Professional Documents
Culture Documents
OBJECTIVE: Our goal was to study the occurrence meningitis as the etiology of persistent postopera-
and source of origin of postcraniotomy headache tive headache. Prevention of postoperative head-
syndrome after removal of vestibular schwannoma ache may include the replacement of bone flap at
via the retrosigmoid approach. the end of surgery, duraplastic instead of direct
METHODS: A retrospective chart analysis was con- dural closure, and prevention of the use of fibrin
ducted of all patients with headache at 3 months glue or extensive drilling of the posterior aspect of
after removal of vestibular schwannoma from Jan- internal auditory canal. (Otolaryngol Head Neck
uary 1981 through March 1997 and with a minimum Surg 2003;128:387-395.)
of 24 months of follow-up. Diagnosis was made ac-
cording to the headache classification and was T he evolution of postoperative care for patients
graded using the HARNER scale. Recovery out- who have a vestibular schwannoma is increasingly
come was compared in selected groups of patients of interest. Before 1960, the possibility of removal
with and without headache. A descriptive statistical of the tumor from the patients was considered a
analysis was used to analyze differences between
success. Now it is recommended that the tumor be
groups.
completely removed without any facial nerve
RESULTS: Of the patients who underwent retrosig-
moid craniotomy for removal of vestibular schwan-
damage and that hearing be preserved whenever
nomas, 52 of 155 patients (34%) reported having possible. The surgical approach to the posterior
severe headache of requiring medication every fossa has changed greatly during the past decades.
day and/or feeling incapacitated 3 months after Different surgical approaches have been success-
surgery. Headache was more prevalent in those fully applied to enable access to the cerebellopon-
who had the bone flap replaced (94% versus 27%), tine angle for tumor removal: the suboccipital, the
if there was duraplastic or direct dura closure (0% middle fossa, the retrosigmoid, and the translaby-
versus 100%). Laboratory-proven aseptic meningi- rinthine approach.1,2 The discussion regarding
tis, most likely due to the use of fibrin glue and which approach may be better continues. The ret-
drilling of posterior aspect of the internal auditory rosigmoid approach to the posterior fossa is a
canal, was mainly associated with postoperative
modification of the traditional neurosurgical sub-
headache (81% versus 2%). In 75% of these cases,
occipital craniotomy. It represents the most versa-
calcifications along the brainstem had been noted.
CONCLUSION: The origin of postoperative head-
tile surgical approach for tumors of the cerebel-
aches after retrosigmoid vestibular schwannoma lopontine angle1,3-8 and can be used for any size
resections is not yet fully understood. Different fac- tumor, from intracanalicular to 40 mm or more
tors may play a role in preventing or reducing from the porus.8 In dealing with larger tumors, the
headache: dural adhesions to nuchal muscles or to skin incision may be lengthened to allow an infe-
subcutaneous tissues and dural tension in the case rior craniotomy, and the superoinferior diameter
of direct dural closure may explain postoperative of the craniotomy may be extended to 50 mm or
headache from dural tension. Intradural drilling and more.8 The wide exposure of the posterior fossa
the use of fibrin glue may be the source of aseptic gives excellent visualization of the tumor bed and
the cranial nerves ranging from nerves V above,
Reprint requests: B. Schaller, MD, Max-Planck-Institute for IX, X, and XI below.8 In addition, posterior fossa
Neurological Research, Gleuelerstrasse 50, D-50931 Co- vessels are well visualized.8 At the present time,
logne, Germany; e-mail, Bernhard.Schaller@pet.mpin- the mortality rate for tumor resection in the cer-
koeln.mpg.de. ebellopontine angle is very low.6-9 Main risks in-
Copyright © 2003 by the American Academy of Otolaryn-
gology–Head and Neck Surgery Foundation, Inc. clude complete loss of auditory function in the
0194-5998/2003/$30.00 ⫹ 0 affected ear, facial nerve paralysis, worsened bal-
doi:10.1067/mhn.2003.104 ance, other neurologic disabilities, infection, and
387
Otolaryngology–
Head and Neck Surgery
388 SCHALLER and BAUMANN March 2003
cerebrospinal fluid (CSF) leak.9,10 Recently, there sured by PTA (average hearing level at 250 and
has been considerable debate regarding the rate of 500 Hz, 1.2 and 4 kHz) and SDS (maximum
postoperative headache resulting from the ret- score) and classified according to Gardner and
rosigmoid approach compared with the translaby- Robertson15: class I, PTA with 0 to 30 dB and SDS
rinthine and other skull base approaches. Several with 70% to 100%; class II, PTA with 31 to 50 dB
authors indicate that the rate of headache is less and SDS with 69% to 50%; class III, PTA with 51
when the translabyrinthine approach is used.11-14 to 90 dB and SDS with 49% to 5%; class IV: PTA
However, the origin of these headaches and the with 91 to maximum loss and SDS with 4% to 1%;
difference in the incidence between the surgical and class V, PTA and SDS with no response.
approaches have not been well characterized and Facial nerve function was classified according to
remain controversial. Despite recent investigations the House and Brackmann grading system.16 Pa-
on the origin and management of postoperative tients were evaluated just before and 1 year after
headache, there remains a lack of complete under- surgery.
standing of their pathophysiology, how to prevent Detailed analysis of headache in all patients’
their occurrence, and how to treat them success- charts included principal location (eg, frontal, oc-
fully. In the present study, we evaluated retrospec- cipital), locality (eg, unilateral or bilateral, alter-
tively the clinical/neurologic factors of postoper- nating), projection (eg, arising from the neck pro-
ative headache persisting 3 months or longer after jecting to frontal), quality (eg, pulsatile, pressing,
resection of vestibular schwannomas via the ret- heightening), precipitation or aggravation (eg,
rosigmoid approach. physical or mental activity, tiredness), and possi-
ble additional symptoms (eg, nausea, sensitivity to
PATIENTS AND METHODS light or noise). In addition, patients’ charts were
Patient Population analyzed to establish whether patients had had
We retrospectively reviewed the case histo- headache before the surgery, and if so, details of
ries of 167 patients with the histologic diagnosis its type and frequency was determined. Further-
of vestibular schwannoma, which were operated more, patients were asked about medication pre-
on by an otoneurosurgical team from January scribed for their headache after surgery and
1981 to March 1997 at the University Hospital whether they were compliant. If headache was
Basel. There were 64 women (51%) and 61 men sufficiently documented in a patient’s chart (fre-
(49%). Their average age was 50 years (SD, 15 quency of at least twice a week), diagnosis was
years; range, 14 to 75 years). Preoperatively, all made according to the headache classification17
patients underwent vestibular testing (eye track- and was graded using the scale of Harner et al12:
ing test, optokinetic reflex, vestibulo-ocular re- grade 1, a relatively minor annoyance; grade 2,
flex), audiologic testing (pure tone audiogram headache present almost every day; grade 3, the
[PTA], speech discrimination score [SDS], patient required medication every day; and grade
brainstem auditory evoked potentials) and neu- 4, the patient felt incapacitated. All headache pa-
roradiologic examinations (computed tomogra- tients with grade 3 and 4 severity (n ⫽ 52) met the
phy [CT], magnetic resonance imaging [MRI], diagnostic criteria for the postcraniotomy head-
or digital subtraction arteriography of cerebral ache syndrome.18 Objectively verified macro-
vessels).6 scopic surgery-related intracranial mass lesion, ep-
Tumor size was measured by the largest diam- ileptic seizures, and bacterial infections or other
eter and categorized into small (⬍20 mm; n ⫽ 38; major complications of surgery were exclusion
30%), medium (20 to 30 mm; n ⫽ 41; 33%), or criteria in 3 cases. Headaches that were present
large (⬎30 mm; n ⫽ 38; 30%). A fourth group before surgery and persisted afterward, such as
consisted of tumors exclusively localized in the migraine headaches, tension headache, and postic-
internal auditory canal (IAC) (n ⫽ 8; 7%). The tal headaches, were not considered as new head-
average tumor diameter was 30 mm (SD, 18 mm; aches and were excluded from the present study
range, 5 to 70 mm). Hearing function was mea- (n ⫽ 9).
Otolaryngology–
Head and Neck Surgery
Volume 128 Number 3 SCHALLER and BAUMANN 389
No. of patients 52 35 10 9
Operating time, average (min) 407 389 365 372
Facial nerve intact 50 (96%) 34 (97%) 10 (100%) 9 (100%)
Cochlear nerve intact 32 (62%) 24 (69%) 8 (80%) 8 (89%)
Complete tumor removal 49 (94%) 34 (97%) 10 (100%) 9 (100%)
Cerebrospinal fluid fistula 3 (6%) 2 (6%) 1 (10%) 0
Aseptic meningitis 42 (81%) 34 (97%) 9 (90%) 8 (89%)
Craniectomy 49 (94%) 32 (91%) 8 (80%) 7 (78%)
Headache onset occurred typically in the imme- Table 2. No difference was noted based on the
diate postoperative period (90%), although 5 pa- skin incision being either trifurcate or sigmoid.
tients (10%) reported onset after 1 month. The There was a difference in postoperative headache
average headache rating was 7, with 10 represent- based on whether the bone flap was or was not
ing the most severe pain experience. The headache replaced (94% versus 27%). In none of the 25
was described as stereotyped by 41 of the 52 cases with duraplastic was there a persistent post-
patients (79%): beginning at the incision scar, operative headache. In all cases with postoperative
extending into the mid-occiput, and frequently headache, there was a direct dura closure, com-
into the posterior mid-partial area of the skull. pared with 76% in the subgroup without postop-
Five patients (10%) complained primarily of fron- erative headache (78 of 103 patients). The place-
tal pain. The other 6 (11%) thought that the pain ment or the duration of intermittent liquor
was “all over the head.” Headache was predomi- drainage showed no correlation to postoperative
nantly sensed as diffuse in the craniotomy group headache. Laboratory-proven aseptic meningitis
(2 of 3 patients; 67%), whereas patients in the (febrile meningeal inflammation characterized by
craniectomy group mostly located their headache CSF mononuclear pleocytosis, normal glucose,
on the operated side (36 of 49 patients; 73%). The mild elevation in protein, and an absence of bac-
quality of headache was predominantly pressing teria on examination and culture), most likely due
(47 of 52 patients; 90%). Of the 52 current head- to the administration of fibrin glue (in 96% of the
ache patients, none reported that their headaches cases) and drilling of the posterior aspect of IAC
interfered with usual activities, such as sports, (in 96% of the cases), was mainly associated with
work, and social functioning. Of those 36 patients postoperative headache (81% versus 2%) (Table
(69%) noted altered sleep patterns with early 3). However, speclet calcifications along the
morning awakening, 42 (81%) reported a negative brainstem had been noted in 75% of the cases with
impact on mood. Exacerbating factors included laboratory-proven aseptic meningitis.
head movements (83%), fatigue (58%), and exer- Furthermore, there was a great variability in
cise (40%). The follow-up of the headache at 3, 6, both prescribed and self-administered medication
12, and 24 months after surgery plus surgical data for posttraumatic headache, including the fre-
are listed (Table 2). There was no correlation quency with which they were applied (eg, only
between operating time, intact facial or cochlear now and these during each attack). Postoperative
nerve, complete tumor removal, or CSF fistula and analgesic medication during hospitalization did
improvement of headache over the months after not differ between craniotomy and craniectomy. In
surgery. It is interesting that 10 of 52 patients the follow-up period, postoperative consumption
(19%) with postoperative headache had a head- of analgesic medication was higher in the crani-
ache before surgery. However, there was a similar ectomy group (33 of 49 patients; 67%) compared
incidence in patients without postoperative head- with the craniotomy group (0 patients).
ache (23 of 103 patients; 22%). Most patients had a broad spectrum of psycho-
The incidence of headache after several varia- logic symptoms, frequently referred to as dys-
tions of the retrosigmoid approach is listed in phonic mood. Psychiatric assistance was given to
Otolaryngology–
Head and Neck Surgery
Volume 128 Number 3 SCHALLER and BAUMANN 391
Table 3. Relation of different skin incisions and different craniotomies to postoperative headache
Headache group* No headache group
Characteristic (n ⴝ 52) (n ⴝ 103)
2 patients with persistent postoperative headache knowledge of both could be helpful for indications
after 12 months; both had a long history of prob- of treatment.
lems preoperatively. One of these 2 had received
therapy for major depression (fulfilled the diag- Incidence
nostic criteria) and required further assistance. In published series of patients with vestibular
One patient had had major anxiety (fulfilled the schwannoma, the incidence of headache lasting
diagnostic criteria) preoperatively and required as- beyond the initial postoperative period ranges
sistance postoperatively. widely from 0% to 65% (Table 4). The reported
rates vary considerably depending on the popula-
DISCUSSION tion of the study, method of data collection, defi-
In the present series, we could show, for the first nition of headache, and duration of follow-up.
time, several different factors that influence the Moreover, preoperative and postoperative head-
occurrence of headache after the removal of ves- ache is not sufficiently documented in most of the
tibular schwannoma via the retrosigmoid ap- previous publications as they may not fulfill the
proach. There have been only a few reports in diagnostic criteria required for the correct diagno-
recent years5,11,12,14,18-23 that discussed possible or- sis of headache. In addition, most of the patients
igins of postoperative headache after use of the have a shorter follow-up than 24 months, as rec-
retrosigmoid approach. Postoperative headache is ommended by AAO-HNS guidelines. However,
not a consideration when the results of vestibular our literal review gives evidence that long-term
schwannoma surgery were evaluated, because it follow-up indicates a gradual improvement in the
does not prolong hospitalization or require addi- severity of headache in most of the patients over
tional hospitalization, both of which were the hall- months but generally gives no convincing reason
marks of surgical complications in most of the for the improvement. The described neurologic
studies. On the other hand, headache represents a symptoms are predominantly of the tension-type
subjective complaint that is difficult to quantify headache, although there may occasionally be
and to document. It can be present for many dif- other forms of headache. The problem of postop-
ferent reasons. The fact that headache exists after erative headache persisting over 3 months is
craniectomy does not necessarily signify a cause- mainly limited to the use of retrosigmoid-suboc-
effect relationship; the individual response to this cipital approach, as there is only a small number of
symptom also varies tremendously. Understanding patients with postoperative headache after the
the nature of postsurgical headache may help us to translabyrinthine and middle fossa approach for
understand its underlying pathophysiology, and the removal of vestibular schwannoma.
Otolaryngology–
Head and Neck Surgery
392 SCHALLER and BAUMANN March 2003
Table 4. Literature report of incidence of postoperative headache (HARNER grade 3 or 4) after removal
of vestibular schwannoma
Incidence (%)
Retrosigmoid-suboccipital approach
Schessel et al,11 1992 58 65 NA NA NA
Glasscock et al,5 1993 147 22 NA NA NA
Harner et al,12 1993 331 23 NA 16 9
Pedrosa et al,23 1994 135 73 NA NA NA
Soumekh et al,20 1996 56 13 NA NA NA
Catalano et al,19 1996 84 30 NA NA NA
Ruckenstein et al,14 1996 35 NA 26 17 NA
Koperer et al,22 1999 29 48 NA 48 NA
Wazen et al,21 2000 60 43 NA NA NA
Jackson et al,18 2000 183 NA 70 44 21
Present series 125 42 NA 13 7
Total 1243 36 63 24 12
Translabyrinthine approach
Parving et al,13 1992 273 14 NA NA NA
Schessel et al,11 1992 40 0 NA 0 0
Pedrosa et al,23 1994 15 53 NA NA NA
Ruckenstein et al,14 1996 18 NA 6 13 NA
Total 346 16 6 3 0
Middle fossa approach
Glasscock et al,5 1993 15 0 NA NA NA
Weber and Gantz,29 1996 49 1 NA NA NA
Total 64 1 NA NA NA
NA, Not available.
undergoing reoperation for tumor recurrence. This necessary after opening the dura. This allows bone
is in concordance with observations during neuro- dust to circulate within the intracranial cavity and
surgical procedures that mechanical or electrical to adhere to arachnoidal surfaces.27 The translaby-
stimulation of the dura mater or blood vessel of rinthine approach, alternatively, involves comple-
conscious humans leads to pain referred to the tion of the major portion of bone removal before
ophthalmic (first) division of the trigeminal opening the dura, and thus the potential for intra-
nerve.25 Similarly, stimulation of dural elements cranial contamination by bone particles is less-
of the posterior fossa whose innervation derives ened. Removal of large tumors by the retrosig-
largely from branches of the C2 root can also lead moid or translabyrinthine approach may block the
to pain in the anterior head.25 According to this dissemination of bone dust intracranially by con-
theory, traction on the dura with activation of the fining it to the operative site and allowing ready
neck muscles stimulates nociceptive fibers with access and removal.27 In addition, large tumors are
resultant pain in the occipital, vertex, frontal, and generally associated with more erosion of the IAC
retro-orbital regions. This is in accordance with with a resultant reduced amount of bone requiring
our observation that many of our patients noted removal and, again, a lessening in the volume of
that coughing or straining aggravated their head- potentially irritating bone dust. Positioning of the
ache. These mechanisms further support the the- patient may also contribute to the accumulation of
ory that increased intracranial pressure associated bone debris: patients in the supine position are
with these maneuvers may cause distention of the more likely to have bone dust collect intracranially
dura and resultant pain. However, the “nuchal- compared with those in the sitting position.27 It is
dural-adhesion theory” does not explain persistent conceivable that chronic low-grade aseptic men-
headache in patients who have translabyrinthine ingitis could be initiated by these bone particles.27
removal of tumor. In contrary to our observation In our patients with postoperative headache, CT
in the present series, some investigators have scans demonstrating calcification along the brain
failed to find any significant reduction in the inci- stem and gadolinium MR images showing dural
dence of postoperative headache after incorporat- enhancement inflammation seem to lend further
ing replacement of calavarina bone or cranioplasty support to this proposed mechanism. We could
in their retrosigmoid procedures.23 These findings further show a strong clinical correlation to these
suggest that some other pathophysiologic distur- radiologic signs and the laboratory-proven aseptic
bance is responsible for the long-term headache meningitis.
that occurs in many patients. Watertight dural closure is a standard compo-
As also seen in the present series, a few studies nent of the retrosigmoid approach. Despite efforts
have noted a seemingly paradoxical inverse rela- to protect the dura from the drying effects of the
tionship between tumor size and the occurrence of microscope light and ambient air, the dura usually
postoperative headache.12 Recently, an attractive contracts after opening, which requires the place-
hypothesis has been proposed to explain the gen- ment of some tension on it to close it tightly.28 In
erally lower incidence of headache associated with addition, the dura is usually sutured up to the bone
the translabyrinthine compared with the retrosig- edges during closure by this approach, which ex-
moid approach, the inverse relationship between erts additional tension on it. Persistent headache
the incidence of headache and tumor size, and the has not been common in patients undergoing re-
relatively uncommon occurrence of headache after section of vestibular schwannoma via the middle
retrosigmoid removal of posterior fossa tumors fossa or translabyrinthine approach, both of which
other than vestibular schwannoma. It has been do not require dural closure.5,14,29 It is possible that
proposed that bone dust trapped within the intra- the tension placed on the dura during closure is the
cranial cavity may cause a protracted inflamma- source of chronic pain. The lack of dural closure
tory response of the meninges, resulting in chronic during translabyrinthine procedures and the lack
headache.19,27 By using the retrosigmoid approach of dural opening in middle fossa approaches to the
to remove vestibular schwannomas, drilling of the IAC may prevent dural tension and resulting head-
posterior part of the internal auditory canal is ache. This hypothesis of dural tension as a source
Otolaryngology–
Head and Neck Surgery
394 SCHALLER and BAUMANN March 2003
REFERENCES 16. House JW, Brackmann DE. Facial nerve grading system.
1. Schaller B, Heilbronner R, Pfaltz CR, et al. Pre- and Otolaryngol Head Neck Surg 1985;93:267-8.
postoperative auditory and facial nerve function in cer- 17. Headache Classification Committee of the International
ebellopontine angle meningioma. Otolaryngol Head Headache Society. Classification and diagnostic criteria
Neck Surg 1995;112:228-34. for headache disorders, cranial neuralgies and facial pain.
2. Pfaltz CR, Gratzl O. The retrosigmoid approach in cer- Cephalagia 1998;8(suppl 7).
ebellopontine angle surgery: otological procedure and 18. Jackson CG, McGrew BM, Forest JA, et al. Comparison
outcome. In: Tos M, Thomsen J, editors. Proceedings of of postoperative headache after retrosigmoid approach:
the First International Conference on Acoustic Neuroma. vestibular nerve section versus vestibular schwannoma
New York; Kugler Publications: 1992. p. 423-7. resection. Am J Otol 2000;21:412-6.
3. Cohen NL, Lewis WS, Ransohoff J. Hearing preservation 19. Catalano PJ, Jacobowitz O, Post KD. Prevention of head-
in cerebellopontine angle tumors surgery. The NY expe- ache after retrosigmoid removal of acoustic tumors. Am J
rience 1974-1991. Am J Otol 1993;14:423-33. Otol 1996;17:904-8.
4. Giannotta SL, Pulec JL. Tumors of the cerebellopontine 20. Soumekh B, Levine SC, Haines SJ, et al. Retrospective
angle: combined management by neurological and onto- study of postcraniotomy headache in suboccipital ap-
logical surgeons. Clin Neurosurg 1988;34:457-66. proach: diagnosis and management. Am J Otol 1996;17:
5. Glasscock ME, Hays JW, Minor LB, et al. Preservation 617-9.
of hearing in surgery for acoustic neuromas. J Neurosurg 21. Wazen JJ, Sisti M, Lam SM. Cranioplasty in acoustic
1993;78:864-70. neuroma surgery. Laryngoscope 2000;110:1294-7.
6. Ebersold MJ, Harner SG, Beatty CW, et al. Current 22. Koperer H, Deinsberger W, Jödicke A, et al. Postopera-
results of the retrosigmoid approach to acoustic neuri- tive headache after the lateral suboccipital approach. Cra-
noma. J Neurosurg 1992;76:901-9. niotomy versus craniectomy. Minim Invas Neurosurg
7. Fischer G, Fischer C, Remond J. Hearing preservation in 1999;42:175-8.
acoustic neurinoma surgery. J Neurosurg 1992;76:910-7. 23. Pedrosa CA, Ahern DK, McKenna MJ, et al. Determi-
8. Cohen NL. Retrosigmoid approach for acoustic tumor nants and impact of headache after acoustic neuroma
removal. Otolaryngol Clin North Am 1992;25:295-310. surgery. Am J Otol 1994;15:793-7.
9. Schaller B, Probst RR, Gratzl O, et al. Different hearing 24. Vijayan N. Postoperative headache in acoustic neuroma.
preservation in surgery of vestibular schwannoma be- Headache 1995;35:98-100.
tween women and men. Acta Neurochir (Wien) 1996; 25. Wolff HG. Headache and other head pain. New York;
138:1275-81. Oxford University Press; 1963.
10. Schaller B, Probst R, Strebel S, et al. Trigemino-cardiac 26. Baldwin RL. Headache prevention in retrosigmoid ves-
reflex during surgery in the cerebellopontine angle. tibular nerve section. South Med J 1996;89:375-9.
J Neurosurg 1999;90:215-20. 27. Driscoll CL, Beatty CW. Pain after acoustic neuroma
11. Schessel DA, Nedzelski JM, Rowed D, et al. Headache surgery. Otolaryngol Clin North Am 1997;30:893-903.
and local discomfort following surgery of the cerebel- 28. Kaur A, Selwa L, Fromes G, et al. Persistent headache
lopontine angle. In: Tos M, Thomsen J, editors. Proceed- after supratentorial craniotomy. Neurosurgery 2000;47:
ings of the First International Conference on Acoustic 633-6.
Neuroma. New York; Kugler Publications; 1992. p. 899- 29. Weber PC, Gantz BJ. Results and complications from
904. acoustic neuroma excision via middle fossa approach.
12. Harner SG, Beatyy CW, Ebersold MJ. Headache after Am J Otol 1996;17:669-75.
acoustic neuroma excision. Am J Otol 1993;14:552-5. 30. Moskowitz MA, Brody M, Liu-Chen LY. In vitro release
13. Parving A, Tos M, Thomson J, et al. Some aspects of life of immunoreactive substance P from putative afferent
quality after surgery for acoustic neuroma. Arch Otolar- nerve endings in bovine pia arachnoid. Neuroscience
yngol Head Neck Surg 1992;118:1061-4. 1983;9:809-14.
14. Ruckenstein MJ, Harris JP, Cueva RA, et al. Pain sub- 31. Warner JS, Fenichel GM. Chronic posttraumatic head-
sequent to resection of acoustic neuromas via suboccip- ache often a myth? Neurology 1996;46:915-6.
ital and translabyrinthine approaches. Am J Otol 1996; 32. Radanov BP, Sturzenegger M, Schnidrig A, et al. Factors
17:620-4. influencing recovery from headache after common whip-
15. Gardner G, Robertson JH. Hearing preservation in uni- lash. BMJ 1993;307:652-5.
lateral acoustic neuroma surgery. Ann Otol Rhinol Lar- 33. Malis LI. Titanium mesh and acrylic cranioplasty. Neu-
yngol 1988;97:55-66. rosurgery 1989;25:351-5.