To the editor

Krause et al.1 have suggested that subclinical and possibly clinical reactivation of VZV serves to boost the humoral immune response in previously immunized healthy children.

We reviewed our prior experience and that from two ongoing prospective studies8,9,10 of rashes occurring more than six weeks after varicella vaccination of healthy subjects. All 72 vaccinees (9 leukemic children, 6 healthy adults, 57 healthy children) with breakthrough varicella, characterized by a generalized maculopapular or vesicular rash, in whom VZV DNA was amplified, the wild-type virus was identified. In the two prospective studies, there were 57 vaccinees who had rashes caused by wild-type VZV, but in addition there were 36 vaccinees with rashes who were PCR-negative for VZV, and 19 vaccinees with rashes whose results were indeterminate. The second two groups indicate that some of the rashes were caused by something other than VZV and that some samples were not properly obtained. The sensitivity of the assay as determined by serial dilutions of extracted VZV Oka & wild-type DNA was previously found to be approximately 100 femtograms10b. In 38 patients with zoster, the vaccine strain was identified in 25 (3 leukemic and 22 healthy children) and the wild-type strain in 13 (2 leukemic and 10 healthy children, 1 healthy adult).

Our results are consistent with those from other studies. Watson3 identified the wild-type virus in five of five cases of breakthrough varicella. In another study, in all cases where an isolate was identified from a vaccinee with a generalized rash beyond the six-week post-vaccination period, it was identified as wild-type VZV.4

Although it has been suggested that reactivation of VZV occurs in the setting of low levels of VZV-specific antibodies, this concept is not in keeping with data from our laboratory or that of others. VZV reactivates in the setting of low cell-mediated immunity.6,11 Low antibody levels are not associated with reactivation,6,12 but are associated with reinfection.6,13 The geometric mean FAMA titer14 of vaccinees who developed breakthrough varicella (9 children with leukemia, 6 healthy adults) at a mean of 5.3 months before rash was 2. That of vaccinees who developed zoster (5 children with leukemia, 1 healthy adult) at a mean of 2.8 months prior to zoster was 5.7 (P = 0.01, Student's t-test).

To our knowledge, there has been only one documented generalized rash in a recipient of varicella vaccine due to vaccine strain VZV.15 This child developed a rash about two weeks after a household exposure to his brother, who had zoster at the time of exposure. Both received the varicella vaccine five months previously. In a personal communication, Brunell proposed that the boy with the generalized rash was re-infected by the vaccine virus following exposure to his brother with zoster (Brunell, personal communication).

The available evidence indicates that generalized rashes due to VZV in previously vaccinated individuals are the result of re-infection with the wild-type virus. Vaccinated individuals with low antibody titers to VZV are at increased risk to become re-infected. In contrast, localized, unilateral rashes are due to reactivation of latent VZV, either wild or vaccine strain. Reactivation occurs in the setting of a low cell-mediated immune response to the virus. If reactivation of the vaccine virus resulting in generalized rash were to occur, it would be expected to occur particularly in immunocompromised vaccinees. It has been our experience (see above) that even in these vaccinees, generalized rash is caused by the wild-type virus.

Whether subclinical reactivation of vaccine strain occurs and boosts immunity to VZV requires further study. As more children are vaccinated, there will be fewer cases of varicella and therefore fewer opportunities for boosting of VZV-specific antibody titers in vaccinees due to exposure to wild-type virus. We would therefore predict that in the absence of subclinical reactivation of the vaccine virus mean antibody titers would decline.

See “Reply to 'Varicella vaccine revisited'” by Krause et al.