The Lewis - Sumner Syndrome
Richard A. Lewis, MD and Arthur K. Asbury, MD, FRCP

In 1982, a paper was published in the journal 'Neurology' entitled "Multifocal demyelinating neuropathy with persistent conduction block", and was authored by Richard A. Lewis, Austin J. Sumner, Mark J. Brown, and Arthur K. Asbury, all based at the Hospital of the University of Pennsylvania (HUP). This work defined the clinical, electrodiagnostic, laboratory and sural nerve biopsy pathological findings of five adults (four women, one man) whom the authors found to have a distinctive, multifocal, motor and sensory, asymmetrical neuropathy.

In each person affected, the disorder began with motor and sensory symptoms relating to one or more upper extremity nerve trunks, usually ulnar or median, with symptoms waxing and in some, waning, but in all having been chronically present from 1.5 to 22 years. Electrophysiological studies showed sites of conduction block that were located more proximally on the affected nerve trunks. Persistence of the conduction block was demonstrated by comparison with prior exams or by serial exams over time in our electrodiagnostic unit. Some of these sites of persistent conduction block were also tender to touch, sometimes thickened, and in three, exhibited a Tinel's sign (tingling and irritation that occurs when the nerve is tapped). In time, other nerve trunks were affected, leading to symptoms in all four limbs in three persons, and to both arms and one leg (peroneal trunk) in the other two.

Optic neuritis with large central scotomas and prolonged visual evoked responses occurred in two persons, one monocular and the other bilateral. Both individuals recovered lost vision partially.

Why the optic nerves, which are part of the central nervous system (brain) were attacked, whereas all the rest of the disorder was in the periph-eral nerves, remains a mystery.

Cerebrospinal fluid examination in all five persons was either normal or in two instances was notable for a mild increase in protein content, less than 100 mgrrao. Oligoclonal bands were not present. (Such bands are characteristic of multiple sclerosis.) The observed spinal fluid formula is concordant with all subsequent experience with this entity; i.e. no bands.

Sural nerve biopsies were performed in three persons, and showed extensive demyelination-remyelination in all, but with highly variable degrees of axonal loss. Biopsy of the sites of conduction block could not be undertaken because of the likelihood of resulting increased neurological deficit. However, in an earlier report of two other cases, biopsy of an affected, thickened cutaneous" nerve from one of these two patients revealed a dramatic picture of hypertrophic changes with onion-bulb formation and perivascular inflammatory cell infiltrates. These two patients had precisely this clinical syndrome, but did not have electrodiagnostic studies. They were evaluated in an era when electrodiagnostic studies were done in only a few places in the world, and where they were evaluated was not one of those places.

To return to the cases reported by us and colleagues, corticosteroid treatment was undertaken in two patients, and was effective in both, but was difficult to reduce in dosage or to stop completely because of relapses. mg was not then available.

There were a number of features of these patients that challenged accepted concepts at the time. Prior to this report, the cause of multifocal attacks on individual nerves, or mononeuropathy multiplex, was thought to be due to an inflammation of blood vessels, technically called `vasculitis' This inflammation specifically af-fects the small blood vessels that allow blood flow to the nerves. The consequent loss of blood flow causes damage to both the axons and myelin. The five persons we reported did not have vasculitis, and the damage was first and foremost to just the myelin. Therefore something else had to be going on. The prevailing idea is that the myelin sheath is attacked because the body defense (immune system) specifically sees something in or on the myelin wrappings (insulation of the nerve fibers) that it considers to be foreign, even though it is not. Why this occurs at just certain sites on certain nerve trunks is a mystery. And why the arms first? Nobody knows. But the effect is blockage of the nerve signals at these sites; it can be likened to short circuiting.

Conduction block, in which the electrophysiologic signal traveling down the nerve is blocked despite the retained integrity of the axon is one of the hallmarks of demyelinating neuropathies. In all previous disorders in which conduction block is prominent, such as in acute nerve trauma or in Guillain Barre Syndrome, the conduction block is repaired within weeks or a few months. But some of our patients had demonstrated block that had persisted for years, one for over 20 years. This observation raised questions as to what actually happens when a nerve displays conduction block.

In 1982, CIDP was just beginning to be discussed and understood. At that time CIDP was thought to be a symmetric disorder of the arms and legs associated with elevated spinal fluid protein. These five persons had an asymmetric disorder, and the spinal fluid protein was either normal or only mildly increased. This report broadened the concepts of CIDP and was the first described variant of CIDP.

Since the 1982 report of Lewis et al, scores of papers have been published that report one or several more cases of this syndrome, and generally corroborate the findings elaborated above. These include five publications that report nerve biopsies of thickened proximal nerves; they substantiate the onion-bulb, hypertrophic changes described previously in 1965, and illustrated later. This pathology has long been thought to indicate long-term, recurrent demyelinationremyelination.

The term "Lewis-Sumner syndrome" (L-SS) had been used at various international meetings but was first proposed as a label for this specific motorsensory, demyelinating multifocal neuropathy in a Japanese language paper in 1990, and soon caught on, particularly in Europe. Many of the other papers and editorials reporting more cases of this entity vigorously debated the classification of these and like cases, and their relation to multifocal motor neuropathy (MMN) and chronic inflammatory demyelinating polyneuropathy (CIDP). Some of these reports of new cases of this particular neuropathy have offered new names for the condition, often in the form of an acronym.

In 1985, patients were reported who also had a multifocal attack on nerves, and had persistent conduction block similar to that seen in the earlier patients. However these patients had no sensory symptoms or signs and often were initially thought to have Amyotrophic Lateral Sclerosis. This disorder, now called Multifocal Motor Neuropathy, was initially considered to be part of a continuum with the other sensorimotor patients. However, subsequent reports have noted significant differences in spinal fluid protein (lower in MMN than in L-SS), the presence of anti-ganglioside antibodies (seen in 35-500/o of MMN and not in L-SS), different pathology at the sites of block (minimal demyelination and hypertrophy in MMN and marked hypertrophic changes in L-SS and the response to corticosteroids (no response or worsening of condition in MMN and good response in many patients with L-SS). Thus many experts currently believe that the two disorders are distinct. L-SS, as a sensorimotor entity with responses to corticosteroids is frequently classified as a multifocal variant of CIDP. MMN, on the other hand, because of some the unique features described above, is considered a distinct entity. However, we await better understanding of the immunological bases for each of these related demyelinating neuropathies.

Treatment for this syndrome remains empirical. Corticosteroids have long been known to reduce pain in this condition (L-SS) and also improve function. Half or more respond well at first, but in the long-term, corticosteroid use often presents problems. At present, high dose intravenous pooled immunoglobulin infusions are the mainstays of therapy for this particular condition, and for MMN and CIDP as well. The other immunosuppressive agents that are considered beneficial in CIDP, usually have similar effects in L-SS.

There are a number of intriguing questions that remain about L-SS. Why do these individuals have such a distinct multifocal pattern which is so different from those persons with more typical CIDP? What is the mechanism of the persistent conduction block? How does this relate to the hypertrophic nerve pathology seen on the nerve biopsy specimens? Why do the sites of nerve pathology persist for years and years? What predisposes particular persons to develop such disorders in the first place? Only further research can answer these questions. References furnished upon request.

Postscript Drs Asbury, Brown and Lewis are members of the GBS/CIDPFI Medical Advisory Board. Drs. Brown and Asbury are still at HUP (2008); Dr. Sumner is Chair Emeritus of Neurology at Louisiana State University, and Dr. Lewis is Vice Chair, Department of Neurology, Wayne State University. Others, beside Dr. Richard A. Lewis, who did their neuromuscular training at HUP and have been active in the debates about multifocal neuropathy with persistent conduction block include, in the order in which they trained, Dr. Gareth J. Parry, Dr. David R. Cornblath, Dr. John D. England, Dr. Ryuji Kaji, and Dr. Shawn J. Bird. An earlier version of this article was presented by Dr. Asbury on the occasion of the 125th anniversary celebration of the Philadelphia Neurological Society on February 20, 2008. et.