|Home | About | Journals | Submit | Contact Us | Français|
The purpose of this systematic review was to compare, from an audiological perspective, hearing preservation rates after fractionated stereotactic radiotherapy (FSRT) and stereotactic radiosurgery (SRS) for patients with vestibular schwannomas (VS). A literature review was conducted via Pubmed with the following phrases: radiosurgery, fractionated radiotherapy, vestibular schwannoma and hearing. Of 57 studies identified, six were included in the analysis based on the inclusion and exclusion factors. Tumor control rates were excellent (>90%) for both treatment options. Regarding hearing preservation, three of the six articles reported a significant advantage of FSRT over SRS, while the other three did not find significant differences. The FSRT method may result in higher hearing preservation in selected clinical settings: larger tumors (>3cc), age >55 years, and/or when the SRS dose prescribed was >13 Gy. However, as limitations exist and these studies were not randomized, further studies are needed which incorporate hearing outcomes using enhanced tools. This systematic review can help physicians and patients make more informed decisions regarding radiation treatment options for VS.
Vestibular schwannomas (VS), also known as acoustic neuromas, are indolent, benign tumors of the myelin-forming sheath cells of the eighth (vestibulocochlear) cranial nerve. This is often a slow-growing tumor caused by overproduction of the Schwann cells which normally wrap around the nerve like an onion skin to help support and insulate the nerve . Due in part to an increase in the incidental diagnosis via magnetic resonance imaging (MRI), the clinical incidence of VS has increased four-fold over the last quarter century from approximately 5 to 20 cases per million .
Patients with VS often report symptoms that include tinnitus, imbalance and unilateral hearing loss on the tumor side. Other possible symptoms include facial numbness, otalgia, or facial pain . Symptoms experienced by patient with VS often lead to a reduced quality of life including a negative health perception, adverse physical functioning, low social functioning, negative mental health function, bodily pain, low energy/fatigue, and physical restrictions to participation related to imbalance and dizziness .
Current treatment options for patients with VS include (a) close observation via serial MRI scans, (b) microsurgery and (c) stereotactic radiation . This review focuses on the option of stereotactic radiation therapy, which involves the precise administration of high, intense dose(s) of radiation to the tumor volume with millimeter accuracy with the goal of arresting tumor growth. This technique is accomplished by immobilizing the patient’s head, defining a coordinate system relative to the cranium, and mapping the tumor volume within the 3D frame of reference. Multiple shaped beams of ionizing radiation are accurately oriented to the center of the lesion, the isocenter, such that they sum to a highly conformal dose at the tumor margin .
Two different methods for delivering stereotactic radiation currently exist. The first, termed stereotactic radiosurgery (SRS), applies a single, intense dose of radiation in a solitary treatment session. The second method, called fractionated stereotactic radiation therapy (FSRT), breaks up the radiation dose into multiple, smaller fractions of treatment sessions, typically over a few weeks. Both treatment options use a stereotactic frame which can accurately localize the tumor in 3D space . While FSRT takes significantly longer to complete than SRS, the potential advantage of FSRT is the biological sparing of the surrounding normal tissues/nerves by using a gentler radiobiologic regimen . Thus, it may be possible that FSRT could potentially preserve hearing and the physiologic function of the auditory system than treatment with SRS.
While there have been no prospective randomized trials directly comparing pre and post treatment hearing for patients treated with SRS and FSRT, there have been a number of clinical studies that compare these techniques and have documented hearing-related treatment outcomes. However, most of these studies have not been analyzed from an audiological perspective. The purpose of this project was to conduct a systematic review of the literature to analyze whether multiple treatment sessions delivered via FSRT yield superior or inferior hearing preservation to that of a single-shot SRS treatment. Findings from this investigation may lead both patients with VS and their physicians to make an informed decision regarding the optimal stereotactic radiation treatment choice with respect to preserving post-operative hearing.
To examine hearing preservation rates, the majority of the research reviewed for this analysis classified hearing according to the standardized Gardner-Robertson (G-R) scale. The Gardner-Robertson scale is comprised of five grades. Each grade represents two scores for each patient: the pure tone average (PTA) and the speech discrimination (SD) score . For studies included in this analysis, the PTA and SD were only calculated for the involved side. A patient is placed into one of five grades based on the lower of the two scores. Grade I denotes “good” hearing in which either the reported PTA is 0-30 dB hearing level (HL) or the SD score is 70-100%. Grade II is described as “serviceable” hearing in which the patient reports either a PTA that ranges from 31-50 dB HL or an SD score of 50-69%. Grade III, “non-serviceable” hearing corresponds to either a PTA of 51-90 dB HL or SD score of 5-49%. Assigned to Grade IV, “poor” hearing, are patients who report either a PTA between 90-100 dB HL or an SD score of 1-4%. Grade V, “deaf”, is reserved for patients who report either a PTA of 0 HL or an SD score of 0. Overall, the studies reviewed for this analysis considered hearing preservation as maintaining a G-R grade I/II hearing following treatment, with hearing preservation rates determined by the number of patients assigned to these G-R grades. One study that was included in this analysis  utilized the American Academy of Otolaryngology – Head and Neck Surgery (AAO-HNS) method, defining hearing preservation as maintaining Class A or B, which corresponds to the G-R grade I or II, respectively.
A database search was conducted via PubMed using the key words “radiosurgery” and “fractionated radiotherapy” and “vestibular schwannoma” and “hearing”. Inclusion criteria for this systematic review included studies that reported on patients who had undergone stereotactic radiation treatment for vestibular schwannoma and for whom pre and post treatment measures of hearing were documented. Studies were excluded from analysis if not published in English, published before the year 2000 (to exclude studies utilizing older radiation techniques), reported on less than 100 patients (to provide larger patient group numbers to increase the statistical power to find potential differences) or did not compare FSRT to SRS (as this was the goal of this analysis). Lastly, studies in which an updated version was subsequently reported were disqualified in order avoid analyzing two studies with similar data sets.
Entering the key word “vestibular schwannoma”, the original PubMed search yielded 57 published investigations. As three fewer manuscripts were generated by using the key term “acoustic neuroma”, this term was not included as a search word. Based on the additional inclusion and exclusion factors described above, a total of six papers were selected to be part of this systematic review. The results from these six papers were reviewed regarding both tumor control and hearing preservation with special attention to the radiation technique used (FSRT vs. SRS). Table 1 provides the key demographic information, tumor size, radiation dose, follow-up data and results (for tumor control and hearing preservation) based on the technique (FSRT vs. SRS) used in each paper. As the tumor control was excellent (>90%) regardless of the technique, a brief synopsis of the hearing outcomes for each article will be provided first, followed by a summary of the findings.
Andrews et al.  reported on a total of 125 participants, who were grouped according to treatment approach (SRS: n=69; FSRT: n=56). Hearing outcomes were classified via the Gardner-Robertson (G-R) scale. Median follow-up was 1.3 years for the FSRT group and 1.2 years for the SRS group. Findings found significantly better hearing preservation rate for patients treated with FSRT (81%) compared to those treated with SRS (33%) (p = .02).
Collen et al.  reviewed data from a total of 119 patients treated for VS (SRS: n=78; FSRT; n=41). While authors found no significant differences in hearing preservation rates (via the G-R scale) at four-year post-treatment, the hearing preservation rate was higher for the FSRT group (82%) when compared to SRS group (59%) (p =.09).
Combs et al.  reported hearing preservation rates via the G-R scale over a 10 year period for 216 patients with VS treated with FSRT and 32 treated with SRS. The hearing preservation rates for patients treated with FSRT (76% at 5 years and 66% at 10 years) were significantly higher than the hearing preservation rates for those treated with SRS (60% at 5 years and 55% at 10 years for the SRS patients) (p =.04). However, for patients treated with an SRS radiation dose of ≤13 Gy, there was no significant difference in hearing preservation rates (via G-R) between groups (p = .4).
Fong et al.  included data from studies (published through 2010) in order to analyze hearing preservation rates (via G-R and/or AAO-HNS scales) for patients treated with either FSRT or SRS. Of note, the authors conducted a meta-analysis across 21 studies, two of which were the previously mentioned investigations by Andrews et al.  and Collen et al. . Because the data collected from these two studies only accounted for 20% of the data set described by Fong et al. , findings from this study were included and reviewed independently for the present systematic review. Of the studies that investigated outcomes of SRS treatment, 400 patients were identified across nine studies. Compiled data suggested an overall hearing preservation rate of 66% with SRS. Regarding the FSRT technique, 629 patients were identified across 12 studies. Compiled data suggested an overall hearing preservation rate of 75% with FSRT, which was significantly higher than the hearing preservation rate among patients treated with SRS (p = .004). Of note, tumor volume and patient age appeared to have potential mediating effects. For patients with larger tumors (>3.0 cc), the hearing preservation rates were significantly higher for those treated with FSRT (94%) than those treated with SRS (71%) (p = .02). For patients 55 years and older, hearing preservation rates were significantly higher for those treated with FSRT (81%) than those treated with SRS (62%) (p < .001).
Kopp et al.  investigated 115 patients (SRS: n= 68; FSRT n=47). Patients in the SRS group received a total radiation dose of 12 Gy and had an average maximum tumor diameter of <1.5 cm versus the FSRT group which had an average maximum tumor diameter of >1.5 cm. Hearing preservation (via G-R classifications) was not reported to be significantly different between the FSRT group (79%) and the SRS group (85%).
Meijer et al.  examined data from 129 patients (SRS: n=49; FSRT: n= 80). It is important to note that hearing in this study was assessed via use of the telephone on the affected side, not by standard audiometric measures that allow for the G-R classification. If the patient could not “discriminate words or could not hear at all”, he/she was scored as “deaf”. When hearing outcomes were assessed in this manner 5 years post-treatment, no significant differences in hearing preservation between treatment groups were found (61% for FSRT vs. 75% for SRS) (p = .42).
In summary, findings across studies meeting the inclusion criteria for this systematic review suggested excellent tumor control rates (>90%) for both treatment types (FSRT or SRS). None of these demonstrated SRS treatment to be significantly better at preserving hearing than FSRT. However, data from three of the six papers (by Andrews et al. , Combs et al.  and Fong et al. ) suggested that, overall, there appears to be a significantly better hearing preservation rate when patients were treated with FSRT. Although the findings of Collen et al.  did not demonstrate a significant difference in hearing preservation rates between the two treatment groups, their data did suggest a trend favoring FSRT over SRS. Two of studies (Kopp et al.  and Meijer et al. ) demonstrated no distinct advantage with the application of one treatment approach over the other in terms of hearing preservation. The FSRT method appeared to result in higher hearing preservation rates in certain clinical situations: larger tumors (>3 cc), age >55 years, and/or when the SRS dose prescribed was >13 Gy.
The optimal management for patients with VS remains controversial. The primary purpose of this systematic analysis was to analyze whether FSRT results in superior or inferior hearing preservation than SRS for patients with VS. This is a critical question that can help patients and physicians evaluate the optimal radiation treatment strategy for this condition. Based on the literature search methodology described above, this systematic analysis included six studies.
Before one can analyze hearing preservation, it is imperative to first discuss treatment efficacy with respect to tumor control. If one radiation strategy improved hearing preservation at the expense of decreased tumor control, this outcome would not be acceptable for patients. Fortunately, all of the studies in this analysis found high rates of tumor control (>90%) for both treatment methods with no statistical differences in tumor control. As the studies consistently demonstrated excellent tumor control rates with either radiation technique, it is reasonable to explore in detail which technique results in better hearing preservation.
To date, there is no level I evidence (or randomized study) that directly compares hearing preservation rates between individuals treated with SRS or FRST . As such, it is not possible to make definitive recommendations. Indeed, caution must be taken when analyzing each of these six papers, which constitute evidence from nonrandomized, controlled studies (i.e., level III evidence) for potential selection factors that may have influenced these results . It is noteworthy that Andrews et al.  was the only study that originally designed a randomized study comparing treatment outcomes between FSRT vs. SRS. However, the authors report that it became too challenging to recruit patients to a randomized study “because of either patient expectation or physician bias” .
Of the six studies included in this systematic review, none of them showed SRS to be superior to FSRT for hearing preservation. However, three of these papers, by Andrews et al. , Combs et al.  and Fong et al. , reported an overall significantly improved hearing preservation rate with FSRT over SRS. While Collen et al.  did not show a significant advantage of FSRT over SRS, the authors did report a trend in this direction (with hearing preservation rates of 82% with FSRT vs. 59% with SRS, p = .09). Two of the six comparative studies by Kopp et al.  and Meijer et al. , however, did not show a significant difference in hearing preservation rates between FSRT and SRS. A closer analysis of these two studies, though, may help explain this finding. In the study by Meijer et al. , hearing outcome was not measured by a standardized hearing scale, such as the G-R and AAO-HNS scales applied in the other studies, but rather each participant listened on the phone and was scored on his or her ability to discriminate speech on the affected side. If the patient could not discriminate words, the patient was scored as “deaf”. By any audiological standard, telephone use would not be considered an accurate test of hearing preservation. Not only does this test not allow for frequency specific information, but the telephone use was not reported to be in an environment with low noise levels, nor was the type of phone quality noted. This less sensitive method may explain the findings that suggested no difference in hearing outcomes between the two radiation techniques.
From collective efforts made by investigators across studies to control for a number of independent variables, three factors appear to have a potential influence on hearing outcomes for patients treated with radiotherapy: tumor volume, patient age and SRS treatment dose. Findings from Fong et al.  suggested that tumor volume may play a mediating role on hearing preservation rates. Their data suggested a significant advantage for FRST over SRS for hearing preservation particularly in those individuals with larger tumor volumes (> 3 cc). Patient age may also be a mediating variable. Fong et al.  found that among patients older than 55 years of age, the hearing preservation rate was significantly higher when treated with FSRT rather than SRS. Another potential mediating variable may be the SRS treatment dose. Based on findings from Combs et al. , no decrement in hearing preservation with SRS was found when doses of <13 Gy were utilized. However, in patients who received higher SRS doses, the authors reported that the hearing preservation rate following this treatment was significantly lower than the rate for patients treated with FSRT. Of note, it appears that SRS doses >13 Gy in this setting are no longer utilized as they were in the past. Indeed, preliminary data suggest that higher radiation therapy doses to the cochlea are associated with higher rates of hearing loss . Future studies need to carefully document and calculate the precise radiation therapy dose to the region of the cochlea.
While the study by Kopp et al.  showed no difference in hearing outcomes between FSRT and SRS, this study focused mostly on patients who had small mean tumor volumes (1.2 cc) and who were treated to a SRS dose of 12 Gy. Thus, based on several potential mediating factors noted above for which FSRT was found to be better at preserving hearing than SRS (e.g., tumor volume >3 cc and SRS dose >13 Gy), it is not unexpected that findings from the Kopp et al.  study found no significant difference in hearing preservation rates between FSRT and SRS.
In addition to the limitations inherent to non-randomized experimental designs, there are other potential limitations that need to be considered when reviewing findings from these studies. In three of the studies, there was a large difference in group sizes between the SRS and FSRT groups. For example, in the papers by Combs et al.  and Fong et al. , there were more patients in the FSRT group. In the Collen et al.  study, on the other hand, there were more patients in the SRS group. These discrepancies in the sizes of the patient groups may be related to underlying (and unknown) selection factors by which patients were separated into the two groups. Such selection factors may have been based, for example, on physician and/or patient preferences and/or other undetermined factors, such as tumor size or patient age, and make comparison between studies difficult.
Another important difference was the mean age of the two groups. In five of the six studies, there was an older mean age in the SRS group compared to the FSRT group. In fact, Meijer et al.  reported that the SRS patients were significantly older than the FSRT patients. One possible factor which may have lead to an age difference in patient preference is that it may have been harder, for example, for older patients to return to the radiation clinic for multiple treatments due to lack of transportation or difficulty leaving the house. Thus, older patients may have more likely opted for the single session SRS treatment strategy. Because age has been shown to be a mediating factor in hearing preservation following radiotherapy, their age might have also contributed to greater hearing damage (i.e. lower hearing preservation rate) as a result. On the other hand, the gender ratios were about equal across the various studies and no significant gender effects were observed.
Another difference was the variation in tumor sizes between patients who received FSRT vs. SRS. Collen et al. , Combs et al. , and Kopp et al.  all reported tumor sizes in the FSRT group that were larger than those of the SRS group. Collen et al.  found statistical significance between the tumor sizes of the two groups. Despite the fact that larger tumor sizes were treated with FSRT, overall, this technique still had the same or better hearing outcomes than SRS. Of note, two papers comparing FSRT to SRS outcomes had no significant tumor size differences and also found better hearing preservation rates for the FSRT group [8, 8].
It is also relevant to recognize the differences in the follow-up times between studies, which may impact the hearing preservation rates (see Table 1). Importantly, within each study, the follow-up times between the two techniques (FSRT vs. SRS) are quite similar. However, this analysis cannot account for such differences between these studies. Similar limitations are noted in an excellent prior review by Murphy and Suh  from 2011, which included three of the six series analyzed here comparing FSRT to SRS for VS [9, 11, 17], of which one prior study  has just recently been updated and included in the current review . Murphy and Suh point out that in the paper by Andrews et al.  there was “a short audiometry follow-up of 38-41 weeks”. Moreover, the “rate of hearing preservation after SRS” in this study “was at the low end of single institution SRS data”. Yet, they acknowledge that others have “reported that most hearing detriment was noted at 6-10 months after treatment” . Of note, our current review includes more recent and updated data with longer follow-up that were not previously available. Overall, the current analysis is in agreement with Murphy and Suh that “one cannot conclude that fractionated [i.e., FSRT] is better”, though at the same time, “the best results have been from FSRT, with 63-94% hearing preservation rate” .
Yet another confounding factor which may influence the findings in all the studies is the likelihood of age-related hearing loss, presbycusis, among the patients in these studies, many of whom were elderly. About 30-35% of older adults, between the ages of 65-75, have hearing loss due to presbycusis. Presbycusis is thought to be caused by noise exposure over time, hereditary factors, and/or various health conditions. Presbycusis can also be caused by changes of blood supply to the ear . Thus, as the population of patients seen for radiation treatment generally ranged from a mean age of 53 to 61 years, it can be difficult to completely separate out whether the hearing loss over time was related to the radiation technique, aging, and/or a combination of these factors.
Besides the study by Meijer et al.  which assessed hearing preservation based on telephone use, all the other studies used a standardized method (mostly the G-R scale) to evaluate hearing outcomes for each patient. Yet, there are many limitations of the current standard scales (such as the G-R and AAO-HNS scales) to accurately assess functional hearing. As pointed out by Gurgel et al. , the two reported scores (i.e., pure tone average, PTA, and speech discrimination, SD) that define the G-R grade may not always be consistent with each other. For example, a particular patient’s PTA may not fall within the same grade level as the patient’s SD score. According to the G-R rating scale, if this situation were to arise, the patient would be placed in the lower of two grades. Furthermore, both the G-R and the AAO-HNS scales use large ranges of scores to define a single grade. This clusters patient scores into a single grouping, when in fact, their scores may be quite dissimilar to each other.
Due to the limitations of these scales, Gurgel et al.  developed a new method of reporting data in which both the PTA and SD scores are maintained separately to give the analyst a better understanding of affected hearing outcomes. Figure 1 illustrates a new system that uses a scattergram to plot both scores. Each patient is grouped into a category and placed into one of 100 spots on the scattergram. Each of the 100 boxes represents a much more specific range by dividing PTAs into ten 10 dB wide categories and dividing SD scores into ten categories, each 10 percentage points wide. If more than one patient falls in a single box, the sum total of all the patients in each box is added up to represent the number of patients that had a particular combination of PTA and SD scores. Another way to present this information, that could be particularly helpful, is to determine if any improvement occurred after treatment. Figure 2 illustrates a post-treatment analysis. This method is evolving into a new standard for reporting hearing preservation rates. This approach should be strongly considered as the new minimal data standard which allows the investigator to categorize the results in a clinically meaningful way and allows for other analysts to easily and thoroughly interpret data.
Another possible method for tracking more accurate and objective changes in hearing may be with otoacoustic emission (OAE) measures. Radiation treatment affects the cochlea and VIIIth nerve via beams of radiation that pass through the cochlea. A measure of OAEs is an objective and quick test which provides information regarding the function of the outer-hair cells in the cochlea, which can be damaged by radiation. An OAE assessment does not rely on the patient’s response to sound, but rather sends a sound in to the ear and assesses for a response that is elicited from the outer-hair cells. With damage to the outer-hair cells, one would expect the response to be low or absent. Thus, in addition to hearing tests, measures of OAEs may be able to improve future studies by objectively measuring the actual damage that is caused to the cochlea when deciding on the best method for treatment . These measures, however, can be easily confounded by the presence of middle ear dysfunction or pre-existing sensory/outer-hair cell loss. Thus, this test would likely serve a complimentary role to other measures of pre and post treatment hearing sensitivity.
In summary, in the absence of randomized studies to date, the evidence for better hearing preservation in VS patients treated with FSRT compared to those treated with SRS is inconclusive. Based on review of the literature, FSRT may result in higher hearing preservation in selected clinical situations: for patients who have larger tumors (>3 cc), are over the age of 55 years, and/or when the planned SRS dose prescribed would be >13 Gy. However, future studies, with fewer methodological limitations, are needed to further assess this important issue. As both FSRT and SRS have excellent tumor control rates, it is important for the physician to discuss the relative benefits and risks of the radiation treatment options for VS (whether FSRT or SRS) with each patient based upon his or her individual clinical situation. In addition to tumor control, future studies should incorporate hearing outcomes, using enhanced tools, as an essential endpoint that needs to be carefully assessed in patients with VS.