Cedars-Sinai Cancer effectively transitions basic science findings into meaningful projects that directly impact clinical care and outcomes.
Converting data and science into medical progress requires a strategic emphasis on work that matters to patients, either through imminent application to clinical practice or potentially transformative basic discoveries.
The rapid translation of findings into fruitful changes in medical practice is not accidental. It stems from convergent science at its best—what Cedars-Sinai Cancer and its biomedical science, medical technology and host of affiliated research programs pursue on a daily basis. From increasing the accuracy and accessibility of breast cancer prognostics to identifying biomarkers that inform new liver cancer treatments, our clinicians and researchers excel at innovating at the intersection of medical science and medical practice.
Turning Biomarker-Based AI Into Prediction Tools
In an effort to identify better prognostic and predictive biomarkers for pancreatic ductal adenocarcinoma (PDAC), researchers at Cedars-Sinai Cancer applied AI technologies to a dataset generated by samples from 74 patients—a small subset of the contributions to a much larger endeavor called the Molecular Twin Precision Oncology Platform (MT-POP).
The computer-generated model for survival prediction used more than 6,300 data points from the molecular and clinical analytes found in patient and tumor DNA, RNA, lipids and proteins. It outperformed the current clinical survival-prediction tool, CA 19-9, and was further simplified (to fewer than 600 analytes) to allow for broader use with fewer samples and resources.
The smaller model still exceeded current clinical standard predictions and was validated using four independent PDAC patient datasets.1
In a separate effort, Cedars-Sinai Cancer and University of Cambridge researchers collaborated to develop PREDICT Breast (v3.0 set to deploy in the near future), an algorithm-based online prognostic model used by clinicians around the world to predict the effectiveness of post-surgical treatments. It shows how various breast cancer therapies may affect survival after surgery for early-onset invasive breast cancer.
The only prognostic tool endorsed by the American Joint Committee on Cancer, PREDICT uses patient-specific data (age, mutations, tumor size, grade and other factors) to generate the anticipated survival benefit of added treatments (including hormone therapy, chemotherapy, bisphosphonates and trastuzumab) alone or in combination. The teams are now collaborating to improve data for better personalized predictions within racial and ethnic subpopulations.
Translating Disparities Into Direction
California’s lung cancer screening rates are the lowest in the nation, at just 1%. Our Community Outreach and Engagement team discovered that Latino, Black and Asian populations in L.A. County—the primary catchment area for Cedars-Sinai—are more likely to be diagnosed with late-stage lung cancer, partly due to low screening rates and barriers to care. The area’s LGBTQIA+ community also has low screening rates, a problem compounded by higher rates of smoking in this community than in other populations. This has prompted strong partnerships with federally qualified health centers in the most-affected areas, community outreach interventions to improve awareness by training community health workers and navigators, as well and providing smoking-cessation support.
Cedars-Sinai Cancer also has secured an institutional commitment for the first mobile low-dose computed tomography (LDCT) scanner on the West Coast to help address these disparities. Scheduled to launch in 2025 and building upon existing partnerships established by the Community Outreach and Engagement team, the LDCT program is one of several direct responses to our community-directed research findings.
A related new effort, the Lung Cancer Screening Program conducts community outreach and follow-up with at-risk patients. In just one year of activity, it increased annual screening rates more than sixfold.
Similarly, Cedars-Sinai has increased screening efforts and research on therapeutic targets for nonalcoholic fatty liver disease and liver cancer patients—another condition disproportionately prevalent in L.A. County’s Latino residents.
In basic and preclinical studies, two mouse models of hepatocellular carcinoma suggested the IL27 cytokine receptor (IL27R) contributes to cancer growth by using signal-induced suppression of tumor-associated cytotoxic natural killer cells and type 1 innate lymphoid cells.2 The team has already launched a clinical trial to evaluate IL27R inhibition as a therapeutic approach.
In addition, Cedars-Sinai Cancer specialists have spent years identifying signals of early cancer that are detectable in the blood rather than via invasive biopsies or specialized ultrasounds. After prior work revealed a higher prevalence of four surface protein markers circulating in the blood of liver cancer patients than in healthy individuals, the interdisciplinary, multi-institutional research team rapidly translated these findings into a biomarker study.
The Phase II trial has already confirmed the blood test has a 90% accuracy in identifying hepatocellular carcinoma.3
These efforts also have led to the establishment and growth of our Liver Cancer Program, which aims to improve diagnosis and treatment options for the condition while reducing disparities in screening and therapy outcomes.
Our clinicians and scientists collaborate closely with a clear mission and continuous focus on improving care for patients in our catchment area and beyond. The size of our program, our access to federal funding and the cohesive network available to our researchers allow for rapid progress toward meaningful change for patients everywhere.
“Profiling of pancreatic adenocarcinoma using artificial intelligence-based integration of multiomic and computational pathology features.” Nat Cancer. In press.
“IL27 signaling serves as an immunologic checkpoint for innate cytotoxic cells to promote hepatocellular carcinoma.” Cancer Discov. PMID: 35723626.
“HCC EV ECG score: An extracellular vesicle-based protein assay for detection of early-stage hepatocellular carcinoma.” Hepatology. PMID: 35908246.