This is the authors' abstract. We don't add key points for this paper.
To the Editor—The interesting article by Lurain et al [1] presents data that are broadly consistent with studies indicating that inflammation caused by various infections can affect the brain. Although the authors mentioned that cytomegalovirus (CMV) is known to infect the central nervous system (CNS), they did not mention that in the immunocompetent host, CMV is a very rare cause of CNS infection. CMV encephalitis is found with any appreciable frequency only in those with severe T-cell–mediated immune deficiency, a condition which is rarely, if ever, found in association with Alzheimer disease (AD). As all of the data presented by Lurain et al relate only to peripheral CMV infection, presumably they are implicating peripheral infection rather than CNS infection. However, it is difficult to understand how CMV, with a tropism for lymphoid and epithelial cells, could result in the loss of neurons within the cerebral cortex observed in AD. This is much more readily explained in terms of the effect of an even more common neurotropic herpes virus, herpes simplex virus type 1 (HSV-1), when it infects the CNS. Several authors, in different laboratories, have found HSV-1 to be present in the brains of a high proportion of the elderly, including patients with AD (see [2] for a review) and, as evidenced by the detection of the synthesis of intrathecal antibodies to HSV-1 [3], to have caused active infection of the brain in AD patients. To verify a role for CMV in causation of CNS disease, Lurain et al would need to determine if there is any evidence of intrathecal antibody synthesis to CMV (ie, evidence of infection within the CNS compartment). Without such evidence, CMV cannot be said to play a direct role in causation of AD. In apparent conflict with a role for HSV-1 in AD, the authors found that infection of cells by CMV—but not by HSV-1—led to the formation of amyloid, a key abnormal molecule of AD brains. However, 3 separate research teams have detected amyloid in HSV-1–infected neural-type cells [4–6], one of which sought and found it also in brain of infected mice [4]. Furthermore, as the authors do not state what doses of the 2 viruses they used, comparison of the viruses’ effects is not feasible, and it seems likely that the difference in the experimental results represents simple differences among in vitro experiments, such as in cell type, as they suggest, or in virus dose. Lurain et al found no association between the level of peripheral blood antibody to HSV-1 and pathological changes in AD, whereas CMV antibody correlated with such changes. For HSV-1, peripheral blood antibody levels change only slightly when HSV-1 reactivates from a dormant state in the body [7], even when such reactivations are very frequent. It is, therefore, to be expected that there is no linkage between HSV-1 antibody levels and AD pathology. However, a relevant linkage between HSV-1 and CMV has been found by Stowe et al [8], whose data led to them to conclude that HSV-1 reactivation is associated with CMV and age, perhaps via CMV-induced immune dysregulation. This conclusion is consistent with several epidemiological studies that have implicated CMV and HSV-1, but the latter to a much greater extent and directly, in cognitive decline and/or AD. Data on the effect of CMV reactivation on CMV antibody levels in peripheral blood are lacking because of the difficulty of observing clinical signs of reactivation of CMV. Exploration of the serum antibody levels of other members of the Betaherpesvirinae subfamily (human herpesviruses 6a, 6b, and 7) in these patients may be instructive. As to the possibility of long-term therapy to prevent the occurrence of symptoms of active CMV infection, the authors state that this is not feasible at present. Fortunately with HSV-1, antivirals that reduce virus damage [9] are very safe, and there is evidence that they could be used long term with no ill effects [10]. Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Itzhaki et al. (2013) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: