FDA 510(k) Clinical Strategy Status: Legit.Health US Version

This document serves as the master tracking record for the clinical strategy and regulatory pathway of the Legit.Health US version, intended for FDA 510(k) clearance. It details the current state-of-the-art, the clinical evidence generation plan, project milestones, and the immediate next steps required to achieve market authorization.

1. Executive Summary

Legit.Health is pursuing FDA 510(k) clearance for its US software-aided adjunctive diagnostic device. The device utilizes Computer Vision and Convolutional Neural Networks (CNN) to evaluate skin lesions suggestive of melanoma, basal cell carcinoma (BCC), and squamous cell carcinoma (SCC). The primary objective is to assist Primary Care Physicians (PCPs) in making referral decisions to dermatologists.

The identified predicate device is DermaSensor (DEN230008, Product Code: QZS). While DermaSensor utilizes Elastic Scattering Spectroscopy (ESS), Legit.Health utilizes clinical imaging. Therefore, our clinical strategy is meticulously designed to demonstrate Substantial Equivalence (SE) and satisfy the Special Controls established under 21 CFR 878.1830, specifically proving superior accuracy of device-aided users compared to unaided users and maintaining a standalone sensitivity of at least 90% for high-risk lesions.

2. Clinical Strategy Overview

To meet the stringent Special Controls and demonstrate that the transition from ESS to CNN-based image analysis does not adversely affect clinical sensitivity or introduce new risks, the following three clinical studies form the core of our regulatory submission:

2.1. Retrospective Standalone Performance Study

• Objective: To demonstrate that the device standing alone achieves a diagnostic sensitivity of at least 90% for high-risk lesions, validating non-random performance (Sensitivity + Specificity > 1).
• Design: A retrospective, cross-sectional study utilizing an enriched dataset (1:1 malignant to benign ratio) to ensure sufficient representation of lower-prevalence subgroups, specifically Fitzpatrick skin phototypes IV-VI and patients aged under 40, since this was one of the main limitations of DermaSensor studies.
• Ground Truth: Consensus histopathology reviewed by a panel of two to five central dermatopathologists.

2.2. Multi-Reader Multi-Case (MRMC) Second-Read Study

• Objective: To demonstrate the superior diagnostic sensitivity of PCPs when aided by the Legit.Health device compared to their unaided clinical assessment (the adjunctive benefit).
• Design: A prospective, blinded, sequential-read MRMC study involving at least 30 board-certified PCPs evaluating a subset of 100 cases (50 malignant, 50 benign) from the standalone study.
• Key Endpoint: Corrective Direction and Clinical Utility, measuring the ratio of correct diagnostic conversions (unaided "Monitor" to aided "Refer") versus incorrect conversions.

2.3. System Reproducibility and Robustness Study

• Objective: To address the unique challenge of image-based AI by proving that the diagnostic sensitivity is not altered by the image capture device (e.g., various smartphones and dermatoscopes).
• Design: A prospective, observational, cross-sectional method-comparison study utilizing a "Repeated Measures" design on at least 50 prospectively captured lesions.
• Key Endpoint: Achieving an Overall Percent Agreement (OPA) point estimate of >= 95% between different validated acquisition systems.

2.4. Key FDA Pre-Submission Questions

To de-risk the 510(k) pathway, a Pre-Submission (Q-Sub) meeting will be held to secure Agency alignment on five critical areas:

1. Acceptability of retrospective data and training data independence.
2. Adequacy of the proposed MRMC study design (30 readers, sequential-read methodology).
3. Appropriateness of the Statistical Analysis Plan (SAP), specifically the use of Generalized Estimating Equation (GEE) models.
4. Validation of the prospective approach to evaluating image acquisition system robustness.
5. Representativeness of the study population regarding U.S. intended use (specifically the enrichment strategy for Fitzpatrick skin phototypes IV-VI).

3. Project Schedule & Milestones (Gantt Chart)

The following schedule represents the high-level roadmap for the 510(k) submission.

Detailed Phase Breakdown

Each phase represents a critical step in building the clinical evidence required for FDA clearance.

• Phase 1: Definition of Clinical Strategy (Completed - March 2026): Consisted of finalizing the core strategy, identifying the DermaSensor predicate, and defining the requirements for the three primary clinical studies (Standalone, MRMC, and Reproducibility).
• Phase 2: Pre-Submission (Q-Sub) Preparation & Submission (In Progress - March to May 2026): Consists of drafting the formal Q-Submission package, including the cover letter, device description, and clinical validation protocols. The goal is to submit by late May 2026.
• Phase 3: FDA Answer to Pre-Submission (June to August 2026): Consists of the 70-75 day waiting period for FDA feedback. This phase includes the potential for a Q-Sub meeting to clarify study designs and ensure alignment on endpoints.
• Phase 4: Finalize Clinical Study Documentation (August to September 2026): Consists of updating the Clinical Study Protocols (CSPs) and Statistical Analysis Plans (SAPs) to incorporate the specific feedback received from the FDA.
• Phase 5: IRB Review and Approval (October 2026): Consists of submitting the finalized protocols and informed consent documents to an Institutional Review Board to ensure ethical compliance for both retrospective and prospective data usage.
• Phase 6: Clinical Studies Execution & Data Collection (November 2026 to February 2027): Consists of the active recruitment phase. This involves dermatopathologist panel reviews, the 30-reader PCP MRMC study execution, and prospective patient recruitment for the hardware robustness study.
• Phase 7: Data Analysis & Final Clinical Reports (March 2027): Consists of locking the clinical databases, running the statistical models (e.g., GEE), and authoring the formal clinical study reports (CSRs) for inclusion in the 510(k).
• Phase 8: FDA 510(k) Compilation and Submission (April to May 2027): Consists of the final regulatory assembly, ensuring all modules (Technical File, Clinical Reports, Labeling, etc.) are complete and submitting the final 510(k) package to the FDA.

4. Current Status

As of March 2026, the project has successfully completed Phase 1. The core clinical strategy has been definitively established, the predicate device (DermaSensor) has been thoroughly analyzed, and the protocols for the three necessary clinical validations have been outlined.

The project is currently entering Phase 2: Pre-Submission Preparation. The focus has shifted from internal strategy definition to drafting the formal regulatory documents required to solicit binding feedback from the FDA regarding our proposed clinical pathway.

5. Immediate Next Steps

Priority	Task	Responsible	Backup/Escalation	Target Completion Date	Status
1	Draft the Pre-Submission Cover Letter and Meeting Request: Formally request a meeting with the FDA to discuss the 5 defined questions.	Saray	[Designate backup if unavailable]	May 15, 2026	In Progress
2	Compile the Device Description Document: Prepare a comprehensive overview of the Legit.Health US version, detailing the CNN algorithm, intended use, and indications for use.	Saray & Jordi	[Designate backup if unavailable]	May 15, 2026	Not Started
3	Formalize Clinical Protocols: Expand the summaries provided in the Clinical Strategy document into formal, structured clinical protocol drafts for the Retrospective Standalone, MRMC, and System Reproducibility studies to include in the Pre-Sub package.	Jordi	[Designate backup if unavailable]	May 15, 2026	Not Started
4	Draft the Statistical Analysis Plan (SAP) Synopsis: Prepare the specific statistical methodologies (e.g., GEE models, Obuchowski-Rockette method) for Agency review.	Jordi	[Designate backup if unavailable]	May 15, 2026	Not Started