Multicenter, Randomized, Phase III Trial of CD8 1
`Tumor-Infiltrating Lymphocytes in Combination
`With Recombinant Interleukin-2 in Metastatic
`Renal Cell Carcinoma
`
`By Robert A. Figlin, John A. Thompson, Ronald M. Bukowski, Nicholas J. Vogelzang, Andrew C. Novick,
`Paul Lange, Gary D. Steinberg, and Arie S. Belldegrun
`
`Purpose: To prospectively evaluate in a multicenter
`randomized trial the antitumor activity of CD81 tumor-
`infiltrating lymphocytes (TILs) in combination with low-
`dose recombinant interleukin-2 (rIL-2), compared with
`rIL-2 alone, after radical nephrectomy in metastatic
`renal cell carcinoma patients.
`PatientsandMethods: Between December 1994 and
`March 1997, 178 patients with resectable primary tu-
`mors were enrolled at 29 centers in the United States
`and Europe. Patients underwent total nephrectomy,
`recovered, and were randomized to receive either CD81
`TILs (5 3 107 to 3 3 1010 cells intravenously, day 1) plus
`rIL-2 (one to four cycles: 5 3 106 IU/m2 by continuous
`infusion daily for 4 days per week for 4 weeks) (TIL/
`rIL-2 group) or placebo cell infusion plus rIL-2 (identical
`regimen) (rIL-2 control group). Primary tumor speci-
`mens were cultured at a central cell-processing center in
`serum-free medium containing rIL-2 to generate TILs.
`
`Results: Of 178 enrolled patients, 160 were random-
`ized (TIL/rIL-2 group, n 5 81; rIL-2 control group,
`n 5 79). Twenty randomized patients received no treat-
`ment after nephrectomy because of surgical complica-
`tions (four patients), operative mortality (two patients),
`or ineligibility for rIL-2 therapy (14 patients). Among 72
`patients eligible for TIL/rIL-2 therapy, 33 (41%) received
`no TIL therapy because of an insufficient number of viable
`cells. Intent-to-treat analysis demonstrated objective re-
`sponse rates of 9.9% v11.4% and 1-year survival rates of
`55% v47% in the TIL/rIL-2 and rIL-2 control groups, respec-
`tively. The study was terminated early for lack of efficacy
`as determined by the Data Safety Monitoring Board.
`Conclusion: Treatment with CD81 TILs did not im-
`prove response rate or survival in patients treated with
`low-dose rIL-2 after nephrectomy.
`J Clin Oncol 17:2521-2529. 1999 by American
`SocietyofClinicalOncology.
`
`APPROXIMATELY 30% OF renal cell carcinoma (RCC)
`
`patients present with metastatic disease, and 20% to
`30% of patients who present with clinically localized disease
`will develop metastatic disease after radical nephrectomy,
`yielding a 10-year disease-free survival rate of approxi-
`mately 50%.1-5 Metastatic renal cell carcinoma (MRCC) is
`associated with a poor prognosis because it
`is highly
`resistant to chemotherapy, hormonal therapy, and radiation
`therapy. The 5-year survival rate varies from 0% to 20%,
`depending on cell type and the extent of disease at the time
`of nephrectomy,4,6 but it is generally less than 2%.7 Clinical
`studies have demonstrated that high-dose intravenous (IV)
`bolus recombinant
`interleukin-2 (rIL-2) (ie, 600,000 to
`720,000 IU/kg every 8 hours) can induce durable complete
`remissions (CRs) in patients with bulky disease and multiple
`visceral metastases, with objective response rates ranging
`from 13% to 20%. In patients treated with high-dose rIL-2,
`1-year survival rates of approximately 55% and 5-year
`survival rates of 10% to 20% have been reported.8-11
`Comparable response rates and survival have also been
`observed with regimens of high-dose continuous IV infusion
`(CIV).12,13
`However, a limitation to the administration of high-dose
`IV rIL-2 is the occurrence of acute toxicity,
`including
`hypotension, oliguria, pulmonary edema, and dyspnea, re-
`lated to capillary leak syndrome. The morbidity associated
`
`with high-dose IV rIL-2 regimens has led to the investiga-
`tion of alternative dosage regimens to determine whether
`durable CRs can be achieved without significant toxicity.
`Administration of low-dose rIL-2 by IV bolus, CIV, or
`subcutaneous (SC) injection either alone or in conjunction
`with recombinant interferon alfa and/or fluorouracil may
`have activity in the treatment of MRCC, and these regimens
`are generally better tolerated.14-18 Moreover, low-dose CIV
`rIL-2 can produce selective expansion of natural killer cells
`in vivo with minimal toxicity.19 This has been described as
`the most physiologic immunotherapeutic strategy to activate
`the anticancer immune response.20 However, further fol-
`low-up is required to determine whether CRs associated
`with low-dose rIL-2 regimens will be as durable as those
`achieved with high-dose rIL-2 regimens.17,18,21-23
`
`From the University of California, Los Angeles, Los Angeles, CA;
`University of Washington, Seattle, WA; Cleveland Clinic Foundation,
`Cleveland, OH; and University of Chicago, Chicago, IL.
`Submitted November 18, 1998; accepted March 19, 1999.
`Supported by Rhoˆne-Poulenc Rorer, Collegeville, PA.
`Address reprint requests to Robert A. Figlin, MD, FACP, University of
`California, Los Angeles School of Medicine, Jonsson Comprehensive
`Cancer Center, 10945 Le Conte Ave, Suite 2333 PVUB, Box 957059,
`Los Angeles, CA 90095-7059; email rfiglin@med1.medsch.ucla.edu.
`1999 by American Society of Clinical Oncology.
`0732-183X/99/1708-2521
`
`JournalofClinicalOncology, Vol 17, No 8 (August), 1999: pp 2521-2529
`
`2521
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216 185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 1 of 9
`
`

`

`2522
`
`One strategy to potentially enhance the efficacy of rIL-2
`therapy is to combine rIL-2 with adoptive immunotherapy,
`using lymphokine-activated killer cells or tumor-infiltrating
`lymphocytes (TILs).12,24,25 TILs are found in high numbers
`in RCC tumors and can be expanded ex vivo in the presence
`of rIL-2, yielding predominantly T lymphocytes.26,27 Murine
`models and clinical studies have suggested that TILs plus
`rIL-2 may act synergistically to activate the cellular immune
`response and mediate tumor regression.28-30 In a phase I/II
`trial at the University of California, Los Angeles, immuno-
`therapy with TILs plus low-dose rIL-2 has produced signifi-
`cant clinical activity in MRCC, with objective response rates
`of 33% to 35% and 1-year survival rates of 65% to 73%.31,32
`In a pilot study involving 55 patients treated with nephrec-
`tomy followed by TILs plus low-dose CIV rIL-2 (2 3 106
`IU/m2/d), 19 patients (34.6%) responded and five (9%)
`achieved a CR.32 Moreover, among 23 patients who received
`CD81 TILs, the overall response rate was 43.5%. Overall,
`the median response duration was 14 months, and the
`actuarial survival rate was 65% at 1 year and 43% at 2 years
`after radical nephrectomy.
`On the basis of this encouraging single-institution study, a
`randomized, multicenter study was conducted to prospec-
`tively compare CD81 TILs plus low-dose rIL-2 (TIL/rIL-2
`group) versus low-dose rIL-2 alone (rIL-2 control group).
`All patients underwent radical nephrectomy, from which
`tissue was obtained for generating CD81 TILs. The rationale
`for selecting CD81 TILs was based on the promising results
`of the pilot study and on previous in vitro characterization of
`TILs, which suggest that the CD81 subset has the greatest
`cytotoxic potential against autologous or allogeneic tumor
`cells.26 The goals of the current study were to investigate the
`safety, efficacy, and feasibility of CD81 TIL therapy in
`conjunction with low-dose CIV rIL-2 in a multi-institutional
`setting.
`
`PATIENTS AND METHODS
`Patient Eligibility
`
`Eligibility criteria included Eastern Cooperative Oncology Group
`(ECOG) performance status (PS) of 0 or 1, histologic or radiologic
`documentation of RCC with the primary tumor suitable for resection,
`bidimensionally measurable metastatic disease, age $ 18 years, willing-
`ness and ability to undergo surgery, willingness and agreement to use
`contraception, and informed consent. Exclusion criteria were prior
`rIL-2 therapy,
`immunotherapy,
`immunosuppressive therapy, radio-
`therapy, or chemotherapy within 4 weeks of screening; significant renal
`dysfunction (ie, serum creatinine level $ 2.0 mg/dL), significant he-
`patic dysfunction (ie, serum total bilirubin level . 1.6 mg/dL,
`ALT . four times normal, and partial
`thromboplastin time . 1.5
`control); inadequate blood counts (ie, hemoglobin count , 8 g/dL,
`granulocyte count # 1,500 cells/mm3, platelet count , 100,000/mm3);
`significant cardiovascular disease (ie, heart failure, ischemia, edema,
`
`FIGLIN ET AL
`
`infarction, or hypertension); CNS disease;
`arrhythmia, myocardial
`pleural effusions or ascites; active infection; active peptic ulcer disease;
`antibodies to human immunodeficiency virus, hepatitis B surface
`antigen, or hepatitis C; only bone or abdominal metastases; prior history
`of malignancy within the last 5 years other than basal cell carcinoma or
`cervical carcinoma-in-situ; serum calcium level greater than 12 mg/dL
`or symptomatic hypercalcemia; use of corticosteroids or calcium
`channel and beta adrenergic blockers; women who were pregnant
`and/or nursing; solitary kidney; significant intercurrent illnesses; and
`New York Heart Association class III or IV.
`
`Study Design
`
`Between December 1994 and March 1997, MRCC patients were
`enrolled onto this phase III, double-blind, randomized study at 29
`centers (19 university hospitals and three community hospitals in the
`United States [U.S.] and seven sites in Europe). The study was
`conducted in compliance with both U.S. Food and Drug Administration
`laws and European Good Clinical Practice Guidelines and approved by
`institutional review boards for U.S. sites and by ethics committees for
`European sites.
`After radical nephrectomy and procurement of $ 10 g of viable
`tumor tissue, and with pathologic confirmation of MRCC, patients were
`randomized to treatment with either CD81 TILs plus rIL-2 (TIL/rIL-2
`group) or control infusion plus rIL-2 (rIL-2 control group). Depending
`on the rate of expansion of the TIL cell cultures, treatment in the
`TIL/rIL-2 group was generally initiated 4 to 7 weeks after nephrectomy.
`Treatment in the rIL-2 control group was initiated approximately 5
`weeks after surgery. Recombinant human IL-2 (Proleukin; Chiron
`Therapeutics, Emeryville, CA) was administered to all eligible patients
`on consecutive days 1 to 4 of each treatment week via CIV using a pump
`dispensed per institutional policy at a daily dose of 5 3 106 IU/m2. One
`treatment cycle consisted of 4 weeks of treatment followed by 2 weeks
`of rest. After at least 2 hours of rIL-2 therapy on the first day of the first
`cycle, patients in the TIL/rIL-2 group received a single IV infusion of
`5 3 107 to 3 3 1010 CD81 TILs, and patients in the rIL-2 control group
`received a placebo (5% human serum albumin) infusion. If fewer than
`5 3 107 cells were available, all harvested cells were infused. Patients
`were hospitalized only during the initial week of the first cycle of
`therapy to permit close monitoring of adverse events. Thereafter, the
`pump was initiated by the nursing staff in the outpatient setting and the
`patient was instructed on how to disconnect it.
`Restaging of measurable disease by computed tomography (CT) scan
`was performed every 6 weeks. According to tolerance and response to
`rIL-2 therapy, patients continued treatment
`to either CR, disease
`progression, dose-limiting toxicity, or a maximum of four cycles (24
`weeks) of therapy. Treatment was withheld in patients with grade 3
`toxicity (excluding granulocytopenia) or grade 2 neurocortical and/or
`cardiac toxicity according to National Cancer Institute Toxicity Criteria.
`Upon reversal to # grade 1 toxicity, treatment with rIL-2 was resumed
`at a reduced dose (80% of the dose at occurrence of toxicity). Further
`reduction to 60% was then allowed as necessary at the investigator’s
`discretion. Patients who experienced grade 4 nonhematologic toxicity
`or grade 3 neurotoxicity or cardiotoxicity were withdrawn from the
`study and entered follow-up evaluation.
`Patients received 650 mg of oral acetaminophen for body tempera-
`ture $ 38.5 C, oral diphenhydramine for rash and pruritus, meperidine
`(25 to 50 mg oral or IV) or morphine sulfate (4 to 6 mg SC or IV) for
`chills, 10 mg of oral prochlorperazine for nausea, and diphenoxylate
`and atropine for diarrhea. Patients were premedicated 30 minutes before
`the start of rIL-2 infusion and as required thereafter. Prophylactic oral
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216.185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 2 of 9
`
`

`

`CD81 TILs PLUS rIL-2 IN METASTATIC RCC
`
`ofloxacin or its equivalent was administered at the start of rIL-2
`administration.
`Patients with documented progressive disease who were previously
`assigned to the rIL-2 control group and who met the rIL-2 eligibility
`criteria were eligible for cross-over to the TIL/rIL-2 group.
`
`TIL Preparation
`
`Surgery specimens were obtained directly from the operating room
`on day 0 in a sterile fashion and sent in cold, sterile saline to the central
`cell-processing center. In the U.S., cell processing occurred at the Ex
`VT cell-processing center in Torrance, CA. In Europe, cells were
`processed at the Zentrallaboratorium Center in Bern, Switzerland. The
`primary tumor was dissected under sterile conditions and digested with
`overnight stirring in a sterile solution containing collagenase (type IV,
`0.1W), hyaluronidase (type V, 0.01), and deoxyribonuclease (type I,
`0.002%). The single-cell suspensions were washed three times in cold
`phosphate-buffered saline and separated (450 3 g, 35 minutes) over a
`Histopaque 1077 (Sigma, St Louis, MO) layer to concentrate the viable
`cells. The cells were again washed three times in phosphate-buffered
`saline and finally resuspended in a serum-free medium (5% human
`serum albumin) containing rIL-2 (1,200 IU/mL). Cells were cultured
`until $ 20% of cells in the unselected cultures were CD81 by
`fluorescent cell-sorter analysis. CD81 lymphocytes were selected using
`CELLector CD8 T-150 culture flasks (Applied Immune Sciences, Santa
`Clara, CA) coated with anti-CD8 antibodies. The CELLector flasks
`positively select adherent CD81 T cells by immunoaffinity. The
`nonadherent CD82 cells were removed by washing. After exposure to
`medium containing rIL-2 (1,200 IU/mL) and phytohemagglutinin for 3
`the activated adherent CD81 cells were removed from the
`days,
`CELLector flasks, transferred to cell culture bags or flasks, and allowed
`to expand in vitro to reach a total cell number of 4 3 109 to 1 3 1010
`cells. The expanded cells were recovered, washed, and resuspended in
`5% human serum albumin. Harvested cells were transported to the
`clinic site only if the cell yield and viability, gram stain, bacterial
`endotoxin, mycoplasma testing, and bacterial sterility cultures met
`release criteria. A sample was retained for testing of phenotype,
`cytokine expression, and cytotoxic activity. Cells collected from
`patients in the rIL-2 control group were cytopreserved for possible
`future use.
`
`Response Assessment
`
`Patients were evaluated for response after each treatment cycle. A CR
`was defined as the complete disappearance of all clinically detectable
`disease for a minimum of 4 weeks. Positive bone scans had to revert to
`normal or show sclerotic healing of lytic metastases, if present. Partial
`response (PR) was defined as a $ 50% decrease in the sum of the
`product of the two greatest perpendicular diameters of all measurable
`marker lesions for at least 4 weeks. Any increase of less than 25% or
`decrease of less than 50% throughout the period of treatment was
`considered stable disease. For both PR and stable disease no simulta-
`neous progression of assessable disease or appearance of any new lesion
`could occur, nor could there be worsening of existing lesions or
`appearance of new ones on bone scan. Progressive disease (PD) was
`defined as a $ 25% increase in the size of one or more marker lesions
`over baseline or over the smallest size observed, or the appearance of
`new lesions. Worsening of existing lesions or the appearance of new
`lesions on bone scan was considered PD. Objective responses were
`confirmed by an independent group of radiologists who reviewed the
`CT scans of responding patients in a blinded fashion. Overall survival
`
`2523
`
`was measured from the time of nephrectomy to the time of death or to
`the last follow-up assessment. Patients who were alive at the time of the
`last follow-up or who were lost to follow-up were censored.
`
`Statistical Methods
`
`The primary efficacy variable was the proportion of patients respond-
`ing to treatment (CR 1 PR). Secondary efficacy variables were time to
`disease progression, durability of response, and survival. An evaluation
`of the relationship between the patient’s clinical response and the
`phenotype, cytokine profile, and cytotoxic activity of the cultured TILs
`was planned.
`The primary efficacy analysis was based on the intent-to-treat (ITT)
`population. The sample size of 166 patients was estimated based on the
`results of the pilot phase I/II trial. Using the 90% confidence interval for
`response from that trial, it was assumed that RCC patients with an
`ECOG PS of 0 or 1 who received treatment with rIL-2 would have a
`complete plus partial response rate of approximately 15%, and that
`patients treated with rIL-2 plus CD81 TILs would have a complete plus
`partial response rate of approximately 36%. This trial was powered to
`detect a difference of this magnitude.
`Logistic regression analysis was used to evaluate the proportion of
`patients responding to treatment. Chi-square statistics (alpha 5 0.048)
`were used to evaluate the hypothesis of conditional independence of the
`treatment group and response, controlling for ECOG PS. Fisher’s exact
`test was used to compare the proportion of responders between the two
`treatment groups. To evaluate the effect of rIL-2 plus CD81 TIL therapy
`on survival, a Cox proportional hazards model was used with treatment
`group and ECOG PS as variables. If survival and ECOG PS were not
`statistically significant variables, survival curves were estimated by the
`Kaplan-Meier method.33
`
`RESULTS
`Patient Characteristics and Disposition
`
`A total of 178 patients presenting with MRCC were
`enrolled. After radical nephrectomy, 160 patients were
`randomized (81 to the TIL/rIL-2 group, 79 to the rIL-2
`control group). Eighteen patients (10%) were not random-
`ized because of pathology other than MRCC (transitional
`cell carcinoma, n 5 5;
`leiomyosarcoma, n 5 3; adrenal
`cortical carcinoma, n 5 1; collecting duct carcinoma, n 5 2;
`neuroectodermal tumor, n 5 1), insufficient tissue available
`for resection (n 5 5), or unresectable tumor (n 5 1) (Fig 1).
`Twenty patients (12.5%) were randomized but did not
`receive rIL-2 treatment because of either surgical complica-
`tions (n 5 4), operative mortality (n 5 2), or failure to meet
`eligibility criteria for rIL-2 therapy after nephrectomy
`(n 5 14). Therefore, 72 patients (88.9%) in the TIL/rIL-2
`group and 68 patients (86.1%) in the rIL-2 control group
`received the first cycle of rIL-2 therapy. Only 30 patients
`(37%) in the TIL/rIL-2 group and 28 patients (35%) in the
`rIL-2 control group were eligible for a second cycle of
`therapy. In the majority of cases, treatment was discontinued
`after the first cycle due to PD; seven patients were removed
`from study at the end of cycle 1 for reasons other than PD,
`including intercurrent illness (n 5 1), voluntary withdrawal
`without adverse events (n 5 3), unacceptable toxicities
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216.185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 3 of 9
`
`

`

`2524
`
`FIGLIN ET AL
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Fig 1. Number of patients en-
`rolled, receiving nephrectomy, ran-
`domized, and receiving treatment.
`Dotted arrows (- - - c) indicate pa-
`tients excluded from treatment. Ab-
`breviations: Nx, nephrectomy; EU,
`Europe; Rx, therapy.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`(n 5 2), and unblinding as a result of mistaken diagnosis of
`PD (n 5 1). Of 81 patients randomized to the TIL/rIL-2
`group, 33 patients (41%) did not receive CD81 TILs due to
`factors related to cell processing,
`including inadequate
`numbers of CD81 TILs or poor cell viability.
`The characteristics of randomized patients in both treat-
`ment groups were comparable for age (mean, 55 to 56
`years), ECOG PS, time from diagnosis to surgery, postopera-
`tive tumor staging, renal vein involvement, and sites of
`metastases (Table 1). There was, however, a greater propor-
`tion of females in the rIL-2 control group (32.9%) compared
`with the TIL/rIL-2 group (13.6%). In the TIL/rIL-2 group
`and the rIL-2 control group, renal vein involvement was
`observed in 25.9% and 29.1% of patients, inferior vena caval
`extension in 11.1% and 21.5%, lymph node involvement in
`39.5% and 36.7%, and multiple organ metastases in 44.4%
`and 57% of patients, respectively. All patients underwent
`radical nephrectomy. Bone metastases were not identified as
`a separate risk category.
`
`Response to Treatment: ITT Analysis
`
`The overall response rate by treatment group and ECOG
`PS is summarized in Table 2. Of eight responders in the
`TIL/rIL-2 group, three patients (7.9%) had an ECOG PS of 0
`and five patients (11.6%) had an ECOG PS of 1, for an
`overall response rate of 9.9%. Of nine responders in the
`rIL-2 control group, five patients (14.3%) had an ECOG PS
`of 0 and four patients (9.1%) had an ECOG PS of 1, for an
`overall response rate of 11.4%. Using a logistic regression
`model,
`the difference in overall response rate was not
`statistically significant between the treatment groups
`(P 5 .753), and ECOG PS was also not predictive of
`response (P 5 .894). The odds ratios were 0.851 for treat-
`ment group and 1.07 for ECOG PS, indicating a similar
`likelihood of response regardless of TIL treatment or ECOG
`PS. However, only 39 (48%) of 81 patients in the ITT
`population who were randomized to the TIL/rIL-2 group
`actually received TIL therapy. Because of the lack of
`efficacy, as determined by the Data Safety Monitoring Board
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216 185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 4 of 9
`
`

`

`CD81 TILs PLUS rIL-2 IN METASTATIC RCC
`
`Table 1. Patient Demographic and Clinical Characteristics
`
`Characteristic
`
`Age, years
`Mean
`Range
`Sex, %
`Male
`Female
`ECOG performance status, %
`0
`1
`Time from Dx to surgery, %
`, 30 days
`30-60 days
`60-90 days
`. 90 days
`Postoperative tumor staging
`Extent of primary tumor, %
`T2
`T3a
`T4
`Renal vein involvement, %
`IVC extension, %
`Lymph node involvement, %
`Tumor completely resected, %
`Sites of metastases, %
`Lung only
`Single organ (not lung)
`Multiple organs
`
`rIL-2 1 TIL
`(n 5 81)
`
`56
`20-77
`
`rIL-2 Control
`(n 5 79)
`
`55
`16-85
`
`86.4
`13.6
`
`46.9
`53.1
`
`32.1
`37.0
`19.8
`9.9
`
`19.8
`46.9
`30.9
`25.9
`11.1
`39.5
`92.6
`
`42.0
`13.6
`44.4
`
`67.1
`32.9
`
`44.3
`55.7
`
`38.0
`31.7
`22.8
`7.6
`
`24.1
`53.2
`21.5
`29.1
`21.5
`36.7
`98.7
`
`34.2
`8.9
`57.0
`
`Abbreviations: Dx, diagnosis; IVC, inferior vena cava.
`
`using the data from 80 patients, the study was terminated
`early. As such, no data are available concerning response
`durations or the number of complete responses.
`
`Survival: ITT Analysis
`
`The 1-year overall survival rates were similar in the
`TIL/rIL-2 group (55%) and the rIL-2 control group (47%)
`(P 5 .551). Median survival was 12.8 months in the TIL/
`rIL-2 group, with 38 patients (46%) censored, versus 11.5
`months in the rIL-2 control group, with 35 patients (44%)
`censored (Fig 2). The ECOG PS was not predictive of
`improved overall survival (P 5 .121). Twenty-nine patients
`
`Table 2. Response (CR 1 PR) by Treatment Group and ECOG PS
`(ITT analysis)*
`
`rIL-2 1 TIL
`(n 5 81)
`
`rIL-2 Control
`(n 5 79)
`
`No. of
`Patients
`
`3/38
`5/43
`8/81
`
`%
`
`7.9
`11.6
`9.9
`
`No. of
`Patients
`
`5/35
`4/44
`9/79
`
`%
`
`14.3
`9.1
`11.4
`
`ECOG PS
`
`0
`1
`Total
`
`*P5 .753, ITT analysis.
`
`2525
`
`in the rIL-2 control group with PD were eligible for
`crossover to the TIL/rIL-2 group; however, because of the
`early study termination, these data were not collected.
`
`TIL Characteristics
`
`The characteristics of the infused TILs are summarized in
`Table 3. The quality of the cell cultures with respect to the
`proportion of CD81 cells varied from 5% to 99% (mean,
`84.8% 6 23.3%); however, the majority of cultures were
`highly enriched for CD31CD81 cells. In general, the TIL
`cell preparations for infusion showed great variability in
`terms of cell numbers and phenotypic characteristics. Unfor-
`tunately, no functional characterization of the TIL cultures
`was performed. The relationship between the characteristics
`of the TIL cultures and patient clinical outcomes was also
`not analyzed because of the large number of cell culture
`failures, which led to reduction in sample size and number of
`responders.
`
`Safety
`
`Postoperative complications, including small bowel ob-
`struction (n 5 1), cardiac arrest (n 5 1), liver failure (n 5 1),
`and cerebrovascular accident (n 5 1), were responsible for
`the exclusion of four patients from treatment with rIL-2.
`Operative mortality, defined as death within 30 days of
`surgery, from any cause, occurred in two patients (1.25%).
`PD associated with clinical deterioration accounted for 14
`patients (7.9%) being excluded from therapy. Adverse
`events occurring in $ 50% of patients in the ITT analysis
`were PD, asthenia, fever, pain, and nausea, with most events
`designated as not serious. The most common serious adverse
`events by body system were related to the body as a whole
`and included PD, asthenia, carcinoma, fever, pain, and
`sepsis (Table 4). Serious hypotension and cardiac side
`effects occurred in 0% to 6% of treated patients. The number
`of serious adverse events was comparable between treatment
`groups. There were no side effects specifically associated
`with TIL therapy.
`The total number of deaths in each of the groups was
`identical (n 5 43). PD accounted for the greatest number of
`deaths in both groups, resulting in 35 of 43 deaths in the
`TIL/rIL-2 group and 43 of 43 deaths in the rIL-2 control
`group. The additional eight deaths in the TIL/rIL-2 group
`were due to cardiac and cardiopulmonary arrests, respiratory
`failure and arrest, pulmonary embolism, and unknown
`cause. One death in the rIL-2 control group was due to
`multiple health problems caused by PD. These deaths were
`designated by the investigator as not directly related to
`treatment.
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216.185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 5 of 9
`
`

`

`2526
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`FIGLIN ET AL
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`for all
`Fig 2. Overall survival
`patients by treatment group, based
`on the ITT analysis.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`DISCUSSION
`Low-dose rIL-2 regimens continue to be of interest in the
`treatment of MRCC for their potential to yield meaningful
`responses with lower toxicity; however, low-dose regimens
`have not yet been demonstrated to induce durable CRs as
`effectively as high-dose rIL-2 therapy. In the current study,
`we treated patients with nephrectomy followed by low-dose
`CIV rIL-2 with and without CD81 TILs. Selection of CD81
`TILs was intended to enhance the proportion of cytotoxic
`cells capable of recognizing major histocompatibility com-
`plex class I-restricted target antigens. In vitro studies have
`demonstrated that TILs cultured for 3 weeks in the presence
`of rIL-2 are primarily CD31CD81 cells and exhibit maxi-
`the proportion of CD81 cells
`mum cytotoxic activity;
`
`Table 3. Characteristics of CD81 TIL Infusion Product
`
`Variable
`
`No. of
`Observations
`
`Total no. of cells, 3 106
`Cell viability, %
`CD31, %
`CD81, %
`CD41, %
`CD561, %
`CD31/CD81, %
`CD81/CD561, %
`
`38
`38
`39
`39
`39
`39
`39
`39
`
`Mean
`
`SD
`
`Minimum Maximum
`
`10,118
`91.0
`91.8
`84.8
`9.0
`30.1
`76.6
`22.3
`
`9,220
`7.1
`17.4
`23.3
`17.8
`22.0
`32.8
`19.6
`
`60
`69.0
`20.8
`5.5
`0.0
`6.7
`0.0
`0.0
`
`49,300
`98.0
`99.1
`99.2
`94.1
`88.9
`99.4
`81.5
`
`thereafter declines with a concomitant decrease in their
`cytolytic potential.26
`In the primary ITT analysis, the overall response rate for
`both groups combined was 10.6%, and the 1-year survival
`rate was 55% in the TIL/rIL-2 group and 47% in the rIL-2
`control group. This multicenter study did not confirm the
`treatment benefit associated with CD81 TILs plus rIL-2
`compared with rIL-2 therapy alone that was observed in the
`pilot study.32 Because the response rates in the two groups
`were comparable, this could be attributed in part to the large
`proportion of patients (41%) in the TIL/rIL-2 group who did
`not receive CD81 TIL therapy because their cell cultures
`failed to yield sufficient numbers of viable cells. In contrast,
`23 (96%) of 24 patients in the pilot study were treated with
`CD81 TILs. The technical limitations related to CD81 TIL
`processing also resulted in great variability in phenotypic
`characteristics and cell numbers between TIL preparations
`(Table 3).
`The outcome of this trial underscores the challenges
`associated with the application of TILs in the broader
`clinical setting. Although some single-institution studies
`have demonstrated objective tumor regression associated
`with TIL plus rIL-2 therapy, results have been variable,
`possibly because of disparity with respect to rIL-2 adminis-
`tration, TIL preparation, or patient selection.24,32,34-37 Further
`randomized studies are necessary to evaluate the clinical
`
`Downloaded from ascopubs.org by Reprints Desk on November 18, 2016 from 216.185.156.028
`
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`NOVARTIS EXHIBIT 2035
`Breckenridge v. Novartis, IPR 2017-01592
`Page 6 of 9
`
`

`

`CD81 TILs PLUS rIL-2 IN METASTATIC RCC
`
`Table 4. Incidence of Serious Adverse Events (grade 3/4) by Body System
`Occurring in More Than 5% of the ITT Population (n 5 160)
`
`Adverse Event
`
`Whole body
`Asthenia
`Fever
`Pain
`Unassessable reaction
`Sepsis
`Cardiovascular
`Embolus
`Digestive
`Anorexia
`Nausea
`Vomiting
`Ileus
`Metabolic/nutritional
`Dehydration
`Hypercalcemia
`Respiratory
`Apnea
`Dyspnea
`Pleural effusion
`Pneumonia
`
`rIL-2 1 TIL
`(%)
`
`rIL-2 Control
`(%)
`
`2
`3
`14
`8
`5
`
`6
`
`2
`3
`6
`3
`
`3
`8
`
`6
`13
`5
`8
`
`14
`8
`8
`5
`10
`
`3
`
`6
`10
`10
`10
`
`6
`6
`
`2
`6
`6
`8
`
`benefit of TIL therapy in combination with standard immu-
`notherapeutic protocols. Such studies must clearly establish
`the technology and delineate its technical challenges, so that
`the technology does not become a dependent variable. In
`practical terms, special care must be applied to standardizing
`the techniques used for tumor handling and transport, tumor
`processing to yield single-cell suspensions, and CD81 TIL
`expansion, selection, and harvesting.
`The substantial number of cell culture failures and lack of
`efficacy based on the data from 80 patients resulted in early
`termination of the trial. Because of the small number of
`responders, the relationship between the phenotypic and
`cytotoxic profile of TIL cultures and the clinical response
`could not be analyzed. Belldegrun et al28 designed a study to
`identify predictors of response to TIL/rIL-2 therapy. Al-
`though no significant differences between responders and
`nonresponders were found with respect
`to the in vitro
`characteristics of tumor TILs (including phenotype), their
`study did suggest that clinical response to TIL/rIL-2 therapy
`may be associated with patients’ natural immune status at
`baseline.28 Many investigators are currently studying differ-
`ent aspects of CD81 TIL function, which should eventually
`advance the clinical application of TIL therapy.28,38-40
`The results of the current study further underscore the
`safety of nephrectomy before systemic rIL-2 therapy. More-
`over, the high incidence of renal vein involvement, inferior
`vena caval extension, and multiple organ metastasis in a
`
`2527
`
`large proportion of patients enrolled on the current study did
`not preclude them from undergoing successful nephrectomy.
`The surgical complication rate was only 5% and, overall,
`only 12.5% of patients were ineligible for rIL-2 therapy after
`nephre