Editor’s Note: Kidney cancer surgical scoring systems provide critical information for treatment planning, surgical method selection, tumor recurrence risk, and patient prognosis. The 39th Annual Congress of the European Association of Urology (EAU24) was held from April 5 to 8, 2024, in Paris, France, where numerous studies and advancements related to kidney cancer surgery were presented. Oncology Frontier invited Professor Jianming Guo from Zhongshan Hospital, Fudan University, to share insights on preoperative and postoperative scoring systems, AI predictions of kidney tumor pathology, and prognosis.
Oncology Frontier: What are the main surgical scoring systems currently used for preoperative applications? What explorations have you made in this area?
Professor Jianming Guo : Urological surgery is akin to “warfare,” requiring meticulous planning beforehand. As urologists, we don’t engage in unprepared battles; thorough preoperative evaluation is half the success. The commonly used international scoring systems, such as the RENAL score and the PAUDA score, which evolved from it, primarily score based on the maximum diameter of the tumor—the higher the score, the greater the surgical difficulty.
Although tumor size is critical in the evaluation, its location is equally significant. Therefore, we developed the first independently developed Chinese kidney cancer surgical scoring system, the “Zhongshan Scoring System (ZS Score),” which evaluates indices related to surgical difficulty, including renal tumor maximum diameter (Ri), tumor location (Location), and tumor depth (Depth) . The higher the cumulative score, the more challenging the surgery; conversely, a lower score indicates shorter operation time, less blood loss, fewer complications, and lower recurrence rates. We also optimized the off-clamp nephron-sparing surgery (NSS) scoring system, proposing the zero ischemia index (ZII), which grades based on the number of renal columns the tumor invades.
Oncology Frontier: There is ongoing debate about the impact of ischemia time on renal function after partial nephrectomy. Current postoperative scoring systems primarily predict tumor recurrence risk. Could you discuss kidney function prediction after partial nephrectomy, incorporating the latest EAU research?
Professor Jianming Guo : Previously, partial nephrectomy mainly focused on ischemia time, with numerous studies conducted in this area. Current postoperative renal cancer prognosis scoring systems include the Mayo Clinic’s SSIGN score and the University of California, Los Angeles’ UISS renal cancer risk classification system. These systems do not specifically target partial nephrectomy but assess recurrence risk and prognosis after renal cancer surgery to guide adjuvant therapy.
With technological advancements, partial nephrectomy has become the mainstream surgery, often performed laparoscopically or robot-assisted. Generally, the shorter the ischemia time, even zero ischemia, the lower the risk of renal function damage. However, a large-scale UroCCR 90 study reported at this EAU (Abstract No. A0624) , involving 3,342 patients undergoing robot-assisted partial nephrectomy (RAPN), found that ischemia time (P=0.04) and anesthetic ASA score ≥II (P=0.007) were significant factors affecting early complications, but ischemia time did not impact patient survival prognosis.
Ischemia time length is usually related to the surgeon’s experience. Another multicenter study (Abstract No. A0629) analyzed 753 laparoscopic and 3,258 robot-assisted partial nephrectomies performed by 119 surgeons, showing no significant correlation between surgical experience and postoperative acute kidney injury (AKI) (P=0.2). A prospective European database study (Abstract No. A0621) involving 1,100 partial nephrectomy patients found that ischemia time did not affect short-term renal function, but underlying conditions like smoking, hypertension, and diabetes were risk factors for AKI.
Therefore, some scholars (Abstract No. P121) suggested using inflammatory factors like C-reactive protein (CRP) to evaluate and predict postoperative renal function damage. Japanese researchers (Abstract No. A0634) used the tumor location index (I index, defined as the distance from the central renal axis to the tumor center + the shortest distance from the renal calyx to the tumor) to classify tumors into high, medium, and low risk groups based on I index (≤15 mm and ≤30 mm), demonstrating high predictive efficiency (AUC=0.858). This “I index” is similar to our “Zhongshan Scoring System (ZS Score),” but our system includes more dimensions, such as tumor location, diameter, and depth. Overall, ischemia time’s long-term impact is minimal, as corroborated by our team’s recent study.
Oncology Frontier: Postoperative scoring systems play a crucial role in guiding follow-up plans. Are there any new research advancements or expert discussions on this topic at the EAU conference?
Professor Jianming Guo : Traditional scoring tools like SSIGN and UISS provide guidance for postoperative follow-up. The 2021 EAU guidelines stratified postoperative renal cancer follow-up into high, medium, and low risk, with corresponding follow-up recommendations. The same year, the “Chinese Consensus on Follow-up Protocols after Urological and Male Reproductive System Tumor Surgery” categorized patients into high, medium, and low risk, with similar follow-up recommendations for medium and high-risk patients.
At this EAU conference, a French multicenter real-world study involving 5,320 postoperative renal cancer patients (including clear cell and non-clear cell renal cancer) showed no significant difference in recurrence risk among low-risk clear cell, low-risk non-clear cell, and medium-risk non-clear cell renal cancer patients, all at 10% (P=0.9). For high-risk non-clear cell renal cancer patients, if imaging was negative at 3 months, their recurrence risk was similar to medium-risk clear cell renal cancer patients (P=0.3). These findings sparked discussions among EAU experts on changing renal cancer follow-up strategies, suggesting that medium-risk clear cell renal cancer might be managed similarly to low-risk cases, and high-risk clear cell renal cancer patients with negative imaging at 3 months could be downgraded to medium-risk follow-up plans.
Oncology Frontier: Artificial intelligence was another major highlight at this year’s EAU conference. You presented a study on AI predicting renal tumor pathology and prognosis using CT. Could you share your thoughts on AI and your research findings?
Professor Jianming Guo : AI was a major highlight at this EAU conference, with dedicated sessions (EAU Guidelines: Artificial intelligence and new clinical insights) and multiple studies on AI applications. Current AI application scenarios include:
Intelligent consultation, such as the widely discussed ChatGPT for disease diagnosis and treatment planning.
Intelligent imaging and pathology, using AI deep learning to enhance diagnostic efficiency and accuracy.
Intelligent surgery, advancing robot-assisted surgery towards “fully automated/unmanned” surgical systems.
Intelligent research, with AI aiding clinical research design and scientific paper writing.
We reported a study (Abstract No. A0458) on using AI to predict renal tumor pathology and clinical outcomes from multi-phase CT, falling under intelligent imaging. Both the model cohort and validation cohorts achieved impressive AUC values (diagnosing benign and malignant diseases AUC: 0.8530.898; identifying invasive renal cancer AUC: 0.7630.792). This diagnostic performance surpasses current radiomics models and RENAL score prediction models, potentially addressing two major clinical challenges: preoperatively determining renal mass benignity to avoid unnecessary nephrectomy and predicting invasive renal cancer to aid in personalized surgical planning.
Our team, in collaboration with Fudan University’s Digital Medicine Center (Researcher Wang Shuo) and Microsoft Research Asia (Researcher Wang Zilong), is conducting cutting-edge research on smart kidney cancer diagnosis and treatment. We have established a kidney cancer pathology-imaging database with over 11,860 whole-slide images and 15,346 CT images from 7 domestic medical centers, involving more than 7,000 patients. Our findings have been accepted by the top-tier international radiology journal Radiology. We welcome further collaboration from domestic and international peers to develop advanced multimodal medical models for kidney cancer research.
Professor Jianming Guo
Professor and Director of the Department of Urology, Zhongshan Hospital, Fudan University
Standing Committee Member, Urology Branch, Chinese Medical Doctor Association
Member of the Oncology Group, Urology Branch, Chinese Medical Association
Vice Chairman and Head of the Oncology Group, Urology Branch, Shanghai Medical Association
Vice Chairman, Urology Tumor Committee, Shanghai Anti-Cancer Association
Vice Chairman, Urology and Andrology Committee, Shanghai Association of Integrative Medicine
Standing Committee Member, Urology Tumor Committee, Chinese Anti-Cancer Association
Standing Committee Member, Male Reproductive System Tumor Committee, Chinese Anti-Cancer Association
Standing Committee Member, Prostate Cancer and Urothelial Cancer Committee, Chinese Society of Clinical Oncology
Vice Chairman, Urology Branch, Chinese Sexology Association
Standing Committee Member and Deputy Secretary-General, World Chinese Urologist Association
Deputy Chairman, Urology Branch, Cross-Strait Medical Exchange Association
Vice Chairman, Urology and Andrology Committee, Chinese Association for Geriatric Health Care
