Phase I/II Study of G17-DT, an Anti-Gastrin Immunogen, in Advanced Colorectal Cancer1

Vol. 6, 4719–4724, December 2000 Clinical Cancer Research 4719

Andrew M. Smith, Timothy Justin, Dor Michaeli, using 100- and 200-␮g doses of G17-DT formulated without and Susan A. Watson2 adjuvant and the emulsifier aluminum monostearate. In addition, the effect of immunizing at different time intervals Academic Unit of Cancer Studies, University of Nottingham, Nottingham NG7 2UH, United Kingdom [A. M. S., T. J., S. A. W.], has been determined. and Aphton Corporation, Woodlands, California 95776 [D. M.] INTRODUCTION Colorectal cancer is the second most common cause of ABSTRACT cancer deaths in the United Kingdom, accounting for ϳ16,000 Gastrin is a growth factor for colorectal cancer, and deaths/year. Despite advances in surgery, chemotherapy, and therefore, anti-gastrin hormone therapy has a potential role radiotherapy, only modest improvements in overall mortality in treatment of this disease. The gastrin immunogen gastrin- have been achieved in the past 30 years. Cancer vaccines offer 17-diphtheria toxoid (G17-DT; Gastrimmune) produces an- a new potential treatment option. ti-G17 antibodies that have been shown to be effective in the Gastrin is a peptide hormone that is important in the treatment of colorectal carcinoma in preclinical models. regulation of acid secretion and the growth of both normal and Fifty patients with advanced colorectal cancer were treated malignant gastrointestinal epithelium (1), and it exists in two with G17-DT in a multicenter, sequential group, open label amidated forms composed of 17 and 34 amino acids, denoted Phase I/II study. Primary injections with two booster doses G173 and G34, respectively. Colorectal tumor cell lines and were given by i.m. injection. The main aim of the study was human cancers may express a number of isoforms of the gastrin/ to assess the safety and efficacy of the production of anti- CCKB receptors (2–8) and produce their own gastrin (9–13). gastrin antibodies. Locally developed and standard WHO Therefore, the tumor cells can potentially respond to circulating toxicity measurements with RIA and Scatchard analysis for endocrine gastrin (14–17) and the locally produced gastrin, antibody assessment were used. One center measured tumor which acts in an autocrine or paracrine manner (18, 19). The response radiologically. Eighty % of patients produced a tumor-associated gastrin is composed mainly of precursor measurable antibody response. Antibodies of high affinity forms, particularly glycine-extended G17, which have also been (median Kd, 0.295 nM; interquartile range, 0.16–0.41 nM) demonstrated to have a proliferative effect (20, 21). were detected between 4 and 12 weeks after primary injec- Elevation of the circulating gastrin level has been shown in ؋ ؊9 tion. The antigen binding capacity was high at 2.8 10 M animal models and humans to enhance the proliferation of the ؋ ؊10 ؋ ؊9 (interquartile range, 5.1 10 to 7.25 10 M). The normal colonic mucosa (20–23). As a result, there is an increase treatment was well tolerated with no systemic side effects in incidence of colorectal cancer in hypergastrinemic patients seen. Myalgia at the injection site was seen in 46% of (24). In addition, studies in the APCMin model of polyposis coli patients with severe pain caused by the formation of a sterile have shown hypergastrinemia induced by omeprazole adminis- abscess seen in 14% of patients. The abscesses were all tration and after infection with Helicobacter pylori promotes the drained under ultrasound guidance, and the patients recov- proliferation and progression of colorectal adenomas (25). Ef- ered fully within 6 weeks. No radiological responses were fective inhibition of gastrin may, therefore, be beneficial as a seen, but two patients had stable disease. G17-DT immuni- therapy in colorectal cancer, potentially in prevention, beneficial zation produces anti-G17 antibodies in patients with ad- as an adjuvant, or even beneficial in advanced disease. vanced colorectal cancer. The antibodies were of an affinity Gastrin/CCKB receptor antagonists have been assessed in high enough to compete with the cholecystokinin B/gastrin patients with gastrointestinal tumors. The antagonists appeared receptor for G17 binding with adequate capacity to neutral- to be ineffectual (26, 27), and it was initially assumed to be ize postprandial gastrin surges. Additional dose-ranging because of their lack of potency, with relatively high concen- studies have been performed in patients with gastric cancer trations required to displace amidated G17 (28, 29). The results of studies assessing tumor receptor profiles indicate that the lack of efficacy is probably also a result of the fact that several receptor subtypes mediate the action of gastrin (2, 8). Received 6/27/00; revised 9/7/00; accepted 9/13/00. A potential method to circumvent the problem created by The costs of publication of this article were defrayed in part by the the presence of numerous gastrin/CCKB receptor isoforms is to 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 Supported by Aphton Corporation, Woodlands, California. 2 To whom requests for reprints should be addressed, at Academic Unit 3 The abbreviations used are: G17, gastrin-17; CCKB, cholecystokinin of Cancer Studies, D Floor West Block, Queen’s Medical Centre, Univer- B; DT, diphtheria toxoid; AMS, aluminum monostearate; nor-MDP, sity Hospital, Nottingham NG7 2UH, United Kingdom. Phone: 44-(0)115- nor-muramyl dipeptide; CEA, carcinoembryonic antigen; CT, comput- 9709248; Fax: 44-(0)115-9709902; E-mail: [email protected]. erized tomogram.

neutralize gastrin before it interacts with its receptor. Gastrim- therapy; (h) no women of childbearing age; (i) no positive mune, an anti-gastrin immunogen, produces in situ neutralizing immediate hypersensitivity reaction to skin testing with antibodies directed against the forms of gastrin that have a Gastrimmune; and (j) adequate hemopoiesis (hemoglobin Ͼ10 proliferative action before they interact with the receptor. g/dl; WBC count Ͼ4.0 ϫ 109/liter; platelets Ͼ100 ϫ 109/liter). Gastrimmune is an immunoconjugate consisting of the nine Written informed consent was obtained according to institu-

NH2-terminal amino acids of human G17 linked via a peptide tional guidelines. spacer to DT. DT acts as the immunogenic carrier, and the G17 Gastrimmune Formulation and Administration. sequence acts as a B-cell epitope. The peptide spacer allows the Gastrimmune was supplied by Aphton Corporation (Wood- gastrin moiety to be orientated spatially in such a way that B lands, CA). Gastrimmune is a sterile, milky white, semiviscous cells recognize the whole sequence and thus raise high affinity, water-in-oil emulsion. The G17 immunogen was dissolved in a neutralizing antibodies (30, 31). PBS aqueous phase that was mixed with a surfactant and oil

Because the immunogen is composed of the NH2-terminal phase to form a water-in-oil emulsion. Three formulations of amino acids, the raised antibodies neutralize both amidated G17 Gastrimmune were used to determine whether they would affect and glycine-extended G17. In an animal model, after priming local tolerance and antibody production. The basic formulation injections, antibody titers were raised that were higher than included an emulsifier, AMS. The second formulation contained those required to neutralize all serum G17, and thus excess an adjuvant in the priming dose, nor-MDP, with the emulsifier unbound antibody was available for neutralization of tumor- AMS. The third formulation was a simple water-in-oil emulsion associated gastrin molecules (32). without adjuvant or emulsifier. By using active immunization with species-specific Gastrimmune was administered as an i.m. injection. The Gastrimmune in a pig model, immunoneutralization of G17 was primary injection and subsequent boosters were given at differ- shown to inhibit acid secretion without having any antiprolif- ent sites. The first immunization was injected into the left thigh, erative effects on the gastrointestinal mucosa, indicating that the second immunization was injected into the right thigh, and normal trophic effects of other gastrin species, such as G34, are the third immunization was injected into the left buttock. Each preserved (33). Antibodies raised by Gastrimmune have been dose of Gastrimmune was drawn up into a syringe immediately evaluated in a number of colorectal tumor models and have been before administration to minimize contact with plastic and pos- shown to exert strong therapeutic effects in primary tumors (34) sible adsorption of the peptide conjugate. and in both hepatic and lung metastasis models (35, 36). Drug Dosage and Treatment Schedule. Each patient After it was demonstrated that Gastrimmune has an anti- was to receive three injections of Gastrimmune. After a 1-week tumor effect in animal models, a Phase I/II study was performed assessment period, a primary injection was to be given, followed in patients with advanced colorectal cancer. The primary aim of by two additional immunizations spaced 4 weeks apart. The this study was to assess the safety and efficacy, in terms of dose level for each group of six patients was to escalate from antibody response, of Gastrimmune in patients with colorectal three injections of 10 ␮g of immunogen for the first group, to adenocarcinoma. Secondary aims were to assess the effect of 100 ␮g for the second group, to 250 ␮g for the third group. The formulation (the dose of Gastrimmune; the addition of an adju- fourth group was to receive 10 ␮g of immunogen with 3 ␮gof vant, nor-MDP, in the primary injection; and the presence of the emulsifier, AMS) on local tolerance and antibody production. nor-MDP adjuvant in the first injection only, and the fifth group ␮ ␮ The effect on tumor parameters was also determined. was to receive 100 g of immunogen with 3 g of nor-MDP adjuvant in the first injection only. An unadjuvanted formulation without the emulsifier AMS was then administered in an esca- PATIENTS AND METHODS lating dose, from 165 ␮g to the sixth group, 330 ␮gtothe Methods. The trial was designed as a multicenter, se- seventh group, 495 ␮g to the eighth group, and 990 ␮gtothe quential group, open label study. It was planned that 54 patients ninth group (Table 1). with advanced colorectal cancer would be admitted to the study Patient Monitoring. The patients were seen in specially in sequential blocks of six. The first patient entered in June designated clinics. Before the treatment commenced, histories 1995, and the last patient enrolled in October 1997. and physical examinations were performed, and the following Patient Selection. Patients, ages 18 years or older, with evaluations were obtained: (a) Karnofsky score; (b) full blood histologically verified adenocarcinoma of the colon or rectum count; (c) urea and electrolytes; (d) liver function tests; (e) 24-h that was either locally advanced, inoperable, recurrent, or met- creatinine clearance; (f) CEA; (g) immunoglobulins and autoan- astatic and that was not amenable to curative surgery and/or tibodies; (h) fasting G17 and anti-gastrin antibody; (i) chest radiotherapy, were candidates for this study. Other eligibility radiograph; and (k) CT scan (Nottingham patients only). criteria included: (a) life expectancy Ͼ3 months; (b) Karnofsky After each immunization, the patient was given a daily index Ն50%; (c) no other neoplasms apart from colorectal diary to assess local tolerance of the injection and interference cancer (except treated basal cell carcinoma of the skin or cancer with the activities of daily life. The patient was required to of the uterine cervix, stages 0–1); (d) no acute intercurrent record the pain at the injection site and how the discomfort from illness; (e) no patients considered to be a poor medical risk the injection interfered with the patient’s normal daily activities because of nonmalignant systemic disease or grossly abnormal on an ordinal scale from 0 to 10. The diaries were checked at laboratory results; (f) no patients who had received any other each visit. The investigator recorded an assessment of the degree anticancer therapy within 3 weeks; (g) no patients immunolog- of discomfort from the injection site and the degree of functional ically compromised, including those on systemic corticosteroid impairment attributable to the injection site that the patient

Table 1 The G17-DT dose groups Table 2 Volumes of individual components used in the The table shows the variation in composition of the immunogen radioimmunoassay of anti-G17 antibodies between groups. Volume (␮l) Immunization Immunogen Adjuvant Emulsifier Volume Dilution Sera Stripped Group schedule (␮g) (␮g) AMS (ml) buffer 125I-G17 dilution sera 1 10 0 Present 0.1 Backgrounda 100 0 0 100 1 2 10 0 Present 0.1 Total CPM 100 100 0 100 3 10 0 Present 0.1 Test sera 0 100 100 0 1 100 0 Present 0.1 Positive serum 0 100 100 0 2 2 100 0 Present 0.1 3 100 0 Present 0.1 a Radioactivity associated with tube assayed directly. 1 250 0 Present 0.25 3 2 250 0 Present 0.25 3 250 0 Present 0.25 1 10 3 Present 0.1 4 2 10 0 Present 0.1 Patients’ sera were diluted in Veronal buffer initially at a 3 10 0 Present 0.1 1:10 dilution and in a 3-fold dilution series thereafter. A positive 1 100 3 Present 0.1 control antiserum (Rabbit antihuman Gastrimmune) was also 5 2 100 0 Present 0.1 3 100 0 Present 0.1 prepared in buffer at the same dilutions. 1 165 0 Absent 0.05 Sera dilutions were then aliquoted into glass tubes that had

6 2 165 0 Absent 0.05 been coated previously with 2.5% polyethylene glycol (Mr 3 165 0 Absent 0.05 10,000; Sigma, Poole Dorset, United Kingdom). Stripped sera 1 330 0 Absent 0.1 7 2 330 0 Absent 0.1 were added to the control tubes. This was prepared from pooled 3 330 0 Absent 0.1 human serum that was added to charcoal that had been previ- 1 495 0 Absent 0.15 ously soaked overnight in 100 ml of 0.02 M Veronal buffer. This 8 2 495 0 Absent 0.15 was mixed and incubated at 4°C for 30 min, after which it was 3 495 0 Absent 0.15 centrifuged at 3000 ϫ g for 10 min. The supernatant was 1 990 0 Absent 0.3 9 2 990 0 Absent 0.3 collected and passed through a C18 Sep-pak filter that had been 3 990 0 Absent 0.3 washed previously with 10 ml of 50% acetonitrile [diluted in 0.02 M PBS (pH 7.2)] and 20 ml of 0.02 M Veronal buffer. After passage through the Sep-pak filter, the serum was collected, filtered through a 0.20 ␮m Millipore filter, and stored frozen at Ϫ20°C until use. The following tubes were set up as described experienced during the period since their last injection. In ad- in Table 2. dition, the investigator inspected the injection site. The tubes described in Table 2 were incubated for 72 h at The patient was seen at 2-week intervals for 12 weeks; 4°C (all tubes were prepared in triplicate). After the incubation, these visits included history and examination, full blood count, 100 ␮l of newborn calf serum (Sigma) and 250 ␮lof25% urea and electrolytes, liver function tests, CEA, and anti-gastrin polyethylene glycol (Mr 8000; Sigma) were added to each tube, antibodies. At 12 weeks, all of the investigations that were which were then vortexed. The tubes were then centrifuged at performed at the beginning of the study were repeated. Further 3000 ϫ g for 30 min at 4°C. The supernatant was aspirated, and follow-up was arranged at 8-week intervals. This allowed the the pellet from each tube was counted on a gamma counter (83% patient to air any problems and gave the investigator an oppor- counting efficiency). The mean of each triplicate was used, and tunity to evaluate long-term antibody response. each result was calculated as the percentage of 125I-labeled G17 This was performed pretreatment and posttreatment. The bound using the following equation: degree and site of disease was recorded. If possible, three to four marker lesions were measured at both times. All of the CT scans Test cpm Ϫ background cpm % 125I-labeled G17 bound ϭ were interpreted by a single consultant radiologist. Standard Total cpm/tube Ϫ background cpm WHO criteria for objective response assessment were used. Partial response was defined as a Ͼ50% reduction in the sum of Titer was evaluated from the titration curve. the products of the largest perpendicular diameters of all measurable disease sites. Progressive disease was indicated by a RESULTS Ͼ25% increase in the size of at least one measurable lesion or Fifty patients entered into the study. Thirty-three (66%) of the appearance of a new lesion. Stable disease was defined as an the patients were male, and 17 (34%) were female. The mean increase in disease measurements of Ͻ25% or a decrease of age at entry was 65.1 years, with a minimum age of 40 and a Ͻ50%. maximum age of 84. Karnofsky scores ranged from 60 to 100%, Anti-Gastrin Antibody Assay. Serum was obtained by with a mean of 84% and a median of 100%. Mean weight was centrifugation at 4°C after the formation of a clot and stored at 66.3 kg, with a minimum weight of 41 kg and a maximum Ϫ70°C until analysis. 125I-labeled G17 (New England Nuclear- weight of 106 kg. DuPont; specific activity, 2200 mCi/mmol) was reconstituted in Six patients died during the study from advanced disease. distilled water to a concentration of 500 Ci/ml. The radiolabeled Thirty-four patients received all three immunizations, 15 pa- G17 was diluted in Veronal buffer [0.02 M (pH 8.4)]. tients received two injections, and 1 patient received only a

Table 3 Number of immunizations received and the pain scores of each groupa Dose (␮g) emulsion 10 100 250 10 ϩ 3 100 ϩ 3 165 330 495 990 Volume (ml) 0.1 0.1 0.25 0.1 0.1 0.05 0.1 0.15 0.3 Group 1 2 3 4 5 6789 No. of patients receiving immunization One 000 1 0 1000 Two 100 1 1 2244 Three 5 6 6 2 5 4422 Local pain or reaction reported 0 5 2 2 4 2241 Maximum mean pain score 0.01 0.03 0.61 0.08 0.33 1.07 0.58 2.08 0.8 Maximum mean interference with daily 0.10 0.01 0.60 0.01 0.48 0.96 0.52 1.39 0.54 activities score No. abscesses/fluid collections 0 0 0 0 2 0140 a Pain score: 0, no pain; 10, worst pain. Interference score: 0, no interference; 10, completely inactive or bedbound.

single immunization. Failure to complete a course of immuni- Table 4 Positive antibody responders by group zations was because of either advancing disease (7 patients) or Dose as a result of a reaction at the local injection site (9 patients). Gastrimmune Adjuvant ␮ ␮ Assessment of Safety Group ( g) ( g) Responder % responder Systemic Effects. The immunogen was well tolerated. 1 10 2/6 34 2 100 5/6 83 Only two patients complained of nausea and vomiting that were 3 250 5/6 83 potentially related to treatment. No other systemic symptoms 4 10 Yes 3/4 75 were noted. Apart from hematological changes that would be 5 100 Yes 5/6 83 expected from advancing colorectal cancer, there were no sig- 6 165 5/6 83 nificant changes in the blood tests throughout the study. There 7 330 6/6 100 8 495 5/6 83 was no effect on renal function, as measured by serial urea and 9 990 4/4 100 electrolytes and by pretreatment and posttreatment 24-h creatinine clearance. Importantly, there was no change in immuno- globulin levels and no development of autoantibodies (parietal cells, smooth muscle, thyroid, intrinsic factor, reticulin, antinu- clear and mitochondrial antibodies, and rheumatoid factor) after recruited to group 9 (990 ␮g). No additional patients were administration of G17-DT. entered into the study. Local Tolerance. Twenty-four patients (46%) experienced a mild myalgia at the injection site after at least one of the Assessment of Efficacy injections. In most cases, the patient experienced mild discom- Anti-Gastrin Antibody Response. Using the criteria fort that lasted between 2 and 5 days and did not interfere with that 10% of 125I-labeled G17 binding was the threshold for a any activities of daily life. The mean pain score for all groups measurable response, 40 of the 50 patients produced antibodies was 0.54 (maximum, 10). There was no significant difference in after immunization with Gastrimmune. The response rate varied the pain scores among the groups or the formulations (Table 3). among the groups, as seen in Table 4. The lowest response rate The onset of the myalgia was variable and did not occur more was seen in group 1 (10 ␮g). frequently after a particular immunization. When patients expe- All patients achieved a Ն10% 125I-labeled G17 binding rienced injection site pain, no abnormality could be found on between weeks 4 and 12. Differences in the formulation of examination in most cases. Gastrimmune did not alter the time to antibody formation.

However, a group of seven patients (14%) developed ster- At week 12, the Kd was 0.295 nM (interquartile range, ile abscesses at an immunization site. Before the detection of the 0.16–0.41 nM). The median antigen binding capacity of the abscess, there was severe pain and limitation of movement of antibodies was 2.8 ϫ 10Ϫ9 mol/liter (interquartile range, 5.1 ϫ Ϫ10 Ϫ9 the affected limb in every patient. The diagnosis was made after 10 to 7.25 ϫ 10 M). There was no significant difference in clinical examination and confirmed by ultrasound. The collec- the characteristics of the antibodies produced by different dose tions were effectively drained percutaneously and contained pus groups. Several patients have been followed-up after the com- in all cases. Microscopy of the sample revealed the presence of pletion of the study. Antibody titers are maintained, up to a pus cells; however, no growth was seen on the culture, and a maximum of 2 years. diagnosis of a sterile abscess was made. After drainage, the We have been able to measure the presence of excess patients made a complete recovery, which took up to 6 weeks. anti-G17 antibodies and excess amounts of bound G17. How- After the development of abscesses in the group receiving ever, measurement of free G17 in the presence of antibodies and adjuvant, no additional patients were recruited to the adjuvant gastrin/antibody complexes is technically difficult. Initially, formulation groups. The abscesses in group 8 (495 ␮g) did not several serum manipulations are required to try to separate the occur until relatively late in the study, and four patients had been free gastrin fraction, which appears to be small. The numerous

manipulations result in an assay that is not reproducible. Advice Although not a primary aim of the study, assessment of the was sought from Professor John Walsh, who believes that meas- efficacy of antibodies raised by Gastrimmune indicated that urement of the free fraction in vitro is technically difficult4 at there was no tumor regression, and stable disease was seen in this time. He suggested that an in vivo bioassay would be more only 2 of 14 patients evaluated. It is likely that gastrin neutral- effective. This has been performed in a rat gastric fistula model. ization may slow growth by the removal of a major proliferative The induction of anti-gastrin antibodies resulted in significant factor, and a slowing of tumor growth would not be measurable inhibition of G17-stimulated acid output. This bioassay con- by methods used in the present study. We believe that G17-DT firms the effectiveness of the antibodies to bind G17 (25). may have a role as an adjuvant in colorectal, gastric, and Disease Response. Although this was a Phase I/II study, pancreatic cancer. In addition, gastrin appears to be an important response data were collected. Four patients had lung metastases growth factor early in the development of colorectal cancer; it evident on their original screening chest radiograph. These had appears to be controlled, in part, by the APC gene. G17-DT, all progressed by the time of the poststudy examination. Two therefore, may have a role in the prevention of colorectal cancer patients developed lung metastases during the study period. in high risk cases. Metastatic burden was assessed by a CT scan. Sixteen patients had pretreatment CT scans, and 14 patients had the The most effective method of evaluating the potential investigation repeated at the end of the study. There were no cytostatic effect of G17-DT is a survival study. Such studies are complete or partial responses, with 12 of 14 patients having currently underway in both gastric and pancreatic adenocarci- progressive disease reflected in elevated CEA levels. Two pa- nomas. These tumors have been chosen because gastrin also has tients had stable disease, both having had progressive disease a proliferative effect in these cancers, and there is a paucity of before study entry. One patient had a local rectal recurrence; the adjuvant therapies that are beneficial. In gastric cancer, G17-DT other patient had liver metastases (data not shown). will be assessed in a prospective, placebo-controlled trial with Overall, the CEA levels rose during treatment in all study patients who have had a potentially curative resection for stage groups. However, in five patients the CEA remained stable, and ⌱⌱/⌱⌱⌱ cancer. Pancreatic cancer patients will be randomized to one patient showed a modest reduction (data not shown). receive either G17-DT or gemcitabine.

DISCUSSION REFERENCES The principal aims of this study were to determine the 1. Walsh, J. Gastrin. In: J. Walsh and G. Dockray (eds.), Gut Peptides, safety and efficacy of the immunogen in terms of an antibody pp. 75–121. New York: Raven Press, Ltd., 1994. response. Gastrimmune immunization was well tolerated sys- 2. Imdahl, A., Mantamadiotis, T., Eggstein, S., Farthmann, E. H., and temically. Only two patients associated nausea and vomiting Baldwin, G. S. Expression of gastrin, gastrin/CCK-B and gastrin/ with immunization. Importantly, immunization did not induce CCK-C receptors in human colorectal carcinomas. J. Cancer Res. Clin. autoantibody production. Oncol., 121: 661–666, 1995. Half of the patients had mild myalgia after immunization. 3. Upp, J. R., Jr., Singh, P., Townsend, C. M., Jr., and Thompson, J. C. This was short-lived and did not interfere with their daily life Clinical significance of gastrin receptors in human colon cancers. Can- activities. Abscess formations occurred in a group of patients cer Res., 49: 488–492, 1989. but were easily treatable, and symptoms were resolved within a 4. Reubi, J. C., Schaer, J. C., and Waser, B. Cholecystokinin(CCK)-A and CCK-B/gastrin receptors in human tumors. Cancer Res., 57: 1377– short period of time. Initially, it was thought that the adjuvant 1386, 1997. nor-MDP was responsible for abscess formation; however, ab- 5. Matsushima, Y., Kinoshita, Y., Nakata, H., Inomoto-Naribayashi, Y., scesses occurred in the formulation without adjuvant or emul- Asahara, M., Kawanami, C., Nakamura, A., Ito, M., Matsui, T., Fuji- sifier. There were no discernible factors that predisposed pa- wara, T., et al. Gastrin receptor gene expression in several human tients to abscess formation. There was also no pattern or timing carcinomas. Jpn. J. Cancer Res., 85: 819–824, 1994. to abscess formation. We are unable to explain why the majority 6. McWilliams, D. F., Watson, S. A., Crosbee, D. M., Michaeli, D., and ␦ of patients tolerate the immunogen well and others develop Seth, R. Coexpression of gastrin and gastrin receptors (CCK-B and CCK-B) in gastrointestinal tumor cell lines. Gut, 42: 795–798, 1998. abscesses. However, unadjuvanted immunogen Ͻ330 ␮g did 7. Singh, P., Xu, Z., Dai, B., Rajaraman, S., Rubin, N., and Dhruva, B. not result in abscess formation. Incomplete processing of progastrin expressed by human colon cancer Excess antibody titers were seen in 40 of 50 (80%) patients cells: role of noncarboxyamidated gastrins. Am. J. Physiol., 266: G459– with advanced colorectal cancer; this is despite only 36 of 50 G468, 1994. patients having a complete immunization course. The Kd of the 8. Singh, P., Owlia, A., Espeijo, R., and Dai, B. Novel gastrin receptors antibodies was such that they would be able to compete for mediate mitogenic effects of gastrin and processing intermediates of gastrin binding with CCKB receptors expressed by colorectal gastrin on Swiss 3T3 fibroblasts. Absence of detectable cholecystokinin (CCK)-A and CCK-B receptors. J. Biol. Chem., 270: 8429–8438, 1995. cancer cells, the affinities of which range from 0.1 to 0.5 nM (37, 38). The median antigen binding capacity was such that it would 9. Kochman, M. L., DelValle, J., Dickinson, C. J., and Boland, C. R. ϳ Post-translational processing of gastrin in neoplastic human colonic be able to absorb a gastrin concentration 47 times greater than tissues. Biochem. Biophys. Res. Commun., 189: 1165–1169, 1992. that seen postprandially (1). 10. Nemeth, J., Taylor, B., Pauwels, S., Varro, A., and Dockray, G. J. Identification of progastrin derived peptides in colorectal carcinoma extracts. Gut, 34: 90–95, 1993. 11. Finley, G. G., Koski, R. A., Melhem, M. F., Pipas, J. M., and Meisler, A. I. Expression of the gastrin gene in the normal human colon 4 J. Walsh, personal communication. and colorectal adenocarcinoma. Cancer Res., 53: 2919–2926, 1993.

12. Ciccotosto, G. D., McLeish, A., Hardy, K. J., and Shulkes, A. 26. Harrison, J. D., Jones, J. A., and Morris, D. L. The effect of the Expression, processing, and secretion of gastrin in patients with colo- gastrin receptor antagonist proglumide on survival in gastric carcinoma. rectal carcinoma. Gastroenterology, 109: 1142–1153, 1995. Cancer (Phila.), 66: 1449–1452, 1990. 13. Van Solinge, W. W., Nielsen, F. C., Friis-Hansen, L., Falkmer, 27. Morris, D. L., Charnley, R. M., Ballantyne, K. C., and Jones, J. A U. G., and Rehfeld, J. F. Expression but incomplete maturation of pilot randomized control trial of proglumide (a gastrin receptor antag- progastrin in colorectal carcinomas. Gastroenterology, 104: 1099–1107, onist) in advanced colorectal cancer. Eur. J. Surg. Oncol., 16: 423–425, 1993. 1990. 14. Okhura, H., Hanafusa, K., Marayama, K., Kitoaka, H., Watanabe, 28. Watson, S. A., Crosbee, D. M., Morris, D. L., Robertson, J. F., S., and Kameya, T. Gastrin enhanced tumor growth of a xenotransplant- Makovec, F., Rovati, L. C., and Hardcastle, J. D. Therapeutic effect of able human gastric carcinoma in nude mice. Jpn. J. Clin. Oncol., 10: the gastrin receptor antagonist, CR2093 on gastrointestinal tumour cell 255–261, 1980. growth. Br. J. Cancer, 65: 879–883, 1992. 15. Winsett, O. E., Townsend, C. M., Jr., Glass, E. J., and Thompson, 29. Watson, S. A., Morris, D. L., Durrant, L. G., Robertson, J. F., and J. C. Gastrin stimulates growth of colon cancer. Surgery (St. Louis), 99: Hardcastle, J. D. Inhibition of gastrin-stimulated growth of gastrointes- 302–307, 1986. tinal tumour cells by octreotide and the gastrin/cholecystokinin receptor 16. Sumiyoshi, H., Yasui, W., Ochiai, A., and Tahara, E. Effects of antagonists, proglumide and lorglumide. Eur. J. Cancer, 28A: 1462– gastrin on tumor growth and cyclic nucleotide metabolism in xenotrans- 1467, 1992. plantable human gastric and colonic carcinomas in nude mice. Cancer 30. Watson, S. A., Michaeli, D., Morris, T. M., Robinson, G., and Res., 44: 4276–4280, 1984. Hardcastle, J. D. Gastrimmune inhibits the primary and secondary 17. Watson, S., Durrant, L., and Morris, D. Gastrin: growth enhancing growth of 2 rat colon carcinomas. Proc. Am. Assoc. Cancer Res., 35: effects on human gastric and colonic tumour cells. Br. J. Cancer, 59: 257, 1994. 554–558, 1989. 31. Makishima, R., Larkin, D., Michaeli, D., and Gaginella, T. S. 18. Hoosein, N. M., Kiener, P. A., Curry, R. C., Rovati, L. C., McGil- Active immunization against gastrin-17 with an N-terminal derived bra, D. K., and Brattain, M. G. Antiproliferative effects of gastrin immunogen inhibits gastric and duodenal lesions in rats. Gastroenterol- receptor antagonists and antibodies to gastrin on human colon carci- ogy, 106: A824, 1995. noma cell lines. Cancer Res., 48: 7179–7183, 1988. 32. Justin, T., Watson, S., Michaeli, D., Hardcastle, J., and Steele, R. 19. Hoosein, N. M., Kiener, P. A., Curry, R. C., and Brattain, M. G. Gastric acid suppression using anti-gastrin17 antibodies produced by a Evidence for autocrine growth stimulation of cultured colon tumor cells gastrin immunogen, Gastrimmune, in an in vivo pig model. Gastroen- by a gastrin/cholecystokinin-like peptide. Exp. Cell Res., 186: 15–21, terology, 108: A125, 1995. 1990. 33. Justin, T. A., Steele, R. J. C., Michaeli, D., Gregson, R., and 20. Wang, T. C., Koh, T. J., Varro, A., Cahill, R. J., Dangler, C. A., Fox, J. G., and Dockray, G. J. Processing and proliferative effects of Watson, S. A. The effect of porcine anti-gastrin-17 antibodies on the pig human progastrin in transgenic mice. J. Clin. Investig., 98: 1918–1929, gastrointestinal mucosa. Gut, 44: A117, 1999. 1996. 34. Watson, S. A., Michaeli, D., Grimes, S., Morris, T., Robinson, G., 21. Koh, T. J., Dockray, G. J., Varro, A., Cahill, R. J., Dangler, C. A., Varro, A., Justin, T., and Hardcastle, J. Gastrimmune raises antibodies Fox, J. G., and Wang, T. C. Overexpression of glycine-extended gastrin that neutralize amidated and glycine-extended gastrin-17 and inhibit the in transgenic mice results in increased colonic proliferation. J. Clin. growth of colon cancer. Cancer Res., 56: 880–885, 1996. Investig., 103: 1119–1126, 1999. 35. Watson, S. A., Smith, A. M., Michaeli, D., Grimes, S., Caplin, M., 22. Renga, M., Brandi, G., Paganelli, G., Calabrese, C., Papa, S., Tosti, and Hardcastle, J. D. Antibodies targeting the amino terminal portion of A., Tomassetti, P., Miglioli, M., and Biasco, G. Rectal cell proliferation the human CCKB/gastrin receptor inhibit the liver invasion of a human and colon cancer risk in patients with hypergastrinaemia. Gut, 41: colonic tumor. Gastroenterology, 114: G2899, 1998. 330–332, 1997. 36. Watson, S. A., Michaeli, D., Morris, T., Clarke, P., Varro, A., 23. Sobhani, I., Lehy, T., Laurent-Puig, P., Cadiot, G., Ruszniewski, P., Griffin, N., Smith, A., Justin, T., and Hardcastle, J. D. Antibodies raised and Mignon, M. Chronic endogenous hypergastrinemia in humans: by Gastrimmune inhibit the spontaneous metastases of a human colo- evidence for a mitogenic effect on the colonic mucosa. Gastroenterol- rectal tumour cell line, AP5LV. Eur. J. Cancer, 35: 1286–1291, 1999. ogy, 105: 22–30, 1993. 37. Singh, P., Walker, J. P., Townsend, C. M., and Thompson, J. C. 24. Thorburn, C. M., Friedman, G. D., Dickinson, C. J., Vogelman, Role of gastrin and gastrin receptors on the growth of a transplantable J. H., Orentreich, N., and Parsonnet, J. Gastrin and colorectal cancer: a mouse colon carcinoma (MC26) in Balb/c mice. Cancer Res., 46: prospective study. Gastroenterology, 115: 275–280, 1998. 1612–1616, 1986. 25. Smith, A. M., Morris, T., Rowlands, B. J., and Watson, S. A. 38. Upp, J. R., Singh, S., Townsend, C. M., and Thompson, J. C. Helicobacter pylori and omeprazole induced hypergastrinaemia pro- Clinical significance of gastrin receptors in human colon cancers. Can- mote colorectal carcinogenesis. Br. J. Surg., 86 (Suppl. 1): 1999. cer Res., 49: 488–492, 1989.

Andrew M. Smith, Timothy Justin, Dor Michaeli, et al.

Clin Cancer Res 2000;6:4719-4724.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/6/12/4719

Cited articles This article cites 33 articles, 11 of which you can access for free at: http://clincancerres.aacrjournals.org/content/6/12/4719.full#ref-list-1

Citing articles This article has been cited by 8 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/6/12/4719.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/6/12/4719. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.