Improved outcomes occur when women with ovarian cancer have their initial surgery with a gynecologic oncologist. ROMA™ (Risk of Malignancy Algorithm) and IVDMIA (In Vitro Diagnostic Multivariate Index Assay [OVA1®]) have been FDA-cleared to aid in the presurgical assessment of women with an adnexal mass. We present here a comparison of ROMA to a second-generation IVDMIA (MIA2G).METHODS:
Both tests were run on a combined cohort (N=993) from two published registries of women undergoing adnexal mass surgery. ROMA was calculated using CA 125-II and HE4 levels with cutoffs derived from the ROMA Roche Fact Sheet. MIA2G was calculated using CA 125-II, HE4, transferrin, FSH and Apolipoprotein A1 (ApoA1) levels with proprietary OVACALC 4.0 software using a single validated cutoff of 5.0 for all subjects. Performance parameters were calculated for the cohort, and sensitivity was further stratified by stage or subtype.RESULTS:
245 subjects (24.7%) had a malignancy, of which 166 (67.8%) were primary ovarian cancers. Mean subject age was 50.3 years (range 18–92), and included 487 (49.0%) postmenopausal subjects. ROMA misclassified 51 malignancies (including 10 high-grade ovarian malignancies) whereas MIA2G misclassified 22 (including 5 high-grade ovarian malignancies). Early stage cancers were more frequently misclassified by ROMA (20 vs 8 cases). The rate of “test-negative” malignancies was significantly higher for ROMA, while the rate of “test-positive” benign cases was significantly higher for MIA2G.CONCLUSION:
Identifying women with ovarian cancer who present with an adnexal mass is critical in optimizing clinical outcomes; significant differences in IVDMIA performance may impact optimal clinical care.