 |
| About Bioline | All Journals | Testimonials | Membership | News |
|
||||||
|
||||||
Journal of Cancer Research and Therapeutics, Vol. 5, No. 1, January-March, 2009, pp. 8-13 Original Article Long-term results of LINAC-based stereotactic radiosurgery for acoustic neuroma: The Greek experience Kalogeridi Maria-Aggeliki, Georgolopoulou Paraskevi, Kouloulias Vassilis, Kouvaris John, Pissakas George Radiotherapy Unit, Medical School, National University of Athens, Greece Code Number: cr09003 Abstract Purpose: To estimate the value of LINAC-based stereotactic radiosurgery (SRS) for the long-term local control of unilateral acoustic neuromas.
Keywords: Acoustic neuroma, linear accelerator, local control, stereotactic radiosurgery Introduction Stereotactic radiosurgery (SRS) has a well-established role as an alternative to microsurgical resection in acoustic neuromas. [1] SRS, using a cobalt unit (gamma knife), was introduced by Leksell in 1969 as a nonsurgical treatment option for acoustic neuromas. [2] Many studies have demonstrated the efficacy of gamma knife for the treatment of neuromas, with tumor control rates of more than 95%. [3],[4],[5] With the introduction of linear accelerators especially adapted for SRS, linac-based SRS has become an alternative to gamma knife radiosurgery. There have been relatively few studies on linac-based SRS for neuromas and, moreover, these studies have been on small samples and have had short follow-up periods; nevertheless, linac-based SRS has been shown to be comparable to gamma knife radiosurgery with respect to local control and complications. [6],[7] Suh et al . reported a 5-year control rate of 94% for 29 neuroma patients treated with linac-based radiosurgery using a median marginal dose of 16Gy. New trigeminal and facial palsy occurred in 15 and 32% of patients, respectively. In the series of Spiegelmann et al . with 44 patients, a 98% tumor growth control rate was achieved with linac-based radiosurgery, using doses ranging from 11 to 20Gy. [8] According to Foote et al ., marginal doses of 12.5Gy are suitable for achieving maximum tumor control with minimal complications. [9] In this paper, we present a single-institution experience in the treatment of acoustic neuromas with linac-based SRS. This article seeks to define local control and toxicity after long-term follow-up when a marginal dose of 12Gy is used. Materials and Methods Between May 2000 and June 2004, 20 patients (5men, 15 women) with acoustic neuromas were treated with linac-based SRS. The age range was 57-80 years (median 66 years). All patients had unilateral tumors. Among them, 13 patients (65%) had tumors with both intracanalicular and cerebellopontine angle components . Patient characteristics are shown in [Table - 1]. Four patients had undergone previous attempts at surgical removal and were referred for radiosurgery for residual or recurrent tumors. Tumors considered for radiosurgery had a maximum diameter of 35mm. Patients with larger tumors were offered surgery or fractionated radiotherapy. The patients were considered to be eligible for treatment if they had documented tumor progression on MRI or CT, progression of symptoms, or both. The maximum tumor diameter ranged from 10 to 32mm. Median tumor volume was 5.95 cm 3 (range:0,44-15,7). In all patients, SRS was performed using an Elekta SL-18 linac converted for radiosurgery with the attachment of an isocentric subsystem (Philips SRS200XK). The whole procedure was carried out over 8-9h. On the treatment day, a Brown-Roberts-Wells stereotactic head frame with four sharp stereotactic pins was screwed onto the patient's skull under local anesthesia. With the invasive stereotactic head frame in place, a contrast-enhanced CT scan was performed using a stereotactic localizer. The patient's entire head was scanned using 1-3mm contiguous slices. Tumor delineation was made on this contrast-enhanced CT scan. Organs at risk, such as the optic nerves, optic chiasm, and brain stem, were outlined on the treatment planning CT. A treatment plan was achieved using 1-8 isocenters. The dose was 11-12Gy to the prescription isodose, which covered 95-100% of the tumor. Marginal tumor dose was prescribed to the 54% isodose (median). High conformality of the treatment dose to the borders of the tumor was established by different combinations of number, span, and weight of noncoplanar arcs. Every effort was made to achieve homogeneity in dose distribution across the tumor while keeping the dose to the stem as low as possible. Treatment duration was 30-60 min once the machine was set up. Once irradiation was completed, the frame was removed from the patient's head. The tumors were divided into three groups: small (< 2.0cm), medium (2.0-3.9cm), and large (≥4cm). [10] Tumor size evaluation was based on the measurement of the largest diameter of the intracanalicular and cerebellopontine angle components of the tumor in the axial, sagittal, and coronal dimensions. These measurements were compared with those noted by the radiologist. Volume changes after treatment were determined by examination of contrast-enhanced CT or MRI. The maximum tumor diameter was measured in three dimensions and compared with the original volume assessment. Before treatment, all patients were evaluated in a standardized fashion. The neurologic evaluation focused on cranial nerve function. None of the 20 patients had serviceable hearing before radiosurgery. Trigeminal nerve function was assessed by examination of touch and pinprick sensation in the territories of the three divisions of the nerve. Sensation was assessed by comparison with that on the unaffected side. At follow-up we looked for any subjective facial numbness, pain, or altered sensation. Facial nerve function was scored using the House-Brackmann facial nerve grading system. [11] Clinical follow-up was obtained from the patients directly or, if they lived a significant distance away from our institution, from their referring doctors. When necessary, patients were contacted by telephone to update their outcomes for the purposes of this study. The primary endpoint was tumor control and the secondary endpoint was reduced cranial nerve toxicity. Our protocol for follow-up included repeat imaging studies and clinical examination with assessment of facial and trigeminal nerve function at 6-month intervals for the first year and yearly thereafter. Verification of tumor control after radiosurgery requires a more extended follow-up period because there is no routinely available imaging modality with which clinicians can assess the tumor's biological viability. [7] An actuarial follow-up of at least 3 years is considered necessary for meaningful conclusions to be drawn. [12] In our study, tumor growth control was defined as absence of permanent increase in tumor dimension by> 2mm or absence of any change requiring surgical resection. The follow-up period in this series ranged from 36 to 84 months (median 55 months). Results Twenty patients were treated and 19 were followed, one patient being lost to follow-up. We judged tumor control rates by assessing tumor diameters over time on contrast-enhanced CT or MRI scans. In most patients, tumors that increase in size slightly either stabilize or regress afterwards. Eleven tumors (58%) decreased in size and eight remained stable (42%). None of the tumors in the long-term follow-up increased in size, thus giving an overall growth control of 100% for the patients in our study. One tumor showed a marginal increase in size on the MRI done 6 months after radiosurgery as compared to the size on the pretreatment MRI; however, a decrease in size was evident at the next radiographic assessment and the tumor remained stable after that. For tumors that showed a decrease in size, the change started 1 year after radiosurgery [Figure - 1]. Eight patients were noted to have a loss of central enhancement on imaging. This is generally accepted as a characteristic change in acoustic neuromas following radiosurgery. [13] This change was first seen 6 months after linac-based SRS. Like others, we too believe that loss of enhancement in the post-treatment period reflects necrosis. [14] Interestingly, this change of central enhancement was not associated with a temporary increase in tumor diameter. The only tumor that showed a transient increase in size on the MRI never manifested this loss in central enhancement. None of the 20 patients had any measurable or useful hearing prior to treatment and no improvement was noted after treatment. This was not surprising - if a nerve has been damaged before treatment starts then it is unlikely that treatment can repair the damage. Eighteen patients (90%) had intact trigeminal nerve function before radiosurgery. In the immediate post-treatment period no treatment-related neuropathy was noticed (neuropathy being defined as subjective decrease in sensation or occurrence of new pain in the ipsilateral trigeminal nerve's distribution). In the long-term follow-up (36-84 months; median 55 months) none of these patients developed new symptoms. One patient out of two with facial numbness before SRS had clear improvement after therapy, while the other remained stable. None of these two patients had undergone prior surgery. Before treatment none of the patients had facial nerve neuropathy. Treatment-related facial neuropathy was defined as temporary or permanent decline in the House-Brackmann facial nerve grade. In the immediate post-treatment period, three patients had transient symptoms that could be related to facial nerve neuropathy. However, the symptoms resolved in a few days time without any medication. At long-term follow up no permanent deficits were recorded. In the immediate post-radiosurgery period, six patients developed new-onset headache which lasted less than 12h and was successfully treated with paracetamol. Most patients attributed it to the head frame pins. All patients were discharged from the hospital immediately after the radiosurgery was completed. Patients were able to return immediately to their everyday activities. None of the 18 patients with KPS (Karnofsky Performance Status) ≥90% experienced any decrease in functional level. Discussion With the advances in imaging modalities, the chances of detecting an acoustic neuroma before it becomes symptomatic have increased. Although acoustic neuromas are usually small tumors that are amenable to either excision or SRS, whether it is necessary to treat asymptomatic tumors is still controversial. [15] For asymptomatic or minimally symptomatic patients, observation through serial imaging studies is an option since slow growth is considered characteristic of these tumors. Rosenberg et al . documented a growth rate of 0.91 mm/year for untreated neuromas. [16] In a postradiosurgery period systematic review that included 26 studies with 1340 patients, the mean annual tumor growth rate was 1.2mm/year (range 0.4-2.9mm/year). The overall frequency of tumor growth during the follow-up period (range 6-64 months) was 46%, while the estimated frequency of tumor regression was 8%. The rate of eventual treatment after conservative management was 18%. [15] The mainstay of treatment for neuromas has traditionally been microsurgical resection. [17] Microsurgery can achieve control rates of up to 90-95% though this is at the cost of treatment-related toxicity and morbidity. [18],[19] However, many authors have reported good results with microsurgical resection, with high rates of long-term preservation of functional hearing. [20] Growth control rates for tumors treated with microsurgery correlate closely with the amount of tumor removed. Gormley et al . reported a growth control rate of 100% for tumors treated with total resection. [21] For tumors that were subtotally resected, a growth rate of 45-90% has been reported. [21],[22],[23] Another therapeutic approach is linac-based Stereotactic radiotherapy (SRT). [24] This technique combines the precision of stereotactic positioning with the radiobiological advantage of fractionation. [25] Studies on stereotactic radiotherapy have demonstrated high tumor control rates, with minimal toxicity. [24],[26],[27],[28],[29] Combs et al . reported the treatment outcome of 106 patients treated with SRT using a median total dose of 57.6Gy. After a median follow-up of 48.5 months, the actuarial local control rate was 93%. Actuarial useful hearing preservation was 94% at 5 years with trigeminal and facial nerve neuropathy of only 3.4% and 2.3%, respectively. [30] In a study published by Selch et al ., the 5-year actuarial local tumor control rate was 100% and the 5-year actuarial rate of preservation of facial and trigeminal nerve function was 97.2 and 96.2%, respectively. [24] Radiosurgery has a long history of use in the treatment of neuromas. It was first used in 1969, when an acoustic neuroma patient was treated at Karolinska Hospital using a gamma knife unit. [2] Since then considerable clinical experience has been gained and several studies have shown high tumor control rates with this technique. Kondiolka et al . reported excellent control rates with tumor doses of the order of 16Gy, though this was at the cost of significantly high rates of subsequent facial weakness (21%), facial numbness (27%), and decreased hearing (49%). [12] With reduction in the marginal dose, a high rate of tumor growth arrest can be achieved with minimal toxicity to the cranial nerves. [3] In a large series by Hasegawa et al ., the 5- and 10-year PFS (Progression-free Survival) rate was 93 and 92%, respectively. Treatment failure (6.9%) developed within 3 years after treatment. When the tumor was treated with a marginal dose of 13Gy or less, the hearing preservation rate was 68%, transient facial palsy developed at a rate of 1%, and facial numbness at a rate of 2%. [17] Other series have also demonstrated diminished facial and trigeminal nerve toxicity. [1],[12],[31] Chopra et al . recently reported on the treatment outcomes of 216 patients treated with gamma knife radiosurgery using marginal tumor doses of 12-13Gy. They reported a 10-year actuarial resection-free control rate of 98.3 ± 1.0%. [32] Ten-year actuarial rates for preservation of facial and trigeminal nerve function were 100% and 94.9 ± 1.8%, respectively. Nowadays, linear accelerators have replaced cobalt units as standard equipment in radiotherapy departments. Linear accelerators have been used since the mid-1980s to treat acoustic neuromas with SRS and have become a cost-effective alternative to gamma knife radiosurgery. [6],[7] Spiegelman et al . showed a 98% control rate of acoustic neuromas treated with linac-based radiosurgery. After a follow-up period ranging from 12-60 months, the actuarial hearing preservation rate was 71%. New transient facial neuropathy developed in 24% and persisted, in a mild degree, in 8% of patients. New trigeminal neuropathy developed in 18% of patients and was always accompanied by facial neuropathy. [7] The mean dose directed to the tumor margin was 1455cGy (range 11-20Gy). Okunaga et al . documented favorable tumor control in 73.8% of patients with acoustic neuromas using a median marginal dose of 14 Gy (range 10-16Gy). This percentage was 81.6 and 100%, respectively, for patients in whom follow-up images were obtained for longer than 2 and 5 years, showing that the tumor decreases in size over time. [33] In the same study, 66.7% of patients retained useful hearing. New trigeminal and facial neuropathy occurred in 2.4 and 4.8%, respectively. In our study, treatment of acoustic neuromas with linac-based SRS has shown favorable tumor control with minimal toxicity. We limited the tumor dose to 11-12Gy, in keeping with the trend initiated at the end of 1980s to lower the radiation dose to the tumor to reduce neuropathic complications. In most cases, favorable tumor control was closely correlated with the length of the follow up period. Tumor size decreased in six patients followed up for 1 year after treatment, in eight patients followed for 2 years, and in all 11 patients with 3 years' follow up. In 18 out of 19 patients, MRI and CT images showed no increase in tumor size after treatment. One tumor showed a marginal increase on the MRI done 6 months after the radiosurgery; however, a subsequent decrease was noticed on the next radiographic assessment and the tumor remained stable from then on. Thus, we obtained a tumor control rate of 100% for the limited number of patients in this study. Eleven out of the 19 tumors that were assessed with CT or MRI after linac-based SRS decreased in size. In our series, no new facial or trigeminal neuropathy developed. The hearing preservation could not be evaluated in our study since no patient had any measurable or useful hearing before radiosurgery. None of the 20 patients needed hospitalization. Six patients complained of headaches in the immediate post-radiosurgery period and were successfully treated with paracetamol. No other acute side effects were noted. These results with linac-based radiosurgery are consistent with previous reports [Table - 2]. Linac-based radiosurgery was developed after the introduction of the gamma knife. There are a lot of series studying the effectiveness and toxicity of gamma knife SRS for neuroma patients, but only a few series with rather short follow-up for neuroma patients treated with linac-based SRS. Although our series has only a small number of patients, the follow-up period is quite long. We assessed all patients with radiological imaging (CT or MRI) for more than 3 years since, as reported by Hasegawa [17] and others, failure usually occurs within 3 years. Thus, an actuarial follow-up of at least 3 years is considered necessary for any meaningful conclusions to be drawn. [12] Conclusion The goal of radiosurgery is not necessarily to cause significant tumor necrosis or to obtain a complete radiographic response, but to halt the tumor's growth permanently through its biological elimination. In our study, tumor doses of 11-12Gy resulted in high rates of tumor growth arrest and low cranial nerve toxicity in long-term follow-up. A larger number of patients than were included in our study is required to accurately assess linac-based radiosurgery in the treatment of acoustic neuromas.References
Copyright 2009 - Journal of Cancer Research and Therapeutics The following images related to this document are available:Photo images[cr09003t1.jpg] [cr09003f1.jpg] [cr09003t2.jpg] |
| |||||||||
