Background. Patients with chiasmal lesions frequently suffer loss of v
ision, but some may be visually asymptomatic and have little, if any,
visual field loss. This study addressed two questions about the nature
of such loss: (1) which tests are best able to detect visual loss and
(2) what is the relation between the size of a lesion and the degree
and type of visual loss. Methods. Visual acuity (VA), Goldmann perimet
ry, pattern evoked potentials, contrast sensitivity (CS), and color vi
sion (CV) were assessed, in addition to a full ophthalmologic examinat
ion. Thirty-five patients were recruited from neurology, endocrinology
, and ophthalmology departments; 31 had pituitary tumors and 4 had tum
ors of other origin (1 secondary) close to the chiasm. For analysis, p
atients were classified (1) according to whether they had visual sympt
oms on presentation and (2) according to lesion size. Results, Sevente
en patients had visual symptoms; all had visual field defects and all
but one had two or more additional abnormal test results. Of the 18 vi
sually asymptomatic patients, 12 had visual field defects and 14 had a
t least 2 abnormal test results; 7 had tumors with less than 5 mm of s
uprasellar extension, of whom 5 had field loss or at least 2 abnormal
test results. CS was abnormal most often (85% of patients), followed b
y Goldmann perimetry (80%), visual evoked potentials (VEPs) (57%), CV
(49%), and VA (26%). Conclusion Our results support previous reports t
hat asymptomatic patients and those with intrasellar tumors or only sm
all amounts of suprasellar extension may have demonstrable visual dysf
unction. Perimetry did not appear to be significantly worse than other
more recently developed tests, which have often been regarded as bein
g more sensitive, for detecting vision loss. However, VA and clinical
observation of pupil reactions had poor sensitivity.