Clinical and immunologic
observations in
patients who stop venom immunotherapy
David B. K. Golden, M.D., Kevin Johnson, B.S., Bonnie I. Addison,
Martin D. Valentine, M.D., Anne Kagey-Sobotka,
Ph.D., and
Lawrence M. Lichtenstein,
M.D., Ph.D. Baltimore, Md.
B.A.,
It is currently recommended that venom immunotherapy (VIT) be continued as long as the
sensitivity persists (indicated by positive venom skin tests or RAST). In this pilot study, we
performed a retrospective survey of the clinical and immunologic effects of stopping VIT. The 82
patients studied had received maintenance VIT for a mean of 14 months and had stopped VIT
a mean of 43 months before evaluation. Subsequent “field” stings in 28 patients caused systemic
reactions in six cases (22%), which is significantly higher than the 1% to 3% systemic reaction
rate in patients who remain on maintenance VIT. The 22% reaction rate is a minimal estimate
caused by loss of venom sensitivity in some patients and residual venom-spectjic IgG antibody
levels in others. Reevaluation of venom skin tests and IgG levels was possible in 43 patients. A
tenfold decline from before VIT skin test results was observed in 27 patients (63%). Skin tests
remained clearly positive in 32143 (74%) became weakly positive in 9143 (21 o/o),and 2143
(5%) became negative. The IgG level declinedfiom typical maintenance levels before stopping
VIT (mean 7.2 ? I .2 pglml) to levels typical of untreated patients at the time of retesting
(mean I .95 ? 0.3 pglml). Despite the marked fall of IgG antibody, one third of the patients
still had levels in the average range observed in patients receiving maintenance VIT. We
conclude that there is a substantial risk of anaphylactic sting reaction if VIT is stopped while
venom sensitivity persists. (J ALLERGYCLIN IMMUNOL77:43542, 1986.)
Systemic allergic reactions to insect stings can be
completely preventedin virtually all patientsreceiving
Hymenoptera venom immunotherapy.I, * Investigations during the past decadehave helped to clarify the
indications for VIT and to define the optimal methods
for induction, maintenance,and monitoring of clinical
protection. However, the required duration of such
treatment is uncertain and is of some concern to patients and physicians alike becauseof the inconvenience and cost of maintenanceVIT during a period
of years or decades.
Although it is currently recommendedthat maintenanceVIT be continued as long as the venom hypersensitivity persists, it has recently becomeevident
From The Johns Hopkins University School of Medicine, Department of Medicine, Division of Clinical Immunology, The Good
Samaritan Hospital, Baltimore, Md.
Supported by Grant AI08270 from the National Institute of Allergy
and Infectious Diseases.
Received for publication Sept. 24, 1984.
Accepted for publication Aug. 1, 1985.
Reprint requests: David B. K. Golden, M.D., Johns Hopkins Wniversity School of Medicine, Dept. of Medicine, Div. of Clinical
Immunology, The Good SamaritanHospital, 5601 Loch Raven
Blvd., Baltimore, MD 21239.
Publication No. 633 from the O’Neill Research Laboratories, The
Good Samaritan Hospital, Baltimore, Md.
Abbreviations used
VIT: Venom immunotherapy
MVV: Mixed vespid venom
PWV: Polistes wasp venom
HBV: Honeybee venom
YJV: Yellow jacket venom
that many patientsare stoppingtreatmentwith or without the consentof their physicians. We know the clinical and immunologic resultsof long-term venomtherapy but not the outcomeof stoppingtreatment.During
5 years of maintenanceVIT, the venom-specific IgG
antibody level is well *maintained,and clinical protection remainsin the rangeof 98%.3.4Venom-specific
IgE antibody levels and skin test sensitivity both decline steadily with time in most patients but rarely
become entirely negative.3-6Since antivenom IgG
levels fall to baseline6 months after a sting, whereas
venom sensitivity persists for years,3.’ the potential
risk of stopping VIT would appear to be great. We
therefore set out to document the clinical and immunologic sequelaeto stopping VIT. We report here
the first phaseof these studies that involved a retrospective survey for clinical and immunologic followup of patients who stoppedtreatment (dropouts). Our
435
436 Golden et al.
J. ALLERGY
TABLE f. Sting reactions after VIT was
stopped
-___
stirlgs
RbsaiQns
patients
N
Systemic
Large local
NORC
2x
6 (22%)
9 (32%)
13 (46%)
36
7 (19%)
II (31%)
18 (507c)
aim was to estimate the risk of sting reaction before
proceeding with a formal prospective study.
N?ETtiOUS
Wc identified 125 patients from our records who had
achievedmaintenancedosesof Hymenopteravenom during
at leastb monthsof treatmentand had subsequentlystopped
treatmentagainstour advice. Eighty-two patientsresponded
tt) a questionnaire, and 43 of these patients were observed
for reevaluation by venom skin testsand a blood samplefor
determinations of venom-specific antibody levels. The 82
patients who respondedto the survey had been treatedwith
full-dose maintenanceVI’J for a meanof 14 months (range
2 to 44 months) and had stopped treatment a mean of 43
months (range 9 to 79 months) before reevaluation. VI?
included all venoms to which the patient originally demonstrated positive skin tests, except that PWV was often
excluded becauseof unavailability (before 1978) and its
cross-reactivity with the other vespid venoms. Thirty-seven
patients had received MVV; 17. MVV and PWV; nine.
HHV; eight, HBV and MVV: four. HBV, MVV. and PWV:
two. HBV. YJV, and PWV: one. HBV and YJV: three.
YJV; and enc. PWV only. The meanageof these82 patients
was 41 years. and the maleifcmale ratio was 3 : 1. Twentyeight ol’ these dropouts had been stung at lcast 12 months
after stopping VIT (mean 36 months, range 12 to 72
months). Thesepatients did not differ significantly from the
patients who were not stung with respectto age, sex ratio,
duration of maintenanceVIT, severity of pretreatmentsting
reaction. or degreeof sensitivity on pretreatmentskin tests.
The 43 patients who presentedto the Allergy Center for
retestingrepresentedsimilar proportions of patientswho had
been stung !15/28) and of those who were not stung (28i
54). The patients who were retesteddid not differ significantly from those who were not available for retesting with
respectto any of the variables listed above.
Intfadermal venom skin tests were performed with all
five commercial Hymenoptera venom protein extracts at
concentrationsup to and including t .Opglml. Venom skin
tests were graded according to the method of Norman.’
Venom-specific I@ antibody levels were measuredby
the StaphyIococcalprotein A solid-phaseradioimmunoassay
previously described.”
RESULTS
Table I illustrates the outcome of insect stings that
occurred in these patients at least 12 months after
stopping VIT. Thirty-six stings occurred in 28 pa-
CLIN. IMMUNOL.
MARCH 1966
tients, causing seven systemic reactions in six of the
patients (22%). To put these data into perspective,
Table II iilustrates a comparison of the reaction rate
in the patients who stopped VIT to that of treated and
untreated patients. In a retrospective survey. concurrent controls could not, of course, be planned. Most
investigators have found a systemic sting reaction rate
of 1% to 3% in patients on long-term maintenance
venom therapy. I. ’ We perform sting challenges every
summer on patient volunteers as part of our monitoring
of long-term venom treatment. For comparison with
the patients who had stopped therapy, we selected the
patients we had deliberately stung in the summer of
this survey who had started treatment between 1978
to 1980 and had therefore been followed for the same
length of time as the dropouts. The systemic reaction
rate in this group of 30 patients who had continued
to receive maintenance immunotherapy for 3 to 5 years
was significantly less than the rate observed in the 28
patients who had stopped treatment (p < 0.04). The
rate of systemic reactions in untreated patients has
been reported in the range of 30% to 60% by many
investigators. The highest estimates of 6 1% systemic
reaction rate in untreated patients, Hunt et al.“’ and
Settipane and Chafee, ’ ’ are signficantly higher than
the rate observed in our dropouts. However, the systemic reaction rates reported by Parker et al.,” Dvorin
et al. ,’ and Blaauw et al.” in their separate investigations were not significantly different from the reaction rate in the patients who stopped venom therapy.
More than one sting episode had occurred in seven
of the patients surveyed. Five had only large local
reactions or no reaction to all stings. One patient
(Table III) had two stings, both of which caused severe
anaphylaxis with unconsciousness 13 months and 35
months after stopping venom therapy. Another patient
(Table IV) had been treated with HBV and MVV
based on his skin test sensitivities, but all his systemic
reactions had been from yellow jacket stings. A honeybee sting 43 months after stopping treatment
caused no systemic reaction. but a yellow jacket sting
45 months after therapy caused a severe systemic rcaction. His skin tests 45 months after therapy was
stopped were negative for honeybee venom and positive for the vespid venoms. It was not possible in
most cases to identify the culprit with absolute certainty. We cannot be sure that all patients were stung
with insects relevant to their sensitivity and treatment.
In four cases (of 28 patients stung) the patient was
quite uncertain of the culprit (two YJ allergic, 1 PW,
and 1 HB/MV). The other 24 patients were confident
of the identification. YJ was indicated by all 14 MVand all six MV/PW-treated patients; four beekeepers
were certain of HB stings.
As illustrated in Table III, there was no consistent
VOLUME
NUMBER
TABLE
77
3
Observations
II. Sting
reactions
in allergic
in patients
who stop venom
immunotherapy
437
patients
N
Patients
Systemic
28
30
VIT stopped
Maintenance VIT
Untreated
Hunt et al.‘”
Settipaneand Chafee”
Parkeret al.‘*
Dvorin et al.’
Blaauw et al.13
6
1
23
119
16
19
85
%
P*
22
3
co.04
14
61
72
61
7
8
44
<0.005
<O.ool
CO.20
CO.20
CO.40
42
21
23
*Comparedto patientswho stoppedVIT by Fisher’sexacttest.
TABLE
111.Systemic
reactions
in patients
who
stopped
VIT
Reaction
Duration*
of VIT
(mo)
4
Timet
of VIT to sting
lmol
Before VIT
35
U&aria, dizziness, dyspnea
Urticaria, dizziness, dyspnea
Urticaria
Urticaria, angioedema,dyspnea, dizziness
Angioedema, throat tightness, dizziness
Urticaria, dyspnea,dizziness,
hypotension
Urticaria, unconsciousness,
dyspnea
13
5
31
45
4
27
2
36
21
23
12
After VIT
Urticaria, dyspnea, unconsciousness
Urticaria, dyspnea, unconsciousness
Dyspnea, lightheadedness
Urticaria
Angioedema, throat tightness, dyspnea,
dizziness
Urticaria, throat tightness
Generalized erythema and pruritus
*Durationof full-dosemaintenance
VIT.
tTime whenVIT wasstoppedto sting.
tendency for the systemic reactions in these six patients to be either less severe or more severe after
therapy than they had been before. Sting reactions
also appeared to be unrelated to the time elapsed after
therapy stopped. The duration of maintenance VIT
was less than 2 years in 22/28 patients stung. No
conclusion can be made about the reactivity of patients
after prolonged VIT, but the two reactors who had
had more prolonged therapy had the mildest posttherapy sting reactions. Fig. 1 presents the severity of sting
reactions after stopping venom therapy as a function
of the pretreatment sting reaction in all 28 patients
who had been reshmg. Susceptibility to systemic sting
reactions after stopping treatment did not demonstrate
any correlation with the severity of reactions that occurred before VIT. The complete characteristics of the
15 patients who had been stung after stopping VIT
and were subsequently available for reassessment are
presented in Table IV. Once again, the severity of sting
reactions before VIT was in no way predictive of the
response to stings after stopping VIT. The results
of retesting for venom-specific IgG antibodies and
venom skin test sensitivity are presented with the time
of retesting (months after therapy was stopped). As
expected, the IgG level was in the low range in two
of the three systemic reactors, and the third systemic
reactor had been stung just a few weeks before testing,
which may have boosted the IgG level. The two patients with the strongest skin test sensitivity both had
systemic reactions when they were stung after stopping VIT; the third systemic reactor had relatively low
skin test sensitivity but had not declined in sensitivity
from the pretreatment level.
In the 43 patients who were available for retesting,
venom skin tests were performed with all five commercial Hymenoptera venoms. All patients had been
originally positive (2 + at 1.0 kg/ml or less) to at
least one of the venoms. We have excluded from the
analysis the results for PWVs, since not all patients
had been originally tested with PWV before its commercial production. Table V presents the skin test
results for the venoms to which each patient had originally been sensitive and treated. Results are also presented for the patient as a whole, in which the skin
438
Golden
J. ALLERGY
et al.
TABLE IV. Characteristics
of patients
stung after VIT was stopped
rt3mtion grade*
Ek#forw MT
Aftw
3
4
2
4
5
4
o*
5
3
3
2
5
5
4
2
Antivenom
Times
WT
0
0
0
0
0
0
0
0
i
1
1
I
2
3
4
Receiving
19
29
13
51
26
37
42
IS
38
33
24
40
36
12
45
“0 -- no reaction (before VIT indicates this was
vascular
symptoms; 4 = severe respiratory or
‘Concentration of venom required to elicit a 21
0 indicates negative.
fh&nths elapsed since VIT was stopped to time
PMonths elapsed since VIT was stopped to time
VIT
25.4
6.1
4.1
NA
NA
15.7
NA
7.2
15.6
2.6
32.8
7.4
16.3
21.3
3.0
CLIN. IMMUNOL.
MARCH 1986
and then reassessed
IgG (Ccglml)
Venom
After VIT
Time5
3.5
4.5
0.9
0.6
2.0
0.9
0.2
1.9
3.4
1.6
22.5
1.1
5.4
3.5
1.3
40
43
46
52
50
39
68
63
39
33
30
54
37
49
46
Before
skin testt
After VIT
VIT
0.1
0.01
0.1
1.0
0.1
0.1
0.1
1.0
0.1
0.1
1.0
0.1
0.01
1.0
1+
I .o
I+
1.0
1.0
0
1.0
1.0
I.0
I+
1.0
0.1
1.0
0.1
1.0
0.01
not primary culprit): I 7 large local: 2 = cutaneous systemic; 3 = mild respiratory or
vascular symptoms; 5 = unconsciousness or respiratory arrest.
grade skin test reaction (microgram per milliterj; 1 + indicates I + grade at 1.O pgiml:
of subsequent sting.
of reevaluation of antivenom IgC; and venom skin tests.
TABLE V. Venom skin test results
Reaction
Hymenoptera
venom
HHV
YJ\.
YHL
WFHV
Total skin tests
Patients’i
N
12
35
33
32
11:
43
10 x -Decline*
II
24
18
21
x4 (75%)
27 (63%)
grade at 1.0 pglml
0
1+
2+
3
8
3
21 (19%)
2 (5%)
4
4
9
7
24 (21%)
9 (21%)
2
28
16
21
67 (60%)
32 (74%)
6
YH =: Yellow hornet: WFH 1 White-faced homer.
*Tenfold decline in sensitivity equals at least a tenfold increase in the venom concentration required to elicit a 2 i
t0verall degree of patient sensitivity equals skin test eliciting strongest reaction.
test eliciting the strongest reaction is taken as an indicator of the patient’s overall degree of sensitivity.
There was at least a tenfold decline in sensitivity (i.e.,
at teast a tenfold increase in the venom concentration
required to elicit a 2 + reaction) in 84 of the 112 skin
tests (75%). These results include skin test results that
had declined to become entirely negative, which occurred in 2 1 of 112 skin tests ( 19%). A decline of at
least tenfold in overall skin test sensitivity was observed in 27 of the 43 patients (63%). The development of a negative skin test to any individual venom
was not uncommon, although only two of 43 patients
reaction.
(5%) had become entirely skin test negative to all
venoms. This decline or loss of sensitivity was most
frequent with honeybee venom whether or not the
patient was sensitive to other venoms, and was somewhat more common with yellow hornet venom than
with YJV or white-hornet venoms. A clear-cut positive reaction of at least 2 + to a concentration of 1.O
kg/ml was observed in 67 of 112 skin tests (60%),
and 32 of 43 patients (74%) maintained at least one
such clear-cut positive skin test. Twenty-one percent
of the skin tests, and patients, became only weakly
positive (1 + at 1 p,g/ml).
VOLUME
NUMBER
77
3
Observations
in patients
who stop venom
e---a
70-
.
50
immunotherapy
Not
439
stung
No reaction
0---o Large loco I
D----0 Systemic
-
30-
.
?
0
2
e
E
IO-
z
3-
:
I
F
z
75
I .o0.7 -
2
PRE-VI
3
4
T REACTION
5
GRADE
Fig. 1 Severity of insect-sting
reactions in 28 individual
patients before beginning VI1 and after VIT was stopped.
Reaction grades: 0, no reaction; 1, large local reaction; 2,
cutaneous systemic reaction; 3, mild respiratory
and/or
vascular systemic reaction; 4, moderately
severe respiratory and/or vascular systemic reaction; 5, severe systemic reaction with unconsciousness
or respiratory arrest.
The fate of venom-specific IgG antibody levels
when immunotherapy was stopped could be determined in 31 patients for whom paired serum samples
were available. Of these, 26 were yellow jacket IgG
measurements(19 patients treated with MVV, five
with MVV and PWV, and two with YJV), and five
were honeybeeIgG levels in patientstreatedonly with
HBV. Three of the vespid venom-treated patients had
also beenimmunized with HBV but had indicated that
all their stings had been from yellow jackets; HBVIgG levels were not available in these three patients.
The first sample in each case was to reflect the level
of IgG antibodies at the time that VIT was stopped,
but blood sampleswere often not available from the
actual date of the last venom injection. The IgG antibody responseto our standardmodified rush therapeutic regimen is known to rise to a peak level within
2 to 4 months and decline somewhat to the steadystate maintenancelevel in 6 to 9 months. This level
is well maintained in most casesduring subsequent
years of maintenancetreatment. Of the 31 paired serum samples analyzed, 23 of the sampleshad been
drawn from patients receiving therapy after 1% to 5
years of treatment and were considered to reflect
steady-statemaintenance levels of IgG; 18 of these
were drawn 0 to 6 months before therapy stopped,
four were drawn from 7 to 12 monthsbefore stopping,
and one was drawn 18 months before stopping (but 3
0.5 0.3-
\
I
,
0
I
I2
TIME SINCE
I
24
VENOM
I
36
THERAPY
I
46
I
60
I
72
STOPPEDhanttd
Fig. 2 Venom-specific
IgG antibody levels (microgram per
milliliter)
in 31 individual
patients listed before VIT was
stopped, and as a function of the time elapsed (months)
after venom therapy was stopped. The reaction to an insect sting after VIT was stopped (if any) is indicated for
each patient.
years into treatment). The other eight samplesfrom
patients receiving therapy were drawn 0 to 6 months
before stopping treatment, but thesewere only 3 to 6
months into maintenancetreatment and may reflect
peak IgG levels in early maintenancetherapy.
The 31 paired IgG antibody results are presented
in Fig. 2 for each of the individual patients and are
plotted as a function of the time elapsed since treatment was stopped. The outcome of stings after VIT,
if any, is presentedfor each patient. A few patients
demonstratedonly a small drop in IgG, but in general
a marked decreasehad occurred. There was no difference in the rate of decline of IgG nor in the mean
IgG level at the time of retesting between those patients who had been stung in the interim and those
who had not been. The rate of decline of venomspecific IgG antibody levels can only be estimated
from these data. The mean IgG levels are plotted in
Fig. 3 for all 31 patients before VIT was stoppedand
for three subsetsof these patients according to the
time elapsed between stopping treatment and retesting. The meanIgG level before treatmentwas stopped
(7.2 ? 1.2 pg/ml) was the sameas the averagelevel
observed during routine maintenance VIT.3, 4 The
440
Golden
J. ALLERGY
et al.
i-
!
c
1
24
12
TIME
SINCE
VENOM
THERAPY
4
1
I
36
48
60
STOPPED
hIWlkr)
Fig. 3 Geometric mean venom-specific IgG antibody levels
( ? SEM) in microgram per milliliter, listed for the whole
group before VIT was stopped, and for three subgroups
of patients after venom therapy had been stopped for a
mean of 24, 42, or 56 months.
Dashed
lines depict two
hypothetical decay curves for IgG.
mean IgG level at the time of retesting of patients not
receiving VIT ( 1.95 2: 0.3 Fg/ml) was typical of untreated sting-allergic patients.’ ” There was no difference in the mean IgG levels in the subgroups of
patients who had not been receiving VIT for a mean
of 24 months. 42 months, or 56 months, indicating
that the IgG levels had already dropped to a new
plateau by the time of retesting.
DISCUSSION
VII‘ is safe and effective for rapid and long-term
protection of susceptible individuals from recurrent
systemic allergic sting reactions. It has been recommended that therapy be continued for as long as the
venom sensitivity persists. It has become evident that
;I great many patients are stopping VIT with or without
the consent of their physicians. These patients give
many different rationales for their actions. Those who
have been stung while receiving treatment often believe that their clinical protection will continue after
they stop treatment. Many other patients believe that
their sensitivity has diminished or disappeared and.
as a result. that there would be no risk of a systemic
reaction to a sting even if they stop treatment. In order
to examine the validity of current recommended practice, we have begun a series of investigations designed
CLIN.
IMMIJNOL.
MARCH 1966
to determine whether, or when, VIT can be safely
stopped. This first, or pilot phase, involved a retrospective survey of patients who had stopped VIT
against our advice (dropouts).
Of the 82 patients who responded to the survey, 28
reported having been stung after a mean of 36 months
after therapy was stopped (the range was 12 to 72
months after treatment). Several points can be made
about the six patients who experienced systemic sting
reactions (Table III). Their reactions demonstrated no
tendency to be either more severe or less severe than
they had been before treatment. This suggests that
VII‘ causes no long lasting decrease or increase in the
reactivity of the individual. In contrast to children who
often have milder reactions on re-stingsI or treated
adults who have only minimal symptoms (if any).
these dropouts had a very appreciable risk of major
anaphylactic reactions. Because of the severity of the
reactions, we questioned these six patients as to why
they had not returned for reassessment and treatment.
Every one of them indicated that the cost and inconvenience of treatment were major deterrents and that
they preferred to take their chances with avoidance
techniques and adrenalin kits.
These systemic sting reactions occurred 12 to 45
months after therapy was stopped, suggesting that the
benefits of treatment are lost shortly after therapy is
stopped. Whether prolonged VIT might confer longer
lasting protection if the treatment is stopped cannot
be assessed from the available data. Among the six
reactors, the four who had received 2 to 5 months of
maintenance VIT had more severe reactions than the
two patients who had had 2 1 and 3 1 months of therapy.
However, only six of the 28 patients who had been
stung had received more than 2 years of maintenance
VIT. This study can therefore not evaluate the reactivity of patients after more prolonged VIT.
As illustrated in Table II, patients who stop VIT
have a significantly greater risk of systemic reactions
than patients who continue to receive treatment for
the same period of time. The only other estimate of
the systemic reaction rate in patients who stop VIT
has been by Dvorin et al.’ in the same article in which
they reported a reaction rate of 42% in untreated patients. Systemic reactions occurred in 42% of the dropouts (24% of the stings) that was not significantly
different from the reaction rate in their untreated group
or in our dropouts. Estimates of the systemic reaction
rate in untreated, but susceptible patients, have ranged
from 27% to 61 G/o.‘,‘-I’ That our estimated 22% risk
of systemic reaction in patients who stop VIT was
less than many of the estimates for untreated patients
is not at all surprising for reasons that have affected
many studies of this disease. The 22% figure is almost
VOLUME
NUMBER
77
3
Observations
certainly an underestimate as a result of the reduced
reactivity of many patients during a period of years
and the likelihood that some stings were by insects to
which the individuals were not allergic at the time.
Our own observation of a 6 1% risk in untreated patients is the highest of many such observations and,
although the observation is significant statistically,
comparison with the minimal estimate of 22% for
dropouts may exaggerate the difference between
these two groups. Furthermore, it is well recognized
that patients may react to some stings and not on other
occasions, which makes it uncertain whether those
who do not react to a challenge sting can be considered
safe in the future. The current data are not sufficient
in this area for any conclusions. We conclude that,
although most had no systemic reaction, the 22% reaction rate in patients who stop venom therapy was
not significantly different from most estimates of reactivity in untreated patients and represents a significantly increased risk of serious systemic reactions over
patients who continue treatment.
The skin test data in Table V provide some insight
into the natural history of the disease. Results are
presented only for venoms to which the patients were
sensitive and immunized. There was a significant decline in sensitivity in 63% of the patients. This decline
occurred in almost all honeybee-sensitive patients and
was more common with yellow hornet venom than
with white-faced hornet venom or YJV. These results
are consistent with the cross-reactivity of the vespid
venoms in which yellow hornet usually demonstrates
the least activity of the three and might therefore be
expected to be the first to decline or disappear. In fact,
the disappearance of skin test reactivity followed the
same tendency to be greater for yellow hornet than
for YJV or white-hornet venoms. The disappearance
of HBV, sensitivity was significantly more frequent
than for YJV (p = 0.005) or white-faced hornet
venom (p = 0.015). Honeybee sensitivity diminished and disappeared as commonly in patients with
isolated honeybee sensitivity as in those with combined HBV and vespid venom sensitivities, suggesting
that the rapid loss of honeybee sensitivity is not simply
caused by the low-level cross-reactivity that has been
recently described between HBV and YJV.16,” Most
importantly, these results demonstrate that many patients would be mistaken to believe that their allergic
sensitivity has disappeared while receiving VIT. We
should note that the observed decline in skin test sensitivity may be unrelated to VIT and has been observed in untreated patients such as those with large
local reactions. ‘* It is unlikely that these changes are
due to batch-to-batch differences in venom potency,
since no such differences have been observed in com-
in patients
who stop venom
immunotherapy
441
parisons of various lots by skin test or RAST inhibition.
The studies of venom-specific IgG antibody levels
were quite revealing. The patients who were available
for follow-up testing were quite typical of VIT patients
in that they had the same mean and range of antivenom
IgG levels before the cessation of VIT.3. 4 Once treatment was stopped, the IgG levels declined markedly
in virtually all patients. The mean IgG level at the
time of reassessment was quite typical of untreated
sting-allergic patients.3. I4 These results indicate that
any clinical benefit derived from the humoral immune
response to VIT is rapidly lost when treatment is
stopped. The rate of decline of IgG antibodies cannot
be determined from these data, since all of the study
patients were reevaluated more than 18 months after
stopping treatment, by which time the mean IgG antibody level had already fallen to the level of untreated
patients (Fig. 3). As suggested by the dashed lines in
Fig. 3, it is not possible to determine whether the fall
in IgG occurred rapidly within the first 6 months after
treatment stopped, as has been observed in the 6
months after an insect sting in untreated patients,“,’
or whether it declined more gradually between 0 and
24 months. In any case, it is clear that many patients
are mistaken if they believe that their immunity will
persist when they stop treatment.
The skin test and antibody results are also important
because they may have some bearing on the observed
systemic reaction rate. Among the 15 patients who
had been stung and were available for retesting, five
patients had skin tests that had decreased to negative
or borderline positive. None of these five patients had
systemic or large local reactions after VIT was
stopped. The systemic reaction rate in dropouts with
persistent sensitivity might therefore be much higher
than the observed 22% rate, since venom sensitivity
becomes borderline or negative in a large proportion.
Venom-specific IgG antibody levels are also important
in this regard. We have previously reported,“ and continue to observe, that patients receiving VIT in whom
the IgG level is >4 pg/ml have a very low risk of
systemic reaction to challenge stings (l%), whereas
patients with IgG levels <3 pg/ml have a significantly
higher risk of reaction of approximately 15%. Examination of Fig. 2 illustrates that despite the marked
decrease of IgG levels after cessation of VIT, eight of
the 31 patients still had IgG levels >4 pg/ml at the
time of reassessment and might therefore have some
residual clinical protection as a result of the persistence of blocking antibodies. In fact, Fig. 2 also illustrates that most of the patients who were not stung
had IgG levels <3 pg/ml. Once again, the systemic
reaction rate might have been distinctly higher if all
442
Golden
et al.
the subjects had been stung. It therefore appearsthat
the 22% reaction rate in the dropouts is a minimal
estimate cauSedby the loss of sensitivity and/or residuaI I@ antibody levels in at least 25% of these
patients.
The results of this retrospective survey lead to several conclusions. The risk of systemic sting reactions
in patients who stop VIT is significantly greater than
in those who fctllow the recommendedmaintenance
treatment schedule. Severe or life-threatening reactions may occur after therapy is stopped. Venom skin
test sensitivity declines in most patientsbut disappears
completely in only 5% of patients retestedafter 3 to
6 years. Venom-specific IgG antibody levels fall to
pretreatment 1eveIsin most patients who stop WT.
This pilot study indicates the need for further investigations to answer several remaining questions. The
patients in this survey had received maintenanceVIT
for an average of 14 months. Since we cannot extrapolate our conclusions to patients who have received prolonged VIT, it would be of interest to study
patients who stop treatment after many years. Prospective study of such patients would permit the description of the decay rate of venom-specific IgG antibodies when treatment is stoppedand would permit
the evaluation of systemicreaction rates in subgroups
of patients on the basis of skin test sensitivity and
venom-specificantibody levels at the time of the posttreatment sting. The current article does indicate that
we should continue to recommendthat VIT be continued for as long as the sensitivity persists.
1. Golden DBK, Lichtenstein LM: Insect sting allergy. In Smith
LH, editor: Cecil textbook of medicine, ed 16. Philadelphia,
1982, WB Saunders Co, p 1805
2. Dvorin DD, Georgitis JW, Reisman RE: Natural history of
insect sting anaphylaxis: evaluation of untreated and incompletely treated patients. J ALLERGY CLIN IMMUNOI. 73:1X8.
1984 (abst)
3. Golden DBK, Kagey-Sobotka A, Hamilton RG. et al: Human
immune response to Hymenoptera venoms. J ALLERGY CLIN
IMMIJNOL.71:140, 1983 (abst)
J. ALLERGYCLIN.
IMMUNOL.
MARCH1986
4. Golden DBK, Meyers DA, Valentine MD, et al: Clinical reevance of the venom-specific immunoglobulin G antibody level
during immunotherapy. J ALLERGY CLIN IMMUNOL 69:489,
1982
5. Graft DF, Schuberth KC, Kagey-Sobotka A, Kwiterovich KA,
Lichtenstein LM, Valentine MD: Immunologic effects of prolonged venom immunotherapy in children. J ALLERGY CLIN
IMMUNOL 71:140, 1983 (abst)
6. Graft DF, Schuberth KC, Kagey-Sobotka A, et al: The dcvelopment of negative skin tests in children treated with venom
immunotherapy. J ALLERGY CLIN IMMUNOI. 73:6I, 1984
7. Light WC, Reisman RE, Shimizu M, Abesman CE: Clinical
application of measurements of serum levels of bee venomspecific IgE and IgG. J ALLERGYCLIN I~~u~01.59:247, 1977
8. Norman PS: Skin testing. In Rose NR, Friedman H, editors:
Manual of clinical immunology, ed 2. Washington, D. C.,
1980. American Society of Microbiology. p 791
9. Hamilton RG, Adkinson NF Jr: Quantitation of antigen-specific
immunoglobulin G in human serum. II. Comparison of radioimmunoprecipitation
and solid-phase radioimmunoassay
techniques for measurement of immunoglobulin G specific for
a complex allergen mixture (yellow jacket venom). J ALLERGY
CI.IN IMMUNOL 67: 14, 198 I
10. Hunt KJ, Valentine MD, Sobotka AK, et al: A controlled trial
of immunotherapy in insect hypersensitivity. N Engl J Med
299:157, 1978
11. Settipane GA, Chafee FH: Natural history of allergy to Hymenoptera. Clin Allergy 9:385, 1979
12. Parker JL, Santrach PJ, Dahlbeg JE, et al: Evaluation of
Hymenoptera-sting sensitivity with deliberate sting challenges:
inadequacy of present diagnostic methods. J ALLERGY CLIN
IMMUNOI.69:2~, 1982
13. Blaauw PJ, van den Brock MJTB, Smithuis LOMJ: The insect-sting challenge as a ultimate indication for venom immunotherapy. Folia Ailergol Immunol Clinica 30: 141, 1983
14. Golden DBK, Langlois J, Valentine MD, et al: Treatment faiiures with whole body extract therapy of insect sting allergy.
JAMA 246:2460, 1981
15. Schuberth KC, Graft DF, Kagey-Sobotka A, et al: Do all
children with insect allergy need venom therapy? J ALLERGY
CLIN IMMUNOL 71:140, 1983 (abst)
16. Hoffman DR, Wood CL: Allergens in Hymenoptera venom
XI. Isolation of protein allergens from Vespulu maculijrons
(yellowjacket) venom. J ALLERGYCLINIMMUNOL 74:93, 1984
17. Reisman RE, Muller UR, Wypych JE, et al: Studies of coexisting honeybee and vespid-venom sensitivities. J ALLERGY
CLN IMMUNOL 73:246, 1984
18. Golden DBK, Johnson K, Addison BJ, et al: Evolution of
Hymenoptcra venom allergy (HVA). J ALLERGYCLIN IMMUNOI.
73:189. 1984 (abst)