Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Dec;163(6):1531-1546.e8.
doi: 10.1053/j.gastro.2022.08.025. Epub 2022 Aug 17.

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

Reetesh K Pai  1 , Imon Banerjee  2 , Sameer Shivji  3 , Suchit Jain  2 , Douglas Hartman  1 , Daniel D Buchanan  4 , Mark A Jenkins  5 , David F Schaeffer  6 , Christophe Rosty  7 , Julia Como  8 , Amanda I Phipps  9 , Polly A Newcomb  9 , Andrea N Burnett-Hartman  10 , Loic Le Marchand  11 , Niloy J Samadder  12 , Bhavik Patel  2 , Carol Swallow  13 , Noralane M Lindor  14 , Steven J Gallinger  15 , Robert C Grant  16 , Thomas Westerling-Bui  17 , James Conner  3 , David P Cyr  13 , Richard Kirsch  3 , Rish K Pai  18
Affiliations

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

Reetesh K Pai et al. Gastroenterology. 2022 Dec.

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.

Keywords: Colorectal Cancer; Image Analysis; Prognosis; Stroma; Tumor Infiltrating Lymphocytes.

PubMed Disclaimer

Conflict of interest statement

Conflict of Interest

David F. Schaeffer reports honoraria from Alimentiv Inc., Pfizer, Merck, Diaceutics, and Astellas; and stock ownership in Satisfai Health Inc outside of the submitted work. Reetesh K. Pai, Christophe Rosty, Richard Kirsch report consulting income from Alimentiv Inc. outside of the submitted work. Rish K. Pai reports consulting income from Alimentiv Inc., Eli Lilly, AbbVie, Allergan, Genentech, and PathAI outside of the submitted work.

Thomas Westerling-Bui is an employee of Aiforia Inc. The remaining authors disclose no conflicts.

Figures

Figure 1.
Figure 1.
Cohort digitization and application of QuantCRC algorithm. (A) Flow chart depicting cohorts used in this study and number of slides digitized from each center. (B) QuantCRC was applied to 6468 images, which segments the image in a stepwise manner. First the image is segmented into carcinoma (green), stroma (light blue), mucin (dark blue), TB/PDC (red), necrosis (brown), smooth muscle (purple), and fat (yellow). Next the stroma is segmented into immature (teal), mature (green), and inflammatory (gray). The carcinoma is segmented into low-grade (purple), high-grade (orange), and signet ring cell (light green). Finally, TILs are recognized as objects (blue dots) within the tumor. After this segmentation, 15 features are calculated from each image as shown. Abbreviations: B, tumor bed; ST, stromal region.
Figure 2.
Figure 2.
RFS in stage I-III CRCs according to quartiles of QuantCRC features using Kaplan-Meier analysis. (A) TB/PDC. (B) Tumor:stroma ratio. (C) TILs per mm2 tumor. (D) Immature stroma within tumor bed. (E) Inflammatory stroma within tumor bed. (F) Mature stroma within tumor bed.
Figure 3.
Figure 3.
Stage III risk groups using the prognostic model. (A) Frequency distribution of predicted 36-month RFS in the training set of stage III tumors. The 33rd and 66th percentiles were used to split tumors into 3 risk groups (dotted lines). (B) These risk group cutoffs were applied to the internal test set of stage III CRCs, and RFS was analyzed by Kaplan-Meier analysis. (C) Similarly, these risk group cutoffs were applied to the external validation cohort of stage III CRCs, and RFS was analyzed by Kaplan-Meier analysis. (D) Scatter plot of 6 QuantCRC features stratified according to stage III risk groups with median and interquartile range.
Figure 4.
Figure 4.
Stage II risk groups using the prognostic model. (A) Frequency distribution of predicted 36-month RFS in the training set of stage II tumors. The 33rd and 66th percentiles were used to split tumors into 3 risk groups (dotted lines). (B) These risk group cutoffs were applied to the internal test set of stage II CRCs and RFS was analyzed by Kaplan-Meier analysis. (C) Similarly, these risk group cutoffs were applied to the external validation cohort of stage II CRCs and RFS was analyzed byKaplan-Meier analysis. (D) Scatter plot of 6 QuantCRC features stratified according to stage II risk groups with median and interquartile range.
See this image and copyright information in PMC

References

    1. Ueno H, Hashiguchi Y, Shimazaki H, et al. Objective criteria for crohn-like lymphoid reaction in colorectal cancer. Am J Clin Pathol 2013;139:434–441. - PubMed
    1. Idos GE, Kwok J, Bonthala N, et al. The prognostic implications of tumor infiltrating lymphocytes in colorectal cancer: a systematic review and meta-analysis. Sci Rep 2020;10:3360. - PMC - PubMed
    1. Williams DS, Mouradov D, Jorissen RN, et al. Lymphocytic response to tumour and deficient DNA mismatch repair identify subtypes of stage II/III colorectal cancer associated with patient outcomes. Gut 2019; 68:465–474. - PubMed
    1. Lee H, Sha D, Foster NR, et al. 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;31:487–494. - PMC - PubMed
    1. Baxter NN, Kennedy EB, Bergsland E, et al. Adjuvant therapy for stage II colon cancer: ASCO Guideline Update. J Clin Oncol 2022;40:892–910. - PubMed

Publication types