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

Autor:	Sinicrope FA; Departments of Medicine and Oncology, Rochester, Minnesota.; Gastrointestinal Research Unit, Mayo Clinic, Rochester, Minnesota., Nelson GD; Division of Clinical Trials and Biostatistics, Mayo Clinic, Rochester, Minnesota.; Department of Quantitative Health Sciences, Mayo Clinic, Rochester, Minnesota., Saberzadeh-Ardestani B; Gastrointestinal Research Unit, Mayo Clinic, Rochester, Minnesota., Segovia DI; Division of Clinical Trials and Biostatistics, Mayo Clinic, Rochester, Minnesota.; Department of Quantitative Health Sciences, Mayo Clinic, Rochester, Minnesota., Graham RP; Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota., Wu C; Division of Medical Oncology, Mayo Clinic, Phoenix, Arizona., Hagen CE; Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota., Shivji S; Department of Pathology, Mount Sinai Hospital, Toronto, Ontario, Canada., Savage P; Mount Sinai Hospital, Toronto, Ontario, Canada., Buchanan DD; Colorectal Oncogenomics Group, Department of Clinical Pathology, The University of Melbourne, Parkville, Victoria, Australia.; University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, Victoria, Australia.; Genetic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, Victoria, Australia., Jenkins MA; Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Melbourne, Victoria, Australia., Phipps AI; Public Health Sciences Division, Fred Hutchinson Cancer Research Center, Seattle, Washington.; Department of Epidemiology, University of Washington, Seattle, Washington., Swallow C; Lunenfeld-Tanenbaum Research Institute, Toronto, Ontario, Canada., LeMarchand L; Department of Epidemiology, University of Hawaii, Honolulu, Hawaii., Gallinger S; Lunenfeld Tanenbaum Research Institute, Mount Sinai Hospital, University of Toronto, Toronto, Ontario, Canada., Grant RC; Division of Medical Oncology and Hematology, Princess Margaret Cancer Centre, University Health Network, Toronto, Ontario, Canada., Pai RK; Department of Pathology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania., Sinicrope SN; University of Chicago Medical Center, Chicago, Illinois., Yan D; Roche Tissue Diagnostics, Tucson, Arizona., Shanmugam K; Roche Tissue Diagnostics, Tucson, Arizona., Conner J; Department of Pathology, Mount Sinai Hospital, Toronto, Ontario, Canada., Cyr DP; Lunenfeld-Tanenbaum Research Institute, Toronto, Ontario, Canada., Kirsch R; Department of Pathology, Mount Sinai Hospital, Toronto, Ontario, Canada., Banerjee I; Department of Radiology and Machine Intelligence in Medicine and Imaging Center (MI-2), Mayo Clinic Arizona, Phoenix, Arizona., Alberts SR; Department of Oncology, Mayo Clinic, Rochester, Minnesota., Shi Q; Division of Clinical Trials and Biostatistics, Mayo Clinic, Rochester, Minnesota.; Department of Quantitative Health Sciences, Mayo Clinic, Rochester, Minnesota., Pai RK; Department of Pathology and Laboratory Medicine, Mayo Clinic, Arizona.
Jazyk:	angličtina
Zdroj:	Cancer research communications [Cancer Res Commun] 2024 May 23; Vol. 4 (5), pp. 1344-1350.
DOI:	10.1158/2767-9764.CRC-24-0031
Abstrakt:	Deep learning may detect biologically important signals embedded in tumor morphologic features that confer distinct prognoses. Tumor morphologic features were quantified to enhance patient risk stratification within DNA mismatch repair (MMR) groups using deep learning. Using a quantitative segmentation algorithm (QuantCRC) that identifies 15 distinct morphologic features, we analyzed 402 resected stage III colon carcinomas [191 deficient (d)-MMR; 189 proficient (p)-MMR] from participants in a phase III trial of FOLFOX-based adjuvant chemotherapy. Results were validated in an independent cohort (176 d-MMR; 1,094 p-MMR). Association of morphologic features with clinicopathologic variables, MMR, KRAS, BRAFV600E, and time-to-recurrence (TTR) was determined. Multivariable Cox proportional hazards models were developed to predict TTR. Tumor morphologic features differed significantly by MMR status. Cancers with p-MMR had more immature desmoplastic stroma. Tumors with d-MMR had increased inflammatory stroma, epithelial tumor-infiltrating lymphocytes (TIL), high-grade histology, mucin, and signet ring cells. Stromal subtype did not differ by BRAFV600E or KRAS status. In p-MMR tumors, multivariable analysis identified tumor-stroma ratio (TSR) as the strongest feature associated with TTR [HRadj 2.02; 95% confidence interval (CI), 1.14-3.57; P = 0.018; 3-year recurrence: 40.2% vs. 20.4%; Q1 vs. Q2-4]. Among d-MMR tumors, extent of inflammatory stroma (continuous HRadj 0.98; 95% CI, 0.96-0.99; P = 0.028; 3-year recurrence: 13.3% vs. 33.4%, Q4 vs. Q1) and N stage were the most robust prognostically. Association of TSR with TTR was independently validated. In conclusion, QuantCRC can quantify morphologic differences within MMR groups in routine tumor sections to determine their relative contributions to patient prognosis, and may elucidate relevant pathophysiologic mechanisms driving prognosis. Significance: A deep learning algorithm can quantify tumor morphologic features that may reflect underlying mechanisms driving prognosis within MMR groups. TSR was the most robust morphologic feature associated with TTR in p-MMR colon cancers. Extent of inflammatory stroma and N stage were the strongest prognostic features in d-MMR tumors. TIL density was not independently prognostic in either MMR group. (© 2024 The Authors; Published by the American Association for Cancer Research.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu