The war against cancer is far from over, but modern medicine has made significant progress in winning some critical battles. Thanks to the rise of new modalities, cancer diagnosis and treatment have seen practice-changing advances. Liquid biopsy, the latest entrant to precision oncology, enables the screening of cancer prevalence and detects multiple cancers with a single blood test. However, there is merit in choosing to tackle one cancer at a time using a multi-pronged approach, says Dr. Jimmy Lin, Chief Scientific Officer of Freenome.
Dr. Jimmy Lin is a seasoned researcher in the field of cancer genomics. He helped perform the first genome-wide sequencing of human cancers and was part of one of the first academic clinical next-generation sequencing efforts. At Freenome, a San Francisco-based startup, his team strives to build a blood-based test for the early detection of colorectal cancer. Before that, he was a Chief Scientific Officer at Natera, a genetic testing company. We interviewed him to learn about the latest progress and challenges in oncology diagnostics.
Dr. Lin has experienced the perils of cancer firsthand. Losing his family members to the deadly disease relatively early in their lives had driven him towards studying it. Additionally, his introduction to diagnostic and therapeutic innovations during his formative years was crucial in shaping his ambitions. His MD-PhD at the Johns Hopkins University School of Medicine allowed him to work under Bert Vogelstein, an eminent researcher renowned for pioneering work in cancer genetics.
“At Johns Hopkins, we performed the whole genome sequencing of breast and colorectal cancer, followed by pancreatic cancer, glioblastoma, and medulloblastoma. It was an amazing time working with some of the world’s leaders in cancer research. That set the stage for me to gain a genomic perspective of cancer and witness the promise of therapeutics and early diagnostics.”
Early Detection of Cancers
When asked how early detection has transpired into clinical practice, Dr. Lin mentioned four cancers currently screened for early diagnosis.
For colorectal cancer screening, multiple modalities like colonoscopy and stool DNA tests are currently available. For prostate cancers, the detection of prostate-specific antigen (PSA) has moderate clinical adoption. Mammography, without a doubt, has helped many women detect breast cancers early. Similarly, low-dose computed tomography is now a recommended screening test for high-risk populations with lung cancer.
“There are several existing modalities that still need improving. Some tests struggle with adherence, such as stool DNA tests and colonoscopy; some of these biomarkers require higher specificity (such as PSA); some test have difficulty in specific populations such as mammography for dense breasts. New and promising technologies have now emerged with increased sensitivity and specificity, higher adherence, and applicability in larger populations” he noted.
Dr. Lin also drew attention to many unscreened cancers like ovarian or pancreatic cancers, which have a very low prognosis when detected late.
“For the unscreened cancers, there need to be tests and performance characteristics that one can aim for and achieve. At present, health economic outcomes studies associated with these currently unscreened cancers have yet to be performed,” he observed.
Single Cancer and Multiomics Approach
Many emerging companies in the liquid biopsy space strive to detect multiple cancers from a single blood sample, but in contrast, Freenome currently focuses solely on colorectal cancer. When asked about this, Dr. Lin explained their distinct approach to cancer diagnosis.
“Among the three exciting things at Freenome, first is the singular focus on one problem. The second is the multi-omics approach that enables problem-solving, and the third is the use of machine learning as an integral part of our endeavors,” he declared.
Singular Focus on Colorectal Cancer
Not all cancers are the same. Therefore, we do not treat every cancer alike. There is a different requirement for each type of cancer in terms of early detection, Dr. Lin emphasized.
“Only very few companies approach the early detection problem in a very focused manner. Our approach is contrary to the one size fits all approach of a pan-cancer test. Instead, we tackle one cancer at a time and we think it’s the right scientific approach. Recently, we published a commentary in JCO Precision Oncology highlighting this very topic,” he said.
Colorectal cancer is well understood in terms of clinical utility. But existing modalities like colonoscopy and stool DNA testing are either invasive or have a low patient adherence. To overcome these barriers, a blood test with much higher adherence is key.
“We selected colorectal cancer because there is already extensive literature and a tremendous body of work showing the health economic impact of colorectal screening,” he argued.
As opposed to perfecting a single technology to decode many questions, Freenome prefers to take one problem and tries to solve it using its multi-omics approach.
Dr. Lin observes that no consolidated set of genes can define even one cancer type. He further explained that there are variations both in terms of genes as well as alteration types. Some individuals may have an early alteration in methylation or an early mutation that leads to protein differences. However, the signal of such early-stage biomarkers is very sparse for many cancer types.
“Since it is only a tiny tumor shedding into the bloodstream, sampling just a small part of it may not yield adequate biomarkers required for assessment. So just looking for mutations in the circulating tumor DNA may be insufficient,” he remarked.
“Even in early-stage cancers or precancerous lesions, the signals look very different. Because of cancer heterogeneity, there is also heterogeneity among different signals such as DNA, RNA, or protein. Therefore it becomes essential to read multiple biological signals,” he explained.
He justifies that for early cancer detection, it is essential to use all the different modalities to magnify the available signal. A multi-omics platform can accomplish that with maximum sensitivity and specificity.
Onus on Machine Learning
Thirdly, he mentioned that the big data generated had accentuated the importance of artificial intelligence (AI). Although AI is now an industry buzzword, he said their team of experts build their own AI machine learning platforms for implementation. In addition, they regularly publish and report new data at relevant conferences.
Cancer detection is just one part of an enormous challenge. Just because cancer is detected early, it doesn’t mean there are available treatments. Similarly, early treatment does not always ensure patient response or improved survival.
“There is a whole field of literature suggesting that early detection of some cancers only results in long treatment cycles, but does not significantly improve overall patient survival. So, we need to demonstrate that early detection can lead to better patient outcomes,” he said.
According to him, understanding the science behind developing these technologies is the main hurdle. So, finding the right biomarker signals is necessary to achieve high sensitivity and specificity, he reiterated. Besides, identifying the right performance characteristics and positive and negative predictive values are critical too.
“In the case of colorectal cancer, even though we have obtained early screening results and better outcomes, our screening rates are still very low. Despite the ACS shooting for a goal of 80% screening, we have only achieved 67%. So, there are multiple stakeholders in play for us to achieve that. But we are excited about what that will mean for patients in terms of early detection, and better survival” he added.
He further stated that health economics is another hurdle. “If we develop a promising technology, we also need to justify its price,” he said. “We need to figure out cost structures of technologies prescribed for a wide population.”
When asked if newer technologies like liquid biopsy can overhaul traditional screening methods, Dr. Lin says it is the eventual goal. The ultimate vision for them is to perform an annual screen not just to detect cancers early but also to identify before they originate. However, that’s a long-term vision that requires surpassing several challenges, including regulatory hurdles.
“To get approval, it is necessary to figure out reimbursement, patient outcomes, and clinical utility,” he stated.
Regulatory Acceptance and Future Prospects
The FDA approval of Guardant Health’s Guardant360 CDx and Foundation Medicine’s FoundationOneLiquid CDx in 2020 marked the acceptance of liquid biopsy tests in the clinic. Dr. Lin said that he welcomes the approval of companion diagnostics that aids therapy selections, but early detection is a different game.
“Regulatory approval always hinges on safety and efficacy, and at least in the US, there is a relationship between approval and reimbursement. We are running preliminary trials to gain FDA approval in the early setting, and it is important in the US context,” he said.
Finally, we asked how he envisioned the evolution of cancer diagnostics in the future. He declared that there are numerous technologies, focusing on different cellular compartments and different bioanalytes. However, the logical extension of that innovation is to find biological relevance.
Many technologies are in the first phase when they are shiny, and everybody is thrilled. The next phase is if the technology can inform about a cancer-specific signal. Only in the last phase, the promise for translation makes a difference in terms of early detection. Then, it will be added to our existing platform.
Interviewer & Editor: Rajaneesh K. Gopinath, Ph.D.
©www.geneonline.news. All rights reserved. Contact: firstname.lastname@example.org