Linear accelerator stereotactic radiosurgery can modulate the clinical course of Hemangioblastoma: Case series and review of the literature

Abstract

Hemangioblastomas (HB) are benign low grade vascular tumors most frequently occurring in the cerebellum, brain stem, and spinal cord. Often associated with Von Hippel Lindau disease (VHL), the lesions are often multifocal requiring complex resection and are difficult to control. Linear Accelerator (LINAC) Stereotactic Radiosurgery (SRS) has been demonstrated to provide additional tumor control.

In this case series, we present our multi-center experience utilizing LINAC SRS in fourteen patients with 23 lesions. We observed a tumor control rate of 87% and found interval changes in the peritumoral enhancement to correlate with treatment outcome. In our study, SRS treatment was also welltolerated in both cystic and noncystic patients with multifocal disease. Disease control was achieved in all but three patients post-resection and no longitudinal radiation-induced secondary malignancy was observed. SRS response correlated highly with lesion size and radiation dose. We conclude that LINAC SRS is safe and effective for patients with HB and should be considered in addition to surgery in asymptomatic, VHL patients, deep seated lesions and isolated lesions.

1. Introduction

Hemangioblastomas (HB) are a rare World Health Organization grade I benign vascular tumor of the central nervous system. These lesions can be either solid or cystic and occur most frequently in the cerebellum, brain stem, and spinal cord and only rarely in the supratentorial regions [6]. These tumors may present as sporadic lesions or in patients with Von Hippel Lindau (VHL) syndrome [11]. HB may also occur as multiple lesions particularly when associated with VHL syndrome. Total surgical resection results in excellent rates of control and has been the treatment of choice for most tumors [7]. Recurrence after subtotal resection is common especially within the VHL subgroup of patients [7]. Although HB are benign vascular tumors, invasion into adjacent imperative neurologic structures can preclude gross total resection, and therefore impair complete recovery.

Several publications have reported stereotactic radiosurgery (SRS) as an effective and safe treatment for intracranial partially-resected, unresectable tumors or in patients who are not surgical candidates (due to other comorbid medical conditions) [8,9]. In this report, we present our experience in treating HB by SRS and conduct a review of the contemporary literature (See Fig. 1).

Fig. 1.

Axial, contrast-enhanced, T1-weighted, magnetic resonance imaging of tumor regression (A–D). Zibly.

2. Materials and methods

2.1. Patient characteristics

Records of patients treated by SRS for brain HB at the Sheba medical center (Israel) and the National Institutes of Health (USA) were retrospectively evaluated. Between 1994 and 2018 fourteen patients (9 men, 5 women) with a total of 23 HB were treated by linear accelerator (LINAC) radiosurgery. Mean age at treatment was 43 years (Min 12, Max 59). Number of patients who were treated for more than a single lesion was 4. Nine patients were diagnosed with VHL (according to the Lonser/Oldfield criteria) [10]. Five of the treated lesions were partially resected before radiosurgery treatment. Mean follow up period was 89 months (min 12, Max 150). At time of SRS, principal clinical symptoms were gait disturbance, vertigo, and cerebellar limb weakness. Treatment method and planning was discussed in cooperation with neurosurgeons, radiation oncologists, and radiologists at multidisciplinary conferences. The study protocol was approved by the institutional review board at both medical institutes. Patient characteristics are presented in Table 1.

Table 1.

Patient and tumor characteristics.

Patient characteristics
Number of patients	14
Number of tumors	23
Mean age years (range)	43(12–59)
Gender Male (%)	9(64.2)
Female (%)	5(35.8)
Mean KPS (range)	89.7(70–100)
Mean follow up period months (range)	89(12–50)
Tumor location Cerebellar (%)	9(39)
Vermis	3(13)
Brain stem	9(39)
Supra tentorial	2(9)
Cystic tumor	7(30)
Mean tumor volume m”l (range)	1.86(0.2–8.76)

Abbreviations: KPS, Karnofsky performance status score.

2.2. Lesion characteristics and follow up

All investigated patients had serial Gadolinium enhanced and non-enhanced magnetic resonance imaging (MRI) studies in proximity to each follow-up visit (every 6–12 months) and on day of treatment. Tumor location and volume based on Gadolinium enhanced T1-weighted MRI were measured and documented. Enhancing tumors were found in the cerebellar hemisphere in 9 cases (39% of lesions), cerebellar vermis in 3 cases (13% of lesions), brainstem in 9 cases (39% of lesions), and supra tentorial in 2 cases (9% of lesions). Mean tumor volume was 1.06 m”l (min 0.2, max 4.66). Seven (30%) of the treated HB had a related cyst. 7 (30.5%) of the lesions had associated surrounding brain edema. Response to treatment was determined by most recent imaging studies as stable, regression or progression (classified as 25% decrease or increase in volume, respectively). Maximal enhancing tumoral length in all three dimensions was used for measurements and calculation of lesion volume. All statistical analyses were performed using SPSS software (IBM SPSS Statistics) (See Fig. 2).

Fig. 2.

Axial, contrast-enhanced, T1/T2 weighted, magnetic resonance imaging of tumor (A, B) and associated cyst regressing (C, D). Zibly.

2.3. Stereotactic radiosurgical treatment

SRS was performed using a 6-MV LINAC unit. Radiation delivery paradigms changed over the years due to technological developments in hardware/software. For the last 20 years, irradiation was achieved using a single isocenter, dynamic conformal radiation arcs paradigm. Treatment plans were generated based on a Gadolinium enhanced magnetic resonance imaging merged and registered with contrast enhances computer tomography (CT) scans obtained following stereotactic frame placement. As a practice, only the enhancing nodule of the tumor was treated. The number of isocenters, maximum radiation dose, prescribed dose to the tumor margins, and isodose normalization were documented. Mean maximum radiation dose of 2175 Gy (range 1750–3445 Gy) was delivered, and mean prescribed dose of 1573.3 Gy (range 1400–2000 Gy) at tumor margins utilizing a mean isodose line of 82% (range 70–80 %). A mean of 5.3 arcs (min 5 Max 9) were used (Table 2).

Table 2.

SRS treatment parameters.

Mean isodose normalization (%) (range)	78.1%(70–80)
Mean maximal dose (Gy) (range)	2321(1824–3444)
Mean Number of arcs (range)	5.9 (5–12)

Abbreviations: SRS, Stereotactic radiosurgery. Gy, Gray.

3. Results

3.1. Tumor control and neurologic assessment

Overall, 23 lesions were treated (14 patients). 11 of the 23 (48%) lesions were coincident with VHL positive patients.

Response to radiosurgery was studied in relationship to the following end points: overall survival, tumor progression (local control) and treatment related side effects. Regression, progression or stable disease were defined as 25% change in size of the solid enhancing lesion as seen on T1W MRI studies. Time to failure for each treated lesion was calculated from the time of treatment. In patients who were lost to follow up, with no evidence of treatment failure, this period was calculated from the time of last follow up visit.

Of the 23 treated HBs, 8(34.8%) showed regression, 12(52.2%) showed no change in volume and 3(13%) demonstrated progression within follow up period. All three lesions that were treated with SRS and have failed were residual tumors (surgical failure) and had to be reoperated.

With regards to tumor volume, the mean tumor volume of failed to be controlled by SRS was 3.18 m”l as matched to 0.97 m”l (min 0.12, max 1.7) for controlled Hemangioblastomas. There was a robust correlation with increasing volume of tumor to increased risk of treatment failure. The lesions that failed to be controlled by SRS received a mean minimum tumor dose of 2063 Gy (min 2125 max 2250 Gy) as compared to 1924 Gy (1750–3445 Gy) with controlled tumors. With regards to lesion with adjacent cyst, we found these to be treatment responsive as were the non-cystic tumors.

3.2. Tumor control and VHL

There was no correlation with diagnosis of Von Hippel Lindau to radiosurgery treatment outcome. The number of lesions that failed to response were similar within the non VHL and the VHL positive patients.

3.3. Neurologic assessment and clinical outcome

Neurologic assessment and serial neuroimaging studies at scheduled intervals of 6–12 months were performed and documented to evaluate response to treatment. The Karnofsky Performance Status Scale (KPS) was used for assesment of patient’s neurologic status. With patients who presented with cranial nerve or brainstem related deficits, we unambiguously documented for any change. Generally, patients remained at their clinical baseline over the study period. The mean KPS pre treatment was 89.7 (min 70, max 100) and at last follow-up was 91.3 (min 60, max 100). Of the 8 treated patients with pre-treatment cranial nerve or brainstem related symptoms, 3(37.5%) experienced clinical improvement after treatment that corresponded with enhancing tumor volume reduction, 3(37.5%) patients were with no neurologic change and two patient (25%) exhibited neurologic deterioration that was associated with tumor growth.

3.4. SRS related adverse effects

Overall, radiation induced radiologic “side effect” was observed in 1(7.14%) patient. This patient was a VHL positive male (tumor volume, 2.3 m”l) treated by a 2000 Gy marginal dose. His imaging studies demonstrated peritumoral edema with no neurological consequences. Thus, we decided not to treat him with steroids and this radiologic finding resolved within 1 year.

4. Discussion

Generally, surgical resection of HBs is offered as first line of treatment (especially within the non VHL group of patients). Resection usually results in immediate resolution of neurologic symptoms [4]. The natural history of HBs, within the VHL patient group, is typically of tumor growth in a two-phase progression pattern [12,11], leading to neurologic symptoms in almost half of the cases [1]. Within the sporadic cases of HBs, the percentage of patients presenting with neurologic symptoms tends to be higher, largely due to late imaging assessment and diagnosis. In spite of the progress in the area of microsurgery, HBs of the central nervous system harbor high risk of mortality and morbidity, particularly for lesions that are located in deep seated regions (e.g., brain stem) or in eloquent brain areas. Furthermore, incomplete resection with residual (even microscopic) margins, often will results in local recurrence [13,5].

Several reports have described the utility of stereotactic radiosurgery in the treatment of central nervous system HBs. Qiu et al. in a mata analysis reported a tumor control rate of up to 91% in a five year follow up period, where a total of 660 HGB were treated with either Gamma knife or Linac SRS concluding that SRS should be offered to patients with limited surgical alternatives [17].

Radiosurgery is of importance especially when approaching VHL patients, potentially reducing the number of surgical resections and avoiding resection of lesions with increased operative risk. Similarly, non-symptomatic HBs that show increased volume can be controlled prior to becoming symptomatic.

With regards to tumor control, in this study the overall rate was 87% during the follow up period. These results are similar to previous studies which reported good short term tumor control rate [13–16]. In our study, we also found that the tumor control of treated lesions with the presence of peritumoral cyst was similar to non-cystic lesions. This finding does not correlate with some of the previous published reports [13,11]. Furthermore, we found a clear correlation between early changes in the enhancing lesion to final outcome, e.g.,. initial change did reflect treatments final outcome.

One of the concerns when treating benign lesions such as HBs with SRS is the long-term risk of developing a de novo secondary intracranial malignancy. In our study, none of the treated patients developed a secondary intracranial malignancy. This finding corresponds with the known low risk of radiation-associated intracranial malignancy after SRS [16]. Moreover, none of the patients in our cohort developed radiation induced injury that necessitated any pharmacologic treatment nor had any neurologic deterioration related to the stereotactic radiosurgery treatment.

5. Conclusion

In conclusion, LINAC based SRS for the treatment of brain HBs is both effective and a safe method of treatment for sporadic lesions or in conjunction with VHL disease. Cystic HBs appears to have similar response as compared to non-cystic lesions, but this was not statistically significant.

Surgical intervention remains the first line treatment of large symptomatic HBs. SRS however, should be considered for patients with residual (post operative) HBs, and in growing lesions that have the potential to become symptomatic as SRS can favorably modulate the clinical course of the disease. SRS should also be offered to VHL patients presenting with multiple HBs or patients with co – morbidities that would result in high surgical morbidity. Furthermore, our study cohort is small, and a larger multi center study should be considered.

Abbreviations:

LINAC

Linear accelerator

SRS

Stereotactic Radiosurgery

VHL

Von Hippel Lindau

HB

Hemangioblastoma