Today’s Molecular Profiling Efforts Will Define the New Decade of GI Cancer Management

Over the past decade in gastrointestinal (GI) cancer treatment, the acknowledgement that cancer is heterogeneous and likely polyclonal has prompted a shift from gene testing for some patients, to many patients. These precision oncology advances have allowed therapeutic strides to be made in the management of GI cancers, and the “prospective incorporation of molecular profiling will continue to transform cancer care” in the years to come, presuming that biomarker discovery efforts continue, according to John L. Marshall, MD.¹

“We have to acknowledge—and I think we are—that cancers are heterogeneous; we can’t [use] a one-size-fits-all approach. Therefore, we need to increase our funding towards biomarker development. We can’t afford to do trial and error medicine anymore,” Marshall said during the 2nd Annual Precision Medicine Symposium, hosted by Physicians Education Resource®, LLC.

A crux of future, and perhaps even ongoing initiatives, to uncover and characterize biomarkers in the GI field is big data and artificial intelligence (AI). “As we look into the decade ahead, the results of these kinds of analyses can bring out new genomic signatures, which in fact can tell us that we should give one regimen versus another based on an individual’s pattern and profile, even with our most basic chemotherapy,” explained Marshall, chief of the Division of Hematology/Oncology at Medstar Georgetown University Hospital and director of the Otto J. Ruesch Center for the Cure of Gastrointestinal Cancer at the Georgetown Lombardi Comprehensive Cancer Center in Washington, DC.

Data-driven methods to share and evaluate genomic insights, such as AI, machine learning, and deep learning specifically, can be expected to define the precision medicine era in GI malignancies.¹ IndivuType, the “first true multi-omics cancer database in the world” developed to facilitate the identification and validation of new molecular drug targets and biomarkers,² exemplifies one existing data-oriented resource that will foreshadow more to come, Marshall contended.

Increasing preliminary patient screening for subsequent study participation will go hand-in-hand with evolving data-based endeavors. “This has to be done collaboratively. When we build networks, we build alliances to feed these new machines,” he said. The phase 2 COLOMATE umbrella study (NCT03765736), short for “Colorectal and Liquid biopsy Molecularly Assigned ThErapy,” is one such “alliance,” according to Marshall.

COLOMATE, which is currently recruiting with an aim to enroll up to 500 patients with metastatic or unresectable colon or rectal cancer, was designed to perform blood-based genomic profiling in patients with treatment-refractory disease. Specifically, investigators seek “to facilitate accrual to molecularly assigned therapies and obtain patient-matched tumor tissue and cell-free DNA from peripheral blood toclinically annotated genomic analyses.”³

To further the COLOMATE study, investigators have “internationally partnered with a group in Japan to drive these rare markers to these arms” in an effort to document therapeutic benefit in a “much more rapid fashion,” Marshall said. “This is how I want to be treated when it's my turn to deal with one of these cancers.”

Although overseas collaboration with other GI cancer experts will indubitably drive biomarker development in the near future, national cooperation in populating clinical trials is no less dispensable, given that as studies necessitating large patient populations are designed, eligible patient subgroups are getting smaller.¹ Every patient contemplating enrolling in a study “needs to be screened to find the few patients that are positive for [the specific] marker,” Marshall noted.

For example, to amass a patient population big enough to study the BRAF mutation—which presents in 10% of colon cancer cases—an enrolling investigator would need to screen 500 patients in order to accrue 50 patients for a trial evaluating this alteration in this disease setting. If the investigator hoped to complete the 50-patient enrollment in 1 year, 41 patients would need to be screened each month. The frequency of screening necessary to achieve this goal signals that clinicians must advocate for patient screening, because even if the patient’s genomic profile would not cause the patient to be a match for the trial at hand, the patient data would be added to a molecular database, allowing the patient to be “available for other trials and outcomes analyses,” Marshall explained.¹

Molecular Strides Set Precedent for Future Advances

In GI malignancies, precision oncology progress has not occurred evenly. Although the field has made “dramatic” strides in improving outcomes in hepatocellular carcinoma (HCC) in the past 5 years with the advent of targeted therapies and immune checkpoint inhibitors, it has “no proven standard markers,” according to Marshall. As such, clinicians do not even biopsy HCC tumors, electing to diagnose them instead via alpha-fetoprotein and imaging “consistent with the tumor.” Despite the demonstrated efficacy of available therapies for HCC, developing markers is a necessary step to amplify therapeutic activity. “My hope is that once we begin to incorporate more and more molecular markers, we will see even further increase in benefit,” Marshall said.¹

By contrast, bile duct malignancies represent a “biomarker rich cancer” for which molecular profiling is typically performed “to follow the leads of the targets that emerge in these diseases.” Pancreatic cancer differs further yet from HCC and even bile duct cancers. Actionable mutations have been identified in pancreatic cancer, with next-generation sequencing efforts consistently showing that 25% or more of pancreatic cancers have “potentially highly actionable molecular biomarkers.”

STK11 and ATM, which are estimated to affect 6% and 5% of pancreatic cancers, respectively, are under investigation in this disease setting, in addition to a host of other molecular markers. “We're learning what targets matter; which ones matter more, and which ones matter less,” Marshall observed.

Each GI malignancy, colon cancer and gastric cancer included, come with distinctive molecular complexities, and sequencing can help clinicians surmount some of these complicating factors by providing insight on a patient’s genomic profile. “Broad molecular profiling in my opinion is the current standard of care for patients with metastatic colorectal cancers [CRCs], and maybe all GI cancers,” Marshall said.

In CRC, all colon cancers are typically screened for mismatch repair and microsatellite instability (MSI). Targets of interest in metastatic CRC include RAS, BRAF, and HER2. At present, standard molecular testing methods in gastric cancer customarily probe for HER2, MSI, PD-L1, and NTRK. Germline analysis can also be performed in this space to determine whether CDH-1, FAP, Lynch syndrome, or other aberrations are present and promoting cancer development.

The targets of priority in these 2 disease types are a microcosm for the molecular diversity in GI malignancies and the various different markers of interest across them. Continuing ongoing biomarker development efforts and spearheading new ones is imperative to improving molecularly guided therapy and inching ever closer to the field’s “overall goal to cure cancer,” Marshall concluded.

References

1. Marshall JL. Gastrointestinal cancers: molecular prediction of response to therapy. Presented at: 2nd Annual Precision Medicine Symposium; September 25-26, 2020.
2. The global multi-omics cancer database. Indivumed. Accessed September 25, 2020. https://www.indivumed.com/services/indivutype​
3. Ciombor KK, Ou F-S, Dodge A, et al. COLOMATE: Colorectal cancer and liquid biopsy screening protocol for molecularly assigned therapy. Presented at: 2019 AACR Annual Meeting; March 29-April 3, 2019. Abstract LB-235/14. doi:10.1158/1538-7445.AM2019-LB-235