| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Departments of Neuroradiology (J.P.C., D.H.), Neurosurgery (C.D., M.J.), and Adult Radiology (L.B., P.B., L.M.), CHRU Bretonneau, 2 bis, Bd Tonnellé, 37044 Tours cedex, France; and the Laboratory of Anatomy, Faculté de Médecine, Tours, France (C.D.). From the 1998 RSNA scientific assembly. Received January 21, 1999; revision requested March 22; revision received July 15; accepted August 12. Address correspondence to J.P.C. (e-mail: cottier@med.univ-tours.fr).
 |
Abstract |
|---|
 TOP Abstract IntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: The MR images obtained in 106 patients (86 female, 20 male; age range, 16–71 years) were reviewed retrospectively by two physicians. The standard-of-reference criteria for invasion were the surgical findings. A 
2 analysis was performed, and the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for nine groups of MR imaging signs were computed.
RESULTS: Invasion of the cavernous sinus was certain (PPV, 100%) if the percentage of encasement of the internal carotid artery (ICA) by tumor was 67% or greater. It was highly probable if the carotid sulcus venous compartment was not depicted (PPV, 95%) or the line joining the lateral wall of the intracavernous and supracavernous ICAs was passed by the tumor (PPV, 85%). It was definitely not invaded (NPV, 100%) if the percentage of encasement of the intracavernous ICA was lower than 25% or the line joining the medial wall of the intracavernous and supracavernous ICAs was not passed by the tumor.
CONCLUSION: The radiologic diagnosis of cavernous sinus invasion by pituitary adenoma remains difficult, but the above-mentioned criteria may be of assistance.
Index terms: Cavernous sinus, 1767.38 • Cavernous sinus, MR, 1767.12141, 1767.12143 • Pituitary, MR, 145.12141, 145.12143 • Pituitary, neoplasms, 145.372
|
Introduction |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Unfortunately, neither computed tomography (CT) nor magnetic resonance (MR) imaging is accurate enough in the preoperative diagnosis of cavernous sinus invasion (1,8). Total encasement of the intracavernous ICA by the tumor is the most reliable MR imaging sign of cavernous sinus invasion, but this sign occurs very late (8). Often, the diagnosis of invasion remains uncertain until surgery because only direct observation can enable one to distinguish compression from real invasion of the cavernous space (9). Nevertheless, the prognosis, surgery-related morbidity, and treatment plan of pituitary tumors may be improved with better preoperative evaluation of cavernous sinus invasion and its extent.
The purpose of this study was to define the MR imaging criteria that provide reliable information on and definitive proof of the presence or absence of cavernous sinus invasion by pituitary adenoma.
|
MATERIALS AND METHODS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The MR examinations were performed with a 1-T MR imaging unit (Magnetom SP; Siemens, Erlangen, Germany). Before contrast material enhancement, we obtained a T1-weighted image through the sella turcica in the coronal plane. The imaging parameters usually were 500/15 (repetition time msec/echo time msec), three signals acquired, 220 x 256 acquisition matrix, 20-cm field of view, and 3-mm-thick sections with a 0.3-mm intersection gap. The acquisition time was 5 minutes 33 seconds for a standard T1-weighted imaging sequence. Postcontrast images were subsequently obtained in the coronal plane after injection of gadoterate meglumine (0.1 mmol/kg gadolinium; Dotarem; Guerbet, Roissy, France).
Each of the 212 images of the cavernous sinus was studied retrospectively. During MR imaging before gadolinium-based contrast material administration, the content of the cavernous sinus is isointense relative to that of the brain and interspersed with small foci of increased signal intensity that correspond to slow blood flow or fat. The intracavernous ICA is easily identified because of its characteristic thin walls surrounding a lumen of low signal intensity, which is reflective of a high-velocity flow void. After gadolinium-based contrast material administration, the venous compartments enhance strongly, and, thus, the depiction of these structures improves. In this study, the cavernous space was divided into five venous compartments with regard to the ICA (Fig 1): the medial compartment, which is between the ICA and the pituitary fossa; the superior compartment, which is above the ICA; the lateral compartment, which is the venous group lateral to the ICA; the carotid sulcus compartment, which is between the ICA and the carotid sulcus of the sphenoid bone; and the inferolateral compartment, which is between the carotid sulcus and the lateral venous compartment.
|
1. Total encasement of the intracavernous ICA—that is, the intracavernous ICA was completely surrounded by the tumor.
2. Displacement of the intracavernous ICA by the tumor.
3. Asymmetry of the right and left cavernous sinuses.
4. Nondepiction of the lateral, superior, inferolateral, and carotid sulcus venous compartments.
5. Lateral bulging of the lateral dural wall of the cavernous sinus (convex contour of the wall).
6. Nondepiction of the medial dural wall of the cavernous sinus.
7. Whether the tumor crossed one of the intercarotid lines (Fig 2) joining the lateral walls (lateral line), center (median line), and medial walls of the intracavernous and supracavernous portions of the ICA.
|
|
Each of the 106 patients was operated on by the same neurosurgeon (M.J.), who directly evaluated the cavernous sinus during the procedure. Each surgical report was retrospectively analyzed by three authors (J.P.C., C.D., L.M.) in consensus, without knowledge of the prior MR imaging analysis results. The standard-of-reference criterion for diagnosis was unequivocal evidence of invasion of the cavernous space, as described in the postoperative report by the neurosurgeon. Invasion corresponded to perforation of the medial wall of the cavernous sinus, with direct visualization of and/or contact with the intracavernous ICA and/or intracavernous trabeculae, both of which were surrounded by tumor. The diagnosis of pituitary adenoma was histologically confirmed in all the patients by using standard techniques and immunohistologic stains for prolactin, growth hormone, adrenocorticotropic hormone, and luteinizing hormone.
The comparison between the MR imaging and surgical findings was conducted by three of the authors (J.P.C., C.D., L.M.) in consensus, and statistical analysis of the results was performed (P.B.) by using the 2 test. Values less than .05 were considered to indicate statistical significance. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) also were computed.
|
RESULTS |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Involvement of the cavernous sinus was unilateral in all cases. Three of the 21 invasive adenomas were microadenomas (ie, <1 cm), and 18 were macroadenomas. Ten secreted prolactin, four secreted follicle-stimulating hormone, three secreted growth hormone, and four were nonsecreting adenomas.
Total removal of the tumor, including the cavernous space invasion, was performed during the operative procedure in 11 cases. In the 10 remaining cases, removal was incomplete, and radiosurgery (six cases) or suppressive drug therapy (four cases) was required.
The main results of the MR imaging examinations, as correlated with the surgical findings, are summarized Table 1. Total encasement of the intracavernous ICA (Fig 4) was observed in five cases of invasion of the cavernous space. The medial wall of the cavernous sinus was never depicted. The intracavernous ICA was significantly displaced laterally by the adenoma (P < .001). The two cavernous sinuses were of unequal size in 16 of the 21 invasions (P < .001). Depiction of the superior venous compartment was a very good sign of noninvasion (NPV, 96%; P < .001). Absent depiction of the carotid sulcus venous compartment had the highest PPV (95%, P < .001). Bulging of the lateral wall of the cavernous sinus was more frequent in invasive adenomas (P < .001). Crossing of the lateral intercarotid line was a very good sign of cavernous sinus invasion (PPV, 85%; NPV, 95%; P < .001). Therefore, in three of 20 cases with crossing of the lateral intercarotid line, there was no intraoperative evidence of invasion of the cavernous sinus (Fig 5). Although extension beyond the medial intercarotid line was observed in all the invading adenomas, it was not helpful for the diagnosis of cavernous sinus invasion because it had a low specificity and PPV (Fig 6).
|
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The anatomic location of the pituitary gland determines its ability to expand laterally toward the cavernous spaces. Because there are no lateral bone boundaries to the pituitary fossa, the expansions develop mainly in these lateral directions, and 6%–10% of pituitary adenomas involve the cavernous sinus (1,2). The pituitary gland is contained in a dural bag attached to the inferior aspect of the diaphragm of sella and surrounded by venous spaces that correspond laterally to the cavernous sinuses (11). The so-called medial wall of the cavernous sinus is the lateral part of this thin dural bag. It forms a convex lateral margin along the pituitary fossa on a coronal or transverse MR section and allows lateral physiologic expansions of the pituitary gland in about 29% of cases (11,12). Invasive disease has to be distinguished from lateral expansion without invasion. In the latter case, one can imagine the growing tumor forming or extending a lateral expansion of the dural bag around the intracavernous ICA similar to the phenomenon that occurs when an inflated balloon is compressed by a rigid tube. As long as the dural bag remains intact, no invasion of the cavernous sinus occurs. It only occurs when a defect appears on the bag.
The standard-of-reference criterion for this study was the intraoperative report of cavernous sinus invasion. Even this criterion may be criticized because it depends on the intraoperative findings of the neurosurgeon—that is, direct visualization of and/or contact with the intracavernous ICA, and the invasion or lack of invasion cannot always be reliably determined, even surgically. Nevertheless, this is the least imprecise criterion that can be applied to living subjects.
The preoperative radiologic diagnosis of invasion is of course more difficult and based on CT and MR imaging findings. MR imaging was proved to be superior to CT in depicting the normal and abnormal pituitary gland and sellar region (8,13). At CT and MR imaging, the medial dural wall cannot be identified. MR imaging is superior to CT in enabling the identification of the intracavernous ICAs and more clearly depicts the lateral dural wall (8). Analysis of the lateral venous spaces may be improved by using dynamic CT scanning (14).
During MR imaging, the absence of invasion can be assumed if a normal pituitary gland is present between the tumor and the intracavernous ICA or if a venous compartment is visible between the tumor and the intracavernous ICA (8,9). Ninety-four (44%) of the 212 cavernous sinuses that we studied did not meet these criteria, but only 21 (22%) of the 94 suspected cases were surgically proved to be invasions by adenomas.
On the other hand, total encasement of the intracavernous ICA is a very specific sign, but it is not very sensitive (8). Our study results confirmed this: Both the specificity and the PPV of total encasement of the intracavernous ICA were 100%, but this finding was present in only five of the 21 cases of invasive tumor, for a sensitivity of 24%.
Preoperative assessment of invasion of the cavernous sinus with MR imaging remains difficult unless the pituitary gland or medial venous compartment is visible or there is circumferential encasement of the intracavernous ICA. Alternative criteria have to be defined by using the limits of the cavernous sinus, its venous contents, and the intracavernous ICA. Table 4 summarizes these criteria.
|
The absence of asymmetry (NPV, 93%) or the absence of lateral bowing of the cavernous sinus (NPV, 91%) makes cavernous sinus invasion unlikely. Nevertheless, these signs, when present, are poor predictors of cavernous sinus invasion (PPVs, 80% and 41%, respectively) because they only reflect the mass effect of the tumor (13,15).
Venous Content of the Cavernous Sinus
Different venous compartments around the intracavernous ICA can be differentiated (12,14). The configuration of these venous compartments depends mainly on the loops of the carotid siphon and its position within the cavernous sinus space (9). In our study, the best radiologic sign of invasion with regard to the venous compartments was obliteration of the carotid sulcus venous compartment (PPV, 95%). This finding corresponds with those of Bonneville et al (14), who concluded that the absence of the carotid sulcus venous compartment was the best sign of cavernous sinus invasion at dynamic CT; these authors found an absent carotid sulcus vein at dynamic CT in the 12 patients who were suspected of having cavernous sinus invasion in their series. In that study, however, the specificity and sensitivity of this sign were unknown because surgical proof of invasion was available for only two of the 12 patients.
If a tumor expands laterally below the intracavernous ICA, the carotid sulcus venous compartment is the first to obliterate. In the present study, lateral expansion with tumoral tissue below the intracavernous ICA (ie, below-the-ICA pattern) was more often associated with invasion of the cavernous sinus (75%) than was expansion above the intracavernous ICA (ie, above-the-ICA pattern) (32%). Conversely, the depiction of this compartment was not a good sign for excluding cavernous sinus invasion (NPV, 50%).
Depiction of the other venous compartments—that is, the superior, lateral, and inferolateral venous groups—was helpful for excluding cavernous sinus invasion (NPVs, 96%, 89%, and 86%, respectively), but the lack of depiction of these compartments did not help (PPVs, 42%, 62%, and 50%, respectively). The absence of one of the venous cavernous compartments at MR imaging has low diagnostic value because it does not necessarily indicate invasion and may be caused by compression from the adenoma, or it may represent a normal anatomic variation in the course of the intracavernous ICA (9).
Intracavernous ICA
In the present study, the cavernous sinus was unlikely to be involved by tumor if the carotid siphon was not laterally shifted by the adenoma (NPV, 88%); conversely, a lateral shift of the ICA was not a very reliable sign of invasion (PPV, 41%).
Total encasement of the intracavernous ICA by tumor is a highly specific but insensitive sign that occurs late. Before total encasement manifests, the adenoma rolls around the intracavernous ICA (16). Thus, most of the tumors initially compress the cavernous sinus and stretch its thin medial wall. Later, infiltration of the medial wall and invasion of the cavernous sinus space may occur and progressively encase the intracavernous ICA. In this study, a percentage of encasement of the intracavernous ICA by adenoma greater than or equal to 67% (8/12) always corresponded to cavernous sinus invasion. Conversely, if the percentage of encasement was less than 25% (3/12), invasion never occurred.
We also used the intercarotid lines described by Knosp et al in 1993 (9) to assess for cavernous sinus invasion. Drawing these lines between the supracavernous and intracavernous ICAs to distinguish the low signal intensity of the ICA from the low signal intensity of the cerebrospinal fluid at T1-weighted imaging requires wide window settings. This analysis is often easier on nonenhanced images because there is better contrast with the window settings normally used, but contrast material–enhanced images are useful for distinguishing tumor from venous spaces and cranial nerves within the cavernous sinus. Moreover, the ICA is not always round on images; it can also be a curved line owing to its compression by the adenoma, and this makes it difficult to draw the lines.
In our study, tumor crossing the lateral intercarotid line was a very helpful sign for the diagnosis of cavernous sinus invasion, with a good PPV (85%) and NPV (95%). Our study results correspond to those of Knosp et al (9), who reported invasion of 12 of 14 cavernous sinus spaces, with tumor extending laterally to the lateral intercarotid tangent. The three false-positive results in our study can be explained by a normal lateral expansion of the pituitary gland, which is present in about 29% of cavernous sinuses (11,12 ). If such a normal glandular variant were present prior to the development of the tumor, the lateral intercarotid line would be crossed earlier, before the actual invasion of the cavernous space.
The absence of crossing of the median or medial intercarotid line excluded cavernous sinus invasion (NPVs, 98% and 100%, respectively), but the crossing of either of these lines was a poor predictor of invasion (PPVs, 54% and 28%, respectively).
In conclusion, the results of this retrospective study suggest that there are several helpful imaging signs for the preoperative MR-based diagnosis of cavernous sinus invasion by adenoma. A percentage of encasement of the intracavernous ICA by adenoma greater than or equal to 67% was the most specific sign; it occurred before the classic but late sign of total encasement. If the percentage of encasement of intracavernous ICA was less than 25% or the tumor did not cross the medial intercarotid line, the cavernous sinus was definitely not invaded. Aside from these two situations, cavernous sinus invasion was very likely if the carotid sulcus venous compartment was obliterated or the lateral intercarotid line was crossed. Conversely, if the median intercarotid line was not crossed, the superior venous compartment was visible, the cavernous sinus was of normal size, or there was no bulging of its lateral dural wall, then invasion of the cavernous sinus space could be excluded reliably.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Author contributions: Guarantors of integrity of entire study, J.P.C., M.J., L.M.; study concepts and design, J.P.C., L.M.; definition of intellectual content, J.P.C., C.D., L.M.; literature research, J.P.C., C.D., L.M.; clinical studies, L.B., J.P.C., C.D., D.H., L.M.; data acquisition, J.P.C., M.J., L.M.; data analysis, J.P.C., C.D., L.M.; statistical analysis, P.B.; manuscript preparation, J.P.C., C.D., L.M.; manuscript editing and review, J.P.C., C.D.
|
References |
|---|
|
TOPAbstractIntroductionMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. Delgrange, T. Daems, J. Verhelst, R. Abs, and D. Maiter Characterization of resistance to the prolactin-lowering effects of cabergoline in macroprolactinomas: a study in 122 patients Eur. J. Endocrinol., May 1, 2009; 160(5): 747 - 752. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A.A.K. A. Razek and M. Castillo Imaging Lesions of the Cavernous Sinus AJNR Am. J. Neuroradiol., March 1, 2009; 30(3): 444 - 452. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gurlek, N. Karavitaki, O. Ansorge, and J. A H Wass What are the markers of aggressiveness in prolactinomas? Changes in cell biology, extracellular matrix components, angiogenesis and genetics Eur. J. Endocrinol., February 1, 2007; 156(2): 143 - 153. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Levy, M. S. Matharu, K. Meeran, M. Powell, and P. J. Goadsby The clinical characteristics of headache in patients with pituitary tumours Brain, August 1, 2005; 128(8): 1921 - 1930. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Levy, H. R. Jager, M. Powell, M. S. Matharu, K. Meeran, and P. J. Goadsby Pituitary Volume and Headache: Size Is Not Everything Arch Neurol, May 1, 2004; 61(5): 721 - 725. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hagiwara, Y. Inoue, K. Wakasa, T. Haba, T. Tashiro, and T. Miyamoto Comparison of Growth Hormone-producing and Non-Growth Hormone-producing Pituitary Adenomas: Imaging Characteristics and Pathologic Correlation Radiology, August 1, 2003; 228(2): 533 - 538. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Nakasu, S. Nakasu, R. Ito, K.-i. Mitsuya, O. Fujimoto, and A. Saito Tentorial Enhancement on MR Images Is a Sign of Cavernous Sinus Involvement in Patients with Sellar Tumors AJNR Am. J. Neuroradiol., September 1, 2001; 22(8): 1528 - 1533. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
