Neuroradiology/Head and Neck Imaging Original Research

Transcription

1 Neuroradiology/Head and Neck Imaging Original Research Lee et al. Visual Defects Noted on MRI Examination of Patients With Pituitary denomas Neuroradiology/Head and Neck Imaging Original Research In Ho Lee 1, 2 Neil R. Miller 3 Elcin Zan 1 Fabiana Tavares 1 ri M. litz 1 Heejong Sung 4 David M. Yousem 1 Michael V. oland 3 Lee IH, Miller NR, Zan E, et al. Keywords: optic chiasm, pituitary adenoma, visual deficits DOI:.2214/JR Received February 7, 215; accepted after revision pril 16, 215. The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as representing the views of the National Institutes of Health. 1 Russell H. Morgan Department of Radiology and Radiological Sciences, The Johns Hopkins Medical Institutions, 6 N Wolfe St, Phipps F, altimore, MD ddress correspondence to D. M. Yousem (dyousem1@jhu.edu). 2 Department of Radiology, Chungnam National University Hospital, Daejeon, Korea. 3 Wilmer Eye Institute, The Johns Hopkins Medical Institutions, altimore, MD. 4 Genometrics Section, Computational and Statistical Genomics ranch, National Human Genome Research Institute, National Institutes of Health, altimore, MD. WE This is a web exclusive article. JR 215; 25:W512 W X/15/255 W512 merican Roentgen Ray Society Visual Defects in Patients With Pituitary denomas: The Myth of itemporal Hemianopsia OJECTIVE. The objective of this study was to test the hypothesis that bitemporal hemianopsia (H) is the most common visual field (VF) defect in patients with pituitary macroadenoma and to assess the degree of optic pathway compression necessary to produce visual defects. MTERILS ND METHODS. We reviewed the MRI findings and medical records of 119 patients with pituitary macroadenoma who had undergone formal assessment of VFs. We then evaluated the degree of optic pathway displacement caused by the pituitary macroadenoma, as observed on MR images. The classifications of optic pathway displacement included no contact, abutment but no displacement, mild displacement (< 3 mm), and moderate displacement ( 3 mm). Qualitative analysis classified VFs as normal or as having defects that were monocular, bitemporal, mixed (bitemporal with additional defects), homonymous, or nonspecific. RESULTS. total of 89 of 115 patients had an abnormal VF. Only one patient had true H. The most common defects were bitemporal or mixed defects (in 49 of 115 patients [42.6%]), likely because more than just the chiasm is often compressed by the pituitary macroadenoma. Classification of optic pathway displacement by the pituitary macroadenoma was as follows: 23 patients had no contact, eight had abutment but no displacement, 27 had mild displacement, and 57 had moderate displacement. In 78 of the 92 patients (84.8%) with pituitary macroadenoma that had contact with the optic pathway, contact was with the optic chiasm and the prechiasmal optic nerve. Of the 49 patients with bitemporal or mixed defects, 42 had moderate displacement of the optic pathway caused by their tumors. CONCLUSION. H is exceedingly uncommon in patients with pituitary macroadenoma. However, although bitemporal and mixed defects are the most common abnormal VF findings, they were found in only 42.6% of patients. Such defects rarely occur if the tumor displaces the optic pathway less than 3 mm from baseline. W hen most neuroradiologists consider the visual field (VF) deficits associated with pituitary adenomas that compress the chiasm, they automatically think of bitemporal hemianopsia (H). This is a VF deficit in which all the vision in the temporal fields of both eyes is lost, leaving only the nasal fields to be perceived. Incomplete bitemporal VF defects are much more common than true hemianopsia and are considered by neuroophthalmologists as a sign characteristic of chiasmal syndrome, which is usually caused by lesions that affect the optic chiasm from below [1]. Pituitary adenomas are the most common of all chiasmal syndrome tumors, followed by other lesions that cause extrinsic optic chiasm compression, such as me- ningiomas, craniopharyngiomas, and aneurysms [2, 3]. The visual deficits associated with pituitary adenoma depend on the size, location, and hormonal activity of the tumor as well as the position of the chiasm as it relates to the sella turcica [4]. ccording to a recent study, the tumor volume also affects the severity of the VF defect [5, 6]. lthough a previous study of 5 patients showed a significant correlation between chiasmal compression and visual disturbances [7], to our knowledge, no MRI-based literature shows the relationship between the degree and symmetry of extrinsic anterior visual pathway compression by pituitary macroadenoma and the pattern of VF defects observed in patients. One goal of this study W512 JR:25, November 215

2 Visual Defects Noted on MRI Examination of Patients With Pituitary denomas was to test the hypothesis that bitemporal VF defects, not H, are the most common defect in patients with pituitary macroadenoma. We also wished to determine what degree of optic chiasm compression is necessary to produce such defects and how often asymmetric visual defects are associated with pituitary macroadenoma that asymmetrically affects the prechiasmal optic nerves, optic chiasm, or postchiasmal optic tracts. Materials and Methods This retrospective study was reviewed and approved by the institutional review board of The Johns Hopkins University School of Medicine. ecause of the retrospective nature of the study, informed consent was not required for review of either the medical records or the MR images. We retrospectively searched our imaging archive for patients who had pituitary macroadenoma diagnosed by MRI between November 29 and October 212. In all patients, at least one dimension of the pituitary macroadenoma measured at least mm. We then reviewed the imaging findings and medical records of the 185 patients who were identified using this strategy. Sixty-six patients were excluded either because they did not have any VF testing results available or because they had other underlying diseases, such as stroke, glaucoma, ocular or intracranial trauma, retinal artery occlusion, other retinal diseases, amblyopia, or unrelated optic neuropathy, all of which could affect the results of VF testing. Next, we reviewed the clinical findings and ophthalmologic records of the remaining 119 patients who were included in the study, including documentation of VF defects and reported visual complaints. ll VF tests were performed using a Humphrey Field nalyzer (Carl Zeiss Meditec). Test patterns were either 24 2 or 3 2, and strategies included use of the Swedish interactive threshold algorithm or full threshold. Qualitative Visual Field nalysis Qualitative analysis of the VF tests was performed by two experienced ophthalmologists (one with years of experience and one with 4 years of experience) who were blinded to the MRI findings. fter assessments were performed independently, the analyses were compared jointly, and any differences in grading were adjudicated by consensus. VF test findings were classified as normal (i.e., no defect), unreliable, or as one of the following types of defect: bitemporal (bitemporal defect only, including H), mixed (bitemporal and additional defects), monocular, homonymous, nonspecific, or other. H, which was included as one type of a bitemporal field defect, was diagnosed when the defect affected the entire outer (or lateral) half of the VF in each eye. bitemporal defect was diagnosed when the defect affected the outer (or lateral) half of the VF in each eye, whereas a mixed defect was diagnosed when the defect involved not only the outer (or lateral) half of the VF of both eyes but also other areas of the VF in one or both eyes. monocular defect was defined as the presence of any VF defect in one eye and a normal VF in the other eye. homonymous defect was defined as a defect that was present in the temporal (outer) field of one eye and the nasal (inner) field of the other eye and, for the purpose of this study, was thought to be consistent with a compressive lesion in the region of the chiasm. Nonspecific defects were defined as defects that could not clearly be attributed to a particular ocular or neurologic process. The other category identified defects that the evaluators thought were not likely caused by the tumor. Finally, unreliable test results were defined as results associated with excessive false-positive responses, false-negative responses, or fixation losses still requiring interpretation. Qualitative analysis of the degree of asymmetry in VF loss between the right and left eyes was also performed for patients with bitemporal, mixed, or homonymous defects. The scale used in such analysis included the following categories of asymmetry: significantly more left, somewhat more left, symmetric, somewhat more right, and significantly more right. The four patients who were found to have unreliable results of VF testing or who received a diagnosis consistent with other ocular or neurologic diseases were excluded from further analysis. nalysis of MR Images Each patient underwent an MRI examination that consisted of a standard protocol of sagittal and coronal T1-weighted and CISS/FIEST (constructive interference in steady state fast imaging employing steady-state acquisition) sequences performed before and after gadolinium contrast agent administration. ll sections were 3 mm or thinner. We evaluated the degree of displacement of the prechiasmal optic nerve, optic chiasm, and postchiasmal optic tract by the pituitary macroadenoma. If there was symmetric displacement, we determined the degree of displacement relative to the expected normal location of the visual pathway. If there was bilateral but asymmetric displacement, we compared the displacement on each side with the expected location of the visual apparatus. Finally, if there was unilateral displacement, we determined the displacement of the affected side relative to the position of the unaffected side. In all cases, the degree of optic pathway displacement was classified as no contact, abutment but not displacement, mild displacement (< 3 mm of displacement), or moderate displacement ( 3 mm of displacement) (Fig. 1). We selected the maximum displacement when more than two parts of the optic pathway were displaced by the pituitary macroadenoma. Two neuroradiologists (one with 8 years of experience and one with 3 years of experience) independently measured displacement with the use of electronic calipers. The mean of the two recorded displacements was used, unless there was a discrepancy of more than 3 mm, in which case an adjudication was performed by a third neuroradiologist (with 25 years of experience), who was blinded to the measurements of the other two neuroradiologists. symmetry from right to left in the degree of compression of the optic pathway on MRI was evaluated subjectively (Fig. 2). djudication was performed in the same manner as it was for displacement. We also recorded changes in signal intensity on T2-weighted or FLIR sequences, contrast enhancement, and atrophy anywhere along the optic pathway and determined the presence of hemorrhage in the pituitary lesions. The imaging parameters used for the sagittal thin T1-weighted sequence were as follows: TR/TE, 45/9.5; matrix, ; FOV, mm; and section thickness, 2 mm. The coronal thin T1-weighted sequence was obtained using the following parameters: TR/TE, 45/9.5; matrix, ; FOV, mm; and section thickness, 2 mm. The parameters used to obtain the axial T2-weighted sequence were TR/TE, 45/89; matrix, ; FOV, mm; and section thickness, 4 mm. FLIR sequence parameters were as follows: TR/TE, 9/5; inversion time, 25 ms; matrix, 32 32; FOV, mm; and section thickness, 4 mm. The contrast-enhanced thin coronal T1-weighted sequence was obtained using the following parameters: TR/TE, 5/9.5; matrix, ; FOV, mm; and section thickness, 2 mm. The contrast-enhanced thin sagittal T1-weighted sequence was obtained with the use of the parameters TR/TE, 45/9.4; matrix, ; FOV, mm; and section thickness, 2 mm. The parameters used to obtain the contrast-enhanced axial T1-weighted sequence were TR/TE, 55/12; matrix, 32 32; FOV, mm; and section thickness, 4 mm. The contrast-enhanced coronal T1-weighted sequence was obtained using the following parameters: TR/TE, 65/9.1; matrix, 32 32; FOV, mm; and section thickness, 4 mm. Results Of the 115 patients (49 women and 66 men; mean [± SD] age, 54.7 ± 6.2 years; age range, JR:25, November 215 W513

3 Lee et al. C Fig. 1 MRI classification of degree of optic pathway displacement in patients with pituitary macroadenoma., MR image shows no contact between optic pathway and pituitary macroadenoma., MR image shows pituitary macroadenoma abutting but not displacing optic pathway. C, MR image shows less than 3 mm of displacement (i.e., mild displacement) of optic pathway caused by compression by pituitary macroadenoma. D, MR image shows displacement of 3 mm or more (i.e., moderate displacement) of optic pathway caused by compression by pituitary macroadenoma. Fig. 2 MRI classification of asymmetry in the degree of compression in the optic pathway in patients with pituitary macroadenoma., MR image shows asymmetric displacement of right visual field of optic pathway caused by compression by pituitary macroadenoma., MR image shows asymmetric displacement of optic pathway caused by compression by pituitary macroadenoma. D years), 7 had pathologically proven pituitary macroadenoma, whereas 45 had pituitary macroadenoma diagnosed on the basis of clinical findings, hormonal analysis, and imaging findings. The height of the 115 lesions ranged from.4 to 5 cm (mean, 2.17 ±.79 cm). Of the patients studied, 57 (49.6%) had moderate displacement (range, 3 21 mm) of the optic apparatus by the pituitary macroadenoma (Figs. 3 and 4), 27 (23.5%) had mild displacement, and eight (7%) had abutment but no displacement. In the remaining 23 patients (2%), the lesion had no contact with any optic pathway structures. Of the 92 patients with pituitary macroadenoma showing contact with parts of the optic pathway, 79 (85.9%) had contact with two or more areas. The areas of contact were the prechiasmal optic nerve and optic chiasm (58 patients); the prechiasmal optic nerve, optic chiasm, and postchiasmal optic tract (2 patients); and the optic chiasm and postchiasmal optic tract (one patient). In the remaining 13 patients, the lesion had contact with only one area: the optic chiasm (nine patients) or the prechiasmal optic nerve (four patients) (Fig. 5). Sixty-three patients had no documented visual complaints. The other 52 patients complained of visual disturbances, including blurred vision (2 patients), diplopia (four patients), or visual change (28 patients). symptomatic Subjects Sixty-three patients had no visual complaints at presentation; in 14 of these patients (22.2%), the tumors had no contact with the optic pathway. In the remainder of the patients, tumor contact with the optic pathway was as follows: four patients (6.3%) had tumor abutment but no displacement, 2 (31.7%) had mild displacement, and 25 (39.7%) had moderate displacement. The formal VF test findings for these asymptomatic patients were classified as follows: 18 patients (28.6%) had normal findings, 14 (22.2%) had bitemporal defects (without H), six (9.5%) had mixed defects, one (1.6%) had homonymous defects, eight (12.7%) had monocular defects, and 16 (25.4%) had nonspecific defects (Fig. 6). Symptomatic Subjects Fifty-two patients had visual complaints at presentation; in nine of these patients (17.3%), the tumor had no contact with the optic pathway. In the remainder of the patients, tumor contact with the optic pathway was as fol- W514 JR:25, November 215

4 Visual Defects Noted on MRI Examination of Patients With Pituitary denomas Fig. 3 Patient with pituitary macroadenoma. and, Measurement of optic pathway displacement at level of prechiasmal optic nerve. Coronal unenhanced () and contrast-enhanced () T1-weighted MR images show 6 mm of displacement of prechiasmal optic nerve. Normal position of optic pathway (bar, ) and distance from normal position (arrow, ) are shown. Fig. 4 Patient with pituitary macroadenoma. and, Measurement of optic pathway displacement at level of optic chiasm. Coronal unenhanced () and contrast-enhanced () T1-weighted MR images show 9 mm of displacement of optic chiasm. Normal position of optic pathway (bar, ) and distance from normal position (arrow, ) are shown. Pituitary Macroadenomas (no.) Prechiasmal Optic Nerve Prechiasmal Optic Nerve and Optic Chiasm Prechiasmal Optic Nerve, Optic Chiasm, and Postchiasmal Optic Tract Optic Nerve and Postchiasmal Optic Tract rea(s) of Contact With Optic Pathway Optic Chiasm Fig. 5 Graph showing relationship between optic pathway and pituitary macroadenoma in 92 patients with pituitary macroadenoma that had contact with optic pathway. lows: four patients (7.7%) had abutment but no displacement, seven patients (13.6%) had mild displacement, and 32 (61.5%) had moderate displacement. For these symptomatic patients, VF test results were graded as follows: eight patients (15.4%) had normal findings, 15 (28.8%) had bitemporal defects (with only one H noted among these 15 patients), 14 (26.9%) had mixed defects, four (7.7%) had monocular defects, and 11 (2.1%) had nonspecific defects (Fig. 7). Visual Field nalysis Overall, 89 patients (77.4%) had abnormal VFs, and 26 patients (22.6%) had normal VFs. Of the 89 patients with abnormal VFs, bitemporal or mixed defects were present in 49 patients (representing 55.1% of this group and 42.6% of all 115 subjects) and were the most common patterns, followed by nonspecific defects (27 patients), monocular defects (12 patients), and homonymous defects (one patient) (Fig. 8). Of the 49 patients with bitemporal or mixed defects, 42 (85.7%) had moderate displacement (range, 4 21 mm), five (.2%) had mild displacement, and two (4.1%) had lesions with no contact with the optic pathway. Of the 49 patients with bitemporal field deficits, only one patient had H. This patient had moderate displacement of the optic pathway. When moderate displacement (optic pathway displacement greater than 3 mm) was used as the criterion for identifying bitemporal or mixed defects, VF testing had a sensitivity of 85.7% (42/49), specificity of 75% (3/4), a positive predictive value of 8.8% (42/52), a negative predictive value of 81.1% (3/37), and accuracy of 8.9% (72/89). When the 27 patients with nonspecific VF defects were excluded, 42 of 49 patients (85.7%) with bitemporal or mixed defects had moderate optic pathway displacement resulting from masses, whereas two of 13 patients (15.4%) with atypical, homonymous, or monocular defects had moderate displacement. symmetry For 41 of 115 patients, asymmetry in the degree of compression of the optic pathway noted on MR images was from right (15 patients) to left (26 patients). symmetry in the degree of VF damage between right and left eyes was also noted in 39 of 49 patients (79.6%) with bitemporal or mixed defects, and symmetry was noted in of 49 patients (2.4%). For predicting asymmetry of the right and left VFs, MRI JR:25, November 215 W515

5 Lee et al. had a sensitivity of 43.6% (17 of 39 patients), specificity of 8% (8 of patients), a positive predictive value of 89.5% (17 of 19 patients), a negative predictive value of 26.7% (8 of 3 patients), and accuracy of 51% (25 of 49 patients) (Fig. 9). One of the 115 patients (.9%) had atrophy of the optic chiasm, and 19 (16.5%) had hemorrhage. With the exception of one patient (.9%) who had high signal intensity noted in the left postchiasmal optic tract on T2-weighted or FLIR images, we did not find any signal change on T2-weighted or FLIR images. There was no optic pathway component enhancement. The patient who had high signal intensity noted on T2-weighted or FLIR images had a mixed VF defect Moderate Mild Discussion Our study confirms that the classic finding of pure H in patients with pituitary macroadenoma is a myth; only one of the 115 patients in our cohort had this defect. In fact, the VF defects in our patients with pituitary macroadenoma were purely bitemporal (even if incomplete) in only 29 of 115 patients (25.2%); we found mixed defects with areas of VF loss outside the temporal fields in 2 of 115 patients (17.4%). These findings are comparable to those of other neuroophthalmologic studies, which found that pure Hs are rare compared with bitemporal defects, with the former occurring in only approximately 1% of patients with pituitary macroadenoma [1, 8, 9]. We also found that patients with macroadenoma may have other defects, including monocular or homonymous defects [5, 12]. Indeed, 4 of the 89 patients with VF defects (44.9%) had nontemporal defects. This may have resulted from compression of prechiasmal optic nerves or postchiasmal tracts. Involvement of prechiasmal optic nerves or postchiasmal tracts was seen in 82 and 21 patients, respectively. Of the patients with preand postchiasmal compression, 79 also had optic chiasm compression. Only nine patients had pure optic chiasm compression alone. Thus, it is more common to have extrachiasmal optic pathway involvement along with compression of the chiasm, rather than just chiasm compression alone. Previous studies did not evaluate extrachiasmal compression, evaluating only chiasmal compression instead. In general, VF defects (whether bitemporal, mixed, homonymous, or monocular) correlated with the degree of optic pathway displacement noted, with 71% of patients with VF defects having moderate optic pathway displacement that ranged from 4 to 21 mm. Eight patients in this study had abnormal VFs, a finding that was thought to be consistent with optic pathway damage resulting from their tumors, although MRI examination of these patients revealed no contact between the tumor and the optic apparatus. It is possible that in some of these patients, the VF defect was spurious or related to a process other than the tumor. lternatively, some studies have theorized that the discovery of butting Tumor Contact With Optic Pathway No Contact Normal Nonspecific defects Monocular defects Homonymous defects itemporal or misdefects Fig. 6 Graph showing relationship between displacement of optic pathway and visual fields in 63 patients without visual complaints Moderate Mild butting Tumor Contact With Optic Pathway No Contact Normal Nonspecific defects Monocular defects Homonymous defects itemporal or misdefects Fig. 7 Graph showing relationship between displacement of optic pathway and visual fields in 52 patients with visual complaints. abnormal VFs in patients who have pituitary tumors that do not appear to be in contact with the optic apparatus may be attributed to previous indentation (and subsequent tumor regression), hormonal influences, intratumor hemorrhage, autonecrosis, or vascular shunting [13]. In our study, the smallest displacement of the optic apparatus from its expected normal position in a patient with a presumed related VF defect was 4 mm on the coronal plane, compared with previous reports indicating a minimum displacement of mm [7, 9]. This discrepancy may W516 JR:25, November 215

6 Visual Defects Noted on MRI Examination of Patients With Pituitary denomas Moderate Mild be attributed to differences in the measurement methods used in our study and previous studies. The studies by Ikeda and Yoshimoto [7] and Schmalisch et al. [9] measured the degree of displacement of the optic chiasm from the upper surface of the internal carotid artery within the cavernous sinus, whereas our study measured the degree of displacement from the expected normal (presumed) location, according to the course of the optic pathway from the optic canal to the optic tract. Even if there are differences in the methods used by the three studies, the notion that optic pathway displacement by pituitary adenomas contributes to visual symptoms is supported by our study. butting Tumor Contact With Optic Pathway No Contact Normal Nonspecific defects Monocular defects Homonymous defects itemporal or misdefects Fig. 8 Graph showing relationship between displacement of optic pathway and visual fields in all 115 study patients Right symmetry Symmetry Left symmetry Results of Visual Field Testing Right asymmetry on MRI Symmetry on MRI Left asymmetry on MRI Fig. 9 Graph showing relationship of asymmetry between MRI and qualitative analysis of visual field test. In this study, a change in the signal intensity noted in the optic apparatus on T2-weighted or FLIR images was exceedingly rare, occurring in only one of 115 patients. This finding is counter to the results reported in a previous study [14]. symmetry in VF abnormalities was also noted and was confirmed by qualitative analysis of the VFs (in 39 of 49 patients [79.6%]) and by MRI (in 41 of 115 patients [35.7%]). This asymmetry has been reported to be a very common finding of VF testing of patients with pituitary adenomas [11, 15]. This asymmetry may also be attributed to a different nerve fiber strain between the nasal and temporal nerve fibers of bilateral eyes [16]. In our study, asymmetry was noted on the MR images of 41 of 115 patients (35.7%), with MRI therefore having 43.6% sensitivity and 8% specificity for predicting asymmetry on the VF test. This sensitivity and specificity may be attributed to multiple factors, including the patient s health status and hormonal status, the follow-up period, and the location and shape of tumor. However, the high positive predictive value (89.5%) of MRI in predicting asymmetry of the right and left VFs suggests that when MRI shows asymmetry to one side, there very frequently is asymmetry in VF damage. There are several limitations of this study. First, not all of the patients had pathologically proven tumors; thus, it is possible (although unlikely) that some of the lesions were not adenomas. Second, there was a wide interval (mean, 48.2 ± 42.8 days; range, 175 days) between MRI examination and VF testing, although there was no documented statistically significant change in the size of pituitary macroadenoma noted on serial MR images. Third, there were three patients without a contrast-enhanced study and two patients without T2-weighted or FLIR images. Finally, there likely is selection bias for patients with visual symptoms to be referred for VF analysis. Conclusion Complete H associated with pituitary macroadenoma is rare, occurring in only one of 115 patients in our cohort. On the other hand, bilateral temporal VF defects, either pure or associated with additional defects, are the most common defects noted in patients with pituitary macroadenoma, occurring in 42.6% of all patients. Most patients had compression of the prechiasmal optic nerves, the postchiasmal tracts, or both, in addition to compression of the chiasm, which may account for these impure VF findings that include areas outside the bitemporal zones. Such defects are typically present in patients with greater than 3 mm of displacement of the optic apparatus. In addition, patients with asymmetric compression of the optic pathway seen on MR images are likely to have asymmetric VF defects. References 1. Schiefer U, Isbert M, Mikolaschek E, et al. Distribution of scotoma pattern related to chiasmal lesions with special reference to anterior junction syndrome. Graefes rch Clin Exp Ophthalmol 24; 242: Foroozan R. Chiasmal syndromes. Curr Opin JR:25, November 215 W517

7 Lee et al. Ophthalmol 23; 14: PJ, Danesh-Meyer HV. Visual acuity and pattern manifesting as homonymous hemianopia. Jpn J 3. Jacobs D, Galetta SL. Neuro-ophthalmology for of visual field loss at presentation in pituitary ad- Ophthalmol 27; 51: neuroradiologists. JNR 27; 28:3 8 enoma. J Clin Neurosci 214; 21: Levy. Pituitary disease: presentation, diagno- 4. Kapoor S. cromegaly. N Engl J Med 27; 9. Schmalisch K, Milian M, Schimitzek T, Lagrèze sis, and management. J Neurol Neurosurg Psy- 356: [author reply, ] W, Honegger J. Predictors for visual dysfunc- chiatry 24; 75:iii47 iii52 5. Lee JP, Park IW, Chung YS. The volume of tumor tion in nonfunctioning pituitary adenomas: impli- 14. Saeki N, Uchino Y, Murai H, et al. MR imaging mass and visual field defect in patients with pituitary macroadenoma. Korean J Ophthalmol 211; 25: Kan E, Kan EK, tmaca, tmaca H, Colak R. Visual field defects in 23 acromegalic patients. Int Ophthalmol 213; 33: Ikeda H, Yoshimoto T. Visual disturbances in patients with pituitary adenoma. cta Neurol Scand 1995; 92: Ogra S, Nichols D, Stylli S, Kaye H, Savino cations for neurosurgical management. Clin Endocrinol (Oxf) 212; 77: Hershenfeld S, Sharpe J. Monocular temporal hemianopia. r J Ophthalmol 1993; 77: Rivoal O, rézin P, Feldman-illard S, Luton JP. Goldmann perimetry in acromegaly: a survey of 37 cases from 1951 through Ophthalmology 2; 7: Nishimura M, Kurimoto T, Yamagata Y, Ikemoto H, rita N, Mimura O. Giant pituitary adenoma study of edema-like change along the optic tract in patients with pituitary region tumors. JNR 23; 24: Poon, McNeill P, Harper, O Day J. Patterns of visual loss associated with pituitary macroadenomas. ust N Z J Ophthalmol 1995; 23: Wang X, Neely J, McIlwaine GG, Lueck CJ. Multi-scale analysis of optic chiasmal compression by finite element modelling. J iomech 214; 47: W518 JR:25, November 215