Published April 1, 1926

ANAPHYLACTIC SHOCK CAUSED BY ANTIBODY IN ANI-

MALS SENSITIZED BY A N T I G E N - - R E V E R S E D
PASSIVE ANAPHYLAXIS.
BYEUGENE L. OPIE, M.D., A~D J. FURTH,M.D.
(From the Henry Phipps Institute of the University of Pennsylvania, Philadelphia.)

(Received for publication, January 15, 1926.)

Manifestations of hypersusceptibility which have received most

Downloaded from jem.rupress.org on April 29, 2010

attention are anaphylactic shock and the local inflammatory reaction
produced by injection of antigen into a sensitized animal (Arthus
phenomenon). Since these phenomena may be reproduced in a
normal animal by injecting into it blood serum obtained from one
which has been actively sensitized, it may be assumed that this sus-
ceptibility to the action of an otherwise harmless substance is caused
by an antibody which finds its way into the blood serum of the sen-
sitized animal.
One of us has found that the usual procedure employed to produce
the Arthus phenomenon by means of passive sensitization may be
reversed) An animal previously treated with horse serum reacts
with acute inflammation when serum of a rabbit immunized against
horse serum is injected into its dermis. Inflammatory edema is
caused by antibody injected into the skin of an animal sensitized by
the corresponding antigen. This observation indicates that acute
inflammation will occur whenever antigen and antibody come into
contact within the tissues.
The experiments which will be described have been undertaken with
the purpose of determining if the meeting of antigen and antibody is
sufficient to produce anaphylacfic shock irrespective of the order of
their introduction into the body. It has seemed probable that further
knowledge concerning the relationship of antigen to antibody might
explain the pathogenesis of both local and general hypersensitiveness.

10pie, E. L., J. Immunol., 1924, ix, 255.

469
Published April 1, 1926

470 R E V E R S E D PASSIVE ANAPIR'YLAXIS

Experiments on Guinea Pigs.

Since guinea pigs have been most frequently used for the study of
anaphylactic shock because of their peculiar susceptibility to it, ex-
periments have been undertaken to determine if this aflimal is sus-
ceptible to reversed passive anaphylaxis as well. I)oerr and Russ 2
were unsuccessful i~ efforts to produce anaphylactic shock by injec-
tion of anti-eel serum into guinea pigs which had received eel serum a
short time before. Similar experiments were performed with anti-
goat serum injected into guinea pigs which had received serum of goat.
In our own experiments anti-horse serum of guinea pig in quantity
only slightly less than that which was fatal for normal animals caused

Downloaded from jem.rupress.org on April 29, 2010

when injected into animals previously treated with horse serum no
symptoms of shock. It is noteworthy that the guinea pig which is
highly sensitive to anaphylactic shock forms antibodies much less
actively than the rabbit. The precipifin titre of the serum used in
this experiment was very low. It is possible that a stronger serum
might cause some reaction.
Since passive anaphylaxis is produced in guinea pigs more readily
with serum of sensitized rabbits than with serum of sensitized guinea
pigs, experiments were performed to determine if reversal of the usual
procedure causes anaphylactic shock when serum of rabbits is used.
These experiments have shown that the serum of immunized rabbits
is much more toxic for normal guinea pigs than the serum of normal
rabbits. Furthermore anti-horse, anti-beef, and anti-egg white serum
of rabbit have been found no more toxic for guinea pigs which have
previously received the corresponding antigen than for untreated
guinea pigs.
Immune sera prepared by use of one mammalian serum such as
horse serum or beef serum precipitates weakly sera derived from other
more or less distantly related species and the possibility suggests itself
by analogy that immune serum may react with protein of the guinea
pig and cause anaphylactic shock. Nevertheless it is noteworthy
that serum prepared by use of an antigen of avian origin, namely egg
white, has been just as toxic as anti-horse or anti-beef serum. This
serum has caused very slight precipitation when mixed with serum of
Doerr, R., and Russ, V. K., Z. ImmuniRitsforsch., Orig., 1909, iii, 706.
Published April 1, 1926

EUGENE L. OPIE A N D ~. FUl~TH 471

guinea pig. A second possibility m a y be considered. In the experi-

ments on guinea pigs the immune serum which has been used has been
collected from several an{reals repeatedly injected with antigen at
intervals of 5 days. In some of these animals antigen may have
persisted in the blood stream so that the injected serum has contained
both antigen and antibody. Subsequent experiments have shown
that mixtures of antigen and antibody may cause shock in rabbits.

Experiments on Rabbits.
Rabbits furnish more favorable conditions for the production of
reversed passive anaphylaxis than guinea pigs. Antibody formation
appears to be more active in the rabbit than in the guinea pig, for

Downloaded from jem.rupress.org on April 29, 2010

passive sensitization of guinea pigs is more readily produced with
serum obtained from actively sensitized rabbits than with that obtain-
able from guinea pigs. Furthermore precipitin formation is much
more active in rabbits than in guinea pigs. When the rabbit is used
and rabbit serum with high antibody content is injected into the cir-
culating blood, the experiment is not complicated b y the toxicity of a
foreign serum. Reversed passive anaphylaxis occurs in rabbits when
a sufficient quantity of strong antiserum is introduced into the vascu-
lar system of animals previously treated with the corresponding an-
tigen. Young rabbits have been used in the experiments because
it has been found impracticable to obtain antiserum in quantity suffi-
d e n t to inject several fully grown an{mals.
Rabbits received 5 cc. of horse serum injected into the peritoneal cavity and
on the next day anti-horse serum of rabbit obtained from four animals was in-
jected into the ear vein. The injection in all experiments has been timed so
that the serum entered the vein at the rate of 10 cc. per minute.

RabbitNo. Weight. Horseserum. Interval. Anti-horse

serulll.
Result.
ht$o CG.

420 5 20 5 Shock.
510 5 20 7.5
580 5 20 9 Death.
500 5 20 10
420 10 No symptoms.

~n{mals which have received 5 cc. of horse serum and 20 hours later
Published April 1, 1926

472 R E V E R S E D PASSIVE A N A P H Y L A X I S

from 5 to 10 cc. of anti-horse serum have exhibited symptoms of shock

or have died immediately following the second injection, whereas a
control animal which has received no preliminary injection of horse
serum has shown no symptoms when injected with anti-horse serum.
In the animals in which shock has occurred there have been pas-
sage of urine and feces and weakness following the injection so that
the animal has rested upon the abdomen with the legs spread out and
flaccid; respiration has been slow and often labored. Recovery has
occurred after 10 or 15 minutes. In the animals which have died there
have been convulsive extension of the legs, passage of urine and feces,
dyspnea with slow respiration, and death within 4 or 5 minutes. The
liver has been found to be engorged with blood and the veins of the

Downloaded from jem.rupress.org on April 29, 2010

intestine have been congested. When "slight shock" is recorded
there has been transient weakness of the extremities; with "severe
shock," convulsions or prolonged prostration with almost complete
loss of reflexes has been followed by recovery.
In a second experiment a smaller quantity of horse serum was in-
jected into the peritoneal cavity.

Rabbit No. Weight. Horse serum. Aa•horse serum. Result.

gm. ¢.6.

6 300 1 5 No s y m p t o m s .
7 320 1 10 Shock.
8 320 1 15 Death.
9 300 10 No s y m p t o m s .
10 320 15 Slight shock.

Preliminary injection of horse serum has sensitized rabbits to the

action of anti-horse serum.
Full grown rabbits have been found to be susceptible to the changes
observed in young animals. Anti-horse serum has been injected into
the ear vein of rabbits which have previously received 5 cc. of horse
serum intravenously.

Rabbit No. Weight. Horse serum, Anti-horse serum. Result.

C6.

11 2170 5 10 Slight shock.

12 2400 5 20 No s y m p t o m s .
13 2400 5 30 Death.
Published April 1, 1926

EUGENE L. OPIE AND ~. ~URTH 473

In Rabbit 13 death occurred preceded by convulsions, passage of

urine and feces, and dyspnea. The liver was engorged and the ves-
sels of the intestine were injected. There was delayed coagulability
of the blood. Other experiments have shown that the quantity of
antiserum used in this experiment has no injurious effect.

Sensitizing Dose of Antigen.

T h e a m o u n t of a n t i g e n u s e d a s a p r e l i m i n a r y i n j e c t i o n h a s b e e n
v a r i e d t o d e t e r m i n e ff s e n s i t i z a t i o n t o a n t i s e r u m s h o w s a n y c o r r e s p o n d -
ing v a r i a t i o n . F i v e t y p i c a l experiments w i t h five s a m p l e s of a n t i -
serum are as follows:

Downloaded from jem.rupress.org on April 29, 2010

Anti-horse
RabbitNo. Weight. Hor~ ~ . Interval. serum Result.
of rsbbit.
gm. g.$. ]~'$.

14 300 5 20 8 Shock.
15 300 0.2 20 8 Severe shock.
16 290 8 N o symptoms.

17 220 0.5 13 9 Severn shock.

18 0.05 13 8 Slight "
19 0.01 13 9
20 8 N o symptoms.
21 10

22 2 20 15 Severe shock.
23 0.1 20 20 Shock.
24 0.02 20 15 No symptoms.
25 0.002 20 15

26 2 20 9 Shock.
27 0.5 20 12 Slight shock.
28 0.2 20 9 Shock.
29 13 Slight shock (?).
30 15 N o symptoms.

31 0.5 20 9 Death.
~c
32 0.05 20 8
33 0.005 20 8 N o symptoms.

T h e s e e x p e r i m e n t s s h o w t h a t v a r i a t i o n of t h e a m o u n t of a n t i g e n
Published April 1, 1926

474 R E V E R S E D PASSIVE A N A P H Y L A X I S

between 2 and 0.05 cc. has no constant effect upon the changes caused
by a subsequent injection of antiserum but smaller quantities of anti-
gen have been followed by less severe symptoms and 0.005 cc. has
been found insufficient to produce any sensitization to antiserum.

Toxic Dose of Antibody.

In Rabbit 32 death was caused by 8 cc. of antiserum following sen-
sitizafion by 0.05 cc. of horse serum; this is the smallest quantity of
antiserum which has produced m a ~ m u m intoxication. Sensitization
has been demonstrated by injection of anti-horse serum in volume 180
times that of antigen. This figure accords approximately with the

Downloaded from jem.rupress.org on April 29, 2010

quantitative relation of antigen to antiserum in the precipitin reaction.
The amount of antiserum necessary to produce death in animals
sensitized by antigen has varied from 8 to 20 cc. and with the same
serum may vary with the weight of the animal.

Interval between Injection of Antigen and Antibody.

When anaphylaxis is produced by the usual method an interval
must elapse between the sensitizing injection of antiserum and the
toxic injection of antigen. It is assumed that this period of incuba-
tion is required to permit penetration of antibody in sufficient con-
centration into cells upon which antigen acts when it produces symp-
toms of shock. Experiments were made to determine if passive sen-
sitization is preceded by a similar interval when antigen is followed
by antiserum; that is, when their usual order of introduction is
reversed.
Four experiments are cited to show the effect of varying the interval
between the sensitizing injection of antigen and the subsequent in-
jection of antiserum. The second injection is said to be "immediate"
when the first injection is made into the marginal ear vein on one side
and is immediately followed by injection of antiserum into the same
vein of the other ear, the interval between the two injections being
less than 30 seconds.
Published April 1, 1926

I¢.UGEI~E L. 0PI~ A N D J . ?LrRTH 475

Anti-horse
tabbit No. i Weight. Horse Interval. serum of Result.
I
serum, rabbit.

gra. ¢c. hrs. co.

34 400 1 20 15 Death.
35 390 1 20 10 "

36 400 1 12 12 Severe shock.

37 400 1 6 12 Death.
38 550 I Immediate. 17 No symptoms.
39 420 ~ -- 15 " "

22 220 O.5 13 9 Severe shock.

40 200 O.5 Immediate. 9 Slight "
25 200 -- -- 8 No symptoms.
26 190 ~ -- 10 Transient weakness.

Downloaded from jem.rupress.org on April 29, 2010

41 340 0. S 20 10 Death.
42 390 O.5 20 15 "
43 360 0.5 2 11 No shock,
44 380 0.5 Immediate. 15 Death.
45 410 0.5 " 12 Transient weakness.
46 340 15 " "
~abbit No. Weight. Beef Interval. Anti-beef Result.
5erln~. sertlin.

gin. c~. hrs. 66.

47 720 1 20 20 Death.
48 770 1 6 20 Transient weakness.
49 770 1 Immediate. 20 " "

E x p e r i m e n t s h a v e been p e r f o r m e d on small groups the size of which

has been determined b y the a m o u n t of available antiserum. I n each
group there h a v e been f r o m one to three animals in which the interval
between the sensitizing injection of antigen a n d the subsequent toxic
dose of a n t i s e r u m has been from 12 to 20 hours a n d in each group there
h a v e been one or m o r e controls in which injection of antiserum has
not been preceded b y injection of antigen. F u r t h e r m o r e , these
groups included animals in which the period between the sensitizing
a n d toxic injection has varied f r o m less t h a n 30 seconds ( " i m m e d i a t e " )
to 6 hours. T h e following table includes all of the animals which
h a v e received 0.1 cc. or m o r e of antigen and a n t i s e r u m in sufficient
q u a n t i t y to produce definite s y m p t o m s when the interval between in-
jection of antigen and a n t i b o d y has been from 12 to 20 hours.
Published April 1, 1926

476 REVERSED P A S S I V E ANAPHYLAXIS

Number of Number N'umber Number

animals with with Number
which severe moderate with slight
shock.
with no
died. shock. shock. symptoms

Controls (no injection of antigen). 14

Interval between injection of antigen

and of antiserum.
Immediate... 1 1 3
2 hrs. 1
1 3
6 " 1
12 " 2
24 " 15 6

Downloaded from jem.rupress.org on April 29, 2010

When control animals have been injected with antiserum alone, in
most instances there have been no symptoms but in a few instances
shock has occurred. It Mll be shown below that this shock is prob-
ably referable to the presence of antigen in the antiserum used for
injection.
When injection of horse serum into one ear vein is followed imme-
diately by injection of anti-horse serum into a vein of the opposite
ear mild shock occurs in most instances but is much less frequent than
in those instances in which an interval of 4 or more hours elapses
between the two injections. In animals in which the interval between
injection of antigen and of antiserum has been from 2 to 6 hours shock
has occurred constantly but has been much less severe than in animals
in which the interval has been from 12 to 20 hours. The experiments
show that the intensity of sensitization increases gradually from the
time of injection and reaches a maximum after 4 hours. After 12
hours there is no further increase.
In the greater part of nineteen experiments anti-horse or anti-beef
serum injected into the ear vein of untreated rabbits has caused no
symptoms but in five instances there has been transient symptom of
shock and in an occasional instance definite shock or death. Ex-
periments of Friedemann, 8 Briot,* and others, have shown that mix-
tures of antigen and antiserum injected into the ear vein of rabbits
may cause anaphylactic shock. Under the conditions of the fore-
8 Friedemann, U., Z. Immunitiitsforsch., Orig., 1909, ii, 591.
* Briot, A., Compt. rend. Soc. biol., 1910, lxviii, 402.
Published April 1, 1926

EUGENE L. OPIE A N D ~. J~URTH 477

going experiments it is not improbable t h a t the injected antiserum in

some instances contained antigen. I n the early stages of immuniza-
tion against horse or beef serum, the antigen makes its appearance in
the serum of the blood but with continued immunization it fails to
enter the blood, s Nevertheless in a few animals in which precipitin
formation is scant antigen m a y persist in the blood serum. I n the
experiments which have been described, serum from three or four or
more animals has been mixed in order to obtain a sufficient q u a n t i t y
of serum. Antigen has doubtless been present in some of the mixtures
of serum injected into normal animals.
Rabbit 50, which had received no preliminary injection of beef serum, received
in the ear vein 20 cc. of anti-beef serum obtained from four rabbits. Following

Downloaded from jem.rupress.org on April 29, 2010

slight restlessness, urine and feces were passed. The animal became very weak
and lay with legs outstretched. The respiration was slow and labored. Recovery
occurred after about 20 minutes. The mixture of serums used for injection of
this animal formed on standing a flocculent sediment. This precipitate was
removed by centrifugalization and 20 cc. of the serum were injected into the
ear vein of a rabbit. Slight transient weakness of the extremities was noted.
Part of the serum used for this injection became slightly turbid on standing.
Rabbits 67 and 68 received in the ear 12 cc. of anti-horse serum obtained from
four rabbits. In Rabbit 67 the injection was followed by weakness of the extremi-
ties so that the animal lay with extremities outstretched and head resting on the
table. Recovery occurred within 5 minutes. In Rabbit 68 there were transient
weakness of extremities and slow respiration. Three of the specimens of serum
used contained no antigen whereas the fourth contained antigen in abundance so
that the mixture of the four formed an abundant precipitate. Vv'hen this pre-
cipitate was removed by centrifugalization the clear serum caused no symptoms
when 12 cc. were injected into the ear vein of each of two rabbits.
In these experiments serum containing b o t h antigen and a n t i b o d y
has caused shock b u t when all of the antigen has been removed there
has been no evidence of shock following injection of the serum.
T h e following experiment in which a strong anti-egg serum has been
used suggests t h a t the interval between injection of egg white and
m a x i m u m sensitization to antl-egg serum m a y be shorter than the
interval in animals tested with horse or beef serum and the correspond-
ing antiserum. T h e serum employed in this experiment was obtained
frown four rabbits, had a precipitin titre of one million, and contained
no antigen.
s Opie, E. L., .1. Immunol., 1923, viii, 55.
Published April 1, 1926

478 I~EVERSED PASSIVE ANAPHYLAXIS

Rabbit No. Anti-egg

Weight. Egg white. Interval. white serum. Result.
gm. Ills.
51 370 0.5 24 25 No symptoms.
52 300 0.45* 24 20
53 360 0.45* 7 20 Slight shock.
54 350 0.5* 4 20
55 310 0.5 Immediate. 20 No symptoms.
~c ~g
56 310 25

* These animals had received on the previous day 0.25 cc. of egg white, diluted
with an equal volume of salt solution.

In the experiment maximum sensitization has been present at 4

and 7 hours after injection of egg white and no evidence of sensitiza-

Downloaded from jem.rupress.org on April 29, 2010

tion has been found imrrtediately after injection and after an interval
of 24 hours.
Desensitization.
The parallel between reversed passive anaphylaxis and the usual
procedure for the production of passive anaphylaxis is emphasized by
the occurrence of desensitization. Animals which have been sensitized
to the action of anti-horse serum of rabbits by an .injection of horse
serum may be desensitized by repeated injection of anti-horse serum
in quantity insufficient to produce symptoms. Young rabbits have
been given the smallest quantity of antigen required to produce effec-
tive sensitization to antiserum. On the day following the injection
of antigen antiserum in quantity much below that which produces
symptoms of shock has been given into the ear vein; from 30 minutes
to 1 hour later, approximately hal/the usual dose required to produce
severe shock or death has been administered. Several hours later
the full toxic dose has been given. For comparison a sensitized rabbit
which has received no preliminary injections of anti-horse serum has
been treated with the same toxic dose.

Result.

I"'J mln. co. krs.

57 260 0.05 20 2 30 17.51 s 15 No symptoms.
5s 26010.051 20 I -- so!- I 3 15 Shock.
Published April 1, 1926

EUGENE L. OPIE A N D J. F U R T H 479

In this experiment 15 cc. of anti-horse serum admluistered to rabbits

sensitized by horse serum have caused shock with passage of urine and
feces, very slow respiration, profound weakness, and partial loss of
reflexes whereas intravenous injection of antiserum in quantity in-
sufficient to produce symptoms has caused complete desensitization
so that 15 cc. of anti-horse serum have caused no symptoms.
To exclude the possibility that serum alone might cause desensitiza-
tion in subsequent experiments an equal quantity of normal serum
has been injected whenever anti-horse serum has been given with the
purpose of causing desensitization.

i,

J]

Downloaded from jem.rupress.org on April 29, 2010

Result.
u

gm. cc. hrs. co.

59 230 0.1 2O 2 45 7.5 4 15 No symptoms.

60 200 0.1 20 45 4 15 Slightshock.
I hrs.
61 350 0.1 20 2 1 7 3½ 13 No symptoms.
62 360 0.1 20 1 3½ 13 Severe shock.

63 1320 0.1 20 2 6 3½ 14 No symptoms.

64 320 0.1 20 3½ t4 Shock.

65 230 0.1 20 2 8 3 20 No symptoms.

66 230 0.1 20 3 2O Severe shock.

Under the conditions which have been described desensitization has

occurred in all experiments.

DISCUSSION.

The experiments which have been described show that anaphylacfic

shock may occur when the usual procedure employed for passive sen-
sitization is reversed. In rabbits which have received antigen the cor-
responding antiserum in sufficient quantity causes anaphylactic shock
and death. It has been shown that the reaction occurs under condi-
tions which reproduce those of passive anaphylaxls. An interval of
approximately 4 hours must intervene between injection of antigen
and injection of antiserum in order to produce maximum shock; ani-
Published April 1, 1926

480 :REVERSED PASSIVE ANAPHYLAXIS

mals sensitized by horse serum may be desensitized to the action of

anti-horse serum by repeated injection of anti-horse serum in quantity
insufficient to cause symptoms. Anaphylactic shock like the specific
inflammatory reaction of the immunized animal or Arthus phenome-
non occurs when antigen and antibody meet and in either instance the
usual order of their introduction may be reversed. In the one instance
they meet within the tissue spaces whereas in the other contact occurs
by way of the circulating blood.
The occurrence of an interval between the injection of a sensitizing
substance (antibody or antigen) and the appearance of maximum
sensitization furnishes evidence in favor of the view that the phe-
nomena of anaphylaxis are referable to changes which occur within

Downloaded from jem.rupress.org on April 29, 2010

the cells. It has been assumed that with passive sensitization, anti-
bodies penetrate from the blood stream into the cells of the body so
that after an interval of approximately 4 hours they have reached
maximum concentration within the cytoplasm. The tissue is thus
prepared for action of antigen. The experiments described in this
paper have shown that tissues may be prepared just as readily by
antigen and then stimulated by antibody. These experiments and
similar observations upon the Arthus phenomenon indicate that the
phenomena of general and local anaphylaxis occur whenever antigen
and antibody meet in sufficient concentration within the tissues, the
resulting changes being dependent upon the peculiar functions of the
affected cells and their susceptibility to stimulation or injury by an-
tigen and antibody.
There is no direct evidence to show how antigen and antibody,
meeting within smooth muscle fibres, bring about contraction as with
anaphylactic shock, or meeting perhaps within the endothelial cells
of vessel walls increase their permeability for fluid and cells as in the
Arthus phenomenon. When antigen and the corresponding anti-
serum are brought together a precipitate is formed but with existing
knowledge it is not possible to determine the relation of precipitin
to the antibody concerned in the production of anaphylactic shock or
of the specific inflammatoryreaction known as the Arthus phenomenon.
Nevertheless there is no longer any reason for doubting that precipita-
tion occurs within the body as well as in vitro whenever antigen and
the corresponding precipitin meet; the reaction is identical when dilu-
Published April 1, 1926

EUGENE L. 0P1-I~. A N D ~. ~FIYRTH 481

tions are made with blood serum or with salt solution. The close
relation which exists between susceptibility to anaphylactic shock and
precipitin has been pointed out by several observers and susceptibility
to the Arthus phenomenon has been found to bear a close if not exact
relation to the precipitin content of the serum.
The investigations which have been cited show that local anaphy-
laxis (susceptibility to specific inflammation or Arthus phenomenon)
occurs under conditions identical with those which induce general
anaphylaxis (susceptibility to anaphylactic shock) save that in the
first instance antigen and antibody are brought together within the
tissue spaces outside of blood vessels and cause the usual phenomena
of inflammation whereas in the latter instance one or other of the two

Downloaded from jem.rupress.org on April 29, 2010

agents is introduced by way of the circulating blood and has the op-
portunity of coming into contact with those tissues which after prep-
aration by the other agent are most susceptible to the two in combina-
tion. Should precipitin be the antibody concerned in the production
of local or general anaphylactic reactions it is essential that precipitin
and precipitinogen meet and react to form precipitate within the tissue
for the introduction of the precipitate formed by their union causes
neither local nor general anaphylaxis. There is n o reason to doubt
that precipitin and precipitinogen meeting within the cytoplasm of
smooth muscle fibre or of endothelial cell would form precipitate and
it is not improbable that the presence of this precipitate within the cell
would produce disturbances such as muscular contraction in one in-
stance and increased permeability in the other.
It is unnecessary to assume the sudden formation of a toxic sub-
stance or anaphylatoxin with the power to elicit the symptoms of
anaphylactic shock. Toxic fluids which reproduce these symptoms
have been formed in vitro by the prolonged action of normal serum
(containing complement) upon a combination of antigen and antibody,
for example upon sensitized red blood corpuscles or upon precipitate
formed by precipitinogen and precipitin. The formation of toxic sub-
stances by the action of various substances such as kaolin, peptone,
agar, etc., upon blood serum does not explain the changes of local or
general anaphylaxis for the phenomena of anaphylaxis are caused by
the meeting of antigen and antibody.
Published April 1, 1926

482 IZEVERSED PASSIVE ANAPHYLAXIS

CONCLUSIONS.

Anaphylactlc shock occurs (in rabbits) when the usual procedure

for the production of passive anaphylaxis is reversed; that is, when an
animal previously treated with antigen receives the corresponding
antiserum by way of the circulating blood.
This susceptibility to the action of anti-horse serum produced by
injection of antigen reaches maximum intensity after an interval of 4
hours presumably required to permit penetration of the antigen in
sufl]clent concentration into the tissues.
Desensitization to the action of a shock-producing dose of anti-
horse serum can be brought about by repeated small doses of the same

Downloaded from jem.rupress.org on April 29, 2010

antiserum.
Anaphylactic shock and local anaphylaxis manifested by the acute
inflammation of an immunized animal when injected with the antigen
used for immunization (Arthus phenomenon) occur under analogous
conditions; that is, when antigen and antibody meet within the tissues.
The peculiar characters of these reactions are dependent upon the
site of entry of the irritating agent, which is the vascular system in one
instance and tissue spaces in the other, and upon the concentration
of antigen and antibody within susceptible tissues.
Meeting of antigen and antibody within susceptible tissues is surlY-
dent to explain the phenomena of local and general anaphylaxis so
that it is unnecessary to assume the sudden formation of a toxic
substance (anaphylatoxin).