Validation of p53 accumulation as a predictor of distant metastasis at 10 years of follow-up in 1400 node-negative breast cancers

Most studies are in favor of a prognostic relevance of p53 accumulation determined by immunohistochemistry in breast cancer, but negative results are not lacking. On a series of 1400 patients with lymph node-negative cancers treated with local-regional therapy alone until relapse and with a median follow-up of 10 years, we validated the prognostic relevance for overall relapse and death of p53 accumulation observed in a pilot study and analyzed its predictivity on different adverse events. p53 protein accumulation was immunocytochemically detected using PAb1801. The case series had also been previously characterized for hormone receptor content [estrogen receptors (ERs) and progesterone receptors (PgRs)] and for cell proliferation [[3H]thymidine labeling index ([3H]dT LI)]. p53 expression, considered as a dichotomous variable with a cutoff value of 5% positive cells, significantly predicted the occurrence of overall relapse, distant metastasis, and death with an interaction with cell proliferation. p53 accumulation, cell proliferation, and their interaction term, along with tumor size and patient age, retained a predictive role for overall relapse, and together with tumor size and PgR, also for overall survival. When considered as continuous variables, we observed that the hazard of metastasis increased linearly with the increase of [3H]dT LI and decreased linearly with the increase of ER and PgR. Conversely, the hazard increased with the increase of p53-positive cells only for tumors with a [3H]dT LI lower than 7.5%. In multivariate analysis, the same prognostic factors for distant metastasis were identified when the biomarkers were considered as continuous or dichotomous variables.

Biological prospectives to define prognosis and treatment strategies in liver metastases from colorectal cancer

Liver metastases arise in about a third of patients with colorectal cancer. Although important clinical results have been obtained by surgical treatment in patients with limited liver involvement, other intra-arterial or systemic therapies do not provide important long term clinical benefits in patients with unresectable liver metastases. Better knowledge of the biology of liver metastases could imply a more appropriate use of the available therapeutic approaches and a retrospective definition the biologic subgroups of patients who benefit from them. Phenotypic and molecular aspects of tumor cells have been investigated and have proven to be important determinants of clinical outcome in patients with different human tumor types. Liver metastases from colorectal cancer have been scarcely studied, but cell proliferation has been shown to be a discriminant of freedom from progression and even more of long-term clinical outcome in subsets in patients treated with radical surgery. Moreover, in patients with resectable liver metastases, DNA and entity of DNA abnormalities are significantly associated with patient survival. A few recent reports have indicated a potential prognostic relevance of abnormal activation or expression of the p53 tumor suppressor gene, bel-2 protein and ras oncogene. In conclusion, prognostic biologic factors are acquiring an important role as indicators of clinical outcome in patients with liver metastases. However, all information is derived from retrospective analyses heterogeneous for patient population and biomarkers analyzed. Therefore, the comparison among results from different studies is difficult, and prospective studies are needed to develop a prognostic classification which integrates biologic and pathologic factors.

Fludarabine as a modulator of cisplatin activity in human tumour primary cultures and established cell lines

The potential of the purine analogue fludarabine (9-beta-D-arabinofuranosyl-2-fluoroadenine-5' monophosphate) as a modulator of cisplatin cytotoxicity was investigated in four established cell lines and 20 primary cultures of human melanoma and ovarian cancer. Tumour cells were exposed to fludarabine and cisplatin, alone or in combination, for 4 h. Fludarabine did not affect the growth of ovarian cancer cell lines, whereas it induced a marked and dose-dependent inhibition of proliferation in melanoma cell lines. In primary cultures of both histotypes, the purine analogue did not induce appreciable antiproliferative effects. Combined cisplatin-fludarabine treatment caused additive effects in all established cell lines. Conversely, a synergistic effect of the combination was seen in 5 of 10 melanoma and 4 of 10 ovarian cancer primary cultures, with a dose-modifying factor ranging from 2.1 to 3.9 for melanomas and from 4.0 to 7.5 for ovarian cancers, respectively. In the remaining cultures, the interaction between fludarabine and cisplatin was additive. The alkaline filter elution analysis performed on primary cultures showed that the synergistic interaction between the two drugs was paralleled by an increase in the extent and persistence of the cisplatin-induced DNA interstrand crosslinks. Our results indicate that fludarabine can enhance cisplatin cytotoxic activity in human tumour primary cultures from ovarian cancer and malignant melanoma. Such an effect may be partially due to an interference by fludarabine on cisplatin-induced DNA adduct metabolism and repair.

Image analysis for the evaluation of p53 expression in human cancers

We compared the results of a subjective microscopic evaluation and an interactive image analysis of p53 expression in different human tumours. Expression of p53 was immunocytochemically detected by using the monoclonal antibody PAb 1801, and the evaluation was made blindly by two observers. Image analysis was performed on the same sections using DISCOVERY. The image was segmented into objects and background by interactive thresholding. Several morphological and densitometric features were selected in order to remove artefacts, lymphocytes, stromal cells and overlapping nuclei from the counting. The study was performed on series of 30 cases for different tumour types: breast, colon, oral cavity, ovary and lung cancer. Spearman's correlation coefficient ranged from 0.72 to 0.95 (P = 0.0001). Image analysis can therefore be considered as a quick and alternative approach to microscopic evaluation of nuclear immuno-determinations.

Modulation of melphalan cytotoxic activity in human melanoma cell lines

The aim of the present study was to potentiate the cytotoxic effects of melphalan through pharmacological and physical modulators. The combination of the cytotoxic agent with ethacrynic acid, a glutathione-S-transferase pi (GST pi) inhibitor, or topotecan, a topoisomerase I inhibitor, or mild hyperthermia was investigated. The selected cell lines exhibited variable levels of expression of GST pi, DNA topoisomerase I and heat-shock proteins. Mild hyperthermia (42 degrees C) alone potentiated melphalan cytotoxicity, especially in the two cell lines exhibiting low basal levels of HSP70 expression. The combination of the GST inhibitor with melphalan resulted in a potentiation of drug cytotoxicity only in JR8 cells, one of the two cell lines which expressed high levels of GST pi mRNA and which were the less responsive to ethacrinic acid alone. A synergistic interaction between topotecan and melphalan was observed only in the cell lines expressing low levels of topoisomerase I even if all cell lines exhibited a comparable sensitivity to this agent. The results support an involvement of GST and DNA topoisomerase in cell defense and response to the alkylating agent. However, the variable potentiation of the cytotoxic effects of melphalan achieved in different cell systems suggests that factors other than the level of expression of the modulation target are responsible of such potentiation.

p53 expression, DNA content and cell proliferation in primary and synchronous metastatic lesions from ovarian surface epithelial-stromal tumours

The aim of this study was to investigate biological heterogeneity between primary and metastatic ovarian cancer lesions from individual patients as a means of elucidating steps in clinical progression. Cancer tissue from 61 untreated patients with ovarian surface epithelial-stromal tumours was examined. p53 expression detected immunocytochemically by the PAb1801 antibody, DNA content evaluated by flow cytometry, and cell proliferation evaluated as the [3H]thymidine labelling index were investigated in primary tumours and corresponding synchronous metastases. The frequency of p53 positivity was similar in primary (62%) and metastatic (66%) sites, with an agreement between the two lesions from the same patient in 97% of the cases. Similarly, aneuploidy frequency (80%) and DNA indices were superimposable in primary and metastatic lesions from the same patient, with a 94% agreement. The frequency of aneuploidy was higher in p53-positive than in p53-negative lesions. An overall poor agreement (rs = 0.44) was observed for proliferative activity of primary and metastatic lesions, due to a heterogeneous profile in omental with respect to primary tumours, which was mainly evident in p53-positive cancers. Conversely, cell proliferation of peritoneal, abdominal and pelvic lesions was qualitatively similar to that of the primary tumour in 88% of patients.

p53 and bcl-2 expression correlates with clinical outcome in a series of node-positive breast cancer patients

BACKGROUND AND PURPOSE

The tumor-suppressor gene TP53 and the proto-oncogene bcl-2 encode, respectively, for a nuclear phosphoprotein and for a mitochondrial protein involved in multiple cellular functions. The proteins provide prognostic information in node-negative breast cancer and are supposed to influence treatment responsiveness. We analyzed the predictive role of p53 and bcl-2 expression, alone and in association with other variables, in postmenopausal women with node-positive, estrogen receptor-positive (ER+) breast cancers treated with radical or conservative surgery plus radiotherapy and adjuvant tamoxifen for at least 1 year.

PATIENTS AND METHODS

On 240 resectable cancers, we determined the expression of p53 and bcl-2, using immunohistochemistry, cell proliferation (3H-thymidine labeling index [3H-dT LI]), and ER and progesterone receptors (PgR).

RESULTS

p53 expression and 3H-dT LI were weakly related to one another and both were unrelated to bcl-2. Relapse-free and distant metastasis-free survival at 5 years were significantly lower for patients with tumors that highly expressed p53 (P = .0001) and for those that weakly expressed or did not express bcl-2 (P = .02). However, p53, but not bcl-2, provided prognostic information independent of tumor size, axillary node involvement, steroid receptors, and 3H-dT LI. Moreover, the simultaneous p53 overexpression and lack of PgR identified patients at maximum risk of relapse, whereas bcl-2 overexpression, associated with a low 3H-dT LI or the presence of PgR, improved the prognostic resolution for low-risk patients.

CONCLUSION

p53 expression appears to be indicative of clinical outcome in postmenopausal patients treated with tamoxifen. Whether p53 overexpression and weak bcl-2 expression are indicators of biologic aggressiveness, regardless of treatment, or of hormone resistance remains to be defined.

P53 expression, DNA ploidy and S-phase cell fraction in operable locally advanced non-small-cell lung cancer

The identification of biomarkers to complement pathological stage for a more accurate prognosis and help clinicians decide on treatment is still an open problem for patients with lung cancer. Expression of P53 protein was detected by an immunohistochemical approach using the monoclonal antibody PAb1801 on paraffin-embedded sections of tumours obtained surgically from 102 stage II - IIIa patients with non-small-cell lung cancer (52 squamous cell carcinomas, 50 adenocarcinomas). [3H]Thymidine labelling index, an indicator of the S-phase cell fraction, was evaluated on histological sections of [3H]thymidine-labelled tumour samples. DNA ploidy was defined by flow cytometric analysis on frozen tumour tissue. The biomarkers, histology and pathological stage were analysed in relation to relapse-free survival in univariate and multivariate analyses. Stage and interaction between [3H]thymidine labelling index and histology provided significant prognostic information for the overall series. [3H]thymidine labelling index was an independent prognostic indicator of 3 year relapse-free survival in patients with adenocarcinoma. The results indicate the importance of cell proliferation to complement prognostic information provided by pathological stage in patients with stage II-IIIa adenocarcinomas.

Effect of ionizing radiation on cell-cycle progression and cyclin B1 expression in human melanoma cells

In the present study we investigated the effect of gamma-irradiation (2.5 and 10 Gy) on cell-cycle progression of a human melanoma cell line, M14, characterized by a moderate radiosensitivity (SF2 = O.5). Flow cytometric analysis showed a dose-dependent S-phase accumulation, which was detectable 8 hr after treatment with 2 and 5 Gy and was still persistent at 12 hr after 10 Gy exposure. Such a delay in S-phase was paralleled or followed by an accumulation of cells in G2M, which was transient at the lowest radiation doses and still persistent at 72 hr after 10 Gy. Such an accumulation was, at least in part, due to a block in G2-M transition, as demonstrated by mitotic index analysis. Bivariate flow cytometric analysis of DNA content and cyclin B1 expression showed that, following 2 and 5 Gy, the fraction of cyclin B1-expressing cells was superimposable upon that of G2M cells. Conversely, in cells treated with 10 Gy, the fraction of cyclin B1-expressing cells was half the G2M cell fraction. Northern-blot analysis indicated that the radiation-induced decrease in cyclin B1 protein expression was accompanied by a reduced cyclin B mRNA level. On the whole, our results indicate a direct inhibitory effect of 10 Gy irradiation on cyclin B1 expression as a possible cause for the persistent G2 block in irradiated M14 cells.

Effect of melphalan and hyperthermia on cell cycle progression and cyclin B1 expression in human melanoma cells

We have investigated the effect of mild hyperthermia (42 degrees C) on the cytotoxic activity of a 1 h melphalan exposure in human melanoma cell lines. Hyperthermia did not affect cell growth of any culture, but it increased, to a different extent, melphalan cytotoxicity in all cell lines, with a reduction in the IC50 of 1.7 to 2.6-fold. Flow cytometric analysis showed that in normal temperature conditions melphalan caused S phase cell accumulation, which was evident only at 24 h in JR8, M14 and 2/21 cell lines and was still persistent at 72 h in 2/60 cells. Moreover, in all cell lines, the delay in S phase was paralleled, or followed, by an accumulation of cells in G2+M, which was transient in JR8 and M14 cells and persisted until 72 h in 2/21 and 2/60 melanoma clones. Hyperthermia caused a stabilization and prolongation of melphalan induced G2+M accumulation in JR8 and M14 cells. Conversely, in 2/21 and 2/60 clones, cell cycle perturbations induced by the drug were similar under normothermic or hyperthermic conditions. Specifically, in JR8, for which the maximum enhancement by hyperthermia on melphalan cytotoxicity was observed, cell accumulation in G2+M was still present 120 h after treatment. The accumulation was accompanied by an inhibition in the G2-M transition, as demonstrated by the significant reduction in the mitotic index of cells exposed to combined treatment compared to controls. Moreover, a bivariate distribution of cells stained for DNA and cyclin B1 showed that, following melphalan and hyperthermia treatment, the fraction of cyclin B1-expressing cells paralleled the fraction of G2+M phase cells, thus indicating that the inability of cells to enter mitosis was not ascribable to a reduction of cyclin B1 expression. On the whole, our results indicate that hyperthermia can stabilize the G2 accumulation induced by melphalan in human melanoma cells. Such a stabilization could contribute to the enhancement of melphalan cytotoxicity by heat, even though a strict correlation was not observed between the magnitude and persistence of the cell cycle perturbations and the extent of melphalan activity.

Detection of P53 expression and S-phase cell fraction in paraffin-embedded tissue by a double-labeling technique

TP53 is a gene that normally regulates cell growth and division. Alterations to it may induce a proliferative advantage and confer an aggressive phenotype. In breast cancer, we observed a poor correlation (rs = 0.17) between P53 expression and proliferative activity evaluated as [3H]-thymidine ([3H]-dT) labeling index and an independent prognostic relevance of the two variables. We used a double-labeling technique to simultaneously evaluate the fraction of P53-positive and [3H]-dT-labeled cells to analyze the degree of association between the two markers on individual cells in order to understand their biological significance. The study was performed on a series of 44 P53-positive (P53+) breast cancers. Histological sections were immunostained for P53 with monoclonal antibody (MAb) PAb1801 and then processed for autoradiography. A weak direct relation between P53 positivity and [3H]-dT incorporation (rs = 0.4) was observed on the overall series of P53+ tumors and was maintained in subgroups defined by several biological and pathological features, except for estrogen receptor-negative tumors. The simultaneous presence of P53 expression and [3H]-dT incorporation was directly and significantly proportional to the fraction of S-phase cells of the tumor (rs = 0.7). Conversely, the fraction of cells expressing only P53 was inversely related to cell proliferation (rs = -0.66). These findings support the hypothesis that P53 has biological functions other than cell cycle regulation.

Contribution of ploidy and cell kinetics from fine-needle aspirates for the diagnosis of breast lesions: study of 606 consecutive cases

A total of 606 consecutive patients with palpable breast lesions underwent physical examination, mammography, and fine-needle aspiration for cytologic and flow-cytometric (FCM) DNA analyses. FCM determinations included DNA ploidy and the fraction of cells in S + G2 + M phases. Aneuploid clones were considered indicative of malignancy; diploid, rapidly proliferating (S+G2+M>12%) clones were considered suggestive of malignancy; and diploid, slowly proliferating clones were considered indicative of benignancy. Sensitivity, specificity, and the predictive accuracy for positive or negative results of FCM information were 82%, 90%, 93.1% and 69.8%, respectively. These values were lower than those observed for the three conventional diagnostic assays owing to the presence of false-positive results (in 23 cases) or unassessable DNA plots (in 90 cases). FCM information alone detected five cases in which the other tests gave inconclusive results and in association to the conventional diagnostic triple test, increased the incidence of positive conclusive cases from 97.9% to 99.2%. However, further refinements are needed before this approach can be used as a routine diagnostic tool.

Changes in biological markers after primary chemotherapy for breast cancers

The profiles of functional (proliferative rate and cell distribution in the cell cycle) and phenotypic (nuclear DNA content and hormone receptor status) biological markers and the expression of P53 and Bcl-2 proteins were prospectively evaluated in breast cancers before and after different regimens of primary chemotherapy. Overall, changes induced on the 2 proliferation indices (3H-thymidine labelling index, 3H-dT LI, and flow-cytometric S-phase fraction, FCM-S) mainly consisted of a decrease for rapidly proliferating tumours and an increase or no change for slowly proliferating tumours. However, when considered as a function of treatment type, changes of 3H-dT LI and FCM-S were superimposable in rapidly proliferating tumours, regardless of the type of treatment, and in slowly proliferating tumours only after anthracycline-including regimens. Conversely, following CMF, FCM-S was increased in 90% of the cases and 3H-dT LI in only 50%. Our data imply that the 2 proliferation indices could reflect different phenomena: an actual variation of proliferative activity by 3H-dT LI and an accumulation of cells in the S-phase by FCM-S. In addition, a higher accumulation of cells in G2-M phases could be detected by FCM after anthracycline-including regimens than after CMF. The fraction of P53-positive cells was reduced by primary chemotherapy in about 50% of P53-positive tumours, whereas Bcl-2 expression was only marginally affected. DNA ploidy and hormone receptor status did not change in about 75% of cases, regardless of the chemotherapeutic regimen.

p53 gene point mutations in relation to p53 nuclear protein accumulation in colorectal cancers

It is known that structural alterations of the p53 tumour suppressor gene cause malignant transformation and tumour progression in colorectal mucosa. In this study, 38 colorectal cancers were analysed for mutations detected in the p53 gene by single-strand conformational polymorphism and DNA sequence analysis, and the results were compared with p53 protein expression detected by immunohistochemistry. A very strict association (P < 0.0001) was found between genetic alterations and protein accumulation, as detected by the PAb 1801 monoclonal antibody. p53 expression and gene mutations were more frequent in rectal than in colonic cancers. No relation was observed with Dukes' stage, even though most of the mutations were at exon 7 in Dukes' A-B cancers and almost all mutations at exon 8 were observed in Dukes' C-D cancers. DNA ploidy was not generally associated with p53 protein expression or gene mutations. However, 83 per cent of cases with exon 5 and 6 mutations were diploid or near-diploid and 71 per cent of cases with mutations at exons 7 and 8 were aneuploid. Tumours with p53 gene mutations at exon 5 had a higher median [3H]thymidine labelling index (17 per cent) than those with mutations at exons 6, 7, and 8 (11.8 per cent).

Biologic and clinicopathologic factors as indicators of specific relapse types in node-negative breast cancer

PURPOSE AND METHODS

We evaluated, in 1,800 patients with node-negative tumors treated with locoregional therapy until relapse, the competitive risks for different types of metastasis by cell proliferation (3H-thymidine labeling index [3H-dT LI]), estrogen receptors (ERs), and progesterone receptors (PgRs), and by the integration of biologic and clinicopathologic information.

RESULTS

Hormone receptor status and proliferative activity of the primary tumor were not indicative of contralateral failures. Hormone receptors failed to predict the 8-year incidence of locoregional recurrence, but they were significant indicators of distant metastasis and overall survival. The latter finding was confirmed even in multivariate analysis. Conversely, cell proliferation predicted both locoregional and distant metastases and survival, regardless of patient age, tumor size, and ER and PgR status. Recursive partitioning and amalgamation analysis ascribed to cell proliferation an important prognostic role for locoregional recurrence together with patient age and tumor size.

CONCLUSION

Biologic markers, in particular cell proliferation, provide information for the different types of relapse and could complement the predictive role of pathologic staging.

Lack of a correlation between p53 protein expression and radiation response in human tumor primary cultures

We investigated the possible relationship between immunohistochemically detected p53 expression and in vitro response to gamma-irradiation in 24 primary cultures of human ovarian cancers and cutaneous melanomas. The frequency of p53-positive tumors was around 60% within each tumor histotype. The range of the surviving fraction at 2 Gy (SF2) was similar in p53-positive (0.10-0.76) and p53-negative (0.23-0.65) tumors, with median values of 0.36 and 0.33, respectively. No differences were observed in the accumulation of DNA-double strand breaks, assessed by neutral filter elution after exposure to 50 Gy, between p53-positive and p53-negative tumors. As regards DNA lesion repair, after 2 h of recovery the percentage of rejoined DNA-double strand breaks ranged from 19% to 99% in the different cultures, but again the distribution of values was similar for p53-positive and p53-negative tumors. Specifically, the median percentage of repaired DNA-double strand breaks was 70% and 74% in the two groups. On the whole, our data do not support the hypothesis that p53 overexpression is a major determinant of in vitro radiation response.

Lack of a correlation between micronucleus formation and radiosensitivity in established and primary cultures of human tumours

The radiation-induced genotoxic damage in three established cell lines and 15 primary cultures of human malignant melanoma and ovarian carcinoma showing different radiosensitivity was tested by the cytokinesis-block micronucleus assay. A dose-related increase in micronucleus frequency was observed in all the cell systems. The mean number of micronuclei per Gy of ionising radiation per binucleated cell was respectively 0.44 +/- 0.0075 and 0.43 +/- 0.04 for M14 and JR8 malignant melanoma cell lines and 0.19 +/- 0.013 for the A2780 ovarian cancer cell line. The number of micronuclei did not rank the cell lines in the same order of radiosensitivity as clonogenic cell survival, which showed a surviving fraction at 2 Gy of 0.38 +/- 0.02 for JR8, 0.34 +/- 0.05 for M14 and 0.22 +/- 0.007 for A2780. As regards primary tumour cultures, no correlation was observed between micronucleus induction and surviving fraction at 2 Gy. In conclusion, the discrepancy we observed between micronucleus formation and cell death raises doubts about the potential of the micronucleus assay as a preclinical means to predict radiosensitivity.

DNA double-strand break repair and radiation response in human tumour primary cultures

The accumulation and repair of radiation-induced DNA double-strand breaks (dsbs) were determined by neutral filter elution on 20 primary cultures obtained from ovarian cancer and malignant melanoma clinical specimens. The initial frequency of DNA dsbs after exposure to 50 Gy gamma-irradiation varied greatly for the individual cultures. However, melanomas were generally more efficient than ovarian cancers in repairing these DNA lesions (mean percentage of DNA dsb rejoined after 2 h: 83 versus 62%). In 13 of 20 cultures radiosensitivity was also assessed by the Courtenay clonogenic assay. The mean +/- SD of the surviving fraction at 2 Gy (SF2) was slightly higher for melanomas (0.56 +/- 0.25) than for ovarian carcinomas (0.43 +/- 0.23). No correlation was observed between SF2 and in vitro plating efficiencies or any biological characteristics of the tumour cell population, such as proliferative activity and DNA ploidy. Similarly, we failed to find any relation between the initial frequencies of DNA dsbs and SF2 in individual tumours. In contrast, a significant and direct relationship (r = 0.86, p < 0.01) was observed between SF2 and the percentages of DNA dsbs rejoined 2 h after irradiation. In agreement with reported data on human tumour established cell lines, our results indicate that the ability to repair DNA dsbs is an important determinant for radiation response even in primary cultures of clinical tumours.

Potentiation of cisplatin cytotoxicity by lonidamine in primary cultures of human ovarian cancer

The ability of lonidamine, an energolytic derivative of indazole-carboxylic acid, to modulate the cytotoxic activity of cisplatin was investigated on primary cultures obtained from 11 human ovarian carcinomas. A 72-h postincubation with lonidamine potentiated the activity of a 1-h cisplatin treatment. Statistical analysis of the dose-effect plots indicated that the interaction between the two drugs was synergistic in 4 tumors and additive in the remaining 7 tumors. The occurrence of the synergistic effect was independent of some biological characteristics of the tumor cell population, such as cell kinetics (as assessed by 3H-thymidine labeling index), DNA content and the expression of the putative factor of cisplatin resistance, glutathione-S-transferase pi.