Stereotactic Image-Guided Microwave Ablation for Malignant Liver Tumors—A Multivariable Accuracy and Efficacy Analysis
Pascale Tinguely, Lorenz Frehner, Anja Lachenmayer, Vanessa Banz, Stefan Weber, Daniel Candinas, Martin Maurer
Abstract
Background: Therapeutic success of thermal ablation for liver tumors depends on precise placement of ablation probes and complete tumor destruction with a safety margin. We investigated factors influencing targeting accuracy and treatment efficacy of percutaneous stereotactic image-guided microwave ablation (SMWA) for malignant liver neoplasms. Materials and methods: All consecutive patients treated with SMWA for malignant liver tumors in three years were analyzed. A computed tomography-based navigation system was used for ablation probe trajectory planning, stereotactic probe positioning, validation of probe positions and validation of ablation zones. Factors potentially influencing targeting accuracy (target positioning error, TPE) and treatment efficacy within 6 months (ablation site recurrence, ASR) were analyzed in a multivariable regression model, including challenging lesion locations (liver segments I, VII and VIII; subphrenic location). Results: Three-hundred-and-one lesions (174 hepatocellular carcinomas, 87 colorectal liver metastases, 17 neuroendocrine tumors, 23 others) were targeted in 191 interventions in 153 patients. The median TPE per ablation probe was 2.9 ± 2.3 millimeters (n = 384). Correction of ablation probe positions by repositioning was necessary in 4 out of 301 lesions (1%). Factors significantly influencing targeting accuracy were cirrhosis (R 0.67, CI 0.22 - 1.12) and targeting trajectory length (R 0.21, CI 0.12 - 0.29). Factors significantly influencing early ASR were lesion size >30 mm (OR 5.22, CI 2.44 - 11.19) and TPE >5 mm (OR 2.48, CI 1.06 – 5.78). Challenging lesion locations had no significant influence on targeting accuracy or early ASR. Conclusions: SMWA allows precise and effective treatment of malignant liver tumors even for lesions in challenging locations, with treatment efficacy depending on targeting accuracy in our model. Allowing for many tumors to be safely reached, this possibly broadens treatment eligibility for patients with otherwise difficult-to-target tumors.