Endoscopic Endonasal Ethmoid-Pterygoid Transsphenoidal Approach to the Cavernous Sinus



Fig. 3.1
MRI after gadolinium. Different tumors should be approached with different approaches, considering their different biological features and paths of growth. In (a, b) (coronal and axial views), the coexistence of a meningioma of lateral wall of right CS and of a clivus chordoma is appreciable, also involving the CS. The meningioma was treated first through a transcranial approach, while the chordomas was removed in the second stage through an EPS endoscopic endonasal approach. In (c, d) (coronal and axial views), the satisfactory surgical results with radical removal of both tumors are demonstrated



Following this criterion, in our experience from 1998 to December 2013, 435 patients with tumors involving the CS have been operated through an endoscopic endonasal approach. Their histology is reported in Table 3.1. PTAs represent the more common histotype, due to the higher incidence of this lesion compared to chordomas, chondrosarcomas, or rarer tumors such as schwannomas, epidermoid cysts, hemangiomas, or inflammatory diseases, such as a case of CS aspergillosis. However, even for PTA, the CS involvement represents a not extremely common condition: indeed, in the same time, 1170 PTAs have been operated; thus CS involvement is present in 32 %. Of these, an EPS was performed in 79 (7 % of the entire series) cases; thus the real indication for this approach is very selected and reserved for tumors lateral to the ICA. On the contrary, for chordomas or chondrosarcomas, the involvement of CS is far more frequent, as in has been observed in 72 % of patients (49 out of 68). For these tumors, and especially for chondrosarcomas, which deriving from the spheno-petrous suture have a lateral to medial direction of CS involvement and thus are more frequently lateral to the ICA, EPS was adopted in 33 cases (48 %). Following the proper patient selection and tailoring the approach on the base of the tumor extension, we obtained a gross tumor removal in 66.3 %, a subtotal tumor removal (with remnant <20 %) in 28.1 %, and a partial tumor removal (with remnant >20 %) in 5.6 % cases of PTA. In the chordomas/chondrosarcomas series, the gross tumor removal was achieved in 51 % of cases, the subtotal tumor removal in 45 %, and partial tumor removal in 4 %. Postoperative definitive VI CN palsy occurred in 3 cases (6 %) after surgery for chordomas/chondrosarcomas, while in PTA series, two definitive (0.4 %) CN palsies were observed (in one case the VI CN and in one the III CN). One case of injury to ICA occurred on in a case of chordomas. Hemorrhage was controlled during surgery through gel foam and nasal packing and immediately after the vessels was closed via endovascular coiling. No neurological sequelae were observed.


Table 3.1
The series of endoscopic endonasal surgeries for CS lesions from 1998 to 2013 at Center of Pituitary and Endoscopic Skull Base Surgery is presented







































Pituitary adenomas

374

86 %

Chordomas/chondrosarcomas

49

11 %

Meningiomas

2

0.5 %

Neurinomas

3

0.75 %

Hemangioma

3

0.75 %

Epidermoid cyst

2

0.5 %

Inflammatory disease (Tolosa-Hunt S.)

2

0.5 %

Total

435

100 %



3.2 Neuroradiology


Magnetic resonance imaging (MRI) is the gold standard to investigate CS involvement [17]. It is preferable to use high-field MRI, such as the 3 T, to obtain adequate images with the contrast enhancement and topographic resolution needed due to small dimension of the structures within it, their deep locations, and uneasy anatomical relationship especially when distorted by a tumoral mass [17]. The MRI study has 3 main goals. The first is to assess a radiological diagnosis, to identify the nature of the tumor. This is crucial in the patient selection phase to evaluate what surgical approach is preferable, in particular according to the selection criteria for the endoscopic endonasal approach. Even if only the histological confirmation can give the definite diagnosis, multiple neuroradiological signs can orientate the preoperative suspect, giving a working hypothesis, basing on which elaborate the treatment strategy. The recognizing of the origin of the tumor can strongly suggest the nature of the tumor. PTA derives from the pituitary gland, for macroadenomas is common to observe an enlarged sella, while only some small remnants of the gland are recognizable pushed in the periphery of the tumor for their evident enhancement after contrast [18]. Chordomas or chondrosarcomas take origin from the clivus or from the spheno-petrosal junction and are associated to erosion of these bony structures. Chondosarcoma are usually brightly hyperintense at T2WI, while chordomas have a more mixed pattern of hypo- and hyperintensity after gadolinium [19]. These tumors, PTA, chordomas, and condrosarcomas, can engulf or totally encase the ICA, but the vessels are of normal caliber, compared to the contralateral one, without any stenosis due to the tumor compression [19]. Conversely, this sign is commonly associated to meningiomas or malignancies which tend to infiltrate and finally to close the vessel. Malignant tumors, such as adenocarcinomas or nasal carcinoma, usually derive from the surrounding paranasal sinuses and occupy the nasal fossa and can present a dural tail as well as meningiomas. These have a uniform and diffuse contrast enhancement and can take origin from the CS, which completely filled by the tumor with a “bombé” or convex aspect [19]. Some exophytic portions can be observed in the sella or other surrounding regions. On the other hand, meningiomas can infiltrate the CS, taking origin in other places, such as the middle fossa, the sellar diaphragm, the anterior or posterior clinoid, or the pterygopalatine fossa. It is important to remark that also nontumoral lesions can involve the CS, and these lesser frequent lesions should be taken in consideration to avoid potentially disastrous event [20, 21]. This is particularly true for intracavernous ICA aneurysm. In these cases the flow sign from the lesion and the relationship with vessels are quite peculiar, but a confirmation through a MRA, CTA, or angiography is mandatory [21]. Finally, inflammatory disorders, such as infections or other idiopathic forms, must be considered for atypical lesions, sometimes enhancing and presenting a dural tail [20]. These patients can present a medical history of immunodepression, hematological disorders, or chemotherapeutic treatments, and they can have alterations in blood examinations. The second goal of the MRI is to determine the involvement of CS, showing where in its wall the tumor overwhelms the dural layer to enter into it; how the vasculo-nervous structures, mainly ICA and III, IV, V1, and VI CNs, are displaced; and what compartments of CS are occupied by the tumor. This evaluation is crucial for the surgical plan, to analyze which corridor can give the more wide and direct access to the tumor, avoiding to cross nervous and vascular structures. To be sufficiently accurate and precise in these determinations, the study should include multiple sequences such as thin (2–3 mm) coronal and sagittal T1W1 with and without contrast administration and thin T2WI multiplanar images. Basing on the information provided by these examinations, it is possible for PTAs to classify the tumor in 4° according to Knosp classification (Table 3.2) [22]. We analyzed in our series of 314 PTAs the reliability of Knosp grade to predict the effective CS invasion (Table 3.3). We found that especially for grade 1 and 2 this classification did not allow reliable conclusions, and in 68 % of cases, only a compression without a medial wall interruption was observed at surgical inspection (Fig. 3.2). Conversely, in 5 sporadic cases with no apparent CS invasion and preoperative imaging according to Knosp classification (Knosp 0), a limited invasion of the medial compartment of CS was revealed. Also for chordomas or chondrosarcomas, the preoperative imaging can lead to inconsistent suspect of CS invasion, as showed in the Fig. 3.3. More innovative sequences, such as fast imaging employing steady-state acquisition (FIESTA) images, promise to assess with a higher accuracy than standard examination the tumor/CS relationship [23]. If the more wide and large experience will confirm these preliminary data, the role of these studies will become in the next years determining in the preoperative assessment of CS tumors. The distortion of the vasculo-nervous structures, and in particularly of the ICA, is crucial in the choice of the surgical approach. When the tumor pushes the artery anteriorly and laterally, mainly occupying the medial and posterosuperior compartments, a midline approach is more appropriate, permitting to approach the tumor from medial to lateral and reducing the chance of injury of the vessels. Conversely, if the ICA is pushed medially or is encased by the tumor, the EPS is preferred to work laterally to ICA with a direct orientation (Fig. 3.4). A few studies can be particularly helpful to evaluate the course of ICA, such as magnetic resonance angiography (MRA), and in particular sequences such as MRA + TOF with contrast, which shows in detail the relationship between tumor and ICA (Fig. 3.5). At present, it is not always evident to evaluate the course of the CNs, and in particular of the VI CN, which running free inside the CS is at major risk to be injured during the tumor removal. Some authors proposed diffusion tensor images (DTI) with tractography or 3D anisotropy contrast to identify with high precision the course of the CNs in the CS [24, 25]. Finally, the third goal of MRI consists in giving some information about the tumor consistency, to allow the surgeon to predict the outcome and assess its surgical strategy and even to plan preoperatively complementary treatments, such as radiosurgery, radiotherapy, or others. For lesions with a cystic component, the T2WI, marking the water content, can clearly identify this aspect, but to determine the consistency in solid tumors, these studies are lesser accurate [26]. Some authors propose to adopt the diffusion-weighted images (DWI), to predict the firmness of the tumors. Indeed, the apparent diffuse coefficient (ADC) is influenced by cellularity and extracellular fibrosis; thus a restricted diffusion is expression of a high reticulin content and thus a higher consistency of the tumor [27]. The efficacy of these analyses seems confirmed by recent studies, and some authors proposed a threshold of ADC value of 1.1 to predict a more favorable tumor removal (Fig. 3.6) [27]. As we have seen, the MRI is more relevant preoperative study of neuroimaging, but also the CT scan and for some extent the angiography still have a role in the preoperative assessments. The CT scan is very helpful to precisely locate the bony landmarks of the approach, and thus a study of the nasal and paranasal anatomy is highly recommended before surgery especially for patients with anatomical variants, complete or partial lack of sphenoidal sinus pneumatization, pediatric cases, re-interventions, where the anatomy can be distorted by the previous approach, and in general for all cases in the early phases of the learning course of each single surgeon. The knowledge of the location of the bony landmarks in the patients can be essential to orientate the surgeon and avoid complications. This can be favored by the adoption of neuronavigation with the preoperative CT scan in the approaching phase of the surgery. The neuronavigation can obtain even a major role, if combined with CT angiography (CTA), which permits to locate the course of ICA, reducing the risk of injuring during the approach and the CS opening (Fig. 3.7). Finally, the angiography can provide information similar to those obtained by MRA, to analyze the course of ICA. Moreover, it is the mail examination that can provide information about the patency of anastomotic circuits by the contralateral ICA or basilar artery through the compression test (Fig. 3.8). This is important in case of ICA injury or pseudoaneurysm to plan a closure through coiling of the vessel and to determine the risk of neurological sequelae.


Table 3.2
The Knosp classification is presented

























ICSI
 

Knosp 0

PTA does not reach the medial aspect of the ICA

Knosp 1

Invasion extending to, but not past, the intercarotid line

Knosp 2

Invasion extending to, but not past, the lateral aspect of the ICA

Knosp 3

Invasion past the lateral aspect of the ICA but not completely filling the CS

Knosp 4

Completely filling the CS both medial and lateral to the ICA



Table 3.3
The comparison between Knosp grade and effective CS invasion at surgical inspection is provided







































ICSI

0

1

2

3

4

Knosp 0

0

5

0

0

0

Knosp 1

87

8

2

0

0

Knosp 2

59

32

27

10

0

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

May 26, 2017 | Posted by in NEUROSURGERY | Comments Off on Endoscopic Endonasal Ethmoid-Pterygoid Transsphenoidal Approach to the Cavernous Sinus

Full access? Get Clinical Tree

Get Clinical Tree app for offline access