`
`Clinical Cancer Research
`
`4697
`
`A Phase II Study of Razoxane, an Antiangiogenic Topoisomerase II
`Inhibitor, in Renal Cell Cancer with Assessment of Potential
`Surrogate Markers of Angiogenesis1
`
`Jeremy P. Braybrooke, Kenneth J. O’Byrne,
`David J. Propper, Andrew Blann,
`Mark Saunders, Nicola Dobbs, Cheng Han,
`Jane Woodhull, Karen Mitchell, Jeremy Crew,
`Kenneth Smith, Ross Stephens,
`Trivadi S. Ganesan, Denis C. Talbot, and
`Adrian L. Harris2
`Imperial Cancer Research Fund Medical Oncology Unit, and
`Department of Urology, Churchill Hospital, Oxford OX3 7LJ, United
`Kingdom [J. P. B., K. J. O., D. J. P., M. S., N. D., C. H., J. W., K. M.,
`J. C., K. S., T. S. G., D. C. T., A. L. H.]; Haemostasis, Thrombosis
`and Vascular Biology Laboratory, University Department of
`Medicine, City Hospital, Birmingham, B18 7QH, United Kingdom
`[A. B.]; and Finsen Laboratory, Righospitalet, DK-2100, Copenhagen,
`Denmark [R. S.]
`
`ABSTRACT
`Renal cell carcinoma (RCC) is an angiogenic tumor
`resistant to standard cytotoxic chemotherapeutic agents. Al-
`though often responsive to immunomodulatory agents in-
`cluding interleukin 2 and IFN-a, the overall results in ran-
`domized Phase III studies are disappointing with only
`modest improvements in overall survival. This Phase II
`study evaluated the efficacy and tolerability of razoxane, an
`antiangiogenic topoisomerase II inhibitor, in 40 patients (32
`men, 8 women; age: range, 31–76 years; median, 58 years)
`with inoperable RCC. Twenty patients received razoxane
`125 mg p.o., twice a day for 5 days each week for 8 weeks
`(one cycle). This was repeated in patients with stable disease
`(StD), but was discontinued after 16 weeks if there was no
`evidence of an objective response. Because minimal toxicity
`was seen, subsequent patients (n 5 20) were treated until
`progressive disease (PD) was documented. Of 38 evaluable
`patients, 11 (29%) had StD for a minimum of 4 months, and
`the remainder had PD. Median overall survival was 7.3
`months. Duration of survival was significantly better in
`patients with StD compared with those with PD (P 5 0.003).
`
`Received 6/27/00; revised 9/28/00; accepted 10/3/00.
`The costs of publication of this article were defrayed in part by the
`payment of page charges. This article must therefore be hereby marked
`advertisement in accordance with 18 U.S.C. Section 1734 solely to
`indicate this fact.
`1 This work was supported by the Imperial Cancer Research Fund
`(ICRF).
`2 To whom requests for reprints should be addressed, at ICRF Medical
`Oncology Unit and Molecular Oncology Laboratory, Institute of Mo-
`lecular Medicine, John Radcliffe Hospital, Oxford, OX3 9DS, United
`Kingdom. Phone: 44-1865-222457; Fax: 44-1865-222431; E-mail:
`aharris.lab@icrf.icnet.uk.
`
`The effect of treatment on six potential surrogate serum/
`plasma (vascular endothelial growth factor (VEGF), basic
`fibroblast growth factor (bFGF), urokinase plasminogen
`activator soluble receptor (uPAsr), E-selectin, vascular cell
`adhesion molecule-1 (VCAM-1) and von Willebrand’s factor
`(vWF) and two urinary (VEGF and bFGF) markers of
`angiogenesis was evaluated before and after 1 cycle of treat-
`ment. Pretreatment serum VEGF and E-selectin levels
`above the median value were associated with a poor prog-
`nosis. Serum VCAM-1 levels and urinary VEGF levels rose
`significantly after one cycle in patients with PD but not in
`those with StD. Serum VEGF, bFGF, VCAM-1 and vWF,
`plasma uPAsr and urinary bFGF levels were significantly
`higher in PD patients compared with StD patients before
`and/or after 1 cycle of treatment. In conclusion, razoxane is
`an antiangiogenic agent that has minimal toxicity and that
`requires further evaluation in combination with other active
`agents in the treatment of RCC. Surrogate serum and uri-
`nary markers of angiogenesis may have a role to play in
`predicting disease response and overall survival in RCC.
`
`INTRODUCTION
`Approximately 25% of patients with RCC3 present with met-
`astatic disease. Although the subsequent progress of the disease is
`highly variable, the overall prognosis is poor with a 5-year survival
`in reported studies of 0–18%. RCC is highly resistant to conven-
`tional cytotoxic chemotherapy. An extensive review of 39 drugs
`used to treat RCC revealed all to have response rates #9%. Ra-
`diotherapy may be used in the palliation of symptoms but likewise
`is associated with a poor response rate (1–6).
`Razoxane (ICRF 159; (6)1,2-di(3,5-dioxopiperazin-1-yl)
`propane) belongs to the family of bis-dioxopiperazines, devel-
`oped in the 1960s as derivatives of the chelating agent EDTA.
`Early in its development it was shown to have potent antian-
`giogenic activity, with vessels at the periphery of tumors revert-
`ing from an abnormal tumor-related vasculature to a normal
`phenotype (7–10). The precise molecular mechanism of this
`antiangiogenic action is not known. More recently razoxane has
`been shown to be a noncleavable inhibitor of topoisomerase II
`(11). Razoxane inhibits metastatic spread of Lewis lung 3LL,
`hamster lymphoma ML, and murine squamous carcinoma G
`
`3 The abbreviations used are: RCC, renal cell cancer; VEGF, vascular
`endothelial growth factor; bFGF, basic fibroblast growth factor; uPAsr,
`urokinase plasminogen activator soluble receptor; VCAM-1, vascular
`cell adhesion molecule-1; vWF, von Willebrand’s factor (antigen); PD,
`progressive disease; StD, stable disease; TTP, time to tumor progres-
`sion; b.d., twice a day; IL, interleukin; ASP, aspartate aminotransferase;
`ALT, alanine aminotransferase; CTC, common toxicity criteria; CR,
`complete response; PR, partial response.
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 1 of 8
`
`
`
`4698 Razoxane in RCC
`
`cells in experimental animals and causes a marked increase in
`the sensitivity of tumors to radiation (reviewed in Refs. 12 and
`13). Initial Phase I and II clinical trials demonstrated modest
`antitumor activity in advanced colorectal and head and neck
`carcinomas,
`lymphomas,
`lymphosarcomas, and acute leuke-
`mias. In combination with radiotherapy, razoxane has activity
`against liver metastases from colorectal, inoperable nonmeta-
`static rectal, bladder, vulval, and lung carcinomas; soft tissue
`and osteosarcomas; and central nervous system tumors includ-
`ing malignant glioma and astrocytoma (reviewed in Refs. 12
`and 13).
`Razoxane is absorbed from the gastrointestinal tract in a
`schedule-dependent manner. Absorption is poor with large sin-
`gle doses but is satisfactory with small divided doses (14). The
`plasma half-life in humans is 3.5 h. It is well tolerated with few
`side effects seen in early clinical trials. However, with high-dose
`razoxane, neutropenia and thrombocytopenia are seen (15) be-
`cause of the topoisomerase activity and inhibition of cell pro-
`liferation. Hematological toxicity was rarely seen with pro-
`longed low-dose schedules and razoxane was well tolerated
`when evaluated in the adjuvant treatment of colorectal carci-
`noma at a dose of 125 mg b.d. p.o. for 5 days every week given
`for up to 2 years, (16, 17). As such, oral razoxane is an
`inexpensive antiangiogenic and antiproliferative agent with few
`side effects.
`Angiogenesis, the process of new microvessel formation, is
`necessary for a tumor to grow beyond 1–2 mm in diameter.
`Neovascularization also plays an important role in the metastatic
`spread of malignant disease (18). New vessel formation, with
`accentuation of capsular vessels, is a feature of RCC (19). High
`microvessel counts in RCC are associated with a poor prognosis
`(20, 21). IFN-a and IL-2 are antiangiogenic. This observation
`may in part explain their antitumor activity in RCC (22, 23).
`Therefore, investigation of inexpensive, established agents with
`antiangiogenic properties requires evaluation in the manage-
`ment of RCC. Angiogenic growth factors and their receptors,
`proteases and protease receptors, and endothelial cell adhesion
`molecules play important roles in the pathogenesis of malignant
`tumors. These factors include VEGF (24), bFGF (25, 26), uPAsr
`(27), E-selectin, VCAM-1 (28), and vWF (29). Elevated levels
`of these proteins may be detected in the blood and urine of
`patients with malignant disease (28 –33).
`The purpose of this study was to evaluate the efficacy of
`prolonged low-dose oral razoxane in the treatment of metastatic
`RCC. Antiangiogenic agents may not produce regression of
`established blood vessels; therefore, we considered StD to be a
`relevant end point. Treatment with antiangiogenic drugs may be
`required over a protracted period of time; therefore, we wanted
`to assess whether it was possible to detect antiangiogenic effects
`early in the course of treatment. We hypothesized that potential
`surrogate plasma/serum and urine markers of angiogenesis
`would be influenced by therapy and would predict subsequent
`disease status.
`
`PATIENTS AND METHODS
`Eligibility Criteria—Single-Center Open-Label Phase
`II Study. Patient
`inclusion criteria included: cytologically
`and/or histologically proven metastatic RCC; age, $18 years;
`
`WHO performance status, #2; expected survival, $3 months;
`hemoglobin, $10 g/dl; white blood count, $3 3 109/liter;
`absolute neutrophil count, $2 3 109/liter; platelet count,
`$100 3 109/liter; bilirubin, #2 3 normal; AST/ALT, #3 3
`normal (unless because of metastases, in which AST/ALT #5 3
`normal is accepted); and creatinine within the normal range for
`our institution (70 –150 mmol/liter, Oxford Radcliffe Hospital).
`Patients with previous or intercurrent malignancies at other
`sites, with the exception of adequately treated cone-biopsied
`carcinoma of the cervix and basal or squamous cell carcinoma
`of the skin were excluded from the study. Other exclusion
`criteria were intensive chemotherapy or radiotherapy ,3 weeks
`before inclusion, pregnancy, or women likely to become preg-
`nant during the trial. Patients of childbearing age had to take
`adequate contraceptive precautions during the trial and for 4
`weeks after completing treatment. The study was approved by
`the Central Oxford Research Ethics committee and conducted
`according to the recommendations of the Declaration of Hel-
`sinki and the Association of British Pharmaceutical Industry
`guidelines for good clinical practice. Informed written consent
`was obtained from each patient prior to entry into the study.
`Patient Evaluation.
`In the 3 weeks prior to commencing
`treatment, each patient had the following assessments: physical
`examination with clinical evaluation of all of the sites of dis-
`ease; a full blood count and renal, liver, and bone biochemistry
`screen; and establishment of a measurable lesion(s) using addi-
`tional investigations as clinically indicated (plain X-rays, ultra-
`sound, computed tomographic imaging, magnetic resonance im-
`aging). On the day of prescription of razoxane, the patient had
`a further clinical evaluation. A full blood count was checked
`every 2 weeks while the patient was on study. Each patient was
`reexamined every 4 weeks and at each time had a biochemistry
`profile.
`Treatment Schedule.
`In the first 20 patients recruited to
`the study, 125 mg razoxane was administered b.d. p.o. for 5 days
`a week for a cycle duration of 8 weeks. After formal disease
`assessment, the cycle was repeated in patients with either StD or
`responding disease and discontinued after 16 weeks (two cycles)
`if there was no evidence of objective disease response. Because
`the agent was well tolerated in these patients, subsequent pa-
`tients were treated until PD was documented.
`Toxicities and Dose Modifications. Toxicity was graded
`according to the Cancer and Leukemia Group B expanded CTC.
`Treatment delay of 1 week’s duration was considered for any $
`grade 2 toxicity apart from anemia, alopecia, and nausea and
`vomiting controlled with antiemetics. If the grade 2 toxicity per-
`sisted for .1 week or if . grade 3 neutropenia occurred, the
`treatment was reduced by 30% with razoxane 125 mg once a day
`on days 1, 3, and 5 and 125 mg b.d. on days 2 and 4 of the 5-day
`schedule. Razoxane treatment was discontinued if a further dose
`reduction was indicated, if grade 4 toxicity (CTC scale) was expe-
`rienced at the reduced dose level, or if a life-threatening event
`occurred that was deemed directly related to razoxane therapy.
`Response Assessment. The assessable lesion/lesions
`were clearly measurable according to CRC recommendations. A
`lesion was considered measurable if it measured $1 cm in
`diameter on computed tomographic or magnetic resonance im-
`aging, or at least 2 cm on plain X-ray or ultrasound, in each of
`two perpendicular dimensions. The established measurable le-
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 2 of 8
`
`
`
`sion(s) were evaluated after each 8-week cycle to assess re-
`sponse. A CR was defined as clearance of all of the measurable
`or evaluable disease for $8 weeks. A PR was defined as a
`reduction in the sum of the product of all of the bidimensional
`measurements of the lesions evaluated at baseline by $50% for
`$8 weeks. Disease was considered stable if the lesions meas-
`ured at baseline reduced in size by ,50% or increased by ,25%
`after two cycles of treatment. PD was defined as an increase of
`$25% in the sum of the product of the bidimensionally meas-
`urable disease sites, or the appearance of new metastases ,16
`weeks (two cycles) after commencing razoxane.
`Surrogate Blood and Urine Markers of Angiogenesis.
`Prior to treatment, two 10-ml blood samples for serum and
`plasma, and a urine sample, were obtained to analyze potential
`surrogate markers of angiogenesis. The blood sample was
`chilled on ice, centrifuged at 2000 rpm for 10 min at 4°C and
`stored in 1.5-ml aliquots at 270°C until analysis. The urine
`samples were either brought in by the patient or taken fresh on
`the morning of treatment and were frozen at 270°C until
`analysis. Additional samples were obtained after 4 and 8 weeks
`(1 cycle) of treatment.
`Serum E-selectin, VCAM-1, and serum and urine bFGF
`and VEGF were measured using ELISA kits supplied by R&D
`systems (Abingdon, United Kingdom). The serum vWF and
`plasma uPAsr levels were analyzed by ELISA using in-house
`methods as described previously (31, 34). The normal range (or
`mean 6 SD) for each assay was as previously published or
`determined as indicated: serum VEGF, 62–707 pg/ml (mean,
`220 pg/ml)4; urinary VEGF, 73–144 ng/g creatinine (32); serum
`bFGF, 0 – 4.4 pg/ml4; urinary bFGF, 0.6 – 8.1 ng/g creatinine
`(33); and plasma uPAsr, 0.83–1.7 ng/ml (31). For uPAsr, a
`control population without cancer (n 5 57; range, 1.09 6 0.76)
`were analyzed at the same time as the razoxane patients sam-
`ples. Analysis of VCAM-1, vWF, and E-selectin was made in
`comparison with a historical control population without cancer
`(n 5 186) previously assayed: serum VCAM-1, 525 6 173
`ng/ml; serum E-selectin, 48.37 6 19.6 ng/ml; serum vWF,
`103 6 30 IU/dl. All of the interassay coefficients of variation
`were ,10% and intra-assay coefficients were ,5%.
`Statistical Methods. The Wilcoxon rank-sum test was
`used to analyze differences in surrogate markers before and after
`one cycle of chemotherapy within the StD and PD groups. The
`Mann-Whitney test was used to test for differences in the
`markers between patients with StD and PD. Fisher’s exact test
`was used for testing relationships between categorical variables
`and Student’s t test to compare pretreatment angiogenic marker
`levels with control values as appropriate. The analysis was
`performed using the Stata statistical software, release 5.0 pack-
`age (Stata Corp., College Station, TX).
`
`RESULTS
`Patients. Between August 1994 and February 1997, 40
`patients with inoperable RCC were recruited to the study. Pa-
`tient characteristics are summarized in Table 1. Only five pa-
`
`4 R&D Systems, personal communication.
`
`Clinical Cancer Research
`
`4699
`
`Table 1 Patients’ characteristics
`
`No. of patients
`40
`58 (31–76)
`32/8
`
`Characteristics
`Total number
`Median age, yr (range)
`Male/Female
`ECOGa performance status
`0
`1
`2
`Patients with primary tumor
`Previous treatment
`Nephrectomy
`Prior chemo- or immunotherapy
`Cytotoxic 1 endocrine
`Endocrine only
`Immunotherapy (IL-2)
`Prognostic status
`Good
`Moderate
`Poor
`Sites of evaluable disease
`33
`Pulmonary metastases
`18
`Renal/Adrenal mass
`22
`Lymphadenopathy
`11
`Bony metastases
`8
`Hepatic metastases
`8
`Soft tissue/Skin metastases
`2
`Ascites
`1
`Pancreatic metastases
`1
`Opposite kidney
`1
`Pleural effusion
`40
`Evaluable for toxicity
`38
`Evaluable for response
`35
`Measurable
`3
`Evaluable
`a ECOG, Eastern Collaborative Oncology Group.
`
`24
`16
`0
`16
`
`24
`5
`4
`1
`1
`
`9
`18
`13
`
`tients had received previous treatment, and of these, only one
`had received cytokine therapy using IL-2.
`Response and Survival. Thirty-eight of the 40 patients
`were evaluable for response. Response was classified at 16
`weeks, those progressing before that time being included in the
`PD group. Although no objective tumor responses were seen, 11
`patients had StD for $16 weeks’ duration. All of these patients
`were either newly diagnosed (n 5 4) or had evidence of objec-
`tive tumor progression in the 6 months prior to starting treat-
`ment (n 5 7). The median survival for all of the 40 patients
`included in the study was 7.3 months. In evaluable patients, the
`median survival for those with StD was 399 days and for the 27
`patients with PD, 127 days (P 5 0.0026). The TTP was 12
`weeks.
`Patients were characterized as having a good, moderate, or
`poor prognosis based on the criteria of Jones et al. This
`involves the application of a simple index based on the presence
`or absence of each of the following three risk factors: perform-
`ance status $1 (rather than 0); time from diagnosis ,2 years;
`and more than one site of metastases. Good-prognosis patients
`have #1 risk factor; moderate-prognosis patients, any two risk
`factors; and poor-prognosis patients, all three risk factors. Of the
`40 patients studied, 9 had good prognostic features, 18 had
`moderate, and 13 had poor. Although the good prognostic group
`had a better overall survival, this was not statistically signifi-
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 3 of 8
`
`
`
`4700 Razoxane in RCC
`
`Table 2 Number of patients developing grade 1, 2, 3, and 4 toxicity
`based on Cancer and Leukemia Group B common toxicity criteria
`(total number of patients evaluated 5 40)
`Toxicity (n 5 40)
`1
`17
`Nausea/Vomiting
`7
`Fatigue
`4
`Neutropenia
`7
`Epistaxis
`3
`Taste
`0
`Hypercalcemia
`2
`Mucositis
`1
`Diarrhea
`CNSa
`0
`0
`Allergic reaction
`a CNS, central nervous system.
`
`4
`0
`0
`0
`0
`0
`0
`0
`0
`0
`0
`
`2
`6
`1
`7
`0
`1
`2
`2
`0
`0
`0
`
`3
`1
`1
`3
`0
`0
`0
`0
`1
`1
`1
`
`cant. Of the 38 evaluable patients, no association was found
`between the prognostic groups and the development of PD.
`Three of 9, 5 of 18, and 3 of 11 patients had StD $16 weeks in
`the good, moderate, and poor prognostic groups, respectively.
`Elevated pretreatment neutrophil counts (.8.25 3 109/liter;
`P 5 0.016), lactate dehydrogenase levels (.170 IU/liter; P 5
`0.016), and platelet counts (.450 3 109/liter; P 5 0.015),
`above the median for the patient group, were associated with a
`poor prognosis as reported previously (4, 35). An unresected
`primary tumor (P 5 0.08), the presence of bone metastases (P 5
`1) or extrapulmonary metastases (P 5 1), or a hemoglobin ,10
`g/dl (P 5 0.366) lacked prognostic significance.
`Toxicity and Dose Modifications. Toxicity is summa-
`rized in Table 2. Grade 3 toxicity was rare and included neu-
`tropenia (three patients), nausea and vomiting (one patient), and
`fatigue and diarrhea (one patient each). One patient had a
`delayed hypersensitivity reaction to razoxane characterized by
`the development of an urticarial rash and arthritis, whereas
`another developed cerebellar symptoms after three cycles of
`treatment. In both cases, the symptoms resolved on withdrawal
`of the drug. Treatment was delayed in 10 patients because of
`neutropenia (8 patients), continuing vomiting (1 patient), and
`mucositis (1 patient). The dose was reduced by 30% in four
`patients on one occasion only, because of grade 3 or persistent
`grade 2 neutropenia.
`Surrogate Blood and Urine Markers of Angiogenesis.
`The results for the plasma/serum and urinary surrogate markers
`of angiogenesis are summarized in Fig. 1 and Table 3. Serum
`VEGF levels were analyzed in 35 patients before treatment.
`Elevated levels above the normal range were seen in 12 (34%).
`Pretreatment levels were significantly higher in patients who
`subsequently developed PD compared with those with StD (P 5
`0.042). No change in overall values, either for the whole group
`or for those with StD or PD, was seen during the first cycle of
`treatment. Patients with pretreatment serum levels above the
`median for the group had a worse prognosis (P 5 0.023). A
`significant correlation was seen between pretreatment serum
`VEGF levels and platelet counts (P , 0.0001; correlation co-
`efficient, r, 0.69). Urinary VEGF levels were analyzed in 34
`patients and were elevated in 10 prior to treatment (30%). A
`significant rise in urinary VEGF levels was seen in patients who
`developed PD (P 5 0.006) but not in those with StD (P 5 0.09).
`Serum bFGF was elevated in 15 (43%) of 35 patients
`
`analyzed prior to razoxane. Although there was no significant
`difference between bFGF levels in StD versus PD patients
`pretreatment, a trend toward a reduction in levels in StD patients
`and an increase in levels in PD patients resulted in a significant
`difference in the values between the groups after treatment (P 5
`0.006). Urinary bFGF levels were elevated in only 1 of 34
`patients pretreatment. Pretreatment levels were higher in pa-
`tients who developed PD as compared with those with StD,
`although this difference was not maintained after treatment.
`uPAsr levels were elevated in patients (mean 6 SD, 1.95 6
`1.4 mg/liter) compared with a normal control population (1.09 6
`0.72 mg/liter; P 5 0.0001) with 12 (37%) of 32 having elevated
`levels pretreatment. There was a trend for levels to increase with
`razoxane therapy in the PD group and to fall in the StD group.
`Levels were significantly higher in the PD as compared with the
`StD group before (P 5 0.04) and after (P 5 0.028) 1 cycle of
`treatment.
`Pretreatment serum VCAM-1 levels were significantly
`higher in patients compared with our control population [717 6
`303 versus 525 6 173 ng/ml (mean 6 SD); P 5 0.0001]. Levels
`rose significantly in patients with PD (P 5 0.039) being higher
`than those in patients with StD after treatment (P 5 0.031).
`Although no significant differences were found in E-selectin
`levels between patients and the normal range [52 6 20 versus
`48 6 19 ng/ml (mean 6 SD); P 5 0.22] or in the StD and PD
`groups before or after treatment, E-selectin levels above the
`median were associated with a worse outcome (P 5 0.022).
`Pretreatment vWF levels were elevated in 6 (17%) of 35 pa-
`tients. No significant difference was seen between vWF levels in
`patients and our controls [107 6 53 versus 103 6 30 IU/dl
`(mean 6 SD); P 5 0.112]. Although no significant change in
`levels was seen on treatment, values in patients with PD were
`significantly higher than those with StD, both before (P 5
`0.026) and after (P 5 0.024) one cycle of razoxane.
`
`DISCUSSION
`Targeting of angiogenesis represents a new approach to the
`management of solid tumors, particularly those that are poorly
`responsive to chemotherapy. In this study, we evaluated razox-
`ane, an antiangiogenic topoisomerase II inhibitor (7–17) in the
`treatment of RCC, a highly angiogenic tumor (19) resistant to
`standard cytotoxic regimens (1). We observed disease stabiliza-
`tion in 11 (29%) of 38 patients evaluable for response with a
`median survival of 399 days for this group. All of these patients
`were either newly diagnosed or had documented PD within the
`6 months prior to razoxane. On the basis of known risk-factor
`stratification, there were no significant differences between the
`patients with StD as compared with those with PD, whose
`median survival was 127 days (3). This suggests that in this
`study, razoxane may be inhibiting tumor growth in patients with
`StD but not causing tumor regression. This must be treated with
`caution based on the sample size and the highly variable growth
`rates seen in RCC.
`This raises the issue as to how we should assess the
`effectiveness of antiangiogenic therapy. The drugs being con-
`sidered for this role are, in general, cytostatic rather than cyto-
`toxic to endothelial cells, inhibiting their proliferation and/or
`elongation (36). It seems reasonable to suppose that tumor
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 4 of 8
`
`
`
`Clinical Cancer Research
`
`4701
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Fig. 1 Surrogate blood and urine markers of angiogenesis. Pre- and posttreatment levels of serum (A) and urinary (B) VEGF, serum (C) and urinary
`(D) bFGF, plasma uPAsr (E), serum VCAM-1 (F), serum E-selectin (G) and serum vWF (H), were measured in patients with RCC as described in
`“Patients and Methods. Individual patient values (Y axis) are shown divided into StD and PD groups before and after razoxane (X axis). Statistical
`analyses for the changes in each marker are indicated, and these results are summarized in Table 3.
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 5 of 8
`
`
`
`4702 Razoxane in RCC
`
`Table 3 Summary of statistical analyses for changes in plasma/serum and urinary surrogate markers of angiogenesis compared with
`patient outcome
`
`Pre PD . StD
`Elevated levels
`Prea pts (%)
`(Mann-Whitney, P)
`Marker
`0.042b
`12/35 (34%)
`Serum VEGF
`0.233
`10/34 (30%)
`Urinary VEGF
`0.814
`15/35 (43%)
`Serum bFGF
`0.035b
`1/34 (3%)
`Urinary bFGF
`0.040b
`12/32 (37%)
`Plasma uPAsr
`0.170
`7/35 (20%)
`Serum VCAM-1
`0.729
`1/19 (5%)
`Serum E-selectin
`0.026b
`6/35 (17%)
`Serum vWF
`a Pre, pretreatment; Post, posttreatment; pts, patients.
`b Significant difference.
`c Significant increase.
`
`Post PD . StD
`(Mann-Whitney, P)
`0.088
`0.313
`0.0057b
`0.491
`0.028b
`0.031b
`0.729
`0.024b
`
`StD Pre vs. Post
`(Wilcoxon rank-sum, P)
`0.646
`0.09
`0.114
`0.059
`0.066
`0.906
`0.223
`0.406
`
`PD Pre vs. Post
`(Wilcoxon rank-sum, P)
`0.738
`0.006c
`0.190
`0.717
`0.211
`0.039c
`0.624
`0.948
`
`Survival, $
`median vs.
`, median (P)
`0.023b
`0.790
`0.780
`0.530
`0.288
`0.333
`0.022b
`0.442
`
`stabilization or reduced tumor-growth kinetics, with a resultant
`prolongation of survival, would be realistic end points in terms
`of tumor response evaluation. If StD is considered a valid end
`point, then the response rate with razoxane in the treatment of
`RCC is similar to those seen in recent studies of other potential
`antiangiogenic drugs. Eighteen of 66 patients recruited to a
`Phase II study of low-dose continuous thalidomide had RCC; of
`these, 3 had a PR and another 3 had StD, giving a 33% response
`rate (CR 1 PR 1 StD; Ref. 37). Likewise in a Phase I dose-
`escalating study of IL-12 (38) and a Phase II study of TNP-470
`(39), 8 (29%) of 28 and 7 (21%) of 33 patients responded to
`treatment.
`These results are comparable with those from three ran-
`domized trials with IL-2 and IFN. In the study comparing IL-2,
`IFN-a, and IL-2 plus IFN-a, the event-free survival of all of the
`3 arms was #10 weeks, whereas the response rate (CR 1 PR 1
`StD) in evaluable patients was 33, 40, and 48%, respectively, at
`10 weeks and 12, 19, and 20% at 25 weeks (5). In the trial
`comparing IFN-a and medroxyprogesterone acetate the progres-
`sion-free survival of each group was 4 and 3 months, respec-
`tively; however, the overall response rates are difficult to deter-
`mine (6). Finally, in the study comparing IFN-g with placebo,
`TTP for both arms was 1.9 months, with a best response at 8
`weeks of 37 and 36%, respectively (40). By comparison, 47, 29,
`and 13% of the patients in our study had StD at 8, 16, and 24
`weeks, respectively, whereas TTP was 12 weeks. These findings
`suggest that StD is an important end point for assessment of
`antiangiogenic agents and that, in this Phase II study, razoxane
`had activity that was comparable with more established treat-
`ments. Additional randomized trials with antiangiogenic agents,
`including razoxane, are warranted.
`The evaluation of surrogate serum, plasma, and urinary
`markers of endothelial function or angiogenesis may have a role
`to play in assessing the efficacy of novel antiangiogenic drugs
`(36). Pretreatment serum VEGF and E-selectin levels above the
`median value were directly associated with a worse prognosis,
`as previously reported in other studies (41, 42). A significant
`rise in urinary VEGF levels was seen in patients with PD (Table
`3), but this rise was not significant when compared with the StD
`patients (Fig. 1B). No correlation was seen between serum and
`urinary VEGF levels, or between urinary levels and the presence
`
`of the primary tumor. This is in keeping with previous work in
`bladder cancer (32).
`Plasma uPAsr levels were higher in patients with RCC than
`in a normal control population. uPAsr was significantly elevated
`in patients who developed PD on razoxane compared with those
`with StD both before treatment and after 8 weeks and, therefore,
`is not serially predictive of which patients will respond. Simi-
`larly vWF, urine bFGF, and serum VEGF were significantly
`different pretreatment in patients who developed PD compared
`with those with StD but this did not alter after treatment and,
`thus, cannot be used as a marker of response to razoxane
`therapy.
`However, serum VCAM-1 and serum bFGF showed no
`difference between patient groups before treatment; but, after
`therapy with razoxane, there was a significant difference be-
`tween patients with PD compared with those with StD. This was
`attributable to a rise in VCAM-1 in patients with PD and a small
`fall in patients with StD (Fig. 1F) and to a rise in bFGF in
`patients with PD (Fig. 1C). These results underline those of
`previous studies showing VCAM-1 to be elevated in patients
`with solid tumors, including breast, renal, colorectal, gastric,
`and ovarian cancer and malignant melanoma, and its association
`with disease progression and a poor prognosis (28, 35, 43– 45).
`The significant rise in serum VCAM-1, urinary VEGF, and
`serum bFGF levels seen in patients with PD indicates that these
`angiogenic growth factors should be further evaluated as poten-
`tial surrogate markers of tumor response in RCC and in malig-
`nant diseases in general. Although these results have to be
`interpreted with caution, markers of endothelial cell prolifera-
`tion and protease activation may be useful in suggesting possi-
`ble molecular mechanisms of action for razoxane on endothelial
`cells. This requires further investigation in the preclinical set-
`ting. Further clinical evaluation of the antiangiogenic effect of
`razoxane could be provided by looking at tissue staining for
`surrogate markers and by using noninvasive radiology (e.g.,
`positron emission tomography scans).
`The prognostic significance of the absolute neutrophil
`count, platelet counts, and lactate dehydrogenase levels seen in
`this study is well established for malignant diseases including
`RCC (4, 46). A significant correlation was found between
`pretreatment platelet counts and serum VEGF levels. Platelets
`
`NOVARTIS EXHIBIT 2032
`Breckenridge v. Novartis, IPR 2017-01592
`Page 6 of 8
`
`
`
`Clinical Cancer Research
`
`4703
`
`transport and, after activation, release VEGF. Intratumoral
`platelet activation may result in the release of high local con-
`centrations of the angiogenic growth factor and induce tumor
`angiogenesis. Therefore, serum VEGF may be an important
`marker for VEGF that could be delivered to tumors by platelets
`(47). In our study, elevated serum VEGF was associated with a
`worse prognosis. VEGF can induce the release of vWF from
`endothelial cells (48). We found no correlation between VEGF
`and vWF levels in our patients, or between vWF and platelet
`counts.
`Razoxane, at the dose sched