3D margin measurement after lung tumor ablation accurately predicts local recurrence risk.

3D margin measurement after lung tumor ablation accurately predicts local recurrence risk.

In thermal ablation of lung tumors, adequate ablation margins (zone of destruction extending beyond the tumor) predict local tumor control. This study developed a 4-step lung-specific deformable image registration algorithm to accurately measure 3D ablation margins from pre- and post-procedural CT scans.

The algorithm achieved a mean target registration error of 0.4 mm. Larger ablation margins were independently associated with longer time to local recurrence (subdistribution HR 0.5 per mm increase). With a 2 mm margin threshold, the 2-year local recurrence rate was just 3%. The AUC for predicting 2-year recurrence was 0.86.

Automated 3D margin quantification could become a standard quality metric for lung ablation procedures.

Key Findings

• 4-step deformable image registration algorithm achieved mean TRE of 0.4 mm
• Larger ablation margins independently associated with lower local recurrence (HR 0.5/mm)
• 2-year local recurrence rate of 3% with ≥2 mm margin threshold
• AUC of 0.86 for predicting 2-year local recurrence
• Algorithm enables automated 3D margin quantification from routine CT scans

Implications

3D margin assessment could become standard practice for quality assurance after lung tumor ablation, enabling immediate feedback on whether adequate margins were achieved and guiding re-treatment decisions. The 2 mm threshold is a clinically useful guideline.

Caveats

Single-institution retrospective study (69 patients, 108 tumors); abstract-only. Algorithm trained and tested at one center—external validation needed. Follow-up limited to 2-year outcomes. Applicable primarily to thermal ablation procedures.

Source: Radiology. Imaging cancer — 2026-05-01