Departments of Neurosurgery (HGD) and Otorhinolaryngology (MSM), Mayo Clinic Scottsdale, Scottsdale, Arizona; Departments of Neurosurgery (MJE, LQ) and Otorhinolaryngology (SGH, CWB), Mayo Clinic Rochester, Rochester, Minnesota; and Departments of Neurosurgery (REW) and Otorhinolaryngology (JDG), Mayo Clinic Jacksonville, Jacksonville, Florida
OBJECTIVE: This study analyzed selection criteria, clinical outcome, and tumor growth rates in patients with acoustic neuromas in whom the initial management strategy was observation.
METHODS: A retrospective review of patients with conservatively managed unilateral acoustic neuromas was conducted. Minimum follow-up was 6 months. Patients with neurofibromatosis Type II were excluded. Differences in tumor growth rates were analyzed by use of the Wilcoxon rank sum test.
RESULTS: Sixty-eight patients (31 men and 37 women) with a mean age of 67.1 years were followed for an average of 3.4 years after diagnosis. The reasons for a trial of observation included advanced age (55%), patient preference (21%), minimal symptoms (9%), poor general medical condition (7%), asymptomatic tumor (4%), and tumor in the
only hearing ear (4%). Fifty-eight patients (85%) were successfully managed with observation alone. Ten patients (15%) ultimately required treatment (nine received microsurgical treatment and one patient underwent radiosurgical intervention) at a mean time interval of 4.0 years after diagnosis. Forty-eight tumors (71%) showed no growth and
20 (29%) enlarged during the study period. The mean tumor growth rate at the 1-year follow-up was significantly higher in the group requiring treatment (3.0 mm) than in the group not requiring treatment (0.36 mm) (P < 0.0001). Thus, the tumor growth rate at the 1-year follow-up was a strong predictor of the eventual need for
treatment.
CONCLUSION: Observation is a reasonable management strategy in carefully selected patients with acoustic neuromas. Diligent follow-up with serial magnetic resonance imaging is recommended, because some tumors will enlarge to the point at which active treatment is required. (Neurosurgery 39:260266, 1996)
Key words: Acoustic neuroma, Magnetic resonance imaging, Vestibular schwannoma
The first successful removal of an acoustic neuroma was performed by Ballance (2) in 1894 and reported in 1907. Reports of acoustic neuroma operations from the early 20th century chronicle the difficulties encountered in treating these tumors. Acoustic tumors were generally not diagnosed until they reached substantial size and caused multiple cranial nerve symptoms, brain stem compression, and an increased intracranial pressure. Procedures were often limited to subtotal removal, and mortality rates of ¾84% were reported (7). Even when the patient survived, the facial nerve was almost always left nonfunctional (13).
During the past 30 years, there have been many advances in the diagnosis and treatment of acoustic neuromas. Development of the operating microscope, refined surgical techniques, sophisticated intraoperative monitoring of cranial nerve function, and improved neuroanesthesia led to dramatic improvements in surgical outcome. Mortality rates of <1% and facial nerve preservation rates >90% have been reported in large series (8, 14, 20). In contemporary practice, hearing preservation is a realistic goal that is being achieved with increasing frequency (8, 9, 11, 12, 18).
During this same period, there have been equally important advances in diagnostic imaging techniques, first with computed tomography (CT) and, more recently, with magnetic resonance imaging (MRI). By use of gadolinium-enhanced MRI with fine axial cuts, the diagnosis of small intracanalicular acoustic neuromas is now possible (24, 28). This technology has resulted in earlier diagnosis of smaller tumors that are more amenable to surgical excision.
Despite these advances, microsurgical treatment still has some risks, and the potential exists for serious complications, even at centers with extensive experience. Furthermore, increased usage of MRI has led to the diagnosis of increasing numbers of small, minimally symptomatic, or even completely asymptomatic, tumors. It is also known that acoustic neuroma is a benign tumor that grows slowly in most patients. Thus, it seems logical that a conservative treatment strategy might be appropriate in selected patients. The authors report a retrospective series of patients with acoustic neuromas initially managed conservatively, with attention to selection criteria, clinical outcome, and tumor growth rates.
The authors conducted a retrospective review of patients with radiographically documented acoustic neuromas seen in the Mayo Clinic system between January 1, 1983, and December 31, 1992. This included patients seen at the Mayo Clinic facilities in Rochester, MN, Jacksonville, FL, and Scottsdale, AZ. Patients in whom the initial management decision was for observation were selected for additional study. The minimal follow-up interval was 6 months. The follow-up was conducted by direct patient examination and review of imaging studies and other records. Patients not accessible for re-evaluation after their initial assessment were excluded from detailed analysis. Individuals with neurofibromatosis Type II were also excluded.
Study parameters
Study parameters included the symptoms, the audiogram results, the type of imaging study, the tumor size, and the reason why no treatment was given. These were recorded for the initial evaluation and for each subsequent patient contact. The duration of follow-up was defined as the interval between the first imaging study that showed the acoustic neuroma and the final imaging study or operation. Tumor size was defined as the mean of the greatest anterior-posterior and medial-lateral dimensions of the tumor (17). During the follow-up period, particular attention was focused on the development of new or progressive symptoms, tumor growth rates, and whether treatment was ultimately required.
Statistical methods
Differences in tumor growth rates were analyzed by use of the Wilcoxon rank sum test.
Of these 32 patients, 15 were contacted; they reported no new symptoms and did not wish to return for further study. Most of these were elderly patients with small tumors and minimal symptoms. Eight patients of the 32 were deceased, six of unrelated cause and two (aged 76 and 90 yr, respectively) of uncertain cause. One patient underwent an acoustic neuroma operation elsewhere. Eight patients could not be located, despite diligent efforts to follow up by telephone and mail, and were therefore considered lost to follow-up. Although some information is available on these 32 patients, all were excluded from additional study because they were seen only one time and no follow-up imaging was available.
There remained 68 patients, with unilateral acoustic neuromas initially managed conservatively, for whom at least a 6-month follow-up was obtained. There were 31 men and 37 women, with a mean age of 67.1 years (range, 3580 yr). The average follow-up period was 3.4 years, with a range from 6 months to 12 years (Fig. 1).
Presenting symptoms are shown in Table 1. Hearing loss on the side of the tumor was present in 85% of patients, vertigo or dysequilibrium in 38%, and tinnitus in 34%. Trigeminal and facial nerve symptoms were uncommon.
| Symptom | No. of Patients |
|---|---|
| (n = 68) (%) | |
| Hearing loss | 58 (85) |
| Vertigo | 26 (38) |
| Tinnitus | 23 (34) |
| Trigeminal neuropathy | 3 (4) |
| Facial neuropathy | 2 (3) |
| Lower cranial nerve (IX-XI) dysfunction | 0 |
The reasons for the initial selection of conservative treatment are shown in Table 2. The majority of patients (55%) were not recommended for surgical treatment because of advanced age. The mean age of this subgroup was 73.4 years. In 14 patients (21%), the decision reflected patient preference. Smaller numbers of patients were selected for observation because they had minimal symptoms, usually with a small tumor (9%), or because they were asymptomatic, having undergone brain imaging for an unrelated concern (4%). Three patients had a tumor in their only hearing ear (4%). Five patients (7%) were managed conservatively, because surgical risks were excessive as a result of poor general medical condition. These five patients included three with severe chronic obstructive pulmonary disease, one with carcinoma of the lung and chronic obstructive pulmonary disease, and one with severe multiple sclerosis.
| Reason | No. of Patients |
|---|---|
| (n = 68) (%) | |
| Advanced age | 37 (55) |
| Patient preference | 14 (21) |
| Minimal symptoms | 6 (9) |
| Poor general medical condition | 5 (7) |
| Asymptomatic tumor | 3 (4) |
| Tumor in only hearing ear | 3 (4) |
The choice of imaging modality reflected the evolution of imaging technology during the study period. MRI was used in 36 patients (57%) and CT in 25 (37%). Seven patients (10%) were imaged with magnetic resonance and CT. CT was used extensively during the early years of the study, but, in recent years, MRI has been used almost exclusively in following these patients. At present, MRI with gadolinium enhancement with fine axial cuts through the internal auditory canals is the imaging procedure of choice for diagnosis and follow-up evaluation of patients with acoustic neuromas. In selected patients, CT may be useful in surgical planning.
The clinical outcomes of these 68 patients are shown in Table 3. Fifty-eight patients (85%) were managed with prolonged observation alone. Ten patients (15%), including three men and seven women, with an average age of 65.0 years, ultimately required treatment. The mean interval between initial diagnosis and treatment was 4.0 years. Nine patients received microsurgical treatment. The reason for surgical intervention was tumor growth with progressive symptoms and signs in six patients, and tumor growth without clinical progression in three. Eight of these nine patients underwent elective operations, and one required an urgent operation. This latter patient had declined surgical treatment after presenting with an 18-mm tumor. She was lost to follow-up and returned 5 years later with a 33-mm tumor, which had caused brainstem compression, obstructive hydrocephalus, and obtundation.
| Outcome | No. of Patients |
|---|---|
| (n = 68) | |
| No treatment | 58 (85%) |
| Microsurgery | 9 (13%) |
| Radiosurgery | 1 (2%) |
| Disposition | |
| Alive | 67 |
| Deceased, unrelated cause | 1 |
| Deceased, postoperative mortality | 0 |
One patient presented with a tumor with a 4-mm extension into the cerebellopontine angle that showed no growth for 6 years. At that point, the tumor began to show gradual enlargement, and the patient subsequently received radiosurgical treatment 8 years after the initial diagnosis. A subgroup of 10 patients with a mean age of 72.9 years had tumor growth but did not undergo treatment. The reasons for continued conservative management in this subgroup included advanced age and minimal symptoms in six patients, poor general health in three patients, and patient preference in one case. Sixty-seven patients remained alive, and one patient in the prolonged observation group died of an unrelated cause (chronic obstructive pulmonary disease). There were no deaths in the treatment group.
Forty-eight tumors (71%) demonstrated no growth and 20 (29%) enlarged during the study period. There were no instances of tumor regression. The mean tumor size at the time of diagnosis was 12.2 mm in the overall group, 12.3 mm in patients not requiring treatment, and 11.7 mm in the group requiring surgical intervention. This compares with a mean tumor size of ~25 mm in patients selected for operation at the time of diagnosis (8). The mean tumor growth rate at the 1-year follow-up was 0.72 mm in the overall group, 0.36 mm in patients not requiring treatment, and 3.0 mm in the group requiring treatment. Patients with higher tumor growth rates at the 1-year follow-up were much more likely to eventually require surgical treatment (Wilcoxon rank sum test, P < 0.0001).
As noted in Figure 1, four patients had follow-up limited to 6 months. Two of these underwent operations after tumor growth was demonstrated at the 6-month interval, and the other two had no tumor growth and no treatment. These latter two patients were excluded from tumor growth-rate studies. Thus, tumor growth-rate calculations included only those patients with a minimum of a 1-year follow-up (66 patients) or treatment at the 6-month interval (2 patients). Subsequent tumor growth was similar to that seen at the 1-year follow-up. The mean tumor growth rate at the 2-year follow-up was 0.55 mm in the overall group, 0.47 mm in patients not requiring treatment, and 1.67 mm in the group requiring active treatment.
During the past 3 decades, there have been dramatic advances in the diagnosis and treatment of acoustic neuromas. For most patients, total excision in a single stage is now the preferred treatment (1). Other surgical options occasionally used include subtotal removal and cerebrospinal fluid diversion for obstructive hydrocephalus. Various forms of radiation treatment are also available, with stereotactic radiosurgical treatment playing an increasing role in the management of these tumors (19). Despite these therapeutic advances, there are patients for whom long-term observation may be preferable to immediate treatment.
Some experienced surgeons expressed the view that an immediate operation is preferable to observation with the possibility of delayed surgical treatment; these surgeons advised operation at the time of diagnosis in virtually all patients (4, 15, 17). Others advocated a more conservative approach of observation or subtotal resection in selected patients (3, 6, 10, 2123, 25, 26). In a retrospective review of 70 patients with acoustic tumors initially managed with observation, Bederson et al. (3) found that 28 tumors (40%) had no demonstrable growth and 4 (6%) showed apparent regression. Conservative treatment was thought to be possible, provided that serial radiological studies are obtained.
Strasnick et al. (23) reviewed 51 patients with conservatively managed acoustic neuromas who were followed for an average of 2.6 years. Thirty-nine patients (76%) required no treatment, 11 (22%) underwent operations, and 1 (2%) received radiosurgical treatment. These authors thought that appropriately selected patients with acoustic tumors could be safely followed with periodic imaging studies. Wiet et al. (27) reported on 53 patients with acoustic tumors managed initially without intervention. Two patterns of growth were identified: no (or slow) growth and a more relentless growth. The authors concluded that a certain population of patients with acoustic neuromas can be managed conservatively regardless of age. Charabi et al. (5) reported on 123 patients with conservatively managed acoustic neuromas. This study included nine patients with neurofibromatosis Type II. Tumor growth was seen in 74%, and the authors thought that the indications for a "wait and see" management strategy were limited.
Present study
The mean age of the patients in the present study was 67.1 years. By comparison, a recent series of patients who underwent microsurgical excision of acoustic neuromas at the time of diagnosis at our institution had a mean age of 50.0 years (8). Thus, the patients selected for observation were significantly older than those who underwent prompt operations. The patients selected for observation also had significantly smaller tumors (a diameter of 12.2 mm versus 25 mm) than the prompt operations group. Presenting symptoms were virtually identical between the two groups.
There were two distinct patterns of tumor growth among those patients initially selected for observation. The majority of patients had slowly growing tumors, and the group successfully managed with prolonged observation alone had a mean growth rate at the 1-year follow-up of 0.36 mm. In contrast, the group of patients ultimately requiring treatment had a mean growth rate at 1-year follow-up of 3.0 mm. Consistent with the findings of other investigators, we found no clinical or radiographic characteristic that would enable the clinician to reliably identify, in a prospective manner, those tumors that would grow and require treatment. Furthermore, in the group of 10 patients who ultimately received treatment, it was not possible to analyze whether tumor growth was linear or erratic, because most of these patients were operated on at the first sign of tumor enlargement. The apparent regression of acoustic neuroma has been reported (3, 16), but we found no examples of this in our series.
The major limitation of this study was the follow-up interval. A 3.4-year mean follow-up period is relatively brief, given the slow growth rate of most of these tumors. Another limitation was the variation in imaging procedures that resulted from the dramatic changes and improvements in acoustic neuroma imaging during the study period. Most of the patients entered in the early 1980s were studied only with CT. In the mid-1980s, nonenhanced MRI became available. It was only in the final 3 years of the study period that patients were consistently evaluated with the current standardMRI with gadolinium enhancement. A prospective study, in which patients are followed systematically by use of MRI with gadolinium enhancement, would be a logical next step to take to further our understanding of the natural history of acoustic neuroma.
Although observation was generally well tolerated, the fact that 29% of tumors enlarged and 15% required some form of treatment, usually microsurgical intervention, during a limited follow-up interval underscores the need for careful monitoring with serial MRI. This need was further highlighted by the patient who required an urgent operation for a 33-mm tumor 5 years after declining surgical intervention and follow-up monitoring for a tumor that initially measured 18 mm. Another patient presented with a 4-mm tumor, which was unchanged during 6 years of observation. The tumor began to grow during the 7th year, and radiosurgical intervention was performed 8 years after diagnosis. This latter patient shows the importance of continued, and, in most patients, life-long follow-up, even in patients whose tumors show no growth for years at a time.
Received, October 30, 1995.
Accepted, February 26, 1996.
Reprint requests: H. Gordon Deen, Jr., M.D., Mayo Clinic Scottsdale, 13400 East Shea Boulevard, Scottsdale, AZ 85259.
In my report of the microsurgical treatment of 410 patients with acoustic neuromas, the mortality rate was 0.5%, and 99% of the patients were independent in their activities (1). However, I noted that there were indications for observation that included a long history of only auditory symptoms in patients of any age with any size tumor, elderly patients with mild symptoms, patients with asymptomatic tumors, some patients in whom the only hearing ear was involved, and the preference of the patient after discussion of the treatment options. As the authors concluded, these patients must be carefully followed indefinitely. The true incidence of growth arrest must await the results of long-term follow-up magnetic resonance imaging (MRI) studies.
Robert G. Ojemann
Boston, Massachusetts
Deen et al. provide information that will assist our patients in selecting from among the different treatment options for acoustic neuromas. The majority of patients with acoustic neuromas harbor a slowly growing tumor that does not pose an immediate threat to life and that affords them the time necessary to consider the various treatment options and to make a decision on the basis of knowledge of the natural history of the tumor and the risks and benefits of each form of treatment. Microsurgical removal of the tumor is clearly the treatment of choice in the majority of patients, but there are patients outlined in this study who can safely be observed without treatment and others in whom radiosurgical treatment will prove to be the method of choice. Fortunately, this tumor provides most patients with the time to collect and consider information about the various treatment options. Observation without treatment is a safe alternative when there is periodic follow-up with clinical examinations and imaging studies. One source of material for patients is the Acoustic Neuroma Association (P.O. Box 12402, Atlanta, GA 30355), which has information helpful before and after surgical intervention. The information presented here further strengthens my thoughts that our best ally in the treatment of neurosurgical disease is a well-informed patient.
Albert L. Rhoton, Jr.
Gainesville, Florida
The natural history of untreated acoustic neuroma is an important subject that is not well studied. Gadolinium-enhanced MRI has now given us a tool to discover and follow acoustic neuromas from their early stages. Most neurosurgeons treating significant numbers of patients with acoustic neuromas have managed a subset of patients conservatively, but the bias favoring surgical treatment, in the referral of patients to surgeons and in the recommendations that surgeons make, is obvious. This article presents, from existing data, a strong case in favor of a more deliberate and organized approach to understanding the natural history of acoustic neuroma in the MRI era.
There are important problems with the information presented here. The minimal follow-up period is 6 months, which is clearly too short to provide meaningful information in the majority of acoustic tumors. Even the 3.4-year average follow-up is short in the lifetime of such a slowly growing lesion. The authors should be strongly encouraged to continue to follow these patients for a much longer time. The loss of 32 patients to follow-up is also a serious problem. The lack of progressive symptoms does not rule out significant growth. One-third of the patients were excluded from the study; this makes generalization of the results difficult.
In assessing studies that attempt to define the prognosis of a given condition, several questions should be asked (1).
1) Was an inception cohort assembled? These authors identified patients at the time of referral to their institution. This creates a potential bias, in that patients may be selected for referral because of factors, such as concern about growth, that might make this group of patients more susceptible than an unselected population to subsequent growth that requires treatment.
2) Was the referral process described? The Mayo Clinic is well known as a tertiary and quaternary referral center. The authors do not provide any information that would allow us to assess factors that might influence the choice of patients to be referred into this study.
3) Was complete follow-up achieved? This problem has been addressed above and is a significant difficulty with the study.
4) Were objective outcome criteria developed and used? The two outcome criteria in this study are growth as measured by imaging studies and the decision to undergo an operation. The measurement of tumor growth by use of computed tomographic and magnetic resonance scans is known to have significant difficulties, and the reproducibility and reliability of the authors' measurements are not clear. Such measurement, however, is significantly more objective than the decision to undergo an operation. Such a decision involves many different factors, including the bias of the surgeon and of the patient. It is quite possible that a different group of surgeons advising the same group of conservatively managed patients might have chosen a quite different group of patients for surgical treatment.
5) Was the outcome assessment blind? The authors do not tell us if the measurements of tumor size were made with or without the knowledge of the mode of management or ultimate decision for surgical treatment. The decision to operate obviously could not be made in a blinded fashion. The rest of the information of the study, however, could have been collected from the records in blinded fashion. We do not know if this was done.
6) Was adjustment for extraneous prognostic factors made? Although the population is well described, it is sufficiently small that multivariable analyses are really impractical, and, therefore, such adjustments were not made.
In summary, the authors' laudable goal has been partially achieved. There are some technical improvements to this study that could be accomplished, but a more comprehensive study with pooling of information from a number of large centers would provide even more precise information. The information presented, however, suggests that untreated patients have a high incidence of lack of tumor progression. Such patients would be considered to have been "successfully treated" had they been given radiosurgical treatment at the inception of the study. These findings strongly suggest that the proper comparison group for patients treated with radiosurgical techniques is a group of untreated patients rather than those treated microsurgically. MRI is a powerful tool that has revealed lesions, such as acoustic neuroma and meningioma, at an early and often asymptomatic stage in their development. Too often such lesions are treated on the basis of the assumption that they will necessarily grow to a size comparable to that which required treatment in the past. Our obligation in working with such powerful imaging instruments is to carefully study the natural history of minimally symptomatic lesions before concluding that surgical intervention is appropriate. The authors started this process and are encouraged to join with others to produce more definitive evaluation.
Stephen J. Haines
Minneapolis, Minnesota
The ability to track the growth rate of vestibular schwannomas has improved dramatically because of MRI. Thus, careful prospective studies of conservatively managed patients with acoustic neuromas followed by serial MRI can arrive at accurate determinations of tumor growth rate. The present study by Deen et al. provides little new information about the average growth rate of such tumors, because it examines only 68 of 626 patients with vestibular schwannomas seen by the authors in a 10-year period. In other words, |mf80% of the patients underwent surgery and were not available for any serial measurements of the tumor and another ~6% were not followed, leaving ~13% available for serial measurements of growth rate. The average follow-up in these 68 of 528 patients was only 3.4 years, which is far too short for meaningful data in a slow-growing benign tumor of this type. Indeed, some of the cases were followed for only 6 to 12 months.
This highly selected group of 68 patients had an average age of 67 years. In most series of patients with such tumors, the average age is many years younger than in the current study.
Thus, the present article provides little information about the growth rate of vestibular schwannomas because a highly selected group of patients was chosen for study and the follow-up interval was too brief for benign tumors with a very slow growth rate.
Charles H. Tator
Toronto, Ontario, Canada
Return to the Neurosurgery Home Page
