Surgery for hormone-secreting adenomas is rare in most European neurosurgical centers participating in the survey. Centers performing > 10 cases per hormonal axis were virtually nonexistent. Casanueva et al. discussed this in their article on the criteria for the definition of a Pituitary Tumor Center of Excellence (PTCOE) [5]. It is not possible to set a limit for each type of hormone-secreting adenoma. Still, the authors support the existence of high-volume centers. They recommend one pituitary center per 2.5-5 million inhabitants. For example, this would support the existence of 16–32 pituitary centers in Germany. Unfortunately, we were not able to cover all the centers performing pituitary surgery. Still, we received completed surveys from 60 centers in Germany. Even more striking data were derived from smaller countries. In all, 13 centers from the Czech Republic participated in the survey and performed pituitary surgery. In all 13 centers from the Czech Republic, participating in the survey, pituitary surgery is performed. The formula of 2.5-5 million Inhabitants per center would call for 2–4 centers in the Czech Republic.
Zamanipoor Najafabadi et al. performed a systematic review and meta-analysis on surgery as a viable alternative first-line treatment for patients with prolactinoma [6]. They included 55 articles on medical treatment (n = 3564 patients) and 25 on transsphenoidal surgery (n = 1836 patients). Long-term disease remission after dopamine agonist withdrawal was 34% and 67% after surgery. Subgroup analysis of microprolactinomas showed 36% disease remission after dopamine agonist withdrawal and 83% after surgery. They concluded that in most prolactinoma patients, disease remission could be achieved through surgery, with low risks of long-term surgical complications. In contrast, disease remission was less often achieved with dopamine agonists. This review may raise questions about whether medical treatment is always the first choice for patients with prolactinoma. In our survey, about 25% of the centers considered surgery as first-line treatment. More often, surgery was seen as a primary treatment modality in western centers than in northern centers.
Timing of pituitary apoplexy surgery in patients on medical therapy for prolactinoma was studied by Rutkowski et al. [7]. In their cohort, 13 patients underwent surgery within 72 hours of symptom onset and 19 underwent surgery > 72 hours after symptom onset. They concluded that the timing of surgical intervention relative to the onset of symptoms does not significantly affect the resolution of neurological or endocrinological deficits. In addition, early versus delayed resection did not significantly improve visual deficits, total visual loss, resolution of oculomotor palsy, recovery from hypopituitarism, or non-neuroendocrine signs and symptoms such as headache and encephalopathy.
In a study from India, Argawal and Mahapatra looked at the timing of pituitary apoplexy surgery [8]. The mean delay between apoplexy and neurosurgical consultation was 10 days (range 4–30 days). All patients showing improvement in vision had been operated on within a week of the apoplectic episode. In addition, the authors found that even completely blind eyes might have a remarkable improvement in vision if surgical decompression of the optic apparatus is undertaken early. In our survey, surgery was done within 48 hours of onset in 72.1% of centers.
Surprisingly, in 20% of the centers, medical pretreatment was often used in patients undergoing pituitary surgery for acromegaly. If this were the case, then first-generation somatostatin analogs were used most often. This finding is quite surprising given that current guidelines do not recommend routine pharmacological pretreatment for acromegaly [9]. Previously, several studies examined whether presurgical pharmacotherapy could improve endocrinological outcomes. Most of them failed to confirm that pretreatment greatly benefits patients (e.g., a better chance to achieve remission following surgery [10, 11]. Others found pharmacotherapy with somatostatin agonists (SSA) beneficial [12]. This could be given by selecting the patients with invasive adenomas or giant adenoma [13, 14]. Several studies have shown that treatment with SSA reduces tumor volume [15, 16], suggesting that such an approach might be applied in selected patients. Besides improved endocrine outcome, a positive impact on anesthesia management was considered a possible indication for presurgical pharmacotherapy with SSA. However, this was not confirmed by Losa et al. [11]. Still, presurgical treatment should be considered in selected cases where improvement in some particular complications of acromegaly (e.g., sleep apnea) is the goal. Besides first-generation SSA, another medication is used in the presurgical management of acromegaly. For the other agents (D2 agonist, pasireotide, pegvisomant), reliable data are scarce. However, when the goal is to improve the anesthetic outcome of patients by decreasing GH secretion/action before surgery, these agents can also be used. The impact of using this medication before surgery on long-term endocrine remission is also debatable. When pasireotide was used in presurgical management, no benefit was seen over immediate surgery [17].
Pretreatment with cortisol lowering medication in patients with Cushing´s is reasonable in severely decompensated individuals in whom improvement of metabolic or cardiac compensation is desired. It has its place when surgery cannot be performed soon and this seems to be a common reason among the centers participating in the survey. Treatment with pasireotide in this setting seems to be a reasonable option because of its dual effect on lowering ACTH and therefore cortisol secretion and its antitumor effect. According to several case reports, it has been successfully used [18]. Treatment with steroid lowering medication (metopirone or ketoconazole) is also widely used in the centers that participated in the survey. It has been confirmed that this management can have a positive impact on surgical outcome with a higher rate of endocrine remission in the group treated with steroid inhibitors [19].
Serum cortisol monitoring in patients with Cushing´s disease in the days following pituitary surgery to assess surgical outcomes has become a widely accepted practice. It helps predict the endocrinological outcome with a serum cortisol level of < 50 nmol/l as a strong predictor for the remission of the disease with an intermediate range from 50 to 140 nmol/l. However, a decrease in serum cortisol is often delayed. Therefore, repeated measurement on consecutive days with close monitoring of the patient’s status to detect clinical signs or symptoms of cortisol deficiency must be planned. In case cortisol remains high or inappropriately normal (e.g., > 140 nmol/l), early repeated surgery to increase chances of remission of the hypercortisolism should be considered [20–22]. Early reoperation is also less demanding than delayed surgery, which may be more complicated due to fibrosis and adhesions. Reversely, repeated surgery may be associated with a higher incidence of complications, such as permanent or transient diabetes insipidus [20] and permanent cortisol deficiency [21]. In our survey, half of the centers considered early reoperation if a sufficient cortisol level decline was not observed.
According to Friedman et al., using a dynamic technique with multiple coronal sequences after intravenous gadolinium injection allows high sensitivity and specificity [23]. A pituitary lesion can be found with this technique in 96% of patients with a biochemical diagnosis of ACTH-dependent Cushing’s disease. In comparison, in suspected Cushing’s disease, a pituitary lesion can be identified with a 50–60% higher diagnostic sensitivity rate than in non-dynamic MRI. Liu et al. studied a combination of dynamic enhanced MRI and high-dose dexamethasone suppression tests and bilateral inferior petrosal sinus sampling in Cushing’s disease [24]. Some 118 patients with Cushing’s syndrome were included. The positive predictive value of the combined pituitary dynamic MRI and high-dose dexamethasone suppression test was 98.6%, higher than that of catheterization. They concluded that patients with both positive findings in dynamic MRI and high-dose dexamethasone suppression test need no further invasive evaluation to establish a definitive diagnosis of Cushing’s disease. Furthermore, they support the application of bilateral inferior petrosal sinus sampling when negative findings are found in either dynamic MRI or a high-dose dexamethasone suppression test. Surprisingly, only half of the centers in our survey use dynamic MRI routinely in diagnostic algorithms for Cushing’s disease. A dynamic MRI is routinely performed more often in high-volume than in low-volume centers.
De Rotte et al. demonstrated that more lesions were detected at 7 T than 1.5 T MRI [25]. In five patients, both the 1.5 T and 7.0 TMRI enabled visualization of a lesion on the correct side of the pituitary gland. In three patients, 7.0 T MRI detected a lesion of the pituitary gland, whereas no lesion was visible at 1.5 T MRI. The authors also reported the magnetic susceptibility effect of air in the sphenoid sinus as a potentially disturbing artifact for pituitary gland imaging. Barisano et al. published a review article on clinical applications for 7 T MRI [26]. They speculated that 7 T MRI might one day become a routine diagnostic technique in patients with MRI negative Cushing’s disease, possibly improving surgical planning and outcomes. However, the current practice in Europe is different, with only two centers (0.8%) in our survey routinely used 7T diagnostic algorithm for Cushing’s disease.
A consensus statement on Cushing’s diagnosis by Arnaldi et al. concludes that if the results of clinical, biochemical and radiologic tests are equivocal or discordant, bilateral sampling of the inferior petrosal sinuses must be performed to confirm the presence of a secreting ACTH pituitary adenoma [27]. A narrative review on pituitary magnetic resonance imaging vs. bilateral inferior petrosal sinus sampling was recently published [28]. The authors are convinced that petrosal sinus sampling is an accurate and safe invasive diagnostic method in expert hands and plays an important role within the decisional algorithm for diagnosing and managing Cushing’s syndrome. The same finding was supported by a systematic review and meta-analysis by Wang et al. [29]. They showed that petrosal sinus sampling has a high diagnostic value for detecting source of ACTH over-secretion in patients with ACTH-dependent Cushing’s syndrome. This result contrasts with our findings: catheterization and petrosal sinus blood sampling are rarely or never used in almost 80% of European neurosurgical centers. Routine catheterization and petrosal sinus blood sampling were performed in 9.8% of high-volume centers and in 2.9% of low-volume centers. Petrosal sinus sampling was more frequently performed in southern than in eastern centers.
There is an ongoing debate on the need for routine perioperative corticoid substitution after pituitary adenoma surgery. For instance, Tothi et al. published a systematic review and meta-analysis on the need for perioperative steroid replacement therapy after pituitary adenoma surgery [30]. They analyzed 18 studies from 11 countries (n = 1224 patients) published between 1987 and 2013. The patients with morning serum cortisol levels of < 60 nmol/l at 3 days after operation were considered adrenal insufficiency and > 270 nmol/l as adrenal sufficient. They found that serum cortisol levels had significantly increased in patients after pituitary adenoma surgery. However, there was also significantly increased postoperative adrenal insufficiency and diabetes insipidus in the supplementation group but not in the no supplementation group. Therefore, they concluded that there is no necessity to receive routine cortisol replacement for patients with normal morning serum cortisol levels. Fridman-Bengtsson et al. evaluated different hydrocortisone treatment strategies in transsphenoidal pituitary surgery [31]. The authors compared three groups: high dose, intermediate dose and low dose hydrocortisone substitution. In the study, all patients received some degree of hydrocortisone substitution. The results indicated that low-dose hydrocortisone therapy favors a better endogenous cortisol production in the early postoperative phase.
De Tommasi et al. conducted a study on transsphenoidal surgery without steroid replacement in patients with preoperative morning serum cortisol < 250 nmol/l [32]. The results were compared to patients with a preoperative morning serum cortisol > 400 nmol/l and another set of patients with morning serum cortisol < 250 nmol/l who received intraoperative cortisol administration. None of the patients experienced a full syndrome of adrenal insufficiency. One patient with a preoperative morning serum cortisol < 250 mol/l had isolated postoperative fatigue and required cortisol replacement. No patient suffered any life-threatening complications. The authors concluded that pituitary adenoma surgery could be performed safely in patients with preoperative morning serum cortisol < 250 nmol/l in closely monitored settings without intraoperative cortisol administration.
Inder and Hunt published guidelines for perioperative assessment and management of glucocorticoid replacement in pituitary surgery [33]. They recommend that perioperative glucocorticoid substitution should not be considered for patients with intact hypothalamic-pituitary axis function and in whom selective adenomectomy is possible. Early postoperative assessment depends on daily clinical assessment of the patient and morning serum cortisol levels.
Wentworth et al. performed a prospective evaluation of a protocol for reduced glucocorticoid replacement in transsphenoidal pituitary adenomectomy [34]. A postoperative morning serum cortisol threshold of 250 nmol/l on days 1–3 was used to guide long-term glucocorticoid requirement in the prospective cohort. In two low-risk cases, long-term glucocorticoid replacement was required despite postoperative cortisol > 250 nmol/l. For the remaining 42 low-risk operations, glucocorticoid was not prescribed on hospital discharge based on a morning serum cortisol level of > 250 nmol/l and no clinical evidence of hypocortisolism. Thus, none of these 42 cases required glucocorticoid treatment for hypocorticolism following surgery.
All the studies mentioned above may call for selective hydrocortisone substitution in patients with no preoperative hypopituitarism and selective adenoma surgery. Still, in almost 60% of the centers participating in our survey, a routine dosage of hydrocortisone substitution was administered.
Study limitations and strengths
The main limitations of our study are related to study design. Every survey suffers from sampling bias due to nonresponse, i.e. nonresponse impacts the representativeness of the survey results, causing estimation bias. In addition, the quality of responses relies on the integrity of the respondents. We also failed to obtain a high response rate from some European countries (see results). Most results came from Germany, Italy, UK, Czech Republic and Spain. We achieved a large sample of data with a very high response rate and 100% completion rate. We want to emphasize that every respondent in our survey represents one neurosurgical center. So far, this study is the largest survey on pituitary adenoma surgery in Europe.