Vaeth JM, Meyer JL (eds): The Present and Future Role of Monoclonal Antibodies in the Management of Cancer. Front Radiat Ther mcii. Basel, Karger, 1990, vol 24, pp 214-224
Treatment of Gastrointestinal Cancer Using Monoclonal Antibodies Arnold M. Markoea, Luther W. Bradya, d, David Woo a, Beatriz Amendolaa, Ulf Karlsson a, Scot Fishera, BizhanMicailya, Michael Rackovera, Stephen Bulovab, Zenon Steplewskib, Hilary Koprowski'
Colorectal carcinoma remains one of the most difficult of the common malignancies to contend with clinically. Undoubtedly, surgery remains the mainstay of extirpative therapy and the results are quite favorable for locoregíonal control and survival in the earliest stages of the disease. However, as the disease progresses, in terms of invasiveness through the intestinal wall and into surrounding paracolic fat and lymphatics, the probability of local control is dramatically reduced and the potential for distant failure is increased. In recent years, the addition of postoperative external beam radiation therapy, in cases of rectosigmoid lesions, has been shown to improve locoregional control rates, but appears to have little impact upon overall survival due to unchanged development of distant metastatic disease, primarily as hepatic or pulmonary metastases. Few advances have been apparent, in recent years, utilizing systemic chemotherapy for the treatment of such disseminated disease. For over two decades the standard by which other chemotherapeutic agents or regimens have been judged with respect to efficacy has been 5-fluοrοuracil (5-FU). Response rates that have been reported for 5-FU, alone, have varied widely from 9.5 to 44%, with the generally accepted response rate being 20% [1]. No other single agent has been demonstrated to be statistically superior to 5-FU. Mitomycin C has been shown to yield an average
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Departments of a Radiation Oncology and b Neoplastic Disease, Hahnemann University Hospital; °The Wistar Institute; d Hylda Cohen-American Cancer Society, Department of Radiation Oncology, Hahnemann University, Philadelphia, Pa., USA
215
response rate of 18.5 % in colorectal carcinoma patients in a large number of clinical series [2]. The nitrosoureas BCNU and methyl-CCNU and CCNU have response rates from 3 to 20 % in patients previously treated by other chemotherapy agents and 10-30 % in patients without prior chemotherapeutic treatment [3-7]. Intensive investigation has been performed using combination chemotherapy for colorectal carcinoma. Although response rates varying from 4 to 50% have been reported using numerous combinations and regimens of 5-FU, BCNU, methyl-CCNU, mitomycin C, vincristine and streptozotocin [8, 9], there has been no survival advantage over single agent 5-FU and median survival from treatment in these studies has been less than 1 year from time of treatment. Similarly, pancreatic carcinoma is associated with a dismal prognosis [ 10]. Surgical extirpation of tumor is the mainstay of successful therapy, but most patients are not operable for cure [ 10, 11 ]. Irradiation by conventional external beam techniques has rarely been successful in effecting disease cure although survival may be increased in treated patients [ 11 ]. Chemotherapy has been attempted by a variety of single agents or multiple agents in combination, with limited success [12]. Most recently, more innovative techniques of irradiation of these cancers have been attempted using interstitial brachytherapeutic approaches [ 13] or intraoperative irradiation of the tumor [14, 15]. These approaches also appear to be of limited success [11, 14]. Any current attempt to improve survival in metastatic colorectal carcinoma remains experimental. One area of investigational promise is the use of monoclonal antibody (MoAb) therapy. MoAbs detect a number of antigens which are specifically or selectively expressed by cancer cells [16]. In early studies, some patients with advanced colorectal or pancreatic carcinoma treated with murine MoAb CO 17-1 A have demonstrated tumor regression [ 17]. The murine MoAb C017-1 A is an IgG2a antibody produced by immunization of mice with an adenocarcinoma cell line. It binds to an antigen which is highly expressed on the surface of cells of human gastrointestinal malignancies [ 18, 19]. Much smaller amounts of the MoAb bind to normal intestinal mucosa. Since MoAb CO17-1A is relatively specific for human gastrointestinal cancer, it has been utilized investigationally for the imaging and treatment of colonic, gastric and pancreatic cancers [16]. Phase I and II trials using MoAb CO17-1A have shown it to be well tolerated and to produce antitumor responses in some patients. However,
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
MoAb Therapy for Gastrointestinal Cancer
Markoe et al.
216
since it is a murine-derived antibody, its efficacy may be limited by several factors. Pharmacokinetic data reveal a half-life of 18-24 h for murine IgG MoAb relative to a half-life of 15-25 days for normal human IgG [20, 21]. Therefore, single doses of CO Ι 7- 1 Α do not produce sustained levels of circulating antibody. In some patients, hypersensitivity reactions occur which can limit the ability to readminister the antibody over weeks or months [22]. It would appear that hypersensitivity reactions are schedulerelated and repeated doses may be tolerated well depending upon the administration schedule chosen [23]. MoAb can be radiolabeled for imaging studies or for attempts at therapy by utilizing to MoAb as a relatively selective delivery vehicle for the isotope utilized. Since 1 251 acts by the emission of an Auger electron, unlike the gammaemitting 131 I, it is a requisite that the antibody be internalized into the target cell. Our laboratories have evidence that this is the actual case in vitro [24]. The internalization appears to be energy dependent since it occurs at 37 ° C, but not at 0 ° C. Moreover, we can demonstrate nuclear chromosomal damage at the former temperature, but not at the latter. This appears to be specific for 1251-C017-1 A, since it does not occur with CO171A alone, X251-labeled irrelevant MoAb or with 1311-labeled CO 17-1A antibody or 1a125 I. The 125 I-C017-1Α does not cause the chromosomal damage in cells not containing the surface antigen to which the MoAb has affinity. The surface antigen density of the target cells does not appear to be altered as a consequence of the internalization of the radiolabeled MoAb. This paper will 'present our initial report of phase Ι trials using MoAbs directed against gastrointestinal malignancies. We have conducted two separate studies, one utilizing radiolabeled MoAb CO 17-1A and the other using one or more MoAbs against gastrointestinal or other tumors administered as `native' MoAb admixed with leukocytes obtained from the patient receiving the antibodies.
MoAbs were supplied by the Wistar Institute as a sterile, nonpyrogenic, colorless solution with each vial containing 100 mg of antibody per 10 ml of isotonic saline. Preparations contained no preservatives. For the radiolabeled MoAb study, MoAb C017-IA was used exclusively. The MoAb is an IgG2a murine MoAb that selectively binds to human gastrointestinal tumor cells. It
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Materials and Methods
217
is produced by a hybridoma cell line isolated after fusion of spleen cells from Balb/c mice, immunized with the human colorectal carcinoma cell line SW 1083, with the mouse myeloma cell line P3 X 63Ág8. All radioiodinations were performed at Hahnemann University. The 125I was obtained as sodium iodine from New England Nuclear (DuPont) (pH 8-10) and the MoAb radioiodinated using the standard indogen method which has yielded ultimate specific activity of 12.5 mCi/mg MoAb using C017-1A. Radiochemical purity is assessed after passage of the radioiodinated MoAb through a Sephadex G-25 column to remove iodide. Instant thin-layer chromatography has shown a single radiolabeled entity with no free iodine. After 1 week of storage of the radiolabeled MoAb, however, about 5 % free iodine is detected. This is due to autoradiolysis of the material. Clinical sterility is assured by passage of the radiolabeled MoAb through a 0.22-µm Millipore filter. Assurance that the radiolabeled MoAb remains pyrogen free is via the Limulus lysate assay. One day prior to administration of radioiodinated MoAb, patients were dosed with 1 ml of Lugol's solution in order to block thyroidal uptake of the radioiodine. Radiolabeled MoAb was injected intravenously. Doses were slowly escalated from 3-5 to 25 mCi per dose. Scheduling was altered from weekly to monthly and patients were clinically assessed for allergic reactions attributable to the induction of a human antimouse antibody (HAMA) response. Patients were assessed for response by radiographic analysis as well as by clinical and/or laboratory assessment. The leukapheresis study utilized not only MoAb C017-1A, but also MoAb GÁ73.3 raised against gastric carcinoma and the 19-9 MoAb against a colorectal tumor-associated antigen which is shed into the circulation. Where appropriate, MoAb BR55-2 against breast carcinoma was also employed. All patients were treated in an ambulatory setting. Leukapheresis was performed by personnel in the Hahnemann University Hospital Blood Donor Center (SB). A Haemonetics Model 150 pheresis unit was employed at a flow rate of 80 mu/min. The patients' blood was collected in transfer packs containing Fenwal ACD Formula A as an anticoagulant. The bully coat from six 600-ml passes was reserved and the remainder of the blood components reinfused into the patient. A total mononuclear white cell count was performed on each collection of autologous leukocytes which approximated 200 ml total volume. Following harvest of the buffy coat, MoAb was added either singly or as a `cocktail' of more than one MoAb. The mixture of MoAb and autologous leukocytes was incubated at room temperature in a platelet rotator for 1 h. The antibody-autologous monocyte mixture was then infused intravenously over a period of 1-2 h. Patients' vital signs were monitored periodically during and after the infusion. Patients were discharged from the Blood Donor Center to home after stability of vital signs for at least 1 h after completion of the infusion or after any reaction had subsided with appropriate antihistaminic or steroid administration. Such pharmaceuticals were kept at hand during all administrations of MoAb in either protocol. Sterile procedures were utilized for all intravenous administrations and safe radioisotope handling procedures were observed for all administrations of radiolabeled MoAb. Most patients with hepatic metastases were given a `conditioning' course of external beam radiation therapy to the involved liver. Such conditioning has been shown to apparently increase the localization of antibody within the tumor [25]. We employed a 10-
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
MoAb Therapy for Gastrointestinal Cancer
Markoe et al.
218
fraction course in 1.5-Gy daily fractions as a conditioning course, delivering 15 Gy to the involved liver with the dose calculated at midplane on the central axis. All patients accessioned into either study had primary gastrointestinal malignancies that had recurred or metastasized. These patients had failed all conventional modalities of therapy. Although we sought patients with Karnofsky Performance Status (KPS) greater than 70%, many of our patients were treated on a compassionate basis and had KPS scores appreciably lower. No patient with shellfish allergy or allergy to iodinated contrast material or prior allergy to muríne proteins were eligible for study. Patients all signed informed consent for participation in an experimental study.
Results Our phase I study utilizing radioiodinated CO 17-1 A accessioned 53 patients of whom 24 had pancreatic carcinoma, 25 had colorectal carcinoma and 1 each had gastric carcinoma, gallbladder carcinoma, primary duodenal adenocarcinoma and carcinoma of the ampulla of Vater. Table 1 shows the characteristics of these patients. Table 2 presents these patients and indicates the types and durations of responses observed. Clearly, all partial or stable responses were short lived, but several of the patients remain alive for times up to or over 8 months with disease stability. There was, in addition to these responses, a very interesting mixed response in a patient treated over a 1-year time span with 17 doses of radioiodinated MoAb. This patient had hepatic metastases from pancreatic carcinoma which was unresectable. Initially, the hepatic metastases progressed with radiolabeled MoAb therapy, but the pancreatic head lesion remained sta-
Site
n
Mean age (range), years
Median
Male/ female
Liver metastasis
Pancreas Colorectal Gallbladder Gastric Duodenum Ampulla of Vater
24 25 1 1 1 1
58.6 (29-77) 59.2 (40-79) 67 72 70 37
63 60
15/9 17/8 1/0 0/1 1/0 1/0
20/24 23/25 0/1 1/1 1/1 1/1
Total
53
59.1 (29-79)
63
35/18
46/53 Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Table 1. Characteristics of patients in the radiolabeled MoAb study
MoAb Therapy for Gastrointestinal Cancer
219
ble. Subsequently, there was stability of both component parts of disease and then the hepatic metastases exhibited reduction in size with continued stability of the primary disease. Finally, the primary disease was seen to progress, but the hepatic metastases remained stable. This patient survived for 13 months after initiation of radiolabeled MoAb therapy after failing initial external beam radiotherapy by CT criteria. Table 3 details the characteristics of the patients entered into the leukapheresis study. Of the 14 patients, 10 had primary gastrointestinal
Table 2. Response of patients to radiolabeled MoAb therapy Site
Median n Rx (range)
Cumulative 125I dose, mCi
CR
PR
MR
Pancreas Colorectal Gallbladder Gastric Duodenum Ampulla of Vater
3(1-17) 3 (1-13) 3 4 5 4
7.1-129.1 11.0-88.0 17.1 42.0 75.0 75.1
0 0
0 0
1 0
Total
ST
PD
4 5
19 20 1
1 1 1 0
1
1
10
41
Length of responses: PR (3 months); mixed R (12 months); stable R (pancreatic: 3, 3+, 8, 8+ months; colorectal: 2, 2, 3+, 4+, 5 months; ampulla of Vater: 8+ months).
Table 3. Characteristics of patients in the leukapheresis MoAb study
Pancreas Colorectal Gastric Vaginal Parotid Lung Mesothelioma Total
n 1 8 1 1 1 1 1 14
Mean age (range) 52 64.5 (59-72) 43 63 78 76 70
Median
63
Male/ female
Liver metastasis
1/0 5/3 1/0 0/1 1/0 1/0 1/0
1/1 7/8 0/1 1/1 1/1 0/1 0/1
10/4
10/15 Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Site
Markoe et al.
220
Table 4. Response of patients to leukapheresis MoAb therapy Site
n
Number of Rx
Cumulative MoAb dose, mg
Pancreas Colorectal Gastric Vaginal Parotid Lung Mesothelioma
1 8 1 1 1 1 1
13 5-13 5 3 3 4 4
2,340 900-2,380 1,960 540 742 720 720
Total
CR
PR
Stable PD
3
0
0
1 5 1
1 1 1 1 3
1
1
9
malignancies, 1 had a treated presumed vaginal primary adenocarcinoma (many years after hysterectomy for benign causes) and subsequently developed hepatic metastases for which she was accessioned, 1 had hepatic metastases from parotid tumor, 1 had mesothelioma and 1 had adenocarcinoma of lung. Of these 14 patients, table 4 documents 3 complete responses (21 %), all in patients with colorectal carcinoma (3/8 = 37.5 %). Two of these complete responses were short-lived. The patient with hepatic metastases from parotid carcinoma achieved a short-lived partial response and the patient with metastases from vaginal adenocarcinoma achieved a long-term stabilization of disease in the liver. Table 5 documents the frequency of toxicity noted in these two phase I studies. Clearly, the radiolabeled MoAb study was associated with significantly less overall toxicity than the leukapheresis study. Only 1.9% of patients experienced any toxic reaction in the radiolabeled MoAb study. An additional patient had tested positive for antimouse antibody by skin testing prior to therapy. This patient was routinely pretreated with decadron and benadryl prior to each of 5 therapeutic sessions and had no adverse reactions at all. On the other hand, 50% of the patients in the leukapheresis MoAb study exhibited adverse reactions, all easily handled by conservative medical management. There were no life-threatening reactions.
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Length of responses: CR (4, 6, 28+ months); PR (4 months); stable (26 months).
MoAb Therapy for Gastrointestinal Cancer
221
Table 5. Tabulation of toxic reactions to MoAb therapy Reaction type
Leukapheresis study
Radioiodine study
Abdominal cramps Low back pain Facial flush Chest pain Anxiety Chills Nausea Urticaria Erythema Dyspnea Tachycardia
2 2 3 3 3 2 1 0 1 1 1
0 1 0 0 0 0 0 0 0 0 0
7 7 50
1 52 1.9
Number of patients with toxic reactions Number of patients with no reactions Frequency of reactions, % of patients
Our phase I results certainly show that there is the potential for MoAb therapy to impact positively on the course of certain patients with gastrointestinal malignancy, even at late stages of disease. The administration of native MoAbs admixed with mononuclear cells from the individual patient appeared to give superior results than the radiolabeled MoAb. Perhaps this is due to the quantum difference in the amount of MoAb administered in each study. The leukapheresis study administered doses of MoAb in no less than 100 mg quantity and patients received from 900 to over 2,300 mg cumulative dose of MoAb in from 3 to 13 dosings. On the other hand, patients in the radioiodinated MoAb study, although receiving from 1 to 17 dosings, had little cumulative MoAb dose. No more than 2 mg MoAb was infused at each dose. Similarly, the low efficacy in the radiolabeled MoAb study could be attributable to the very low doses of radioisotope administered per dosing in this phase I trial. With such low doses, the fact that any positive or stable responses were found is quite encouraging. The differences in toxicity manifestations may also reflect on the quantitative difference in amount of MoAb utilized in each study. Certain-
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Discussion
222
ly, it would appear that the study associated with the greater toxicity had the higher individual and cumulative doses of MoAb infused. It would be interesting to think that the lower doses of MoAb was associated with less toxicity because the human antimouse antibody (ΗΑΜΑ) response did not develop at these lower MoAb doses. Yet, this is probably not the case since doses of ΜοΑb CO 17-1 Α from 15 to 200 mg still evoke a ΗΑΜΑ response in a very high percentage of patients [26]. Assuming that ΗΑΜΑ responses are equivalent in the two phase I trials, the only significant difference is the fact that the study having the greater toxic reactions utilized reinfusion of the patients mononuclear cells with MoAb, rather than MoAb itself. It is certainly possible that this difference leads to a higher probability of having patient discomfort and overt clinical reactions. Obviously, in the radioiodinated MoAb study, the fact that injection was by slow intravenous `push' did not cause significant toxicity, but only small quantities of MoAb were infused per dose. In the leukapheresis study, MoAbs were infused over a 1- to 2-hour time span. Single ΜοΑb dose was usually not above 200 mg. However, when `cocktails' of MoAb were administered, up to 600 mg total MoAb could be administered. Thus, the infusion rate could have been as high as 200-600 mg MoAb per hour at each dosing. There is evidence in the literature that dosing at less than 1,000 mg/h (i.v. infusion rate
Treatment of Gastrointestinal Cancer Using Monoclonal Antibodies Arnold M. Markoea, Luther W. Bradya, d, David Woo a, Beatriz Amendolaa, Ulf Karlsson a, Scot Fishera, BizhanMicailya, Michael Rackovera, Stephen Bulovab, Zenon Steplewskib, Hilary Koprowski'
Colorectal carcinoma remains one of the most difficult of the common malignancies to contend with clinically. Undoubtedly, surgery remains the mainstay of extirpative therapy and the results are quite favorable for locoregíonal control and survival in the earliest stages of the disease. However, as the disease progresses, in terms of invasiveness through the intestinal wall and into surrounding paracolic fat and lymphatics, the probability of local control is dramatically reduced and the potential for distant failure is increased. In recent years, the addition of postoperative external beam radiation therapy, in cases of rectosigmoid lesions, has been shown to improve locoregional control rates, but appears to have little impact upon overall survival due to unchanged development of distant metastatic disease, primarily as hepatic or pulmonary metastases. Few advances have been apparent, in recent years, utilizing systemic chemotherapy for the treatment of such disseminated disease. For over two decades the standard by which other chemotherapeutic agents or regimens have been judged with respect to efficacy has been 5-fluοrοuracil (5-FU). Response rates that have been reported for 5-FU, alone, have varied widely from 9.5 to 44%, with the generally accepted response rate being 20% [1]. No other single agent has been demonstrated to be statistically superior to 5-FU. Mitomycin C has been shown to yield an average
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Departments of a Radiation Oncology and b Neoplastic Disease, Hahnemann University Hospital; °The Wistar Institute; d Hylda Cohen-American Cancer Society, Department of Radiation Oncology, Hahnemann University, Philadelphia, Pa., USA
215
response rate of 18.5 % in colorectal carcinoma patients in a large number of clinical series [2]. The nitrosoureas BCNU and methyl-CCNU and CCNU have response rates from 3 to 20 % in patients previously treated by other chemotherapy agents and 10-30 % in patients without prior chemotherapeutic treatment [3-7]. Intensive investigation has been performed using combination chemotherapy for colorectal carcinoma. Although response rates varying from 4 to 50% have been reported using numerous combinations and regimens of 5-FU, BCNU, methyl-CCNU, mitomycin C, vincristine and streptozotocin [8, 9], there has been no survival advantage over single agent 5-FU and median survival from treatment in these studies has been less than 1 year from time of treatment. Similarly, pancreatic carcinoma is associated with a dismal prognosis [ 10]. Surgical extirpation of tumor is the mainstay of successful therapy, but most patients are not operable for cure [ 10, 11 ]. Irradiation by conventional external beam techniques has rarely been successful in effecting disease cure although survival may be increased in treated patients [ 11 ]. Chemotherapy has been attempted by a variety of single agents or multiple agents in combination, with limited success [12]. Most recently, more innovative techniques of irradiation of these cancers have been attempted using interstitial brachytherapeutic approaches [ 13] or intraoperative irradiation of the tumor [14, 15]. These approaches also appear to be of limited success [11, 14]. Any current attempt to improve survival in metastatic colorectal carcinoma remains experimental. One area of investigational promise is the use of monoclonal antibody (MoAb) therapy. MoAbs detect a number of antigens which are specifically or selectively expressed by cancer cells [16]. In early studies, some patients with advanced colorectal or pancreatic carcinoma treated with murine MoAb CO 17-1 A have demonstrated tumor regression [ 17]. The murine MoAb C017-1 A is an IgG2a antibody produced by immunization of mice with an adenocarcinoma cell line. It binds to an antigen which is highly expressed on the surface of cells of human gastrointestinal malignancies [ 18, 19]. Much smaller amounts of the MoAb bind to normal intestinal mucosa. Since MoAb CO17-1A is relatively specific for human gastrointestinal cancer, it has been utilized investigationally for the imaging and treatment of colonic, gastric and pancreatic cancers [16]. Phase I and II trials using MoAb CO17-1A have shown it to be well tolerated and to produce antitumor responses in some patients. However,
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
MoAb Therapy for Gastrointestinal Cancer
Markoe et al.
216
since it is a murine-derived antibody, its efficacy may be limited by several factors. Pharmacokinetic data reveal a half-life of 18-24 h for murine IgG MoAb relative to a half-life of 15-25 days for normal human IgG [20, 21]. Therefore, single doses of CO Ι 7- 1 Α do not produce sustained levels of circulating antibody. In some patients, hypersensitivity reactions occur which can limit the ability to readminister the antibody over weeks or months [22]. It would appear that hypersensitivity reactions are schedulerelated and repeated doses may be tolerated well depending upon the administration schedule chosen [23]. MoAb can be radiolabeled for imaging studies or for attempts at therapy by utilizing to MoAb as a relatively selective delivery vehicle for the isotope utilized. Since 1 251 acts by the emission of an Auger electron, unlike the gammaemitting 131 I, it is a requisite that the antibody be internalized into the target cell. Our laboratories have evidence that this is the actual case in vitro [24]. The internalization appears to be energy dependent since it occurs at 37 ° C, but not at 0 ° C. Moreover, we can demonstrate nuclear chromosomal damage at the former temperature, but not at the latter. This appears to be specific for 1251-C017-1 A, since it does not occur with CO171A alone, X251-labeled irrelevant MoAb or with 1311-labeled CO 17-1A antibody or 1a125 I. The 125 I-C017-1Α does not cause the chromosomal damage in cells not containing the surface antigen to which the MoAb has affinity. The surface antigen density of the target cells does not appear to be altered as a consequence of the internalization of the radiolabeled MoAb. This paper will 'present our initial report of phase Ι trials using MoAbs directed against gastrointestinal malignancies. We have conducted two separate studies, one utilizing radiolabeled MoAb CO 17-1A and the other using one or more MoAbs against gastrointestinal or other tumors administered as `native' MoAb admixed with leukocytes obtained from the patient receiving the antibodies.
MoAbs were supplied by the Wistar Institute as a sterile, nonpyrogenic, colorless solution with each vial containing 100 mg of antibody per 10 ml of isotonic saline. Preparations contained no preservatives. For the radiolabeled MoAb study, MoAb C017-IA was used exclusively. The MoAb is an IgG2a murine MoAb that selectively binds to human gastrointestinal tumor cells. It
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Materials and Methods
217
is produced by a hybridoma cell line isolated after fusion of spleen cells from Balb/c mice, immunized with the human colorectal carcinoma cell line SW 1083, with the mouse myeloma cell line P3 X 63Ág8. All radioiodinations were performed at Hahnemann University. The 125I was obtained as sodium iodine from New England Nuclear (DuPont) (pH 8-10) and the MoAb radioiodinated using the standard indogen method which has yielded ultimate specific activity of 12.5 mCi/mg MoAb using C017-1A. Radiochemical purity is assessed after passage of the radioiodinated MoAb through a Sephadex G-25 column to remove iodide. Instant thin-layer chromatography has shown a single radiolabeled entity with no free iodine. After 1 week of storage of the radiolabeled MoAb, however, about 5 % free iodine is detected. This is due to autoradiolysis of the material. Clinical sterility is assured by passage of the radiolabeled MoAb through a 0.22-µm Millipore filter. Assurance that the radiolabeled MoAb remains pyrogen free is via the Limulus lysate assay. One day prior to administration of radioiodinated MoAb, patients were dosed with 1 ml of Lugol's solution in order to block thyroidal uptake of the radioiodine. Radiolabeled MoAb was injected intravenously. Doses were slowly escalated from 3-5 to 25 mCi per dose. Scheduling was altered from weekly to monthly and patients were clinically assessed for allergic reactions attributable to the induction of a human antimouse antibody (HAMA) response. Patients were assessed for response by radiographic analysis as well as by clinical and/or laboratory assessment. The leukapheresis study utilized not only MoAb C017-1A, but also MoAb GÁ73.3 raised against gastric carcinoma and the 19-9 MoAb against a colorectal tumor-associated antigen which is shed into the circulation. Where appropriate, MoAb BR55-2 against breast carcinoma was also employed. All patients were treated in an ambulatory setting. Leukapheresis was performed by personnel in the Hahnemann University Hospital Blood Donor Center (SB). A Haemonetics Model 150 pheresis unit was employed at a flow rate of 80 mu/min. The patients' blood was collected in transfer packs containing Fenwal ACD Formula A as an anticoagulant. The bully coat from six 600-ml passes was reserved and the remainder of the blood components reinfused into the patient. A total mononuclear white cell count was performed on each collection of autologous leukocytes which approximated 200 ml total volume. Following harvest of the buffy coat, MoAb was added either singly or as a `cocktail' of more than one MoAb. The mixture of MoAb and autologous leukocytes was incubated at room temperature in a platelet rotator for 1 h. The antibody-autologous monocyte mixture was then infused intravenously over a period of 1-2 h. Patients' vital signs were monitored periodically during and after the infusion. Patients were discharged from the Blood Donor Center to home after stability of vital signs for at least 1 h after completion of the infusion or after any reaction had subsided with appropriate antihistaminic or steroid administration. Such pharmaceuticals were kept at hand during all administrations of MoAb in either protocol. Sterile procedures were utilized for all intravenous administrations and safe radioisotope handling procedures were observed for all administrations of radiolabeled MoAb. Most patients with hepatic metastases were given a `conditioning' course of external beam radiation therapy to the involved liver. Such conditioning has been shown to apparently increase the localization of antibody within the tumor [25]. We employed a 10-
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
MoAb Therapy for Gastrointestinal Cancer
Markoe et al.
218
fraction course in 1.5-Gy daily fractions as a conditioning course, delivering 15 Gy to the involved liver with the dose calculated at midplane on the central axis. All patients accessioned into either study had primary gastrointestinal malignancies that had recurred or metastasized. These patients had failed all conventional modalities of therapy. Although we sought patients with Karnofsky Performance Status (KPS) greater than 70%, many of our patients were treated on a compassionate basis and had KPS scores appreciably lower. No patient with shellfish allergy or allergy to iodinated contrast material or prior allergy to muríne proteins were eligible for study. Patients all signed informed consent for participation in an experimental study.
Results Our phase I study utilizing radioiodinated CO 17-1 A accessioned 53 patients of whom 24 had pancreatic carcinoma, 25 had colorectal carcinoma and 1 each had gastric carcinoma, gallbladder carcinoma, primary duodenal adenocarcinoma and carcinoma of the ampulla of Vater. Table 1 shows the characteristics of these patients. Table 2 presents these patients and indicates the types and durations of responses observed. Clearly, all partial or stable responses were short lived, but several of the patients remain alive for times up to or over 8 months with disease stability. There was, in addition to these responses, a very interesting mixed response in a patient treated over a 1-year time span with 17 doses of radioiodinated MoAb. This patient had hepatic metastases from pancreatic carcinoma which was unresectable. Initially, the hepatic metastases progressed with radiolabeled MoAb therapy, but the pancreatic head lesion remained sta-
Site
n
Mean age (range), years
Median
Male/ female
Liver metastasis
Pancreas Colorectal Gallbladder Gastric Duodenum Ampulla of Vater
24 25 1 1 1 1
58.6 (29-77) 59.2 (40-79) 67 72 70 37
63 60
15/9 17/8 1/0 0/1 1/0 1/0
20/24 23/25 0/1 1/1 1/1 1/1
Total
53
59.1 (29-79)
63
35/18
46/53 Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Table 1. Characteristics of patients in the radiolabeled MoAb study
MoAb Therapy for Gastrointestinal Cancer
219
ble. Subsequently, there was stability of both component parts of disease and then the hepatic metastases exhibited reduction in size with continued stability of the primary disease. Finally, the primary disease was seen to progress, but the hepatic metastases remained stable. This patient survived for 13 months after initiation of radiolabeled MoAb therapy after failing initial external beam radiotherapy by CT criteria. Table 3 details the characteristics of the patients entered into the leukapheresis study. Of the 14 patients, 10 had primary gastrointestinal
Table 2. Response of patients to radiolabeled MoAb therapy Site
Median n Rx (range)
Cumulative 125I dose, mCi
CR
PR
MR
Pancreas Colorectal Gallbladder Gastric Duodenum Ampulla of Vater
3(1-17) 3 (1-13) 3 4 5 4
7.1-129.1 11.0-88.0 17.1 42.0 75.0 75.1
0 0
0 0
1 0
Total
ST
PD
4 5
19 20 1
1 1 1 0
1
1
10
41
Length of responses: PR (3 months); mixed R (12 months); stable R (pancreatic: 3, 3+, 8, 8+ months; colorectal: 2, 2, 3+, 4+, 5 months; ampulla of Vater: 8+ months).
Table 3. Characteristics of patients in the leukapheresis MoAb study
Pancreas Colorectal Gastric Vaginal Parotid Lung Mesothelioma Total
n 1 8 1 1 1 1 1 14
Mean age (range) 52 64.5 (59-72) 43 63 78 76 70
Median
63
Male/ female
Liver metastasis
1/0 5/3 1/0 0/1 1/0 1/0 1/0
1/1 7/8 0/1 1/1 1/1 0/1 0/1
10/4
10/15 Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Site
Markoe et al.
220
Table 4. Response of patients to leukapheresis MoAb therapy Site
n
Number of Rx
Cumulative MoAb dose, mg
Pancreas Colorectal Gastric Vaginal Parotid Lung Mesothelioma
1 8 1 1 1 1 1
13 5-13 5 3 3 4 4
2,340 900-2,380 1,960 540 742 720 720
Total
CR
PR
Stable PD
3
0
0
1 5 1
1 1 1 1 3
1
1
9
malignancies, 1 had a treated presumed vaginal primary adenocarcinoma (many years after hysterectomy for benign causes) and subsequently developed hepatic metastases for which she was accessioned, 1 had hepatic metastases from parotid tumor, 1 had mesothelioma and 1 had adenocarcinoma of lung. Of these 14 patients, table 4 documents 3 complete responses (21 %), all in patients with colorectal carcinoma (3/8 = 37.5 %). Two of these complete responses were short-lived. The patient with hepatic metastases from parotid carcinoma achieved a short-lived partial response and the patient with metastases from vaginal adenocarcinoma achieved a long-term stabilization of disease in the liver. Table 5 documents the frequency of toxicity noted in these two phase I studies. Clearly, the radiolabeled MoAb study was associated with significantly less overall toxicity than the leukapheresis study. Only 1.9% of patients experienced any toxic reaction in the radiolabeled MoAb study. An additional patient had tested positive for antimouse antibody by skin testing prior to therapy. This patient was routinely pretreated with decadron and benadryl prior to each of 5 therapeutic sessions and had no adverse reactions at all. On the other hand, 50% of the patients in the leukapheresis MoAb study exhibited adverse reactions, all easily handled by conservative medical management. There were no life-threatening reactions.
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Length of responses: CR (4, 6, 28+ months); PR (4 months); stable (26 months).
MoAb Therapy for Gastrointestinal Cancer
221
Table 5. Tabulation of toxic reactions to MoAb therapy Reaction type
Leukapheresis study
Radioiodine study
Abdominal cramps Low back pain Facial flush Chest pain Anxiety Chills Nausea Urticaria Erythema Dyspnea Tachycardia
2 2 3 3 3 2 1 0 1 1 1
0 1 0 0 0 0 0 0 0 0 0
7 7 50
1 52 1.9
Number of patients with toxic reactions Number of patients with no reactions Frequency of reactions, % of patients
Our phase I results certainly show that there is the potential for MoAb therapy to impact positively on the course of certain patients with gastrointestinal malignancy, even at late stages of disease. The administration of native MoAbs admixed with mononuclear cells from the individual patient appeared to give superior results than the radiolabeled MoAb. Perhaps this is due to the quantum difference in the amount of MoAb administered in each study. The leukapheresis study administered doses of MoAb in no less than 100 mg quantity and patients received from 900 to over 2,300 mg cumulative dose of MoAb in from 3 to 13 dosings. On the other hand, patients in the radioiodinated MoAb study, although receiving from 1 to 17 dosings, had little cumulative MoAb dose. No more than 2 mg MoAb was infused at each dose. Similarly, the low efficacy in the radiolabeled MoAb study could be attributable to the very low doses of radioisotope administered per dosing in this phase I trial. With such low doses, the fact that any positive or stable responses were found is quite encouraging. The differences in toxicity manifestations may also reflect on the quantitative difference in amount of MoAb utilized in each study. Certain-
Downloaded by: Université de Paris 193.51.85.197 - 1/20/2020 11:21:00 AM
Discussion
222
ly, it would appear that the study associated with the greater toxicity had the higher individual and cumulative doses of MoAb infused. It would be interesting to think that the lower doses of MoAb was associated with less toxicity because the human antimouse antibody (ΗΑΜΑ) response did not develop at these lower MoAb doses. Yet, this is probably not the case since doses of ΜοΑb CO 17-1 Α from 15 to 200 mg still evoke a ΗΑΜΑ response in a very high percentage of patients [26]. Assuming that ΗΑΜΑ responses are equivalent in the two phase I trials, the only significant difference is the fact that the study having the greater toxic reactions utilized reinfusion of the patients mononuclear cells with MoAb, rather than MoAb itself. It is certainly possible that this difference leads to a higher probability of having patient discomfort and overt clinical reactions. Obviously, in the radioiodinated MoAb study, the fact that injection was by slow intravenous `push' did not cause significant toxicity, but only small quantities of MoAb were infused per dose. In the leukapheresis study, MoAbs were infused over a 1- to 2-hour time span. Single ΜοΑb dose was usually not above 200 mg. However, when `cocktails' of MoAb were administered, up to 600 mg total MoAb could be administered. Thus, the infusion rate could have been as high as 200-600 mg MoAb per hour at each dosing. There is evidence in the literature that dosing at less than 1,000 mg/h (i.v. infusion rate
