Written by: Lori Lombardo
VP & Therapeutic Area Leader, Dermatology and General Medicine
Q: How is photography influencing dermatology clinical trials today?
A: Photography has become indispensable in dermatology clinical trials, enhancing data quality and enabling more reliable assessments. Recently, at Caidya’s webinar Dermatology Imaging and Innovation: Raising the Quality Bar, we explored how imaging technologies, coupled with advances in artificial intelligence (AI), are revolutionizing how clinical trials are conducted. These tools help improve trial efficiency, support objective evaluation, and ensure regulatory compliance, setting new standards for dermatology research.
Q: What are the key regulatory considerations for using photography in clinical trials?
A: Regulators require that clinical trial protocols and statistical analysis plans clearly specify how images will be captured, evaluated, and analyzed. This is critical to ensure consistency across multiple sites. Standardization of factors like lighting, camera angles, distance, and equipment is necessary to reduce variability. When images serve as efficacy or safety endpoints, blinded assessments are essential to maintain objectivity.
Additionally, photographs are treated as source data and must be managed with strict data integrity controls—this means images must be attributable, unaltered, audit-trailed, and securely stored in line with regulations such as 21 CFR Part 11 in the US and Annex 11 in Europe. Patient privacy is paramount, with informed consent and adherence to laws like HIPAA and GDPR mandatory. Transparency with patients about how images will be used, stored, and shared is a regulatory expectation.
Typically, sponsors develop a photography charter or standard operating procedure (SOP) that defines equipment, image capture settings, staff training, and quality control. This document becomes part of the trial master file and is foundational to regulatory acceptance.
Q: What challenges does remote photography present, and how is innovation addressing these?
A: Remote photography can reduce patient burden by minimizing travel and allowing data collection from remote locations. However, it’s generally limited to early-phase studies or simple endpoints like confirming wound healing because of concerns around image quality, consistency, and maintaining patient anonymity and chain of custody.
While emerging technologies such as 3D imaging and dermatoscopic photography improve diagnostic precision in telemedicine, they currently cannot replace thorough in-person clinical examinations, particularly for complex diseases like psoriasis. Patient-captured images, such as selfies, often lack standardization and validation, which limits their use as primary endpoint data in clinical trials. Therefore, validated photography systems remain necessary to ensure data reliability.
Q: How is dermatoscopy being integrated into clinical trials?
A: Dermatoscopy provides magnified and illuminated views of the skin, revealing structures beneath the surface, such as collagen fibers, hair follicles, and blood vessels. Traditionally used to evaluate potentially malignant lesions, dermatoscopy is gaining traction in inflammatory skin disease trials.
Although validated scales for dermatoscopic assessment of inflammatory conditions are not yet established, emerging research suggests this technique may deepen our understanding of disease mechanisms. However, clinical trials still primarily rely on traditional endpoints assessing visible clinical improvement, and the full role of dermatoscopy in endpoint determination is still evolving.
Q: What role does artificial intelligence (AI) currently play in dermatology clinical trials?
A: AI is increasingly integral to dermatology trials, primarily by improving trial management processes. It provides real-time feedback to study coordinators, enhances data quality, and reduces the burden of data reconciliation. AI also aids in identifying potential patient health information (PHI) on images to ensure compliance with privacy laws.
Looking forward, AI holds promise to transform endpoint assessments by bringing objectivity to traditionally subjective measures. It can quantify lesion characteristics such as size and redness, turning descriptive clinical evaluations into precise, numerical data. Deep learning models are also being developed to classify lesion types, improve diagnostic accuracy, assess severity, and discover novel disease associations, which could profoundly impact future trial designs and clinical practice.
Q: What regulatory frameworks guide the use of AI and imaging tools in clinical trials?
A: The FDA and EMA have issued guidance to ensure AI and machine learning tools meet stringent validation standards. The FDA focuses on validated systems compliant with 21 CFR Part 11, emphasizing accuracy, auditability, data integrity, risk-based validation, algorithm transparency, bias mitigation, and continuous performance monitoring.
Similarly, the EMA requires compliance with Good Clinical Practice (GCP) principles, including thorough documentation of image collection, metadata retention, and validation processes. Both agencies expect sponsors to provide detailed information on AI architectures, training datasets, and performance metrics to demonstrate reproducibility and data control.
Conclusion
The integration of advanced imaging and AI technologies in dermatology clinical trials is redefining the landscape, enabling higher data quality and greater trial efficiency. However, these innovations must be carefully balanced with rigorous standardization, validation, and regulatory compliance to maintain scientific integrity and protect patient privacy. As the field continues to evolve, it is imperative that sponsors establish robust protocols and adopt emerging best practices to raise the quality bar for dermatologic endpoints, ultimately advancing the development of safe and effective treatments.