Journal of Biological Response Modifiers
`9345—354 © 1990 Raven Press, Ltd., New York
`
`Antimelanoma Monoclonal Antibody-Ricin A Chain
`Immunoconjugate (XMMME—OOl-RTA) Plus
`Cyclophosphamide in the Treatment of Metastatic
`Malignant Melanoma: Results of a Phase II Trial
`
`R. Oratz, J. L. Speyer, J. C. Wemz, H. Hochster, M. Meyers, *R. Mischak, and
`*L. E. Spitler
`
`Kaplan Cancer Center, New York University Medical Center, New York, New York, and *XOMA Corporation,
`Berkeley, California, U.S.A.
`
`
`
`Summary: Prior studies with the XMMME-OOl-RTA immunoconjugate com-
`posed of an antimelanoma monoclonal antibody and ricin A chain demon-
`strated some antitumor activity. However, almost all patients studied devel-
`oped human antimurine antibodies and antiricin antibodies. In an effort to
`abrogate these host anti—immunotoxin immune responses and thus enhance
`antitumor activity, we treated 20 patients with the immunoconjugate plus a
`single dose of intravenous cyclophosphamide. An overall response rate of 20%
`was observed—predominantly in pulmonary and soft tissue nodules. There
`was no diminution in antibody responses against either the murine antibody or
`the ricin moiety. Further studies to elucidate the role of cyclophosphamide in
`monoclonal antibody therapy are planned. Key Words: Immunoconjugate—
`Antimelanoma monoclonal antibody—Ricin A chain—Human antimurine an-
`tibodies—Cyclophosphamide.
`
`
`We have tested an antimelanoma monoclonal an-
`tibody conjugated to ricin A chain in patients with
`metastatic malignant melanoma. XMMME-
`001-RTA, an Ing murine monoclonal antibody,
`recognizes two high molecular weight antigens of
`220 kDa and greater than 500 kDa, and is conju-
`gated to purified ricin A chain by SPDP reaction.
`Prior animal and phase I studies have demonstrated
`the safety of this agent (1,2). Phase II studies sug-
`gested potential clinical usefulness of this immuno-
`toxin after a single course, with a small number of
`patients achieving durable partial remissions (3).
`In these previous studies, all patients tested
`
`Received October 24, 1989; accepted February 26, 1990.
`Address correspondence and reprint requests to Dr. R. Oratz
`at NYU Medical Center, Old Bellevue Administration Building,
`Division of Oncology, 462 First Avenue, Room 224, New York,
`NY 10016, U.S.A.
`
`mounted a host antibody response against both mu-
`rine immunoglobulin and ricin A chain moieties of
`the immunotoxin. Cyclophosphamide given with or
`shortly after sensitization to a new antigen has been
`shown in animal (4,5) and human studies (6) to blunt
`humoral responses to new antigens. In an animal
`model designed to test the effect of various immu-
`nosuppressive drugs on antibody responses, Santos
`et a1.
`immunized rats with sheep red blood cells
`(SRBCs) and at various times in relation to immu-
`nization-administered cyclophosphamide, meth-
`otrexate, or 6-mercaptopurine (6). The animals
`were then bled periodically and peak anti-SRBC an-
`tibody titers were measured. Cyclophosphamide
`was a powerful inhibitor of the humoral response
`particularly when administered one or several days
`prior to immunization. In a comparison of the im-
`munosuppressive capacity of drugs, cyclophospha-
`
`345
`
`,,
`
`—
`
`IMMUNOGEN 2306, pg. 1
`Phigenix v. Immunogen
`
`|PR2014-9(:§7§
`
`
`
`
`
`IMMUNOGEN 2306, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`346
`
`R. ORATZ ETAL.
`
`mide and methotrexate were far superior to 6-
`mercaptopurine. In an early clinical study (7), pa—
`tients were immunized with either the V1 antigen (a
`purified polysaccaride) or a Pasturella tularensis
`vaccine. Immunosuppressive drugs including cyclo-
`phosphamide and 6-mercaptopurine were adminis-
`tered either prior to or following immunization. All
`patients who received cyclophosphamide (7 mg/kg
`i.v. daily for 7 days prior to antigen challenge)
`showed no rise in antibody titer while under obser-
`vation.
`
`More recent animal studies (8) used a Balb/c
`mouse model in which animals were sensitized with
`
`thymus, and
`alloantigens comprised of spleen,
`lymph node cells from C3H mice. Two dosages
`(20% LD50 and 60% LDSO) of a number of different
`immunosuppressive agents were administered at
`various times in relation to antigen challenge. Cy-
`clophosphamide in doses of 102 and 306 mg/kg both
`showed strong suppression of antibody responses
`when given with or shortly after the immunizing
`antigen.
`In an effort to enhance antitumor responses and
`abrogate the host anti-immunotoxin antibody re-
`sponse, cyclophosphamide was added to this an—
`timelanoma immunotoxic protocol. A single large
`dose of cyclophosphamide (1,000 mg/m2 i.v.) given
`immediately following immunotoxin infusion was
`selected based on preclinical and animal studies, so
`that the immunosuppressive agent was given essen—
`tially along with the potential antigen.
`
`MATERIALS AND METHODS
`
`Patients
`
`Patients over age 18 years with histologically doc—
`umented malignant melanoma with measurable
`metastatic disease were eligible for study. Other el-
`igibility criteria included Karnofsky performance
`status 280% and life expectancy of at least 12
`weeks. Patients were required to have adequate
`bone marrow, renal, and liver function. Prior sys-
`temic therapy for metastatic melanoma was al-
`lowed. Patients with resected, irradiated brain me-
`tastases and stable head computed tomography
`(CT) scan were eligible. This study was approved
`by the New York University Medical Center IRB
`and all patients signed written informed consent
`prior to treatment. Patients were ineligible if they
`had previously been treated with murine monoclo-
`nal antibodies or ricin A chain containing toxins.
`
`J Biol Response Mod. Vol. 9, N0. 4, I990
`
`Immunotoxin and Cyclophosphamide
`
`The immunotoxin XMMME-OOl-RTA was pro-
`vided by XOMA Corporation (Berkeley, CA,
`USA.) (9). The immunotoxin is a murine mono-
`
`clonal antibody of the Inga subclass (MW
`150,000:) to which ricin toxin A chain (RTA) (MW
`30,000) is covalently coupled. The conjugation tech-
`nique has been previously described (1,9). The ricin
`A chain is purified by affinity chromatography using
`an anti-ricin B chain column. Briefly, the antibody
`is activated with N-succinimidyl-3-(2-pyridyl-
`dithio)pr0prionate followed by addition of affinity-
`purified ricin A chain that has been reduced with
`dithiothreitol. The immunotoxin is then purified by
`gel chromatography. It is provided in a sterile, py-
`rogen-free formulation at a concentration of 1.0 mg/
`ml in 0.9% phosphate-buffered saline solution, pH
`7.0. The binding specificity of the antibody as de-
`termined by enzyme-linked immunosorbent assay,
`radioimmunoassay, flow cytometry, and immuno-
`peroxidase techniques demonstrates binding with
`all melanomas tested by frozen section and the ma-
`jority of melanoma cell lines. There is no binding
`with other tumors or normal tissues with the excep-
`tion of pigmented nevi and some cytoplasmic bind-
`ing of vascular endothelium. Specificity of the ricin-
`conjugated immunotoxin is identical to that of the
`unlabeled antibody and binding activity is only min-
`imally reduced by conjugation. Lots of 30 mg were
`shipped in 10 mg vials. Each patient was treated
`with a single lot of immunotoxin.
`Commercially prepared cyclophosphamide was
`used.
`
`Treatment Plan
`
`Prior to each treatment, each patient had both a
`skin test and, if negative, an intravenous (i.v.) test
`dose. Skin tests were performed by subcutaneously
`injecting 0.1 ml of saline containing 0.01 mg of im-
`munotoxin. Skin tests were considered to be nega-
`tive if erythema and induration at the site were less
`than 5 mm in greatest diameter at 30 min. After a
`negative skin test, patients received an intravenous
`challenge with 1 ml containing 0.2 mg of immuno-
`toxin. Patients were monitored closely and vital
`signs were recorded for 30 min following the i.v.
`test dose. If no adverse reaction developed, pa-
`tients were then treated with an intravenous infu-
`
`sion of immunotoxin at a dose of 0.4 mg/kg in 150 cc
`of normal saline over 30—60 min. Thirty minutes
`
`IMMUNOGEN 2306, Pg. 2
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2306, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`IMMUN0C0N]UGA TE FOR MALIGNANT MELANOMA
`
`347
`
`after the completion of immunotoxin infusion, pa-
`tients received intravenous cyclophosphamide at a
`dose of 1,000 mg/m2.
`Physical examination, with measurement of indi-
`cator lesions, and laboratory tests were performed
`at baseline and weekly for 1 month on an outpatient
`basis and less frequently thereafter. Patients were
`followed for tumor response for a minimum of 8
`weeks after the initiation of treatment. Laboratory
`tests included blood samples analyzed for complete
`blood count with white blood cell differential, plate-
`let count, and serum chemistry tests including elec-
`trolytes, urea nitrogen, creatinine, and liver en-
`zymes as well as albumin and total serum protein.
`Serum samples for quantitative determinations of
`human antimurine and antiricin immunoglobulins
`were obtained before treatment and at weekly in-
`tervals thereafter.
`
`Assay for Human Antimurine and
`Antiricin Antibodies
`
`The antibody response to immunotoxin compo-
`nents was measured in all patients by a previously
`described enzyme immunoassay method (1). Pa-
`tients’ sera were obtained prior to treatment and at
`weekly intervals thereafter. Appropriate serial dilu-
`tions (1:10 + 12105) were prepared and added to
`microtiter plates that contained either the adsorbed
`murine antimelanoma monoclonal antibody or ad-
`sorbed ricin A chain. The plates were washed and
`incubated for 1 h at room temperature with goat
`anti-human IgG antibody or anti-human IgM conju-
`gated to alkaline phosphatase (Zymed Laborato-
`ries, South San Francisco, CA, U.S.A.). Another
`wash was followed by addition of p-nitrophenyl
`phosphate (Sigma Laboratories, St. Louis, MO,
`U.S.A.). This reaction produced a color titration
`measured spectrophotometrically at 405 nm. Titra-
`tion curves were generated for each serum sample
`and immune responses were expressed as a re-
`sponse ratio: the ratio of the end-point dilution of
`the serum sample showing maximum response to
`the end-point dilution of the pretreatment serum
`sample.
`
`Evaluation of Tumor Responses
`
`Patients were examined weekly for 4 weeks and
`then every 2 weeks for 4 weeks following treatment
`in order to evaluate tumor response. Palpable dis-
`ease was assessed by weekly examination and di-
`
`rect measurement of the perpendicular diameters of
`all measurable nodules. CT scans and chest radio-
`graphs were used to evaluate visceral disease and
`were obtained every 4 weeks. Complete response
`was defined as the disappearance of all measurable
`tumor. Partial response was defined as a reduction
`of all measurable tumors by at least 50% of the sum
`of the product of the two greatest diameters
`present, in the absence of any new lesions or any
`tumor enlargement. Mixed response was a reduc-
`tion in size of some measurable tumors by at least
`50%, but either no change or progressive disease in
`other tumors. Minimal response was a reduction in
`size of less than 50% in some tumors. Stable disease
`was no objective change in all measurable tumors.
`Progressive disease was an increase in size of mea-
`surable tumors by at least 25% or the appearance of
`new lesions. The duration of response was defined
`from the date of therapy until the date of progres-
`sive disease, most recent follow-up, or death. All
`responses were required to persist for at least 30
`days.
`
`RESULTS
`
`Patients
`
`Twenty patients were entered. Their characteris-
`tics are detailed in Table 1. The median age was
`58.5 years (range of 38—73 years). Twelve were male
`and 8 were female. Nine patients (45%) had re-
`ceived no prior treatment for metastatic melanoma,
`whereas 11 patients (55%) had been previously
`
`TABLE 1. Patient characteristics
` No. (%)
`
`58.5
`38—73
`
`Age
`Median (years)
`Range
`Sex
`Male
`Female
`No prior treatment
`Prior treatment
`Chemotherapy
`Radiotherapy
`Immunotherapy
`Sites of metastatic disease
`9 (45)
`Soft tissue/subcutaneous/lymph nodes
`7 (35)
`Lung
`4 (20)
`Liver
`5 (25)
`Spleen
`2 (10)
`Brain
`l (5)
`Adrenal
`—_——_——-—
`
`12 (60)
`8 (40)
`9 (45)
`11 (55)
`7 (35)
`5 (25)
`6 (30)
`
`No. of patients = 20.
`
`J Biol Response Mod, Vol. 9, N0. 4, I990
`
`,
`
`,
`
`IMMUNOGEN 2306, pg. 3
`Phigenix v. Immunogen
`
`|PR2914-9“§i§
`
`
`
`IMMUNOGEN 2306, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`
`
`348
`
`R. ORATZ ET AL.
`
`treated with chemotherapy, radiation therapy, im-
`munotherapy, or a combination of these. Most pa-
`tients had more than one site of metastatic disease.
`
`included
`The predominant areas of involvement
`skin and soft tissue, lung, and liver. Two patients
`with resected and irradiated brain metastases were
`
`treated on this protocol.
`
`Toxicity
`
`All patients were evaluable for toxicity. Overall,
`the combination treatment was well tolerated and
`
`toxicity was manageable. Patterns of toxicity are
`outlined in Table 2. No patient had a positive reac-
`tion to the skin test dose. No patient developed
`hypotension, tachycardia, rash, hives, or wheezing
`during the intravenous test dose of immunotoxin.
`One patient had an episode of sneezing during the
`i.v. test dose with no other symptoms. After receiv-
`ing approximately 30 cc of the intravenous infusion
`dose, he developed facial flushing, increased lacri-
`mation, and swelling of the lower lip. He had no
`dyspnea, wheezing, stridor, rash, hypotension,
`tachycardia, or fever. The infusion was discontin-
`ued and the patient was given 50 mg of diphenhydr-
`amine by i.v. bolus. His symptoms resolved, and
`the immunotoxin infusion was resumed. The treat-
`
`ment was completed at a slower infusion rate and
`was well tolerated.
`
`Other immediate toxicities from this regimen in-
`cluded nausea and vomiting in 18/20 (90%) patients
`in the first 24—48 h (mild in 13 patients, moderate in
`5 patients), which was felt to be due to cyclophos-
`phamide. Low-grade fevers were seen in 4/20 (20%)
`patients during the first 72 h after treatment. Most
`patients complained of constitutional symptoms
`
`consisting of fatigue, malaise, myalgias, and arthral-
`gias during the first several days after treatment.
`This was reflected in a general decline in perfor-
`mance status by at least 10—20%, and resolved with
`return to baseline performance status by the middle
`of the second week after treatment.
`
`Within the first 2 weeks following treatment, hy-
`poalbuminemia was noted in 15/20 (75%) patients,
`manifested by decreases in serum albumin by less
`than 0.2 g/dl in 5 patients, 20.2—05 g/dl in 3 pa-
`tients, and greater than 0.5 g/dl in 7 patients. Only
`five of these patients developed clinically evident
`peripheral edema:
`three patients had mild ankle
`swelling, one patient had lower extremity edema
`that was treated with oral furosemide, and one pa-
`tient developed significant swelling of the left arm
`that had been the site of his primary melanoma and
`left axillary lymphadenectomy. The upper extrem-
`ity edema was managed with oral diuretics, arm el-
`evation, and an elastic arm stocking and resolved
`within 1 week. Three patients reported mild to mod-
`erate dyspnea—but pulmonary edema was not doc-
`umented on chest radiographs and on no occasion
`were rales or wheezing appreciated on auscultation.
`One patient became acutely ill within 24 h of
`treatment. She had a history of resected and irradi-
`ated brain metastases. A CT scan of the brain done
`
`less than 4 weeks prior to treatment showed no ev-
`idence of involvement, and her pretreatment neu-
`rologic exam was unremarkable. Nonetheless, she
`developed grand mal seizures on the evening fol-
`lowing treatment. A repeat CT scan the next day
`revealed the presence of multiple new brain me-
`tastases. Furthermore, this patient remained hospi-
`talized and developed grade 4 neutropenia, grade 3
`anemia, fever, and sepsis. She had been heavily
`
`TABLE 2. Toxicity results
`
` Grade None (0) Mild (1) Moderate (2) Severe (3) Life-threatening (4)
`
`
`
`
`
`
`
`
`
`
`
`0
`0
`1
`3
`16
`Fever
`0
`4
`7
`5
`4
`Fatigue
`0
`0
`0
`9
`ll
`Malaise
`0
`0
`2
`8
`l0
`Myalgias
`0
`l
`0
`l
`18
`Arthralgias
`0
`0
`I
`2
`17
`Dyspnea
`0
`l
`l
`3
`15
`Edema
`0
`7
`3
`5
`5
`Decreased albumin
`0
`0
`5
`l3
`2
`Nausea/vomiting
`l
`1
`5
`6
`7
`Neutropenia
`0
`1
`l
`1
`l7
`Anemia
`1
`3
`8
`6
`2
`Decrease in performance status
`0
`0
`l
`0
`19
`Seizures"
`—_————_—
`
`" See the discussion in the Toxicity section.
`
`J Biol Response Mod, Vol. 9, No. 4, 1990
`
`IMMUNOGEN 2306, pg. 4
`Phigenix v. Immunogen
`|PR2014-00676
`
`
`IMMUNOGEN 2306, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`IMMUN0CONJ UGA TE FOR MALIGNANT MELANOMA
`
`349
`
`pretreated with multiple cytotoxic and myelosup-
`pressive regimens. Despite a measureable partial
`response in her non-central nervous system tumor,
`the patient expired 43 days following treatment, re-
`lated to nadir sepsis and brain metastases. There
`were no other deaths during the study period.
`
`Clinical Responses
`
`Clinical outcomes of responding patients are de-
`tailed in Table 3. There were no complete responses
`to treatment. Four (20%) partial responses were
`seen. One patient achieved a partial response as
`measured by disappearance of several pulmonary
`nodules and reduction of all others by at least 50%.
`Early response was noted on the chest radiograph
`within 1 month after treatment, and improvement
`continued over 2—3 months. This patient received a
`second immunotoxin treatment 4 months after his
`initial treatment and the lesions then stabilized for a
`
`total response duration of 1 year. During this time,
`the patient was clinically well, and worked regu-
`larly. A second patient had a partial response of
`skin, soft tissue, and lymph node disease including
`multiple tumors involving the gastrointestinal tract.
`He reported tenderness of the responding subcuta-
`neous and soft tissue nodules during the first week
`after treatment. These became erythematous and
`warm, and over the next 2—3 weeks gradually be-
`came softer and smaller. Some nodules disappeared
`completely. The duration of response was 10
`weeks. Two other patients had partial responses in
`subcutaneous and soft tissue nodules of 6 and 15
`weeks duration. Of interest, responding nodules in
`general did not grow at the time of disease progres-
`sion. Progressive disease was usually characterized
`
`by growth of nonresponding nodules or appearance
`of new metastases. One patient (5%) had a minor
`response characterized by a 30% reduction in a soft
`tissue pelvic mass of greater than 56 weeks dura-
`tion. She has required no further treatment of her
`melanoma. Two patients (10%) had mixed re-
`sponses in which there was a 50% decrease in the
`size of soft tissue metastases but progressive dis-
`ease in visceral lesions. Three patients (15%) had
`stable disease throughout the study period and 10
`patients (50%) had progressive disease.
`
`Immune Responses
`
`In 13 patients, pre- and posttreatment titers of
`human antimurine antibodies and antiricin antibod-
`
`ies were determined separately, and in seven pa-
`tients antibody titers against the complete immuno-
`toxin were measured. In all instances but one, an-
`
`tibody titers against the mouse immunoglobulin,
`ricin A chain, and whole immunotoxin rose after
`treatment. Patient #122 did not mount an antibody
`response against the ricin moiety of the immuno-
`conjugate but did produce a response ratio of 7.5 in
`the human antimurine antibody (HAMA) response.
`Baseline, maximum end-point titers, and day 28 ti-
`ters are shown in Tables 4—6. The maximum re-
`
`sponse ratios are displayed graphically in Fig. 1.
`The median HAMA response ratio was 1.25 (range
`of 3-100). The median response ratio to the ricin A
`chain component was 32 (range of 1—250). In the
`seven patients so tested, the median response ratio
`to the whole immunotoxin was 13.3 (range of 4.5—
`62.5).
`The single dose of 1,000 mg/m2 of cyclophospha—
`mide used in this study neither abrogated the pro-
`
`TABLE 3. Clinical features of responding patients
`_—_—_—__———————_———
`
`Duration (weeks)
`Response
`Prior therapy
`Sites of disease
`Age (years)/sex
`Patient no.
`__________________—_.—————————
`
`121
`124
`
`130
`
`73/M
`45/M
`
`59/F
`
`138
`125
`133
`
`135
`
`73/F
`58/F
`46/M
`
`SW
`
`Lung
`Soft tissue, skin,
`lymph nodes, GI tract
`Soft tissue, skin,
`lymph nodes, brain
`Soft tissue, lymph nodes
`Lymph nodes
`Lymph nodes
`
`Soft tissue, skin,
`GI tract
`
`15
`56 +
`6 (progression in spleen,
`continued response in
`lymph nodes)
`15 (progression in GI tract
`but continued response in
`skin and soft tissue)
`_—____—___—____————_——
`
`None
`Chemotherapy
`
`Chemotherapy/RT
`
`PR
`PR
`
`PR
`
`53
`10
`
`6
`
`None
`None
`None
`
`None
`
`PR
`Minor
`Mixed
`
`Mixed
`
`
`
`
`
`PR, partial response.
`
`J Biol Responxe Mod, Vol. 9, No. 4, I990
`
`IMMUNOGEN 2306, pg. 5
`Phigenix v. Immunogen
`
`
`
`IMMUNOGEN 2306, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`
`
`350
`
`R. ORATZ ETAL.
`
`TABLE 4. Human antimurine antibody (HAMA) responses
`Maximum
`Response ratio
`Response ratio
`
`
`
`
` Patient no. Baseline titer end—point titer Day 28 titer max./basline day 28/base1ine
`
`
`
`
`
`100
`100
`10,000
`10.000
`100
`120
`33
`33
`3,300
`3,300
`100
`121
`7.5
`7.5
`12,000
`12,000
`1,600
`122
`18
`18
`30,000
`30,000
`1,666
`123
`12.5
`12.5
`10,000
`10,000
`800
`124
`20
`20
`20,000
`20,000
`1,000
`125
`3.5
`4
`14,000
`16,000
`4,000
`126
`10
`10
`35,000
`35,000
`3,500
`127
`4.4
`4.4
`14,000
`14,000
`3,182
`128
`50
`75
`100,000
`150,000
`2,000
`129
`31.3
`37.5
`50,000
`60,000
`1,600
`130
`7.5
`7.5
`30,000
`30,000
`4,000
`131
`
`
`
`
`
`800 24,000 24,000 3.0132 3.0
`
`duction of these antibodies nor decreased the titers
`
`fact that the dose of immunotoxin administered was
`
`of the responses. The median and overall response
`ratios reported here are not significantly different
`than those previously published for treatment with
`the immunotoxin alone (1). Despite the appearance
`of the antibody response, one patient underwent
`repeat skin test and iv. challenge that were both
`negative, and he was successfully retreated with im—
`munotoxin without an allergic reaction.
`
`DISCUSSION
`
`This study was conducted to determine whether a
`single large dose of cyclophosphamide given imme—
`diately after immunotoxin administration would ab-
`rogate the immune response to the immunotoxin
`components. Suppression of the immune response
`was not achieved. An unexpected clinical observa-
`tion, however, was that the tumor response rate in
`patients receiving this combination was as good as
`or better than that previously observed, despite the
`
`one-fifth of that used in previous studies. This sug-
`gests the possibility of a synergistic action between
`the immunotoxin and cyclophosphamide.
`In a previous phase I trial,
`1 complete response
`was observed in 21 evaluable patients (1). In a sub-
`sequent phase II study, there were 3 partial re-
`sponses in 43 patients (3). Thus, in these studies,
`there were 4 responses in 64 patients as compared
`with 4 responses in 20 patients in the current study.
`This difference is of borderline statistical signifk
`cance (p = 0.07); but it suggests that the addition of
`cyclophosphamide might improve the efficacy of
`the immunotoxin. It is unlikely that cyclophospha-
`mide in the dose and schedule used in this trial re-
`
`sulted in the antitumor activity seen. However, it is
`possible that the cyclophosphamide modulated host
`lymphocyte subsets such that, rather than blunting
`humoral responses against the immunotoxin, it en-
`hanced immune responses against melanoma. In a
`recent study by Uekun et al., the cyclophospha-
`
`TABLE 5. Human antiricin antibody (HARA) responses
`
`Maximum
`Response ratio
`Response ratio
`
`Patient
`Baseline titer
`end-point titer
`Day 28 titer
`max./basline
`day 28/baseline
`
`32
`32
`3,200
`3,200
`100
`120
`20
`20
`2,000
`2,000
`100
`121
`1
`1
`100
`100
`100
`122
`40
`40
`8,000
`8,000
`200
`123
`100
`100
`10,000
`10,000
`100
`124
`250
`250
`50,000
`50,000
`200
`125
`32
`32
`64,000
`6,400
`200
`126
`1.3
`1.3
`800
`800
`615
`127
`25
`25
`5,000
`5,000
`200
`128
`12.5
`12.5
`10,000
`10,000
`800
`129
`100
`100
`40,000
`40,000
`400
`130
`42.7
`42.7
`25,600
`25,600
`600
`131
`24
`24
`24,000
`24,000
`1,000
`132
`—————_——_——__—
`
`J Biol Response Mod, Vol. 9, No. 4, I990
`
`IMMUNOGEN 2306, pg. 6
`Phigenix v. Immunogen
`|PR2014-00676
`
`IMMUNOGEN 2306, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`IMM UN0C0NJ UGA TE FOR MALIGNANT MELANOMA
`
`35 I
`
`TABLE 6. Human anti-immunotoxin antibody responses
`Maximum
`Response ratio
`Response ratio
`
`
`
`
`
` Patient no. Baseline titer end-point titer Day 28 titer max./baseline day 28/base1ine
`
`
`
`32
`32
`12,800
`12,800
`400
`133
`4.5
`6.4
`9,000
`12,800
`2,000
`134
`30
`30
`24,000
`24,000
`800
`135
`12
`12
`60,000
`60,000
`5,000
`136
`8
`8
`24,000
`24,000
`3,000
`137
`13.3
`13.3
`40,000
`40,000
`3,000
`138
`
`
`
`
`
`800 50,000 50,000 62.5139 62.5
`
`mide congener, mafosfamid, markedly enhanced
`the target cell cytotoxicity of a ricin-conjugated
`monoclonal antibody in an in vitro clonogenic assay
`(10). The authors found that mafosfamid appeared
`to shorten the lag period seen in the immunotoxin
`inactivation of protein synthesis. They speculated
`that the mafosfamid effect might be related to (a)
`alteration of the chemical processing of the immu-
`
`notoxin by target cells, (b) action at the ribosomal
`level, (0) increased sensitivity of the neoplastic cells
`because of a decline in aldehyde dehydrogenase ac-
`tivity, or (d) immunotoxin-induced acceleration of
`mafosfamid to reactive mustard and acrolein metab-
`
`olites. Our study was not designed to examine such
`effects and therefore we did not collect data di-
`
`rected at addressing these questions. Further stud-
`
`300
`
`250
`
`200
`
`I50
`
`100
`
`
`
`RESPONSERATIO
`
`
`
`50
`
`WHOLE
`IMMUNOTOXIN
`
`(N=7)
`
`FIG. 1. Response ratios of maximum end-point titers of human antimurine antibodies (HAMA), human antiricin antibodies, and human
`anti-immunotoxin antibodies.
`
`J Biol Response Mod, Vol. 9, No. 4, I990
`
`IMMUNOGEN 2306, pg. 7
`Phigenix v. Immunogen
`
`IP -
`
`IMMUNOGEN 2306, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`352
`
`R. ORATZ ETAL.
`
`ies of this combination of immunotoxin plus cyclo-
`phosphamide on host antitumor immune responses
`are needed to confirm the activity seen in our study
`and to elucidate potential mechanisms of action.
`Cyclophosphamide is known to have effects on
`the cell membrane and it is possible that this re-
`sulted in enhanced efficacy of the immunotoxin per-
`haps by increasing entry of the immunotoxin into
`the cell or by making the cell more susceptible to
`the toxic activity of the ricin A chain.
`It
`is noteworthy that a delayed antitumor re-
`sponse occurred in some patients suggesting that
`mechanisms in addition to direct ricin A chain tox-
`
`icity may be relevant in the antitumor activity of
`this immunotoxin. In a number of instances, soft
`tissue deposits became erythematous and tender in
`the first week after treatment and then gradually
`became softer and smaller, mimicking an inflamma-
`tory reaction. Furthermore, in one patient with mul-
`tiple pulmonary metastases, and several with mul-
`tiple soft tissue nodules, we observed continued,
`gradual response over a period of several weeks.
`An excisional biopsy of a responding lymph node in
`one patient revealed necrosis and small amounts of
`residual tumor with dense lymphocytic infiltration,
`11 months after treatment (Fig. 2A and B). This
`type of response is similar to the inflammation de-
`scribed at tumor sites during treatment with the R24
`antibody.
`This time course of response suggests that mech-
`anisms other than direct ricin activity may be rele—
`vant in the antitumor activity of this immunotoxin.
`It has been proposed that monoclonal antibodies
`induce tumor regressions by directly facilitating an-
`tibody- (ADCC) or complement-dependent cellular
`cytotoxicity (11,12). It is also possible that, as has
`been demonstrated in a number of preclinical ani-
`mal studies, monoclonal antibodies interact with
`host effector cells, stimulating them to recognize
`and kill tumor cells (13—19). The precise nature of
`the interactions of the XMMME-OOl-RTA mono-
`
`clonal antibody with host effector cells, whether
`lymphocytes or macrophages, has not yet been de-
`fined. More detailed studies perhaps in combination
`with lymphokines [such as interleukin-2 (IL—2)] and/
`or monocyte stimulators ('y-interferon) might be
`conducted in order to elucidate these immune ef-
`fects.
`
`Also of interest in our study was the pattern of
`responses. Significant tumor regressions were ob-
`served in pulmonary lesions, subcutaneous nod-
`ules, and lymph nodes, whereas visceral sites of
`
`J Biol Response Mod, Vol. 9, N0. 4. 1990
`
`
`
`metastases rarely responded. Antigenic heterogene-
`ity of tumor cells may provide one explanation for
`this variability in sites of response. It is known that
`melanoma cells vary widely both at the quantitative
`level and with respect to qualitative patterns of ex-
`pressed antigens (20,21). It is possible that a single
`monoclonal antibody directed against a specific tu-
`mor-associated antigen would bind only a portion of
`and not all tumor cells—thereby producing regres-
`sions in only some but not all clinically apparent
`metastatic deposits. Of interest, in our patients we
`saw little if any progression in nodules that had ini-
`tially responded to treatment—progressive disease
`was almost exclusively noted either in nonrespond-
`ing metastatic sites—or in the development of new
`metastases. This pattern of response has been pre-
`viously noted in earlier clinical trials with immuno-
`toxin XMMME-RA-OOI, and seems to mirror re-
`
`sponse patterns observed with chemotherapy and
`other immunotherapies such as a-interferon, i.e.,
`increased responsiveness of lymph nodes, soft tis-
`sue, and pulmonary metastases rather than visceral
`or bone disease (1,22).
`In this protocol, we added cyclophosphamide as
`an immunosuppressive agent in an effort to abro-
`gate the host antibody response against both the
`murine antibody and ricin A chain components of
`the immunotoxin. Cyclophosphamide is known to
`have selective and potent actions in inhibiting B cell
`function. The dose and schedule of administration
`
`relative to antigenic stimulus are critical in the ef-
`fect of cyclophosphamide on the immune response.
`When given in moderate doses prior to antigenic
`stimulus,
`immune responses are potentiated (23—
`25). Treatment following antigenic stimulus may
`lead to immune suppression (4—8). In this trial, titers
`and response ratios of human antimurine immuno-
`globulin antibodies and antiricin antibodies. were
`similar to those seen in earlier studies using
`XMMME-OOl-RTA (1). In the dose and schedule
`employed in this trial, we were unable to demon-
`strate any suppression of host anti-immunotoxin
`immune response.
`Other investigators working with XMMME-
`001-RTA have studied the effects of a variety of
`immunosuppressive regimens on the antibody re-
`sponses of treated patients. Preliminary data have
`been reported in abstract form (26,27). Patients re-
`ceiving aziathioprine/prednisone combinations had
`more immunosuppression of anti-immunotoxin an-
`tibodies than patients receiving moderate dose cy-
`clophosphamide (250 mg/m2 p.o. X 5 days) with
`
`IMMUNOGEN 2306, pg. 8
`Phigenix v. Immunogen
`IPR2014-00676
`
`IMMUNOGEN 2306, pg. 8
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`[MM UN0CONJ UGA TE FOR MALIGNANT MELANOMA
`
`353
`
`FIG. 2. Excised lymph node 11
`months following treatment with
`XMMME—OOl—RTA immunocon—
`jugate and cyclophosphamide. (A)
`9.7x magnification; (B) 38.6x
`magnification.
`
`
`
`
`
`prednisone (100 mg/day X 5 days). Data on the high
`dose cyclophosphamide (400 mg/m2 p.o. X 5 days)
`with prednisone regimen are not yet available. In 3%:
`patients given cyclophosphamide at 100 mg/m2 p.o.
`on days 1—14, antibodies to murine immunoglobulin
`and ricin A chain components were suppressed.
`Trials with these agents are ongoing.
`In conclusion, we found the combination of the
`XMME—OOl-RTA immunotoxin and cyclophospha-
`mide to be safe and more effective than either agent
`
`used singly in the treatment of metastatic malignant
`melanoma. Further in vitro and clinical studies
`should be directed towards a more detailed defini-
`tion of the tumoricidal effects of the immunotoxin—
`
`cyclophosphamide combination—with specific at-
`tention to possible interactions with host effector
`cells. Administration with IL-2 and lymphokine-
`activated killer cells, if ADCC is found to be a pri-
`mary mechanism, or y-interferon in order to aug-
`ment monocyte activity, might further enhance the
`
`JBiol Response Mod, Vol. 9, No. 4, I990
`
`IMMUNOGEN 2306, Pg. 9
`Phigenix v. Immunogen
`
`|PR2014-99§7§
`
`IMMUNOGEN 2306, pg. 9
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`354
`
`R. ORATZ ET AL.
`
`antitumor effect of this regimen. The role of cyclo-
`phosphamide in modulating lymphocyte subpopula-
`tions in this regimen requires clarification. More ef-
`fective means of abrogating host immune response
`must be explored.
`
`REFERENCES
`
`10.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`21.
`
`22.
`
`23.
`
`24.
`
`25.
`
`26.
`
`27.
`
`hibit human tumor growth through interaction with effector
`cells. Proc Natl Acad Sci USA 1982;79:4761—65.
`Ben'nstein N, Levy R, Treatment of a murine B-cell lym-
`phoma with monoclonal antibodies and