I N 1955 Chamlin, Davidoff and Feiring ~

Transcription

1 SYMPOSIUM ON PITUITARY TUMORS--II OPHTHALMOLOGIC CRITERIA IN DIAGNOSIS AND MANAGEMENT OF PITUITARY TUMORS* MAX CHAMLIN, M.D., AND LEO M. DAVIDOFF, M.D. Departme~ts of Ophthalmology and Neurosurgery, Albert Einstein College of Medicine, Yeshiva University, Bronx Municipal Hospital Center and the Mo~tefiore Hospital, New York, New York (Received for publication June 1, 1961) I N 1955 Chamlin, Davidoff and Feiring ~ made a study of the ophthalmologic changes in 156 cases of pituitary tumor. They found that visual-field defects (bitemporal hemianopia or some variant thereof) were found in 86 pcr cent of all patients with pituitary tmnors and craniopharyngiomas, while ii1 cases of chromophobe adenomas they were foumt in 96 per cent. Optic atrophy was present in 50 per cent and loss of visual acuity in 3~ per cent. Other ophthahnologic changes were of a relatively infrequent nature. For example, palsy of extraocular muscles was found in 5 per cent, and pupillary changes, involvement of 5th nerve, papilledema and proptosis, in even lesser numbers. Since the true existence of optic atrophy can be substantiated only by disclosing a visual-field defect, and since the loss of visual acuity depends on the loss of central vision, it may well be said that both optic atrophy and loss of visual acuity are functions of visual-fieht impairment. Visual-fieht loss, therefore, stands out as the most frequent and outstanding ophthahnologic feature of pituitary tumors, particularly chromophobe adenomas. Visual-field defects may indeed be found in the absence of other convineing neurologic or cndocrinoh)gie symptoms and signs or even roentgen evidence of pituitary tumors. In 1950, Charnlin and Davidoff 3 studied the importance of the 1/~000 field in cases of chiasmal interference. At thal time they * Presented in part at meeting of the IIarvey Cushing Society, Mexico City, Mexico, April ls, found that the 1/~000 field was the first to show bitemporal defect and the last to lose it after successful treatment, either surgical or roentgenologic. They also found that, whatever the peripheral defect was, the 1/~O0O field always showed at least as much involvement and usually much more. Finally, it was shown that the 1/~0OO field practically always showed some defect before central visual acuity was affected. In a review of the existing literature at that time, it was found that many of these seeming discoveries were only re-discoveries of facts that were already known to Clifford B. Walker in 191,5. In his classic paper entitled '% Contribution to the Study of Bitelnporal Hemianopsia he pointed out very nmch the same findings. At that time he urged the use of a wide range of test objects in order to pick out the extent of involvement of the various portions of the field, and benmaned the fact that most ophthahnologists were not using enough test objects, largely because of the time and effort involved. Today, virtually the same principles hohl true and most ophthahnologists still seem to be nluch too busy to do a complete set of visual fields with each of many test objects. Therefore, in 195r the present authors 4 described the use of a series of test objects with which to qualify a defective field, without necessarily repeating the entire visualfield study for each of these test objects. When a visual field was found to be defective for the usual small test object such as r

2 10 MAX CHAMLIN AND LEO M. DAVIDOFF FIG. 1. A. Use of multiple isopters. Peripheral fields for 5/330, 3/330, ~/330 and 1/330. Central fields for 5/~000, ~/~000 and 1/~000; in right eye also for 10/~000. B. Same fields utilizing only one isopter for the periphery (~/330) and one isoptcr for the central field (1/~000) and qualifying the defective areas of the field. (These fields do not represent a case study, but are hypothetical and used only to demonstrate the method.) white, or even 1/~000 white, instead of using a larger test object and again outlining the entire visual field, larger and larger test objects were displayed in the defective areas only until one was found that could be visualized in this area, and this was so recorded. While this may not show all the minute details of the quality in each meridian, for practical or clinical purposes, it gives as much information as is necessary for the proper control of cases of chiasmal interference (Fig. 1). When the peripheral fields, for example ~/830 white, and the central fields, for example l/e000 white, have thus been studied and the defective areas qualified, one may then, in most cases, be satisfied that one has an adequate evaluation of the entire visual field. However, in some cases of very minimal chiasmal interference, it is sometimes not sufficient to limit oneself to these two isopters. Thus, it has been found that if the isopter for 1/~000 does not extend beyond 10 ~ a temporal hemianopic defect, particularly above, may be picked up with ~/~000 white, a slightly more peripheral isopter. Also, while ~/330 may not disclose a defect, 1/330, a slightly less peripheral isopter, may do so (Figs. ~ and 3). In other words, the intermediate field somewhere between the overlapping limits for the tangent screen and the perimeter may disclose a minimal defect. The defect may be small enough so that minimal techniques such as simultaneous stimulation may be necessary to bring them out. 1 The full extent of the field as well as the quality of the defective field is important not only from a diagnostic point of view, but also as a means of following the cases clinically and evaluating the progress in regard to the response to treatment, either surgical or radiological. Thus, in a case of chiasmal interference such as is caused by chromophobe adenoma, in order to know whether surgery or more particularly roentgen-ray or cobalt therapy has produced a reasonable degree of improvement, visual-field studies are important because it is largely by the changes in the visual field that one can judge the effect of therapy. The use of colored test objects, while helpful to corroborate the presence of a defect or to qualify the depth of a defect in comparison to another area, is not as reliable for reproducibility and exact estimation of size of field, especially when such exact measure-

3 H'HTHA LMOI,t)t ~ l(' ASPE(~'I'S OF PITUITARY TUMORS 11 FIG. ~2. H.F. Fields of 1-~9-57 showing defect in left central field for ~/~000 white and not with 1/~000 white. Field of right peripheral fields showing defects for 1/880 white and not for ~/880 white; defect for ~/~000 white and not for 1/~000 white. ments arc necessary for the purpose of evaluating whether or not radiotherapy is helping in a particular case. FOLLOW-UP OF PATIENTS UNDER RADIATION THERAPY When a patient is receiving radiotherapy, it is important to follow the progress of the case by measuring the visual acuity and visual fields periodically. Since visual acuity is only an indication of the state of the very central portion of the visual field, one may then say that the study of the entire visual field assumes a major role in following cases of chiasinal interference under radiation therapy. When it has been decided by the neurosurgeon that a patient with a chromophobe adenoma is to receive radiotherapy, the patient is examined immediately prior to treatment, and visual acuity and visual fields are plotted with ~/330 white for the periphery and 1/~000 white for the central field. The defective areas are qualified as shown before and the patient is then sent for his first treatment. About ~4 to 48 hours after the first treatment, the patient is re-examined to see if there is any immediate effect. Occasionally improvement is noted after only a single treatment. If, however, there has been a suppression or loss in any of the areas, the radiotherapist is advised, and it is suggested that the next treatment be withheld until the lost fieht is regained, and when restarted the initial dose should be smaller than the previous one. During the course of treatment, which may last several weeks, the patient is examined at intervals of 3 to 7 days, depending on how severely visual function had been affected. If there is improvement or no remarkable change, treatment is carried on as planned. If, on the other hand, there is a definite loss at any time during treatment, it is advisable to wait until Fro. 5. A.S. Peripheral fields show upper temporal hemianopic defects for 1/330 and not for ~/380. Central fields show small defects for 1/~000 white but the defects also are definitely present for ~2/~2000 white and even more pronounced in right eye.

4 1~ MAX CHAMLIN AND LEO M. DAVIDOFF this loss has been regained before subjecting the patient to further treatment. It is not within the scope of this paper to discuss exact dosages or which patients should receive radiotherapy and which should be operated on. This decision belongs to the neurosurgeon and the radiotherapist after they have evaluated the case. However, once it has been decided that a patient is to receive radiotherapy, the routine as set forth above has been found most useful. If one does not see improvement during treatment, but if the patient is found to tolerate the treatment without loss, one may still be hopeful of finding improvement well after treatment has been completed. Thus, it has been shown that, in many cases, demonstrable improvement does not occur during or immediately after treatment, but within several months after treatment has been started or even a year or more after treatment has been completed. As a matter of fact, in a study of 18 cases in which the patient had responded particularly favorably to radiotherapy, it was found that most improvement did not begin until treatment had been completed and even then the improvement came on slowly over a period of about 6 months. In other cases a good deal of the improvement came within 1 year or even after ~ or 3 years following completion of treatment (Fig. 4). If, during the course of treatment, the visual field should become depressed and if, after withholding treatment for a while, this depression does not disappear or becomes even worse, this is usually enough indication for the neurosurgeon to operate rather than to wait for more improvement from radiation. On the other hand, if there is no improvement but vision remains stationary, and especially in older people when the visual loss has not been excessive and when life expectancy is not very long, one might well content himself if the visual status is controlled without further loss (Figs. 5, 6, 7 and 8). DIFFERENTIAL DIAGNOSIS AND PITFALLS IN THE VISUAL-FIELD DIAGNOSIS OF CHIASMAL INTERFERENCE Optic Neuritis. One of the common clinical pictures confused with chiasmal interference is that of optic neuritis in the form of retrobulbar neuritis. Thus if a patient loses vision in only one eye, and optic atrophy develops Fin. 4. Effect of radiotherapy in treatment of ehromophobe adenomas. Stippled areas represent time during which treatment was administered. Each white circle indicates an episode of improvement, while the small white lines indicate a time when examination was made but no improvement was found. An episode of improvement was gauged by a substantial gain in visual acuity or visual field. It is noted that 44 per cent of the episodes of improvement took place particularly within the first 6 months after radiotherapy had been administered, 37 per cent during the ~nd year after treatment had been started, and 14 per cent during the 3rd year after treatment had been started. There were even several episodes of improvement (5 per cent) during the 4th year.

5 ()PHTHAI,MOI,()GIC ASPECTS OF PITUITARY TUMORS 13 FIG. 5. J.A. Showing maintenance of fields 4 months, as well as 5 years after beginning of radiation therapy. FIG. 6. E.W. Showing progressive loss of field despite continued radiotherapy. Surgery had to be resorted to eventually, The tumor was very extensive.

6 14 MAX CHAMLIN AND LEO M. DAVIDOFF FIG. 7. G.G. Typical bitemporal hemianopia for 1/~2000 white with normal peripheral fields. At the end of 9 months there was improvement in quality of temporal paracentral fields. Six months after radiation central fields filled out; 189 years, as well as 3 years after radiation, fields were practically full. Fields still maintained now, 7 years after radiation. FIG. 8. J.C. Very little improvement even after 1 year following radiation. However, between the 1st and ~nd year extensive improvement both in left peripheral field and left central field.

7 OPHTHALMOLO(;I(" ASPECTS OF PITUITARY TUMORS 15 in that eye, a diagnosis of retrobulbar neuritis might well be made if careful visual fields are not plotted. Perhaps even more important than the visual field is a very carefully taken history. It is often found that if one really takes the time and trouble to exact a very detailed history from the patient, one is often rewarded with an illuminating set of facts that may well lead to a diagnosis of pituitary tumor even before the visual fields are plotted. Aside from the other symptoms and signs of disturbances of the pituitary gland, the mode of onset of loss of vision is significant. A very rapid or sudden loss of vision is more in favor of an inflammatory disease such as retrobulbar neuritis whereas with a pituitary tumor the onset is slow and insidious. However, one must guard very carefully against a sudden awareness of visual loss, as distinct from an actual sudden loss. In differentiating the loss of vision in pituitary disease from that in retrobulbar neuritis on the basis of visual fiehts, one must remember that in optic neuritis, including retrobulbar neuritis, visual-fieht loss is caused by a defect of a nerve-fiber bundle. Thus, if one is to make a diagnosis of retrobulbar neuritis, one should find a visual-field defect corresponding reasonably to an anatomic nerve-fiber bundle or bundles. Central visual acuity need not necessarily be involved in retrobulbar neuritis 2 or in chiasmal interference. One of the commonest types of visual-field defects in retrobulbar neuritis is the so-called cecocentral scotoma. Since the greatest part of the cecocentral scotoma lies between the area ceca or the blind spot and fixation, the area occupied by the scotoma is, in effect, a temporal defect. This is particularly true when there is little to no involvement nasal to fixation. Therefore, if a patient has had an attack of retrobulbar neuritis in one eye and then in the other eye, he is likely to have bilateral cecocentral scotomas. As pointed out above, these cecocentral scotomas occupy a largely temporal position in the visual field and therefore such defects may be mistaken for bitemporal hemianopia. The crux of the matter however is that the cecocentral scotoma of optic neuritis or retrobulbar neuritis, while it may lie mostly temporally and may be considered as a temporal defect, is not really "hemianopic." In other words the central limit of the scotoma may end around the fixation point but it does not end directly on and along the vertical midline of the field of vision. Nor is there any good anatomic or physiologic reason for such defect to end in the hemianopic midline. Such cases are illustrated in :Fig. 9. At this point it would be very comforting to say that cecocentral scotomas do not occur in chiasmal interference and this may be said to be generally true. However, during the past few years we have encountered 3 eases of central seotomas proved to be the result of chiaslnal interference. Two of these patients showed scant and unconvincing evidence of henlianopie features while the third showed a definite eeeoeentral seotoma in one eye only (Fig. 10). This experience is apparently shared by Walsh 8 who also claims to have seen 1 ease of unilateral central seotoma caused by a pituitary tumor. On the other hand, if a peripheral or central visual field shows true hemianopie features, bitemporal or some variant thereof, one may conclude that there is ehiasmal interferen ee. To carry this point further, once it is suspected that chiasm is involved, one should consider the lesion as mass (rather than inflammation) until proved otherwise. Enlarged Blind Spots. In some eases of increased intraeranial pressure, and particularly when there is or has been papilledema, the blind spots will be enlarged and when these blind spots enlarge enough they will encroach on the central area, thereby produeing a defect in the temporal (paraeentral) field. However, here again the defect is not hemianopic and should not be confused with the bitemporal hemianopia of ehiasmal interferenee. It is our impression that some eases of dilated 3rd ventricle that is said to press on the ehiasm and produce bitemporal defects may very well be showing defects caused by enlargement of the blind spots encroaching on the bitemporal fields but without true

8 16 MAX CHAMLIN AND LEO M. DAVIDOFF FiG. 9. B.P. Ceeocentral scotomas caused by retrobulbar neuritis. The temporal location of the scotomas with hardly any encroachment on nasal field gave character of "bitemporal defect." Note that 1/~000 field overlaps temporal area and shows no hemianopic features. C.S. Cecocentral scotomas resulting from past episodes of optic neuritis. Again the temporal location of the scotomas simulates "bitemporal defects," but 1/~000 fields are involved nasally and, therefore, the defects are not truly hemianopic. hemianopic features. The field defects, therefore, are not really evidence of chiasmal interference, but of long-standing elevated intracranial pressure. Developmental Anomalies of the Optic Disc. In 1946, Rueker ~ described 6 cases in which congenital anomalies of the optic discs caused depressed areas in the central fields, resem- FIa. 10. F.W. Right eye showed ceeocentral scotoma and peripheral defect but no hemianopic features. Eventually turned out to be caused by chromophobe adenoma and after some roentgen-ray treatment and even more so after surgery, true hemianopic features with bitemporal defect showed up. N.E. (Central fields extremely poor, visualizing 60/~000 red but without any hemianopie features.) Peripheral fields for ~/880 white show left upper temporal hemianopic defect and right lower nasal hemianopic defect. Patient turned out to have a chromophobe adenoma. M.G. Mass lesion involving chiasm. Peripheral defects showed no hemianopic features. Right central field showed a true cecocentral scotoma, but left central field, while extremely poor, did show a hemianopic feature with 40/~000 red.

9 t)phtitai~m()lo(,ic ASPE(;TS OF PITUITARY TUMORS 17 bling bitemporal hemianopia. These visualfield defects were easily detected with 1/~000 or 2/~000 white objects and were found mostly in the temporal field and above (Fig. 11). However, there was no sharp division at the vertical midline, and the defect gradually passed over to involve the nasal field as well. The peripheral field is normal in such cases. The optic discs usually are tilted downward and usually show some anomalous position of the optic cups. There may or may not be an inferior seleral crescent adjacent to the discs. In any event, there is some disturbanee, probably of the peri-papillary lower nerve-fiber bundles, to produce these defects. In these eases again one should hesitate lo make the diagnosis of chiasmal interference, particularly in the absence of a history supporting such disease. However, most important is the fact that the defects, while bitemporal, are not at all hemianopic and the lack of a hemianopic feature speaks very strongly against the interference being in the area of the chiasm. At this point it may be well to repeat that the contrary is true. In other words, if there is a defect in the visual field and if the visual-field defect shows strong evidence of hemianopic features, then one nmst very definitely suspect that there is chiasmal interference. The high incidence of mass as a cause of chiasmal interference has already been mentioned. Fro. 11. E.G. Patient diagnosed as having "bitemporal hemianopia." Note normal peripheral fields and definite temporal defects for 1/~2000 white, but lack of hemianopic features and involvement of nasal fields for 1/~000 as well. Patient had anomalous discs. N.D. Similar findings with bitemporal defect but lack of hemianopie features in central field. Case of anomalous discs. A.G. Also showed anomalous discs with oblique nerve entrances. Note defects in central field largely temporal and also upper nasal. E.B. Fields in 1951 showed temporal as well as upper nasal defect for 1/!000 white. Visual fields 10 years later showed diminution in over-all size of central field with more involvement of upper nasal area. Visual acuity and peripheral fields remained unchanged. Patient had distortion of discs in that the tapering shallow portion of the physiological cups were pointed downward rather than temporally and there was sonm inferior eonus showing below the discs.

10 18 MAX CHAMLIN AND LEO M. DAVIDOFF SUMMARY The finding of a bitemporal hemianopic defect is most important in the diagnosis of chiasmal tumors. The bitemporal type of defect is seen first in the intermediate field and usually is detected earliest with the 1/~000 test, and only later in the peripheral field. When improvement occurs, as by surgical or radiation treatment, the peripheral field returns to normal first but the 1/~000 field may continue to show defect indefinitely or the defect may disappear more gradually until complete improvement takes place. In plotting the fields it is not enough to indicate the defective area for a single test object but, once such defective area is found, this area must be qualified; that is, larger and larger test objects should be used until the smallest one just visible in the area is found and this information is then recorded. The use of such test objects to qualify the defective field is a very valuable clinical basis for comparison from time to time and obviates the need for a complete visual-field study with each of many test objects, thus avoiding mutually exhausting cfforts on the part of the examiner and the patient, the subject of Dr. Clifford Walker's lament. In the differential diagnosis of chiasmal interference one must be particularly alert to the occurrence of retrobulbar neuritis which may simulate chiasmal interference or vice versa. A cecocentral scotoma which may also occupy a "bitemporal" position is not truly hemianopic and must be recognized and differentiated from true hemianopic defects of chiasmal interference. Enlargement of the blind spots will also encroach on the bitemporal areas but again do not have truly hemianopic features and should, therefore, be differentiated rather easily from true chiasmal interference. Anomalous optic-disc entrances may cause distortions in the optic nerve and its fibers, thereby resulting in field defects which may be largely bitemporai. However, these defects are not only not truly hemianopic but actually reach over and involve the upper nasal paracentral fields as well. When it has been decided that a case is truly that of chiasmal interference and radiotherapy is to be given, the ophthalmologist should establish the visual status of the patient in terms of visual acuity and visual field, including qualifying of the defective areas. As radiotherapy is given the ophthalmologist should follow the progress of the patient and report his findings to the neurosurgeon and radiotherapist who use these findings as indications for continuing treatment, diminishing dosage or frequency of treatments, or, if loss of field continues despite very little to no treatment, resorting to surgery. One must remember that improvement after radiotherapy need not necessarily come immediately after the radiotherapy but may be delayed until after treatment has been completed, coming on often during the first 6 months after radiation and continuing, perhaps, for a year or even more after completion of treatment. REFERENCES 1. C~AMLIN, M. Minimal defects in visual field studies. Arch. Ophthal., 1949, 2nd s. 42: CHAMLIN, M. Visual field changes in optic neuritis. Arch. Ophthal., Chicago., 1953, 2nd s. 50: CHAMLIN, M., and DAVIDOFF, L. M. The 1/2000 field in chiasmal interference. Arch. Ophthal., Chicago, 1950, 2nd s. 44: CHAMLIN, M., and DAVlDOFF, L. M. Choice of test objects in visual field studies. Amer. J. Ophthal., 1952, s.3 35: CHAMLIN, N[., DAVlDOFF, L. M., and FEIRING, E. ti. Ophthalmologic changes produced by pituitary tumors. Amer. J. Ophthal., 1955, s.8 40: RUCKER, C. W. Bitemporal defects in the visual fields resulting from developmental anomalies of the optic disks. Arch. Ophthal., Chicago, 1946, 2nd s. 35: WALKER, C. B. A contribution to the study of bitemporal hemianopsia with new instruments and methods for detecting slight changes. Arch. Ophthal., N.Y., 1915, 44: WALSH, F.B. Clinical neuro-ophthalmology. Baltimore: Williams & Wilkins Co., 1957, 2nd ed., xvi, 1294 pp.