Nuclear overexpression of p53 protein does not correlate with gene mutation in primary peritoneal carcinoma

Tumor suppressor gene, cell surface adhesion molecule, and multidrug resistance in Müllerian serous carcinomas: clinical divergence without immunophenotypic differences

OBJECTIVES

We hypothesize that differences in the expression of selected tumor suppressor genes, cell surface adhesion molecules, and multidrug resistance glycoproteins could account for some of the reported differences between uterine serous carcinoma (USC) and extrauterine serous carcinomas (ESC), including ovarian and primary peritoneal carcinoma (OSC and PSC, respectively).

METHODS

We studied the expression of the following antigens in 20 USCs, 20 OSCs, and 10 PSCs: p53 and mdm-2 (tumor suppressor genes), CD44 and CD44v6 (cell surface adhesion molecules), and the p-glycoprotein (a multidrug resistance protein recognized by two antibodies, C494 and JSB1). We further studied chemotherapeutic drug resistance by examining reports prepared using the Oncotech Extreme Drug Resistance Assay from 24 of the 50 study patients. Clinical data were obtained from medical record review.

RESULTS

USC, OSC, and PSC patients were similar with respect to mean age at diagnosis, mean gravidity, mean parity, personal history of breast cancer, percentage treated with chemotherapy, and survival at 3 and 5 years postdiagnosis. Significant clinical differences included a high prevalence of nulliparity in OSC (P = 0.05), a low prevalence of Caucasian race in USC (P = 0.008), a paucity of stage I patients in OSC and PSC (P = 0.03), a high prevalence of familial breast cancer in OSC (P = 0.06), and superior 2-year survival in OSC (P = 0.02). Seventy-five percent of USCs, 52% of OSCs, and 60% of PSCs expressed p53. Five percent of USCs, 19% of OSCs, and 0% of PSCs expressed mdm-2. Forty percent of USCs, 33% of OSCs, and 10% of PSCs expressed CD44. None of the USCs, OSCs, or PSCs expressed CD44v6. Sixty-one percent of USCs and OSCs and 82% of PSCs expressed C494 while 17% of USCs, 19% of OSCs, and 20% of PSCs expressed JSB1. None of these apparent differences was statistically significant. USC, OSC, and PSCs patients did not demonstrate significant differences with respect to extreme drug resistance. However, the following trends were noted (P = 0.06): more prevalent low drug resistance for cyclophosphamide in OSC compared with USC and more prevalent extreme drug resistance for etoposide in OSC compared with USC.

CONCLUSIONS

Therefore, despite significant clincial differences, the USCs and ESCs in our series do not differ significantly with respect to the expression of the tumor suppressor genes, cell surface adhesion molecules, and drug resistance proteins studied. It is premature, however, to recommend that USCs and ESCs should be treated identically.

Topoisomerase II alpha: prognostic predictor and cell cycle marker in surface epithelial neoplasms of the ovary and peritoneum

Immunohistochemistry for Topoisomerase II alpha (TopoIIa), a nuclear protein important for the separation of chromosomes and deoxyribonucleic acid replication, provides insight into the molecular events in the cell cycle and the response to chemotherapeutic agents, which target TopoIIa. We test the hypothesis that the percentage of TopoIIa immunoreactive nuclei (TopoIIaI) aids in the treatment and prognostic evaluation of ovarian and primary peritoneal surface epithelial neoplasms (SENs) and correlates with established cell cycle control markers: p53, p21WAF1/CIP1 (p21), and Ki67. Paraffin sections from a retrospective surgical series of 108 SENs were immunostained with anti-TopoIIa, anti-p53, anti-p21, and anti-Ki67. The TopoIIaI, the Ki67 proliferation index (Ki67PI), and the immunoreactivity score for p53 and p21 (IMS: S1, S2, S3 < 10%, 10 to 50%, > 50% of strong staining cells, respectively) were evaluated manually. TopoIIaI and Ki67PI ranged from 5 to 84% and 4 to 88% (mean/median: 31/30 and 44/46%), respectively, and were correlated (coefficient 0.62, p < 10(-11)). IMS of 108 SENs was as follows: p53 50% + (2S1, 52S3) and p21 66% + (38S1, 12S2, 21S3). The TopoIIaI associated directly with p53 (p < 10(-5) and inversely with p21 (p < 0.005) IMS. TopoIIaI correlated with SEN architectural/nuclear grade (p < 10(-5)/10(-7)), but not histologic type. Sixty-seven patients had disease at last follow-up, 55 were dead from disease at 2 to 67 months (mean/median 24/21), and 14 were alive with disease at 31 to 230 months (mean/median 73/59). Forty-one patients were disease free at 5 to 228 months (mean/median 75/54). TopoIIaI correlated with presence of disease (p < 0.01) and poor survival (p < 1 x 10(-9), even when only 93 invasive SEN cases are considered (p < 0.005). TopoIIaI correlates with poor prognosis and other cell cycle control markers. The patients in this retrospective series of SEN were treated primarily with platinum-based chemotherapy. These data may suggest further prospective studies in which patients with SENs exhibiting high TopoIIaI are treated with chemotherapy targeted against TopoIIa (e.g., etoposide). In this retrospective series, high SEN TopoIIaI predicted poor survival when treated with platinum-based chemotherapy, which does not target TopoIIa.

Loss of heterozygosity in P53, BRCA1, and estrogen receptor genes and correlation to expression of p53 protein in ovarian epithelial tumors of different cell types and biological behavior

Loss of heterozygosity (LOH) of tumor suppressor genes (TSGs) in ovarian epithelial tumors of differing cell types and biological behavior has not been thoroughly investigated. Moreover, there have been conflicting reports correlating LOH of the p53 gene to overexpression of p53 protein. This study evaluated 34 formalin-fixed, paraffin-embedded ovarian epithelial tumors for LOH by polymerase chain reaction (PCR) for the following microsatellite markers: TP53(17p13.1/p53 gene), D17S579(17q/BRCA1 gene), and ESR (6q24-27/estrogen receptor gene). LOH of the TP53 marker was detected in 4 (44%) of 9 informative serous cystadenocarcinomas (SCa) but in 0 of 4 informative clear cell carcinomas (CCa) and 0 of 5 informative serous tumors of low malignant potential (SLMP). LOH of the BRCA1 marker was detected in 5 (83%) of 6 informative SCa, but in 1 (13%) of 8 informative CCa and 1 (14%) of 7 informative SLMP. LOH of the ESR marker was detected in 4 (50%) of 8 informative SCa, but in 0 of 4 informative CCa and 1 (16%) of 6 informative SLMP. p53 protein overexpression was present in 8 of 12 SCa but did not correlate to TP53 LOH. LOH for TP53, D17S579/ BRCA1, and ESR is common in ovarian SCa, and is observed in primary tumors as well as metastases. In contrast, these genetic alterations are less common in CCa and in the biologically less aggressive SLMP tumors. These data suggest different mechanisms of oncogenesis in ovarian epithelial tumors of different cell types and biological behavior.

Immunohistochemical markers of cell cycle control applied to ovarian and primary peritoneal surface epithelial neoplasms: p21(WAF1/CIP1) predicts survival and good response to platinin-based chemotherapy

Immunohistochemistry for p53, p21(WAF1/CIP1), and Ki-67 provides insight into the molecular events controlling the cell cycle. We tested the hypothesis that these cell cycle markers will aid in the clinical evaluation of ovarian and primary peritoneal surface epithelial neoplasms (SENs). Paraffin sections from a retrospective surgical series of 117 SENs were immunostained with anti-p53 (clone DO7, Novacastra Laboratories, UK), anti-p21(WAF1/CIP1) (clone EA10, Oncogene Science, Cambridge, MA), and anti-Ki-67 (clone MIB-1, Immunotech, Westbrook, ME). The Ki-67 proliferation index (Ki-67PI) and immunoreactivity were evaluated. One hundred seventeen SENs reacted as follows: p53 50%+ and p21(WAF1/CIP1) 65%+. Ki-67PI ranged from 4% to 88% (mean/median = 44/46%). p53 reactivity associated with transitional cell histology, decreased p21(WAF1/CIP1) staining, increased Ki-67PI, architectural/nuclear grade, and stage (P < .05, 1 x 10(-7), .01, .05/.0001, .001,). p21(WAF1/CIP1) staining was associated with endometrioid/clear cell histology, decreased Ki-67PI, architectural/nuclear grade, and stage (P < 05/.05, .05, .01/1 x 10(-8), 1 x 10(-5)). Ki-67PI associated with increased architectural/nuclear grade but not mucinous histology (P < 1 x 10(-5)/1 x 10(-6), .01). Sixty-seven patients had disease at last follow-up; 53 were dead of disease at 0 to 67 months (mean/median, 21/18), and 14 were alive with disease at 12 to 224 months (mean/median, 56/40). Fifty patients were disease free at 5 to 214 months (mean/median, 59/41). Predictors of survival include decreased Ki-67PI, stage, architectural/nuclear grade (P < 1 x 10(-6), 1 x 10(-10), 1 x 10(-10)/.005) and p21(WAF1/CIP1) IMS (multivariate P < 1 x 10(-6)). p21(WAF1/CIP1), a potent inhibitor of cyclin-dependent kinases necessary for cell cycle progression, functions as a key checkpoint in cell cycle control. Immunoreactivity for p21(WAF1/CIP1) provides prognostic information independent of other histological and clinical predictors, p53 IMS, and Ki-67PI in this series of 117 PTs with SENs. Our preliminary data suggest an interrelationship between p21(WAF1/CIP1) expression and an effective clinical response to platinin-based chemotherapy, both associated with apoptosis. Further investigation seems warranted.

Loss of p53 function in uterine papillary serous carcinoma

In contrast to endometrioid carcinoma, uterine papillary serous carcinoma (UPSC) is an aggressive type of endometrial cancer. Loss of p53 function is critical for the molecular pathogenesis of UPSC. Both UPSC and its putative precursor, endometrial intraepithelial carcinoma (EIC), show abnormal p53 overexpression in most tumors. To further assess the nature of p53 alterations in UPSC, we systematically reevaluated a subset of our previous cohort of UPSC patients. In the current study, we correlate mutations of the p53 gene as detected by direct sequencing of exons 5 through 8 with p53 accumulation and expression of Waf-1 in 32 UPSC tumors. Waf-1 is a downstream effector of p53-mediated G1 arrest after DNA damage and, thus, an indicator of p53 functionality. Although 78% of tumors exhibited strong nuclear p53 immunoreactivity in 100% of tumor cells, we were able to detect p53 mutations in 53%. As expected, all p53 mutant tumors (17 cases) exhibited p53 overexpression. Seventy percent of those (12 tumors) showed concomitant lack of Waf-1 expression consistent with transcriptionally inactive p53, whereas the other five tumors showed Waf-1 staining in only a minor fraction of tumor cells consistent with p53-independent Waf-1 expression. In contrast, 47% (15 cases) of tumors failed to exhibit p53 mutations; interestingly, more than half of those (eight cases) showed strong nuclear p53 accumulation in all tumor cells but lacked concomitant Waf-1 expression. These findings are consistent with a mutation-dependent and -independent type of p53 inactivation in UPSC that are both associated with nuclear overexpression. Our findings suggest that the combined immunocytochemical analysis of p53 and Waf-1 is a valuable means of assessing the functional status of p53. In summary, p53 alterations are common in UPSC and probably responsible for its aggressive biological behavior.