Population Pharmacokinetics and Exposure‐Response Analyses for Telisotuzumab Vedotin in Patients With c‐Met Protein Overexpressing Tumors

ABSTRACT

Telisotuzumab vedotin (Teliso‐V) is a c‐Met‐directed antibody‐drug conjugate that delivers a cytotoxic microtubule inhibitor monomethyl auristatin E (MMAE) payload to c‐Met‐expressing tumor cells. It received accelerated approval from the US FDA for the treatment of adults with locally advanced or metastatic non‐squamous non‐small cell lung cancer (NSCLC) with high c‐Met protein overexpression (≥ 50% of tumor cells with strong [3+] staining) at a dose of 1.9 mg/kg every 2 weeks (Q2W; maximum 190 mg for patients ≥ 100 kg) as an intravenous infusion. Population pharmacokinetic (PK) modeling used pooled data from a phase 1 (N = 35) and phase 2 study (N = 269) to describe the Teliso‐V conjugate and unconjugated MMAE PK and evaluate the impact of intrinsic and extrinsic factors on exposures in patients with solid tumors. Body weight, race, albumin, and anti‐drug antibody status were identified as significant covariates on Teliso‐V conjugate clearance, but did not result in clinically meaningful changes in exposure. The exposure‐response evaluations for efficacy (based on the pivotal phase 2 study) showed significant correlations between conjugate exposure and overall response rates. Higher conjugate exposures were also correlated with improved progression‐free survival and overall survival, demonstrating meaningful clinical benefit with the 1.9 mg/kg Q2W dosing regimen. Exposure‐safety evaluations showed significant relationships between conjugate exposures and grade ≥ 2 and grade ≥ 3 peripheral neuropathy, and grade ≥ 2 corneal epitheliopathy. Unconjugated MMAE payload exposures were correlated with a greater probability of grade ≥ 3 treatment‐emergent adverse events. The 1.9 mg/kg Q2W dose maximized efficacy while balancing adverse events in patients with c‐Met overexpressing NSCLC.

Study Highlights

• What is the current knowledge on the topic?

  • ○
    Teliso‐V has compelling efficacy at the 1.9 mg/kg every 2‐week regimen in c‐Met overexpressing non‐squamous non‐small cell lung cancer (NSCLC). Population pharmacokinetic (PK) analyses integrating data across trials, including covariate assessments and ER evaluations from the pivotal LUMINOSITY phase 2 study to inform dosing, have not been published to date.
• What question did this study address?

  • ○
    This analysis characterized the PK of Teliso‐V conjugate and unconjugated MMAE and assessed the impact of intrinsic and extrinsic factors on PK. ER analyses for efficacy and safety were used to evaluate the dosing regimen of Teliso‐V.
• What does this study add to our knowledge?

  • ○
    Positive relationships between higher Teliso‐V conjugate exposure and efficacy in the LUMINOSITY phase 2 study support the benefit of Teliso‐V 1.9 mg/kg every 2 weeks in patients with c‐Met protein overexpressing NSCLC as the approved dose.
• How might this change drug discovery, development, and/or therapeutics?

  • ○
    The Teliso‐V approved dose of 1.9 mg/kg Q2W is appropriate for the approved indication. The findings indicate that for Teliso‐V, efficacy is correlated with ADC exposure, while safety is influenced by both ADC and MMAE payload exposure. Higher exposures correlate with improved efficacy but also increased safety risks, underscoring the need to carefully balance therapeutic benefit with tolerability to achieve optimal clinical efficacy with manageable safety for Teliso‐V.

1. Introduction

Lung cancer remains a leading cause of cancer‐related mortality in the United States [1], with non‐small cell lung cancer (NSCLC) comprising approximately 85% of all diagnoses [2, 3, 4, 5]. The MET proto‐oncogene encodes the c‐Met protein (also known as MET protein and hepatocyte growth factor receptor), a receptor tyrosine kinase that mediates cell proliferation, survival, and angiogenesis and is expressed on various normal and malignant cells [6]. Dysregulated MET signaling has been shown to be involved in the proliferation, angiogenesis, invasion, and survival of several types of cancer cells through gene amplification, mutation, or protein overexpression [4, 5, 7, 8]. The overexpression of c‐Met has been found to occur in 20%–50% of solid tumors and 25% of non‐squamous epidermal growth factor receptor (EGFR) wild‐type NSCLC [1, 9, 10, 11]. Additionally, it is associated with a poor prognosis in patients with NSCLC [9, 12, 13]. In recent years, MET‐targeting therapies have emerged as treatment options for patients with NSCLC, including tyrosine kinase inhibitors, monoclonal antibodies, and antibody‐drug conjugate (ADC) treatments [4].

Telisotuzumab vedotin (Teliso‐V; Emrelis) is a c‐Met‐directed ADC comprised of the monoclonal antibody telisotuzumab (ABT‐700) conjugated to the microtubule inhibitor monomethyl auristatin E (MMAE) [14, 15, 16]. It targets c‐Met‐expressing tumor cells with high specificity and high affinity, mediating the delivery of MMAE (a potent antimitotic agent that inhibits cell division by preventing the polymerization of tubulin) directly to the tumor [15, 17]. Once engaged, Teliso‐V is internalized, and MMAE is released into the tumor cell, where it binds tubulin, leading to cell death [14, 15]. Unlike MET tyrosine kinase inhibitors, which are approved for the 2%–4% of patients with NSCLC harboring MET exon 14 skipping mutations, Teliso‐V has shown activity in the ~25% of patients with non‐squamous EGFR wild‐type NSCLC with c‐Met overexpression [18].

Teliso‐V has been evaluated as a monotherapy in a phase 1 study (NCT02099058) using every 2‐week (Q2W; 1.6 to 2.2 mg/kg) and every 3‐week (Q3W; 0.15 to 3.3 mg/kg) dosing in patients with solid tumors. Initial dose escalation was conducted in c‐Met unselected patients with advanced solid tumors, while dose expansion was conducted in c‐Met‐positive NSCLC patients (as defined by an H‐score ≥ 150, determined by immunohistochemistry). The monotherapy Q2W dosing regimen was selected for further development based on higher observed clinical efficacy (overall response rate [ORR]) in c‐Met positive NSCLC patients, higher minimum concentrations (C_min) with Q2W dosing, and similar tolerability profiles compared to Q3W dosing [19]. A phase 2 study (LUMINOSITY; NCT03539536) in patients with NSCLC with overexpression of c‐Met protein (defined as ≥ 25% tumor cells with 3+ staining [high: ≥ 50% 3+; intermediate: ≥ 25%–< 50%] for patients with non‐squamous disease) who received previous lines of therapy evaluated the efficacy and safety of Teliso‐V at 1.9 mg/kg Q2W. A supplemental cohort evaluated Teliso‐V at 1.6 mg/kg Q2W [18]. The c‐Met cutoffs for staining intensity and percent positivity were selected based on statistical analyses that evaluated the association with the best objective response in the phase 1 study [20]. ORR for the 1.9 mg/kg dose was 29% in all patients with c‐Met overexpressing non‐squamous EGFR wild‐type NSCLC, increasing to 35% among patients with c‐Met high overexpression. The most common treatment‐emergent adverse events (TEAEs) were peripheral sensory neuropathy (30%), peripheral edema (16%), and fatigue (14%) [18].

Here, we describe the population pharmacokinetics (PK) of the Teliso‐V conjugate and unconjugated MMAE payload, the effect of intrinsic and extrinsic factors on exposures, as well as the relationships of both the Teliso‐V conjugate and unconjugated MMAE payload exposure to the clinical efficacy and safety variables using data from the phase 1 and phase 2 studies that supported the recommended and approved dosing regimen in patients with c‐Met overexpressing NSCLC.

2. Methods

2.1. Study Designs and Assessments

The studies included in these analyses were conducted in accordance with Good Clinical Practice guidelines and the ethical principles that have their origin in the Declaration of Helsinki. The protocol and informed consent forms were approved by the institutional review boards or ethics committees for each study site, and participants provided written informed consent before any study‐related procedures were performed.

These analyses included dose‐ranging data from patients with advanced solid tumors likely to express c‐Met enrolled in the phase 1 study of Teliso‐V (Q2W and Q3W dosing; NCT02099058) and patients with locally advanced or metastatic NSCLC overexpressing c‐Met enrolled in the phase 2 LUMINOSITY study (NCT03539536) of Teliso‐V 1.6 mg/kg or 1.9 mg/kg Q2W [18, 21]. Additional details regarding the study designs, objectives, and assessments for the population PK and efficacy‐response analyses are provided in Table S1.

2.2. Population Pharmacokinetic Analyses

The population PK analysis used Teliso‐V conjugate and unconjugated MMAE data from Teliso‐V monotherapy patients with solid tumors enrolled in the phase 1 study and patients with NSCLC enrolled in the LUMINOSITY study. The model was developed using non‐linear mixed effects modeling based on NONMEM (Version 7.5.1) compiled with the GNU Fortran compiler (Version 7.5.0) to describe the concentration‐time profiles of Teliso‐V conjugate and unconjugated MMAE payload.

The conjugate and unconjugated MMAE models were developed independently since independent models show similar performance compared to integrated models [22, 23]. Additional potential covariate‐parameter relationships were evaluated. Non‐linear clearance (CL) was assessed and did not improve the model fit based on the objective‐function value. A deconjugation rate (K_a) was investigated for the unconjugated MMAE payload model, which was described using a first‐order release of MMAE from the ADC structure [22]. Additionally, interindividual variability in PK parameters was modeled using a multivariate log‐normal distribution. Proportional residual error models and a combination of additive and proportional residual error were tested and compared for the best description of the residual unexplained variability. See the Supporting Information for more details.

Continuous covariates were normalized to the median of the overall population and incorporated into the model using a power function. Categorical covariates were included multiplicatively to obtain the proportional change between the tested categorical groups. Covariates were investigated for influence on Teliso‐V PK via stepwise forward inclusion and backward elimination for CL and volume of distribution (V_c) in the Teliso‐V conjugate model and unconjugated MMAE payload model as implemented in Perl Speaks NONMEM (PsN Version 5.3.1). Forward inclusion and backward elimination steps were conducted using the likelihood ratio test at significance levels of 0.01 and 0.001, respectively. A list of the covariates of interest can be found in Table S2.

After identifying the significant covariates, simulations were performed to explore the impact of relevant covariate effects on Teliso‐V conjugate steady‐state average serum concentration and unconjugated MMAE payload steady‐state average plasma concentration (C_avg), the area under the concentration‐time curve within a dosing interval (AUC_tau), and the geometric mean (GM) of the maximum concentration (C_max) exposures. Predicted Teliso‐V conjugate and unconjugated MMAE exposure from the simulations were compared to the corresponding reference group (Table S3) by calculating the geometric mean (GM) ratio and its 95% confidence interval (CI) and summarized graphically using a forest plot.

The final population PK model was evaluated using goodness‐of‐fit plots and visual predicted checks (VPCs). For VPCs, 500 simulated replicates of the dataset were generated using NONMEM. Subsequently, the simulated predictions were compared to the observed data by superimposing the observed data on selected percentile intervals (median, 5th and 95th percentiles with their 90% CIs) of the simulated data and stratified by dosing frequency.

2.3. Exposure‐Response Analyses

The exposure‐efficacy analysis population included non‐squamous EGFR wild‐type c‐Met + NSCLC patients with c‐Met high (defined as having ≥ 50% of tumor cells with strong [3+] staining by an immunohistochemistry) and c‐Met intermediate (defined as ≥ 25% to < 50% of tumor cells with strong [3+] staining by immunohistochemistry) who received Teliso‐V 1.6 mg/kg or 1.9 mg/kg Q2W and were enrolled in the LUMINOSITY study. Efficacy variables included ORR as determined by an Independent Central Review [ICR], duration of response (DoR), disease control rate (DCR), progression‐free survival (PFS) by ICR, and overall survival (OS).

The exposure‐safety analysis population included patients with NSCLC receiving monotherapy from the phase 1 study and all patients enrolled in the LUMINOSITY phase 2 study. Data from patients enrolled in both studies were pooled and included in the exposure‐safety analysis. Safety variables included grade ≥ 2 and grade ≥ 3 peripheral neuropathy, grade ≥ 2 and grade ≥ 3 corneal epitheliopathy, all‐grade and grade ≥ 2 interstitial lung disease (ILD)/pneumonitis, dose interruptions or discontinuations due to an adverse event (AE), and grade ≥ 3 TEAEs. Relative dose intensity was computed as the ratio of the actual dose received by each patient relative to the planned dose and assessed through rainbow plots of relative dose intensity, showing the percent of patients receiving the target dose level for each cycle relative to the start of treatment.

Individual exposure metrics for Teliso‐V conjugate and unconjugated MMAE payload were estimated based on post hoc estimates from the population PK models, utilizing actual doses received by patients. Exposure metrics included C_avg (up to the time of the event or up to the end of treatment if no event occurred), the average concentration of cycle 1 (C_avgC1; calculated as the average concentration for the first 2 and 3 weeks depending on the assigned dosing interval of Q2W or Q3W, respectively), the GM of the minimum observed concentration (C_mingm; calculated as the GM of all pre‐dose values up to the time of event), and the GM of C_max up to event (C_maxgm for exposure‐safety only).

Exposure‐response (ER) relationships for efficacy and safety were evaluated using quartile plots. Only variables with > 10 events that indicated trends were evaluated further using ER models for exposure‐dependent increase in efficacy responses or incidence of safety events. Linear and logarithmic logistic regression analyses for the efficacy and safety parameters were evaluated to characterize the relationship between Teliso‐V conjugate or unconjugated MMAE payload exposures and the different variables. The effect of Teliso‐V on conjugate and unconjugated MMAE payload exposures on the DoR, PFS, and OS was evaluated using Kaplan–Meier curves and Cox proportional hazard models. The ER models were developed in R (Version 4.3.1) using the SUSE Linux Enterprise Server (15 SP3) operating system.

Covariates were evaluated for their effect on all regression models for safety and efficacy endpoints in a stepwise forward selection and backward elimination procedure with significance levels of α = 0.01 and α = 0.001, respectively, and were only evaluated if the exposure was statistically significant (p < 0.05). A list of the covariates of interest can be found in Table S3.

2.4. Model‐Based Simulations

Logistic regression models developed for efficacy and safety variables were used to conduct simulations to predict the probabilities of the efficacy and safety variables at different dosing regimens (1.6 and 1.9 mg/kg Q2W). The simulations incorporated interindividual variability in Teliso‐V PK and covariate distributions by resampling 500 patients (exposures and covariates) from the populations of both studies for each treatment group. The variability associated with the uncertainty of the parameter estimates in the logistic regression models was also included in the simulations by sampling 1000 parameters from a multivariate random normal distribution with means and covariance terms that are estimated in the logistic regression models. The percentage of simulated patients experiencing the efficacy or safety outcome of interest was then calculated across all parameter sets.

3. Results

Demographic and baseline characteristics for all analyses can be found in Tables S4–S6.

3.1. Population Pharmacokinetics

A total of 304 patients from the phase 1 study (N = 35) and the LUMINOSITY phase 2 study (N = 269) were included in the population PK analysis (see Supporting Information).

The conjugate PK was best described by a two‐compartment PK model with linear elimination and interindividual variability on CL and V_c, while the unconjugated MMAE payload PK was best described by a one‐compartment PK model with first‐order deconjugation rate from the Teliso‐V conjugate (K_a), linear elimination, and interindividual variability on CL, V_c and K_a with the inclusion of correlation between CL and V_c (Figure S1). Residual variability was best described by combined additive and proportional error terms. A non‐linear clearance was tested for the conjugate PK as well but was not found to be significant and did not improve the model accuracy.

The goodness‐of‐fit plots indicated no systematic model bias or misspecifications (Figures S2 and S3). The VPCs stratified by dosing frequency for the Teliso‐V conjugate and unconjugated MMAE payload concentrations demonstrated that the models adequately captured the observed concentration‐time courses and variability of the observed data (Figures S4 and S5).

Teliso‐V conjugate concentrations and unconjugated MMAE payload concentrations were adequately described by the final population PK models (Table S7). All structural conjugate PK parameters in the model were estimated with a relative standard error of ≤ 23.1% and payload parameters were estimated with a relative standard error of ≤ 29.3%. The shrinkages for Teliso‐V conjugate CL and V_c were 4.03% and 13.0%, respectively, and the shrinkages for unconjugated MMAE payload CL, V_c, and K_a were 4.64%, 14.2%, and 18.2%, respectively.

Among the covariates evaluated for Teliso‐V conjugate, a combined effect of body weight on V_c and volume of distribution of the peripheral compartment and body weight on CL and intercompartmental clearance was found to be significant. Anti‐drug antibody (ADA) status, baseline albumin, and race were significantly correlated with CL; and age, baseline albumin, and sex were significantly correlated with V_c. The effect of these covariates on Teliso‐V conjugate exposures (C_max and AUC_tau) was within −15.4% to 56% (Figure 1A).

FIGURE 1.

Forest plots of (A) Teliso‐V Conjugate and (B) Unconjugated MMAE Payload Exposures (C_max and Steady‐State AUC_tau). Effect of covariates on Teliso‐V conjugate and unconjugated MMAE payload exposures. Points represent the geometric mean ratio, and error bars represent 95% CIs of model‐predicted exposures relative to reference groups. ADA, anti‐drug antibody; AUC_tau, area under the concentration‐time curve within a dosing interval; CI, confidence interval; C_max, maximum observed concentration; MMAE, monomethyl auristatin E; Teliso‐V, telisotuzumab vedotin.

Baseline albumin and renal impairment status were significantly correlated with MMAE CL; age, baseline albumin, and race were significantly correlated with K_a; and body weight was significantly correlated with V_c in the MMAE population PK model. The effect of body weight and age on MMAE exposures (C_max and AUC_tau) was generally within 25% (Figure 1B).

Model‐predicted PK parameters for 1.9 mg/kg and profiles (VPCs) for all dosing regimens are shown in Figure S5.

3.2. Exposure‐Response Analyses

A total of 193 patients with c‐Met protein overexpressing EGFR wild‐type non‐squamous NSCLC from the LUMINOSITY phase 2 study were included in the exposure‐efficacy analysis, and 284 patients of advanced solid tumor types from the phase 1 and LUMINOSITY phase 2 studies were included in the exposure‐safety analysis. PK data for one patient were imputed for the ER analysis using population estimates due to study discontinuation prior to the collection of any PK samples.

Exposure‐efficacy analyses demonstrated that higher conjugate exposures (C_avg) were strongly correlated with ORR per ICR (Figure 2). Similar relationships were observed between higher conjugate exposure (all metrics) and improved DCR. Efficacy was not correlated with MMAE exposures; flat ER relationships for ORR were observed with MMAE. In addition to ORR, significant positive correlations between other efficacy variables (DoR, PFS per ICR, and OS) and C_avgC1 ADC and C_mingmADC (for PFS per ICR and OS) were observed (p < 0.05) across the evaluated range of exposures (Figure 3).

FIGURE 2.

Logistic Regression for Exposure‐Overall Response Rate per ICR based on EGFR Wild‐Type Patients. Number of patients with a response. Nominal p = 8.55 × 10–6. The figure shows observed data in solid symbols, prediction from the model as a solid line, and 95% CI of the prediction in the shaded region. ADC, antibody‐drug conjugate; C_avg, average serum concentration; EGFR, epidermal growth factor receptor; ICR, Independent Central Review; ORR, overall response rate.

FIGURE 3.

Kaplan–Meier Curves of Time to (A) Duration of Response, (B) Progression‐Free Survival, and (C) Overall Survival by Teliso‐V Conjugate Exposure (C_avgC1) Quartiles. Q1‐4: First to fourth quartile (in μg/mL), colors indicating increasing exposure‐quartiles up to event. The numbers at risk table shows the number of patients per exposure‐quartiles that have not experienced an event or censoring at their corresponding point of time. C1, cycle 1; C_avg, average serum concentration; ICR, Independent Central Review; ORR, overall response rate.

Exposure‐safety evaluations showed that higher conjugate exposures were correlated with a higher probability of experiencing grade ≥ 2 and ≥ 3 peripheral neuropathy and grade ≥ 2 corneal epitheliopathy (Figures 4 and 5). Significant positive correlations between grade ≥ 2 and grade ≥ 3 peripheral neuropathy, grade ≥ 2 corneal epitheliopathy, dose interruption or discontinuation due to an AE and C_avgADC, C_avgC1 ADC, C_mingmADC, and C_maxgmADC were observed (p < 0.05) across the evaluated range of exposures. There was a significant positive correlation between grade ≥ 2 and all‐grade ILD/pneumonitis and C_avgADC (p < 0.05) across the evaluated range of exposures. There was also a significant positive correlation between grade ≥ 3 TEAEs and C_avgMMAE, C_avgC1 MMAE, C_mingmMMAE, and C_maxgmMMAE (p < 0.05) across the evaluated range of exposures. Grade ≥ 3 corneal epitheliopathy events were not evaluated in the exposure‐safety analyses due to a very small number of events (N = 2).

FIGURE 4.

Exposure‐Safety Analyses using Logistic Regression for Grade ≥ 3 Peripheral Neuropathy (Left) and Grade ≥ 2 Corneal Epitheliopathy (Right) based on Patients with NSCLC. Nominal p‐value (peripheral neuropathy) = 2.98 × 10–4; Nominal p‐value (corneal epitheliopathy) = 3.69 × 10–8. Figures show observed data in solid symbols, prediction from the model as a solid line, and 95% CI of the prediction in the shaded region. The dots and error bars represent median and 95% binomial CIs of binned observed responses. ADC, antibody‐drug conjugate; C_avg, average serum concentration; NSCLC, non‐small cell lung cancer; Q2w, every 2 weeks.

FIGURE 5.

Observed and Model‐Predicted Percentage of Patients with Grade ≥ 3 Treatment‐Emergent Adverse Events by Unconjugated MMAE Payload Exposures. The solid lines represent the median predicted response, and the shaded areas represent 95% CIs of the response. The dots and error bars represent median and 95% binomial CIs of binned observed responses. C1, cycle 1; C_avg, average serum concentration; CI, confidence interval; C_maxgm, geometric mean of the maximum concentration; C_mingm, geometric mean of the minimum observed concentration; MMAE, monomethyl auristatin E.

No covariates were found to have a significant effect on efficacy or safety. Overall, the selected models adequately described the ER relationships for efficacy and safety and were used to predict probabilities of response or safety event for the dosing regimens of 1.6 and 1.9 mg/kg Q2W (Table 1).

TABLE 1.

Predicted efficacy and safety based on simulations.

Dosing regimen (Q2W)	Efficacy: median probability of ORR (95% CI)	Safety: median probability of grade ≥ 3 peripheral neuropathy (95% CI)	Safety: median probability of grade ≥ 2 corneal epitheliopathy (95% CI)
1.6 mg/kg	22.2 (15.8, 29.8)	6.10 (3.20, 10.2)	7.80 (4.60, 12.4)
1.9 mg/kg	31.4 (24.2, 39.0)	9.60 (6.00, 14.4)	15.0 (10.4, 20.6)

Abbreviations: CI, confidence interval; ORR, overall response rate; Q2W, every 2 weeks.

4. Discussion

This population PK analysis used data from 2 clinical trials (phase 1 study [NCT02099058] with solid tumors and LUMINOSITY phase 2 study [NCT03539536] in patients with NSCLC) to adequately characterize the PK for Teliso‐V conjugate and unconjugated MMAE.

Body weight, sex, and treatment‐emergent ADAs were significant covariates on Teliso‐V conjugate CL, but not MMAE CL. None of these covariates resulted in clinically relevant changes in C_max and AUC_tau based on model‐predicted steady‐state exposures (< 16% change). Although unconjugated MMAE payload exposure was lower in patients with higher baseline albumin, this is not predicted to impact efficacy, given that Teliso‐V conjugate exposure (not unconjugated MMAE payload) is correlated with efficacy based on ER analyses.

Black or African American race was a significant covariate on Teliso‐V conjugate exposure, resulting in a model‐predicted increase in AUC_tau and C_max of ~56% and ~23%, respectively. However, the median and individual Teliso‐V conjugate exposures for these patients were within the variability of exposures in White patients (n = 198), and the small number of Black or African American patients (n = 8) limits conclusions on relevance. Asian patients (n = 98) had similar Teliso‐V conjugate exposures (within 10%) compared to White patients, but ~25% lower unconjugated MMAE exposures, though the median and individual unconjugated MMAE payload exposures were also within the exposure range of White patients. Since Teliso‐V conjugate exposure (not unconjugated MMAE payload) was correlated with efficacy based on ER analyses, this lower unconjugated MMAE payload exposure is not expected to impact efficacy. Additionally, given that the probability of grade ≥ 3 TEAEs was correlated with higher unconjugated MMAE payload exposures, lower unconjugated MMAE payload exposure is not expected to negatively impact the safety profile. These data suggest that dose adjustments based on race are not required.

Though identified as a significant covariate on MMAE CL, all categories of renal impairment did not result in clinically relevant changes in unconjugated MMAE payload exposure. However, conclusions regarding severe renal impairment on unconjugated MMAE payload exposure are limited, given the small number of patients with severe renal impairment (n = 2). Lastly, mild hepatic impairment was not identified as a significant covariate on Teliso‐V conjugate or free MMAE payload clearance and resulted in similar Teliso‐V conjugate and unconjugated MMAE exposures compared to patients with no hepatic impairment. No patients with moderate or severe hepatic impairment were included in the analysis, so conclusions on that population are limited. Overall, these results support that no dose adjustment is needed in patients with mild or moderate renal impairment or mild hepatic impairment. The impact of severe renal impairment or moderate to severe hepatic impairment on the PK of Teliso‐V conjugated or unconjugated MMAE payload is unknown, given the small number of patients included in the analysis.

The exposure‐efficacy analyses in patients with c‐Met protein overexpression EGFR wild‐type non‐squamous NSCLC showed that higher Teliso‐V conjugate exposures were strongly correlated with ORR (per ICR), DCR, DoR, PFS (per ICR), and OS. Teliso‐V conjugate exposure metrics (C_avg, C_avgC1, and C_mingm) correlated better with these efficacy variables compared to unconjugated MMAE payload exposure metrics. At the median simulated Teliso‐V conjugate C_avg, the 1.9 mg/kg Q2W regimen showed a predicted ORR of 31.4% (24.2%–39.0%), which is generally consistent with the observed ORR of 29% in the LUMINOSITY phase 2 study [18]. Higher Teliso‐V conjugate exposure (C_avgC1) was associated with longer DoR, PFS per ICR, and OS. None of the covariates evaluated were found to be significant for any of the exposure‐efficacy models.

The exposure‐safety analyses showed strong trends of Teliso‐V conjugate exposure‐dependent increases in the following safety events: grade ≥ 2 and grade ≥ 3 peripheral neuropathy, grade ≥ 2 corneal neuropathy, all‐grade ILD/pneumonitis, grade ≥ 2 ILD/pneumonitis, and drug interruptions or discontinuations due to AEs. Grade ≥ 3 TEAEs were correlated with unconjugated MMAE payload exposure metrics only. The positive exposure‐safety relationships are managed in the clinical setting through dose modifications and/or dose reductions [16]. The predicted event rates at the median simulated Teliso‐V conjugate C_avg for 1.9 mg/kg Q2W were 9.60% (grade ≥ 3 peripheral neuropathy), 15.0% (grade ≥ 2 corneal epitheliopathy), and 10.2% (any grade ILD/pneumonitis). None of the evaluated covariates were found to be significant for any of the exposure‐safety models.

A limitation of the efficacy analyses is the limited sample size of the 1.6 mg/kg Q2W dose (N = 25) compared to the 1.9 mg/kg Q2W (N = 168) dose. Additionally, enrollment into these dose groups occurred in a non‐randomized fashion due to which study results may be subject to potential bias and confounding. As such, randomization of dosing arms enables robust dose–response evaluation and benefit–risk characterization [24].

In summary, the population PK modeling showed no clinically relevant differences in PK of the Teliso‐V conjugate and MMAE based on key intrinsic factors and supports no dose adjustment in those populations. Additionally, ER analysis indicated that efficacy was greater with higher Teliso‐V conjugate exposures, and safety is influenced by ADC exposure and MMAE payload exposure. Simulations based on the ER analyses predict that the 1.9 mg/kg Q2W provides higher efficacy over a lower dose, with a safety profile that is generally managed by dose modification or reduction. The ER analyses supported the label‐approved 1.9 mg/kg Q2W dose of Teliso‐V as appropriate for achieving the desirable antitumor activity while balancing safety.

Author Contributions

P.B., H.B., and V.S. wrote the manuscript. All authors reviewed and edited the manuscript. A.P., S.M., R.M.M., and B.E. designed the research. H.B., V.S., B.E., P.B., C.R., and A.P. performed the research. H.B., V.S., and B.E. analyzed the data.

Funding

This work was supported by AbbVie Inc. AbbVie participated in the design, study conduct, analysis, and interpretation of data, as well as the writing, review, and approval of the publication.

Conflicts of Interest

H.B., P.B., B.E., V.S., C.R., S.M., R.M.M., and A.P. are employees of AbbVie and may hold AbbVie stock, stock options, and/or patents.

Supporting information

Data S1: Supporting Information.