Table 1 Overview of selected multigene signatures for breast cancer
From: Developing a new generation of breast cancer clinical gene expression tests
Assay; platform, clinical material | Training parameter | Approval or endorsement | Analytical validity: published assay validation | Clinical validity: prognosis validation | Predicting treatment benefit using randomized clinical trials | Randomized prospective trials | ||||
|---|---|---|---|---|---|---|---|---|---|---|
Tamoxifen | Herceptin | Chemotherapy versus no chemotherapy | Specific agent: anthracycline | Specific agent: taxane | ||||||
Breast Cancer Index; RT-PCR, FFPE (central) | Outcome (ER+, pN0, endocrine-treated women) MGI component – biology (tumor grade related genes) H:I component – outcome (recurrence in tamoxifen-treated women) | No | No | ATAC [13], Stockholm [17], multiple nonrandomized trial cohorts | Noa | No | No | No | No | No |
EndoPredict; RT-PCR, FFPE (distributed) | Outcome (distant recurrence in endocrine-treated ER+/HER2− pN0/pN+ women) | CE Mark | No | No | No | No | GEICAM/9906 [1] (failed to predict benefit) | No | ||
IHC4; IHC, FFPE (distributed) | Outcome (distant recurrence in ER+ endocrine-treated women) | No | No | No | No | No | No | No | No | |
MammaPrint; microarray, fresh and FFPE (central) | Outcome (5-year metastasis rate in pN0 women) | FDA (fresh): risk for distant metastasis, <61 years, stage I and II, tumor ≤5 cm and node-negative | No | Multiple nonrandomized trial cohorts including RASTER | No | No | No | No | No | MINDACT prognosis validation (to report 2015) |
Mammostrat; IHC, FFPE (central) | Outcome (unselected cohort of breast cancer patients) | No | No | NSABP-B14, NSABP-B20 [15] multiple nonrandomized trial cohorts | No | No | NSABP-B20 (±CMF) [15] (all women benefit – high risk benefit the most) | No | No | No |
Oncotype DX; RT-PCR, FFPE (central) | Outcome (recurrence in mainly tamoxifen-treated ER+, pN0 women) | NCCN, ASCO, St. Gallen (role for identifying women that may benefit from chemotherapy) | Yes [21] | NSABP-B14 [9], NSABP-B28 [22], SWOG8814 [23], multiple nonrandomized trial cohorts | NSABP-B14 [24] (largest benefit in quantitative ER high/recurrence risk low patients) | No | NSABP-B20 (±CMF) [25], SWOG8814 (±CAF) [23] (benefit from chemotherapy with high recurrence score) | No | NSABP-B28 [22] (failed to predict a benefit) | TAILORx (node-negative, to report 2015) RxPONDER (one to three positive nodes, recruiting) |
PAM50 (research based assay); RT-PCR and microarray, FFPE and fresh (distributed) | Biology (identification of major molecular subtypes) | N/A research assay | No | NCIC-MA5 [26], NCIC-MA12 [27], GEICAM/9906 [28], multiple nonrandomized trial cohorts | NCIC-MA12 [27] (luminal subtype predicts benefit) | NOAH [29] (HER2-enriched benefits the most) | No | NCIC-MA5 [26] (CMF vs. CEF; epirubicin benefit in HER2-enriched subtype only) | GEICAM/9906, CALGB/9342 and CALGB/9840 [30] (low proliferation score predicts weekly paclitaxel benefit) | No |
Prosigna; nCounter, FFPE (distributed) | Biology (subtype); outcome (ROR score) | CE Mark, Health Canada, FDA: prediction of 10-year DRFS in ER+, node 0 to 3, postmenopausal women treated with endocrine therapy | Yes [31] | No | No | No | No | No | RxPONDER (one to three nodes, recruiting; embedded additional analysis) | |