Hemangiomas of the vertebral bodies are commonly encountered, being found in 10–12% of patients in autopsy series (1). The vast majority of these are solitary, are discovered incidentally, and are seldom symptomatic. Symptoms, when they do occur, can vary from chronic, poorly defined pain to cord compression and paraplegia. These undesirable consequences are usually caused by compression fracturing, hematoma, epidural extension of a hemangioma, or bony expansion of the vertebral body (1, 2).
A variety of treatment options exists for addressing symptomatic hemangiomas. Surgery is well established, and has been proved effective. A surgical approach is technically challenging because of the markedly vascular nature of these congenital vascular malformations. The introduction of presurgical embolization has facilitated surgery significantly, and is now commonly performed (2). Scattered reports of treatment of symptomatic lesions with embolization alone exist; however, treating these lesions exclusively with embolization is generally inadequate in the majority of cases. Radiotherapy also has been used in treatment of these lesions, usually as a complement to surgery. Radiation therapy alone is often inadequate.
In 1994, Heiss et al reported two patients with spinal cord compression who improved appreciably after direct percutaneous injection of alcohol into vertebral body hemangiomas (3), and then they reported on seven patients 2 years later (4). In this issue of the AJNR, Goyal et al (page 1091) report their experience with 14 patients they treated with this technique, considerably expanding the reported number of patients in the literature. The authors found that 86% of their patients demonstrated appreciable clinical improvement in a manner analogous to those reported by Heiss et al.
Direct alcohol injection has several advantages. It is readily performed using conventional bone biopsy equipment and CT guidance without requiring the facilities of an operating room or the sophisticated equipment needed in an interventional angiography suite or radiotherapy unit. Only intravenous sedation and analgesia are required. The procedure is less invasive than open surgery, and does not entail the risks of significant blood loss that may be encountered with invasive therapy. Embolization, although extremely helpful in preoperative treatment of these lesions, cannot directly reach or destroy the hemangioma because of the presence of an intervening capillary bed separating feeding arteries from the hemangioma (2). The risks of embolization to the spinal arterial supply, as well as surrounding normal tissue, is well recognized and a dreaded, if fortunately infrequent, complication of embolization (2). Radiation therapy carries the risk of possible toxicity to the spinal cord. Direct injection of alcohol, on the other hand, destroys the hemangioma without these problems.
Goyal et al, however, omit an important caveat. The authors encountered two significant complications. The first was penetration of the pleural space in treatment of a thoracic hemangioma, with complicating empyema. A less acute problem was also noted in that a treated vertebral body underwent collapse subsequent to ethanol ablation; interestingly, Heiss encountered the same problem (4). This is likely attributable to the osteonecrosis that can occur with injection of alcohol. Although data is inadequate to be certain, it may be that larger hemangiomas or hemangiomas treated with larger volumes of alcohol may be more prone to developing this complication, as osteonecrosis may be more likely to develop in those situations.
All patients, except one, were being treated for progressive neurologic dysfunction, and all patients had transient worsening of neurologic status. We presume this is related to alcohol-induced inflammatory changes, including swelling of the lesion. Steroids were administered empirically to all patients with possibly reduced inflammatory response. For those considering this treatment for their patients, neurologic worsening will have to be an acceptable and accepted event that will usually get better—although with increasing experience there may be instances when a neurologic improvement will not occur. Therefore, treating patients who present for reasons other than neurologic compromise may be difficult to justify.
Regarding technique, the injection of alcohol need not be “a blind procedure” as the authors have indicated. Without significantly changing the absolute nature of the alcohol, we have made alcohol radioopaque by mixing it in metrizamide powder when using it for other vascular lesions. This might allow for a more precise volume administration and reduce the “subjective assessment” of the rate of opacification.
The precise role for percutaneous alcohol ablation remains unclear. Unlike embolization, this technique allows actual destruction of the hemangioma. This presumably would make the risk of recurrence after treatment less likely, although in the current series recurrent (or more likely residual) hemangioma was noted in one instance. Because it appears that collapse of the vertebral body may be a significant complication of this procedure, its use as the sole treatment for symptomatic hemangiomas is, at the present time, questionable. It may,however, play an important role as an adjunct to surgery, allowing devascularization of the hemangioma without entailing the risks of angiography.
Another rapidly expanding technique is that of percutaneous vertebroplasty where image-guided injection of methyl methacrylate bone cement directly into the hemangioma is performed (5). Experience with this technique is increasing rapidly as injection of cement does have the advantage of providing structural support at the site of the ablated lesion. Long-term follow-up with this technique also is presently lacking. Cement, of course, creates a permanent incompressible cast of the hemangioma, which may be a problem in patients who have extensive expansion of the hemangioma with ballooning of the vertebral body or involvement of surrounding soft tissues or epidural space.
- Copyright © American Society of Neuroradiology