AI-measured preoperative body composition predicts complications after radical cystectomy

An artificial intelligence (AI)-based assessment of preoperative body composition can determine patients who might develop early complications following radical cystectomy (RCx), a study has shown.

A group of researchers assessed the cystectomy registry of a tertiary referral centre for patients who underwent RCx between 2009 and 2017 for bladder cancer. They identified 843 RCx patients with CT imaging within 90 days of preceding surgery, allowing for the extraction of body composition parameters by AI.

The researchers evaluated complications within 90 days of surgery, including wound, infectious, and major complications; readmission; and death. They also used multivariable logistic regression to examine the association between pre-RCx measures and post-RCx complications.

Increasing subcutaneous adipose tissue correlated with more wound complications, while increasing visceral adipose tissue tended to increase infectious-related complications. Following adjustments for patient characteristics, each 10-cm² increase in fat mass appeared to increase the odds of infectious (odds ratio [OR], 1.04; p=0.002) and wound (OR, 1.06; p<0.001) complications.

Multivariable analysis revealed the association between a higher preoperative skeletal muscle index and a reduced likelihood of major complications (OR, 0.75 for every 10 cm²; p=0.008). Conversely, a higher intramuscular adipose tissue predicted a greater chance of major complications (OR, 1.93; p=0.008).

“These measurements, in addition to patient (Eastern Cooperative Oncology Group performance status and smoking status) and surgical (robotic approach and continent diversion) characteristics, can then be used to individualize patient counselling and facilitate triage of nutritional and rehabilitation efforts,” the researchers said.