≤ 25 or > 25) in 93% of cases. Although further validation of this model is warranted, preliminary evidence supports use of ERS to reliably identify those patients most likely to benefit from including ODX RS results in therapeutic decision-making.

JH cohort ODX RS
≤ 25
ODX RS
> 25
Total External
cohort
ODX RS
≤ 25
ODX RS
> 25
Total
ERS < 21 191 9 200 ERS<21 227 17 244
ERS(21-30) 52 30 82 ERS(21-30) 31 22 53
ERS(> 30) 1 18 19 ERS(>30) 2 27 29
Total 244 57 301 Total 260 66 326

Disclaimer

This material on this page is ©2024 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org

Abstract Details

Meeting

2014 ASCO Annual Meeting

Session Type

Poster Session

Session Title

Breast Cancer - HER2/ER

Track

Breast Cancer

Sub Track

ER+

Citation

J Clin Oncol 32:5s, 2014 (suppl; abstr 559)

DOI

10.1200/jco.2014.32.15_suppl.559

Abstract #

559

Poster Bd #

23

Abstract Disclosures