- Email: [email protected]
Immunologic studies in subjects with a serum sickness-like illness after immunization with human diploid cell rabies vaccine
lmmunol ic studies in with a sickness-like illness after immun n with human diploid cell rabies vaccine Richard J. Warrington, M.B., B.S., Ph.D., F.R.C.P.(C), Catheryn J. Martens, R.N., Margaret Rubin, R.N., B.Sc., W. John Rutherford, B.Sc., and Fred Y. Aoki, M.D., F.R.C.P.(C) Winnipeg, Manitoba, Canada
Ten patients developing a serum sickness-like hypersensitivity reaction to human diploid cell rabies vaccine were studied and compared with control subjects matched for previous vaccination history and level of rabies virus-specific IgG response to immunization. The clinical reaction consisted of delayed onset, generalized urticaria, and angioedema, with some arthralgias. Skin biopsy specimens demonstrated a leukocytoclastic vasculitis. Individuals reacting to the vaccine possessed lgE antibodies to human diploid cell rabies vaccine, to mock vaccine lacking viral antigen, and to fetal calf serum (FCS), a vaccine trace contaminant. Increased levels of lgG antibodies to FCS, mock vaccine, and fl-propiolactone-modified FCS, and human serum albumin were also found. Such humoral responses to vaccine components other than rabies virus might be responsible for the hypersensitivity reactions developing after rabies vaccination. (J ALLERGY CLIN IMMUNOL 1987;79:605-10.)
The reactivated HDCRV used in these studies is a lyophilized and stabilized suspension of rabies virus inactivated by 13-propiolactone. It is prepared in human diploid cell cultures of the MRC-5 embryonic lung cell line and contains trace amounts of FCS, human albumin, neomycin, vitamins, and 13-propiolactone.* Adverse reactions to the vaccine have previously been limited and mild, either local reactions at the injection site or systemic reactions of headache, malaise, nausea, and dizziness. In 1976, Cox and Schneider ~ reported two patients with generalized urticaria and angioedema after booster immunization with the vaccine. This present study was initiated in 1983 when three of our patients developed similar reactions suggestive of a hypersensitivity response that appeared somewhat unusual in the light of previously reported clinical experience with the vaccine. 2"4 During this survey, 10/485 patients (2.06%)
From the Departmentsof Medicine, Immunologyand Medical Microbio|ogy, Universityof Manitoba, and the Section of Allergy and Clinical Immunology,Health Sciences Center, Winnipeg, Manitoba, Canada. Received for publication Jan. 23, 1986. Accepted for publication Sept. 20, 1986. Reprint requests: Richard J. Warrington, M.B., Sectionof Allergy and Clinical Immunology,C319, Health Sciences Centre, 700 William Ave., Winnipeg, Manitoba R3E 0Z3. *Fontaine R. Personalcommunication.
Abbreviations used
HDCRV: FCS: HSA: HKCRV: OD:
Human diploid cell rabies vaccine Fetal calf serum Human serum alLnmam Hamster kidney cell rabies vaccine Optical density
developed urticaria and angioedema IO to 12days-af~er receiving a booster immunization with ~ R V . In an attempt to determine the cause of this hypersensitivity reaction, immune responses to the vaccine components were studied in the patient group and in a matched control group of immunized subjects.
Subjects A total of 485 subjects received primary "and booster immunizations with HDCRV (Institute Merieux, Lyon. France) during a period of 2 years, in a study direemd by F. Y. A. to produce hyperimmmae subjects for the isolation of human rabies-specific imrrmnoglob~lin. Subjects were individuals at increased risk of exposure to ra~ies virus from the provinces of Manitoba and Sa~katchewaa. Most vaccine recipients had previously been immunized with HKCRV.
Rabies virus vaccine The HDCRV was prepared by Institute Merieax and was supplied by Rhone-Poulenc Pharma Inc.. Montreal. Que6O5
606
W a r r i n g t o n et al.
J. ALLERGY CLIN. ]MMUNOL. APRIL 1987
TABLE I. Characteristics of patients reacting to HDCRV Patient
Past vaccination history
Mode of administration
Time of onset
Abnormal clinical tests
1 2 3 4 5 6 7 8 9 10
Ha/Hu Ha/Hu Hu Hu Ha/Hu Du/Ha/Hu Du/Hu Ha/Hu Ha/Hu Ha/Hu
SC SC ID SC SC ID SC ID SC ID
7 10 8 11 12 12 12 8 10 10
Nil lgM ~, C $ C Nil Nil C4 ~, IgM ~, C Nil Nil Nil IgM ~,/C ~,
Ha = hamster kidney vaccine; Hu = human diploid cell vaccine; Du = duck embryo vaccine; SC = subcutaneously; and ID = intradermally.
bec. An identical "mock" vaccine without the rabies virus was supplied to us by Institute Merieux.
Clinical evaluation All subjects developing a reaction to the vaccine were observed by one of us (R. J. W.) within 1 to 3 days of the reaction and clinically evaluated. When it was necessary, appropriate treatment with antihistamines and with corticosteroids was administered. Laboratory investigations included complete blood count and differential count, sedimentation rate, quantitative immunoglobulin levels of IgE, IgG, IgM, and IgA, complement components C3 and C4, and urinalysis. In addition, serum samples were collected for the immunologic tests described below. A skin biopsy was performed at the height of the reaction on one patient for light microscopy and immunofluorescencestudies of the cutaneous lesions. Skin tests for immediate hypersensitivity to the vaccine and its components were carried out. Similar evaluations of control subjects matched on the basis of vaccination history and previous IgG antibody response to rabies virus were also undertaken.
Skin tests Tests were performed by prick test by use of HDCRV or the mock vaccine (1:10 dilution and undiluted), FCS (1 : 10 and undiluted; Gibco, Grand Island Biological Co., Grand Island, N. Y.), and neomycin (1 x 10 -2, Sigma Chemical Co., St. Louis, Mo.) approximately 2 to 6 weeks after the reaction and antihistamines had been stopped. A positive reaction was considered to be a wheal with flare 2 mm or larger than the control response to diluent, read at 12 minutes. All skin tests were carried out and read by a single individual (C. J. M.).
Determination of rabies virus neutralizing antibody levels in serum Determination was undertaken according to the method of Smith et al2 with a rapid fluorescent-focus inhibition
technique with BHK-21 cells infected with the tissue culture-adapted CVS-11 rabies virus. Results are expressed as the titer of the serum causing a 50% reduction in the number of low-power fields containing fluorescing cells.
ELISA for IgG antibodies to FBS FBS was coupled to a 96-well Linbro Titertek plate (Flow Laboratories, Mississauga, Ontario) in 200 ixl per well for 48 hours at 37~ C at a concentration of 200 ng/ml in sodium carbonate Coupling buffer (1.59 gm NaCO2, 2.93 gm NaHCO3, 0.2 gm NaN3 per liter, pH 9.6). Excess sites were blocked by use of a 1 : 100 dilution of normal sheep serum for 24 hours at 37~ C. Fresh serum samples were diluted in phosphate-buffered saline containing Tween 20 (Fisher Scientific, Toronto, Ontario) and 200 ~1 per well was added to the plate. After incubation for 1 hour at 37~ C, the plate was washed and an additional 1-hour incubation with a 1:1000 dilution of alkaline phosphatase-coupled goat antihuman IgG antibody or antihuman IgM antibody (Tago Laboratories, Burlingame, Calif.) was undertaken. The plate was then washed and incubated at room temperature for 30 minutes with substrate 104 (Sigma Chemical Co.). Color was read at 410 nm on a Dynatech minireader MR 590 (Flow Laboratories). Sera from patient and control groups and normal unimmunizedcontrol subjects were studied. Results were expressed as OD at 410 nm, and mean -+ SD were calculated foraU groups, To examine the effects of substituting 13-propiolactone-modified FBS in the ELISA, a 1:10 dilution of 13-propiolactone (1.15 gm/ml, Sigma Chemical Co.) was added to each well 24 hours after the initiation of antigen coupling. Blocking of excess reactive sites was carried out 24 hours later with a 1:100 dilution of normal sheep serum.
ELISA for circulating immune complexes ELISA was carried out by a Clq solid-phase assay by use of the same technique as described for the FCS ELISA but substituting 200 ng/rni of purified Clq. 6 Serum samples
VOLUME 79 NUMBER 4
Serum sickness after i m m u n i z a t i o n
607
TABLE II. Results of prick skin testing for immediate hypersensitivity to HDCRV components No. of patients with positive t est s (%)
Group
No.
HDCRV*
Mock HDGRV*
FCS*
Neomycin1"
Patients Control subjects
(10) (7)
6 (60) 0
2,/7 (29) 0/7 (0)
.3 (30):!: 2 (29)
(~
i)
*Tested as undiluted and 1710 dilution in buffered saline diluent. ~Tested at 1 • 10 --~ mol/L solution. ~:Skin test positive also to HDCRV.
for the Clq binding assay were collected and frozen immediatelyl Sera from patient and control groups and from normal unimmunized control subjects were studied. Sera from patients with active systemic lupus erythematosus and heat-aggregated lgG were used as positive control subjects.
ELISA for ~ to mock vaccine, HSA, ~ I ~ : m p i o l a c t e n e - m o d i f i e d HSA ELISA was carried out as described above substituting 200 ng/ml of the "mock" vaccine, HSA (Pentax; Miles Laboratory Inc., Rexdale, Ontario), or 13-pmpiolactonemodified HSA for coupling to the plate. Sera from patient and control groups and normal unimmunized control subjects were studied. SUlthaies The results from patient and control groups were compared by Student's t test.
RESULTS Ten (three female and seven male subjects) of 485 subjects receiving HDCRV booster immunization, developed a reaction at 10 to 12 days after vaccination characterized by generalized urticaria and angioedema, with occasional arthralgias but n o arthritis, lasting for 2 to 4 days, and requiring treatment in nine of 10 subjects with antihistamines and with corticosteroids (Table I). A punch skin biopsy specimen of a recent urticarial lesion demonstrated a leukocytoelastic vasculitis with neutrophil and lymphocyte infiltrates and fibrin and Clq granular deposits on immunofluorescence, All these individuals had (1) received previous immunizations with HKCRV 6 and/or HDCRV, 9 (2) either intradermal (0.1 ml) or intramuscular (1.0 ml) booster HDCRV immunization, and (3) high (>2.5 IU) IgG antibody responses to the rabies virus. Initially 10 control subjects were selected, matched for the above characteristics (1 to 3), and were booster immunized with 0.1 ml of HDCRV intradermally. Three of the 10 subjects subsequently developed the characteristic hypersensitivity reaction to the vaccine.
When the remaining seven control subjects (three female and four male subjects) were compared with the allergic subjects, no differences were found after immunization in regard to complete blood counts, sedimentation rates, quantitative immunoglobutins, urinalysis, and IgG antibodies to rabies virus ( > 2 . 5 IU). The mean age of the patient group was 34.9 • 9.1 (SD) years, compared to 38.7 • 12.5 years tor the control group. Skin t e s t s The patient group was skin tested 26. I • t4 days after immunization, whereas the control group was studied 44 • 1.5 days after immunization. The delay for the latter was required in order to ensure that an appropriate increase in lgG antibodies to rabies virus had occurred after immunization. As can be observed in Table II. a total of 60% of the patients demonstrated a positive prick test to the HDCRV (undiluted or al 1 : 10) as compared to none of the control subjects. A smaller proportion, 29%, reacted to the rm~ck vacone. Approximately one third of patients was positive to FBS (these individuals were also positive to HDCRV), whereas 29% of control subjects was positive to FCS but not to HDCRV. The positive skin tests to FCS in the patient group were larger than those in the control group (mean diameter 4 m m compared to 2 mm tested with undiluted FCS), and the patients demonstrated increased sensitivity (a mean reaction of 3 mm to a l : l0 dilution of FCS compared to a negative reaction to FCS in the control subjects). No subject in either group reacted to neomycin.
Circulating immune complines In both patient and control groups, there were ~bund to be increased low levels of circulating imnmne complexes m some subjects as measured by Clq solidphase ELISA, but the differences between the two groups were marginally significant only at the 10% level (Fig. 1). The s h a d e d area in the figure indicates the normal (immunized) range :• 2 SD. The
608 Warrington et al.
J. ALLERGYCLIN.IMMUNOL. APRIL 1987
60
180 9
p <0.1 160
50 E t-
140
40
O T-
d.
p < 0.0125
120
30
o
20
8O |
10 I
PATIENTS
0
I
CONTROL
RG. 1. CIq binding ELISA of sera from patients and immunized control subjects. Mean SEM is illustrated. H a t c h e d area indicates mean -+ 2 SD for normal unimmunized control subjects.
mean OD 410 nm ___ SEM for the patients was 29.06 + 10.4 as compared to that of control subjects, 20.86 + 11. IgG antibodies to FBS
All patients developing reactions to the HDCRV (with one exception) demonstrated an increase in IgG antibodies to FCS by ELISA (Fig. 2) as compared to unimmunized and immunized control subjects (p < 0.0125). Thus, in part, the IgG antibodies to mock vaccine could be directed against FCS contaminants in the vaccine. When 13-propiolactone-incubated FCS was substituted for the FCS (Fig. 3), an overall increase in binding of IgG antibody to the plate was noted (p < 0.0005). This was not due to a nonspecific change in "stickiness" of the HSA, since an increase in control level binding was not observed. IgG antibodies to mock vaccine
There was a significant difference in the presence of IgG antibodies to mock vaccine as detected by ELISA (Fig. 4) when patients and matched control subjects were compared. The patient group had increased levels of antibody to nonviral components of the vaccine (OD 410 nm 68 --- 49.5) compared to immunized control subjects (20.4 -4- 5.4). This difference was significant (p < 0.0125) by Student's t test. A significant elevation of IgM antibodies to the mock vaccine was not found. IgG antibodies to HSA
It was found that 6/9 subjects demonstrated an increase in IgG binding to 13-propiolactone-modified HSA as compared to HSA itself (p < 0.0025). This
40 2O 0
i
PAI"IBCrS
i
COIOROI~
RG. 2. ELISA for IgG antibodies to "mock" vaccine, identical to HDCRV but lacking the rabies virus. Samples from patients and immunized control subjects are compared. Mean -+ SEM for the groups are illustrated. Hatched area indicates mean _+ 2 SD for normal unimmunized control subjects.
increase in binding was not due to a nonspecific effect of the 13-propiolactone, since a similar increase did not occur with the control samples. Serial absorption studies were carried out with two sera that demonstrated high IgG binding to the mock vaccine and to 13-propiotactone-modified HSA. It was found that 30% to 50% of the IgG binding activity for mock vaccine was removed by two absorptions on 13-propiolactone-modified HSA-coated plates. However, 90% to 100% of IgG binding activity was removed by two absorptions on the mock vaccine; <10% of activity was removed on modified FCS. DISCUSSION
HDCRV, introduced into North America in 1980, is prepared by 13-propiolactone inactivation of rabies virus grown in tissue culture in human embryonic lung cells. Vaccine contaminants of the Institute Merieux version of HDCRV potentially include FCS, HSA, cellular components, neomycin, vitamins, and/3-propiolactone. Two cases of urticaria and angioedema after booster immunization with the vaccine were reported from Europe in 1976,~ whereas additional cases of systemic allergic reactions have recently been described in the United States after primary or booster preexposure immunization. 7 In the present study, 10 of 485 subjects receiving
VOLUME 79 NUMBER
Serum sickness after immur~izatiop,
609
4
12s~ I
Ht~AN SERUM ALB~
1,0! lOG to FC$
!
150 i
! lO0~-
130!
P< 0.0125 E o
110~" 100~
E r5~-
5~t
9
O
T
6 4o~!
5oI
20
p,
0 [
0~ , d r
PATIENTS
CONTROLS
FIG. 3. ELISA for 10G antibodies to FCS. Samples from patients and immunized control subjects are compared. Mean +_ SEM for the groups are illustrated, Hatched area indicates mean +- 2 SD for normal unimmunized control subjects.
M
"
t
~-PROPIOLACTONE
--
~
4-
,
.......
-L-------
--
4"
t= < 0,0025
t N.S,
PATIENTS
CONTROLS
FIG. 4. ELISA for IgG antibodies to HSA and (~-propiolactone-modified HSA. Samples from-patients and immunized control subjects are illustrated, whereas results with HSA and ~-propiolactone-modiftedHSA are compared, Mean ~- SEM for the groups are illustrated,
120
FOC~/aL CALF j)
100
H ~ R V booster immunizations developed generalized ttrticaria and angioedema 10 to 12 days after vaccination. Symptoms lasted for 2 to 4 days and usually responded rapidly to antihistamines and corticosteroids. Although no significant changes in complement components or circulating immune complexes (by Clq binding assay) occurred and urinalysis was normal, skin biopsy specimen findings typical of a leukocytoclastic vasculitis were demonstrated with fibrin and C3 granular deposits on immunofluorescence. None of the patients developing reactions were IgA deficient, whereas one subject had an elevated IgE level. Other laboratory parameters were normal. The only common features noted in the reactor group were (1) usually a previous immunization with HKCRV, and (2)high (>2.5 IU) IgG antibody levels to rabies vaccine. Although the overall reaction rate was only 2%, when 10 subjects were selected to act as matched control subjects in the study (based on vaccination history and previous antibody response to HDCRV), three of these 10 patients developed the characteristic hypersensitivity reaction after reimmunization. Therefore, the above criteria may define a high-risk group. The reactors to HDCRV demonstrated on skin testing an increased frequency of immediate skin test
E c o
/
/
/
/
80
6O
6 4O
20 I
~ .II~IOPIOLACTONE
i
--
.p
t :<0.0005
PATIENTS
J ~
l
--
.p
t N.S.
CONTROLS
FIG. 5. ELISA for tone-modified P nized control sul; FCS and I~-propiolaCtone-modified FCS are compared, Mean _+ SEM for the groups are iltustrated~
reactivity to the HDCRV and to the mock vaccine, as compared to control subjects. A l ~ h the frequency of immediate skin reactions to FCS in the two groups was the same, reactivity was greater in those with sensitivity to the vaccine, that is, reactions were larger and positive at a greater dilution of the FCS. Since
610 W a r r i n g t o n et al.
J. ALLERGYCLIN. )MMUNOL. APRIL 1987
60% of patients reacted to the HDCRV, it is likely that in part this reactivity was to a vaccine component other than the FCS. The skin reactions to mock vaccine would suggest that this response was not to the rabies virus itself but could be to FCS and other vaccine components. The clinical presentation of these patients is suggestive of a serum sickness-like reaction involving a primary immune response to a vaccine component. In the reactors group, a number of patients demonstrated elevated levels of IgG antibodies to FCS in addition to possessing IgE antibodies to the FCS. Thus, in some subjects, the presence of IgE and IgG antibodies to FCS might be responsible for the hypersensitivity reaction. The patient group demonstrated evidence for both an IgE and an IgG response to vaccine components other than the rabies virus and FCS. There were other possible candidates for such a response, these being (1) cellular or membrane components of the diploid MRC-5 cell line used to propagate the virus, and (2) modified protein components, e.g., fl-propiolactonemodified HSA or BSA. In six of nine samples tested, it appeared that IgG antibodies to [3-propiolactoneexposed HSA were present, suggesting that in some cases modified normal serum components may become antigenic. In addition, some increase in IgG binding to [3-propiolactone-modified FCS was observed with sera from the patient group. It was not possible for us to test reactivity to potential cellular contaminants in the vaccine. However, the absorption studies would suggest the presence of IgG antibodies to antigens other than [3-propiolactone-modified HSA. In no case could we demonstrate significantly elevated levels of IgM antibodies directed to the vaccine components. Shortly before completion of this study, we also became aware of two other investigations into the pathogenesis of this hypersensitivity reaction to the vaccine. Deutsch et al. s demonstrated immediate skin test reactivity to HDCRV in six patients with a similar serum sickness-like reaction after booster immunization, whereas Baer et al.9 found IgE antibodies to [3-propiolactone-modified HSA and to the whole vaccine by RAST in four sera from patients with allergic reactions to HDCRV. It would therefore appear that both IgE and IgG antibodies to vaccine components
other than the virus may be responsible for the hypersensitivity reactions observed on booster immunization. The development of a leukocytoclastic vasculitis with Clq deposits might indicate that IgE antibody-induced permeability changes facilitate deposition of complement-activating IgG antibodies to FCS or modified HSA, producing the urticarial vasculitis observed in this condition. It is obviously of some importance to define and eliminate, if it is possible, the causative antigens of this reaction or to develop predictive criteria for subjects at high risk of developing the hypersensitivity response. If such antigens are rendered immunogenic by reagents used in viral inactivation, it may be necessary to use alternative methods for this process, since regular immunizations are essential to ensure protection of individuals in high-risk occupations. We acknowledge the secretarial assistance of Mrs. S. Gustavson and Mrs. T. Coulson and the cooperation and collaboration of Institute Merieux and Rhone-Poulenc Pharma Inc. REFERENCES
I. Cox JH, Schneider LG. Prophylactic immunization of human against rabies by intradermal inoculation of human diploid cell culture vaccine, J Clin Microbiol 1976;3:96. 2. Nicholson KG, Turner GS, Aoki FY. Immunizationwith a human diploid cell strain of rabies virus vaccine:two-year results. J Infect Dis 1978;137:783. 3. AndersonLJ, Winkler WG, Hafkin B, et al. Clinical experience with a human diploid cell rabies vaccine. JAMA 1980;244:781. 4. Bernard KW, Roberts MA, Summer J, et al. Human diploid cell rabies vaccine: effectiveness of immunization with small intmdermal or subcutaneous doses. JAMA 1982;247:1138. 5. Smith JS, Yager PA, Baer GM. A rapid reproducible test for determining rabies neutralizing antibody. Bull WHO 1973; 48:535. 6. Pohl DA, Gibbons JJ Jr, Tsai CC, Roodman ST. Isolation and purification of human Clq from plasma. J Immunol Methods 1980;36:13. 7. SchurrengergerE Dreesen DW, Brown J, et al. Systemic reactions following immunizations with human diploid cell rabies vaccine. MMWR 1984;33:185. 8. Deutsch AJ, Dreesen DW, Bernard KW. Serum sickness-like reaction associated with skin-sensitizing antibody following booster dose of rabieshuman diploid cell vaccine [Abstract]. J ALLERGYCLINIMMUNOL!985;75:108. 9. Baer H, Anderson MD, Bernard K, Quirman G. Beta-propiolactone-treated human serum albumin (BPL-HSA):an allergen for humans receiving rabies vaccine lAbstraCt]. J Clin Immunol 1985;75:137.
Ten patients developing a serum sickness-like hypersensitivity reaction to human diploid cell rabies vaccine were studied and compared with control subjects matched for previous vaccination history and level of rabies virus-specific IgG response to immunization. The clinical reaction consisted of delayed onset, generalized urticaria, and angioedema, with some arthralgias. Skin biopsy specimens demonstrated a leukocytoclastic vasculitis. Individuals reacting to the vaccine possessed lgE antibodies to human diploid cell rabies vaccine, to mock vaccine lacking viral antigen, and to fetal calf serum (FCS), a vaccine trace contaminant. Increased levels of lgG antibodies to FCS, mock vaccine, and fl-propiolactone-modified FCS, and human serum albumin were also found. Such humoral responses to vaccine components other than rabies virus might be responsible for the hypersensitivity reactions developing after rabies vaccination. (J ALLERGY CLIN IMMUNOL 1987;79:605-10.)
The reactivated HDCRV used in these studies is a lyophilized and stabilized suspension of rabies virus inactivated by 13-propiolactone. It is prepared in human diploid cell cultures of the MRC-5 embryonic lung cell line and contains trace amounts of FCS, human albumin, neomycin, vitamins, and 13-propiolactone.* Adverse reactions to the vaccine have previously been limited and mild, either local reactions at the injection site or systemic reactions of headache, malaise, nausea, and dizziness. In 1976, Cox and Schneider ~ reported two patients with generalized urticaria and angioedema after booster immunization with the vaccine. This present study was initiated in 1983 when three of our patients developed similar reactions suggestive of a hypersensitivity response that appeared somewhat unusual in the light of previously reported clinical experience with the vaccine. 2"4 During this survey, 10/485 patients (2.06%)
From the Departmentsof Medicine, Immunologyand Medical Microbio|ogy, Universityof Manitoba, and the Section of Allergy and Clinical Immunology,Health Sciences Center, Winnipeg, Manitoba, Canada. Received for publication Jan. 23, 1986. Accepted for publication Sept. 20, 1986. Reprint requests: Richard J. Warrington, M.B., Sectionof Allergy and Clinical Immunology,C319, Health Sciences Centre, 700 William Ave., Winnipeg, Manitoba R3E 0Z3. *Fontaine R. Personalcommunication.
Abbreviations used
HDCRV: FCS: HSA: HKCRV: OD:
Human diploid cell rabies vaccine Fetal calf serum Human serum alLnmam Hamster kidney cell rabies vaccine Optical density
developed urticaria and angioedema IO to 12days-af~er receiving a booster immunization with ~ R V . In an attempt to determine the cause of this hypersensitivity reaction, immune responses to the vaccine components were studied in the patient group and in a matched control group of immunized subjects.
Subjects A total of 485 subjects received primary "and booster immunizations with HDCRV (Institute Merieux, Lyon. France) during a period of 2 years, in a study direemd by F. Y. A. to produce hyperimmmae subjects for the isolation of human rabies-specific imrrmnoglob~lin. Subjects were individuals at increased risk of exposure to ra~ies virus from the provinces of Manitoba and Sa~katchewaa. Most vaccine recipients had previously been immunized with HKCRV.
Rabies virus vaccine The HDCRV was prepared by Institute Merieax and was supplied by Rhone-Poulenc Pharma Inc.. Montreal. Que6O5
606
W a r r i n g t o n et al.
J. ALLERGY CLIN. ]MMUNOL. APRIL 1987
TABLE I. Characteristics of patients reacting to HDCRV Patient
Past vaccination history
Mode of administration
Time of onset
Abnormal clinical tests
1 2 3 4 5 6 7 8 9 10
Ha/Hu Ha/Hu Hu Hu Ha/Hu Du/Ha/Hu Du/Hu Ha/Hu Ha/Hu Ha/Hu
SC SC ID SC SC ID SC ID SC ID
7 10 8 11 12 12 12 8 10 10
Nil lgM ~, C $ C Nil Nil C4 ~, IgM ~, C Nil Nil Nil IgM ~,/C ~,
Ha = hamster kidney vaccine; Hu = human diploid cell vaccine; Du = duck embryo vaccine; SC = subcutaneously; and ID = intradermally.
bec. An identical "mock" vaccine without the rabies virus was supplied to us by Institute Merieux.
Clinical evaluation All subjects developing a reaction to the vaccine were observed by one of us (R. J. W.) within 1 to 3 days of the reaction and clinically evaluated. When it was necessary, appropriate treatment with antihistamines and with corticosteroids was administered. Laboratory investigations included complete blood count and differential count, sedimentation rate, quantitative immunoglobulin levels of IgE, IgG, IgM, and IgA, complement components C3 and C4, and urinalysis. In addition, serum samples were collected for the immunologic tests described below. A skin biopsy was performed at the height of the reaction on one patient for light microscopy and immunofluorescencestudies of the cutaneous lesions. Skin tests for immediate hypersensitivity to the vaccine and its components were carried out. Similar evaluations of control subjects matched on the basis of vaccination history and previous IgG antibody response to rabies virus were also undertaken.
Skin tests Tests were performed by prick test by use of HDCRV or the mock vaccine (1:10 dilution and undiluted), FCS (1 : 10 and undiluted; Gibco, Grand Island Biological Co., Grand Island, N. Y.), and neomycin (1 x 10 -2, Sigma Chemical Co., St. Louis, Mo.) approximately 2 to 6 weeks after the reaction and antihistamines had been stopped. A positive reaction was considered to be a wheal with flare 2 mm or larger than the control response to diluent, read at 12 minutes. All skin tests were carried out and read by a single individual (C. J. M.).
Determination of rabies virus neutralizing antibody levels in serum Determination was undertaken according to the method of Smith et al2 with a rapid fluorescent-focus inhibition
technique with BHK-21 cells infected with the tissue culture-adapted CVS-11 rabies virus. Results are expressed as the titer of the serum causing a 50% reduction in the number of low-power fields containing fluorescing cells.
ELISA for IgG antibodies to FBS FBS was coupled to a 96-well Linbro Titertek plate (Flow Laboratories, Mississauga, Ontario) in 200 ixl per well for 48 hours at 37~ C at a concentration of 200 ng/ml in sodium carbonate Coupling buffer (1.59 gm NaCO2, 2.93 gm NaHCO3, 0.2 gm NaN3 per liter, pH 9.6). Excess sites were blocked by use of a 1 : 100 dilution of normal sheep serum for 24 hours at 37~ C. Fresh serum samples were diluted in phosphate-buffered saline containing Tween 20 (Fisher Scientific, Toronto, Ontario) and 200 ~1 per well was added to the plate. After incubation for 1 hour at 37~ C, the plate was washed and an additional 1-hour incubation with a 1:1000 dilution of alkaline phosphatase-coupled goat antihuman IgG antibody or antihuman IgM antibody (Tago Laboratories, Burlingame, Calif.) was undertaken. The plate was then washed and incubated at room temperature for 30 minutes with substrate 104 (Sigma Chemical Co.). Color was read at 410 nm on a Dynatech minireader MR 590 (Flow Laboratories). Sera from patient and control groups and normal unimmunizedcontrol subjects were studied. Results were expressed as OD at 410 nm, and mean -+ SD were calculated foraU groups, To examine the effects of substituting 13-propiolactone-modified FBS in the ELISA, a 1:10 dilution of 13-propiolactone (1.15 gm/ml, Sigma Chemical Co.) was added to each well 24 hours after the initiation of antigen coupling. Blocking of excess reactive sites was carried out 24 hours later with a 1:100 dilution of normal sheep serum.
ELISA for circulating immune complexes ELISA was carried out by a Clq solid-phase assay by use of the same technique as described for the FCS ELISA but substituting 200 ng/rni of purified Clq. 6 Serum samples
VOLUME 79 NUMBER 4
Serum sickness after i m m u n i z a t i o n
607
TABLE II. Results of prick skin testing for immediate hypersensitivity to HDCRV components No. of patients with positive t est s (%)
Group
No.
HDCRV*
Mock HDGRV*
FCS*
Neomycin1"
Patients Control subjects
(10) (7)
6 (60) 0
2,/7 (29) 0/7 (0)
.3 (30):!: 2 (29)
(~
i)
*Tested as undiluted and 1710 dilution in buffered saline diluent. ~Tested at 1 • 10 --~ mol/L solution. ~:Skin test positive also to HDCRV.
for the Clq binding assay were collected and frozen immediatelyl Sera from patient and control groups and from normal unimmunized control subjects were studied. Sera from patients with active systemic lupus erythematosus and heat-aggregated lgG were used as positive control subjects.
ELISA for ~ to mock vaccine, HSA, ~ I ~ : m p i o l a c t e n e - m o d i f i e d HSA ELISA was carried out as described above substituting 200 ng/ml of the "mock" vaccine, HSA (Pentax; Miles Laboratory Inc., Rexdale, Ontario), or 13-pmpiolactonemodified HSA for coupling to the plate. Sera from patient and control groups and normal unimmunized control subjects were studied. SUlthaies The results from patient and control groups were compared by Student's t test.
RESULTS Ten (three female and seven male subjects) of 485 subjects receiving HDCRV booster immunization, developed a reaction at 10 to 12 days after vaccination characterized by generalized urticaria and angioedema, with occasional arthralgias but n o arthritis, lasting for 2 to 4 days, and requiring treatment in nine of 10 subjects with antihistamines and with corticosteroids (Table I). A punch skin biopsy specimen of a recent urticarial lesion demonstrated a leukocytoelastic vasculitis with neutrophil and lymphocyte infiltrates and fibrin and Clq granular deposits on immunofluorescence, All these individuals had (1) received previous immunizations with HKCRV 6 and/or HDCRV, 9 (2) either intradermal (0.1 ml) or intramuscular (1.0 ml) booster HDCRV immunization, and (3) high (>2.5 IU) IgG antibody responses to the rabies virus. Initially 10 control subjects were selected, matched for the above characteristics (1 to 3), and were booster immunized with 0.1 ml of HDCRV intradermally. Three of the 10 subjects subsequently developed the characteristic hypersensitivity reaction to the vaccine.
When the remaining seven control subjects (three female and four male subjects) were compared with the allergic subjects, no differences were found after immunization in regard to complete blood counts, sedimentation rates, quantitative immunoglobutins, urinalysis, and IgG antibodies to rabies virus ( > 2 . 5 IU). The mean age of the patient group was 34.9 • 9.1 (SD) years, compared to 38.7 • 12.5 years tor the control group. Skin t e s t s The patient group was skin tested 26. I • t4 days after immunization, whereas the control group was studied 44 • 1.5 days after immunization. The delay for the latter was required in order to ensure that an appropriate increase in lgG antibodies to rabies virus had occurred after immunization. As can be observed in Table II. a total of 60% of the patients demonstrated a positive prick test to the HDCRV (undiluted or al 1 : 10) as compared to none of the control subjects. A smaller proportion, 29%, reacted to the rm~ck vacone. Approximately one third of patients was positive to FBS (these individuals were also positive to HDCRV), whereas 29% of control subjects was positive to FCS but not to HDCRV. The positive skin tests to FCS in the patient group were larger than those in the control group (mean diameter 4 m m compared to 2 mm tested with undiluted FCS), and the patients demonstrated increased sensitivity (a mean reaction of 3 mm to a l : l0 dilution of FCS compared to a negative reaction to FCS in the control subjects). No subject in either group reacted to neomycin.
Circulating immune complines In both patient and control groups, there were ~bund to be increased low levels of circulating imnmne complexes m some subjects as measured by Clq solidphase ELISA, but the differences between the two groups were marginally significant only at the 10% level (Fig. 1). The s h a d e d area in the figure indicates the normal (immunized) range :• 2 SD. The
608 Warrington et al.
J. ALLERGYCLIN.IMMUNOL. APRIL 1987
60
180 9
p <0.1 160
50 E t-
140
40
O T-
d.
p < 0.0125
120
30
o
20
8O |
10 I
PATIENTS
0
I
CONTROL
RG. 1. CIq binding ELISA of sera from patients and immunized control subjects. Mean SEM is illustrated. H a t c h e d area indicates mean -+ 2 SD for normal unimmunized control subjects.
mean OD 410 nm ___ SEM for the patients was 29.06 + 10.4 as compared to that of control subjects, 20.86 + 11. IgG antibodies to FBS
All patients developing reactions to the HDCRV (with one exception) demonstrated an increase in IgG antibodies to FCS by ELISA (Fig. 2) as compared to unimmunized and immunized control subjects (p < 0.0125). Thus, in part, the IgG antibodies to mock vaccine could be directed against FCS contaminants in the vaccine. When 13-propiolactone-incubated FCS was substituted for the FCS (Fig. 3), an overall increase in binding of IgG antibody to the plate was noted (p < 0.0005). This was not due to a nonspecific change in "stickiness" of the HSA, since an increase in control level binding was not observed. IgG antibodies to mock vaccine
There was a significant difference in the presence of IgG antibodies to mock vaccine as detected by ELISA (Fig. 4) when patients and matched control subjects were compared. The patient group had increased levels of antibody to nonviral components of the vaccine (OD 410 nm 68 --- 49.5) compared to immunized control subjects (20.4 -4- 5.4). This difference was significant (p < 0.0125) by Student's t test. A significant elevation of IgM antibodies to the mock vaccine was not found. IgG antibodies to HSA
It was found that 6/9 subjects demonstrated an increase in IgG binding to 13-propiolactone-modified HSA as compared to HSA itself (p < 0.0025). This
40 2O 0
i
PAI"IBCrS
i
COIOROI~
RG. 2. ELISA for IgG antibodies to "mock" vaccine, identical to HDCRV but lacking the rabies virus. Samples from patients and immunized control subjects are compared. Mean -+ SEM for the groups are illustrated. Hatched area indicates mean _+ 2 SD for normal unimmunized control subjects.
increase in binding was not due to a nonspecific effect of the 13-propiolactone, since a similar increase did not occur with the control samples. Serial absorption studies were carried out with two sera that demonstrated high IgG binding to the mock vaccine and to 13-propiotactone-modified HSA. It was found that 30% to 50% of the IgG binding activity for mock vaccine was removed by two absorptions on 13-propiolactone-modified HSA-coated plates. However, 90% to 100% of IgG binding activity was removed by two absorptions on the mock vaccine; <10% of activity was removed on modified FCS. DISCUSSION
HDCRV, introduced into North America in 1980, is prepared by 13-propiolactone inactivation of rabies virus grown in tissue culture in human embryonic lung cells. Vaccine contaminants of the Institute Merieux version of HDCRV potentially include FCS, HSA, cellular components, neomycin, vitamins, and/3-propiolactone. Two cases of urticaria and angioedema after booster immunization with the vaccine were reported from Europe in 1976,~ whereas additional cases of systemic allergic reactions have recently been described in the United States after primary or booster preexposure immunization. 7 In the present study, 10 of 485 subjects receiving
VOLUME 79 NUMBER
Serum sickness after immur~izatiop,
609
4
12s~ I
Ht~AN SERUM ALB~
1,0! lOG to FC$
!
150 i
! lO0~-
130!
P< 0.0125 E o
110~" 100~
E r5~-
5~t
9
O
T
6 4o~!
5oI
20
p,
0 [
0~ , d r
PATIENTS
CONTROLS
FIG. 3. ELISA for 10G antibodies to FCS. Samples from patients and immunized control subjects are compared. Mean +_ SEM for the groups are illustrated, Hatched area indicates mean +- 2 SD for normal unimmunized control subjects.
M
"
t
~-PROPIOLACTONE
--
~
4-
,
.......
-L-------
--
4"
t= < 0,0025
t N.S,
PATIENTS
CONTROLS
FIG. 4. ELISA for IgG antibodies to HSA and (~-propiolactone-modified HSA. Samples from-patients and immunized control subjects are illustrated, whereas results with HSA and ~-propiolactone-modiftedHSA are compared, Mean ~- SEM for the groups are illustrated,
120
FOC~/aL CALF j)
100
H ~ R V booster immunizations developed generalized ttrticaria and angioedema 10 to 12 days after vaccination. Symptoms lasted for 2 to 4 days and usually responded rapidly to antihistamines and corticosteroids. Although no significant changes in complement components or circulating immune complexes (by Clq binding assay) occurred and urinalysis was normal, skin biopsy specimen findings typical of a leukocytoclastic vasculitis were demonstrated with fibrin and C3 granular deposits on immunofluorescence. None of the patients developing reactions were IgA deficient, whereas one subject had an elevated IgE level. Other laboratory parameters were normal. The only common features noted in the reactor group were (1) usually a previous immunization with HKCRV, and (2)high (>2.5 IU) IgG antibody levels to rabies vaccine. Although the overall reaction rate was only 2%, when 10 subjects were selected to act as matched control subjects in the study (based on vaccination history and previous antibody response to HDCRV), three of these 10 patients developed the characteristic hypersensitivity reaction after reimmunization. Therefore, the above criteria may define a high-risk group. The reactors to HDCRV demonstrated on skin testing an increased frequency of immediate skin test
E c o
/
/
/
/
80
6O
6 4O
20 I
~ .II~IOPIOLACTONE
i
--
.p
t :<0.0005
PATIENTS
J ~
l
--
.p
t N.S.
CONTROLS
FIG. 5. ELISA for tone-modified P nized control sul; FCS and I~-propiolaCtone-modified FCS are compared, Mean _+ SEM for the groups are iltustrated~
reactivity to the HDCRV and to the mock vaccine, as compared to control subjects. A l ~ h the frequency of immediate skin reactions to FCS in the two groups was the same, reactivity was greater in those with sensitivity to the vaccine, that is, reactions were larger and positive at a greater dilution of the FCS. Since
610 W a r r i n g t o n et al.
J. ALLERGYCLIN. )MMUNOL. APRIL 1987
60% of patients reacted to the HDCRV, it is likely that in part this reactivity was to a vaccine component other than the FCS. The skin reactions to mock vaccine would suggest that this response was not to the rabies virus itself but could be to FCS and other vaccine components. The clinical presentation of these patients is suggestive of a serum sickness-like reaction involving a primary immune response to a vaccine component. In the reactors group, a number of patients demonstrated elevated levels of IgG antibodies to FCS in addition to possessing IgE antibodies to the FCS. Thus, in some subjects, the presence of IgE and IgG antibodies to FCS might be responsible for the hypersensitivity reaction. The patient group demonstrated evidence for both an IgE and an IgG response to vaccine components other than the rabies virus and FCS. There were other possible candidates for such a response, these being (1) cellular or membrane components of the diploid MRC-5 cell line used to propagate the virus, and (2) modified protein components, e.g., fl-propiolactonemodified HSA or BSA. In six of nine samples tested, it appeared that IgG antibodies to [3-propiolactoneexposed HSA were present, suggesting that in some cases modified normal serum components may become antigenic. In addition, some increase in IgG binding to [3-propiolactone-modified FCS was observed with sera from the patient group. It was not possible for us to test reactivity to potential cellular contaminants in the vaccine. However, the absorption studies would suggest the presence of IgG antibodies to antigens other than [3-propiolactone-modified HSA. In no case could we demonstrate significantly elevated levels of IgM antibodies directed to the vaccine components. Shortly before completion of this study, we also became aware of two other investigations into the pathogenesis of this hypersensitivity reaction to the vaccine. Deutsch et al. s demonstrated immediate skin test reactivity to HDCRV in six patients with a similar serum sickness-like reaction after booster immunization, whereas Baer et al.9 found IgE antibodies to [3-propiolactone-modified HSA and to the whole vaccine by RAST in four sera from patients with allergic reactions to HDCRV. It would therefore appear that both IgE and IgG antibodies to vaccine components
other than the virus may be responsible for the hypersensitivity reactions observed on booster immunization. The development of a leukocytoclastic vasculitis with Clq deposits might indicate that IgE antibody-induced permeability changes facilitate deposition of complement-activating IgG antibodies to FCS or modified HSA, producing the urticarial vasculitis observed in this condition. It is obviously of some importance to define and eliminate, if it is possible, the causative antigens of this reaction or to develop predictive criteria for subjects at high risk of developing the hypersensitivity response. If such antigens are rendered immunogenic by reagents used in viral inactivation, it may be necessary to use alternative methods for this process, since regular immunizations are essential to ensure protection of individuals in high-risk occupations. We acknowledge the secretarial assistance of Mrs. S. Gustavson and Mrs. T. Coulson and the cooperation and collaboration of Institute Merieux and Rhone-Poulenc Pharma Inc. REFERENCES
I. Cox JH, Schneider LG. Prophylactic immunization of human against rabies by intradermal inoculation of human diploid cell culture vaccine, J Clin Microbiol 1976;3:96. 2. Nicholson KG, Turner GS, Aoki FY. Immunizationwith a human diploid cell strain of rabies virus vaccine:two-year results. J Infect Dis 1978;137:783. 3. AndersonLJ, Winkler WG, Hafkin B, et al. Clinical experience with a human diploid cell rabies vaccine. JAMA 1980;244:781. 4. Bernard KW, Roberts MA, Summer J, et al. Human diploid cell rabies vaccine: effectiveness of immunization with small intmdermal or subcutaneous doses. JAMA 1982;247:1138. 5. Smith JS, Yager PA, Baer GM. A rapid reproducible test for determining rabies neutralizing antibody. Bull WHO 1973; 48:535. 6. Pohl DA, Gibbons JJ Jr, Tsai CC, Roodman ST. Isolation and purification of human Clq from plasma. J Immunol Methods 1980;36:13. 7. SchurrengergerE Dreesen DW, Brown J, et al. Systemic reactions following immunizations with human diploid cell rabies vaccine. MMWR 1984;33:185. 8. Deutsch AJ, Dreesen DW, Bernard KW. Serum sickness-like reaction associated with skin-sensitizing antibody following booster dose of rabieshuman diploid cell vaccine [Abstract]. J ALLERGYCLINIMMUNOL!985;75:108. 9. Baer H, Anderson MD, Bernard K, Quirman G. Beta-propiolactone-treated human serum albumin (BPL-HSA):an allergen for humans receiving rabies vaccine lAbstraCt]. J Clin Immunol 1985;75:137.