Dear Editor, We read with interest the article by Vilardell et al. [1] where lymph node metastasis was missed in a patient with a cytokeratin-19 (CK19)-negative breast cancer. Sentinel node biopsy (SNB) is currently the standard practice for staging the axilla [2], with one-step nucleic acid amplification (OSNA) becoming more widely accepted [3] and used in clinical practice. OSNA detects metastasis by quantifying the presence of CK19 mRNA, in theory irrespectively whether this is expressed or not. There are limited data regarding the true incidence of CK19-negative tumours. The authors estimate this to be about 3 %, which is probably similar to anecdotal data in our practice. Our institution was second in the UK to adopt OSNA analysis routinely in its practice. Since May 2010, nearly 600 such procedures have been performed. However, in contrast with Vilardell, we identified a patient who, despite both the core biopsy and the subsequent breast excision specimen testing negative for CK19 expression, was found to have SNB micrometastasis on OSNA analysis. The tumour biopsy confirmed an oestrogen and progesterone receptor-positive ductal carcinoma (ER8/8, PR 6/8), with negative HER2 immunohistochemistry. Retrospective Ki67 proliferative index was found at 10 %, confirming the tumour’s luminal A immunophenotype. CK19 was negative despite repeated testing in the presence of positive controls. Intraoperative full-node OSNA analysis of two sentinel nodes showed one negative lymph node and one with micrometastasis (820 copies of CK19 mRNA). As per hospital practice at the time, the patient proceeded to completion axillary lymph node dissection. Final histology confirmed a 10-mm grade 2 invasive ductal carcinoma, with few foci of ductal carcinoma in situ. Ten further lymph nodes were identified, with no evidence of metastatic disease. Testing for CK19 expression on the excision specimen was negative. Despite being used for a few years now, the OSNAmethod is still relatively new, and updated figures are published regularly. The discrepancy between the OSNA and the imprint cytology results in the case of Vilardell could also be due to tissue allocation bias or due to the false-negative rate of the OSNA method. The latter however is limited to less than 2 % [3]. Moreover, the size of the “malignant cells” on imprint cytology has not been stated suggesting that overinterpretation has not been discounted as a possibility. Our patient’s result may only be representing detection of extremely low-volume disease. It may however be falsely positive. Again, the OSNA false-positivity rate is similarly low, but, importantly, true rates are very difficult to assess. Indeed, as the authors suggest, more documentation is needed to assess the expression of CK19 in breast cancer and its implication on the OSNA results. However, a study to look at the presence or absence of CK19 mRNA, rather than its expression, as detected by immunohistochemistry, may be more appropriate. We suspect that CK19 mRNAnegative tumours will be a small fraction of the CK19negative tumours, which are already rare, and therefore may make routine CK19 testing prior to OSNA processing unnecessary. D. D. Remoundos (*) :M. Joshi : F. Ahmed :G. H. Cunnick Department of Breast Surgery, Wycombe Hospital, Buckinghamshire Healthcare NHS Trust, Queen Alexandra Road, High Wycombe HP11 2TT, UK e-mail: dr223@cantab.net
Read full abstract