Chiasmal lesions

The principal cause of chiasmal dysfunction is pituitary tumor, the anatomical relationship between the optic chiasm and the pituitary gland making the chiasm susceptible to compression by lesions expanding from the pituitary fossa. The classic triad of neuro-ophthalmic signs in pituitary tumors of reduced visual acuity, visual field defects, and optic atrophy arises from suprasellar tumor extension with resulting compression of the chiasm. Other types of tumor, aneurysm, inflammation, demyelination, and trauma can also affect chiasmal function. A bitemporal hemianopia is the classic visual field defect due to a disturbance of the decussating fibers from the nasal retinae but occurs in less than 50% of patients with pituitary tumors and visual loss.^42,78 Other types of visual field defect can result, including central scotoma. Approximately 13% of patients present with unilateral visual loss.²⁵

The investigation of choice in patients with suspected chiasmal compression is neuroradiology, either magnetic resonance imaging (MRI) or high-resolution computed tomography (CT) scanning. There are probably two roles for electrophysiological testing. The first is in the initial assessment and diagnosis of patients with visual symptoms; the delayed or misdiagnosis of chiasmal dysfunction can result in severe irreversible visual loss, and it is therefore of critical importance that the correct diagnosis be reached promptly. Reports of misdiagnoses in the literature include (atypical) retrobulbar neuritis, glaucoma, cataract, hysteria, macular degeneration, refractive error, choroidal sclerosis, and vascular lesions.^31,41,59,65 The second role for electrophysiology is in the follow-up and management of patients with radiologically confirmed lesions that might or might not show suprasellar extension and signs of visual pathway dysfunction and encompasses postoperative monitoring.

The likely involvement of the crossing fibers enables the use of hemifield stimulation in the visual evoked potential (VEP) assessment of chiasmal function, but adequate consideration of registration parameters is critical to VEP interpretation. A hemifield VEP abnormality may be the most sensitive electrophysiological index of early chiasmal involvement, but some patients with reduced visual acuity have difficulty in maintaining accurate fixation, and it might not be possible to perform hemifield stimulation adequately in all patients. Full-field stimulation also gives accurate localization of chiasmal lesions but is slightly less sensitive.^15,28 Multichannel recording is indicated; assessment of chiasmal function should not be attempted with a single midline channel.

An understanding of the results of hemifield pattern stimulation in normal individuals is important to accurate interpretation of the electrophysiological abnormalities in chiasmal dysfunction. Use of a large hemifield stimulus, for example, greater than a 12-degree radius, gives “paradoxical” lateralization of the normal P100 component of the pattern-reversal VEP ipsilateral to the stimulated hemifield.⁶ There is a contralateral N105/P135 complex. However, as the size of the stimulus field is progressively reduced, the P100 firstly becomes bilateral in distribution and then contralateral with a small hemifield stimulus (e.g., 2.5-degree radius³⁶). Similar changes occur in patients with hemifield defects³⁹ (see below). In general, when Fz is used as a reference, a small-field, small-check stimulus will show anatomical lateralization, whereas a large-check, large-field stimulus will show paradoxical lateralization. Bipolar recordings using ipsilateral hemisphere reference electrodes do not show paradoxical lateralization with any stimulus parameters, and the contribution to paradoxical lateralization of the signal recorded via the Fz “reference” is thus apparent.

Following the initial report by Muller⁵² that the flash VEP (FVEP) could be of abnormal latency in chiasmal dysfunction, other workers noted that the maximum FVEP abnormality was localized contralateral to the visual field defect.^27,43,48,73 The first reports using contrast stimuli appeared in 1976. Van Lith's group⁷⁶ used both full-field and hemifield steady-state (8 Hz) stimulation in six patients with bitemporal hemianopia due to tumor, and found both phase and amplitude abnormalities contralateral to the stimulated eye.

The first detailed report of transient pattern VEP (PVEP) was that of Halliday's group.³⁴ Using a 16-degree radius, 50-minute check stimulus, they found markedly asymmetrical scalp distribution in ten patients with chiasmal dysfunction. In particular, they described the “crossed” asymmetry typical of chiasmal lesions in which the findings from one eye are more abnormal over one hemisphere but the distribution of abnormality changes such that findings from the fellow eye are more abnormal over the other hemisphere. Unexpectedly, the maximum abnormality was localized ipsilateral to the visual field defect, that is, the “paradoxical” lateralization referred to above (figure 78.1). PVEP abnormalities were present from some eyes with normal (kinetic) visual fields. The findings were contrasted to those in demyelination, in which preservation of waveform, a generally greater latency delay, and symmetry across the scalp were much more frequent. The use of hemifield stimulation was further elaborated in another publication by the same group.⁸

Figure 78.1.  

Crossed VEP asymmetry in a 48-year-old male with bitemporal hemianopia from a suprasellar mass (16-degree radius, 50-minute checks). A, With full-field stimulation, the normal P100 component is recorded over the right hemisphere when the left eye is stimulated and over the left hemisphere when the right eye is stimulated, that is, contralateral to the impaired temporal visual field and showing the phenomenon of paradoxical lateralization. B, The use of hemifield recording demonstrates that the full-field responses reflect preservation of the responses to the preserved nasal fields. (From Halliday AM, Barrett G, Blumhardt LD, et al: The macular and paramacular subcomponents of the pattern evoked response. In Lehman D, Calloway E (eds): Human Evoked Potentials: Applications and Problems. New York, Plenum, 1975, pp 135–151. Used by permission.)

Holder³⁸ confirmed this “crossed” asymmetry in ten patients, but when full-field stimulation (11-degree full-field, 26-minute checks, bipolar recording) was used, the maximal PVEP abnormality was always contralateral to the stimulated eye (figure 78.2). Although apparently contradictory, these findings are in fact consistent with those of Halliday's group, the alternate abnormality lateralization reflecting the use of a smaller stimulating field/check size (see above). The abnormality lateralization was enhanced with a 4-degree radius, 13-minute check stimulus. It was confirmed that the asymmetrical scalp distribution was atypical for demyelination and that abnormal VEPs could occur in eyes with full visual fields. Equally, normal PVEPs could occur in eyes with field defects. Latency delays were a frequent occurrence.

Figure 78.2.  

Crossed VEP asymmetry in a 48-year-old male with a nonfunctioning chromophobe adenoma showing crossed asymmetry of the VEPs. Use of a small field (11 degrees), small check stimulus (26 minutes), gives “anatomical” rather than “paradoxical” distribution of abnormality. The preoperative findings from the right eye show increased latency in the left hemisphere traces, in keeping with dysfunction of the decussating fibers from the right eye to the left hemisphere. Preoperative stimulation of the left eye shows an overall longer latency compared to the right eye, in keeping with a degree of optic nerve dysfunction; the right hemisphere traces are more abnormal than the left, in keeping with the chiasmal compression. Note the improvement after surgery such that the right eye findings no longer show any abnormality and the left eye findings now show no delay and less interhemispheric asymmetry.

Those findings were extended in a study of 34 patients with histologically confirmed nonfunctioning chromophobe adenomas.⁴¹ The PVEP results were compared with clinical, radiological, and surgical findings. There were four eyes with normal PVEPs; one had a full visual field, one had a paracentral scotoma, and two had superior temporal quadrant defects. It is of interest that FVEPs in the latter two eyes were abnormal. Full fields but abnormal PVEPs occurred in two eyes. The PVEPs often indicated marked functional asymmetry when the neuroradiology (CT scan) suggested symmetrical midline suprasellar extension. The PVEPs were usually more sensitive than the conventional clinical tests of visual acuity and visual fields.

A number of other studies reported PVEP findings in chiasmal dysfunction, mostly (those using multichannel recording techniques) confirming the “crossed” PVEP asymmetry to be pathognomonic of chiasmal dysfunction but describing clinical and electrophysiological findings in varying degrees of detail.^{13–16,28,32,33,51,55,58,68,71} Gott and colleagues examined 83 patients with tomographically demonstrated pituitary tumors.³² Most were intrasellar and had normal fields and PVEPs. Suprasellar extension was radiologically demonstrated in 12 cases; all had abnormal PVEPs, but visual fields were normal in eight patients. The abnormality was usually an increased P100 latency, but asymmetrical scalp amplitude distribution was also observed (22-degree full field, abnormality ipsilateral to the field defect).

The ability of the PVEP to influence management was noted by Stark and Lenton,⁶⁸ who cite one case with a radiologically confirmed pituitary tumor but unreliable clinical testing in which an abnormal PVEP prompted surgical intervention. Haimovic and Pedley³³ found a delayed P100 (19 × 13.5-degree hemifield, 31-minute checks, abnormality ipsilateral to the field defect) in one of 15 patients with hemifield stimulation but in four patients when full-field stimulation was used. This illustrates the difficulty in component identification with large-field stimuli, which can lead to spurious “delays.” Blumhardt⁷ forcefully argued this point. Others concluded that the VEP was not a suitable means of detecting subtle field defects following a study of eight patients⁵¹ with 5-degree hemifield, 50-minute checks: two patients who were normal, four with ipsilateral abnormality, and two with no lateralization. This failure to reveal abnormalities may relate to the choice of stimulus parameters and emphasizes the importance of this factor. The two patients with normal PVEPs were presumably postoperative because the visual field defects had “resolved.” There was, however, subjective desaturation to red. Flanagan and Harding²⁸ carefully examined the effects of various stimulus parameters in nine patients with pituitary tumors; hemifield stimulation with a large-check, large-field stimulus was more sensitive than full-field stimulation in the early detection of chiasmal dysfunction. This observation was later confirmed.^13,15

Optimal use of medical therapy, such as bromocriptine, for pituitary lesions is aided by a sensitive, objective assessment of chiasmal function. Wass et al.⁷⁴ first described PVEP improvement during bromocriptine therapy in patients with large pituitary tumors but did not supply full details. Pullan and colleagues examined hemifield PVEPs in five nonfunctioning and five functioning tumors (prolactinomata) before and after bromocriptine treatment.⁶⁰ Suprasellar extension on CT scan was a criterion for patient selection. All patients with radiological evidence of tumor shrinkage showed PVEP improvement, as did one patient without evident radiological change. The author's laboratory has also monitored patients with intrasellar lesions (unpublished data). Changes in the VEP may be the first indicator of functional involvement of the chiasm, preceding field loss, and thus precipitate a change from medical to surgical management. Serial postoperative VEP recording can also monitor the functional state of the optic nerves and chiasm in a patient following tumor excision (figure 78.3) and may help detect tumor recurrence prior to deterioration in visual fields or acuity. Similarly, some patients decline surgery when offered, and additional objective evidence of increasing visual pathway dysfunction may help them to reconsider.

Figure 78.3.  

Serial VEPs in a patient with recurrence of a nonfunctioning chromophobe adenoma (11-degree full-field stimulus; 26-minute checks). The patient was aware of the recurrence but declined further surgical intervention. Initial findings from the right eye show a P100 component that is markedly delayed and is better seen in the ipsilateral hemisphere traces than the right in keeping with the lateralization expected with a small-field, small-check stimulus. The PVEP had become undetectable by March 1984 but without change in visual acuity. Right eye visual acuity dropped to 6/60 approximately one year later. The initial findings from the left eye show a well-formed PVEP in the ipsilateral hemisphere traces but marked abnormality in the right hemisphere traces in keeping with dysfunction of the decussating chiasmal fibers. The latency of the P10 component in the ipsilateral hemisphere traces increases by ∼20 ms over a 4-year period with no deterioration in visual acuity. Note the continuing interhemispheric asymmetry. Visual fields were abnormal throughout but showed no significant deterioration. Neuroradiological investigation (CT scan) showed tumor expansion during the period of follow-up.

VEP recording has also been used to monitor chiasmal function during surgery.^{2,17,20,26,54,61,77} There is no consensus in relation to the contribution of intraoperative recording to surgical outcome. There are inevitable limitations of the technique owing to the need for diffuse flash stimulation. Chiasmal hypoplasia or aplasia can also be detected by using VEP techniques.^4,70

The PERG has been suggested to be a useful prognostic indicator for visual outcome in the preoperative assessment of optic nerve compression in pituitary tumor.^45,63 That has been confirmed in the author's laboratories.⁵⁷ An abnormal PERG correlates with a lack of postoperative recovery, presumably by demonstrating significant retrograde degeneration to the retinal ganglion cells.

The VEP is also of major importance in the demonstration of abnormal chiasmal routing in patients suspected of albinism.²³ That issue is addressed elsewhere in this volume (see chapter 25).