Tumour suppressor p21 WAF1/CIP1 is the main downstream effector gene mediating p53-induced cell cycle arrest, up-regulated by normal wild-type p53 (El-Deiry et al, 1993). In addition to direct transcriptional induction by the tumour suppressor p53, various other signals can induce p21 expression in the absence of wildtype p53 (Michieli et al, 1994;Zeng and El-Deiry, 1996). Harper et al (1993) showed that the ability of the p21 gene product to arrest the cell cycle is based on its virtue to bind and inhibit active cyclin/CDK complexes needed in DNA replication and cell cycle progression. p21 can also inhibit DNA replication by directly binding to the proliferating cell nuclear antigen (PCNA) molecule, thus inhibiting the ability of DNA polymerase δ to extend new DNA chains but still allowing the DNA repair function to continue (Flores-Rozas et al, 1994;Waga et al, 1994;Podust et al, 1995).In human melanoma cell lines, induction of p21 is independent of wild-type p53 expression (Jiang et al, 1995;Vidal et al, 1995), and elevated expression of p21 is associated with melanoma differentiation, growth arrest and metastatic suppression (Jiang et al, 1995). However, the knowledge of p21 expression in clinical materials of cutaneous malignant melanoma is rather limited. Maelandsmo et al (1996) revealed a significant correlation between elevated p21 expression and increasing tumour thickness in superficially spreading melanoma, but no correlation between p21 and survival was found. Elsewhere, a significant overexpression of p21 has been demonstrated in primary and metastatic melanomas (Trotter et al, 1997).As far as the authors are aware, no previous study has compared the relationships between p21, p53 and PCNA expression in primary stage I cutaneous melanoma. In the present study, we used immunohistochemistry to analyse the above-mentioned relationships in a cohort of 369 patients with long-term follow-up data. In addition, our aim was to analyse whether p21 protein levels are related to clinical data, histological parameters (Clark and Breslow levels) and prognosis.
MATERIALS AND METHODS
PatientsThe retrospective study consists of a consecutive series of 369 clinical stage I cutaneous malignant melanoma patients with complete clinical and histopathological data available, who were diagnosed and treated in the district of Kuopio University Hospital, Finland, between 1974 and 1989. The clinical staging of all tumours was carried out according to UICC (UICC, 1987). Because of insufficient tumour material, pigment or technical artefacts, there were 267 valid immunostainings for p21, 284 for p53 and 219 for PCNA respectively. Patient records were reviewed and the pertinent clinical and histopathological data of the patients are shown in Table 1. The mean follow-up time of all 369 patients was 6.4 years (range 0.2-18 years). When the analysis was restricted to patients with p21, p53 or PCNA data, the mean follow-up times were 6.3 years (range 0.5-18 years), 6.3 years (range 0.5-18 years) or 7.2 years (range 0.5-18 years) re...