• 
• 

Three Part Question

In [a neonate born to a mother with chickenpox around the time of delivery] does [varicella-zoster immunoglobulin] effectively [prevent or ameliorate chickenpox]?

Clinical Scenario

On a routine neonatal check on the postnatal ward you come across a neonate whose mother developed chickenpox 3 days before delivery. You are aware that according to UK guidelines a baby should be given prophylactic varicella-zoster immunoglobulin (VZIG) if its mother develops chickenpox during the period that spans 7 days prior to delivery to 7 days after (and that the US Red Book recommends VZIG for neonates born to mothers with onset of chickenpox 5 days prior to delivery to 2 days after). The mother, who is keen for her baby to have this preventive treatment, asks you how likely it is that this measure will prevent the disease. You are unsure and consult the Department of Health (UK) publication Immunisation against Infectious Diseases (the Green Book), which provides limited information on this aspect, citing one reference only. The British National Formulary for Children (2007) offers even fewer details, and refers you to the former publication. Therefore, you wonder what evidence exists for the effectiveness of VZIG in preventing or modifying chickenpox in this setting.

Search Strategy

Search of the Cochrane Library database using the terms "chickenpox" and "varicella" yielded no relevant reviews. Medline search using the PubMed interface (1950 to date/no limits set): (1) using the MeSH terms "infant, newborn", "chickenpox" and "varicella-zoster immune globulin" joined by the Boolean operator AND revealed 27 publications, only one of which was relevant, and (2) using (congenital OR neonatal OR perinatal) AND (varicella OR chickenpox) AND zoster immunoglobulin yielded 199 publications, only five of which were relevant.3–7 Search of the ISI Web of Science using the second strategy yielded 36 publications, but no further relevant publications were identified. Case reports and case series with less than five patients were excluded. In addition, the UK, US, Australian, Canadian and German immunisation guidelines were searched for relevant references, but no further references were identified. Search of multiple trials registers (MRC, NHS, Wellcome Trust, NIH and ISRCTN) identified no relevant trials. All relevant publications were searched for further references, but no additional reports were identified.
The search date was 1 August 2008.

Search Outcome

5 were relevant

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Miller et al, 1999,	167 neonates whose mother developed chickenpox between 7 days prior to and 7 days after delivery.Dose of VZIG: 100 or 250 mg.	Prospective, non-randomised study(level 4)	Seroconversion (VZV-IgM)andclinically-apparent chickenpox. Death	109 (65%) seroconverted indicating infection:12 (7%) asymptomatic,79 (47%) mild chickenpox,18 (11%) severe chickenpox.58 (35%) no serological or clinical evidence of infection. No deaths.	Part of a larger study population of neonates exposed during the peripartum period. Definition of mild and severe disease not provided.
Carter et al, 1986,	5 neonates whose mother developed chickenpox between 10 days prior to and 2 days after delivery.Doses of VZIG used: 150 or 500 mg.	Case series (level 4)	Clinically-apparent chickenpox. Death	1 (20%) asymptomatic,2 (40%) mild disease,2 (40%)	Part of a larger case series of neonates exposed during the peripartum period. Definition of mild and severe disease not provided. All cases additionally received acyclovir, which may have prevented disease.Unclear whether "skin desquamation" indicates infection.
Hanngren et al, 1985,	41 neonates whose mother developed chickenpox between 4 days prior to and 2 days after delivery. Dose of VZIG: 1 ml.	Prospective, non-randomised study (level 4)	Clinically-apparent chickenpox. Death	20 (49%) asymptomatic,13 (32%) very mild symptoms,6 (15%) mild/normal symptoms,2 (5%) severe chickenpox. No deaths.	Part of a larger study population of neonates exposed during the peripartum period. Very mild disease defined as <20 pocks and apyrexial. Definition of mild/normal and severe disease not provided. The 2 neonates classified as severe disease were not profoundly unwell according to the case descriptions. Seroconversion was only evaluated in a small subgroup of patients.
Evans et al, 1980,	11 neonates whose mother developed chickenpox postnatally (timing not specified).Only 9 received VZIG (VZIG group), 2 did not (non-VZIG group).Dose of VZIG: 100 mg.	Case series (level 4)	Clinically-apparent chickenpox. Death	VZIG group:3 (33%) asymptomatic,6 (67%) mild chickenpox.Non-VZIG group:2 (100%) chickenpox.p = 1.00 for difference between groups* No deaths	Part of a larger cohort, which included older infants and immunocompromised patients.2 patients received VZIG with a significant delay (between 4–6 days after exposure).Report does not provide a detailed description of the severity of illness.
Winsnes R, 1978,	7 neonates whose mother developed chickenpox up to 14 days prior to delivery(5 within 4 days before delivery).Dose of VZIG: 1 ml.	Case series (level 4)	Clinically-apparent chickenpox. Death	5 (71%) asymptomatic,2 (29%) very mild disease. No deaths.	Part of a larger cohort, which included immunocompromised patients. Definition of mild and severe disease not provided. Some patients received VZIG with a delay of >3 days; it is unclear whether this included any of the neonates in this study.

Comment(s)

Primary varicella-zoster virus (VZV) infection in pregnancy is a relatively rare event. However, a questionnaire-based survey in the UK suggests that the incidence of chickenpox in pregnant women is at least 1 per 2000 pregnancies. Based on epidemiological data it has been estimated that there are a total of 2000 cases per year in England and Wales (Nathwani et al).

Primary maternal VZV infection in late pregnancy can result in congenital varicella in the neonate, defined as clinically apparent infection at birth or up to 10 days after delivery. Clinical disease after the 10th day of life is termed neonatal varicella and results from postnatal exposure to chickenpox in contrast to the trans-placental infection that causes congenital varicella.

As a result of the higher viral load associated with trans-placental infection, congenital chickenpox is associated with a relatively high risk of progression to severe disease and an overall mortality rate of around 10%.13 Older reports suggest that newborns of mother who develop chickenpox in the period between 4 days prior to delivery and 2 days after delivery are at greatest risk, with an estimated mortality rate of approximately 20–30% (Meyers, DeNicola). However, based on data from the Public Health Laboratory Service (now the Health Protection Agency), it has been suggested this may be an overestimate (Miller). Also, with modern intensive care support and prompt treatment with acyclovir, it is likely that mortality rates would be lower today.

After the seminal report by Brunell et al in 1969 demonstrating the efficacy of VZIG in preventing chickenpox in healthy children, the inclusion of untreated individuals or placebo controls in subsequent trials in high-risk individuals (including neonates) became unethical. In this study, two non-immune siblings in families with an index case with chickenpox were randomised such that one received normal immunoglobulin (n = 6) and the other VZIG (n = 6). All in the former group developed chickenpox, but none of the children who had received VZIG developed disease. These results were not statistically analysed in the original paper, but the difference between the two groups is highly statistically significant (two-sided Fisher’s exact test: p = 0.0022). A subsequent report showed that VZIG contains up to 30 times higher varicella antibody titres than normal immunoglobulin products (Brunell), providing an explanation for the greater effectiveness of VZIG.

Only two of the reports identified in the literature search were prospective studies; the remaining three were relatively small case series. None of the studies included randomisation or a placebo control group (for reasons outlined above). However, one report included two neonates who did not receive VZIG, both of whom developed chickenpox (Evans).

All five reports recorded the incidence of clinically apparent chickenpox. The proportion of neonates who developed no clinical symptoms after receiving VZIG ranged from 33% to 71% in individual reports. The two prospective studies reported by Miller et al and Hanngren et al, which also included the largest number of patients and are therefore most likely to provide the most reliable data, reported that clinically apparent disease was prevented in 42% and 49% of the neonates, respectively. Serological testing in the former study revealed that 7% of the neonates had experienced subclinical infection.

Of those neonates who developed chickenpox despite VZIG prophylaxis, the majority had only mild disease. Only two studies – those by Hanngren et al and Miller et al – reported cases with severe disease, accounting for 5% and 11% of the study population, respectively. Although classified as "severe", both children in this category in the former report had relatively minor symptoms without evidence of organ involvement, and did not require intensive support. None of the publications included here reported any fatalities

In addition to the absence of control groups, there are further limitations that complicate the interpretation of these studies. Firstly, the studies used varying definitions for "perinatal chickenpox" and therefore different inclusion criteria, with the onset of maternal chickenpox ranging from up to 14 days prior to delivery to only 4 days prior to delivery. Secondly, different doses and preparations of VZIG were used in different studies and sometimes even within the same study. Finally, none of the studies provided a precise definition of disease severity.

In view of the ongoing global shortage of VZIG, the indications for use and the potential benefits for at-risk groups need careful evaluation. Despite a number of limitations, when compared with the findings from historical reports, the data from the studies summarised here provide considerable evidence that VZIG prevents chickenpox in a large proportion of perinatally exposed neonates. In addition, they provide reasonable evidence that VZIG ameliorates the course of the disease in cases in which the infection is not fully prevented. However, it is important to highlight that several case reports have described neonates with severe congenital varicella despite VZIG prophylaxis, predominately manifesting as respiratory compromise associated with interstitial pneumonitis, with fatal outcome in some cases (Bakshi, Reynolds, Holland). Therefore, parents need to be made aware that VZIG may not prevent disease and that they should seek medical attention immediately if their baby becomes unwell. VZIG can prolong the incubation period for up to 30 days6 and therefore vigilance should be maintained for this duration.

Editor Comment

*Significance calculated with two-sided Fisher’s exact test.

Clinical Bottom Line

Varicella-zoster immunoglobulin (VZIG) prevents clinically-apparent chickenpox in approximately half of neonates born to mothers with chickenpox around the time of delivery. (Grade C)

VZIG ameliorates the course of chickenpox in most neonates in whom disease is not prevented. (Grade C))

A small number of cases of severe congenital varicella despite VZIG prophylaxis have been reported. (Grade A)

References

1. Salisbury D, Ramsay M, Noakes K, eds. Immunisation against Infectious Diseases . London: Department of Health 2006:421–2.
2. American Academy of Pediatrics. The Red Book: 2006 report of the Committee on Infectious Diseases . American Academy of Pediatrics 27th edn. Elk Grove Village, IL: , 2006:711–25.
3. Miller E, Cradock-Watson JE, Ridehalgh MK. Outcome in newborn babies given anti-varicella-zoster immunoglobulin after perinatal maternal infection with varicella-zoster virus. Lancet 1989;2:371–3.
4. Carter PE, Duffy P, Lloyd DJ. Neonatal varicella infection. Lancet 1986;2:1459–60.
5. Hanngren K, Grandien M, Granstrom G. Effect of zoster immunoglobulin for varicella prophylaxis in the newborn. Scand J Infect Dis 1985;17:343–7.
6. Evans EB, Pollock TM, Cradock-Watson JE, et al.. Human anti-chickenpox immunoglobulin in the prevention of chickenpox. Lancet 1980;1:354–6.
7. Winsnes R. Efficacy of zoster immunoglobulin in prophylaxis of varicella in high-risk patients. Acta Paediatr Scand 1978;67:77–82.
8. Centers for Disease Control (The Advisory Committee on Immunization Practices (ACIP)). Recommendations for the use of VZIG 2007. MMWR Morb Mortal Wkly Rep 2007;56(RR-4):1–40.
9. Department of Human Services. The Australian Immunisation Handbook . 9th edn. Canberra, ACT: DHS, 2008.
10. Public Health Agency of Canada. Canadian immunization guide. 7th edn. Ottawa: PHA, 2006.
11. Ständige Impfkommission Deutschland (STIKO)/Robert Koch Institut. Varizellen (Windpocken), Herpes zoster (Gürtelrose). Available from: www.rki.de/cln_100/nn_504678/DE/Content/Infekt/EpidBull/Merkblaetter/Ratgeber__Mbl__Varizellen.html (accessed 24 April 2009).
12. Nathwani D, Maclean A, Conway S, et al. Varicella infections in pregnancy and the newborn. A review prepared for the UK Advisory Group on Chickenpox on behalf of the British Society for the Study of infection. J Infect 1998;36(Suppl 1):59–71.
13. Meyers JD. Congenital varicella in term infants: risk reconsidered. J Infect Dis 1974;129:215–17.
14. DeNicola LK, Hanshaw JB. Congenital and neonatal varicella. J Pediatr 1979;94:175–6.
15. Brunell PA, Ross A, Miller LH, et al. Prevention of varicella by zoster immune globulin. N Engl J Med 1969;280:1191–4.
16. Hanngren K, Falksveden L, Grandien M, et al Zoster immunoglobulin in varicella prophylaxis. A study among high-risk patients. Scand J Infect Dis 1983; 15:327–34.
17. Bakshi SS, Miller TC, Kaplan M, et al. Failure of varicella-zoster immunoglobulin in modification of severe congenital varicella. Pediatr Infect Dis 1986;5:699–702.
18. Reynolds L, Struik S, Nadel S. Neonatal varicella: varicella zoster immunoglobulin (VZIG) does not prevent disease. Arch Dis Child Fetal Neonatal Ed 1999;81:F69–70.
19. Holland P, Isaacs D, Moxon ER. Fatal neonatal varicella infection. Lancet 1986;2:1156