An investigator-initiated phase I study of ONX-0801, a first-in-class alpha folate receptor targeted, small molecule thymidylate synthase inhibitor in solid tumors

2503

Background: ONX-0801 is a first-in-class alpha folate receptor (AFR) targeted thymidylate synthase inhibitor, engineered to differentially accumulate 6000-fold in AFR overexpressing cancer cells. Methods: A 3+3 dose escalation design was used and two IV schedules were explored. Schedule A, weekly dosing (QW) and schedule B, once every 2 weeks dosing (Q2W). A cycle consisted of 4 weeks and treatment was stopped after 6 cycles in both schedules. An expansion cohort to evaluate clinical activity in patients with AFR overexpressing high grade serous ovarian cancer (HGSOC) was planned. Results: 21 patients each were treated in schedule A and B exploring doses ranging from 1-6 mg/m2 and 2-12 mg/m2, respectively. The dose limiting toxicity on schedule A was G3 cellulitis; no dose limiting toxicity was seen on schedule B. The most common toxicities were fatigue 15/42 (36%), nausea 9/42 (21%) and dysgeusia 5/42 (12%). Within schedule A at 4 mg/m2, 2 patients developed suspected drug-related changes on pulmonary function tests (drop in Dlco > 15%) at cycles 5 and 6, respectively. No cases of suspected drug-related drop in Dlco were noted in patients treated in schedule B. No grade 3-4 diarrhea, mucositis or neutropenia were seen in either cohort. The Cmax, AUC and half-life at 12 mg/m2 were 4952 ng/mL, 85170 h*ng/mL and 26 h, respectively. Pre-clinical PK-PD modelling aimed to achieve concentrations between 0.05-1 µM and this was achieved for periods of 48 h at doses of 4 mg/m2and above. Based on safety and PK, the recommended phase II dose (RP2D) of ONX-0801 was 12 mg/m2 Q2W and an expansion in patients with HGSOC is ongoing. 5 patients with HGSOC had partial responses (PRs) in the dose escalation cohort. In the current expansion cohort in patients with HGSOC, 5/11 patients had PRs. Archival samples have been analyzed from 8/11 patients in the expansion cohort. 4/4 AFR+ve and 4/4 AFR-ve patients did and did not have a PR following treatment with ONX-0801, respectively. Conclusions: The RP2D of ONX-0801 is 12 mg/m2 Q2W. At the RP2D, multiple patients with AFR overexpressing HGSOC have had PRs and further randomized biomarker prespecified phase II trials are warranted. Clinical trial information: NCT02360345.

A first in human phase I study of AZD8186, a potent and selective inhibitor of PI3K in patients with advanced solid tumours as monotherapy and in combination with the dual mTORC1/2 inhibitor vistusertib (AZD2014) or abiraterone acetate

2570

Background: Loss of PTEN function leads to increased PI3Kβ signalling. AZD8186 (AZD) exhibits significant anti-tumour activity in PTEN-deficient preclinical models, particularly when combined with anti-androgens or the dual mTORC1/2 inhibitor vistusertib (AZD2014). Here we report on the dose-finding part of this Phase 1 study. Methods: AZD single agent was administered twice daily (BD) in 3 different schedules (5 days on/ 2 days off, 2 days on/ 5 days off and continuous). Escalating doses of AZD were evaluated in cohorts of 3-6 patients treated until progression, unacceptable toxicity, or consent withdrawal. Accrual is ongoing in the combination arms with vistusertib or abiraterone acetate. Results: As of 16 Jan 2017, 87 patients have received AZD at doses of 30–360 mg BD, with 28 confirmed as PTEN deficient (IHC). The selected RP2D for the 5 days on/2 days off monotherapy schedule is 60 mg BD. PK parameters show that systemic exposures to AZD and its major active metabolite increase in a dose proportional manner. 69 serious adverse events (SAEs) were reported in 31 patients on AZD monotherapy with 23 SAEs considered possibly related to AZD. In the 5/2 schedule: 5 dose limiting toxicities (G3 rash with ≥G2 fever and/or chills) were observed in 5 patients at doses of 120-360mg. Adverse events ≥G1 in > 20% included diarrhoea, nausea, fatigue, LFT elevations and decreased appetite. 20 patients remained on study for > 100 days. Dose-dependent target inhibition has been demonstrated in surrogate tissue (platelets). Evaluation of direct tumour target engagement in paired biopsies is currently ongoing. Preliminary efficacy: Confirmed PRs seen in a CRPC patient (BRCA2 and androgen receptor mutant) treated in combination with vistusertib (on study for 411 days) and in one ongoing monotherapy PTEN-deficient colorectal cancer patient (on study > 329 days). Updated data will be presented. Conclusions: AZD has potential for treatment of PTEN-deficient tumours. Investigation of the safety/tolerability and preliminary efficacy in combination with vistusertib or abiraterone acetate is continuing. Clinical trial information: NCT01884285.

Detecting NTRK, ROS1, and ALK gene fusions in gastrointestinal tumor patients

619

Background: The STARTRK-2 trial is a potentially registration-enabling Phase 2 global basket trial of the tyrosine kinase inhibitor entrectinib in patients with solid tumors harboring NTRK1, NTRK2, NTRK3, ROS1, or ALK gene fusions. Phase 1 studies of entrectinib reported a 79% ORR across multiple histologies in patients with gene fusions of these targets who were naïve to inhibitors of these targets, received an effective dose, and had extracranial disease. Patients harboring these fusions are typically rare in the cancer population ( < 3%); however, they have been seen across a wide variety of GI tumors. In this presentation we report on the occurrence of these fusions in GI tumors, based on Ignyta’s internal and partner testing. We discuss the development of an assay for GI samples and immunohistochemistry (IHC) screening efficiency rates in a clinical cohort of 157 GI samples. Methods: The occurrence of NTRK, ROS1, and ALK gene fusions in GI tumors was studied from Ignyta internal and partner collaborations. We developed a 2-step diagnostic test to identify fusions in FFPE clinical specimens. IHC screening using a pan-receptor tyrosine kinase cocktail of antibodies precedes an RNA-based anchored multiplex PCR next generation sequencing (NGS) assay. Results: IHC positivity rates ranged from 0 to 24 percent depending on histology and averaging 8% positivity rate across 15 tumor locations. The screen is particularly effective in cholangiocarcinoma, colon, esophagus, gastric, peritoneum, rectum and small bowel. It was highest in pancreas (24%). Out of 157 GI samples, no instances were detected where the IHC was screened negative and gene fusions were observed by NGS. The IHC cocktail enriches detection of the patient population for gene fusions over 11-fold and has a 100% negative predictive value. Conclusions: This 2-step assay is an efficient method for detection of gene rearrangements in both high-volume clinical testing and studies of archival formalin-fixed paraffin-embedded specimens. Depending on how it is implemented in clinical trials, the staged approach can substantially decrease the costs and tissue expended for NGS. Screening efficiency may be improved even further as additional diagnostic antibodies become available. Clinical trial information: NCT02568267.

Neuroendocrine and epithelial phenotypes in small-cell lung cancer: implications for metastasis and survival in patients

BACKGROUND

Small-cell lung cancer (SCLC) has a very aggressive clinical course with early metastasis. This study investigated how the distinctive neuroendocrine characteristics contribute to disease progression and invasion in human SCLC.

METHODS

The neuroendocrine phenotype (pro-opiomelanocortin (POMC)) was quantified by ELISA in blood samples from 43 SCLC patients. The neuroendocrine (POMC, chromogranin A, neuron-specific enolase, NCAM) and epithelial (cytokeratin and E-cadherin) phenotypes were investigated, using ELISA and immunocytochemistry/immunohistochemistry.

RESULTS

In SCLC patients, 16% had elevated circulating POMC, which was associated with significantly worse survival (P=0.02) and liver metastases (P=0.004). In addition, POMC correlated with epithelial-positive circulating tumour cells (P=0.0002). In a panel of SCLC cell lines, all POMC-secreting cell lines expressed cytokeratin (40% of total). Even after cloning, DMS 79 cells expressed both neuroendocrine and epithelial markers. DMS 79 xenografts secreted POMC into the blood, which mirrored the tumour volume. These xenografts expressed both neuroendocrine and epithelial phenotypes in all tumours, with both phenotypes prevalent in cells invading the surrounding tissue.

CONCLUSION

Both neuroendocrine and epithelial phenotypes coexist in human SCLC tumours in vitro and in vivo and this persists in invading tumour cells. In patients, POMC secretion predicts poor survival and liver metastases, suggesting a crucial role of the neuroendocrine phenotype.

Phase I study to determine the bioavailability and tolerability of a tablet formulation of the PARP inhibitor olaparib in patients with advanced solid tumors: Dose-escalation phase

3051

Background: We previously reported the comparative bioavailability of the olaparib tablet (TAB) up to 200 mg BID, with the initial capsule formulation; gmean AUC0–T following 200 mg BID TAB was ~20% lower than 400 mg BID capsule (CAP; Molife et al ASCO 2010). Olaparib 200 mg BID TAB had an acceptable tolerability profile suggesting further dose escalation might be justified. Methods: Study NCT00777582 was amended to include a dose-escalation phase, to define the maximum tolerated dose (MTD) and safety profile of the TAB, and 2 expansion phases to compare safety and efficacy of TAB doses vs 400 mg BID CAP. Expansion phase data are reported separately. Here, we report preliminary data from the dose-escalation phase where patients (pts) with advanced solid tumors, ECOG PS 0–2 and adequate organ function were assigned to treatment with increasing olaparib BID TAB doses in 28-day continuous dosing cycles. Pharmacokinetic (PK) sampling was performed on days 1, 8, 15, 29, 57. Results: 30 pts (M:F, 3:27; median age 54 yrs [range 19–70]) were enrolled in the dose-escalation phase and received treatment at 5 dose levels: 250, 300, 350, 400 and 450 mg (each dose level, n=6). Overall, the most common AEs were nausea (80%), fatigue (73%) and diarrhea (36%). The majority of AEs were mild to moderate (CTC grade [G] 1/2). Hematologic toxicity in terms of, mostly mild, anemia, neutropenia and thrombocytopenia, appeared to increase at doses of 300 mg BID or higher. Two pts had dose-limiting toxicities at 450 mg (G3 thrombocytopenia and G3 anemia); the TAB MTD was 400 mg BID. Exposure increased proportionally with increasing dose: mean exposure after 400 mg BID was double that following the previously reported 200 mg BID dose. Following 300 and 400 mg doses, gmean Cmax ss, AUC0–T and Cmin ss matched or exceeded the 400 mg BID CAP dose. Conclusions: The MTD of olaparib TAB was 400 mg BID. Based on PK analysis, the 300 and 400 mg BID doses were selected for evaluation in the MTD expansion phase to further define safety and preliminary efficacy.

Phase I dose-finding study of golvatinib (E7050), a c-Met and Eph receptor targeted multi-kinase inhibitor, administered orally QD to patients with advanced solid tumors

3030

Background: Golvatinib is a highly potent, small molecule ATP-competitive inhibitor of the c-Met receptor tyrosine kinase and multiple members of the Eph receptor family as well as c-Kit and Ron, based on isolated kinase assays. Golvatinib showed preclinical evidence of anti-tumor activity. This phase 1 study was performed to determine the MTD, safety, PK, PD and preliminary activity of golvatinib. Methods: Patients (pts) with advanced solid tumors, ECOG PS 0-1, ≥ 18 years (yrs) and adequate organ function were eligible. Golvatinib was administered orally, once daily (QD), continuously. Blood samples for PK and PD analysis were collected at multiple time-points. Mandatory tumor biopsies for PD analysis were taken pre and post Cycle 1 in an expanded MTD cohort. Results: 34 pts (M/F: 21/13; median age 63.5yrs [range 32-78]) were treated at 6 dose levels: 100, 200, 270, 360, 450 and 400 mg. Tumor types were colorectal (n=15), lung (n=4), renal (n=4), esophageal (n=2), melanoma (n=2) and others (n=7). Three DLTs were observed: Gr3 GGT and alkaline phosphatase (n=1; 200mg) and repeated Gr2 (n=1) and Gr3 (n=1) fatigue, both at 450mg. The MTD was determined to be 400 mg QD. Frequently occurring adverse events ([AEs]; all grades) were fatigue: 68% (Gr3: 15%), diarrhea: 65%, nausea: 62%, vomiting: 53%, decreased appetite: 47% (Gr3: 9%), ALT increase: 38% and AST increase: 23%. No Gr4 AEs were observed. Best response was stable disease in 6 pts lasting >84 days. PK showed high variability and plasma concentration increased with dose. The Cmax was reached within a median time of 4 hours. Plasma PD analysis showed an increase in soluble c-Met and Ang 2 levels after golvatinib treatment. Tumor PD analysis in 5 pts at 400 mg demonstrated a baseline elevated MET gene copy number, with c-Met overexpression and post treatment decline in phospho(p)-c-Met expression in 1 pt; post-treatment decline in p-c-Met in a 2nd pt, and post-treatment decline in p-ERK in a 3rd pt. Conclusions: Golvatinib at an MTD of 400 mg QD has manageable toxicity. Preliminary PD analysis demonstrated evidence of c-Met target modulation. Further evaluation will continue in phase 2 combination studies.

First-in-human phase I trial of the dual mTORC1 and mTORC2 inhibitor AZD2014 in solid tumors

3004

Background: AZD2014 is a potent, dual mTORC1/mTORC2 inhibitor with clear activity in in vivo and in vitro experimental models. Methods: This 2-part study consisted of "rolling six" dose escalation (Part A) and expansion (Part B) phases. Part A: 3–6 pts per cohort received an oral solution of AZD2014 BD starting at 50 mg. A further 6 pts were treated in Part A below the MTD to study changes in pharmacodynamic (PD) biomarkers. Part B: additional pts were dosed at the MTD, including a group of ER+/PR+ or HER2+ patients with breast cancer. Primary endpoint: safety and tolerability; secondary endpoints: pharmacokinetics (PK), PD and efficacy. Biomarkers assessed: mTORC1: pS6 (S235/236) and p4EBP1 (T37/46); mTORC2: pAKT (S473). Results: 50 pts have been enroled in this ongoing study and interim data are reported: Part A, n=23 (25 mg, n=6; 50 mg, n=8; 70 mg, n=5; 100 mg, n=4; all BD); Part B, n=27. The MTD was 50 mg BD. DLTs were seen at both 100 mg (Gr 2 and 3 lethargy/fatigue, n=4/4) and 70 mg (Gr 3 mucositis, Gr 2 lethargy, n=2/4); no DLTs were seen at 25 mg or 50 mg. The most common AEs in order of incidence were fatigue, stomatitis, decreased appetite, nausea and diarrhea. Seven SAEs (nausea, vomiting, lethargy, abdominal pain, mucositis) reported by 3 pts were considered ‘possibly related’ to the study drug by investigators. AZD2014 is rapidly absorbed following single and BD multiple doses with a short t1/2 of ~3 h. At 50 mg (n=32), preliminary data show geometric mean AUCss=7.4 µg.h/ml and Cmax ss=1.7 µg/ml. One pt with acinar pancreatic cancer had a RECIST partial response. pAKT and p4EBP1 reductions were observed between 2–8 h in platelet-rich plasma and peripheral blood mononuclear cells respectively. Target modulation in paired tumor biopsies was seen at the MTD. Reduction in the phosphorylation of S6 and 4EBP1 was evident in 8/10 and 3/9 paired biopsies respectively. pAKT was reduced in 3/6 evaluable paired biopsies. As opposed to rapalogs, pAKT was not upregulated in any of the evaluable post-treatment biopsies. Conclusions: The MTD for AZD2014 is 50 mg BD. Further clinical evaluation of AZD2014 is now warranted based on the safety, PK and proof-of-mechanism PD data, as well as its preliminary clinical activity. Updated results will be presented.

A small molecule inhibitor of XIAP induces apoptosis and synergises with vinorelbine and cisplatin in NSCLC

Background:

Evasion of apoptosis contributes to the pathogenesis of solid tumours including non-small cell lung cancer (NSCLC). Malignant cells resist apoptosis through over-expression of inhibitor of apoptosis proteins (IAPs), such as X-linked IAP (XIAP).

Methods:

A phenylurea-based small molecule inhibitor of XIAP, XIAP antagonist compound (XAC) 1396-11, was investigated preclincally to determine its ability to sensitise to clinically relevant cytotoxics, potentially allowing dose reduction while maintaining therapeutic efficacy.

Results:

XIAP protein expression was detected in six NSCLC cell lines examined. The cytotoxicity of XAC 1396-11 against cultured NSCLC cell lines in vitro was concentration- and time-dependent in both short-term and clonogenic assays. XAC 1396-11-induced apoptosis was confirmed by PARP cleavage and characteristic nuclear morphology. XAC 1396-11 synergised with vinorelbine±cisplatin in H460 and A549 NSCLC cells. The mechanism of synergy was enhanced apoptosis, shown by increased cleavage of caspase-3 and PARP and by the reversal of synergy by a pan-caspase inhibitor. Synergy between XAC 1396-11 and vinorelbine was augmented by optimising drug scheduling with superior effects when XAC 1396-11 was administered before vinorelbine.

Conclusion:

These preclinical data suggest that XIAP inhibition in combination with vinorelbine holds potential as a therapeutic strategy in NSCLC.

The effect of homogenates and excretory/secretory products of Nippostrongylus brasiliensis and of acetylcholinesterase on the amplitude and frequency of contraction of uninfected rat intestine in vitro

Whole worm homogenates and excretory/secretory (E/S) products of adult Nippostrongylus brasiliensis significantly decreased the amplitude of contractions of segments of uninfected rat intestine maintained in an organ bath, with significantly larger volumes of E/S products being required to bring about a similar decrease to that caused by the whole worm homogenate. Boiled samples of homogenate and E/S products significantly decreased the amplitude of contractions of uninfected rat intestine, but larger volumes were needed than with unboiled samples. Frequency of contraction was unaltered by homogenates or E/S products. When electric eel AChE was injected into the lumen of segments of uninfected rat intestine maintained in an organ bath there was no significant decrease in the amplitude of contractions. These results suggest that substances present in the E/S products of N. brasiliensis significantly reduce the amplitude of contractions of uninfected rat intestine in vitro and that the biochemical holdfast responsible for this phenomenon may not be AChE.