compromise, thus frequent testing of postoperative acuity and confrontation fields is necessary.

Immediate reexploration may be required.

2. CSF leakage can be sometimes difficult to detect, as postoperative nasal drainage is common. If

suspected, the best confirmatory test is to assay the collected fluid for tau protein (beta transferrin),

which is found in CSF but not in nasal secretions or blood. If positive, treatment options include the

insertion of a lumbar drain to lower CSF pressure for a few days and allow spontaneous closure of the

leak, but oftentimes the optimal course is reexploration and repacking of the tumor bed with closure

of the fistula.

3. Infection in the immediate postoperative period is rare. Prophylactic antibiotic coverage is

maintained for 24 hours, or until the nasal packing is removed. Delayed fever, especially with signs of

meningitis, requires evaluation with a lumbar puncture; an occult CSF leak should be suspected.

Patients may be at increased risk of sinus infection for many months.

4. Bloody nasal secretions are common, and usually clear with nasal irrigation. Frank epistaxis is rare,

usually occurs 2 or 3 weeks after surgery, and is most often caused by delayed hemorrhage from a

septal branch of the spheno-palatine artery. It can sometimes be managed with nasal packing, but

may require cauterization or clipping, or rarely embolization.

5. Endocrine management. Postoperative endocrine management is best done in conjunction with an

endocrinologist. Monitoring for pituitary insufficiency, especially hypoadrenalism, is necessary. Some

centers prophylactically treat with replacement steroids until adequate cortisol levels can be proven;

early discharge (on POD #1 or #2) may require outpatient steroid supplementation until testing can

be completed. Others monitor for adrenal insufficiency without replacement, though this is safest as

an inpatient. After successful surgery for Cushing disease, immediate and prolonged (6 to 12 months)

hypoadrenalism is expected if remission is achieved. Postoperative steroid replacement will be

required until normalization of the pituitary–adrenal axis is documented. Complete postoperative

hormone testing is usually performed at 6 to 8 weeks in order to establish the presence of remission

of hormone excess and reassess the patient’s pituitary function, including the possible development of

new pituitary hormone deficiencies (occurring in 5% to 10% of patients). Conversely, pre-existing

pituitary hormone deficiencies may resolve in 30% to 40% of patients postoperatively. Transient

diabetes insipidus is not uncommon (about 10% to 20% of cases), but permanent DI is unlikely

(approximately 1% to 2%). It is important to monitor urine output, urine-specific gravity, and serum

sodium levels in the immediate postoperative period, and treat with desmopressin (vasopressin

analog devoid of pressor activity) if needed. Mild cases can be treated with ad lib fluid intake if the

patient’s thirst mechanism is intake, as it usually resolves spontaneously within a few days. Sodium

levels should be monitored intermittently as an outpatient for 10 to 14 days, as SIADH can develop in

a delayed fashion, and a triphasic pattern (DI → SIADH → DI) has been described. Patients with

SIADH who are symptomatic from hyponatremia may present with headaches and nausea, and

neurologic dysfunction – confusion and rarely seizures – can occur if the hyponatremia is severe

(serum sodium <120 to 125 mEq/L). Mild SIADH usually responds to fluid restriction and salt

supplements; symptomatic patients may require 3% hypertonic saline. It is important not to correct

the hyponatremia too quickly, as central pontine myelinolysis can occur.

RESULTS AND COMPLICATIONS

Reviews of the literature on the outcome after transsphenoidal surgery require recognition of a number

of factors potentially affecting the interpretation of the results. Endocrine diagnostic techniques,

biochemical criteria for remission for secretory tumors, and surgical techniques have changed over the

years, which can make comparison difficult. There may be a publication bias favoring newer approaches

over old. Meta-analyses can be useful, but need to be interpreted with an eye to the underlying data,

and most of the meta-analyses review retrospective case series rather than randomized studies. A recent

meta-analysis reviewed short-term remission and recurrence data over the past 25 years and reported

overall short-term surgical remission rates of 61.7% in prolactinomas, 44.4% in nonfunctioning tumors,

60.9% in acromegaly, and 72.7% in Cushing disease, with no clear-cut improvement over the interval.

Because of the historical time frame, the majority of the series used the microscopic approach. The risk

of recurrence per patient year at risk was reported to be 3.4% for prolactinomas, 2.2% for

nonfunctioning tumors, 0.7% for acromegaly, and 2.3% for Cushing disease.29

Much of the recent literature describes the results with endoscopic approaches, often in comparison to

historical microscopic results; no major randomized comparison has been published. One recent review

2220

included nine pooled studies; they reported a 78% rate of gross tumor removal (GTR), 81% remission

rate in Cushing disease, 84% remission in acromegaly, and 82% in prolactinomas, with a 2% incidence

of CSF leak and 1% incidence of permanent DI. There were two deaths (0.24%), both from vascular

injury.30 This review was overweighted by a single large study which contributed 39% of the secretory

cases. An additional 12 endoscopic studies were included in a second meta-analysis, with an overall rate

of GTR 68%; 72% remission in acromegaly, 75% remission in Cushing disease, and 78% remission in

prolactinomas.31 Meta-analyses comparing the microscopic and endoscopic approaches now exist, often

using the same underlying data but reaching differing conclusions. Three of these are summarized in

Table 78-5; one of the three shows a resection benefit with endoscopy, while two do not. In addition, a

number of reviews have specifically analyzed complication rates between endoscopic and microscopic

techniques.32–34 Three of these reviews report that, in comparison to a retrospective analysis of

traditional sublabial approaches, the endoscopic approach appears to show decreased operating time,

length of hospital stay, risk of diabetes insipidus, nasal complications, and patient pain and discomfort.

The largest study incorporated 38 studies with greater than 10 patients; a total of 5,643 patients were

included.35 This study found no significant difference in the incidence of complications with the

exception of a statistically significant increased risk of vascular complications with endoscopic

procedures, perhaps due to the wider exposure possible within the endoscope. The overall incidence of

typical complications for both endoscopic and microscopic procedures is summarized in Table 78-6,

abstracted from this review.

Table 78-5 Comparison of Endoscopic and Microscopic Approaches

Table 78-6 Comparison of Complication Rates in Endoscopic Versus Microscopic

Surgery Based on a Meta-Analysis of 38 Studies35

Prolactinomas

Given the efficacy of dopamine agonist treatment, surgery for prolactinomas is rarely indicated.

Nonetheless, it remains effective in selected cases, especially in smaller tumors. A large surgical series

reported initial remission of 91.3% in microadenomas, with a recurrence rate of 7.1%. The overall

remission rate was 53.2% including giant prolactinomas.37

Acromegaly

Surgery remains the most effective treatment for acromegaly, despite the availability of medical

therapy. Multiple surgical series in the literature report remission rates on the order of 50% to 65% for

macroadenomas, and 70% to 90% for microadenomas (Table 78-7). Results from specialized centers are

2221