Longevity of rabies antibody titre in recipients of human diploid cell rabies vaccine Deborah
J. Briggs*$ and James R. Schwenkel
Sera from individuals across the USA and Peace Corps Volunteers (PCV) were assayed for neutralizing antibody to human diploid cell rabies vaccine (HDCV). A pairwise comparison between intramuscular (i.m.) and intradermal (i.d.) administration, PCV and non-PCV, revealed statistically signtjicant diflerences between i.m. and i.d. for PCV and non-PCV, and between PCV and non-PCV for i.d. and i.m. administration. Survival analysis andfrequency tables employing x2 tests determined that the percentage probability of maintaining an adequate titre from 152.0 years after primary vaccination was 99% in non-PCV i.m., 93% in non-PCV i.d., 88% in PCV i.m., and 64% in PCV i.d. Keywords:
Rabies
; vaccine ; intramuscular;
intradermal; longevity
INTRODUCTION Licensure of human diploid cell rabies vaccine (HDCV) and rabies immune globulin of human origin (RIG) has resulted in virtually complete protection in individuals who have had an exposure to rabies’. HDCV has proven to be more immunogenic, producing fewer allergic reactions than reported with earlier rabies vaccines and can be administered before or after potential exposure to rabies virus2-4. Pre-exposure vaccination consists of a series of three immunizations (1 .Oml intramuscular (i.m.) or 0.1 ml intradermal (i.d.)). In non-immunized individuals, postexposure vaccination is given after exposure to a known or suspect rabid animal in a series of five i.m. injections including RIG at the time of the first injection. In individuals previously immunized with HDCV, postexposure treatment consists of two i.m. booster vaccinations. One booster vaccination of HDCV is also recommended when the virus neutralizing antibody (VNA) level falls below an acceptable level (the Centers for Disease Control (CDC) currently specify a 1: 5 titre by the rapid fluorescent focus inhibition test as adequate3). Questions concerning the immunogenicity of i.d. administered HDCV have initiated several controlled experiments to investigate its adequacy5-7. Results indicate that when HDCV is administered i.d. by experienced medical personnel, seroconversion is at or very near 100%. Other investigators have reported lower than expected VNA titres after vaccination with HDCV in military personnel and Peace Corps Volunteers *Department of Veterinary Diagnosis, Kansas State University, Manhattan, KS 66606, USA. +Department of Statistics, Kansas State University, Manhattan, KS 66606, USA. *To whom all correspondence should be addressed. (Received 6 July 1991; revised 30 August 1991; accepted 3 September 1991)
026440x/92/020125 0 1992 Butterworth-Heinemann
Ltd
(PCV) *vg. In a dose-response study by Fishbein et al., the amount of vaccine administered i.m. or i.d. had an effect on the immunological response7. It is possible that less than the recommended amount of HDCV may be received if individuals are vaccinated by personnel inexperienced in administering i.d. vaccines. This may be related to vaccine leakage through the needle tract when the syringe is withdrawn or to administration in an unapproved manner, that is subcutaneously instead of by the correct i.d. method. In experimental studies determining the immunogenicity of HDCV in veterinary students at various Colleges of Veterinary Medicine, VNA has been reported to be present for 2-3 years postvaccination5~i0. However, a survey of the longevity of VNA after vaccination with HDCV i.d. or HDCV i.m. in the general public where the possibility that a less than satisfactory vaccination protocol may have been used has not been conducted. The laboratory in the Department of Veterinary Diagnosis at Kansas State University was able to conduct such a survey on serum samples received for postvaccination VNA evaluation from individuals across the continental United States and from PCV overseas in the years 1984-1989. Survival analyses were used to estimate the longevity of VNA in the general public and in PCV after vaccination with Imovax Rabies or Imovax Rabies ID (Merieux Institute Inc., Miami, Florida, or Institut Merieux, France). SUBJECTS
AND
METHODS
Serum samples Between 1984 and 1989,875 serum samples submitted to the Department of Veterinary Diagnosis at Kansas State University from vaccinees across the United States and from PCV were screened for antibody to HDCV using the rapid fluorescent focus inhibition test
Vaccine, Vol. 10, Issue 2, 1992
125
Longevity of rabies antibody titre: D.J. Briggs and J.R. Schwenke Table Peace
1 Numberof subjectssurveyedaccordingto routeof vaccination, Corps Volunteerstatus, and sex PCV" ~Female (15) ( 3 2 ) ~ ~ Male (17)
Intramuscular (212)~
/
Female (84)
Non-PCV ( 1 8 0 ) ~ Male (96) j~PCV
Intradermal
(663)~
( 3 4 9 ) ~
~Non-PCV
(314)~
Female (176) Male (173) Female (143) ~~ Male (171)
"PCV, PeaceCorps Volunteer
( R F F I T ) 11. Serum samples included in the study met the following criteria : all were from individuals who had received pre-exposure or postexposure prophylaxis with H D C V produced by Merieux Institute (Miami, Florida), all were accompanied by a complete vaccination history, and no vaccinees had a history of receiving duck embryo vaccine. M a n y PCV were vaccinated with Imovax Rabies or Imovax Rabies I D produced by Institut Merieux in France.
Epidemiological data Epidemiological data were collected from the information included on the case history submitted with each specimen. Vaccination history included the route of inoculation (i.m. or i.d.), number of doses, date of vaccination, date on which the specimen was obtained, sex, and whether the vaccinee was a PCV. The age of the subject was included with 270 of the samples. The range of ages was as follows: 21 samples were from subjects 2 - 2 0 years of age, 146 samples were from subjects 21 40 years of age, 79 samples were from subjects 4 1 - 6 0 years of age, and 24 samples were from subjects 61-74 years of age. Only individuals that received three pre-exposure vaccinations or five postexposure vaccinations of H D C V were included. Serum from individuals that received one or more booster vaccinations prior to serum submission were eliminated.
Antibody determination Rabies VNA was measured by the R F F I T 11. Each time the R F F I T was conducted, a reference serum containing two International Units ( I U ) of rabies YNA was included to assure quality control. Tests results were evaluated using the criteria recommended by CDC, i.e. complete virus neutralization at a 1:5 serum dilution was considered adequate. Due to cytotoxicity that sometimes occurs at low serum dilutions, samples were initially screened at a 1 : 10 dilution, 20 microscopic fields were examined for virus neutralization and results were reported as < 1:5 if three or more virus infected cells were present. If fewer than three infected cells were present at a 1:10 dilution and incomplete virus neutralization occurred at a 1:50 dilution results were reported as > 1 : 5. When complete virus neutralization occurred at a 1 : 10 and at a 1 : 50 dilution results were reported as > 1 : 50.
126 Vaccine, Vol. 10, Issue 2, 1992
Statistical analyses Data on 875 vaccinated subjects were analysed. Subjects were divided into specific groups according to data supplied by their vaccination history (Table 1 ).,The time from initial vaccination to serum withdrawal was calculated and expressed in half-year increments to make data presentation and analysis tractable. Frequency distribution of the R F F I T titre for each treatment group summed over the number of vaccinations was determined for each half year. The time to seronegativity (defined as a R F F I T of < 1 : 5) was considered to be the response variable for analysis. Because the data were collected by survey instead of by designed experiment, the time to seronegativity was overstated. For a subject no longer seropositive, the time to seronegativity indicated that the vaccination titre decreased on or before the recorded time. However, because subjects were obtained randomly, the overstating of time should affect each treatment group equally. The survey maintained validity because the data identified subjects that had an adequate titre as well as subjects that did not have an adequate titre during each half year increment of time. Time to seronegativity was analysed utilizing survival analysis techniques. A subject that maintained an adequate titre of at least > 1 : 5 was considered a censored observation. The route of injection (i.m. or i.d.), sex (M or F), and PCV or non-PCV defined eight treatment groups. A Kaplan Meier survival curve was constructed for each treatment group and comparisons among groups made using the Wilcoxon Z 2 t e s t . The number of injections and age, when recorded (n = 270), were considered as covariates with respect to all treatment groups. Pairwise comparisons were considered among the levels of each treatment factor. Statistical significance was determined at the 0.05 level.
RESULTS Kaplan Meier survival estimates detected differences in the expected longevity of VNA after vaccination with H D C V in the following four treatment groups : non-PCV i.m., non-PCV i.d., PCV i.m. and PCV i.d. Between 1.5 and 2.0 years after primary vaccination an adequate titre was maintained by 99% of non-PCV i.m., 93% of non-PCV i.d., 88% of PCV i.m. and 64% of PCV i.d. Six months later, between 2.0 and 2.5 years after primary vaccination, the percentage of each treatment group that maintained an adequate titre dropped to 93% of non-PCV i.m., 74% of non-PCV i.d., 79% of PCV i.m. and 51% of PCV i.d. Of the 875 serum samples analysed, 180 were non-PCV i.m., 314 were non-PCV i.d., 32 were PCV i.m., and 349 were PCV i.d. The majority of subjects received three doses of H D C V i.d. as pre-exposure rabies prevention (n = 834). The distribution of the R F F I T titres that were analysed in this report is displayed in Table 2. The majority of serum samples originating from non-PCV was received 2 2.5 years postvaccination. This is probably due to the Immunization Practices Advisory Committee ( A C I P ) recommendation that a vaccination booster be given or a quantitative antibody test be conducted at 2 years postvaccination 3. Initial analysis comparing all treatment groups detected significantly different K a p l a n - M e i e r survival estimates (p ~< 0.0001 ). The effect of the number of injections on the Kaplan Meier survival estimates was not significant (p = 1.000). However, the number of subjects included
Longevity of rabies antibody titre: D.J. Briggs and J.R. Schwenke Table 2 (HDCV)
Number of subjects at each level of titre response per half-year increment of time after vaccination with human diploid cell rabies vaccine
Level of titre response
Intramuscular Time (half-year increments)
1:5 >1:50 5 2 1 3 11 7 12 0 1 2 1 0 1 0 0 0
50 1 6 3 23 9 17 2 2 1 2 0 2 0 1 0
Intradermal
1:5
>1:50
0.2895) or when treatment groups were combined (p = 0.5437). Due to these analyses, further comparisons among treatment groups were made for M and F combined with no adjustment for age or number of injections. The combination of M and F response data created the four treatment groups: non-PCV i.m., non-PCV i.d., PCV i.m. and PCV i.d. Additional analyses on the combined response data detected significantly different K a p l a n - M e i e r survival estimates among the four treatment groups (p ~< 0.0001 ). Significantly different responses were detected between i.d. and i.m. administrations for both PCV (p = 0.0041 ) (Fiyure 1) and non-PCV (p~1:5 >1:50 10 7 3 22 57 6 2 7 7 1 0 0 2 0 0 0
PCV >1:5
>1:50
1:5
>1:50
34 20 10 8 12 4 0 0 1 0 0 0 1 0 0 0
67 14 9 13 15 10 0 0 0 0 0 0 1 0 0 0
102 1 5 8 9 3 2 0 0 0 0 0 0 0 0 0
34 26 17 17 59 7 4 2 3 2 2 1 1 1 0 0
83 23 13 43 87 23 16 7 8 3 1 0 4 0 0 0
190 11 23 23 71 17 25 6 9 3 5 0 3 0 1 1
0.8
-__
0.6
0,4
t
I
0.2
0.0
I I
i 2
I 3 Time
I 9 after
I 5
I 6
vaccination
l 7
I 8
(years)
Figure 1 Kaplan-Meier survival function estimates of the probability of maintaining an adequate titre of 1 : 5 in Peace Corps Volunteers after vaccination with human diploid cell rabies vaccine administered intradermally ( ) or intramuscularly ( - - ) 1.0
DISCUSSION This survey was conducted on subjects that had received either three pre-exposure or five postexposure injections of H D C V . The sex, age and number of injections were not significant in estimating the longevity of VNA titres after vaccination with H D C V . Although the longevity of VNA after pre-exposure and postexposure vaccination treatment regime did not differ, there is no indication that the current recommendation for postexposure treatment should be modified. These data did not analyse the initial vaccination response of a three versus five injection series and the efficacy of the recommended postexposure treatment regime has been reported elsewhere lz. Two factors influenced the K a p l a n - M e i e r survival estimates and the expected longevity of VNA in recipients of H D C V , that is, whether the vaccinee was a PCV and the route of vaccine administration (i.m. or i.d.). PCV and recipients of H D C V i.d. consistently had lower VNA survival estimates with the lowest survival estimates occurring in PCV i.d. The highest K a p l a n - M e i e r survival estimates were detected in non-PCV i.m.
36 6 12 12 33 3 4 4 7 2 3 0 1 0 0 0
J. Briggs*$ and James R. Schwenkel
Sera from individuals across the USA and Peace Corps Volunteers (PCV) were assayed for neutralizing antibody to human diploid cell rabies vaccine (HDCV). A pairwise comparison between intramuscular (i.m.) and intradermal (i.d.) administration, PCV and non-PCV, revealed statistically signtjicant diflerences between i.m. and i.d. for PCV and non-PCV, and between PCV and non-PCV for i.d. and i.m. administration. Survival analysis andfrequency tables employing x2 tests determined that the percentage probability of maintaining an adequate titre from 152.0 years after primary vaccination was 99% in non-PCV i.m., 93% in non-PCV i.d., 88% in PCV i.m., and 64% in PCV i.d. Keywords:
Rabies
; vaccine ; intramuscular;
intradermal; longevity
INTRODUCTION Licensure of human diploid cell rabies vaccine (HDCV) and rabies immune globulin of human origin (RIG) has resulted in virtually complete protection in individuals who have had an exposure to rabies’. HDCV has proven to be more immunogenic, producing fewer allergic reactions than reported with earlier rabies vaccines and can be administered before or after potential exposure to rabies virus2-4. Pre-exposure vaccination consists of a series of three immunizations (1 .Oml intramuscular (i.m.) or 0.1 ml intradermal (i.d.)). In non-immunized individuals, postexposure vaccination is given after exposure to a known or suspect rabid animal in a series of five i.m. injections including RIG at the time of the first injection. In individuals previously immunized with HDCV, postexposure treatment consists of two i.m. booster vaccinations. One booster vaccination of HDCV is also recommended when the virus neutralizing antibody (VNA) level falls below an acceptable level (the Centers for Disease Control (CDC) currently specify a 1: 5 titre by the rapid fluorescent focus inhibition test as adequate3). Questions concerning the immunogenicity of i.d. administered HDCV have initiated several controlled experiments to investigate its adequacy5-7. Results indicate that when HDCV is administered i.d. by experienced medical personnel, seroconversion is at or very near 100%. Other investigators have reported lower than expected VNA titres after vaccination with HDCV in military personnel and Peace Corps Volunteers *Department of Veterinary Diagnosis, Kansas State University, Manhattan, KS 66606, USA. +Department of Statistics, Kansas State University, Manhattan, KS 66606, USA. *To whom all correspondence should be addressed. (Received 6 July 1991; revised 30 August 1991; accepted 3 September 1991)
026440x/92/020125 0 1992 Butterworth-Heinemann
Ltd
(PCV) *vg. In a dose-response study by Fishbein et al., the amount of vaccine administered i.m. or i.d. had an effect on the immunological response7. It is possible that less than the recommended amount of HDCV may be received if individuals are vaccinated by personnel inexperienced in administering i.d. vaccines. This may be related to vaccine leakage through the needle tract when the syringe is withdrawn or to administration in an unapproved manner, that is subcutaneously instead of by the correct i.d. method. In experimental studies determining the immunogenicity of HDCV in veterinary students at various Colleges of Veterinary Medicine, VNA has been reported to be present for 2-3 years postvaccination5~i0. However, a survey of the longevity of VNA after vaccination with HDCV i.d. or HDCV i.m. in the general public where the possibility that a less than satisfactory vaccination protocol may have been used has not been conducted. The laboratory in the Department of Veterinary Diagnosis at Kansas State University was able to conduct such a survey on serum samples received for postvaccination VNA evaluation from individuals across the continental United States and from PCV overseas in the years 1984-1989. Survival analyses were used to estimate the longevity of VNA in the general public and in PCV after vaccination with Imovax Rabies or Imovax Rabies ID (Merieux Institute Inc., Miami, Florida, or Institut Merieux, France). SUBJECTS
AND
METHODS
Serum samples Between 1984 and 1989,875 serum samples submitted to the Department of Veterinary Diagnosis at Kansas State University from vaccinees across the United States and from PCV were screened for antibody to HDCV using the rapid fluorescent focus inhibition test
Vaccine, Vol. 10, Issue 2, 1992
125
Longevity of rabies antibody titre: D.J. Briggs and J.R. Schwenke Table Peace
1 Numberof subjectssurveyedaccordingto routeof vaccination, Corps Volunteerstatus, and sex PCV" ~Female (15) ( 3 2 ) ~ ~ Male (17)
Intramuscular (212)~
/
Female (84)
Non-PCV ( 1 8 0 ) ~ Male (96) j~PCV
Intradermal
(663)~
( 3 4 9 ) ~
~Non-PCV
(314)~
Female (176) Male (173) Female (143) ~~ Male (171)
"PCV, PeaceCorps Volunteer
( R F F I T ) 11. Serum samples included in the study met the following criteria : all were from individuals who had received pre-exposure or postexposure prophylaxis with H D C V produced by Merieux Institute (Miami, Florida), all were accompanied by a complete vaccination history, and no vaccinees had a history of receiving duck embryo vaccine. M a n y PCV were vaccinated with Imovax Rabies or Imovax Rabies I D produced by Institut Merieux in France.
Epidemiological data Epidemiological data were collected from the information included on the case history submitted with each specimen. Vaccination history included the route of inoculation (i.m. or i.d.), number of doses, date of vaccination, date on which the specimen was obtained, sex, and whether the vaccinee was a PCV. The age of the subject was included with 270 of the samples. The range of ages was as follows: 21 samples were from subjects 2 - 2 0 years of age, 146 samples were from subjects 21 40 years of age, 79 samples were from subjects 4 1 - 6 0 years of age, and 24 samples were from subjects 61-74 years of age. Only individuals that received three pre-exposure vaccinations or five postexposure vaccinations of H D C V were included. Serum from individuals that received one or more booster vaccinations prior to serum submission were eliminated.
Antibody determination Rabies VNA was measured by the R F F I T 11. Each time the R F F I T was conducted, a reference serum containing two International Units ( I U ) of rabies YNA was included to assure quality control. Tests results were evaluated using the criteria recommended by CDC, i.e. complete virus neutralization at a 1:5 serum dilution was considered adequate. Due to cytotoxicity that sometimes occurs at low serum dilutions, samples were initially screened at a 1 : 10 dilution, 20 microscopic fields were examined for virus neutralization and results were reported as < 1:5 if three or more virus infected cells were present. If fewer than three infected cells were present at a 1:10 dilution and incomplete virus neutralization occurred at a 1:50 dilution results were reported as > 1 : 5. When complete virus neutralization occurred at a 1 : 10 and at a 1 : 50 dilution results were reported as > 1 : 50.
126 Vaccine, Vol. 10, Issue 2, 1992
Statistical analyses Data on 875 vaccinated subjects were analysed. Subjects were divided into specific groups according to data supplied by their vaccination history (Table 1 ).,The time from initial vaccination to serum withdrawal was calculated and expressed in half-year increments to make data presentation and analysis tractable. Frequency distribution of the R F F I T titre for each treatment group summed over the number of vaccinations was determined for each half year. The time to seronegativity (defined as a R F F I T of < 1 : 5) was considered to be the response variable for analysis. Because the data were collected by survey instead of by designed experiment, the time to seronegativity was overstated. For a subject no longer seropositive, the time to seronegativity indicated that the vaccination titre decreased on or before the recorded time. However, because subjects were obtained randomly, the overstating of time should affect each treatment group equally. The survey maintained validity because the data identified subjects that had an adequate titre as well as subjects that did not have an adequate titre during each half year increment of time. Time to seronegativity was analysed utilizing survival analysis techniques. A subject that maintained an adequate titre of at least > 1 : 5 was considered a censored observation. The route of injection (i.m. or i.d.), sex (M or F), and PCV or non-PCV defined eight treatment groups. A Kaplan Meier survival curve was constructed for each treatment group and comparisons among groups made using the Wilcoxon Z 2 t e s t . The number of injections and age, when recorded (n = 270), were considered as covariates with respect to all treatment groups. Pairwise comparisons were considered among the levels of each treatment factor. Statistical significance was determined at the 0.05 level.
RESULTS Kaplan Meier survival estimates detected differences in the expected longevity of VNA after vaccination with H D C V in the following four treatment groups : non-PCV i.m., non-PCV i.d., PCV i.m. and PCV i.d. Between 1.5 and 2.0 years after primary vaccination an adequate titre was maintained by 99% of non-PCV i.m., 93% of non-PCV i.d., 88% of PCV i.m. and 64% of PCV i.d. Six months later, between 2.0 and 2.5 years after primary vaccination, the percentage of each treatment group that maintained an adequate titre dropped to 93% of non-PCV i.m., 74% of non-PCV i.d., 79% of PCV i.m. and 51% of PCV i.d. Of the 875 serum samples analysed, 180 were non-PCV i.m., 314 were non-PCV i.d., 32 were PCV i.m., and 349 were PCV i.d. The majority of subjects received three doses of H D C V i.d. as pre-exposure rabies prevention (n = 834). The distribution of the R F F I T titres that were analysed in this report is displayed in Table 2. The majority of serum samples originating from non-PCV was received 2 2.5 years postvaccination. This is probably due to the Immunization Practices Advisory Committee ( A C I P ) recommendation that a vaccination booster be given or a quantitative antibody test be conducted at 2 years postvaccination 3. Initial analysis comparing all treatment groups detected significantly different K a p l a n - M e i e r survival estimates (p ~< 0.0001 ). The effect of the number of injections on the Kaplan Meier survival estimates was not significant (p = 1.000). However, the number of subjects included
Longevity of rabies antibody titre: D.J. Briggs and J.R. Schwenke Table 2 (HDCV)
Number of subjects at each level of titre response per half-year increment of time after vaccination with human diploid cell rabies vaccine
Level of titre response
Intramuscular Time (half-year increments)
1:5 >1:50 5 2 1 3 11 7 12 0 1 2 1 0 1 0 0 0
50 1 6 3 23 9 17 2 2 1 2 0 2 0 1 0
Intradermal
1:5
>1:50
0.2895) or when treatment groups were combined (p = 0.5437). Due to these analyses, further comparisons among treatment groups were made for M and F combined with no adjustment for age or number of injections. The combination of M and F response data created the four treatment groups: non-PCV i.m., non-PCV i.d., PCV i.m. and PCV i.d. Additional analyses on the combined response data detected significantly different K a p l a n - M e i e r survival estimates among the four treatment groups (p ~< 0.0001 ). Significantly different responses were detected between i.d. and i.m. administrations for both PCV (p = 0.0041 ) (Fiyure 1) and non-PCV (p~1:5 >1:50 10 7 3 22 57 6 2 7 7 1 0 0 2 0 0 0
PCV >1:5
>1:50
1:5
>1:50
34 20 10 8 12 4 0 0 1 0 0 0 1 0 0 0
67 14 9 13 15 10 0 0 0 0 0 0 1 0 0 0
102 1 5 8 9 3 2 0 0 0 0 0 0 0 0 0
34 26 17 17 59 7 4 2 3 2 2 1 1 1 0 0
83 23 13 43 87 23 16 7 8 3 1 0 4 0 0 0
190 11 23 23 71 17 25 6 9 3 5 0 3 0 1 1
0.8
-__
0.6
0,4
t
I
0.2
0.0
I I
i 2
I 3 Time
I 9 after
I 5
I 6
vaccination
l 7
I 8
(years)
Figure 1 Kaplan-Meier survival function estimates of the probability of maintaining an adequate titre of 1 : 5 in Peace Corps Volunteers after vaccination with human diploid cell rabies vaccine administered intradermally ( ) or intramuscularly ( - - ) 1.0
DISCUSSION This survey was conducted on subjects that had received either three pre-exposure or five postexposure injections of H D C V . The sex, age and number of injections were not significant in estimating the longevity of VNA titres after vaccination with H D C V . Although the longevity of VNA after pre-exposure and postexposure vaccination treatment regime did not differ, there is no indication that the current recommendation for postexposure treatment should be modified. These data did not analyse the initial vaccination response of a three versus five injection series and the efficacy of the recommended postexposure treatment regime has been reported elsewhere lz. Two factors influenced the K a p l a n - M e i e r survival estimates and the expected longevity of VNA in recipients of H D C V , that is, whether the vaccinee was a PCV and the route of vaccine administration (i.m. or i.d.). PCV and recipients of H D C V i.d. consistently had lower VNA survival estimates with the lowest survival estimates occurring in PCV i.d. The highest K a p l a n - M e i e r survival estimates were detected in non-PCV i.m.
36 6 12 12 33 3 4 4 7 2 3 0 1 0 0 0
