Introduction
Clear cell renal cell carcinoma (RCC) is a prevalent form of kidney cancer, affecting a significant number of individuals worldwide. The paper discusses the development of a novel multimodal recurrence scoring system that integrates clinical, genomic, and histopathological data to improve the predictive accuracy for localized RCC recurrence.
Methods
In a retrospective analysis and validation study, researchers developed a histopathological whole-slide image (WSI)-based score using deep learning and digital scanning of conventional haematoxylin and eosin-stained tumor tissue sections. They also incorporated a six single nucleotide polymorphism (SNP)-based score derived from paraffin-embedded tumor tissue samples and the Leibovich score based on clinicopathological risk factors. These modalities were combined to construct a multimodal recurrence score in a training dataset. The performance of the multimodal recurrence score was then validated in independent datasets.
Findings
The multimodal recurrence score demonstrated significantly higher predictive accuracy compared to the three single-modal scores and clinicopathological risk factors. It accurately predicted the recurrence-free interval (RFI) in both the training and validation datasets. The areas under the curve at 5 years were consistently higher for the multimodal score compared to the single-modal scores and clinicopathological risk factors (p<0.05). Notably, the multimodal recurrence score successfully stratified patients into distinct risk groups, even within the same TNM stage or pathological grade. Patients classified as high-risk stage I and II according to the multimodal score had a shorter RFI compared to low-risk stage III patients (p<0.0001). Similarly, the high-risk grade 1 and 2 group had a shorter RFI compared to the low-risk grade 3 and 4 group (p<0.0001).
Significance
The multimodal recurrence scoring system developed in the study represents a practical and reliable predictor for localized RCC recurrence after surgery. It offers significant added value to the current staging system, enabling more accurate prediction of recurrence risk. This integrated approach, combining clinical, genomic, and histopathological information, can better inform treatment decisions regarding adjuvant therapy. By identifying patient subgroups with high recurrence risk, the multimodal scoring system facilitates a more targeted approach to the selection of adjuvant therapies.
Research Context and Added Value
Previous studies explored multimodal prognostic models for RCC, incorporating histopathological whole-slide images and molecular profile data. However, the molecular signatures of those models were primarily derived from fresh-frozen specimens and lacked validation in multiple independent cohorts. The current study overcomes these limitations by developing and validating a multimodal recurrence scoring system using a large sample size of localized clear cell RCC cases. The inclusion of independent cohorts from multicenters in China and the Cancer Genome Atlas dataset strengthens the robustness and generalizability of the findings.
Implications and Clinical Relevance
The study's findings have implications for clinical practice. The multimodal recurrence scoring system serves as a practical and reliable prognostic tool for localized RCC. Its compatibility with routine paraffin-embedded tumor tissue and haematoxylin and eosin-stained sections facilitates easy translation into clinical application. By supplementing the existing staging system, the multimodal score enables more accurate prediction of recurrence after surgery. Consequently, it supports informed treatment decisions and the design of adjuvant therapy trials, ultimately improving patient outcomes in clear cell RCC.
Reference
Comments