As in many other idiopathic illnesses, the evolution of knowledge has identified the cause of some cases of Bell palsy and has reduced the number of patients for whom the cause is unknown. Thus, 40 of 299 (13%) cases were shown by serologic tests to be associated with varicella zoster virus infection, that is, a form of "zoster sine herpete" or the Ramsay Hunt syndrome without herpetic skin lesions [7]. Lyme disease and human immunodeficiency virus (HIV) infection have recently been identified as causes of facial paralysis. Serologic studies have provided evidence for possible involvement by many other viruses, including herpes simplex virus, cytomegalovirus, Epstein-Barr virus, rubella, mumps, adenovirus, and enterovirus, although the number of cases allegedly caused by any one agent has been small and the results are inconsistent [8]. The finding of brain stem lesions and persistently elevated serum interferon levels has led some investigators to suggest a relation between Bell palsy and multiple sclerosis [9]. Others have proposed that Bell palsy might arise from a peripheral nerve demyelination reaction associated with one or more infectious agents. This proposal is supported by the following: evidence that the disease is part of a cranial polyneuropathy [4], the occurrence of a simultaneous outbreak of Bell palsy and the Guillain-Barre syndrome in Hawaii [3], the response of peripheral blood mononuclear cells to peripheral nervous system basic protein p_1L [10], and the occurrence of spinal fluid abnormalities in one third of patients, including elevation of the myelin-associated enzyme 2', 3'-cyclic nucleotide 3'-phosphohydrolase [11].

Since McCormick [12] first hypothesized in 1972 that herpes simplex virus could be a cause of Bell palsy, herpes simplex virus infection has been a popular but unsubstantiated theory. The low frequency of seroconversion to this virus (4%) in association with Bell palsy excludes primary herpes simplex virus infection. On the other hand, evidence suggests an increased prevalence and increased mean serum antibody titers to herpes simplex virus [8]. Herpes simplex virus could be involved in Bell palsy by the reactivation of latent infection in the geniculate ganglion and the spread of the virus through epineural cells within the temporal bone, leading to inflammation and compression of the facial nerve in the fallopian canal. Herpes simplex virus type 1 has been identified by the polymerase chain reaction and in situ hybridization in a high percentage of geniculate ganglia from adults, at autopsy [13, 14] and by isolation in tissue culture from a biopsy specimen of the facial nerve of one patient during an operation for nerve decompression [15]. Inoculation of the tongue of mice with herpes simplex virus leads to infection of the geniculate ganglion, showing a potential source of neural infection through the facial nerve's sensory fibers for taste [16]. Schwann cells have been shown to support the growth of herpes simplex virus both in vitro and in vivo [17, 18]. Although mucocutaneous herpes simplex virus lesions are not common features of Bell palsy, it is possible that the retrograde spread of herpes simplex virus through sensory fibers for taste might result in asymptomatic virus excretion into the oral cavity or an unusual clinical presentation such as inflammation of the papillae of the tongue [1].

At this time, the various features of Bell palsy cannot yet be assembled into a coherent pathogenetic hypothesis. Unlike in Bell palsy, a virus-associated immune-mediated peripheral demyelinating disease such as the Guillain-Barre syndrome is generally symmetrical and diffuse. Nervous system disease may immediately follow primary virus infection, for example, postinfectious encephalomyelitis and the Guillain-Barre syndrome. In cases of chronic virus infection such as subacute sclerosing panencephalitis, there is a relentless, progressive course. Neither of these patterns resemble Bell palsy. As reviewed above, herpes simplex virus reactivation from the geniculate ganglion and direct injury to the facial nerve are more plausible. However, unlike herpes simplex labialis, which is the common result of reactivation of herpes simplex virus infection of the trigeminal ganglion, Bell palsy is not associated with peripheral herpes simplex virus lesions, is usually a one-time rather than a recurring illness, and is not exacerbated in frequency and severity by immunosuppression.

Antiviral therapy for this disease might be considered. In trials of such therapy, the optimal protocol should exclude patients with serologic evidence of varicella zoster virus reactivation, Lyme disease, and HIV infection. Clinical neurophysiologic techniques are now available that can differentiate nerve block from Wallerian degeneration and can further define the anatomic site of the lesion [19]. Drug efficacy should be studied in patients who are herpes simplex virus type 1-seropositive by Western blot [20]. A thorough history of herpes simplex labialis should be obtained from the patient, including the location, frequency, and severity of episodes. Careful intraoral examination for viral lesions, with biopsies done as necessary, and measurement of the excretion of herpes simplex virus in the oral cavity during the acute phase of the illness would add to our understanding.