| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Clinical Trials |
Laboratory of Molecular Biology [L. H. P-S., I. P.], Department of Nuclear Medicine, Warren G. Magnuson Clinical Cancer Center [J. A. C., C. P., M. W.], Medicine Branch [D. P., M. H., C. A.], Chemistry Section, ROB [O. G., M. B.], National Cancer Institute, NIH, Bethesda, Maryland 20892; Food and Drug Administration, Bethesda, Maryland 20892 [K. W.]; and Department of Pathology, Wake Forest University, Winston-Salem, North Carolina 27157 [M. C. W.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this study, we determined whether mAb B3, a murine IgG1
that
reacts with the LewisY carbohydrate epitope (B3 antigen;
Ref. 13
), would serve as a target for radioimmunotherapy. This epitope
is present on a large number of glycoproteins and is abundantly and
uniformly expressed by most carcinomas, including >95% of colorectal
cancer, 80% of breast cancer, and 60% of non-small cell lung cancer
as well as esophageal, gastric, pancreatic, ovarian, and bladder
carcinomas. In contrast, mAb B3 has limited reactivity with normal
tissues. mAb B3 has been chemically linked to a truncated form of
Pseudomonas exotoxin to form immunotoxin B3-LysPE38 (LMB-1).
In a Phase I clinical trial, LMB-1 was given to 38 cancer patients with
tumors that react with mAb B3. In that study, five objective responses
(1 complete remission, 1 partial remission, and 3 minor
responses) were observed (14)
, indicating that the
B3 antigen can be used as target for cancer therapy.
To explore the usefulness of B3 antigen as a target for radioimmunotherapy, preclinical experiments were conducted. When injected into immunodeficient mice bearing a human epidermoid carcinoma that expresses the B3 antigen, 111In-B3 showed selective and progressive accumulation at the tumor site (15) . These results and preclinical biodistribution studies using 88Y-mAb B3 (16 , 17) indicated that radiolabeled B3 warrants further clinical evaluation.
Whereas the largest number of radioimmunotherapy studies have focused
on using 131I-labeled antibodies, several limitations have
been identified, including rapid dehalogenation (18
, 19)
and emission of high-energy 
-rays, which imposes certain radiation
safety constraints. 90Y has been evaluated as an
alternative to 131I for radioimmunotherapy because of its
ready availability from a 90Sr/90Y generator
(20)
and its physical and biological characteristics
(21, 22, 23, 24)
. Whereas 90Y has favorable
characteristics for therapy [t1/2 =
64 h; pure ß-emission (Emax = 2.28
MeV)], the lack of -ray emission makes it suboptimal for imaging
and assessing biodistribution (12)
. To trace the
biodistribution of 90Y, 111In has been used as
a surrogate marker because it has similar coordination chemistry
(25
, 26)
and metabolic handling (19
, 22)
. In
this study, we carefully compared the differences in biodistribution
between 111In- and 90Y-labeled B3. Prior
studies using weaker chelates have shown significant differences
between 111In and 90Y (23
, 24)
.
Newer chelates have greater in vitro and in vivo
stability (16
, 27
, 28)
. Nevertheless, even with improved
chelates, some differences between 111In and
90Y have been observed (17
, 24)
. Using a
different antibody labeled with 111In and 90Y
via the 1B4M chelate (also known as Mx-diethylenetriamine
pentaacetic acid; Ref. 27
) in patients with adult T-cell leukemia
(3)
, we have previously shown that there are differences
in biodistribution between these two radiolabels, although these
differences were small (29)
. Other investigators have also
used the same chelate conjugate to label other antibodies
(30, 31, 32)
; nevertheless, this is the first detailed
pharmacokinetic comparison of these two isotopes using this chelate in
epithelial tumors.
This study presents the results of a clinical trial in patients with advanced carcinomas that express the B3 antigen. We studied the ability of 111In-1B4M-mAb B3 to image known metastasis and performed a Phase I trial to determine the toxicities, pharmacokinetics, and the MTD of 90Y-1B4M-mAb B3.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Conjugation and Labeling.
The B3 mAb was conjugated to
2-(4-isothiocyantobenzyl)-6-methyl-diethylenetriamine pentaacetic acid
(1B4M-diethylenetriamine pentaacetic acid; Ref. 27
).
Radiolabeling was performed with pharmaceutical grade 111In
(DuPont New England Nuclear, Wilmington, DE) for imaging and/or
pharmaceutical grade 90Y for therapy (DuPont New England
Nuclear). In brief, 1.0–1.2 mg of conjugated mAb B3 was put into a
polypropylene vial that served as the reaction vessel. For
111In, 10.2–23.2 mCi were added to the reaction vessel and
allowed to react for 1 h. For 90Y labeling, the
starting amount of radioactivity and antibody dose depended on the dose
level to be used. Typically, 8.9–53.3 mCi of 90Y were
incubated with 1.08–4.31 mg of the conjugate for 15 min. After the
initial eight patients’ 90Y labeling, the method
was modified to add ascorbate (11 mg/0.05 ml) as a radioprotectant
during the incubation with 90Y. Excess DTPA
(10-4 M) was then added to the incubation
mixtures to form complexes with unreacted ionic isotope. The mAb
B3-bound fraction was separated by preparative size-exclusion
high-performance liquid chromatography (3)
. Purification
resulted in a final product with >99% antibody-bound
111In or 90Y. The 90Y fraction was
mixed with 25% HSA to yield a 2.5% HSA solution. Purity was
determined by instant thin-layer chromatography using silica
gel-impregnated glass fiber sheets (2:2:1, 10% ammonium formate in
water/methanol/0.2 M citric acid) and paper chromatography
using saline solvent and Whatmann #1 paper pretreated with 5% HSA. The
final product was filtered using a sterile 0.22-µm
low-protein-binding filter (Millex-GV; Millipore, Inc., Bedford, MA).
The specific activities of the 111In-mAb B3 doses
(n = 26) ranged from 4.3–13.0 mCi/mg (7.3 ± 1.8
mCi/mg). The total 111In activity injected ranged from
3.5–5 mCi. The specific activities of the 90Y-mAb B3 doses
(n = 23) ranged from 5.3–12.9 mCi/mg (8.4 ± 1.7
mCi/mg), with individual doses of 5–25 mCi of 90Y-mAb B3.
All products passed sterility and pyrogen testing. The
111In-labeled products were injected within 72 h of
preparation. Twenty-two of 23 90Y-mAb B3 doses were
injected the day of labeling, whereas one product was injected the next
day (24 h). The dose injected the day after labeling was retested
before injection and showed similar protein-bound radioactivity.
The immunoreactivity of the radiolabeled products was tested using a modification of the cell-binding assay described by Lindmo et al. (33) . In brief, an increasing number of A431 cells were incubated in 6-well plates in cell numbers ranging from 4 x 105 to 1 x 106. A small fixed amount of the radiolabeled B3 was added to the wells and incubated for 2 h. After incubation, the wells were washed, and the percentage of activity bound to cells was determined. Overall, the cell binding assay for the 111In and 90Y preparations was not significantly different, with a mean ± SD of 69 ± 10% and 69 ± 17% for 111In- and 90Y-mAb B3, respectively. When the immunoreactivity from 90Y preparations labeled in the absence of ascorbic acid were compared with those labeled in the presence of ascorbic acid, a significant difference was observed (58 ± 10% and 74 ± 18%, respectively).
Patient Selection.
Adult patients with metastatic gastrointestinal tract, breast,
non-small cell lung, bladder, and ovarian carcinoma who had failed
standard therapy were eligible for this study. Tumors expressed B3
antigen on 
30% of the tumor cells as determined by
immunohistochemistry. Tumor histology was confirmed by a NIH
pathologist. Other eligibility criteria included: (a)
advanced unresectable disease; (b) failed conventional
chemotherapy; (c) Eastern Cooperative Oncology Group
performance status of 
2; (d) a minimum life expectancy of
3 months; (e) serum creatinine < 1.6 mg/dl;
(f) serum bilirubin < 1.5 mg/dl; (g)
absolute neutrophil count (ANC) > 2,000/mm3; and
(h) platelets > 100,000/mm3. Patients with
clinically significant cardiac disease (New York Heart Association
grade 3 or 4) were excluded, as were those with infectious disease that
required antibiotic therapy, brain metastasis, prior exposure to murine
antibodies, pregnancy, or lactation. Patients who had received bone
marrow transplant therapy, more than 3 chemotherapy regimens,
pelvic radiation, or local radiation to more than one site were
excluded. The clinical protocol and the consent form were approved by
the Institutional Review Board of the National Cancer Institute.
Informed consent was obtained from all patients before participation in
this study.
Study Design.
An outline of the protocol design and the number of patients entered in
each group are shown in Table 1
. In the
initial portion of this study, three patients received 5 mCi of
111In-mAb B3. The objective was to determine whether there
would be any gross unexpected sites of localization of the antibody or
acute toxicities that would prompt us not to proceed with the therapy
portion of the study. To provide potential therapeutic benefit, these
initial patients were offered therapy with 5 mCi of 90Y-mAb
B3, if they had tumor imaging, no HAMAs, and no toxic side effects or
unexpected, undesirable tissue accumulation as compared with other
111In-labeled antibodies. In the second portion of the
study, we evaluated whether the amount of antibody resulted in
dose-dependent changes in biodistribution. Three groups of three
patients each received 111In (5 mCi)- and
90Y-mAb B3 (5 mCi) mixed together with 5, 10, or 50 mg of
unlabeled mAb B3. In the third and main part of the protocol, the MTD
of 90Y was determined. Groups containing three patients
each received escalating doses of 90Y-mAb B3 for therapy
coinfused with 5 mCi of 111In and a total of 10 mg (based
on a lack of dose-dependent changes at the higher amounts of mAb B3).
The 90Y doses were escalated in 5-mCi intervals. Patients
with hematological toxicity < grade 3 were eligible for
retreatment with the same dose of 90Y if they had no
evidence of disease progression and remained HAMA negative.
|
Pharmacokinetics.
Intravascular kinetics were determined by counting 111In or
90Y radioactivity in blood and plasma aliquots obtained at
the following times after the end of infusion: 5 min, 30 min, 1 h,
2 h, 6 h, 1 day, and daily for up to 7 days after the end of
the infusion. The % ID/ml was obtained by comparing the counts to a
standard of the injected dose. The plasma and blood volumes were
estimated at each time of treatment using a nomogram based on body
surface area (34)
. Using the latter estimated volumes and
the % ID/ml, the total % ID in the blood and plasma volume was
calculated. Because the infusion time was short compared with the
disposition half-life (t1/2), the intravascular
data were treated similar to an i.v. bolus. The % ID/ml of blood or
plasma was fitted to a biexponential curve to obtain both the 
and
ß phase t1/2 using a least-squares fit
algorithm. Conventional pharmacokinetic parameters were then derived
(35)
. The AUCs for the blood or plasma curves were
calculated in two steps. First, the AUC from the end of antibody
infusion (To) to 168 h was obtained by trapezoidal
integration of the decay-corrected blood and plasma data, and then the
terminal AUC was estimated using the terminal clearance rate to
extrapolate from the activity retained at the last measured time point.
Using this data, we then estimated additional pharmacokinetic
parameters, including volume of distribution, clearance, and
half-life (35)
. Serial 24-h urine collections were
obtained for up to 96 h so that we could compare the urinary
excretion of the two tracers. Whole body clearance of 111In
was determined from the imaging data (see below).
Imaging.
Scintillation camera images were first recorded up to six times with a
large field of view dual-headed gamma camera starting within 
2 h of
the end of the infusion and daily for up to 6 or 7 days. Analogue and
digital images of anterior and posterior whole body as well as spot
views (5–10 min/image) were obtained. For quantitative imaging, a 20%
window centered over the 247 keV photopeak of 111In was
obtained. A scatter correction method was utilized. The images were
corrected for attenuation using a 99mTc flood source and
allowing for the energy differences and the sensitivity of the gamma
camera. A geometric mean image was then generated (in units of
µCi/pixel). For visual interpretation, individual anterior and
posterior images were reviewed. SPECT of the chest, abdomen, and pelvis
was recorded using a medium energy collimator; SPECT was typically
performed 4 to 5 days after injection. The 111In SPECT
images were obtained for visual assessment utilizing a 20% window
centered over the 174 and 247 keV photopeak of 111In; the
images were reconstructed using a Hamming filter with a high cutoff
frequency of 0.75 cycle/cm. All images were interpreted by one
experienced nuclear medicine physician.
Two quantitative image regions of interest were drawn over the liver, spleen, and L4 vertebral body on the geometric mean images. The integrated radioactivity in the organs (AUC) was then determined by trapezoidal integration up to the last time point imaged (typically 7 days), and the remaining AUC was determined by extrapolation using the terminal t1/2 of clearance. The whole body t1/2 was obtained by fitting the geometric mean concentration of 111In activity of the anterior and posterior whole body scans. Using the residence times obtained from the data above and the individual organ size, as determined from CT, the medical internal radiation dose method was used to calculate organ dosimetry (36) . In brief, the integral of radioactivity in a given organ, blood, or tissue was divided by its weight (µCi ·h/g) and multiplied by the mean energy emitted per nuclear transition of the ß particle from 90Y (1.99 g ·Rad/µCi/h), assuming an absorbed dose fraction of 1. The activity in L4 was determined from the region of interest analysis and integrated over time. This activity was normalized by the grams of bone marrow in L4 as estimated in standard man (37) . The dose to the bone marrow was then estimated as described above.
Counting Methods.
Dual isotope counting of 111In and 90Y was
performed on the patient samples. The 111In -ray peaks
were counted in a gamma counter using a 100–500 keV energy setting.
Because 90Y is counted with <4% efficiency in a gamma
counter, Cerenkov counting in a beta counter was also utilized.
Cerenkov counting was performed using an energy range of 0–200 keV
(A4530D Packard, Downes Grove, IL) in a beta counter. Because Cerenkov
counting is sensitive to quench and geometry, all samples were
processed in a similar and reproducible manner as described previously
(29)
, which included solubilizing the samples with SDS and
bleaching them with 30% hydrogen peroxide to minimize quench. The
counts in the samples were referred back to a standard of the injected
dose that had been prepared in a similar fashion using the patients’
plasma or blood to mimic the quench. The counts obtained in the gamma
and beta counters were corrected for cross-talk and decay.
Bone Marrow Biopsies.
Twenty-five patients underwent bone marrow biopsy of the posterior
iliac spine within 5–7 days after mAb injection. Thirteen biopsies
were performed 6 days after therapy, eight were performed at 5 days
after therapy, and four were performed at 7 days after initial therapy.
The biopsy cores were weighed on an analytical balance and put in a
conical tube with 10 ml of PBS for 1 h and analyzed as described
previously (29)
. The core was broken with a jagged-edged
glass rod. This was centrifuged for 10 min at 640 x g,
and the supernatant was removed and counted (saline fraction). The
pelleted core was broken with a jagged-edged glass rod and mixed with
0.5 ml of 10% SDS. The core was heated to 56°C for 30 min in an
attempt to remove any cell-bound activity. After the sample cooled, 0.4
ml of 30% hydrogen peroxide was added as bleach, and the mixture was
incubated at 56°C for 1 h to bleach the sample. Ten ml of
distilled water were added, and the sample was again centrifuged for 10
min. The supernatant was separated for counting (SDS fraction).
Perchloric acid (0.2 ml) was then added to the remaining bone chips,
and the mixture was incubated at 56°C until the bone was dissolved
(bone fraction). This sample was again treated with hydrogen peroxide
as described above. After cooling, the sample was transferred to a
counting vial with 10 ml of distilled water. All samples were then
counted in the gamma and beta counters with the appropriate decay and
cross-talk corrections.
To attempt to find parameters that predict bone marrow toxicity, we correlated the dose to bone marrow based on imaging, bone marrow biopsy, and blood retention of 90Y. The dose to the blood was calculated as described above, and the dose to the marrow was then determined by multiplying times the red marrow:blood ratio as described by Sgouros et al. (38) . The dose from marrow biopsies was determined by calculating the activity concentrated in the marrow by gamma counting of a precisely weighted specimen. We assumed that there was no biological clearance from the marrow, based on gamma camera imaging. The dose was then calculated as described above, with no corrections for cortical bone. In addition, we correlated toxicity to the bone marrow with administered activity or administered activity corrected for body weight.
HAMA Assays.
HAMA assay was performed as described previously using a
high-performance liquid chromatography method (39)
. More
than 10% complex formation using a 125I isotype matched
nonspecific mAb (BL-3) and 125I-labeled B3 was considered
positive.
Statistics.
To compare independent data, 111In and 90Y
patient data obtained from the initial dual-injection study were used.
Paired t test or Wilcoxon rank signed test (when data were
not normally distributed) was performed to assess the differences in
biodistribution between the two radiolabels. Pearson’s correlation
coefficient was used to evaluate the relationship between nominal data,
and Spearman correlation coefficient was used with ordinal data.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
There were 12 men and 14 women (age range, 39–73 years; mean age, 58
years). Twenty patients had colorectal cancer, 2 patients had
esophageal cancer, 1 patient had gastric carcinoma, 1 patient had
carcinoma of the ampulla of Vater, 1 patient had breast cancer, and 1
patient had bronchioalveolar cancer. Prior therapy and sites of
metastasis are shown in Table 2
. Immunohistochemical staining of B3 was
performed on paraffin-embedded tumor blocks from the primary surgery in
all cases. Homogeneous B3 expression was found in 23 of 26 patients. In
3 of 26 cases, specimens were poorly preserved (tightly fixed). However
in all three cases, sections indicated that in the better preserved
areas, more than 30% of the tumor cells were clearly positive for B3
antigen.
|
Imaging Results.
Twenty-one of 26 patients (85%) had localization in known sites of
disease. Sites of disease imaged included liver (Fig. 1)
, lung (Fig. 2)
, bone, soft tissue mass, and lymph
nodes. Table 3
shows the radiographic
results and 111In-mAb B3 imaging results of individual
patients.
|
|
|
; whereas three patients had hot lesions.
Eight patients were considered negative and had either no localization
or lesions that appeared cold and had some filling in on delayed
imaging but were never hotter than the normal liver.
Normal sites of 111In accumulation included the liver,
spleen, and bone marrow (Fig. 1)
. Excretion into the urinary bladder
and occasional mild bowel uptake were seen. Bowel activity moved over
time, indicating that it represented intraluminal activity. No gall
bladder uptake was visualized. Delayed images showed progressive
clearance of the blood pool and persistent accumulation in the liver
and bone marrow. Lesions were better visualized on SPECT scans than on
planar images (Figs. 1
and 2)
. No gross differences in imaging were
observed at the various dose levels to suggest dose-dependent changes
in biodistribution.
Pharmacokinetics.
The administered mass of mAb B3 did not have a significant effect on
pharmacokinetic parameters for 111In (Table 4)
. No significant differences in plasma
pharmacokinetics were observed between 111In- and
90Y-mAb B3 (Table 5)
. The
blood and plasma clearance was very similar for both 111In-
and 90Y-mAb B3 (Fig. 3)
. The
amount of antibody remaining in the plasma volume at the end of
infusion was 80.7 ± 6.1, 86.7 ± 10.6, and 90 ± 7.6%
ID for the 5, 10, and 50 mg dose of 111In-mAb B3,
respectively (ANOVA, P = 0.43). At 168 h after
infusion, the amount of 111In-mAb B3 remaining in the
plasma volume was 12.3 ± 2.0, 11.8 ± 3.0, and 12 ±
1.6% ID, for the 5, 10, and 50 mg dose, respectively (ANOVA,
P = 0.96).
|
|
|
The amount of 90Y injected did not affect the urinary
excretion (0–96 h) of 90Y. Patients receiving 5
(n = 3), 10 (n = 3), 15
(n = 3), or 20 mCi (n = 6) showed no
significant differences in urinary excretion of 90Y with
8.5 ± 2.2, 11.1 ± 1.0, 18.9 ± 15.1, and 13.3 ±
3.7% ID excreted in 96 h, respectively (ANOVA, P = 0.35).
The urinary excretion of 111In and 90Y in the
interval collected is shown in Table 6
.
The urinary excretion of 90Y in the first 24 h was
greater than that of 111In [3.9 ± 1.6% compared
with 3.4 ± 1.3% ID, respectively (P = 0.016)].
However, at 48–96 h, the urine excretion of 90Y was less
than that of 111In [8.0 ± 4.8% ID excreted compared
with 12.3 ± 6.3% ID, respectively (P = 0.016)].
|
. Hematologic DLT for
this study was defined as an ANC < 500/µl for >5 days or
platelets < 25,000/µl for >5 days. At doses of 5 and 10 mCi of
90Y-mAb B3, minimal to no hematological toxicity was
observed. Grade 2 and grade 3 thrombocytopenia were observed at 15 and
20 mCi of 90Y-mAb B3. One instance of DLT occurred in one
patient at 20 mCi of 90Y (patient 21); as a result, a total
of six patients was treated at this dose level. Dose-limiting
neutropenia and thrombocytopenia occurred in two patients who received
25 mCi of 90Y (patients 25 and 26). Grade 4 neutropenia
(absolute granulocyte count < 500) was observed at day 28 after
dosing and lasted 5 and 18 days. One patient required hospital
admission and i.v. antibiotics for fever. Both patients received
granulocyte colony-stimulating factor. Grade 4 thrombocytopenia
(platelets = 25,000/µl) requiring platelet transfusion was
observed at 18 and 25 days after dosing, lasting 42 days in one patient
and 11 days in another patient.
|
. With the exception of estimates of
bone marrow dose from imaging, the majority of the parameters showed a
good correlation with changes in peripheral blood counts. The
correlation of drop in blood counts or bone marrow toxicity based on
the dose (mCi) administered correlated in the same range as the other
parameters that corrected the dose (mCi) based on weight. The estimates
of bone marrow dose based on the blood radioactivity or the
90Y accumulation in the bone marrow biopsy were in the same
range and were not better predictors of toxicity than the administered
activity.
|
The concentration of radioactivity in the bone marrow did not vary depending on the dose of 90Y administered. The mean activities of 111In in the bone marrow for the 5, 10, 15, 20, and 25 mCi doses were not significantly different with a mean of 0.0039, 0.0029, 0.004, 0.0027, and 0.0044% ID/g, respectively (ANOVA, P = 0.123). Similarly the % ID/g of 90Y in the bone marrow was not affected by the administered dose (ANOVA, P = 0.56).
The concentration of 111In in the bone marrow was lower
than that for 90Y in the 17 patients receiving 10 mg of
antibody with escalating doses of 90Y. The concentration of
111In was 0.0034 ± 0.0015% ID/g, whereas that for
90Y was 0.0047 ± 0.0015% ID/g. The distribution of
the 111In and 90Y in the compartments of the
bone measured (saline, SDS, and bone) were different for
111In and 90Y (Fig. 4)
.
|
Dosimetry.
The doses to liver, spleen, bone marrow, and tumor were estimated based
on the 111In biodistribution. The mean dose to the liver
and spleen was 19.4 ± 4.5 and 22.2 ± 8.6 rad/mCi,
respectively. The maximum total dose to the liver in this study was 576
rad, and the maximum total dose to the spleen was 716 rad. The dose to
the bone marrow based on the AUC of blood was 4.8 ± 1.6 rad/mCi,
and the marrow dose estimated from the biopsy was 8.3 ± 3.1
rad/mCi. Many of the tumors visualized were very extensive and
overlapped with other tumor or normal structures; in some cases, their
size could not be determined. The doses to tumors were variable and
ranged from 7.7–65.1 rad/mCi. The mean dose to tumor was 25.1 ±
18.3 rad/mCi.
Tumor Response.
No tumor responses were observed in this study. Stable disease was seen
in 6 of 24 patients. No patient could be retreated because all patients
developed HAMA response within 5–6 weeks after dosing.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have shown here that 111In-mAb B3 has good sensitivity
for imaging metastatic epithelial carcinomas that express the
LewisY antigen. Twenty-one of 26 patients had positive
uptake of 111In-mAb B3, as shown by planar and SPECT
imaging. Sites imaged included lung, liver, bone, soft tissue masses,
and lymph nodes. The ability of 111In-mAb B3 to detect
known sites of metastasis was most effective when the tumor was
localized in soft tissues, in particular, the peritoneal cavity and the
abdominal wall. In this location, strong, localized uptake of
111In-mAb B3 was observed in eight of eight patients. This
is particularly important if the agent is to be used in the setting of
occult recurrences in colorectal carcinomas, ovarian carcinomas, and
other gastrointestinal malignancies. Although the accumulation of
111In in the normal liver tissue makes it difficult to
develop tumor
background ratios high enough to detect the malignancy,
clear positive uptake was observed in 9 of 17 patients with liver
metastasis. Four different patterns were observed. In some cases, the
liver metastasis presented as hot lesions, whereas others had lesions
with a hot rim and cold centers. Some lesions that were considered
negative had either no localization or had lesions that appeared cold;
however, on delayed imaging, these lesions showed some filling in but
were never hotter than the normal liver. It is possible that these
represent areas of necrosis and/or poor vascularization
(42, 43, 44)
. Failure of 111In-mAb B3 to detect
all tumors in this study could be due to nonhomogeneous expression of
the B3 antigen in the metastatic sites or poor vascularization. It is
also possible that a number of lesions detected by radiographic imaging
represent reactive lymph nodes or fibrotic tissue.
The usefulness of mAb B3 as an imaging agent would need to be further evaluated in larger clinical studies in patients with small volume disease (e.g., patients with elevated carcinoembryonic antigen or CA-125 with no radiographic evidence of disease). However, the tumor localization of B3 observed in this first trial indicates that the use of this antibody as a radiotherapeutic agent warrants further investigation.
The use of 90Y-labeled antibodies directed against lymphoma has resulted in significant objective tumor responses (3 , 31 , 32 , 45) . Response in epithelial tumors has been more difficult to attain. Several radioimmunotherapy trials using 90Y have been reported previously, although the majority are preliminary reports or reports on a smaller number of patients than seen in our trial.
90Y-mAb B3 is well tolerated at doses up to 20 mCi, with neutropenia and thrombocytopenia as the dose DLTs. The MTD is in the range observed by others (3 , 46 , 47) . Our dose estimates based on marrow biopsy were in a range consistent with that at which bone marrow toxicity would be expected at the 25-mCi dose level. Toxicity could be correlated to various 90Y parameters including administered dose, dose/kg, dose/m2, and dosimetry based on blood or biopsy. The correlation to dose based on imaging was the poorest, perhaps due to the difficulty in assessing bone marrow concentration using imaging techniques. If the bone marrow were not dose limiting, our normal organ dosimetry in liver would suggest that significantly higher doses could be utilized, if one assumes that toxicity for external beam to the liver will be similar, because the TD50/5 for liver is 2000 rad. The dosimetry estimates to tumor suggest that we could deliver doses of up to 80 mCi. The bone marrow concentration observed, based on bone marrow biopsy, is in the range of that observed with murine anti-Tac (29) . The 111In concentrations from mAb B3 and mAb anti-Tac in bone marrow were 0.00378 and 0.00293% ID/g, respectively (P = 0.243 paired t test). In addition, the concentration of 90Y in the bone marrow from 90Y-B3 and anti-Tac were 0.0047 and 0.00494% ID/g, respectively (P = 0.777, paired t test).
The blood clearance of 111In- and 90Y-mAb B3 is similar to that reported previously with 1B4M chelate and other second-generation chelates (29 , 48) ; in contrast, chelates that are less stable for 90Y have shown less concordance in preclinical models (16) or in clinical trials (49 , 50) .
The tumor doses that we delivered are in the general range of those reported by others using other 90Y-labeled antibodies (49 , 51) . The doses to other organs are generally similar, although some antibodies may have a higher concentration in the liver and therefore result in slightly higher liver doses (51 , 52) . These higher liver doses may be related to the presence of circulating antigen and complex formation (51) , which was not present in our system.
Halpern et al. (53) have previously shown that the distribution of each mAb is unique unto itself with regard to its response to the mass effect. Because this "carrier effect" can alter not just the serum half-time of an antibody but also the organ tumor distribution of the radionuclide, we administered increasing amounts of mAb B3 to patients. No dose-dependent differences in pharmacokinetics were observed at doses of 5–50 mg of mAb B3, nor were there differences in urine excretion that related to the dose administered. Pharmacokinetic analysis indicated that there were no significant differences in plasma clearance of 111In- and 90Y-mAb B3.
Nevertheless, significant differences in urinary excretion of 111In and 90Y were observed. These consisted of slightly faster excretion of 90Y at the early times after injection but much higher excretion at later times. This probably reflects different handling of the radiometals, with possible excretion of some 111In as catabolism occurs, which would contrast with retention of 90Y, in particular, in the bone. These findings are similar to those we have reported previously with two different 90Y-radiolabeled mAbs (29 , 48) . The preferential accumulation of 90Y was evident as reflected by the bone marrow biopsy, which showed a mean of 1.4 times more 90Y than 111In. Using our coarse bone marrow fractionation method, it was clear that the greatest portion of 90Y retained was in the bone fraction.
In summary, we have shown here that specific targeting of the LewisY antigen using radionuclide linked to mAb B3 is feasible. It is clear that to achieve an objective antitumor effect, higher doses of radionuclide will need to be delivered. Strategies to improve the therapeutic index include peripheral autologous stem cell rescue to overcome the myelosuppression (54 , 55) and i.v. infusion of EDTA to prevent binding of free 90Y to bone (56) . As in external beam radiotherapy, the use of fractionated multiple doses for delivery of 90Y should also be explored. Humanized mAb B3 is presently being produced for evaluation in radioimmunotherapy. This should overcome the problem with HAMA formation. A clinical trial to explore the use of a higher dose of 90Y-mAb B3 with autologous stem cell support is being planned at the National Cancer Institute.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 To whom requests for reprints should be
addressed, at Laboratory of Molecular Biology, National Cancer
Institute, Building 37, Room 4E16, 37 Convent Drive MSC 4255, Bethesda,
MD 20892-4255. Phone: (301) 496-4797; Fax: (301) 402-1344. 
2 The abbreviations used are: mAb, monoclonal
antibody; HSA, human serum albumin; % ID, percentage of injected dose;
HAMA, human antimouse antibody; AUC, area under the curve; MTD, maximum
tolerated dose; SPECT, single-photon emission computed tomography; ANC,
absolute neutrophil count; CT, computed tomography; DLT, dose-limiting
toxicity.
Received 6/ 8/99; revised 2/ 7/00; accepted 2/16/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-expressing adult T-cell leukemia with yttrium-90-labeled anti-Tac. Blood, 86: 4063-4075, 1995.This article has been cited by other articles:
|
|
 |
|
  A. Khaibullina, B.-S. Jang, H. Sun, N. Le, S. Yu, V. Frenkel, J. A. Carrasquillo, I. Pastan, K. C.P. Li, and C. H. Paik Pulsed High-Intensity Focused Ultrasound Enhances Uptake of Radiolabeled Monoclonal Antibody to Human Epidermoid Tumor in Nude Mice J. Nucl. Med., February 1, 2008; 49(2): 295 - 302. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Scott, N. Tebbutt, F.-T. Lee, T. Cavicchiolo, Z. Liu, S. Gill, A. M.T. Poon, W. Hopkins, F. E. Smyth, C. Murone, et al. A Phase I Biodistribution and Pharmacokinetic Trial of Humanized Monoclonal Antibody Hu3s193 in Patients with Advanced Epithelial Cancers that Express the Lewis-Y Antigen Clin. Cancer Res., June 1, 2007; 13(11): 3286 - 3292. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Kelly, F. T. Lee, F. E. Smyth, M. W. Brechbiel, and A. M. Scott Enhanced Efficacy of 90Y-Radiolabeled Anti-Lewis Y Humanized Monoclonal Antibody hu3S193 and Paclitaxel Combined-Modality Radioimmunotherapy in a Breast Cancer Model J. Nucl. Med., April 1, 2006; 47(4): 716 - 725. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Jhanwar and C. Divgi Current Status of Therapy of Solid Tumors J. Nucl. Med., January 1, 2005; 46(1_suppl): 141S - 150S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. I. Milowsky, D. M. Nanus, L. Kostakoglu, S. Vallabhajosula, S. J. Goldsmith, and N. H. Bander Phase I Trial of Yttrium-90--Labeled Anti--Prostate-Specific Membrane Antigen Monoclonal Antibody J591 for Androgen-Independent Prostate Cancer J. Clin. Oncol., July 1, 2004; 22(13): 2522 - 2531. [Abstract] [Full Text] [PDF]
|
|
|
|
|
); SDS fraction
(
); and the intravascular fraction (
). The data are expressed as
the mean concentration from patients undergoing paired
111In and 90Y studies.