Visual Evoked Potentials in Occipital Lobe Lesions Leopold J. Streletz, MD; Sung Ho Bae, MD; Robert M. Roeshman, DO; Norman J. Schatz, MD; Peter J. Savino, MD \s=b\ Recording of visual evoked potentials (VEPs) to pattern reversal is considered to be a reliable diagnostic procedure for examining patients with anterior visual pathway lesions (optic nerves and chiasm). Less consistent results have been reported in studies of more posterior lesions. The VEPs were recorded in 20 patients with occipital lobe lesions. A maximal VEP response (P94) was recorded at the scalp electrodes situated over the involved occipital lobes and contralateral to the hemianoptic visual field defect, indicating a positive correlation of unilateral occipital lobe lesions, homonymous visual field loss, and the VEP abnormali- ty. (Arch Neurol 38:80-85, 1981) "Vrisual evoked potential (VEP) re' cordings are recognized as a relia¬ ble diagnostic procedure in the inves¬ tigation of lesions of the anterior visual pathways.15 Geniculocalcarine lesions accompanied by hemianoptic visual field defects also show asym¬ metrical VEPs.2 We report our experi¬ ence with VEPs to pattern reversal stimuli in patients with occipital lobe lesions. MATERIALS AND METHODS Twenty patients (seven men and 13 women) with occipital lobe lesions were Accepted for publication April 24, 1980. From the Department of Neurology, Jefferson Medical College, Thomas Jefferson University, Philadelphia (Drs Streletz, Bae, and Roeshman); Neuro-ophthalmology Unit, Wills Eye Hospital, Philadelphia (Drs Schatz and Savino); and the Departments of Neurology and Ophthalmology, University of Pennsylvania, Philadelphia (Drs Schatz and Savino). Reprint requests to Department of Neurology, Thomas Jefferson University Hospital, Room 9606 NH, 11th and Walnut streets, Philadelphia, PA 19107 (Dr Streletz). selected for VEP studies to pattern rever¬ sal stimuli. Ages ranged from 22 to 71 years (mean, 56 years). All patients under¬ neuro-ophthalmologic examination, including Goldmann perimetry. A comput¬ erized axial tomographic (CT) scan was performed for anatomical verification of the site of the lesion on all but one patient. Similarly, 20 (ten men and ten women) normal volunteers were studied according to the following technique. With the inter¬ national 10-20 system of electrode place¬ ment, five scalp electrodes were located transversely in the temporo-occipital re¬ gion (T,, 0„ 0„ 02, and Te) with a frontal (F„) reference site (Fig 1). Electrodes were attached to the scalp with collodion jelly. Impedance was maintained between 1,000 and 3,000 . The volunteers and patients were comfortably seated in a chair before a went translucent screen on which a black-andwhite checkerboard pattern was back-pro¬ jected (Digitimer Pattern Generator 4330). They were instructed to fixate on a red dot at the center of the screen while the responses were being recorded. Each eye was tested separately. A stimulator (Grass S-8) rhythmically reversed this pattern by deflecting a small mirror system twice per second. The slew rate of the mirror system was 2.4 ms. The patterned screen sub¬ tended 30° of the visual angle. Each indi¬ vidual square made up Io 15' of arc. The luminance of the screen measured 60 footlamberts. The VEP signal was amplified (Grass 8A5) with a bandpass of 1 to 1,000 cps. A four-channel averaging computer (Nicolet 1072) was triggered by the stimulator. The analysis time was 200 ms, and 128 responses were summated. The sampling rate was 1.25 kHz. The resulting VEP signals were recorded on an X-Y plotter (Hewlett-Packard 7035 B). The averaging process was continuously displayed on an oscilloscope (Tektronix RM 504) to monitor artifacts of electrical or biological origin. Individual VEPs were superimposed to aid in differentiating time-locked from ran¬ dom activity. The peak latency, amplitude, and distribution of the VEP were com¬ pared with the patients. Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 RESULTS Full-field stimulation of 20 normal volunteers resulted in a VEP response with maximal amplitude in the midline (Oz). This response was a triphasic wave form with negative-positivenegative components. The peak laten¬ cies (mean ± SD) for these compo¬ nents were 71.0 ± 3.6, 94.0 ± 5.6, and 134.0 ± 9.5 ms. The major positive potential was the most consistent and largest component with a mean ampli¬ tude of 14.7 ± 4.6 µ measured peak to peak from the preceding negative potential. Measureable amplitude asymmetrices at lateral occipital re¬ cording sites (0, and 02) were said to have a predominance on the side of greatest amplitude. The mean differ¬ ence in amplitude for 20 subjects was 0.5 ± 2.7 µ , which suggested no defi¬ nite predominance of either side. (The mean difference amplitude between the ! and 02 recording sites was calculated as follows: [ µ (O, lead) YßV (02 lead) OS] + [ 'µ (O, lead) 'µ (02 lead) OD]/2 ampli¬ tude difference in a subject [P94 peakpeak].) In all 20 patients with occipital lobe abnormalities (18 occipital infarcts, two biopsy-proved tumors), some form of VEP abnormality was detected — = - (Table 1). All lesions were associated with congruous homonymous visual field defects and two patients (No. 9 and 20) displayed Riddoch's phenome¬ non.6 In two patients (No. 19 and 20) with bilateral occipital lobe infarctions, grossly abnormal VEP wave forms were obtained that made determina¬ tion of predominance impossible. In patients 17 and 18, the peak latency of the major positive component was delayed in the absence of demonstra- ble abnormality of anterior visual pathways. Both patients had left occi¬ pital lobe infarctions, but no definite amplitude asymmetry was detected. In 15 of 18 patients (patients 1 through 16) with unilateral occipital lobe lesions and homonymous visual field defects, the P94 component had the greatest amplitude ipsilateral to the lesion and contralateral to the visual field defect. This predominance applied whether the right or left occip¬ ital lobe was involved (Fig 2). Predom¬ inances were also observed in two cases (No. 8 and 12) with superim¬ posed ocular and optic nerve lesions. All three groups (normal subjects and the two groups of unilateral left [six] and right [12] occipital lobe lesions) demonstrated highly signifi¬ cant intergroup variation with regard to predominance of the P94 component (Fig 2). With the use of the Newman- Keuls method, the groups were found to differ significantly among them¬ selves (P < .01 for normal vs right and for left vs right and < .05 for normal vs left). Although differences in amplitude were dissimilar between groups, this measurement was not effective in distinguishing individual patients with unilateral occipital ab¬ normalities from normal subjects (Ta¬ ble 2). Despite this limitation under some clinical circumstances when an occipi- Fig 1.—Normal computerized tomographic scan indicating location of recording elec¬ trodes in relation to occipital lobes (top). Visual evoked potentials show symmetrical distribution of P.,., component (bottom). tal lesion was suspected and objective laboratory confirmation was needed, this technique proved to be a useful adjunct to the neuro-ophthalmologic evaluation (Table 1, patients 6, 9, and 10). REPORT OF CASES Case 15.-A 64-year-old woman noted the sudden onset of inability to see to on the right three months prior to examination. This remained unchanged and was unac¬ companied by neurologic or visual symp¬ toms. Neuro-ophthalmic examination re¬ vealed visual acuity of 6/6 bilaterally. The results of pupillary and ophthalmoscopic examinations were normal. An exquisitely congruous right homonymous hemianoptic scotoma was detected on visual field test¬ ing. Neurologic examination gave normal results. Potassium perchlorate brain scan and skull x-ray films were normal. An EEG was normal, but VEP testing revealed a left-sided predominance. Computerized to¬ mographic scan revealed a low-density area in the left occipital lobe consistent with infarction (Fig 3, left). Case 5.—A 46-year-old man had a convul¬ sive seizure disorder. A large occipital meningioma was discovered and removed. He remained well for four years, when he had another generalized seizure. Following removal of a recurrent occipital meningio¬ ma, he complained of decreased vision. His visual acuity was 6/6 bilaterally. Perimetry revealed a left inferior homonymous paracentral scotoma. Pupillary reaction and results of ophthalmoscopy were normal. The CT scan revealed encephalomalacia and focal dilation of the occipital horn of the right lateral ventricle. The VEP dis¬ closed predominance of the PS4 component over the right temporo-occipital region (Fig 3, right). COMMENT In studies of cerebral lesions pro¬ ducing homonymous visual field de¬ fects, flash stimulation VEPs have demonstrated reduction in amplitude of early components of the response recorded from occipital leads ipsilater- al to the side of the lesion.79 There are, however, considerable difficulties between patients and within individu¬ al patients in distinguishing such changes from normal variations with the flash technique. Earlier investiga¬ tions with full-field pattern reversal techniques have not identified consis¬ tent abnormalities in posterior visual pathway lesions.'·' One difficulty in the interpretation of pattern-reversal VEP findings is the parodoxical lateralization of the normal half-field response in trans¬ verse temporo-occipital scalp leads. Barrett et al reported that the halffield responses in referential and bipolar recordings appeared asym¬ metrical and had a potential maxi- Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 Table 1.—Twenty Patients With Occipital Lobe Lesions yr/Sex Diagnosis Scan Visual Evoked Potential 1/47/M Occipital lobe infarction (R) Low-density area R occipital lobe Predominance R side 2/70/F Occipital lobe infarction (R) Low-density area R occipital lobe Predominance 3/67/F Occipital lobe infarction (R) Low-density area R occipital lobe Predominance R side 4/36/M Occipital lobe metastasis (R) Low-density area R occipital lobe Predominance R side 5/46/M Occipital lobe Low-density area R occipital lobe Predominance Occipital lobe infarction (R) Normal Predominance R side 7/65/F Occipital lobe infarction (R) Low-density area R occipital lobe Predominance R side 8/71/M Optic atrophy (L), occipital Low-density area R occipital No response (OS) Patient/Age, Clinical meningioma (R) 6/59/F lobe infarction Computerized Tomographie lobe (R) R side R side mm ASfÜíí" Predominance R side (OD) 9/54/F Occipital lobe infarction (R) Normal Predominance R side 10/22/F Occipital lobe Normal Predominance infarction Visual Field R side (R), migraine Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 ¿ \^.' 'v;Í?r Table 1.—Twenty Patients With Occipital Lobe Lesions (cont) Visual Evoked Potential Clinical Computerized Tomographie Diagnosis Scan 11/67/F Occipital lobe infarction (R) Low-density area R occipital lobe (?), cortical atrophy Predominance R side 12/51/F Occipital lobe infarction (R), maculopathy (R), and aque- Ventricular Predominance R side Occipital lobe infarction (L) Low-density area L occipital lobe, cortical atrophy Predominance 14/57/F Occipital lobe infarction (L) Low-density area L occipital lobe Predominance L side 15/70/F Occipital lobe infarction (L) Low-density area L occipital lobe Predominance L side 16/67/F Occipital lobe infarction (L) Low-density area L occipital lobe Predominance L side 17/64/M Occipital lobe infarction (L) Low-density area L occipital lobe Symmetrical but delayed 18/52/M Occpital lobe infarction (L) Not done Symmetrical but delayed 19/63/M Bilateral Low-density area both occpital lobes, Grossly Patient/Age, yr/Sex ductal stenosis 13/42/F occipital lobe infarctions 20/57/F Bilateral occipital lobe infarctions dilation, R occipital porencephaly L side abnormal wave forms R > L Low-density areas both occipital lobes, Grossly hydrocephalus Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 abnormal wave forms Visual Field Table 2.—Predominance of Component of Visual Evoked Potential for the major positive compo¬ nent (Pno) on the side ipsilateral to the mum ,,, Occipital Lobe Lesions Normal L R Mean difference, µ _0.5 SD_±2.7 20 3.0 -3.3 ±2.0 6 ±2.1 V2~ stimulated side of the visual field.'" They proposed a simple dipole model to explain their findings and located the cortical generators for this compo¬ nent of the response in the posteromedial surface of the occipital lobes. In this model, the scalp electrodes located in the midline and lateral sites in the temporo-occipital region ipsilateral to the stimulated half field are best Fig 2.—Correlation between side of lesion and visual evoked potential predominances in 18 patients and 20 control subjects. Scattergram shows mean differences (PM) in amplitude (microvolts) in patients with right (top) and left (middle) occipital lobe lesions compared with control subjects (bottom). aligned to record from the generators in the contralateral visual cortex. This finding was confirmed in nine healthy subjects and was used to predict the type of response asymmetry to be found in cases of homonymous hemianopsia with full-field stimulation. The full-field response in patients with hemispherectomy and occipital lobectomy have subsequently shown the predicted lateralization of the re¬ sponse to the side of the surgical removal.-'11'2 Thus, unilateral occipital lobe lesions paradoxically leave unaf¬ fected responses recorded on the side of clinical involvement that are gener¬ ated in the intact contralateral occipi¬ tal lobe. The VEP findings in our patients with occipital lobe abnormalities are consistent with this hypothesis. A demonstrable predominance of the P94 component was present over the involved side in 83% of our patients with unilateral occipital lobe lesions and a homonymous hemianoptic visual field defect. Similar findings have been reported by Carlow and Rodri¬ guez13 in studies of posterior cerebral dysfunction with this technique (tran¬ sient pattern response). We confirm a positive correlation between occipital lobe damage, ho- Fig 3.—Left, Right homonymous hemianoptic scotoma, low-den¬ sity lesion of left occipital lobe, and left-side predominance of P9( component of visual evoked potential (case 15). Right, Left homonymous visual field defect, low-density lesion of right occi¬ pital lobe with craniotomy de¬ fect, and right-sided predomi¬ nance of visual evoked potential (case 5). Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 monymous visual field defects, and abnormalities of the VEP, especially a predominance of the P94 component ipsilateral to unilateral lesions. The presence of asymmetry on VEP alone is not prima facie evidence of occipital lobe damage, since normal subjects have asymmetry in some cases. How¬ ever, when this finding is combined with a visual field deficit, the asym¬ metry (with increased amplitude over the involved occipital lobe) is highly suggestive of occipital lobe damage. We caution that our patient popula¬ tion comprised exclusively destructive lesions, and recent reports indicate that irritative lesions may show a reversal of the patterns we de¬ scribe.'3 The abnormal wave forms detected in our two patients with bilateral oc¬ cipital lobe lesions probably arise from grossly disorganized generators in preserved calcarine cortex or visual association areas. The delayed peak latencies and symmetrical appearance of responses in two patients with left occipital lobe lesions is more difficult to explain. Ashworth et al" reported such VEP delays in an autopsy case of bilateral visual pathway abnormality (astrocytoma of corpus callosum) with no anterior visual pathway lesions. The usefulness of the VEP observa¬ tions in this instance seems limited by the fact that single-channel recording was obtained without comparison to control subjects from their laboratory. Nevertheless, the multichannel re¬ cordings of the ß4 component in our two patients (cases 17 and 18) indicate definite delays (110 and 130 ms, respectively), but none of our patients have had complete neuropathologic examination. We cannot exclude sub¬ clinical anterior visual pathway in¬ volvement in our patients. Supplementation of full-field VEP techniques with half-field steady state and/or flash stimulation"-131516 may increase the specificity and accu¬ racy of this technique as a diagnostic test. Nevertheless, it is clear that clin¬ ically useful VEP abnormalities are not confined to the anterior visual pathways, and further study of geniculocalcarine lesions will clarify the role of this technique in evaluating pathologic abnormalities of the poste¬ rior visual pathways. Hyman Menduke, PhD, provided valuable help in the statistical analysis of the data. Miriam Heebner Kogan and Ruth Buley did the art preparation and typed the manuscript. References 1. Halliday AM, Halliday E, Kriss A, et al: The pattern-evoked potential in compression of the anterior visual pathways. Brain 99:357-374, 1976. 2. Blumhardt LD, Barrett G, Halliday AM: The asymmetrical visual evoked potential to pattern reversal in one half field and its significance for the analysis of visual field defects. Br J Ophthalmol 61:454-461, 1977. 3. Asselman P, Chadwick DW, Marsden CD: Visual evoked responses in the diagnosis and management of patients suspected of multiple sclerosis. Brain 98:261-282, 1975. 4. Celesia GG, Daly RF: VECA: A new electrophysiologic test for the diagnosis of optic nerve lesions. Neurology 27:637-641, 1977. 5. Holder GE: The effects of chiasmal compression on the pattern visual evoked potential. Electroencephalogr Clin Neurophysiol 45:278-280, 1978. 6. Riddoch G: Dissociation of visual perception due to occipital injuries with especial reference to appreciation of movement. Brain 40:15, 1915. 7. Vaughan HG, Katzman R, Taylor J: Altera- tions of visual evoked responses in the presence of homonymous visual defects. Electroencephalogr Clin Neurophysiol 15:737-746, 1963. 8. Kooi KA, Guvener AM, Bagchi BK: Visually evoked responses in lesions of the higher optic pathways. Neurology 15:841-854, 1965. 9. Oosterhuis HJGH, Ponsen L, Jonkman EJ, et al: The average visual response in patients with cerebrovascular disease. Electroencephalogr Clin Neurophysiol 27:23-34, 1969. 10. Barrett G, Blumhardt LD, Halliday AM, et al: A paradox in the lateralization of the visual evoked response. Nature 261:253-255, 1976. 11. Blumhardt LD, Barrett G, Halliday AM, et al: The effect of experimental 'scotomata' on the ipsilateral and contralateral responses to pattern reversal in one half field. Electroencephalogr Clin Neurophysiol 45:376-392, 1978. 12. Hoeppner TJ, Bergen D, Morrell F: Visual evoked potentials after partial occipital lobecto- Downloaded From: http://archneur.jamanetwork.com/ by a Oakland University User on 06/09/2015 my, abstracted. Electroencephalogr Clin Neurophysiol, to be published. 13. Carlow TJ, Rodriguez M: Localization of cerebral dysfunction by visual evoked responses, abstracted. Ann Neurol 4:176, 1978. 14. Ashworth BA, Maloney FJ, Townsend RA: Delayed visual evoked potentials with bilateral disease of the posterior visual pathways. J Neu- rol Neurosurg Psychiatr 41:449-451, 1978. 15. Celesia G, Kuroiwa Y: Quantification of transient and steady-state evoked potentials during hemifield stimulation and its significance for the detection of visual field defects, abstracted. Electroencephalogr Clin Neurophysiol, to be pub- lished. 16. Wildberger HGH, van Lith GHM, Wijngaarde R, et al: Visually evoked cortical potentials in the evaluation of homonymous and bitemporal visual field defects. Br J Ophthalmol 60:273-278, 1976.