In radioimmunoguided surgery (RIGS), a radiolabeled antibody is given i.v. before surgery and a hand-held gamma-detecting probe is used to locate tumor in the operative field. The rapid blood clearance and good tumor penetration of single-chain Fv antibodies (scFv) offer potential advantages over larger antibody molecules used previously for RIGS. A Phase I clinical trial is reported on RIGS with scFv (MFE-23-his) to carcinoembryonic antigen (CEA). Thirty-four patients undergoing surgery for colorectal carcinoma (17 primary tumors, 16 liver metastases, and 1 anastomotic recurrence) and 1 patient with liver metastases of pancreatic carcinoma received 125I-labeled MFE-23-his scFv(125I-MFE-23-his) 24, 48, 72, or 96 h before operation. 125I-MFE-23-his showed biexponential blood clearance with α and β half-lives of 0.32 and 10.95 h,respectively. The abdomen was scanned during surgery with a hand-held gamma detecting probe (Neoprobe Corp.). 125I-MFE-23-his showed good tumor localization; comparison with histology showed overall accuracy of 84%. Highest median ratios for tumor:normal tissue and tumor:blood were recorded 72 or 96 h after scFv injection for patients undergoing resection of liver metastases. High levels of radioactivity were found in the kidneys. Five patients had grade 1 fever, and three had a grade 1 rise in blood pressure according to the Common Toxicity Criteria. There was a significant correlation between these ratios and those measured in excised tissues using a laboratory gamma counter (P < 0.001). MFE-23-his scFv antibody localizes in CEA-producing carcinomas. The short interval between injection and operation, the lack of significant toxicity, and the relatively simple production in bacteria make MFE-23-his scFv suitable for RIGS.

Antibody targeting has potential for selective imaging or delivery of therapy. scFvs,3 composed of the variable heavy and light chains of immunoglobulin linked by a synthetic peptide, have a molecular weight of Mr ∼27,000 compared with Mr 150,000 for a whole IgG antibody. The low molecular weight gives rapid plasma clearance and faster tumor penetration, resulting in high tumor:blood ratios at early time points in animal models and in humans (1, 2).

Complete local resection of colorectal cancer is essential to prevent local recurrence and for long-term survival (3). Therefore, accurate determination of the presence of metastases and the absence of tumor in resection margins is important to obtain complete tumor resection or to identify patients with unresectable tumor for systemic or regional therapy. Conventional imaging and operative assessment by inspection and palpation frequently fail to detect small deposits of tumor, and yet these deposits are probably treated most effectively when they are small.

RIGS is based on the preoperative injection of a radiolabeled antitumor antibody and the intraoperative use of a hand-held GDP(Neoprobe 1000 instrument; Neoprobe, Columbus, OH) to detect radioactivity localized selectively in tumor deposits. The RIGS system has the potential to detect small tumor deposits because the proximity of the probe to tumor deposits exploits the inverse square law. The technique was first described at the beginning of the 1980s (4), and various antibodies have been used since then for RIGS including B72.3 and CC49 (5, 6), which are directed against tumor-associated glycoprotein-72 and A5B7, an anti-CEA antibody (7). Antibody localization was found in 60–100% of tumors; RIGS findings resulted in up to 50% in alteration of surgery. However, the slow blood clearance of whole antibodies requires an interval of up to 4 weeks before tumor:blood ratios are high enough to allow accurate discrimination of tumor from normal tissue, resulting in a delay before surgical resection of the cancer.

scFvs may provide a more satisfactory alternative because of their rapid uptake and clearance. MFE-23-his, the anti-CEA scFv used for this trial, is derived from a bacteriophage library (8) and has been shown to localize successfully in tumor deposits in a radioimmunoscintigraphy trial in humans (2). The aim of this study was to investigate the performance of this antibody in humans, the optimal time interval between administration and scanning, and its efficacy in localizing primary and metastatic tumor deposits in patients with colorectal cancer using RIGS.

Patients.

Thirty-five patients (17 males and 18 females; ages 41–81 years; median age, 62 years) scheduled to undergo surgery for primary or recurrent adenocarcinoma of the colon, rectum, or pancreas were entered into the study. Preoperative diagnosis was based on barium enema, colonoscopy, ultrasound, and computer tomography. Histology was obtained where appropriate. Seventeen patients underwent resection of their primary tumor, 1 patient underwent resection of an anastomotic recurrence, and 17 patients underwent resection of liver metastases. Details of the primary tumors and the anastomotic recurrence are given in Table 1, whereas details about patients undergoing resection of liver metastases are summarized in Table 2. Twenty patients had performance status 0, 12 had performance status 1, 2 had performance status 2, and 1 had performance status 3 (Eastern Cooperative Oncology Group).

Patients were excluded from the protocol if they had a history of atopic asthma or eczema, evidence of allergy to iodine,mouse protein, ampicillin and polymyxin, a positive response to intradermal administration of the antibody (wheal >5 mm in diameter),and if pregnancy could not be excluded. Written informed consent was obtained from the patient or his/her legal guardian prior to antibody administration. The thyroid was blocked with 50 mg of potassium iodide three times daily for 10 days, starting on the day before the antibody injection, and 200 mg of potassium perchlorate four times daily on the day of injection.

A license from the Administration of Radioactive Substances Advisory Committee (ARSAC) and a Doctors and Dentists Exemption from the Medicines Control Agency at the United Kingdom Department of Health were obtained. The study was approved by the Ethical Practices Sub-Committee.

Antibody.

MFE-23-his was selected from a phage library, cloned,and expressed in Escherichia coli(8). Purification was performed by immobilized metal affinity and size exclusion chromatography. A polymyxin gel was used to remove endotoxins (9). Clinical grade material was prepared and tested according to the Cancer Research Campaign operation manual (10, 11).

MFE-23-his was radiolabeled with 125I(median, 119 MBq; range, 57–285 MBq; Amersham, Little Chalfont, United Kingdom) using the Iodogen method (Pierce and Warriner, Chester, United Kingdom); unbound iodine was removed by gel-filtration chromatography. CEA-binding capacity was tested in samples administered to 32 patients using a CEA coupled Sepharose 4B column and incorporation of 125I by thin layer chromatography (TLC) (9). All samples of labeled antibody showed CEA binding (median, 77%; range, 54–93%) and incorporation of 125I incorporated into the protein (median,98.7%; range, 81.7–99.8%).

Detection of Radioactivity.

The GDP and the control unit were provided by Neoprobe Corp. Detection of radioactivity is based on a cadmium-telluride crystal,which works as a detector and a preamplifier. It converts the detected radioactivity into a digital readout on the portable control unit and into an audible signal. The audible signal is set at a threshold of 3 SD above normal tissue, which allows the surgeon to locate positive areas. The threshold for the control tissues is calculated by the control unit after taking a 5-s count (squelching).

Study Protocol and Patient Recruitment.

Patients received 1 mg of 125I-labeled MFE-23-his. The interval between injection and surgery was 24 h (five primary tumors), 48 h(five primary tumors, one anastomotic recurrence, and three liver metastases), 72 h (three primary tumors and seven liver metastases), and 96 h (four primary tumors and seven liver metastases).

Precordial counts were taken prior to surgery. At laparotomy the abdomen was scanned with the GDP after visual inspection and manual palpation of intraabdominal viscera. Counts that were 3 SD above normal adjacent tissue were regarded as positive. Tumor:normal tissue ratios for solid organs were determined by using the normal tissue of the investigated organ as reference (e.g., liver metastasis:normal liver), whereas aorta acted as background for the region of the suprapancreatic aorta, celiac axis, hepatoduodenal ligament, infrapancreatic aorta, aorta/vena cava below superior mesenteric artery, small bowel mesentery, right and left iliac hypogastric bifurcation, and pouch of Douglas. These areas were regarded as potentially lymph node-bearing areas. Blood pool radioactivity at the bifurcation of the aorta was taken to give the estimate of blood background levels least influenced by surrounding tissues. If 2-s counts at the bifurcation of the aorta were>25% higher than the precordial values, counts were taken in the region of the iliac arteries. Squelching was done on each organ separately. Tissue obviously infiltrated by tumor, as detected by palpation and inspection, and tissue with positive counts were resected or biopsied if possible.

Biodistribution counts were obtained from the primary tumor and normal colon (if applicable) and subsequently from the following organs: right and left liver (zone I); suprapancreatic aorta,pancreas, celiac axis, hepatoduodenal ligament, and spleen (zone II);infrapancreatic aorta, aorta/vena cava below superior mesenteric artery, small bowel mesentery, and right and left kidneys (zone III);and bladder, right and left iliac hypogastric bifurcations, pouch of Douglas, right and left ovaries and uterus (zone IV). After removal of the tumor, additional counts of the resection margins were taken. Approximately 20 min were allowed for the perioperative scanning procedure in each patient.

Blood was taken from each patient 2, 10, and 30 min and 1, 4,24, 48, 72, and 96 h after injection of the antibody and during the operation to calculate the percentage of injected activity/kg of blood and the half-life of the radiolabeled scFv. These values were used to determine the clearance half-lives of 125I-labeled MFE-23-his by fitting a biexponential model to the data using a nonlinear optimization method.

Specimens.

All specimens were sent for routine pathological examination. Results obtained by the GDP were compared with histology. In addition, 10-s ex vivo counts of tumor and adjacent normal tissue were taken with the GDP immediately after resection. Blood and tissue samples were weighed and counted in duplicate by a laboratory gamma counter (1470 Wizard; Wallac, Milton Keynes, United Kingdom). The percentage of injected activity/kg tissue and tumor:normal tissue and tumor:blood ratios were determined.

Storage phosphor plate technology (12) was used to assess the distribution of the antibody in sections of tumor. Three-μm formalin-fixed paraffin sections and 0.5–1-cm-thick slices of tumor were exposed to phosphor plates for 28 days and digitized with a phosphor plate reader (Model 425 Phosphorimager; Molecular Dynamics,Chesham, United Kingdom). Quantitation of the distribution was performed using Image Quant (Molecular Dynamics) and Interactive Data Language Software.

Statistics.

The Pearson’s correlation coefficient was calculated to compare the data obtained by laboratory gamma counting and RIGS. A Students t test was used for the comparison of two groups.

Biodistribution and Clearance: Tumor and Normal Tissue Radioactivity from RIGS.

Thirty-four patients had RIGS examination of the abdominal cavity; the remaining patient (patient 35) did not proceed to operation because of a respiratory infection. This patient was expected to undergo resection of hepatic metastases. The abdominal survey gave estimates of concentration of radioactivity for groups of patients studied at 24, 48, 72, and 96 h after administration in the primary colon and rectal carcinomas, liver metastases, and normal intra-abdominal organs and blood vessels (2-s counts/MBq injected). The median values for the group at each time point are given in Fig. 1, and ranges are in . Median tumor values of radioactivity exceeded those for blood background at all time points. They also exceeded those for the normal tissue in which they were located (colon or liver).

The highest levels of radioactivity were found in the kidney at 24 h, consistent with renal filtration of scFvs. Kidney levels fell below those in liver metastases but not in colonic primary tumor by 48 h. Liver, spleen, and pancreas had levels of radioactivity similar to or above those of tumor for 48 h, but these then fell more rapidly. They exceeded blood levels throughout. Other normal tissues followed blood clearance except where they were close to the kidney (celiac axis and hepatoduodenal ligament) or bladder (uterus and pouch of Douglas).

There was a significant correlation between blood radioactivity, as determined by RIGS counts, close to large blood vessels (the bifurcation of the aorta, hepatoduodenal ligament,suprapancreatic aorta, infrapancreatic aorta, aorta/vena cava below superior mesenteric artery, celiac axis, right iliac hypogastric bifurcation, left iliac hypogastric bifurcation, and small bowel mesentery) and counts at the main tumor site (P <0.0005). Counts in the spleen, pancreas, normal liver, normal colon,and kidneys also correlated with blood activity (P <0.0005). We also found a correlation between tumor counts and blood pool background, which was most significant at 24 h(P < 0.0025), became less pronounced at 48 h(P < 0.025) and 72 h (P < 0.01),and was no longer evident at 96 h.

RIGS Findings at the Main Tumor Site Compared with Histology.

Results of RIGS findings at the site of the primary tumor, anastomotic recurrence, and liver metastases in comparison with histology are summarized in Table 3. Eighty-two % of the lesions at the main tumor site were true positive,2% true negative, and 16% false negative. There were no false-positive findings, one adenoma with moderate to severe dysplasia was true positive (patient 31), and one mildly dysplastic adenoma was false negative (patient 2). Sensitivity for all time points taken together was 84%, with a predictive value of a positive 100% and accuracy 84%. Specificity and predictive value of a negative could not be calculated in a meaningful way because there was only one true negative site at which tumor was suspected and shown not to be present because of unexpected complete removal at colonoscopy prior to operation (patient 19). Comparison of the in vivo and ex vivo GDP counts showed a significant correlation(P < 0.001). The operation of one patient in the 96-h group (patient 15, Dukes’ stage B) was delayed to 144 h for medical reasons. This tumor was RIGS negative, whereas histology showed adenocarcinoma. In two patients, no specimen was obtained. One patient(patient 30) with an inoperable primary tumor and liver metastases received a defunctioning ileostomy, and in a second patient (patient 22), resection of liver metastases was abandoned after a frozen section of a lymph node in the region of the celiac trunk showed tumor infiltration. This lymph node was RIGS positive.

Results of probing the resection margin are available for 30 patients, 3 patients did not proceed to the planned resection, and in 2 patients, counting of the resection margin was not possible because of an emergency during the operation. The resection margin was RIGS negative in 27 patients. Three patients undergoing resection of the primary tumor showed a RIGS-positive resection margin, which was not confirmed by histology (patients 2, Dukes’ stage D; 8, Dukes’ stage B; and 9, Dukes’ stage B). The relevance of this remains unclear at the moment.

Findings in Lymph Nodes.

Comparison of histology and RIGS was possible for lymph node biopsies performed in four patients. Results are shown in Table 4. Two lesions were suspicious by palpation, whereas two biopsies were done because of difficulties in interpreting the figures obtained by RIGS. In three additional patients, the percentage of injected activity/kg of the involved lymph node and the lymph node:normal tissue ratio were determined (Table 4).

Laboratory Gamma Counting.

In the blood, the fractional blood clearance of 125I-labeled MFE-23-his was calculated by fitting a biexponential function to the levels of radioactivity determined in all of the blood samples from the patients (Fig. 2). The α and β phases were 0.32 and 10.59 h, respectively. In tumor and normal tissues, median cpm/g/MBq injected in tumor and normal tissue for each time point are shown in Fig. 3. Tumor selectivity of antibody is presented by expressing gamma counts of tissues as tumor:normal tissue ratios; this also permits comparison with RIGS data (Table 5). Tumor:blood, tumor:normal tissue, and normal tissue:blood ratios obtained by RIGS and the laboratory gamma counter were significantly correlated (P < 0.001).

The highest median tumor:blood and tumor:normal tissue ratios for patients undergoing resection of liver metastases rose to a peak of 7.6:1 and 4.3:1, respectively, at 96 h (Table 5). Median tumor:blood ratio of patients undergoing resection of liver metastases at 72 and 96 h was higher compared with primary tumors. However,there were only 5 patients with primary tumors included at these time points, whereas 13 had a resection of liver metastases. The tumor:blood ratio of the patient undergoing surgery 144 h after injection of the antibody was 0.6:1, the tumor:normal tissue ratio was 1.2:1, normal tissue:blood was 0.5:1, and the percentage of injected activity/kg tumor was 0.04.

Phosphor Imaging of Tissue Specimens.

Radioactivity in histological tissue sections used for phosphorimaging was found to be too weak to give useful results. However, when the whole specimen was exposed, highly radioactive regions correlated well with tumor-infiltrated areas of the specimen(Fig. 4).

Drug-related Toxicity.

Five patients developed fever grade 1 according to the CTC,and three patients were found to have an increase of blood pressure(CTC grade 1) within 70 min after injection of the antibody. One patient developed a generalized itch (CTC grade 2) 5 days after injection. It was considered that this was unlikely to be linked with the antibody injection.

The hypothesis that a high-affinity scFv antibody would offer advantages over whole antibody depends on rapid clearance from plasma and retention in tumor relative to normal tissues. The biexponential plasma fractional clearance showed a rapid α phase with a half-life of 0.32 h. The great majority of circulating radioactivity had cleared before the earliest time of operation at 24 h. The βhalf-life of 10.59 h is consistent with a minor degree of nonspecific binding of MFE-23-his to serum proteins and with reequilibration between extravascular spaces and plasma. Probe counts over intraabdominal tissues and vessels confirmed that the rapid blood clearance and showed retention in tumor relative to blood and other normal tissues was greater; ratios increased up to 96 h after injection. Higher counts in organs compared with the blood pool are likely to be attributable to a larger distribution volume.

As reported previously for radioimmunoscintigraphy (2), activity was retained in the kidney at levels similar to those of the tumors. This is consistent with glomerular filtration of the scFv, reabsorption, and retention in renal tubules. Other normal tissues retained antibody longer than in the blood, but in all cases tumor:normal tissue ratios increased up to 96 h. This selective uptake made RIGS possible from 24 to 96 h, and the high overall sensitivity and predictive value of a positive formed a sound basis for further studies aimed at locating occult disease at surgery.

Antibodies diffuse into the necrotic center of tumors with time (12), and scFv antibodies penetrate and clear rapidly because of their low molecular weight but have a reduced binding capacity because of their monovalency, which would favor early time points (13). However, retention in necrotic areas seems less critical for imaging than for treatment than high tumor:normal tissue ratios. Therefore, 72 and 96 h appear to be the most favorable times. This contrasts with intervals of 4–10 days, which were necessary to allow discrimination of tumor and normal tissue when whole anti-CEA antibody was used (7) and 3–4 weeks in studies using whole anti-tumor-associated glycoprotein antibodies like CC49 (14) and CC83 (15). The images obtained by radioluminography (Fig. 4) show good antibody localization in the primary tumor but also in lymph node metastases. However, we also detected some nonspecific uptake of MFE-23-his in lymph nodes, as shown in patient 14, who was found to have a higher percentage of injected activity/kg of uninvolved lymph node than in the primary tumors (Table 4). There were no RIGS data for this node.

Areas such as the suprapancreatic aorta, celiac axis,and hepatoduodenal ligament were regarded as potentially lymph node-bearing areas and were compared with counts taken at the bifurcation of the aorta. We detected a high frequency of counts 3 SD above the blood pool background at the bifurcation of the aorta. In a previous study using the whole IgG antibody CC83 (15),histological examination of RIGS-positive lymph nodes in the periportal region frequently failed to confirm tumor infiltration, although the positive predictive value in this study, which included patients with recurrent colorectal cancer, was 69%. In our study, these findings are most likely attributable to the aforementioned pharmacokinetics of the scFv antibody. The conventional criteria for RIGS described in studies for whole IgG are therefore not suitable for scFv antibodies, and we suggest that a more appropriate alternative would be to use these areas as their own reference, i.e., squelching is performed in each area, and hot spots are regarded as suspicious and biopsied.

This study shows for the first time that an anti-CEA scFv localizes selectively at the site of primary colorectal cancer and metastases, which can be detected by RIGS at intervals of 24–96 h between injection and scanning. MFE-23-his clears more rapidly from blood than from normal tissue and is retained in tumor. The data also support the use of MFE-23 as a targeting moiety of therapeutic molecules (16). However, further research is warranted to investigate the technique to allow detection of otherwise occult disease.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

        
1

This work was supported by the Cancer Research Campaign, the Neoprobe Corporation, the European Society for Medical Oncology, and the Ronald Raven Chair in Clinical Oncology Trust.

                
3

The abbreviations used are: scFv, single-chain Fv; RIGS, radioimmunoguided surgery; GDP, gamma-detecting probe; CEA,carcinoembryonic antigen; CTC, Common Toxicity Criteria.

Fig. 1.

Median 2-s RIGS counts/injected MBq in tumor and various normal tissues at 24, 48, 72, and 96 h. The range,including 70% of the maximum and minimum for each time point, is shown in .

Fig. 1.

Median 2-s RIGS counts/injected MBq in tumor and various normal tissues at 24, 48, 72, and 96 h. The range,including 70% of the maximum and minimum for each time point, is shown in .

Close modal
Fig. 2.

Fractional blood clearance of 125I-labeled MFE-23-his. Blood samples were measured by a laboratory gamma counter. The percentage of injected radioactivity/kg blood was calculated, and a biexponential function was fitted to these values. The α (0.32 h) and β (10.59 h) half-lives were derived from this line.

Fig. 2.

Fractional blood clearance of 125I-labeled MFE-23-his. Blood samples were measured by a laboratory gamma counter. The percentage of injected radioactivity/kg blood was calculated, and a biexponential function was fitted to these values. The α (0.32 h) and β (10.59 h) half-lives were derived from this line.

Close modal
Fig. 3.

Median cpm/g tissue/MBq injected in tumor,normal tissue, and blood at 24, 48, 72, and 96 h measured by laboratory gamma counting. Bars, 70% of the maximum and minimum. TC, tumor colon; COL, colon; BLO, blood; LM, liver metastases; LIV,liver.

Fig. 3.

Median cpm/g tissue/MBq injected in tumor,normal tissue, and blood at 24, 48, 72, and 96 h measured by laboratory gamma counting. Bars, 70% of the maximum and minimum. TC, tumor colon; COL, colon; BLO, blood; LM, liver metastases; LIV,liver.

Close modal
Fig. 4.

a and b,localization of 125I-labeled MFE-23-his in a hepatic metastasis exposed to a phosphorimager compared with photographs of the specimen. c and d, localization of 125I-labeled MFE-23-his in primary tumor with lymph node metastases exposed to a phosphorimager compared with photographs of the specimen.

Fig. 4.

a and b,localization of 125I-labeled MFE-23-his in a hepatic metastasis exposed to a phosphorimager compared with photographs of the specimen. c and d, localization of 125I-labeled MFE-23-his in primary tumor with lymph node metastases exposed to a phosphorimager compared with photographs of the specimen.

Close modal
Table 1

Patient characteristics: primary tumors/anastomotic recurrence

No. of patients
Location  
Cecum 
Ascending colon 
Transverse colon  
Descending colon 
Sigmoid colon 
Rectum 11 
Dukes’ stage  
Adenoma 
No tumor 
Histological grading  
Well differentiated 
Moderately differentiated 12 
Poorly differentiated 
Adenoma 
No tumor 
No specimen (bypass) 
Preoperative CEA values  
Normal 13 
Raised 
Not done 
No. of patients
Location  
Cecum 
Ascending colon 
Transverse colon  
Descending colon 
Sigmoid colon 
Rectum 11 
Dukes’ stage  
Adenoma 
No tumor 
Histological grading  
Well differentiated 
Moderately differentiated 12 
Poorly differentiated 
Adenoma 
No tumor 
No specimen (bypass) 
Preoperative CEA values  
Normal 13 
Raised 
Not done 
Table 2

Patient characteristics: liver metastases

No. of patients
No. of lesions  
1–3 
>3 
Histological grading  
Well differentiated  
Moderately differentiated 12 
Poorly differentiated  
Not done/no specimen 3/2 
CEA values  
Normal 
Raised 13 
No. of patients
No. of lesions  
1–3 
>3 
Histological grading  
Well differentiated  
Moderately differentiated 12 
Poorly differentiated  
Not done/no specimen 3/2 
CEA values  
Normal 
Raised 13 
Table 3

Summary of results (RIGS compared with histology)

True positive, No. (%)True negative, No. (%)False positive, No. (%)False negative, No. (%)
24 h     
PTa 4 (80)   1 (20) 
48 h     
PT 3 (33.4) 1 (11.1)   
AR    1 (11.1) 
LM 2 (22.2)   2 (22.2) 
72 h     
PT 3 (13.1)    
LM 19 (82.6)   1 (4.3) 
96 h     
PT 2 (14.4)   1 (7.1) 
LM 9 (64.3)   1 (7.1) 
144 h     
PT    1 (100) 
All time points 42 (82) 1 (2)  8 (16) 
True positive, No. (%)True negative, No. (%)False positive, No. (%)False negative, No. (%)
24 h     
PTa 4 (80)   1 (20) 
48 h     
PT 3 (33.4) 1 (11.1)   
AR    1 (11.1) 
LM 2 (22.2)   2 (22.2) 
72 h     
PT 3 (13.1)    
LM 19 (82.6)   1 (4.3) 
96 h     
PT 2 (14.4)   1 (7.1) 
LM 9 (64.3)   1 (7.1) 
144 h     
PT    1 (100) 
All time points 42 (82) 1 (2)  8 (16) 
a

PT, primary tumor; LM, liver metastasis; AR, anastomotic recurrence.

Table 4

Findings in lymph nodes

Patient/Tissue originIn vivo findingsLaboratory gamma counting
Suspicious by palpationRIGSHistology% injected dose/kg primary tumor% injected dose/kg lymph nodeTumor: normal tissue ratio, primary tumorTumor: normal tissue ratio, lymph node
Pat 1/Mesentery Yes Negative Negative 0.39 NA NAa 
Pat 10/Right iliac hypogastric bifurcation No Negative Negative 0.47 NA 5.4 NA 
Pat 11/Mesentery Yes (ex vivoNA Positive 0.26 0.31 1.7 1.6 
Pat 12/Hilum of kidney No Negative Negative 0.48 NA 1.4 NA 
Pat 13/Mesentery Yes (ex vivoNA Positive 0.33 0.94 0.8 
Pat 14/Mesentery NA NA Negative 0.49 0.68 1.56 2.5 
Pat 22/Celiac trunk Yes Positive Positive NA NA NA NA 
Patient/Tissue originIn vivo findingsLaboratory gamma counting
Suspicious by palpationRIGSHistology% injected dose/kg primary tumor% injected dose/kg lymph nodeTumor: normal tissue ratio, primary tumorTumor: normal tissue ratio, lymph node
Pat 1/Mesentery Yes Negative Negative 0.39 NA NAa 
Pat 10/Right iliac hypogastric bifurcation No Negative Negative 0.47 NA 5.4 NA 
Pat 11/Mesentery Yes (ex vivoNA Positive 0.26 0.31 1.7 1.6 
Pat 12/Hilum of kidney No Negative Negative 0.48 NA 1.4 NA 
Pat 13/Mesentery Yes (ex vivoNA Positive 0.33 0.94 0.8 
Pat 14/Mesentery NA NA Negative 0.49 0.68 1.56 2.5 
Pat 22/Celiac trunk Yes Positive Positive NA NA NA NA 
a

NA, not applicable.

Table 5

Comparison of RIGS findings and laboratory gamma counting

Time pointMedian tumor:blood ratio (range)Median tumor:normal tissue ratio (range)Median normal tissue:blood ratio (range)
RIGSLaboratory gamma countingRIGSLaboratory gamma countingRIGSLaboratory gamma counting
24 h       
Tumor colon 1.18 (1.04–1.38) 1.04 (0.95–1.27) 1.6 (1.26–2.22) 1.06 (0.93–1.9) 0.79 (0.47–0.97) 1.06 (0.58–2.06) 
48 h       
Tumor colon 1.41 (0.93–2.07) 1.45 (1.28–5.12) 1.56 (1.1–2.17) 1.72 (1.4–2.95) 1.03 (0.6–1.54) 0.91 (0.84–1.73) 
Liver metastases 1.17 (0.96–2.9) 1.8 (1.6–2) 1.06 (0.72–2.47) 1.94 (1.68–2.19) 1.17 (1.1–1.33) 0.91 (0.52–0.96) 
72 h       
Tumor colon 1.55 (1.52–1.58) 2.3 (2–2.6) 1.62 (1.5–2.85) 1.7 (1.56–1.86) 0.83 (0.56–1.01) 1.38 (1.07–1.69) 
Liver metastases 4.62 (1.75–9.68) 5.54 (3.33–17.62) 2.13 (1.46–4.4) 3.9 (1.4–5.4) 1.92 (1.21–4.6) 1.73 (1.02–4.5) 
96 h       
Tumor colon 1.46 (1.03–1.9) 1.48 (1.27–2) 1.78 (1.1–2.09) 1.98 (0.76–2.82) 0.91 (0.81–1.06) 1.01 (0.52–1.67) 
Liver metastases 5 (2.44–11.13) 7.56 (1.52–17.75) 3.33 (1.66–5.21) 4.29 (2.34–9.3) 1.79 (1.47–2.1) 1.87 (0.21–2.35) 
Time pointMedian tumor:blood ratio (range)Median tumor:normal tissue ratio (range)Median normal tissue:blood ratio (range)
RIGSLaboratory gamma countingRIGSLaboratory gamma countingRIGSLaboratory gamma counting
24 h       
Tumor colon 1.18 (1.04–1.38) 1.04 (0.95–1.27) 1.6 (1.26–2.22) 1.06 (0.93–1.9) 0.79 (0.47–0.97) 1.06 (0.58–2.06) 
48 h       
Tumor colon 1.41 (0.93–2.07) 1.45 (1.28–5.12) 1.56 (1.1–2.17) 1.72 (1.4–2.95) 1.03 (0.6–1.54) 0.91 (0.84–1.73) 
Liver metastases 1.17 (0.96–2.9) 1.8 (1.6–2) 1.06 (0.72–2.47) 1.94 (1.68–2.19) 1.17 (1.1–1.33) 0.91 (0.52–0.96) 
72 h       
Tumor colon 1.55 (1.52–1.58) 2.3 (2–2.6) 1.62 (1.5–2.85) 1.7 (1.56–1.86) 0.83 (0.56–1.01) 1.38 (1.07–1.69) 
Liver metastases 4.62 (1.75–9.68) 5.54 (3.33–17.62) 2.13 (1.46–4.4) 3.9 (1.4–5.4) 1.92 (1.21–4.6) 1.73 (1.02–4.5) 
96 h       
Tumor colon 1.46 (1.03–1.9) 1.48 (1.27–2) 1.78 (1.1–2.09) 1.98 (0.76–2.82) 0.91 (0.81–1.06) 1.01 (0.52–1.67) 
Liver metastases 5 (2.44–11.13) 7.56 (1.52–17.75) 3.33 (1.66–5.21) 4.29 (2.34–9.3) 1.79 (1.47–2.1) 1.87 (0.21–2.35) 
Appendix I.

APPENDIX I.

TissueMedian 2-s RIGS counts/MBq injectedRange including 70% of minimum and maximum
Tumor colon   
24 h 18.3 13.1–19.5 
48 h 4.3 2.4–9.1 
72 h 2.5 1.9–3.1 
96 h 1.3 0.7–1.6 
Liver metastases   
24 h — — 
48 h 8.8 4.4–18.22 
72 h 3.4 2.8–10.7 
96 h 2.4 1.3–4.6 
Bifurcation of the aorta   
24 h 17.5 12.5–19.9 
48 h 3.4 1.7–6.4 
72 h 1.4 0.7–3.6 
96 h 0.5 0.4–0.8 
Colon   
24 h 13.9 7.6–16.1 
48 h 3.3 1.7–5.6 
72 h 1.5 1–3.7 
96 h 0.4 0.3–1.3 
Liver   
24 h 18.7 12.6–21.5 
48 h 5.2 1.7–9 
72 h 1.9 1–7.5 
96 h 0.9 0.6–1.4 
Spleen   
24 h 14.1 11.9–15.6 
48 h 6.3 4.4–7.7 
72 h 1.7 1.1–5.2 
96 h 1.1 0.8–1.3 
Pancreas   
24 h 19.8 12.9–24.8 
48 h 7.3 3.4–11.5 
72 h 1.5 0.7–7.9 
96 h 0.7 0.5–1.1 
Kidney   
24 h 26.1 15.2–40.3 
48 h 8.2 5.1–14.4 
72 h 3.3 1.8–7.4 
96 h 1.7 0.6–2.1 
Hepatoduodenal ligament   
24 h 18.8 14–31.9 
48 h 4.8 2.4–8.5 
72 h 1.4 1–3.4 
96 h 0.5–1.8 
Suprapancreatic aorta   
24 h 21.1 11.5–26 
48 h 4.8 1.4–12.4 
72 h 1.9 0.8–11.7 
96 h 0.9 0.6–2.1 
Celiac axis   
24 h 20.1 12.4–30 
48 h 5.6 1.3–13.3 
72 h 1.5 0.7–8.1 
96 h 0.7 0.6–1.5 
Infrapancreatic aorta   
24 h 16.2 13.4–23.1 
48 h 3.8 2.2–7.4 
72 h 1.4 0.8–3.9 
96 h 0.7 0.5–0.9 
Small bowel mesentery   
24 h 14.5 7.2–17.6 
48 h 3.8 1.6–5.9 
72 h 1.2 0.6–3.3 
96 h 0.4 0.3–0.8 
Aorta/vena cava   
24 h 18.6 14–20.4 
48 h 3.3 2–5.4 
72 h 1.4 0.8–3.1 
96 h 0.5 0.4–1.5 
Right iliac hypogastric bifurcation   
24 h 15.6 10.2–20.3 
48 h 3.3 1.7–5.4 
72 h 0.6–3.2 
96 h 0.4 0.3–0.8 
Left iliac hypogastric bifurcation   
24 h 14.2 8.8–17.2 
48 h 2.9 1.6–5.7 
72 h 1.2 0.7–2.9 
96 h 0.5 0.4–3 
Uterus   
24 h 19.1 12.8–25.3 
48 h 7.5 5.3–9.7 
72 h 1–1.6 
96 h 0.7 0.6–0.7 
Right ovary   
24 h 13.5 8.7–18.3 
48 h 4.6 3.1–6.1 
72 h 0.9 0.8–1 
96 h 0.4 0.4–0.5 
Left ovary   
24 h 13.5 9.9–16.9 
48 h — — 
72 h 0.9 0.6–1.1 
96 h 0.4 0.4–0.5 
Bladder   
24 h 15.7 12.3–25.4 
48 h 5.6 4.5–8.1 
72 h 1.4 0.8–3.4 
96 h 0.6 0.5–1.1 
Douglas pouch   
24 h 13.5 10.1–17.9 
48 h 6.1 3.9–8.3 
72 h 1.1 0.7–1.5 
96 h 0.5 0.4–1.8 
TissueMedian 2-s RIGS counts/MBq injectedRange including 70% of minimum and maximum
Tumor colon   
24 h 18.3 13.1–19.5 
48 h 4.3 2.4–9.1 
72 h 2.5 1.9–3.1 
96 h 1.3 0.7–1.6 
Liver metastases   
24 h — — 
48 h 8.8 4.4–18.22 
72 h 3.4 2.8–10.7 
96 h 2.4 1.3–4.6 
Bifurcation of the aorta   
24 h 17.5 12.5–19.9 
48 h 3.4 1.7–6.4 
72 h 1.4 0.7–3.6 
96 h 0.5 0.4–0.8 
Colon   
24 h 13.9 7.6–16.1 
48 h 3.3 1.7–5.6 
72 h 1.5 1–3.7 
96 h 0.4 0.3–1.3 
Liver   
24 h 18.7 12.6–21.5 
48 h 5.2 1.7–9 
72 h 1.9 1–7.5 
96 h 0.9 0.6–1.4 
Spleen   
24 h 14.1 11.9–15.6 
48 h 6.3 4.4–7.7 
72 h 1.7 1.1–5.2 
96 h 1.1 0.8–1.3 
Pancreas   
24 h 19.8 12.9–24.8 
48 h 7.3 3.4–11.5 
72 h 1.5 0.7–7.9 
96 h 0.7 0.5–1.1 
Kidney   
24 h 26.1 15.2–40.3 
48 h 8.2 5.1–14.4 
72 h 3.3 1.8–7.4 
96 h 1.7 0.6–2.1 
Hepatoduodenal ligament   
24 h 18.8 14–31.9 
48 h 4.8 2.4–8.5 
72 h 1.4 1–3.4 
96 h 0.5–1.8 
Suprapancreatic aorta   
24 h 21.1 11.5–26 
48 h 4.8 1.4–12.4 
72 h 1.9 0.8–11.7 
96 h 0.9 0.6–2.1 
Celiac axis   
24 h 20.1 12.4–30 
48 h 5.6 1.3–13.3 
72 h 1.5 0.7–8.1 
96 h 0.7 0.6–1.5 
Infrapancreatic aorta   
24 h 16.2 13.4–23.1 
48 h 3.8 2.2–7.4 
72 h 1.4 0.8–3.9 
96 h 0.7 0.5–0.9 
Small bowel mesentery   
24 h 14.5 7.2–17.6 
48 h 3.8 1.6–5.9 
72 h 1.2 0.6–3.3 
96 h 0.4 0.3–0.8 
Aorta/vena cava   
24 h 18.6 14–20.4 
48 h 3.3 2–5.4 
72 h 1.4 0.8–3.1 
96 h 0.5 0.4–1.5 
Right iliac hypogastric bifurcation   
24 h 15.6 10.2–20.3 
48 h 3.3 1.7–5.4 
72 h 0.6–3.2 
96 h 0.4 0.3–0.8 
Left iliac hypogastric bifurcation   
24 h 14.2 8.8–17.2 
48 h 2.9 1.6–5.7 
72 h 1.2 0.7–2.9 
96 h 0.5 0.4–3 
Uterus   
24 h 19.1 12.8–25.3 
48 h 7.5 5.3–9.7 
72 h 1–1.6 
96 h 0.7 0.6–0.7 
Right ovary   
24 h 13.5 8.7–18.3 
48 h 4.6 3.1–6.1 
72 h 0.9 0.8–1 
96 h 0.4 0.4–0.5 
Left ovary   
24 h 13.5 9.9–16.9 
48 h — — 
72 h 0.9 0.6–1.1 
96 h 0.4 0.4–0.5 
Bladder   
24 h 15.7 12.3–25.4 
48 h 5.6 4.5–8.1 
72 h 1.4 0.8–3.4 
96 h 0.6 0.5–1.1 
Douglas pouch   
24 h 13.5 10.1–17.9 
48 h 6.1 3.9–8.3 
72 h 1.1 0.7–1.5 
96 h 0.5 0.4–1.8 

We thank Drs. R. B. Pedley, L. Robson, and D. A. Read for advice and technical assistance for the preparation of MFE-23-his. We are grateful to Drs. A. Green and M. Napier for the calculations for the Administration of Radioactive Substances Advisory Committee license and C. Pagonis, K. Bellenger, and the staff of the Drug Development Office, Cancer Research Campaign, for assistance in protocol development, data monitoring, and data analysis.

1
Verhaar M. J., Chester K. A., Keep P. A., Robson L., Pedley R. B., Boden J. A., Hawkins R. E., Begent R. H. J. A single chain Fv derived from a filamentous phage library has distinct tumor targeting advantages over one derived from a hybridoma.
Int. J. Cancer
,
61
:
497
-501,  
1995
.
2
Begent R. H. J., Verhaar M. J., Chester K. A., Casey J. L., Green A. J., Napier M. P., Hope-Stone L. D., Cushen N., Keep P. A., Johnson C. J., Hawkins R. E., Hilson A. J. W., Robson L. Clinical evidence of efficient tumor targeting based on single-chain Fv antibody selected from a combinatorial library.
Nat. Med.
,
2
:
979
-984,  
1996
.
3
Enker W. E., Laffer U. T., Block G. E. Enhanced survival of patients with colon and rectal cancer is based upon wide anatomic resection.
Ann. Surg.
,
190
:
350
-360,  
1979
.
4
Aitken D. R., Hinkle G. H., Thurston M. O. A gamma detecting probe for radioimmunodetection of CEA-producing tumors: successful experimental use and clinical case report.
Dis. Colon Rectum
,
27
:
279
-282,  
1984
.
5
Cohen A. M., Martin E. W., Lavery I., Daly J., Sardi A., Aitken D., Bland K., Mojzisik C., Hinkle G. Radioimmunoguided surgery using iodine 125 B72.3 in patients with colorectal cancer.
Arch. Surg.
,
126
:
349
-352,  
1991
.
6
Arnold M. W., Hitchcock C. L., Young D. C., Burak W. E., Bertsch D. J., Martin E. W. Intra-abdominal patterns of disease dissemination in colorectal cancer identified using radioimmunoguided surgery.
Dis. Colon Rectum
,
39
:
509
-513,  
1996
.
7
Dawson P. M., Blair S. D., Begent R. H. J., Kelly A. M. B., Boxer G. M., Theodorou N. A. The value of radioimmunoguided surgery in first and second look laparotomy for colorectal cancer.
Dis. Colon Rectum
,
34
:
217
-222,  
1991
.
8
Chester K. A., Begent R. H. J., Robson L., Keep P., Pedley R. B., Boden J. A., Boxer G., Green A., Winter G., Cochet O., Hawkins R. E. Phage libraries for generation of clinically useful antibodies.
Lancet
,
343
:
455
-456,  
1994
.
9
Casey J. L., Keep P. A., Chester K. A., Robson L., Hawkins R. E., Begent R. H. J. Purification of bacterially expressed single chain Fv antibodies for clinical applications using metal chelate chromatography.
J. Immunol. Methods
,
179
:
105
-116,  
1995
.
10
Begent R. H. J., Chester K. A., Connors T., Crowther D., Fox B., Griffiths E., Hince T. A., Ledermann J. A., McVie J. G., Minor P., Secher D. S., Schwartsmann G., Thorpe R., Wilbin C., Zwierzina H. Cancer Research Campaign operation manual for control recommendations for products derived from recombinant DNA technology prepared for investigational administration to patients with cancer in Phase I trials.
Eur. J. Cancer
,
29A
:
1907
-1910,  
1993
.
11
Mayer A., Chester K. A., Bhatia J., Pedley R. B., Read D. A., Boxer G. M., Begent R. H. J. Exemplifying guidelines for preparation of recombinant DNA products in Phase I trials in cancer: preparation of a genetically engineered anti-CEA single chain Fv antibody.
Eur. J. Cancer
,
34
:
968
-976,  
1998
.
12
Flynn A. A., Green A. J., Boxer G. M., Casey J. L., Pedley R. B., Begent R. H. J. A novel technique, using radioluminography, for the measurement of uniformity of radiolabelled antibody distribution in a colorectal cancer xenograft model.
Int. J. Radiat. Oncol. Biol. Phys.
,
43
:
183
-189,  
1999
.
13
Yokota T., Milenic D. E., Whitlow M., Wood J. F., Hubert S. L., Schlom J. Microautoradiographic analysis of the normal organ distribution of radioiodinated single-chain Fv and other immunoglobulin forms.
Cancer Res.
,
53
:
3776
-3783,  
1993
.
14
Arnold M. W., Schneebaum S., Berens A., Petty L., Mojzisik K., Hinkle G., Martin E. W. Intraoperative detection of colorectal cancer with radioimmunoguided surgery and CC49, a second-generation monoclonal antibody.
Ann. Surg.
,
216
:
627
-632,  
1992
.
15
Burak W. E., Schneebaum S., Kim J. A., Arnold M. W., Hinkle G., Berens A., Mojzisik C., Martin E. W. Pilot study evaluating the intraoperative localization of radiolabeled monoclonal antibody CC83 in patients with metastatic colorectal carcinoma.
Surgery
,
118
:
103
-108,  
1995
.
16
Michael N. P., Chester K. A., Melton R. G., Robson L., Nicholas W., Boden J. A., Pedley R. B., Begent R. H. J., Sherwood R. F., Minton N. P. In vitro and in vivo characterisation of a recombinant carboxypeptidase G(2): anti-CEA scFv fusion protein.
Immunotechnology
,
2
:
47
-57,  
1996
.