Protective Heat and Cold Shields Make Tumor Ablation Near Nerves Dramatically Safer

For patients with painful tumors embedded in bone or soft tissue, the promise of thermal ablation has always come with a terrifying caveat: what happens when the tumor sits millimeters from the spinal cord, a major nerve, or the bowel? A new decade-spanning study from Washington University in St. Louis delivers the most reassuring answer yet. In a retrospective analysis of 364 percutaneous ablation procedures performed between January 2013 and August 2023, researchers found that cryoablation and radiofrequency ablation (RFA) of musculoskeletal tumors carried overall complication rates of just 4 percent and 3.4 percent respectively, with zero procedure-related deaths. Even more striking, not a single complication occurred among patients whose radiofrequency procedures included active thermoprotection, a finding that could reshape how interventional radiologists think about the boundaries of what is safely treatable.

The research, published open access in CVIR Oncology, tackles a question that has long haunted the field: how close is too close? Tumors metastatic to the spine and pelvis frequently press against the spinal cord, the cauda equina, peripheral nerves, and visceral organs, and the thermal energy that destroys tumor cells cannot distinguish friend from foe. The magnitude of unintended injury depends on the absolute temperature achieved, the duration of the thermal dose, the temperature sensitivity of neighboring tissues, and the distance between the at-risk structure and the edge of the ablation zone. Yet despite two decades of advances, the safe distance between a lesion and a vital structure had never been firmly established, leaving clinicians to rely on intuition and case-by-case judgment.

The study team, led by Carl Stokes and colleagues at the Mallinckrodt Institute of Radiology, analyzed 100 cryoablation procedures in 91 patients and 264 RFA procedures, all performed by attending musculoskeletal radiologists with five to twenty years of experience. Roughly four in five ablations targeted metastatic lesions, with the remainder treating primary tumors. The cryoablation cohort ranged from 5 to 82 years old, while RFA patients ranged from 4 to 89, underscoring how broadly these techniques are deployed across age groups and disease states. Lesions treated with cryoablation were predominantly osseous, most often in the pelvis or chest wall, while the overwhelming majority of RFA cases, 87.1 percent, involved the spine.

The technical arsenal documented in the study reads like a catalog of modern interventional radiology. Cryoablation was performed with the Visual-ICE system using a family of cryoprobes selected for target size and location, with standard two freeze-thaw cycles and CT images acquired every two to three minutes to track the growing ice ball. Skin warming and tissue displacement using carbon dioxide gas or hydrodissection fluid, either saline, iodinated contrast, or 5 percent dextrose, were the most common protective maneuvers. In select high-stakes cases near neural structures, transcranial electrical motor evoked potentials and somatosensory evoked potentials were monitored by certified neuroelectrophysiologic technologists with more than a decade of experience each, providing a real-time electrical readout of spinal cord and nerve function during the freeze.

RFA followed a different logic. Most procedures used a bipolar probe system under fluoroscopic guidance, with built-in thermocouples tracking tissue impedance and temperature throughout ablation. In 82.8 percent of RFA cases, patient biofeedback was the sole protective mechanism, meaning an awake patient reporting sensations of heat, burning, or tingling served as the early warning system. In the remaining 17.2 percent, clinicians added active measures such as carbon dioxide dissection or hydrodissection to physically push vulnerable structures out of harm’s way. The distinction between passive protection, which monitors for impending injury, and active protection, which physically prevents it, proved central to the study’s conclusions.

The numbers tell a compelling story. Thermoprotective techniques measurably widened the safety margin in every anatomic category. For cryoablation lesions near the spinal cord or cauda equina, the mean distance increased from 1.0 centimeter to 1.3 centimeters after protection. Near peripheral nerves, the mean distance grew from 0.9 to 1.3 centimeters. The most dramatic gains appeared near visceral structures such as bowel and bladder, where the average distance of a mere 0.3 centimeters, essentially touching, was expanded to 1.3 centimeters. In the RFA cohort, lesions near the spinal canal went from a mean of 0.4 centimeters to 0.9 centimeters, and lesions adjacent to peripheral nerves more than tripled their separation, from 0.4 to 1.2 centimeters.

Complications, when they occurred, were mostly transient or self-limited. In the cryoablation group, the most common event was grade-2 transient weakness in two patients, followed by one case each of grade-3 hypertensive urgency and grade-4 hemorrhage during ablation of a chest wall hemangiopericytoma, which required emergency embolization of an intercostal artery. The single neurological event involved transient left leg weakness that resolved within 24 hours without intervention, occurring in a lesion that remained less than one centimeter from the thoracic spinal cord even after protection. In the RFA group, complications included arrhythmia, nausea, post-procedural pain requiring hospital admission, hypertensive urgency, and one venous spinal cord infarction, all in patients receiving passive protection only. Zero complications were recorded in the 44 patients who received active thermoprotection during RFA.

Two risk patterns emerged with particular clarity. First, procedural duration correlated with complications: patients who experienced adverse events averaged 217.5 minutes of cryoablation time and 137.4 minutes of RFA time, longer than cohort norms, likely reflecting case complexity. Second, the number of cryoprobes used proved a striking predictor. Complicated cases averaged 13.5 probes compared with a cohort average of 4.1, more than a threefold difference, confirming a trend previously described in the literature. Notably, concomitant cementoplasty, biopsy, and neurophysiologic monitoring added no meaningful procedure time or complication risk, and the exothermic polymerization of bone cement was insufficient to cause tumor necrosis at the bone-cement interface, corroborating earlier work.

The study’s most consequential message is a reframing of contraindications. Because no complications occurred in RFA with active protection, and only one transient deficit appeared in cryoablation even when lesions sat within a centimeter of the spinal cord, the authors argue that thermoprotection itself, rather than strict adherence to any distance threshold, is the dominant factor in mitigating risk. The clinical significance of adding a centimeter of separation remains uncertain in absolute terms, but the pattern across hundreds of cases, including 15 extremity cryoablations near the skin, a site where prior studies report skin burn rates around 2.8 percent, suggests protective techniques work. No dermal complications, infections, or superficial burns occurred in this cohort.

The authors are candid about limitations: the retrospective single-institution design, the low event rate that prevented statistically powered multivariate analysis, a single reader for image review, and the absence of recorded time spent on protective maneuvers. Still, the scale of the series, spanning more than a decade and hundreds of procedures, lends weight to its central recommendation: prophylactic use of active and passive thermoprotection whenever critical structures are at risk, and a firm rejection of proximity alone as a reason to deny patients ablation. For the growing number of patients with painful musculoskeletal tumors, many of them metastatic and with limited options, the message is that the no-go zones around the spine and nerves may be far smaller than once feared, provided the right shields are in place.

Subject of Research: Safety of thermoprotective techniques during image-guided cryoablation and radiofrequency ablation of musculoskeletal tumors

Article Title: Too close for comfort? Assessing the impact of thermoprotective techniques in image-guided ablation of musculoskeletal tumors

Article References: Stokes, C., Ly, M., Tomasian, A., Imaoka, R., Vander Velde, T. L., Northrup, B. E., Cusworth, B., & Jennings, J. W. (2025). Too close for comfort? Assessing the impact of thermoprotective techniques in image-guided ablation of musculoskeletal tumors. CVIR Oncology, 1(1), Article 18. https://doi.org/10.1007/s44343-025-00014-2

Image Credits: AI Generated

DOI: 10.1007/s44343-025-00014-2

Keywords: cryoablation, radiofrequency ablation, musculoskeletal tumors, thermoprotection, interventional radiology, spinal metastases, hydrodissection, neurophysiologic monitoring, pain palliation, interventional oncology, patient safety, image-guided ablation