Findings from the COVER-ALL trial suggest that software-based assessment of the minimal ablative margin during percutaneous thermal ablation significantly improves treatment precision and may reduce local tumour progression in liver cancer patients.
Published in The Lancet Gastroenterology & Hepatology, the randomised phase 2 trial represents the first prospective evaluation of a software-driven margin confirmation method, which integrates artificial intelligence (AI)-based autosegmentation and biomechanical deformable image registration.
Led by Bruno Odisio at The University of Texas MD Anderson Cancer Center in Houston, USA, the trial enrolled 100 patients with up to three liver tumours undergoing computed tomography (CT)-guided thermal ablation. Participants were randomised 1:1 intraprocedurally to receive either standard visual assessment or the experimental software-based margin analysis tool. The primary endpoint of the study was the post-ablation intraprocedural minimal ablative margin.
Patients were enrolled and treated with thermal ablation between 15 June 2020 and 5 October 2023. The interim analysis showed that the software-based assessment group achieved a mean ablative margin of 5.9mm, compared to 2.2mm in the visual assessment group, which prompted early halting of enrolment in the control group. Only 15% of patients in the control group reached the optimal 5mm threshold, versus 75% in the experimental arm.
Odisio and colleagues subsequently enrolled a further 50 patients to the experimental group. In this non-randomised cohort, the average margin further increased to 7.2mm, with 84% achieving optimal coverage.
Secondary endpoints, including the two-year cumulative incidence of local tumour progression, also favoured the software-based group, with a progression rate of 5%, compared to 16% in the visual assessment group. Although the authors note that this difference did not reach statistical significance, they believe that the trend is suggestive of the software’s potential long-term benefit.
“This is, to the best of our knowledge, the first randomised trial to establish the feasibility and efficacy of software-based assessment as an intraprocedural tool for optimising the minimal ablative margin during the thermal ablation of liver tumours,” Odisio et al write. The researchers assert that their study adds to the “growing evidence that this approach allows accurate quantification of the minimal ablative margin, which is predictive of local tumour progression”.
The COVER-ALL system uses biomechanical deformable image registration and AI-based autosegmentation algorithms to quantify the shortest distance between the tumour and ablation zone during the procedure. It provides visual feedback and real-time margin calculations to guide overlapping or repeat ablations if necessary, the authors describe.
Odisio and colleagues conducted a clinical usability survey using the Likert scale, which found that the software was rated “highly” by interventional radiologists, with 91% finding it helpful for evaluating technical success, and 89% agreeing it improved understanding of applicator positioning relative to the target tumour.
The use of software guidance was associated with a greater frequency of overlapping and repeat ablations, reflecting increased precision rather than complexity. Odisio et al note that procedure times were longer in the experimental arm, but without increased adverse events. The rate of grade 1–3 adverse events was low overall (5%) and comparable between groups.
The study authors note that historical overestimation of the minimal ablative margin using visual methods may explain the high variability in local control outcomes seen across institutions. By introducing a reproducible, objective approach, they believe that the COVER-ALL platform offers a means of “standardising intraprocedural decision-making”.
While acknowledging the study’s limitations— including its single-centre setting and small cohort— the researchers maintain that the evidence supports broader implementation of the software. “These results highlight the potential of integrating software-based assessment into the standard-of-care for liver thermal ablation, given its ability to improve the minimal ablative margin and potentially reduce rates of local tumour progression,” Odisio and colleagues conclude.
