Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer

Yuli Chen; Haojia Li; Andrew Janowczyk; Paula Toro; Germán Corredor; Jon Whitney; Cheng Lu; Can F. Koyuncu; Mojgan Mokhtari; Christina Buzzy; Shridar Ganesan; Michael D. Feldman; Pingfu Fu; Haley Corbin; Aparna Harbhajanka; Hannah Gilmore; Lori J. Goldstein; Nancy E. Davidson; Sangeeta Desai; Vani Parmar; Anant Madabhushi

doi:10.1038/s41523-023-00545-y

Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer

Yuli Chen, Haojia Li, Andrew Janowczyk, Paula Toro, Germán Corredor, Jon Whitney, Cheng Lu, Can F. Koyuncu, Mojgan Mokhtari, Christina Buzzy, Shridar Ganesan, Michael D. Feldman, Pingfu Fu, Haley Corbin, Aparna Harbhajanka, Hannah Gilmore, Lori J. Goldstein, Nancy E. Davidson, Sangeeta Desai, Vani ParmarAnant Madabhushi

Cancer Institute of New Jersey (CINJ), Division of Molecular Oncology

Rutgers, The State University

Research output: Contribution to journal › Article › peer-review

Abstract

Prognostic markers currently utilized in clinical practice for estrogen receptor-positive (ER+) and lymph node-negative (LN−) invasive breast cancer (IBC) patients include the Nottingham grading system and Oncotype Dx (ODx). However, these biomarkers are not always optimal and remain subject to inter-/intra-observer variability and high cost. In this study, we evaluated the association between computationally derived image features from H&E images and disease-free survival (DFS) in ER+ and LN− IBC. H&E images from a total of n = 321 patients with ER+ and LN− IBC from three cohorts were employed for this study (Training set: D1 (n = 116), Validation sets: D2 (n = 121) and D3 (n = 84)). A total of 343 features relating to nuclear morphology, mitotic activity, and tubule formation were computationally extracted from each slide image. A Cox regression model (IbRiS) was trained to identify significant predictors of DFS and predict a high/low-risk category using D1 and was validated on independent testing sets D2 and D3 as well as within each ODx risk category. IbRiS was significantly prognostic of DFS with a hazard ratio (HR) of 2.33 (95% confidence interval (95% CI) = 1.02–5.32, p = 0.045) on D2 and a HR of 2.94 (95% CI = 1.18–7.35, p = 0.0208) on D3. In addition, IbRiS yielded significant risk stratification within high ODx risk categories (D1 + D2: HR = 10.35, 95% CI = 1.20–89.18, p = 0.0106; D1: p = 0.0238; D2: p = 0.0389), potentially providing more granular risk stratification than offered by ODx alone.

Original language	English (US)
Article number	40
Journal	npj Breast Cancer
Volume	9
Issue number	1
DOIs	https://doi.org/10.1038/s41523-023-00545-y
State	Published - Dec 2023

All Science Journal Classification (ASJC) codes

Oncology
Radiology Nuclear Medicine and imaging
Pharmacology (medical)

Access to Document

10.1038/s41523-023-00545-y

Cite this

Chen, Y., Li, H., Janowczyk, A., Toro, P., Corredor, G., Whitney, J., Lu, C., Koyuncu, C. F., Mokhtari, M., Buzzy, C., Ganesan, S., Feldman, M. D., Fu, P., Corbin, H., Harbhajanka, A., Gilmore, H., Goldstein, L. J., Davidson, N. E., Desai, S., ... Madabhushi, A. (2023). Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer. npj Breast Cancer, 9(1), Article 40. https://doi.org/10.1038/s41523-023-00545-y

@article{4eb8d0c6ba604aaca40e821a49c837b1,

title = "Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer",

abstract = "Prognostic markers currently utilized in clinical practice for estrogen receptor-positive (ER+) and lymph node-negative (LN−) invasive breast cancer (IBC) patients include the Nottingham grading system and Oncotype Dx (ODx). However, these biomarkers are not always optimal and remain subject to inter-/intra-observer variability and high cost. In this study, we evaluated the association between computationally derived image features from H&E images and disease-free survival (DFS) in ER+ and LN− IBC. H&E images from a total of n = 321 patients with ER+ and LN− IBC from three cohorts were employed for this study (Training set: D1 (n = 116), Validation sets: D2 (n = 121) and D3 (n = 84)). A total of 343 features relating to nuclear morphology, mitotic activity, and tubule formation were computationally extracted from each slide image. A Cox regression model (IbRiS) was trained to identify significant predictors of DFS and predict a high/low-risk category using D1 and was validated on independent testing sets D2 and D3 as well as within each ODx risk category. IbRiS was significantly prognostic of DFS with a hazard ratio (HR) of 2.33 (95% confidence interval (95% CI) = 1.02–5.32, p = 0.045) on D2 and a HR of 2.94 (95% CI = 1.18–7.35, p = 0.0208) on D3. In addition, IbRiS yielded significant risk stratification within high ODx risk categories (D1 + D2: HR = 10.35, 95% CI = 1.20–89.18, p = 0.0106; D1: p = 0.0238; D2: p = 0.0389), potentially providing more granular risk stratification than offered by ODx alone.",

author = "Yuli Chen and Haojia Li and Andrew Janowczyk and Paula Toro and Germ{\'a}n Corredor and Jon Whitney and Cheng Lu and Koyuncu, {Can F.} and Mojgan Mokhtari and Christina Buzzy and Shridar Ganesan and Feldman, {Michael D.} and Pingfu Fu and Haley Corbin and Aparna Harbhajanka and Hannah Gilmore and Goldstein, {Lori J.} and Davidson, {Nancy E.} and Sangeeta Desai and Vani Parmar and Anant Madabhushi",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s).",

year = "2023",

month = dec,

doi = "10.1038/s41523-023-00545-y",

language = "English (US)",

volume = "9",

journal = "npj Breast Cancer",

issn = "2374-4677",

publisher = "Nature Publishing Group",

number = "1",

}

Chen, Y, Li, H, Janowczyk, A, Toro, P, Corredor, G, Whitney, J, Lu, C, Koyuncu, CF, Mokhtari, M, Buzzy, C, Ganesan, S, Feldman, MD, Fu, P, Corbin, H, Harbhajanka, A, Gilmore, H, Goldstein, LJ, Davidson, NE, Desai, S, Parmar, V & Madabhushi, A 2023, 'Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer', npj Breast Cancer, vol. 9, no. 1, 40. https://doi.org/10.1038/s41523-023-00545-y

TY - JOUR

T1 - Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer

AU - Chen, Yuli

AU - Li, Haojia

AU - Janowczyk, Andrew

AU - Toro, Paula

AU - Corredor, Germán

AU - Whitney, Jon

AU - Lu, Cheng

AU - Koyuncu, Can F.

AU - Mokhtari, Mojgan

AU - Buzzy, Christina

AU - Ganesan, Shridar

AU - Feldman, Michael D.

AU - Fu, Pingfu

AU - Corbin, Haley

AU - Harbhajanka, Aparna

AU - Gilmore, Hannah

AU - Goldstein, Lori J.

AU - Davidson, Nancy E.

AU - Desai, Sangeeta

AU - Parmar, Vani

AU - Madabhushi, Anant

PY - 2023/12

Y1 - 2023/12

N2 - Prognostic markers currently utilized in clinical practice for estrogen receptor-positive (ER+) and lymph node-negative (LN−) invasive breast cancer (IBC) patients include the Nottingham grading system and Oncotype Dx (ODx). However, these biomarkers are not always optimal and remain subject to inter-/intra-observer variability and high cost. In this study, we evaluated the association between computationally derived image features from H&E images and disease-free survival (DFS) in ER+ and LN− IBC. H&E images from a total of n = 321 patients with ER+ and LN− IBC from three cohorts were employed for this study (Training set: D1 (n = 116), Validation sets: D2 (n = 121) and D3 (n = 84)). A total of 343 features relating to nuclear morphology, mitotic activity, and tubule formation were computationally extracted from each slide image. A Cox regression model (IbRiS) was trained to identify significant predictors of DFS and predict a high/low-risk category using D1 and was validated on independent testing sets D2 and D3 as well as within each ODx risk category. IbRiS was significantly prognostic of DFS with a hazard ratio (HR) of 2.33 (95% confidence interval (95% CI) = 1.02–5.32, p = 0.045) on D2 and a HR of 2.94 (95% CI = 1.18–7.35, p = 0.0208) on D3. In addition, IbRiS yielded significant risk stratification within high ODx risk categories (D1 + D2: HR = 10.35, 95% CI = 1.20–89.18, p = 0.0106; D1: p = 0.0238; D2: p = 0.0389), potentially providing more granular risk stratification than offered by ODx alone.

AB - Prognostic markers currently utilized in clinical practice for estrogen receptor-positive (ER+) and lymph node-negative (LN−) invasive breast cancer (IBC) patients include the Nottingham grading system and Oncotype Dx (ODx). However, these biomarkers are not always optimal and remain subject to inter-/intra-observer variability and high cost. In this study, we evaluated the association between computationally derived image features from H&E images and disease-free survival (DFS) in ER+ and LN− IBC. H&E images from a total of n = 321 patients with ER+ and LN− IBC from three cohorts were employed for this study (Training set: D1 (n = 116), Validation sets: D2 (n = 121) and D3 (n = 84)). A total of 343 features relating to nuclear morphology, mitotic activity, and tubule formation were computationally extracted from each slide image. A Cox regression model (IbRiS) was trained to identify significant predictors of DFS and predict a high/low-risk category using D1 and was validated on independent testing sets D2 and D3 as well as within each ODx risk category. IbRiS was significantly prognostic of DFS with a hazard ratio (HR) of 2.33 (95% confidence interval (95% CI) = 1.02–5.32, p = 0.045) on D2 and a HR of 2.94 (95% CI = 1.18–7.35, p = 0.0208) on D3. In addition, IbRiS yielded significant risk stratification within high ODx risk categories (D1 + D2: HR = 10.35, 95% CI = 1.20–89.18, p = 0.0106; D1: p = 0.0238; D2: p = 0.0389), potentially providing more granular risk stratification than offered by ODx alone.

UR - http://www.scopus.com/inward/record.url?scp=85159652136&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85159652136&partnerID=8YFLogxK

U2 - 10.1038/s41523-023-00545-y

DO - 10.1038/s41523-023-00545-y

M3 - Article

SN - 2374-4677

VL - 9

JO - npj Breast Cancer

JF - npj Breast Cancer

IS - 1

M1 - 40

ER -

Computational pathology improves risk stratification of a multi-gene assay for early stage ER+ breast cancer

Abstract

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this