Safety and immunogenicity of varicella vaccine in patients with juvenile rheumatic diseases receiving methotrexate and corticosteroids


  • Gecilmara Salviato Pileggi,

    1. School of Medicine of Ribeirão Preto, University of São Paulo, Ribeirão Preto, Brazil
    Search for more papers by this author
  • Cleonice Barbosa Sandoval de Souza,

    1. School of Medicine of Ribeirão Preto, University of São Paulo, Ribeirão Preto, Brazil
    Search for more papers by this author
  • Virgínia Paes Leme Ferriani

    Corresponding author
    1. School of Medicine of Ribeirão Preto, University of São Paulo, Ribeirão Preto, Brazil
    • Department of Pediatrics, School of Medicine of Ribeirão Preto, University of São Paulo, Avenida Bandeirantes, 3900, 14049-900 Ribeirão Preto, São Paulo, Brazil
    Search for more papers by this author



To evaluate the safety and immunogenicity of varicella vaccine (VV) in susceptible patients with juvenile rheumatic diseases receiving methotrexate and corticosteroids.


Twenty-five patients with juvenile rheumatic diseases (ages 2–19 years) and 18 healthy children and adolescents (ages 3–18 years) received a single dose of VV. All 25 patients were receiving methotrexate; 13 were also receiving prednisone and 5 were also receiving other disease-modifying antirheumatic drugs. None of the vaccinated patients or controls had a previous history of varicella. Anti–varicella-zoster virus IgG antibody (anti–VZV-IgG) titers were measured by enzyme-linked immunosorbent assay immediately before, 4–6 weeks after, and 1 year after vaccination. The patients were monitored prospectively for adverse reactions related to the vaccine, exposure, and occurrence of varicella. Disease activity was assessed 3 months before and 3 months after VV.


Twenty patients and all of the controls had negative preimmunization titers of VZV-IgG, and 5 patients had equivocal levels. Positive VZV-IgG titers were detected in 10 (50%) of 20 seronegative patients and 13 (72.2%) of 18 controls 4–6 weeks after VV (P = 0.2). One year after vaccination, 8 of 10 patients maintained positive VZV-IgG titers. No overt varicella episodes and no severe adverse reactions were observed during the followup period. No worsening of clinical parameters and no flares of juvenile rheumatic diseases or changes in doses of medications used were detected after vaccination. In fact, the number of active joints in patients with juvenile idiopathic arthritis was significantly lower after VV (P = 0.009).


VV appears to be safe in patients with juvenile rheumatic diseases receiving methotrexate, as long as continuous prospective vigilance for side effects is performed.


Varicella (chickenpox), the primary varicella-zoster virus (VZV) infection, is a highly contagious disease that occurs worldwide and is endemic in many countries (1). Control measures aiming to reduce varicella infection, including universal vaccination, are important to prevent serious complications known to occur even in immunocompetent subjects, but especially in children with chronic diseases such as juvenile rheumatic diseases. The effective drugs used to treat juvenile rheumatic diseases, including disease-modifying antirheumatic drugs (DMARDs), corticosteroids, and biologic agents, although promoting control of inflammation and autoimmune mechanisms, usually induce a variable degree of immunosuppression (2).

Although there is no doubt that treatment with high-dose systemic corticosteroids results in increased vulnerability to infections, particularly viruses such as chickenpox, there is no consensus on the minimum dose likely to cause clinically significant immunosuppression. Similarly, for methotrexate (MTX) there is no published data on the dose at which immunosuppression is likely to occur (3).

This lack of evidence forces manufacturers to advise caution and recommend that live vaccines should be withheld in any patient receiving MTX, regardless of dose (2). However, many patients receiving MTX and other immunosuppressive drugs have no immunity to varicella and remain at risk.

Patients with juvenile rheumatic diseases exposed to VZV can develop macrophage activation syndrome (4) and also invasive group A streptococci infections with necrotizing fasciitis when continually receiving nonsteroidal antiinflammatory drugs (5). In this context, the question arises as to whether varicella vaccine (VV) should be avoided in these patients.

Although there are case reports on serious adverse events after live vaccine administration to immunocompromised hosts, interestingly, none are related to patients with rheumatic disease. In addition, the risk–benefit relationship of patients with immunizing cancer was proven to be favorable, leading to the routine immunization of oncology patients against VV in several countries (6).

In a recent American College of Rheumatology (ACR) position paper on recommendations to patients with rheumatoid arthritis receiving DMARDs, the authors do not formally contraindicate the use of live attenuated vaccines (7) in those receiving nonbiologic drugs. Moreover, the Advisory Committee on Immunization Practice guidelines state that zoster vaccination (a live attenuated vaccine) is indicated in patients with inflammatory disorders who are receiving low-dose MTX, azathioprine, or corticosteroids in dosages lower than 20 mg/day (8).

Until now, there has been no information on the use of live VV in children and adolescents with juvenile rheumatic diseases, and it is not known if the benefits of the vaccine would outweigh the disadvantages in these patients.

In the present study, we evaluated the safety and immunogenicity of VV in susceptible patients with juvenile rheumatic diseases receiving MTX and corticosteroids at our center.

Patients and Methods

This was an open-label prospective study conducted at the Paediatric Rheumatology Unit of the Clinical Hospital of the School of Medicine of Ribeirão Preto, University of São Paulo. The study was approved by the Institutional Board of Ethics in Medical Research and written consent was obtained on the first visit from all of the participants and legal guardians after a full explanation of the study.

Patients with juvenile rheumatic diseases and no history of chickenpox followed at our center from August 2005 to June 2007 who met the following inclusion criteria were invited to receive VV: between ages 1 and 20 years; diagnosis of juvenile idiopathic arthritis (JIA), juvenile dermatomyositis (DM), scleroderma, or vasculitis according to published criteria; current use of stable doses of MTX, cyclosporin A, leflunomide, D-penicillamine, or prednisone (maximum 20 mg/day) for at least 30 days before vaccination; negative or equivocal serum levels of antibodies to VZV; adequate lymphocyte count in peripheral blood (>0.7 × 109/liter); and negative test for pregnancy in female adolescents.

The subjects were excluded if they had received any blood products in the last 3 months or any other vaccine; cyclophosphamide, pulse methylprednisolone, or anti–tumor necrosis factor (anti-TNF) drugs in the last 30 days; and also if they were exposed to VZV in the month before immunization or were known to be hypersensitive to vaccine components.

Healthy children and adolescents with ages and social conditions similar to the patients and no previous history of varicella were included as controls.

Patients and controls received a single dose (0.5 ml) of live attenuated VZV containing >1,000 plaque-forming units of the VZV (Oka strain) by subcutaneous route into the right upper arm. The vaccine was stored, handled, and administered according to the information on the package insert.

Followup and vaccine safety assessment.

The patients were carefully monitored for vaccine side effects (including fever, site reactions, and skin rash) through phone calls (weekly) and outpatient visits to the hospital (biweekly) during a period of 40 days after vaccination. The parents received standardized diaries to record adverse reactions and were encouraged to bring children to the hospital if any significant clinical symptoms developed.

Disease activity, always assessed by the same pediatric rheumatologist (GSP), was compared during the period of 3 months before and after varicella vaccination. For patients with JIA, the clinical status was assessed using the core set of criteria for disease activity proposed by the ACR (the ACR Pediatric 30 criteria [ACR Pedi 30]) (9, 10). The physician's global assessment of disease activity and the parents' assessment of overall well-being, both measured on a 10-cm visual analog scale, and the erythrocyte sedimentation rate were used for patients with other rheumatic diseases. Active disease was considered when the physician's global assessment of disease activity was >1 (11).

Secondary outcome parameters were flare occurrence for patients with JIA and changes in current therapeutic plan (increasing doses of already received medications or introduction of new drugs) for the entire group of patients. A flare was defined as a worsening of ≥40% in ≥2 disease activity parameters of the ACR Pedi 30 core set without a simultaneous improvement of ≥30% in ≥2 of the remaining parameters (12).

Assessment of the serologic response to VV.

Serum samples were obtained immediately before, between 4 and 6 weeks after, and 1 year after vaccination. The samples were frozen and stored at −20°C until serologic analysis. Concentration of VZV-specific IgG antibody (VZV-IgG) in serum was measured using a commercial enzyme-linked immunosorbent assay (Dade-Behring Enzygnost Immunoassay kit) as already described. Corrected absorbance values were converted to log10 mIU/ml and the results were classified as negative (values <50 mIU/ml), equivocal (between 50 and 100 mIU/ml), and positive (>100 mIU/ml) according to the manufacturer's recommendations (13).

Statistical analysis.

VZV-IgG titers of patients and controls obtained at 4–6 weeks after vaccination were compared using Wilcoxon's signed rank test, and vaccine response rates in patients and controls were compared using a Fisher's exact test. Mean values of each individual disease activity parameter and the number of flares during the period of 3 months before and after VV were compared using the Wilcoxon's signed rank test for paired samples. A P value of less than 0.05 was considered as statistically significant.


Baseline characteristics of the study population.

Twenty-five patients (13 girls) between ages 2 and 19 years (median 7.2 years) were vaccinated: 17 with JIA, 4 with juvenile DM, 3 with juvenile scleroderma, and 1 with vasculitis. All of the subjects were receiving MTX at baseline (mean dosage of 16.4 mg/m2/week, range 10–27). Thirteen of them were also receiving 0.1–0.7 mg/kg/day of prednisone (mean total dosage of 4.2 mg/day, range 3–20), and 5 also received other DMARDs: 3 patients were receiving cyclosporine (3–3.5 mg/kg/day), 1 was receiving penicillamine (13 mg/kg/day), and 1 was receiving leflunomide (10 mg/day).

Twenty patients had negative preimmunization VZV-IgG levels; 5 patients had equivocal levels and were excluded from the vaccine response analysis, but were kept in the safety evaluation. Eighteen VZV-seronegative healthy children and adolescents (ages 3–18 years, median age 9.3 years) were vaccinated as controls.

The baseline characteristics of the 25 vaccinated patients and the responsiveness data of the 20 seronegative patients are shown in Table 1.

Table 1. Clinical characteristics, medications used, response to varicella vaccine, and occurrence of vaccine-related rash in 25 vaccinated patients*
PatientAge, yearsDisease typeDisease activityMedication usedAnti–VZV-IgG titersVaccine-related rash
MTX, mg/m2/weekPrednisone, mg/dayOther4–6 weeks1 year
  • *

    MTX = methotrexate; anti–VZV-IgG = anti–varicella-zoster virus IgG; DM = dermatomyositis; JIA = juvenile idiopathic arthritis; ND = not done.

  • Active disease was considered when the physician's global assessment of disease activity was >1.

  • Interpretation of the response to the varicella vaccine: anti–VZV-IgG titers >100 mIU/ml = responders, between 50 and 100 mIU/ml = equivocal response, and <50 mIU/ml = nonresponders.

  • §


  • Patients with equivocal anti–VZV-IgG titers at baseline not included in the responsiveness evaluation.

13Juvenile DMNo122 78400§No
212VasculitisYes2010 <5055No
32Juvenile sclerodermaYes207.5 <50<50No
47Juvenile DMYes1510Cyclosporine (3 mg/kg/day)<50<50No
514Oligoarticular JIANo120 60<50No
64Polyarticular JIAYes1510 947§290§No
713Systemic JIAYes255Leflunomide (10 mg/day)192§330§Yes
85Polyarticular JIANo150 120§<50No
95Polyarticular JIANo130 6256No
104Polyarticular JIAYes205 120§<50No
115Polyarticular JIAYes16.60 567§485§No
128Polyarticular JIAYes200 <50<50No
137Systemic JIAYes209Cyclosporine (3 mg/kg/day)1,458§1,300§Yes
144Systemic JIANo150 130§2,800§No
158Polyarticular JIANo150 635§140§No
166Systemic JIAYes1810 <50<50No
1712Polyarticular JIAYes150 6455No
1815Polyarticular JIANo130 <50<50No
196Juvenile sclerodermaYes2520 947§210§No
204Oligoarticular JIANo190 335§820§Yes
215Juvenile sclerodermaYes16.67.5Penicillamine (13 mg/kg/day)947§340§No
229Polyarticular JIAYes220 335§150§No
2311Systemic JIANo180 3,316§NDNo
2413Juvenile DMYes2010 982§NDNo
2519Juvenile DMNo135Cyclosporine (3.5 mg/kg/day)2,230§NDNo

Safety evaluation: adverse events.

No overt varicella episodes were observed after vaccination in the entire group of 25 patients and 18 controls. None of them developed local reactions to the vaccine. Fever related to VV occurred in 1 patient and 1 control; both had low-grade fever episodes (<38°C) starting on day 2 after VV and persisting for 1 day.

Three patients (15%; patients 7, 13, and 20 in Table 1) developed a mild self-limited varicella-like rash with few vesicular lesions (6–12 per patient) during the first 2 weeks postvaccination. The rash was not accompanied by any other symptoms and lasted 5–7 days. All of the patients had a good outcome, and the use of acyclovir was not necessary. Interestingly, the patient receiving the higher dose of corticosteroids and MTX did not have any side effects and showed good response to the vaccine (patient 19, Table 1).

Exposures and followup.

The median followup period after vaccination for the 25 patients was 32 months (range 24–36 months). Sixteen patients could precisely report about exposure to wild VZV, and 8 of them reported household or classmate contact in this period. Two of those 8 patients (both nonresponders to VV) developed chickenpox. One was receiving anti-TNF therapy at the time of the contact and developed severe varicella, complicated with pneumonia and probable macrophage activation syndrome. None of the other 23 patients developed herpes zoster or chickenpox during the followup, including those who had had close contact with the wild virus in the household environment. None of the 15 children in the control group who had completed at least 2 years of followup developed chickenpox.

Disease activity and medication use.

No worsening of any disease activity parameter was observed in the entire group of 25 patients over the period of 3 months after vaccination when compared with 3 months before. In fact, a significant improvement in the mean number of active joints over the period of 3 months after receiving the vaccine (P = 0.009) was observed in the group of 17 JIA patients (Table 2). In addition, no flares were detected during the period of 3 months after the vaccination. Regarding medication use, the mean doses of MTX, other DMARDs, and oral steroids were not increased, and it was not necessary to introduce new drugs during the period after VV in the entire group of patients.

Table 2. Disease activity parameters assessed 1 and 3 months before and after varicella vaccination in 17 patients with juvenile idiopathic arthritis*
 Before, mean (95% CI)After, mean (95% CI)P
  • *

    95% CI = 95% confidence interval. C-HAQ = Childhood Health Assessment Questionnaire; ESR = erythrocyte sedimentation rate.

  • By Wilcoxon's signed rank test.

  • Significant, P < 0.05.

  • §

    Measured on a 10-cm visual analog scale.

No. of joints with active arthritis3.2 (1.4–5)1.8 (0.8–2.8)0.009
No. of joints with limited range of motion0.8 (0.3–1.3)0.7 (0.3–1.2)0.94
C-HAQ score0.4 (0.1–0.6)0.3 (0.1–0.4)0.19
Parent's global assessment of overall well-being§2.1 (0.8–3.5)1.6 (0.6–2.7)0.23
Physician's global assessment of disease activity§2.4 (1–3.8)1.7 (0.7–2.8)0.077
ESR, mm/hour26.6 (18–35)24.3 (18–31)0.56

Immunogenicity to the vaccine.

Positive VZV-IgG titers were reached at 4–6 weeks after vaccination in 10 (50%) of 20 previously seronegative patients and in 13 (72.2%) of 18 controls. The response was equivocal in 4 patients (20%) and 3 controls (16.6%). Vaccine response rates and median postimmunization VZV-IgG titers were not different when patients and controls were compared (P = 0.2 and P = 0.74, respectively) (Figure 1). One year after receiving VV, 8 (80%) of 10 seroconverted patients maintained positive VZV-IgG titers.

Figure 1.

Median values (thick lines), interquartile ranges (boxes), and minimum and maximum values (whiskers) of anti–varicella-zoster virus IgG (anti–VZV-IgG) titers (mIU/ml) measured in patients with juvenile rheumatic diseases and controls 4–6 weeks after vaccination. P = 0.74 by Wilcoxon's signed rank test.


To the best of our knowledge, this is the first study evaluating the safety and immunogenicity of VV in children and adolescents with juvenile rheumatic diseases receiving MTX and corticosteroids. The vaccine proved to be safe in this group of patients. No overt varicella episodes and no severe adverse effects related to the VV were observed during the followup period, even in patients receiving the highest doses of corticosteroids and MTX. The frequency of varicella-like rash was higher in vaccinated patients (15%) in this study as compared with rates described in the literature for healthy children (5%) (14), but none of our patients required acyclovir treatment. Furthermore, no flares of rheumatic diseases were detected in our study, and changes in therapeutic plans were unnecessary after vaccination. In fact, the number of active joints was lower after VV in patients with JIA. The question of whether vaccination could induce the onset of symptoms or trigger flares in patients with rheumatic disease is a matter of debate (15, 16).

There are only 2 published studies regarding the safety of using live vaccines in juvenile rheumatic diseases. Both of them showed that the measles, mumps, and rubella booster vaccination was safe in JIA patients, including those who were taking prednisone, MTX, or etanercept (17, 18). Studies of effects of inactivated vaccines, including influenza vaccine (19, 20), hepatitis B vaccine (21), and meningococcal C vaccine (22), have not shown changes in disease activity or an increase in the occurrence of flares after vaccination in patients with JIA.

Regarding responsiveness to the VV observed in the present study, both patients and controls had a lower seroconversion rate when compared with those of a previous large study with the same vaccine that reported seroconversion rates in healthy children greater than 90% (23). However, a recent study showed a seroconversion rate in susceptible healthy children of 76% after one dose of VV (24), suggesting that one dose of this vaccine could lead to primary vaccine failure. In fact, the last published guidelines on VV from the Centers for Disease Control and Prevention (June 2007) advocated the use of 2 doses of the vaccine for all children after 1 year of age, independently of the immunologic situation (14).

One caveat of our study is the low number of patients susceptible to varicella included, but this reflects the rigorous criteria used for the selection of patients, leading to a small but fairly homogeneous group followed in a single reference center. Multicenter studies with larger numbers of patients are clearly needed to estimate the VV responsiveness in patients with juvenile rheumatic diseases.

In conclusion, our results show that VV appears to be safe in susceptible children and adolescents with rheumatic diseases taking MTX. As long as continuous prospective vigilance for side effects is performed, we believe that the risk of being exposed to the varicella disease in these patients is higher than the risk of receiving the vaccine.

More research in this field is necessary for the development of specific immunization guidelines for patients with rheumatic and other autoimmune diseases receiving treatment with immunosuppressive agents who are still susceptible to preventable infectious diseases.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ferriani had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Pileggi, Ferriani.

Acquisition of data. Pileggi, de Souza, Ferriani.

Analysis and interpretation of data. Pileggi, de Souza, Ferriani.