Designing ILD trials to reflect real-world clinical practice

Q. How can inclusion and exclusion criteria be designed to capture a more representative ILD population while maintaining trial safety and data integrity?

One of the biggest challenges of ILD trials may not be the intervention, but who we choose to include and how that impacts the real-world utilization of medications once approved. ILD trials generally have overly narrow eligibility criteria, but in practice, our patients are older, have multiple comorbidities, limited mobility, and often fall into diagnostic gray zones. Rather than excluding these patients, a more pragmatic approach is to include them and then stratify by key variables, such as background therapy and disease severity. The rationale for setting lower forced vital capacity (FVC) and diffusing capacity of the lungs for carbon monoxide (DLCO) thresholds is to mitigate safety risks associated with enrolling patients with more advanced disease. However, the disease trajectory of patients currently included in trials varies when compared to the disease trajectory in real-world cohorts.

Safety can still be preserved through thoughtful guardrails, such as closer early monitoring, or predefined stopping rules, rather than blanket exclusions. This allows us to maintain internal validity while improving generalizability.

Q. How can trial designers account for the low prevalence and diagnostic delays associated with many ILD subtypes, and what strategies help ensure adequate recruitment in rare-disease populations?

Recruitment in ILD is inherently challenging because, although the incidence is increasing, it is rare and often misdiagnosed or diagnosed late. Early engagement of the physician community and enhancing education on ILD can reduce diagnostic delay. Leveraging established networks and national registries can significantly expand reach and accelerate enrollment.

At the same time, advancing trial designs such as platform trials, including REMAP-ILD (Randomized, Embedded, Multifactorial, Adaptive Platform trials in interstitial lung disease), allows multiple therapies to be evaluated simultaneously within a single master protocol, reducing the number of patients assigned to placebo and improving overall efficiency. By enabling adaptive randomization, shared control groups, and the ability to add or remove treatment arms over time based on emerging results, these designs streamline evidence generation and make it possible to identify effective therapies more rapidly while minimizing patient exposure to ineffective interventions. Decentralized trial models, which incorporate telemedicine, local testing, and remote consent, can also help reduce geographic and logistical barriers.

Q. What strategies can trial designers use to align background standard-of-care therapy with real-world practice, also taking into consideration differences in global accessibility?

In real-world ILD care, patients are rarely treatment-naïve, so requiring washout periods or restricting background therapy can make trials less relevant and harder to enroll. A more practical approach is to allow stable standard-of-care therapies, whether antifibrotics or immunosuppressants, and then stratify accordingly. It is also important to recognize that access to therapies varies globally, so protocols should allow flexibility across regions and adjust analyses for geographic variation to address this heterogeneity in care. Pragmatic and platform designs, such as REMAP-ILD, further embed trials within routine clinical practice, while use of real-world evidence during design ensures SOC definitions are globally feasible and clinically relevant.

Q. How can endpoints be chosen or adapted to reflect outcomes that matter most in routine ILD care?

While FVC decline has been a cornerstone endpoint and has been validated in multiple studies, it does not fully capture what matters most to patients. Increasingly, there is a shift toward composite and hierarchical endpoints that reflect the broader disease experience such as progression, hospitalization, acute exacerbations, and survival. Just as importantly, incorporating patient-reported outcomes like dyspnea and quality of life helps ensure that trials are measuring meaningful benefit.

Q. How can variability in site capabilities be addressed to ensure data consistency and real-world relevance across global research settings?

Global ILD trials inevitably involve sites with varying levels of expertise and infrastructure, and rather than viewing this as a limitation, it can enhance external validity if managed well. During protocol design, it is important to consider site capabilities and distinguish between essential assessments required for successful trial implementation and those that are desirable but not critical, ensuring feasibility while maintaining scientific rigor. At the same time, centralization of key elements like imaging review, pulmonary function test over-reading, and endpoint adjudication helps maintain consistency. Standardized training can further reduce variability.

Q. How can trial monitoring schedules be structured to reflect realistic clinical workflows while still ensuring robust data collection?

Trial visit schedules are often more intensive than routine care, which can create unnecessary burden for both patients and sites. Aligning study visits with typical clinical follow-up intervals such as every 3 to 6 months can make participation more feasible. This can be complemented by hybrid approaches, including virtual visits and phone calls. While we are yet to validate the use of home-based monitoring tools, like spirometry or wearable devices, as primary endpoints, they can provide meaningful data and should be incorporated when feasible. Ultimately, the goal is to design trials that integrate seamlessly into clinical workflows while still capturing high-quality, reliable data.