Monoclonal Antibody KS 1 I 4-Methotrexate
`lmmunoconjugate Studies in Non-Small Cell
`Lung Carcinoma
`
`DARLENE J. ELIAS, LAWRENCE E. KLINE, BRUCE A. ROBBINS, HENRY C. L. JOHNSON, JR.,
`KATHERINE PEKNY, MITZI BENZ, JAMES A. ROBB, LESLIE E. WALKER, MICHAEL KOSTY,
`and ROBERT 0. DILLMAN
`
`Department of Molecular and Experimental Medicine, Divisions of Chest and Critical Care Medicine, Hematology and Oncology,
`and Gastroenterology, Departments of Pathology and Immunology, Ida M. And Cecil H. Green Cancer Center, Scripps Clinic
`and Research Foundation, La jolla, California
`
`Th.e antigen reactive with murine monoclonal antibody (MAb) KS1/4 is expressed on epithelial malignan(cid:173)
`cies and some normal epithelial tissues. Studies were undertaken to evaluate KS1/4-methotrexate (KS1/4-
`MTX) immunoconjugate in patients with advanced non-small cell carcinoma of the lung. Eleven patients
`in two different groups received KS1/4-MTX in two different escalating dose infusion schedules with a max(cid:173)
`imal tolerated dose of 1,750 mg/M2 and a cumulative dose of MTX of 40 mg/M2• Toxicities were similar
`in both groups and included fever, anorexia, nausea, vomiting, diarrhea, abdominal pain, guaiac positive
`stool, and hypoalbuminemia. Two patients had an associated aseptic meningitis. One patient had a 50%
`decrease in two lung nodules without a change in lymphangitic infiltrates. This patient received a second
`course of treatment and developed an immune complex-mediated arthritis and serum sickness. Four pa(cid:173)
`tients mounted a human antimouse antibody response. Post-treatment tumor biopsies documented bind(cid:173)
`ing of MAb KS1/4. These studies document the feasibility and potential usefulness of a MAb directed against
`tumor-associated antigens with the targeting of chemotherapeutic drugs in patients with non-small cell
`lung carcinoma. Elias OJ, Kline LE, Robbins BA, Johnson HCL Jr, Pekny K, Benz M, Robb JA, Walker
`LE, Kosty M, Dillman RO. Monoclonal antibody KS1/4-methotrexate immunoconjugate studies in
`non-small cell lung carcinoma. Am J Respir Crlt Care Med 1994;150:1114-22.
`
`lung cancer is the leading cause of cancer death in the United
`States and throughout the world. Of the different histologic types,
`non-small cell comprises about 80% of lung carcinomas and in(cid:173)
`cludes the subtypes of squamous, adenocarcinoma, and large
`cell carcinoma. Surgical resection has proven to be the best ther(cid:173)
`apeutic option for patients with non-small cell lung carcinoma.
`However, for the majority of patients, at the time of diagnosis, the
`disease has spread to regional lymph nodes or to distant sites.
`Once non-small cell carcinoma spreads beyond the site of ori(cid:173)
`gin, the disease is incurable. Aggressive chemotherapy regimens
`and radiation therapy have met with only limited success.
`These investigations are part of ongoing efforts to devise new
`therapeutic approaches for the treatment of malignancy, particu(cid:173)
`larly for non-small cell lung carcinoma. From a large body of im(cid:173)
`munologic data, a rationale has been developed during the last
`decade to suggest that the immune system may have a role in
`
`(Received in original form june 16, 1992 and in revised form February 16, 1994).
`Supported in part by National Institutes of Health General Clinical Research
`Center Grant RR00833, by Scripps Clinic and Research Foundation, Depart(cid:173)
`ment of Medicine Clinical Research Grant 89-01, and University of California
`Tobacco-related Disease Research Program Research Grant RT339.
`Correspondence and requests for reprints should be addressed to Darlene
`1. Elias, M.D., Division of Chest and Critical Care Medicine, Scripps Clinic and
`Research Foundation, 10666 North Torrey Pines Road, La Jolla, CA 92037.
`Am J Respir Crit Care Med Vol 150. pp 1114-1122, 1994
`
`the treatment of cancer. Tumor-associated antigens have been
`an important part of new approaches in immunotherapy of can(cid:173)
`cer. These antigens appear as cell surface-associated structures,
`possibly as a consequence of normal cell surface antigens being
`altered or expressed by the process of malignant transformation.
`Monoclonal antibodies (MAb) directed against tumor-associated
`antigens have the potential to specifically target chemotherapeu(cid:173)
`tic drugs as a therapeutic modality in the treatment of cancer.
`The murine monoclonal KS1/4 is an lgG2a antibody that recog(cid:173)
`nizes a carcinoma-associated cell surface antigen which is found
`on a variety of neoplastic tissues and is expressed by most, if not
`all, adenocarcinomas (1). The antigen is known to be expressed
`on a number of normal epithelial cell types, suggesting that it
`represents an epithelial cell-derived carcinoma marker (2). The
`sequence of eDNA clones that code for the KS1/4-reactive anti(cid:173)
`gen has been determined (3, 4). These data suggest that the anti(cid:173)
`gen is a 40 kD polypeptide that has heterogeneity in glycosyla(cid:173)
`tion, is susceptible to specific proteases and contains a cysteine(cid:173)
`rich domain. The antigen that is recognized by MAb KS1/4 may
`be a suitable target antigen for antibody-directed therapy of
`non-small cell lung carcinoma (5, 6).
`We have previously reported a clinical comparative trial of MAb
`KS1/4 and KS1/4-methotrexate (KS1/4-MTX) immunoconjugate in
`patients with non-small cell lung carcinoma (7). Doses and toxic(cid:173)
`ities were defined and compared in both the KS1/4 alone and the
`KS1/4-MTX patient groups. In general, the infusions were well toler(cid:173)
`ated, and similar toxicities were experienced by both groups. Bi-
`
`IMMUNOGEN 2011, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Elias, Kline, Robbins, eta/.: MAb-Drug Conjugate Clinical Trial in Non-Small Cell Lung Carcinomas
`
`1115
`
`opsies of carcinoma taken during and after the course of im(cid:173)
`munotherapy showed selective localization and binding of KS1/4
`to the tumor. We found a dose-response relationship between
`the quantity of administered antibody or immunoconjugate and
`the subsequent binding of antibody to carcinoma. The KS1/4-
`reactive antigen is known to be expressed on normal colonic
`mucosa and post-treatment colonic mucosal biopsies showed
`binding of MAb KS1/4.
`The current studies were undertaken, using higher doses and
`differing schedules of administration to further evaluate the safety,
`toxicities, and clinical response of MAb KS1/4-MTX immunocon(cid:173)
`jugate when administered to patients with non-small cell lung car(cid:173)
`cinoma. The cross-reactivity of MAb KS1/4 with normal epithelial
`tissue was further investigated.
`
`METHODS
`Preparation of KSl/4-MTX lmmunoconjugate
`
`Unconjugated KS1/4 was prepared by Brunswick Biotech nics (San Diego,
`CA) and was covalently conjugated to methotrexate as previously described
`(7, 8). As measured spectrophotometrically at 410 nm, the conjugation
`ratio was 6 mol of methotrexate/mol of antibody (17 mg of methotrexate/g
`of KS1/4).
`
`Patients
`
`Eleven patients with advanced non-small cell carcinoma of the lung who
`had received or declined conventional therapies were selected to receive
`KS1/4-MTX. Eligibility criteria required that each patient's carcinoma ex(cid:173)
`press the KS1/4-reactive antigen by immunoperoxidase staining. All pa(cid:173)
`tients had measurable or evaluable disease, a performance status of 0-2
`on the Eastern Cooperative Oncology Group scale, carcinoma accessi(cid:173)
`ble for repeat biopsy, no chemotherapy or radiation therapy for at least
`30 d before entry into the trial, and adequately preserved hematologic
`(hemoglobin> 10 g/dl, white blood cell count [WBC) > 3,000/mm3, plate(cid:173)
`let count > 100,000/mm3), renal (creatinine < 2.0 mg/dl), and hepatic
`(bilirubin< 2.5 mg/dl, serum glutamic-oxaloacetic transaminase< 70 units,
`alkaline phosphatase< 300 units) parameters. Informed consent was ob(cid:173)
`tained based on protocols on file with the Human Subjects Committee
`of Scripps Clinic and Research Foundation.
`
`Study Plan
`
`The first group of six patients received KS1/4-MTX as a 1-mg test dose,
`followed by an escalating infusion schedule of 50, 100, 250, 350, 500, and
`500 mg/M2. The duration of each infusion was 24 h. The longer time of
`infusion, compared with our initial comparative clinical trial (7), was cho(cid:173)
`sen on the basis of data that suggest toxicities, particularly respiratory
`symptoms, with high doses of MAb infusions can be lessened or avoided
`by longer infusion times. The infusions were three times a week for a total
`of 2 wk. The shorter duration of 2 wk compared with the 3 wk was chosen
`in an attempt to complete the administration of MAb before the mounting
`of a maximal human anti mouse response. This new protocol administered
`a maximal dose of 1,750 mg/M2 for a cumulative dose of methotrexate
`of 40 mg/M2.
`A second group of patients received KS1/4-MTX in a different infu(cid:173)
`sion schedule. Two patients received 350 mg/M2 in 24 h for three con(cid:173)
`secutive 24-h periods, for a total dose of 1,050 mg/M2 in 72 h. Two pa(cid:173)
`tients received 425 mg/M2 for a total dose of 1,275 mg/M2 and one patient
`received 500 mg/M2 for a total dose of 1,500 mg/M2. This second infusion
`schedule was designed to possibly avoid some of the observed gastroin(cid:173)
`testinal (GI) toxicity. We hypothesized that with the initial infusion of
`KS1/4-MTX, there was binding to Gl mucosal epithelium. The previous
`schedule allowed recovery periods, and it was postulated that the regu(cid:173)
`lar, normal cellular turnover of the Gl epithelium resulted in newly avail(cid:173)
`able Gl epithelial receptors available for retargeting with the next infu(cid:173)
`sion, and that this targeting and retargeting might play a role in the
`escalating and cumulative Gl symptoms. This, coupled with the large sur(cid:173)
`face area of the Gl tract when compared with tumor surface area, presented
`a situation where the Gl tract might be bound in preference to the tumor.
`
`It was thought that with higher daily doses and a shorter total infusion
`time, Gl epithelium might be saturated early, promote subsequent bind(cid:173)
`ing to the carcinoma, and lessen Gl side effects. Also, a continuous infu(cid:173)
`sion schedule may favor passive diffusion of macromolecules from the
`vascular and interstitial space into tumor tissue (9, 10).
`After the course of treatment in both groups, biopsies of the tumor
`were obtained after the last dose of KS1/4-MTX and examined for evi(cid:173)
`dence of in vivo binding of antibody. Because of the known presence of
`the antigen on epithelial derived cell surfaces (1, 2) and the previously
`reported Gl toxicities (6, 7), patients underwent examination of their up(cid:173)
`per Gl tracts pre- and post-treatment with upper gastrointestinal (UGI) en(cid:173)
`doscopy. Esophagus, stomach, and duodenum were visually examined,
`photographed and duodenal biopsies were performed pre- and post(cid:173)
`treatment. The antrum was biopsied if the epithelial mucosa was visually
`abnormal. Pretreatment duodenal biopsies were examined for the pres(cid:173)
`ence of the KS1/4-reactive antigen. Post-treatment duodenal biopsies were
`examined for KS1/4 antibody binding and deposition of complement. En(cid:173)
`doscopically observed mucosal abnormalities were graded on a scale
`of 0 through 4. Normal mucosa was graded as 0, change including ery(cid:173)
`thema only was graded as 1, erosions with friability and exudate were
`graded as 2, frank ulcer craters were graded as 3, and complete mucosal
`destruction was graded as 4. Serial blood samples were obtained at mul(cid:173)
`tiple time points during and following treatment for determination of KS1/4
`serum levels and human anti mouse levels. Evaluation of the status of the
`carcinoma were based on radiographic evaluation and physical exami(cid:173)
`nation performed at time of entry and 4 wk after initiation of treatment.
`
`Toxicity Monitoring
`
`A clinical research nurse monitored patients closely for any untoward reac(cid:173)
`tions. Grading and monitoring of toxicities were based on a modified tox(cid:173)
`icity scale (based upon the Biological Response Modifier Program), and
`approved by the Human Subject Committee of Scripps Clinic and Re(cid:173)
`search Foundation.
`
`Immunohistochemical Staining of Tissue
`
`Detection of KS1/4-reactive antigen was performed using a biotin-avidin
`immunoperoxidase technique on paraffin-embedded and fresh frozen tis(cid:173)
`sue blocks. Briefly, paraffin sections were deparaffinized, incubated with
`0.3% hydrogen peroxide for 15 min followed by 0.25% porcine trypsin for
`30 min at room temperature (RT). Fresh frozen cryostat sections were air(cid:173)
`dried, and then fixed in acetone at RT for 10 min. All sections were washed
`in phosphate-buffered saline (PBS), incubated for 30 min at RT with MAb
`KS1/4 in PBS (1% bovine serum albumin [BSA)), washed again and in(cid:173)
`cubated for 30 min with biotinylated horse anti mouse lgG, 3 ~g/ml (Vec(cid:173)
`tor, Burlingame, CA). The slides were then rinsed and incubated with horse(cid:173)
`radish peroxidase (HRP) conjugated avidin D 151!g/ml (Vector). Sections
`were rinsed and overlaid with 3-amino-9-ethyl carbazole (AEC) peroxidase
`chromagen (Biomeda, Foster City, CA) for 10 min, rinsed and counter(cid:173)
`stained in aqueous Mayer's hematoxylin. Positive controls were immune(cid:173)
`stained with antihuman keratin, AE1/AE3, 1 ~/ml (Boehringer Mannheim,
`Indianapolis, IN). Negative controls were immunostained with lgG(K)MOPC
`21, 11lQ/ml (Organon Teknika, West Chester, PA), a myeloma lgG1, with
`no known hapten or antigen binding activity.
`In vivo binding of KS1/4 antibody to tumor and normal gastrointestinal
`mucosa was detected by staining for mouse lgG with an indirect im(cid:173)
`munoperoxidase technique. Cryostat sections were air-dried, washed with
`PBS, covered for 30 min at RT with biotinylated horse anti mouse lgG (Vec(cid:173)
`tor) at 3 11giml in PBS. Negative controls were stained with biotinylated
`goat anti rabbit lgG, 1 ~/ml (Tago, Burlingame, CA). Sections were rinsed
`in PBS and covered with HAP-conjugated Avidin D for 30 min. Bound
`antibody was visualized with AEC chromagen (Biomeda) and counter(cid:173)
`stained with aqueous hematoxylin.
`A biotin-avidin immunoperoxidase procedure was used to detect com(cid:173)
`plement deposition at the antibody-antigen binding site on duodenal and
`carcinoma specimens. After biopsy, specimens were placed in a citrate
`buffer (Zeus Scientific, Raritan, NJ) which contains an antiautolytic agent,
`n-ethyl-maleimide, and a fixative, (NH4hS04• Specimens were then washed
`with citrate buffer solution (Zeus), covered with Tissue.:rek O.C.T. freezing
`compound (Miles, Elkhart, IN) and snap frozen in isopentane cooled in
`
`IMMUNOGEN 2011, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`1116
`
`AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 150 1994
`
`SUMMARY OF THE CLINICAL FEATURES, TOTAL DOSES OF KS1/4-METHOTREXATE IMMUNOCONJUGATE,
`HUMAN ANTIMOUSE RESPONSES, MAXIMAL TOXICITY, AND CLINICAL RESPONSE
`
`TABLE 1
`
`Patient
`Entry No.* and
`Treatment
`Schedule
`
`02-01
`02-02
`02-03
`02-04
`02-05
`02-06
`03-01
`03-02
`03-03
`03-04
`03-05
`
`Histologic
`Cell Type
`
`Adenocarcinoma
`Adenocarcinoma
`Large cell
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`Adenocarcinoma
`
`Age
`
`76
`59
`53
`45
`71
`63
`64
`75
`59
`82
`66
`
`Previous
`Treatment
`
`Surgery
`None
`Chemo
`Chemo
`None
`XRT, chemo
`Chemo
`Surgery
`Chemo
`Chemo, XRT
`Surgery
`
`Total
`Dose
`(mg)
`
`608
`2510
`2660
`1330
`1520
`1250
`1920
`1500
`2490 (1992)§
`2160
`1332
`
`Dose/M2
`
`331
`1755
`1750
`722
`850
`1250
`1050
`1050
`1270 (1015)§
`1270
`780
`
`Human
`Anti mouse
`Antibodies
`
`Maximal
`Toxicityt
`
`Clinical
`Response:!:
`
`+
`+
`
`+
`
`+
`
`2
`
`1
`3
`2
`2
`2
`2
`3
`2
`3
`
`SD
`PD
`PD
`PD
`SD
`SD
`SD
`SD
`MR
`SD
`PD
`
`Definition of abbreviations: chemo = chemotherapy; XRT = radiation therapy; SD = stable disease; PD = progressive disease; MR = minimal response.
`• Patients 02.{)1 through 02-06 received KS1/4- MTX as an escalating dose infusion schedule on alternate days over 2 wk. Patients 03-01 through 03-05 received KS1/4-Mtx as a continuous
`3-d infusion.
`t Degree of toxicity (0 = none, 1 = minimal, 2 = mOderate, 3 = severe, 4 = life-threatening).
`* Clinical response.
`
`§ Retreatment doses.
`
`liquid nitrogen. Fresh frozen cryostat sections of pre- and post-treatment
`biopsies were air-dried for 1 h, washed in PBS, incubated 30 min with
`either rabbit anti-C3d (4.711Q/ml) or rabbit anti-C4c (20 11g/ml) in PBS (1%
`BSA). Next, the slides were rinsed with PBS, incubated for 30 min with
`biotinylated goat anti rabbit lgG (1 llQiml) in PBS and rinsed. Sections were
`incubated another 30 min with HRP conjugated to avidin D (15 11g/ml),
`rinsed and incubated 10 min with AEC peroxidase chromagen. Negative
`controls were incubated with normal rabbit serum (1:1,000 in PBS, 1%
`BSA). Normal human tonsil sections were used as positive controls.
`
`Determination of KS1/4 Serum Levels and Human
`Antimouse Antibodies
`
`Monoclonal antibody KS1/4 serum levels were measured by enzyme-linked
`immunosorbent assay (ELISA) as previously described (7). Human an(cid:173)
`timouse antibodies were measured by ELISA. Aliquots of serum were in(cid:173)
`cubated for 1 h in 96-well microliter plates which had been coated with
`MAb KS1/4 (1 ~/well). After washing, plates were incubated for 1 h with
`HAP-goat antihuman lgA, lgG, lgM 0.1 ~/ml (Kirkegaard & Perry Labora(cid:173)
`tories, Gaithersburg, MD). Reaction with 0-phenylene diamine as the chro(cid:173)
`magen and absorbance reading at 490 was performed. Based upon a
`standard curve, the absorbance values (11g/ml) human antimouse anti(cid:173)
`body were calculated.
`
`Toxicity
`There were no differences in maximal tolerated doses and toxici(cid:173)
`ties between the two treatment infusion schedules (Table 1). Tox(cid:173)
`icities are summarized in Table 2. Some degree of toxicity occurred
`in each patient. Fever, anorexia, nausea, vomiting, diarrhea, ab(cid:173)
`dominal pain, guaiac positive stool, hypoalbuminemia, mild ane(cid:173)
`mia (hemoglobin 9.5-10.9), and brief increases in liver transami(cid:173)
`nases were seen. One patient experienced a mild pancreatitis.
`Allergic reactions such as pruritus, urticaria, and anaphylaxis were
`not seen.
`One patient experienced an immune complex-mediated ar(cid:173)
`thritis and serum sickness. Patient 03-03 had a clinical response
`at the first post-treatment evaluation and received a second se(cid:173)
`ries of KS1/4-MTX infusions 6 wk after completing the initial infu(cid:173)
`sions. About 36 h into the second infusion, having received a to(cid:173)
`tal dose of 2,490 mg previously and the current dose of 1,160 mg,
`
`RESULTS
`Clinical Response
`The clinical features, total administered doses of KS1/4-MTX im(cid:173)
`munoconjugate, human antimouse response, maximal toxicity,
`and clinical response are summarized in Table 1. Clinical response
`to treatment was evaluated by comparison of measurements of
`tumor with imaging studies and physical examination made at time
`of entry and 4 wk after initiation of therapy. Of the 11 patients com(cid:173)
`pleted, one patient (03-03) had a minimal response at the first post(cid:173)
`treatment evaluation. This patient had a 50% decrease in size
`of two midlung zone nodules without change in surrounding lym(cid:173)
`phangitis infiltration. This patient underwent retreatment with
`KS1/4-MTX. Repeat evaluation showed continued slight decrease
`in the same areas of nodularity, while other areas were unchanged.
`Progressive disease was subsequently documented in this pa(cid:173)
`tient. Three patients had progressive disease at the first post(cid:173)
`treatment evaluation and six patients had stable disease. Ten pa(cid:173)
`tients have died, one patient is alive with metastatic disease.
`
`Toxicities
`
`Fever
`Rigor/chills
`Anorexia
`Nausea/vomiting
`Diarrhea
`Abdominal pain
`Guaiac positive stool
`Hypoalbuminemia
`Anemia
`Transaminasemia
`Pancreatitis
`Aseptic meningitis
`Serum sickness
`Anasarca
`Urticaria
`Pruritus
`Dyspnea
`Bronchospasm
`Hypotension
`Anaphylaxis
`
`TABLE 2
`
`SUMMARY OF TOXICITIES
`
`Patients
`(No.)
`
`Total
`
`%of Total
`
`4
`0
`11
`10
`9
`10
`8
`7
`
`1
`2
`
`1
`0
`0
`0
`0
`0
`0
`
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`11
`
`36
`0
`100
`90
`82
`90
`73
`64
`9
`9
`9
`18
`9
`9·
`0
`0
`0
`0
`0
`0
`
`IMMUNOGEN 2011, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Elias, Kline, Robbins, eta!.: MAb-Drug Conjugate Clinical Trial in Non-Small Cell Lung Carcinomas
`
`1117
`
`the patient had fever and an acute, severe, polyarthritis and sy(cid:173)
`novitis involving the small joints of the feet, ankles, and knees.
`There was no rash. No infectious etiology was identified. Erythro(cid:173)
`cyte sedimentation rate (ESR) was elevated at 55, creatine phos(cid:173)
`phokinase (CPK) was normal. Testing for antinuclear antibodies
`(ANA) was positive at 1:640 in a speckled pattern. Raji cell assay
`for circulating immune complex was elevated at 225 (reference
`range < 100 llg aggregated human gamma globulin). Quantita(cid:173)
`tive cryoglobulins were negative. C4d/C4 ratio was 1.5 (normal
`0-1.1). These results are consistent with activation of the classic
`pathway of complement. KS1/4-MTX infusion was held, the pa(cid:173)
`tient was treated with corticosteroids and rapidly improved.
`KS1/4-MTX infusion was reinstituted and the patient completed
`the scheduled infusion for a total retreatment dose of 1,992 mg.
`There were no further fever or joint symptoms. He did experience
`gastrointestinal toxicity. Human anti mouse antibody (HAMA) re(cid:173)
`sponse was positive upon entry for retreatment at 180 J.lg/ml. The
`HAMA response at 6 wk after the retreatment was 450 J.lg/ml.
`Two patients had aseptic meningitis. Patient 02-04 developed
`headache, photophobia, and fever which were temporally related
`to the immunoconjugate infusions. The cerebrospinal fluid (CSF)
`was abnormal with a total WBC of 152 (normal 0-5) with 98% lym(cid:173)
`phocytes and 2% monocytes; protein was elevated to 62 mg/dl
`with a glucose of 47 mg/dl (concurrent serum glucose was 80
`mg/dl). CSF cytology was benign. CSF cultures were negativE!Jor
`bacteria, fungus, acid-fast bacillus (AFB), and viruses. MAb KS1/4
`was undetectable in CSF by ELISA. A magnetic resonance imag(cid:173)
`ing (MRI) scan of the head was normal. The patient improved with(cid:173)
`out treatment and returned 1 wk later with recurrent headache
`but no fever. Repeat analysis of CSF was improved with a total
`WBC of 40 with 96% lymphocytes and 4% monocytes. Protein
`had normalized to 24 mg/dl and glucose was 49 mg/dl (concur(cid:173)
`rent serum glucose was 92 mg/dl). Repeat CSF cytology was again
`benign and repeat cultures were negative.
`Patient 03-05 developed lethargy and fever. The cerebral fluid
`was abnormal with a total WBC of 176 with 78% polymorpho(cid:173)
`nuclear leukocytes, 2% lymphocytes, and 20% monocytes. Pro(cid:173)
`tein was elevated at 79 mg/dl with a glucose of 178 mg/dl (concur(cid:173)
`rent serum glucose was 250 mg/dl). CSF cytology was benign.
`
`CSF cultures were negative for bacteria, fungus, AFB, and viruses.
`MAb KS1/4 was undetectable in CSF by ELISA. A CT scan of the
`head was normal. MAb infusion was discontinued, antiemetics
`were held, and the patient improved over 12 h. Lumbar puncture
`was repeated several days later and total WBC had declined to
`52 with 94% lymphocytes, 5% monocytes, and 1% macrophage.
`Repeat CSF cytology and all cultures were negative. MAb KS1/4
`was, again, undetectable in CSF by ELISA.
`Total administered doses, maximal clinical toxicity, and the find(cid:173)
`ings on upper Gl endoscopy and the histopathologic abnormali(cid:173)
`ties of the duodenal biopsies are summarized in Table 3. There
`was no significant difference in maximal tolerated doses or toxic(cid:173)
`ities between the two treatment infusion schedules. Doses at which
`Gl toxicity occurred were as low as 608 mg and as high as 2,660
`mg. All patients recovered symptomatically after withdrawal or
`completion of therapy.
`Post-treatment endoscopic findings were variable. The duode(cid:173)
`num was abnormal in 6 of 11 patients. Mucosal changes charac(cid:173)
`terized by erosions with friability and exudates were seen in four
`patients, erythema alone was seen in two patients. The antrum
`was abnormal in two patients. The characteristic histopathologic
`abnormality of the post-treatment duodenum was a mild chronic
`inflammatory change with increases in lymphocytes, plasma cells,
`and eosinophils within the lamina propria. There was a correla(cid:173)
`tion between the endoscopic and histopathologic abnormalities
`(Table 3). Two patients (02-01, 02-02) underwent UGI endoscopy
`with duodenal biopsies 2 wk after completion of therapy and the
`duodenum had returned to normal, both visually and histopatho(cid:173)
`logically.
`
`In vivo Localization of MAb KSl/4 and Complement Fragments
`Pretreatment immunoperoxidase staining of each patient's carci(cid:173)
`noma demonstrated expression of the KS1/4-reactive antigen (Fig(cid:173)
`ure 1). Post-treatment carcinoma biopsies were examined for evi(cid:173)
`dence of in vivo binding of MAb KS1/4. In vivo binding was
`documented at all doses.
`Pretreatment immunohistochemical stains of normal duodenal
`biopsies uniformly demonstrated expression of KS1/4-reactive an(cid:173)
`tigen and post-treatment duodenal biopsies in all patients docu-
`
`SUMMARY OF TOTAL DOSES, MAXIMAL TOXICITIES, AND UPPER GASTROINTESTINAL
`ENDOSCOPIC FINDINGS AND HISTOPATHOLOGY OF DUODENAL BIOPSIES
`
`TABLE 3
`
`Patient
`Entry No. • and
`Treatment
`Schedule
`
`02-01
`02-02
`02-03
`02-04
`02-05
`02-06
`03-01
`03-02
`03-03
`03-04
`03-05
`
`Total
`Dose
`(mg)
`
`608
`2510
`2660
`1330
`1520
`2250
`1920
`1500
`2490 (1992)11
`2160
`1332
`
`Maximal
`Clinical
`Toxicityt
`
`Endoscopic/Mucosal
`Abnormalities:!:
`
`Histopathologic
`Abnormalities§
`
`2
`
`1
`3
`2
`2
`2
`2
`3
`2
`3
`
`2
`2
`0
`2
`
`1
`0
`0
`2
`0
`0
`
`2
`1
`0
`0
`
`0
`2
`1
`0
`0
`
`• Patients 02-01 through 02-06 received KS114-MTX as an escalating dose infusion schedule on alternate days over 2 wk. Patients 03.01 through
`03-05 received KS114-MTX as a continuous 3-day infusion.
`t Degree of toxicity (0 = none, 1 = minimal, 2 = moderate, 3 = severe, 4 = life-threatening).
`:!= Post-treatment endoscopically observed duodenal abnormalities (0 = normal, 1 = change including erythema only, 2 = erosions with friability
`and exudate, 3 = frank ulcer craters, 4 - complete mucosal destruction).
`§ Post-treatment histopathological grading (0 = normal, 1 = minimal, 2 = mild, 3 = moderate, 4 = severe).
`II Retreatment dose.
`
`IMMUNOGEN 2011, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`1118
`
`AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL150 1994
`
`Figure 1. Carcinoma histopathology and immunoperoxidase staining. (Top panel) Pretreatment transbronchial
`biopsy shows expression of KS1/4-reactive antigen by adenocarcinoma. There is extension into the alveolar walls,
`and nests of malignant cells are present within alveolar spaces. Normal surrounding lung and fibrous tissue is
`not stained. lmmunostained with MAb KS1/4. Magnification: x45. (Bottom panel) Post-treatment lymph node bi(cid:173)
`opsy immunostained with horse antimouse antibody and documenting that MAb KS1/4 is bound to metastatic
`carcinoma. Magnification: x95.
`
`mented intense staining of bound KS1/4 to the duodenal mucosa
`(Figure 2). The histopathologic features of pretreatment duodenal
`biopsies were normal. Post-treatment, in some patients, the vil(cid:173)
`lous architecture was focally flattened with a loss of goblet cells.
`A mild increase in lymphocytes, plasma cells, polymorphonuclear
`leukocytes, and eosinophils could be seen throughout the lam(cid:173)
`ina propria (Figure 2).
`Post-treatment duodenal biopsies were evaluated for the pres(cid:173)
`ence of complement activation. C3d and C4c deposition were not
`specifically associated with the surface mucosa and did not corre(cid:173)
`late with the site of MAb KS1/4 deposition post-treatment.
`
`Pharmacokinetics and Human Antimouse Response
`To assess the concentration of the circulating MAb achieved with
`
`the escalating dose infusions, serum levels of MAb KS1/4 were
`measured at baseline before the initiation of each infusion and
`during the last hour of each infusion. In the first group of patients,
`serum levels increased with each dose escalation of antibody (in
`J.lg/ml: 50 mg, 11.6 ± 2.1 [SE]; 100 mg, 31.6 ± 5.9; 250 mg,
`64.6 ± 9.8; 350 mg, 76.2 ± 6.9; 500 mg, 89.5 ± 23.3; 500 mg,
`132.3 ± 27.7). In the second group of patients, serum levels in(cid:173)
`creased at the second 24-h time points and then decreased or
`stabilized by the third time point (in J.lg/ml, x: 350 mg/M2 , 75, 89,
`46; 425 mg/M 2 , 94, 129, 130). For the one patient who completed
`the first infusion at 500 mg/M2, the maximal serum level achieved
`was 125 J.lg/ml' at the termination of the infusions and rapidly
`declined over 72 h (data not shown). Only 4 of 11 patients demon(cid:173)
`strated a human antimouse antibody response (Table 1).
`
`IMMUNOGEN 2011, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Elias, Kline, Robbins, et a/.: MAb-Drug Conjugate Clinical Trial in Non-Small Cell Lung Carcinomas
`
`1119
`
`IMMUNOGEN 2011, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`1120
`
`DISCUSSION
`
`Despite advances in diagnosis, surgical techniques, radiation ther(cid:173)
`apy, and chemotherapy, most non-small cell lung cancers are in(cid:173)
`curable once they have spread beyond their site of origin. There
`is an ongoing need to devise new therapeutic approaches for
`metastatic non-small cell lung cancer.
`One of the major problems with modern systemic radiation ther(cid:173)
`apy and chemotherapy for malignancy is their lack of tumor spec(cid:173)
`ificity. Normal tissues may be injured and systemic toxicities can
`be significant. The specificity of MAbs directed against tumor(cid:173)
`associated antigens offers the potential of targeting cytotoxic
`agents to tumor cells and sparing normal tissues. Tumor-specific
`antigens have proven elusive; however, many tumor-associated
`antigens have been identified.
`The antigen reactive with MAb KS1/4 is expressed on epithe(cid:173)
`lial malignancies and some normal epithelial tissue (1, 2). The
`function of the antigen is unknown, but its high density on the
`cell surface and homogeneous expression on non-small cell lung
`carcinoma has suggested that it might be a suitable target for MAb(cid:173)
`directed therapy (6, 7). We previously reported a comparative clin(cid:173)
`ical trial of MAb KS1/4 and KS1/4-MTX in patients with non-small
`cell lung carcinoma (7). The maximal tolerated dose in both treat(cid:173)
`ment groups was 1,661 mg with similar side effects.
`In this report we have shown that two different infusion sched(cid:173)
`ules, the first an escalating schedule over 2 wk and the second
`a continuous 3-d infusion, resulted in similar maximal tolerated
`doses and toxicities. Toxicities included fever and gastrointesti(cid:173)
`nal side effects. These side effects are similar to those seen in
`earlier MAb KS1/4 clinical trials (6, 7) and in other MAb clinical
`trials (11-27). Allergic reactions such as puritis, urticaria, and
`anaphylaxis were not seen. Two patients developed aseptic menin(cid:173)
`gitis with symptoms of fever, headache, and altered level of con(cid:173)
`sciousness. Neither infectious nor malignant etiology was identi(cid:173)
`fied and both patients recovered after withdrawal of treatment.
`The etiology of this side effect is unknown. KS1/4 does not cross(cid:173)
`react with meningeal tissue and like all murine MAbs does not
`cross the blood brain barrier in the absence of meningeal inflam(cid:173)
`mation. At the height of symptoms there was no MAb KS1/4 de(cid:173)
`tected in CSF. Aseptic meningitis, also of unknown etiology, has
`been described following administration of MAb anti-CD3 (OKT3)
`(23-28). OKT3 is directed against T cells and is widely used to
`treat graft rejection following renal, hepatic, and cardiac trans(cid:173)
`plantation. Aseptic meningitis has not been previously described
`in association with MAbs directed against-Carcinoma-associated
`antigens.
`There was one possible clinical response. This patient, at the
`first post-treatment evaluation, had a 50% decrease in the size
`of two midlung zone nodules without a change in the surround(cid:173)
`ing