Previously submitted to: JMIR Cancer (no longer under consideration since Jul 09, 2026)
Date Submitted: Oct 28, 2025
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
A recent study on the dominance of artificial intelligence in malignancy treatment and follow-up modalities: ChatGPT-5's performance in kidney cancer
ABSTRACT
Background:
The rapid development of artificial intelligence (AI) has facilitated access to medical information. The most well-known AI model today is ChatGPT, developed by OpenAI, and its use in various medical disciplines has become increasingly widespread. The reliability of this system in providing disease-specific clinical guidance in urologic oncology remains uncertain.
Objective:
To investigate the performance of ChatGPT-5 on kidney cancer treatment, staging, and follow-up modalities using a real patient population.
Methods:
A total of 211 patients who underwent radical and partial nephrectomy at our clinic between 2014 and 2025 were retrospectively analyzed. Preop symptoms, tumor size, location, R.E.N.A.L. scores, planned and performed surgical method, tumor pathological type, size, TNM staging, and Leibovich scores for clear cell tumors were recorded. All data were then imported into ChatGPT-5, and the results were recorded and compared for their compatibility with actual clinical outcomes.
Results:
Among 211 patients (mean age: 58.2 years; 65.8% male), 113 underwent radical nephrectomy and 98 underwent partial nephrectomy. ChatGPT-5 recommended radical nephrectomy in 57 cases and partial nephrectomy in 68 cases, demonstrating moderate agreement with clinical decisions (Cohen's Kappa = 0.410, p<0.001). It demonstrated 85.9% accuracy in TNM classification and 76.3% consistency in Leibovich scoring for clear cell RCC. AI tended to favor nephron-sparing approaches more frequently than clinicians, particularly in mid-pole tumors and in cases with low R.E.N.A.L. scores.
Conclusions:
ChatGPT-5 demonstrates promising potential as a support tool for the treatment, follow-up, and staging of kidney cancer. However, the variability in surgical recommendations highlights the need for continuous improvement and physician oversight. Before full clinical integration, multicenter, comparative studies between different AI models are required.
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.