The prognostic significance of c-erbB-2 expression was studied in paraffin wax embedded colorectal cancer tissue using a monoclonal antibody. One hundred and sixty-four patients with Dukes' B disease were studied. Membranous staining was not detected in any case. Cytoplasmic c-erbB-2 staining was seen in 55 cancers (33.5%). Cytoplasmic staining was unrelated to patient age (P = 0.31), sex (P = 0.69), tumour site (P = 0.69), size (P = 0.57), histological grade (P = 0.42) or ploidy status (P = 0.21) but was found more frequently in obstructing cancers (P = 0.03). Mean follow up of the patient population was 6.3 years. Five-year-survival estimated by the Kaplan-Meier life-table method was 47% for those with cytoplasmic c-erbB-2 staining and 77% for those without (log rank analysis; P < or = 0.0001). Stepwise regression analysis identified c-erbB-2 staining (relative risk, 2.51; P = 0.0005) and bowel obstruction (relative risk, 1.99; P = 0.015) as independent predictors of survival. It is suggested that cytoplasmic c-erbB-2 expression may provide a useful marker of tumour behaviour in Dukes' B colorectal cancer.

Ninety-four H & E-stained slides of malignant melanoma were circulated to 6 pathologists in 2 university departments. For each slide, the growth phase of the lesion, Breslow thickness, and Clark level were determined by each observer. The aims of the study were to evaluate agreement between nonspecialist pathologists in identifying the vertical growth phase in malignant melanoma and to compare agreement for the growth phase with agreement for Breslow thickness and the Clark level. Our results show that although overall agreement for the growth phase is moderate, agreement between experienced observers is good. In fact agreement for the growth phase among this group was equal to the agreement for Breslow thickness. Overall agreement for Breslow thickness also was good but for the Clark level was only fair. These findings suggest that if the predictive value of the vertical growth phase proves to be robust, it will be used with an acceptable level of accuracy in routine diagnostic practice.

Aims-To compare differences in cell proliferation indexes and apoptotic indexes between cutaneous basal and squamous cell carcinomas, in an attempt to suggest an explanation for the differences in their biological behaviour. Methods-Forty cases of cutaneous basal cell carcinoma (BCC) and 40 cases of moderately and well differentiated squamous cell carcinoma (SCC) were retrieved from the archives. Sections, 4 gm thick, were cut from formalin fixed, paraffin wax embedded tissue in each case and stained with haematoxylin and eosin. These were then examined for mitotic and apoptotic figures per 1000 cells. Sections from the same cases were also immunostained with the mouse monoclonal antibody Ki67 (MIBI); positive nuclear staining was counted per 1000 cells. Results-No significant differences were found between the mitotic indexes and apoptotic indexes in these tumours. There was, however, a significant difference in Ki67 (MIBI) staining, with greater staining in the squamous cell carcinomas. Conclusion-Estimation of the mitotic and apoptotic indexes did not reveal any differences between these two tumour types. The proliferation indexes, assessed by Ki67 immunostaining, did differ. This may be one of the factors underlying the more aggressive behaviour of SCC. ( Clin Pathol 1996;49:549-551) Keywords: skin, neoplasm, proliferation. Department of

No abstract

Hepatomegaly was noted in a 63-year-old man who presented with an exacerbation of chronic pulmonary disease. A diagnosis of hepatic leiomyosarcoma was made by fine needle aspiration biopsy. Intensive investigations failed to reveal a primary source. The patient was treated conservatively.

The fragile histidine triad (FHIT) gene encompasses the common chromosomal fragile site FRA3B. Human papilloma virus (HPV), which is the main aetiological agent in cervical cancers, has been found to be able to integrate its genes into the chromosome 3 fragile site of cultured cells, deleting a piece of DNA which includes the FHIT gene. Eighty-six microdissected archival cervical LLETZ biopsies comprising cases of cervical intraepithelial neoplasia (CIN) 1 (n=27), CIN3 (n=30) and microinvasive carcinoma (n=29) were evaluated for HPV infection and FHIT gene loss of heterozygosity (LOH). FHIT gene LOH was detected by polymerase chain reaction (PCR) using fluorescently labelled intragenic microsatellite markers D3S1300 and D3S4103. PCR products were analysed on a semi-automated DNA sequencer using Fragment Manager(trade mark) software to determine allele loss. The HPV status of the lesions was determined by PCR using generic and type-specific primers in conjunction with restriction endonuclease digestion. The results were analysed using Epi-Info and SPSS-PC statistical analysis software. Haematoxylin and eosin-stained sections from the 86 cases were profiled for six histopathological features, some of which have been previously shown to be associated with microinvasive cancer. FHIT gene LOH was found in 36% of CIN1 cases, 52% of CIN3 cases and 73% of microinvasive cases (p=0.029). HPV 16 DNA was found in 68% of CIN3 cases and 93% of microinvasive cases (p<0.001). The second most prevalent HPV type found was HPV 31, which was present in only four lesions, three of which had FHIT gene LOH. When FHIT gene LOH was evaluated versus HPV 16 and 31 infection using the chi-square test, a statistically significant relationship was found (p=0.014). FHIT gene LOH was found to be independent of the histopathological features evaluated. The finding of a statistically significant relationship between FHIT gene LOH and oncogenic HPV infection suggests a link between the integration of viral DNA and subsequent gene deletion in the progression of cervical cancer. FHIT gene anomalies may prove to be excellent markers of progression in early uterine cervical cancers.

This study evaluates the feasibility of DNA analysis of cervical intraepithelial neoplasia III (CIN III) lesions on cervical smear and formalin-fixed paraffin-embedded tissue (FFPET) blocks with a view to extending this type of analysis to milder grades of dyskaryosis. DNA ploidy was determined by image analysis using a CAS 200 Image Analyser. Seventeen patients with a diagnosis of CIN III were studied. Results show that all smear and tissue samples were non-diploid with nine aneuploid and eight tetraploid lesions. In 6/7 patients whose smears and corresponding biopsies were examined there was complete agreement as to the DNA profile. We conclude that DNA quantification is technically feasible in archival, routinely prepared cervical smears. This technique should now be applied to CINI and CINII cervical smears to determine if it is of value in identifying those lesions that will progress to CIN III. This study is particularly timely with the possibility in the near future of estimation of ploidy by image analysis using instruments such as the Highly Optimized Microscope Environment (HOME) system.

No abstract