CLASSIC IMMUNOTOXINS WITH PLANT OR MICROBIAL TOXINS - Fusion Protein Technologies for Biopharmaceuticals: Applications and Challenges - page 276

remain unclear. The pathogenesis of endothelial cell (EC)
damage is complex and includes activation or damage to
ECs and leukocytes, release of cytokines and of inflamma-
tory mediators, and change in cell-cell and cell-matrix
adhesion. VLS is a syndrome observed with IL-2, monoclo-
nal antibodies, some chemotherapy drugs such as taxotere,
and many immunotoxins. Higher incidence of VLS is found
in immunotoxin trials than in other protein-based therapy
trials. This may be related to the presence of potential VLS
motifs in classic immunotoxins. VLS motifs, (x)D(y) where
x

and AUC (area under the curve) for immunotoxins. Further,
treatment of immunotoxins boosts antitoxin antibody pro-
duction in most patients. Interestingly, in seven out of eight
patients treated with A-dmDT390-bisFv(UCHT1), antidrug
antibody titers declined below an acceptable level for
retreatment within a year (unpublished data). In this trial,
a patient with an antidrug antibody titer of 61.1 m g/mL has a
durable PR lasting 22 months. These results are encour-

aging because patients can be retreated yearly with DT
immunotoxins. In addition, 67.6% of children and adoles-
cents have anti-DT antibody titers

þ

V, L, or S, are commonly found in
toxins, ribosome-inactivating proteins, and IL-2 [202]. VLS
restricts the doses of classic immunotoxins and necessitates
the cessation of therapy in severe cases. Owing to VLS,
congestive heart failure (CHF) is a contraindication to the
treatment with classic immunotoxins. A total of 65% of CHF
patients have clinically unrecognized fluid overload [203].
Under clinical setting, fluid infusion for adequate hydration
and immunotoxin treatment could induce fluid overload and
VLS, aggravating CHF. It is also difficult to distinguish
between VLS and CHF because symptoms are overlapping.
One way to reduce the VLS incidence is to use prophylactic
steroids. Premedication of steroids improved tolerability of
DAB 386 IL2 without compromising the clinical response
[166]. In one trial, the VLS incidence was significantly
reduced from 27% to 13% as compared to the Phase III
studies [166]. In addition to steroid, theophylline, terbuta-
line, thalidomide, and immunoglobulins are used to treat
systemic capillary leak syndrome (SCLS) [204-207]. Pro-
phylactic use of theophylline, terbutaline, and thalidomide
may enhance tolerability of immunotoxins. Since interac-
tions of the toxin moiety with endothelial cells and/or cell
matrix appear to be a major cause of VLS, integrin-binding
motif peptides of fibronectin and vitronectin as antagonists
may also reduce the VLS incidence. Clinical experience
with VLS can also minimize the risk of the VLS complica-
tions due to early intervention.

¼ L, I, G, or V and y ¼ 2 m g/mL, which did not

block in vivo efficacy of DT 388 -GMCSF [208].
There are three ways to overcome the blocking effect of
the preexisting antitoxin antibodies in human sera. One is to
modify the toxin structure for deimmunization, another is to
neutralize the antibodies with nontoxic toxin mutants, and
the other is to use two or three different immuntoxins with
different toxin moieties but the same ligand. Many efforts
have been made for humanization or deimmunization of the
toxins. In particular, PE and DT have been modified to
minimize immune responses. The deimmunized DT by
extensive mutagenesis on T-cell epitopes of DT [209] and
the deimmunized PE by eight point mutations on B-cell
epitopes of PE38 [51] will be tested in clinical trials in near
future. Time to patient retreatment with deimmunized DT
and PE immunotoxins may be shorter as compared to
immunogenic DT and PE immunotoxins. The other method
is to neutralize existing antitoxin antibodies by infusion of
nontoxic toxin mutants such as CRM197 (DT mutant). This
hypothesis has been tested in rhesus monkeys using FN18-
CRM9 chemical conjugate (a monkey equivalent of anti-
CD3 DT immunotoxin) [210]. 100-fold excess CRM197
(20mg/kg) before FN18-CRM9 (0.2mg/kg) is infused in
two rhesus monkeys with preexisting anti-DT antibodies. In
one monkey, this treatment produces a similar T-cell deple-
tion effect of FN18-CRM9 conjugate as observed in rhesus
monkeys with no anti-DT antibodies. But the other monkey
died because of multiple kidney infarcts. This could be
related to either formation of immune complex precipitates
or weak toxicity of CRM197 or both. Although CRM197 is
known as a nontoxic DT, further studies demonstrated that
CRM197 has about 10 6 the toxicity of wild-type DT
because of weak ADP-ribosylating activity. In addition, a
large dose of CRM197 (50mg/kg) caused death of DT-
sensitive animals [211]. Therefore, it would be worth testing
safety and neutralizing efficacy of different nontoxic DT
mutants in animals. The last approach is to use two or three
different immuntoxins with different toxin moieties but the
same ligand. For example, the first course of immunotoxin
therapy for CTCL patients would use an anti-CD3 DT
immunotoxin, the second course would use an anti-CD3
PE immuntoxin, and the third course of immunotoxin
therapy would use an anti-CD3 saporin immunotoxin.

<

18.5.2 Immunogenicity

Immunogenicity of immunotoxins hinders repeated dosing
so that a small number of immunotoxins have produced
objective clinical efficacy in clinical trials. Immunotoxins
have two parts, ligands and toxins. Except for monoclonal
antibodies and antibody fragments, all ligands are derived
from human. Monoclonal antibodies and antibody frag-
ments can be humanized. The immunogenicity problem
of the immunotoxins is attributed to the toxin parts. Most
individuals are also immunized with diphtheria toxoid as
children. Some patients have subclinical Pseudomonas
infection. And patients exposed to castor oil may have
anti-ricin antibodies. These circulating antibodies reduce
the half-life, C max (maximum plasma concentration of drug)