Vidit Sharma, MD, MS, was one of the recipients of the 2024 Urology Care Foundation™ Research Scholar Awards. These awards provide $40,000 annually for mentored research training for clinical and postdoctoral fellows or early-career faculty. The Indian American Urological Association sponsored Dr Sharma’s award.

Renal cell carcinoma (RCC) can form a venous tumor thrombus (VTT) in approximately 10% of cases, and these patients present unique surgical challenges given the large tumor size, need for inferior vena cava (IVC) control, possible hepatic mobilization, and possible cardiac involvement. We presented data¹ at the AUA 2024 Annual Meeting in San Antonio, Texas, examining 111,785 radical nephrectomies from 2016 to 2020 in the National Inpatient Sample–an administrative hospital discharge database. We found that approximately 3.3% (N = 3691) of cases had significant IVC manipulation, and 0.3% (N = 335) had cardiac involvement. Although the risk of major complications decreased with annual hospital IVC VTT caseloads, even patients receiving surgery at high-volume centers had a nearly 10% risk of major complications (Figure 1). In addition, at the same meeting, we presented data (N = 786) from our own center to show that over 50% of patients with RCC and IVC VTT (Mayo Thrombus Level >0) will develop metastases on prolonged follow-up after radical nephrectomy (Figure 2).² Thus, there is a significant unmet need to improve near-term surgical outcomes and long-term oncologic outcomes for patients with RCC and VTT.

As the principal investigator of the Mayo Clinic Nephrectomy Registry, these cases have grown into a focus of my clinical practice and academic pursuits. My long-term goal is to broadly improve outcomes for these patients, and I was glad to see this topic selected for the AUA 2024 Early Career Investigator and subsequently the AUA Urology Care Foundation™ Research Scholar Award. Under the mentorship of Dr Brad Leibovich, I hope to use these opportunities to improve outcomes for RCC patients with VTT using artificial intelligence. We are collaborating with Dr Tim Kline (a data scientist well versed in artificial intelligence applications to imaging kidneys), Dr Andrew Rule (a nephrologist with extensive funding studying postnephrectomy renal functional outcomes and in applying artificial intelligence models to renal imaging), and Dr Jonathan Morris (a radiologist and head of the Mayo Clinic 3D anatomical modeling laboratory). Together with the above team, Dr Abhinav Khanna and I have developed an automated segmentation model for renal tumors and nonneoplastic renal parenchyma with high accuracy (Figure 3).³ Using the Research Scholar Award, I hope to extend our prior efforts of automated renal tumor segmentation to patients with VTTs through the following aims.

In Aim 1, we will create an artificial intelligence algorithm to automatically segment VTTs in renal tumors. This computer-aided detection will allow for improved identification of the presence and extent of VTTs in patients with renal tumors. We will then train another algorithm based on these segmentations to predict acute postoperative complications from the Nephrectomy Registry. This will enable individualized complication prediction based on the unique anatomy of the renal tumors and VTTs for each patient during preoperative counseling. Next, full-scale 3D-printed models (such as Figure 4) of the patient’s RCC and VTT will be printed using the automated segmentation algorithm to determine if this improves the surgeon’s understanding of the anatomy and patients’ comfort with the procedure.