NSCLC Clinical Pathways in the Era of Molecular Complexity, AI, and Value-Based Care

Key Takeaways 

• As non-small cell lung cancer (NSCLC) becomes increasingly segmented by molecular subtype, tools such as artificial intelligence (AI) can enhance guideline navigation, but individualized treatment decisions—particularly around toxicity tolerance and patient preferences—remain dependent on physician expertise.
• While radiographic progression remains the standard for detecting resistance, emerging technologies such as minimal residual disease assays and liquid biopsy hold promise for earlier detection and improved sequencing strategies, though tissue biopsy remains essential in many relapse settings.
• Proactive supportive care models, alongside research into cost-effective sequencing strategies and broader access to biomarker-driven therapies, are critical to sustaining high-quality, equitable cancer care in an era of increasingly expensive targeted treatments.

Jorge J Nieva, MD: My name is Dr Jorge J Nieva. I am an academic medical oncologist with 25 years of experience in oncology. I trained in San Diego and began my career at Scripps Clinic, where I served as Director of Translational Research for the Scripps Cancer Center. I later joined Billings Clinic in Montana, where we introduced advanced technologies—including Phase I clinical trials and gene therapy trials—to a community cancer center in the Rocky Mountain West. For the past 11 years, I have been based in Los Angeles at the USC Norris Comprehensive Cancer Center, where I serve as Section Head of Thoracic and Head and Neck Tumors.

Given your extensive work in EGFR- and ALK-mutated NSCLC, how should clinical pathways evolve to keep pace with increasing molecular subtypes while remaining practical and scalable for real-world care delivery?

Dr Nieva: We are fortunate to be practicing in an era in which artificial intelligence (AI) has significantly improved the ability to search clinical guidelines and identify the most appropriate recommendations for individual patients—far more efficiently than even 2 years ago. Although AI has limitations, it is particularly effective as a search tool. Therefore, physicians must remain adept at validating AI-generated information by confirming primary sources and contextualizing recommendations within established models of cancer care.

Cancer care is inherently complex, as is medicine more broadly. However, modern tools allow us to manage this complexity more effectively. At the same time, guidelines must be applied in a patient-centered manner that ensures an appropriate balance between efficacy and toxicity. Not all patients can tolerate the same degree of treatment-related toxicity. It is the physician’s responsibility to determine which patients can undergo intensive therapy and which may require a more conservative approach.

This level of clinical judgment is difficult to replicate algorithmically and must begin with a comprehensive clinical assessment, including functional status, comorbidities, and individual patient preferences.

How should clinical pathways better account for resistance monitoring, re-biopsy, and optimal sequencing to avoid ineffective treatment delays?

Dr Nieva: At present, therapeutic resistance is best determined radiographically. Blood-based assays have not yet been validated as reliable indicators of therapeutic futility. However, I am optimistic that we are not many years away from identifying resistance at a molecular level before overt radiographic progression occurs.

Future clinical pathways may enable real-time detection of drug resistance, but we are not yet at that stage. In early-stage disease, minimal residual disease (MRD) assays show promise as risk-stratification tools for recurrence. As liquid biopsy technologies continue to improve, we may eventually reach a point at which tissue biopsy at relapse is no longer necessary. Currently, tissue biopsy remains essential to identify small cell transformation and other resistance mechanisms that may not be reliably detected with existing liquid biopsy platforms.

Regarding sequencing, we aim to use the most appropriate therapy first. Importantly, the “best” therapy differs across patients. For some, it is the most efficacious agent; for others, it may be the least toxic option. Determining the optimal strategy requires individualized clinical judgment based on the patient in front of us.

What are some ways that pathways could better incorporate proactive toxicity management to improve adherence, outcomes, and overall value of care?

Dr Nieva: We are practicing in an era in which many highly effective therapies are also associated with substantial toxicity. It is imperative that, as these agents are developed, comprehensive systems for toxicity mitigation are implemented concurrently.

Some pharmaceutical companies have recognized this responsibility. For example, Johnson & Johnson’s Cocoon program represents a comprehensive approach to addressing toxicity associated with certain targeted therapies. The program includes clinical trials focused on supportive care strategies aimed at preserving quality of life for patients receiving EGFR and MET co-inhibition. Although other companies have also implemented strong supportive care initiatives, the Cocoon program serves as a particularly robust model. It provides a useful framework for improving tolerability as new therapies enter clinical practice.

How do you recommend NSCLC clinical pathways be designed to support value-based decision-making while ensuring equitable access to biomarker-driven therapies across diverse patient populations?

Dr Nieva: The cost of cancer therapeutics remains a significant burden on the healthcare system. Drug pricing affects not only patients but also the sustainability of care delivery more broadly. There are ongoing policy efforts within the United States aimed at addressing how the costs of these therapies are distributed internationally so that the financial burden does not fall disproportionately on American patients.

Additionally, as newer targeted agents lose patent protection and become available in generic form, it is important to evaluate whether sequential use of lower-cost options may provide overall survival outcomes comparable to those achieved with high-cost first-line therapies. Clinical research should prioritize strategies that maximize survival and quality of life while maintaining cost-effectiveness. Value-based care must balance innovation, accessibility, and sustainability.