A perspective on integrating digital pathology, proteomics, clinical data and AI analytics in cancer research
(2025) In Journal of Proteomics 320.- Abstract
Nearly 40 % of individuals will be diagnosed with cancer in their lifetime, translating to an estimated 20 million new cases annually. Despite remarkable therapeutic advances, only 15–20 % of patients achieve durable responses to immunotherapy, and the high cost of treatment (illustrated by immune checkpoint inhibitors like pembrolizumab and nivolumab, totaling roughly $191,000 per year) remains a formidable global challenge. The convergence of digital pathology, high-throughput molecular profiling, and advanced computational strategies has the potential to transform cancer research. By integrating high-resolution morphological data with proteomic, transcriptomic, and spatial molecular insights, we can elucidate the complex interplay... (More)
Nearly 40 % of individuals will be diagnosed with cancer in their lifetime, translating to an estimated 20 million new cases annually. Despite remarkable therapeutic advances, only 15–20 % of patients achieve durable responses to immunotherapy, and the high cost of treatment (illustrated by immune checkpoint inhibitors like pembrolizumab and nivolumab, totaling roughly $191,000 per year) remains a formidable global challenge. The convergence of digital pathology, high-throughput molecular profiling, and advanced computational strategies has the potential to transform cancer research. By integrating high-resolution morphological data with proteomic, transcriptomic, and spatial molecular insights, we can elucidate the complex interplay between tumor cells and their microenvironment. In this perspective, we review how emerging techniques, from AI-driven image analysis to deep visual proteomics, can accelerate biomarker discovery, refine patient stratification, and ultimately improve clinical outcomes. We illustrate these principles with a case study in melanoma, where the integration of digital pathology and deep proteomic profiling uncovered a molecular signature predictive of recurrence in early-stage disease. As these technologies evolve, we foresee a future of precision oncology characterized by the seamless integration of morphological, clinical, and molecular data enabled by AI-driven analytics. Significance: This perspective represents a pivotal step toward transforming cancer research by bridging the gap between traditional histopathological evaluation and modern molecular analytics. By integrating digital pathology with spatial proteomics and advanced AI-driven analytics, our approach provides a multidimensional view of tumor biology that captures both morphological nuances and molecular heterogeneity. This comprehensive framework not only enhances our understanding of the tumor microenvironment but also facilitates the discovery of robust biomarkers for disease recurrence and therapeutic response. Ultimately, our findings underscore the potential of precision oncology to tailor treatment strategies to individual patient profiles, thereby improving clinical outcomes and guiding the next generation of personalized cancer care.
(Less)
- author
- Guedes, Jéssica
LU
; Woldmar, Nicole
LU
; Szasz, A. Marcell
; Wieslander, Elisabet
LU
; Pawłowski, Krysztof
LU
; Horvatovich, Peter
LU
; Malm, Johan
LU
; Szadai, Leticia
; Németh, István Balázs
; Marko-Varga, György
LU
and Gil, Jeovanis
LU
(Less) - organization
- publishing date
- 2025-09
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- AI, Cancer biomarkers, Deep visual proteomics, Digital pathology, Laser capture microdissection, Machine learning, Melanoma, Patient stratification, Proteomics, Spatial proteomics, Tumor microenvironment
- in
- Journal of Proteomics
- volume
- 320
- article number
- 105493
- publisher
- Elsevier
- external identifiers
-
- pmid:40680990
- scopus:105010690299
- ISSN
- 1874-3919
- DOI
- 10.1016/j.jprot.2025.105493
- language
- English
- LU publication?
- yes
- id
- 79ca34bb-6661-4fa0-9683-726ca2703f85
- date added to LUP
- 2025-11-04 08:45:39
- date last changed
- 2025-11-05 03:25:13
@article{79ca34bb-6661-4fa0-9683-726ca2703f85,
abstract = {{<p>Nearly 40 % of individuals will be diagnosed with cancer in their lifetime, translating to an estimated 20 million new cases annually. Despite remarkable therapeutic advances, only 15–20 % of patients achieve durable responses to immunotherapy, and the high cost of treatment (illustrated by immune checkpoint inhibitors like pembrolizumab and nivolumab, totaling roughly $191,000 per year) remains a formidable global challenge. The convergence of digital pathology, high-throughput molecular profiling, and advanced computational strategies has the potential to transform cancer research. By integrating high-resolution morphological data with proteomic, transcriptomic, and spatial molecular insights, we can elucidate the complex interplay between tumor cells and their microenvironment. In this perspective, we review how emerging techniques, from AI-driven image analysis to deep visual proteomics, can accelerate biomarker discovery, refine patient stratification, and ultimately improve clinical outcomes. We illustrate these principles with a case study in melanoma, where the integration of digital pathology and deep proteomic profiling uncovered a molecular signature predictive of recurrence in early-stage disease. As these technologies evolve, we foresee a future of precision oncology characterized by the seamless integration of morphological, clinical, and molecular data enabled by AI-driven analytics. Significance: This perspective represents a pivotal step toward transforming cancer research by bridging the gap between traditional histopathological evaluation and modern molecular analytics. By integrating digital pathology with spatial proteomics and advanced AI-driven analytics, our approach provides a multidimensional view of tumor biology that captures both morphological nuances and molecular heterogeneity. This comprehensive framework not only enhances our understanding of the tumor microenvironment but also facilitates the discovery of robust biomarkers for disease recurrence and therapeutic response. Ultimately, our findings underscore the potential of precision oncology to tailor treatment strategies to individual patient profiles, thereby improving clinical outcomes and guiding the next generation of personalized cancer care.</p>}},
author = {{Guedes, Jéssica and Woldmar, Nicole and Szasz, A. Marcell and Wieslander, Elisabet and Pawłowski, Krysztof and Horvatovich, Peter and Malm, Johan and Szadai, Leticia and Németh, István Balázs and Marko-Varga, György and Gil, Jeovanis}},
issn = {{1874-3919}},
keywords = {{AI; Cancer biomarkers; Deep visual proteomics; Digital pathology; Laser capture microdissection; Machine learning; Melanoma; Patient stratification; Proteomics; Spatial proteomics; Tumor microenvironment}},
language = {{eng}},
publisher = {{Elsevier}},
series = {{Journal of Proteomics}},
title = {{A perspective on integrating digital pathology, proteomics, clinical data and AI analytics in cancer research}},
url = {{http://dx.doi.org/10.1016/j.jprot.2025.105493}},
doi = {{10.1016/j.jprot.2025.105493}},
volume = {{320}},
year = {{2025}},
}