Abstract
In 50 adult necropsies, measurements of linear
dimensions and volume of the normal pituitary gland were
correlated with corresponding measurements of the sella turcica
obtained from radiographs. Though the female pituitary gland
was significantly larger than the male, there was no significant
sex difference in sellar size. Dimensions of gland and fossa
were correlated separately for each sex. Correlation between
gland volume and estimated sellar volume was good. The length
of the sella and the width of its floor correlated well with the
length and-width of the gland, but correlation between
measurements of sellar height and gland height was poor,
especially in women.
Detailed studies of the radiological anatomy of these
glands and fossae showed wide variation in the configuration of
the diaphragma sellae, particularly in the position of its central
portion relative to bony landmarks. In several a sub-diaphragmatic
extension of the subarachnoid space was found. The shape of
the gland varied markedly and was frequently distorted by
compression by the carotid arteries. Bony configuration was
also altered by these arteries which in some cases eroded the
lateral margins of the dorsum sellae, undercutting the posterior
clinoid processes. Some other variations in sellar contour are
illustrated and discussed.
To evaluate radiological criteria of abnormality in
the light of such wide normal variation, radiographs of the
pituitary fossa of 140 acromegalic patients were assessed. Over
half were grossly abnormal, showing marked bony erosion and
enlargement. In a quarter, a double sellar contour was present,
but, since dimensions of the inner contour were normal or only
slightly enlarged, abnormality was often overlooked. The
remainder (18%) presented particular diagnostic difficulty. On
plain films, commonly accepted criteria of abnormality
discriminated poorly between this group and controls.
Measurement of sellar dimensions, because of the wide normal
range, was also unhelpful. However, tomography of the sella
demonstrated significant abnormality in almost all and frequently
revealed downward extension of the pituitary tumour unsuspected
on plain films. Combined assessment by plain films and
tomography showed sellar abnormality in nearly 99% of the
140 patients.
To demonstrate suprasellar anatomy, tomography
in two planes was combined with pneumoencephalography in
37 patients suffering from various types of pituitary tumour.
Examinations were successful in all but one, showing a wide
variation in the position of the superior aspect of the tumour.
Suprasellar extensions were not predictable on the basis of bony
configuration of the sella and frequently failed to produce visual field defects. Subarachnoid extensions into the sella were also
demonstrated. Attention is drawn to the implications of these
variations and the importance of their demonstration before
therapeutic procedures on the pituitary gland are undertaken.
