TABLE 4.

Criteria for causality and strength of evidence linking HHV-6A/B to mesial temporal lobe epilepsy

Disease causation criterion	Evidencea
HHV-6A/B nucleic acid is present in diseased tissue, in most cases, in higher abundance than in nondiseased tissue (by qPCR or other means).	Positive evidence:  HHV-6A/B infects cells in the hippocampus (32, 71, 94, 108–111).  Over 60% of hippocampus and temporal lobe specimens resected to treat MTLE have evidence of active infection with HHV-6A/B (32, 71, 109, 127–131).  HHV-6A/B DNA is found only rarely in the same anatomic areas in other forms of temporal lobe epilepsy (32, 109).
	Negative evidence: One study did not find HHV-6A/B in resected tissue (132).
The amount of HHV-6A/B nucleic acid in diseased tissue or blood, and/or antibody levels, correlates with the severity of the disease.	No evidence yet.
HHV-6A/B nucleic acid is demonstrated in cells relevant to disease pathology.	HHV-6A/B can infect adult neurons (230, 232), adult astrocytes (109, 230, 231), and microglial cells (30, 229, 230).
HHV-6A/B mRNA (by RT-PCR or other means) and antigens by immunohistochemistry are present in diseased tissue.	Viral mRNA detected by RT-PCR (71) and viral antigen (32) are found more frequently in the hippocampus of people with MTLE.
Exposure to and then presence of the viruses and their gene products in affected tissue precede the development of the disease (temporal relationship).	95% of humans are infected with HHV-6B in very early childhood. The number of humans infected with HHV-6A, and the typical age of primary infection, is less clear.
Infectious agents other than HHV-6A/B are not detected in diseased tissue in a substantial number of cases.	HHV-6A/B is present in resected tissue more often than other human herpesviruses (71).
There are cellular and/or humoral immune responses to HHV-6A/B in diseased tissue and/or in blood, and these responses correlate with the severity of the disease.	No evidence yet.
HHV-6A/B affect cellular function in diseased tissue in a manner known to cause or augment the disease pathology (in vitro or in vivo studies).	HHV-6A/B infect astrocytes in the temporal lobe and hippocampus and cause hippocampal and mesial temporal lobe sclerosis (42, 71, 109, 123, 124), anatomic correlates of mesial temporal lobe epilepsy.
Specific antiviral therapy both reduces viral load in diseased tissue or blood and is followed by clinical improvement.	No evidence yet.

^aAll evidence cited is positive evidence in support of the assertion unless specifically identified as negative evidence.