HSC Section 8_April 2017

the end of the operation in 95% (35) of cases following radiation and in 100% of all control subjects. In two pre- viously irradiated cases, the anatomical continuity of the facial nerve was lost while dissecting thinly splayed nerve fibers. The single patient with a nongrowing tumor who underwent salvage surgery for intractable trigeminal neuralgia experienced symptom resolution following surgery. Complications The rate of postoperative cerebrospinal fluid (CSF) leak requiring reoperation in radiated cases was 11% (n 5 4). One case occurred in a delayed fashion over 12 months following surgery and developed subsequent meningitis, requiring return to the operating room. Two additional cases of early postoperative CSF wound leaks from the craniotomy incision occurred and were success- fully managed with suture reinforcement of the skin clo- sure. One patient with prior radiation (3%) suffered brainstem stroke following salvage surgery, with resid- ual neurologic deficits including mild hemiparesis and ataxia, but currently lives independently, whereas none of the control subjects experienced postoperative stroke. Overall complication rates were not significantly differ- ent compared to matched controls (Table III). The median length of radiologic follow-up after sal- vage surgery was 26.4 months (range 3–114 months), which was not statistically significantly different than duration of follow-up for controls ( P 5 0.7). At last follow-up, no recurrences in cases of GTR, or progressive growth of tumor remnants in less than GTR resection cases, were diagnosed on imaging in either group. Facial nerve outcomes are summarized in Table IV and Figure 2. At last follow-up, good facial nerve function (HB I-II) was observed in 27 (73%) of the patients who underwent primary radiation therapy compared to 28 (76%) controls ( P 5 0.8). Notably, when only analyzing patients with normal preoperative facial nerve function, excluding the two study subjects with post-SRS facial nerve paresis, 27 of 35 (77%) retained good long-term postoperative facial nerve function following salvage surgery. After adjusting for age, tumor size, extent of resection, and duration of follow-up using logistic regression modeling, there was no difference in rate of satisfactory facial Facial Nerve Function and Tumor Control Following Salvage Surgery

TABLE III. Surgical Approach, Extent of Resection, and Complications Following Salvage Surgery of Previously Irradiated VS Compared to Matched Controls.

Primary RT n (%)

Matched Control n (%)

Feature

P Value

Surgical Approach Retrosigmoid

23 (62%)

30 (81%)

0.1

Translabyrinthine

14 (38%)

7 (19%)

Extent of Resection Gross total

18 (49%)

30 (81%)

0.01*

Near total

10 (27%)

4 (11%)

Subtotal

9 (24%)

3 (8%)

Complications Stroke

1 (3%)

0

1.0 1.0

Hydrocephalus

0

1 (3%)

CSF leak

4 (11%)

4 (11%)

1.0

Meningitis

2 (5%)

0

0.5

CSF 5 cerebrospinal fluid. RT 5 radiation therapy; VS 5 vestibular schwannoma.

underwent surgery 26 months after radiosurgery for treatment of medically intractable trigeminal neuralgia. The median time interval between primary treatment and salvage surgery was 36 months (range 9.6–153 months). The median tumor size at time of salvage sur- gery was 2.0 cm (range 0.6–4.1 cm), and the median amount of posttreatment tumor growth was 0.7 cm (range 0.3–1.8 cm). At the time of tumor progression, no patients had purely intracanalicular tumors; seven (19%) demon- strated primarily cystic tumor growth; two (5%) had asso- ciated brainstem edema; and 14 (38%) had varying degrees of brainstem compression on imaging, which was not different from controls. Table II summarizes the clini- cal characteristics of patients at the time of salvage sur- gery with comparison to matched controls. Patients presenting for salvage therapy had poorer hearing com- pared to matched controls ( P 5 0.001); otherwise, there were no differences between groups prior to microsurgery. Surgical Salvage Operative data are summarized in Table III. Previ- ously radiated patients underwent either a translabyrin- thine (14; 38%) or retrosigmoid (23; 62%) approach for surgical salvage. Eighteen (49%) patients underwent gross total, 10 (27%) near total, and nine (24%) subtotal resection. The most common indication for near- or sub- total resection was to preserve facial nerve integrity. In contrast, the majority of matched controls underwent GTR (30; 81%), a difference that was statistically signifi- cant ( P 5 0.01). Intraoperative facial nerve electroprognostic testing was performed using proximal minimum stimulation thresholds or the supramaximal stimulation technique described previously. 14 Facial nerve dissection was noted to be subjectively more difficult secondary to adherent and or poorly defined surgical planes at the nerve–tumor interface. The facial nerve was anatomically intact at

TABLE IV. Surgical outcomes. Postoperative House-Brackmann Score

Primary RT* n (%)

Matched Control n (%)

Feature

P Value

I–II

27 (73%)

28 (76%)

III–IV

8 (22%)

8 (22%)

0.8

V–VI

2 (5%)

1 (3%)

*Includes two patients who had preoperative facial nerve paresis fol- lowing prior radiation therapy. RT 5 radiotherapy.

Wise et al.: Surgical Salvage for Recurrent VS

186