JMIR Publications

ABSTRACT

Hereditary papillary renal carcinoma (HPRC) is an extremely rare autosomal dominant cancer syndrome that predisposes carriers to bilateral and multifocal papillary renal cell carcinoma later in life. Expert opinion recommends ongoing cross-sectional renal surveillance and nephron-sparing surgery when a renal lesion reaches 3 cm [1–3]. However, the rarity of this condition makes evidence supporting management very limited and derived from small observational hereditary renal-cancer cohorts with limited numbers of HPRC members. A 59-year-old male physician who was diagnosed with HPRC 10 years ago requests assistance in prevention and management of his disease. After his initial diagnosis, he enrolled in the NIH/NCI hereditary renal cancer protocol where he has undergone serial MRI surveillance studies [4]. These studies have uncovered a right lower-pole lesion that has enlarged slowly from 0.4 cm in 2016 to 0.7 cm in 2025. The patient’s current management is MRI surveillance every 2 years with anticipation of nephron-sparing surgery for lesions reaching 3 cm. The patient is seeking rational, low-risk options to delay cancer development and improve his prognosis. At age 59, he is on the steepest portion of his variant-associated age-risk curve, which contributes a clinically meaningful anxiety burden. Given the rarity of HPRC and the absence of prevention evidence, can AI-assisted evidence synthesis use the biology of MET-driven tumor neogenesis to identify plausible, theoretical low-risk adjunctive strategies to mitigate this patient’s cancer risk? The strongest current strategy remains surveillance and timely nephron-sparing surgery at the 3 cm threshold. Lifestyle measures have favorable general-health evidence and are likely to improve surgical resilience. Several medications intersect with MET-adjacent signaling nodes while also carrying independent indications relevant to this patient: dyslipidemia, obstructive sleep apnea, insulin resistance/metabolic risk, Crohn's disease, and healthspan. Newer GLP-1 literature in IBD is observational but generally reassuring, and prospective tirzepatide trials in Crohn's disease are now underway. Rapamycin has the most direct mechanistic fit but the highest personalized risk and monitoring burden. Several candidate agents carry independent clinical indications or healthspan rationales that materially change their risk–benefit balance for this patient and, in several cases, make adopting them considerably easier than the thin HPRC-specific evidence alone would suggest. This case illustrates how AI can organize a rare-cancer decision problem by integrating germline risk, family history, serial imaging, comorbidities, medications, tumor biology, and graded evidence. AI tools were highly effective in synthesizing the literature, developing an evidence-based model for tumor development, and providing guidance for managing a complex rare disease and integrating treatment options with multiple chronic conditions.