Blink reflex recovery in facial weakness

Methods.

Seven healthy volunteers (age range, 35 to 60 years) and 12 patients with a history of Bell’s palsy (age range, 33 to 67 years) gave written informed consent, in accordance with the Declaration of Helsinki, for the study protocol, which was approved by the institutional review board. The degree of facial weakness was graded as 0, 1, 2, 3, 4, or 5, according to the House–Brackmann grading system.7 At the time of the study all patients were divided into two groups: 1) patients demonstrating complete or almost complete recovery of facial weakness on clinical examination (House–Brackmann grade 0 or 1); and 2) patients demonstrating persistent facial weakness on clinical examination (House–Brackmann grade 2 or higher).

None of the patients demonstrated grade 4 or 5 weakness. All patients were diagnosed with Bell’s palsy by a neurologist who evaluated for possible alternative underlying conditions.

Patients were seated comfortably in a quiet room during the study and were asked to look at given reading materials during the test to maintain a relatively uniform level of alertness. A Counterpoint electromyograph (Dantec Medical, Inc., Allendale, NJ) was used for all studies. Direct orbicularis oculi compound muscle action potentials were obtained by stimulation of the facial nerve anterior to the mastoid process. Blink reflexes were evoked by electrical stimulation of the supraorbital nerve using a bar electrode. The cathode was placed over the supraorbital notch, and the anode was placed 3 cm superiorly and laterally with adjustment as necessary to minimize stimulus artifact, and was then taped securely. The R2 threshold stimulus intensity and the sensory perceptual threshold stimulus intensity were measured. The R2 threshold intensity was defined as the stimulus intensity necessary to evoke a consistent bilateral 50-μV R2 response. The sensory threshold for the supraorbital nerve was defined as the minimum stimulus intensity detected by the patient on at least three of six trials. The blink reflex was evoked with a 0.2-msec square-wave pulse at three times the R2 threshold intensity.6

Blink responses were obtained using paired pulses delivered at the following interstimulus intervals: 160, 300, 500, 700, and 1,000 msec. These intervals were pseudorandomized in blocks of five, with a minimum of eight blocks recorded on each side. To prevent habituation, consecutive paired pulses were delivered at an interval of 20 to 25 seconds. The blink reflexes were recorded from orbicularis oculi muscles bilaterally using tin-plated electrodes with shielded cables. The high-frequency filter setting was 1,000 Hz and the low-frequency filter setting was 30 Hz. The active recording electrode was placed on the lower orbicularis oculi muscle directly below the pupil, and the reference recording electrode was placed laterally on the temple.8 The evoked responses were sampled at a frequency of 10 kHz using a Macintosh computer with LabView software (version 4.0) and boards (National Instruments Corporation, Austin, TX). Peak area and amplitude of the R2 response was measured within a window from 32 to 90 msec6,9 from an average of eight rectified traces.

The R2 amplitude and area were expressed as a ratio of the second R2 (R2_b) to the first R2 (R2_a). This R2_b-to-R2_a ratio was plotted against the interstimulus intervals to construct a blink reflex recovery curve. The blink reflex recovery curves were compared between patients and control subjects, and within patients between those that had recovered strength (House–Brackmann grade 0 or 1) and the group with persistent weakness (House–Brackmann grade 2 or 3).

Statistical analysis consisted of analysis of variance to determine the effects of groups (patients versus control subjects, residual weakness versus recovered strength), side of stimulation (stimulation of the supraorbital nerve on the paretic side versus the nonparetic side), and side of recording (electromyographic [EMG] activity from the paretic versus the nonparetic side). A log transformation of the R2_b-to-R2_a ratio was performed to normalize the distribution.9,10 Paired comparisons using f-tests with Bonferroni corrections were used for post hoc comparisons of individual intervals.

Results.

Of the 12 patients with a history of Bell’s palsy, six had recovered from their weakness (mean age, 54.3 years) and six were left with residual weakness (mean age, 57.2 years). All patients were tested at least 3 months after Bell’s palsy. None of the patients showed clinical signs of blepharospasm or hemifacial spasm. There was no significant side-to-side difference in the threshold stimulus intensity to evoke an R2 response in control subjects or Bell’s palsy patients. There was no side-to-side difference in the threshold stimulus intensity for sensory perception in control subjects or Bell’s palsy patients.

In patients as well as control subjects, paired pulse stimulation produced suppression of the R2 component of the second blink reflex (R2_b). Suppression was maximal at the shortest interstimulus interval tested and recovered gradually. Patients with residual facial weakness had less suppression of the R2_b compared with normal control subjects. This is illustrated for a normal control subject and a patient with persistent facial weakness (figure 1). In contrast, patients who recovered strength were not different from control subjects. Recovery curves (figure 2) for the three groups were obtained by pooling all recordings at each interval for each person and plotting the means of subject means. Recovery curves from patients with persistent facial weakness (House–Brackmann grade 2 or 3) were significantly different from control subjects (p < 0.001). On post hoc analysis this difference was found to consist of enhanced recovery of the R2_b at the shorter interstimulus intervals. The mean R2_b-to-R2_a ratio at an interstimulus interval of 160 msec was 0.55 for patients with residual weakness versus 0.18 for control subjects, and at 300 msec was 0.58 for patients with residual weakness versus 0.33 for control subjects. At interstimulus intervals longer than 300 msec, the patient’s R2_b-to-R2_a ratio was not different from the control subjects. Recovery curves of patients who recovered facial strength (House–Brackmann grade 0 or 1) were not significantly different from control subjects.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Blink reflex recordings using paired electrical stimulation of the supraorbital nerve at selected interstimulus intervals from a normal control subject (A) and a Bell’s palsy patient with residual weakness (B). Dotted lines show the portion of the trace that was used to calculate the R2 area.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Blink reflex recovery curves (mean and SEM) for normal control subjects (•) and Bell’s palsy patients with (▪) or without (▴) recovery of facial strength.

In patients with persistent weakness, the enhanced R2 recovery occurred in facial muscles of the paretic as well as the nonparetic side, but was greater on the paretic side (p = 0.04).

The side of stimulation also affected the blink reflex recovery curves of patients with persistent facial weakness. The recovery curves obtained by stimulation of the trigeminal nerve on the paretic side were significantly different from those obtained by stimulation of the trigeminal nerve on the nonparetic side (p = 0.03; figure 3). The mean R2_b/R2_a value for each interstimulus interval was higher when stimulating the paretic side. Although the more stringent criteria for post hoc analysis did not allow us to pinpoint individual intervals with enhanced recovery, the curves appear most different at the shortest intervals. For example, the difference between the mean R2_b-to-R2_a ratio at an interstimulus interval of 160 msec was 0.35 versus 0.26 when stimulating the paretic versus the nonparetic side (p = 0.034), and at 300 msec the mean R2_b-to-R2_a ratio was 0.42 versus 0.26 (p = 0.031). Stimulation of either side produced a significantly higher R2_b-to-R2_a ratio at an interstimulus interval of 160 msec compared with control subjects (p = 0.0001); but within the group of patients with residual weakness, the magnitude of the enhanced recovery exhibited a dependence on the side of stimulation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3. Comparison of blink reflex recovery curves (mean and SEM) for patients with residual weakness after Bell’s palsy with stimulation of the supraorbital nerve on the weak side (▪) and the unaffected side (•).

Discussion.

We found enhanced recovery of the blink reflex in patients with unilateral facial weakness. Earlier blink reflex recovery was recorded in the orbicularis oculi muscles bilaterally and was more prominent when evoked by stimulating the supraorbital nerve on the paretic side compared with stimulating the supraorbital nerve on the normal side. This trend toward lateralization of the enhanced excitability is most easily explained on the basis of alterations in the sensory pathways and premotor circuits mediating the R2 component of the blink reflex. Studies11,12 of patients with brainstem lesions indicate that the R2 component of the blink reflex is produced through polysynaptic pathways in the lower brainstem. The afferent loop of the R2 component of the blink reflex is mediated via the supraorbital branch of the trigeminal nerve, with inputs relayed to the spinal trigeminal nucleus ipsilateral to the side of stimulation and subsequently to bulbopontine R2 interneurons. These project bilaterally to the facial motor nucleus in the pons. Thus several anatomic loci are candidate sites for plastic changes. The enhanced sensitivity of the blink reflex pathway is probably not mediated at the level of the trigeminal afferents because we did not find a significant difference in either the magnitude of the R2 threshold or the sensory threshold between the supraorbital nerve on the paretic and the normal side. Although there was somewhat greater enhancement of R2 recovery in paretic facial muscles, the enhancement occurred bilaterally in facial muscles. Because the enhanced blink reflex recovery occurs to a greater extent when stimulating the affected side, we conclude that the site of sensitization of the blink reflex is most likely to occur within the trigeminal complex or the R2 interneurons rather than both facial nuclei.

This finding of enhanced blink reflex excitability is consistent with previous electrophysiologic studies. Valls–Solé13 found a larger R2 response on the paretic side of patients recovering from facial palsy, suggesting hyperexcitability of either facial motoneurons or their inputs. Contralateral R1 responses have also been found in patients with Bell’s palsy and may reflect unmasking of preexisting trigeminofacial reflex pathways.14,15

Adaptive gain modification occurs when the relationship between the magnitude of a stimulus and the amplitude of a reflex response is changed to compensate for a disturbance of the amplitude of the reflex response. Previous kinematic studies also indicated that unilateral facial weakness causes adaptive enhancement of the blink reflex. Studies of blink main sequence slope (relationship of blink peak velocity versus blink amplitude) and eyelid peak velocities during blink in patients recovering from facial weakness suggest that the adaptive gain mechanisms are bilateral.1,16 Studies of the eyelid peak velocity versus amplitude in monkeys following unilateral weakness induced by botulinum toxin also found bilateral enhancement of the main sequence slope. Additional studies to assess the etiology of this enhancement in rabbits used isotonic weights to impede eyelid closure. Weights produced an increase in the force and EMG activity generated by the orbicularis oculi. Impeding eyelid closure to different degrees showed that the final position of the eyelid at the end of a blink was the driving factor for enhanced orbicularis oculi force, rather than a specific amplitude of blink.17

In the current study, enhanced blink excitability was limited to patients with residual weakness and did not occur in the group of patients who recovered from their facial palsy. This suggests that the enhanced excitability of the blink reflex recovery is dependent on ongoing weakness. The feedback mechanism is not clear. The supraorbital nerve has been shown to have a critical role in maintaining the adaptive gain modification of the blink in animal models. Rabbits exhibiting normal adaptation when their blinks were impeded lost the ability for adaptive gain modification after sectioning of the supraorbital nerve.17 The lack of eyelid closure caused by weakness itself may provide abnormal afferent signals that maintain abnormal gain of the blink reflex. Alternatively, sensory input from dry or inflamed eyes may cause enhancement of the blink reflex in certain situations leading to abnormal eye closure.

Inputs from the cerebral cortex and basal ganglia are known to modulate the blink reflex. Cortical input has been implicated by the findings of depressed unilateral and bilateral R2 components following focal hemispheric lesions.18 Diseases of the basal ganglia affect the excitability of the blink reflex. Huntington’s disease is associated with increased habituation of the blink reflex, whereas in PD the blink reflex is hyperactive. Enhanced sensitivity of the cortex or basal ganglia is an unlikely cause for the enhancement of the blink reflex noted in the current study because supranuclear sensitization would not be expected to result in a lateralized sensitivity of the blink reflex. Additionally, the adaptive gain modification in decerebrate rats was similar to healthy alert rats.17

Blepharospasm has been proposed to be caused by an abnormal excitatory drive from the basal ganglia to the facial and other motor brainstem nuclei. This hypothesis is supported by the presence of enhanced R2 recovery in patients with blepharospasm.6,9,19,20 Although some cases of blepharospasm may be secondary to lesions in the brainstem or basal ganglia, the majority of cases of blepharospasm are idiopathic. The etiology of the increased excitatory output of the facial motor neurons in these patients is not clear. It is clear that the vast majority of patients with blepharospasm do not have a clinically obvious facial palsy to account for their dystonia, but subtle abnormalities and maladaptive responses involving the afferent and efferent pathways may be important in the etiology of some dystonias.

Recent work on rats21 proposed a two-step process underlying the pathophysiology of blepharospasm. In the first step, mild striatal dopamine depletion reduced the inhibition of the blink reflex. An additional lesion, slight weakening of the facial nerve, caused blepharospasm. Either of the two lesions alone caused increased excitability of the blink reflex, but spontaneous blepharospasm was induced only with the combination of the two lesions. The development of blepharospasm in only a small fraction of patients with facial weakness2,22 might be explained on the basis of a relative dopaminergic deficiency, perhaps caused by an underlying acquired insult or a genetic susceptibility. There may be similar mechanisms operative in at least some of the cases of essential blepharospasm in humans. The contribution of a dopaminergic deficit in producing blepharospasm can also be inferred from the normalization of the blink reflex recovery curve in patients with essential blepharospasm after the infusion of apomorphine.23

In four patients with blepharospasm22 there was a disparity in the main sequence relationship between the two eyelids, suggesting divergent influence on the blink reflex as a result of conflicting adaptive needs in a patient with both blepharospasm and Bell’s palsy. Although kinematic studies cannot distinguish whether adaptation occurs on the sensory or motor side of the reflex, it is interesting that patients with blepharospasm are noted frequently to display enhanced sensitivity to sensory stimuli such as light. Light is a frequent trigger for blepharospasm. Abnormal ophthalmologic symptoms were present in 57% of 272 patients with blepharospasm. Inflammation of the lid margin and the Meibomian gland margin is diagnosed frequently in early blepharospasm. Some blepharospasm patients with local ocular pathology experience resolution of their eyelid spasms after treatment of their local ocular pathology.

Given the alternative possibilities that the increased excitability of the blink reflex in Bell’s palsy could be caused by enhanced supranuclear drive or by abnormal sensory inputs, we would favor the latter explanation in view of the relative lateralization to the afferents on the paretic side. Abnormal sensory input could be caused by eyelid weakness causing incomplete eye closure or the dry, inflamed eyes that result from corneal exposure.