Clinical trial designs are transforming dramatically in order to expedite the discovery and validation of new predictive biomarkers for patients with nonâ€“small cell lung cancer and other types of solid tumors.
Primo N. Lara Jr, MD
Clinical trial designs are transforming dramatically in order to expedite the discovery and validation of new predictive biomarkers for patients with non—small cell lung cancer (NSCLC) and other types of solid tumors, according to a presentation by Primo N. Lara Jr, MD, at the 16th Annual International Lung Cancer Congress (ILCC) held in Huntington Beach, California.1
“If you look at it on the basis of tumor types for which there is an established biomarker in the clinic for us to use to select patients, in non—squamous non–small cell lung cancer, there’s EGFR and ALK,” said Lara, associate director for Translational Research at the UC Davis Comprehensive Cancer Center. “You can see in the other solid tumors, as well as in GIST, there are predictive biomarkers, but there’s not enough, even though they’ve been hyped up over the past decade.”
Biomarkers, which Lara defined as any molecular feature that is indicative of a clinically relevant phenomenon, have proven difficult to uncover, due to small populations of patients, a limited number of drugs, and intra/inter-tumoral genetic complexity. Adding to the dilemma are the antiquated methods used for biomarker testing, which can easily lead to misinterpretation, Lara explained.
“IHC and light microscopy date back to the 16 and 17 hundreds. We need to move beyond just IHC and light microscopy,” he said. “There are several ways we can order next-gen sequencing or molecular phenotyping tests. There are 7 commercial labs that do this.”
Once next-generation sequencing and other molecular profiling is conducted, clinical trials that randomize patients based on their genetics become more plausible. To this end, in the past 5 years, basket and umbrella protocols have been formed and the NCI has enhanced the study of exceptional responders. These steps are essential to accelerating biomarker development, Lara added.In an umbrella trial, upfront next-generation sequencing is conducted and patients are randomized based on the results to smaller individualized arms or substudies. Each smaller group explores a targeted therapy matched to the patient’s biology. This type of trial generally explores various markers within a single histology and can grow to be very large, with 20 or more arms.
“There are difficulties and costs associated with opening this type of trial, and unfortunately a bunch of those arms will not succeed,” Lara pointed out. The umbrella trial strategy is a hallmark of the Lung-MAP trial, which is a multi-drug, multi-arm, biomarker- driven clinical trial for patients with advanced squamous cell NSCLC. This trial uses genomic profiling to match patients to one of several therapies targeted against PIK3CA, CCND1/2/3, CDK4, FGFR, and other markers. Immunotherapy will be administered for those without an apparent molecular driver and chemotherapy will be used as a comparator (NCT02154490).In a basket trial, patients with various types of cancer are enrolled based on the presence of a biomarker, such as a genetic alteration or a distinct molecular signature. This style of trial takes the emphasis off histology and focuses it on the underlying genetics of each tumor. Typically, these protocols explore a single therapy across various types of cancer.
“The disadvantage here is that biology is not always histology agnostic. An example would be BRAF in colorectal cancer, which didn’t quite pan out,” Lara noted.
In the recently launched NCI-MATCH basket trial, 1000 adult patients with progressive advanced solid tumors or lymphomas will be assigned to 1 of 10 groups based on their molecular profiles and regardless of histology. While 6 of the agents being explored in this trial have gained approvals for other indications, the remaining 4 are still investigational. The NCI anticipates the need to screen 3000 patients in order to enroll the required number of participants.
“NCI-MATCH is a classic basket study. I encourage everyone to open this trial at your center,” Lara noted. “The NCI tells us that it will take 10 to 14 days for the genetic report, which they have ensured us is feasible even in the context of an NCI trial.”A number of basket and umbrella trials are open to enrollment, outside of the NCI-MATCH and Lung-MAP investigations. While these study designs are not new, they have begun to gain greater attention in recent years. However, in past trials that used these designs, the results were not always positive and all of the intended arms frequently had trouble enrolling, Lara cautioned. “Not all basket and umbrella strategies work out, sometimes it is a lot of resources and it doesn’t pan out,” he said. “This is not new, this is an accepted drug development pathway.”
In 2014 the NIH/NCI launched a study focused on exceptional responders to cancer therapy. Under this paradigm, the genetics of those who demonstrate an unusually robust response will be examined using deep sequencing to uncover potential molecular drivers. The findings from these investigations could be used to help tailor therapy for other individuals who also display these characteristics.
“This takes advantage of a real world experience with that agent in that patient population. It also takes advantage of luck,” Lara noted.
Training and stakeholder agreement from all those involved is required in order to put molecular testing into practice at a community center, especially regarding reimbursement, Lara noted. Additionally, he encouraged the establishment of a set decision-making strategy, once results were returned. Once molecular screen is in place, referrals could easily be made to the appropriate trials, Lara advised.Despite the ever growing need for new targets, a number of highly effective next-generation therapies are currently in development for patients with known oncogene-driven NSCLC, specifically those with alterations in EGFR, ALK, ROS1, NTRK, or a variety of other genes, Tony Mok, MD, explained during a keynote address at ILCC.2
“At this moment, we have a good number of driver mutations that we already know, and there are a lot of exciting developments,” explained Mok, a professor in the Department of Clinical Oncology at the Chinese University of Hong Kong. “Even with all of these, there are still a lot of questions that will drive our next studies, such as sequences.” The NCI’s ALK Master Protocol is currently exploring various ALK inhibitors in the first- and second-line setting. In the study, patients will be randomized to receive one of four ALK inhibitors in the first-line setting followed by a second agent upon disease progression. This study could provide hints at an optimal sequence for these therapies.
In addition to the more established ALK and EGFR-driven subtypes, there are a number of therapies currently in development for patients with rare alterations, including ROS1 or NTRK (Table 1).
Lovly, C, Horn L, Pao W. Molecular Profiling of Lung Cancer. My Cancer Genome http://www.mycancergenome.org/content/disease/lung-cancer/ (Updated June 17, 2015).
In for those with ROS1 alterations, crizotinib has received a breakthrough therapy designation from the FDA for patients. Additionally, for those with NTRK alterations, the Trk family TKI entrectinib has demonstrated promise in early phase clinical trials.
In February 2015, entrectinib received an orphan drug designation from the FDA as a treatment for patients with TrkA, TrkB, TrkC, ROS1, and ALK-positive NSCLC. In joint findings from the phase I STARTRK-1 and ALKA-372-001 studies,3,4 which were presented at the 2015 ASCO Annual Meeting, treatment with entrectinib demonstrated promising activity in this same molecularly defined subset of patients.
In 11 patients across both phase I studies who had not received a prior ALK and/or ROS1 inhibitor and who harbored an NTRK1/2/3, ROS1, or ALK alteration, treatment with entrectinib at 400 mg/m2 demonstrated an objective response rate of 91%. This dose and patient population has been selected for future study in a phase II trial. In the 11 patients who met these criteria, all those with ROS1 alterations had NSCLC (n = 6), and those with ALK alterations had NSCLC (n = 1) and another solid tumor (n = 1). Patients with NTRK1/2/3 fusions had NSCLC (n = 1), colorectal cancer (n = 1; CRC), and acinic cell cancer (n = 1). No drug-related serious adverse events were seen with entrectinib in the study.
The phase II STARTRK-2 trial, which is a basket study, is enrolling patients with NSCLC, CRC, and other solid tumors based on molecular profiles (alterations in ALK, ROS1, and NTRK1/2/3). Tumors will be analyzed using a potential companion diagnostics that has been developed by the drug’s developer, Ignyta (NCT02097810).
Tony Mok, MD
“This is actually a registrational study. This is the first time the FDA will allow a basket study as a registration study,” explained Mok. “This is very exciting, in a way we’re transitioning from the tissue-based to the molecular- based for treatment selection.”For immunotherapy, the recent approval of pembrolizumab (Keytruda) with a companion diagnostic for PDL1 has signaled a turning point in the biomarker discussion, since many researchers have questioned the validity of PD-L1 as a true biomarker for these treatments. In a talk at ILCC,5 Roy Herbst, MD, PhD, said that true precision medicine with these agents could only come from biomarker-driven frontline regimens delivered to the right patients. Herbst, chief of Medical Oncology, Yale Cancer Center and Smilow Cancer Hospital at Yale-New Haven, ranked PD-L1 as biomarker as “okay, at best.”
“As someone who treats lung cancer and speaks with patients, we have therapies that are working, but it’s going to be up to us to find and use biomarkers to find that group that benefits most so that they can get these therapies early on, and identify the group that isn’t benefiting so that they can get combination therapies and other approaches,” he said at the meeting.
PD-L1 expression is heterogeneous and varies, based on the specific antibody used for testing. Other factors that play a role with using PD-L1 as a biomarker include the interval between biopsy and treatment, use in the primary versus metastatic disease setting, staining conditions, and the definition of PD-L1—positivity.
Factors involved in the varying definitions of PD-L1— positivity, according to Herbst, include the cell type expressing PD-L1 (immune cell vs tumor vs both), the location of expression (cell surface vs intracellular vs stromal), intensity (percent of cells “positive”), and distribution (patchy vs diffuse, intratumoral vs peripheral).
“We’ve learned that the biomarker doesn’t actually have to be on the tumor,” said Herbst. “If the T cells aren’t there, it doesn’t matter how much PD-L1 you have on the tumor.”
According to Herbst, it is only through refining use of PD-L1, discovering other biomarkers, and identifying patient characteristics linked to improved response that true precision medicine with immunotherapy in NSCLC can be reached. These mechanisms will lead to “better responses and better activity early on by selecting the right patients or combining the right drugs and thereby increasing the number of patients who make it to longterm survival,” Herbst said.
Roy Herbst, MD, PhD
In many ways, the recent approval of nivolumab (Opdivo) for patients with nonsquamous NSCLC better characterizes an optimal scenario for PD-L1. In this situation, the test was approved as a “complementary” diagnostic but not a “companion.” This indicates the test could be used to add important information but that it was not required to prescribe the drug, as is the case with pembrolizumab.
As it stands now, multiple pharmaceutical companies are developing anti—PD-1/PD-L1 therapies each using different assays to measure expression levels. This disjoined development process could lead to future confusion, when selecting treatments, experts believe.
These fears have led to many of the leading companies collaborating to analyze the differing assays. The drugmakers working on the project are Bristol-Myers Squibb, Merck, Genentech, and AstraZeneca, which are developing nivolumab, pembrolizumab, atezolizumab (MPDL3280A), and durvalumab (MEDI4736), respectively (Table 2). Two diagnostic companies that have helped develop the assays used in clinical trials are conducting the analyses: Dako, which is owned by Agilent Technologies, and Ventana Medical Systems, which is a member of the Roche Group along with Genentech.
mAB indicates monoclonal antibody; BMS, Bristol-Myers Squibb; ORR, objective response rate; PD-L1, programmed death ligand-1. Complexities in personalized medicine: harmonizing companion diagnostics across a class of targeted therapies. Presented at: FDA-AACR-ASCO Public Workshop; Washington, DC; March 24, 2015.
The project has been moving forward after the companies jointly announced a blueprint for the assay analysis during a workshop that the FDA, along with leaders from ASCO and AACR, hosted in March. This initial talk was followed by a larger workshop on the complexities of personalized medicine.6
“Different pharmaceutical companies and diagnostic companies don’t often get together premarket and discuss their issues,” Elizabeth Mansfield, PhD, an FDA official who co-chaired the workshop told Precision Medicine in Oncology. “The proposal is intended to establish how the tests compare to each other given the same sample. What will come out of this will be a set of data that shows you that, if you use the same specimen from one patient and test it with all the different tests, the way [results] would read out with all the different tests.”
Mansfield anticipates that the industry group will report on its findings next year. However, she said it is not yet clear whether the results of this assay comparison would apply to cancer types besides NSCLC. “We hope that it will at least partially inform and will at least set up a framework if we need to do this again in other tissue types,” she said.