ERS5750. Screening Outcomes with Contrast-Enhanced Mammography (CEM): Initial Experience in a South Florida Multi-Center Breast Practice
Authors * Denotes Presenting Author
  1. Cedric Pluguez-Turull; University of Miami/Jackson Health System
  2. Antoine Hamedi-Sangsari *; University of Miami/Jackson Health System
  3. Aazim Arif; University of Miami/Jackson Health System
  4. Taylor Schwartz; University of Miami/Jackson Health System
  5. Matthew Alfonso; University of Miami/Jackson Health System
  6. Jose Net; University of Miami/Jackson Health System
This study aims to explore the performance outcomes after implementation of screening CEM in a multicenter breast imaging practice in South Florida.

Materials and Methods:
This IRB-approved and HIPPAA compliant retrospective study included CEM exams since implementation of this technology at our breast imaging centers since 2021. Data on demographic composition, study indication, BIRADS assessment, biopsy recommendation, pathologic outcome after core needle biopsy and interval cancers was collected. Biopsy rate, positive predictive value (PPV) for malignancy, sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), cancer incidence, cancer detection rate (CDR) and interval cancer rate were calculated.

258 patients with median age of 59 years (26 – 83 years) and a total of 347 CEM performed for a screening indication were accrued with a median follow-up time of 17.5 months (5 - 40 months). Screening indication was varied and included: past medical history of breast cancer, dense breasts, prior high risk lesion, above average calculated overall lifetime risk and breast cancer associated gene carrier or untested with affected relative. Biopsy rate was 5.5% (19/347) and the overall positive predictive value for malignancy (PPV3) after core needle biopsy was 42.1% (8/19). Overall cancer incidence was 2.3% (8/347), including DCIS and invasive cancer, for a CDR of 23.1/1000. The interval cancer rate in this cohort was 5.8 per 1,000 patients (2/347). Both interval cancers were identified on high risk screening MRI. The first was performed due to strong family history where MRI performed 9 months later revealed 1.3 cm area of nonmass enhancement, biopsy confirmed ductal carcinoma in situ; no finding was visible in retrospect. The second interval cancer was on a BRCA2+ patient, MRI performed 7 months later revealed a new 1.2 cm mass, biopsy confirmed triple negative invasive ductal carcinoma; no mass was present on CEM in retrospect. The overall sensitivity was 80% (44.4-97.5%), specificity was 96.4% (93.6 – 98.2%), PLR was 21.9 (11.34 – 42.26) and NLR was 0.21 (0.06 - 0.72).

CEM improves cancer detection compared to both digital mammography and digital breast tomosynthesis, and shows similar performance to Breast MRI in the diagnostic setting. Our analysis revealed high PPV for malignancy on CEM (42.1% and high CDR (23.1/1000), but also an unexpectedly high interval cancer rate based on 2/347 interval cancers, both which were identified on subsequent high risk screening breast MRI- one small focus of DCIS and one triple negative invasive cancer. While emerging data is promising, CEM requires further validation for screening. This study aims to increase the available literature on screening CEDM particularly in the classically underrepresented Hispanic-enriched population of South Florida.