Predictors of outcome in surgery for sporadic cerebellar hemangioblastomas.

Hemangioblastomas are benign neoplasms, consisting of stromal cells surrounded by dense capillaries (3). Hemangioblastomas are most likely to be seen in adults, accounting for 2% of intracranial tumors and 7%–10% of posterior fossa tumors (9). Microsurgical resection is the treatment of choice for virtually all hemangioblastomas, and it is usually curative, but complete removal of large tumors with a rich vascular supply or deeply embedded in the brainstem may be challenging. External-beam radiation therapy may be useful in the management of these tumors. However, the tendency of hemangioblastomas to progress, recur, and generate new lesions suggests that this method of radiation may not be ideal because of the potential for cumulative neuronal damage. Several studies have suggested that radiosurgery is effective for solid hemangioblastomas, resulting in shrinkage or stabilization with low complication rates 4, 6, 8 and 11. Nevertheless, as Wanebo et al. (10) reported in their study, hemangioblastomas often had multiple periods of tumor growth alternated with periods of quiescence; this means that many untreated tumors sometimes remain the same size for several years. Because of the relatively short follow-up period after radiation reported in almost all of these studies, it is still difficult to determine whether the absence of tumor growth is attributable to a quiescent phase or to a positive response to the radiation therapy. However, when surgery is unsuitable, radiosurgery may be the preferred treatment for residual or recurrent and newly diagnosed solid hemangioblastomas with a diameter <3 cm and without mass effect. Hemangioblastomas may have a solid or cystic configuration, and their clinical behavior is often different. For this reason, solid hemangioblastomas have been investigated as a separate clinical entity 9 and 14. Hemangioblastomas can arise either in the setting of von Hippel-Lindau (VHL) disease or, more frequently, as a solitary sporadic lesion (1). Despite similar histologic characteristics, sporadic and syndromic hemangioblastomas have different biologic and clinical features 7 and 12. Nevertheless, in the most recently published data, clinical studies have grouped together sporadic hemangioblastomas and hemangioblastomas associated with VHL disease or addressed specifically syndromic hemangioblastomas despite their lower incidence. In this issue of WORLD NEUROSURGERY, Fukuda et al. provide a concise discussion about clinical and biological behavior of sporadic cerebellar hemangioblastomas. They reviewed a series of 36 surgically treated patients to evaluate factors that may predict early and long-term outcome. The study demonstrated a significant correlation between solid configuration of tumor and outcome of the patients. The finding that patients with solid hemangioblastomas showed poor outcomes more frequently than patients with cystic tumors has been previously reported. However, to date, there are insufficient data in the literature about this specific issue in terms of both outcome and surgical removal. In the light of my experience, I believe that the present article supports the general theme that in brain tumors (especially in benign ones), outcome mainly depends on the extent of resection. Considering the results of the study, among the solid tumor group, 5 of 19 patients (26%) had a partial or subtotal removal, 4 of 19 (21%) developed postoperative hematoma (3 of the patients had hemorrhage from the residual tumors), and 4 of the 5 patients with partial or subtotal removal had poor long-term outcome. In hemangioblastomas, factors affecting extent of resection or complications are primarily deep location, rich vascular supply, and association with VHL disease. The last-mentioned is important because, despite similar histologic characteristics, syndromic hemangioblastomas may show aggressive development through rapidly evolving or multifocal lesions. Because solid hemangioblastomas have usually an abundant blood supply, surgical risk during the removal of solid hemangioblastomas is the uncontrollable bleeding. Some authors suggested that selective preoperative embolization of feeding arteries may reduce intraoperative blood loss and assist the en bloc resection of tumors (13). However, because embolization involves potential complications such as ischemia, bleeding, and increased intracranial pressure, the use of preoperative embolization is still controversial. In my opinion, an adequate microsurgical technique and a better understanding of the vascular pattern of this type of tumor can favor a complete removal with acceptable mortality and morbidity. Location of the tumors may influence the extent of the resection. Le Reste et al. (5) reported surgical results of 38 patients harboring sporadic hemangioblastomas (79% located in the cerebellum). They concluded that surgery of sporadic hemangioblastomas provides good quality of life and tumor control except in cases of tumors located in the medulla oblongata, which are the most difficult to be removed completely. Solid hemangioblastomas mostly occur in the brainstem, cerebellar vermis, and spinal cord, whereas cystic tumors are predominantly located in the cerebellar hemispheres (9). The involvement of brainstem or cranial nerves is easy to document with magnetic resonance imaging. What is difficult to understand preoperatively is how tightly the tumor is adherent to these structures. The latter is the true key point to achieving a radical resection of these tumors. In my personal experience, all patients who received subtotal resection or reported complications had large tumors with a rich vascular supply or that were deeply embedded in the brainstem (unpublished data F. Tomasello, M.D.). Another important topic in hemangioblastomas is tumor recurrence. The most common cause of morbidity and mortality in such tumors is related to frequent tumor recurrence and incomplete resection at first surgery. Conway et al. (2) reported that partial resection led to recurrence in 8 of 40 patients (20%) and additional morbidity in 6 of 40 patients (15%). There have been 2 histologic hemangioblastoma variants recognized: reticular and cellular. Cellular hemangioblastomas are less common than the reticular type, and they recur more frequently (4). The recurrence rate of patients harboring sporadic hemangioblastomas ranged from 5%–17% 2 and 9, whereas patients with VHL disease had a recurrence rate that ranged from 17%–75% 2 and 5. In their article, Fukuda et al. report a recurrence rate of 13% (4 of 30 patients); in addition, among 4 recurrent tumors, 3 had a solid configuration and poor long-term outcome. Recurrence of hemangioblastomas seems to be related to both extent of the resection and biologic behavior of these tumors. In conclusion, the study by Fukuda et al. demonstrates well that patients with solid hemangioblastomas had a worse prognosis. Solid hemangioblastomas are more often located in the brainstem compared with cystic tumours (13) and have a more plentiful blood supply (9). The chances of total resection of such tumors are decreased, whereas the recurrence rate is increased. Taking into account that association does not always prove causation, I believe that the findings reported by Fukuda et al. are adequate to explain why solid configuration of hemangioblastomas may influence the prognosis; however, they may be insufficient to establish whether the solid configuration of the tumor per se or the feasibility of a radical surgery influences patient outcome. I believe that, along with biologic behavior, radical resection of these tumors, although sometimes quite tricky, remains the most meaningful prognostic factor. Nevertheless, owing to the small number of cases in the published series and the still unclear biologic behavior, it is difficult at the present time to draw firm conclusions in these issues. However, the study by Fukuda et al. provides useful information to improve our prognostic evaluation in this specific type of brain tumor.