Radiography of the Sella Turcica

Abstract

The present interest in endocrine disorders and the apparent increasing frequency with which the roentgenologist is called upon to assist in the diagnosis of these conditions have caused the roentgen demonstration and interpretation of changes in the sella turcica to become of considerable importance. The value, however, of the roentgen interpretation in these cases is dependent on a knowledge of the normal structure as revealed not only by roentgenologic studies on normal individuals, but on actual anatomic specimens as well. Unfortunately, the latter prerequisite was underestimated by early workers and as a result there have arisen certain conclusions which cannot be substantiated by anatomic facts. I refer particularly to the so-called “bridged sella,” which has been cited from time to time as a cause for diabetes insipidus, epilepsy, and migraine. This anatomic variation, which consists of a bony union between the anterior and posterior clinoid processes with occasional inclusion of the middle clinoid process, occurs in about 5 per cent of necropsy specimens. When the relation of the anterior clinoid processes to the pituitary fossa is considered and especially the fact that they always pass lateral to this fossa and never over it, the theory of a union between the anterior and posterior clinoid processes producing pressure on the hypophysis, infundibulum or associated blood vessels, seems quite improbable. Roentgenograms of this anatomic variation are quite misleading, as the shadow of the bony union which is in reality just lateral to the pituitary fossa is cast over the outlet of the sella and seems to enclose it and diminish its size. Aside from existing as an anatomic possibility, the appearance of a bridged sella may be produced on the roentgenogram by the superimposition of the shadows of the anterior and posterior clinoid processes. This may occur when the sagittal plane of the head is not placed parallel with the film, or when the tube is incorrectly centered. An incorrect lateral position not only distorts the images of the clinoid processes but the contour of the sella as well, and under such conditions the pituitary fossa may appear unusually small or shallow. These are not uncommon faults of roentgenograms of this region and many of the so-called bridged sellæ are not true anatomic variations but artifacts due to technical errors. The intimate relation of the sella turcica to the pituitary gland is no doubt responsible for the erroneous idea that all changes in its contour are the result of pituitary disease. It is true, with certain exceptions, that pituitary tumors produce changes in the contour rather characteristic in type, but there are so many other causes of sellar deformities that failure to bear them in mind may result in an erroneous diagnosis with subsequent disappointment for the surgeon and perhaps disaster for the patient.