Antibody Delivery and Effector Cell Activation in a Phase II Trial Of

[CANCER RESEARCH 48, 2568-2573, May 1, 1988] Antibody Delivery and Effector Cell Activation in a Phase II Trial of Recombinant 7-Interferon and the Murine Monoclonal Antibody CO 17-1A in Advanced Colorectal Carcinoma1

Louis M. Weiner,2 Philip J. Moldofsky, Robert A. Gatenby, Joyce O'Dwyer, Jacqueline O'Brien, Samuel Litwin, and Robert L. Comis Departments of Medical Oncology [L. M. W., J. O. D., J. O. B., R. L. C.J, Radiology (P. J. M., R. A. GJ, and Biostatistics [S. L.], Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111

ABSTRACT when labeled with n'I, can concentrate in colorectal carcinoma lesions in these nude mice and in many patients (8, 9). Murine monoclonal antibodies of the immunoglobulin G2a isotype These characteristics formed the basis for a series of clinical interact with human effector cells to mediate antibody-dependent cellular trials, initiated in 1980, using this antibody. Hundreds of pa cytotoxicity (ADCC) directed against malignant cells which express antigens recognized by these antibodies. f-Interferon enhances these tients have been treated worldwide; while occasional objective effects in vitro. In a Phase I trial of murine monoclonal antibody CO17- clinical responses have been reported in patients with metastatic 1A and recombinant -y-interferon (r!FN--y), we demonstrated that low colorectal or pancreatic carcinomas, no consistent patterns of doses of rIFN-7 were superior to high doses in augmenting ADCC response have been identified (10-17). These results have fo mediated by treated patients' monocytes. These results formed the basis cused our attention on the obstacles to success of this therapy, for a Phase II trial of CO17-1A combined with low dose rlFN-f. Nineteen with emphasis upon evaluating therapeutic antibody delivery patients with metastatic colorectal carcinoma were treated with four and potentiating the cellular effector mechanisms mediated by consecutive daily infusions of 1.0 x 1(1"II /nr rIFN-T, with 150 mg of this antibody. CO17-1A administered on days 2, 3, and 4. Therapy was tolerated well. rIFN-7 enhances CO17-lA-directed ADCC mediated by hu Peripheral blood mononuclear cells were purified from patient samples man monocytes against the SW1116 colorectal carcinoma cell obtained at baseline and at 1, 4, or 24 h following the start of the first rIFN-7 infusion and were tested for their ability to lyse '"In-labeled cells line (5, 18). These results led to a Phase I clinical trial of combined therapy with rIFN-7 and CO17-1A in patients with of the SW1116 colorectal line. Enhancement of monocyte ADCC was seen by 24 h, while lymphocyte ADCC and natural killer activity directed advanced gastrointestinal carcinoma. In this study, it was de against K562 cells were enhanced to a lesser extent. Nonspecific lysis of termined that peripheral blood monocytes from patients treated SVVII16 cells by effectors was not seen at the time points examined. with low doses of this cytokine (<1.0 x IO6 IU/m2/day for 4 While CO17-1A antigen expression was observed in most biopsies, ' "I- consecutive days) augmented ADCC directed against SW1116 labeled CO17-1A imaged positively in less than one-half of the organs cells significantly more than monocytes from patients treated known to harbor mÃ©tastases,and therapeutic antibody delivery was not with higher doses (19). High dose interferon therapy caused always demonstrated by immunoperoxidase staining techniques of tissue more nonspecific monocyte cytotoxicity than low dose therapy, obtained following therapy. In antigen-positive lesions, tissue p<>.â€¢levels thus partially offsetting this ADCC advantage (20). However, appear to identify lesions which would image positively. No objective the enhanced target cell specificity and minimal toxicity seen responses were seen. Our findings suggest that prolonged therapy with low doses of rIFN--x potentiates ADCC but that physiological obstacles with low dose therapy formed the rationale for the current to therapeutic antibody delivery are significant. In order to evaluate the clinical study, a Phase II trial of low dose rIFN-7 therapy validity of this therapeutic approach, measures to enhance antibody administered with CO 17-1A in patients with advanced colorec delivery are needed, starting with systematic evaluations of therapy with tal carcinoma. escalating doses of CO17-1A combined with low dose rIFN--x therapy. In addition to efficacy, we studied the effects of rIFN--y on the cellular effector mechanisms of monocyte and lymphocyte ADCC and lymphocyte NK activity. We also examined the INTRODUCTION relationship of antigen expression to labeled antibody imaging, The murine monoclonal antibody CO 17-1A binds to a cell tumor oxygen tension, and therapeutic antibody delivery. surface glycoprotein expressed preferentially by adenocarcino- mas of the gastrointestinal tract (1-3). This antibody, which is MATERIALS AND METHODS of the IgG2a isotype, does not fix complement but participates in ADCC3 mediated by human monocytes and lymphocytes Patients. Nineteen patients with histologically documented met directed against carcinoma cells which express the target anti astatic colorectal carcinoma were treated. All patients had measurable gen (4-6). Injection of the antibody into nude mice bearing disease and Eastern Cooperative Oncology Group performance status appropriate human tumor xenografts induces significant of 0 (n = 14) or 1 (n = 5). Five had received no prior systemic therapy, growth delays of the grafts and under selected conditions may while 14 had been treated with one regimen, but not for at least 4 cure the animals (7). The antibody or its F(ab')2 fragments, weeks. The biperpendicular products of all measurable lesions in each patient were added to yield an estimate of relative tumor burden; the values ranged from 2.2 to 181.3 cm2 with the median sum of these Received 6/9/87; revised 1/14/88; accepted 2/3/88. The costs of publication of this article were defrayed in part by the payment products being 66.6 cm2 in the 19 patients. Pertinent features are of page charges. This article must therefore be hereby marked advertisement in described in Table 1. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' This study was supported by funds from the Frank Strick Foundation. Treatment. Patients were hospitalized for treatment. Recombinant 2 Recipient of a National Cancer Institute Clinical Investigator Award. To 7-interferon (Biogen, Cambridge, MA), 1.0 x IO6 IU/m2, was admin whom requests for reprints should be addressed, at Department of Medical istered by a 4-h i.v. infusion daily for 4 consecutive days. The murine Oncology, Fox Chase Cancer Center, Central and Shelmire Avenues, Philadel monoclonal antibody CO 17-1A (Centocor, Malvern, PA) was admin phia, PA 19111. 3The abbreviations used are: ADCC, antibody-dependent cellular cytotoxicity; istered at a dose of ISO mg over 2 h at the conclusion of each interferon rIFN-f, recombinant -y-interferon; NK, natural killer cell; CT, computer-assisted infusion on days 2, 3, and 4 (cumulative dose, 450 mg). One 4-day tomography. course of therapy was given to each patient. 2568

Table 1 Patient characteristics of all In perpendicular products of all Patient measurable le code28293031323334353637383940414243444546Age(yr)44394963583650645753645753446057444956SexMFMMFFMMMFFMFFMFMFFECOGP.S."000001000111000000!PriortherapyNoneNoneXRTIL-2 ofmÃ©tastasesLungsLiverLiver(L) sions(cm2)2.21877.019.962.572.0*181.34.066.640.761.4104.829.27.049.2135.284.6140.4105.6

+IFN-05-FUraXRT5-FUraXRT5-FUraNone5-FUraNone5-FUra5-FUraTMX Supraclavicu-lar and (L) axil larynodesLiverPeripancreaticnodesLungs,

(L)supraclavicularnodesLiverLiverLiverPeriaorticliver,

nodesLiver, abdominalwall, (L)supra-clavicular node(R) LungLungs, +5FUraNoneNoneFdUrd liverLiverLiver,

nodesPeripancreaticnodesLiver,celiac +MMC5-FUra

+MMCNoneIL-2 nodesLiverLiverLungsLiverSumpelvic

+IFN-0TMX +5-FUraXRTIL-2

+IFN-0TMX +5-FUraSites

* XRT, radiation therapy; IL-2, recombinant interleukin 2; IFN-/J, recombinant (i-interferon; 5-FUra, 5-fluorouracil; TMX, trimetrexate; FdUrd, floxuridine; MMC, miteirmein C; ECOG P.S., performance status criteria of the Eastern Cooperative Oncology Group; L, left; R. right. * Multiple small pulmonary mÃ©tastasesnot included in this sum.

Monitoring Studies. Whole blood (100 ml) was drawn from all The midsection of a tumor and all sites of measurements were localized patients at the start of therapy and at two or three other time points with CT scanning using 8-mm sections. The sensing electrode is nega during the first and second days of the treatment course. Normal control tively polarized (â€”0.75V) with a reference electrode which is placed samples were obtained for each day of assay. Five ml of plasma were just under the skin near the area of interest to allow electrical continuity obtained at baseline, 30 min, 1 h, 2 h, and 4 h following the start of with the sensing electrode. the first interferon infusions and at the beginning and end of the The sensing electrode is then inserted into the tumor, using CT interferon infusions of the subsequent day. Daily complete blood counts guidance for approximately 5 min to form a "tissue membrane" over and blood chemistries were performed, as were baseline electrocardi the platinum wire. Current readings are obtained by passing the elec ographs, and chest X-rays. Tumor measurements were obtained by trode through the tumor at approximately 5-mm increments. To pre physical examination, X-rays, or CT scans and were obtained monthly vent artifact from tissue damage, measurements are taken only when until there was evidence of progressive disease. the needle is being inserted and never while being withdrawn. Because Immunohistochemistry. Fine needle aspirations of accessible tumor the electrode motion creates artifact, the probe remains in position at deposits were performed in a number of patients and sent for cytological each site until the current output is stable (this rarely requires more evaluation; all demonstrated adenocarcinoma. Aliquots of tissue were than 30 s). The site of each measurement was documented by localizing snap-frozen and embedded in O.C.T. Compound (Tissue-Tek II; Miles the tip of the sensing electrode with the CT scan. Generally, three Laboratories, Naperville, IL), and frozen sections were made for im- passes were performed separated by at least 1 cm so that a total of 7 to munohistological analysis. A slightly modified immunoperoxidase 25 measurements were obtained for each tumor, depending on the staining method was used (Vectastain ABC kit; Vector Laboratories, tumor size. Current readings in normal adjacent tissue were also meas Burlingame, CA). All samples were evaluated for binding by (a) no ured. After each study was completed, the electrode was calibrated at primary antibody; (b) Pj supernatant from MOPC-21 myeloma cells; 37Â°Cin normal saline, using 0 and 20% oxygen levels so that specific and (c) CO 17-1A. P<>.-values could be assigned to each current reading. With each batch of patient slides, negative and positive controls were Monoclonal Antibody Imaging. CO 17-1A antibody or its fragments evaluated for binding of P3 and the monoclonal antibody 19-9 (Cento- were labeled as described (7). Clinical imaging was performed by using cor), respectively, to slides coated with SW1116 cells. Results were previously reported techniques (22). After assuring a negative i.v. test interpreted as 0 (similar to negative control), 2 (similar to positive dose, '"I F(ab')2 fragments were administered via an i.v. line established control), or 1 (intermediate stain intensity); only the areas of specimens with a 15 nun infusion in 0.9% NaCI solution. Images were acquired with histolÃ³gica!preservation of the cells were considered interpretable. with either a Picker Dynacamera 4/15 equipped with parallel-hole CT Scan-guided Oxygen Probes. Direct measurement of tissue oxygen medium-energy collimator or with the GE Maxicam 400 SPECT tension in tumors was performed by using a CT-guided probe as system. Images were stored and processed on Medical Data Systems described previously (21). The sensing electrode is a platinum wire (200 A2computer. Known or suspected tumor deposits were imaged at 72 h /Â¿mindiameter) insulated with glass and mounted in a 22-gauge needle. postadministration. Usually, three views (anterior, posterior, and right 2569

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1988 American Association for Cancer Research. PHASE II TRIAL OF rlFN-> AND MONOCLONAL ANTIBODY 17-1A lateral) of abdomen/pelvis were obtained. A total of 100,000 to 500,000 obstruction secondary to peritoneal carcinomatosis. counts per view were collected. During the imaging session, RBC were One patient (L. B.) died at home suddenly, 3 weeks after labeled with 2 mCi "Te in vivo as in routine blood pool imaging. A "Tc-RBC image corresponding to each 13IIimage was obtained for completing therapy. While the history suggests a cardiac event or a pulmonary embolus, a postmortem examination was not subsequent background subtraction with the patient remaining in the same position. "'I scatter into the technetium energy window is found obtained. The other 18 patients are Ã©valuableforresponse with a median follow-up of 8.0 months. to be minimal, as required for valid background subtraction. Typically, 2% of the counts representing the tumor in ""Te image are contributed Efficacy. Of the 18 Ã©valuablepatients, 17 have progressed by scatter from the I3II present in the patient. Subtraction was per objectively, all within 2 months. No objective responses, of formed on the computer by placing a region-of-interest outline over the either partial or less than partial nature, have been observed. heart (blood pool) in corresponding WmTCand I3II images and normal One patient has had stable disease for 2 months at the time of izing to make the total count in this region equal on both images prior manuscript submission. to subtracting *9nTC from 131I.Images were evaluated both with and without blood pool subtraction. Antibody Delivery to Tumor Deposits Quantitative techniques were applied to determine the amount of ' " I Sixteen patients were imaged with I31l-labeled CO17-1A localized in tumor (23). To avoid false positive image interpretations F(ab')2 fragments or whole antibody. Images which demon caused by simple blood pool activity, a scan was called positive for specific localization of antibody fragments only if (a) the site was visible strated focal uptake in areas of known tumor in eight patients on the unprocessed I31IF(ab'>2 image and (ft) the site was less active on were positive in both unaltered and computer-subtracted images blood pool image than on I3II F(ab')2. Thus, decisions were made on which corrected for blood-pooling artifacts. Most positive im the unprocessed images, with subtracted images being used only to ages were associated with liver mÃ©tastases;7 of 11 clinically eliminate false positives. involved livers imaged positively. Lesions in other organ sites Laboratory Studies. Monocytes were purified by reversible adherence were identified in seven patients but only a pulmonary lesion to gelatin-autologous plasma-coated plastic flasks as described (24). imaged positively. 131Idelivery to image-positive tumor masses The nonadherent cells which were obtained during monocyte purifica ranged from 0.001 to 0.016% of the administered dose of I3'I/ tion were characterized as lymphocytes by staining with a standard ml tumor, and tumor/liver ratios ranged from 1.0 to 8.0. The panel of antileukocyte antibodies (data not shown). percentage of administered I3II which was delivered to total Monocytes and lymphocytes were assayed for their ability to lyse cells of the colorectal carcinoma cell line SW1116 by using an 18-h image-positive tumor volumes ranged from 0.00051 to 4.0% '"In release assay as described (19). Since this cell line expresses the with a median of 0.867% and a mean value of 1.10981 Â± target antigen recognized by CO 17-1A, these cytotoxicity assays were 0.385% (SE). performed in the presence or absence of 100 Mg/ml CO 17-1A. Lower concentrations of antibody also mediate ADCC, but antibody was used Detection of Antigen Expression by Immunohistology in excess to ensure that Fc receptors and tumor antigens were saturated in all assays. ADCC values were calculated by subtracting the percent Antigen expression was determined by biopsy of accessible age of specific lysis mediated by an individual effector cell population deposits in 13 patients. Six of the eight image-positive lesions from that of an identical population incubated with CO 17-1A prior to expressed the target antigen of CO 17-1A; the other two lesions addition of the labeled targets. were not biopsied. Seven lesions which did not image were Natural killer cell activity of the purified lymphocytes was evaluated biopsied, and 6 of 7 lesions expressed the target antigen as well. by using the human erythroleukemia cell line KS62 as the target in an Only one lesion, a left supraclavicular nodal mass, failed to 18-h '"In release assay performed as described above. express the CO 17-1A antigen in this series. Response Evaluation. Responses were evaluated by using standard Eastern Cooperative Oncology Group criteria (25) and were categorized Assessment of Therapeutic Antibody Delivery by Immunohistol as either complete (complete regression of all measurable and Ã©valuable disease), partial (greater than 50% decrease of biperpendicular diameter products of measurable lesions for at least 1 month), stable (less than Nine antigen-positive lesions were biopsied at the conclusion 50% decrease of biperpendicular diameter products of measurable of therapy on day 4; seven of these lesions had been imaged lesion and less than 25% increase when compared to pretreatment successfully. Antibody delivery (stain intensity, 1 or 2) was measurements), or progressive (greater than 25% increase of measura detected in 3 of 7 image-positive lesions. Hepatic mÃ©tastases ble disease) disease. which were antigen positive at the first biopsy in three patients Statistics. Except where noted, all data comparisons were made using the nonparametric Mann-Whitney two sample test (26, 27). did not accumulate murine immunoglobulin after antibody therapy. However, these lesions were antigen negative at the second biopsy; since two of these lesions were image positive, RESULTS biopsy-sampling errors due to antigenic heterogeneity within tumor masses may account for these findings because the 17- Clinical Results 1A antigen does not modulate. The image-negative lesions did Toxicity. No treatment-related hematological toxicity was not accumulate murine immunoglobulin, except for Patient 44, noted. One patient developed mild symptomatic hypotension whose tumor and normal liver both accumulated high levels of after the second day of treatment which was due to fluid antibody. The relationship of antigen expression to imaging depletion; this reversed rapidly with i.v. hydration and the and therapeutic antibody delivery is displayed in Table 2. patient completed her 4 days of therapy without dosage reduc Determination of Oxygen Tension of Biopsied Lesions tion or further difficulty. All patients experienced Grade I or II fevers and mild constitutional symptoms. Six patients experi Physiological barriers to antibody penetration into tumor enced transient elevations of transaminases, which returned to masses may be important obstacles to therapeutic antibody baseline upon completion of therapy. Four patients developed delivery. Accordingly, tumor pO2 levels were prospectively de anemia which required RBC transfusions. One patient devel termined in II lesions from 10 patients in this study. Mean oped transient abdominal pain of unclear etiology; another pO2 in each tumor ranged from 1 to 34 mm Hg (Table 3). In patient developed abdominal pain due to a partial small bowel antigen-positive lesions, lesions were imaged successfully when 2570

Table 2 Relationship of 17-IA antigen expression to "'Â¡-17-IA imaging and therapeutic antibody delivery

17-IA antigen % of 131I Therapeutic Patient expression dose/ml of Tumorliver antibody code Site(s) (0, 1, 2)Â° Imaging tumor ratio delivery (0, 1, 2)*

Antigen positive, image positive 29 Liver I Posc 0.004 2.5 1 30 Liver 1 Pox 0.007 1.8 0 36 Liver l Pos 0.004 3.6 V 45 Liver 1-2 Pos NT NT 0-1 46 Liver l Pos 0.001 1.0 1â€”2 41 Peripancreatic mass l N. Eval. 0.004 2.0 0" Antigen positive, image negative 32 Liver 1 Neg NT 37 (L) SCN 2 Neg 0 Abdominal mass 1 Neg NT 38 Liver 1-2 Neg NT 43 Liver 1-2 Neg 0" 44 Liver 1-2 Neg 1' Antigen negative, image negative 31 (L) SCN 0 Neg NT Â°Fine needle aspirate frozen sections of accessible tumor deposits prior to therapy were evaluated by an immunoperoxidase staining method with CO 17-1A antibody. Intensity of staining was interpreted as 0 (negative), 1 (positive), or 2 (strongly positive) as described in the text. When possible, lesions which imaged positively were biopsied. *The same lesions were biopsied posttherapy and evaluated for antigen expression using CO 17-IA and therapeutic antibody delivery using a conjugated antimouse antibody only. Results were interpreted as described above and in the test. ' Pos, positive; Neg, negative; NT, not tested; N. Eval., probably positive, but gastric activity overlaid mass seen on CT scan; SCN, supraclavicular node. 'The samples in which these studies were performed were 17-IA antigen negative as well. ' Negative image, but significant delivery of antibody to both tumor and normal liver noted.

Table 3 Influence of tumor pO2 on antibody imaging results and 5 of therapy only. Accordingly, we sampled peripheral blood at earlier time points in this study. All patients had Patient pO2 code3132343637374143444546Siteoflesion(L) (mm/Hg)2-3125202-32-325333234RemarksAntigen-negative*Overlyingbaseline determinations, 15 were evaluated at 1 h, 14 were supraclavicularLiverLiverLiver(L) evaluated at 4 h, and 10 were evaluated 24 h following the start of the first interferon infusion. The effects of interferon therapy on monocyte and lympho supraclavicularAbdominal cyte cytotoxicity directed against SW1116 cells in the presence wallPeripancreaticLiverLiverLiverLiverImageNegNegNDPosNegNegN. or absence of 17-IA incubation in vitro, as a function of time, Eval.NegNegPosPosLesionactivityobscuredgastric region of in were studied. Interferon therapy did not have significant effects terest on nonspecific monocyte and lymphocyte cytotoxicity. In con trast, monocyte and lymphocyte ADCC appeared to rise 24 h after starting therapy (for monocytes, specific lysis is 13.1 Â± 4.1% versus 5.4 Â±1.4%; P = 0.07) (Fig. 1). " L, left; Neg, negative; ND, not determined; Pos, positive; N. Eval., probably When total cytotoxicity (i.e., ADCC plus nonspecific effector positive. * All lesions were antigen positive except where noted. cell cytotoxicity) was evaluated, the trend toward enhanced tumor cell lysis at the 24-h time point continued for monocytes tumor pO2 was >20 mm Hg (P < 0.05 by Yates' correction of and lymphocytes; monocyte ADCC at 24 h exceeded nonspe X2.Successful therapeutic antibody delivery was documented in cific cytotoxicity of baseline monocytes (18.2 Â±4.8% versus 2.8 only two of the patients who had oxygen probe analysis of Â±1.1%; P = 0.03). The bulk of the enhanced cytotoxicity seen tumor pO2; therefore, correlations between these parameters with monocytes and lymphocytes was contributed by the ADCC cannot be made. rather than nonspecific cytotoxicity components. The most striking example of the concept that antigen expres Thus, therapy with low doses of 7-interferon does not en sion does not predict absolutely for antibody image positivity hance nonspecific cytotoxicity, using our assay system at the or therapeutic antibody delivery is found by examination of the time points examined. In contrast, effector cell ADCC is en findings in Patient 37. This 53-year-old woman's left supra- hanced at 24 h; with monocytes, therapy with low doses of the clavicular modal metastasis was strongly 17-IA antigen posi interferon is superior to in vitro incubation of purified mono tive. However, it did not image with labeled 17-IA and no cytes with the cytokine at a concentration which we have found antibody was detected in the mass at the completion of therapy. to be optimal for enhancing ADCC in our assay system in The mean pO2 of the lesion was 2-3 mm Hg (Table 3), and previous studies (data not shown). hypoxia appeared more important than antigen expression in Natural killer cell activity of nonadherent cells was examined determining the accessibility of the malignant cells to the anti as a function of interferon therapy to permit evaluation of this body. well-described consequence of interferon administration in con cert with our other in vitro studies (Fig. 1). The pattern of NK Effector Cell Cytotoxicity Studies activity paralleled the ADCC temporal patterns closely, inas In previous studies, we showed that monocyte cytotoxicity much as enhanced NK activity was not seen until 24 h. While and ADCC directed against the colorectal carcinoma cell line the enhancement at 24 h may have exceeded baseline values (P SW1116 were enhanced by 7-interferon therapy, using the same = 0.075), NK activity at this time point exceeded that seen at schedule used here; however, effects were evaluated on Days 3 either 1 (P = 0.03) or 4 (P = 0.02) h. 2571

30-1 -30 does not ensure detectable therapeutic antibody delivery is in general agreement with the imaging results. Nonimmunological constraints to antibody, such as poor tissue oxygÃ©nation,may explain these results. Tissue oxygÃ©nationis depressed in many malignant deposits and may be viewed as a covariate of tissue pH, necrosis, vascular supply or, more likely, a combination of 20- -20 Â«e. these features (21). Although we have analyzed relatively few patients with oxygen probes and antibody delivery in this study, x a tissue pO2 >20 mm Hg appears necessary for detection of O C o D antibody binding by either radioimmunoimaging or immuno- o" o histological methods. >< 10- -io -1 Further studies are needed to extend our observations that local tumor characteristics exert significant influence on ulti mate antibody delivery, even with detectable antigen expression. If these findings are confirmed, strategies designed to alter local tumor characteristics will be needed. For example, pretherapy tumor irradiation and chemotherapy have been used in a series of clinical immunotherapy trials using radiolabeled antiferritin (17) (9) (19) (10) (8) (8) Monocvte LymphocvleADCC' NK antibodies (32), but the effects of these maneuvers on tumor ADCf Activity oxygen tension are not known. Fig. 1. Monocytes and lymphocytes were purified from peripheral blood as The dose of 7-interferon used in this study was selected based described in the text. The samples were assayed for lysis of SW1116 or K562 cells using an '"In release assay. To test ADCC against the SW1116 line, 100 upon previous observations that several days of therapy in this fig/ml of CO17-1A were added to test wells prior to the assay. Monocyte and dose range provided optimal enhancement of monocyte ADCC. lymphocyte ADCC directed against SU 1116 cells at baseline (il) and following The effects of a single 4-h infusion on this and associated 24 h of therapy (^) are measured by the left ordinate, while lymphocyte natural parameters during the first day of treatment were not known. killer activity at the same time points against K562 erythroleukemia cells is displayed by the right ordinate. Natural killer activity is presented as lytic units Our results show that enhanced monocyte ADCC is not found contained in 10*effector cells using 30% specific lysis of targets as the standard. until 24 h following the start of treatment and that a similar Numbers in parentheses, number of patients contributing to each mean value. although lesser trend is seen with lymphocyte ADCC. Induction of NK cell lysis of K562 cells follows the ADCC pattern, as it DISCUSSION is increased by the 24-h time point. In contrast, nonspecific It is likely that successful therapy with unconjugated murine lymphocyte and monocyte cytotoxicity are not enhanced at any monoclonal antibodies will require the following, at a mini of the time points studied. Using the SW1116 cell line, which mum: (a) binding with high affinity to structurally accessible is relatively resistant to rIFN-7-induced monocyte lysis, our antigens expressed by the malignant cells to be targeted; (b) two clinical studies with rIFN-7 and CO 17-1A suggest that isotypes which permit association with human effector cell Fc prolonged therapy with low doses of rIFN-7 is required to receptors, human complement, or both, with activation of these enhance in vitro ADCC and NK activity maximally. effector mechanisms; (c) antibody delivery to a critical number The modest enhancement of effector cell mechanism by of antigen-expressing target cells; and (d) administration with rIFN-7 did not translate to therapeutic effectiveness of CO 17- acceptable host toxicity. In addition, selection of appropriate 1A in our studies. However, it must be noted that the dose of antigenic targets which affect target cell function (28, 29) or CO 17-1A in these studies was chosen arbitrarily, and that no are tumor specific (30) may be crucial; however, antibody ther systematic Phase I evaluations which use acceptable dose esca apy directed against tumor-associated antigens can be successful lation schemes have been performed using this antibody. The as well (31). dependence of therapeutic antibody delivery on administered While CO 17-1A has been demonstrated to possess all of the dose has been demonstrated with an anti-melanoma antibody biological characteristics thought to be necessary for successful (33). It is possible that significant dose escalations of CO 17-1A antibody therapy in preclinical studies, the results of this clinical will be required to overcome the physiological barriers which trial do not support previous reports of the clinical efficacy of limit the binding of this antibody to antigen-bearing tumor CO 17-1A (17). The investigations of effector cell activation cells. A Phase I trial of prolonged low dose 7-interferon therapy and antibody delivery in our patients offer insights which may with escalating doses of CO 17-1A is needed so that the validity prove valuable for designing future antibody therapy strategies. of this therapeutic approach can be tested in an appropriate Remarkably little antibody is delivered to antigen-expressing Phase II study. malignant cells following i.v. administration. Using a relatively insensitive immunoperoxidase technique, most biopsied lesions ACKNOWLEDGMENTS expressed the 17-1A antigen. However, only approximately 1% of the total administered I31Iconjugated to 17-1A or its F(ab')2 The authors are grateful to Catherine Janus, Christopher Berry, fragment was detected in those tumor deposits at 72 h; this is Caroline Zarou, Michael Wintjen, Perry Watts, and Noreen McCann in agreement with previous imaging studies using this antibody for their valuable assistance in this study. CO 17-1A and rIFN-7 were (7,22, 23). Even allowing for the 72-h delay and dehalogenation provided by the Cancer Therapy Evaluation Program of the National of the antibody-iodine complex, these results are discouraging, Cancer Institute. particularly since reticuloendothelial system uptake is not a significant problem with this antibody (7). Antigenic heteroge REFERENCES neity and biopsy sampling error may explain the poor correla 1. Ross, A. H., LÃ¼beck,M., Steplewski, Z., and Koprowski, H. Identification tion between label delivery per ml tumor and therapeutic anti and characterization of the CO 17-1A carcinoma-associated antigen. Hybrid- body delivery. However, our finding that antigen expression oma, 5 (Suppl. 1): S21-S28, 1986. 2572

2. Herlyn, I>.,and Koprowski, H. IgG2a monoclonal antibodies inhibit human mouse IgG2a and IgG3-dependent cellular cytotoxicity in human monocytic tumor growth through interaction with effector cells. Proc. Nati. Acad. Sci. cells (U937) by immune interferon. Cancer Res., 44: 5127-5131, 1984. USA, 79:4761-4765, 1982. 19. Weiner, L. M., Steplewski, Z., Koprowski, H., Sears, H. F., Litwin, S., and 3. Ross, A. H., Herlyn, D., Iliopoulos, D., and Koprowski, H. Isolation and Comis, R. L. Biologic effects of -y-interferon pre-treatment followed by characterization of a carcinoma-associated antigen. Biochem. Biophys. Res. monoclonal antibody 17-1A administration in patients with gastrointestinal Commun., 735: 297-303, 1986. carcinoma. Hybridoma, 5 (Suppl. 1): S65-S77. 1986. 4. Steplewski, Z., LÃ¼beck,M. D., and Koprowski, H. Human macrophages 20. Weiner, L. M., Steplewski, Z., Koprowski, H., Litwin, S., and Comis, R. L. armed with murine immunoglobulin G2a antibodies to tumors destroy human Divergent dose-related effects of -y-interferon therapy on in vitro antibody- cancer cells. Science (Wash. DC), 221: 865-867, 1983. dependent cellular and nonspecific cytotoxicity by human peripheral blood 5. Steplewski, Z., Herlyn, D., LÃ¼beck,M., Kimoto, Y., Herlyn, M., and Ko monocytes. Cancer Res., 48: 1042-1046, 1988. prowski, H. Mechanisms of tumor growth inhibition. Hybridoma, 5 (Suppl. 21. Gatenby, R. A., Kessler, H. B., Rosenblum, J. S., Coia, L. A., Moldofsky, P. 1):S59-S62, 1986. J., Hartz, W. H., and Broder, G. J. Oxygen distribution in squamous cell 6. Herlyn, D., Herlyn, M., Steplewski, Z., and Koprowski, H. Monoclonal carcinoma mÃ©tastasesand its relationship to outcome of radiation therapy. antibodies in cell-mediated cytotoxicity against human melanoma and colo- Radiology, in press, 1988. rectal carcinoma. Eur. J. Inumino!., 9: 657-659, 1979. 22. Moldofsky, P. J., Powe, J. E., and Hammond, N. D. Monoclonal antibodies 7. Herlyn, D., Steplewski, Z., Herlyn, M., and Koprowski, H. Inhibition of for radioimmunoimaging: current perspectives. In: L. M. Freeman and H. S. growth of colorectal carcinoma in nude mice by monoclonal antibody. Cancer Weissman (eds.), Nuclear Medicine Annual 1986, pp. 57-103. New York: Res., 40:717-721, 1980. Raven Press, 1986. 23. Hammond, N. D., Moldofsky, P. J., Beardsley, M. R., and Mulhern, C. B. 8. Moldofsky, P. J., Powe, J., Mulhern, C. B., Hammond, N., Sears, H. F., External imaging techniques for quantitation of distribution of I-131-F(ab')Ã¯ Gatenby, R. A., Steplewski, Z., and Koprowski, H. Metastatic colon carci noma detected with radiolabeled F(ab')2 monoclonal antibody fragments. fragments of monoclonal antibody in humans. Med. Phys., II: 778-783, Radiology, 149: 549-555, 1983. 1984. 9. Powe, J., Herlyn, D., Alavi, A., MÃ¼nz,D., Steplewski, Z., and Koprowski, 24. Freundlich, B., and Avdolovic, N. Use of gelatin/plasma coated flasks for H. Radioimmunodetection of human tumor xenografts by monoclonal anti isolating human peripheral blood monocytes. J. Immunol. Methods, 62: 31- bodies correlates with antibody density and affinity. In: K. Button and E. 37, 1983. Donato (eds.), Immunoscintigraphy, pp. 139-156. London: Gordon & 25. Oken, M. M., Creech, R. H., and Tormey, D. C., et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am. J. Clin. Oncol., 5: Breach Scientific Publishers, 1985. 649-655, 1982. 10. Sears, H. F., Atkinson, B., Herlyn, D., Ernst, C., Matties, J., Steplewski, Z., 26. Siegel, S. Non-parametric statistics. Mann-Whitney Statistics. New York: Hayry, P., and Koprowski, H. The use of monoclonal antibody in a Phase I clinical trial of human gastrointestinal tumors. Lancet, I: 762-765, 1982. McGraw Hill Book Co., 1956. 27. INSL Library. Mann-Whitney Statistics, 9: 2, November 1984. 11. Sears, H. F., Steplewski, Z., Herlyn, D., and Koprowski, H. Effects of monoclonal antibody immunotherapy on patients with gastrointestinal ade- 28. Waldmann T. A., Longo, D. L., Leonard, W. J., Depper, J. M., Thompson, nocarcinoma. J. Biol. Response Modif., 3:138-150, 1983. C. B., Kronke, M., Goldman, C. K., Sharrow, S., Bongiovanni, K., and Greene, W. C. Interleukin 2 receptor (Tac antigen) expression in HTLV-I- 12. Sears, H. F., Herlyn, D., Steplewski, Z., and Koprowski, H. Phase II clinical associated adult T-cell leukemia. Cancer Res., 45:4559s-4562s, 1985. trial of a murine monoclonal antibody cytotoxic for gastrointestinal adeno- 29. Cuttitta, F., Carney, D. N., and Mulshine, J. L., et al. Bombesin-like peptide carcinoma. Cancer Res., 45: 5910-5913, 1985. can function as an autocrine growth factor in human small cell lung cancer. 13. Sears, H. F., Herlyn, D., Steplewski, Z., and Koprowski, H. Initial trial use Nature (Lond.), 316: 823-826, 1985. of murine monoclonal antibodies as immunotherapeutic agents for gastroin 30. Miller, R. A., Maloney, D. G., Warnke, R., and Levy, R. Treatment of B- testinal adenocarcinoma. Hybridoma, 5 (Suppl. 1): S109-S115, 1986. cell lymphomas with monoclonal anti-idiotype antibody. N. Engl. J. Med., 14. Lobuglio, A. F., Salen, M., Peterson, L., Wheeler, R., Carrano, R., Huster, 306:517-522, 1981. W., and Khazaeli, M. B. Phase I clinical trial of CO 17-1A monoclonal 31. Houghton, A. N., Mintzer, D., Cordon-Cardo, C., Welt, S., Fliegel, B., antibody. Hybridoma, 5 (Suppl. 1): SI 17-S123, 1986. Vadhan, S., Carswell, E., Melamed, M. R., Oettgen, H. F., and Old, L. J. 15. Sindelar, W. F., MÃ¤her,M. M., Herlyn, D., Sears, H. F., Steplewski, Z., and Mouse monoclonal IgG3 antibody detecting GD3 ganglioside: a Phase I trial Koprowski, H. Trial of therapy with monoclonal antibody 17-1A in pan in patients with malignant melanoma. Proc. Nati. Acad. Sci. USA, 82:1242- creatic carcinoma: preliminary results. Hybridoma, 5 (Suppl. I): S125-S132, 1246, 1985. 1986. 32. Order, S. E., Stillwagon, G. B., Klein, J. L., Leichner, P. K., Siegelman, S. 16. Tempero, M. A., Pour, P. M., Uchida, E., Herlyn, D., and Steplewski, Z. S., Fishman, E. K., Ettinger, D. S., Haulk, T., Kopher, K., Finney, K., Monoclonal antibody CO 17-1A and leukopheresis in immunotherapy of Surdyke, M., Self, S., and Leibel, S. Iodine 131 antiferritin, a new treatment pancreatic cancer. Hybridoma, 5 (Suppl. 1): S133-S138, 1986. modality in hepatoma: a Radiation Therapy Oncology Group study. J. Clin. 17. Douillard, J. Y., Lehur, P. A., Vignoud, J., Blottiere, H., Maurel, C., Thedrez, Oncol., 3: 1573-1582, 1985. P., Kremer, M., and Le Mevel, B. Monoclonal antibodies specific immuno 33. Schroff, R. W., Alton, C. M., Woodhouse, C. S., Arams, P. G., Farrell, M. therapy of gastrointestinal tumors. Hybridoma, 5 (Suppl. 1): S139-S149, M., Carpenter, B. E., Oldham, R. D., and Foon, K. A. Monoclonal antibody 1986. therapy in malignant melanoma: factors effecting in vivo localization. J. Biol. 18. Akiyama, Y., LÃ¼beck,M., Steplewski, Z., and Koprowski, H. Induction of Response Modif., 0:457-472, 1987.

2573

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1988 American Association for Cancer Research. Antibody Delivery and Effector Cell Activation in a Phase II Trial of Recombinant γ-Interferon and the Murine Monoclonal Antibody CO17-1A in Advanced Colorectal Carcinoma

Louis M. Weiner, Philip J. Moldofsky, Robert A. Gatenby, et al.

Cancer Res 1988;48:2568-2573.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/48/9/2568

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/48/9/2568. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.