The “split-leg” syndrome in ALS: specific or variable?

The “split hand” in amyotrophic lateral sclerosis (ALS), first described by Wilbourn (1), is characterized by predominant atrophy and weakness of the thenar muscles, with relative sparing of the hypothenar muscles. This clinical observation was confirmed by neurophysiological studies, evaluating compound muscle action potential (CMAP) amplitude and the number of motor units (MUs) (2–4). A split-hand index has been proposed (4), with a cutoff value for the differential diagnosis between ALS versus non-ALS. The physiological basis for the split-hand phenomenon is not clear, but both cortical and peripheral mechanisms may be important (5). More recently, a “split-leg phenomenon” has been described, with a greater reduction in the number of MUs and CMAP in plantar flexors compared with dorsiflexors (6). However, subsequent studies have described predominant dorsiflexor weakness, contrary to the original report (7). We here describe an ALS patient with “splitleg” features in one leg, but with an inverted splitleg phenomenon in his opposite leg, implying a measure of inconsistency of the predominant pattern of distal lower limb (LL) involvement. A 63-year-old man followed in our ALS clinic in Lisbon for progressive LL weakness and atrophy presented 6 years ago with right foot-drop. LL weakness progressed proximally, with fasciculation and cramps in the legs, without pain or sensory loss. He is now able to walk only with foot orthoses and a cane. He has no bulbar, respiratory, or upper limb (UL) symptoms. On recent examination, proximal (MRC 4) and distal (MRC 0-2) segments of the LLs were weak and atrophic. In the right leg plantar flexion was MRC 2, but foot dorsiflexion was absent. In the left leg foot dorsiflexion was MRC 2, but plantar flexion was absent. The knee jerks were brisk (þþþ). The right ankle jerk was just present, but the left ankle jerk and both plantar responses were absent. Nociceptive and large fiber-dependent sensibility in LLs was normal. In the ULs there was no weakness or atrophy, but fasciculations were present in proximal and distal muscles bilaterally, and the tendon reflexes were brisk (þþþ). A Hoffman sign was detected on the right side. The bulbar region was normal. Brain and spinal-cord MRI and CSF analysis were unremarkable. Laboratory investigation excluded autoimmune disorders, and respiratory tests have been repeatedly normal. Current electrophysiological investigation shows normal motor conduction velocities were normal in ULs (both ulnar nerves), and normal sensory potentials in UL and LLs (ulnar, sural, and peroneal nerves, bilaterally). LL motor conduction studies are summarized in Table 1. Needle electromyography (EMG) of the ULs confirmed the presence of fasciculation potentials in proximal (biceps) and distal (first dorsal interosseous) muscles, associated with chronic neurogenic changes in first dorsal interosseous muscles. In the LLs, frequent complex fasciculation potentials and very unstable MUs were detected in both vastus medialis muscles, with marked neurogenic changes. Fibrillation and sharp-waves were detected in tibialis anterior and gastrocnemius medialis bilaterally, but motor unit potentials were not recruited in right tiabilis anterior, extensor digitorum brevis, or in left gastrocnemius medialis and abductor hallucis. However, a few very polyphasic and unstable motor units with very high