Clinical Study Synopsis

Transcription

1 Clinical Study Synopsis This Clinical Study Synopsis is provided for patients and healthcare professionals to increase the transparency of Bayer's clinical research. This document is not intended to replace the advice of a healthcare professional and should not be considered as a recommendation. Patients should always seek medical advice before making any decisions on their treatment. Healthcare Professionals should always refer to the specific labelling information approved for the patient's country or region. Data in this document or on the related website should not be considered as prescribing advice. The study listed may include approved and non-approved formulations or treatment regimens. Data may differ from published or presented data and are a reflection of the limited information provided here. The results from a single trial need to be considered in the context of the totality of the available clinical research results for a drug. The results from a single study may not reflect the overall results for a drug. The following information is the property of Bayer AG. Reproduction of all or part of this report is strictly prohibited without prior written permission from Bayer AG. Commercial use of the information is only possible with the written permission of the proprietor and is subject to a license fee. Please note that the General Conditions of Use and the Privacy Statement of bayer.com apply to the contents of this file.

2 Study no Page: 1 of 13 Date of study report: 31 MAY 2016 Study title: Sponsor s study number: NCT number: Sponsor: Clinical phase: Study objectives: A Phase 1 study of BAY (phosphatidylinositol-3 kinase inhibitor) in combination with gemcitabine (Treatment A) or cisplatin plus gemcitabine (Treatment B) in subjects with advanced solid malignancy NCT Bayer I The primary objectives of this study were to: Determine the safety, tolerability, and recommended Phase II dose (RP2D) of copanlisib in combination with gemcitabine (1000 mg/m 2 ) Determine the safety, tolerability, and RP2D of copanlisib in combination with cisplatin (25 mg/m 2 ) and gemcitabine (1000 mg/m 2 ) Characterize the pharmacokinetics (PK) of copanlisib, cisplatin (25 mg/m 2 ), and gemcitabine (1000 mg/m 2 ) when administered concomitantly The secondary objectives of this study were to: Assess the clinical benefit and response to the combination drugs Explore biomarkers of phosphatidylinositol 3 -kinase (PI3K), alpha serine / threonine-protein kinase (AKT), and mammalian target of rapamycin (mtor) pathway and transforming protein p21 (RAS), mitogen-activated protein kinase (MAPK) pathways that may predict response to the combination drugs Test drug (1): Copanlisib (BAY ) Name of active ingredient(s): BAY Dose: Treatment A: - Level 1 (starting dose) : 0.6 mg/kg - Level 2: 0.8 mg/kg Treatment B: Maximum tolerated dose (MTD) of Treatment A

3 Study no Page: 2 of 13 Test drug (2): Name of active ingredient(s): Dose: Test drug (3): Name of active ingredient(s): Dose: Route of administration: Gemcitabine Gemcitabine 1000 mg/m 2 Cisplatine Cisplatine 25 mg/m 2 Intravenous (IV) infusion Duration of treatment: Treatment A: Gemcitabine and Copanlisib were administered at a fixed dose on Days 1, 8 and 15 every 28 days until disease progression or dose limiting toxicity. Treatment B: Cisplatin, gemcitabine and copanlisib (Treatment A MTD) were administered at fixed doses on Days 1 and 8 every 21 days for 24 weeks (8 Cycles). After 24 weeks of treatment, gemcitabine plus copanlisib, without cisplatin, could be continued at the discretion of the Investigator until disease progression or limiting toxicity if a clinical benefit was noted. Indication: Diagnosis and main criteria for inclusion: Advanced solid malignancy Main inclusion criteria: Subjects, at least 18 years of age, with advanced or refractory solid tumors in whom gemcitabine (Treatment A) or cisplatin plus gemcitabine (Treatment B) was appropriate medical therapy as determined by the treating physician Histological or cytological documentation of non-hematologic, malignant solid tumor, excluding primary brain or spinal tumors, with no current involvement in the central nervous system (CNS) At least one measurable lesion or evaluable disease, as per Response Evaluation Criteria in Solid Tumors version 1.1 (RECIST 1.1) Eastern Cooperative Oncology Group (ECOG) Performance Status Assessment of 0 or 1 Main exclusion criteria: Current diagnosis of diabetes Active clinically serious infection > grade 2 Poorly controlled hypertension Use of strong CYP 3A4 inducers and inhibitors

4 Study no Page: 3 of 13 Study design: Methodology: Multi-center, open-label, non-randomized, dose-escalation, single group assignment The study consisted of a screening period (within 14 days before the first IV infusion on cycle1 day1 (C1D1)), a treatment period with repeated 28- day or 21-day treatment cycles starting on C1D1 (with the duration of combined treatment limited by toxicity to the subject), and an End of Treatment Visit ( within 14 days of the last dose) and post-treatment follow-up at least 30 days after the last study drug administration. Subjects were enrolled in Treatment A first. Enrollment into Treatment B began as soon as the MTD in Treatment A was determined. When both Treatment A and B were simultaneously enrolling, subjects were assigned to Treatment B until enrollment in Treatment B was complete. Subjects who are not eligible for Treatment B were assigned to Treatment A. Once enrollment in Treatment B was complete, subjects were enrolled into Treatment A until enrollment was complete. Once the toxicity was determined to be acceptable with Treatment B, enrollment in the biliary tract cancer (BTC) expansion cohort (Cohort BTE) began. Treatment Treatment A: Gemcitabine was administered at a fixed dose on Days 1, 8 and 15 every 28 days. Copanlisib was administered 1 hour after completing gemcitabine infusion on Days 1, 8 and 15 every 28 days. Treatment B: Cisplatin (25 mg/m 2 ), gemcitabine (1000 mg/m 2 ) and copanlisib (Treatment A dose) were administered at fixed doses on Days 1 and 8 every 21 days for 24 weeks (8 Cycles). After 24 weeks of treatment, gemcitabine plus copanlisib, without cisplatin, could have continued at the discretion of the Investigator Efficacy Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) of all anatomic regions involved with the disease was performed within 7 days of the end of every second cycle The RECIST 1.1 criteria were used. Biomarker evaluation Biomarker evaluation (from plasma, tumor, and/or blood) may have included the following: Analysis of tumor mutations (e.g. PIK3CA and KRAS) using deoxyribonucleic acid (DNA) derived from blood, plasma or tumor tissue (if available) Analysis of gene copy number (e.g. PIK3CA and KRAS) in tumor

5 Study no Page: 4 of 13 tissue specimens (if available) Protein quantification (e.g. phosphorylated AKT [pakt] and PTEN) in formalin fixed/paraffin-embedded (FFPE) tumor tissue via immunohistochemistry; or in fresh tumor tissue (if available) Analysis of ribonucleic acid (RNA) (e.g. messenger RNA [mrna] expression in tumor tissue) Quantification of plasma proteins via enzyme-linked immunosorbent assay (ELISA) or multiplex immunoassay Pharmacokinetics Serial PK samples were collected for Treatment A for the Cycle 1 Day 1 dosing: IV infusion of gemcitabine followed by PK assessment; IV infusion of copanlisib followed by PK assessment. Study center(s): Publication(s) based on the study (references): 3 study centers in the United States None at the time of report creation Study period: Study Start Date: 18 NOV 2011 Study Completion Date: 18 DEC 2015 Early termination: Not applicable Number of subjects: Planned: 70 subjects Analyzed: 50 subjects total Criteria of evaluation: Efficacy: Secondary efficacy variable: - Tumor response using RECIST 1.1 criteria Clinical pharmacology: Primary pharmacokinetic parameters : - Cmax and AUC(0-tlast) of copanlisib and its metabolite M-1, gemcitabine and its metabolite dfdu as well as total and free (unbound) platinum - AUC(0-25) of copanlisib and its metabolite M-1

6 Study no Page: 5 of 13 Secondary pharmacokinetic parameters: - tmax and t1/2 of copanlisib and its metabolite M-1, gemcitabine and its metabolite dfdu as well as total and free (unbound) platinum Pharmacodynamic: - PD biomarkers evaluation - PTEN analysis by immunohistochemistry Safety: Primary: - Adverse events (AEs) - Dose limiting toxicities (DLTs) Other: Blood / urine laboratory parameters (hematology, clinical chemistry, urinalysis); ECOG performance status; Vital signs (blood pressure, pulse, respiratory rate, body temperature); Physical examination; Electrocardiogram (ECG); Left ventricular ejection fraction (LVEF) Statistical methods: Substantial protocol changes: Efficacy, safety/ tolerability, PD and PK parameters were summarized using frequency tables for qualitative data and descriptive statistics for quantitative data. All statistical analyses were explorative. A confirmatory statistical analysis was not intended. Medical history and AE data were coded using the Medical Dictionary for Regulatory Activities (MedDRA), Version There were 3 amendments to the original protocol (CSP Version 1.0), dated 18 APR 2011: Protocol Version 2.0, dated 27 JUL 2011, was implemented before any subject had been screened for study participation and reflects changes to the original protocol as requested by the FDA. Protocol Version 3.0, dated 01 MAR 2012, clarified entry criteria for diabetics, added capillary glucose measurements, and clarified DLT criteria for hypertension Protocol Version 4.0, dated 18 DEC 2012, added the BTC expansion cohort

7 Study no Page: 6 of 13 Subject disposition and baseline In this study, 65 adult male or female subjects with advanced solid malignancies were screened at 3 study centers in the US. There were 15 subjects who failed screening and 50 subjects who were enrolled and assigned to treatment in 1 of 4 cohorts. Of the 50 subjects who received treatment, 22 (44.0%) were male and 28 (56.0%) were female. Of the 50 subjects assigned to treatment, 16 subjects were assigned to dose escalation with Treatment A, which was weekly infusions on Days 1, 8, and 15 of a 28-day cycle of 1000 mg/m 2 gemcitabine and the following escalating doses of copanlisib by cohort: Cohort A-1 (n=8): 0.6 mg/kg copanlisib Cohort A-2 (n=8): 0.8 mg/kg copanlisib The remaining 34 subjects were randomized to Treatment B, which used the Treatment A MTD dose level (0.8 mg/kg copanlisib) with a schedule of weekly infusions on Days 1 and 8 of a 21-day cycle of 1000 mg/m 2 gemcitabine plus 25 mg/m 2 cisplatin Cohort B (n=14): Treatment B Cohort BTE (n=20): Treatment B expansion cohort in subjects with biliary tract cancer (BTC) In the overall population, the most common types of cancer were intrahepatic bile duct cancer in 16 (32.0%) subjects and gallbladder cancer in 7 (14.0%) subjects, primarily due to subjects in the BTC expansion cohort, followed by breast cancer and pancreatic adenocarcinoma in 5 (10.0%) subjects each, NSCLC in 4 (8.0%) subjects, and esophageal carcinoma and peritoneal carcinoma in 2 (4.0%) subjects each. Discontinuation of treatment in the 50 treated subjects was primarily due to disease progression [26 (52.0%) subjects], followed by occurrence of an AE [16 (32.0%) subjects]. Further primary reasons for discontinuation included withdrawal of consent by subject [6 (12.0%) subjects], death [1 (2.0%) subject], and physician decision [1 subject (2.0%)]. All 50 subjects assigned to treatment were valid for the analysis of safety (SAF: N=50) and efficacy (ITT: N=50), and 47 subjects were valid for the analysis of pharmacokinetics (PKS: N=64). Efficacy / clinical pharmacology evaluation The best overall responses in all 50 subjects assigned to treatment who achieved at least stable disease were complete response (CR) in 1 (2.0%) subject, partial response (PR) in 4 (8.0%) subjects, and stable disease in 19 (38.0%) subjects. Objective response (CR or PR) was achieved by 5 (10.0%) subjects overall. Of these 5 responders, one belonged to Cohort A and 4 to Cohort B. The response rate in Cohort A was 6.3% (1/16 subjects). The 1 responder (PR) was a subject diagnosed with primary peritoneal carcinoma in Cohort A-1. The response rate in Cohort B was 11.8% (4/34 subjects). The 4 responders were subjects diagnosed with gallbladder cancer (1 subject with a CR) or intrahepatic biliary cancer (3 subjects with a PR).

8 Study no Page: 7 of 13 Two of the subjects who achieved PR were in Cohort BTE, which had a response rate of 10.0% (2/20 subjects). Biomarkers analysis PTEN by immunohistochemistry No conclusion could be drawn about baseline tumor PTEN IHC in relation to response due to the small number of subjects in BTC and other tumor types. Additional biomarker analyses were performed and are reported in a separate biomarker report. Pharmacokinetic evaluation Forty-seven subjects were valid for PK analysis [7 (14.9%) subjects in Cohort A-1, 7 (14.9%) subjects in Cohort A-2, 13 (27.7%) subjects in Cohort B, and 20 (42.6%) subjects in the biliary tract expansion cohort]. Three subjects were excluded from the PKS because of missing PK samples and/or incomplete dosing. Pharmacokinetic parameters of primary interest in this study were Cmax and AUC(0-tlast) of copanlisib and its metabolite M-1, gemcitabine and its metabolite dfdu as well as total and free (unbound) platinum. Additionally, AUC(0-25) of copanlisib was determined for comparison of results with previous studies. PK parameters were derived from plasma samples collected after the first dose of copanlisib, gemcitabine and cisplatin on C1D1. Copanlisib and metabolite M-1 Selected PK parameters for copanlisib and its metabolite M-1 are shown in Table 0-1 and Table 0-2, respectively. After an IV infusion of copanlisib on C1D1 over 1 hour at doses of 0.6 and 0.8 mg/kg in combination with 1000 mg/m² gemcitabine (Cohorts A-1 and A-2) or in combination with 1000 mg/m² gemcitabine and 25 mg/m² cisplatin (Cohort B and expansion cohort), the geometric mean Cmax of copanlisib in plasma ranged between 249 and 414 µg/l and were reached at a median tmax of about 1.0 hour which corresponds to the end of the infusion. After reaching Cmax, the plasma concentration vs. time profiles of copanlisib declined in a multiphasic manner. Geometric mean AUC(0-tlast) for copanlisib were between 1563 and 2617 µg*h/l over the different dosing groups and geometric mean AUC(0-25) were between 984 and 1623 µg*h/l. Mean terminal elimination half-life was 25 to 26 hours. Inter-subject variability of copanlisib exposure, as measured by PK parameters Cmax, AUC(0-tlast), and AUC(0-25) was moderate to high with %CV ranging from 33.8% to 87.7%. Dose-normalized data show evidence for dose-proportionality of maximum plasma concentrations and overall exposure of copanlisib following IV dosing of 0.6 to 0.8 mg/kg. For M-1, mean maximum plasma concentrations between 5.24 and 9.31 µg/l were observed at a median tmax of 2.0 to 4.0 hours (range: 1.0 to 49 hours) after start of copanlisib infusion. Plasma concentrations of M-1 declined with an almost similar half-life as seen for the parent compound. In general, inter-subject variability of metabolite M-1 exposure, as measured by PK parameters Cmax, AUC(0-tlast), and AUC(0-25) was high with %CV ranging from 65.8% to 115%.

9 Study no Page: 8 of 13 The metabolite ratios of AUC(0-tlast) and AUC(0-25) of metabolite M-1 to copanlisib had a high inter-individual variability but seemed to be independent of the copanlisib dose and co-medication with individual ratios from to 22.1% and from 1.53 to 18.9%, respectively. The amount of copanlisib and metabolite M-1 excreted into urine during the 25-h urine collection interval was low with about 4.9% and 0.87% of the administered dose, respectively. Co-administration of gemcitabine and cisplatin did not influence the PK of copanlisib and its metabolite M-1. Table 0-1: Selected pharmacokinetic parameters of copanlisib [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of copanlisib (PKS) Treatment A Treatment B Parameter Unit n Cohort A-1 n Cohort A-2 n Cohort B n Cohort BTE Cmax µg/l / / / / 73.3 ( ) ( ) ( ) ( ) AUC(0-tlast) µg h/l / / / / 44.4 ( ) ( ) ( ) ( ) AUC(0-25) µg h/l / / / / 42.0 ( ) ( ) ( ) ( ) Cohort A1 = 0.6 mg/kg copanlisib mg/m² gemcitabine Cohort A2 = 0.8 mg/kg copanlisib mg/m² gemcitabine Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Table 0-2: Selected pharmacokinetic parameters of copanlisib metabolite M-1 [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of copanlisib in the Cohorts A-2 and B and in the expansion cohort (PKS) Treatment A Treatment B Parameter Unit n Cohort A-2 n Cohort B n Cohort BTE Cmax µg/l / 87.7 ( ) / 92.2 ( ) / 98.7 ( ) AUC(0-tlast) µg h/l / 115 ( ) / 88.9 ( ) / 82.9 ( ) AUC(0-25) µg h/l / 84.9 ( ) / 65.8 ( ) / 80.0 ( ) Cohort A2 = 0.8 mg/kg copanlisib mg/m² gemcitabine Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Gemcitabine and metabolite dfdu Selected PK parameters for gemcitabine and its metabolite dfdu are shown in Table 0-3 and Table 0-4, respectively. Following an IV infusion of 1000 mg/m² gemcitabine on C1D1, the geometric mean Cmax ranged between 11.1 and 16.8 mg/l and were observed at a median time of about 0.5 hours, i.e. at the end of the infusion. Inter-subject variability of gemcitabine PK parameters Cmax and AUC(0-tlast) was

10 Study no Page: 9 of 13 low to high (CV 19.4 to 77.2%). After reaching Cmax, concentration vs. time curves declined rapidly within the first 2 hours after start of infusion and concentrations of gemcitabine in plasma were not measurable beyond 3 h following start of infusion. T1/2 is not reported because there is not enough data to calculate and report a reliable value. For the metabolite dfdu, the mean maximum plasma concentrations ranged between 31.4 and 38.2 mg/l and were reached at a median tmax of 0.5 to 1 hour after start of gemcitabine infusion. Metabolite dfdu concentrations in plasma were above the LLOQ for all subjects up to 24 hours after start of gemcitabine infusion. Inter-subject variability of dfdu PK parameters Cmax and AUC(0-tlast) was low to moderate (CV 15.1 to 43.3%). T1/2 is not reported because there is not enough data to calculate and report a reliable value. Co-administration of gemcitabine with copanlisib and cisplatin did not influence the PK of gemcitabine and metabolite dfdu. Table 0-3: Selected pharmacokinetic parameters of gemcitabine [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of gemcitabine (PKS) Treatment A Treatment B Parameter Unit n Cohort A-1 n Cohort A-2 n Cohort B n Cohort BTE Cmax mg/l / 65.6 ( ) / 40.2 ( ) / 22.5 ( ) / 77.2 ( ) AUC(0-tlast) mg h/l / 52.6 ( ) / 34.2 ( ) / 19.4 ( ) / 56.6 ( ) Cohort A1 = 0.6 mg/kg copanlisib mg/m² gemcitabine Cohort A2 = 0.8 mg/kg copanlisib mg/m² gemcitabine Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Table 0-4: Selected pharmacokinetic parameters of gemcitabine metabolite dfdu [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of gemcitabine (PKS) Treatment A Treatment B Parameter Unit n Cohort A-1 n Cohort A-2 n Cohort B n Cohort BTE Cmax mg/l / 32.2 ( ) / 15.1 ( ) / 17.4 ( ) / 23.3 ( ) AUC(0-tlast) mg h/l / 28.8 ( ) / 27.4 ( ) / 18.0 ( ) / 43.3 ( ) Cohort A1 = 0.6 mg/kg copanlisib mg/m² gemcitabine Cohort A2 = 0.8 mg/kg copanlisib mg/m² gemcitabine Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin

11 Study no Page: 10 of 13 Total and free (unbound) platinum Selected PK parameters for total and unbound platinum are given in Table 0-5 and Table 0-6, respectively. Following an IV infusion of 25 mg/m² cisplatin on C1D1 over 1 hour, mean maximum plasma concentrations of total platinum was about 1.3 mg/l, and was observed at a median tmax of 1 hour, which corresponds to the end of cisplatin infusion. Concentrations of total platinum in plasma were measurable up to 24 hours in all subjects. Inter-subject variability of total platinum PK parameters Cmax and AUC(0-tlast) was low (CV 10.2 to 21.7%). T1/2 is not reported because there is not enough data to calculate and report a reliable value. For free platinum, mean maximum plasma concentrations between and mg/l were also observed at a median time of 1 hour after start of cisplatin infusion. After reaching Cmax, concentrations of free platinum declined rapidly and were only measurable for a median tmax of about 3 hours after start of cisplatin infusion. Inter-subject variability of free platinum PK parameters Cmax and AUC(0-tlast) was low to moderate (CV 17.3 to 37.7%). T1/2 is not reported because there is not enough data to calculate and report a reliable value. The PK of total and free (unbound) platinum was consistent when co-administered with copanlisib and gemcitabine in Cohorts B and BTE. Table 0-5: Selected pharmacokinetic parameters of total platinum [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of cisplatin in both expansion cohorts (PKS) Parameter Unit n Cohort B n Cohort BTE Cmax mg/l / 10.2 ( ) / 21.4 ( ) AUC(0-tlast) mg h/l / 10.4 ( ) / 21.7 ( ) Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Table 0-6: Selected pharmacokinetic parameters of free (unbound) platinum [geometric mean / %CV (range)] following the Cycle 1, Day 1 administration of cisplatin (PKS) Parameter Unit n Cohort B n Cohort BTE Cmax mg/l / 17.3 ( ) / 24.9 ( ) AUC(0-tlast) mg h/l / 24.0 ( ) / 37.7 ( ) Cohort B = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Biliary tract expansion (BTE) = 0.8 mg/kg copanlisib mg/m² gemcitabine + 25 mg/m² cisplatin Safety evaluation All 50 (100%) subjects included in the SAF population experienced at least 1 TEAE, 47 (98.0%) experienced at least 1 TEAE related to copanlisib, 49 (98.0%) experienced at least 1 TEAE related to gemcitabine, and 34 (100%) experienced at least 1 TEAE related to cisplatin, irrespective of seriousness, severity, and causality. A total of 48 (96.0%) subjects experienced at least 1 TEAE of CTCAE v4.0 Grade 3: 26 (52.0%) subjects experienced at least 1 TEAE of Grade 3, 20 (40.0%) subjects experienced at least 1 TEAE of Grade 4, and 2 (4.0%) subjects experienced at least 1 TEAE

12 Study no Page: 11 of 13 of Grade 5. A total of 43 (86.0%) subjects experienced at least 1 TEAE of Grade 3 related to copanlisib, 42 (84.0%) experienced at least 1 TEAE of Grade 3 related to gemcitabine, and 29 (85.3%) experienced at least 1 TEAE of Grade 3 related to cisplatin. Serious TEAEs were reported for 26 (52.0%) subjects. In the 50 subjects treated in this study, the most common TEAEs of any CTCAE v4.0 grade by MedDRA PT, occurring in 20% of subjects were nausea in 43 (86.0%) subjects, hyperglycemia in 40 (80.0%) subjects, fatigue in 38 (76.0%) subjects, neutropenia in 34 (68.0%) subjects, decreased appetite and anemia in 32 (64.0%) subjects each, thrombocytopenia in 27 (54.0%) subjects, hypertension and vomiting in 25 (50.0%) subjects each, diarrhea in 21 (42.0%) subjects, platelet count decreased and constipation in 15 (30.0%) subjects each, rash in 14 (28.0%) subjects, leukopenia, hyponatremia and pyrexia in 13 (26.0%) subjects each, and weight decreased and chills in 12 (24.0%) subjects each. In the 50 subjects treated in this study, the most common TEAEs of any CTCAE v4.0 grade related to copanlisib by MedDRA PT occurring in 20% of all 50 subjects in the SAF population were hyperglycemia in 38 (76%) subjects, nausea in 37 (74.0%) subjects, fatigue in 33 (66.0%) subjects, decreased appetite in 28 (56.0%) subjects, neutropenia in 26 (52.0%) subjects, hypertension in 22 (44.0%) subjects, anemia in 21 (42.0%) subjects, thrombocytopenia and vomiting in 18 (36.0%) subjects each, diarrhea in 15 (30.0%) subjects, platelet count decreased in 13 (26.0%) subjects, and leukopenia in 12 (24.0%) subjects. In the 50 subjects treated in this study, the most common TEAEs of CTCAE v4.0 Grade 3 by MedDRA PT, occurring in 5% of subjects were neutropenia in 26 (52.0%) subjects, hypertension in 21 (42.0%) subjects, thrombocytopenia in 15 (30.0%) subjects, hyperglycemia in 12 (24.0%) subjects, fatigue in 11 (22.0%) subjects, leukopenia in 8 (16.0%) subjects, lymphocyte count decreased, nausea, neutrophil count decreased, hyponatremia, and anemia in 6 (12.0%) subjects each, abdominal pain, platelet count decreased, pneumonia, and dehydration in 4 (8.0%) subjects each, and dyspnea, febrile neutropenia, hypokalemia, hypophosphatemia, lymphopenia, and pulmonary embolism in 3 (6.0%) subjects each. Dose modifications (dose interruption and / or dose reduction) of copanlisib and/or gemcitabine and/or cisplatin due to TEAEs were necessary for 45 (90.0%) subjects. Among those 45 subjects with dose modifications, 26 (52.0%) subjects had a dose modification due to copanlisib-related TEAEs, 37 (74.0%) subjects due to gemcitabine-related TEAEs, and 20 (58.8%) subjects due to cisplatin-related TEAEs. Overall, 15 (30.0%) subjects permanently discontinued copanlisib due to 1 or more TEAEs: 22 TEAEs leading to permanent discontinuation of copanlisib were reported in total. The most common TEAE leading to permanent discontinuation of copanlisib was fatigue in 3 subjects. There were 5 deaths reported in this study, including 2 subjects who died during treatment or within 30 days after permanent discontinuation of study treatment and 3 subjects who died later than 30 days after permanent discontinuation of study treatment. The primary cause of death in the 2 subjects who died on-study were AEs associated with clinical disease progression (respiratory failure) or otherwise unrelated to treatment (cardiac failure). The primary cause of death in the 3 subjects who died later than 30 days after permanent discontinuation of study treatment were either

13 Study no Page: 12 of 13 due to AEs associated with clinical progression of disease or disease progression. The investigators assessed all deaths as not related to study treatment. Out of 50 subjects treated in this study, 26 (52.0%) subjects experienced at least 1 serious TEAE. Serious TEAEs by MedDRA PT experienced by 5% of the 50 subjects were pneumonia in 5 (10.0%) subjects, and dehydration and abdominal pain in 3 (5.5%) subjects each. Pneumonitis CTCAE v4.0 Grade 3 was reported in 1 (2.0%) subject (Cohort BTE), diagnosed after completion of Cycle 11.Following treatment pneumonitis improved to Grade 2, and the subject continued study treatment. Only 1 (2.0%) subject (Cohort A-1) had a TEAE that was assessed by the investigator as a DLT and serious. The subject with resected ER + /PR - /HER2 + breast cancer experienced posterior reversible encephalopathy syndrome of CTCAE v4.0 Grade 3. Copanlisib and gemcitabine were both permanently discontinued, remedial therapy was given, and the event resolved in approximately 8 days. Laboratory toxicities Common laboratory toxicities (CTCAE v4.0 Grade 1) reported for 50% of subjects with postbaseline laboratory data were WBC count decreased (90.0%), neutrophil count decreased and hyperglycemia (85.7% each), platelet count decreased (81.6%), anemia (79.6%), as well as decreased lymphocyte count (69.4%). Most hematological and biochemical toxicities were of CTCAE v4.0 Grade 1 or Grade 2. Infusion-related increases in plasma glucose and plasma insulin were observed in nearly all subjects. Since in this study few subjects (18.0%) received systemic corticosteroids as concomitant medication, the observed glucose increase can likely be attributed to copanlisib. ECOG performance status All 50 treated subjects had an ECOG performance status of 0 (fully active) or 1 (restricted active) at baseline. In most subjects (24 subjects, 64.9%), the ECOG performance status did not change during treatment. Vital signs / physical examination No clinically relevant changes were observed in vital signs except for blood pressure. Increases in BP post-dose were more commonly seen in systolic BP and were easily manageable. No clinically relevant changes in weight from baseline were observed. 12-lead ECG and LVEF No subject had a clinically significant abnormal ECG recorded during this study. A decrease in LVEF >10% occurred in 1 of 32 subjects with baseline and (last) post-baseline measurements. However, the decreased values remained within the normal LVEF limits. In this study, no TEAE of reduced LVEF were reported. One subject with extensive cardiac history died of cardiac failure, not assessed as drug-related by the investigator

14 Study no Page: 13 of 13 Overall conclusions No indication for any clinically-relevant PK interactions between copanlisib and gemcitabine or cisplatin was observed in the present study. Individual subjects showed clinical efficacy: out of the 16 subjects treated with copanlisibgemcitabine, 1 subject with primary peritoneal carcinoma achieved a PR. Out of 34 subjects treated with the triplet copanlisib-gemcitabine-cisplatin, 1 subject with gallbladder cancer achieved CR and 3 subjects with intrahepatic biliary cancer achieved PR. Two of the subjects who achieved PR were in Cohort BTE The safety profile of the copanlisib-gemcitabine-cisplatin combination treatment was generally consistent with the previously observed toxicities of copanlisib given as single agent (e.g. hyperglycemia, nausea, rash, diarrhea, and hypertension) as well as the wellknown safety profile of gemcitabine (e.g. hematological toxicities) and cisplatin (e.g. hematological toxicities, nausea). Due to the combination, the frequency of hematological toxicities was higher than reported with copanlisib alone. Most toxicities were manageable through dose modifications and remedial drug treatment. No new or unexpected safety signals were observed. The RP2D for the combination of copanlisib and gemcitabine was defined as copanlisib 0.8 mg/kg + gemcitabine 1000 mg/m 2 and the RP2D for the triplet combination is defined as copanlisib 0.8 mg/kg + gemcitabine 1000 mg/m mg/m 2 cisplatin on Days 1 and 8 of a 21-day treatment cycle.