Transsphenoidal Surgery of Invasive Corticotroph Adenomas: Results in A Series of 86 Consecutive Patients

Objective: Surgery is rst-line treatment for corticotroph adenomas. Although most of corticotroph adenomas are noninvasive microadenomas that show expansive growth to surrounding tissues, a small subset of them is locally invasive and dicult to manage. The aim of this study was to evaluate surgical outcome of invasive corticotroph adenomas from a single-center. Patients and Methods: The clinical features and outcomes of CD patients who underwent transsphenoidal surgery (TSS) between January 2000 and September 2019 at Peking Union Medical College Hospital were collected from medical records. The clinical, endocrinological, radiological, histopathological, surgical outcomes and a minimum 12-month follow-up of 86 consecutive CD patients with invasive corticotroph adenomas were retrospectively reviewed. Results: Eighty-six patients with invasive corticotroph adenomas were included in the study. The average age at TSS was 37.7 years (range, 12 to 67 years), with a female-to-male ratio of 3.1:1 (65/21). The median duration of symptoms was 52.6 months (range, 1.0 to 264 months). The average of maximum diameter of tumor was 17.6 mm (range, 4.5–70 mm). All 86 patients with invasive corticotroph adenomas were performed TSS by microscopic or endoscopic approach. Gross-total resection was achieved in 63 patients (73.3%), subtotal resection in 18 (20.9%), and partial resection in 5 (5.8%). After surgery, the overall postoperative immediate remission rate was 48.8% (42/86), 51.2 % (44/86) of patients maintained persistent hypercortisolism. In 42 patients with initial remission, 16.7 % (7/42) of them experienced a recurrence. In these patients with persistent disease and recurrent CD, data about further treatment was available for 30 patients. The radiotherapy was used for 15 patients, and 4 (26.7%) of them achieved biochemical remission. Repeat TSS was performed in 5 patients, and none achieved remission. Medication was administrated


Introduction
Cushing disease (CD) is de ned as chronic hypercortisolemia, which is mainly caused by corticotroph adenomas. CD is a relatively rare with an incidence of estimated at 0.7-2.4 cases per million persons/year (1). If untreated or incompletely controlled, CD is associated with increased morbidity and higher mortality rate compared to general population (2). Transsphenoidal surgery (TSS) is considered the rst-line treatment for patients with CD. However, the remission rate after TSS has been reported to vary widely from 59 to 94% (3). Most of corticotroph adenomas are non-invasive microadenomas with tumor diameter less than 8 mm (4,5). The surgical remission rates in these patients with noninvasive visible microadenomas frequently exceed greater than 80% (6). However, surgery for invasive corticotroph adenomas has represented challenge, and the remission rates in patients with extensive invasion into parasellar regions, such as the cavernous sinus, the skull base, or intracranially have tended to be relatively lower. Because invasive corticotroph adenomas are relatively rare, there is still no comprehensive study on large-scale cases of invasive corticotroph adenomas. To date, surgical outcome and perioperative complications of invasive corticotroph adenomas have not been clearly reported (7).
The purpose of this study was to assess the surgical outcome and perioperative complications of patients with invasive corticotroph adenomas.

Patients And Methods
We conducted a retrospective analysis of consecutive patients with invasive corticotroph adenomas who underwent TSS at Peking Union Medical College Hospital (PUMCH) between January 2000 and September 2019. According to the guidelines of the PUMCH Medical Ethical Committee, clinical information on age, sex, clinical manifestation, preoperative biochemical parameters and imaging results, operative ndings, postoperative hormone levels and pathology, long-term outcomes and follow-up were collected for analysis.

Inclusion and exclusion criteria
The inclusion criteria for this study were: 1) The CD was diagnosed based on the patient's history and results of a physical examination, laboratory tests, magnetic resonance imaging (MRI), and surgical ndings; 2) a corticotroph adenomas was con rmed on immunohistopathological studies; 3) the invasion was recognized by intraoperatively ndings; 4) with complete medical records and a minimum 12-month follow-up. According to the Knosp grade classi cation (8), if tumors were classi ed as Knosp grade of 1 or 2 based on preoperative MRI results, but the invasion was con rmed intraoperatively, these patients were also included into this study. If the tumors were classi ed as Knosp grade of 3 or 4 according to the preoperative MRI results, but no invasion was recognized intraoperatively, these patients were excluded from the study. Patients were excluded if the surgery was aborted because of carotid artery bleeding, or surgery was undergoing via transcranial approach, or the medical records were not complete, or lack of pathological report or follow-up data.

Diagnosis of Corticotroph Adenomas
Most patients were diagnosed with invasive corticotroph adenomas by a pituitary multidisciplinary team (MDT) including members from the Departments of Neurosurgery, Endocrinology, Radiology and Pathology. As we described in previous study (9).All invasive corticotroph adenomas were diagnosed based on clinical manifestation, endocrine examinations, radiological examinations, intraoperative observation, and pathological results. In our center, all CD patients underwent comprehensive biochemical tests, including morning serum cortisol and ACTH, 24h-urine free cortisol (UFC), and a combined low-dose and high-dose dexamethasone suppression test (LDDST and HDDST). The methods of LDDST and HDDST has been described in detail in our previous study (10). If 24-h UFC and cortisol were not suppressed on the LDDST but was suppressed on the HDDST, CD was diagnosed. Based on our previous research results, the sensitivity of combined LDDST and HDDST to diagnose CD is 88.53% in our center (11).
Preoperatively, all patients underwent pituitary contrast-enhanced MRI and/or dynamic gadoliniumenhanced MRI. Images were independently reviewed by one neuroradiologist and at least one neurosurgeon, to evaluate the invasiveness of tumors, and ensure that agreement was achieved on Knosp grade which was descripted previously(8). Adenomas of Knosp Grade 0-2 were de ned as noninvasive PAs, and adenomas of Knosp Grade 3-4 were de ned as invasive PAs, respectively. According to the maximal diameter, adenomas were categorized as microadenoma (<1 cm), macroadenoma (<4 cm) and giant adenoma (≥4 cm).

Surgical procedures
All TSSs were performed in Department of Neurosurgery in PUMCH. Almost all the included patients were operated on by the same two experienced surgeons (Renzhi Wang and Ming Feng). The adenomas were removed by transsphenoidal microscopic, endoscopic approach or combination of them. The extended transsphenoidal approach assisted by multiple techniques was performed in some invasive macroadenomas or giant adenomas as previously described (12). In surgery, standard transsphenoidal PAs resection via microscopic approach or endoscopic approach was performed after broadly opening the sella oor. The tumor localized in the intrasellar and suprasellar region was removed rst, then the residual tumor in the CS was inspected. The medial wall of the CS was explored under direct vision, soft tumor tissue was removed using suction and ring curettes. After the location of the ICA using intraoperative mini-Doppler. The anterior wall of the CS was opened carefully, then the soft tumors located lateral to the ICA was removed by suction and ring curettes. If the tumor was exposed insu ciently, the ICA could be gently pushed to increase visualization for further exposure. If the tumor was too hard to removed, or tumor adhesion to the ICA or nerves rmly, leave the tumor to avoid injury ICA or nerves. The residual tumors will be addressed by postoperative radiotherapy. Finally, the sella was reconstructed in all patients to avoid postoperatively cerebrospinal uid (CSF) leakage and CNS infection.

Classi cations of intraoperative observed Invasion
The cavernous sinus invasion was determined by the experienced neurosurgeons during surgery based on degree of invasion of the adenoma into the CS wall, the presence or absence of adenomas within the CS. After tumor removal, if the medial wall of CS was intact and smooth, absence of invasion was deemed. If the medial wall was not intact, intracavernous ligaments, ICA or nerve bers were visible, the invasion was deemed.

Extent of tumor resection
The extent of tumor resection was evaluated through MRI examinations within 3 days and on 3 months after initial surgery. The gross total resection (GTR) was de ned as no residual tumor was identi ed, neartotal resection. The subtotal resection (STR) was de ned as more than 90% volume compared with the preoperative volume was removed, or when a de nite residual tumor volume less than 10% compared with the preoperative volume was identi ed. Partial resection was de ned as residual tumor volume greater than 10% (13).

Histological analysis
The resected specimens including pituitary adenomas were performed using routine H & E histological stains and immunohistochemical tests for ACTH. The diagnosis of CD was con rmed if pathology under light microscopy demonstrated an adenoma with immunohistological staining positive for ACTH.

Surgical outcome Measures
After TSS, the endocrinological tests were performed within the rst 7-day, 1 month, 3 months, and 12 months, respectively. As we described previously (14), immediate remission was de ned as a serum cortisol level <5 µg/dL, and/or 24hUFC level fell below 20 µg (56 nmol) within the first 7-day after TSS. Persistent disease was de ned in patients as elevated postoperative cortisol levels and a need for additional therapy. Recurrence was de ned in patients who achieved initial remission, followed by a rise in serum cortisol or 24hUFC and associated with clinical symptoms of CD. To assess the degree of tumor resection, postoperative MRI was done within 3 days and on 3 months after the surgery.

Statistical Analysis
Statistical signi cance for continuous variables was determined using the Student t test or analysis of variance. Categorical variables were analyzed using either the Pearson x 2 test or Fisher exact test. Pearson x 2 testing for determination of statistical signi cance was performed by using either a 2 · 2 contingency table or test for independence if greater than 2 outcomes were being analyzed. Fisher exact test was used if any expected values were less than or equal to 5. Statistical signi cance was de ned as a P value less than.05. All statistical analyses were completed by using SPSS software, version 22 (SPSS, Inc, Chicago, Illinois).

Demographic and clinical characteristics of included patients
From January 2000 to September 2019, 1381 patients with CD had TSS at the PUMCH, the intraoperative observed invasive corticotroph adenomas accounted for 7.5 % (104/1381) of overall patients. Of these, 86 invasive corticotroph adenomas patients with complete medical records, pathological con rmation and a minimum 12-month follow-up were included this study. The detailed information was listed in Table 1. There are 21 male patients and 65 female patients, and the median duration of symptoms was 52.6 months (range, 1.0 to 264 months). The average age at TSS was 37.7 years (range, 12 to 67 years), and the mean follow-up time was 38.0 months (range, 12.0-147 months). morning serum cortisol and ACTH were 721.4 µg (range 71.3-3705.6), 28.9 µg/dL (range7.4-75.0) and 142.5 ng/L (range 23.1-1250.0), respectively. In these cases, 66 patients underwent rst TSS, and 20 patients underwent repeated TSS, respectively. In 83.7% (72/86) patients, LDDST was not suppressed but HDDST was suppressed in the combined LDDST and HDDST test.

Prediction of intraoperative cavernous sinus invasion by preoperative MRI-based Knosp grade
The radiological results represent the rst essential step to evaluate invasiveness of PAs. However, the results of invasion evaluated using Knosp grade based on preoperative MRI are not always consistent with intraoperatively observed invasiveness. In this study, 1273 patients were identi ed as Knosp Grade 0, 1 and 2, and 108 patients were classi ed into Knosp Grade 3 and 4 respectively based on preoperative coronal MRI (   The remission rate in patient performed with GTR was signi cantly higher than that in the patients with STR and RR( all P<0.05).
TSS can be performed by microscopic or endoscopic approach, whether endoscopic approach yielded higher remission rate over the microscopic approach is still controversial. In this study, there were 31 patients underwent microscopic approach and 55 patients underwent endoscopic approach respectively. However, no signi cant difference in the remission rate was observed between the two techniques in patients with invasive corticotroph adenomas (P = 0.61).
To further analyze the effect of the number of operations on the surgical outcome of invasive corticotroph adenomas, patients were subclassi ed into two groups based on the status of rst or repeated TSS. In this study, 66 patients underwent rst TSS and 20 patients underwent repeated TSS respectively. In the patients who underwent repeated TSS, only 20.0% (4/20) of them achieved immediate remission, which is signi cantly lower than the remission rate (57.6%, 38/66) in the patients who underwent rst TSS (p = 0.003).

Perioperative complications
The surgery for invasive corticotroph adenomas usually was related to the high incidence of complications. In this series, the most common complication was intraoperatively CSF leakage, which occurred in in 24.4 % (21/86) patients ( Three patients showed visual deterioration after TSS, and two of them experienced various degrees of improvement of visual dysfunction. One patient who was undergoing repeated TSS experienced ICA rupture, the rupture was controlled by direct compression, and underwent digital subtraction angiography immediately. The ICA rupture was blocked by the stent, and the patient recovered nally without neurological de cits.   Further treatments for persistent and recurrent CD For those patients with persistent disease (44) and recurrent CD (7), data about further treatment were available for only 30 of 51 patients. Radiotherapy was used for 15 patients, and 4 (26.7%) of them achieved biochemical remission. Repeat TSS was performed in 5 patients, but none achieved remission. Medication was administered in 4 patients, and only one obtained disease control. Adrenalectomy was performed in 6 patients, and 5 (83.3%) achieved biochemical remission. At the last follow-up, 33.3 % (10 of 30) patients were in remission, and 20 patients still had persistent disease (Table 6).  (19). A recent Meta-Analysis also reported that the prevalence of CSI radiographically (43%) was much higher than that (18%) intraoperatively, and the radiologic criteria of Knosp 3-4 had the highest correlation with intraoperative CSI (20). Therefore, although MRI-based Knosp grade can reliably de ne the degrees of CSI in  larger tumors, it is often unreliable to de ne the absence of dural invasion in Knosp 0-2 microadenomas.
To de ne more accurately invasion beyond the lateral tangential line between ICA segments respectively, Knosp updated the original grading system of invasion in PAs by establishing the subtypes of grade 3a and 3b PAs in 2015 (15). However, up to 80% of CD patients present with a microadenoma, and there is no reliable grading system for microadenomas that predicts accurately CSI. Thus, more reliable grading systems of invasion for corticotroph adenomas are need.
For invasive corticotroph adenomas, complete surgical resection of tumor is di cult. Thus, lower remission rates of TSS have been reported in the patients with invasive corticotroph adenomas. However, 48.8 % of patients with invasive corticotroph adenomas in our center achieved remission after TSS, which was higher than that reported in previous studies (7,21). The main reasons for the higher remission rate may include the experienced neurosurgeon, intraoperative multiple techniques assistance and aggressive surgical procedure.
Surgery for invasive corticotroph adenomas has always been a challenge because of the highly complex anatomy of the CS and di cult in CS dissection. Thus, experienced neurosurgeon is essential to achieve complete tumor resection, biochemical remission, and avoid perioperative complications. In our center, almost all the surgery for CD patients were performed by Pro. Renzhi Wang and Ming Feng, who had experience with more than one thousand pituitary surgeries as previously described (22). Each year, more than one hundred of CD patients undergo pituitary surgery in our center, and a large part of them are invasive macroadenomas and recurrent CD (23). The number of surgical patients with CD may be one of the largest centers. Therefore, large-scale surgical patients with CD have accumulated rich surgical experience in management of patients with invasive corticotroph adenomas. Unfortunately, even in the hands of experienced surgeons, only about half of patients with invasive corticotroph adenomas could achieve remission after TSS. There are several studies also found that biochemical remission rate is related to the number of years of neurosurgical experience. Yap and colleagues reported that the rst decade of neurosurgery experience was associated with lower remission rates than that with the second and third decade of neurosurgery experience (24). A recent meta-analysis also demonstrated that the possible association of neurosurgeons' experience with remission rates in CD patients (25). Therefore, neurosurgical experience may be one of main reasons for the higher remission rate of TSS in patients with invasive corticotroph adenomas.
Application of multiple techniques assistance maybe another important factor for higher remission rate of TSS in patients with invasive corticotroph adenomas. In this study, multiple techniques including neuronavigation and intraoperative Doppler ultrasonography were used intraoperatively for surgical assistance in most patients, which resulted in maximum tumor removal and a relatively low rate of perioperative complications. For larger and recurrent invasive corticotroph adenomas, neuronavigation and Doppler ultrasonography was used to determine the exact location of the ICA. These techniques can provide references for locating some important structures including the ICA, brainstem, and optic canal in real time during surgery(26). Precisely location these structures can prevent to injuries them, thereby decrease the frequency of perioperative complications. T. J. Owen also reported that using the neuronavigation system for localization the rostral and caudal margins of the pituitary fossa during TSS may decrease morbidity and surgical time (27). Doppler ultrasonography also has been shown to determine the exact location of the ICA and whether it exist an aneurysm, thereby avoid to injury ICA the during surgery (28). Therefore, these multiple techniques assistance used in surgery facilitate tumor resection, the operation safe and decrease perioperative complications in patients with corticotroph adenomas.
In our center, aggressive procedure was used to pursue a maximum safe removal of tumor. During the surgery, in order to accurately assess the dural invasion and to remove the tumor maximally, it is critical to widely expose of the anterior and inferior sella dura, even the medial dural wall of the cavernous sinuses. Thus, endoscopic extended transsphenoidal surgery was used for most lager invasive corticotroph adenomas, which provide direct visualization for resection of the tumors invading the CS and suprasellar (28). Recently, the incision of the CS wall also has been performed for PAs invading the CS, which resulted in a higher GTR rate (12). Therefore, rich surgical experience, wide exposure by endoscopic extended transsphenoidal surgery and incision of the CS, and intraoperative assistance of combined neuronavigation and intraoperative Doppler ultrasonography are the main reasons for the higher remission rate in our center.
Identi cation of the factors affecting surgical outcomes is very is important for predicting the prognosis of patients with invasive corticotroph adenomas. In this study, we found that the remission rate (75%) in patients with Knosp grade 2 adenomas is higher than that in patients with Knosp  Other studies also demonstrated an inverse correlation between remission rates and tumor size in patients with CD (33,34). However, most of previous studies reported that the effect of tumor diameter on remission rate in patients with CD, but not patients with invasive corticotroph adenoma. Therefore, more studies on potential factors predicting surgical outcomes in patients with invasive corticotroph adenomas are needed.
In this study, we found that the repeated TSS for invasive corticotroph adenomas has been shown signi cant lower remission rates compared with the rst TSS. The main reasons for this result maybe that the destruction of the original anatomy and scar formation within the recurrent tumor. These factors make the surgery more di cult and dangerous. Valderrábano and colleagues (35) also demonstrated that the repeat TSS for CD is associated to a lower remission rate and a higher risk of recurrence, which is consistent to our results. However, the large-scale clinical study on results of repeat TSS for invasive corticotroph adenomas is limited, and further studies are needed.
Depending on neurosurgeon's preference, TSS for invasive corticotroph adenomas could be performed by microscopic or endoscopic approach. Compared with microscopic approach, the endoscopic approach provides a broader surgical view of pituitary region, including lateral edges of the sella and cavernous sinuses. However, in this study, no signi cant difference was found in the remission rates among the patients with invasive corticotroph adenomas who underwent microscopic TSS and endoscopic TSS. The microscopic and endoscopic techniques were used in combination for subset of patients with invasive corticotroph adenomas, which might explain why endoscopic versus microscopic technique yielded the similar remission rate in our center. This result in accordance with one recent meta-analysis, which indicated that comparisons of remission rates by endoscopic versus microscopic technique yielded the same results(36). However, another meta-analysis demonstrated that the endoscopic TSS reaches comparable results for microadenomas, and probably better results for macroadenomas than microscopic TSS for CD patients (37). To date, the data about comparisons of remission rates by endoscopic versus microscopic technique in invasive corticotroph adenomas is limited, more studies on comparison of the two techniques in patients with invasive corticotroph adenomas at the same institution are needed.
It is di cult to manage patients with persistent or recurrent invasive corticotroph adenomas. Therapy options include radiation therapy, repeated TSS, medical therapy, and as a nal step bilateral adrenalectomy. These treatments have their own advantages and disadvantages, however, there is no consensus on which treatment is preferable. In our center, an individual-based comprehensive treatment was discussed by a multidisciplinary team (MDT) with collaborating experts. However, even if comprehensive treatments were used, the prognosis of these patients is still poor, and more effective treatment are need.

Limitations
Although this study is a larger patient cohort on the results of TSS on invasive corticotroph adenomas in a single institution, there are some limitations that deserve mention. First, because of the retrospective nature of the study, it is di cult to collect completely clinical information and long-term follow-up data for all patients. Some data on long-term follow-up and further treatments were missing, which may result in a bias of the results and conclusions on long-term outcomes and recurrence rate. Additional, information of survival or death for half of the included patients cannot be obtained, thereby the mortality results are not included and analyzed. Another limitation is that referral or selection bias must be taken into consideration because this was a single-center cohort study.

Conclusions
In summary, the remission rate of CD patients with invasive corticotroph adenomas was unsatisfactory due to incomplete resection of invasive and/or a large adenoma. Extent of tumor resection and number of operations were associated with surgical remission rate in invasive corticotroph adenomas. In contrast, Knosp grade, tumor size and surgical approach did not affect the remission rate of TSS. In addition, further treatment including radiotherapy, repeated operation, medical therapy, and even bilateral adrenalectomy are required for persistent or recurrent invasive corticotroph adenomas. Declarations