Cytokeratin 19 immunohistochemistry and assessment of depth of myometrial invasion in stage I endometrial cancer

F Alexander-sefre, MB CHB, MRCOG, FRCS, N Singh MD, MRCPath, H Salvesen MD, I Jacobs MD, MRCOG

Institue of Cancer Research, St Bartholomew's Medical school, london, london United Kingdom

Background: Approximately 50% of endometrial cancer deaths occur in patients who initially presented with stage I. In stage I disease depth of myometrial invasion is the most important risk factor for lymph node metastasis and mortality. In addition lymphovascular space (LVSI) invasion has been shown to have poor prognostic value in stage I EC. Assessment of depth of invasion and LVSI are currently done by light microscopic examination of H&E stained sections. Tumour cells can evade detection by light microscopy if they are isolated and vastly outnumbered by myometrial, inflammatory or normal circulating cells. In non-gynaecological cancer patients, immunohistochemical staining of tumour cells has been proven to be of value in detecting isolated infiltrating tumour cells at surgical margins. It has also been shown to correlate well with tumour recurrence. Cytokeratin 19 (CK19) belongs to a family of 20 related polypeptides that are constituents of the intermediate filaments of epithelial cells. Of these CK19 is the most frequently expressed cytokeratin endometrial cancer. Objective: To investigate the role of CK-19 immunohistochemistry (IHC) in assessing depth of myometrial invasion in stage I EC and to determine its clinical importance. Methods: A total of 111 stage I EC patients with minimum 4 years follow-up were selected. Representative H&E stained sections were re-examined by a pathologist to identify blocks containing tumour with maximum myometrial invasion. Four micron sections from the selected blocks were stained for CK-19 using a standard avidin-biotin method. The immunostained sections were then examined by the same pathologist and the results compared with H&E findings. Cases suspected of LVSI were dual stained for CD31 vascular endothelial marker and pancytokeratin to confirm true vascular space invasion as opposed to retraction artefact. Results: Endometrial tumour cells showed positive though variable immunoreactivity for CK-19. Immunohistochemical assessment of 111 cases revealed deeper myometrial invasion than detected on H&E staining in 7 cases (6%). In 5 of these 7 cases isolated tumour cells surrounded by inflammatory cells were noted 0.2 to 1.2mm deeper within the myometrium than detected by H&E staining. In the remaining 2 cases the deeper extension seen was the result of examining serial levels through the tumour block; in these cases deeper infiltration would have been apparent on H&E sections. The surgical staging remained unaltered in all 7 cases. Thirty four cases (30%) were suspected to have LVSI on CK19 IHC that were not detected on H&E sections. Dual staining has so far been conducted in 23 of these cases and has confirmed the presence of tumour cells within lymphovascular space in 21 cases. There were 3 cases in which LVSI was suspected on H&E only and these are awaiting confirmation on dual staining. Conclusions: CK-19 IHC can detect isolated infiltrating tumour cells. It therefore has potential advantage over routine histology in accurate assessment of myometrial invasion in stage I EC. CK-19 IHC identifies lymphovascular invasion with much greater sensitivity than H&E staining. We aim to correlate the presence of isolated infiltrating cells and LVI identified solely by CK-19 IHC with known adverse clinico-pathological features, disease recurrence and overall survival.

KEY WORDS: Lymphovascular Invasion, Vascular invasion.

For more information, contact

Paper presented at the International Symposium on Predictive Oncology and Intervention Strategies; Paris, France; February 9 - 12, 2002; in the section on Diagnostic Markers.