Does viral disease underlie ALS?

A viral etiology for ALS has long been hypothesized because of the selective vulnerability of motor neurons to certain viruses.1 In the 1960s and 70s, poliovirus was the darling of those proposing a viral etiology2 because of its relatively selective infection of motor neurons, the occasional reports of antecedent poliomyelitis in patients with ALS, and the frequent occurrence of chronic progressive weakness many years after acute poliomyelitis, now recognized as the post-polio syndrome (PPS).3 Most patients with PPS have only lower motor neuron disease, but a small percentage also have upper motor neuron signs and present with features suggestive of ALS.3,4⇓ Although prior attempts to isolate polioviruses and other enteroviruses from the CNS of patients with ALS have failed,5 newer technologies may prove valuable in helping to establish a possible viral etiology.6

In this issue of Neurology, seven cases of ALS syndromes accompanying HIV infection are reported.7,8⇓ The retroviruses, like HIV, are ideal candidates to cause classic ALS because they replicate by synthesizing a DNA copy (the provirus) that can persist by integration into the host’s cellular DNA. For example, it has long been known that murine neurotropic retroviruses cause motor neuron disease, although the pathology is different than that seen in ALS.9 In the 1980s it was recognized that human T-cell lymphotrophic virus type I (HTLV-I), the retrovirus responsible for HTLV-I–associated myelopathy/tropical spastic paraparesis (HAM/TSP), causes an ALS syndrome.10 Currently at least eight of these cases have been reported. In the one autopsy, there were not only pyramidal tract and anterior horn cell lesions, but also prominent lymphocytic inflammation characteristic of HAM/TSP.11 HTLV also causes lymphoma, and because there is an increased frequency of lymphoma in patients with ALS,12 a link between ALS and HTLV is further suggested.

In this issue, Moulignier et al.7 retrospectively identified six cases of ALS syndromes among 1700 patients with HIV with neurologic manifestations who were seen over a 13-year period. In the population at large, the annual incidence of ALS was 0.4 to 1.76 per 100,000. Thus, the occurrence of this CNS complication of HIV infection is rare but 27-fold more than expected. All six patients have upper and lower motor neuron dysfunction, but only one met the El Escorial criteria for definite ALS.13 The other five patients in the Moulingnier et al. study were classified as having probable or possible ALS, but it is not clear how many were in each category. Unlike patients with classic ALS, an inexorably progressive disorder, these six patients either stabilized or partially recovered from their ALS syndromes after starting antiretroviral therapy. Similarly, the patient with definite ALS by El Escorial criteria in the accompanying report by MacGowan et al.8 demonstrated a dramatic clinical response after treatment with highly active antiretroviral therapy (HAART). A significant reduction in both plasma and CSF viral load was also observed. Curiously, this patient had white matter abnormalities of the brachium pontis and medulla on MRI, which also resolved after HAART. The underlying pathology was never determined, and the basis for the ALS-like manifestations is not clear.

These HIV-associated ALS syndromes differ from classic ALS in some respects. First, they occurred in younger patients than is typical for classic ALS. Second, they were unusually rapidly progressive. Third, the disorder did not progress inexorably, but improved after institution of antiretroviral therapy. The latter suggests an etiologic association with HIV. However, HIV is not a neurotropic virus. It rarely, if ever, infects neurons, but occurs predominately in microglial cells within the CNS.14 Therefore, if HIV is the cause of this motor neuron disorder, one might envision a selective damage to motor neurons by neurotoxic viral proteins or by cytokines (e.g., COX-215) and chemokines elaborated as a consequence of the HIV infection. Alternatively, one could argue that an unknown, opportunistic virus that directly infects motor neurons caused these ALS syndromes, which then improved after immunologic recovery attending antiretroviral therapy.

During the last 15 years, more than a dozen cases of HIV-associated ALS syndromes have been reported.7,8,10⇓⇓ Of the two cases that have come to autopsy, only one was demonstrated to have had motor neuron loss.16 Although some familial forms of ALS have an abnormality in superoxide dismutase suggesting a neurodegenerative process as the underlying cause of sporadic ALS, these HIV-related cases suggest that a viral etiology may underlie some forms of the disorder. Continued investigations of this possibility are clearly indicated.

Many causes of ALS syndromes have been recognized,17 some of which are treatable: hyperparathyroidism, hyperthyroidism, heavy metal toxicity, hexosamimidase deficiency, remote effects of cancer, immune-mediated antiganglioside disease, and cervical spondylitic myelopathy. Based on the accompanying studies,7,8⇓ when a patient presents with the classic clinical signs of ALS, a viral cause should also be considered because HIV-ALS syndromes are treatable.