ORIGINAL RESEARCH Two phase I studies of BI 836880, a vascular endothelial growth factor/ angiopoietin-2 inhibitor, administered once every 3 weeks or once weekly in patients with advanced solid tumors C. Le Tourneau1*, H. Becker2, R. Claus2,3, E. Elez4,5, F. Ricci1, R. Fritsch2, Y. Silber3, A. Hennequin6, J. Tabernero4,7, G. Jayadeva8, D. Luedtke9, M. He10 & N. Isambert6y 1 Department of Drug Development and Innovation (D3i), Institut Curie, INSERM U900 Research Unit, Paris-Saclay University, Paris & Saint-Cloud, France; 2Department of Medicine I (Hematology, Oncology and Stem Cell Transplantation), Medical Center e University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg; 3 Medical Faculty Augsburg University, Augsburg, Germany; 4Vall d’Hebron Barcelona Hospital Campus and Vall d’Hebron Institute of Oncology (VHIO), Barcelona; 5 Autonomous University of Barcelona, Barcelona, Spain; 6Centre Georges-François Leclerc, Dijon, France; 7UVic-UCC, IOB-Quiron, Barcelona, Spain; 8Boehringer Ingelheim International GmbH, Ingelheim; 9Boehringer Ingelheim Pharma GmbH and Co KG., Biberach, Germany; 10Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, USA Available online 13 September 2022 Background: BI 836880 is a humanized bispecific nanobody® that inhibits vascular endothelial growth factor and angiopoietin-2. Here, we report results from two phase I, nonrandomized, dose-escalation studies (NCT02674152 and NCT02689505; funded by Boehringer Ingelheim) evaluating BI 836880 in patients with confirmed locally advanced or metastatic solid tumors, refractory to standard therapy, or for which standard therapy was ineffective. Patients and Methods: Patients aged 18 years, with an Eastern Cooperative Oncology Group performance status of 0-2 and adequate organ function received escalating intravenous doses of BI 836880 once every 3 weeks (Q3W; Study 1336.1) or once weekly (QW; Study 1336.6). Primary objectives were maximum tolerated dose (MTD) and recommended phase II dose of BI 836880, based on dose-limiting toxicities (DLTs) during the first cycle. Results: Patients received one of five dosages of 40-1000 mg Q3W (29 patients) or 40-240 mg QW (24 patients). One DLT occurred with Q3W treatment [Grade (G) 3 pulmonary embolism (1000 mg)]. Five DLTs occurred in four patients treated QW [G2 proteinuria (120 mg); G3 hypertension (180 mg); G3 proteinuria and G3 hypertension (240 mg); and G4 respiratory distress (240 mg)]. All patients experienced adverse events, most commonly hypertension with Q3W treatment (89.7%; G3 41.4%), and asthenia with QW treatment (62.5%). Two patients treated Q3W (both 1000 mg) and three patients treated QW (120 mg, 2 patients; 180 mg, 1 patient) experienced partial response. Conclusions: The MTD of BI 836880 was 720 mg Q3W and 180 mg QW. BI 836880 was generally manageable and demonstrated preliminary efficacy. Clinical trial registration: ClinicalTrials.gov NCT02674152; https://clinicaltrials.gov/ct2/show/NCT02674152 and NCT02689505; https://clinicaltrials.gov/ct2/show/NCT02689505 Key words: Vascular endothelial growth factor, angiopoietin-2, advanced solid tumors, phase I, nanobody INTRODUCTION overexpressed in cancer.3 Targeting VEGF is an established Angiogenesis, the process by which new blood vessels are anticancer therapeutic approach. The anti-VEGF-A mono- formed from pre-existing vasculature, is one of the hall- clonal antibody, bevacizumab, was first approved over 15 marks of cancer.1,2 Vascular endothelial growth factor years ago, and is now used in the treatment of a range (VEGF) is a proangiogenic growth factor that is often of solid tumor types.4 Other agents targeting angiogenesis, such as the antivascular endothelial growth factor receptor-2 (VEGFR-2) antibody ramucirumab, the recom- *Correspondence to: Professor Christophe Le Tourneau, Department of Drug Development and Innovation (D3i), Institut Curie, INSERM U900 Research unit, binant fusion protein aflibercept, and VEGF receptor Paris-Saclay University, Paris & Saint-Cloud, France tyrosine kinase inhibitors, are also approved for the E-mail: christophe.letourneau@curie.fr (C. Le Tourneau). treatment of a variety of advanced cancers.5,6 y Present address: Centre Hospitalier Universitaire de Poitiers, Pôle Régional While the activity of VEGF and the therapeutic value of de Cancérologie, Poitiers, France. VEGF-signaling inhibition have been established, the contri- 2059-7029/© 2022 The Author(s). Published by Elsevier Ltd on behalf of European Society for Medical Oncology. This is an open access article under the bution of other proangiogenic pathways and the clinical CC BY license (http://creativecommons.org/licenses/by/4.0/). impact of selective inhibition of those other targets are less Volume 7 - Issue 5 - 2022 https://doi.org/10.1016/j.esmoop.2022.100576 1 ESMO Open C. Le Tourneau et al. well described. One such target is the angiopoietin/Tie were also required to have an Eastern Cooperative signaling pathway, which plays a role in vascular stability and Oncology Group performance status of 0-2; adequate organ acts in a complementary manner to VEGF.7 Angiopoietin-2 function; a life expectancy of 3 months; and had recov- (Ang-2) inhibits angiopoietin-1-induced Tie-2 signaling and ered from all reversible adverse events (AEs) of previous promotes vessel remodeling, which is necessary for VEGF- anticancer therapies to baseline (except alopecia, sensory induced angiogenesis.8,9 There is also crosstalk between peripheral neuropathy, and those not considered clinically the two pathways as Ang-2 enhances VEGF signaling, and significant) or Common Terminology Criteria for Adverse VEGF upregulates Ang-2 expression.9,10 Events (CTCAE) Grade 1. VEGF and Ang-2 also have distinct immunosuppressive Exclusion criteria included: known hypersensitivity to the effects in the tumor microenvironment.11 Thus, inhibiting study drug; prior treatment with any systemic anticancer VEGF and Ang-2 may also enhance the tumor microenvi- therapy within 28 days; serious concomitant disease; sur- ronment to support T cell function, providing a rationale for gery/major injury within 4 weeks; QT prolongation and/or combining with agents that target the programmed death long QT syndrome; or symptomatic brain metastases. protein-1 (PD-1) pathway. Considering the mechanism of action, patients with signif- Clinical studies have demonstrated that patients with icant cardiovascular or cerebrovascular diseases (i.e. un- tumors harboring high expression levels of both VEGF and controlled hypertension, unstable angina, history of Ang-2 have poorer prognoses than patients with high infarction within the past 6 months, congestive heart failure expression of only one of these proteins.12,13 Therefore, it classification of at least New York Heart Association class II); has been suggested that dual inhibition of VEGF and Ang-2 a severe hemorrhagic or thromboembolic event in the past may improve clinical activity in comparison to the inhibition 12 months; and patients who required full-dose anti- of either pathway alone. This is supported by data from coagulation were also excluded. preclinical models.7,14 BI 836880 is a humanized bispecific nanobody® comprising Treatment two single variable domains that bind to all splice variants of In both studies, patients received BI 836880 via intravenous VEGF (VEGF-165, VEGF-121, and VEGF-189) and Ang-2, infusion at a starting dose of 40 mg (Q3W in 1336.1 and QW inhibiting binding of VEGFR-2 and Tie-2. It also contains an in 1336.6). Dose escalation was guided by a Bayesian lo- additional albumin module that extends half-life in vivo.9 gistic regression model (BLRM) utilizing escalation with BI 836880 potently and selectively neutralizes VEGF and overdose control (EWOC). Ang-2 and showed superior antitumor activity to VEGF or The BLRM estimated the MTD by updating estimates of Ang-2 inhibition alone in preclinical models of cancer.9 the probability of observing a dose-limiting toxicity (DLT) in Here we report results from two phase I studies of the MTD evaluation period for each dose level as data BI 836880 in patients with advanced or metastatic solid accumulated. The dose of BI 836880 was escalated in co- tumors. Study 1336.1 assessed BI 836880 given once every horts at predefined provisional dose levels based on a 3 weeks (Q3W), and Study 1336.6 assessed BI 836880 given maximum escalation of 200% for doses 120 mg Q3W and once weekly (QW). The primary objective of each study was 40 mg QW and 100% for doses >120 mg Q3W or QW. to define the maximum tolerated dose (MTD) and recom- Successive cohorts of patients received increasing doses of mended phase II dose (RP2D) of BI 836880 in patients with BI 836880 until the MTD was reached. advanced or metastatic solid tumors. Once the MTD had been reached, this dose level was expanded to further examine safety and further endpoints METHODS at this dose. Treatment was continued until disease pro- Study design and patients gression or unacceptable toxicity. Both studies were phase I, noncontrolled, nonrandomized, open-label, dose-escalation studies. Study 1336.1 Endpoints and assessments (NCT02674152) was conducted at three centers in France The primary endpoint in both studies was the MTD of and Germany (January 2016 to September 2018) and Study BI 836880 and the number of patients experiencing DLTs 1336.6 (NCT02689505) was conducted at two centers in during the MTD evaluation period (first 3-week cycle). France and Spain (May 2016 to July 2019). Both studies Secondary endpoints included drug-related AEs (DRAEs) were sponsored by Boehringer Ingelheim. leading to dose reduction or discontinuation; exposure The studies were conducted in accordance with the measures [area under the plasma concentrationetime Declaration of Helsinki and Good Clinical Practice guidelines curve from time zero to the last quantifiable concentra- as defined by the International Conference on Harmoniza- tion (AUC0etz)]; and disposition kinetic measures [elimina- tion; all patients provided written informed consent for tion half-life (t1/2)] after the first dose. Further endpoints study participation. included best overall response. Eligible patients in both studies were aged 18 years, AEs were graded according to CTCAE version 4.03. DLTs with confirmed locally advanced or metastatic solid tumors are defined in Supplemental Materials S1, available at that were either refractory after standard therapy or for https://doi.org/10.1016/j.esmoop.2022.100576. MTD was which standard therapy was not reliably effective. Patients defined as the highest dose with <25% risk of the true DLT 2 https://doi.org/10.1016/j.esmoop.2022.100576 Volume 7 - Issue 5 - 2022 C. Le Tourneau et al. ESMO Open rate being above 0.33 during the MTD evaluation period, treatment. All patients had discontinued the QW treatment and could be considered reached if there was a sufficiently schedule [mainly due to progressive disease (n ¼ 17)]. large probability that the true DLT rate was in the target Across both studies, no patients withdrew consent interval of 0.16-0.33. RP2D was selected based on an (Supplementary Figure S1, available at https://doi.org/10. analysis of overall safety, and pharmacokinetic (PK) and 1016/j.esmoop.2022.100576). pharmacodynamic (PD) data. Baseline characteristics for patients treated with the Serial blood samples were collected to determine plasma Q3W and QW schedules are presented in Table 1. In pa- levels of BI 836880, free VEGF-A and free/total levels of tients receiving treatment Q3W, the median (range) age Ang-2 and antidrug antibodies (ADAs). BI 836880 plasma was 57.0 (28-79) years and the majority of patients were concentrations were quantified using a validated bio- female (62.1%). In patients receiving the QW treatment analytical assay. Plasma biomarker analyses of free/total schedule the median (range) age was 58.5 (31-74) years and VEGF-A and Ang-2 were performed using immunoassays the majority of patients were also female (70.8%). and the presence of ADAs to BI 836880 was assessed Median duration of treatment was 22.0 days for the Q3W via a tiered approach using a validated electro- schedule (range 1-786) and 66.5 days for the QW schedule chemiluminescence assay (MSD QuickPlex SQ 120 Reader; Meso Scale Discovery® Model SQ 120). Tumor response was assessed every 6 weeks and evalu- ated by the investigator, per Response Evaluation Criteria in Table 1. Baseline characteristics for both studies (treated set) Solid Tumors (RECIST version.1.1) and was defined as the Q3W QW best response at any time from the date of the first (n [ 29) (n [ 24) administration of the study drug until progression. Median age at baseline, years (range) 57.0 (28-79) 58.5 (31-74) Patients were followed up until disease progression, start Gender, n (%) of a new anticancer therapy, death, end of the study, or lost Male 11 (37.9) 7 (29.2) Female 18 (62.1) 17 (70.8) to follow-up. Race, n (%) White 9 (31.0) 0 (0) Hispanic/Latino 0 (0) 14 (58.3) Statistical analysis Missinga 20 (69.0) 10 (41.7) Baseline ECOG performance status, n (%) Dose escalation was guided by a two-parameter BLRM with 0 11 (37.9) 17 (70.8) overdose control. The estimated probability of a DLT at each 1 18 (62.1) 7 (29.2) dose level from the model was summarized using the 2 0 (0) 0 (0) following intervals: under toxicity [0.00, 0.16]; targeted Median time from first diagnosis, years 3.5 (0.3-22.6) 3.8 (0.5-12.6) (range) toxicity [0.16, 0.33]; and over toxicity [0.33, 1.00]. The Primary cancer diagnosis, n (%) BLRM-recommended dose for the next cohort was the level Colon/rectum 5 (17.2) 10 (41.7) with the highest posterior probability of the DLT rate falling Pancreas 6 (20.7) 3 (12.5) Breast 4 (13.8) 1 (4.2) in the target interval [0.16, 0.33] among the doses fulfilling Esophagus 3 (10.3) 1 (4.2) the EWOC criterion. According to the EWOC criterion, it Ovary 1 (3.4) 3 (12.5) should be unlikely (<25% posterior probability) that the DLT Cavum 2 (6.9) 0 (0) Choroid melanoma 2 (6.9) 0 (0) rate at that dose is 0.33. Thymus 2 (6.9) 0 (0) Analyses of AEs were descriptive and the relationship Other 4 (13.8)b 6 (25.0)c between the study drug and the AE was assessed by the Prior anticancer therapy, n (%) Chemotherapy 29 (100.0) 24 (100.0) investigator. The best overall response and ADA levels were Immunotherapy 4 (13.8) 4 (16.7) analyzed descriptively. Radiotherapy 18 (62.1) 8 (33.3) PK parameters were calculated using noncompartmental Hormone therapy 5 (17.2) 2 (8.3) analysis via Phoenix WinNonlin. A population PK and inte- Other 10 (34.5) 1 (4.2) Number of prior systemic therapy regimens, 2 (1.0-4.0)d 1.5 (1.0-3.0)e grated PD model was developed based on data from pa- median (range) tients treated Q3W to support dose selection for future Prior antiangiogenic treatment, n (%) 14 (48.3) 14 (58.3) studies. Population PK and PK/PD analyses for repeated- Baseline conditions, n (%) Hypertension 16 (55.2) 8 (33.3) measures endpoints were conducted via nonlinear mixed Anemia 9 (31.0) 0 (0) effects modeling with a qualified installation of the ECOG, Eastern Cooperative Oncology Group; Q3W; once every 3 weeks; QW, once nonlinear mixed effects modeling (NONMEM) software. weekly. a Race and ethnicity data were not recorded for patients treated in France, as it was prohibited by local law. b RESULTS This category includes one patient each with the primary site of left iliac fossa (neuroectodermal tumor), proximal jejunum, uterus, and unknown origin. c This category includes one patient each with the primary site of cervix, Patients and treatment endometrium, gastric, kidney, peritoneum, and trachea. d Overall, 29 patients received BI 836880 Q3W, and 24 pa- A total of 8 patients had received one prior regimen, 11 patients had received two prior regimens, 4 patients had received three prior regimens, and 6 patients had tients received BI 836880 QW. At the time of data cut-off, received four prior regimens. e 27 patients had discontinued Q3W treatment [mainly due A total of 12 patients received one prior regimen, 9 patients had received two prior regimens, and 3 patients had received three prior regimens. to progressive disease (n ¼ 19)] and 2 patients remained on Volume 7 - Issue 5 - 2022 https://doi.org/10.1016/j.esmoop.2022.100576 3 ESMO Open C. Le Tourneau et al. (range 7-524). Mean duration of treatment was 105.5 days Safety and tolerability for Q3W treatment and 112.4 days for QW treatment. All 29 patients treated Q3W reported 1 AE during the on- Patients received BI 836880 at five dose levels from 40 to treatment period; the median number of AEs per patient 1000 mg Q3W (40 mg, n ¼ 3; 120 mg, n ¼ 2; 360 mg, n ¼ 2; was 7 (range 2-36). The most common AEs were hyper- 720 mg, n ¼ 17; 1000 mg, n ¼ 5), and at five dose levels tension (89.7%), asthenia (51.7%), and nausea (44.8%). The from 40 to 240 mg QW (40 mg, n ¼ 2; 120 mg, n ¼ 5; most common Grade 3 AE was hypertension (41.4%). 150 mg, n ¼ 3; 180 mg, n ¼ 11; 240 mg, n ¼ 3). Three patients (10.3%) had AEs leading to death (two pa- tients with metastatic pancreatic cancer and one patient with esophageal cancer); all of these patients died due to Determination of the MTD progressive disease, and none were considered related to In patients treated Q3W, one DLT occurred during the MTD treatment. Five patients discontinued the study drug due to evaluation period (Grade 3 pulmonary embolism that led an AE; one was a DLT [drug-related Grade 3 pulmonary to discontinuation in a patient with pancreatic adenocar- embolism (1000 mg dose cohort)]. cinoma in the 1000 mg group). Although the BLRM A total of 22 (75.9%) patients had DRAEs; the most allowed expanding the 1000 mg dose for MTD determi- common were hypertension (51.7%), anemia, and asthenia nation (posterior probability of the true DLT rate was (20.7% each; Table 2). Of the patients with hypertension, six <0.25), the 720 mg Q3W dose was defined as the MTD received antihypertensive medication. Grade 3/4 DRAEs and RP2D, based on an overall safety data review and PK/ occurred in 10 patients (34.5%). The most common Grade 3 PD data. DRAE was hypertension (8 patients, 27.6%). There was one In patients treated QW, five DLTs were reported in four Grade 4 DRAE (dyspnea; 720 mg cohort). Two patients had patients during the MTD evaluation period and all were DRAEs leading to discontinuation of study drug [Grade 3 drug-related serious AEs: Grade 2 proteinuria, which led to myocarditis (720 mg dose cohort) and Grade 3 pulmonary discontinuation in a patient with ovarian cancer (120 mg embolism (reported as a DLT in the 1000 mg dose cohort)]. cohort); Grade 3 hypertension in a patient with tracheal All 24 patients treated QW reported 1 AE during the on- sarcoma (180 mg cohort); Grade 3 hypertension and Grade treatment period; the median number of AEs per patient 3 proteinuria, which resulted in discontinuation in a patient was 7 (range 1-34). The most common AEs were asthenia with colic adenocarcinoma of the colon (240 mg cohort); (62.5%), decreased appetite (37.5%), and abdominal pain, and Grade 4 respiratory distress (not associated with constipation, and diarrhea (25.0% each). No patients had pulmonary embolism), which resulted in discontinuation in AEs leading to death. a patient with adenocarcinoma of the rectum (240 mg Overall, 17 (70.8%) patients had DRAEs (Table 2); the cohort). Therefore BI 836880 180 mg QW was determined most common were asthenia (37.5%) and hypertension as the MTD and RP2D. (20.8%). Of the patients who experienced hypertension, Table 2. Most common DRAEs occurring in ‡5% of patients treated with BI 836880 administered either Q3W or QW (treated set) AE, n (%) Q3W (n [ 29) QW (n [ 24) All Grades Grade 1 Grade 2 Grade 3 Grade 4 All Grades Grade 1 Grade 2 Grade 3 Grade 4 Any DRAE 22 (75.9) 5 (17.2) 7 (24.1) 9 (31.0) 1 (3.4) 17 (70.8) 4 (16.7) 7 (29.2) 5 (20.8) 1 (4.2) Hypertension 15 (51.7) 1 (3.4) 6 (20.7) 8 (27.6) 0 (0) 5 (20.8) 1 (4.2) 2 (8.3) 2 (8.3) 0 (0) Asthenia 6 (20.7) 3 (10.3) 2 (6.9) 1 (3.4) 0 (0) 9 (37.5) 4 (16.7) 5 (20.8) 0 (0) 0 (0) Anemia 6 (20.7) 3 (10.3) 3 (10.3) 0 (0) 0 (0) 1 (4.2) 1 (4.2) 0 (0) 0 (0) 0 (0) Vomiting 5 (17.2) 5 (17.2) 0 (0) 0 (0) 0 (0) 2 (8.3) 2 (8.3) 0 (0) 0 (0) 0 (0) Infusion-related reaction 5 (17.2) 4 (13.8) 1 (3.4) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Nausea 5 (17.2) 5 (17.2) 0 (0) 0 (0) 0 (0) 1 (4.2) 1 (4.2) 0 (0) 0 (0) 0 (0) Proteinuria 2 (6.9) 1 (3.4) 1 (3.4) 0 (0) 0 (0) 4 (16.7) 0 (0) 3 (12.5) 1 (4.2) 0 (0) Diarrhea 2 (6.9) 1 (3.4) 0 (0) 1 (3.4) 0 (0) 3 (12.5) 3 (12.5) 0 (0) 0 (0) 0 (0) Hypersensitivity 2 (6.9) 1 (3.4) 1 (3.4) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Increased AST 2 (6.9) 2 (6.9) 0 (0) 0 (0) 0 (0) 1 (4.2) 1 (4.2) 0 (0) 0 (0) 0 (0) Lymphopenia 2 (6.9) 1 (3.4) 1 (3.4) 0 (0) 0 (0) 1 (4.2) 0 (0) 1 (4.2) 0 (0) 0 (0) Decreased appetite 1 (3.4) 1 (3.4) 0 (0) 0 (0) 0 (0) 4 (16.7) 3 (12.5) 1 (4.2) 0 (0) 0 (0) Dyspnea 1 (3.4) 0 (0) 0 (0) 0 (0) 1 (3.4) 3 (12.5) 1 (4.2) 1 (4.2) 1 (4.2) 0 (0) Headache 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 3 (12.5) 3 (12.5) 0 (0) 0 (0) 0 (0) Ejection fraction decreased 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (8.3) 0 (0) 0 (0) 2 (8.3) 0 (0) Cardiac failure 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (8.3) 0 (0) 1 (4.2) 1 (4.2) 0 (0) Peripheral edema 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (8.3) 0 (0) 1 (4.2) 1 (4.2) 0 (0) Pleural effusion 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (8.3) 2 (8.3) 0 (0) 0 (0) 0 (0) Upper abdominal pain 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 2 (8.3) 1 (4.2) 1 (4.2) 0 (0) 0 (0) Grade 4 DRAE in patients treated QW was respiratory distress. There were no Grade 5 DRAEs. AE, adverse event; AST, aspartate aminotransferase; DRAE, drug-related adverse event; Q3W, once every 3 weeks; QW, once weekly. 4 https://doi.org/10.1016/j.esmoop.2022.100576 Volume 7 - Issue 5 - 2022 C. Le Tourneau et al. ESMO Open four received antihypertensive medication. Related Grade 1336.6 was ongoing at this time and no data were available 3/4 events occurred in six patients (25%); the most com- to include in the model). A two-compartment PK model mon Grade 3 events were hypertension and decreased provided a good description of BI 836880 plasma concen- ejection fraction [two patients each (8.3%)]. There was one trations at the population and individual levels. A popula- Grade 4 DRAE (respiratory distress which was reported as tion PK/PD model was developed to evaluate the effect of DLT; 240 mg cohort). Five patients had DRAEs leading to BI 836880 on plasma concentrations of Ang-2 for several discontinuation of study drug [three patients with DLTs: doses that Boehringer Ingelheim considered to use for the proteinuria (n ¼ 2) and respiratory distress (n ¼ 1); one next study. VEGF measurements were not considered for patient with Grade 3 abdominal infection and one patient this model, as VEGF-A levels were inhibited >90% for all with Grade 1 hemoptysis]. doses and all samples collected. A mass action model in which BI 836880 reversibly bound Ang-2 and sequestered it Pharmacokinetics and pharmacodynamics in a circulating BI:Ang-2 complex provided a good descrip- tion of free Ang-2 and total Ang-2 levels. Within this anal- Plasma concentrations of BI 836880 increased with ysis, it was shown that Ang-2 would be completely blocked increasing dose in patients treated with both Q3W and QW (>90% inhibition) over the entire dosing interval for doses schedules. Plasma concentrationetime profiles are shown of 360, 500, and 720 mg Q3W, in 85.7, 95.2, and 100% of in Figure 1 and Supplementary Figure S2, available at patients, respectively. Hence for the large majority of the https://doi.org/10.1016/j.esmoop.2022.100576. Further- patients on these two doses (360 and 500 mg) complete more, the maximum measured plasma concentration (Cmax) Ang-2 binding would be reached. Based on the results, and AUC0etz both increased in a dose-proportional manner Boehringer Ingelheim decided to include the 500 mg dose over the entire dose range with Q3W and QW treatment into the combination study with the PD-1 inhibitor eza- (Supplementary Table S1, available at https://doi.org/10. benlimab (NCT03468426 and NCT03697304). 1016/j.esmoop.2022.100576). In patients treated Q3W, blood VEGF-A and Ang-2 were almost completely inhibited (>90%) by BI 836880 at the Immunogenicity assessment first postdose sampling point (8 hours) of Cycle 1 irre- Of the 24 patients treated with the Q3W schedule who were spective of the dose administered for VEGF-A and for doses evaluable for ADAs, 15 (62.5%) showed pre-existing BI 836880 360 mg Q3W for Ang-2 (Figure 2). VEGF-A and Ang-2 ADAs before the start of treatment. From the nine remaining remained completely inhibited over the entire dosing in- patients, eight patients (33.3%) developed treatment- terval and all subsequent cycles. Results from QW dosing induced BI 836880 ADAs and only one patient (4.2%) did were similar (Supplementary Figure S3, available at https:// not develop ADAs.The time of onset of ADA development was doi.org/10.1016/j.esmoop.2022.100576). rather fast, usually within 24 days, and only two patients A population PK model for BI 836880 was developed developed BI 836880 ADAs at a later time point. using quantifiable postdose BI 836880 plasma concentra- Of the 23 patient treated with the QW schedule who tions from patients treated Q3W in Study 1336.1 (Study were evaluable for ADAs, 13 (56.5%) showed pre-existing 40 mg (n = 3) 120 mg (n = 2) 360 mg (n = 2) 1000 720 mg (n = 17) 1000 mg (n = 5) BI 836880 plasma conc. (μg/ml) 100 10 1 0 50 100 150 200 250 300 350 400 450 500 550 Time (h) Figure 1. gMean plasma concentrationetime profiles after first infusion (Cycle 1) of BI 836880 administered Q3W. conc, concentration; gMean, geometric mean; Q3W, once every 3 weeks. Volume 7 - Issue 5 - 2022 https://doi.org/10.1016/j.esmoop.2022.100576 5 ESMO Open C. Le Tourneau et al. 40 mg (n = 3/2/–/–) 120 mg (n = 2/–) 360 mg (n = 2/–/–/–) 720 mg (n = 17/15/7/3) 1000 mg (n = 5/3/3/3) A 120 Free Ang-2 bound to BI 836880 (%) 100 80 60 40 20 0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Time (h) B 105 90 Free VEGF-A bound to BI 836880 (%) 75 60 45 30 15 0 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 Time (h) Figure 2. Mean binding-time profiles of (A) Ang-2 and (B) VEGF-A to BI 836880 after multiple Q3W infusions. Ang-2, angiopoietin-2; Q3W, once every 3 weeks; VEGF-A, vascular endothelial growth factor A. BI 836880 ADAs before the start of treatment, including 2 induced BI 836880 ADAs and the PK of BI 836880 was not patients (8.7%) with ADA-positive samples at baseline only. influenced by the development of ADAs (Supplementary Of the remaining 10 patients, 8 patients (34.8%) developed Figure S4, available at https://doi.org/10.1016/j.esmoop. treatment-induced BI 836880 ADAs, and 2 patients (8.7%) 2022.100576). did not develop ADAs. The onset of ADA development occurred rather quickly, usually within the first or second treatment cycle (21-45 days); only two patients developed Antitumor activity BI 836880 ADAs at a later time point (during Cycle 3 or A waterfall plot of best percentage decrease from baseline Cycle 4). in the sum of the target lesions and a spider plot of tumor In patients treated both Q3W and QW, there was no dose response for patients treated with the Q3W and QW dependency observed in the development of treatment- treatment schedules are shown in Figure 3. 6 https://doi.org/10.1016/j.esmoop.2022.100576 Volume 7 - Issue 5 - 2022 C. Le Tourneau et al. ESMO Open A C 100 40 mg 120 mg 360 mg 100 Planned treatment 720 mg 1000 mg 40 mg Decrease from baseline (%) 80 80 120 mg (% change from baseline) 60 60 360 mg 720 mg 40 40 1000 mg Tumor size 20 20 0 0 –20 –20 –40 –40 –60 –60 –80 * * –80 –100 –100 1 3 5 7 9 11 13 15 17 19 21 0 84 168 252 336 420 504 588 672 756 Patient index sorted by maximum decrease (%) Days B D 100 40 mg 120 mg 150 mg 100 Planned treatment 180 mg 240 mg 40 mg Decrease from baseline (%) 80 80 120 mg (% change from baseline) 60 60 150 mg 180 mg 40 40 240 mg Tumor size 20 20 0 0 –20 –20 * –40 –40 * * –60 –60 –80 –80 –100 –100 1 3 5 7 9 11 13 15 17 19 21 23 0 84 168 252 336 420 504 588 Patient index sorted by maximum decrease (%) Days Figure 3. Waterfall plots of best percentage decrease from baseline in sum of target lesion diameter of indicator lesions in patients treated (A) Q3W and (B) QW, and spider plots of target tumor response by dose level in patients treated (C) Q3W and (D) QW. Q3W, once every 3 weeks; QW, once weekly. Positive values represent tumor growth; negative values represent tumor shrinkage. One patient treated Q3W had an unconfirmed partial response (who only had one postbaseline tumor measurement). The asterisk indicates patients with best overall response of confirmed partial response. Of the 22 patients treated Q3W who were evaluable for carcinoma (120 mg cohort; duration of response was not response, two patients had confirmed partial responses available as there were no further imaging data); and one (PRs; both patients received BI 836880 1000 mg). One pa- had mucinous colorectal adenocarcinoma (180-mg cohort; tient (previously treated with sorafenib) had cavum carci- duration of response: 12.1 weeks). Nine patients had SD, noma (duration of response: 12.1 weeks) and the other including three patients with SD lasting 6 months [breast patient (previously treated with bevacizumab) had ovarian invasive ductal carcinoma (150 mg cohort), rectal adeno- adenocarcinoma (duration of response: 5.6 weeks). In carcinoma (180 mg cohort), and gastric adenocarcinoma addition, one patient with HER2-negative breast cancer in (240 mg cohort)]. the 720 mg cohort had an unconfirmed PR (the patient had only one postbaseline tumor measurement). Nine patients had stable disease (SD), including three with SD lasting DISCUSSION 6 months [mucinous adenocarcinoma of unknown origin Here, we summarize results from two phase I studies of (360 mg cohort), sigmoid adenocarcinoma (720 mg cohort), BI 836880, a VEGF/Ang-2 inhibitor, which aimed to define and neuroectodermal tumor of the left iliac fossa (1000 mg the MTD and RP2D of BI 836880 administered Q3W or QW cohort)]. The patients with mucinous adenocarcinoma and in patients with advanced or metastatic solid tumors. The neuroectodermal tumor were on treatment at data cut-off, MTD was determined as BI 836880 720 mg Q3W and as with no evidence of progressive disease (last tumor mea- BI 836880 180 mg QW. There was no evidence of dose- surement on Days 782 and 571, respectively). Progressive dependent increases in safety risk in patients treated disease was the best response for 10 patients; confirmed either Q3W or QW, while responses tended to be observed response was not evaluable due to missing data in 8 pa- in the higher-dose cohorts. Based on population PK/PD tients (including the patient with an unconfirmed PR). model simulations, most patients were predicted to reach Of 23 patients treated QW who were evaluable for >90% free Ang-2 inhibition over the complete treatment response, 3 patients had confirmed PRs. One patient had cycle at steady-state at the 500 mg Q3W and 720 mg Q3W endometrial adenocarcinoma (120 mg cohort; duration of dose levels (91.4% and 95.6% of patients, respectively). response: 84.6 weeks); one had ovarian papillary serous VEGF-A measurements were not considered in the Volume 7 - Issue 5 - 2022 https://doi.org/10.1016/j.esmoop.2022.100576 7 ESMO Open C. Le Tourneau et al. population PK/PD model, as free VEGF-A levels were VEGF and Ang-2 also have distinct immunosuppressive inhibited >90% for all doses and all samples collected. effects in the tumor microenvironment. VEGF inhibits den- Treatment with BI 836880 was generally well tolerated and dritic cell maturation and T cell function, and promotes the there was no dose-dependent increase in the frequency of activity of regulatory T cell and myeloid-derived suppressor DRAEs. The most common DRAEs were hypertension and cells.11,23 Ang-2 increases the recruitment and adhesion of asthenia in patients treated Q3W and QW, and anemia in pa- neutrophils and Tie-2-expressing macrophages (TEMs) to tients treated Q3W.The AE profile was as expected for an agent the endothelium, and increases their conversion to the M2- targeting VEGF and Ang-2. Hypertension is a well-recognized like macrophage phenotype. It also stimulates TEMs to side-effect of VEGF inhibition,15 and has been documented secrete interleukin-10, which promotes regulatory T-cell in studies of bevacixumab16-20 and vanucizumab.21 expansion and the inhibition of effector T cells.11 As such, BI 836880 showed typical PK/PD for a nanobody. Plasma there is interest in combining antiangiogenic agents with concentrations of BI 836880 increased with increasing dose, PD-1 or programmed death-ligand 1 inhibitors. Combination and plasma concentrationetime profiles were similar of anti-VEGF therapy with PD-1 pathway inhibition has been among doses. In line with the mechanism of action of shown to be effective in several tumor types.24,25 As Ang-2 BI 836880, VEGF-A and Ang-2 were almost completely leads to immune suppression by a different mechanism to bound (>90%) by BI 836880 at the first postdose sampling VEGF,11 it is anticipated that addition of Ang-2 inhibition point of Cycle 1 and remained completely bound over the may further improve antitumor activity. entire dosing interval and all subsequent cycles. Similarly, in a study investigating the bispecific Ang-2 and VEGF antibody Conclusion vanucizumab, levels of both biomarkers decreased rapidly In the current studies, the MTD of BI 836880, a humanized following treatment.21 bispecific nanobody® comprising two single variable do- The majority of patients developed ADAs to BI 836880 mains that inhibit VEGF and Ang-2, was determined as 720 (95.8% and 91.3% of patients treated Q3W and QW, mg Q3W and 180 mg QW. Therefore 720 mg Q3W was respectively). These ADAs were most likely not directly chosen as the RP2D for the monotherapy studies. BI 836880 targeted against the paratopes, and rather were directed was generally manageable and demonstrated preliminary against the C terminus of BI 836880, as suggested by the antitumor activity; the findings from these phase I, single- rapid onset of production and the high percentage of pa- arm, nonrandomized studies therefore warrant further tients with pre-existing ADAs. Further characterization of investigation. Ongoing studies will examine the safety the domain specificity might be required to confirm this and antitumor activity of BI 836880 and other anticancer interpretation. agents given in combination with the PD-1 inhibitor eza- Early signs of clinical activity were observed in the two benlimab in advanced solid tumors (NCT03468426 and current studies that enrolled a heterogeneous population of NCT03697304).26-30 As ezabenlimab is given in a Q3W patients, all of whom had advanced solid tumors refractory schedule, and the QW schedule of BI 836880 provided no to available therapy. Two patients treated Q3W and three advantage over the Q3W schedule in terms of target patients treated QW achieved PRs. Of note, both patients engagement in the current studies, BI 836880 Q3W will be who received the Q3W schedule had previously been investigated. Based on overall assessment of efficacy, safety, treated with antiangiogenic therapies (patient with carci- and PK/PD from the current studies, the following doses noma of the cavum, sorafenib; patient with ovarian were selected for the dose-finding phase of the combina- adenocarcinoma, bevacizumab). Furthermore, durable SD tion study: 360, 500, and 720 mg Q3W. was observed in some patients, with two patients treated Q3W remaining on treatment at the time of data cut-off. FUNDING Symptomatic progression was reported after the cut-off date for the final report: progression-free survival was This work was supported by Boehringer Ingelheim Inter- 1578 days (360 mg) and 1321 days (1000 mg), respectively. national GmbH. Medical writing support for the develop- In a previous study investigating vanucizumab in 42 pa- ment of this manuscript, under the direction of the authors, tients with advanced solid tumors, two patients achieved was provided by Hannah Simmons, MSc, of Ashfield Med- PR, 19 patients had SD, and 10 patients were without dis- Comms, an Ashfield Health company and funded by Boeh- ease progression for 6 months.21 However, in a phase II ringer Ingelheim. The authors received no direct payment study in patients with previously untreated metastatic for development of the manuscript. colorectal cancer, vanucizumab plus chemotherapy did not improve progression-free survival versus bevacizumab plus DISCLOSURE chemotherapy.22 Furthermore, vanucizumab was associated CLT reports consulting fees from Merck Sharp & Dohme, with increased rates of antiangiogenic toxicity, such as hy- Bristol Myers Squibb, Merck Serono, Nanobiotix, Seattle pertension and peripheral edema. Development of vanuci- Genetics, GlaxoSmithKline, Roche, Rakuten, and Celgene. zumab as a monotherapy was discontinued following this HB reports receiving honoraria from Bristol Myers Squibb, study.22 Vanucizumab is a bispecific monoclonal antibody, Roche, Novartis, Servier, Pierre Fabre, and Merck Sharp & whereas BI 836880 is a nanobody. Dohme. EE reports being part of an advisory council or 8 https://doi.org/10.1016/j.esmoop.2022.100576 Volume 7 - Issue 5 - 2022 C. Le Tourneau et al. ESMO Open committee for Hoffman La Roche, Bristol Myers Squibb, 5. Syed YY, McKeage K. Aflibercept: a review in metastatic colorectal Servier, Amgen, Merck Serono, Array Biopharma, and cancer. Drugs. 2015;75(12):1435-1445. 6. Qin S, Li A, Yi M, Yu S, Zhang M, Wu K. Recent advances on anti- Sanofi; receiving honoraria from Array Biopharma, Merck angiogenesis receptor tyrosine kinase inhibitors in cancer therapy. Sharp & Dohme, AbbVie, Amgen, GlaxoSmithKline, Astra- J Hematol Oncol. 2019;12(1):27. Zeneca, Merck Sharp & Dohme Corp., Bristol Myers Squibb, 7. Hashizume H, Falcón BL, Kuroda T, et al. Complementary actions of Novartis, Boehringer Ingelheim, and Hoffman La Roche; inhibitors of angiopoietin-2 and VEGF on tumor angiogenesis and receiving consulting fees from Hoffman La Roche, Bristol growth. Cancer Res. 2010;70(6):2213-2223. 8. Hu B, Cheng SY. Angiopoietin-2: development of inhibitors for cancer Myers Squibb, Servier, Amgen, Merck Serono, Array Bio- therapy. Curr Oncol Rep. 2009;11(2):111-116. pharma, and Sanofi; receiving grants or funds from Roche, 9. Hofmann I, Baum A, Hilberg F, et al. Dual targeting of angiogenesis Amgen, Merck Serono, Sanofi/Aventis, Bristol Myers, Serv- pathways: combined blockade of VEGF and Ang2 signaling. Data pre- ier, and Array Biopharma. RF reports serving on safety sented at the 8th Euro Global Summit on Cancer Therapy. November monitoring committee of follow-up combination trial and 3-5, 2015; Valencia, Spain. 10. Reginato S, Gianni-Barrera R, Banfi A. Taming of the wild vessel: pro- receiving honoraria for this role from Boehringer Ingelheim. moting vessel stabilization for safe therapeutic angiogenesis. Biochem JT reports receiving honoraria for his educational collabo- Soc Trans. 2011;39(6):1654-1658. ration with Imedex, Medscape Education, MJH Life Sci- 11. Rahma OE, Hodi FS. The intersection between tumor angiogenesis and ences, and PeerView Institute for Medical Education and immune suppression. Cancer Res. 2019;25(18):5449. Physicians Education Resource (PER); and receiving consul- 12. Tsutsui S, Inoue H, Yasuda K, et al. Angiopoietin 2 expression in invasive ductal carcinoma of the breast: its relationship to the VEGF ting fees for scientific consultancy role for Array Biopharma, expression and microvessel density. Breast Cancer Res Treat. AstraZeneca, Avvinity, Bayer, Boehringer Ingelheim, Chugai, 2006;98(3):261-266. Daiichi Sankyo, F. Hoffmann-La Roche Ltd, Genentech Inc, 13. Andersen S, Donnem T, Al-Shibli K, et al. Prognostic impacts of HalioDX SAS, Hutchison MediPharma International, Ikena angiopoietins in NSCLC tumor cells and stroma: VEGF-A impact is Oncology, IQVIA, Lilly, Menarini, Merck Serono, Merus, strongly associated with Ang-2. PLoS One. 2011;6(5):e19773. 14. Brown JL, Cao ZA, Pinzon-Ortiz M, et al. A human monoclonal anti- MSD, Mirati, NeoPhore, Novartis, Orion Biotechnology, ANG2 antibody leads to broad antitumor activity in combination Peptomyc, Pfizer, Pierre Fabre, Samsung Bioepis, Sanofi, with VEGF inhibitors and chemotherapy agents in preclinical models. Seattle Genetics, Servier, Taiho, Tessa Therapeutics, and Mol Cancer Ther. 2010;9(1):145-156. TheraMyc. GJ, DL, and MH report being employed with 15. Hayman SR, Leung N, Grande JP, Garovic VD. VEGF inhibition, hyper- Boehringer Ingelheim. NI reports being part of an advisory tension, and renal toxicity. Curr Oncol Rep. 2012;14(4):285-294. 16. Ogita S, Tejwani S, Heilbrun L, et al. Pilot Phase II trial of bevacizumab council or committee for Eisai, Pfizer, and Daichi; receiving monotherapy in nonmetastatic castrate-resistant prostate cancer. ISRN honoraria from Amgen, Transgene, and Ipsen. The remain- Oncol. 2012;2012:242850. ing authors have declared no conflicts of interest. 17. Kreisl TN, Zhang W, Odia Y, et al. A phase II trial of single-agent bev- acizumab in patients with recurrent anaplastic glioma. Neuro Oncol. 2011;13(10):1143-1150. DATA SHARING 18. Cannistra SA, Matulonis UA, Penson RT, et al. Phase II study of bev- To ensure independent interpretation of clinical study re- acizumab in patients with platinum-resistant ovarian cancer or peri- toneal serous cancer. J Clin Oncol. 2007;25(33):5180-5186. sults and enable authors to fulfill their role and obligations 19. Burger RA, Sill MW, Monk BJ, Greer BE, Sorosky JI. Phase II trial of under the ICMJE criteria, Boehringer Ingelheim grants all bevacizumab in persistent or recurrent epithelial ovarian cancer or external authors access to relevant clinical study data. In primary peritoneal cancer: a Gynecologic Oncology Group Study. J Clin adherence with the Boehringer Ingelheim Policy on Trans- Oncol. 2007;25(33):5165-5171. parency and Publication of Clinical Study Data, scientific and 20. Cobleigh MA, Langmuir VK, Sledge GW, et al. A phase I/II dose- escalation trial of bevacizumab in previously treated metastatic medical researchers can request access to clinical study data breast cancer. Semin Oncol. 2003;30:117-124. after publication of the primary manuscript in a peer- 21. Hidalgo M, Martinez-Garcia M, Le Tourneau C, et al. First-in-hu- reviewed journal, regulatory activities are complete, and man phase I study of single-agent vanucizumab, a first-in-class other criteria are met. Researchers should use the https:// bispecific anti-angiopoietin-2/anti-VEGF-A antibody, in adult pa- vivli.org/ link to request access to study data and visit tients with advanced solid tumors. Clin Cancer Res. 2018;24(7): 1536-1545. https://www.mystudywindow.com/msw/datasharing for 22. Bendell JC, Sauri T, Gracián AC, et al. The McCAVE trial: vanucizumab further information. plus mFOLFOX-6 versus bevacizumab plus mFOLFOX-6 in patients with previously untreated metastatic colorectal carcinoma (mCRC). Oncol- ogist. 2020;25(3):e451-e459. REFERENCES 23. Yang J, Yan J, Liu B. Targeting VEGF/VEGFR to modulate antitumor 1. Li T, Kang G, Wang T, Huang H. Tumor angiogenesis and anti-angiogenic immunity. Front Immunol. 2018;9:978. gene therapy for cancer. Oncol Lett. 2018;16(1):687-702. 24. Finn R, Qin S, Ikeda M, et al. Atezolizumab plus bevacizumab in 2. El-Kenawi A, El-Remessy A. Angiogenesis inhibitors in cancer therapy: unresectable hepatocellular carcinoma. N Engl J Med. 2020;382(20): mechanistic perspective on classification and treatment rationales. Br J 1894-1905. Pharmacol. 2013;170(4):712-729. 25. Socinski MA, Jotte RM, Cappuzzo F, et al. Atezolizumab for first-line 3. Fukumura D, Kloepper J, Amoozgar Z, Duda DG, Jain RK. Enhancing treatment of metastatic nonsquamous NSCLC. N Engl J Med. cancer immunotherapy using antiangiogenics: opportunities and 2018;378(24):2288-2301. challenges. Nat Rev Clin Oncol. 2018;15(5):325-340. 26. Girard N, Wermke M, Barlesi F, et al. Phase Ib study of BI 836880 4. Garcia J, Hurwitz HI, Sandler AB, et al. Bevacizumab (avastin) in cancer (VEGF/Ang2 nanobody) plus ezabenlimab (BI 754091; anti-PD-1 anti- treatment: a review of 15 years of clinical experience and future body) in patients (pts) with solid tumors. J Clin Oncol. 2021;39(suppl outlook. Cancer Treat Rev. 2020;86:102017. 15):Abstr2579. Volume 7 - Issue 5 - 2022 https://doi.org/10.1016/j.esmoop.2022.100576 9 ESMO Open C. Le Tourneau et al. 27. Girard N, Wermke M, Barlesi F, et al. PD1-1-1 Phase Ib study of BI 29. Hussein MA, Bendell JC, Arkenau HT, et al. Platform trial of BI 836880 (VEGF/Ang2 nanobody®) plus ezabenlimab (BI 754091, anti- 754091, an anti-PD-1 antibody, in patients with previously treated PD-1 antibody) in patients with solid tumors. Ann Oncol. 2021;32:S325. advanced solid tumors: combination with BI 836880, a VEGF/ 28. Hussein MA, Percent IJ, Bendell JC, et al. Platform trial of ezabenli- Ang2-blocking nanobody. J Clin Oncol. 2021;39(suppl 3):Abstr mab (BI 754091), an anti-PD-1 antibody, in patients (pts) with pre- TPS152. viously treated advanced solid tumors: combination with BI 836880, 30. Patel M, Johnson M, Winer I, et al. 542P Ezabenlimab (BI 754091) a VEGF/Ang2-blocking nanobody. J Clin Oncol. 2021;39(suppl 15): monotherapy in patients (pts) with advanced solid tumours. Ann Abstr2582. Oncol. 2021;32(5):S361-S1344. 10 https://doi.org/10.1016/j.esmoop.2022.100576 Volume 7 - Issue 5 - 2022