Indium 111 ZCE-025 Immunoscintigraphy in Occult Recurrent Colorectal Cancer With Elevated Carcinoembryonic Antigen Level Ralph J. Doerr, MD; Hani Abdel-Nabi, MD, PhD; Bruce Merchant, MD, PhD \s=b\ We investigated the utility of scanning with indium 111 labeled to monoclonal antibody in 13 patients after curative resection of colorectal cancer who had elevated carcinoembryonic antigen levels and negative results of clinical workup. Each patient received 1 mg of anti\p=n-\carcinoembryonicantigen monoclonal antibody type ZCE 025 labeled with 5.5 mCi of 111In, plus 9 to 39 mg of the same antibody unlabeled. Patients underwent scanning 3 to 7 days after infusion by planar and emission computed tomography. ZCE-025 monoclonal antibody imaging detected tumor recurrence or metastasis in 11 of 13 patients. In one patient the monoclonal antibody scan gave a true-negative result, and in one patient the monoclonal antibody scan failed to disclose a metachronous cecal primary. Tumor sites identified were the pelvis (2 patients), abdominal wall (2), retroperitoneum (1), lymph nodes (3), liver (2), bone (2), and lung (1). The accurate localization of colorectal carcinoma recurrences by means of 111In ZCE-025 monoclonal antibody demonstrates the usefulness of this diagnostic agent in the setting of elevated carcinoembryonic antigen level and negative results of clinical and radio-

logic workup. (Arch Surg. 1990;125:226-229)

gastrointestinal carcinoembryonic antigen That (CEA)-producing by intravenous tract tumors can be localized

exter¬

nal scintigraphy after injection of radiolabeled monoclonal antibody directed against CEA has been estab¬ lished." The ultimate clinical efficacy of this technique rests on this detection of minimum tumor deposits when conven¬ tional examinations fail to identify the lesion. This circumstance often exists in early recurrent colorectal carcinoma after curative resection. The opportunity to inter¬ vene surgically before clinically evident relapse may translate into prolonged survival" or enhanced palliation. Prospectively, we examined the sensitivity of tumor localization ofindium Ill-labeled anti-CEA monoclonal antibody (MoAb) ZCE-025 in patients who had previously undergone curative resection for colorectal cancer but who presented with elevated CEA levels and negative results of conventional diagnostic evalua¬ tion. With this antibody, we were able to localize colorectal carcinoma recurrences accurately. PATIENTS AND METHODS Thirteen patients, 5 female and 8 male, aged 46 to 78 years, were studied. All patients had undergone curative resection of adenocarcinoma of the colon and rectum before the study. The anatomic site of the tumor, the Astler-Collier modified Dukes' classification at the time of resection, and the interval after resection appear in Table 1. Inclusion in the study required that patients be asymptomatic, with

Accepted for publication October 14,1989. From the Departments of Surgery (Dr Doerr) and

Nuclear Medicine (Dr Abdel-Nabi), Veterans Administration Medical Center, Buffalo, NY, Buffalo General Hospital, and State University of New York at Buffalo; and Hybritech Inc, San Diego, Calif (Dr Merchant). Read before the annual meeting of the Society of Surgical Oncology, San

Francisco, Calif, May 24,1989. Reprint requests to Buffalo General Hospital, 100 High St, Buffalo, NY 14203 (Dr Doerr).

elevated serum CEA level (range, 3.6 to 36 ng/mL), and have nega¬ tive results of conventional workup, including physical examination, colonoscopy, abdominal computed tomographic (CT) scan, and chest roentgenogram. In all patients, serum levels of CEA were measured with CEA kits (Hoffmann-LaRoche Inc, Nutley, NJ) (normal values, *s2.5 ng/mL in 97% of healthy nonsmokers) and liver, kidney, and bone marrow functions were chemically evaluated before antibody infusion and 3 to 7 days after infusion. Anti-CEA MoAb ZCE-025 labeled with mIn (Hybritech Inc, San Diego, Calif) is derived from the same clone as MoAb 35.8,9 ZCE-025 is a murine antibody of the IgG subclass and was supplied (in vials) already bound to the chelating agent diethylenetriaminepenta-acetic acid by a modification of the bifunctional chelating technique.10 mIncitrate (antibody grade) was added to the vial, and the reaction was allowed to proceed for 30 minutes at room temperature, followed by the addition of 4.0 mL of neutralizing buffer solution. Commonly, 80% or more of the radioactivity was bound to the antibody, as judged

by thin-layer chromatography. This study was conducted under a US investigational new drug permit held by Hybritech Inc. Institutional review board approval and informed consent were obtained in all patients. All patients received 1 mg of ZCE-025 MoAb labeled with approximately 5.5 mCi of U1ln. Varying amounts of unlabeled ZCE-025 (9 to 39 mg) were

coinfused with the labeled MoAb in 100 mL of saline. Scintillation-camera planar images were obtained 72 to 96 hours after infusion with a large-field-of-view camera interfaced with a dedicated computer system. With a medium-energy collimator and the two photon peaks of mIn (174 and 247 keV), approximately 800 000 counts per view were collected. Data were stored in a 256 x 256 word-mode format. Single photon emission CT scans of the abdomen and pelvis with lnIn-ZCE-025 were obtained in all patients also within 72 to 96 hours after infusion. Acquisition time ranged from 40 to 60 minutes. Coronal and sagittal data sets were generated by pixel reorganization. Results of MoAb scans were made available to the surgeon before surgery. Operation was performed 4 to 21 days after completion of MoAb imaging. Resected specimens were examined by conventional histopathologic

techniques.

RESULTS No adverse reactions

were

observed after intravenous in¬

jection of niIn-labeled MoAb ZCE-025 in the 13 patients studied. No clinically significant alteration in bone marrow, liver, and kidney function was detected in the immediate (1 to 3 days) or remote (3 to 6 weeks) periods after ZCE-025 injection. In the 13 patients with elevated CEA levels after curative resection of colo-rectal adenocarcinoma, MoAb scans were positive for recurrence or metastasis in 11 patients, detecting more than one lesion in 2 patients. Results of MoAb scanning

and methods of confirmation of tumors are shown in Table 2. Second-look laparotomy 1 to 3 weeks after MoAb scanning confirmed the results of MoAb scanning in 6 patients (5 truepositive results and 1 true-negative result), for a sensitivity of 88%. Tumor was identified at the following locations: pelvis (two patients), abdominal wall (two), retroperitoneum (one),

Downloaded From: http://archsurg.jamanetwork.com/ by a Karolinska Institutet University Library User on 06/02/2015

Table 1 .—Patients With Colorectal Cancer After Curative Resection Time After Resection

Dukes'

Patient/Age, y/Sex

Site

Staging

1/49/M

Sigmoid

C1

2/66/F

Cecum

B2

3/76/M

Right colon

B2

4/51/M

Rectum

C1

9 mo

5/46/M

Rectum

B2

4

mo

6/50/F

Left colon

C1

10

mo

7/56/F

3

mo

C2

15

mo

9/71/M

Sigmoid Right colon Sigmoid

B2 C1

27

mo

10/55/M

Left colon

C2

24 mo

11/63/M

Right colon

12/55/F

Rectum

B1

22 y 5 mo

13/78/F

Rectosigmoid

8/63/M

3y 4y 5y

15y

Table 2.—Indium Ill-Labeled MoAb ZCE-025 Imaging in Patients With Colorectal Cancer After Curative Resection*

Patient 1

CEA at Time of MoAb Scan, 7.6

MoAb Scan Results Portal LN

CT scan

4.0

Retroperitoneal

CT-guided biopsy

ng/mL

Confirmation

mass

13.3 7.1

11.0

L4-5 Lesion

CT scan/bone scan

Pelvic recurrence, liver métastases

Intraoperative biopsy

Negative

Cecal adenocarcinoma,

Thoracic

CT scan

intraoperative resection

6

3.6

paraspinal LN 10.9

Pelvic recurrence, liver métastases

Intraoperative biopsy

16.7

Parapancreatic

Intraoperative biopsy

Abdominal wall

Intraoperative biopsy

LN

36

recurrence

10

72

Abdominal wall

Intraoperative resection

recurrence

biopsy

35.9

Bone métastases

Bone scan,

12

10.6

Negative

No tumor on

13

22

exploration Lung métastases Malignant pleural effusion

*MoAb indicates monoclonal antibody; CEA, carcinoembryonic antigen; LN, lymph node; and CT, computed tomography.

lymph nodes (three), liver (two), bone (two), and lung (one). Figure 1 demonstrates a MoAb-positive periportal lymph node in patient 1. Figure 2 shows interval CT scans of patient 1. Figure 3 shows pelvic recurrence and liver metastasis by MoAb imaging in patient 4.

Liver métastases were detected as focal areas of increased MoAb accumulation (hot spots) in two patients. Two patients with elevated CEA levels and negative results of clinical workup also had negative MoAb scans. At second-look explo¬ ration in one of these patients (patient 12), no evidence of locoregional recurrence or metastasis was identified despite

Fig 1.—Patient 1. Periportal lymphadenopathy positive for tumor (ar¬ row) by ZCE-025 monoclonal antibody scan.

positive results of conventional studies (CT, ultrasound). The resected right lobe of the liver in this patient was negative for tumor. The MoAb scan was false-negative in one patient (patient 5); adenocarcinoma of the cecum was found at subse¬ quent exploration, probably representing a metachronous primary tumor. Of the 13 patients in this study, 6 had MoAb scans positive for recurrent tumor and did not undergo second-look laparotomy. In 3 patients, tumors were confirmed by biopsy or needle aspiration cytologie examination. In the remaining 3 patients, MoAb-directed conventional imaging procedures (bone scan, CT scan) were subsequently consistent with re¬ currence in portal and mediastinal lymph nodes and lumbar vertebrae. Concurrent and subsequent CT scans remained negative for liver metastasis during 1 year of follow-up. COMMENT

The use of radiolabeled antibodies to detect cancer deposits is not new. More than 40 years ago, Pressman and Keighley11 used iodine 131-labeled antiserum from rabbits immunized by rat kidney cells to demonstrate antibody activity in nephrotoxic rat kidneys. In 1957, Pressman et al12 showed tumor localizing antibodies by paired labeling but were hampered by having to use polyclonal antisera and poorly defined tumor

antigens.

With the discovery by Gold and Freedman13 of CEA, the possibility of making an antibody to this tumor marker opened

for radiolabeled localization. clearer understanding of patterns of recur¬ rence after curative resection of gastrointestinal tract malig¬ nant neoplasms was being acquired. In colorectal cancer, Cass et al" reported that the degree of tumor anaplasia and the depth ofbowel wall penetration influenced the local recur¬ rence rate. After resection of adenocarcinoma ofthe colon and rectum, 60% of their patients presented with local recurrence alone, 26% had distant métastases as the initial site of recur¬ rence, and 14% had concomitant local recurrence and distant metastasis. Olson et al15 found a higher distant metastasis rate (46%) as the first evidence of recurrence after curative colo¬ rectal cancer resection, with more than half presenting as hepatic metastasis. Tumor size, stage, and site ofthe primary a new avenue

Concurrently,

Downloaded From: http://archsurg.jamanetwork.com/ by a Karolinska Institutet University Library User on 06/02/2015

Fig 2.—Patient 1. Left, Computed tomographic scan obtained in conjuction with monoclonal antibody Portal lymph node measures 1.3 cm (arrow), interpreted as negative. Right, Repeated computed tomographic scan obtained 6 months later shows Interval progression, with lymph node measuring 2 cm (arrow). tumor were predictive of relapse. Sugarbaker et al16 reported scan.

a cohort of patients with colon and rectal cancer thought to be at high risk for recurrence, ie, Dukes' class C cancers, perforated lesions, adjacent organ involvement, and age less than 30 years. He performed intensive follow-up studies. A CEA assay was performed monthly for 3 years, and abdominal CT scans were obtained every 4 months during that period. Half of this group of patients developed recur¬ rence, with 60% at more than one anatomic site. Eight-four percent had recurrences within the abdominal cavity, and 85% had had relapses within 30 months, with a median time of 17 months. A progressive rise in serial CEA level was the first

data from

and most sensitive indicator in 67% ofthe recurrences. Unfor¬ tunately, abdominal CT scans yielded little useful information as a

screening device, although

once

tumor

was

suspected,

the sensitivity was 85%, with a high false-positive rate (45%). Others7,17,18 have observed that elevated levels of CEA after curative resection of colorectal carcinoma were strongly pre¬ dictive of subsequent tumor recurrence. Refinements of the interpretation of CEA increase include both slope of rise analysis and serial monthly increments (as an absolute per¬ centage). Staab et al17 found that patients with CEA incre¬ ments with a flat ascending slope had local recurrences, whereas a steep ascending slope reflected liver metastasis. While there is ample confirmation of the value of serial postoperative CEA assays, controversy exists regarding therapeutic decisions based on an elevated CEA value. Given the magnitude of the problem of colorectal cancer in the United States, the question arises whether detecting a color¬ ectal recurrence earlier will impact significantly on survival. Second-look laparotomy on the basis of elevated postopera¬ tive CEA level was first proposed by Martin et al6 and Minton et al.19 Of their asymptomatic patients with a serial rise in CEA level, 55% were found at second-look laparotomy to have a resectable recurrence; 40% had unresectable tumors. Of the 5% (seven patients) in whom no tumor was found at exploration, six patients subsequently had recurrences. The 5-year disease-free survival was 25.8%. In 1985, a nonrandomized multicenter study corroborated this experience by re¬ porting 5-year survival of 30%20 for second-look laparotomy directed by elevated CEA levels. Taking this concept further, Sardi et al21 advocated multiple reoperations for recurrent colorectal carcinoma and demonstrated significant median survival improvement after second, third, fourth, and even fifth recurrences. Again, elevated CEA levels initiated the algorithm that eventuated in repeat surgery. Although oth-

Fig 3.—Patient 4. ZCE-025 monoclonal antibody scan demonstrating liver metastasis (arrow). ers7,22 have shown less survival benefit after second-look lapa¬ rotomy for matched-stage recurrences, it remains clear that

serial CEA measurement is the earliest indicator of tumor recurrence.

In this preliminary clinical trial, we combined elevated CEA levels with anti-CEA monoclonal antibody scintigraphy. The setting was asymptomatic patients with colorectal carcinoma undergoing curative resection who postoperatively demonstrated elevated CEA levels but negative results of physical examination and radiologie and colonoscopic

findings. In postoperative patients with colorectal cancer, we were able to detect multiple sites of tumor recurrence with planar and single photon emission CT MoAb imaging in 88% of surgically confirmed lesions. In three patients, positive MoAb scans initiated or directed more thorough conventional workup, resulting in tumor confirmation within 6 months of the original MoAb scan. In the two patients with negative MoAb scans, there was no evidence of recurrence or metasta¬ sis in one patient at exploration, and the other patient had a

Downloaded From: http://archsurg.jamanetwork.com/ by a Karolinska Institutet University Library User on 06/02/2015

false-negative result,

with

a

metachronous primary tumor

subsequently found in the cecum.

The formation of human antimouse antibody response was study. ZCE-025 induces antimouse re¬ sponse in approximately 40% of patients after a single infu¬ sion.23 Recently, we reported that repeated infusion of mInZCE-025 did not lead to an increased risk of adverse reactions or interfere with successful tumor detection, even in the presence of elevated human antimouse antibody titers.24 Mor¬ ton et al26 found an artifactual elevation of CEA level due to human antimouse antibodies. Methods to solve the problem of human antimouse antibody interference with serum CEA measurement have been devised.26 As has been previously observed,3 liver lesions can appear not addressed in this

on MoAb scans as

filling defects, as positive accumulations, or

combination of both. Efficacy for the detection of hepatic metastasis may be lessened by the intrinsic activity of mIn in the liver.27 However, liver lesions appearing as "hot spots" have been reported to occur in 25% to 40% of patients after the administration of varying amounts of cold unlabeled anti¬ body.2&i0 Recently, we reported the ability to detect liver metastasis as hot lesions after repeated infusions of this mon¬ oclonal antibody.81 The degree of differentiation of the tumor (whether mucin is produced or necrosis is present) also im¬ pacts on imaging accuracy. Abdel-Nabi et al,3 in reporting on mIn-ZCE-025 MoAb scanning for colorectal primary tumors, noted that metastatic involvement of pericolic lymph nodes was often missed due to the small size of lymph nodes and their proximity to the intensely imaged primary tumor. In our study we found that retroperitoneal and mesenteric meta¬ static adenopathy readily yielded positive scans. Particular as a

in the ability to detect tumor as a local Conventional imaging techniques (abdominal CT and magnetic resonance imaging) have been disappointing in their detection rates, particularly in the pelvis, where artifact with postoperative distortion and scarring can mimic

benefit

was seen

recurrence.

recurrence.

In this study we did not address the question of a therapeu¬ tic benefit from discovering colorectal cancer relapse by MoAb imaging before results with conventional diagnostic modalities become positive. Previous data on the sites of recurrence of colorectal cancer and the median time to devel¬ opment of recurrent disease in high-risk patients may require scrutiny in the face of a more sensitive tumor imaging test, such as MoAb scintigraphy. In accord with other studies,6"7,19,20 we believe that the earlier recurrent colorectal cancer is detected in the asymptomatic patient, the greater the num¬ ber of patients who will be disease free after reoperative surgery. The accurate localization of colorectal carcinoma recur¬ rences with mIn-labeled ZCE-025 MoAb demonstrates the usefulness of this diagnostic agent in the face of an elevated CEA level and negative clinical and radiologie workup. Mono¬ clonal antibody scintigraphy further offers the advantage of a whole-body image with a single examination. When the safety and efficacy of multiple dosings of radiolabeled monoclonal antibodies become firmly established, MoAb imaging will assume a place in the algorithm of the follow-up of the highrisk postoperative patient with colorectal cancer. The lead time to diagnosis may be shortened in this surveillance and could result in improved disease-free survival or earlier

palliation.

References Goldenberg DM, DeLand F, Evishin K, et al. Use of radiolabeled antibodcarcinoembryonic antigen for the detection and localization of diverse cancers by external photo scanning. N Engl J Med. 1978;298:1384-1388. 2. Dykes PW, Hine KR, Bradwell AR, et al. Localization of tumor deposits by external scanning after injection of radiolabeled anti-carcinoembryonic antigen. Br MedJ. 1980;280:220-222. 3. Abdel-Nabi HH, Schwartz AN, Goldfogel G, et al. Colorectal tumors: scintigraphy with In-111 anti-CEA monoclonal antibody and correlation with surgical, histopathologic, and immunohistochemical findings. Radiology. 1.

ies to

1988;166:747-752. 4. Mach FP, Carrel S, Forni M, et al. Tumor localization of radiolabeled antibodies against carcinoembryonic antigen in patients with carcinoma. N Engl J Med. 1980;303:5-9. 5. Attiyeh FF, Stearns MW. Second-look laparotomy based on CEA elevations in colorectal cancer. Cancer. 1981;47:2119-2125. 6. Martin EW, Minton JP, Carey LC. CEA-directed second-look surgery in the asymptomatic patient after primary resection of colorectal carcinoma. Ann Surg. 1985;202:310-317. 7. Steele G, Zamcheck N, Wilson R, et al. Results of CEA-initiated second\x=req-\ look surgery for recurrent colorectal cancer. Am J Surg. 1980;139:544-548. 8. Buchegger F, Haskell CM, Schreyer M, et al. Radiolabeled fragments of monoclonal anti-CEA antibodies for localization of human colon carcinoma

grafted into nude mice. J Exp Med. 1983;158:413-427. 9. Delaloye B, Bischof-Delaloye A, Buchegger F, et al. Detection of colorectal carcinoma by emission computerized tomography after injection of I-123 labeled Fab or F(ab1)2 fragments from monoclonal anti-carcinoembryonic antigen antibodies. J Clin Invest. 1986;77:301-311. 10. Frejcarek GE, Tucker KL. Covalent attachment of chelating groups to macromolecules. Biochem Biophys Res Commun. 1977;77:581-585. 11. Pressman D, Keighley G. The zone of activity of antibodies as determined by the use of radioactive tracers: the zone of activity of nephrotoxic antikidney serum. J Immunol. 1948;59:141-146. 12. Pressman D, Day ED, Blau M. The use of paired labeling in the determination of tumor-localizing antibodies. Cancer Res. 1957;17:845-850. 13. Gold P, Freedman SO. Specific carcinoembryonic antigens of the human digestive system. J Exp Med. 1965;122:467-481. 14. Cass AW, Million RR, Pfaff WW. Patterns of recurrence following surgery alone for adenocarcinoma of the colon and rectum. Cancer.

1976;37:2861-2865. 15. Olson RM, Perencevich NP, Malcolm AW, et al. Patterns of recurrence

following curative resection of adenocarcinoma of the colon and rectum.

Can-

1980;45:2969-2974. 16. Sugarbaker PH, Gianola FJ, Dwyer A, et al. A simplified plan for follow\x=req-\

cer.

up of patients with colon and rectal cancer supported by prospective studies of laparotomy and radiologic test results. Surgery. 1987;102:79-87. 17. Staab HJ, Anderer FA, Stumpf E, Fischer R. Slope analysis of the postoperative CEA time course and its possible application as an aid in the diagnosis of disease progression in gastrointestinal cancer. Am J Surg.

1978;136:322-327.

18. Denstuian F, Rosen L, Khubchandani IT, et al. Comparing predictive decision rules in postoperative CEA monitoring. Cancer. 1986;58:2089-2095. 19. Minton JP, James KK, Hurtubise PE, et al. The use of serum CEA determinations to predict recurrence of carcinoma of the colon and time for a second-look operation. Surg Gynecol Obstet. 1978;147:208-210. 20. Minton JP, Hoehn JL, Gerer DM, et al. Results of a 400 patient carcinoembryonic antigen second-look colorectal cancer study. Cancer. 1985;55:1284\x=req-\ 1290. 21. Sardi A, Minton JP, Nieroda C, et al. Multiple reoperations in recurrent colorectal carcinoma. Cancer. 1988;61:1913-1919. 22. Griffen WO, Humphrey L, Sosin H. The prognosis and management of recurrent abdominal malignancies. Curr Probl Surg. 1969;6:4-43. 23. Smith LM, Unger N, Bartholomew R. Anti-mouse responses to monoclonal antibody (MoAb) therapy in human patients. J Nucl Med. 1986;27:942. 24. Abdel-Nabi H, Smith L, Unger B, et al. Patterns of HAMA development following repeated infusions of In-111 ZCE-025 MoAb in recurrent colorectal carcinoma patients. J Nucl Med. 1989;5:906. 25. Morton BA, O'Connor-Tressel M, Beatty BG, et al. Artifactual CEA elevation due to human anti-mouse antibodies. Arch Surg. 1988;123:1242-1246. 26. Hansen HJ, LaFontaine G, Newman ES, et al. Solving the problem of antibody interference in commercial 'sandwich'-type immunoassays of carcinoembryonic antigen. Clin Chem. 1989;35:146-151. 27. Moldofsky PJ, Powe J, Hammond N. Monoclonal antibodies for radioimmunoimaging: current perspectives. In: Freeman LM, Weissman HS, eds. Nuclear Medicine Annual 1986. New York, NY: Raven Press Inc; 1986:57-105. 28. Abdel-Nabi H, Schwartz AN, Higano CS, et al. Colorectal carcinoma: detection with indium-111 anticarcinoembryonic antigen monoclonal antibody ZCE 025. Radiology. 1987;164:617-621. 29. Patt YZ, Lamki LM, Hayme TM, et al. Improved tumor localization with increasing dose of indium-111 labeled anti-carcinoembryonic antigen monoclonal antibody ZCE 025 in metastatic colorectal cancer. J Clin Oncol.

1988;6:1220-1230. 30. Lamki LM, Murray JL, Rosenblum MG,

et al. Effect of unlabeled monoclonal antibody (MoAb) on biodistribution of In-111 labeled MoAb. Nucl Med Commun. 1988;9:553-564. 31. Abdel-Nabi H, Doerr R, Roth SC, et al. Recurrent colorectal carcinoma detection with repeated infusions. J Nucl Med. 1989;30:748-749.

Downloaded From: http://archsurg.jamanetwork.com/ by a Karolinska Institutet University Library User on 06/02/2015