• EMG
• GUILLAIN-BARRE SYNDROME
• MRI
• NEUROMUSCULAR
• ULTRASOUND

In early Guillain-Barré syndrome (GBS), including in its axonal forms, the most frequent electrophysiological findings are abnormalities of late responses (H reflex and F response) pointing to dysfunction in the proximal segments of the peripheral nerves.1 Ultrasonography and MRI are the imaging techniques of choice in any pathology of the peripheral nerve trunks.2 We report nerve imaging findings in a patient with early axonal GBS.

An 18-year-old woman was admitted with a 1-day history of lower limb weakness and calf myalgias. Two weeks prior, she had an episode of profuse diarrhoea of 4-day duration attended at home by her family physician; stool culture was not carried out. Initial examination revealed a waddling gait, weakness of proximal lower limb muscles mainly involving ilio-psoas (Medical Research Council grade 4/5), lower limb areflexia and absence of sensory loss. Upper limb reflexes were brisk. In the next few days the patient's lower limb weakness progressed, so that on day 6 after onset she was unable to walk, her muscle power being 2–3/5 proximally and 3–4/5 distally. Treatment with standard intravenous immunoglobulin (2 g/kg in 5 days) was administered. There has been progressive improvement; 3 months after onset, examination revealed mild waddling gait, paresis of hip abductors/flexors (3–4/5) and foot dorsiflexors (4/5), and lower limb areflexia. No serum IgG antiganglioside antibodies were detected. Campylobacter jejuni serology was not available. Cerebrospinal fluid showed a protein content of 70 mg/dL and no cells.

Sequential nerve conduction studies were performed on days 4, 12 and 42 after onset. On day 4, the only positive finding was the absence of H waves, the remaining conduction parameters of median, ulnar, peroneal, tibial, sural and superficial peroneal nerves being normal (figure 1A). On days 12 and 42, there was a marked amplitude reduction of compound muscle action potentials (CMAPs) of the tibial nerve (figure 1B) and, to a lesser degree, of the peroneal nerve; nerve conduction parameters in other explored nerves were normal. Electromyography of tibialis anterior showed denervation potentials at the latest examination.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

(A) Tibial nerve motor conduction velocity (MCV) study on day 4 after onset showing normal compound muscle action potential (CMAP) morphology, both on distal (ankle to abductor hallucis (AH) muscle) and proximal stimulation (popliteal fossa (PF) to AH); proximal CMAP amplitude is 14.4 mV, and distal CMAP amplitude is 18.9 mV (normal ≥3); MCV is 47 m/s (normal ≥41). (B) On day 12, note severe CMAP amplitude reduction to 1.7 mV, both on distal and proximal stimulation. Comparatively with the previous study, at this stage there was a mild slowing of MCV (passing from 47 to 39.4 m/s) and prolongation of F-wave latency (55.5 ms; normal ≤55), which seems to be proportional to the observed CMAP reduction. Postcontrast sagittal (C) and coronal (D) T1-weighted, fat-saturation MRIs of the lower thoracic and lumbosacral spine, performed on day 4, showing diffusely thickened cauda equina (arrowheads), which in the axial image (L1 level) selectively involves the anterior roots (E, arrows); CM indicates conus medullaris. (F) Sagittal sonogram of the right ventral rami of C6–C7 cervical nerves (callipers) performed on day 3; note their characteristic homogeneous hypoechoic texture with partial loss of the surrounding perineural hyperechoic rims. (G) Short-axis sonogram of the right ventral ramus of the C7 spinal nerve showing the cross-sectional area (dotted green tracing) measuring 27.47 mm² (control 12.29±5.3 mm²).2 Note blurred margins and the absence of the physiological hyperechoic rim. Asterisks indicate the posterior tubercle of the seventh transverse vertebral process.

MRI of the lumbosacral spine showed thickening of the cauda equina (figure 1C–E). Ultrasonography showed normal findings both in upper limb nerves (ulnar, median and radial) and lower limb nerves (peroneal, tibial and sural). Conversely, sonograms of the ventral rami of the fifth to seventh cervical spinal nerves showed widespread changes (figure 1F, G).

The patient presented with paraparetic GBS,3 which could be categorised within acute motor axonal neuropathy (AMAN) with negative serology for gangliosides. The initial electrophysiological test, on day 4 after onset, revealed completely normal findings with the exception of absent H waves, a pattern that may occur in 16% of early GBS cases.1 The selective absence of late responses, either H reflex or F waves, in the early stage of illness, has been interpreted as a dysfunction in the proximal nerve segments caused by demyelinative conduction block, Wallerian degeneration, conduction failure at the nodes of Ranvier, or impaired excitability of the motor neuron or proximal axon.1 On serial examination, marked CMAP amplitude reduction of tibial nerve appeared, a feature giving support to the notion that the electrophysiological classification of GBS would be better determined based on sequential findings rather than only on the results of an initial study.

Our MRI study corroborates that in pure motor GBS, there may be a characteristic enhancement of only the anterior nerve roots (see ref 2 for further details). In the seminal series of 50 patients with autopsied GBS, with 32 of them having died between 2 and 10 days after symptomatic onset, Haymaker and Kernohan described the observed topography of lesions as follows:4 “As a whole, the observed pathological changes were usually more prominent in the region where the motor and sensory roots join to form the spinal nerve”. Intriguingly, in AMAN, most of the initial lesions are in the spinal roots, rather than in peripheral nerves; in some cases the process of Wallerian-like degeneration being at a more advanced stage in the roots than in the nerves.5 In keeping with these pathological studies, we have reported that nerve ultrasonography in early demyelinating and axonal GBS may show selective and frequent changes in the ventral rami of spinal cervical nerves, consisting of blurred boundaries, nerve enlargement or both.2 Our imaging study indicates that in early AMAN, outstanding changes may also occur in anterior spinal roots and spinal nerves, though such changes might be, to some degree, subclinical.

In early GBS, the brunt of pathology, consisting of inflammatory oedema, may rely on spinal nerves.2 ,4 Starting from the observed imaging features, this notion is applicable to the patient with current AMAN. Conversely to intrathecal spinal roots, covered by a multicellular root sheath derived from the arachnoid, spinal nerves possess epiperineurium, rendering these structures susceptible to increase in endoneural pressure causing constriction of transperineurial vessels, which diminishes nerve blood flow facilitating nerve injury.2 The consequences of such early pathological events would be focal nerve conduction failure resulting in, as described here, selective absent H waves. Interestingly, electrical root stimulation may reveal early attenuated M responses in GBS cases showing preserved conventional nerve conduction studies (see ref 2 for further details).

We conclude that in early axonal GBS, electrophysiology may demonstrate selective absence of H waves, which correlates with the observed imaging enlargement of intrathecal anterior spinal roots and spinal nerves.