Stereotactic Radiosurgery for Cavernous Malformations: Kjellberg's Experience with Proton Beam Therapy in 98 Cases at the Harvard Cyclotron

To the Editor: Amin-Hanjani et al. (1) retrospectively evaluated their experience with proton beam therapy in the treatment of surgically inaccessible cavernous malformations (CMs) in 98 patients at the Harvard Cyclotron. The study is well written and is analyzed in great detail. We would like, however, to share a few observations. First, how was "surgically inaccessible" defined? Figure 1, for example, shows a lobar lesion that seems "surgically accessible." Were some lesions that were considered inaccessible early in the study later considered surgically accessible as surgical techniques and stereotactic guidance improved?FIGURE 1: A, T1-weighted MR image of a 40-year-old urologist who presented with hiccups, hemiparesis, and hypesthesia, revealing a lateral medullary CM, for which the patient underwent a far-lateral craniotomy. B, 5-year follow-up MR image showing no evidence of residual or recurrent lesion.Second, the authors state that "unlike angiographic obliteration for arteriovenous malformations, there are no radiographic criteria for therapeutic success in the treatment of cavernous malformations." This may be true for lesions treated by stereotactic radiosurgery. After surgical resection, however, radiographic evidence of the obliteration of these lesions is clear (Fig. 1). Third, the authors' formula for calculating the annual hemorrhage rate (17.4%/yr) may actually reflect the "rehemorrhage rate." To calculate the actual bleed rate would require knowing the total number of years the lesions had been present (2). Using the first bleed as a marker for the start of the observation period and then counting the number of subsequent hemorrhages would decrease the denominator (the number of years of observation during which the patient harbors a CM), thereby increasing the hemorrhage rate. Referral bias could be another confounding factor. Thus, if the annual hemorrhage rate of 2.3% after proton beam therapy is compared with prospective data in the literature (also 2.3%) (3), it is unclear whether the natural history of the disease would be altered. Furthermore, the complication rate associated with the Bragg-peak proton beam therapy (26%) was substantial. To determine whether treatment reduces the number of hemorrhages, patients should be observed at least (on average) as long after treatment as before treatment. In our experience, the interval between hemorrhages can be extremely sporadic; some patients can have three hemorrhages in a week, whereas others may not subsequently hemorrhage for decades. We have found predicting the time of subsequent hemorrhagic episodes to be challenging. Finally, the single patient who underwent autopsy showed no evidence of CM (only radiation necrosis). This finding is significant because the literature (4-6) reflects a lack of histopathological changes in irradiated CM 2 years after stereotactic radiosurgery. Randall W. Porter Paul W. Detwiler Patrick P. Han Robert F. Spetzler Phoenix, Arizona