Tricortical Skull Anatomy Is a Potential Contraindication for MR-guided Focused Ultrasound

The human thalamus

Transcranial magnetic resonance–guided focused ultrasound (MRgFUS) is now considered safe and effective for treating essential tremor (ET) in adults. However, several factors affect the feasibility of MRgFUS, including the skull density ratio (SDR), and the ratio between the radiodensities of marrow and cortical bone.

The manufacturer of the NeuroAblate 4000 device (InSightec, Haifa, Israel) cites SDR <0.45 as a relative contraindication for MRgFUS. Yet multiple centers have performed successful thalamotomies in patients with lower SDRs.

Physicians at Brigham and Women’s Hospital have successfully treated >90 patients with borderline SDRs (0.32–0.45) and 50 patients with a form of cranial overgrowth called hyperostosis frontalis interna, characterized by nodules bilaterally located on the inner lamina of the frontal bone.

Joshua D. Bernstock, MD, PhD, a resident in the Department of Neurosurgery, G. Rees Cosgrove, MD, director of Epilepsy and Functional Neurosurgery, and colleagues recently described unsuccessful thalamotomy in a patient who displayed hyperostosis calvariae diffusa—expansion of the frontal endocranium in addition to the bones within the skull—as well as tricortical osseous anatomy. In Stereotactic and Functional Neurosurgery, the authors explain why tricortical anatomic variations should be considered possible exclusion criteria when evaluating patient eligibility for MRgFUS.

Case Report

The patient, a 74-year-old right-handed male with ET since his thirties, presented with severe bilateral upper extremity tremor despite maximal medical management. On examination, he displayed a 2+ postural tremor and 3+ intention tremor that was worse on the right than the left. His writing sample and attempted drawing of a spiral were barely legible.

Plain CT showed hyperostosis calvariae diffusa plus the tricortical anatomy—two separate layers of cancellous bone sandwiched between three layers of cortical bone. The SDR was 0.37.

Eight sonications were performed, with neurologic status assessed after each one. No clinical improvement in tremor was noted. Correspondingly, MRI during and after MRgFUS showed no lesion within the left thalamus.

The total sonication time was 50 seconds, the maximum energy was 49.3 kJ, and the maximum temperature reached at the thalamic target was 47°C. The patient reported head pain during the prolonged pulses that was abnormally severe for MRgFUS.

At the time of this writing, the patient was considering deep brain stimulation.

Systematic Review

The team identified a randomized controlled trial, a cohort study and a case report that described three other patients with hyperostosis for whom MRgFUS failed. Their SDRs ranged from 0.37 to ≥0.45. The Brigham and Women’s case was distinct concerning the diffuse hyperostosis encountered and its tricortical configuration.

Two of the three patients identified in these reports had some improvement in their tremor after MRgFUS, but no clinically relevant brain lesion was achieved for any of them. The authors of the cohort study reported excluding four patients with severe hyperostosis frontalis from MRgFUS, citing inherent unreliability of SDR measurement in such patients.

Applying the Findings to the Clinic

The SDR represents an average—it is calculated for multiple subsections of the cranial vault. As a predictive metric, the SDR partially accounts for the effects of variation in the cortical and cancellous layers.

However, the SDR doesn’t account for interface effects on transmission, attenuation, and reflection of ultrasound that will differ between a three-layer model of the skull (inner table, diploe, outer table) and a five-layer model (inner table, diploe, middle table, diploe, outer table).

Thus, the SDR may be of limited use for determining the feasibility of MRgFUS for patients with tricortical osseous anatomy. Furthermore, this anatomy may be impossible to overcome using current techniques (e.g., maximizing the number of ultrasound elements and minimizing no-pass regions) and clinically acceptable energy dosing.

Leave a Reply