We considered diphtheria, because of its characteristic short incubation time, to be a typical primary toxic disease. This is also confirmed by the curative effect of its immune serum. However, the disease has a manifestation in which we recognize an allergic nature: diphtheria paralysis. With a usual obligatory incubation period of about 8 days, and always of more than 6 days, diphtheria paralysis is a typical allergic condition. Once present, it resists diphtheria immune serum, yet it can be efficiently prevented by the same serum. In animals such as guinea pigs and hamsters, such paralysis can be induced and its allergic character clearly recognized. The heating of toxin at 56°C reduces its direct toxicity without impairing its antigenic properties. Administration of even huge amounts of heated toxin does not produce any immediate toxic effect. Yet the heated toxin, even though it has lost its toxic effect, induces paralysis.

Classically, this paralysis has been related to a hypothetical thermo stabile fraction of the toxin, with a high incubation time. This view does not agree with the results of our experiments. When different amounts of the same heated toxin are injected in guinea pigs of the same sex, age and weight, the incubation time for paralysis, although always above 6 days, changes, becoming paradoxically longer when higher doses are used. With great amounts of the heated toxin, corresponding to 20,000 lethal doses of the nonheated toxin, incubation time becomes as long as 14-17 days in contrast to 8-9 days for relatively small amounts. Table XIII shows this relationship.

Table XIII. Changes In The Incubation Time Of Paralysis Induced By Different Amounts Of Heated Diphtheria Toxin

A mount used

Incubation timeó average of 4 animals

.5 LD

8.33 days *

1 LD


5 LD


20 LD


100 LD


1000 LD


5000 LD


20000 LD


* One in four animals did not show paralysis.

If the paralysis were induced by direct action of a thermostabile fraction of the toxin, then higher doses of the fraction should reduce, or at least, not increase the incubation time. This paradoxical fact can be explained simply through the mechanism of allergic pathogenesis. It is a fact common to immunological reactions that an organism has greater difficulty in manufacturing any antibody when very large amounts of antigen are present than when smaller amounts are involved. This difficulty is translated into a longer time necessary for the appearance of the antibodies. As seen in our experiments, in the case of an allergic reaction, this difficulty in the manufacture of coagulant antibodies would result in a longer incubation time.

The localization of the allergic manifestation as paralysis can be explained in part through the affinity of toxin as antigen for nerves and in part through the participation of the nerves in the allergic reaction. The levels at which the diphtheria toxin acts seem to be tissular, organic and systemic, with preference for the adrenals, inducing characteristic supra renalitis. When coagulant antibodies appear, no manifest systemic allergic reaction will occur with the antigen still present in the blood. The allergy will be manifest, however, at the lower tissue level and especially in the nearby nerves. Antigen must be present in the nerve at the moment of appearance of the coagulant antibodies if paralysis is to occur. This can be demonstrated by using sensitizing and triggering injections of toxin in animals.

We sensitized guinea pigs to heated and unheated toxin by injecting relatively small amounts intravenously. On the sixth or seventh day, another small quantity of the same toxin was injected, this time near the sciatic nerve. The total amount of toxin was far below the lethal dose. Two or three days later, paralysis developed in the injected limb in a high proportion of animals while no such paralysis could be observed in animals injected only intravenously or with the same total amount of toxin at once in the limb. In other experiments, the daily injection of small amounts of toxin, whether heated or nonheated, near the sciatic nerve, induced paralysis although the total quantity of toxin was much lower than that which ordinarily would induce paralysis in any similar animal. Paralysis appeared in these cases after an incubation period of about 12 days. The animals sensitized by one or more injections of heated toxin responded to the non heated as triggering injections and vice versa, indicating that antigenic properties were responsible for the paralytic allergic manifestation.

In humans, anti diphtheria serum, effective against toxic manifestations, had no effect upon paralysis once it had appeared but is very effective in preventing it. The same was true in animal experiments. Injected 24 hours before the appearance of paralysis, the serum had a consistent preventive action. This fact confirms again the allergic pathogenesis of the paralysis.

In another experiment, we showed that administration of cortisone, with its anti allergic action, also reduces the incidence of paralysis without having the same effect upon the direct toxic action. It can be noted, too, that among small laboratory animals, diphtheria paralysis can be induced readily in guinea pigs, less readily in hamsters, and not at all in adult rats and mice. In addition to the sensitivity of guinea pigs to diphtheria toxin, this can be related to the great capacity of these animals in general to produce allergic antibodies and thus to be subject to anaphylactic reaction. Based on these considerations, we can classify diphtheria paralysis as a typical localized allergic reaction.