Diabetic Lumbosacral Polyradiculoneuropathies.

The pathogenic basis and treatment of diabetic polyradiculoneuropathy is a source of recent controversy as there may be two or more distinct forms of diabetic polyradiculoplexopathy. We believe that the following two categories of diabetic polyradiculoneuropathy can be made on the basis of clinically differences: 1) the more common asymmetric, painful polyradiculoneuropathy; and 2) the rare symmetric, painless, polyradiculoneuropathy. The asymmetric, painful form (also known as diabetic amyotrophy) may have an autoimmune basis, but the etiology is not clear. The natural history for diabetic amyotrophy is spontaneous improvement. Nevertheless, various immunotherapies (eg, corticosteroids and intravenous immunoglobulin (IVIg) have been tried with subsequent improvement in symptoms. Treatment is reserved only for patients with severe ongoing pain, given the significant side effects of these medications in those patients with diabetes. Prednisone and IVIg may help alleviate the pain associated with diabetic amyotrophy. Relief of pain can help patients begin physical therapy earlier, however, there are no prospective, blinded, controlled studies that demonstrate that these treatments lead to an earlier and better recovery of muscle strength compared with the natural history of the disorder. The symmetric, painless form of diabetic polyradiculoneuropathy may in fact represent chronic inflammatory demyelinating polyneuropathy (CIDP) occurring in a patient with diabetes mellitus (DM). Patients with idiopathic CIDP may improve various immunomodulating therapies, including corticosteroid treatment, plasma exchange (PE), and IVIg. In this regard, patients with the symmetric, painless, proximal diabetic polyradiculoneuropathy may also respond to corticosteroids, plasma exchange, IVIg, azathioprine, or cyclophosphamide. However, as with diabetic amyotrophy, some patients improve spontaneously without treatment. In still other patients, the neuropathy appears unresponsive to immunotherapy. In such patients, this polyradiculoneuropathy might be caused by metabolic dysfunction associated with DM. Unfortunately, from a clinical, laboratory, and electrophysiologic standpoint, it is impossible to distinguish the patients with a symmetric, painless diabetic polyradiculoneuropathy who might respond to therapy. A trial of PE can be useful in identifying patients who might have a polyradiculoplexopathy that is responsive to immunotherapy. If patients respond to PE, they may continue to receive intermittent exchanges or be switched over to prednisone or IVIg.