Download Peritoneal Lavage Boosts Monoclonal Antibody Uptake in Tumors and more Exercises Medicine in PDF only on Docsity!
BASIC SCIENCES
Improved Radiolabeled Monoclonal Antibody
Uptake by Lavage of Intraperitoneal
Carcinomatosis in Mice
Richard L. Wahl and Monica Liebert
Division ofNuclear Medicine, Department oflnternal Medicine, University of Michigan
Medical Center, Ann Arbor, Michigan
Theeffectof pentoneallavagewith salineon tumorandsystemicuptakeof intraperitoneally
administeredtumor-specific(131l-5G6.4)and nonspecific(125I-UPC-10)radiolabeledmonoclonal
antibodieswas evaluatedin a nudemousemodelof humanintraperitonealovarian
carcinomatosis(lP3 model).Pentoneallavageat 2 or 6 hr postintrapentonealantibody injection significantly improves intraperitoneal tumor/nontumor uptake ratios of specific antibody apparently by limiting systemic exposure to antibody. This enhancement tends to be more dramatic if lavage is performed within 2 hr, rather than 6 hr, of intrapentoneal antibody administration,thoughbothtimesresultinsignificantimprovementsintarget/background ratios over no lavage. Twenty-four-hour tumor/nontumor ratios for specific antibody 5G6.
generallyare 1.5-fourfoldhigherfollowinglavagethanthoseachievedin controlanimals,
withoutdecreasingabsolutetumoruptakeof specificradiolabeledantibody.By contrast,
nonspecific antibody UPC-10 binding is lower in tumor and normal tissues following Iavage, with no Iavage-inducedimprovementin tumor/nontumorratiosseen. Pentoneallavageis a
simplemethodto allowfor specificantibodybindingto accessibleintrapentonealtumorsyet
to limitsystemicexposurethus increasingthe therapeuticmargin.This methodmay have
considerableapplicabilityin the enhancementof intraperitonealimmunoconjugatedeliveryto
intrape,itoneal tumors.
J Nucl Med 30: 60-65, 1989
he use of radiolabeled antibodies as diagnostic or
therapeutic agents is hindered by the relatively low
target/background ratios and absolute tumor delivery
achieved following i.v. administration (1). While a va
riety of methods have been proposed to deal with this
problem including computerized background subtrac
tion (2, 3), antibody fragmentation (particularly using
F(ab')2) (4, 5), and antibody hapten conjugates (6); for
localized diseases, such as those limited to the lymphat ics or body cavities, the possibility of regional antibody
delivery exists, which potentially can circumvent many
of the problems of i.v. antibody delivery.@(7,8). Two
malignancies that commonly are limited to the perito
neal cavity include ovarian and colonic cancer.
Intraperitoneal delivery of monoclonal antibodies
results in significantly higher exposure ofthe peritoneal
cavity to radioantibody than does intravenous admin
Received May 13, 1988; revision accepted Aug. 31, 1988. For reprints contact: Richard L. Wahl, MD, University of Michigan, Medical Center, 1500 East Medical Center Dr., Ann Arbor, MI 48109-0028.
istration (9). This regional delivery advantage translates
to superior specific monoclonal delivery to isolated
ascites cells postintraperitoneal injection, and high solid tumor/blood ratios soon after intraperitoneal injection (10, 11). This can also translateto higherabsolute tumor uptake and less systemic exposure at early times
postinjection for intact tumor-specific radiolabeled an
tibody (12—14).At present, better quantitative results
have been seen in patients with i.p. delivery in colon
cancer than in patients with ovarian cancer (11, 12).
Following i.p. antibody injection blood levels of ra
dioantibody eventually rise due to absorption. With
high i.p. doses, this may result in blood antibody levels
that could result in toxicity (15). This systemic uptake
may limit the utility of this approach for high level
therapy ofisolated accessible i.p. tumor foci. Certainly,
however, this systemic delivery may be necessary if
tumors are not accessible from the peritoneal space,
and in certain instances, the delivery routes may be
complementary (12).
One approach to limiting systemic exposure in cases
60 WahIandLiebert The Journalof NuclearMedicine
where only intraperitoneal tumor (accessible from the
peritoneal cavity) is present, would be to rapidly clear
from the blood the radiolabeled antibody that has left
the peritoneal cavity. We have demonstrated the feasi
bility of this approach in a nontumor bearing animal system through the use of systemically-administered
polyclonal anti-mouse antibodies (16). We hypothe
sized that a similar reduction (or lack of rise in) blood
radioactivity levels might be seen ifthe peritoneal cavity
is lavaged with saline postintraperitoneal injection.
While there would be a rapid drop in the peritoneal
fluid radioactivity level, we hypothesized that specific
antibody would be firmly attached to intraperitoneal
tumor, and that only nonspecific binding would drop,
improving tumor/background ratios. Lavage of radio
active albumin from the peritoneal cavity has been
successfully performed in humans and lavage following
i.p. injection of monoclonal antibodies in humans has
briefly been described (13, 17).
The present study was performed to determine
whether peritoneal lavage would limit the systemic
exposure to radioantibody, yet maintain intraperitoneal
tumor uptake of specific antibody given intraperito
neally in a system with intraperitoneal tumor present (HTB77 IP3 ovarian cancer model) (14). The concept
explored is one of regional (i.p.) delivery of specific
antibody to the regional (i.p.) tumor at high concentra
tion, allowing it to bind, and then removing the Un
bound antibody to prevent unwanted systemic and
peritoneal exposure to radioactivity.
METhODS Antibodies 506.4 is a murine IgO2a kappa and binds to most ovarian cancers (18). It localizes specifically when labeled to ovarian
carcinoma xenografts(14, 19). UPC-10 is a murine IgG2ak
myeloma protein without known specificity (Bionetics, Inc.,
Charleston,S.C.).Generally, 100 @gof purifiedantibody are
@@ c
@ ‘@ p @‘ @@ .$@;‘@‘@ @.
p
@@.
@@@. ::‘@z@L@:,,: - @ ;*.,@,•
@@@ .J. @,*-@
@@@ @-‘t:@,@..
@@ -&...-@-
labeled using the lodogen method(Pierce Chemical Company,
Rockford,)by reactionwith 1 mCi of radioiodine(ICN, Inc.)
with 60—80%efficacy ofincorporation (20). Separation of free from bound iodine is through the use of anion exchange chromatography. Immunoreactivities are measured by a 1-hr
direct cell-bindingassayto HTB-77ovarian carcinomatarget
cells (ATCC) (21).
Animal Model The HTB-77 1P3 model is a model we have developed of human intraperitoneal ovarian carcinomatosis that grows well intraperitoneally in the nude mouse and mimics human ovar ian carcinoma (22). Athymic Swiss Nu/Nu mice first receive
0.5 cc of pristane i.p. (AldrichChemicalCompany, Milwau
kee, WI). One week later, they are innoculated i.p. with 10 million HTB77 IP3 ovarian carcinoma cells. Small tumors grow and attach to bowel, peritoneum, and invade the dia phragm (Fig. 1). The mean tumor size in this study was #@ 0.4 g.
PeritonealLavage
The study was divided into two parts, an initial feasibility study, with kill time immediately after lavage to determine the completeness of lavage, and a later study allowing the animals to survive nearly 24 hr postlavage. In the first part of the study, two groups of four nude mice, each with HTB-
1P3tumors, were injected intraperitoneallywith 0.5 cc of a
dual-label mixture composed of 9 @Ciof iodine-l31 (‘@‘I) 5G6.4 and 13 g@Ciof iodine-125 (‘25I)UPC-lO. The experi mental animals were lavaged four times with 2 cc of saline/ wash beginningat 1.75 hr following intraperitoneal antibody injection. The animals were injected i.p. with saline and then the peritoneal cavity, after allowing for brief mixing, was drained in the prone position while the mouse was suspended by the neck and tail. Dose calibrator readings on the whole animal were performed before and after each lavage. The multiple lavages were completed by 2.4 hr following antibody injection. The control and experimental mice were then killed with tissues and fluids weighed and counted.
In the secondportion of the study 32 HTB77 IP3 bearing
nude mice were studied to determine ifsystemic exposure was lessened by lavage, whether early or later lavage was superior,
FIGURE 1
MIcrOScOpesec@ondemonstratinga
focus of HTB77 lP3 ovarian cancer attheperitonealsurface.
.-‘. .,.-.
@ -,. ‘—a- ,@:
Volume30 •Numberi •Januaryi 989 (^61)
2Effect TABLE hrFollowing of Pentoneal Lavage with Saline at 2 and 6 Intraperitoneal AntibOd@@Administration (Dual Label Mixture of [131I]5G6.4animalswith and [ @IJUPC-10)to HTB-77 lP3 tumors, with Kill Time 24 hr AfterInitialAntibody
Installation[131IJ5G6.
[125l](Jp@10(Specific)
(Irrelevant)
04@ II@ V (a 0.@ 0.
ol dl ZI V I V I ‘II 0.1 0.1 0.
C,) ice ice' 0 10 10 @ 0 +1 @H—J. @@@ 0)Co ,@.
@ u@
@ .@ •
@2 J @
Results:(killtime24 hr followingintraperitonealantibodyinjec tion):expressedas % kg injecteddose/g±i s.e.m.(32animals studied). NotethatwithIavage,thenormalorganuptakeforbothanti bodiesdropssignificantlyversuscontrols(exceptfor the spleen), as wellas forthetumorwithUPC-10(p generally<0.01). With 5G6.4,nodropintumoruptakeis seenwithlavage.In the 1-hr bindingassayin vitroto HTB-77ovariancarcinomacells,<1% of UPC-i0 inputcountsbound,while30—50%ofinput5G6.4counts bound.
ControlsTumor 0.004Thigh i.p.0.063 ±0.0050.063 ± 0.0040.036±0.007Blood0.072 muscle0.016 ± 0.009Liver0.021 ±0.0070.148 ± 0.001Spleen0.016 ±0.0020.031 ± 0.0010.03±0.002Lung0.040 ± 0.0062-hr ±0.0040.083 ± LavageTumor 0.002Thigh i.p.0.061 ±0.0160.034 ± 0.0010.014±0.001Blood0.024 muscle0.005 ± 0.006Uver0.013 ±0.0030.075 ± 0.002Spleen0.015 ±0.OOi0.02 ± 0.001Lung0.014 ±0.0020.018 ± 0.0026-hr ±0.OOi0.034 ± LavageTumor 0.006Thigh i.p.0.056 ±0.0060.05 ± 0.001Blood0.048 muscle0.009 ±0.0010.016 ± 0.0060.190±0.007Liver0.012 ± ±0.001Spleen0.015 ±0.0010. 0.001Lung0.02 ±0.0020.021 ± ±0.0020.045 ±0.
(I) E
C@) 0@ N.N. I-
v@ 04 In peritoneal cavity. In the largest tumors the uptakes of
5G6.4 with lavage were slightly lower than in their
control group; though even in these, peritoneal lavage
enhanced tumor/nontumor uptake. Nontarget 5G6.
uptakes in the lavaged animals were in general far less
than control values (except for the spleen). There was
also a significant drop in the tumor and systemic uptake
of nonspecific antibody (UPC-10) following lavage in
the same system, which was larger than the drop seen
when kill time was at 2.4 hr (immediately postlavage)
(46% vs. 29%).
The mean specific antibody (5G6.4) tumor/nontu
mor ratios, shown in Table 3, clearly demonstrate the
enhancement of relative tumor uptake of the 506.
antibody by the lavage approach. The enhancement
tended to be greater (p < 0.1) when lavage was at 2 hr
postinjection, though both 2 and 6 hr time points
generally showed significant (p < 0.01—0.05)improve
ments in the tumor/nontumor ratios over control ani
Volume30 •Number1 •Januaryi 989 63
,- ‘ c@JIn I- 0 CV) 01* 0 0 0
@@ 01 @+I @+I (^) ll@@+I II @ t@;-i a@-i C @@@ 0 C')
C1J U) ZI z H I
1@ @Ifl @ (^) 0 i;. 10 @ 0 +1—a (^10) I0 0) 0 01
U) z
01 0 @11— 01
F,. 0 0 @H .11—
@Th
N.
U) In z z II II
01
Co @11—I.
01
0
z
I
I
mals (Table 3). mean 2-hr lavage tumor/blood ratios of 6.3/1 are significantly better than would typically be achieved with the intact 506.4 monoclonal antibody given intravenously at this time point. UPC-10 tumor/ nontumor ratios following lavage versus controls (Table
- are virtually identical, indicating that the lavage decreases all tissues uptake of nonspecific antibody proportionately. Thus the lavage process is enhancing the specificity of the labeled monoclonal antibody for
tumor.
DISCUSSION
The regional nature ofdisease such as ovarian cancer and some colon cancers, with spread mainly in the
peritoneal cavity, should lend themselves well to re
gional antibody delivery. A variety of pre-clinical and clinical studies have shown this antibody delivery ap proach to be useful for colon cancer and in some
instances in ovarian cancer (8, 10—14).It is also clear
that the intraperitoneal approach alone is not the an
swer for all i.p. tumors because if there is not access of
antibody to antigen, then binding may not occur or may be superior by the vascular delivery route (12).
Results with the HMFG-2 antibody in ovarian cancer
in humans have shown that intraperitoneal delivery of
that antibody may, in fact, result in less delivery to tumor foci than i.v. delivery (14). Despite these concerns, in our animal model of in
traperitoneal human ovarian carcinomatosis, it is clear
that peritoneal lavage enhances tumor/background ra
tios without, on average, decreasing absolute tumor
uptake of antibody. The mechanism of this phenome
non most likely is due to binding of antibody to intra peritoneal tumor, followed by removal of non-bound
intraperitoneal antibody before it is absorbed and dis
tributed systemically.
Systemic exposure is reduced up to 50% by this maneuver. This enhancement in relative tumor uptake isachieved,on average,without compromisingabsolute tumor dose of specific antibody. In the larger (0.7 g or
greater) tumors, specific antibody uptake appeared to
drop somewhat with lavage, but no drop was seen with the 0.06 g tumors and on average there was no statisti cally significant alteration in tumor binding. The bind ing of nonspecific antibody UPC-lO to tumor was
dropped by lavage, with the drop more apparent at 1
day postlavage than immediately postlavage. This greater drop in the experiments carried to 24 hr than
those at 2.4 hr postinjection may be related to ongoing
loss of weakly-attached nonspecific antibody over the
time of observation. There was no difference between the 2- and 6-hr lavage times for UPC-lO.
From these studies it is apparent that through the use
oflavage it should easily be possible to increase several
fold the amount of specific radioantibody given intra
peritoneally, without increasing systemic exposure over
a low i.p. dose without lavage, and thus increase abso
lute tumor uptake of radiolabeled antibody to small, accessible, antigen positive intraperitoneal tumor foci. Naturally, careful attention would need to be paid to the possibility of bowel radiotoxicity should higher ra
dioantibody doses be administered; however, the cu
mulative dose to bowel should not be greater than when no lavage is performed. This may be therapeutically
valuable as intraperitoneal radioimmunotherapy is su
perior to intravenous in an animal model of aggressive
intraperitoneal adenocarcinoma of the colon we have been studying (24). Certainly, for tumor foci not acces sible by intraperitoneal antibody delivery (i.e., where
vascular delivery is essential, including many large and
subserosal lesions), this lavage approach will not en hance antibody delivery as both vascular and intraper
itoneal delivery are needed (12). It may, however, be
possible to combine intravenous and intraperitoneal delivery plus lavage to optimally deliver radioantibody
to both types oftumor (12). Regional chemotherapy to
the peritoneal cavity is most valuable in low bulk disease
and it is likely that this would be the case for regional
antibody delivery to the pentoneal cavity followed by lavage (25).
In conclusion, peritoneal lavage is a simple method
to enhance the specificity of antibody binding to acces
sible, antigen positive, intraperitoneal tumors and to limit systemic exposure. These effects may allow for higher doses of radiolabeled antibody to be given with
the result being more antibody reaching the accessible
tumors. Since ovarian cancer spreads initially by intra
peritoneal dissemination, such an approach seems ra
tional particularly in early disease. Peritoneal lavage,
through increasing the therapeutic index, may have
considerable applicability in the enhancement of intra peritoneal radioimmunodiagnosis and radioimmuno therapy, as well as in the delivery ofother immunocon
jugates, though validation of the technique in patients
will be necessary.
ACKNOWLEDGMENTS
The authors acknowledge the technical assistance of Gayle Jackson, Joseph Wissing, Martin Strnat, Sue Fisher and Phil Sherman. The word processing skill of Michele Curro is also appreciated.
This work was supported by CA 41531 awarded by the
PHS.
REFERENCES
- Goodwin DA. Pharmacokinetics and antibodies. J NuclMed 1987;28:1358—1362.
- Goldenberg DM, DeLand F, Kim E, et al. Use of
64 WahIandLiebert The Journalof NuclearMedicine
