Are We Getting Closer to the Treatment of Rabies?

Rodney E Willoughby Jr


Future Virology. 2009;4(6):563-570. 

In This Article


While there are many animal models investigating immunogenicity and protection by rabies vaccines, there are few models that investigate its pathophysiology using wild-type viruses. We are not aware of studies addressing peripheral neurophysiology, cerebral metabolism or perfusion, while a single study applied electroencephalography to wild-type rabies infection.[29] Gene-expression studies of wild-type rabies have been reported but do not readily explain pathophysiology.[30,31] Animals will occasionally survive wild-type rabies, but I am unaware of further studies to characterize their immunology, recovery or even possible relapse of rabies. Medically, the mechanisms behind our improvised, multifactorial treatment of rabies are unknown and cannot justify a clinical trial. We learned informally, while advising on specific patients being treated by motivated physicians and the patient's families. Given that rabies is 100% fatal, it is difficult to recover from therapeutic mis-steps. This learning came at a considerable cost.

We published our protocol online at.[102] Online dissemination was key because it was quickly learned that there are large differences in time zones between rabies-endemic areas and us, and that most rabies patients die within 2–3 days of first hospitalization. An International Rabies Treatment Registry that tabulates the simple outcome of survival was set up. The pace of attempts at the MP has been surprisingly slow. There has been significant resistance to this approach, and intensive care is economically prohibitive for most residents in rabies-endemic countries.[32]

Our first survivor had discordant progress along milestones during her rehabilitation, particularly involving involuntary movements, which raised the question of an acquired or unmasked metabolic disorder. We diagnosed deficiencies of BH4 and coenzyme Q10 (ubiquinone), both electron-transport cofactors. Ubiquinone deficiency was not confirmed in subsequent rabies patients. As noted earlier, BH4 is essential for the synthesis of dopamine and serotonin; all three metabolites were deficient in the five consecutive patients with rabies studied to date.[14,33] BH4 deficiency also occurs as an inborn error of metabolism, and sapropterin (synthesized BH4) has been licensed by the US FDA for replacement therapy of phenylketonuria. There is considerable overlap in symptoms between congenital BH4 deficiencies and rabies, but we have not been able to assess symptomatic improvement with BH4 deficiency during early rabies while patients remain under heavy sedation. Neurotransmitter metabolites are restored to normal range with sapropterin supplementation.[13]

Tetrahydrobiopterin is also a cofactor for nitric oxide synthase (NOS). Endothelial and neuronal isoforms of NOS contribute differentially to systemic, pulmonary and cerebrovascular tone. In the brain, 50% of blood flow is controlled by the neuronal isoform.[34] The neuronal isoform of NOS is affected by the highly neurotropic rabies virus.[35] While loss of 50% cerebral blood flow (CBF) is not intrinsically fatal, it does move the patient to the limits of pressure autoregulation of CBF. Evidence for vasospasm has been prospectively evaluated in three rabies patients by transcranial Doppler (TCD) ultrasonography. Vasospasm of the large conduit vessels, with normal resistance, was documented on hospital during days 6–8.[36] In rabies, vasospasm was associated with transient loss of EEG amplitude, seizures and subsequent cerebral edema. The timing of vasospasm corresponds closely to when coma supervenes in human rabies.[36] Vasospasm was generalized, in keeping with innervation of internal carotids and more distal arteries through the pterygopalatine ganglia.[34] Blood velocities normalize, then a second form of vasospasm (type 2) follows 7 days after type 1 vasospasm. The type 2 spasm is peripheral, characterized by very low blood velocities, low intracranial pressure, high resistive indices and complete loss of EEG activity.[36] An autopsy revealed that the cerebral microvasculature was normal. Administration of BH4 was temporally associated with improvements in type 1 spasm but did not prevent type 2 vasospasm. The timing of type 1 spasm in rabies is similar to vasospasm complicating subarachnoid hemorrhage. Nimodipine is often used prophylactically after subarachnoid hemorrhage. We adopted prophylaxis using nimodipine (a calcium-channel blocker), 'downstream' in the cellular signaling pathway from BH4 and NOS, in MP version 2.

We have virological evidence that ribavirin, amantadine and ketamine, administered together, are ineffective at reducing salivary viral load in human rabies. Ribavirin, a broad-spectrum antiviral drug active in vitro against rabies virus, was intially used despite evidence that it poorly penetrated the blood–brain barrier acutely and had not worked therapeutically in animals and humans.[26] It might provide virological benefit in the heart, given that 20% of rabies patients die of early dysrhythmias and rabies virus has been demonstrated in the conduction system of the heart. Unfortunately, ribavirin is also an immunosuppressant, sharing the same mechanism of action as mycophenolate, a drug in common use during organ transplantation.[37] In replicating the original, improvised protocol that included ribavirin (MP version 1), we noticed unusual week-long hiatuses in the normally exponential development of antibody titers to rabies. This was mystifying until we noted the same phenomenon reported in macaque monkeys treated for the Lassa virus.[38] Ribavirin was removed from MP version 2, and we have since noted much higher rabies neutralizing antibody titers [WILLOUGHBY R, UNPUBLISHED DATA]. The use of barbiturates was proscribed for similar reasons.[39]