Artificial intelligence (AI) assessment of total kidney volume (TKV) on computed tomography (CT) scans may emerge as a significant prognostic marker for patients being treated with 177Lu-PSMA-I&T therapy for metastatic castration-resistant prostate cancer (mCRPC).
For a retrospective study, recently published in Radiology, researchers reviewed automated TKV measurements from deep learning segmentation (Total Segmentator) of CT scans in 121 patients with mCRPC. All patients in the cohort had at least four cycles of 177Lu-PSMA radioligand therapy (Pluvicto, Novartis), according to the study.
The study authors found that a decrease of TKV of 10 percent or greater at six months had a 90 percent area under the receiver operating characteristic curve (AUC) for predicting a 30 percent or greater decline of the estimated glomerular filtration rates (eGFR) at one year.
The aforementioned TKV marker also had an 85 percent sensitivity rate and a 96 percent specificity rate for predicting one-year decline of eGFR of 30 percent or greater, according to the researchers.
“Automated TKV assessment on standard-of-care CT images predicted deterioration of renal function 12 months after 177Lu-PSMA-I&T initiation in metastatic castration-resistant prostate cancer,” wrote lead study author Lisa Steinhelfer, M.D., who is affiliated with the Institute for Diagnostic and Interventional Radiology at the School of Medicine at the Technical University of Munich in Germany, and colleagues.
Three Key Takeaways
1. AI-derived TKV as a prognostic marker. Automated AI-based total kidney volume (TKV) assessment on CT scans can serve as a strong prognostic marker for predicting renal function decline in patients undergoing 177Lu-PSMA-I&T therapy for metastatic castration-resistant prostate cancer (mCRPC).
2. High predictive accuracy for renal function decline. A decrease in TKV of 10 percent or more at six months had a 90 percent AUC for predicting a 30 percent or greater decline in estimated glomerular filtration rate (eGFR) at one year with high sensitivity (85 percent) and specificity (96 percent).
3. Potential clinical implications for treatment monitoring. Given the high accuracy of TKV as a predictive tool, AI-assisted TKV assessment may help identify patients at risk for radiation-induced nephropathy, potentially guiding early interventions and treatment modifications in mCRPC management.
Noting that the mean age of the cohort was 76 and the commonly delayed presentation of radiation-induced nephropathy, the researchers maintained the study findings are particularly relevant given the increasing research evaluating the use of 177Lu-PSMA radioligand therapy to treat less advanced prostate cancer.
“The volume decrease observed in our patient cohort far exceeds physiologic decline, strongly suggesting radiation-induced renal atrophy as the primary cause,” emphasized Steinhelfer and colleagues.
(Editor’s note: For related content, see “Could Pluvicto Have a Role in Taxane-Naïve mCRPC?: an Interview with Oliver Sartor, MD?,” “What a New Meta-Analysis Reveals About PET/CT Radiotracers for csPCa” and “What SPECT/CT May Reveal About Early Treatment Response for Pluvicto in Patients with mCRPC.”)
Beyond the inherent limitations of a retrospective single-center study, the authors noted the lack of a control group precluded analysis of confounding factors that may have contributed to TKV decreases. They also acknowledged that the timing of the eGFR assessments may not have fully reflected dynamic changes in renal function.
