Objective. To determine the long-term prevalence of varicella infection and herpes zoster after kidney transplantation and to assess the effectiveness of varicella immunization with the Oka attenuated strain.
Methods. This study involved 704 children and adolescents who received a kidney graft in our institution from 1973 to 1994 and had routinely been given varicella vaccine beginning in 1980 in preparation for transplantation.
Results. After vaccination 62% of these patients still had varicella/zoster (VZ) antibodies at 1 year and 42% after 10 years. After transplantation the incidence of varicella was significantly lower, 26/212 (12%), in patients who received immunization than in those who did not and had no history of varicella, 22/49 (45%). The disease was also significantly less severe in the vaccinated patients (three deaths among naive patients vs none among vaccinees). In the vaccinees, varicella infection was observed only in those who did not develop or lost VZ antibodies; in addition, 21 patients of this subgroup had an asymptomatic seroconversion. Four of the 415 patients with a history of varicella had another episode of benign varicella after grafting. Herpes zoster was observed in 76 of the 704 patients included in the study. The prevalence differed according to VZ status at the time of grafting: 13% in patients with a history of varicella, 7% in the vaccinees, and 38% in the naive patients at grafting who developed varicella. Three rejection episodes occurred in association with a varicella episode and four with a zoster episode, but graft function was only transiently impaired, and as a whole varicella or zoster did not significantly affect graft function or survival.
Conclusion. Naive VZ patients with a kidney graft are at risk to develop severe varicella and this may be effectively prevented by available immunization.
Varicella, most often a benign disease in the normal child, may be severe and even fatal in the immunocompromised child or in children on corticosteroids. Herpes zoster, a reactivation of varicella/zoster (VZ) virus usually observed in older patients associated with the spontaneous lowering of immunity, is also known to occur in patients receiving immunosupressive treatment. Little information is available on the occurrence of varicella and herpes zoster after kidney transplantation in children.1,2 Although immunization against varicella with an attenuated strain was first reported as effective and safe in 1974,3 there are few data on its effectiveness in the prevention of the disease in children with a kidney graft, except a preliminary report from our group4and another in a small group of patients.5 Our purpose was to assess the prevalence of varicella and of herpes zoster in a large population of children and adolescents after kidney transplantation and to determine the effectiveness of immunization with the Oka strain that we have been performing systematically in naive patients before transplantation beginning in 1980.
All the files of patients receiving grafts in our institution from January 1973 to December 1992 were reviewed for this report. During this period 760 children and adolescents received 855 grafts. Fifty-six patients died or lost their graft during the first 3 months and were excluded (none developed varicella during this time). The remaining 704 patients, with a follow-up of 4 months to 22 years, are the subject of this study. The immunosupressive protocol varied with time and has been reported in several publications.6-8
VZ serology was assessed by different methods according to time: first by the fluorescence antibody to membrane antigen (FAMA) test9 and second beginning in July 1984 by the enzyme-linked immunosorbent assay (ELISA)10 with a VARICELLA-STAT kit (Biowhittaker, Inc, Walkersville, MD). The titers were expressed as optical density (OD) patient × 100/OD cut-off. For the calculation of mean values only the measurements performed by ELISA were considered.
All candidates for grafting have been routinely tested for VZ serology since 1980, and all seronegative patients were given one injection of 1000 plaque-forming units of attenuated Oka strain (SmithKline Beecham Biologicals Laboratories, Rixensart, Belgium). Among patients who remained seronegative, 11 received a second dose of the vaccine. Safety of the vaccine was excellent, except in 3 patients who, 8 days after vaccination, developed a mild disease with a limited rash (grade I as defined in this section) without fever. No graft was performed within the 3 months after immunization.
The VZ status at the time of grafting of the 704 patients retained for this study distinguished several groups: Group A: 49 naive patients who had no history of varicella and did not receive immunization; most of them were transplanted before 1980, and all were negative for the VZ virus. Group B: 415 patients who had a history of varicella and positive serology. Group C: 212 patients who received varicella immunization. A group of 28 patients for whom clear information was lacking were not considered in this study.
For the purpose of this study, the severity of varicella was estimated according to an arbitrary scale from I to IV used in Table 2: I: benign varicella, less than 10 vesicles, no fever; II: moderate varicella, typical eruption, fever less than 38.5°; III: marked varicella, diffuse vesicles with several crops, fever more than 38.5° lasting more than 3 days; IV: malignant varicella with visceral involvement, thrombopenia, hepatitis, pancreatitis, pneumopathy, neuritis, etc, and death.
The severity of herpes zoster was also assessed according to an arbitrary scale from I to III: I: benign zoster, limited number of vesicles, no fever; II: moderate zoster, numerous vesicles, low grade fever; and III: severe zoster, extensive vesicles, fever more than 38°, evolution longer than 15 days, persistent pain.
The Kaplan Meier formula was used for the calculation of antibody survival, Student's t test and χ2for comparison of groups.
Follow-up of VZ Serology After Vaccination
Due to incomplete recovery of data, the actuarial proportion of patients with persistent VZ antibodies after vaccination was assessed from a subgroup of 186 patients with a follow-up longer than 1 year (Fig 1). Sixty two percent still had antibodies after 1 year; this proportion declined slightly with time since 42% remained protected after 10 years.
In only 178 patients were antibodies checked at the time of grafting; 84 had no antibody and 94 had VZ antibodies. One hundred seventeen patients were tested by ELISA; 60 had no antibody detectable and 57 had a mean titer of 347 ± 183, not significantly different from the mean titer of 113 patients of group B (404 ± 148) who had natural immunity also assessed at the time of grafting by ELISA. It should be noted that among the 11 patients who were given a second dose of the vaccine only 5 developed antibodies and are included in the 94 protected patients. After transplantation, 19 of these 94 patients had lost their VZ antibodies at the time of the study, ie, a cumulative proportion of 7.4% after 1 year, 11.1% after 2 years, and 24% after 5 years. Considering only the patients tested by ELISA, the subgroup of 15 patients who lost antibodies had a mean titer at the time of transplantation of 296 ± 126 vs 364 ± 196 in the subgroup of the 43 also regularly checked by ELISA and who kept their antibodies after a follow-up of 1 to 8 years.
In comparison, among patients who had antibodies after natural varicella, the cumulative proportion who lost these antibodies after grafting was 0.4% after 1 year, 2.8% after 2 years, and 4.5% after 4 years, with no further change thereafter.
Varicella After Transplantation
Of the 49 naive patients of group A, 22 developed varicella (45%). The mean age at time of the disease was 13.7 ± 7 years and the mean time elapsed between transplantation and the disease was 4.5 ± 3.8 years (range 2 months to 22 years). Concerning the severity of the 19 documented cases, 4 patients had grade I, 5 grade II, 7 grade III, 3 grade IV and died.
Of the 415 patients of group B with a history of varicella, 4 had another episode of varicella after grafting. Two patients developed the disease on days 20 and 21 after transplantation after nosocomial exposure in hospital; the severity was grade II in both cases. Two other patients also developed another varicella 1 and 4 years after grafting; in the latter case follow-up had shown a loss of VZ antibodies.
Of the 212 patients of group C who were vaccinated two subgroups may be considered: the C1 subgroup of 75 patients with persisting immunity and the C2 subgroup of 103 patients grafted without VZ antibodies despite immunization, plus the 19 who lost VZ antibodies sometime after grafting.
Only the 103 patients of the C2 subgroup were at risk to develop varicella. Of them 26 patients contracted varicella and 21 had asymptomatic seroconversion. The overall prevalence of varicella in this group was 12% significantly lower than in the group of naive patients (χ2; P < .001). The mean age at the time of disease was 11.4 years (range, 3.8 to 20) and the mean time elapsed from the graft was 2.5 years (range, 1 month to 7 years), not significantly different from group A patients. Importantly, the grade of the disease was less severe in the 25 documented cases, as shown in Fig 2, with 11 of grade I, 9 of grade II, 5 of grade III, and none of grade IV (χ2; P < .04).
Once the first symptoms of varicella were identified, azathioprine was stopped in all patients for a mean duration of 10 days in group A patients and 6 days in group C2 patients. In addition, different treatments were given according to time: hyperimmune gamma globulins in 10 patients, one of whom died, IV acyclovir in 22 other patients, half of whom also received hyperimmune globulins; one patient also died in this group. Two patients of group C2 received only oral acyclovir. Note that one patient of group A did not receive any specific treatment and died.
Incidence on Rejection and Graft Function
Three patients developed a rejection episode after varicella and stopping azathioprine (all of group C2); all three had complete recovery. Plasma creatinine increased nonsignificantly from baseline during the disease and returned to a similar level after 3 to 6 months in the three groups (Table).
Of the 704 patients included in this study 76 exhibited herpes zoster. The apparent prevalence of herpes zoster differed among the groups of patients: 14% (7/49) in group A, 13% (54/415) in group B, and 7% (15/212) in group C. In fact all the patients who developed herpes zoster had previously had varicella or asymptomatic seroconversion and it may be argued that the prevalence should be calculated taking into account only the patients at risk. In group A, only 22 patients had had varicella and because three died from this disease the prevalence was actually 38% (7/19). It remained 13% in group B. For the C1 subgroup, of the 75 patients who retained antibodies, 13 developed herpes zoster, a prevalence of 17%; and in subgroup C2, of the 26 patients who had varicella, 3 subsequently developed herpes zoster, ie, a prevalence of 11%.
Considering all cases together, the mean age at time of herpes zoster was 15.6 years (range, 5 to 22) and the mean time elapsed from grafting was 3.3 years (range, 2 months to 11 years).
Clinically, the localization of herpes zoster was variable: 46 on the thorax, 7 on the arm, 4 on the neck or the face, 2 ophthalmic, 6 on the thigh, and 7 on the abdomen (unknown in 4).
The severity of the disease was analyzed according to the immunosuppression administered during the preceding months. Of 69 documented cases, 29 were mild, 29 moderate, and 11 severe and of long duration. The patients receiving alternate day steroids tended to have a milder disease. Cyclosporine was not associated with greater severity. Severity was not dependent on the time of varicella before or after transplantation, or on whether immunization was previously performed or not.
Six patients had relapsing zoster, most often severe, all but one in different localizations. Five patients were on azathioprine and prednisone and one was also receiving cyclosporine.
Except in 10 patients who had a mild form of herpes zoster, azathioprine was stopped at the time of diagnosis for a mean duration of 10 days (range, 3 to 50). The treatment also varied with time and included hyperimmune gamma globulins and/or acyclovir, generally per os and sometimes intravenously. The effectiveness of these different treatments is difficult to analyze because they were not uniformly started during the first 2 or 3 days of the disease.
Effect on Rejection and Graft Function
Four patients had a rejection crisis after zoster and stopping azathioprine; all subsequently regained previous graft function. Plasma creatinine increased nonsignificantly from 133 ± 62 to 149 ± 57 μmol/L during the disease and 148 ± 67 μmol/L after 3 to 6 months.
The risk of severe varicella in patients on immunosuppresive treatment and especially on corticosteroids is well established. The role of corticosteroid seems particularly severe when given at the time of exposure.11 Consequently patients with a kidney graft may develop a severe disease if they have no history of varicella. Most large series of transplantation in children reported some cases of malignant varicella.1 This was recently stressed in a report of eight patients with varicella observed over a 12-year period in a series of 83 children with a kidney graft: 4 had visceral localizations and 2 died despite administration of intravenous acyclovir, begun after the first 2 days of the disease, 1 having also received hyperimmune globulins on the day of exposure.2Another recent study in 67 children with liver transplantation reported 14 cases of varicella, again despite prophylaxis with γ-globulins and intravenous acyclovir; 2 of them died.12 The data reported confirm the vital risk of varicella after kidney transplantation in children who had no history of the disease before the availability of the vaccine, because 3 of the 22 patients who developed varicella in this situation died, one despite hyperimmune globulins at day 7 after exposure, and another despite intravenous acyclovir.
Given this background, active immunization was expected to avoid the permanent threat of severe varicella in children on continuous immunosuppression after transplantation. A live attenuated varicella vaccine was developed in Japan in the early 1970s.3 This vaccine was shown to be safe, immunogenic, and efficacious after a single dose in healthy children, and despite the fact that some of them developed chickenpox, the disease was always mild. White et al13 reported a 96% response after vaccination in healthy children with persistent antibodies after 1 year and an incidence of varicella after household exposure of 12%, vs 87% in historical nonimmunized children. Similar data were reported by Johnson et al14 with a varicella prevalence of 7% after a 3-year follow-up. Varicella in vaccine recipients was clinically modified and significantly less severe in both healthy15 and leukemic children,16 and none of the breakthrough varicella reported in immunocompromized patients was fatal.
As soon as the vaccine was available we included routine active immunization against varicella in preparing seronegative children for kidney transplantation, and reported encouraging results in 1985.4 At this time the antibody response detected by the FAMA technique was observed in 20 of the 23 patients within the first 3 months after vaccination. Most of the FAMA titers were in the low range but the protection was effective in comparison with the historical control group. Our study gives more information on the effectiveness of this vaccination 10 years later.
It was considered safer not to administer this attenuated live virus although the patient was under immunosuppressive drugs, and none of the patients of this series was transplanted before an interval of 3 months after vaccination. This cautious approach is perhaps not entirely justified considering the good tolerance reported in leukemic patients17 and in transplanted children5 who receive the vaccine although under chemotherapy or immunosuppressive treatment. This point deserves further study.
Only the results of ELISA-assessed antibody responses are reported. This response was different from that after natural infection, because half of the vaccinees had no detectable antibodies or lost their antibodies. Nevertheless, those who kept these antibodies at the time of transplantation had a mean level similar to those with natural immunity. It is noteworthy that breakthrough chickenpox was observed only in patients who had no detectable antibodies at the time of grafting and in the additional ones who lost them after grafting. Finally, of 212 vaccinees only 12% developed varicella, and as previously reported the disease remained mild or moderate in comparison with that in the nonimmunized children.
There is little information on the occurrence of zoster after kidney transplantation in children. Of the 415 patients who had had varicella before grafting 13% developed zoster. Taking the number of patient-years as a denominator, there was 1 first zoster per 46 patient-years. The effectiveness of varicella vaccine for preventing zoster was reported in leukemic patients.18 In our 1985 study4 we reported apparent efficacy of the vaccine against zoster, but the follow-up was too short to allow a firm conclusion. In our study the proportion of vaccinees who developed zoster was 7%, with 1 first zoster for 68 patient-years. Expressed as standardized morbidity ratio, the relative risk of herpes zoster after transplantation in patients with natural immunity as compared with the vaccinees was calculated to be 1.9.
Interestingly, the prevalence of zoster was the highest, 39%, in the group of naive patients at time of transplantation and who had varicella although receiving immunosuppressive drugs. They also had a shorter interval between varicella and zoster. Some patients had relapsing zoster with no clear explanation for relapses.
Because there is some evidence that the closely related cytomegalovirus can induce allograft injury,19 it was important to seek graft dysfunction in the large series reported herein. After the withdrawal of azathioprine during a mean period of 10 days in the group of naive patients developing varicella, 3 patients had a rejection crisis, but after 3 to 6 months plasma creatinine returned to baseline and no graft was directly lost. Similar findings were observed in zoster for which azathioprine was also interrupted for a mean period of 10 days. Although four patients (5%) had a rejection crisis, again no graft loss followed.
In conclusion, this study shows the usefulness of varicella immunization in children before transplantation. Due to the slight reduction of immunocompetence related to uremia, it would be appropriate to perform this vaccination in any child with a renal problem. It could also be appropriate to give another dose of vaccine in patients who did not develop detectable antibodies. This strategy should allow normal activities on the part of transplanted children, and if the risk of chickenpox is not completely abolished in these patients, it will no longer be severe.
- Received December 4, 1995.
- Accepted February 28, 1996.
Reprint requests to (M.B.) Department of Pediatric Nephrology, Hôpital Necker-Enfants Malades, 149 rue de Sèvres 75743, Paris, Cedex 15, France.
- VZ =
- varicella/zoster •
- FAMA =
- fluorescence antibody to membrane antigen (test) •
- ELISA =
- enzyme-linked immunosorbent assay •
- OD =
- optical density
- Lynfield R,
- Herrin JT,
- Rubin RH
- Broyer M,
- Boudaillez B
- Broyer M,
- Gagnadoux MF,
- Guest G,
- Beurton D,
- Niaudet P,
- Habib R,
- Busson M
- Fishaut M,
- Murphy D,
- Yanogihara R,
- Mc Intosh K
- Dowell JS,
- Bresee JS
- McGregor RS,
- Zitelli BJ,
- Urbach AH,
- Malatack JJ,
- Gartner JC
- White CJ,
- Kuter BJ,
- Hildebrand CS,
- et al.
- Johnson C,
- Rome LP,
- Stancin T,
- Kumar ML
- Bernstein HH,
- Rothstein EP,
- Watson BM,
- et al.
- Gershon AA,
- Steinberg SP
- Arbeter AM,
- Granowetter L,
- Starr SE,
- Lange B,
- Wimmer R,
- Plotkin SA
- Copyright © 1997 American Academy of Pediatrics