ATLANTA (GenomeWeb) – The American Association for Cancer Research's annual meeting kicked off in Atlanta on Sunday with a call for more personalized care for patients through a more efficient use of technology and a greater degree of diversity in research cohorts. Chairman John Carpten of the University of Southern California Keck School of Medicine said this year's conference will feature the most diverse and youngest group of researchers yet presenting their work to the tens of thousands of assembled attendees.
Building on the theme of the 2018 AACR opening plenary of interdisciplinary science, this year's meeting commenced with presentations that showed how a convergence of technologies could help advance the personalization of cancer care.
In a joint presentation, for example, USC physicist Peter Kuhn and USC Norris Comprehensive Cancer Center researcher Jorge Nieva described collaborative efforts between their groups to study cancer in the context of the complexity of human life. They built models to quantify the pathways of disease in multiple dimensions and reflect the complexity of cancer spread.
This digitization of patients and their cancers can reduce biases in the delivery of care, Nieva said. But it could also help increase participation in clinical trials, which would benefit both patients and the wider oncology community. As of right now, there are more than 1,100 anti-cancer compounds in clinical development, but only 3 percent of patients participate in clinical trials.
"We've done a good job of getting to more precise definitions of cancer," Nieva said. "We now have a wealth of new treatments that have been developed in the past 20 years, and each have their own complexities. Picking the right therapies is incredibly important for patients, and understanding how cancer evolves from one line to the next is important to picking the right treatment."
Part of the process of getting more people into cancer trials involves digitizing the cancer and the patient simultaneously in order to track the cancer's evolution over time and in the space of the body, and understanding the relationship of the genome to the proteome, Kuhn said.
For example, he presented the case of a breast cancer patient whose disease his group had tracked from 2012 to 2017. She'd had 11 treatments, three biopsies, and 17 liquid biopsies. In looking at individual circulating tumor cells' copy number profiles and CTC proteomic profiles, the researchers found that her cancer had three clones from the outset that stayed the same throughout progression of the disease, though in different ratios. But in 2017, a fourth clone emerged, at which point the team was able to pinpoint the exact mutation attached to that clone. In all, they ended up with 700 single-cell whole genomes for this patient.
But while such data may be useful in the lab, Nieva said, in the clinic a doctor needs a binary decision: therapy A or therapy B? Nieva also noted that cancer care is about unique people, who can each have characteristics, such as frailty, that can affect treatment decisions. This is particularly important in a disease like lung cancer. One third of patients don't get treatment at all in lung cancer, Nieva said. While some of this is due to bias, much of it is due to the frailty of the patient. Further, patients can sometimes disagree with their doctors on their own performance, leading to suboptimal treatment decisions. So, there's a need to properly quantify the performance of patients, and make sure that patients and doctors agree about performance and health.
Many clinical trials, about 38 percent, close for poor accrual because inclusion criteria are very stringent. But if we could stratify patients and include people with all levels of frailty, we could broaden inclusion and hopefully develop a greater number of targeted therapies, he added.
In this way, digitizing patients and their health can deconvolute cancer health disparities, Kuhn noted.
Memorial Sloan Kettering Cancer Center's David Solit also spoke about the importance of including more patients in clinical trials, and showed examples of how a convergence of technologies is helping to advance researchers' understanding of cancer.
Tumor genomic profiling is part of the standard management of a variety of solid tumors, but most cancer patients are still treated based on the site of the tumor rather than receiving personalized treatment according to the molecular alterations present in their cancers, he said. However, tumor profiling and the measurement of cell-free DNA can be used to guide routine clinical care and enrollment into clinical trials.
He provided insights from MSK's institution-wide tumor profiling initiative, which aims to define the genetic driver of every cancer for every patient that comes to the center. The institute primarily uses the MSK-IMPACT targeted next-generation sequencing test, but is also introducing complementary assays such as IMPACT Heme for blood cancers, and other tests for cell-free DNA and RNA sequencing.
All the data from these assays gets uploaded into a portal that's open to the research community, Solit said. So far, MSK teams have analyzed more than 38,000 tumors from 33,000 patients. The hope is that this data will yield new insights, and so far, it has. For example, MSK researchers looking to identify somatic mutations that confer endocrine therapy resistance in breast cancer found that intrinsic mutations cause intrinsic resistance whereas acquired mutations cause acquired resistance, he noted.
Solit also pointed to research in which investigators looked at 17,000 patients sequenced with the MSK-IMPACT assay who had loss of heterozygosity in both their germline and somatic BRCA1 and BRCA2 genes. They found that that LOH is common in BRCA-associated cancer types, and PARP inhibitors were not effective as treatments in non-BRCA-associated cancers. Solit noted that this could mean that BRCA could be a biomarker like BRAF, where lineage really matters. The MSK team is now doing several retrospective studies with this dataset, but the researchers are also using the insights to help patients.
Solit did emphasize that next-generation sequencing is not a therapeutic modality, but a diagnostic tool, and that we need more drugs in order to help patients in a more personalized manner.
In the following talk, Eli Lilly Senior Vice President Levi Garraway said new classes of personalized drugs are indeed being developed, thanks to more precise clinical trials and molecular profiling of tumors.
Version 1.0 of precision oncology, as Garraway termed it, has focused on actionable mutations — one drug for one mutated gene. But some cancers can have a so-called "long tail" of actionable gene mutations, he added. They're sometimes hard to find, but some are useful in terms of treating patients.
For example, Garraway said, when it comes to kinase mutations, long-tail mutations are rare, but some are recurrent. If we are to go beyond Precision Oncology 1.0, we must realize that the vast majority of cancers don't have actionable mutations, he added. And that's where treatment modalities like immuno-oncology or non-oncogene dependence — where key cancer genes that are not mutated are targeted — come in. In this latter vein, Lilly is developing an Aurora A kinase inhibitor for RB-null or MYC-driven cancers, even though the AURKA gene which encodes this kinase isn't mutated in these cancers. Garraway also noted that there are at least 1,800 clinical trials currently ongoing for immuno-oncology agents.
Overall, he said, although there are challenges, precision oncology is moving forward and these approaches offer the best potential for treating patients. Figuring out these patients' genetic or molecular determinants of response will be the next challenge, and will need to extend beyond their genes, and into their macrophages, ligands, other molecular determinants in the tumor.