The prevalence and clinical significance of chromogranin A and secretogranin II immunoreactivity in colorectal adenocarcinomas

Colorectal adenocarcinomas may display features of endocrine differentiation, shown by argyrophil stains and by the expression of endocrine markers such as chromogranin A. We investigated chromogranin A and secretogranin II immunoreactivity in a series of 208 carcinomas of the large bowel to assess the prevalence and clinical significance of endocrine differentiation. Tumours expressing endocrine markers were classified as low expressors (< than 1 immunoreactive tumour cell/mm²) and high expressors (> than 1 immunoreactive tumour cell/mm²). There were 33 (16%) carcinomas showing both chromogranin A and secretogranin II immunoreactivity: 11 tumours (5%) were high expressors. Endocrine differentiation was not related to the disease stage, tumour location, grade, DNA ploidy and p53 protein accumulation. In the entire series chromogranin A immunoreactivity did not provide prognostic information using univariate and multivariate analysis. A worse overall survival (P=0.048) was demonstrated for the stage III patients with high expressor tumours, but there were only five patients in this group. The results of our investigation suggest that chromogranin A immunoreactivity is not a useful variable in the prognostic assessment of colorectal adenocarcinomas.     

Proliferating cell nuclear antigen in gallbladder carcinoma

Proliferating cell nuclear antigen (PCNA) expression was studied in 103 gallbladder carcinomas and 23 metastatic lesions as well as in 25 control non-neoplastic gallbladder specimens. Positive nuclear staining was observed in 88% of controls, in 92% of carcinomas and in 70% of metastases. The mean number of positive cells was 21.2% in controls, 44.1% in primary carcinomas and 32% in metastatic cancer cells. Differences which were significant were control v . primary tumour, P <0.000001; control v . metastasis, P <0.01 and primary tumour v . metastasis, P <0.006. In 57 (60%) of the primary tumours there was positive staining in over 40% of tumour cells. We were not able to demonstrate any relationship between macroscopic or microscopic features and PCNA expression. However, tumours confined to the mucosa expressed PCNA more frequently than did more advanced tumours.     

Abnormalp53 immunoreactivity and prognosis in node-negative breast carcinomas with long-term follow-up

Summary The expression of thep53 gene product was investigated immunocytochemically in a retrospective series of 164 formalin-fixed paraffin-embedded invasive breast carcinomas with pathologically proven negative lymph nodes. Overall, 78 tumors (48%) showed a variable degree ofp53 immunoreactivity. Among these, 38 cases were low expressors (1–10%p53 immunoreactive tumor cells), 21 moderate expressors (10–50% immunoreactive cells) and 19 high expressors (> 50% immunoreactive cells). Abnormalp53 expression correlated significantly with tumor size, histological and nuclear grade, DNA ploidy, mitotic rate and proliferation index, and with the lack of estrogen receptors. Disease-free and adjusted survival analysis of the 124 node-negative patients with long term (more than 10 years) follow-up, however, did not reveal an independent prognostic role for p53 expression. These data suggest that the evaluation ofp53 immunoreactivity may only play a role in a multiparametric prognostic assessment of node-negative breast carcinoma.     

Proliferating cell nuclear antigen in breast carcinomas

Proliferating cell nuclear antigen (PCNA), was examined by immunohistochemistry in 509 breast carcinomas. The immunoreactivity was found to be independent of the length of fixation when the tissue sections were microwaved before incubation with the primary antibody. The PCNA immunoreactivity was assessed by two semi-quantitative methods, which were correlated but not exchangeable. The comedo type of intraductal carcinomas and invasive ductal carcinomas had a higher PCNA score than other types. Lymph node metastases had a significantly higher PCNA score than primary carcinomas. High PCNA immunoreactivity was correlated with the presence of lymph node metastases, absence of tubule formation, numerous mitoses, severe nuclear pleomorphism, high histological grade and absence of progesterone receptors (PgR). PCNA in lymph node positive tumours was correlated with tumour type, especially with ductal carcinomas, absence of tubule formation, high histological grade and absence of PgR, whereas PCNA in lymph node negative tumours was correlated with large tumour size, numerous mitoses, severe nuclear pleomorphism and high histological grade. Number of mitoses and nuclear pleomorphism were the two most important factors in predicting the PCNA score; the absence of PgR and nuclear pleomorphism were important in lymph node negative and positive tumours, respectively. In a univariate analysis high PCNA score was found to be correlated with shorter relapse-free period and poorer over-all survival.     

Co-expression of endothelial cell and macrophage antigens in Kaposi's sarcoma cells

Mutations in the p53 tumour suppressor gene, with consequent accumulation of the p53 protein, are frequently observed in non-small cell lung cancer (NSCLC). Little is known, however, about the timing of their appearance or their maintenance through cancer progression and metastatic spread. We have examined the normal epithelium and a panel of bronchial lesions, including dysplastic, neoplastic, and metastatic leisons, for p53 immunoreactivity and for expression of proliferating cell nuclear antigen (PCNA). No p53 immunoreactivity was found in normal and hyperplastic epithelium, nor in squamous metaplastic lesions. Twenty out of 30 invasive tumours and 13 out of 17 in situ carcinomas adjacent to an invasive tumour showed p53 immunoreactivity. There was a strict correlation between the level of p53 expression in the non-invasive and the invasive components of the tumours. Five out of eight pairs of primary tumours and matching metastases expressed p53, at identical levels in both compartments. These data indicate that p53 overexpression can occur in the earliest recognized phase of NSCLC and that the alteration is maintained during progression from in situ to invasive carcinoma and metastatic spread. PCNA expression increased from early to advanced phases of NSCLC. High PCNA immunoreactivity was observed in tumours expressing high p53 levels. A significant association was observed for PCNA expression between preinvasive and invasive lesions.     

Expression of E-cadherin (E-CD) as related to other prognostic factors and survival in breast cancer

A series of 208 breast cancer biopsies were analysed immunohistochemically for expression of E-cadherin (E-CD). Altogether, 72 per cent of the tumours showed E-CD positivity in over 50 per cent of cells, the staining being heterogeneous in nearly all tumours. In only 16 per cent of ductal carcinomas was positive staining seen in less than 1 per cent of cells. Expression of E-CD was not related to tumour diameter, nodal status, metastasis at diagnosis, histological grade, DNA ploidy, S-phase fraction, nuclear area, mitotic frequency, or PR content. There was a significant relationship between expression of E-CD, histological type (P=0.01), the proportion of intraductal growth (P=0.008), the density of tumour-infiltrating lymphocytes (P=0.0007), ER content (P=0.012), and morphometric nuclear factors (P=0.02). Expression of E-CD showed a weak association with a high survival probability (P=0.02), while the relation to recurrence-free survival was not significant (P=0.11). In axillary lymph node-negative tumours, E-CD expression was not related significantly to survival (P=0.11) or to recurrence-free survival (P=0.06). In multivariate analysis, E-CD expression had no independent prognostic value, while the axillary lymph node status, tumour diameter, patient age, and mitotic frequency were independent prognostic factors. The results indicate that E-CD expression is related to several histological features in breast cancer, but has no independent prognostic value over standard prognostic factors.     

Immunohistochemical and molecular analysis of p53, MDM2, proliferating cell nuclear antigen and Ki67 in benign and malignant peripheral nerve sheath tumours

A series of 26 malignant peripheral nerve sheath tumours (MPNST) and 24 benign peripheral nerve sheath tumours (BPNST) were analysed immunocytochemically for p53 expression and the cell proliferation markers proliferating cell nuclear antigen (PCNA) and Ki67 (with MIB1). In 23/26 MPNST, 5%–65% of the tumour cell nuclei were immunoreactive for Ki67 with MIB1 while none of the 24 BPNST had nuclear staining exceeding 5%. Greater than 50% nuclear PCNA staining was detected in 25/26 MPNST compared with 8/24 BPNST; 17/26 MPNST showed 5–100% nuclear staining for p53 (13/26>20%), whereas none of the BPNST had nuclear staining exceeding 1%. The Ki67, PCNA and p53 immunostaining results correlated significantly with benign versus malignant (P<0.001, P<0.001 and P<0.005, respectively) as well as mitotic rate (P<0.001, P<0.05 and P<0.05). Ki67 immunostaining results correlated significantly with PCNA and p53, as did p53 with Ki67 and PCNA (P<0.001 in both). Stepwise (logistic regression forward) multivariate analysis of the variable, benign versus malignant, revealed the strongest correlations with PCNA (P=0.007) and Ki67 (P=0.021). Direct confirmation of the presence of p53 protein was obtained by western blot analysis of 3 MPNST and 5 BPNST. Two MPNST, showing 90% and 30% immunoreactivity, were positive for p53, while one MPNST with 5% immunoreactivity and all 5 BPNST were negative. Southern blot analysis performed on the two MPNST with high p53 protein levels revealed no amplification of the MDM2 gene, suggesting that high p53 levels in MPNST are likely to be due to mutation. The results also indicate that PCNA and Ki67 are potentially useful in distinguishing BPNST from MPNST, particularly in problematic cases of cellular schwannoma versus MPNST. The detection of p53 in a large percentage of cells of a plexiform neurofibroma giving rise to MPNST and Ki67 in 5% and 25% of cells of two similar cases suggests that malignant transformation may be detected in some cases by p53 and proliferation markers prior to overt histological evidence of malignancy.     

Ki-67 expression in primary breast carcinomas and their axillary lymph node metastases: clinical implications

Proliferative activity of tumour cells assessed by immunohistochemical Ki-67 expression is one of several prognostic indicators in breast cancer. The major objective of this study was to investigate the prognostic impact of Ki-67 proliferative activity in the axillary lymph node metastases and in the matched primary breast carcinoma from 194 patients. There was a statistically significant up-regulation of Ki-67 protein in the metastatic deposit compared to where the primary tumour was found (p = 0.001). A low Ki-67 index in both the primary and the metastatic tumours was a favorable prognostic factor. A high index in both primary and metastatic lesion and an up-regulation from a low index in the primary tumour to a high index in the metastatic deposit represented an unfavorable prognostic factor. Multivariate analysis showed that Ki-67 expression in the metastases was a superior independent prognostic factor of clinical outcomes compared to that in the primary tumours. Ki-67 expression in ≥10% of carcinoma cells in the primary tumours and ≥15% in the nodal metastases seems to be optimal cut-off levels. Ki-67 is of value as an independent prognostic factor in breast cancer.     

Clinicopathological value of somatostatin type 2A and estrogen receptor immunoreactivity in human breast carcinoma

Somatostatin type 2A (sstr2A) and estrogen receptor (ER) are interrelated regulatory receptors present in normal breast epithelium and in a population of breast carcinomas. ER mediates growth stimulatory effects of estrogens whereas sstr2A mediates growth inhibitory actions of somatostatin. However, much work has been devoted to elucidate the biological role of ER, little is known about sstr2A in breast cancer. In the present study we examined immunoreactivity of sstr2A and ER in 64 breast carcinomas in correlation with tumor size and histological grade (HG), presence of lymph node metastasis (LNM), Nottingham prognostic index (NPI), and the patients' age. ER and sstr2A immunoreactivity were present in 78% and 63% of the breast carcinomas, respectively. Ninety percent of tumors immunoreactive for sstr2A were simultaneously immunoreactive for ER. ER immunoreactivity correlated significantly with lower HG (p=0.03) and better NPI (p=0.02). sstr2 A immunoreactivity correlated significantly with lower HG (p=0.012) but not with NPI (p=0.26). There was no correlation of sstr2A immunoreactivity and tumor size, patients' chronological age or LNM. The results confirm prognostic value of ER immunohistochemistry in breast carcinoma. However sstr2A cannot substitute ER for prognostic evaluation, sstr2A immunoreactivity being significantly associated with lower HG seems to represent an independent prognostic factor in breast carcinoma.     

Preliminary results of prognostic significance of proliferating cell nuclear antigen expression in advanced primary larynx carcinomas and lymph node metastases

Introduction

The aim of this study was to investigate the prognostic significance of proliferating cell nuclear antigen (PCNA) expression in laryngeal carcinoma in relation to clinicopathological features. Special emphasis was placed on examining the relationship of PCNA expression in the primary tumour and PCNA expression in corresponding lymph node metastases obtained from the same patients.

Material and methods

The study included 60 patients with advanced larynx carcinoma who had received treatment and follow-up for at least 5 years. Sixty laryngeal carcinoma specimens and metastatic lymph nodes from 24 patients were examined for immunohistochemical PCNA expression.

Results

The percentages of PCNA positive cells were significantly higher in the primary tumours which developed lymph node metastases than in those without metastases. The fraction of PCNA immunolabelled cells in metastatic lymph nodes increased significantly when compared with the PCNA positive cell score in their corresponding primary tumours obtained from the same patient. There was a significant difference in PCNA index score in primary tumours between the group of patients who survived a 5-year period and those who died within 5 years after treatment.

Conclusions

Our data demonstrate that a high proliferation index in primary larynx tumours is retained and increased in corresponding lymph node metastases. Measurement of the fraction of cancer cells stained for PCNA in primary larynx carcinomas can be helpful in selecting tumours with high aggressiveness potential that are more likely to develop neck metastases and thereby in identifying patients who need elective lymph node dissection or additional treatment.     

Prognostic value of proliferating cell nuclear antigen in gastric carcinoma.

A new monoclonal antibody to proliferating cell nuclear antigen (PCNA), PC10, which can be used on routinely processed tissue, was applied to 93 cases of gastric carcinoma. Significant intra-tumoural variation in staining occurred. In addition to a PCNA index (percentage of positive cells per 1000 tumour cells), a semiquantitative PCNA grading system was devised, based on estimates of less than or more than 50% of positive tumour cells in whole sections. Neither PCNA index nor PCNA grade showed any correlation with established histological variables, tumour stage, or the presence of lymph node metastases. No significant correlation was observed between PCNA index and S + G2M phase fraction measured by flow cytometric analysis. To analyse survival tumours with PCNA indices above and below the median level (41%) were compared. Those with a higher index tended to have a worse prognosis, but when PCNA grade was considered, it was found to have definite independent prognostic value, tumours of low grade surviving better than those of high grade. The ability of semiquantitative PCNA grading to allow for intra-tumoural variation suggests it may have advantages over absolute counting, which is prone to sampling error when tumour heterogeneity is a major factor. The prognostic value of PC10 staining in gastric carcinoma is therefore promising.     

PROGNOSTIC INFLUENCE OF p53 EXPRESSION IN GASTRIC CANCER

The presence of the nuclear phosphoprotein p53 was investigated in a series of 120 consecutive gastric carcinomas. This immunohistochemical study on formalin-fixed, paraffin-embedded material found p53 expression in 43 per cent ( n =51) of carcinomas using a monoclonal antibody (DO-1), whereas no immunoreactivity for p53 was present in tumour-associated non-neoplastic gastric mucosa or tumour stroma. There was no statistically significant correlation with known prognostic parameters such as extent of tumour growth (pT state), nodal involvement (pN state), or tumour grade. The same applied for association with patient age and sex or pathological parameters such as tumour size, localization, or growth pattern according to histological classification. Kaplan–Meier analysis revealed marginal statistically significant differences in survival times between patients with p53-positive tumours with more than 35 per cent of p53-positive tumour cells and those with less than 35 per cent of p53-positive tumour cells or p53-negative tumours ( P =0·04). However, by multivariate analysis, p53 immunoreactivity did not turn out as an independent prognostic parameter. p53 expression can easily be detected in a variety of human malignancies including gastric cancer by immunohistochemical methods, but its prognostic significance and possible role as an independent marker of poor prognosis still have to be confirmed by further studies.     

Immunoreactivity of proliferating cell nuclear antigen compared with bromodeoxyuridine incorporation in normal and neoplastic rat tissue.

Monoclonal antibodies (MoAbs) against proliferating cell nuclear antigen (PCNA) represent a potentially useful tool for cell kinetic analysis of tumours. Because in paraffin-embedded tissue the relationship between PCNA immunoreactivity and tumour cell proliferation is not well characterized, we have compared PCNA positivity as detected by the PC10 MoAb with the bromodeoxyuridine labelling index (BrdUrd-LI) in two different transplantable hormone-dependent rat mammary tumours. Together, these two tumour models (MCR-83 and EMR-86) cover a wide range of S-phase fractions. Evaluating 31 methacarn-fixed tumours, a strong but non-linear relationship (r = 0.98) was obtained. PCNA-positive fractions were invariably higher than corresponding BrdUrd-LIs and also higher than the estimated growth faction: growth fractions as determined by continuous BrdUrd labelling of the tumour and stromal cell population in EMR-86 carcinomas were 12 and 26 per cent lower than PCNA-positive fractions, implying that a certain fraction of non-cycling cells can also express PCNA. A dramatic disturbance in the relationship of PCNA positivity and the BrdUrd-LI was observed in the EMR-86 model after growth arrest induced by hormonal ablation: PCNA immunoreactivity remained detectable for at least 3 days, whereas the BrdUrd-LI decreased almost immediately. In comparison, PCNA immunoreactivity persisted for a much shorter period in small intestinal cells that had stopped DNA replication when moving from the crypt towards the villus. It is concluded that although differences in PCNA expression exist between various tissues, PCNA as detected by the PC10 MoAb may be used in tumours as an operational marker for the growth fraction.(ABSTRACT TRUNCATED AT 250 WORDS)     

Delay in fixation does not affect the immunoreactivity of proliferating cell nuclear antigen (PCNA).

The effect of delayed fixation on the immunoreactivity of proliferating cell nuclear antigen (PCNA) was investigated using eight breast carcinomas. Topologically shuffled samples of each tumour were immersed in fixative at times of 0.5, 1, 2, 4, 6, 18, and 24 h after surgical removal. In addition to a PCNA index (percentage of positive cells per 1200 tumour cells), a semi-quantitative PCNA grading system was used, based on estimates of more than or less than 50 per cent of positive tumour cells at each time interval. The PCNA index of six tumours increased by a mean of 10 per cent with a fixation delay of 24 h. The PCNA grade of all eight tumours showed no change with delayed fixation.     

Expression of nuclear insulin receptor substrate 1 in breast cancer

BACKGROUND: Insulin receptor substrate 1 (IRS-1), a cytoplasmic protein transmitting signals from the insulin and insulin-like growth factor 1 receptors, has been implicated in breast cancer. Previously, it was reported that IRS-1 can be translocated to the nucleus and modulate oestrogen receptor alpha (ERalpha) activity in vitro. However, the expression of nuclear IRS-1 in breast cancer biopsy specimens has never been examined. AIMS: To assess whether nuclear IRS-1 is present in breast cancer and non-cancer mammary epithelium, and whether it correlates with other markers, especially ERalpha. Parallel studies were carried out for the expression of cytoplasmatic IRS-1. METHODS: IRS-1 and ERalpha expression was assessed by immunohistochemical analysis. Data were evaluated using Pearson's correlation, linear regression and receiver operating characteristic analysis. RESULTS: Median nuclear IRS-1 expression was found to be low in normal mammary epithelial cells (1.6%) and high in benign tumours (20.5%), ductal grade 2 carcinoma (11.0%) and lobular carcinoma (approximately 30%). Median ERalpha expression in normal epithelium, benign tumours, ductal cancer grade 2 and 3, and lobular cancer grade 2 and 3 were 10.5, 20.5, 65.0, 0.0, 80 and 15%, respectively. Nuclear IRS-1 and ERalpha positively correlated in ductal cancer (p<0.001) and benign tumours (p<0.01), but were not associated in lobular cancer and normal mammary epithelium. In ductal carcinoma, both nuclear IRS-1 and ERalpha negatively correlated with tumour grade, size, mitotic index and lymph node involvement. Cytoplasmic IRS-1 was expressed in all specimens and positively correlated with ERalpha in ductal cancer. CONCLUSIONS: A positive association between nuclear IRS-1 and ERalpha is a characteristic for ductal breast cancer and marks a more differentiated, non-metastatic phenotype.     

Proliferating cell nuclear antigen (PCNA) immunolocalization in paraffin sections: an index of cell proliferation with evidence of deregulated expression in some neoplasms

Proliferating cell nuclear antigen (PCNA) is a 36 kD nuclear protein associated with the cell cycle. A monoclonal antibody, PC10, that recognizes a fixation and processing resistant epitope has been used to investigate its tissue distribution. Nuclear PCNA immunoreactivity is found in the proliferative compartment of normal tissues. PCNA immunoreactivity is induced in lectin stimulated peripheral blood mononuclear cells in parallel with bromodeoxyuridine incorporation and the number of cells with PCNA immunoreactivity is reduced by induction of differentiation in HL60 cells. In non-Hodgkin's lymphomas a linear relation between Ki67 and PCNA staining was demonstrated. These data suggest that in normal tissues and lymphoid neoplasms, PCNA immunolocalization can be used as an index of cell proliferation. However, in some forms of neoplasia, including breast and gastric cancer and in vitro cell lines, the simple relation between PCNA expression and cell proliferation is lost. In some breast and pancreatic tumours there is apparent deregulation of PCNA with increased expression in tissues adjacent to the tumours. The over-expression in some tumours and in adjacent morphologically normal tissue may represent autocrine or paracrine growth factor influence on PCNA gene expression.     

Expression of p53 protein has no independent prognostic value in breast cancer

A series of 392 female breast carcinomas was analysed immunohistochemically for expression of p53 protein with special emphasis on the role of p53 as an independent prognostic factor. Altogether, 54·8 per cent of the carcinomas expressed p53 protein, with the mean [standard error (SE)] fraction of positive nuclei being 17·1 per cent (1·2 per cent). Expression of p53 protein was independent of tumour metastasis at diagnosis, axillary lymph node status, tumour diameter, histological type, tubule formation, proportion of intraductal growth, margin formation, necrosis, DNA ploidy, and S-phase fraction. A high fraction of p53-positive nuclei was significantly related to patient age under 70 years, high grade, severe nuclear pleomorphism, dense infiltration of tumour by lymphocytes, high mitotic index, and high apoptotic index (for all, P<0·05). Impaired survival probability in the entire cohort (P=0·05) and in the axillary lymph node-positive (ANP) tumours (P=0·015) was associated with a fraction of p53-positive nuclei less than 25 per cent, while in the axillary lymph node-negative (ANN) tumours, expression of p53 had no prognostic value. In multivariate analysis, independent prognostic predictors included axillary lymph node status, tumour diameter, and mitotic index. In the ANN tumours, tumour diameter, fraction of p53-positive nuclei, and tumour grade were independent prognostic factors, whereas in the ANP tumours, diameter and mitotic index were the two independent prognostic factors. The results suggest that abnormal expression of p53 protein is only a weak independent prognostic factor in female breast cancer.     

The biological and prognostic significance of cell polarity and E-cadherin in grade I infiltrating ductal carcinoma of the breast

Grading of breast cancer based on the modified Scarff, Bloom, and Richardson system provides invaluable prognostic information. Recent evidence suggests that most tumours do not usually progress between grades and that groups of tumours within each grade are biologically distinct. This study has explored one potential aspect of biological tumour heterogeneity within grade by examining the relationship between cell polarity, the cell adhesion molecule E-cadherin, a major effector of cell polarity, and outcome, in 149 grade I infiltrating ductal breast carcinomas. Polarity was evaluated by studying the degree to which three features of polarized epithelial cells—nuclear ordering, basal positioning of nuclei within cells, and apical snouting/blebbing—were present in these tumours. E-cadherin expression was investigated using the antibody HECD-1. A low degree of tubule formation was correlated with poor nuclear ordering ( p< 0·01). The three histological features—nuclear ordering, basal nuclei, and apical blebbing—were all correlated with each other (all p< 0·0001). Polarity measurements did not correlate with survival. E-cadherin expression did not correlate with polarity and negative tumours were still able to form tubules. Surprisingly, strong E-cadherin immunostaining correlated with poor survival, tumour size, and nodal status. On univariate parametric (Weibull) survival models, high E-cadherin scores and tumour size were both significant predictors of survival in this group. Copyright © 1999 John Wiley & Sons, Ltd.     

Proliferation in malignant mesothelioma as determined by mitosis counts and immunoreactivity for proliferating cell nuclear antigen (PCNA)

In order to assess its discriminating and prognostic value, we studied immunoreactivity for proliferating nuclear cell antigen (PCNA) in human malignant mesothelioma (31 cases) and in human non-neoplastic mesothelium (33 cases with reactive mesothelium and 20 cases of normal mesothellum) using the murine monoclonal antibody PC 10. We also compared it with mitosis counts expressed as the mitotic volume index (MV index). There were differences between malignant mesothelioma, reactive mesothelium, and normal mesothelium for percentage of PCNA immunoreactive cells (mean ± SD; 27 ± 9, 9·5 ± 5·1, and 3·6 ± 1·6, respectively) and for their MV index (20·3 ± 4·5, 9·4 ± 2·1, and 3·6 ± 0·6, respectively). The median actuarial survival was 10·1 months for patients with less than 25 per cent PCNA immunoreactive cells, 9·4 months for patients with less than 20 mitoses per mm² of tumoural tissue, 5·9 months for patients with more than 25 per cent PCNA immunoreactive cells, and 5·3 months for patients with more than 20 mitoses per mm² of tumoural tissue. Our results suggest that PCNA immunoreactivity is useful in discriminating between neoplastic and non-neoplastic mesothelium and that it may have prognostic value in malignant mesothelioma.