Quantitative Pathologic Analysis of Digitized Images of Colorectal Carcinoma Improves Prediction of Recurrence-Free Survival

Overview

Journal Gastroenterology

Specialty Gastroenterology

Date 2022 Aug 19

PMID 35985511

Authors

Reetesh K Pai

Imon Banerjee

Sameer Shivji

Suchit Jain

Douglas Hartman

Daniel D Buchanan

Mark A Jenkins

David F Schaeffer

Christophe Rosty

Julia Como

Amanda I Phipps

Polly A Newcomb

Andrea N Burnett-Hartman

Loic Le Marchand

Niloy J Samadder

Bhavik Patel

Carol Swallow

Noralane M Lindor

Steven J Gallinger

Robert C Grant

Thomas Westerling-Bui

James Conner

David P Cyr

Richard Kirsch

Rish K Pai

Affiliations

Soon will be listed here.

Abstract

Background & Aims: To examine whether quantitative pathologic analysis of digitized hematoxylin and eosin slides of colorectal carcinoma (CRC) correlates with clinicopathologic features, molecular alterations, and prognosis.

Methods: A quantitative segmentation algorithm (QuantCRC) was applied to 6468 digitized hematoxylin and eosin slides of CRCs. Fifteen parameters were recorded from each image and tested for associations with clinicopathologic features and molecular alterations. A prognostic model was developed to predict recurrence-free survival using data from the internal cohort (n = 1928) and validated on an internal test (n = 483) and external cohort (n = 938).

Results: There were significant differences in QuantCRC according to stage, histologic subtype, grade, venous/lymphatic/perineural invasion, tumor budding, CD8 immunohistochemistry, mismatch repair status, KRAS mutation, BRAF mutation, and CpG methylation. A prognostic model incorporating stage, mismatch repair, and QuantCRC resulted in a Harrell's concordance (c)-index of 0.714 (95% confidence interval [CI], 0.702-0.724) in the internal test and 0.744 (95% CI, 0.741-0.754) in the external cohort. Removing QuantCRC from the model reduced the c-index to 0.679 (95% CI, 0.673-0.694) in the external cohort. Prognostic risk groups were identified, which provided a hazard ratio of 2.24 (95% CI, 1.33-3.87, P = .004) for low vs high-risk stage III CRCs and 2.36 (95% CI, 1.07-5.20, P = .03) for low vs high-risk stage II CRCs, in the external cohort after adjusting for established risk factors. The predicted median 36-month recurrence rate for high-risk stage III CRCs was 32.7% vs 13.4% for low-risk stage III and 15.8% for high-risk stage II vs 5.4% for low-risk stage II CRCs.

Conclusions: QuantCRC provides a powerful adjunct to routine pathologic reporting of CRC. A prognostic model using QuantCRC improves prediction of recurrence-free survival.

Citing Articles

Automated deep learning-based assessment of tumour-infiltrating lymphocyte density determines prognosis in colorectal cancer.

Millward J, He Z, Nibali A, Mouradov D, Mielke L, Tran K J Transl Med. 2025; 23(1):298.

PMID: 40065354 PMC: 11892243. DOI: 10.1186/s12967-025-06254-3.

Causal debiasing for unknown bias in histopathology-A colon cancer use case.

Correa-Medero R, Pai R, Ebare K, Buchanan D, Jenkins M, Phipps A PLoS One. 2024; 19(11):e0303415.

PMID: 39576760 PMC: 11584097. DOI: 10.1371/journal.pone.0303415.

Predictors of Local Invasion in Infiltrative Basal Cell Carcinoma: Tumour Budding Outperforms the WHO Subtyping.

Fernandez-Figueras M, Perez-Munoz N, Puig L, Posada-Caez R, Ballester Victoria R, Henriquez M Acta Derm Venereol. 2024; 104:adv40172.

PMID: 38956962 PMC: 11247513. DOI: 10.2340/actadv.v104.40172.

Prediction of recurrence risk in endometrial cancer with multimodal deep learning.

Volinsky-Fremond S, Horeweg N, Andani S, Barkey Wolf J, Lafarge M, de Kroon C Nat Med. 2024; 30(7):1962-1973.

PMID: 38789645 PMC: 11271412. DOI: 10.1038/s41591-024-02993-w.

Use of Deep Learning to Evaluate Tumor Microenvironmental Features for Prediction of Colon Cancer Recurrence.

Sinicrope F, Nelson G, Saberzadeh-Ardestani B, Segovia D, Graham R, Wu C Cancer Res Commun. 2024; 4(5):1344-1350.

PMID: 38709069 PMC: 11114095. DOI: 10.1158/2767-9764.CRC-24-0031.

References

Sirinukunwattana K, Domingo E, Richman S, Redmond K, Blake A, Verrill C . Image-based consensus molecular subtype (imCMS) classification of colorectal cancer using deep learning. Gut. 2020; 70(3):544-554. PMC: 7873419. DOI: 10.1136/gutjnl-2019-319866. View

Ueno H, Kanemitsu Y, Sekine S, Ishiguro M, Ito E, Hashiguchi Y . Desmoplastic Pattern at the Tumor Front Defines Poor-prognosis Subtypes of Colorectal Cancer. Am J Surg Pathol. 2017; 41(11):1506-1512. DOI: 10.1097/PAS.0000000000000946. View

Williams D, Mouradov D, Jorissen R, Newman M, Amini E, Nickless D . Lymphocytic response to tumour and deficient DNA mismatch repair identify subtypes of stage II/III colorectal cancer associated with patient outcomes. Gut. 2018; 68(3):465-474. DOI: 10.1136/gutjnl-2017-315664. View

Collins G, Reitsma J, Altman D, Moons K . Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD statement. BMJ. 2015; 350:g7594. DOI: 10.1136/bmj.g7594. View

Alsop K, Mead L, Smith L, Royce S, Tesoriero A, Young J . Low somatic K-ras mutation frequency in colorectal cancer diagnosed under the age of 45 years. Eur J Cancer. 2006; 42(10):1357-61. DOI: 10.1016/j.ejca.2006.02.023. View

Bychkov D, Linder N, Turkki R, Nordling S, Kovanen P, Verrill C . Deep learning based tissue analysis predicts outcome in colorectal cancer. Sci Rep. 2018; 8(1):3395. PMC: 5821847. DOI: 10.1038/s41598-018-21758-3. View

Bilal M, Raza S, Azam A, Graham S, Ilyas M, Cree I . Development and validation of a weakly supervised deep learning framework to predict the status of molecular pathways and key mutations in colorectal cancer from routine histology images: a retrospective study. Lancet Digit Health. 2021; 3(12):e763-e772. PMC: 8609154. DOI: 10.1016/S2589-7500(21)00180-1. View

Phipps A, Alwers E, Harrison T, Banbury B, Brenner H, Campbell P . Association Between Molecular Subtypes of Colorectal Tumors and Patient Survival, Based on Pooled Analysis of 7 International Studies. Gastroenterology. 2020; 158(8):2158-2168.e4. PMC: 7282955. DOI: 10.1053/j.gastro.2020.02.029. View

Shivji S, Conner J, Barresi V, Kirsch R . Poorly differentiated clusters in colorectal cancer: a current review and implications for future practice. Histopathology. 2020; 77(3):351-368. DOI: 10.1111/his.14128. View

10.

Yamashita R, Long J, Longacre T, Peng L, Berry G, Martin B . Deep learning model for the prediction of microsatellite instability in colorectal cancer: a diagnostic study. Lancet Oncol. 2021; 22(1):132-141. DOI: 10.1016/S1470-2045(20)30535-0. View

11.

Trinh A, Ladrach C, Dawson H, Ten Hoorn S, Kuppen P, Reimers M . Tumour budding is associated with the mesenchymal colon cancer subtype and RAS/RAF mutations: a study of 1320 colorectal cancers with Consensus Molecular Subgroup (CMS) data. Br J Cancer. 2018; 119(10):1244-1251. PMC: 6251036. DOI: 10.1038/s41416-018-0230-7. View

12.

Lee H, Sha D, Foster N, Shi Q, Alberts S, Smyrk T . Analysis of tumor microenvironmental features to refine prognosis by T, N risk group in patients with stage III colon cancer (NCCTG N0147) (Alliance). Ann Oncol. 2020; 31(4):487-494. PMC: 7372727. DOI: 10.1016/j.annonc.2020.01.011. View

13.

Pai R, Pai R . A Practical Approach to the Evaluation of Gastrointestinal Tract Carcinomas for Lynch Syndrome. Am J Surg Pathol. 2016; 40(4):e17-34. DOI: 10.1097/PAS.0000000000000620. View

14.

Bokhorst J, Blank A, Lugli A, Zlobec I, Dawson H, Vieth M . Assessment of individual tumor buds using keratin immunohistochemistry: moderate interobserver agreement suggests a role for machine learning. Mod Pathol. 2019; 33(5):825-833. PMC: 7190566. DOI: 10.1038/s41379-019-0434-2. View

15.

Kirsch R, Messenger D, Riddell R, Pollett A, Cook M, Al-Haddad S . Venous invasion in colorectal cancer: impact of an elastin stain on detection and interobserver agreement among gastrointestinal and nongastrointestinal pathologists. Am J Surg Pathol. 2012; 37(2):200-10. DOI: 10.1097/PAS.0b013e31826a92cd. View

16.

Pai R, Hartman D, Schaeffer D, Rosty C, Shivji S, Kirsch R . Development and initial validation of a deep learning algorithm to quantify histological features in colorectal carcinoma including tumour budding/poorly differentiated clusters. Histopathology. 2021; 79(3):391-405. DOI: 10.1111/his.14353. View

17.

Ueno H, Hashiguchi Y, Shimazaki H, Shinto E, Kajiwara Y, Nakanishi K . Objective criteria for crohn-like lymphoid reaction in colorectal cancer. Am J Clin Pathol. 2013; 139(4):434-41. DOI: 10.1309/AJCPWHUEFTGBWKE4. View

18.

Huijbers A, Tollenaar R, v Pelt G, Zeestraten E, Dutton S, McConkey C . The proportion of tumor-stroma as a strong prognosticator for stage II and III colon cancer patients: validation in the VICTOR trial. Ann Oncol. 2012; 24(1):179-85. DOI: 10.1093/annonc/mds246. View

19.

Echle A, Grabsch H, Quirke P, van den Brandt P, West N, Hutchins G . Clinical-Grade Detection of Microsatellite Instability in Colorectal Tumors by Deep Learning. Gastroenterology. 2020; 159(4):1406-1416.e11. PMC: 7578071. DOI: 10.1053/j.gastro.2020.06.021. View

20.

Weisenberger D, Levine A, Long T, Buchanan D, Walters R, Clendenning M . Association of the colorectal CpG island methylator phenotype with molecular features, risk factors, and family history. Cancer Epidemiol Biomarkers Prev. 2015; 24(3):512-519. PMC: 4355081. DOI: 10.1158/1055-9965.EPI-14-1161. View