Hypofractionated stereotactic radiotherapy for acoustic
neuromas: safety and effectiveness over 8 years of experience.
Sakanaka, Katsuyuki; Mizowaki, Takashi; Arakawa, Yoshiki;
Araki, Norio; Oya, Natsuo; Takahashi, Jun A; Mikuni,
Nobuhiro; Miyamoto, Susumu; Hashimoto, Nobuo; Hiraoka,
International journal of clinical oncology / Japan Society of
Clinical Oncology (2011), 16(1): 27-32
Issue Date 2011-02
RightThe final publication is available at www.springerlink.com
Type Journal Article
KURENAI : Kyoto University Research Information Repository
Title: Hypofractionated stereotactic radiotherapy for acoustic neuromas: safety and
effectiveness over 8-year experience
Jun A. Takahashi5
1Department of Radiation Oncology and Image-applied Therapy, Kyoto University
Graduate School of Medicine, 54 Sho-goin Kawahara-cho, Sakyo-ku, Kyoto, Japan
2Department of Neurosurgery, Kyoto University Graduate School of Medicine, 54 Sho-
goin Kawahara-cho, Sakyo-ku, Kyoto, Japan
3Department of Radiology, National Hospital Organization Kyoto Medical Center, 1-1,
Fukakusa Mukaihata-cho, Fushimi-ku, Kyoto, Japan
4Department of Radiation Oncology, Faculty of Life Sciences, Kumamoto University,
1-1-1 Honjo, Kumamoto, Japan
5Department of Neurosurgery, Kitano Hospital, 2-4-20 Ohgimachi,Kita-ku, Osaka,
6National Cardiovascular Center, 5-7-1 Fujishiro-dai, Suita, Osaka
Corresponding author: Takashi Mizowaki, MD, PhD,
Mailing address: Department of Radiation Oncology and Image-applied
Therapy, Kyoto University Graduate School of Medicine, 54
Sho-goin Kawahara-cho, Sakyo-ku, Kyoto, 606-8507, Japan
Background: Little information is available about long-term outcomes of
hypofractionated stereotactic radiotherapy (hypo-FSRT) for acoustic neuromas. In this
study, the safety and effectiveness of hypo-FSRT for unilateral acoustic neuroma were
reviewed over 8-year experience of our institution.
Methods: Between May 1998 and October 2006, 27 patients were consecutively treated
by linear accelerator-based hypo-FSRT. Two patients were excluded from this study
because they were lost to follow-up within 12 months. The median follow-up period for
the rest was 59 months (range, 24-133). Two types of treatment schedule were adopted.
Thirteen patients received 30-39 Gy, given in 10-13 fractions (regimen A), while after
July 2003, twelve patients received 20-24 Gy, given in 5-6 fractions at the tumor
periphery (regimen B). These treatments were scheduled to be delivered in three
fractions per week (Monday, Wednesday, Friday).The median planning target volume
was 2.0 mL, 1.7 mL (range, 0.7-10.6) in regimen A and 5.2 mL (range, 0.9-9.3) in
regimen B. Seven patients had serviceable hearing (Gardner-Robertson Class 1-2) in the
pre-treatment audiogram; two patients in regimen A and five in regimen B.
Results: Local control rates were 100% with regimen A and 92% with regimen B,
respectively. Serviceable hearing was preserved in four of five patients in regimen B but
no patients in regimen A at the last follow-up. No permanent facial and trigeminal nerve
morbidity were newly observed. No salvage surgery was needed.
Conclusions: Hypo-FSRT for acoustic neuromas achieved a high local control rate with
minimal facial and trigeminal nerve morbidity.
Hypofractionated stereotactic radiotherapy for acoustic neuromas achieved a high local
control rate with minimal facial and trigeminal nerve morbidity.
Key words: acoustic neuroma, stereotaxic techniques, radiotherapy, dose fractionation.
Acoustic neuroma is a benign intracranial tumor, commonly arising from the
vestibular portion of the 8th cranial nerve. They are typically slow growing, but can
cause several symptoms, such as hearing loss, facial paresis, facial numbness, and
hydrocephalus 1, 2.
Surgery has been the main treatment option for small-to-large acoustic neuromas, but
the morbidity rates can not be ignored, even with an experienced team 3. As an
alternative, stereotactic radiosurgery (SRS) has been adopted, but initial reports failed to
achieve favorable cranial nerve preservation rates. To improve cranial nerve morbidities
with a high local control rate, two different radiotherapeutic strategies have been
developed; dose-reduced SRS and stereotactic radiotherapy (SRT). Dose-reduced SRS
had shown excellent local control rates and minimal morbidity in the treatment of small-
to-medium acoustic neuromas in long-term follow-up 4, 5. Conventionally fractionated
SRT has also demonstrated high hearing preservation rates and local control, with
relatively short follow-up periods 6-8.
Our institution started to treat acoustic neuromas with hypofractionated stereotactic
radiotherapy (hypo-FSRT) in May 1998. Various schedules of hypofractionation have
been developed not only for the intracranial lesions but also the extracranial lesions 9-11.
The merits of hypo-FSRT are that it should reduce normal tissue damage, versus SRS
and is more convenient than conventionally fractionated SRT, because it involves fewer
irradiation sessions. Recent studies have reported that hypo-FSRT provides a high local
control rate and acceptable morbidity for small and even large acoustic neuromas 12-15.
However, little information is available as to whether high tumor control rates and
minimal morbidity are maintained for an extended period with a hypofractionated
regimen. The aim of this study was to review the safety and effectiveness of linear
accelerator-based hypo-FSRT for acoustic neuroma over 8-year experience.
Materials and Methods
Between May 1998 and October 2006, 27 patients (27 tumors) with acoustic
neuroma underwent hypo-FSRT at Kyoto University Hospital. Two patients who were
lost to follow-up within 12 months were excluded from the analysis. For hypo-FSRT,
we sought patients with continuously enlarging tumors who were not candidates for
surgery. All patients gave informed consent for the treatment. Eleven patients had
previously undergone at least one surgical tumor resection and were histologically
confirmed to have acoustic neuroma; the median interval between the last surgical
procedure and hypo-FSRT was 45 months (range, 4-120). The remaining 14 patients
who were diagnosed with acoustic neuroma by imaging studies received hypo-FSRT as
an initial treatment. No patient suffered from neurofibromatosis type 2. Patients were
regularly followed by physical examination, audiometric testing, and contrast-enhanced
MR imaging studies, every 3-6 months for 3 years and every 12 months thereafter.
All tumors were treated by linear accelerator-based hypo-FSRT, using 6 MV X-ray
beams generated by a Clinac-2300c linear accelerator (Varian Inc., Palo Alto, CA) with
head immobilization using a Gill-Thomas-Cosman relocatable stereotactic frame.
Treatment planning was carried out using the X-knife system (Radionics Inc.,
Burlington, MA). In this study, all patients were treated with not using multi-leaf
collimator, but using collimating cylinder. From May 1998 to May 2003, twelve
patients were treated with a regimen of 39 Gy in 13 fractions at the edge of the PTV
(3.2-3.75 Gy per fraction at the isocenter). One patient was treated with 30 Gy in 10
fractions (3.75 Gy per fraction at the isocenter). Starting in July 2003, the regimen was
changed by reference to the other institutional report 13, and 11 patients received 20 Gy
in five fractions (5 Gy per fraction at the isocenter). One patient was treated with 24 Gy
in six fractions (5 Gy per fraction at the isocenter). We define the regimen of 30-39 Gy
in 10-13 fractions as regimen A and 20-24 Gy in 5-6 fractions as regimen B. The
median diameter of extrameatal portion was 15 mm: 13 mm (range, 5-34) in regimen
A and 19 mm (range, 5-27) in regimen B. The planning target volume (PTV) was
defined as the contrast-enhancing area of the tumor with a 2 mm margin in helical CT
images of 2.5 mm slice thickness, fused with previously generated MR images. The
median PTV was 2.0 mL: 1.7 mL (range, 0.7-10.6) in regimen A and 5.2 mL (range,
0.9-9.3) in regimen B. All patients were treated with a single isocenter and the
collimator diameter ranged from 17.5 to 32.5 mm.
Three-dimensional multiple arc therapy, with a median arc number of five (range, 4-
7), was used. These treatments were scheduled to be delivered in three fractions per
week (Monday, Wednesday, Friday). The median follow-up period was 59 months; 87
months (range, 24-133) in regimen A and 39.5 months (range, 24-64) in regimen B.
In this study, local tumor control was defined as no continuous tumor enlargement on
serial MR images. Tumor size was estimated according to the Committee of Hearing
and Equilibrium guidelines for the evaluation of hearing preservation in acoustic
neuroma, in which the diameter of the tumor is defined as the square root of (long
diameter × short diameter 16.
The Gardner-Robertson class system was used to classify the hearing 17(Table 2).
Seven patients with serviceable hearing (Gardner-Robertson class I-II) in the pre-
treatment audiogram were included for the evaluation of hearing preservation. The
median pure tone average before hypo-FSRT was 32 dB (range, 31-33) in regimen A
and 18 dB (range, 13-26) in regimen B respectively. The median audiometric follow-up
period was 43 months (range, 30-52): 37 months (range, 30-44) in regimen A and 43
months (range, 34-52) in regimen B. Facial nerve dysfunction was scored by using
House-Brackmann grading system18 (Table 3). The characteristics of 25 patients are
outlined in Table 1.
Local control rates in the follow-up period were 100% (13/13) in regimen A and
92% (11/12) in regimen B. The transient tumor increase was seen in 48% (12/25) of
treated tumors but the continuous tumor enlargement was observed in the only one case
(Figure. 1). Six tumors, the diameters of the extrameatal portion of which were larger
than 25 mm, were well-controlled. The median time when the tumor was maximally
enlarged after hypo-FSRT was 13 months (range, 8-35). No patient needed salvage
surgery within the follow-up period.
The changes of pure tone average after hypo-FSRT in patients with serviceable
hearing are shown in Figure 2. The deterioration of pure tone average by more than 20
dB was observed in 5 of 7 patients during the follow-up period. The median pure tone
average at the last follow-up was 59.5 dB (range, 59-60) in regimen A and 40 dB
(range, 9-65) in regimen B. Serviceable hearing (Gardner-Robertson class I- II) was
preserved in 4 of 5 patients in regimen B but no patients in regimen A at the last follow-
No early toxicity requiring medication occurred. Late toxicity (3 months after hypo-
FSRT) requiring treatment was seen in two patients (8%): one with temporary facial
paresis (House-Brackmann grade IV) 5 months after the hypo-FSRT in regimen A and
one with transient symptomatic non-communicating hydrocephalus 12 months after
hypo-FSRT in regimen B. Facial paresis was resolved in 12 months with oral steroid
medication. Symptoms related with non-communicating hydrocephalus disappeared in a
month with intravenous steroid and glycerol, without permanent cerebrospinal fluid
shunting or operative resection of tumor. No patient developed trigeminal nerve
dysfunction or dysequilibrium requiring medication in the acute or late period. No
radiation-induced secondary tumor was observed.
We found that hypo-FSRT for acoustic neuromas had the excellent rate of local control
without severe cranial nerve morbidity over 8-year experience. The findings of this
study support the growing body of literature of hypo-FSRT regarding safety and
effectiveness. In previous studies, the rate of local control and severe cranial nerve
morbidity were reported as 94-100% and 0-3%, respectively 12-15 (Table 4) and
comparable to this data.
In this study, the transient enlargement of tumor volume was observed in 48% of
treated tumors but the continuous enlargement of tumor volume was observed in only
one case over a follow-up period. The reasons for the transitional enlargement remain
unclear but were supposed to associate with intratumoral hemorrhages or increased
vascular permeability induced by radiotherapy 19, 20. The rate of the transient tumor
enlargement after SRS was reported as 14 to 45.2% 21-23 and the time when the transient
enlargement occurred was said to be within 2 years after radiotherapy 21, 22. Pollock
suggested that surgical resection should be delayed until continuous enlargement was
confirmed with serial imaging over 2 or more years 23. In this study, all patients were
followed for more than 2 years and our data was compatible with those reports.
Cautious follow-up could be helpful to distinguish the transient or progressive
enlargement and to avoid an unnecessary operation due to a misinterpretation of natural
history of tumor after radiotherapy. The salvage surgery for an irradiated tumor could
cause the poor patient outcomes 24, 25. Hypo-FSRT was effective to control tumor
growth over a long-term follow-up period. Watch and wait policy is important in the
follow-up of irradiated acoustic neuromas.
Six tumors, the diameters of the extrameatal portion of which were larger than 25
mm, were well-controlled locally with minimal morbidity in this study. Previously
favorable outcomes of radiotherapy for large acoustic neuromas were reported in a short
follow-up. Fourteen patients with tumors greater than or equal to 3.0 cm, received 30
Gy given in 10 fractions. No patient had growth of acoustic neuromas or developed
facial weakness with the median clinical follow-up 1.8 years 13. In another report,
tumors greater than or equal to 3 cm (16 patients) received 20 Gy in 5 fractions. All
tumors were controlled with median radiographic follow-up 20 months. No patient
developed trigeminal nerve symptoms after treatment nor did any patient require
surgery for treatment failure 26. In the treatment of large acoustic neuroma, a high local
control rate with minimal morbidity was achieved by maximal removal and adjuvant
radiotherapy 27, 28, however we believe that the treatment by radiotherapy alone is still
needed for some patients who are unfit for a surgical treatment due to medical reasons.
This study suggests that radiotherapy alone with hypo-FSRT is one of the acceptable
treatment options in the treatment for large acoustic neuromas.
It is reported that the annual rate of hearing loss was slower than accelerated after
radiotherapy but the tendency of deterioration in hearing level was observed even after
radiotherapy 29 30. In this study, the continuous deterioration of hearing was observed
after hypo-FSRT; serviceable hearing was preserved in all patients one year after hypo-
FSRT but 4 of 7 in 2-year after hypo-FSRT. The similar type of deterioration was seen
after SRS and therefore the rates of hearing preservation were lower in the long-term
follow-up reports; 44.5% ± 10.5% at 10 years 31 and 55% at 9 years in the reports of
SRS 32. The wide range of the rates of hearing preservation from 66 to 100% was
reported in hypo-FSRT. Those results may not reflect the change of continuous
deterioration of hearing level after radiotherapy because those reports suffered from a
relatively short follow-up period 12-15. The length of time required for stabilization of
hearing after radiotherapy is unclear. The number of literatures of hypo-FSRT was
growing but their follow-up period was short yet. The longer follow-up study is
required for evaluating the rates of hearing of hypo-FSRT, to reflect the delayed change
of hearing level.
The rate of facial nerve preservation and severe trigeminal nerve morbidity were
reported as 97 to 100% and 0 to 3% in hypo-FSRT 12-15 (Table 4). The rate of cranial
nerve morbidity was comparable to those of other radiotherapeutic techniques: gamma
knife 31, 33, linac-based SRS 32, 34, and conventionally fractionated SRT 8, 35. The findings
in this study substantiated the safety of hypo-FSRT for acoustic neuromas.
The local control rate for small-to-medium-sized acoustic neuromas is approaching
satisfactory, but the rate of hearing preservation is not. From the rationale of a linear-
quadratic model, fractionated SRT is expected to have radiobiological benefits for
normal tissue, versus SRS. From clinical experience, fractionated SRT did lead to a
higher rate of serviceable hearing preservation than SRS 7. Hypo-FSRT has the
possibility of being less time-consuming than conventionally fractionated SRT and of
better radiobiological benefit to adjacent normal tissues, compared with SRS. Excellent
local control with minimal morbidity were demonstrated in this study, however our data
weren’t far from adequate. This study was retrospective in nature; the population was
heterogeneous with regard to initial treatment, recurrence, tumor size, and hearing
ability. Additional studies are needed to evaluate the effectiveness of hypo-FSRT
especially for hearing preservation.
In conclusion, hypo-FSRT for acoustic neuromas achieved a high local control rate
with minimal facial and trigeminal nerve morbidity.