Comparison of sequential and joint nonlinear mixed effects modeling of tumor kinetics and survival following Durvalumab treatment in patients with metastatic urothelial carcinoma

Standard endpoints such as objective response rate are usually poorly correlated with overall survival (OS) for treatment with immune checkpoint inhibitors. Longitudinal tumor size may serve as a more useful predictor of OS, and establishing a quantitative relationship between tumor kinetics (TK) and OS is a crucial step for successfully predicting OS based on limited tumor size measurements. This study aims to develop a population TK model in combination with a parametric survival model by sequential and joint modeling approaches to characterize durvalumab phase I/II data from patients with metastatic urothelial cancer, and to evaluate and compare the performance of the two modeling approaches in terms of parameter estimates, TK and survival predictions, and covariate identification. The tumor growth rate constant was estimated to be greater for patients with OS ≤ 16 weeks as compared to that for patients with OS > 16 weeks with the joint modeling approach (k_g= 0.130 vs. 0.0551 week^-1, p-value < 0.0001), but similar for both groups (k_g = 0.0624 vs.0.0563 week^-1, p-value = 0.37) with the sequential modeling approach. The predicted TK profiles by joint modeling appeared better aligned with clinical observations. Joint modeling also predicted OS more accurately than the sequential approach according to concordance index and Brier score. The sequential and joint modeling approaches were also compared using additional simulated datasets, and survival was predicted better by joint modeling in the case of a strong association between TK and OS. In conclusion, joint modeling enabled the establishment of a robust association between TK and OS and may represent a better choice for parametric survival analyses over the sequential approach.

Nonlinear Population Pharmacokinetics of Anifrolumab in Healthy Volunteers and Patients With Systemic Lupus Erythematosus

We characterized the population pharmacokinetics of anifrolumab, a type I interferon receptor–blocking antibody. Pharmacokinetic data were analyzed from the anifrolumab (intravenous [IV], every 4 weeks) arms from 5 clinical trials in patients with systemic lupus erythematosus (SLE) (n = 664) and healthy volunteers (n = 6). Population pharmacokinetic modeling was performed using a 2‐compartment model with parallel linear and nonlinear elimination pathways. The impact of covariates (demographics, interferon gene signature [IFNGS, high/low], disease characteristics, renal/hepatic function, SLE medications, and antidrug antibodies) on pharmacokinetics was evaluated. Time‐varying clearance (CL) was characterized using an empirical sigmoidal time‐dependent function. Anifrolumab exposure increased more than dose‐proportionally from 100 to 1000 mg IV every 4 weeks. Based on population pharmacokinetics modeling, the baseline median linear CL was 0.193 L/day in IFNGS‐high patients and 0.153 L/day in IFNGS‐low/healthy volunteers. After a year, median anifrolumab linear CL decreased by 8.4% from baseline. Body weight and IFNGS were significant pharmacokinetic covariates, whereas age, sex, race, disease activity, SLE medications, and presence of antidrug antibodies had no significant effect on anifrolumab pharmacokinetics. Anifrolumab at a concentration of 300 mg IV every 4 weeks was predicted to be below the lower limit of quantitation in 95% of patients ≈10 weeks after a single dose and ≈16 weeks after stopping dosing at steady state. To conclude, anifrolumab exhibited nonlinear pharmacokinetics and time‐varying linear CL; doses ≥300 mg IV every 4 weeks provided sustained anifrolumab concentrations. This study provides further evidence to support the use of anifrolumab 300 mg IV every 4 weeks in patients with moderate to severe SLE.

Exposure-response analysis for selection of optimal dosage regimen of anifrolumab in patients with systemic lupus erythematosus

OBJECTIVES

The randomized, double-blind, phase 2 b MUSE study evaluated the efficacy and safety of the type I IFN receptor antibody anifrolumab (300 mg or 1000 mg every 4 weeks) compared with placebo for 52 weeks in patients with chronic, moderate to severe SLE. Characterizing the exposure-response relationship of anifrolumab in MUSE will enable selection of its optimal dosage regimen in two phase 3 studies in patients with SLE.

METHODS

The exposure-response relationship, pharmacokinetics (PK) and SLE Responder Index (SRI(4)) efficacy data were analysed using a population approach. A dropout hazard function was also incorporated into the SRI(4) model to describe the voluntary patient withdrawals during the 1-year treatment period.

RESULTS

The population PK model found that type I IFNGS-high patients, and patients with a higher body weight, had significantly greater clearance of anifrolumab. Stochastic clinical simulations demonstrated that doses <300 mg would lead to a greater-than-proportional reduction in drug exposure owing to type I IFN alpha receptor-mediated drug clearance (antigen-sink effect, more rapid drug clearance at lower concentrations) and suboptimal SRI(4) responses with wider confidence intervals.

CONCLUSIONS

Based on PK, efficacy and safety considerations, anifrolumab 300 mg every 4 weeks was recommended as the optimal dosage for pivotal phase 3 studies in patients with SLE.

Pharmacokinetic and Pharmacodynamic Modeling of Tezepelumab to Guide Phase 3 Dose Selection for Patients With Severe Asthma

Tezepelumab is a human monoclonal antibody that blocks thymic stromal lymphopoietin, an epithelial cytokine involved in asthma pathogenesis. In the phase 2b PATHWAY study (ClinicalTrials.gov identifier: NCT02054130), tezepelumab significantly reduced exacerbations in adults with severe, uncontrolled asthma. We used pharmacokinetic (PK) and pharmacodynamic (PD) modeling to guide tezepelumab dose selection for phase 3 trials in patients with severe asthma. PK data from 7 clinical studies were used to develop a population PK model. Population PK-PD models were developed to characterize the relationship between tezepelumab PK and asthma exacerbation rate (AER) and fractional exhaled nitric oxide (FeNO) levels (using phase 2b PD data only). Tezepelumab PK were well described by a 2-compartment model with first-order absorption; PK parameter estimates were consistent with those of other immunoglobulin G2 antibodies. PK-PD models predicted that subcutaneous dosing at 210 mg every 4 weeks was associated with ≈90% of the maximum drug effect of tezepelumab on AER and FeNO; further dose increases were not expected to result in additional, clinically meaningful treatment benefit. No clinically significant covariates of treatment effects on AER and FeNO were identified. Population PK simulations, exposure-response relationships and safety profiles of tezepelumab at doses up to 280 mg every 2 weeks suggested that no dose adjustment based on body weight or for adolescents was required. These results support the selection of 210 mg every 4 weeks subcutaneously as the dose for phase 3 studies of tezepelumab in adults and adolescents with severe asthma.

A Novel Pharmacodynamic Biomarker and Mechanistic Modeling Facilitate the Development of Tovetumab, a Monoclonal Antibody Directed Against Platelet-Derived Growth Factor Receptor Alpha, for Cancer Therapy

Tovetumab (MEDI-575) is a fully human IgG2κ monoclonal antibody that specifically binds to human platelet-derived growth factor receptor alpha (PDGFRα) and blocks receptor signal transduction by PDGF ligands. The affinity of tovetumab determined using surface plasmon resonance technology and flow cytometry demonstrated comparable binding affinity for human and monkey PDGFRα. In single and repeat-dose monkey pharmacokinetic-pharmacodynamic (PK-PD) studies, tovetumab administration resulted in dose-dependent elevation of circulating levels of PDGF-AA, a member of the PDGF ligand family, due to displacement of PDGF-AA from PDGFRα by tovetumab and subsequent blockade of PDGFRα-mediated PDGF-AA degradation. As such, PDGF-AA accumulation is an indirect measurement of receptor occupancy and is a novel PD biomarker for tovetumab. The nonlinear PK of tovetumab and dose-dependent increase in circulating PDGF-AA profiles were well described by a novel mechanistic model, in which tovetumab and PDGF-AA compete for the binding to PDGFRα. To facilitate translational simulation, the internalization half-lives of PDGF-AA and tovetumab upon binding to PDGFRα were determined using confocal imaging to be 14 ± 4 min and 30 ± 8 min, respectively. By incorporating PDGFRα internalization kinetics, the model not only predicted the target receptor occupancy by tovetumab, but also the biologically active agonistic ligand-receptor complex. This work described a novel PD biomarker approach applicable for anti-receptor therapeutics and the first mechanistic model to delineate the in vivo tri-molecular system of a drug, its target receptor, and a competing endogenous ligand, which collectively have been used for optimal dose recommendation supporting clinical development of tovetumab.

Predicting overall survival (OS) and overall response (OR) following durvalumab treatment in patients with multiple cancer types using a hybrid modeling strategy

e15161

Background: This study aimed to predict OS/OR and identify key predictors in patients with diverse cancer types treated with durvalumab, a PD-L1 targeting monoclonal antibody, using a hybrid modeling strategy that combines population pharmacodynamic (PD) modeling and machine learning (ML) algorithms. Methods: Individual longitudinal tumor size measurements and OS/OR data were available for 855 patients who received durvalumab therapy (10 mg/kg Q2W or 20 mg/kg Q4W; NCT01693562). Nine cancer types included non-small cell lung cancer (NSCLC), bladder cancer (BC), microsatellite instability-high (MSI-H) cancer, hepatocellular carcinoma (HCC), squamous cell carcinoma of the head and neck (SCCHN), gastroesophageal cancer (GEC), ovarian cancer (OC), pancreatic adenocarcinoma (PDAC) and triple-negative breast cancer (TNBC). A tumor kinetic model was developed to characterize diverse temporal profiles using a population-based modeling approach. Individual estimated tumor kinetic model parameters and patient demographic/physiological factors were used as inputs for predicting OS/OR using several ML approaches. Results: The final tumor kinetic model with liver metastasis (LM), neutrophil/lymphocyte ratio (NLR), tumor size at baseline (TBSL) and cancer types as covariates characterized the temporal tumor size data well. HCC and MSI-H cancer have the slowest tumor growth rate constant (kg), while GEC, SCCHN and TNBC have the fastest kg. BC, NSCLC and OC have the highest tumor killing rate constant. The most important predictors of OS identified by ML approach were tumor kinetic parameters (kg, fraction of drug-sensitive cells, time-delay in immune response), along with baseline disease factors, including hemoglobin (HGBBL), albumin (ALB), and NLR. Decision tree-based algorithms showed the best performance in predicting OR with accuracy above 90%. In addition to tumor kinetic parameters, PD-L1 expression on tumor cells (TC) and ALB were the most important predictors of OR. Conclusions: A combined population PD/ML approach showed good predictions of OS/OR in patients with different cancer types treated with durvalumab. LM, NLR,TBSL and cancer types were found to be important factors for tumor kinetics. In addition to tumor kinetic parameters, HGBBL, ALB, and NLR were found to be important predictors of OS, and TC and ALB were found to be important predictors of OR. These findings could provide a guidance on patient selection in future clinical trials.

Overall survival modeling and association with serum biomarkers in durvalumab-treated patients with head and neck cancer

6549

Background: Optimal patient selection for immunotherapy remains a challenge as most patients fail to respond. We aim to assess baseline factors for association with long-term survival from durvalumab treatment in patients with recurrent/metastatic head and neck squamous cell carcinoma (HNSCC)1,2. Methods: Pooled longitudinal tumor size, survival, and dropout data from four trials (1108: NCT01693562, CONDOR: NCT02319044 , HAWK: NCT02207530, and EAGLE: NCT02369874) involving 467 HNSCC patients were used to develop tumor size-driven hazard models. A panel of 66 serum protein biomarkers at baseline and 4 relevant clinical markers from 346 out of 413 patients treated with durvalumab (all studies except 1108) were initially screened to select a pool of 21 candidate covariates. The criteria for dimensionality reduction comprised correlation strength between biomarkers and pharmacological hypotheses pertaining to a prior analysis3 (inflammation, immunomodulation, tumor burden and angiogenesis). Results: The final tumor model highlighted that high tumor burden, elevated LDH and neutrophil-lymphocyte ratio were associated with faster tumor growth while patients with lower baseline tumor burden had an increase in net tumor shrinkage. For overall survival, the model suggested that high levels of immunomodulators (IL23, Osteocalcin), low inflammation (IL6, NLR), low tumor burden, and low angiogenesis factors (von Willebrand factor (vWF), plasminogen activator inhibitor-1 (PAI-1)) were associated with survival benefits for patients treated with durvalumab. Specifically, these patients had baseline serum IL23 > 2.1 pg/mL and Osteocalcin > 32 pg/mL or serum PAI-1 < 229 pg/mL and serum IL6 < 5.4 pg/mL which corresponded to a hazard ratio estimate (HR and 95%CI) of 0.36 (0.27- 0.47), logrank p-value: 2.3x10−14. The median (n, 95%CI) overall survival time for the patients with favorable biomarker profile was 14.6 months (n = 129, 11.2-21.4) vs. 4.4 months (n = 217, 3.6-5.3). Conclusions: Our results corroborate the prior hypothesis highlighting the prognostic value of inflammation, disease burden, tumor angiogenesis, and immunomodulatory factors on the clinical outcomes of HNSCC patients treated with durvalumab3. Collectively, we identified a serum biomarker profile of HNSCC patients with median survival times exceeding 1 year which may potentially be used for patient enrichment following further validation in prospective studies. References: 1Yanan CPT 2017, 2Baverel, 2018 ENA, 3Guo, X, 2019 Asco P6048 Clinical trial information: NCT01693562, NCT02319044, NCT02207530, NCT02369874 .

Population pharmacokinetics, efficacy, and safety of moxetumomab pasudotox in patients with relapsed or refractory hairy cell leukaemia

AIMS

To characterize the pharmacokinetics (PK) of moxetumomab pasudotox, an anti-CD22 recombinant immunotoxin, in adults with relapsed or refractory hairy cell leukaemia, we examined data from a phase 1 study (Study 1001; n = 49) and from the pivotal clinical study (Study 1053; n = 74).

METHODS

Data from both studies were pooled (n = 123) to develop a population PK model. Covariates included demographics, disease state, liver and kidney function, prior treatment, and antidrug antibodies (ADAs). Exposure-response and exposure-safety were analysed separately by study. A 1-compartment model with linear elimination from the central compartment and 2 clearance (CL) rates was developed.

RESULTS

Moxetumomab pasudotox was cleared more rapidly after cycle 1, day 1 (CL₁ = 24.7 L/h) than subsequently (CL₂ = 3.76 L/h), with high interindividual variability (116 and 109%, respectively). In Study 1053, patients with ADA titres >10 240 showed ~4-fold increase in CL. Higher exposures (≥median) were related to higher response rates, capillary leak syndrome and increased creatinine (Study 1053 only), or grade ≥3 adverse events (Study 1001 only). Clinical benefits were still observed in patients with lower exposure or high ADA titres.

CONCLUSION

Despite a high incidence of immunogenicity with increased clearance, moxetumomab pasudotox demonstrated efficacy in hairy cell leukaemia.

1557. Population Pharmacokinetics of Suvratoxumab (MEDI4893), an Extended Half-life Staphylococcus aureus Alpha Toxin-Neutralizing Human Monoclonal Antibody, in Healthy Adults and Patients on Mechanical Ventilation in Intensive Care Units

Background

Suvratoxumab (suvra), an extended half-life (~80 days), Staphylococcus aureus (SA) alpha toxin-neutralizing IgG monoclonal antibody, is under investigation for prevention of SA pneumonia in patients on mechanical ventilation (MV). We characterized the serum PK of suvra using population pharmacokinetics (popPK) in both healthy volunteers and MV patients and quantified the proportion of patients reaching the serum target of 211 μg/mL at 30 days post-dose.

Methods

The popPK analysis included 1,368 serum samples from two early phase studies (NCT02296320; EudraCT 2014-001097-34): (1) Phase 1 study in 26 healthy adults receiving single IV suvra doses ranging from 0.225g to 5g, with PK sampled up to 360 days; and (2) Phase 2 study in MV patients with PCR-confirmed SA colonization of lower respiratory tract receiving one suvra IV dose of 2g (n = 15) or 5g (n = 96), with PK sampled up to 100 days.

Results

A two-compartment linear model with weight-based scaling of the PK parameters adequately described the serum PK data (Figure 1). MV status, number of days on MV, and age impacted the PK of suvra. A moderate between-subject variability (<45% CV) was estimated for key PK parameters. An estimated two-fold increase in MV patients’ volume of distribution parameters compared with healthy volunteers explained the observed C_max differences between the two groups (1145±369 μg/mL vs. 1783±396 μg/mL) (Figures 2 and 3). Although age, MV status and days on MV post-dose appeared to be associated with higher systemic clearance (CL) in the model, this estimate could be biased due to limited PK data available for only one half-life (~90 days) of the drug in MV patients (Figure 2). More patients achieved suvra levels above the PK target following the 5 g (73.5%; 50/68) vs. 2 g dose (7.6%; 1/13) at 30 days post-dose.

Conclusion

MV status, post-dose duration on MV, body weight, and age were identified as statistically significant covariates influencing the PK of suvra. Serum PK and popPK analyses support the 5g dose for future studies with suvra in MV patients.

Disclosures

All authors: No reported disclosures.

Comparison of autoinjector with accessorized prefilled syringe for benralizumab pharmacokinetic exposure: AMES trial results

OBJECTIVE

We compared the pharmacokinetic exposure following a single subcutaneous dose of benralizumab 30 mg using either autoinjectors (AI) or accessorized prefilled syringes (APFS). APFS and AI functionality and reliability for at-home benralizumab delivery have been demonstrated in the GREGALE and GRECO studies, respectively.

METHODS

In the open-label AMES study (NCT02968914), 180 healthy adult men and women were randomized to one of two device (AI or APFS) and three injection site (upper arm, abdomen, or thigh) combinations. Randomization was stratified by weight (<70 kg, 70-84.9 kg, and ≥85 kg). Blood eosinophil counts were measured on Days 1, 8, 29, and 57.

RESULTS

Benralizumab pharmacokinetic exposure was similar between AI and APFS. Geometric mean ratios (AI/APFS) (90% CI) were 92.8% (87.4-98.6) and 94.5% (88.2-101.2) for two area under the concentration‒time curve measurements (AUC_last and AUC_inf). Benralizumab exposure was approximately 15-30% greater for thigh vs. abdomen or upper arm administration. Exposure was slightly greater for APFS vs. AI regardless of injection site or weight class. These differences were unlikely to be clinically relevant, as eosinophil depletion was achieved consistently with both devices at all injection sites. No device malfunctions were reported. No new or unexpected safety findings were observed.

CONCLUSION

Benralizumab pharmacokinetic exposure was similar between AI and APFS, with consistent blood eosinophil count depletion observed with both devices. These results support benralizumab administration with either AI or APFS, providing patients and physicians increased choice, flexibility, and convenience for potential at-home delivery.

A QSP Model for Predicting Clinical Responses to Monotherapy, Combination and Sequential Therapy Following CTLA-4, PD-1, and PD-L1 Checkpoint Blockade

Over the past decade, several immunotherapies have been approved for the treatment of melanoma. The most prominent of these are the immune checkpoint inhibitors, which are antibodies that block the inhibitory effects on the immune system by checkpoint receptors, such as CTLA-4, PD-1 and PD-L1. Preclinically, blocking these receptors has led to increased activation and proliferation of effector cells following stimulation and antigen recognition, and subsequently, more effective elimination of cancer cells. Translation from preclinical to clinical outcomes in solid tumors has shown the existence of a wide diversity of individual patient responses, linked to several patient-specific parameters. We developed a quantitative systems pharmacology (QSP) model that looks at the mentioned checkpoint blockade therapies administered as mono-, combo- and sequential therapies, to show how different combinations of specific patient parameters defined within physiological ranges distinguish different types of virtual patient responders to these therapies for melanoma. Further validation by fitting and subsequent simulations of virtual clinical trials mimicking actual patient trials demonstrated that the model can capture a wide variety of tumor dynamics that are observed in the clinic and can predict median clinical responses. Our aim here is to present a QSP model for combination immunotherapy specific to melanoma.

Abstract 3158: Prediction of overall survival in urothelial cancer patients using tumor sizes and baseline risk factors: longitudinal modeling approach for durvalumab and durvalumab + tremelimumab

Background: Durvalumab [D] is a human mAb that binds to PD-L1 and blocks its interaction with PD-1 and CD80. Tremelimumab [T] blocks the inhibitory effects of CTLA-4, and therefore enhances T-cell activation. The objectives of this analysis were to develop a model linking overall survival (OS) to baseline risk factors and changes in tumor size during treatment to identify key factors impacting response to D or D +T.

Methods: The analysis dataset included UC patients from two clinical trials: Study 1108 (D 10 mg/kg Q2W; n=201) and Study 10 (D 20 mg/kg Q4W + T 1 mg/kg Q4W for 4 doses, followed by D 20 mg/kg Q4W alone; n=168). Longitudinal tumor size data were analyzed using a nonlinear mixed effect model with key parameters describing tumor growth, tumor killing, and delay in immune response. Subsequently, a parametric survival model was developed to link baseline risk factors and predicted percent change in tumor size at week 8 to OS.

Results: Tumor kinetic model adequately described the longitudinal tumor size data from UC patients. Baseline tumor size (p&lt;0.01) and PD-L1 status (p&lt;0.01) were identified as significant covariates for tumor killing rate. The most influential factor associated with faster tumor growth was liver metastasis (p&lt;0.01), while higher hemoglobin levels (p&lt;0.01) were associated with decreased tumor growth rate. Based on parametric survival modeling, liver metastasis (~34% decrease in OS, p&lt;0.0001), albumin (~ 1-fold increase in OS per 1g/dL increase, p&lt;0.0001), and percent change in tumor size at week 8 (~52% increase in OS with 30% tumor shrinkage at week 8, p&lt;0.0001) were found to be significant and clinically relevant predictors of OS.

Conclusions: The parametric survival model coupled with tumor kinetic model adequately described clinical outcomes in UC patients treated with D or D+T and enabled identification of key factors potentially impacting response to immune therapy in UC. This approach can be a useful tool for guiding patient selection/enrichment strategies and optimizing trial designs for immuno-oncology (IO) therapies. Further validation and prospective evaluation of this model may be conducted in other IO trials.

Citation Format: Mei Tang, Yu Jiang, Han Si, Yanan Zheng, Chen Gao, Guozhi Gao, Natasha Angra, Shaad Abdullah, Brandon Higgs, Lorin Roskos, Rajesh Narwal. Prediction of overall survival in urothelial cancer patients using tumor sizes and baseline risk factors: longitudinal modeling approach for durvalumab and durvalumab + tremelimumab [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3158.

A Computational Model of Neoadjuvant PD-1 Inhibition in Non-Small Cell Lung Cancer

Immunotherapy and immune checkpoint blocking antibodies such as anti-PD-1 are approved and significantly improve the survival of advanced non-small cell lung cancer (NSCLC) patients, but there has been little success in identifying biomarkers capable of separating the responders from non-responders before the onset of the therapy. In this study, we developed a quantitative system pharmacology (QSP) model to represent the anti-tumor immune response in human NSCLC that integrated our knowledge of tumor growth, antigen processing and presentation, T cell activation and distribution, antibody pharmacokinetics, and immune checkpoint dynamics. The model was calibrated with the available data and was used to identify potential biomarkers as well as patient-specific response based on the patient parameters. The model predicted that in addition to tumor mutational burden (TMB), a known biomarker for anti-PD-1 therapy in NSCLC, the number of effector T cells and regulatory T cells in the tumor and blood is a predictor of the responders. Furthermore, the model simulated a set of 12 patients with known TMB and MHC/antigen-binding affinity from a recent clinical trial (ClinicalTrials.gov number, NCT02259621) on neoadjuvant nivolumab therapy in resectable lung cancer and predicted an augmented durable response in patients with adjuvant nivolumab treatment in addition to the clinical trial protocol of neoadjuvant nivolumab treatment followed by resection. Overall, the model provides a valuable framework to model tumor immunity and response to immune checkpoint blockers to enhance biomarker discovery and performing virtual clinical trials to aid in design and interpretation of the current trials with fewer patients.

Detecting bliss synergy in in vivo combination studies with a tumor kinetic model

Linear models are generally reliable methods for analyzing tumor growth in vivo, with drug effectiveness being represented by the steepness of the regression slope. With immunotherapy, however, not all tumor growth follows a linear pattern, even after log transformation. Tumor kinetics models are mechanistic models that describe tumor proliferation and tumor killing macroscopically, through a set of differential equations. In drug combination studies, although an additional drug-drug interaction term can be added to such models, however, the drug interactions suggested by tumor kinetics models cannot be translated directly into synergistic effects. We have developed a novel statistical approach that simultaneously models tumor growth in control, monotherapy, and combination therapy groups. This approach makes it possible to test for synergistic effects directly and to compare such effects among different studies.

Correlation of angiogenic and immunomodulatory proteins with clinical outcomes of durvalumab (anti-PDL1) in recurrent/metastatic head and neck squamous cell carcinoma

6048

Background: The potential for durvalumab, a PD-L1 blocking monoclonal antibody, to treat head and neck squamous cell carcinoma (HNSCC) is being evaluated in multiple clinical trials. We assessed circulating protein biomarkers in HNSCC patients prior to treatment to better understand pathways related to clinical outcomes and potentially relevant for targeting in combination with durvalumab. Methods: Sixty-six selected serum proteins were measured by multiplex immunoassay at baseline in HNSCC patients receiving durvalumab treatment: 106 patients with high PD-L1 (≥25% tumor cells; SP263 assay) in phase II HAWK trial and 52 patients with low/no PD-L1 in phase II CONDOR trial. Results: Multivariate Cox modeling demonstrated that higher baseline concentrations of angiogenic, pro-inflammatory, and myeloid-associated proteins (ANGPT2, CRP, IL6, S100A12) were associated with shorter overall survival (OS), while higher concentration of a bone formation marker and immunostimulatory hormone (BGLAP) correlated with longer OS in 158 durvalumab-treated HNSCC patients ( P< 0.05). These 4 proteins also showed higher baseline levels in patients with progressive disease (PD) compared to stable disease (SD) and partial or complete responses (PR/CR), while BGLAP had lower levels in PD compared to SD or PR/CR (Mann-Whitney P< 0.05). The 5 proteins remained significantly associated with OS in a multivariate model including PD-L1, ECOG, tumor size, and neutrophil count. Bone metastasis status had no impact on the association of BGLAP with OS, which has not been reported before in HNSCC. Interestingly, ANGPT2 level above median showed the highest hazard ratio (HR = 2.2, P <0.001) among all evaluated variables. Furthermore, higher levels of VWF, an angiogenesis-related protein, correlated with shorter OS by univariate survival analysis ( P < 0.001). Conclusions: Our results suggested an important role of angiogenesis in the resistance of HNSCC patients to durvalumab treatment, and ANGPT2 may have predictive utility for durvalumab combination with an anti-angiogenic agent. The predictive value of BGLAP remains to be evaluated in a randomized clinical study. Clinical trial information: NCT02207530; NCT02319044.

In silico simulation of a clinical trial with anti-CTLA-4 and anti-PD-L1 immunotherapies in metastatic breast cancer using a systems pharmacology model

The low response rate of immune checkpoint blockade in breast cancer has highlighted the need for predictive biomarkers to identify responders. While a number of clinical trials are ongoing, testing all possible combinations is not feasible. In this study, a quantitative systems pharmacology model is built to integrate immune-cancer cell interactions in patients with breast cancer, including central, peripheral, tumour-draining lymph node (TDLN) and tumour compartments. The model can describe the immune suppression and evasion in both TDLN and the tumour microenvironment due to checkpoint expression, and mimic the tumour response to checkpoint blockade therapy. We investigate the relationship between the tumour response to checkpoint blockade therapy and composite tumour burden, PD-L1 expression and antigen intensity, including their individual and combined effects on the immune system, using model-based simulations. The proposed model demonstrates the potential to make predictions of tumour response of individual patients given sufficient clinical measurements, and provides a platform that can be further adapted to other types of immunotherapy and their combination with molecular-targeted therapies. The patient predictions demonstrate how this systems pharmacology model can be used to individualize immunotherapy treatments. When appropriately validated, these approaches may contribute to optimization of breast cancer treatment.

Exposure-Response Analysis of Overall Survival for Tremelimumab in Unresectable Malignant Mesothelioma: The Confounding Effect of Disease Status

Tremelimumab, an anti‐cytotoxic T‐lymphocyte antigen‐4 monoclonal antibody that enhances T‐cell activation, was evaluated in a randomized, double‐blind, placebo‐controlled, phase IIb study (NCT01843374) in patients with unresectable malignant mesothelioma. The study demonstrated no clinically meaningful differences in overall survival (OS). The objective of this analysis was to evaluate the relationship of exposure with OS. A population pharmacokinetic (PK) model adequately described the PK data. Three factors (sex, C‐reactive protein, and baseline tumor size) were identified as statistically significant PK predictors (P < 0.05 on clearance). A positive association between exposure and OS was observed. However, an association between key baseline factors with OS (regardless of treatment) and imbalances in prognostic factors favoring patients with higher exposure (upper vs. lower PK quartile) was seen. Taken together, these results suggest that the exposure OS relationship observed for tremelimumab in mesothelioma is likely spurious rather than a true association of exposure with efficacy.

Relationship Between Benralizumab Exposure and Efficacy for Patients With Severe Eosinophilic Asthma

We evaluated the relationship between benralizumab (30 mg every 4 and 8 weeks (Q4W, Q8W)) pharmacokinetic (PK) exposure and end points of asthma exacerbation rates (AERs) and change from baseline in prebronchodilator forced expiratory volume in 1 second (FEV₁ ) for patients with severe, uncontrolled eosinophilic asthma in the SIROCCO/CALIMA phase III trials. In empirical assessment, AER ratios in SIROCCO were similar across PK quartiles. However, the lowest PK quartile in CALIMA had reduced efficacy; low CALIMA placebo AER possibly confounded this result. In population modeling, estimated benralizumab 90% effective concentration for AER reduction was 927 ng/mL, below the Q8W dosage steady-state average PK concentration (1,066 ng/mL). Benralizumab treatment resulted in more rapid FEV₁ improvement vs. placebo (estimated half-maximum time: 7.6 vs. 18 days); this response was greater for patients with greater baseline eosinophil counts. These results confirmed 30 mg Q8W is the optimal benralizumab dosage for patients with severe eosinophilic asthma.

A computational multiscale agent-based model for simulating spatio-temporal tumour immune response to PD1 and PDL1 inhibition

When the immune system responds to tumour development, patterns of immune infiltrates emerge, highlighted by the expression of immune checkpoint-related molecules such as PDL1 on the surface of cancer cells. Such spatial heterogeneity carries information on intrinsic characteristics of the tumour lesion for individual patients, and thus is a potential source for biomarkers for anti-tumour therapeutics. We developed a systems biology multiscale agent-based model to capture the interactions between immune cells and cancer cells, and analysed the emergent global behaviour during tumour development and immunotherapy. Using this model, we are able to reproduce temporal dynamics of cytotoxic T cells and cancer cells during tumour progression, as well as three-dimensional spatial distributions of these cells. By varying the characteristics of the neoantigen profile of individual patients, such as mutational burden and antigen strength, a spectrum of pretreatment spatial patterns of PDL1 expression is generated in our simulations, resembling immuno-architectures obtained via immunohistochemistry from patient biopsies. By correlating these spatial characteristics with in silico treatment results using immune checkpoint inhibitors, the model provides a framework for use to predict treatment/biomarker combinations in different cancer types based on cancer-specific experimental data.

Population Pharmacokinetics of Durvalumab in Cancer Patients and Association With Longitudinal Biomarkers of Disease Status

The objectives of this analysis were to develop a population pharmacokinetics (PK) model of durvalumab, an anti-PD-L1 antibody, and quantify the impact of baseline and time-varying patient/disease characteristics on PK. Pooled data from two studies (1,409 patients providing 7,407 PK samples) were analyzed with nonlinear mixed effects modeling. Durvalumab PK was best described by a two-compartment model with both linear and nonlinear clearances. Three candidate models were evaluated: a time-invariant clearance (CL) model, an empirical time-varying CL model, and a semimechanistic time-varying CL model incorporating longitudinal covariates related to disease status (tumor shrinkage and albumin). The data supported a slight decrease in durvalumab clearance with time and suggested that it may be associated with a decrease in nonspecific protein catabolic rate among cancer patients who benefit from therapy. No covariates were clinically relevant, indicating no need for dose adjustment. Simulations indicated similar overall PK exposures following weight-based and flat-dosing regimens.

Characterisation of anti-alpha toxin antibody levels and colonisation status after administration of an investigational human monoclonal antibody, MEDI4893, against Staphylococcus aureus alpha toxin

Objectives

MEDI4893 is a novel, long‐acting human monoclonal antibody targeting Staphylococcus aureus (SA) alpha toxin (AT). This report presents the results of the exploratory analyses from a randomised phase 1 dose‐escalation study in healthy human subjects receiving single intravenous MEDI4893 doses or placebo.

Methods

Anti‐AT antibodies and AT expression were measured as described previously. Nasal swabs were analysed by culture and PCR. Data were summarised by treatment groups and visits by using SAS System Version 9.3.

Results

Subjects receiving 2250 or 5000 mg of MEDI4893 had the highest serum anti‐AT neutralising antibody (NAb) levels: approximately 180‐ to 240‐, 70‐ to 100‐ and sevenfold to 10‐fold higher than respective baseline levels at peak, 30 and 360 days, respectively. In these subjects, levels of serum anti‐AT NAbs were >3.2 International Units (IU) mL⁻¹ for at least 211 days. In the upper respiratory tract, anti‐AT NAb levels increased with MEDI4893 dose. No apparent effect of MEDI4893 on SA nasal colonisation, hla gene sequence or AT expression was observed. Five AT variants were detected, their lytic activity was fully neutralised by MEDI4893.

Discussion

Our results indicate that (1) MEDI4893 administration at 2250 and 5000 mg would provide effective immunoprophylaxis against systemic SA disease; (2) MEDI4983 distributes to the upper respiratory tract and retains neutralising activity against AT; and (3) potential for emergence of MEDI4893 resistance is low.

Conclusion

Intravenous administration of MEDI4893 maintained levels of anti‐AT NAbs in serum and nasal mucosa that may provide effective immunoprophylaxis against SA disease and support continued clinical development of MEDI4893.

Population Modeling of Tumor Kinetics and Overall Survival to Identify Prognostic and Predictive Biomarkers of Efficacy for Durvalumab in Patients With Urothelial Carcinoma

Durvalumab is an anti‐PD‐L1 monoclonal antibody approved for patients with locally advanced or metastatic urothelial carcinoma (UC) that has progressed after platinum‐containing chemotherapy. A population tumor kinetic model, coupled with dropout and survival models, was developed to describe longitudinal tumor size data and predict overall survival in UC patients treated with durvalumab (NCT01693562) and to identify prognostic and predictive biomarkers of clinical outcomes. Model‐based covariate analysis identified liver metastasis as the most influential factor for tumor growth and immune‐cell PD‐L1 expression and baseline tumor burden as predictive factors for tumor killing. Tumor or immune‐cell PD‐L1 expression, liver metastasis, baseline hemoglobin, and albumin levels were identified as significant covariates for overall survival. These model simulations provided further insights into the impact of PD‐L1 cutoff values on treatment outcomes. The modeling framework can be a useful tool to guide patient selection and enrichment strategies for immunotherapies across various cancer indications.

Model-Based Discovery and Development of Biopharmaceuticals: A Case Study of Mavrilimumab

Drug development is a lengthy, costly process with low probability of success. Biopharmaceuticals are highly specific molecules, with efficacy and safety closely tied to target biology and pharmacology. The “learning−predicting−confirming” continuum by translational and clinical modeling and simulation (M&S) was implemented at every decision point for mavrilimumab, a human monoclonal antibody in development for rheumatoid arthritis (RA). This tutorial uses mavrilimumab as an example to demonstrate rational discovery, preclinical development, clinical study design, and dose selection of biotherapeutics by M&S.

Abstract 4531: Systems pharmacology to predict cellular biomarkers and optimize mono- and combination-therapy regimens: Focusing on immune checkpoint targets PD-1, PD-L1 and CTLA-4

Cancer immunotherapy focuses on stimulating and promoting the immune system to recognize and eliminate cancer cells, with several FDA approvals in recent years. However, identifying patients best suited for specific immune therapies, and determining optimal treatment regimens continue to be a clinical challenge. Using a molecular-detailed computational systems pharmacology model to identify cellular biomarkers and optimize regimens, we may be able to predict the efficacy of regimens in specific patient populations, and expedite drug development for cancer treatment. We developed a cell/receptor-based multi-compartment systems pharmacology model focusing on the immune response against a growing tumor, with the intent to test the effects of immune checkpoint inhibitors against PD-1, PD-L1 and CTLA-4 administered as mono- and combination therapies. Additionally, the model also allows for testing of other immuno-therapies, such as adoptive cell therapies, which can be combined with the checkpoint inhibitors. The model was designed and developed using the SimBiology plug-in in MATLAB. Simulations were performed with parameters that define the immune response at particular tumor stages of melanoma and NSCLC. All results were qualitatively and quantitatively compared to experimental pre-clinical and clinical data for model validation, or used for the generation of predictions suitable for further experimental testing. In silico, we have identified that administrations of the prescribed doses of 1-10 mg/kg of anti-CLTA-4 (based on binding kinetics) effectively saturates the receptors on the T cells, and promotes both an extended life span of the antigen presenting cells (APCs), and the maximum attainable activation levels of the effector T cells. The model further predicts that the effectiveness of anti-CTLA-4 therapy is limited by the immunogenicity of the system (i.e., the antigen intensity level and number of APCs presenting the antigens) in a monotonic fashion. Furthermore, injecting activated APCs without therapy would show a temporary tumor response and a subsequent recovery by the tumor to its original growth trajectory, while raising the antigen intensity had a sustained effect on tumor response. Other simulations indicate that, despite the lack of apparent tumor response, a sustained immune attack may be ongoing in the body; however, the immune activity is proportionally limited by the tumor and regulatory cells. Lastly, several dose-responses and clinical trials were simulated for both combination and monotherapies, and correlated with published clinical trial data. Future work will focus on uncovering the cellular biomarkers responsible for such results, experimentally validating them, as well as simulating optimal combination treatment regimens for future evaluation.

Citation Format: Oleg Milberg, Chang Gong, Bing Wang, Paolo Vicini, Rajesh Narwal, Lorin Roskos, Aleksander Popel. Systems pharmacology to predict cellular biomarkers and optimize mono- and combination-therapy regimens: Focusing on immune checkpoint targets PD-1, PD-L1 and CTLA-4 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4531. doi:10.1158/1538-7445.AM2017-4531

Abstract 5045: Pharmacokinetics and pharmacodynamics of MEDI0680, a fully human anti-PD1 monoclonal antibody, in patients with advanced malignancies

Background: MEDI0680 (AMP-514) is a humanized immunoglobulin gamma 4, kappa (IgG4κ) monoclonal antibody (mAb) specific for human programmed cell death-1 (PD-1), developed for the treatment of cancer. The primary objectives of this analysis were to (a) describe the pharmacokinetics (PK) of MEDI0680 and quantitate the impact of patient/disease characteristics on PK variability (b) to compare body weight (WT)-based and fixed dosing regimens of MEDI0680 and (c) to characterize PK-pharmacodynamic (receptor occupancy) relationship.

Methods: A total of 905 serum concentration records from 58 patients in Phase 1 study (D6020C00002) designed to evaluate safety, tolerability and PK following 0.1, 0.5, 2.5, 10, and 20 mg/kg every 3 weeks (Q3W), every 2 weeks (Q2W) or weekly doses (QW) as intravenous (IV) infusion of MEDI0680 were included in this analysis. The population PK analysis was performed using a non-linear mixed effects modeling approach in NONMEM (version 7.2) software. Impact of patient demographics, clinical indices and biomarkers on PK parameters were explored. The appropriateness of the final model was tested using visual predictive check (VPC). A sequential PK-PD analysis was performed using receptor occupancy (RO) data from 35 subjects.

Results: MEDI0680 PK profiles were best described using a 2-compartment model with linear clearance. The clearance (CL), volume of distribution (Vc) were 0.27 L/day, 5.07 L with a modest between-subject variability of 30% and 19%, respectively. None of the evaluated covariates showed any impact on PK parameters except a minor (not clinically relevant) impact of body weight on volume of distribution. VPC results demonstrated good predictability of the final population PK model. A direct Emax model described the PK-PD relationship of MEDI0680. The estimate of EC50 was approximately 9.3 µg/mL. PK/PD simulations indicate that following 20 mg/kg Q2W dose, &gt;90% receptor occupancy can be maintained in all subjects. Based on preclinical/clinical PK, PD, and safety data, a dose of 20 mg/kg Q2W was selected for phase 2 studies.

Conclusions: A population PK model of MEDI0680 was developed and validated. Modeling results indicate no need for dose adjustment based on patient/disease characteristics. Similar PK is expected following both WT-based and fixed dosing regimens. PK/PD findings support the dose of 20 mg/kg Q2W. Clinical studies are ongoing in various tumor types.

Citation Format: Xuyang Song, Xizhe Gao, Bo Zheng, Chelsea Black, Matthew Gribbin, Joyson Karakunnel, Lorin Roskos, Rajesh Narwal. Pharmacokinetics and pharmacodynamics of MEDI0680, a fully human anti-PD1 monoclonal antibody, in patients with advanced malignancies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5045. doi:10.1158/1538-7445.AM2017-5045

Abstract 5046: Exposure-efficacy (OS) analysis of tremelimumab in unresectable malignant mesothelioma

Purpose: Tremelimumab is a fully human anti-CTLA-4 IgG2 monoclonal antibody that enhances human T-cell activation. Tremelimumab was evaluated in a Phase IIb (DETERMINE), randomised, double-blind, placebo-controlled study in patients with unresectable pleural or peritoneal malignant mesothelioma, randomised (2:1) to receive either tremelimumab (10 mg/kg, seven doses Q4W followed by Q12W) or placebo. The study demonstrated no clinically meaningful differences in overall survival (OS). The primary objectives of this analysis were to evaluate the relationship of exposure with OS, and the impact of potential confounders.

Methods: A population PK model was developed to estimate and derive PK exposure metrics (area under the curve at steady state [AUCss] or clearance [CL]) for exposure-OS analysis. Impact of potential confounders was evaluated using graphical and exploratory approaches. Factors including body weight, age, gender, race, ECOG status, anatomical site (pleural or peritoneal), line of therapy, EORTC status, tumour histology, baseline tumour size, LDH, and CRP were evaluated. The analyses were performed using NONMEM 7.2 and R software.

Results: The population PK included 376 patients and 1328 post-first dose PK concentrations. PK was consistent with previous knowledge and low incidence of anti-drug antibodies was observed. A 2-compartment linear PK model adequately described the data. Tremelimumab CL and volume of distribution (V1) were 310 mL/day and 3.85 L, with moderate variability of ~38% and ~32%, respectively. There was an apparent exposure-OS relation when stratified by AUCss. However, at least 3 factors (gender, CRP, and baseline tumour size) were statistically significant PK predictors (p&lt;0.05 on CL) indicating multi-dimensional confounding effect. Higher baseline tumor size, higher CRP levels and males were associated with lower PK exposure of tremelimumab.

Conclusions: The observed apparent exposure-OS relationship is the result of imbalance in prognostic factors impacting OS rather than a true association of exposure with efficacy.

Citation Format: Paul Baverel, Lorin Roskos, Manasa Tatipalli, Nancy Lee, Paul Stockman, Maria Taboada, Paolo Vicini, Kevin Horgan, Rajesh Narwal. Exposure-efficacy (OS) analysis of tremelimumab in unresectable malignant mesothelioma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5046. doi:10.1158/1538-7445.AM2017-5046

Exposure-efficacy and safety analysis of durvalumab in patients with urothelial carcinoma (UC) and other solid tumors

2568

Background: Durvalumab is a human monoclonal antibody that binds to PD-L1 and blocks its interaction with PD-1 and CD-80. The objective of this analysis was to evaluate the relationship between durvalumab PK exposure with efficacy and safety following 10 mg/kg Q2W durvalumab. Methods: Data from Study 1108 (Phase 1/2; all tumor types) and ATLANTIC (Phase 2; NSCLC) were used for exposure-safety analysis for Study 1108 UC cohort, Study 1108 all patients and ATLANTIC patients, respectively, whereas the exposure-efficacy analysis was performed using data from Study 1108 UC cohort. The observed PK exposure metrics included PK concentrations after the first, second or steady state doses. Efficacy endpoints used were objective response rate (ORR) and best percentage change in target lesion from baseline per BICR assessment. Safety endpoints included Grade 3+ AE (any AE, drug-related AE, AESI, and drug-related AESI) and AE leading to treatment discontinuation. Results: Overall, no association of PK exposure with efficacy or safety was observed. Distribution of PK metrics were similar between responders and non-responders. The probability of objective response was similar in all quartiles of exposure (p-value ranged from 0.37 to 0.67; n = 96) with no obvious trends between PK exposures and change in tumor size. For Grade 3+ AE (all types) and AE leading to treatment discontinuation, higher PK exposure was not associated with an increased risk of AE (p-value ranged from < 0.00005 to 0.88; n = 158, 929 and 434 for 1108 UC cohort, 1108 all patients and ATLANTIC all patients, respectively). A few inverse trends were observed, likely due to confounding effect of ECOG or albumin since covariate analysis demonstrated that both variables correlated with PK and AEs. In addition, the association of ECOG and albumin versus PK exposure were also observed in the population PK modeling. Conclusions: The exposure-efficacy and exposure-safety analyses suggested that 10 mg/kg IV Q2W regimen was an appropriate dose for durvalumab as single agent in UC patients. Overall, no relationship of PK exposure with either the efficacy or safety was observed following 10 mg/kg IV Q2W regimen. Clinical trial information: NCT02087423 and NCT01693562.

Tumor shrinkage and increased overall survival are associated with improved albumin, neutrophil lymphocyte ratio (NLR) and decreased durvalumab clearance in NSCLC and UC patients receiving durvalumab

3035

Background: Progression of cancer is often associated with biomarkers of cancer inflammation, cachexia, and increased protein catabolism. Anti-PD1 and PD-L1 therapy have demonstrated durable responses across a number of tumor types. Durvalumab is a human monoclonal antibody that binds to PD-L1 and blocks its interaction with PD-1 and CD80. The primary objective of this analysis was to prospectively assess potential correlations of longitudinal changes in ALB and NLR and durvalumab clearance (CL) rate to maximum decrease in tumor size and overall survival (OS) in patients (pts) with NSCLC and UC receiving durvalumab. Methods: Longitudinal target lesion size, serum chemistry, hematology and pharmacokinetic data were obtained from 3L+ NSCLC pts (n = 418) in study ATLANTIC and 2L+ UC pts (n = 182) in study 1108 during durvalumab treatment. Nonparametric correlations (Spearman’s rho) were evaluated between OS, maximum percent change in target lesion size, and the maximum percent change from baseline observed in ALB, NLR, and CL. Results: In NSCLC, maximum decrease in tumor size was correlated with increased ALB (r = 0.46, p < 0.0001), decreased NLR (r = 0.44, p < 0.0001), and decreased CL (r = 0.66, p < 0.0001). OS was similarly correlated with increased ALB (r = 0.47, p < 0.0001), decreased NLR (r = 0.41, p < 0.0001), and decreased CL (r = 0.76, p < 0.0001). In UC, decreased tumor size also correlated with increased ALB (r = 0.43, p < 0.0001), decreased NLR (r = 0.38, p < 0.0001), and decreased CL (r = 0.65, p < 0.0001). OS in UC also correlated with increased ALB (r = 0.50, p < 0.0001), decreased NLR (r = 0.33, p < 0.0001) and decreased CL (r = 0.82, p < 0.0001). Conclusions: In NSCLC and UC pts receiving durvalumab, tumor shrinkage and longer survival are associated with increased ALB, decreased NLR and decreased clearance of durvalumab. These findings support the hypothesis that durvalumab may be associated with a decrease in protein catabolism, inflammation and cachexia among pts who benefited from therapy. Additional biomarkers of cancer, inflammation and cachexia will be evaluated for relationships to clinical outcomes.

Tumor kinetic modeling and identification of predictive factors for tumor response to durvalumab in patients with non-small cell lung cancer (NSCLC)

11555

Background: Durvalumab is a human monoclonal antibody that binds to PD-L1 and blocks its interaction with PD-1 and CD80. The primary objectives of this analysis were to describe the longitudinal tumor size profiles and identify the key factors predicting tumor growth and regression following durvalumab. Methods: Longitudinal tumor size data obtained from NSCLC patients in study 1108 (all lines of therapy) and ATLANTIC (third line and beyond) following durvalumab treatment were modeled using nonlinear mixed effect modeling. Tumor kinetics were described by four key parameters: tumor growth and killing rate constants, fraction of durvalumab-sensitive tumor cells, and delay time for tumor killing. Potential predictive factors for tumor growth and regression were evaluated in a multi-variable covariate analysis. The model was used to simulate response rates at different tumor PD-L1 expression cutoffs. Results: Tumor kinetic modeling accurately described the longitudinal tumor response profiles from NSCLC patients in both studies. The factors associated with more rapid tumor growth were liver metastases, ECOG score > 0, high neutrophil-to-lymphocyte ratio and EGFR/ALK mutation. Tumor cell PD-L1 expression, baseline tumor size and smoking history were identified as significant predictive factors for tumor killing or the fraction of sensitive tumor cells. Simulations using the tumor kinetic model showed increased response rates in patients with higher tumor cell PD-L1 expression (increased by 9-11% and 10-14% with 25% and 50% cutoff, respectively), patients receiving durvalumab as first-line therapy (increased by 12% vs. 2nd line/above), and patients with smoking histories (increased by 4-5% vs. non-smokers). Conclusions: Tumor kinetic modeling identified factors that predict tumor progression and response following durvalumab in NSCLC patients. The multivariate analysis accounts for various predictive factors within predictive biomarker strata, allowing better interpretation of different biomarker cutoffs. The modeling technique can potentially guide patient selection/enrichment, clinical trial design strategies and tumor biology. Clinical trial information: NCT02087423 and NCT01693562.

Population pharmacokinetics of durvalumab and fixed dosing regimens in patients with advanced solid tumors

2566

Background: Durvalumab is a human monoclonal antibody that binds to PD-L1 and blocks its interaction with PD-1 and CD80. The objectives of this analysis were to develop a population pharmacokinetics (PK) model of durvalumab, to quantitate the effect of patient/disease characteristics on PK, and to compare weight (WT)-based versus fixed dosing regimens. Methods: Data were pooled from two studies: Study 1108 (Phase 1/2; various tumor types) and ATLANTIC (Phase 2; NSCLC). A total of 1324 patients provided data following 0.1 to 20 mg/kg IV durvalumab. The population PK was performed using a non-linear mixed effects modeling approach in NONMEM software. The impact of demographics, clinical indices, and biomarkers on PK was explored. Results: Durvalumab PK was best described using a 2-compartment model with both linear and non-linear clearances. The mean (between-patient variability) linear clearance (CL) and central volume of distribution (V1) were 226 mL/day (~29%) and 3.51 L (~21%), respectively. Although population PK analysis identified a few statistically significant covariates (WT, sex, CrCL, post-baseline ADA, ECOG performance status, LDH, sPDL1 levels, tumor type, and albumin), none were found to be clinically relevant (effect on PK parameters < 30%), indicating no need for dose adjustment. Simulations indicated similar overall PK exposures following WT-based (10 mg/kg Q2W) and fixed dosing regimens (1500 mg Q4W or 750 mg Q2W); with all regimens expected to maintain target trough exposure of ~50 µg/mL in ≥95% patients. In a post-hoc analysis, durvalumab clearance was found to decrease slightly over time, with a mean maximal reduction from baseline value of 15.5%. The decrease in CL was associated with tumor shrinkage, decreased LDH, increased albumin and decreased neutrophil to lymphocyte ratio. The small decrease in CL was not considered relevant to PK exposure or dosing. Conclusions: A population PK model of durvalumab was developed and validated. No dose adjustments were needed based on any patient or disease characteristics. The analysis demonstrated the feasibility of switching to a fixed dose regimen. Clinical trial information: NCT02087423 and NCT01693562.

Evaluation of assay interference and interpretation of CXCR4 receptor occupancy results in a preclinical study with MEDI3185, a fully human antibody to CXCR4

Background

Receptor occupancy (RO) assays provide a means to measure the direct interaction of therapeutics with their cell surface targets. Free receptor assays quantify cell‐surface receptors not bound by a therapeutic while total receptor assays quantify the amount of target on the cell surface.

Methods

We developed both a flow cytometry‐based free RO assay to detect free surface CXCR4, and a total surface CXCR4 assay. In an effort to evaluate potential displacement interference, we performed in vitro experiments to compare on‐cell affinity with the IC₅₀ values from in vitro and in vivo from the free CXCR4 assay. We determined free and total surface CXCR4 on circulating blood cells in cynomolgus monkeys dosed with MEDI3185, a fully human monoclonal antibody to CXCR4.

Results

We devised an approach to evaluate displacement interference during assay development and showed that our free assay demonstrated little to no displacement interference. After dosing cynomolgus monkeys with MEDI3185, we observed dose‐dependence in the magnitude and duration of receptor occupancy and found CXCR4 to increase on lymphocytes, monocytes, and granulocytes. In a multiple dose study, we observed time points where surface CXCR4 appeared fully occupied but MEDI3185 was not detectable in serum. These paradoxical results represented a type of assay interference, and by comparing pharmacokinetic, ADA and total CXCR4 results, the most likely reason for the free CXCR4 results was the emergence of neutralizing anti‐drug antibodies (ADA). The total CXCR4 assay was unaffected by ADA and provided a reliable marker of target modulation in both in vivo studies. © 2015 The Authors Cytometry Part B: Clinical Cytometry Published byWiley Periodicals, Inc.

Abstract 30: Medi3622, a monoclonal antibody to ADAM17, inhibits tumor growth by inhibiting EGFR- and non-EGFR-mediated pathways

ADAM17 is the primary sheddase for HER pathway ligands. We report the discovery of a potent and specific ADAM17 inhibitory antibody, MEDI3622, which induces tumor regression or stasis in many EGFR-dependent tumor models. The inhibitory activity of MEDI3622 correlated with EGFR activity both in a series of tumor models across several indications as well as in a focused set of head and neck patient derived xenograft models. Cynomolgus monkey and rat PK/PD assays showed MEDI3622 inhibited TNFα shedding. Toxicity observed in cynomolgus monkey and rat was similar to EGFR inhibitor-induced rash. However, the antitumor activity of MEDI3622 was superior to that of EGFR/HER pathway inhibitors in OE21 head and neck and COLO205 colorectal xenograft models suggesting additional activity outside of the EGFR pathway. Combination of MEDI3622 and cetuximab in the OE21 model was additive and eradicated tumors. Proteomics analysis revealed novel ADAM17 substrates which function outside of the HER pathways and may contribute towards the antitumor activity of the monoclonal antibody.

Citation Format: Darrin Sabol, Jonathan RiosDoria, Jon Chesebrough, David Stewart, Kevin Schifferli, Raymond Rothstein, Ching Ching Leow, Jenny Heidbrink-Thompson, Li Cheng, Qun Du, Linda Xu, Xiaofang Jin, Ravinder Tammali, Chanshou Gao, Jay Friedman, Brandy Wilkinson, Melissa Damschroder, Andrew Pierce, MunMun Patnaik, Rong Zeng, Yuling Wu, Susan Spitz, Gabriel Robbie, Lorin Roskos, Robert Hollingsworth, David Tice, Emil Michelotti. Medi3622, a monoclonal antibody to ADAM17, inhibits tumor growth by inhibiting EGFR- and non-EGFR-mediated pathways. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 30. doi:10.1158/1538-7445.AM2015-30

Choice of Starting Dose for Biopharmaceuticals in First-in-Human Phase I Cancer Clinical Trials

BACKGROUND

First-in-human (FIH) trials of low-molecular-weight anticancer agents conventionally derive a safe start dose (SD) from one-tenth the severely toxic dose in 10% of rodents or one-sixth the highest nonseverely toxic dose (HNSTD) in nonrodent species. No consensus has been reached on whether this paradigm can be safely applied to biotechnology-derived products (BDPs).

MATERIALS AND METHODS

A comprehensive search was conducted to identify all BDPs (excluding immune checkpoint inhibitors and antibody drug conjugates) with sufficient nonclinical and clinical data to assess the safety of hypothetical use of one-sixth HNSTD in an advanced cancer FIH trial.

RESULTS

The search identified 23 BDPs, of which 21 were monoclonal antibodies. The median ratio of the maximum tolerated or maximum administered dose (MTD or MAD) to the actual FIH SD was 36 (range, 8-500). Only 2 BDPs reached the MTD. Hypothetical use of one-sixth HNSTD (allometrically scaled to humans) would not have exceeded the MTD or MAD for all 23 BDPs and would have reduced the median ratio of the MTD or MAD to a SD to 6.1 (range, 3.5-55.3). Pharmacodynamic (PD) markers were included in some animal toxicology studies and were useful to confirm the hypothetical SD of one-sixth HNSTD.

CONCLUSION

One-sixth HNSTD would not have resulted in unacceptable toxicities in the data available. Supporting its use could reduce the number of dose escalations needed to reach the recommended dose. A low incidence of toxicities in animals and humans underscores the need to identify the pharmacokinetic and PD parameters to guide SD selection of BDPs for FIH cancer trials.

IMPLICATIONS FOR PRACTICE

Start dose (SD) for biotechnology-derived products (BDPs) can be safely derived from one-sixth the highest nonseverely toxic dose in nonrodent species and may reduce the number of dose escalations needed to reach the recommended dose in first-in-human studies while limiting unnecessary exposure to high drug levels in humans. The use of this type of SD could improve the design of phase I studies of BDPs by making them more efficient. The role of preclinical pharmacodynamic markers was useful in confirming the hypothetical SD, and attempts should be explored in future animal studies to identify such parameters.