In the JAVELIN Solid Tumor studies (NCT01772004), the use of avelumab was evaluated against a variety of different malignancies. In a phase 1a dose escalation trial, the experimental treatment was tested at four different dosing levels (1 mg/kg, 3 mg/kg, 10 mg/kg, and 20 mg/kg) with cohort expansions based on these dose levels. In this manner, additional information was gleaned regarding safety, target occupancy, and pharmacokinetics. In this phase 1a portion of the study, patients with several different cancers were included.
In a separate phase 1b cohort for trial for JAVELIN Solid Tumor, avelumab was assessed in progressive or platinum-resistant metastatic recurrent NSCLC. Interestingly, genetic biomarkers were not utilized for exclusion from this portion of the study. Notably, patients were not excluded from participation based on their PD-L1 expression levels or the presence of other biomarkers (e.g., KRAS mutation, ALK translocation, or EGFR status). These portions of the study were directed by James Gulley, MD, PhD, Head, Immunotherapy Section, NCI. The results of these trials have been published recently (Phase 1a study: Lancet Oncology 2017;18:587-598; Phase 1b study: Lancet Oncology 2017;18:599-610).
Avelumab is a fully human anti-PD-L1 immunoglobulin G1 (IgG1) monoclonal antibody that disrupts the immunosuppressive PD-1/PD-L1 interaction. When elaborating on the novelty of this antibody, Gulley commented, “In addition to reducing the PD-1/PD-L1 immunosuppression of CD8+ T cells, avelumab, by virtue of its unmodified Fc portion, which interacts with CD16 expressed on the surface of natural killer (NK) cells, can potentially mediate antibody-dependent cellular cytotoxicity (ADCC). This provides an additional mechanism by which the patient's immune system can target the cancer.”
One potential concern prior to these trials was the issue of the antibody targeting the PD-L1 ligand, which, in addition to being expressed on the surface of tumor cells, is also expressed on the surface of lymphocytes. “This was a clear cause of concern; however, it was hoped that the tumor cells could be targeted selectively, based on their overexpression of PD-L1 relative to those levels found on the lymphocytes,” Gulley explained.
On March 23, 2017, the FDA approved the use of avelumab for use in adult and pediatric Merkel cell carcinoma patients, regardless of prior therapies, largely on the basis of results obtained in the phase II JAVELIN Merkel 200 clinical trial.
JAVELIN Phase 1a Study
This portion of the JAVELIN trial was a single-center, open-label, phase 1a dose-escalation study that evaluated avelumab against a variety of different malignancies at four different dosing levels (1-, 3-, 10-, and 20 mg/kg). A standard 3+3 cohort design was employed, which sequentially assigned patients at trial entry using a 3+3 dose escalation algorithm and was dependent upon the dose-limiting toxicities during the first 3-week (primary endpoint) assessment. The primary endpoint, which was determined during the initial phase of the study, was the presence of dose-limiting toxicities. Secondary endpoints of safety (treatment-emergent/related adverse events (AEs)), pharmacokinetic and pharmacodynamic profiles, best overall response (RECIST v1.1), and antidrug antibody formation were assessed in those patients enrolled in the dose-escalation part of the trial.
Key eligibility criteria included the following:
- age ( ≥ 18 years);
- ECOG performance status of 0-1;
- a 3-month or greater life expectancy;
- cytologically- or histologically-confirmed locally advanced or metastatic tumor for which the standard therapy had failed or no standard therapy exists;
- measurable disease using RECIST v1.1 criteria; and
- appropriate end-organ functioning (as defined by these laboratory values: creatinine clearance > 50 mL/min, white blood cell count of ≥ 3x109 cells/L, absolute neutrophil count of ≥ 1.5x109 cells/L, lymphocyte count of ≥ 0.5x109 cells/L, platelet count of ≥ 1.0x1011 platelets/L; hemoglobin concentration of 9 g/dL, total bilirubin concentration ≤ 1.5x upper limit of normal (ULN), and alanine aminotransferase and aspartate aminotransferase levels ≤ 2.5x ULN).
Patients were not excluded from this portion of the study based on the presence of any known biomarkers (e.g., EGFR, ALK translocation, KRAS, etc.). Exclusion criteria included prior use of a T cell targeting checkpoint inhibitor, CNS metastases, another cancer diagnosis within the previous 5 years, known autoimmune disorder or sensitivity to monoclonal antibodies, prior stem cell or organ transplant, pregnancy or lactating, or progressive disease. Additionally, other cancer therapies were not allowed during the study, with the exception of palliative bone-directed radiotherapy or for metastatic castration-resistant cancer, the use of androgen-deprivation therapy.
Phase 1a Study Results
From January 2013 to October 2014, 53 patients were enrolled in this portion of the study from the Center for Cancer Research of the NCI at the following dose levels: 1 mg/kg (n=4); 3 mg/kg (n=13); 10 mg/kg (n=15); 20 mg/kg (n=21). Patients were given the experimental treatments intravenously every 2 weeks. The dose-limiting toxicity analysis set consisted of 18 patients, three each for the 1- and 3 mg/kg dosing levels and six each for the 10- and 20 mg/kg levels.
When asked about the results from those 18 patients, Gulley responded, “Since only one patient displayed a dose-limiting toxicity at the highest dosing level (20 mg/kg), the maximum tolerated dose was not reached. This grade 3 AE (elevated blood creatine phosphokinase levels) was observed in a metastatic thymoma patient.”
The safety analysis group consisted of all 53 enrollees. The most common AEs encountered were fatigue (40%), influenza-like symptoms (21%), fever (15%), and chills (11%).
“There were nine grade 3-4 treatment-related AEs. The most frequent of these were autoimmune disorder (n=3; two patients at 10 mg/kg and one at 20 mg/kg), increased blood creatine phosphokinase levels (n=2; two patients at 20 mg/kg), and aspartate aminotransferase (n=2; one patient each at 1 mg/kg and 10 mg/kg),” Gulley noted.
Serious treatment-related AEs were noted in six of the 53 patients, three each at the 10- and 20 mg/kg dosing levels. The serious treatment-related AEs observed in the three 10 mg/kg patients included autoimmune disorder (n=2), fatigue (n=1), influenza-like illness (n=1), and lower abdominal pain (n=1), while the serious treatment-related AEs observed in the three 20 mg/kg patients included autoimmune disorder (n=1), dysphonia (n=1), increased amylase (n=1), and myositis (n=1).
In addition to the safety data obtained, some evidence of clinical activity was observed in these patients.
“Despite the fact that this was a heavily-pretreated patient population with a number of different malignancies, partial confirmed or unconfirmed responses were noted in four of 53 patients, while 30 patients displayed stable disease (57%),” Gulley elaborated.
The data obtained from pharmacokinetic (PK) analyses (n=86; consisting of the 53 patients in the phase 1a dose escalation cohort plus 33 patients from the same institution with thorough PK sampling) showed dose-proportional exposures between the 3- and 20 mg/kg dosing levels. For the highest dosing levels (10- and 20 mg/kg), a half-life of 95-99 hours was obtained.
Pharmacodynamic profiling included the assessment of target occupancy of PD-L1 on CD3+ T cells by avelumab in 21 patients on day 15 (last day) of cycle 1, but before the second dose. The mean PD-L1 target occupancy levels were as follows: 1 mg/kg–76 percent (n=3); 3 mg/kg–90 percent (n=7); 10 mg/kg–93 percent (n=4); 20 mg/kg–87 percent (n=7).
During the course of these studies, two of the 3 mg/kg patients developed antidrug antibodies. For one patient, there was a single occurrence, while for the other, the antibodies were observed at several time points.
One concern prior to these studies was the potential effect of avelumab on lymphocytes, which also express the PD-L1 ligand. “The experimental treatment appeared to have no negative effects on the lymphocytes in this portion of the study; there were no decreases in the numbers of immune cell counts in these patients, thus addressing this potential concern,” Gulley revealed.
As a result of the data obtained from safety and pharmacokinetic/pharmacodynamics studies, the 10 mg/kg dosing with a 2-week dosing schedule was selected for use in subsequent trials.
JAVELIN Phase 1b Study
This portion of the JAVELIN Solid Tumor clinical trial was a dose expansion cohort of an open-label, phase 1b, multi-center evaluated the use of avelumab in platinum-treated, platinum-resistant or progressive, metastatic or recurrent NSCLC patients.
Key eligibility criteria included the following:
- age (≥18 years);
- ECOG performance status of 0-1;
- a 3-month or greater life expectancy;
- no presence or history of CNS metastases;
- MRI- or CT-measurable disease using RECIST v1.1 criteria; and
- cytologically- or histologically-confirmed stage IIIB or IV NSCLC having either squamous- or non-squamous histology and disease progression after previous platinum-based doublet therapy for metastatic disease.
Although patients were not excluded based on the presence of biomarkers (e.g., PD-L1 expression level, EGFR or KRAS mutations, ALK translocations), a biopsy or archival sample was required for biomarker analysis. The same renal, hematological, and hepatic functioning criteria described for the prior 1a cohort also applied to these patients as well. The primary endpoint of this study was the assessment of safety and tolerability. Additionally, patient response and survival data were obtained and stratified according the PD-L1 expression levels.
When describing this patient population, Gulley commented, “We did not exclude patients from this cohort on the basis of their biomarker status; we wanted to include as many patients as possible, even including those who had undergone more than 2 prior chemotherapy regimens.”
Phase 1b Study Results
From September 2013 to June 2014, 184 NSCLC patients were enrolled at 58 academic hospitals or cancer treatment centers across the U.S. Patients were dosed with avelumab at 10 mg/kg IV once every 2 weeks until either disease progression or toxicity was noted.
The most commonly-encountered treatment-related AEs of any grade were fatigue (n=46; 25%), infusion-related reaction (n=38; 21%), and nausea (n=23; 13%). Treatment-related AEs of grade 3 or worse were noted in 23 of the 184 patients (13%), and of these, the most frequently-observed were infusion-related reaction (n=4) and increased lipase levels (n=3). Serious treatment-related AEs occurred in 16 patients (9%) and, of these, the most common were infusion-related reaction (n=4) and dyspnea (n=2).
In this cohort, 22 of 184 patients showed a confirmed objective response and, of these, one patient had a complete response while 21 had a partial response. In addition, 70 patients displayed stable disease. These combined results showed disease control (confirmed objective response + stable disease) in 92 of the total 184 patients (50%).
As part of these studies, objective response (OR), progression-free survival (PFS), and overall survival (OS) were compared for patients of differing PD-L1 expression levels on their tumor- and tumor-associated immune cells (n=142).
For the PD-L1 cutoff at ≥1 percent for tumor cells, for those patients at or above this limit (n=122) the following data were obtained: OR-17 (14%); median PFS-12.0 weeks; PFS at 48 weeks-21 percent; median OS-8.9 months; OS at 12 months-39 percent. Those patients below this limit (n=20) afforded the following results: OR-2 (10%); median PFS-5.9 weeks; PFS at 48 weeks-5 percent; median OS-4.6 months; OS at 12 months-36 percent. With this cutoff PD-L1 level, the hazard ratio (HR) for median PFS was 0.45, while for median OS, the HR was 0.64.
With the PD-L1 cutoff at ≥25 percent for tumor cells, for those patients at or above this limit (n=53), analysis yielded the following results: OR-9 (17%); median PFS-11.9 weeks; PFS at 48 weeks-28 percent; median OS-8.44 months; OS at 12 months-28 percent. Those patients below this limit (n=89) had the following results: OR-10 (11%); median PFS-10.79 weeks; PFS at 48 weeks-15 percent; median OS-8.57 months; OS at 12 months-48 percent. With this cutoff PD-L1 level, the HR for median PFS was 0.79, while for median OS, the HR was 1.14.
With the PD-L1 cutoff at ≥10 percent for tumor-associated immune cells, those patients at or above this limit (n=27) had the following results: OR-4 (15%); median PFS-8.4 weeks; PFS at 48 weeks-not evaluable; median OS-8.5 months; OS at 12 months-30 percent. Those patients below this limit (n=115) yielded the following results: OR-15 (13%); median PFS-11.3 weeks; PFS at 48 weeks-19 percent; median OS-8.9 months; OS at 12 months-41 percent. With this cutoff PD-L1 level for the immune cells, the HR for median PFS was 1.19, while for median OS, the HR was 1.20.
When discussing the results of these two cohorts, Gulley noted, “One theoretical concern that was settled in these studies was the potential targeting of lymphocytes by this antibody; there were no decreases seen in the levels of immune cells for these patients.
“In addition to assisting a T cell-mediated immunological tumor response, this antibody, by virtue of its Fc portion which binds to CD16 on NK cells, also has the potential to initiate ADCC, and this has been demonstrated in vitro; however, further analyses need to be done to verify if this in fact occurs clinically.” This sets it apart from other anti-PDL1 antibodies and provides additional rationale for combination studies that can augment NK activity.
“Currently, there are two ongoing phase III trials for avelumab in lung cancer, and there is an interesting study being done by Arun Rajan evaluating this antibody in thymoma patients,” Gulley noted. “Another ongoing phase 1b trial in solid tumors is evaluating the use of avelumab with NHS-IL 12, an antibody-cytokine conjugate that is designed to deliver a genetically-modified version of IL 12 to the tumors (NCT02994953).”
When asked about other combination studies, Gulley replied, “One particularly interesting trial is evaluating the combination of avelumab, an adenoviral CEA vaccine, FOLFOX (5-fluorouracil, leucovorin, and oxaliplatin), and bevacizumab in previously-untreated metastatic colorectal cancer patients (NCT03050814).”
Richard Simoneaux is a contributing writer.