In his presentation on translational research in non–small cell lung cancer during the <em>20th Annual </em>International Lung Cancer Congress®, Scagliotti, a professor of oncology at the University of Torino in Italy, discussed the promising evolution of therapeutic options, pointing to steps being taken to create a larger precision medicine ecosystem.
Giorgio V. Scagliotti MD, PhD
Giorgio V. Scagliotti MD, PhD
The rapid introduction of new biomarkers, cancer detection strategies, immunotherapies, and targeted therapies brings a growing need to share information across research settings and the community to translate new findings into clinically relevant improvement in care.
“In the digital era, the 5P’s of healthcare will predominatepreventive,predictive,personalized,participatory, andpertinent,” said Giorgio V. Scagliotti MD, PhD, adding that a more personalized research strategy can lead to greater evolutions in research for biomedicine and prevention, as well as clinical medicine.1
In his presentation on translational research in nonsmall cell lung cancer (NSCLC) during the20th AnnualInternational Lung Cancer Congress®, Scagliotti, a professor of oncology at the University of Torino in Italy, discussed the promising evolution of therapeutic options, pointing to steps being taken to create a larger precision medicine ecosystem.
In this type of ecosystem, the shared goal of developing more effective therapies and mitigating challenges associated with trial-and-error or ineffective treatments represents a uniting thread.
“Translational research has led to a lot of changes in our clinical practice. [Advances in tumors with]EGFRmutations,ALKtranslocations,RETfusions, andROS1fusions are just a few examples of the clinical implications of research,” Scagliotti said in an interview withTargeted Therapies in Oncology. “There are different layers of complexity because the integration of [...] genomics, transcriptomics, and proteomics will generate knowledge that will escalate clinical implication of research to a higher level.”
A preview of the next steps in the clinical treatment paradigm may be gleaned by examining the latest strategies being tested and implemented in advanced research settings.
The use of liquid biopsies to detect potential biomarkers for therapy, such as cell-free DNA (cfDNA), circulating tumor cells (CTCs), and proteins and cytokines detected in plasma, are under development. However, the utility of these approaches has not been fully realized in daily practice due to limitation in accessibility, reproducibility, sensitivity, and cost. Rapid discoveries into these methods for use in NSCLC and other cancers will necessitate cooperation across specialties. (FIGURE).2
“There are benefits and limitations to all of these methods. You can get a lot [of information] from a liquid biopsy; we can look to genomic alterations or transcriptional changes in lung cancer and different solid tumors,” Scagliotti said.
Detection of certain mutations by cfDNA may be indicative of immunotherapy response in NSCLC. There is evidence thatSTK11mutations are associated with a limited response to immunotherapy. Additionally, tumors with high microsatellite instability could be detected by cfDNA. However, it will be important to filter out mutations associated with clonal hematopoiesis because they can lead to misleading assay results.
Until now, testing a patient’s PD-L1 expression level by immunohistochemistry on tissue samples was the only validated companion diagnostic for frontline immunotherapy in NSCLC. The use of cfDNA to detect tumor mutational burden (TMB) has attracted attention as a promising biomarker for response to immunotherapies in NSCLC and other tumor types but is still a moving target. Plasma assays for determining TMB are becoming more diagnostically relevant; what is important now is figuring out the respective values for determining high levels of TMB in each tumor type, because it has been determined that each disease may have its own cutoff for high TMB.
“There is a general idea that tumor mutational load is predictive of survival after immunotherapy across different tumor types, but we are still searching for the ideal cutoff value in the tissue and in the blood to separate [out] those patients that [will have] a high-relapse rate versus a low-relapse rate.” Scagliotti said.
CTCs and Plasma Proteins
Expression of PD-L1 may be more accurately assessed using CTCs than tissue samples because they are derived from more than 1 tumor site. Despite technical issues associated with biomarker testing using CTCs, persistence of PD-L1positive cells in patients treated with immunotherapy has been associated with a poorer prognosis. Investigators led by Paul Hofman, MD, PhD, called for more prospective clinical trials for assessing the use of PD-L1 expression on CTCs for initial treatment and monitoring of disease.3
Studies have also correlated the level of PD-1/L1 proteins detected in the plasma with immunotherapy treatment efficacy. Additionally, varying levels of IL-8 and angiopoietin-2 in the plasma have been associated with immunotherapy efficacy, introducing the possible validity of these as biomarkers for treatment stratification.3
“We can make the hypothesis that liquid biopsy will be a nice tool to trace minimal residual disease and also the recurrence of disease earlier than [with] morphological changes,” Scagliotti said.
A precision medicine ecosystem could bridge clinicians, laboratories, research enterprises, and clinical informationsystem developers in novel ways.3This can include clinical-decision support from various parties and case-level databases and biobanks that receive data from clinical and research workflows.
In this ecosystem, researchers and clinical labs are able to use all these data sources and also contribute new information. Factors should be considered to assess the impact of clinical development trends, Scagliotti said. This includes digital health and mobile technologies, an increased focus on patient-reported outcomes (PROs), the emergence of curated real-world data sources, use of predictive analytics and artificial intelligence (AI), shifts in the classes of agents being evaluated, availability of biomarker assays, changed in the regulatory landscape, and the availability of pools of prescreened patients or direct-to-patient recruitment.
Mobile technologies will also affect clinical trial designs, as they have already transformed clinical development, Scagliotti explained. He cited telemedicine and virtual physician visits, connected biometric sensors, consumer mobile apps, disease management apps, consumer wearables, in-home connected virtual assessments, and web-based interactive programs.
These tools have led to increased data sourcescontinuous data, contextual metadata, real-time data, and electronic PRO data—before evolving to facets like novel endpoints, digital biomarkers, companion apps, virtual trials and patient-centric designs, patient safety and centralized monitoring, virtual electronic consent, direct-to-patient recruitment, and work burden.
One forward-thinking strategy is the Translational Platform for de-orphaning malignant pleural MESOthelioma (TOPMESO), a biobank and patient samples that, through analyses, could lead to more effective therapeutic strategies for clinical studies and determine minimally invasive biomarkers.4 The biobank and patient samples would comprise established primary lines, short-term cultures, malignant pleural mesothelioma tissue, immuno-organoids, and patient-derived xenografts. Through that, researchers evaluate omics layers and molecular subtypes; exome next-generation sequencing, global RNA sequencing, and the Methylation EPIC assay; and marker analysis for drug screening to be able to gain a deeper understanding of genomics, for example.
Scagliotti said primary and secondary prevention will remain key in significantly reducing cancer mortality.
“There are 3 tumors, including lung, in which there is a relevant number of genomic abnormalities in the content of the normal tissue,” Scagliotti said. He indicated that RNA-sequencing analysis has revealed macroscopic somatic clonal expansion across normal tissues and that mutation load is associated with age and tissue-specific proliferation rates.
In an analysis led by Keren Yizhak, PhD, and colleagues, genetic diversity of individuals and how it contributes to the disease process was explored by using The Cancer Genome Atlas samples and normal tissue from the Genotype- Tissue Expression project to detect mutations by RNA sequencing. “Studying the genetic makeup of a tumor when it is already fully developed limits our ability to uncover how and which somatic mutations accumulate in normal tissues in the stages preceding cancer initiation,” they wrote.5
In their study, the investigators developed a method by which they could identify somatic mutations using tissue-derived RNA samples and its matched-normal DNA and found multiple somatic mutations, including known cancer genes, in almost all individuals. They concluded that somatic mutationcarrying genetic clones are detected across normal tissues, with the highest rates seen in sun-ex- posed skin, esophagus mucosa, and the lungs.2
Based on these results, both Scagliotti and the investigators on the study called for higher-resolution studies of normal tissues and precancerous lesions for elucidating factors that lead to early cancer development.2
Artificial Intelligence BecomesPertinent
Scagliotti emphasized the role of AI, which is being applied to medical imaging by high-lighting suspicious regions in images, detecting indeterminate nodules, and addressing high false-positive rates and overdiagnosis. It can also contribute to characterization of tumors and tumor monitoring that go behind traditional techniques. “AI promises to make strides in the qualitative interpretation of cancer imaging, including volumetric assessment of tumors over time, extrapolation of the tumor genotype and biological course from its radiographic phenotype, and prediction of outcomes,” he said.
The advent of AI may mean better detection of incidental pulmonary nodules, which can be used to predict future risk of cancer development, differentiate between benign and cancerous nodules, and/or differentiate between aggressive tumors and indolent cancers.
Another step forward involves the International Association for the Study of Lung Cancer’s cloud-based screening registry, Early Lung Imaging Confederation, developed in 2018.6 Still in its early phases, the registry stems from prior studies that demonstrated an association between lung cancer screening and reduced lung cancer mortality.4 The system connects lung cancer screening images with biomedical data to assist with fewer screening backlogs, but with a larger picture to increase reliability of clinical decision support with computed tomography images and improve the development of precise quantitative disease biomarkers.
Due to the ever-changing treatment landscape, it is increasingly important to keep up with the latest advancements in the precision ecosystem for its eventual applicability to clinical practice.
“This is just the beginning of the story,” Scagliotti concluded.