AKC'H Comparative Dual Label Study of First and Second … · with a K, of 2.5 X lo%-'. CC49 and...

[CANCER RkSL AKC'H 57. 271-278. January 15, 19931

Comparative Dual Label Study of First and Second Generation Antitumor-associated Glycoprotein-72 Monoclonal Antibodies in Colorectal Cancer patients1

Steven Gallinger, Raymond M. Reilly, Joel C. Kirsh, Robert D. Odze, Beverly J. Schmocker, Kazy Hay, Jeannie Polihronis, Manzur T. Damani, Baruch S h p i t ~ , ~ and Hartley S. Stern3 Dep(irttnents o f Slri;qen (S. G.. B. J. S., B. S., K. H., H. S. S.), Radiological Sciences, Division o f Nuclear Medicine ( J . C. K., M. T. D.), and Pathology (R. D. O.), Mount Sinai Hospitcil, trtzcl Department of Rctdiological Sciences, Division o f Nuclear Medicine (R. R., J. P ) , The Toronto Hospital, Universio of Toronto, Toronto, Ontario. Cr~nud(r. MSG I X 5

ABSTRACT

Radiolabeled first-generation anti-tumor-associated glycoprotein-72 (TAG-72) monoclonal antibody (MAb), B72.3, has proven useful in de- tecting primary and secondary colorectal carcinoma. It has been antici- pated that the development of second-generation, higher affinity, anti- TAG-72 MAbs, CC49 and CC83, would be of greater use in cancer detection and of value in radioimmunotherapy of human cancer. We compared the pharmacokinetics, biodistribution, and immune responses of 13'1-labeled CC49 and CC83 to 1251-labeled B72.3 by preoperatively coinjecting dual-labeled MAbs into 16 colorectal cancer patients. The imaging properties of radiolabeled CC49 and CC83 were also assessed. Pharmacokinetics of all three MAbs were identical, and there were no differences in the uptake of any of three MAbs in tumor and normal tissues. Maximum tumor uptake was 0.0041% of the injected doselg for lZ5I-B72.3, 0.0024% for l3l1-CC49, and 0.0029% for '3'I-CC83. Radiola- beled CC49 and CC83 detected most known tumor sites on scintigrams without any clear advantage for either MAb. Nonspecific splenic and testicular uptake was frequently observed. Anti-idiotypic human anti- mouse antibody responses were seen more frequently with B72.3 than with CC49 or CC83. We conclude that higher affinity, radiolabeled anti- TAG-72 MAbs can detect colorectal cancer but do not penetrate these tumors more effectively than B72.3. Improvements in tumor detection and radioimmunotherapeutic strategies will likely require the administration of smaller fragments of MAb molecules or novel delivery systems rather than the continued development of higher affinity MAbs.

INTRODUCTION

Approximately 150,000 new patients with colorectal cancer will be managed in the United States in 1992 (1). In the majority of these patients. pretherapeutic investigations will involve a series or combination of external imaging modalities including ultrasound, CT4 and magnetic resonance imaging. These investigations are considered es- sential as staging modalities to determine whether surgery is indicated and the type of surgery, as well as potential adjuvant chemotherapy and radiotherapy (2). Recently, there have been a number of reports suggesting clinical utility of immunoscintigraphy with MAbs for the management of patients with colorectal cancer (reviewed in Ref. 3). These initial results have provided impetus for the development of

Received 7/28/92 accepted 1 1/2/92.

higher affinity MAbs which could, at least theoretically, improve tumor imaging and permit their use in radioimmunotherapy of colorectal cancer.

Some of the best characterized and most extensively tested MAbs have been those directed against the TAG-72 antigen which is present on the majority of gastrointestinal, ovarian, breast, and non-small cell lung carcinomas (4, 5) . The prototype "first-generation" anti-TAG-72 MAb, B72.3, was generated by Colcher et al. (6) following the im- munization of mice with a membrane-enriched fraction from a human breast cancer metastasis. Clinical trials with radioiodine and indium- 1 1 1 -labeled B72.3 have provided evidence that (a) >70% of biopsy- proven tumors, some < 1.5 cm, are detectable by immunoscintigraphy or with an intraoperative hand-held gamma probe; (b) radiolocalization indices for malignant tissue are usually much higher than those for normal tissue; (c) elevated serum TAG-72 levels do not interfere with tumor imaging; and (d) radiolabeled B72.3 can be administered safely with few adverse effects (7-10).

Following the initial clinical successes with radiolabeled B72.3, the TAG-72 large molecular mass mucin (>lo6 daltons) was purified, characterized, and used to produce a number of second-generation anti-tumor MAbs referred to as CCMAbs (11-13). It was subse- quently shown that ( a ) each CCMAb recognizes a different epitope on TAG-72; (b) two of the CCMAbs, CC49 and CC83, possess affinity constants (K, ) 8- to 10-fold >B72.3; (c) CC49 and CC83 stain a greater percentage of tumor cells within more tumors than B72.3; and (d) in dual label studies, radiolabeled CCMAbs demonstrate higher RIs in human tumor xenografts in nude mice than B72.3 (13-16). The high affinity of CC49 for TAG-72 has been exploited in the successful radioimmunotherapy of LS174T tumors in mice using '77Lutetium and '"Iodine labeled CC49 (1 7, 18). Based on the results of tumor imaging with B72.3, it was expected that the in vitro and animal studies demonstrating the superiority of CC49 and CC83 would trans- late into improved imaging and radiolocalization of colorectal cancer lesions with these MAbs (18).

The purpose of our study was to compare radiolabeled second- generation anti-TAG-72 MAbs CC49 and CC83 with B72.3 in colorectal cancer patients. The clinical trial was designed to study the pharmacokinetics, biodistribution, and radiolocalization of the MAbs by preoperatively coinjecting radioiodine-labeled MAbs followed by blood sampling, scintigraphy, and tissue sampling during surgery. The

The costs of publication of this article were defrayed in part by the payment of page data obtained from this study are intended to clarify the current utility charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. of these antibodies in colorectal cancer and provide directions for

I Supported in part by the Dow Chemical Corp. Presented in part at the American future investigations with this modality. Association for Cancer Research Annual Meeting, San Diego, CA, 1992. ' B.S. is supported by the Lederman Cancer Research Foundation and an American

Physician Fellowship. MATERIALS AND METHODS 'To whom requests for reprints should be addressed, at Suite 449, Mount Sinai

Hospital. 600 I!niversity Ave., Toronto, Ontario, Canada M5G 1x5. Patients. All individuals recruited into the study were surgical candidates The abbreviations used are: CT, computerized tomography; MAb(s), monoclonal and had either primary or secondary colorectal cancer. ~h~~~ patients were

antibody(ies1; TAG-72, tumor-associated glycoprotein-72; CEA, carcinoembryonic antigen; K,, BSM, bovine submaxillary mucin; HAMA, human anti-mouse

being investigated for increasing serum CEA levels as pan of routine surveil-

antibody; Rl(s). radiolocalization index(es); %ID/g, % injected doselg ; FPLC, fast protein lance following a previous colon resection and Prior to ''second-look" surgery ltq~11 ' ,hromatography (19). There were 14 males and 2 females in the study (age range, 26-73 years).

Research. on September 10, 2021. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

COMPARISON OF ANTI-TAG-72 MONOCLONAL ANTIBODIES

Routine preoperative radiological investigations including CT scans, ultra- computer (Elscint 009). Overlapping images were obtained from the neck to sonography, magnetic resonance imaging, and barium studies were conducted to determine the extent of disease and to assess resectability. Informed consent was obtained from each patient, and the protocol was approved by the Human Ethics Committee of the University of Toronto. No patients had prior exposure to murine monoclonal antibodies.

Antibodies. Monoclonal antibodies B72.3, CC49, and CC83 were phar- maceutical grade and obtained from the National Cancer Institute, National Institutes of Health, under investigational drug protocol NCI T90-0167. B72.3 is a murine IgGl monoclonal antibody which recognizes the TAG-72 antigen with a K, of 2.5 X lo%-'. CC49 and CC83 are also IgGl antibodies which exhibit higher affinity for the TAG-72 antigen ( K , of 1.6 and 2.8 X 10" M-I, respectively) (1 3).

Radiolabeling of MAbs and Quality Control. Radiolabeling of B72.3 with lz5I and CC49 or CC83 with I3'I was performed using the Iodo-Gen technique (20). Briefly, the antibodies (1.1-1.6 mg) were incubated for 10 min at room temperature with 185-370 MBq of radioiodine (Frosst Radiopharma- ceuticals, Dorval, Quebec, Canada) in a sterile glass tube precoated with 20 yg of Iodo-Gen (1,3,4,6-tetrachloro-3a,6a-diphenylglyc0uri1; Sigma Chemical Co., St. Louis, Mo.). Radioiodinated antibody was purified from free radioiodine on a PD-10 Sephadex G-25M column (Pharmacia, Baie d'urfe, Quebec, Canada) eluted with 1% human serum albumin in saline. The purified, radioiodinated antibody was sterilized by filtration through a 0.22-pm Millex-GV filter (Millipore, Bedford, MA). The radioiodinated antibodies were tested retrospectively for sterility by the USP sterility test and for pyrogens by the limulus amebocyte lysate technique and by the USP pyrogen test. All antibody preparations were sterile and pyrogen free.

Radiochemical purity of Iz5I-B72.3 and 13'1-CC49 or CC83 was determined by size exclusion FPLC on a Superose- 12 HR column (Pharmacia) eluted with 0.05 M phosphate buffer, pH 6.5, at a flow rate of 0.4 mllmin and by paper chromatography (Whatman No. 1; Maidstone, England) in 85% methanol. The radiochemical purity of all antibody preparations was >95% with <5% immunoglobulin aggregates as determined by FPLC analysis.

Immunoreactivity of the labeled antibodies at infinite antigen excess was measured using polyvinylchloride microenzyme-linked immunosorbent assay plates (Immulon 2; Dynatech, Chantilly, VA) coated with BSM (Sigma). Ap- proximately 20 ng of labeled antibody were added to wells containing increasing amounts (0.5-10 pg) of BSM. After incubation for 4 h at 37"C, the wells were washed with 0.1 M Tris buffer, pH 8.3 and then cut and placed in counting tubes for gamma scintillation counting. The immunoreactive fraction was determined from a plot of total counts applied/counts bound versus IIBSM concentration according to the method of Lindmo et al. (21). The immunoreactive fraction ranged from 0.4-0.7.

Administration of MAbs. Twice daily, 5 drops of a saturated solution of potassium iodide were administered to patients starting 48 h prior to injection of the radioiodinated antibodies and continuing for 7 d after injection. Two i.v. lines were established, one for antibody administration and the other for blood sampling. Patients received a slow (3-5 min) i.v. injection of a mixture of IL5I-B72.3 and I3lI-labeled CC49 or CC83 in 2-3 mL of 1% human serum albumin diluted in 20 ml of NaCl solution. No adverse effects occurred following antibody injection.

Pharmacokinetics. Blood samples were obtained at 15, 45, 90, and 120 min and at 3.5,48,72, and 168 h postinjection. Aliquots of serum and a sample of the injectate were counted in an automatic gamma counter (Auto Gamma 5650; Packard, Downer's Grove, IL) using windows for lZ5I (15-80 keV) and "'I (260-470 key). Counts in the lZ5I window were corrected for crossover from I3'I. The concentration of Iz5I or 13'1 (kBq/liter) present in the serum at various times postinjection was corrected for radioactive decay back to the time of injection. The serum concentrations of the two radionuclides were then fit to a two-compartment pharmacokinetic model using PCNONLIN software (22). Serum samples were also analyzed by size exclusion chromatography on a Sephadex G-50 minicolumn eluted with NaCl solution to determine the percentage of radioiodine bound to protein at various times postinjection. Plasma volume was calculated from a nomogram (0.065 literkg of body weight) (23).

Immunoscintigraphy. On the day of MAb injection and again soon before surgery (6-8 days postinjection), the patients were imaged on a large field view camera (G. E. Maxicamera 500) interfaced to a dedicated nuclear medicine

27

the pelvis in both anterior and posterior projections, and lateral views were obtained as required. In some patients with known pelvic tumors, subtraction images were produced using 9"Tc-pentacarboxymethly diethylenetriamine to localize the bladder.

Images (128 X 128 matrix) were obtained for 10 minlview except for the abdominal images obtained at 1 week which were acquired for 20 minlview. Countslview ranged from 100,000 to >1,000,000. For display purposes, all images were normalized to a maximum of 255 countslpixel. A 9-point spatial filter was also applied. Images were assessed in a nonblinded fashion by a nuclear medicine physician.

Circulating TAG-72 Antigen Levels and Human Anti-Mouse Antibody Response. Serum samples obtained prior to injection of the radiolabeled antibodies were tested for levels of circulating TAG-72 antigen using the CA72-4 radioimmunoassay kit (Centocor, Malvern, PA) and for baseline HAMA levels using the TruQuant HAMA radioimmunoassay kit (Biomira, Edmonton, Alberta, Canada). Additional testing for HAMA was carried out on serum samples obtained at 2-6 weeks postinjection. The TruQuant HAMA kit is capable of measuring both anti-idiotypic and anti-isotypic HAMA response.

Tissue Biodistribution of MAbs. At the time of surgery, small biopsies of known or suspected tumor sites and normal tissues were obtained. The tissue samples were weighed, placed in gamma-counting tubes and Iz5I and I3'I radioactivity measured in an automatic gamma scintillation counter. The lZ5I counts were correcteci for crossover from I3lI and %ID/g present in each tissue sample calculated. RIs were calculated by dividing the %1D/g for tumor samples by the %ID/g for normal tissues.

Immunostaining. Immunohistochemical staining was performed on mul- tiple 5-pm thick formalin-fixed, paraffin-embedded sections of the primary tumor using the avidin-biotin-complex method as described previously (24). Representative sections of archival autopsy spleen and testes tissue were also analyzed by this technique. Briefly, serial sections from the paraffin-embedded blocks of tissue were deparaffinized in xylene, rehydrated in graded ethanol solutions, and treated with 0.3% hydrogen peroxide in 100% methanol for 15 min to block endogenous peroxidase activity. MAbs B72.3 or CC83 were incubated on the tissue sections for 30 min at room temperature (25°C) at dilutions of 11100, and 111000, respectively. A biotinylated horse anti-mouse IgG (H & L) secondary antibody was used, followed by the avidin-biotin- complex (Vectastain ABC Mouse IgG Kit; Vector Laboratories, Burlingame, CA). 3,3'-Diaminobenzidine tetrahydrochloride (Sigma) was used as a chro- magen. Representative sections of a colorectal mucinous adenocarcinoma, known to be positive for both antibodies, served as a positive control. A negative control omitting the primary antibody and one without any antibodies was performed in each case. Immunoreactivity of the tumor samples was scored according to the following grading scheme: 0, no tumor cell staining; +, occasional ( < I % ) staining; ++, 1-2596 staining; +++, 25-50% staining; and ++++, >50% staining. Each case was evaluated by a single gastrointestinal pathologist unaware of the immunoscintigraphy results.

RESULTS

Patient Characteristics. Sixteen patients with primary or secondary colorectal cancer were entered in the study. Their pertinent clinical data are summarized in Table 1. Eight patients were received I3'I- CC49 and Iz5I-B72.3, and 8 patients received '311-CC83 and Iz5I- B72.3.

Pharmacokinetics. In order to directly compare first- and second- generation anti-TAG-72 MAbs, we believed that it was critical to coinject dual-labeled MAbs in order to eliminate interpatient variability which could skew the pharmacokinetic evaluation of any of the three individual MAbs. The plasma concentration versus time profile for each of the three MAbs B72.3, CC49, and CC83 was best described by a two-compartment pharmacokinetic model with i.v. bolus input. In dual isotope studies, there were no differences between the MAbs in the values for various pharmacokinetic parameters (Table 2). The mean distribution half-life ranged from 1.3-2.0 h for the three MAbs. The mean elimination half-life was 48.3-50.2 h for B72.3,

2




Table 1 Patient characteristics, antibodies administered, and clinical data

Age Antibodies Patient (yr) Sex injected" Clinical data

1 69 M CC49/B72.3 Hepatic metastases 2 60 M CC49/B72.3 Pelvic recurrence and lung metastases 3 73 M CC49/B72.3 Rectal and splenic flexure primaries, liver

metastases 4 73 M CC49/B72.3 Cecal primary and liver metastases 5 64 M CC49/B72.3 Pelvic recurrence 6 50 M CC49/B72.3 Increasing CEA 22 mo after colon resection 7 28 M CC49/B72.3 Solitary liver metastasis 8 72 M CC49/B72.3 Increasing CEA 9 mo after colon resection 9 26 M CC83/B72.3 Rectosigmoid primary

10 41 M CC83/B72.3 Increasing CEA 12 mo after colon resection 11 38 M CC83/B72.3 Pelvic recurrence and liver metastases 12 49 F CC83/B72.3 Rectal primary 13 55 M CC83/B72.3 Rectal primary 14 64 F CC83/B72.3 Rectal and sigmoid primaries 15 51 M CC83/B72.3 Massive cecal primary 16 65 M CC83/B72.3 Solitary liver metastasis a B72.3 radiolabeled with 125I; CC83 and CC49 radiolabeled with 131I.

Table 2 Pharmacokinetics of 1251-B72.3, 1311-CC49, and 13t!-CC83

Volume distribution (liters/kg) Systemic

Distribution Elimination Central clearance half-life (h) half-life (h) compartment Steady state (ml/min)

B72.3 CC49 B72.3 CC83

2 .0 • 50.2• 1 6 . 5 0.060• 0.080-+0.016 1.7• 2.0 • 3.0 44.9 • 10.6 0.053 • 0.008 0.073 • 0.009 1.7 • 0.6 1.6 • 1.4 48.3 • 8.8 0.058 - 0.013 0.071 • 0.015 1.3 _+ 0.3 1.3 • 1.2 53.6 • 14.3 0.059 _+ 0.010 0.079 • 0.015 1.3 + 0.7

a Mean • SD.

44.9 h for CC49, and 53.6 h for CC83. The systemic clearance ranged

from 1.3-1.7 mL/min. Elevated levels of circulating TAG-72 had no

effect on the plasma disappearance rate of B72.3, CC49, or CC83. The

volume of distribution of the central compartment was very similar to

the plasma volume in each patient, indicating that the MAbs distrib-

uted initially in the plasma compartment. The volume of distribution

at steady-state was 22-37% larger than the volume of distribution of the central compartment, indicating that at later times the MAbs were

distributed in a second peripheral compartment. Although there were

minor apparent differences between the three MAbs when comparing each group of eight patients, the dual-labeled MAbs displayed almost

identical pharmacokinetics within each patient. In fact, the plasma

elimination curves were parallel and often superimposable for each of two MAbs within each patient.

Size exclusion chromatography of plasma samples using a Sepha-

dex G-50 minicolumn demonstrated that > 9 5 % of plasma radioac-

tivity was protein bound at all times after injection. l m m u n o s c i n t i g r a p h y . Although t25I-B72.3 and t3JI-CC49 or

13tI-CC83 were coinjected into patients, only the immunoscinti-

graphic properties of the second-generation antibodies were evalu-

ated. This was carried out using a gamma camera calibrated for the 364-keV photopeak of 13~I. The very low energy 28- to 35-keV

photopeaks of ~25I are not suitable for gamma camera imaging. Scin-

tigrams were acquired on the day of injection and 6-8 days later. These times were chosen based on the relatively long half-life of ~31I

(8 days) and our previous observations of the optimal time for imaging patients treated with ~3tI-B72.3 (25). MAb uptake in various tumors

and normal organs was evaluated from the scans and graded as fol-

lows: - , no uptake; +, moderate uptake; ++, high uptake.

Scans obtained on the day of MAb injection showed only the "blood pool." In fact, in patients with liver metastases, the metastases were usually photon deficient on the day of MAb injection similar to that seen with standard 99mTc-sulfur colloid liver-spleen scans. On

days 6-8 postinjection, primary and secondary tumor sites were often

clearly detectable with both 13JI-CC49 and 13~I-CC83. As shown in

Table 3, there were 3 true positive, 2 true negative, and two false-

negative scans in 7 evaluable patients within each group. One patient

was not evaluable in each group. Patient 5 had a large pelvic recur-

rence, and a focal area of increased tracer accumulation in the lower

abdomen was seen on his scan, but it could not be conclusively

determined whether this region represented tumor or radioactivity

within the bladder. The other patient (patient 16) was only scanned on

the third day after MAb injection because of scheduling difficulties.

Overall, liver metastases were most easily demonstrated on scintigrams with either 13II-CC49 or 131I-CC83 (patients 3, 4, and 11, Fig.

1, a and b); however, a 2-cm liver metastasis was not seen with CC49

(patient 7). Primary colorectal cancers were detected in both patients 3 and 4 with ~31I-CC49. ~31I-CC83 detected primary tumors in two

patients (patients 9 and 15, Fig. lc). However, rectosigmoid cancers

were not imaged in two other patients with CC83 (patients 13 and 14).

Extrahepatic secondary tumor was seen in patient 11 with CC83 but

not in patient 2 with CC49.

True negative scans occurred in 4 patients overall (patients 6, 8, 10, and 12). In three instances, these patients were being evaluated for

increasing serum CEAs as part of routine postoperative surveillance.

Tumor was not seen on the scintigrams, and during second-look

surgery, no intraabdominal cancer was found. The fourth patient with

a true negative scan (patient 12) was found to have a large villous

adenoma without invasive cancer. There were no clearly false-positive scans with t31I-CC49 or 1311-

CC83, although faint outlines of the transverse colon were occasion-

ally seen (Fig. lb). Of particular significance was the fact that the

spleen was easily seen on almost all scans with both MAbs (Fig. 2a). Similarly, pronounced testicular uptake was noted in scans from 10 of

14 male patients (Fig. 2b). Despite attempts to block thyroid uptake

with Lugol 's iodine, the thyroid gland was also visualized in some

patients. Although scintigrams often demonstrated areas of known tumor

sites based on clinical examination and other imaging modalities, the

MAb scans did not affect surgical decision making or actual patient

management .

Table 3 Radioimaging of colorectal cancer with 1311-CC49 or 131I-CC83

Uptake in normal tissues a

Patient b Liver Spleen Testes Thyroid Uptake in tumor c

131 I-CC49 1 + + + + - TP 2 + ++ ++ + FN 3 - + - ++ TP 4 + + + - TP 5 + ++ ++ - Not evaluable 6 + ++ ++ + TN 7 - ++ - - FN 8 + ++ ++ - TN

131 I-CC83 9 + ++ ++ + TP

10 + + + - TN 11 + + + - - TP 1 2 + + + + + T N

13 + + + + - F N

14 + + - FN 15 + + + - ++ TP 16 ++ + + - Not evaluable

a--, no uptake; +, moderate uptake; ++, high uptake. b Patients !-8 received 131 I-CC49 and 125I_B72.3; patients 9-16 received 131I.CC83

and 125I-B72.3. c Interpretation of nuclear scan with respect to tumor sites detected at laparotomy or

using other imaging modalities. TP, true positive; TN, true negative; FN, false negative,

273




a

b Fig. 2. a, posterior immunoscintigram of patient 11 (as in Fig. la) showing marked

splenic uptake (straight arrow) and uptake in liver metastases (curved arrow), b, anterior image from immunoscintigram of patient 13 given 13 q-CC83 showing intense activity in the testes (arrows).

Table 4 Biodistribution and radiolocalization of antibodies in tumor and normal tissues

Tumor a Bowel Omentum Muscle Peritoneum

%ID/g x 10 --4 (mean • SEM)

B72.3 12.7• 5.3_+1.3 1.8• 4 .8• 8.5_+6.6 CC49 10.0_+7.0 5.2• 3.1 b 4 .6• 1.6 13.0 CC83 12.0 • 2.4 4.6 _+ 1.0 0.9 • 0.2 3.9 • 0.9 6.4 • 2.8

Radiolocalization indices (mean • SEM)

B72.3 1.0 3.0 • 0~ 10.2 • 3.5 4.5 • 2.3 3.5 • 1.8 CC49 1.0 2.3 • 0.9 7.8 2.6 • 1.3 1.8 CC83 1.0 3.9• 1.5 13.9• 6.1 • 4 .0• 1.7

a "Tumor" tissue counted was from primary colectal tumor except patient 16 for whom tumor was a secondary liver metastasis.

b Single sample only.

normal serum TAG-72 levels. Moreover, as described earlier, the 4 false-negative scans were in patients with normal TAG-72 levels (<4 units/ml).

Human Anti-Mouse Antibody Responses. HAMA was evaluated in 12 patients in this dual-isotope study and in three other patients given 131I-B72.3 and two patients given 13tI-CC49 from an earlier pilot study (Table 5). The mass of MAb administered ranged from

0.4-2.8 mg for B72.3, 1.0-2.8 mg for CC49, and 0.4-2.6 mg for CC83. Anti-isotypic HAMA (directed primarily at the Fc region of murine immunoglobulin) developed in 7 of 17 patients (41%) tested. Anti-isotypic HAMA developed in 2 of the 3 patients given only B72.3 but in neither of the two patients treated with only CC49.

In 15 patients to whom B72.3 was administered, anti-idiotypic HAMA, directed primarily against the variable region of murine immunoglobulin, developed against the B72.3 MAb in 12 patients. In 5 of these, there was an anti-idiotypic response only, without a concom- itant anti-isotypic response. In the other 7 patients, both anti-isotypic and anti-idiotypic HAMA were detected. Of 10 patients given CC49, anti-idiotypic HAMA developed in only 3. Two of 4 patients given CC83 developed an anti-idiotypic response.

There appears to be a degree of cross-reactivity between the MAbs with respect to their potential to elicit anti-idiotypic HAMA. This is based on the observation of anti-idiotypic HAMA against CC49 in one patient given only B72.3. Similarly, in another patient given only CC49, anti-idiotypic HAMA was detected against B72.3. Moreover, three patients given ~25I-B72.3 and 13~I-CC49 developed anti-idiotypic HAMA against CC83.

Immunostaining. Both B72.3 and CC83 stained surgically excised primary and secondary colorectal cancers to variable degrees. As shown in Table 6, more tumors were positively stained with the second-generation MAb CC83 than with B72.3. In general, within each tumor examined there were a greater number of CC83-positive than B72.3-positive cells. Serum TAG-72 levels did not correlate with the number of positive cancer cells within the specimens examined. Interestingly, tumors excised from patients with false-negative nuclear scans were all positive for TAG-72 by immunostaining, and in 3 of 4 cases, CC83 immunostaining was ++++ within these samples. We retrospectively examined the original primary tumor excised from 3 patients being investigated for increasing CEA levels during postoperative follow-up (patients 6, 8, and 10). In one of 3 tumors, immunostaining was +++ when stained with B72.3 but negative otherwise. However, CC83 immunostaining was at least ++ for all 3 tumors studied.

Representative sections of three normal spleens and three testes were immunostained with B72.3 and CC83. None of the spleens stained positively with either antibody. However, two of three testes showed focal positivity of interstitial (Leydig) cells and occasional endothelial cells lining small capillaries.

Table 5 HAMA responses

Antibody Anti-isotypic administered n HAMA

B72.3 3 2 CC49 2 0 B72.3 + CC49 8 2 B72.3 + CC83 4 3 Total 17 7(41%)

Anti-idiotypic HAMA

B72.3 CC49 CC83

3 1 0 1 0 0 7 3 3 2 2 2

13(76%) 6(35%) 5(29%)

Table 6 lmmunostaining of tumors

% positive tumors when stained with:

Degree of positivity a B72.3 CC83

Negative 23 8 + 8 8 ++ 38 8 +++ 31 46 ++++ 0 30

'~ +, occasional positive cells (<1%); ++, 1-25% positive cells; +++, 25-50% positive cells; ++++, >50% positive cells.

275




DISCUSSION

In this study, we have compared first- and second-generation radiolabeled MAbs with respect to pharmacokinetics, biodistribution, and HAMA response in 16 colorectal cancer patients. Additionally, the tumor-imaging properties of the second-generation MAbs labeled with 131I were studied by nuclear scanning prior to surgery. Our results demonstrate that the pharmacokinetics of all three MAbs were identical in terms of plasma half-lives, systemic clearances, and vol- umes of distribution. The uptake of all three MAbs in colorectal cancer tissues were not different, and uptake of the three antibodies in the normal tissues sampled were also similar. As a result, radiolocalization indices for all three MAbs were almost identical. In general, administration of B72.3 generated more frequent anti-isotypic HAMA responses than CC49 and more frequent anti-idiotypic HAMA responses than CC49 and CC83. Immunoscintigraphy with ~3~I_CC49 or 131I-CC83 detected known tumor sites in most patients without any clear advantage for either of the two second-generation MAbs. Liver metastases were the lesions most clearly demonstrated.

The use of anti-TAG-72 MAbs for the detection of colorectal cancer was first described in a number of earlier clinical studies demonstrating that first-generation anti-TAG-72 MAb, B72.3, localizes reliably to tumor tissues in patients with advanced colorectal cancer (7-10). Since these reports, it has been assumed that the discovery of MAbs with greater affinity for TAG-72 would result in improved tumor imaging. Following the purification of the TAG-72 antigen, Schlom et

al. (12-16) produced and characterized a number of second-generation MAbs. CC49 and CC83 demonstrated Ka values considerably higher than B72.3 and localized more effectively to human colon cancer xenografts, as well as demonstrating greater reactivity to formalin-fixed human tumors. The next step in the analysis of CCMAbs required clinical evaluation of their efficacy in humans with cancer. Therefore, the major purpose of the present study was to assess the ability of two second-generation CCMAbs to bind, in vivo, to human colorectal cancers and to compare their binding with the prototype first-generation MAb, B72.3. We designed the study to specifically compare each CCMAb to B72.3 by coinjecting the MAbs labeled with different isotopes of iodine into colorectal cancer patients. This type of analysis eliminated interpatient variability and permitted more accu- rate comparisons of pharmacokinetics and biodistribution of the MAbs.

The plasma pharmacokinetics of the three MAbs, B72.3, CC49, and CC83, labeled with radioiodine were almost identical in patients with colorectal cancer. These findings contrast with those obtained in pre- clinical studies in nude mice bearing LS174T human colon cancer xenografts in which the blood clearance of CC49 was found to be more rapid than B72.3 or CC83 (15).

The results of our biodistribution studies were somewhat disap- pointing. Despite affinity constants of one order of magnitude greater for CC49 and CC83 compared with B72.3, in vivo tumor uptake in humans with colorectal cancer was the same for the first- and second- generation MAbs. These results are in sharp contrast to that which has been observed in nude mice bearing LS 174T human colon carcinoma xenografts. In the animal studies, > 10% of the injected dose of the second-generation MAbs was present per gram of tumor tissue. Also, the second-generation MAbs, CC49 and CC83, exhibited tumor uptake approximately 2-fold that of B72.3 with significantly lower distribution in normal tissues. As a result, it was found that in dual- labeling studies RIs in mice given CC49 or CC83 were approximately 5- to 10-fold higher than those for B72.3 (15). Our findings demonstrate the mean %ID/g of all three MAbs to be in the 10 -3-10 -4 range, as has been noted in other clinical studies with B72.3 alone (7, 8). It

appears that the uptake of a murine MAb into a s.c. tumor xenograft in a nude mouse bears little relevance to penetration of the same MAb into a solid tumor in vivo in humans. Tumor heterogeneity, necrosis, tumor size, interstitial pressure, and tumor vasculature have been suggested as playing a large role in preventing MAbs from penetrating human tumors to the same degree as in experimental xenograft models (26, 27). In our study, affinity for the tumor by the MAb does not appear to be the crucial factor in this process.

The generation of HAMA in humans given anti-tumor MAbs has important ramifications with respect to their potential immunothera- peutic role. Dose fractionation through repeated injections of MAbs will likely be necessary to reduce the toxicity of radioimmunotherapy (reviewed in Ref. 28). The development of HAMA has previously been shown to cause altered pharmacokinetics in subsequent MAb administrations (29). This altered pharmacokinetics will likely result in faster whole body elimination of radioactivity, decreased tumor uptake, and lack of therapeutic response in patients treated with radiolabeled MAbs. Studies of individual MAbs are therefore important in order to evaluate their specific immunogenicity. It has been shown that each anti-TAG-72 MAb recognizes distinct epitopes on the TAG-72 molecule (14). Therefore, the development of anti-idiotypic HAMA may be different for each MAb. The data from the present study suggest that B72.3 is somewhat more immunogenic than CC49 or CC83. Anti-idiotypic HAMA against B72.3 was found more com- monly than anti-idiotypic HAMA against CC49 or CC83. This finding is supported by recent clinical studies of chimeric B72.3 in which it was found that antibodies against the murine V-region of the chimeric B72.3 were frequent (7 of 12 patients), resulting in rapid elimination of MAb following a second injection (30). The development of anti- idiotypic HAMA directed against B72.3 in patients given only CC49 and vice versa suggests that anti-TAG-72 MAbs possess a degree of cross-reactivity in their ability to generate HAMA.

The ability of radiolabeled anti-TAG-72 MAbs to detect primary and secondary colorectal cancer is a potentially important application of hybridoma technology (7-10). Our study is the first to demonstrate the tumor-imaging properties of radiolabeled CC49 and CC83. Al- though primary and secondary tumors were seen on nuclear scans, the images were not qualitatively any clearer than those seen in our previous study with ~3~I-B72.3 (25). The lack of increased tumor uptake by CC49 or CC83 compared to B72.3 as shown in our tissue- sampling data is presumably one of the main reasons for this negative ~ finding. We also found that tumors were detected most readily and with greatest clarity in patients with high serum TAG-72 levels. Therefore, circulating TAG-72 antigen does not appear to hamper the ability to detect cancer sites with radiolabeled anti-TAG-72 MAbs, and in fact the opposite appears to be true.

The high level of splenic uptake noted on the scans in this study with ~ 31 I-CC49 and ~3 ~ I-CC83 was unexpected. Immunoscintigrams with ~3~I-B72.3 occasionally demonstrate nonspecific splenic uptake; however, we observed qualitatively greater splenic activity in the scans with CC49 and CC83 compared to that seen in our previous study with B72.3 (25). Other investigators have reported RIs >3 in histologically normal spleen using B72.3 and have hypothesized that this may be due to high levels of circulating immune complexes (7, 8). In our study, we did not detect immune complexes by FPLC in the serum of patients given anti-TAG-72 MAbs. In some of our patients, testicular uptake was also seen with the second-generation MAbs. We cannot definitively explain this finding; however, it is possible that the second-generation anti-TAG-72 MAbs recognize a "normal" antigen on testicular tissue, specifically on Leydig cells. Molinolo et al. (16) showed that endothelial cells occasionally stain positively with B72.3 and CC49, and this may partly explain the splenic and testicular

276




uptake in our study. Moreover, in the same report, testicular staining was also described with B72.3 and CC49. We cannot accurately quantify the amount of testicular uptake by the MAbs in our patients, but any significant binding of anti-TAG-72 MAbs to testicular tissue may have deleterious effects if these MAbs are to be used as a radioimmunotherapeutic approach to colorectal cancer.

Since the plasma elimination of the 13J I-labeled second-generation MAbs, CC49 and CC83, was very similar to radioiodinated B72.3, it is a reasonable assumption that the whole body radiation-absorbed dose associated with administration of these MAbs will also be similar to that of B72.3. We have previously estimated this to be 1.70 rads/ mCi (31). Severe toxicity to the bone marrow occurs when the whole body radiation-absorbed dose exceeds 200 rads. Thus, the dose of ~31i_labeled second-generation MAb administered for tumor therapy would be limited to no more than 117 mCi. The tumor uptake of the second-generation MAb observed in this study (no more than 0.004% ID/g) is not significantly higher than that observed for B72.3 or for most other murine anti-colon cancer MAbs currently under investi- gation. Most solid tumors require a radiation-absorbed dose to the tumor of at least 6000 rads for effective therapy (32). It has been estimated that a radiation-absorbed dose of 6000 rads will require a tumor uptake of at least 0.05% ID/g of 13 q-labeled MAb (32). Based on our observed tumor uptake data, it is likely that radioimmunotherapy with ~31I_CC49 or I3li-cc83 will not deliver adequate radiation- absorbed doses to tumor nodules and will therefore be no more effective in cancer therapy than other MAbs with lower in vitro

affinities. Advances in radioimmunotherapy are more likely to be realized through reducing toxicity to the bone marrow through alternative MAb delivery strategies. One possible strategy is the use of antibody fragments (Fab', F(ab')2, and Fv) which exhibit faster blood clearance than the whole immunoglobulin molecule (33, 34). Another approach is to deliver radioactivity to the tumor using a bifunctional antibody and radiolabeled hapten (35, 36).

In conclusion, we have compared the pharmacokinetics, biodistribution, and radioimaging properties of first- and second-generation anti-TAG-72 MAbs in colorectal cancer patients. We found no differences in pharmacokinetic, biodistribution, or radioimaging parameters between the three MAbs. It appears that the critical variables affecting the use of these MAbs may be factors other than antigen-antibody interactions as measured by affinity constants. Strategies to modify the

WIAbs, use different radioisotopes, or use alternative MAb delivery systems are necessary before they will be widely applicable in the clinical management of colorectal cancer patients.

ACKNOWLEDGMENTS

We thank Susan Levitt for excellent secretarial assistance and our patients for their participation and cooperation. Appreciation is extended to Dr. J. Schlom for assistance with this study and helpful advice.

REFERENCES

1. Silverberg, E. A., Boring, C., and Squires, T. Cancer Statistics, 1990. CA-Cancer J. Ctin., 40: 9-26, 1990.

2. Moertel, C. G., Fleming, T. R., MacDonald, J. S., et al. Levamisole and fluorouracil for adjuvant therapy for resected colon cancer. N. Engl. J. Med., 322: 352-358, 1990.

3. Goldenberg, D. M., and Larson, S. M. Radioimmunodetection in cancer identification. J. Nucl. Med., 33: 803-814, 1992.

4. Johnson, V. G., Schlom, J., Paterson, A. J., Bennett, J., Magnani, J. L., and Colcher, D. Analysis of a human tumor-associated glycoprotein (TAG-72) identified by monoclonal antibody B72.3. Cancer Res., 46: 850-857, 1986.

5. Thor, A., Ohucbi, N,, Szpak, C. A., Johnston, W. W., and Schlom, J. Distribution of oncofetal antigen tumor-associated glycoprotein-72 defined by monoclonal antibody B72.3. Cancer Res., 46: 3118-3124, 1986.

6. Colcher, D., Horan Hand, E, Nuti, M., and Schlom, J. A spectrum of monoclonal antibodies reactive with human mammary tumor cells. Proc. Natl. Acad. Sci. USA, 78: 3199-3203, 1981.

277

7. Esteban, J. M., Colcher, D., Sugarbaker, R, Carrasquillo, J. A., Bryant, G., Thor, A., Reynolds, J. C., Larson, S. M., and Schlom, J. Quantitative and qualitative aspects of radiolocalization in colon cancer patients of intravenously administered MAb B72.3. Int. J. Cancer, 39: 50-59, 1987.

8. Colcher, D., Esteban, J. M., Carrasquillo, J. A., Sugarbaker, P., Reynolds, J. C., Bryant, G., Larson, S. M., and Schlom, J. Quantitative analyses of selective radiolabeled monoclonal antibody localization in metastatic lesions of colorectal cancer patients. Cancer Res., 47- 1185-1189, 1987.

9. Martin, E. W., Jr., Mojzisik, C. M., Hinkle, G. H., Jr., Sampsel, J., Siddiqi, M. A., Turtle, S. E., Sickle-Santanello, B., Colcher, D., Thurston, M. O., Bell, J. G. Farrar, W. B., and Schlom, J. Radioimmunoguided surgery using monoclonal antibody. Am. J. Surg., 156: 386-392, 1988.

10. Doerr, R. J., Abdel-Nabi, H., Krag, D., and Mitchell, E. Radiolabeled antibody imaging in the management of colorectal cancer. Ann. Surg., 214: 118-124, 1991.

11. Katari, R. S., Fernsten P. D., and Schlom, J. Characterization of the shed form of the human tumor associated glycoprotein (TAG-72) from serous effusions of patients with different types of carcinomas. Cancer Res., 50: 4885-4890, 1990.

12. Sheer, D. G., Schlom, J., and Cooper, H. L. Purification and composition of the human tumor-associated glycoprotein (TAG-72) defined by monoclonal antibodies CC49 and B72.3. Cancer Res., 48: 6811-6818, 1988.

13. Muraro, R., Kuroki, M., Wunderlich, D., Poole, D. J., Colcher, D., Thor, A., Greine, J. W., Simpson, J. E, Molinolo, A., Noguchi, P., and Schlom, J. Generation and characterization of B72.3 second generation monoclonal antibodies reactive with the tumor-associated glycoprotein 72 antigen. Cancer Res., 48: 4588-4596, 1988.

14. Kuroki, M., Fernsten, E D., Wunderlich, D., Colcher, D,, Simpson, J. E, Poole, D. J. and Schlom, J. Serological mapping of the TAG-72 tumor-associated antigen using 10 distinct monoclonal antibodies. Cancer Res., 50: 4872-4879, 1990.

15. Colcher, D., Minelli, M. E, Roselli, M., Muraro, R., Simpson-Milenic, D., and Schlom, J. Radioimmunolocalization of human carcinoma xenografts with B72.3 second generation monoclonal antibodies. Cancer Res., 48: 4597-4603, 1988.

16. Molinolo, A., Simpson, J. E, Thor, A., and Schlom, J. Enhanced tumor binding using immunohistochemical analyses by second generation anti-tumor-associated glycoprotein 72 monoclonal antibodies versus monoclonal antibody B72.3 in human tissue. Cancer Res., 50: 1291-1298, 1990.

17. Schlom, J., Siler, K., Milenic, D. E., Eggensperger, D., Colcher, D., Miller, L. S., Houchens, D., Cheng, R., Kaplan, D., and Goeckeler, W. Monoelonal antibody-based therapy of a human tumor xenograft with a 1771utetium-labeled immunoconjugate. Cancer Res., 51: 2889-2896, 1991.

18. Schlom, J., Eggensperger, D., Colcher, D., Molinolo, A., Houchens, D., Miller, L. S., Hinkle, G., and Siler, K. Therapeutic advantage of high-affinity anticarcinoma radio- immunoconjugates. Cancer Res., 52: 1067-1072, 1992.

19. Wilking, N., Petrelli, N. J., Hen'era, L., Holyoke, E. D., and Mittelman A. Abdominal exploration for suspected recurrent carcinoma of the colon and rectum based upon elevated carcinoembryonic antigen alone or in combination with other diagnostic methods. Surg. Gynecol. Obstet., 162: 465-468, 1986.

20. Fraker, P. J., and Speck, J. C. Protein and cell membrane iodination with a sparingly soluble chloramide 1,3,4,6-tetrachloro-3a,6a-diphenylglycouril. Biochem. Biophys. Res. Commun., 80: 849-857, 1978.

21. Lindmo, T., Boven, E., Cuttitta, E, Fedorko, J., and Bunn, P. A. Determination of the immunoreactive fraction of radiolabelled monoclonal antibodies by linear extrapola- tion to binding at infinite antigen excess. J. Immunol. Methods, 72: 77-89, 1984.

22. Metzler, C. M., and Weiner, D. L. PCNONLIN version 3.0. Lexington, KY: Statistical Consultants Inc., 1989.

23. Wyngaarden, J. B. Common laboratory values of clinical importance. In: J. B. Wyngaarden, W. McDermott, and P. B. Beeson (eds.), Cecil Textbook of MediCine, p. 2347. Philadelphia, PA: W. B. Saunders, 1979.

24. Hsu, M., Raine, L., and Fauger, H. Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabelled antibody (PAP) procedures. J. Histochem. Cytochem., 29: 577-580, 1981.

25. Stem, H., Reilly, R., Gallinger, S., Kirsh, J., DeAngelis, C., Papa, M., Hay, K., and Schmocker, B. Radioimmunodetection of colorectal cancer metastases with 1-131 labelled monoclonal antibody B72.3: a pilot study to determine efficacy of detection, and pharmacokinetics. Cancer Invest., 11: 129-134, 1993.

26. Jain, R. K. Physiological barriers to delivery of monoclonal antibodies and other macromolecules in tumors. Cancer Res., 50(Suppl.): 814s-819s, 1990.

27. Gallinger, S., Papa, M., Reilly, R. M., Xiang, J., Kirsh, J., Mullen, B. M., Stem, H. S., Hozumi, N., and Roder, J. C. Comparative biodistribution and antibody-dependent cellular cytotoxicity of native and heavy chain chimeric antibody. Mol. Biother., 3: 197-203, 1991.

28. Reilly, R. M. Radioimmunotherapy of malignancies. Clin. Pharm., I0: 359-375, 1991.

29. Stewart, J. S. W., Hird, V., and Snook, D. Intraperitoneal radioimmunotherapy for ovarian cancer: pharmacokinetics, toxicity and efficacy of I-131 labelled monoclonal antibodies. Int. J. Radiat. Oncol. Biol. Phys., 16: 405-413, 1989.

30. Khazaeli, M. B., Saleh, M. N., Liu, T. P., Meredith, R. F., Wheeler, R. H., Baker, T. S., King, D., Secher, D., Allen, L., Rogers, K., Colcher, D., Schlom, J., Shochat, D., and LoBuglio, A. E Pharmacokinetics and immune response of 13~ I-chimeric mouse/ human B72.3 (human 3,4) monoclonal antibody in humans. Cancer Res., 51: 5461- 5466, 1991.

31. Reilly R. M., Kirsh, J., Gallinger, S., Thiessen, J. J., Damani, M., Hay, K., Polihronis, J., Schmocker, B., Odze, R., Houle, S., and Stem, H. Compartmental analysis of the pharmacokinetics of radioiodinated monoclonal antibody B72.3 in colon cancer patients. Nucl. Med. Biol., in press, 1992.

32. Vaughan, A. T. M., Anderson, R, Dykes, E W., Chapman, C. E., and Bradwell, A. R.




Limitations to the killing of turnouts using radiolabelled antibodies. Br. J. Radiol., 60: 567-572, 1987.

33. Milenic, D. E., Yokota, T., Filpula, D. R., Finkelman, M. A. J., Dodd, S. W., Wood, J. E, Whitlow, M., Snoy, E, and Schlom, J. Construction, binding properties, metab- olism, and tumor targeting of a single-chain Fv derived from the pancarcinoma monoclonal antibody CC49. Cancer Res., 51: 63634371, 1991.

34. Bucheggar, E, Pelegrin, A., Delaloye, B., Bischof-Delaloye, A., and Mach, J-E Iodine-131-1abeled MAb F(ab')2, fragments are more efficient and less toxic than intact anti-CEA antibodies in radioimmunotherapy of large human colon carcinoma

grafted in nude mice. J. Nucl. Med., 31: 1035-1044, 1990. 35. Stickney, D. R., Anderson, L. D., Slater J. B., Ahlem, C. N., Kirk, G. A., Sch-

weighardt, S. A., and Frincke, J. M. Bifunctional antibody: a binary radiopharma- ceutical delivery system for imaging colorectal carcinoma. Cancer Res., 24: 6650- 6655, 1991.

36. Paganelli, G., Magnani, P., Zito, F., Villa, E., Sudati, E, Lopalco, L., Rossetti, C., Malcovati, M., Chiolerio, F., Seccamani, E., Siccardi, A. G., and Fazio, E Three-step monoclonal antibody tumor targeting in carcinoembryonic antigen-positive patients. Cancer Res., 51: 5960-5966, 1991.

278



1993;53:271-278. Cancer Res Steven Gallinger, Raymond M. Reilly, Joel C. Kirsh, et al. in Colorectal Cancer PatientsAntitumor-associated Glycoprotein-72 Monoclonal Antibodies Comparative Dual Label Study of First and Second Generation

Updated version

http://cancerres.aacrjournals.org/content/53/2/271

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/53/2/271To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



AKC'H Comparative Dual Label Study of First and Second … · with a K, of 2.5 X lo%-'. CC49 and...

Documents

Transcript of AKC'H Comparative Dual Label Study of First and Second … · with a K, of 2.5 X lo%-'. CC49 and...