RESPONSE TO VARICELLA IMMUNIZATION IN PEDIATRIC LIVER TRANSPLANT RECIPIENTS

Abstract

Background. 

Individuals undergoing therapeutic immunosuppression are at risk of severe varicella-zoster virus (VZV) infection, particularly those without evidence of previous infection.

Methods. 

Eleven children, median age 10 months (range 5.5 months to 7 years and 9 months) received one dose of varicella vaccine (Varilrix, SmithKline Beecham plc, UK) before liver transplantation (median interval 95 days, range 40–289 days). The serological response to varicella vaccine was evaluated retrospectively and matched with the outcome and management of any subsequent exposures to VZV.

Results. 

Three children responded postimmunization, six children showed no response, and in two children the outcome was difficult to interpret having received blood products. Four children required varicella-zoster immunoglobulin prophylaxis posttransplantation, two of whom developed mild chickenpox.

Conclusions. 

Only 3 of 11 children developed a clear antibody response to varicella vaccine. Administration of varicella vaccine did not affect the management of subsequent VZV exposures.

Varicella-zoster virus (VZV) infection may cause significant morbidity and mortality in the immunocompromised host. In particular, malignant varicella is still an important cause of death in pediatric solid organ transplant recipients, despite the use of varicella-zoster immunoglobulin (VZIG) prophylaxis and aciclovir therapy (¹). The rising number of pediatric transplant recipients, their ages at transplantation and the use of therapeutic immunosuppression has highlighted problems in the management of exposure to chickenpox in this group. In the UK, most chickenpox infections are acquired by 10 years of age and therefore posttransplantation contact with chickenpox at nurseries and schools is commonplace.

Baseline VZV antibody testing should be carried out pretransplantation to determine susceptibility to infection. After transplantation, those susceptible children who have had a significant exposure to VZV should be offered prophylaxis. Previous exposure to VZV or recent administration of blood products warrants reevaluation of the antibody status to exclude subclinical infection or the presence of passively acquired antibody.

In 1975, Takahashi and colleagues (²) developed a live attenuated varicella-zoster virus vaccine (Oka strain) which has been shown to be safe and effective in healthy populations (³), and in specific groups with impaired immune function, notably children with acute lymphoblastic leukaemia (⁴). Immunization against varicella using a vaccine based upon the Oka strain is now recommended in the United States of America as part of the routine pediatric schedule.

Successful immunization against chickenpox is clearly desirable for individuals who are expected to receive therapeutic immunosuppression. This would avoid the need for repeated assessment regarding post-VZV contact prophylaxis in addition to protecting against unrecognized exposures. Experience with varicella vaccine in patients requiring solid organ transplantation is mainly limited to its use in children prior to renal transplantation (⁵).

Varicella vaccine (Varilrix, SmithKline Beecham plc, Welwyn Garden City, Herts, UK) has been given to a number of children listed for liver transplantation at King’s College Hospital, London. The aim of this retrospective study was to evaluate the postimmunization VZV antibody response together with the clinical outcome after subsequent exposure to VZV posttransplantation.

MATERIALS AND METHODS

Data were collected retrospectively on the 11 children eligible for the study. The inclusion criteria were administration of one dose of varicella vaccine (Varilrix, SmithKline Beecham plc) before liver transplantation, no history of chickenpox, a negative VZV IgG result before immunization, and availability of data regarding use of blood and blood products and immunosuppressive drug regimens and sera collected preimmunization and pre- and posttransplantation. All children received varicella vaccine containing a minimum of 10^3.3 plaque forming units. Eight children had extrahepatic biliary atresia (patients 1, 4, 5, 7, 8, 9, 10, and 11), two had progressive familial intrahepatic cholestasis (patients 2 and 6), and one had cryptogenic cirrhosis (patient 3). The median age at immunization was 9.5 months, with a range of 5.5 months to 7 years and 9 months, and the median interval from immunization to transplantation was 95 days, with a range of 40 to 289 days. Full contact histories had been collected, predominantly from patients’ relatives. Sera were stored at −20°C.

Varicella zoster antibody levels were measured using an ELISA format (Diamedix Corp. FL) according to the manufacturer’s instructions. Sera were tested in two lots. All of the samples from any one patient were tested in the same lot and several samples were tested in both to demonstrate reproducibility.

The following arbitrary units (AU) according to the manufacturer’s criteria were used to determine the antibody result: Less than 15 AU—VZV IgG not detected; between 15 and 19 AU—VZV IgG equivocal; more than or equal to 20 AU—VZV IgG positive.

A response to varicella vaccine was defined as a result of more than or equal to 20 AU for a serum sample taken within the first 6 to 10 weeks postimmunization in the absence of documented administration of blood products.

RESULTS

The data for the 11 children including the individual VZV antibody results, age at immunization, and timing of immunization in relation to transplantation are shown in Table 1. No major adverse effects of immunization were recorded.

Table 1: Varicella zoster vaccine response data

Of the 11 children included in the study, 3 showed evidence of a vaccine response (patients 7, 8, and 9), 6 showed no clear evidence of a vaccine response (patients 1–6), and the inability to distinguish between antibody generated by immunization and VZV antibody which may have been acquired passively from blood products resulted in two indeterminate outcomes (patients 10 and 11). The antibody levels in those children with evidence of a vaccine response were generally low and did not persist longer than 40 weeks postimmunization. A fluctuating level of VZV antibody was seen in consecutive samples collected from three children (patients 1, 4, and 9).

Four of the 11 children required VZIG prophylaxis posttransplantation. These included two nonresponders to vaccine (patients 3 and 5) and two children with indeterminate outcomes postimmunization (patients 10 and 11). Child 5 and child 11 developed chickenpox, at 51 weeks and 3.5 years postimmunization, respectively.

DISCUSSION

This retrospective study revealed a postimmunization response in only 3 of 11 children. In comparison, more than 90% of healthy children usually develop vaccine induced antibody, often within 6 weeks of immunization, which is maintained in the majority for at least 1 year (⁶). Three children (patients 1, 4, and 9) showed fluctuating antibody levels without an apparent cause, that may be accounted for by incomplete blood product data, subclinical infection or, if the results were equivocal, by nonspecific assay reactivity. Specifically, the results for child 9 showed a possible response to vaccination, an initial decline in antibody titer which may have been due to immunosuppression and then a reappearance of antibody at 24 weeks postimmunization. In the absence of documented administration of blood products or VZV exposure, this reappearance of high level VZV antibody may have resulted from a subclinical exposure to VZV and an anamnestic antibody response in the presence of vaccine-mediated protection against disease. In addition, for child 6, the detection of VZV antibody in samples collected at 17 and 60 weeks postimmunization was difficult to interpret because blood product administration data were not available for that time period.

There are several possible explanations for the poor vaccine response in our patients. Six children were younger than 1 year of age at immunization (patients 1, 3, 4, 7, 8, and 9). This has been identified as a risk factor for breakthrough varicella in healthy children (⁷). The poor response to varicella vaccine in infants is consistent with the observation of early subclinical VZV reactivation after natural VZV infection in that age group (⁸). Despite this, the older children in our study did not consistently show a better response to immunization.

Antibody passively acquired both pre- and postimmunization may interfere with the immune response to viral vaccines. Although there is no specific data for varicella vaccine, guidelines suggest that, if possible, blood products that may contain VZV antibody should not be given 5 months before, or 3 weeks after, immunization. There is no clear evidence for this effect in the children shown, although patients 5, 7, and 10 received blood products within this time period.

Initiation of therapeutic immunosuppression close to the time of immunization may be responsible for the observed early loss of varicella vaccine-induced antibody and the initial low antibody titers. Four children were transplanted within 8 weeks of receiving vaccine (patients 6, 7, 9, and 10). Children undergoing chemotherapy for various malignancies can show an antibody response to vaccine within 6 weeks (⁹) but the duration and level of this response may be reduced.

Alternatively, the underlying end stage liver disease of these children may be the important factor in vaccine failure. There are little published data for comparison, but a brief report including 18 children requiring liver transplantation showed an antibody response rate of 50% to a single dose of varicella vaccine (¹⁰). Comparison with this study is difficult as individual patient details were not given. However, none of the children were younger than 18 months of age. Patients awaiting renal transplantation have responded well to varicella immunization (⁵), including some individuals who had asymptomatic seroconversion and a reduced severity of varicella posttransplantation despite the absence of a detectable antibody response to vaccine.

These results have implications for the clinical management of this group of children after a significant exposure to VZV. Our experience has been that the use of single dose varicella vaccine has not greatly influenced either the management of exposure to VZV or the outcome of these events. Four children (patients 3, 5, 10, and 11) have already required VZIG posttransplant, two of whom developed mild varicella (patients 5 and 11). It is unclear whether the mild nature of the varicella infection in these two children resulted from a varicella vaccine effect, VZIG prophylaxis, or therapy with aciclovir. Protection against severe varicella may remain after loss of vaccine-induced antibody, as cell-mediated immunity stimulated by varicella infection or immunization is thought to be more important in protection against, and recovery from, disease. Despite this and given our findings, we believe assessment of immunity and targeted prophylaxis to be the safest option in our group, regardless of immunization history.

Because VZIG may be in short supply, and is often given repeatedly to the same patient, an effective pretransplantation varicella immunization protocol is desirable. These initial findings have highlighted the need for a prospective study of varicella vaccine efficacy in children listed for liver transplantation. Ideally, any future research would involve the evaluation of booster doses and assessment of both the antibody and cell-mediated immune response to varicella vaccine. In the presence of clear evidence of seroconversion after varicella immunization, protective immunity could be assumed and the contact observed with the option of early use of antiviral therapy if required. Alternatively, the VZ antibody status could be determined and VZIG or aciclovir prophylaxis given as necessary.