Read time: 7 minutes.
Lung cancer sits at the forefront of precision medicine, with a large number of targeted therapies approved to treat patients based primarily on driver mutations in their tumor’s DNA. However, precision medicine is now leveraging other data beyond genomics to better tailor the right treatments, for the right patients, at the right time. Further, liquid biopsy is also being increasingly leveraged not only in directing how cancer is treated, but how it can be detected in the first place.
The 33rd Annual Precision Med TRI-CON brought experts from the pharmaceutical and diagnostics industries, and research thought leaders, together in San Francisco to discuss:
- Ongoing advances in precision medicine
- Key considerations for implementation
- Ongoing challenges of getting existing targeted treatments to the right patients today
The Evolution of AI-Driven, Multimodal Precision Medicine
Precision medicine has shifted the way cancer and other diseases are treated, from one-size-fits-all approaches to personalized treatments better tailored to the individual. Still, not all patients with a given biomarker will necessarily respond to a matched targeted therapy, and those who do respond may eventually develop resistance. To address these challenges, drug and diagnostic developers are seeking previously undiscovered biomarkers associated with treatment response and resistance mechanisms, to improve matching of patients to effective and durable personalized therapies.
To do this, conference attendees discussed leveraging artificial intelligence (AI) to analyze large amounts of data from multiple sources to reveal previously undiscovered biological patterns associated with clinical outcomes, including treatment response and resistance. Experts describe this process as “reverse translation,” a strategy wherein data from different sources (e.g., clinical trial data, imaging data, electronic health records (EHRs), etc.) are leveraged to inform precision medicine research and development. This approach reveals new patterns of disease development and distinct patient subpopulations that may be missed by single-modality approaches.
A key bottleneck to the use of AI-enabled tools in precision medicine research and development revolves around robust data. The optimal performance of an AI tool depends upon the quality of the data used to train it, and some datasets may be “noisy” and poorly curated, particularly certain real-world data sources like EHRs.
Stakeholders at the conference noted, though, the improving ability of other AI tools to organize and extract information from datasets in more structured ways. High-quality data must also be accessible, though datasets are often kept proprietary and siloed. A key challenge will be improving data sharing while also maintaining incentivization of its collection and consolidation. Furthermore, while some companies are open to sharing data with certain collaborators, the data privacy rights of patients are a critical consideration.
These AI-enabled tools can help identify new predictive biomarkers for companion diagnostic development. In fact, experts discussed a shift to more multimodal strategies in precision medicine, wherein AI-assisted analysis of multiple data sources linked to a patient can be used to tailor treatment decisions. Discussions focused on the need for clear pathways to market for these diagnostics, and their regulation and reimbursement will hold unique considerations.
Regulators must ensure AI-driven diagnostics are robust, with lifecycle management of device performance as opposed to one-time validation. The need for appropriate regulatory flexibility was also noted, with stakeholders citing the usefulness of tools like predetermined change control plans for device modifications. Stakeholders also stressed the importance of generating adequate clinical validity and utility data for these novel tools to not only support insurance coverage decisions, but build trust with providers and drive adoption in guidelines.
The Advancing Use of Liquid Biopsy in Precision Medicine
The conference also illuminated how advances in precision medicine are increasingly leveraging analytes from liquid biopsies, particularly from the blood. Molecular residual disease (MRD) testing in solid tumors is a rapidly evolving area, with use cases including guiding of treatment decisions in the adjuvant setting, surveilling for recurrence, and assessing response to therapy. Participants noted the promise MRD may hold in serving as a surrogate endpoint for clinical outcomes in trials assessing new treatment regimens for solid tumors which, provided sufficient evidence and validation, could transform the pace and conduct of clinical trials.
The potential of MRD testing to enable adjusting a patient’s therapy sooner than radiologically evident progression was also raised. Stakeholders pointed out the importance of demonstrating that switching therapy based on an MRD test result prior to radiologic findings results in improved patient outcomes, aligning with a key concern raised in the April 30th meeting of the US Food and Drug Administration’s (FDA) Oncologic Drug Advisory Committee (ODAC) related to detection of treatment resistance mutations in the blood.
The conference also highlighted recent advances in the development of blood-based multi-cancer early detection (MCED) and single-cancer early detection (SCED) tests. Biomarkers being developed for these assays have expanded beyond just circulating tumor DNA (ctDNA), with tests being developed to detect proteins, protein modifications, extracellular vesicles, and other analytes as cancer signals. Multiomic approaches to blood-based cancer detection are also on the horizon, with combinations of different analytes potentially facilitating more accurate early cancer detection. However, participants stressed the importance of prioritizing clinical validity, noting that combining analytes may compromise test specificity.
Experts noted that blood-based screening assays are not meant to compete with more established and more sensitive cancer screening programs, noting the primary goal of MCEDs being to expand the reach of screening to more cancer types lacking screening programs. The importance of robust clinical validation of MCEDs and SCEDs was also stressed, as was the importance of clinical utility data to support coverage and adoption, though clinical utility studies may be very difficult and costly to run. Clear reimbursement pathways must also be established for effective new blood-based cancer detection assays to ensure not only that they are appropriately covered, but that such coverage will not compromise reimbursement of subsequent diagnostic testing.
Ongoing Challenges in Precision Medicine Implementation
The conference highlighted that implementation challenges are not unique to new technologies. Despite the remarkable strides made in precision oncology, its implementation in cancer care today remains far from the ideal. A 2022 study showed that two-thirds of patients with non-small cell lung cancer (NSCLC) who are eligible for precision oncology treatments are not benefiting from them. While some hurdles are related to the conduct of biomarker testing itself (e.g., lack of available tissue), suboptimal treatment decisions are sometimes made regardless of testing results.
The growing number and complexity of biomarkers as precision medicine becomes more multimodal could make it even more complicated and difficult for healthcare providers to make the best treatment decisions for their patients.
Meeting participants stressed the importance of improving communication channels between pathologists and treating physicians, as well as the importance of molecular tumor boards (MTBs) to drive greater adoption of precision medicine. Stakeholders advocated for greater investment in MTBs and in their broader implementation, noting most only exist at large medical centers and are typically reserved for “more difficult” cases.
LUNGevity’s Perspective
As greater amounts of data and data sources are leveraged in precision medicine research and development, LUNGevity stresses the importance of adherence to data privacy protections for patients.
We strongly advocate for the robust validation of diagnostic tests developed to direct treatment decisions for patients, as well as for screening purposes, particularly the lung cancer SCEDs becoming increasingly clinically available. While clinical utility studies of these tests may be difficult to conduct, they are vital to instill confidence that their use will improve outcomes for patients. We also support the robust validation of MRD-related biomarkers, particularly as potential surrogate endpoints for long-term clinical benefit, which could accelerate pathways to market—particularly for early-stage lung cancer therapies.
LUNGevity is very familiar with the present challenges to the implementation of precision medicine and is active in the development of solutions to overcome them. We stand as a leader in driving patient education, awareness, and understanding of personalized medicine, advocating for the appropriate coverage of biomarker testing, and driving practical solutions to improve precision medicine implementation through multi-stakeholder collaborations.
Looking Ahead
The use of AI, applications of more multimodal approaches, and greater leveraging of blood-based biomarkers are ushering cancer precision medicine into the future. While this year’s Precision Medicine TRI-CON spotlighted key advancements in the field, it also highlighted important caveats and key considerations related to their validation and clinical implementation.
The increasing personalization of treatment strategies, coupled with concerted efforts to remove hurdles to their implementation, will bring the field closer to what some attendees highlighted as the ultimate goal: that every cancer patient be treated according to the molecular profile of their disease, and that precision medicine be simply referred to as medicine.