43 MP2—The Next-Generation MarginProbe Device: Incremental Performance Gains in an Area of Persistent Unmet Need

Background

Despite widespread adoption of consensus margin guidelines, positive margins after breast-conserving surgery (BCS) remain common, resulting in reexcision rates of approximately 20% for invasive cancer and up to 30% for ductal carcinoma in situ (DCIS). Positive margins are associated with increased local recurrence and additional patient and health-system burden. Since its introduction, the first-generation MarginProbe (MP) has been evaluated in multiple peer-reviewed studies over the past decade, consistently demonstrating a 40% to 50% reduction in positive margin rates, with particularly strong effects in DCIS. However, there remains a paucity of FDA-approved, noninvasive technologies for real-time intraoperative lumpectomy margin assessment. MP2 represents a next-generation iteration incorporating multiple miniaturized radiofrequency spectroscopy sensors and updated algorithms designed to improve diagnostic performance.

Methods

This prospective, multicenter, single-arm, open-label study (NCT05377229) included distinct learning and pivotal phases. Women with histologically confirmed breast cancer undergoing BCS were enrolled. MP2 measurements were performed intraoperatively on freshly excised lumpectomy specimens (ex vivo). The learning phase was used to optimize device software and signal processing. Primary performance end points in the pivotal phase were sensitivity and specificity for margin assessment at the specimen face level. Performance was benchmarked against goals derived from the MP1 postapproval study, with a predefined noninferiority margin of 10%. Margin positivity prompting reexcision was defined according to consensus margin guidelines.

Results

A total of 117 patients were enrolled, with 114 completing the pivotal phase. One-hundred-nineteen lumpectomy specimens contributed 706 margin faces for final performance analysis. Using 5-15 measurements per face, MP2 demonstrated a sensitivity of 76.8% and a specificity of 49.1% (Table). With a measurement restriction simulation (maximum 5 readings per face), sensitivity was 74.1% and specificity improved to 58.9%. All primary performance end points met or exceeded predefined success criteria. No device-related adverse events were observed. Based on these results, MP2 received FDA approval in September 2025 without regulatory restriction.

Conclusion

In a clinical landscape with limited FDA-approved options for intraoperative lumpectomy margin assessment, MP2 demonstrates incremental but meaningful improvements in diagnostic performance over its predecessor. These gains build upon a substantial body of prior evidence showing significant reductions in positive margins with MP1, particularly in DCIS. The multisensor architecture and optimized algorithms of MP2 improve discrimination between malignant and benign tissue, including dense breast tissue, while aligning margin detection depth with current invasive cancer and DCIS guidelines. Together, these findings support MP2 as a regulatory validated, technically mature advancement addressing a persistent and clinically relevant gap in BCS.