Search

The aim of the present study was to test the polymerase chain reaction (PCR) as a tool to identify human papillomavirus (HPV) in routine cytological samples scraped from the uterine cervix. Moreover, attention has been focused on the correlation between HPV types and early intraepithelial lesions. The study involved 586 women who had undergone conventional Pap test. Analysis of HPV infection was performed by PCR and HPV typing by dot blot. In a group of 78 cases histologically diagnosed as high-grade squamous intraepithelial lesions (HSILs), the cytological diagnosis was correct in 92.3% and the HPV test was positive in 89.8% of cases; combined positivity at Pap and/or HPV tests raised this figure to 99.0%. In a group of 67 cases histologically diagnosed as low-grade squamous intraepithelial lesions (LSILs), the cytological diagnosis was correct in 73.1% and the PCR-based HPV test was positive in 64.2%; combined positivity at Pap and/or HPV tests raised this figure to 91.0%. This study confirms the limitations of screening programs based on Pap test only. Our results suggest, in fact, that adding the HPV test to primary screening could increase the yield of preinvasive cervical lesions.

Histological detection of axillary lymph node metastases is still the most valuable prognostic parameter for breast cancer, but about 30% of node-negative patients relapse within five years, suggesting that current methods are inadequate for identifying metastatic disease. More sensitive, PCR-based methods for the detection of metastatic cells are now available, enabling the amplification of cancer cell-specific mRNA messages by the RT-PCR assay. An ideal tumour marker, consistently expressed in tumour samples and not at all in normal lymph nodes, remains to be identified. The present study first investigated the expression of seven mRNA markers, CEA, CK19, c-Met, mammaglobin, MUC-1, β1→GalNAc-T and p97, selected on the basis of their previously reported specificity for breast cancer cells. Eighteen lymph nodes were examined from patients without tumours. Only mammaglobin mRNA and CEA mRNA were not expressed in normal nodes. All of the other markers showed a band of expression in 17%-55% of cases, indicating that they are not breast cancer-specific. CEA mRNA and mammaglobin mRNA expression could be detected in 15/20 (75%) and 19/20 (95%) primary breast carcinomas, respectively. The expression of mammaglobin mRNA and CEA mRNA was then compared in axillary lymph nodes from 248 consecutive breast cancer patients, 89 with histologically documented lymph node metastasis and 159 without histological evidence of metastatic disease. Ninety-seven per cent of the patients with histologically involved nodes showed expression of mammaglobin mRNA, whereas CEA mRNA was expressed in 79% of these cases. In the group of patients with histologically negative lymph nodes, 46 (29%) and 32 (20%) were found to be positive for mammaglobin and CEA expression, respectively, indicating the presence of metastases not detected by routine histological examination of one lymph node section. These results show that both mammaglobin RT-PCR and CEA RT-PCR are useful tools for the detection of breast cancer metastases in axillary lymph nodes. The detection sensitivity of the mammaglobin RT-PCR is far superior to that of the CEA RT-PCR, allowing the diagnosis of occult metastases in nearly one-third of cases. Copyright © 2001 John Wiley & Sons, Ltd.