✓ Radiosurgical treatment with the gamma knife for venous angiomas was used as an alternative to microsurgical removal in order to avoid abrupt cessation of venous drainage, which may be shared by the venous angioma and important parts of the brain. Thirteen cases of venous angioma were treated between 1977 and 1991. In two cases cavernous angiomas were also present and in one case a distant arteriovenous malformation (AVM) was also found. In two cases the angioma shared the venous drainage with an adjoining AVM; this is the first description of such pathology. For venous angiomas irradiation was prescribed to cover at least the convergence of the medullary veins. For AVM's close to a venous angioma the treatment was exclusively prescribed to the AVM nidus. After treatment, complete obliteration of the venous angioma was observed in one case, partial obliteration was observed in three cases, and five venous angiomas were unaffected by the treatment. Undue effects of radiation occurred in four cases: one focal edema and three radionecroses. Extirpation of the radionecrotic tissue 6 months after radiosurgery was necessary in one case. In the other three cases, the venous angioma was observed to be completely or partially obliterated, or unaffected by the treatment (one case each). In two cases of combined AVM and venous angioma, complete obliteration of the treatment AVM nidus was obtained. It is concluded that radiosurgery for venous angioma, although conceptually attractive, still does not fulfill the rigid criteria of minimal risk which must be set for the treatment of a lesion with a benign natural history.
Christer Lindquist, Wan-Yio Guo, Bengt Karlsson, and Ladislau Steiner
Report of two cases
Bengt Karlsson, Kaj Ericson, Lars Kihlström, and Per Grane
✓ In a series of 22 patients treated with gamma knife surgery for brain metastasis in whom biopsy specimens were obtained via stereotactically guided procedures before the radiosurgical treatment was administered, two cases with evidence of tumor seeding were observed on subsequent follow-up examination. These findings contradict the opinion that the risk for tumor spread after a biopsy is negligible. This evidence may be explained by the fact that radiosurgery leaves the surrounding tissue unaffected by the treatment, which results in preserved anatomy around the tumor. This allows the surgeon to define the previous biopsy channel and, consequently, whether a distant tumor recurrence may have resulted from tumor seeding related to the biopsy procedure. Additionally, radiosurgical treatment leaves tumor cells that may have been spread as a result of the biopsy unaffected, giving them the potential to divide and develop into a new tumor. In contrast to this, microsurgical removal of the tumor will affect the surrounding tissue, making it impossible to detect whether new metastases are resulting from seeding. Furthermore, conventional fractionated radiation therapy will sterilize tumor cells that may have spread, thus making it impossible for these cells to regrow.
The authors conclude that the risk for tumor seeding following a stereotactically guided biopsy may be higher than previously assumed.
Patrick Hanssens, Bengt Karlsson, Tseng Tsai Yeo, Ning Chou, and Guus Beute
The aim of this study was to assess the order of micrometastases that can be detected with high-resolution MR imaging at the time of Gamma Knife surgery (GKS), and to estimate the impact this has on the time until and incidence of distant recurrences.
A consecutive series of 835 patients with brain metastases treated with GKS in a 7-year period, excluding patients in whom earlier brain metastases were treated with other modalities, were retrospectively analyzed. In all patients GKS was based on high Gd–dose (0.3 mmol/kg), high-resolution stereotactic MR imaging. These images were compared with the standard pretreatment MR images, and the difference in number of metastases found was analyzed. The distant recurrence rate following GKS was compared with that found in a prospective randomized study (Aoyama et al.) comparing radiosurgery to radiosurgery plus prophylactic whole-brain radiation therapy.
New tumors were diagnosed in 40% (95% CI 36%–43%) of all patients as well as in the majority of patients with multiple lesions found on the diagnostic scan. The more tumors there were on the diagnostic scan, the higher the likelihood of detecting additional lesions with high-resolution imaging. It was calculated that approximately 50% of the micrometastases present at the time of GKS could be diagnosed with high-resolution imaging, which decreased the incidence of and delayed the time for the development of distant recurrences.
Additional brain metastases can be diagnosed in 40% of patients by using high-resolution imaging. Thus, radiosurgical treatments based on high-resolution stereotactic MR imaging decrease the incidence of and lengthen the time to distant recurrences.
Ladislau Steiner, Christer Lindquist, Wayne Cail, Bengt Karlsson, and Melita Steiner
Bengt Karlsson, Lars Kihlström, Christer Lindquist, Kaj Ericson, and Ladislau Steiner
Object. The authors examined 22 patients with cavernous malformations (CMs) who had undergone gamma knife radiosurgery (GKRS) to assess the value of this procedure in treating these lesions.
Methods. At the Karolinska Hospital, GKRS was used to treat 23 patients with CMs during the period of 1985 through 1996. One of the patients was lost to follow up and the treatment results of the 22 remaining patients were analyzed. In the first half of the series, the CMs were treated with high doses of radiation (> 15-Gy dose to the periphery); in the second half of the series, lower doses were used.
Nine of the 22 patients suffered a post-GKRS hemorrhage and six developed a radiation-induced complication (two of these patients experienced both). Some time after GKRS was performed, surgical removal of the CM had to be undertaken in four patients because of hemorrhage and in two patients because of radiation-induced complications. Four of the nine patients who experienced no post-GKRS hemorrhage or radiation-induced complication were treated before 1990; recent magnetic resonance imaging revealed a decrease in the size of the CM in three of these individuals and no size change in the other.
The annual post-GKRS hemorrhage rate was 8% in this group. There was a trend in the hemorrhage rate to decrease 4 years postsurgery. There was also a trend for higher radiation doses administered to the periphery of the lesion to result in a lower risk of posttreatment hemorrhage. However, it could not be concluded whether GKRS affects the natural course of a CM. The incidence of radiation-induced complications was approximately seven times higher than that expected if the same number of patients had been treated by GKRS with the same radiation dose distributions for arteriovenous malformations instead of CMs.
Conclusions. The high incidence of radiation-induced complications does not seem to justify the limited protection the treatment may afford in only exceptional cases. A prospective randomized study is needed before the role of radiosurgery in the management of these lesions can be defined. Until such a study has proved differently, a caveat must be raised for the treatment of CM with GKRS.
Bengt Karlsson, Hidefumi Jokura, Masaaki Yamamoto, Michael Söderman, and Ingmar Lax
The results of a novel radiosurgical approach to treat large arteriovenous malformations (AVMs) with repeated radiosurgery are presented and discussed.
The outcome was studied following repeated Gamma Knife surgery (GKS) for large AVMs, defined as a nidus volume of 9 ml or more. The philosophy was to treat the whole AVM with a low dose of radiation (≥ 10 Gy), and to repeat the treatment if the AVM shrank but was not obliterated. The study included 133 patients with AVMs treated at one of three different institutions. Clinical information was available for all patients, and complete radiological follow-up was available in 89 patients after the first treatment, and in 19 after the second treatment.
The estimated obliteration rate following repeated GKS was 62%. Four patients (3%) developed neurological deficits caused by the radiation, whereas five others (4%) developed cystic changes. The annual incidence of hemorrhage was high (7%), of which 35% occurred within the 1st year after the first treatment.
Repeated radiosurgery seems to be a viable option for some AVMs considered to be too large for conventional radiosurgical treatment. The incidence of posttreatment hemorrhages seems to be a larger clinical problem than radiation-induced complications.
Bengt Karlsson, Patrick Hanssens, Robert Wolff, Michael Söderman, Christer Lindquist, and Guus Beute
The aim of this study was to analyze factors influencing survival time and patterns of distant recurrences after Gamma Knife surgery (GKS) for metastases to the brain.
Information was available for 1855 of 1921 patients who underwent GKS for single or multiple cerebral metastases at 4 different institutions during different time periods between 1975 and 2007. The total number of Gamma Knife treatments administered was 2448, an average of 1.32 treatments per patient. The median survival time was analyzed, related to patient and treatment parameters, and compared with published data following conventional fractionated whole-brain irradiation.
Twenty-five patients survived for longer than 10 years after GKS, and 23 are still alive. Age and primary tumor control were strongly related to survival time. Patients with single metastases had a longer survival than those with multiple metastases, but there was no difference in survival between patients with single and multiple metastases who had controlled primary disease. There were no significant differences in median survival time between patients with 2, 3–4, 5–8, or > 8 metastases. The 5-year survival rate was 6% for the whole patient population, and 9% for patients with controlled primary disease. New hematogenous spread was a more significant problem than micrometastases in patients with longer survival.
Patient age and primary tumor control are more important factors in predicting median survival time than number of metastases to the brain. Long-term survivors are more common than previously assumed.
Lars Kihlström, Wan-Yuo Guo, Bengt Karlsson, Christer Lindquist, and Melker Lindqvist
✓ The authors report outcomes in 18 patients with arteriovenous malformations (AVMs) who were treated with gamma knife radiosurgery and in whom magnetic resonance (MR) imaging was obtained a mean of 14 years (range 8–23 years) after treatment and 10 years (range 4–17 years) after confirmed obliteration of the AVM. All patients were asymptomatic after radiosurgery and during the time of the study. In five patients (28%), cyst formation was observed that corresponded to the site of the obliterated AVM. Cyst formation and contrast enhancement on MR imaging could not be statistically correlated to the radiation dose. In 11 (61%) of the 18 patients, contrast enhancement that was not related to a recanalization of the nidus was observed in the target area. In three patients (17%), an increased T2-weighted signal was detected at the site of previous AVM; this was interpreted as gliosis or demyelination, which appeared to be dose dependent. The study illustrates that cyst formation, contrast enhancement, and an increased T2-weighted signal can be observed in asymptomatic patients in the area that was targeted for AVM radiosurgery up to 23 years after the procedure. The report provides new and essential information about long-term effects on normal tissue after radiosurgery and provides a basis for the interpretation of MR studies in the follow up of small AVMs treated by radiosurgery.
Bengt Karlsson, Ingmar Lax, Masaaki Yamamoto, Michael Söderman, Hidefumi Jokura, Charles Rosen, and Julian Bailes
The authors sought to assess the relationship between obliteration rate and different dose parameters following fractionated radiotherapy for arteriovenous malformations (AVMs). A comparison of the results of radiosurgery and radiotherapy for AVMs was made to calculate the best fit α/β value, which would then be used as a model for predicting the treatment outcome, independent of the number of fractions applied.
Data from 1453 patients were analyzed: 1154 treated with radiosurgery and 300 with fractionated radiotherapy. The relationships between dose and obliteration rate after 3 years were calculated, and the best fit curve to the empirical results was defined. The higher the dose per fraction, biologically effective dose, and the lower the total dose, the higher the obliteration rate. The isoeffective doses when comparing radiotherapy and radiosurgery independent of the α/β value could not be defined. The dose per fraction had the best predictive value, independent of the number of fractions.
Dose per fraction seems to be the decisive parameter for the treatment response following both radiotherapy and radiosurgery. A larger number of fractions did not increase the obliteration rate. The data indicate that higher doses per fraction should be used when irradiating AVMs.
Michael Söderman, Göran Edner, Kaj Ericson, Bengt Karlsson, Tiit Rähn, Elfar Ulfarsson, and Tommy Andersson
The aim of this study was to assess the clinical efficacy of gamma knife surgery (GKS) in the treatment of dural arteriovenous shunts (DAVSs).
From a database of more than 1600 patients with intracranial arteriovenous shunts that had been treated with GKS, the authors retrospectively and prospectively identified 53 patients with 58 DAVSs from the period between 1978 and 2003. Four patients were lost to follow-up evaluation and were excluded from the series. Thus, this study is based on the remaining 49 patients with 52 DAVSs. Thirty-six of the shunts drained into the cortical venous system, either directly or indirectly, and 22 of these were associated with intracranial hemorrhage on patient presentation. The mean prescription radiation dose was 22 Gy (range 10–28 Gy).
All patients underwent a clinical follow-up examination. In 41 cases of DAVS a follow-up angiography study was performed. At the 2-year follow-up visit, 28 cases (68%) had angiographically proven obliteration of the shunt and in another 10 cases (24%) there was significant flow regression. Three shunts remained unchanged.
There was one immediate minor complication related to the administration of radiation. Furthermore, one patient had a radiation-induced complication 10 years after treatment, although she recovered completely. There was one posterior fossa bleed 2 months after radiosurgery; a hematoma, as well as a lesion, was evacuated, and the patient recovered uneventfully. A second patient had an asymptomatic occipital hemorrhage approximately 6 months postradiosurgery.
The clinical outcome after GKS was significantly better than that in patients with naturally progressing shunts (p < 0.01, chi-square test); figures on the latter have been reported previously.
Gamma knife surgery is an effective treatment for DAVSs, with a low risk of complications. Major disadvantages of this therapy include the time elapsed before obliteration and the possibility that not all shunts will be obliterated. Cortical venous drainage from a DAVS, a risk factor for intracranial hemorrhage, is therefore a relative contraindication. Consequently, GKS can be used in the treatment of both benign DAVSs with subjectively intolerable bruit and aggressive DAVSs not responsive to endovascular treatment or surgery.