PROLONGED EFFICACY OF RHESUS-HUMAN REASSORTANT ROTAVIRUS VACCINE

Abstract

Annually rotavirus gastroenteritis causes 600 000 deaths in children in developing countries and leads to substantial medical interventions in developed countries.1 An effective vaccine against rotavirus gastroenteritis, the rhesus-human reassortant rotavirus tetravalent (RRV-TV) vaccine, has recently been licensed in the US.2-4 One of the pivotal studies for the licensure of the RRV-TV vaccine was conducted in 1993 to 1995 in Finland. In that study 2398 infants received three doses of RRV-TV vaccine or placebo between ages 2 and 7 months and were followed for one or two rotavirus epidemic seasons.5 It was observed that the vaccine efficacy for severe rotavirus diarrhea remained sustained in the two seasons (95 and 90% in the first and second rotavirus season, respectively), whereas efficacy for any rotavirus diarrhea was better in the first (83%) than in the second (63%) season. In the first multicenter trial of RRV-TV vaccine in the US the follow-up time was two rotavirus seasons.6 In that study the efficacy of RRV-TV vaccine against any rotavirus diarrhea was 69% in the first and 57% in the second season, whereas the efficacy of monovalent RRV-S1 vaccine was lower during the second season.6 The second multicenter trial of RRV-TV vaccine, in which a dose of 4 × 105 plaque-forming units (PFU) was used, had a follow-up of one rotavirus season only.7 In studies of single serotype rhesus-human reassortant rotavirus vaccines, in which the follow-up time lasted two rotavirus seasons, the efficacy was lower in the second season.8, 9 Rotavirus can cause morbidity in the third rotavirus epidemic season of a child's life,10 and, therefore, it would be important to know whether the efficacy of RRV-TV vaccine will last beyond two rotavirus seasons. We have evaluated the duration of protective immunity of RRV-TV vaccine against severe rotavirus gastroenteritis for three and, in part, four rotavirus seasons. After the end of the surveillance of two rotavirus seasons in 1995,5 we continued the follow-up of gastroenteritis cases treated in the hospital clinic or admitted to hospital for two more rotavirus seasons, until the end of June, 1997. The code of the vaccine study was broken in May, 1996, and the surveillance of the fourth rotavirus season was unblinded. Methods. The study protocol and informed consent form complied with the US Public Health Service, Food and Drug Administration guidelines and were approved by appropriate ethics committees in Finland. The RRV-TV vaccine was produced by Wyeth-Ayerst Research (Philadelphia, PA). The vaccine was a lyophilized product containing 105 PFU of each of the rhesus-human reassortant rotaviruses, corresponding to G-types 1, 2 and 4, and of rhesus rotavirus, for a total of 4 × 105 PFU. A placebo preparation (tissue culture medium) was also produced by Wyeth-Ayerst Research. The study vaccine, in a volume of 2.5 ml, was administered orally in three doses between ages 2 and 7 months. The study started in September, 1993, in 100 well-baby clinics, and infants were recruited until September, 1994.5 A total of 2398 infants were randomized and received at least one dose of vaccine or placebo. After the end of the trial in June, 1995, data were entered into the database until the code was broken in May, 1996. The active surveillance of gastroenteritis in the participating children during the first and second rotavirus seasons until the end of June, 1995, was described previously.5 In this period medical intervention for gastroenteritis occurred through the health care system's normal channels. If a hospital clinic was needed children were referred to Tampere University Hospital for treatment in the outpatient clinic and, when required, for hospital admission. During the third and fourth rotavirus season, from the beginning of July, 1995, until the end of June, 1997, the follow-up was passive and only gastroenteritis episodes treated in the clinic of Tampere University Hospital or admitted to hospital were recorded. This hospital was the only one where children with gastroenteritis were admitted in the third and fourth rotavirus season of follow-up. In the third season the parents were also asked to keep a diary card and collect one stool sample of cases treated at home. Furthermore hospital patient records were checked daily to detect all study children who had visited the pediatric clinic or were admitted to the hospital. In the fourth rotavirus season, when the surveillance was already unblinded, only clinical information retrieved from the hospital patient records was collected. No stool samples were collected for the study purposes, but the results of rotavirus studies of the hospitalized children were available. For the evaluation of the vaccine efficacy in the third and fourth season, it was assumed that all children who completed the primary efficacy analysis were still living in the area during the prolonged surveillance. It was not known exactly how many families might have moved out of the study area. In the primary efficacy analysis period 15 families moved out and 6 interrupted the follow-up for other reasons. Assuming the same rate of moving up to 30 more families (1 to 2% of the study population) might have migrated out of the area. Rotavirus antigen testing and G-typing by reverse transcriptase polymerase chain reaction were performed as described previously.5 Results. Of the children who had received all 3 doses of vaccine and were included in the primary analysis, 1073 (RRV-TV 536, placebo 537) were present in the first rotavirus season (until June 30, 1994), and 1200 (RRV-TV 592, placebo 608) children were only present in the second rotavirus season from July 1, 1994, to June 30, 1995. Likewise the rotavirus seasons of prolonged follow-up, 1995 to 1996 and 1996 to 1997, were defined by the same cutoff date and were, respectively, the third and fourth follow-up year for those children who had been present in the first season and the second and third follow-up year for those infants who had not been present in the first season of study (Table 1). The first follow-up year of a child was not a complete calendar year but started 13 days after the third dose of vaccine and ended at the end of the ongoing rotavirus season. A total of 1128 children in the vaccine group and 1145 in the placebo group entered their third follow-up year and 536 infants in the vaccine group and 537 infants in the placebo group entered their fourth follow-up year.TABLE 1: Hospital outpatient clinic visits and hospital admissions because of rotavirus gastroenteritis during 4 years of follow-up after vaccination, and protective efficacy of RRV-TV vaccine In the primary efficacy analysis period of the original efficacy trial there were 13 hospitalizations related to rotavirus gastroenteritis in the placebo group vs. none in the RRV-TV vaccine group.5 During the extended follow-up, 6 more hospital admissions associated with rotavirus gastroenteritis were recorded in the placebo group vs. none in the RRV-TV vaccine group. Of the hospitalizations 4 were in the 1995 to 1996 season and 2 in the 1996 to 1997 season. However, relative to the time of enrollment in the trial, all 6 hospital admissions were either in the second or third year of follow-up of the child in question. Therefore, altogether 19 children in the placebo group (1.66% of the group) vs. 0 in the vaccine group were hospitalized because of rotavirus gastroenteritis during the entire follow-up, for vaccine efficacy of 100% [95% confidence interval (CI) 83 to 100, P < 0.001]. Likewise all the 7 outpatient clinic visits, as a result of confirmed rotavirus gastroenteritis detected during the extended follow-up, were in the second or third year of follow-up of any given child (Table 1). During the first year of follow-up one child in the vaccine group vs. 25 in the placebo group were seen in the clinic or admitted to hospital, for a vaccine efficacy of 96%. During the second and third year of follow-up no child in the vaccine group vs. 19 and 5, in the 2 years, respectively, in the placebo group were seen, for a vaccine efficacy of 100%. For all 4 follow-up years the vaccine efficacy was 98% (Table 1). In the first rotavirus season, 1993 to 1994, 73 rotaviruses were detected; 54 (74%) were G1 and 19 (26%) were G4. In the second rotavirus season, 1994 to 1995, 181 rotaviruses were detected, and 168 (94%) were G1 and 9 (5%) were G4. Of the 11 rotaviruses detected in the 1995 to 1996 season 10 were G-typed and all were G1. In the 1996 to 1997 season stool samples were not available for G-typing. During the first, second, third and fourth follow-up years there were a total of 129 cases of any gastroenteritis treated in the hospital or in the clinic, caused by rotavirus or not. The number of cases by the follow-up year, respectively, were 12, 16, 5 and 1 in the vaccine group, vs. 40, 39, 13 and 3 in the placebo group, for vaccine efficacies of 70% (CI 42 to 84, P < 0.001), 58% (CI 25 to 76, P = 0.002) and 60% (CI −9 to 86, P = 0.095) in the first 3 years. Because of the small number of cases the efficacy was not determined for the fourth follow-up year. Vaccine protection against any gastroenteritis seen in the hospital was 64% (CI 47 to 74, P < 0.001) in the entire follow-up time. Discussion. The principal finding of importance in the prolonged surveillance was that 6 (32%) more cases of rotavirus gastroenteritis requiring hospital admission and 7 (23%) more cases requiring a clinic visit were detected in the placebo recipient children and none in the RRV-TV vaccine recipient children, after the completion of the original follow-up of this vaccine trial.5 It is apparent that the present follow-up was of sufficient length to detect most, if not all, cases of severe rotavirus gastroenteritis to occur in this cohort of children, because no further cases were detected during the fourth year of follow-up of any child (even though not all children completed the fourth year of follow-up). Thus it can be concluded that the efficacy of the RRV-TV vaccine against severe rotavirus gastroenteritis remains sustained for 3 years after vaccination. The protection is likely to be a combined effect of the three doses of RRV-TV vaccine plus natural exposure to rotavirus during the epidemic seasons, resulting in a booster effect. In other studies a decline of rotavirus vaccine efficacy against any rotavirus gastroenteritis has been observed in the second on third year of follow-up. In the study of Vesikari et al.8 the combined efficacy of two rhesus-human rotavirus reassortant vaccines (DxRRV and DS1xRRV) was 92% in the first season and 59% in the second season. In the study of Madore et al.,9 in which infants received either rhesus rotavirus vaccine (RRV) or G1 reassortant (DxRRV), the surveillance lasted for three rotavirus seasons. The efficacy of RRV vaccine was 66% in the first season and 51% in the three seasons combined, and that of DxRRV vaccine was 77% in the first rotavirus season and 67% for the three seasons. Also in our present trial vaccine efficacy against all rotavirus gastroenteritis declined from the first to the second year but could not be determined in the third and fourth years. In conclusion this study has demonstrated that (1) rotavirus causes significant morbidity in the third year of life, corresponding to the third rotavirus season in the life of a child, but not thereafter and (2) vaccination with three doses of RRV-TV vaccine between 2 and 7 months of age protects the child against severe rotavirus disease in the third year of life or third rotavirus epidemic season. These conclusions were reached in a situation where G1 rotavirus was predominant in each of the three seasons and should be validated for the case when the challenging rotavirus is of different G-type. Jaana Joensuu, M.D. Eeva Koskenniemi, M.Sc. Timo Vesikari, M.D. University of Tampere; Medical School; Tampere, Finland (JJ, TV) The National Public Health Institute; Helsinki, Finland (EK)