House Clinic Perspective on Acoustic Neuroma Therapy

In the earlier part of the 20th century, many patients with vestibular schwannomas presented with debilitating neurological sequelae.(1) Olivecrona (1967) reported facial paresis at the time of tumor diagnosis in 54.7% of his patients.(2) This fact, combined with the emphasis placed on preservation of life, certainly explain Dandy’s (1941) statement. "Paralysis of the facial nerve must usually be accepted as a necessary sequel of the operation.".(3) Pennybacker and Cairns (1950), in their review of 130 cases of vestibular schwannoma, stated that "complete facial paralysis seems a small price to pay for relief from an acoustic tumour".(4)

Neural preservation in vestibular schwannoma surgery, like surgical mortality, has undergone tremendous evolution over the past century. With the introduction of the operating microscope and the revival of the middle cranial fossa and translabyrinthine approaches by William House in 1961, mortality rates dropped dramatically, and the focus of vestibular schwannoma removal shifted to the preservation of neural function. Patients undergoing vestibular schwannoma surgery today expect that they will survive surgery and are most concerned with preservation of facial function and, when appropriate, auditory function.

The total number of vestibular schwannomas removed at the House Clinic now approaches 5000. With experience and refinements of technique, the results have progressively improved. Following is an up-to-date review of our past and present series with regard to postoperative outcomes.

Middle Cranial Fossa Approach
Hearing preservation is the major reason we use this approach in appropriately selected patients. Our hearing preservation rates described in five series from 1989 to the present demonstrate preservation of hearing to within 15 dB speech reception threshold (SRT) and 15% word recognition score (WRS) in between 59% and 71% of operated patients.(5-8) Our most recent published series demonstrated 60% Class A or B hearing according to the reporting guidelines of the American Academy of Otolaryngology-Head and Neck Surgery and we have seen the same success in the management of NF2.(5,9) These results compare favorably with those of recently reported gamma knife immediate post-treatment hearing preservation results.(10) We recently reviewed the duration of hearing preservation postoperatively and found 89% of patients retained useful hearing with a minimum of 5 years of postoperative follow-up. This is in sharp contrast to the radiation therapy literature demonstrating significant hearing deterioration over time.

The middle cranial fossa approach has, in the past, been described as an approach placing the facial nerve at undo risk. This has not been our experience. Our most recent reviews demonstrate excellent postoperative facial nerve function (House-Brackmann Grade I or II) in over 94% of our patients.(7,8)

The middle cranial fossa approach provides an unobstructed view of the entire IAC and its contents. This allows complete tumor removal under direct vision. In a review of 500 cases with several years of follow-up, we identified only one case of residual tumor (0.2%).

We have not seen a permanent neurological complication following this approach. A single patient experienced a postoperative seizure that responded to Dilantin. After one year of seizure prophylaxis the Dilantin was stopped and no further seizures have occurred. We feel our low rates of complications are the result of short temporal lobe retraction times (average 90 minutes).

Translabyrinthine Approach A recent review from our database of vestibular schwannoma cases provides data from 1302 patients who underwent a translabyrinthine approach between 1982 and 1993. Their mean age was 50.0 years, and 46 per cent were male and 54 per cent were female. Tumor size varied from 0.5 to 6.5cm, with a mean size of 2.4cm. Operating time averaged 3.3 hours. Three (0.2 per cent) deaths occurred in this series.

Data on long-term (6 months) facial nerve function as determined by the House-Brackmann scale were available on 889 cases, with a mean follow-up time of 2.1 years. Of these, 58.2 per cent had a grade I function; 12.6 per cent, grade II; 13.2 per cent, grade III; 7.8 per cent, grade IV; 3.3 per cent, grade V; and 5.1 per cent, grade VI. In this same group of patients undergoing surgery since the advent of facial nerve monitoring (1988), 59% were grade I, 15.4% were grade II, 9.3% were grade III, 7.7% were grade IV, 4.2%, were grade V and 4.5%, grade VI. The vast majority of the poor facial nerve outcomes occurred in larger (greater than 4 cm) tumors. When comparing the risk of facial nerve paralysis in comparable tumors (

Postoperative Follow-Up
In our experience, vestibular schwannomas rarely recur after translabyrinthine removal. Our recurrence rate for unilateral tumors removed through the translabyrinthine approach treated between 1961 and 1995 was 0.3%.(11) The average interval to recurrence was ten years. Based upon these findings we have recommended a single gadolinium-enhanced MRI five years postoperatively.

Stereotactic Irradiation
Radiosurgery is a term originally popularized by Leksell referring to a procedure that delivers three-dimensional stereotactic external-beam irradiation to a specified tissue volume in a single session. This term is now applied to various techniques, including gamma-units using 60Co photons, protons, helium ions, and neutron beams and modified linear accelerator units (linac). Stereotactic external-beam irradiation differs from conventional external-beam radiation therapy in several important ways: 1) Treatment of small tissue volumes in the range of 1 to 30 cm3, 2) A single fraction of radiation is typically delivered. Some institutions are exploring fractionated therapy, but, for the treatment of vestibular schwannomas, long-term data with significant numbers of patients is lacking, and 3) High-dose gradients at field edges minimize dose deposition outside the target volume. Since its introduction in 1969, the gamma unit has been used to treat vestibular schwannomas based upon these principles. Initially utilizing high marginal doses and CT planning, the complications reported from experienced centers far exceeded those reported from microsurgical centers of excellence (12). With the improved technology of MR imaging for treatment planning, and the gradual empirical reduction in marginal dosing, the complication rates have reached levels comparable to those reported by microsurgical centers of excellence.

As discussed previously, the evolution of goals in the treatment of vestibular schwannomas has shifted from solely the preservation of life, to include the preservation of function (facial and auditory). What must be common to any form of intervention, should one be chosen, is the goal of life-long control.

The concern over the management of these benign tumors with stereotactic irradiation lies partially in this principle of life-long control. Studies have clearly demonstrated a dose response relationship with single fraction radiation. This relationship, at the current doses employed to the tumor margin, does not demonstrate complete cell kill. This fact is based upon the fundamentals of radiation induced cell death. Irradiated cells are most sensitive to injury during metaphase. Benign tumors have few cells in this phase of the cell cycle in comparison to malignant tumors. The ultimate result of irradiation is DNA damage secondary to the direct effects of the gamma rays, or strand breaks from oxygen radicals that are generated. There are also indirect tumor effects that result from radiation induced damage to nutrient vessels. The cells with damaged DNA either undergo apoptosis (programmed cell death) due to the load of DNA strand breaks, or, they undergo cell cycle arrest. The total and marginal doses currently prescribed have been done so empirically, with insufficient biological evidence that the cells will all undergo apoptosis. These cells will not remain in cell cycle arrest indefinitely. They will either enter an apoptotic pathway, or they will harbor radiation/repair-induced mutations that will allow them to re-enter the cell cycle and grow.

The time that it will take to see the fraction of tumors with cells re-entering the cell cycle can only be estimated as there is no long-term (20 to 30 years) follow-up. All of the irradiated vestibular schwannomas seen at our institution, and in the literature, demonstrate enhancement up to a decade after treatment suggesting a significant fraction of cells that have not undergone apoptosis.

As dosing has recently undergone significant reduction in an effort to reduce complications, a brief review of current data from experienced centers is provided. Flinkinger et al. recently provided data on one hundred ninety nine patients receiving gamma knife radiotherapy (median marginal dose of 13Gy) for their vestibular schwannomas with a median follow-up of 30 months.(13) The range of follow-up with the number of patients at each year was not provided. Additionally, the authors state that they had "more than one year MR follow-up" in 147 patients. Of those followed, to their knowledge, five patients (3.4%) have required microsurgical removal for continued tumor enlargement. In this same group of patients they identified MR evidence of 1 to 2 mm of tumor growth in nearly 10%.

In a recent series of patients treated with the gamma knife (marginal doses of approximately 13 Gy), Prasad et al. reviewed 95 patients with between 5 and 10 years of follow-up and found tumor growth in 6.3%.(14) When analyzing their complications, they noted that despite hearing preservation in many of the patients immediately after treatment, approximately 60% lost their preserved hearing over time. These figures support those of the Marsailles experience outlined in the preceding chapter. We also agree with the Marsailles group that cystic tumors should be considered for microsurgery due to the risk of post-irradiation complications.

We believe that all of these series lack sufficient follow-up. This is best illustrated by the experience with radiotherapy in another benign intracranial neoplasm, pituitary adenomas. In a study of radiotherapy for nonfunctional pituitary adenomas, Breen and co-authors demonstrated actuarial tumor control rates of 77.6±6.3% and 64.7±12.9% at 20 and 30 years, respectively (15). From these studies it is clear that to completely assess tumor "control" with radiotherapy it will require several decades of follow-up on these newer dosing strategies. As mentioned above, these irradiated tumor cells may re-enter the cell cycle due to induced mutations, potentially in genes regulating growth control resulting in neoplasia. Several authors, in contrast to the findings of Breen et al., state that radiation failures will occur early. This has not been our experience in the thirty cases of microsurgical management of radiation failures performed at our institution. We have seen several cases of failure beyond three to five years.

As described above radiation induced mutations can occur in other growth control genes leading to neoplasia. Studies in primates exposed to single dose radiation, similar to those given with gamma knife for vestibular schwannomas, demonstrated a 50% incidence of malignant brain tumor in the decade following irradiation.(16) In study of neoplasia in patients receiving radiation therapy for adeno-tonsillar hyperplasia, approximately one-third developed thyroid gland tumors and many were malignant.(17) In this same cohort of patients there were 66 neural tumors (2.2%), 67 salivary tumors (2.3%) and 93 carcinomas (3.1%) of multiple tissue types. In a follow-up study of 3013 of this same cohort receiving 7.5 Gy in three divided doses over three weeks, 85 neural tumors were identified in 70 patients (2.3%).(18) This included fifteen meningiomas, twenty-seven cranial nerve schwannomas and one malignant neural sheath tumor. The incidence of vestibular schwannoma in this group of patients was 0.7%, nearly 1000-fold greater than the normal population. One of the most striking features of this study was that new tumors were being identified more than 30 years after the initial treatment. This underscores the need for long-term follow-up not only for re-growth of the treated vestibular schwannoma, but, the appearance of radiation induced neoplasia. To date, there have been several reports in the literature of malignant tumors appearing after radiation therapy for vestibular schwannomas(19-22). Based upon the findings of Breen described earlier for radiated pituitary adenomas an actuarial rate for developing a radiation-induced neoplasm of 2.7±1.9% at 10 and 30 years, and the data sited above, we will likely be seeing similar results after vestibular schwannoma radiation therapy.

Conclusions
Based upon our current rates of hearing preservation, facial nerve functional preservation and cure (over 99%), we recommend microsurgery for those patients undergoing therapy for their vestibular schwannomas. We strongly recommend against radiation therapy for NF2 based upon our results and the unknown effects of irradiation in a condition known to result from a mutation in a growth control gene(19).

As no mode of therapy is without risk, we thoroughly discuss tumor observation with patients that have this as an option. Until studies demonstrating long-term tumor control (a minimum of three decades) with less than the expected rate of malignant transformation (approximately 3%) are available, we will only recommend radiation therapy for patients requiring treatment that are not surgical candidates either by personal choice or infirmity. Often, these are older patients.