48
The increased rate of hearing preservation over
time is likely due to a combination of technical
improvements as well as increased experience,
thanks to the growing number of smaller tumors
found on imaging [ 45 ]. Several authors have
found that preserved hearing in patients who
underwent middle fossa tumor resection deteri-
orates over time. One study published in 1990
found that 56% of patients followed for at least
3 years had a significant loss of hearing (mean
loss of 2% SDS and 12 dB SRT) [ 46 ]. In a
recently published series of 57 patients followed
for 5 years, 55% of patients had serviceable
hearing in the immediate postoperative period
and 75% of those maintained it at their 5-year
follow-up. We hypothesize that this improve-
ment is due to a change from packing the IAC
defect with temporalis muscle to abdominal fat,
which likely results in less fibrosis and pre-
served cochlear blood supply [ 47 ].
Facial Nerve Function
Many consider the preservation of facial nerve
function during acoustic neuroma resection as
second in importance only to tumor removal,
particularly when the patient does not have ser-
viceable hearing preoperatively. In our recent
series of 78 patients with acoustic neuroma
removed via middle fossa approach, 90% had
facial nerve function of House-Brackmann (HB)
grade I or II [ 44 ]. This is consistent with the
long-term outcomes published in previous stud-
ies [ 47 – 49 ]. Slattery et al. [ 50 ] reported 95% of
151 patients were HB grade I or II at year 1, and
Woodson et al. found that 96% of their 49
patients were HB grade I and 4% were HB grade
II at 2 years from surgery [ 51 ]. In a systematic
review including 35 studies of over 5,000
patients who underwent surgical removal of
acoustic neuromas in which facial nerve dys-
function was defined as HB grade III or higher
at last follow-up, facial nerve dysfunction was
seen in 16.7% in those treated with a middle
cranial fossa approach and 4% of those treated
with a retrosigmoid approach (P < 0.001) with
intracanalicular tumors. When tumors were
≤1.5 cm (but extended out of the IAC), facial
nerve dysfunction was seen in 3.3%, 7.2%, and
11.5% of the middle fossa, retrosigmoid, and
translabyrinthine groups, respectively. The mid-
dle fossa approach was associated with signifi-
cantly lower rates of facial nerve dysfunction
than the translabyrinthine approach (P < 0.001),
but no difference was seen between retrosig-
moid approach and the other two approaches.
When tumors were 1.5–3.0 cm, facial nerve
dysfunction was seen in 17.3% of those treated
with a middle cranial fossa approach, 6.1% of
those treated with a retrosigmoid approach, and
15.8% of those treated with a translabyrinthine
approach (P < 0.001). In the group with tumors
>3 cm in diameter, 30.2% of those who under-
went a retrosigmoid approach and 42.5% of
those who underwent translabyrinthine approach
had facial nerve dysfunction (P < 0.001) [ 19 ].
As one would expect, as tumor size increases, so
does the risk to the facial nerve during surgical
resection. In cases of large tumors (defined as
≥2.5 cm maximal or extrameatal cerebellopon-
tine angle diameter), subtotal resection can be
considered in order to increase the chance of
facial nerve preservation [ 52 , 53 ]. However,
such an approach must be weighed against the
risk of tumor recurrence [ 52 ]. In another sys-
tematic review evaluating facial nerve outcomes
following the resection of tumors ≥2.5 cm, good
facial nerve outcomes (HB grade I or II) were
seen in 62.5% of the 555 translabyrinthine
approaches and in 65.2% of the 601 retrosig-
moid approaches (P > 0.05). When broken down
by degree of resection, good facial nerve out-
comes were seen in 92.5% who underwent sub-
total resection (STR; n = 80), compared with
74.6% (n = 55) and 47.3% (n = 336) of those
who underwent near-total (NTR) and gross-total
resections (GTR), respectively (P < 0.001) [ 54 ].
In a recently published study prospectively eval-
uating 73 patients with large acoustic neuromas
with at least 1 year of follow-up, 14 (19%) cases
showed regrowth on MRI, 1 who underwent
GTR (8.3%), 2 who underwent NTR (9.1%),
and 11 who underwent STR (28.2%). This dif-
ference was statistically significant (P = 0.01)
[ 53 ].J.C. Sowder et al.