The utility of a real-time RT-PCR assay for the detection of metastases greater than 0.2 mm in sentinel lymph nodes of breast cancer patients confirmed by detailed histological analysis
628 Background: Analysis using real time RT-PCR for the detection of metastases in lymph nodes (LN) increases sampling but is associated with a risk of too much sensitivity as compared to histological analysis. Determining the appropriate cutoff value is important in introducing this system to routine clinical practice. Our study confirms the reliability of the cutoff values of a real-time RT-PCR assay (GeneSearch, Veridex LLC) to detect metastases larger than 0.2 mm by detailed 0.2 mm frozen section histological diagnosis. Methods: 129 sentinel and non-sentinel lymph nodes were obtained by sentinel LN biopsy from 79 breast cancer patients. All LNs were cut in half. One half of each LN was used for routine intra-operative diagnosis. The other LN half had sections cut every 0.2 mm for histological analysis. All frozen sections were H&E stained and examined by on-site pathologists. These sections were then re-stained immunohistochemically using a pancytokeratin antibody (AE1/AE3) for detecting submicrometastses. All remaining tissue between the histology sections was assayed by the real-time RT-PCR assay using the genetic markers mammaglobin (MG) and cytokeratin-19 (CK-19) (note: tissue processing method was investigational and off-label). Cutoff values were pre-set in a large US study (n = 304). Results: Compared to the histological diagnosis using 0.2 mm interval frozen sections, the real-time RT-PCR results were as follows; sensitivity of 100.0% (34/34), specificity of 93.7% (89/95), and overall accuracy of 95.3% (123/129). Submicrometastases were recognized by IHC in 3 of 95 samples (2 solitary and 1 multiple sites) diagnosed as negative by H&E-stain sections, and 3 samples were negative for RT-PCR. Conclusions: Results of 0.2 mm interval histological analysis suggest a near perfect sensitivity of the real time RT-PCR assay, allowing reliable detection of LN metastases larger than 0.2 mm. In addition, at least three samples with submicrometastases detected by immunohistochemistry for pancytokeratin were above the preset cutoff values of this real time RT-PCR assay. [Table: see text]