Abstract
Purpose: This phase I trial assessed the safety, tolerability, and preliminary antitumor activity of lifastuzumab vedotin (LIFA), an antibody–drug conjugate of anti-NaPi2b mAb (MNIB2126A) and a potent antimitotic agent (monomethyl auristatin E).
Patients and Methods: LIFA was administered to patients with non–small cell lung cancer (NSCLC) and platinum-resistant ovarian cancer (PROC), once every 3 weeks, by intravenous infusion. The starting dose was 0.2 mg/kg in this 3+3 dose-escalation design, followed by cohort expansion at the recommended phase II dose (RP2D).
Results: Overall, 87 patients were treated at doses between 0.2 and 2.8 mg/kg. The MTD was not reached; 2.4 mg/kg once every 3 weeks was selected as the RP2D based on overall tolerability profile. The most common adverse events of any grade and regardless of relationship to study drug were fatigue (59%), nausea (49%), decreased appetite (37%), vomiting (32%), and peripheral sensory neuropathy (29%). Most common treatment-related grade ≥3 toxicities among patients treated at the RP2D (n = 63) were neutropenia (10%), anemia (3%), and pneumonia (3%). The pharmacokinetic profile was dose proportional. At active doses ≥1.8 mg/kg, partial responses were observed in four of 51 (8%) patients with NSCLC and 11 of 24 (46%) patients with PROC per RECIST. All RECIST responses occurred in patients with NaPi2b-high by IHC. The CA-125 biomarker assessed for patients with PROC dosed at ≥1.8 mg/kg showed 13 of 24 (54%) had responses (≥50% decline from baseline).
Conclusions: LIFA exhibited dose-proportional pharmacokinetics and an acceptable safety profile, with encouraging activity in patients with PROC at the single-agent RP2D of 2.4 mg/kg.
Translational Relevance
NaPi2b (SLC34A2) is a multitransmembrane, sodium-dependent phosphate transporter with high expression in non–small cell lung cancer (NSCLC) and ovarian cancer. We tested lifastuzumab vedotin (DNIB0600A), an antibody–drug conjugate (ADC) comprising an anti-NaPi2b antibody conjugated to monomethyl auristatin E, a potent microtubule inhibitor. In this phase I study, results showed limited efficacy in patients with NSCLC, but were promising for patients with platinum-resistant ovarian cancer where radiographic and serologic responses were demonstrated. Notably, objective responses were limited to patients with high tumor NaPi2b expression, supporting the hypothesis that NaPi2b target expression may be important for clinical benefit from an anti-NaPi2b ADC treatment. With their mechanism of selective delivery of cytotoxic agents to tumor cells leading to increased therapeutic index of cell-killing agents for treating cancer, ADC drug formats can potentially offer improved tolerability and therapeutic window.
Introduction
Non–small cell lung cancer (NSCLC) is the most common cause of cancer-related death in the United States in men and women, and ovarian cancer is the fifth most common cause of cancer-related death in women (1). In both cancer types, the majority of patients are diagnosed at advanced stage and succumb to their disease. For NSCLC, newer targeted therapies based on genomic alterations and PD-L1 status have changed the treatment landscape (2). Initial treatments for ovarian cancer include platinum-based combination chemotherapy; retreatment with platinum-based doublets following progression may be beneficial for patients with long chemotherapy-free intervals (3). Unfortunately, recurrence or progression following initial therapy occurs in both diseases. In addition, therapies are often associated with significant systemic toxicity. Hence, treatments that can provide meaningful clinical benefit with acceptable, if not superior, safety profiles remain a significant unmet need for patients with NSCLC or platinum-resistant ovarian cancer (PROC).
NaPi2b, also known as SLC34A2, is a multitransmembrane, sodium-dependent phosphate transporter (4) that has demonstrated high frequency of expression in NSCLC and PROC patient tumor samples (5–7). Lifastuzumab vedotin (DNIB0600A) is an antibody–drug conjugate (ADC) that comprises a humanized IgG1 anti-NaPi2b mAb (MNIB2126A) and a potent antimitotic agent [monomethyl auristatin E (MMAE)]; a protease-labile linker [maleimidocaproyl-valine-citrulline-p-aminobenzyloxycarbonyl] conjugates the antibody to drug. Upon binding its target antigen, the ADC–antigen complex is internalized and MMAE is released intracellularly, resulting in cell death through the inhibition of cell division and growth (8–10). In vivo xenograft models and in vitro cell models of NaPi2b-expressing cancer have demonstrated the antiproliferative effects of anti-NaPi2b ADC (5). Furthermore, the tolerability and therapeutic window of MMAE was improved in the ADC format with selective toxicity directed toward NaPi2b-positive cells (5). The pharmacokinetics for anti-NaPi2b ADC was linear in rodent and nonhuman primate studies, driven mainly by the anti-NaPi2b antibody and consistent with other humanized mAbs (5). On the basis of favorable preclinical data, we performed a clinical study of lifastuzumab vedotin (LIFA) in previously treated advanced NSCLC and PROC.
Patients and Methods
Study design
Study NCT01363947 (ClinicalTrials.gov) was conducted in accordance with Good Clinical Practice guidelines and the Declaration of Helsinki. Written informed consent was obtained from all patients prior to enrollment, in agreement with approved protocols from ethics committees at all study sites. This was a phase I, multicenter, open-label, dose-escalation study of LIFA administered as a single agent to patients with nonsquamous NSCLC, or patients with nonmucinous PROC. The starting dose for LIFA was 0.2 mg/kg administered by intravenous infusion every 3 weeks. The primary objectives were to evaluate the safety and tolerability of LIFA, to determine the maximum tolerated dose (MTD) and dose-limiting toxicities (DLTs), and to identify the recommended phase II dose (RP2D) at or below the MTD. Secondary objectives included assessments for pharmacokinetics, efficacy, and immunogenicity of LIFA. The study employed a traditional 3+3 dose-escalation design (11) to determine the MTD, followed by cohort expansion at the RP2D to further characterize safety and activity of LIFA. Patients were monitored for DLTs after the first infusion during the DLT assessment window (cycle 1, days 1–21). A minimum of three patients were enrolled at each dose level. In the absence of a safety signal, successive higher-dose cohorts were enrolled with dose increments of 100%, 75%, 71%, 50%, 33%, 33% again, and 31%. The dose increment was reduced to ≤50% in the presence of a safety signal (a DLT in the first cycle or ≥2 patients in any dose level cohort experiencing grade ≥2 study drug−related toxicities).
Patients
Eligible patients were 18 years of age or older with histologic documentation of incurable, locally advanced, or metastatic disease (nonsquamous NSCLC or nonmucinous PROC) that had progressed on or following prior chemotherapy and for which no standard therapy existed. Other inclusion criteria were: Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1; availability and willingness to provide an adequate archival tumor sample; measurable disease; absolute neutrophil count ≥1,500/μL, hemoglobin ≥9 g/dL, and platelet count ≥100,000/μL; total bilirubin ≤1.5 × the upper limit of normal (ULN); aspartate transaminase and alanine transaminase levels ≤2.5 × the ULN; serum creatinine ≤1.5 × the ULN or creatinine clearance ≥50 mL/minute based on a 24-hour urine collection; international normalized ratio ≤1.5 × the ULN; and activated partial thromboplastin time ≤1.5 × the ULN. Exclusion criteria were: major surgical procedure or antitumor therapy (chemotherapy, biologic, experimental, or hormonal) within 4 weeks prior to day 1; active infection; current grade >1 toxicity (except alopecia, anorexia, and fatigue) from prior therapy or grade >1 neuropathy from any cause; history of severe allergic or anaphylactic reactions to mAb therapy; clinically significant history of liver disease; untreated or active central nervous system metastases; evidence of significant uncontrolled concomitant diseases including cardiac, nervous system, pulmonary, renal, hepatic, endocrine, or gastrointestinal disorders; and serious nonhealing wound or fracture.
Safety
Adverse events were graded according to the NCI Common Terminology Criteria for Adverse Events, v4.0. The following adverse events considered by the investigator to be related to LIFA and occurring during days 1–21 of cycle 1 in dose-escalation cohorts were considered a DLT: grade ≥3 nonhematologic toxicity not attributable to disease progression or other identifiable cause [excluding grade 3 diarrhea that responded to standard-of-care (SOC) therapy; grade 3 nausea or vomiting, in the absence of premedication that responded to SOC; and grade 3 infusion reactions, in the absence of premedication that returned to baseline with SOC treatment within 24 hours], grade ≥4 neutropenia lasting >5 days or associated with fever, grade ≥4 anemia, and grade 3 thrombocytopenia complicated by clinically significant bleeding requiring medical intervention or grade ≥4 thrombocytopenia. Dose modification and retreatment guidelines were described in detail in the protocol. Given the predictive nature of lung toxicity based on preclinical studies, this risk was addressed in patients on a case-by-case scenario according to pulmonary signs and symptoms, and pulmonary function tests were performed as clinically indicated.
Pharmacokinetics and immunogenicity
Pharmacokinetic parameters of LIFA were derived from the serum and plasma concentration–time profiles, including exposure (AUC), maximum concentrations (Cmax), minimum concentrations (Cmin), clearance, half-life (t1/2), and volume of distribution. Key analytes were measured using validated assays to characterize LIFA pharmacokinetics: total antibody (fully conjugated, partially deconjugated, and fully unconjugated antibody), conjugate (evaluated as antibody-conjugated MMAE), and unconjugated MMAE (12, 13). Baseline prevalence and postbaseline incidence of antidrug antibodies (ADA) to LIFA were assessed using validated assays using a tiered testing strategy (12, 14, 15).
Biomarkers
For determination of NaPi2b protein expression, a fully automated IHC assay was developed using the anti-NaPi2b (10H1) mouse mAb (Genentech, Inc.) and ultraView DAB IHC Detection (Ventana Medical Systems, Inc.). Formalin-fixed, paraffin-embedded archival tissues were processed whereby NaPi2b membranous staining level was scored according to the following algorithm with ≥50% stained cells to qualify as positive in each category: IHC = 3+ for samples with predominantly strong staining intensity; IHC = 2+ for samples with predominantly moderate staining intensity; IHC = 1+ for samples with predominantly weak staining intensity; and IHC = 0 for samples with very weak or no staining in >50% of tumor cells. NaPi2b H score was defined as [1 × (% cells staining at IHC 1+)] + [2 × (% cells staining at IHC 2+)] + [3 × (% cells staining at IHC 3+)].
Pharmacodynamic measurements of serum cancer antigen 125 (CA125) were assessed in patients with PROC with use of the Gynecologic Cancer Intergroup criteria. Patients with baseline CA125 that was ≥2 × ULN were considered evaluable for response, defined as when the CA125 level had decreased by ≥50% and confirmed a minimum of 28 days following the first documented decrease. Normalization of the CA125 level was designated a complete CA125 response.
Antitumor activity
Response assessments at protocol-defined time points were assessed via modified RECIST 1.1 (16) and included objective response outcomes [complete response, partial response (PR), stable disease, progressive disease (PD), not evaluated, and duration of objective response (DoR)]. Progression-free survival was defined as the time from treatment day 1 to disease progression or death (whichever occurred first) within 30 days of the last study drug administration, and was censored at the day of last tumor assessment showing no progression within 30 days of the last study drug administration.
Statistical analysis
The planned enrollment for this study was 20–40 patients in the dose-escalation cohorts, up to 50 patients in the two lung cancer dose-expansion cohorts, and up to 16 in an ovarian cancer dose-expansion cohort. The final analysis was based on patient data collected through the last patient's discontinuation. All analyses included patients who received any amount of LIFA according to the assigned dose level. Safety data were assessed through summaries of adverse events, changes in laboratory test results, and changes in vital signs. Response outcomes and progression endpoints were assessed for all patients and by indication using RECIST v1.1.
Results
Patient population
From June, 2011, to August, 2015, a total of 87 patients (57 NSCLC and 30 PROC) were enrolled in dose-escalation (0.2–2.8 mg/kg) and dose-expansion (2.4 mg/kg) cohorts (Fig. 1). Patient demographic and baseline characteristics are summarized in Table 1. The most common histopathologic subtype was adenocarcinoma (68%) for patients with NSCLC and serous (70%) for patients with PROC. Pretreatment CA-125 levels for patients with PROC ranged from 6–19,304 U/mL, with a median of 588. Patients with NSCLC received a median of two prior systemic therapies in metastatic setting (range, 0–9), and patients with PROC received a median of four prior systemic therapies in metastatic setting (range, 0–11).
Study schema.
Baseline patient characteristics.
Study drug exposure
Across all dose cohorts, the median number of cycles was 4.0 (range, 1–28) and the median number of days on treatment was 63 (range, 0–582). The duration of treatment in days was similar between NSCLC (median, 63; range, 0–553 days) and PROC (median, 67; range, 0–582). Reasons for discontinuation included disease progression (64%), physician decision (15%), adverse events (14%), patient withdrawal (3%), and death (2%). The reason for discontinuation was missing for one patient (2%).
At 2.4 mg/kg (RP2D), patients with PROC received a median of 5.5 cycles of treatment (range, 1–18) during a median of 95 days of treatment (range, 0–399); patients with NSCLC received a median of four cycles of treatment (range, 1–26) during a median of 64 days of treatment (range, 0–553).
Safety
All 87 patients experienced at least one adverse event regardless of their relationship to the study drug (Supplementary Table 1). A total of 77 (89%) patients experienced adverse events related to the study drug (Table 2), the most common of which were fatigue (52%), nausea (38%), decreased appetite (33%), peripheral sensory neuropathy (29%), and vomiting (24%). A total of 41 (47%) patients experienced adverse events of grade 3–5 regardless of attribution to study drug (Supplementary Table S2), the most common of which were neutropenia and pneumonia (each: 8%). Among all patients, one DLT (grade 3 dyspnea) was reported in a patient with NSCLC treated at a dose of 1.8 mg/kg. Study treatment was discontinued because of adverse events in 10 (16%) patients treated at 2.4 mg/kg.
Adverse events related to study drug occurring in 10 or more patients overall.
The MTD was not reached on this study and the maximum administered dose (MAD) was 2.8 mg/kg. Upon evaluation of the safety data of the six patients treated at the MAD, the dose of 2.4 mg/kg was established as the RP2D. Even though no DLTs were reported at the 2.4 or 2.8 mg/kg dose levels, the 2.4 mg/kg was selected for an overall better tolerability of the regimen, in particular in relation to treatment-related events of fatigue, peripheral sensory neuropathy, and gastrointestinal toxicities (nausea, vomiting, and decreased appetite). In addition, 2.4 mg/kg had been selected as the RP2D for numerous other Genentech MMAE-ADCs with similar drug-antibody ratios.
Because of the mechanism of action of MMAE, peripheral neuropathy is a potential risk of LIFA. One reason that peripheral neuropathy events were likely not reported at doses less than 1.8 mg/kg (Table 2) may be due to the small number of patients dosed at those levels, and potential early treatment withdrawal. Any grade adverse events of peripheral neuropathy and associated terms (identified using the broad Standardized MedDRA Queries for peripheral neuropathy) regardless of attribution to study drug were reported in 36 (41%) patients. These adverse events included peripheral sensory neuropathy (29%), paresthesia (9%), hypoesthesia and muscular weakness (each: 5%), peripheral neuropathy (2%), and peripheral motor neuropathy, gait disturbance, neurotoxicity, and peroneal nerve palsy (each: 1%). Three patients experienced grade 3 neuropathy events regardless of attribution, including peripheral sensory neuropathy (2%), and peripheral sensory neuropathy and peripheral motor neuropathy (1%). Peripheral neuropathy resulted in the discontinuation of four patients (three NSCLC and one PROC) at 2.4 mg/kg and 1 patient with PROC at 2.8 mg/kg. There appeared to be no correlation between prior exposure to taxane and the risk of developing peripheral neuropathy on study. Of the 61 patients who received prior taxane, 43% of patients experienced events of peripheral neuropathy and associated terms; among the 26 patients who did not receive prior taxane, 38% of patients experienced events of peripheral neuropathy and associated terms.
Because of the expression of NaPi2b in normal lung tissue and based on results from Good Laboratory Practice multiple dose toxicity study in cynomolgus monkeys (data not shown), pulmonary toxicity is another potential risk of LIFA. Nonclinical studies performed with NaPi2b-MMAE suggested a toxicity profile that was consistent with microtubule inhibitors, with possible on-target toxicity in lung tissue. There was no clear dose dependency, although lung inflammation was seen in one of six animals treated at the highest preclinical dose (6 mg/kg). To identify adverse events potentially suggestive of pulmonary toxicity clinically, all adverse events reported in DNIB0600A trials were reviewed and medically adjudicated for inclusion into a dedicated analysis. Using this methodology, 54 patients (63%; NSCLC 43% and PROC 20%) experienced grade 1–3 pulmonary toxicity regardless of attribution (Supplementary Table S3). Pulmonary events of grade ≥3 were reported in 11 (13%) patients (Supplementary Table S4), including adverse events of dyspnea, pneumonia, and upper respiratory infection considered related to study drug in three (3%) patients with NSCLC. One of these patients with NSCLC an 80-year-old white male, who died on study day 89 from grade 5 respiratory failure considered related to study treatment, after treatment with four doses of DNIB0600A; the patient had extensive metastatic disease in the lung, including pleural effusion at baseline. Other pulmonary events considered related to the study drug were grade 1–2 and were manageable with observation or supportive care.
Clinically, it appeared that patients with NSCLC may have had an increased risk of pulmonary toxicity. There were 13 pulmonary toxicities that were serious adverse events occurring in 11 patients (nine patients with NSCLC and two patients with ovarian cancer). The events were considered related in three patients (pneumonia and upper respiratory tract infection in one patient, and dyspnea and respiratory failure in one patient each). All events resolved prior to discontinuing from the study, except for the respiratory failure event, which resulted in death. There was no clear evidence that the pulmonary toxicity was cumulative.
Pharmacokinetics
The pharmacokinetic properties of the two analytes of LIFA conjugate, evaluated as antibody-conjugated MMAE (Supplementary Table S5) and total antibody (Supplementary Table S6), were generally similar (Fig. 2), as characterized by the relatively low estimate of volume of distribution under steady state (VSS), long t1/2, and low total clearance of drug (CL), suggesting that the disposition of the MMAE component of the ADC appeared to be driven by the antibody component. Maximum concentrations of antibody-conjugated MMAE and total antibody were reached at the end of infusion, followed by a multi-exponential decline. Comparison of mean dose-normalized AUCinf and Cmax values for both the antibody-conjugated MMAE and total antibody analytes also revealed linear pharmacokinetic behavior across the tested doses. The Cmax for unconjugated MMAE occurred approximately 2–4 days after dosing across all the doses tested (Supplementary Table S7). This trend is suggestive of a delayed formation of unconjugated MMAE after intravenous dosing of LIFA.
Pharmacokinetic profiles of total antibody, ADC, and unconjugated drug in patients with ovarian cancer (dash lines) and patients with NSCLC (solid lines) at dose level of 2.4 mg/kg.
Immunogenicity
Eight patients (9%) had evidence of ADA at pretreatment baseline. Of these patients, two had treatment-enhanced ADA responses and six were treatment unaffected. Eighteen (21%) patients had evidence of ADA at postbaseline samples across all treatment groups. Of these 18 patients, 16 had treatment-induced ADA responses and three patients experienced adverse events potentially related to immunogenicity. However, the relationship between the presence of ADAs and the observed events is unknown. ADA status showed no apparent impact of ADA response on LIFA drug exposure.
Antitumor activity (RECIST and CA125)
All 87 patients were evaluable for efficacy (Supplementary Table S8). Confirmed objective responses were observed in patients with ovarian and lung cancer at doses ≥1.8 mg/kg, including in the expansion cohort at the RP2D level of 2.4 mg/kg. In the NSCLC cohorts, PRs were seen in one of four (25%) and in three of 45 (7%) patients at 1.8 mg/kg and 2.4 mg/kg, respectively (Fig. 3); DoR ranged from 132 to 317 days (median, 161; Fig. 4). In PROC, PRs were seen at 1.8 mg/kg in one of two (50%), at 2.4 mg/kg in seven of 18 (39%), and at 2.8 mg/kg in three of four (75%) patients (Fig. 3); DoR ranged from 43 to 561 days (median, 342.0; Fig. 4).
Investigator-assessed best radiographic response in patients with ovarian cancer (top) and patients with NSCLC (bottom). IHC assessments of NaPi2b expression were used to categorize patients according to NaPi2b IHC levels of 0, 2, and 3. SD, stable disease; PD, progressive disease.
Time on study for patients with ovarian cancer (top) and patients with NSCLC (bottom). IHC assessments of NaPi2b expression were used to categorize patients according to NaPi2b IHC levels of 0, 2, and 3. Reasons for discontinuation from study are indicated [progressive disease (PD), adverse event (AE), physician decision (Physician), and subject withdrawal (Subject)]. Yellow diamonds represent PRs.
Among the 30 patients with PROC, 25 were considered evaluable for CA125 response and five patients with baseline CA125 levels measured below 2 × ULN were unevaluable. No CA125 responses were identified in three evaluable patients treated at dose levels of 0.2–1.2 mg/kg. For patients treated at dose levels of 1.8–2.8 mg/kg, one of two patients at 1.8 mg/kg had a complete CA125 response, eight of 16 patients had responses at 2.4 mg/kg including three patients with complete CA125 response, and three of four patients responded at 2.8 mg/kg including one complete CA125 response (Supplementary Figure S1).
Among the patients with NSCLC on this study, 28 (65%) had been treated previously with a taxane. Given that MMAE is mechanistically similar to a taxane, this may in part explain the low response rate seen in the patients with NSCLC. In addition, the median of number of prior therapies for the patients with NSCLC was two (range, 0–9). Late line lung cancer is not a very chemo-sensitive disease. Historically, the response seen in third-line and beyond (3L+) patients treated with single-agent chemotherapy is around 10%, which also may explain the low response rate seen in this study.
Predictive biomarkers
On the basis of the mechanism of action of ADCs, we hypothesized that NaPi2b expression may be required for antitumor activity from LIFA. In the phase I trial, NaPi2b expression was evaluated retrospectively. The IHC cutoff for expression was selected on the basis of preclinical data; Supplementary Fig. S2 is a representative photomicrograph of IHC scores. In the expansion cohorts, tumor NaPi2b expression was evaluable in 18 of 18 patients with PROC showing 6% IHC = 0, 39% IHC = 2+, and 56% IHC = 3+, and in 26 of 30 patients with NSCLC showing 23% IHC = 0, 23% IHC = 2+, and 54% IHC = 3+. While all responses occurred in patients with PROC and NSCLC with IHC = 2+ or IHC = 3+ only tumor NaPi2b expression (Fig. 3), this study did not enroll an adequate number of IHC = 0 patients to formally test the diagnostic hypothesis. Of the 48 patients with NaPi2b IHC = 2+ or IHC = 3+ treated at dose levels of 1.8–2.8 mg/kg, 14 patients had a confirmed PR: three of 26 (12%) patients with NSCLC and 11 of 22 (50%) patients with PROC, respectively. However, larger studies would be needed to assess whether the IHC score is a valuable positive or negative predictor of response to anti-NaPi2b therapy. For patients with NSCLC, approximately 75% of patients enrolled had three or more previous therapies, and limited data suggest a potential trend for less activity in patients with more prior treatment. For patients with PROC, the median number of prior therapies was five; however, there was no apparent trend for activity in relation to prior treatment.
Discussion
Results from this phase I study demonstrated that LIFA has an acceptable safety profile when administered by intravenous infusion every 3 weeks in patients with NSCLC or PROC. On the basis of the protocol-specified criteria, an MTD for single-agent LIFA was not reached and the MAD was 2.8 mg/kg. On the basis of overall safety and tolerability, the RP2D for LIFA was determined to be 2.4 mg/kg. One DLT of grade 3 dyspnea was reported in a patient with NSCLC at the dose level 1.8 mg/kg. The most common LIFA-related adverse events across dose levels were fatigue, nausea, decreased appetite, peripheral sensory neuropathy, vomiting, and alopecia. Neutropenia was the most frequent grade ≥3 adverse event. Adverse events consistent with peripheral neuropathy were the most frequent adverse events leading to study treatment discontinuation.
The pharmacokinetic properties of antibody-conjugated MMAE and total antibody were generally similar. This suggests that the disposition of antibody-conjugated MMAE appears to be driven by the antibody component, given the formation rate-limited kinetics of MMAE. LIFA demonstrated linear pharmacokinetic in the tested dose range of 0.2–2.8 mg/kg at the every 3 weeks schedule for both the antibody-conjugated MMAE and total antibody analytes. On the basis of our data, slight accumulation of antibody-conjugated MMAE and total antibody is expected upon repeated administration of LIFA at every 3 weeks dosing schedules. At the 2.4 mg/kg every 3 weeks dose, the pharmacokinetic parameters for the three analytes were comparable in both expansion cohorts in patients with NSCLC and ovarian cancer. Unconjugated MMAE exposures generally increased with increasing dose, with delayed formation, and had relatively low exposure compared with antibody-conjugated MMAE and relatively long t1/2, suggestive of formation rate-limited kinetics.
To evaluate the correlation between tumor expression of NaPi2b and clinical efficacy, we developed an IHC assay for NaPi2b that was analytically validated before and implemented during this study. Tissue was analyzed with the IHC assay retrospectively. Objective responses were only observed in NaPi2b-high patients (IHC 2+/3+); while NaPi2b target expression may be important for clinical benefit from LIFA treatment, larger studies are needed to provide statistically significant evidence to that effect.
In conclusion, LIFA demonstrates an encouraging safety profile, a reasonable pharmacokinetics profile, and evidence of antitumor activity. While efficacy was limited in patients with NSCLC, patients with PROC demonstrated responses in 11 of 30 (36.7%) patients with robust pharmacodynamic substantiation of CA-125 target modulation that would merit additional studies. Evaluation of NaPi2b expression as a predictive biomarker to further enrich efficacy of LIFA in future PROC studies would also be warranted. However, the development of this molecule has now been halted for reasons unrelated to safety (17), and no further studies are planned.
Disclosure of Potential Conflicts of Interest
D.E. Gerber is a paid consultant for Samsung Bioepsis; reports receiving commercial research grants from Karyopharm, BerGenBio, and AstraZeneca; reports receiving other commercial research support from ArQule, Boehringer-Ingelheim, BerGenBio, Celgene, Genentech, ImmunoGen, Karyopharm, and Peregrine; has ownership interest in Gilead Sciences, Inc and is an unpaid consultant/advisory board member for Boehringer-Ingelheim, Bristol-Myers-Squibb Company, MS, Celgene, Genentech, Guardant, Peregrine, Samsung, Synta, and AstraZeneca. J.R. Infante is a consultant/advisory board member for BioMed Valley Discoveries and Armo Biosciences. M.S. Gordon reports receiving other commercial research support from BMS, Genentech, Roche, Eli Lilly, Celgene, BeiGene, Syndax, Merck, AbbVie, BluePrint, Deciphera, Celldex, Daiichi, Corcept, Endocyte, ESSA, Five Prime, Genocea, Macrogenics, MedImmune, OncoMed, Pfizer, Aeglea, Astex, Calithera, Gilead, ImaginAB, Incyte, Novartis, Plexxikon, Serono, SynDevRx, Tolero, and Tracon and reports unpaid consultant or advisory board relationships with Agenus, Tracon, and Deciphera. S.B. Goldberg reports receiving commercial research grants from AstraZeneca and is a consultant/advisory board member for AstraZeneca, Genentech, Eli Lilly, Boehringer Ingelheim, Bristol-Myers Squibb, Amgen, and Spectrum. M. Martín reports receiving commercial research grants from Novartis and Roche and is a consultant/advisory board member for Roche, Pfizer, Taiho, AstraZeneca, Eli Lilly, Pharmamar, and Novartis. E. Felip reports receiving speakers bureau honoraria from Abbvie, AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Eli Lilly, Merck KGaA, Merck Sharp & Dohme, Novartis, Pfizer, Roche, and Takeda and is a consultant/advisory board member for Abbvie, AstraZeneca, Blue Print Medicines, Boehringer Ingelheim, Bristol-Myers Squibb, Celgene, Eli Lilly, Guardant Health, Janssen, Medscape, Merck KGaA, Merck Sharp & Dohme, Novartis, Pfizer, Roche, Takeda, and Touchtime. M. Martinez Garcia is a consultant/advisory board member for Roche. J.H. Schiller is an employee of Genentech and a paid consultant/advisory board member of Merck and AstraZeneca. D.R. Spigel is a consultant/advisory board member for Genetech/Roche, Novartis, Celgene, Bristol-Myers Squibb, AstraZeneca, Pfizer, Boehringer Ingelheim, Abbvie, Foundation Medicine, GlaxoSmithKline, Eli Lilly, Merck, Moderna, Nektar, Takeda, Amgen, TRM Oncology, Precision Oncology, Evelo, Illumina, and PharmaMar and reports that his employer receives commercial research grants from Genentech/Roche, Novartis, Celgene, Bristol-Myers Squibb, Eli Lilly, AstraZeneca, Pfizer, Boehringer Ingelheim, UT Southwestern, Merck, Abbvie, GlaxoSmithKline, G1 Therapeutics, Neon Therapeutics, Takeda, Foundation Medicine, Nektar, Celldex, Clovis, Daichi, EMD Serono, Acerta, Oncogenex, Astellas, Grail, Transgene, Aeglea, Tesaro, Ipsen, Amgen, and Millenium. J. Cordova is an employee of Roche. V. Westcott is an employee of and has ownership interest (including patents) in Genentech/Roche. Y. Wang is an employee of and has ownership interest (including patents) in Genentech, Inc. D.S. Shames holds ownership interest (including patents) in Roche. Y. Choi is an employee of and has ownership interest (including patents) in Genentech/Roche. R. Kahn is an employee of Kite Pharma/Gilead and holds ownership interest (including patents) in Roche. R.C. Dere is an employee of and has ownership interest (including patents) in Genentech/Roche. D. Samineni is an employee of and has ownership interest (including patents) in Genentech/Roche. J. Xu is an employee of Klus Pharma, is a former employee of Genentech, and has ownership interest in Genentech/Roche. K. Lin is an employee of and has ownership interest (including patents) in Genentech/Roche. K. Wood is an employee of BeiGene Ltd. S. Royer-Joo is an employee of Roche. V. Lemahieu holds ownership interest (including patents) in Genentech. E. Schuth holds ownership interest (including patents) in Genentech. A. Vaze is an employee of and has ownership interest (including patents) in Genentech/Roche. D. Maslyar holds ownership interest (including patents) in Genentech. E.W. Humke is an employee of and has ownership interest (including patents) in Genentech/Roche. H. Burris reports that his employer receives commercial research grants from Genentech/Roche, Bristol-Meyers Squibb, Incyte, AstraZeneca, MedImmune, Macrogenics, Novartis, Boehringer-Ingelheim, Lilly, Seattle Genetics, Merck, Celgene, Agios, Jounce Therapeutics, Moderna Therapeutics, CytomX Therapeutics, GlaxoSmithKline, Verastem, Tesaro, Immunocore, Takeda, Milennium, BioMed Valley Discoveries, TG Therapeutics, Loxo, Vertex, eFFECTOR Therapeutics, Janssen, Gilead Sciences, BioAtla, CicloMed, Harpoon Therapeutics, Jiangsu Hengrui Medicine, Arch, Kyocera, Arvinas, and Revolution Medicines. No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
Conception and design: D.E. Gerber, J.R. Infante, M.S. Gordon, D.S. Shames, Y. Choi, R. Kahn, K. Lin, D. Maslyar, E.W. Humke, H.A. Burris III
Development of methodology: J.R. Infante, D.S. Shames, K. Lin, D. Maslyar, E.W. Humke
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): D.E. Gerber, J.R. Infante, M.S. Gordon, S.B. Goldberg, M. Martín, E. Felip, M. Martinez Garcia, J.H. Schiller, D.R. Spigel, Y. Wang, K. Wood, E.W. Humke
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): D.E. Gerber, J.R. Infante, M.S. Gordon, M. Martín, J.H. Schiller, D.R. Spigel, Y. Wang, D.S. Shames, Y. Choi, R. Kahn, R.C. Dere, D. Samineni, J. Xu, K. Lin, S. Royer-Joo, V. Lemahieu, E. Schuth, A. Vaze, D. Maslyar, E.W. Humke, H.A. Burris III
Writing, review, and/or revision of the manuscript: D.E. Gerber, J.R. Infante, M.S. Gordon, S.B. Goldberg, M. Martín, E. Felip, M. Martinez Garcia, J.H. Schiller, D.R. Spigel, V. Westcott, Y. Wang, D.S. Shames, Y. Choi, R. Kahn, R.C. Dere, D. Samineni, J. Xu, K. Wood, S. Royer-Joo, V. Lemahieu, E. Schuth, A. Vaze, D. Maslyar, E.W. Humke, H.A. Burris III
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Choi, S. Royer-Joo, V. Lemahieu
Study supervision: J.R. Infante, M. Martín, J. Cordova, R. Kahn, A. Vaze, D. Maslyar
Acknowledgments
We thank the patients, study investigators, and staff who participated in this study. Editing and writing support was provided by A. Daisy Goodrich (Genentech, Inc.) and was funded by Genentech, Inc.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Footnotes
Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).
Clinical Trial registration ID: ClinicalTrials.gov NCT01911598
Clin Cancer Res 2020;26:364–72
- Received December 5, 2018.
- Revision received April 2, 2019.
- Accepted September 18, 2019.
- Published first September 20, 2019.
- ©2019 American Association for Cancer Research.
References
Volume 26, Issue 2
