The 723,000 people who succumbed to gastric cancer in 2012 make it clear that the global burden of this disease continues to weigh heavily, despite valiant clinical efforts otherwise. Confronted by multiple failed trials of targeted agents, the gastric cancer field is finally embracing the realization that the diverse collection of tumors made up of myriad genetic and molecular subtypes deserves a much more nuanced approach to treatment than previous endeavors.
As attendees learned at the June 3 Education Session “HER2, VEGFR, and Beyond: Genomic Profiling of Upper Gastrointestinal Tract Cancers and the Future of Personalized Treatment,” better insight into past failures in gastric cancer, along with emerging genetic and molecular information, hold good solutions to solving the current treatment conundrum for this disease.
Moving the Targeted Therapy Needle in Gastric Cancer
Patients with metastatic gastric cancer face a very grim outlook of less than 1 year to live with currently available therapies—a bar that has moved little over the past decade. As Jeeyun Lee, MD, PhD, of the Samsung Medical Center, South Korea, explained, gastric cancer has not benefited from the promise of targeted therapy realized in many other malignancies. Of 17 major phase II and phase III trials of targeted therapies that have been conducted in gastric cancer since 2010, only three trials succeeded, giving the field trastuzumab in the first-line setting and ramucirumab in the second- and third-line settings.
Although trials such as TYTAN, SHINE, and MEGA may have several reasons for not living up to their grand names, Dr. Lee remarked that the most plausible explanation is that “gastric cancer is not one disease.”
Dr. Jeeyun Lee
Emerging research underscores the marked molecular heterogeneity between tumors from different patients—even when they share a common targetable biomarker. Moreover, the molecular drivers change over time in an individual patient as disease metastasizes from the primary lesion. As a result, hitting just one driver mutation in the setting of numerous other dynamic concomitant aberrations is likely to exert little damage to the tumor.
To most effectively account for all the interpatient and intrapatient heterogeneity in gastric cancers in future studies, Dr. Lee believes it is crucial to learn as much as possible about targeted therapies in early-phase trials by designing both subtype- and target-based studies.
“Try to learn everything you need to know for specific targeted agents early on, and do not wait until a phase III trial,” she emphasized.
Dr. Lee noted that an umbrella trial offers good prospects for moving the field forward and described the VIKTORY (Targeted Agent Evaluation in Gastric Cancer Basket Korea) trial as a prime example. Within this study, patients with metastatic gastric cancer in whom first-line cytotoxic therapy failed are assigned to different treatment options within the confines of phase II studies based on the molecular profile of their disease. For example, patients with RAS mutations or amplification and either low MEK or high MEK expression are being assigned to treatment with selumetinib plus docetaxel to determine which subsets demonstrate marked selumetinib sensitivity. Thereafter, larger confirmatory trials can be performed in the appropriate patient subsets to validate the findings.
Given limitations in the availability of tumor samples for genetic and molecular analysis, Dr. Lee also encourages researchers to look to other technologies to evaluate intrapatient tumor heterogeneity, such as analysis of cell-free DNA, and more feasible methods for tumor subtyping, such as finding ways to subject small samples of formalin-fixed, paraffin-embedded archival tissue to gene expression profiling.
Practical Issues Concerning HER2
Amid the numerous trials of targeted therapy in advanced gastric cancer that have faltered, a few rare gems stand out. Jaffer A. Ajani, MD, of the University of Texas MD Anderson Cancer Center, focused on one of these trials—ToGA (Trastuzumab for Gastric Cancer)—and the current approach to HER2-directed therapy that has been adopted in the trial’s wake.
Dr. Jaffer A. Ajani
A small proportion of gastric adenocarcinomas generate an overabundance of HER2—the familiar EGFR family member known to fuel tumor growth and survival in several cancers, making it an alluring therapeutic target. However, among the many logical contenders that could suppress HER2 activity (e.g., lapatinib, ado-trastuzumab emtansine), only trastuzumab emerged as a viable therapeutic option based on ToGA, which ushered in the use of trastuzumab in combination with chemotherapy as a standard of care for patients with HER2-positive metastatic gastric adenocarcinoma.
Although only a subset of individuals will be eligible for trastuzumab, Dr. Ajani emphasized that all patients with unresectable, advanced gastric cancer suitable for chemotherapy should undergo HER2 testing. Current guidelines from the National Comprehensive Cancer Network advocate HER2 screening by immunohistochemistry. Results of 0 or 1+ clearly denote negative HER2 expression, whereas a result of 3+ signals clear overexpression, with no further testing required in either scenario. A 2+ result is indeterminate and requires further work-up using fluorescent in situ hybridization for analysis of ERBB2 amplification (the gene encoding HER2).
These recommendations may well change in the near future. ASCO, the College of American Pathologists, and American Society for Clinical Pathology have been hard at work developing formal HER2 testing guidelines that are slated for publication later in 2016.
Pending the publication of these guidelines, Dr. Ajani provided much-needed answers to three key clinical questions concerning HER2. First, he indicated that there is no evidence to support repeat HER2 testing in gastric adenocarcinoma. Second, no data indicate that trastuzumab should be continued following disease progression. Third, chemotherapy should not be held while awaiting the results of HER2 testing, which can take up to 2 weeks to receive. Most patients are symptomatic and require immediate therapy, and trastuzumab can be added later if HER2 is positive, Dr. Ajani explained.
Other questions still remain unanswered, such as the number of samples that should be studied during HER2 testing and whether HER2-positive status provides prognostic information. Data informing both of these issues remain mixed.
Despite the potential utility of trastuzumab use among appropriate patients, the actual benefit of HER2 inhibition remains a moving target. The survival gain observed in ToGA with the addition of trastuzumab to chemotherapy has attenuated over time, shrinking from 2.5 months at the second interim analysis that led to regulatory approval (13.5 vs. 11.0 months; HR 0.73, 95% CI [0.60, 0.91]) to a margin of only 1.4 months with longer follow-up (13.1 vs. 11.7 months; HR 0.80, 95% CI [0.67, 0.97]). In light of this decrease, Dr. Ajani notes that “we need to describe novel strategies” to overcome such setbacks.
Novel Targets Key to Future Research
The current paradigm in molecular medicine of understanding what drives tumor growth and then attacking the driver keeps coming up short in gastric cancer, according to Adam Joel Bass, MD, of the Dana-Farber Cancer Institute, largely because there may be assorted drivers behind the wheel—and some at the same time.
Dr. Adam Joel Bass
“We really need to move beyond this idea of ‘the right drug for the right patient’ that came from studies in [chronic myeloid leukemia] and lung cancer,” he said.
Dr. Bass explained that a critical tenet moving forward will be to leverage genomic data to allow researchers to better categorize gastric cancers, uncover candidate therapeutic targets, and identify biomarkers to guide how best to apply current therapies.
Enter The Cancer Genome Atlas (TCGA), a consortium under the auspices of the National Institutes of Health that has collected a bevy of untreated, high-quality, primary gastric tumor samples from institutions around the world. By subjecting these samples to a gamut of genomic and molecular analyses, the TCGA has succeeded in developing four molecular classes of gastric cancers with distinct features and prognoses: (1) tumors positive for Epstein-Barr virus, (2) tumors with microsatellite instability, (3) genomically stable tumors, and (4) tumors with chromosome instability.
A rich spectrum of candidate targets has emerged across the four classification groups. For example, the PI3K and PD-1 pathways appear to be attractive targets among the Epstein-Barr virus group of gastric cancers, whereas ERBB2, ERBB3, and PIK3CA offer targetable options in microsatellite instability cancers.
Despite the information gleaned about gastric cancer from the TCGA, Dr. Bass warned that genome-guided therapy is not a panacea: “Cancers are smart; they’re a tough enemy.” He explained that instead of focusing on acquired resistance in gastric cancer, a more pressing problem is de novo resistance that prevents a tumor from responding to a targeted agent despite bearing the biomarker of interest.
“This is really critical moving forward. It’s not so much finding the targets; it’s about finding how to make the targets work,” Dr. Bass remarked. Toward this end, he thinks researchers will have to uncover both genetic and nongenetic factors that influence the response to therapy.
“As we move forward, we have great opportunities to leverage these data to develop new therapies, but we should not oversimplify the challenge at hand,” Dr. Bass said. “The key will be to combine our knowledge of the genome with tumor biology to learn how to ‘play chess’ with cancer—that is, to develop effective rationale combination strategies to target key cancer genes and pathways.”
– Kara Nyberg, PhD