Abstract
CONTEXT: Live vaccines usually provide robust immunity but can transmit the vaccine virus.
OBJECTIVE: To assess the characteristics of secondary transmission of the vaccine-strain varicella-zoster virus (Oka strain; vOka) on the basis of the published experience with use of live varicella and zoster vaccines.
DATA SOURCES: Systematic review of Medline, Embase, the Cochrane Library, Cumulative Index to Nursing and Allied Health Literature, and Scopus databases for articles published through 2018.
STUDY SELECTION: Articles that reported original data on vOka transmission from persons who received vaccines containing the live attenuated varicella-zoster virus.
DATA EXTRACTION: We abstracted data to describe vOka transmission by index patient’s immune status, type (varicella or herpes zoster) and severity of illness, and whether transmission was laboratory confirmed.
RESULTS: Twenty articles were included. We identified 13 patients with vOka varicella after transmission from 11 immunocompetent varicella vaccine recipients. In all instances, the vaccine recipient had a rash: 6 varicella-like and 5 herpes zoster. Transmission occurred mostly to household contacts. One additional case was not considered direct transmission from a vaccine recipient, but the mechanism was uncertain. Transmission from vaccinated immunocompromised children also occurred only if the vaccine recipient developed a rash postvaccination. Secondary cases of varicella caused by vOka were mild.
LIMITATIONS: It is likely that other vOka transmission cases remain unpublished.
CONCLUSIONS: Healthy, vaccinated persons have minimal risk for transmitting vOka to contacts and only if a rash is present. Our findings support the existing recommendations for routine varicella vaccination and the guidance that persons with vaccine-related rash avoid contact with susceptible persons at high risk for severe varicella complications.
- ALL —
- acute lymphocytic leukemia
- CINAHL —
- Cumulative Index to Nursing and Allied Health Literature
- FAMA —
- fluorescent antibody to membrane antigen
- HZ —
- herpes zoster
- vOka —
- vaccine-strain varicella-zoster virus (Oka strain)
- VZV —
- varicella-zoster virus
Varicella, caused by the varicella-zoster virus (VZV), is highly contagious. Although generally a mild childhood disease, serious complications (including death) can occur, most commonly in infants, adults, and immunocompromised persons.1 Varicella was an important cause of morbidity and mortality in the United States before the introduction of varicella vaccination (1996), with ∼4 million cases, 10 500 to 13 000 hospitalizations, and 100 to 150 deaths annually.2 In the 1970s, 30% of children with leukemia who contracted varicella developed viscerally disseminated disease, with 7% mortality.3 In the era of antiviral therapy, prognoses improved with treatment administered early in the course of illness, but deaths continue to occur. Varicella during pregnancy can cause congenital disease in the fetus or severe varicella in the newborn or can lead to herpes zoster (HZ) in early childhood.4 Nosocomial transmission of VZV often disrupts health care facilities’ operations.5,6 VZV exposures among patients and health care personnel can be time consuming and costly even when they do not result in transmission.
Vaccination effectively prevents disease transmission in community and health care settings. To address the burden of varicella, a live attenuated vaccine was developed in the 1970s7 and licensed in the United States in 1995. Varicella vaccines are contraindicated for immunocompromised persons or pregnant women, but these individuals may benefit indirectly from vaccination of household contacts and health care workers who care for them.2 Several formulations of varicella vaccines, all containing the live attenuated virus, are available worldwide but are recommended for routine use in only a limited number of countries. Since 2006, a live attenuated zoster vaccine, which contains the same VZV strain as varicella vaccines but at a higher potency, has been available to prevent HZ.
Live attenuated vaccines usually provide more robust immunity than vaccines containing inactivated viruses but can, in rare instances, transmit the vaccine virus. Millions of persons are vaccinated with varicella vaccine each year in the United States, including routine vaccinations and vaccination of close contacts of persons at high risk for severe varicella; therefore, tracking the characteristics of secondary transmission of vaccine-strain VZV (Oka strain; abbreviated as vOka) is important. We reviewed the experience reported in the literature with the use of live VZV vaccines administered to healthy and immunocompromised persons to address this question.
Methods
We defined secondary transmission of the vaccine strain as transmission of vOka from a person who had ever received a vaccine containing live attenuated VZV to a contact. Transmission was determined by the authors of the articles by documentation of (1) laboratory confirmation of vOka from the rashes of the vaccinated index case and secondary case(s), (2) laboratory confirmation of the virus from the secondary case’s rash similar to that in the vaccine received by the index case, or (3) seroconversion of contacts of vaccine recipients in the absence of a rash or other VZV exposure of the contact. Some authors reported transmission based only on temporal association between vaccination and the disease in the contact of a vaccine recipient. Those reports were classified in this article as unsubstantiated transmission because no or insufficient laboratory testing was performed to document vOka; additionally, in some reports, the epidemiological assessment did not provide support for transmission from the vaccine recipient or did not accord with the biology of varicella, especially the incubation period.
We searched Medline, Embase, the Cochrane Library, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) databases for articles published, in any language, from database inception through December 31, 2018. The complete search strategy is described in Supplemental Table 3. Two authors reviewed each title and abstract to determine if the article included any information on transmission of vOka. For articles that passed this initial screen, we reviewed the full text. We also reviewed references that described transmission of vOka identified from the reference section of the articles retrieved by the database search. Only articles that included original reports of vOka transmission were retained.
Two authors abstracted data on disease presentation and immune status of the index case (ie, varicella vaccine recipient), number of secondary cases, transmission setting, how transmission was determined, interval between vaccination and rash onset in the index case, interval between rash onset in the index and secondary case(s), characteristics of the disease and immune status of the secondary case(s), whether the index and secondary case(s) were laboratory confirmed, and vaccine manufacturer. We describe vOka transmission by immune status of the index patient (immunocompetent or immunocompromised), type (varicella or HZ) and severity of illness this patient had, and whether vOka transmission was laboratory confirmed.
Results
We screened 378 nonduplicate articles and identified 98 for full-text review; this included 1 study known by the authors to be published after the literature search was completed.8 After excluding 72 articles, 26 met inclusion criteria for our review (Fig 1). Six articles reported interim results on 2 studies. To describe instances of secondary transmission, we included only the last reports from these 2 studies; when specific information of interest was not included in the last publications, we abstracted data from the earlier publications. A total of 20 articles were included in our review.
Study selection.
Articles originated from the United States (n = 15), Australia (n = 2), China (n = 1), Japan (n = 1), and Europe (n = 1; country not reported); publication years were 1981–2019. Most articles described transmission of vOka after receipt of Varivax (Merck varicella vaccine), which is licensed in the United States, likely because large clinical trials among immunocompromised and immunocompetent children were conducted in the United States before licensure of the vaccine, and the United States has the longest-running routine varicella vaccination program. Transmission of vOka from recipients of other varicella vaccines (SmithKline and/or GlaxoSmithKline or Biken) also occurred.9–12 Of note, all available varicella vaccines except the vaccine used in Korea contain live attenuated VZV derived from the same parental seed stock of vOka. All reports of vOka transmission followed receipt of a first dose of the varicella vaccine; no transmission was reported from recipients of the live zoster vaccine.
Overall, we identified 13 cases of confirmed transmission of vOka from 11 immunocompetent vaccine recipients with a rash.8–17 One additional case was not considered direct transmission from a vaccine recipient, but the mechanism was uncertain.18 In the largest study of immunocompromised children, the rate of transmission of vOka was 23% from vaccine recipients who developed rash after vaccination and 0% if the vaccine recipient did not develop a rash.19
vOka Transmission From Immunocompetent Varicella Vaccine Recipients
Transmission Resulting in Clinical Varicella Among Contacts
Five articles reported 6 instances of transmission of vOka from an immunocompetent varicella vaccine recipient who developed a varicellalike rash soon after vaccination (Table 1).10,14–17 In 5 instances, transmission occurred in the household; 1 occurred in a chronic care facility. Eight secondary cases occurred from the 6 vaccinated individuals, all with laboratory-confirmed vOka. Of the vaccinated individuals, 2 were adults and 4 were children, and all developed a vesicular rash 12 to 24 days after vaccination; rash was mild (2–40 lesions) with 1 exception (>500 lesions). Contacts who developed varicella were more variable in age (range 4 months–39 years) and with no reported immunocompromising conditions. Rash in the contacts developed 16 to 21 days after rash onset in the index cases and was mild (median 25–30 lesions; range 10–100; 1 with >100 lesions).
vOka Transmission From Immunocompetent Varicella Vaccine Recipients: Transmission From a Varicella Vaccine Recipient Who Developed a Varicella-like Rash Soon After Vaccination
Five articles reported 5 instances of transmission of vOka from an immunocompetent varicella vaccine recipient who developed vOka HZ after vaccination (Table 2).8,9,11–13 In 4 instances, transmission occurred in the household; 1 occurred in school. Five secondary varicella cases occurred. No immunocompromising conditions were reported in the infected contacts. Index cases were children (age 20 months–5 years) who developed HZ 8 months to 2 years after varicella vaccination. Contacts who developed varicella ranged in age from 2 to 35 years. Rash in the contacts developed 14 to 19 days after HZ rash onset in the index cases, and the number of lesions ranged from 10 to 20 to “uncountable” (median 50–100); 1 contact had mild meningismus but no focal neurologic signs.8
vOka Transmission From Immunocompetent Varicella Vaccine Recipients: Transmission From a Varicella Vaccine Recipient Who Developed vOka HZ After Vaccination
Seroconversion Among Contacts of Varicella Vaccine Recipients as a Proxy for Transmission
In clinical trials of the varicella vaccine, seroconversion in siblings of healthy vaccinated children was examined. The siblings were varicella susceptible and vaccinated with a placebo.20 Six (1%) of 439 placebo recipients seroconverted without having developed a rash; none of their varicella-vaccinated siblings developed a rash. Serological data of the vaccine recipient and sibling who received the placebo suggested that 3 of 6 seroconverters received the vaccine mistakenly in lieu of their sibling.20 An alternative explanation is that the serological test used yielded false-positive results, not an uncommon occurrence.21,22
Other Mechanism of Transmission
Kluthe et al18 reported a case of neonatal varicella (50–60 maculopapular and vesicular lesions) with vOka diagnosed 22 days after maternal postpartum vaccination. The mother did not have a rash, but the newborn was in the room when the mother was vaccinated; the mechanism of transmission remains undetermined, but the most plausible mode of transmission was deemed aerosolization when the vaccine-filed syringe was cleared of air bubbles rather than transmission from the mother.
vOka Transmission From Immunocompromised Varicella Vaccine Recipients Resulting in Clinical Varicella or Seroconversion Without Rash in the Contact
Several publications described case reports or observational studies of transmission of vOka after vaccination of immunocompromised children.17,19,23–25 In these studies, authors attributed disease or seroconversion without rash among unvaccinated siblings to vOka, although laboratory confirmation of vOka causing the rash was not always obtained (or could not be obtained in cases of seroconversion). Exposures to varicella or HZ were closely monitored, and if no other known exposures were reported, rash in the vaccinated sibling was considered the likely source for the rash or seroconversion in the contacts.
Because immunocompromised persons are at high risk for severe varicella, the US National Institutes of Health supported a clinical trial of varicella vaccine in immunocompromised children during the 1980s26; 575 children with acute lymphocytic leukemia (ALL) in remission were vaccinated.19 A vaccine-associated rash (mean 77 lesions) occurred a median of 30 days (range 7–41) after the first dose in 40% of children in whom chemotherapy was suspended for 1 week before and 1 week after vaccination and 10% of those who had completed their chemotherapy before vaccination.17 Of 93 varicella-susceptible (laboratory-confirmed) healthy siblings exposed to a vaccine recipient with leukemia with a vaccine-associated rash, transmission was reported in 21 (23%); of these, 16 developed a rash 14 to 22 days after rash onset in the vaccinated sibling (vOka was isolated from 4), and 5 had only seroconversion by fluorescent antibody to membrane antigen (FAMA) assay.19 The rash among contacts was mild, with a mean of 38 lesions (median 12; range 1–200) and no systemic symptoms. Transmission of vOka correlated with the number of skin lesions in the vaccinated sibling; vaccine recipients who transmitted the virus to susceptible exposed siblings had a mean of 195 lesions (median 110) compared with 43 lesions (median 15) among those who did not transmit (P = .006).17 No spread of vOka was documented (by FAMA) among 124 siblings exposed to a vaccine recipient without rash after the first dose27; there was no evidence of transmission to 72 siblings exposed to a vaccine recipient who received a second dose, including 10 siblings exposed to a vaccine recipient with a rash.27
Similar experiences were reported by a case report and 2 smaller studies (Supplemental Table 4).23–25
Reports of Unsubstantiated vOka Transmission
Six instances of alleged but unsubstantiated transmission of vOka after vaccination of immunocompetent persons have been reported (Supplemental Table 5).17,25,28–31
Discussion
Spread of vOka as documented in the literature illustrates both potential for vaccine virus transmission and characteristics associated with the spread, with applicability in clinical practice. We found that transmission of the vaccine virus from healthy vaccine recipients is rare. To date, only 13 cases from 11 immunocompetent vaccine recipients have been reported, most commonly among household contacts. An additional few alleged instances of vOka transmission were suspected but not confirmed by appropriate laboratory testing. These rare instances of transmission occured in the context of large numbers of varicella vacine used; in the United States alone, >140 million doses of varicella vaccine were distributed during 1995–2013.32 Transmission of vOka resulting in clinical varicella was reported only from vaccine recipients who developed a rash after vaccination. Putative mechanisms other than rash (eg, aerosolization of the vaccine during administration to another patient) would be even more rare. Secondary cases of varicella caused by vOka have been typically mild.
Rash after varicella vaccination is uncommon; ∼4% of children and 6% of adolescent and adult vaccine recipients develop a rash after vaccination, and not all are vaccine associated.33 The rarity of the rash and subsequent transmission from vaccinated healthy persons support use of varicella vaccine for routine immunization. They also indicate the safety of vaccination of close contacts of susceptible persons at high risk for severe varicella who themselves have contraindications for vaccination. Considering the high infectivity of the wild-type virus (secondary household attack rate of ∼87%), assessing evidence of immunity to varicella of family members and other close contacts (eg, health care workers) of susceptible immunocompromised persons and subsequent vaccination of contacts without evidence of immunity is recommended.2,34 This scenario will occur more frequently as more conditions, including cancer, organ transplant comorbidities, and autoimmune conditions, are treated with immunosuppressive therapy. The benefits of herd immunity for immunocompromised persons vastly outweigh the risk of secondary transmission of vOka to those patients. It may, however, be prudent to avoid preparation of varicella vaccine with a susceptible immunocompromised person, pregnant woman, or newborn present in the room with the individual being immunized.18 In general, if a susceptible, immunocompromised person is exposed to a person who has a vaccine-related rash, passive immunization is not needed because disease associated with the vaccine virus is expected to be mild.2 Some experts, however, advise that passive immunization can be considered on a case-by-case basis in the face of vOka exposure depending on the degree of immunocompromise for susceptible immunocompromised persons.
Transmission of vOka from vaccine recipients without a rash has been exceedingly rare, possibly nonexistent. This is consistent with the rarity of respiratory spread of wild-type VZV. To date, there is no evidence of respiratory spread of the vaccine virus, and cultures of nasopharyngeal and throat specimens from vaccine recipients have not yielded vOka.26,35,36
Although varicella vaccine is not recommended for immunocompromised persons, experience from the clinical trial of varicella vaccine in immunocompromised children is instructive. In this trial, transmission occurred only if the vaccine recipients had a rash and it was positively correlated with vaccine recipients’ number of skin lesions; the more skin lesions were present, the more likely transmission occurred.17 Among healthy persons, varicella-related rash after vaccination besides being uncommon typically manifests with only a few lesions (median 2–5).33 Because of the low number of lesions, the risk for vOka transmission is thus expected to be low for healthy vaccine recipients even if they develop a rash. In instances of transmission of vOka from varicella-like rashes postvaccination in immunocompetent persons, vaccine recipients tended to have more lesions (median 30), but 2 had 2 and 12 lesions, respectively. These findings support the guidance that because of potential for transmission, vaccine recipients with a vaccine-related rash avoid contact for the duration of the rash with persons without evidence of varicella immunity at high risk for severe complications.2,34
When transmission occurred from healthy vaccine recipients, the secondary cases developed a rash 14 to 21 days after rash onset in the vaccine recipient. Disease presentation was mild with fewer lesions (median of 50 in the 13 secondary patients), shorter duration of illness, and no or low fever. One secondary patient had mild meningismus but no focal neurologic signs.8 These characteristics support infection with attenuated virus; in immunocompromised persons who were vaccinated, infection with the vaccine virus has also been reported to be milder than infection with wild-type virus.27 Oka strain has not been demonstrated to revert to wild-type VZV.17,24 Similarly, in immunocompromised contacts of vaccinated persons, if the virus is transmitted, illness is unlikely to be severe unless the patient is severely immunocompromised.37 Nonetheless, persons with immunocompromising conditions, regardless of degree of immunosuppression, should avoid contact with vaccine recipients who develop postvaccination rash.2
Live attenuated VZV in the varicella vaccine can establish latent infection and subsequently reactivate as HZ. Risk for HZ development is lower after vaccination than after natural VZV infection among both healthy and immunocompromised children; 1 large study reported a 78% lower HZ risk in healthy vaccinated children than in those unvaccinated.38 In varicella vaccine recipients, HZ can be caused by vaccine or wild-type strains. We identified 5 cases of secondary transmission of vOka from immunocompetent vaccine recipients with HZ, indicating that vaccine recipients who later develop HZ may rarely transmit vOka to susceptible children or adults. Therefore, no matter what strain of virus is involved, patients with HZ should be considered contagious and infection control precautions followed.39
Our study has several limitations. We report all cases of vaccine-strain transmission as they were classified by the authors. Some published reports relied on seroconversion data alone as a proxy for vaccine-strain transmission, although we found few case reports based on seroconversion in the absence of rash. Additionally, we classified as unsubstantiated transmission the few putative instances of vOka transmission not confirmed by appropriate laboratory testing. Because of the limited number of cases and absence of a denominator, we were unable to calculate rates of vOka transmission from healthy vaccine recipients. It is likely that other cases remain unreported in the literature; however, were this an important phenomenon, we believe more signals would have been present.
Pre- and postlicensure studies have shown that the varicella vaccine is safe, immunogenic, and effective in healthy persons and certain groups of immunocompromised patients.40,41 Routine varicella vaccination has significantly reduced morbidity and mortality. In the United States, during 2 decades of program implementation, incidence declined by 93% to 98% and hospitalizations and deaths by >93%.37,42,43 Globally, most studies have documented >80% reduction in hospitalizations and deaths44 after varicella vaccination. Review of the data suggests that healthy, vaccinated persons have minimal risk for transmitting the vaccine virus to contacts, particularly in the absence of vaccine rash in the vaccine recipient. Transmission occurred from vaccine recipients who developed a varicella-like rash or HZ after vaccination and was documented in close-contact settings (homes, long-term care facilities, and schools). Our findings support the existing recommendations for routine varicella vaccination and the guidance that persons with vaccine-related rash, particularly health care personnel and household contacts of immunocompromised persons, avoid contact for the duration of the rash with persons without evidence of immunity who are at high risk for severe complications.2,34
Acknowledgments
We thank Centers for Disease Control and Prevention colleagues Joanna Taliano for performing the literature search and Mary Ann Hall, MPH, for editorial review of the article.
Footnotes
- Accepted June 3, 2019.
- Address correspondence to Mona Marin, MD, Centers for Disease Control and Prevention, 1600 Clifton Rd NE, Mail Stop H24-5, Atlanta, GA 30333. E-mail: zsn8{at}cdc.gov
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention or the US Department of Health and Human Services.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
POTENTIAL CONFLICT OF INTEREST: Dr Gershon receives National Institutes of Health funding (R01DK03094) and has a contractual relationship with Merck through the Varicella Zoster Virus Identification Program; Drs Marin and Leung have indicated they have no potential conflicts of interest to disclose.
References
- Copyright © 2019 by the American Academy of Pediatrics
In this issue
