Phase I Study of Monoclonal Antibody-Ricin a Chain Immunotoxin Xomazyme-791 in Patients with Metastatic Colon Cancer

(CANCER RESEARCH 49, 6153-6160, November I. 1989] Phase I Study of Monoclonal Antibody-Ricin A Chain Immunotoxin XomaZyme-791 in Patients with Metastatic Colon Cancer

V. S. Byers, R. Rodvien, K. Grant, L. G. Durrani, K. H. Hudson, R. W. Baldwin, and P. J. Scannen XOMA Corporation, Berkeley, California [V. S. B., K. H. H., P. J. S.J; Cancer Research Campaign Laboratories, University of Nottingham, Nottingham, England [V. S. B., L. G. D., R. W. B.I; and Pacific Medical Center, San Francisco, California [R. R., K. G., K. H. H.]

ABSTRACT images primary and metastatic ovarian and colorectal cancers (9-11). Monoclonal antibody 791T/36, recognizing a M, 72,000 antigen on The MoAb has been used to construct an immunotoxin, the surface of colon carcinoma cells, has been used to construct an XomaZyme-791, by conjugation with RTA chain (12, 13). In immunotoxin by conjugating to it the ribosomal inhibitor protein, ricin vitro studies indicate that the immunotoxin retains over 70% toxin A chain. The antibody 791T/36 has been shown to bind to mem of its binding to cells carrying the M, 72,000 antigen, and when branes of freshly disaggregated tumor cells from human colon tumors, tested on 79IT target cells using a [75Se]methionine incorpo and to localize in tumors in vivo. Subacute toxicology testing in rats receiving immunotoxin i.v. showed, at highest doses, weight loss, de ration assay, there was more than a 1000-fold difference be creased serum albumin, and hepatocyte vacuolization without elevation tween the molarity of immunotoxin and free RTA necessary to in liver function tests. attain 50% inhibition of tumor cell growth in vitro (12). It A Phase I dose escalation study was carried out in which 17 patients specifically and effectively inhibits growth of human tumor with metastatic colorectal cancer were treated with doses of immunotoxin xenografts (13). On the basis of these findings, animal toxicol ranging from 0.02 to 0.2 mg/kg/day in 1-h i.v. infusions for a 5-day ogy studies were carried out, and XomaZyme-791 was then course. Side-effects included a composite of signs and symptoms thought tested in a Phase I clinical trial for the treatment of metastatic to be generic to ricin A chain immunotoxins, including decreased serum albumin, mild fever, and flu-like symptoms, all being reversible. Two colorectal cancer. additional findings, reversible proteinuria and mental status changes, were also noted which may be characteristic of this immunotoxin. By 10-20 days after therapy, most patients developed IgM and IgG MATERIALS AND METHODS antibodies against both the ricin toxin A chain and the immunoglobulin portion of the immunotoxin, which were asymptomatic. A strong anticom- Monoclonal Antibody. The generation of the hybridoma producing 79IT/36 MoAb (IgG2b) has been previously described (12, 14). The bining site antibody response was seen. Biological activity manifest as MoAb used for these clinical trials was produced by XOMA Corpora mixed tumor regression was seen in five patients. tion from murine ascites and purified by affinity adsorption on Sepha- rose-protein A (13). The homogeneity of the IgG2b preparations, eval uated by SDS-PAGE, HPLC, and double immunodiffusion indicated INTRODUCTION preparations had purities of greater than 95%. Reactivity of the MoAb MoAbs1 directed against human tumor-associated antigens on normal tissues was assessed by immunoperoxidase staining on frozen sections of a range of normal adult and fetal tissue from five cadaveric have allowed drug targeting to be explored as a therapeutic donors using a modification of the ABC technique (15). modality in cancer (1-3). Cytotoxic moieties, when coupled to To measure reactivity with hematopoietic progenitor cells, mature MoAbs, can be directed specifically to the relevant target cell. T-lymphocytes were removed from human bone marrow aspirates by One such moiety, RTA, has been used to construct several first treating the cells with soybean lectin and removing the resultant immunotoxins (1,4, 5). RTA is a ribosomal inhibiting protein agglutinates, then forming E rosettes and removing them. This tech which functions as an RNA /V-glycosidase specific for the 28-S nique produces a 4 logic depletion of T-lymphocytes (16). Binding to ribosomal subunit (6). Since RTA alone is poorly internalized, the remaining cells was assessed by indirect immunofluorescence, meas it is functionally inactive as a free agent. When coupled to ured by flow cytometry. monoclonal antibodies, however, it can be targeted to tumor The purified MoAb is free of xenotropic and ecotropic viruses as well as the twelve murine viruses measured by the mouse antibody cells, internalized, and is cytotoxic to the cells. production test (a test in which contamination with 12 murine viruses MoAb 79IT/36 recognizes a M, 72,000 antigen present on is evaluated by injection of the test article into mice and determination tumor cells derived from ovarian, colorectal, and osteogenic of antibody production to the viruses of interest2). sarcoma tissues (7-9). Flow cytometric analysis of tumor cells Ricin Toxin A Chain. RTA was purified from castor beans by a series from colorectal and ovarian tumors, derived by enzymatic dis- of column based separations, including immunoaffinity chromatogra aggregation of the tumors, demonstrated that 79IT/36 MoAb phy (17). The RTA was greater than 95% pure as judged by SDS- binds to the majority of cells from over 80% of both types of PAGE, and contained no detectable ricin, or ricin toxin B chain by any tumors (8, 9), indicating the antigen is expressed on the cell assay including immunoprecipitation. The IC!0 level of the purified surface. In vivo, this MoAb localizes in tumors, as demonstrated RTA as measured by a reticulocyte lysate assay (17) was less than 10 by studies in which the antibody, labeled with 13II or '"In, pM. This assay measures inhibition of protein synthesis in a cell free system. In a mouse toxicity assay, RTA injected into BALB/c mice at Received 9/9/88; revised 2/16/89, 5/17/89; accepted 8/4/89. 10 mg/kg produced no deaths. The costs of publication of this article were defrayed in part by the payment Immunotoxin. Immunotoxin XomaZyme-791 was prepared for clin of page charges. This article must therefore be hereby marked advertisement in ical trials by conjugating purified ricin A chain to the murine monoclo accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' The abbreviations used are: MoAb, monoclonal antibody; RTA, ricin A nal antibody 79IT/36 by means of /V-succinionidyl-3-(2-pyridyldi- chain; LDH, lÃ¡clate dehydrogenase: SCOT, serum glutamic oxaloacetic acid thio)propionate reagent, forming a disulfide bond (4,12). It was purified transaminase; CT, computerized tomography; CEA, carcinoembryonic antigen; by gel filtration. Each lot was subjected to a series of tests prior to PBS, phosphate buffered saline: XomaZyme-791. immunotoxin made from release. The free IgG level and amount of immunotoxin present in the 791T/36 MoAb (clinical material); CPK, creatinine phosphokinase: BUN. blood urea nitrogen; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electro- immunotoxin preparation was determined by size exclusion HPLC. phoresis; HPLC, high pressure liquid chromatography; EEC. electroencephalo Binding of the conjugate to 79IT target cells was compared to that of gram: ELISA, enzyme-linked immunosorbent assay; FITC, fluorescein isothio- cyanate. 2 Xoma Corporation, unpublished observations. 6153

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1989 American Association for Cancer Research. MoAb-RlCIN A CHAIN IMMUNOTOXIN XomaZyme-791 the native antibody by flow cytometry analysis as previously described PBS. The plates were incubated for 1 h at room temperature with serial (18). The cytotoxicity of the immunotoxin against relevant and irrele dilutions (1/10 to 1/10,000) of patient's serum diluted in 50 mM sodium vant target cells was determined using an in vitro assay in which target citrate buffer, pH 4.5, containing 5% bovine serum albumin. Following cell survival is determined by [3HJthymidine incorporation. Target cells extensive washing, the plates were incubated for 1 h at room tempera (791T) or the erythroleukemia cell line, Molt-4 (ATCC Bethesda, MD) ture with a 1:1,000 dilution of alkaline phosphatase conjugated goat which does not carry the M, 72,000 antigen recognized by this MoAb, anti-human IgG or anti-human IgM (Sigma, Poole, UK) antiserum. were plated at 4 x 10" cells/well in 100 ml of RPMI 1640 (GIBCO, After washing, the assay was developed with /7-nitrophenolphosphate Grand Island, NY) with 10% fetal calf serum (Hyclone Labs, Logan, (Sigma, Poole, UK) as the alkaline phosphatase substrate. The optical UT). After incubation at 37Â°Cfor 3 h, various concentrations of densities of each well were read at 405 nM and serum titers determined immunotoxin ranging from 20 to 2000 ng/ml were added in 100-/J as the serum dilution producing 50% of the maximum ELISA value aliquots to triplicate wells. After 48 h, 1 //Ci of [3H]thymidine was (19). Ami-combining site antibodies were detected using a flow cytom added to each well and 72 h later the wells were harvested and incor etry assay in which the capacity of patient's serum to block binding of porated radioactivity determined. Results are expressed as amount of FITC conjugated 79IT/36 (791T/36-FITC) to target cells was deter immunotoxin/ml producing 50% inhibition (IC50). These tests were mined (20). This was expressed as titer of serum which produced 50% repeated at monthly intervals after immunotoxin production with min inhibition of the maximum 791T/36-FITC binding to target cells. imal change in reactivity. XomaZyme-791 was prepared for clinical use at a concentration of 1 mg/ml in phosphate buffered saline, pH 7.3. It RESULTS was clear to visual inspection and was filtered through a low protein binding 0.22 /mi filter into about 100 ml of normal saline just prior to Reactivity of MoAb and Immunotoxin infusion. Animal Toxicology Studies. The LD50level was assessed on BALB/c Apart from tumor cells, reactivity of the MoAb assessed by mice; subacute toxicology studies were performed in Sprague-Dawley immunoperoxidase staining is primarily with stromal (noncel- rats, both from Charles River laboratories. Mice were injected i.v. lular) tissue, although there is cytoplasmic staining in the region through the tail vein and observed for 5 days. After at least 7 days of the juxtaglomerular apparatus and occasional reactivity with quarantine, three groups of rats, 12 per group, were dosed i.v. with pulmonary epithelium and isolated kidney glomeruli in some either saline, or 1 mg/kg or 5 mg/kg of 791T/36-RTA (RTA:MoAb sections. There is no detectable binding to progenitor cells by ratio of 4.3:1; 3.7% free RTA) through the tail vein daily for 10 days flow cytometry.2 Other studies have found weak antigen binding (study days 1-10). Each group of animals was weighed daily, and three mitogen stimulated (but not resting) lymphocytes (21). Two in each group were bled, sacrificed, and necropsied on days 6, 11, or 17. Other Sprague-Dawley rats (three rats per group, Simonsen Labs) preparations of immunotoxin were used for clinical trials (Table 1). Analysis by SDS-PAGE indicated several species of immu- were given i.v. injections of 0.2, 1.0, or 5 mg/kg of RTA alone or of saline (2 ml/kg) for S sequential days. Animals were bled and necropsied noconjugate were present with antibody:RTA ratio of 1:1 to on day 5. Serum chemistries included SCOT, serum glutamic pyruvate 1:5. Less than 10% aggregates were present by weight. The transaminase, bilirubin, BUN, creatinine, total protein, albumin, CPK, intrinsic variation of the binding assay is 15% and that of the uric acid, LDH, glucose, and electrolytes. Hematology included indices cytotoxicity assay is 50%; monthly analysis during the time the and platelets were estimated. lots were in clinical use indicated all variations were within this Patient Population and Treatment Plan. Seventeen patients were range. No increase in free antibody or change MoAb:RTA ratio entered in the trial. All had at least one measurable lesion. No patient was noted. had received a murine monoclonal antibody prior to this therapy. No While both lots fell within the accepted ranges, the first had patient studied had significant organ dysfunction; i.e., neurological, a higher MoAb:RTA ratio than the other and correspondingly cardiologica!, and pulmonary functions were within normal ranges. less free antibody. The binding to target cells was decreased as Signed informed consent was obtained from all patients prior to entry into the study which was conducted under a U. S. FDA investigational compared to the lot with lower conjugation ratios but the in new drug exemption notice. All patients signed informed consent. vitro cytotoxicity was similar. XomaZyme-791 was given as 1-h daily i.v. infusions for 5 days, with Patient Characteristics. The characteristics and sites of dis the ability to postpone doses for up to 3 days if suspected side effects ease of the 17 cancer patients (10 females and 7 males) evaluated intervened. Immunotoxin doses, from 0.02 to 0.2 mg/kg/day, were in the immunotoxin study are summarized in Table 2. The age infused over 1 h. Most patients were skin tested prior to the first dose range was 30-70 years. Sixteen patients had colorectal cancer; with 100 /ig of unconjugated antibody; some received an i.v. challenge one patient (Patient 14) had the diagnosis of colorectal cancer of the equivalent amount of diluted immunotoxin, with infusions pro later revised to ovarian cancer after laparotomy. Sixteen had ceeding 15 min later if no reaction was seen. No adverse reactions to liver mÃ©tastasesdocumented by CT scan; one patient (Patient the test dose were noted. Physical exams and laboratory evaluation, 4) did not. Ten patients also had pulmonary mÃ©tastases.All including hematological and serum chemistry panel and urinalyses, measurable lesions were less than 12 cm in size. Most patients were done daily through study day 6, and then at study days 15, 28, had had the primary tumor removed. Some had received other and 60. Prothrombin time, partial thromboplastin time, complement therapy such as 5-fluorouracil chemotherapy of IL2-LAK cell levels and electrocardiograms were carried out on study days 0 and 6. Where indicated, EEC and CT examinations of the head were per immunotherapy no less than one month prior to immunotoxin formed. Patients were evaluated by sequential chest X-rays or CT scans of the abdomen, CEA levels, blood chemistries, and urinalyses for up Table 1 Characteristics of clinical lots of XomaZyme-791 to 6 months after completion of therapy. In most patients proteinuria was quantitated by dipstick where 1+ = 30 mg/dl and 4+ = 2000 mg/ against cell line' dl. One patient with 4+ proteinuria had quantitation of the urinary molar (IC50)791T antibody9MoAb:RTAsubstitution reactivity%*62 protein over a 24-h period and identification of the protein by urine Lot12Free ratio"1:3.3 41 Molt electrophoresis. 3 ng 62 fig Assessment of Human Immune Response to Immunotoxin. The IgG 1:2Antibody 93Cytotoxicity12ng ND and IgM antibody response to the immunoglobulin moiety or the RTA " Determined by SDS-PAGE. moiety of the immunotoxin were tested on days 0, 13-16, 21-23, 35- * Relative to antibody 79IT/36: determined by competitive inhibition of bind ing of MoAb 791T/36-FITC to 791T cells. 40, and 60 using an ELISA assay. ELISA microplates were coated with c Determined by in vitro cytotoxicity of immunotoxin for cultured target cells. purified 79IT/36 (250 ng/well in PBS) or RTA (2.5 Mg/well in PBS) Cell survival determined by [3H]thymidine incorporation. Values expressed in or purified myeloma lgG2. or IgG2b(Sigma, Poole, UK, 250 ng/ml) in terms of amount of immunotoxin/ml necessary to produce 50% inhibition. 6154

Table 2 Patient population treated with immunotoxin XomaZyme- 791

since other (mg/dl)Site function tests at entry treatment PatientSex1 (months)041None451NoneNone152None174NoneUnknown12NoneLiverdisease*Liver,of W205618172248202233437208158386280519595482118199381SCOT12-45'40243221663561265858401205923303048 F2 lymph nodes, perito neum,lungLiver, F3 spine.lungLiver,lymph nodes,

F4 lung.periaorticlymph nodes, massRight F5 peritoneal mass, peri tonealmetsLiver, F6 lymphnodesLiver,pelvic mass,

M7 lungsLiver, M8 nodesLiver,lymph M9 lungsLiver,lymph nodes, F10 nodesLiver,lung, lymph M11 lungLiver, M12 lungLiver,lymph nodes, F13 lungs.presacrailymph nodes, massLiver, M14 lungs,retroperitonealmass, boneLiver, F15 periaortic nodes, pel vicmassLiver, F16 lymphnodes,peritoneum, boneLiver, F17 tumorLiverBili0.1-1.3'0.30.10.70.30.40.71.70.91.80.50.50.80.70.30.50.40.5LDH50-2recurrent rectal MAge4960306032586967685345586236405070Time " Patients received no other therapy (none) or radiation, chemotherapy (principally 5-fluorouracil) or IL-2/LAC cells (Patient 5). * All except Patient 14 had diagnosis of colorÃ©ela]cancer made histologically. Patient 14 was subsequently rediagnosed as having ovarian cancer. ' Normal range. treatment. Karnofsky scores were greater than 70; all had ated with other renal abnormalities was first noted on days 5- normal levels of blood urea nitrogen and serum creatinine; and 10 of study and increased through day 15 (Fig. \b). Decreased no more than 1+ proteinuria by dipstick. Complement and serum albumin and weight gain occurred before onset of this coagulation parameters were within normal limits as were he delayed proteinuria. In all cases but one, the proteinuria re moglobin and white blood cell counts. Of the 17 patients, five solved to 1+ or less after 30-45 days. The one patient (Patient had normal liver function tests, 10 patients had mildly elevated 3) in whom proteinuria persisted had an accompanying urinary LDH levels, and seven had elevation of SCOT. Most patients tract infection. This delayed proteinuria was noted in 11/11 had values less than twice normal for each test; one had values patients treated with doses at or greater than 0.1 mg/kg/day, less than three times normal, and two had mild bilirubin ele and in three of five patients treated with 0.05 mg/kg/day or vations less than 30% above normal. Serum albumin was nor less. Protein was quantitated in one of two patients with 4+ mal in all. None had other serious diseases or tumors apart proteinuria (receiving a dose of 0.1 mg/kg/day) and totalled from colon (or in one case ovarian) carcinoma. Fifteen patients 2 g protein in a 24-h period. Urine protein electrophoresis received a full course of five doses of immunotoxin; doses were demonstrated that the protein was primarily albumin. Urine temporarily postponed in two patients (Patients 9 and 13) sediment in all patients was unremarkable. In no case was there because of neurological events (increased unilateral tremor and any decrease in serum complement levels (C3, C4, or CH50). transient mental status change) thought possibly due to drug No patient's serum albumin level decreased after the onset of (Table 3). One patient (Patient 6) received only one infusion proteinuria (Fig. 1). No increase in serum creatinine or BUN because of an anaphylactoid reaction consisting of periorbital was seen in any patient. In most patients there was a slight edema. His skin test was negative prior to immunotoxin treat decrease in platelet counts within the first 6 days of an average ment. Another patient (Patient 17) received only four doses of 79,000 which was not related to dose. The lowest count because of mental status change which required more than 3 reached was 146,000 in one patient. Counts returned to baseline days to resolve. or above within 15-20 days in all patients. There was no Clinical Observations. Clinical observations associated with decrease in white blood cells or red cells and no change in immunotoxin therapy are summarized in Table 3. Patients prothrombin time, partial thromboplastin time, or fibrinogen generally tolerated the immunotoxin well. Decrease in serum levels. albumin levels was noted in all patients, beginning during the On study day 6, only one patient (Patient 12) had a significant 5 days of infusion. This occurred with all doses of immunotoxin, increase in any liver function test; this was a twofold increase and serum albumin levels fell to 12-48% of the starting level. in the total bilirubin value which had returned to base line by There was no obvious correlation between dose of immunotoxin day 15 occurring in a patient with liver mÃ©tastases.Over the and degree of albumin drop. In all cases the albumin levels initial 28-day study period, one patient had SCOT and bilirubin either stabilized or began to increase by day 15 (Fig. la). Weight values increasing by at least twofold. gain of 5% or greater was noted in seven of 17 patients and was In one patient, there was worsening of a prÃ©existanttremor manifest as peripheral edema. Pulmonary edema was not seen. of the left hand and four of 17 patients had reversible mental One patient developed fever of 39Â°Cduring or within 6 h of status changes; all had been treated at a dose of 0.1 mg/kg or immunotoxin infusion. Asymptomatic proteinuria not associ higher (Table 3). In three patients, mild fatigue, slurred speech, 6155

Table 3 Clinical observations in patients treated with XomaZyme-791 drop gain Patient dose in serum albu % (peak number123456789IO11121314151617Dose(mg/kg)0.02e0.02e0.050.05e0.050.1"0.10.10.1'0.10.1e0.1e0.1e0.10.150.150.2Total(mg)5.05.013.011.014.56.842.537.525.038.040.026.527.727.445.061.452.8Maximummin(%)1213282445182640452629262228362748Weightday)2(15)2 eventsFevers*37.237.137.938.639.437.938.137.8+

(4)1 (6)3 (6)5 (5)0 (2)4(16)1

(5)12(16)2(15)10(16)12(15)8 37.938.4+

38.338.3+

(8)7(15)4(17)4 38.137.637.8+

(5)14(10)Neurological" 37.8+ 38.5 ' Headache, dementia, or tremor. 4 During therapy. c RTA:MoAb ratio of 2.0:1; other patients treated with lots with RTA: MoAb ratio of 3.5:1. * Received only one dose. ' Postponed 4th and 5th dose study day 2-4.

were most compatible with a mild toxic encephalopathy. The patient who developed dementia had a normal head CT scan and no other etiology for the dementia. Nausea with vomiting was noted in four other patients during therapy; headaches were seen in two. Biological Activity. Although this was a Phase I dose escala tion study, observations concerning antitumor activity were made. Of 16 patients with hepatic mÃ©tastases(Table 4), two (Patients 10 and 16) had objective evidence by abdominal CT scans of decreasing size in large mÃ©tastasesand disappearance of smaller lesions. These changes were seen at 3 months without additional intervening treatment. In another case (Patient 7), new calcification of the hepatic mÃ©tastases,with stabilization of growth, was noted at 2 months, but there was increased size at 6 months. Three patients (Patients 8, 11 and 14) had fixed supraclavicular nodes which decreased in size or disappeared by study day 30. Three also had liver mÃ©tastaseswhich increased in size or number over the 2-month follow-up period. One patient (Patient 6) had both liver and pulmonary mÃ©tastases; the liver mÃ©tastasesincreased in size but the lung mÃ©tastases decreased in size 5 months after therapy. These patients re ceived no additional chemotherapy after the immunotoxin ther apy. Although there were transient decreases in the CEA values of some patients (Table 4), these could not be correlated with decreased tumor size. Of the three patients (Patients 7, 10, and 10 100 14) who had calcification or decrease in size of the hepatic mÃ©tastases,one had CEA values that decreased during the Study Day period of observation. Fig. I. a, serum albumin levels in colorÃ©ela!cancer patients treated with immunotoxin XomaZyme-791 (dose range, 0.02-0.2 mg/kg/day). Study days are Immunological Studies. Thirteen patients were skin tested plotted on a log scale to allow comparison between serum albumin and protein- with the unmodified antibody prior to initiation of therapy and uria, and patients are identified by dose in mg/kg; 0.02 (â€¢),0.05 (D). 0.15 (â€¢), 0.1 (A), and 0.2 (O). A, proteinuria assayed by dipstick, scale 1-4 in colorectal all tests were negative. Between days 10 and 15, onset of patients treated with XomaZyme-791. erythema and induration at the skin test site was noted in 12 of 13 patients. The exception was one patient (Patient 4), irritability, or expressive aphasia were noted. These events treated at 0.05 mg/kg/day. In all cases the reaction lasted 3-5 usually began about study day 4 and were largely resolved within days and then resolved. 2 days although complete resolution could take seven days. One Humoral antibody responses to murine 791T/36 immuno- patient treated at the highest dose (0.2 mg/kg) became frankly globulin and RTA components of the immunotoxin were ob demented; the patient received steroids and this conditon re served in all but one patient, including Patient 6 who received versed after 3 days. EEGs done on the four patients with mental a single injection of immunotoxin (Table 5). Most patients status changes after therapy revealed diffuse slowing and/or produced IgM and IgG responses to 79IT/36 immunoglobulin; paroxysmal bursts, and both the clinical examination and EEC the one patient who did not (Patient 15) was only tested as late 6156

Table 4 Biological responses in patients treated with immunotoxin XomaZyme-791 (ng/ml)Change CEA in Pa liver in tient mÃ©tastases pulmonary in other initials12345678910111213Dose0.020.020.050.050.050.1Â»0.10.10.10.10.10.10.1Changeat 2mos.StableIncreasednumberStableN/AIncreasednumberEnlargedStablemÃ©tastasesIncreased mÃ©tastasesN/AN/AN/APeritoneal sizeIncreased numberStableN/AN/AIncreased

mÃ©tastasesin sizeIncreasedcreased in peritonealmassN/AN/ATransientsize,

numberat 2monthsbut decreasedsize at5monthsN/AN/AStableNo

withcalcificationIncreasednumberStableDecreasednumberIncreasednumberIncreased

sizeL decrease in nodeN/AN/AResolutionsupraclavicular

new,somedecreased insizeAppearance

ofnew supraclavicularnodeN/AIncreased lesionsat2 mos.Increased sizeIncreased numberIncreased sizeSerum sizePre13419220181183285804522>10002256967TherapyPost"19315571901794425971024>100046521000Changeretroperitoneal adenopathy 14 0.1 Increased size N/A 17 18 Resolution supraclavicular mass 15 0.15 N/D N/A 142 383 N/A 16 0.15 Decreased size N/A 1000 1000 N/A 17 0.2 Stable N/A 498 374 N/A "Study day 15. * Received only one dose. as study day 19. The IgM response was first detected between 79IT cells. Anti-combining site antibodies were detected in sera days 5 and 20 of study, with maximum responses around day from 14 of 17 patients (Table 5 and Fig. 3). Peak serum titers 30. The overall pattern was that of a rapid rise in the IgM were variable ranging from up to 1:5 in two patients to 1:100 antibody response and then a fall in antibody levels up to day or greater in 12 patients. The kinetics of this response in 80 of study. Peak IgM responses against the MoAb portion of patients is illustrated in Fig. 3 showing that the response begins the immunotoxin ranged in titer from 0 to 1:500. about days 10 to 30. In most instances, antibody levels remained The IgG response to 79IT/36 immunoglobulin was markedly elevated up to day 70. In three patients there was a pronounced greater than the IgM response, with peak antibody titers of fall of the anti-combining site antibody titer. 1:7000 or greater observed in four patients. In comparison, the Antibody responses to RTA were detected in sera from 15 maximum IgM antibody titer was 1/500 (Patient 12). The patients (Table 5 and Fig. 4). One patient (Patient 15) receiving pattern of the IgG response was different from the IgM response 0.15 mg/kg/day of immunotoxin (total dose 45 mg) did not with IgG antibodies continuing to rise from approximately days produce antibody to either RTA or 79IT/36 immunoglobulin. 10 to 20 and remaining elevated for up to 80 days of study. A second patient (Patient 16) treated at the 0.15 mg/kg dose (Fig. 2). (total dose 61.4 mg) also produced only minimal responses to Sera were screened for antibodies recognizing 79IT/36 im RTA (and also 79IT/36). Significant titers of IgM antibodies munoglobulin (IgG2b) in comparison with nonspecific mouse to RTA were observed in 12 patients, titer range 1/50 to IgG2b and IgG2a immunoglobulins to determine if the predom 1/16,600 (Table 5). IgG anti-RTA antibodies were detected in inant response was anti-mouse IgG, anti-subclass, or anti-vari 13 patients with titers ranging from 1/50 to 1/316,000. The able region (Table 5 and Fig. 2). The predominant response kinetics of the anti-RTA IgG response (Fig. 4) indicates a rapid was specific for the 79IT/36 immunoglobulin, where peak rise in antibody titer between study day 10 to 20, this being 5 serum titers greater than 1:1000 were obtained in samples from to 15 days after completion of immunotoxin treatment. In one 11 patients. In 13 of 17 patients the anti-79IT/36 response was of the patients (Patient 9) who generated a very pronounced at least twice as high as the response to mouse myeloma IgG2b response, the anti-RTA antibody levels decreased over a period or IgG2a. of 70 days. In most patients, however, the IgG antibody level The pronounced antibody response to 79IT/36 immunoglob remained elevated throughout the 80-day period of investiga ulin (IgG2b), compared to normal mouse IgG2b and IgG2a sug tion. gested that the patients generated antibody to the monoclonal Animal Toxicology antibody 79IT/36 combining site. This was confirmed using a flow cytometry assay which measured the capacity of patients' The LD50 level of the immunotoxin assessed in BALB/c mice serum to block binding of fluorescein labeled 79IT/36 (FITC- receiving a single i.p. injection of doses from 10 to 100 mg/kg, 79IT/36) with the target antigen on tumor 79IT cells (18, 20). given in a volume of 0.5 ml, was calculated to be 81 mg/kg. In Titers were determined as the dilution of patients' serum which the subacute toxicology studies, rats receiving the highest dose produces a 50% inhibition of FITC 79IT/36 binding to target of immunotoxin had a significant decrease in body weight from 6157

Table 5 Antibody responses to immunotoxin XomaZyme-791 in colorectal cancer patients against*Patient1234567891011121314151617Immunotoxin Peak antibody titers

tested dose per kg/ thru study IgGa IgGc. anticombining total0.02/5.00.02/5.00.05/13.00.05/11.00.05/14.50.1/6.80.1/42.50.1/37.50.1/25.00.1/38.00.1/40.00.1/26.50.1/27.70.1/27.40.15/45.00.15/61.40.2/52.8Seraday:2836283723146041761431274331192172XMMCO-791Skintest"IgM+ IgG18016070500750250251,0001,780404,0005001,25010,0000101,600MurineIgG160606311028070152801,70004,000801,6003,2000101,400RTAIgM2,1000021,70016,6001,780350250100802,5000500550ND5,000IgG500400201,6005202,50010,000316,000158,5006,3002803,5502,6600ND630Peaksitetiter5,00010051001,00033010020006002,5003,1003,20000140 125+ 40+ 2050NT7 125+ 10+ 30-1- 140+ 100+ 100+ 64NT 500NT 70NT 80+ 0+ 0+ 80IgG3501705002,7804,7501,0001,0007,0004,0002007,0002,0004,70010,00001007,000Murine " Late erythrema and induration appearing at site of skin test. Patients 5, 12, 13. and 14 received i.v. challenge instead of skin test. * Expressed as reciprocal of the serum dilution. c NT, not tested.

101 !

10'

10* cc LU

CE LU M

10Â° kâ€”A

TIME! days I Fig. 2. Kinetics of the IgG and IgM antibody responses to 79IT/36 immunoglobulin (â€¢),mouse myeloma IgG2b (O), and mouse myeloma IgG2a (A) in -10 0 10 20 30 40 50 60 70 80 two patients; nos. 17 (a, b) and 7 (c, d). Serum titers determined as serum dilution producing 50% of maximum reaction in ELISA. IgG responses (a, c); IgM responses (ft, d). STUDY DAYS Fig. 3. Serum anticombining site antibody titers in patients treated with XomaZyme-791. Symbol, individual patients. day 3 through day 10 of the study followed by a weight gain toxin although this effect was not seen until day 11. This from day 10 through the end of the study. They also had a persisted through day 17 of the study; i.e., 1 days after the last significant increase in absolute neutrophil counts to seven times injection. Mild renal tubular necrosis was seen at day 6 in the higher than that seen in the control animals; this had returned group receiving the highest dose; this was reversing by day 11. to baseline by day 11 (data not shown). Serum albumin was significantly decreased in the high dose group on day 6 and 11 DISCUSSION of study but had returned to baseline by day 17 (Fig. 5). There were no changes in other serum chemistries including liver This Phase I study on the administration of XomaZyme-791 function tests, CPK, LDH, or renal function tests. immunotoxin in patients with metastatic colorectal cancer has Histopathologically, there was mild vacuolization in hepato- established the side-effects and clinical responses in patients cytes in groups receiving either 1 or 5 mg/kg/day of immuno- receiving doses ranging from 0.02 mg/kg/day for 5 days 6158

106 petite has been noted with other immunotoxins (4, 23). Delayed asymptomatic proteinuria was also a common finding associ ated with therapy with XomaZyme-791 in these patients. This generally began on study days 10-15 and reversed by days 30- 105 45. It began in the presence of a serum albumin level which was either stable or increasing, and thus could not be implicated as a cause of the early drop in serum albumin. 79IT/36 MoAb is 104 - known to bind to stromal (noncellular) elements of normal tissues, and it is possible that the antibody portion of the immunotoxin has a longer dwell time in organs such as the skin and kidneys because of its stromal binding. The temporal association between skin test flare, onset of proteinuria, and CC 103 - UJ humoral immune response suggests that the onset of protein uria could be related to the onset of the immune response causing a mild and reversible antigen-antibody complex disease. oc 10* LU There was poor correlation between antibody titers and pro (A teinuria in individual patients, and two patients (Patients 1 and 4) with antibodies did not develop proteinuria. However, the phenomena has been reproduced in a rat model, showing that 101 â€¢¿ the onset of proteinuria occurs at the time that antibody re sponses to immunotoxin components are detected, and is ac celerated by passive transfer of antibodies to the 79IT/36 MoAb.3 10Â° Other observations were more subtle. A decreased platelet count was noted in most patients, but this was not clinically significant since the platelet count remained above 100,000. There was no indication in the preclinical studies that this 10 antibody bound to human platelets or progenitor cells, and the -10 O 10 20 30 40 50 60 70 80 effect was not seen in the animal toxicology studies. The neu rological abnormalities noted with XomaZyme-791 have not STUDY DAYS been seen with other immunotoxins, but may have been over Fig. 4. Serum IgG antibody responses to ricin A chain in patients treated with looked due to the mildness of the abnormality. XomaZyme-791. Symbols, individual patients. The same symbols are used as in Figs. 3. Neither the decreased serum albumin, constitutional changes, or protein urea were dose limiting. One patient treated at the highest dose developed a marked mental status change and as a result only received 4 doses of drug. This patient was also the oldest subject treated and had several documented episodes of atrial fibrillation during the 3 days of observation after termi nation of therapy. However, he had no prior evidence of mental status change and no evidence by computerized tomography of the head of any abnormalities and it was possible that this was a true effect of drug. Additional patients were thus accrued at lower doses but although subtle mental status changes were noted the finding was not repeated. Thus a dose range of 0.1- 0.15 mg/kg/d for 5 days has been established as safe, and it is 2.0 17 felt doses above this should be monitored carefully. All but one of the patients produced IgG and IgM responses Fig. 5. Albumin levels of Sprague-Dawley rats dosed i.v. with saline or 1 mg/ to the immunoglobulin and ricin A chain components of the kg or 5 mg/kg XomaZyme-791 for 10 days. There are 12 animals per dosage immunotoxin, the anticombining site response being a predom group and each point represents three animals bled before death and necropsy. D, control; Â®,1.0 mg/kg; B, 5.0 mg/kg. inant feature of the antiimmunoglobulin response. Both the antiisotypic and anticombining site antibodies would be ex pected to alter the biodistribution and pharmacokinetics of the through 0.2 mg/kg/day for up to 5 days. immunotoxin. The influence of antiisotypic antibodies on the Animal toxicology studies predicted the decreased albumin pharmacokinetics of XMMCO-791 antibody has been demon which has been seen in animals treated with RTA alone, and in strated in rats (24). That study showed that the blood survival animals and humans treated with other immunotoxins (4, 22). of l25I-791T/36 as well as normal mouse IgG2b was markedly Clinically this results in a mild weight gain, and is not dose reduced with an associated liver uptake in rats producing anti- limiting. Hepatocyte vacuolization was seen in animals treated 79IT/36 antibody. The influence of antibodies reacting with with immunotoxin, but there was no increase in liver function the combining site of 79IT/36 immunoglobulin has also been tests, nor was there evidence of hepatic dysfunction in the demonstrated using BALB/c mice immunized with 791T/36- patients, even though most had liver mÃ©tastasesand abnormal RTA. In this case the biodistribution of 79IT/36 immunoglob liver function tests prior to immunotoxin therapy. Weight loss, ulin (but not normal mouse IgG2b) was modified in immunized commonly seen in animals receiving higher doses, was not seen in humans receiving this immunotoxin although decreased ap 3 V. S. Byers et a/., unpublished observations. 6159

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1989 American Association for Cancer Research. MoAb-RICIN A CHAIN IMMUNOTOXIN XomaZyme-791 mice (25). It is expected that anti-ricin A chain antibodies will 10. Armitage, N. C., Perkins, A. C., Pimm, M. V., Farrands, P. A., Baldwin, R. W., and Hardcastle, J. D. The localisation of an anti-tumour monoclonal also influence immunotoxin biodistribution and pharmacoki- antibody (79IT/36) in gastrointestinal tumours. Br. J. Surg., 71: 407-412, netics. This issue is currently being investigated and it has been 1984. shown that murine monoclonal anti-RTA antibodies increase 11. Armitage, N. C., Perkins, A. C., Pimm, M. V., Wastie, M. L., Baldwin, the hepatic uptake of '"I-RTA.4 R. W., and Hardcastle, J. D. Imaging of primary and metastatic colorectal cancer using an 111-In-labelled ant Â¡tumourmonoclonal antibody (79 IT/36). Evidence of biological activity was seen in the three patients NucÃ.Med. Commun., 6: 623-631, 1985. 12. Embleton, M. J., Byers, V. S., Lee, H. M., Scannon, P.. Blackhall, N. W., with hepatic mÃ©tastases,who had initial calcification of lesions and Baldwin, R. W. Sensitivity and selectivity of ricin toxin A chainâ€” followed by resolution of the smaller lesions, and fading of the monoclonal antibody 79IT/36 conjugates against human tumor cell lines. larger lesions. One of the patients with pulmonary mÃ©tastases Cancer Res., 46: 5524-5528, 1986. had a decrease in size of the lesions; another with a supracla- 13. Byers, V. S., Pimm, M. V., Scannon, P. J., Pawluczyk, I. Z. A., and Baldwin, R. W. Inhibition of growth of human tumor xenografts in athymic mice vicular node had resolution of that lesion but concomitant treated with ricin toxin A chainâ€”monoclonal antibody 79IT/36 conjugates. Cancer Res., 47: 5042-5046, 1987. growth of others. 14. Embleton, M. J., Gunn, B., Byers, V. S., and Baldwin, R. W. Antitumour Pharmacokinetic studies have demonstrated that ricin A reactions of monoclonal antibody against a human osteogenic sarcoma cell chain immunotoxins are rapidly taken up by the liver, since the line. Br. J. Cancer, 43: 582-587, 1981. 15. Falini, B., and Taylor, C. R. New developments in immunoperoxidase ricin A chain contains an oligosaccharide which is recognized techniques. Arch. Pathol. Lab. Med., 107:105-122, 1983. by receptors on the hepatic Kupffer cells (26-28). Improve 16. Reisner, Y., Kapoor, N. K., O'Reilly, R. J., and Good, R. Allogeneic bone ments in immunoconjugate design should prolong blood cir marrow transplantation using stem cells fractionated by lectins: VI, in vitro analysis of human and monkey bone marrow cells fractionated by sheep red culation time and tumor localization. Methods to control the blood cells and soybean agglutin. Lancet, 2: 1320-1324, 1980. immune response to immunotoxin are under development, and 17. Kernan, N. A., Knowles, R. W., Burns, M. J., Broxmeyer, H. E., Lu, L., Lee, H. M., Kawahata, R. T., Scannon, P. J., and Dupont, B. Specific inhibition will allow retreatment. Both these measures should lead to of in vitro lymphocyte transformation by an anti-pan T cell (gp67) ricin A increased efficacy of the immunotoxin. chain immunotoxin. J. Immunol., 133:137-146, 1984. 18. Roe, R., Robins, R. A., Laxton, R. R., and Baldwin, R. W. Kinetics of divalent monoclonal antibody binding to tumour cell surface antigens using flow cytometry: standardization and mathematical analysis. Mol. Immunol., 22:11-21, 1985. REFERENCES 19. Durrani, L. G., Byers, V. S., Scannon, P. J., Rodvien, R., Grant, K., Robins, R. A., Marksman, R. A., and Baldwin, R. W. Humoral immune response to XMMCO-791-RTA immunotoxin in colorectal cancer patients. Clin. Exp. 1. Baldwin, R. W., and Byers, V. S. Monoclonal antibodies and immunocon- Immunol., 75: 258-264, 1989. jugates for cancer treatment. In: H. M. Pinedo, D. L. Longo, and B. A. 20. Robins, R. A., Laxton, R. R., Garnett, M., Price, M. R., and Baldwin, R. W. Chabner (eds.). Cancer Chemotherapy and Biological Response Modifiers Annual, Vol. 9, pp. 409-431. Amsterdam: Elsevier Science Publishers B.V., Measurement of tumour reactive antibody and antibody conjugate by com petition, quantitated by flow cytofluorimetry. J. Immunol. Meth., 90: 165- 1987. 172, 1986. 2. Neville, D. M., Jr. Immunotoxins: current use and future prospects in bone 21. Price, M. R., Campbell, D. G., and Baldwin, R. W. Identification of an anti- marrow transplantation and cancer treatment. CRC Crit. Rev. Ther. Drug human osteogenic sarcoma monoclonal-antibody-defmed antigen on mito- Carrier Systems, 2: 329-351, 1986. gen-stimulated peripheral blood mononuclear cells. Scand. J. Immunol., IS: 3. Vitetta, E. S., Fulton, R. J., May, R. D., Till, M., and Uhr, J. W. Redesigning nature's poisons to create anti-tumor reagents. Science (Wash. DC), 238: 411-420, 1983. 22. Harkonen, S., Stoudemire, J., Mischak, R., Spitler, L. E., Lopez, H., and 1098-1104. 1987. Scannon, P. Toxicity and immunogenicity of monoclonal antimelanoma 4. Spitler, L. E., Del Rio. M.. Kentigan, A., Wedel, N. I., Brophy, N. A., Miller, antibody-ricin A chain immunotoxin in rats. Cancer Res., 47: 1377-1382, L. L., Harkonen, W. S., Rosendorf. L. L., Shannon. C. E., Lee, H. M., 1987. Mishak, R. P., Kawahata, R. T., Stoudemire, J. B., Fradkin, L. B., Bautista, 23. Byers, V.. Henslee, P., Kernan, N., Blazar, B., Gingrich, R., Phillips, G., E. E., Dellio, C. L., Mendell, S. C, and Scannon, P. J. Therapy of patients Antin, J., Mischak, R., O'Reilly, R., and Scannon, P. Therapeutic response with malignant melanoma using a monoclonal antimelanoma antibody-ricin to a pan T lymphocyte monoclonal antibody-ricin A chain immunotoxin in A chain immunotoxin. Cancer Res., 47: 1717-1723, 1987. steroid refractory graft versus host disease (GVHD) (Abstract). Blood, 5. Thorpe, P. E., Wallace, P. M.. Knowles, P. P., Reif, M. G., Brown, 70(suppl. 1):304A, 1987. A. N. F., Watson, G. J., Knyba, R. E., Wawrzynczak, E. J.. and Blakey, 24. Pimm, M. V., Perkins, A. C., Durrani, L. G., and Baldwin, R. W. A rat D. C. New coupling agents for the synthesis of immunotoxins containing a model for imaging the effecl of anli-mouse anlibody responses on Ihe hindered disulfide bond with improved stability in vivo. Cancer Res., 47: biodislribulion of radiolabeled mouse monoclonal anlibodies. Eur. J. NucÃ. 5924-5930, 1987. Med., 14: 507-511, 1988. 6. Endo, Y., and Tsurugi, K. RNA N-glycosidase activity of ricin A chain. 25. Pimm, M. V., Durrani, L. G., and Baldwin, R. W. The influence of syngeneic J. Biol. Chem., 262: 8128-8130, 1987. anliidiotypic anlibody on Ihe biodislribulion of an ant i tumor monoclonal 7. Campbell, D. G., Price, M. R., and Baldwin, R. W. Analysis of a human antibody in BALB/c mice. Int. J. Cancer, 43: 147-151. osteogenic sarcoma antigen and its expression on various human tumour cell 26. Byers, V. S., Pimm, M. V., Pawluczyk, I., Lee, H. M., Scannon, P. J., and lines. Int. J. Cancer, 34: 31-37, 1984. Baldwin, R. W. Biodislribulion of ricin loxin A-chain monoclonal anlibody 8. Durrani, L. G.. Robins, R. A., Armitage, N. C., Brown, A., Baldwin, R. W., 79IT/36 immunoloxin and Ihe influence of hepalic blocking agenls. Cancer and Hardcastle, J. D. Association of antigen expression and DNA ploidy in Res., 47: 5277-5283, 1987. human colorectal tumors. Cancer Res., 46: 3543-3549. 1986. 27. Bourrie, B. J. P., Casellas, P.. Blylhman, H. E., and Jansen, F. K. Sludy on 9. Powell. M. C., Perkins, A. C., Pimm, M. V., Jetaily, A., Wastie, M. L., Ihe plasma clearance of anlibody-ricin A chain immunoloxins. Evidence for Durrani, L., Baldwin, R. W., and Symonds, E. M., Diagnostic imaging of specific recognition siles on the A chain lhat mediate rapid clearance of Ihe gynaecological tumors using the monoclonal antibody 79IT/36. Am. J. immunoloxin. Eur. J. Biochem., 755: 1-10, 1986. Obstet. Gynecol., 157: 78-84, 1987. 28. Blakey, D. C., Walson, G. J., Knowles, P. P., and Thorpe, P. E. Effecl of chemical deglycosylation of ricin A chain on the in vivo fate and cytoloxic activily of an immunotoxin composed of ricin A chain and anli-Thy 1.1 ' M. V. Pimm, unpublished observations. anlibody. Cancer Res., 47: 947-952, 1987.

6160

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 1989 American Association for Cancer Research. Phase I Study of Monoclonal Antibody-Ricin A Chain Immunotoxin XomaZyme-791 in Patients with Metastatic Colon Cancer

V. S. Byers, R. Rodvien, K. Grant, et al.

Cancer Res 1989;49:6153-6160.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/49/21/6153

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://cancerres.aacrjournals.org/content/49/21/6153. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.