Nistor, Raymond F.
To the Editor: I read with great interest the recent report by Knosp et al. (7) on the interesting and still controversial issue of magnetic resonance imaging (MRI) of parasellar pituitary adenoma. I was astonished by the fact that the authors ignored some basic studies and publications on this issue (3,5,7,9,10), even in their native language. These studies described some years ago the MRI patterns of parasellar tumor growth and the MRI-surgical classification of pituitary adenomas. There are several serious points of criticism, of which I think some have to be discussed here.
I think it is somehow hypothetical to try to suggest a classification of the parasellar development of pituitary adenomas based on a quite limited experience of 25 examined cases, of which more than the half measured 25 mm in diameter or larger. The prediction of intracavernous development becomes more complicated in smaller tumors. The statement that microadenomas were not encountered in the 25-case series should not exclude this type of adenoma from the “classification.” In a personal series (9) of 79 tumors that developed toward the cavernous sinus, 17 patients harbored microadenomas, of which 6 invaded the medial wall of the cavernous sinus.
It is well known that the rate of surgically detected invasion into the cavernous sinus is a quite subjective parameter, which is directly proportional with the surgeon's operative experience. The correlation of this variable parameter with the information provided by MRI seems to me dangerous, especially because neuroradiologists still are tending to overevaluate the prediction capability of MRI on this issue. The general experience of the most versatile authors (4,6,8) shows, even with high-resolution MRI techniques, that there still is no strong correlation between the prediction capability of invasion by MRI and the intraoperative findings of skilled, experienced pituitary surgeons. Because the authors analyzed that mainly large adenomas with generalized invasion tendency (Types 3 and 4) versus tumors (Type 2), which, after reasonable pituitary surgery criteria, should not be classified as being parasellar expansions, the value of their discussion becomes questionable.
On the other hand, using the carotid arteries (the intercarotid line) as landmarks for a classification system of parasellar developed pituitary adenomas is an arbitrary and extremely variable criterion. This imaginary line and its angulation related to the midline in a coronal projection shows a great variability because parasellar tumor growth is seldom as symmetrical as shown in the article. And even if this line would be theoretically constant, the main problem in predicting invasion is to be able to detect exactly the lateral confines of the tumor and the medial sinus cavernosus wall, which often is difficult even when using gadolinium diethylene-triamine-pentaacetic acid. The intracavernous bicarotid distance is still a better parameter for suspecting invasion, as compared with the relationship of the tumor/intercarotid line. Only the future developments in “superselective neurovascular MRI techniques” are promising a better preoperative prediction capability of MRI in parasellar pituitary adenomas.
Despite the classifications introduced by Hardy and Wilson, no exact differentiation of parasellar development of pituitary adenomas has been demonstrated, based on both surgical findings and modern imaging procedures. From this point of view, I agree perfectly with the authors that a new classification should be taken into consideration. In parasellar tumor extension toward the cavernous sinus, we should differentiate the following types: Type 1, tumor with localized lateral extension causing a circumscribed displacement of the medial wall of the cavernous sinus; Type 2, subcavernous extradural development (1), with the adenoma extending along the basal dura into the subcavernous area; Type 3, tumors causing generalized displacement and compression of the intracavernous structures; Type 4, tumors perforating the medial wall of the cavernous sinus with two varieties–localized invasion (4a) and with diffuse invasion (4b) of the intracavernous structures; Type 5, another type that deserves special attention is the supracavernous, supraclinoidal intradural extension (5a), with the tumor developed between the clinoid process and the disrupted, elevated dura.
If there is a disruption of the continuity of the lateral wall of the cavernous sinus, as frequently seen in the so-called “giant adenomas” with generalized invasion tendency (5b), there should be no doubt on the invasive growth. This classification takes into consideration the anatomomorphological aspects of tumor growth of pituitary adenomas, with special regard to the operative approach.
The infracavernous type described by Ahmadi et al. (1) is not, as stated in the article, an exceptional entity; as in larger series, it represents 18% of the noninvasive parasellar and 8% of all types of parasellar developed adenomas (1,9). This type of tumor development is clearly defined as being located extradurally, displacing the cavernous sinus above and laterally, and was not confused with tumor invasion into the inferior compartment of the cavernous sinus.
In suprasellar developed adenomas, the tumor may extend laterally, above the upper confines of the cavernous sinus. This is the supraclinoidal intradural parasellar type, with two different variants: 1) the growth between the cavernous sinus and the supraclinoidal segment of the ICA; or 2) between the internal carotid artery and the optic chiasm. This tumor type, which unfortunately is not described by the authors of the article, is also an invasive intradural tumor that is not necessarily invading the cavernous sinus, but also must be defined as being parasellar (and subtemporal). This type of tumor is of interest because transsphenoidal surgery is contraindicated in such cases (3).
In an analogy to Bonneville's dynamic computed tomographic scan of the sella region (2), demonstrating the involvement of the cavernous sinus by the obliteration of the veins of the carotid sulcus or of the medial group, an MRI is able to demonstrate in a simpler and more rapid manner, and without the application of contrast medium, the same phenomenon (void of paradoxical enhancement). As this pattern appears constantly in parasellar-developed adenomas, its presence should be evaluated as a pathognomonic magnetic resonance sign for the involvement of the cavernous sinus; although, it does not allow the differentiation between compressive and invasive development of the pituitary adenomas (9). The classical angiographic signs, which now are demonstrable because of the angiographic capability of MRI, as well as the clear-cut delineation of the relationship between the intra- and parasellar tumoral aspects, allow a fairly good preoperative distinction between invasion and displacement.
As a result of these imaging possibilities, the neurosurgeon has the advantage of more selective definite criteria for transsphenoidal and transcranial surgery. Based on this findings and supported by modern “superselective neurovascular MRI imaging techniques,” a new MRI-surgical classification of pituitary adenomas should be considered, not as a dogmatic result of the evaluation of 25 cases, but in consent and with the approval of most of the versatile pituitary surgeons.
1. Ahmadi J, North C, Segall HD, Zee CS, Weiss MH: Cavernous sinus invasion by pituitary adenomas. AJNR 6:893–898, 1985.
2. Bonneville JF, Cattin F, Bonchareb M, Potelon P, Boulard D, Racle A: The veins of the cavernous sinus: Usefulness of their demonstration for the diagnosis of invasive pituitary adenomas, in Landolt AM, Heitz PU, Zapf J, Girard J, Del Pozo E (eds): Advances in Pituitary Adenoma Research. Advances in the Biosciences, Oxford, Pergamon Press, 1988, vol 69, pp 151–153.
3. Buchfelder M, Fahlbusch R: Transsphenoidal surgery of pituitary adenomas developed towards the cavernous sinus, in Dolenc VV (ed): The Cavernous Sinus. A Multidisciplinary Approach to Vascular and Tumorous Lesions. Wien, Springer-Verlag, 1987, pp 404–414.
4. Daniels DL, Pech P, Mark L, Pojunas K, Williams AL, Haughton VM: Magnetic resonance imaging of the cavernous sinus. AJNR 6:187–191, 1985.
5. Fahlbusch R, Nistor R, Buchfelder M, Huk W: Magnetic resonance imaging (MRI) in the preoperative diagnosis of pituitary adenomas: A comparison with CT, in Landolt AM, Heitz PU, Zapf J, Girard J, Del Pozo E (eds): Advances in Pituitary Adenoma Research. Advances in the Biosciences. Oxford, Pergamon Press, 1987, vol 69, pp 117–126.
6. Huk WJ, Gademann G, Friedmann G: Magnetic Resonance Imaging of the Central Nervous System Diseases. Berlin, Springer-Verlag, 1989.
7. Knosp E, Steiner E, Kitz K, Matula CH: Parasellar classification of pituitary adenomas. Neurosurgery 33:610–618, 1993.
8. Kucharezyk W: The pituitary and sella turcica, in Brant-Zawadzki M, Norman D (eds): Magnetic Resonance Imaging of the Central Nervous System. New York, Raven Press, 1987, pp 187–208.
9. Nistor R, Fahlbusch R, Buchfelder M, Huk W: Magnetic resonance imaging of parasellar developed pituitary adenomas: New consequences for pituitary surgery, in Samii M (ed): Surgery of the Sellar Region and Paranasal Sinuses. Berlin, Springer-Verlag, 1991, pp 199–204.
10. Nistor R, Huk W, Fahlbusch R: Kernresonanztomographie von Hypophysenadenomen-Bedeutung für Diagnose und Therapie. München, Quintessenz-Verlags, 1990.