gms | German Medical Science

The ultimate goal of the surgical management of patients with a vestibular schwannoma is radical tumour removal without any adverse sequelae for the patient. However, even in the hands of the most experienced surgeons, this goal is not always achieved, especially not in the larger tumours. The aim of this study was to present an update of the surgical results of the authors since 1982 with emphasis of some aspects regarding complete versus incomplete tumour removal and follow-up by MRI.

Data were, partly retrospectively and partly prospectively, collected from all patients with a vestibular schwannoma operated upon within the period 1982–2003. Three hundred-twenty-eight tumours in 317 patients were included. There were twelve tumours in 10 NF-2 patients. One-hundred-eighty patients were female and 137 were male. The mean age was 49 years, range: 10–86 years. There were 179 right-sided and 149 left-sided tumours. Tumour size was measured by the largest extra-meatal diameter in an axis perpendicular to the petrous bone. Compared to other series, in which often the largest diameter in any direction is measured, this may underestimate somewhat the tumour size in this series. Tumours smaller than 15 mm were classified as small, between 15 and 30 mm as medium and larger than 30 mm as large. There were 94 (29%) small, 136 (41%) medium and 98 (30%) large tumours. The retrosigmoid approach was used in all but three tumours (translabyrinthine). The amount of tumour removal was classified according to the subjective assessment by the surgeon. Near-total meant that a tiny piece of tumour capsule, not more than 1 or 2 mm, was left on the facial nerve. In a subtotal removal a small nodule was left behind. Partial meant everything else. Tumour removal was total in 209 (64%), near-total in 45 (14%), subtotal in 30 (9%) and partial in 44 (13%) patients. The House-Brackmann classification was used for assessing the facial nerve outcome and follow-up was at least one year. Hearing preservation was attempted only in patients with small tumours and class A or B hearing (AAO-HNS classification). Detailed MRI follow-up was available in 188 patients. A three-months MR scan was available in a limited number of cases.

In the early years, the yearly number of operated tumours gradually increased. Thereafter, it remained more or less constant, in spite of our policy shifting towards more observation and stereotactic radiotherapy or radiosurgery. When the period was divided in two halfs, the proportions of small, medium and large tumours were not significantly different between the first 167 tumours and last 161 tumours.

However, during the last three years the proportion of small tumours significantly decreased in “favour” of the larger tumours. This probably reflects the growing role of observation and radio-surgery.

The subjective assessment of the amount of tumour removal by the surgeon could be verified by the three-months MRI-scan. In 63 total removals, three-months MRI showed no nodular gadolinium enhancement in 46 patients, point-like enhancement in 13 patients and small nodular enhancement in four patients (4, 3, 5 and 4 mm). In the near-total removals (27), there was no nodular enhancement in half of the patients (13). In the other 14 patients nodular enhancement was observed, varying in size from point-like to 4 mm. In the subtotal removals (22), the residual tumour was not visible on the three-months MRI in two patients. In the other 20 patients the tumour remnants varied in size from 2 to 13 mm. On the three-months follow-up MRI of partial removals varying sizes of tumour remnants were observed.

One year follow-up of facial nerve functional outcome was available in 298 patients. Small tumours: 97% grade I and 3% grade II. Medium sized tumours: 73% grade I, 10% grade II, 11% grade III, 3% grade IV and 3% grade VI. Large tumours: 54% grade I, 1% grade II, 16% grade III, 5% grade IV, 5% grade V and 19% grade VI. Over-all results: 75% grade I, 6% grade II, 9% grade III, 3% grade IV, 1% grade V and 6% grade VI.

The outcome of facial nerve function was compared between two periods: early and late in the series. In order to exclude the learning curve, the first three years of tumour surgery were not included. Moreover, in the late period a second tandem of surgeons, ENT-man and neurosurgeon, joint the team. So, a learning curve, although less steep, is included in the second period. Most likely, this excludes the effect of the learning curves on these results. The results in the period 1986–1994 concerning 117 patients were: 65% grade I, 7% grade II, 15% grade III, 3% grade IV, 2% grade V and 8% grade VI. The results of the last 155 patients from 1995–2002 were: 84% grade I, 5% grade II, 5% grade III, 2% grade IV and 4% grade VI. The outcome significantly improved in the second period, a satisfactory result (grade I or II) increasing from 72% to 89%. What is the explanation for this? Tumour sizes did not significantly differ between the two periods. However, in the second period, the number of complete tumour removals significantly decreased from 85% to 56% and this should explain the better facial nerve outcome in that period.

Full audiological data were available in 207 patients. Considering all tumour sizes, class A or class B hearing was present in 75 patients (36%): small tumours 36, medium sized tumours 32 and large tumours 7 patients. According to the early experience of the authors and on the basis of data from the literature, hearing preservation was considered to be a realistic goal in small tumours only. Consequently, hearing preservation was attempted in patients with small tumours and a class A or class B hearing level only. Of 36 patients, 18 (50%) had their hearing preserved. These patients had a mean follow-up of 4.6 years (range: 6 months–14.8 years). During the follow-up period five patients lost their class A or B hearing, so that at last follow-up hearing preservation was only 36% (13 patients).

One hundred-sixteen patients with complete tumour removal were followed by MRI for more than three years (mean: 6.9 years; range 3–19.9 years). MRI showed a recurrence in four patients (3.4%): 3 mm after 7 years, 3 mm after 3.8 years, 5 mm after 5 years and 7 mm after 4 years. In the first two patients with a tumour recurrence of 3 mm, the lesion was of a triangular shape and this type of lesion has been described as scar tissue.

The question is then, why it appeared so late after the operation. Stable nodular enhancement was observed in 14 patients (12.1%). The question in these patients is, whether these lesions are stable recurrences or scar tissue.

Thirty-four patients with a near-total removal had a mean follow-up of 5.6 years (range: 11 months–15 years). Growth of the tumour remnant was observed in three patients (8.8%): 31 mm after 13.6 years, 10x 15 mm after 6.7 years and 7x 10 mm after 4.8 years. A stable tumour remnant was observed in 11 patients and no nodular enhancement at all in 16 patients. Regression of the tumour remnant was observed in 4 patients. Twenty patients with a subtotal removal had a mean follow-up of 4.5 years (range: 1.2–12.4 years). Tumour growth was observed in two patients (10%): 10 mm after 5.1 years and 28 mm after 7 years. A stable remnant was observed in 16 patients, no nodular enhancement in one patient and tumour regression in another patient. Eighteen patients with a partial removal had a mean follow-up by MRI of 6.4 years (11 months–17.5 years). Six patients (33%) showed tumour growth. Eleven patients showed no tumour growth and in one patient the tumour had regressed in size.

Altogether, in 188 patients, closely followed with MRI for more than three years, there were 15 (7.9%) either true or false recurrences. It should be noted, however, that most of these recurrences were radiological recurrences. Only four patients (2.1%) of the whole series (317) had a clinical recurrence requiring re-operation. Of the other, radiological recurrences, one patient was re-operated, eight patients were radiated and two patients were observed. Of patients with a remnant tumour, five patients were “up-front” treated by stereotactic radiotherapy, without further observation.

An incidental finding on MRI was the occlusion of the sigmoid sinus in three patients without clinical signs or symptoms.

The results of this series regarding the preservation of the facial and cochlear nerves are comparable with data from the literature. The facial nerve results compiled from ten series [Ref. 1], [Ref. 2], [Ref. 3], [Ref. 4], [Ref. 5], [Ref. 6], [Ref. 7], [Ref. 8], [Ref. 9], [Ref. 10] showed a good outcome (grade I or II) in 70% of 3540 patients, and in only 42% of 785 patients with large tumours. It is suggested that the better results of this series are due to the policy of leaving smaller or bigger tumour remnants behind if the facial nerve is at risk. The authors, therefore, believe that facial nerve preservation should have a higher priority than total tumour removal. It has been demonstrated by this study that most small tumour remnants, during the follow-up period, do not grow and, if they do, they can be treated by stereotactic radiotherapy or radiosurgery. For bigger tumour remnants (subtotal and partial removals) it may be advisable to treat them up-front by radiation without further observation. In view of the findings of this study, the issue of follow-up by MRI may be raised: how often and how long? Although the number of patients in whom hearing preservation was attempted is relatively small, the long-term results are poor and the question is whether radiation does better.

Finally, the authors realise that this is a somewhat “old-fashioned” series, which will not be extended in the same manner. This is because the authors have recently adopted a new management protocol with more emphasis on a policy of observation and radiation. This means, that the young generation of neurosurgeons and otologists will not be able to acquire the same surgical skills as those of the older generation. However, this makes the conclusions of this study all the more valid.