Long-term significance of urinary neopterin in ovarian cancer: a study by the Austrian Association for Gynecologic Oncology (AGO).

BACKGROUND: Neopterin is produced by activated macrophages upon stimulation with interferon-γ (IFN-γ) and thus, elevated neopterin concentrations in patients indicate cellular inate immune response. Most studies in patients with malignant diseases found an association between higher neopterin concentrations and reduced survival and impaired prognosis. Nevertheless, neopterin is not a classical tumor marker since it is not produced by the cancer cells themselves. PATIENTS AND METHODS: In a study conducted by the Austrian Gynecologic Oncology Group (AGO) in 114 patients with ovarian cystadenomas and 223 patients with invasive ovarian cancer, patients' urinary neopterin was determined before and after primary therapy. The relevance of neopterin in long-term median follow-up was assessed. RESULTS: Elevated levels (cut-off 250 µmol/mol creatinine) were found less frequently in women with benign ovarian cystadenomas (24%) than in patients with malignant disease (58%). After 10 years, only 57% of ovarian cancer patients with elevated urinary neopterin levels survived without disease progression following primary therapy when compared with 86% of women with normal levels (P < 0.001). Along with residual tumor, FIGO stage, age and histological type, neopterin was significantly associated with overall survival (OS) and progression-free survival (PFS). The median PFS was 52 and 12 months and the median OS was 81 and 24 months for patients with normal and elevated neopterin, respectively, P < 0.001. In a multivariate Cox regression analysis, only residual tumor, neopterin and age were independently associated with OS, while only residual tumor was predictive for PFS. Thirty patients with early-stage invasive ovarian cancer (FIGO I and II) were analyzed separately. Of 3 patients with elevated neopterin, 2 died of disease in contrast to 2 out of 27 patients with normal neopterin (P = 0.004). CONCLUSION: In ovarian cancer, the negative impact of elevated urinary neopterin levels indicates a detrimental effect of cancer-associated inflammatory reaction.

[Treatment of Chronic Functional Constipation during Pregnancy and Lactation]. / Behandlung der chronischen funktionellen Obstipation in der Schwangerschaft und Stillperiode.

Natural fibres (bulk-forming agents), docusate sodium (stool-softener), mineral oils (lubricant laxatives), macrogol (polyethylene glycol, PEG), sugars and sugar alcohols (osmotic laxatives) and anthraquinones and diphenolic laxatives (stimulant laxatives) seem to be safe medicaments regarding teratogenicity and lactation. The US Food and Drug Administration (FDA) risk categories for these substances taken during pregnancy and lactation are often the result of the lack of studies than of evidence-based information. So risk categories do not help in the decision-making for the right laxative. Alternative solutions such as proposals of the American College of Gastroenterology's Committee on FDA related matters, (ACG-FDA) and the Motherisk Programme try to improve decision-making. For newer compounds such as chloride-channel-activators and procinetics no data regarding safe use in pregnancy and during breast-feeding are available as yet. We suggest the use of macrogol and lactulose as the first-line therapy in treating chronic constipation during pregnancy. Macrogol shows some advantages, such as faster onset of bowel action and fewer flatulences. If this treatment does not work or starts but then stops working, we recommend in the second and third trimenon a second-line treatment with diphenolic laxatives such as bisacodyl and and sodium picosulfate. During pregnancy the decision on the application of these laxatives is largely determined by the side-effects of tenesmus associated with preterm births. During lactation we recommend macrogol (preferable to lactulose due to the lack of data), lactulose, bisacodyl and sodium picosulfate, according to the nature of the conditions.

Computerized in vivo classification of methylene blue stained fallopian tube mucosal damage: preliminary results.

OBJECTIVE: Fertiloscopy is a simple minimal invasive method which allows salpingoscopy and microsalpingoscopy in order to examine the mucosa of the fallopian tubes of patients with unexplained infertility. Infectious tubal damage is a common cause of tubal infertility. In 1998 it was demonstrated that nuclear staining of cellular nuclei during microsalpingoscopy with methylene blue provides a simple in vivo method to evaluate cellular damage of the tubal epithelium. The purpose of this study was to introduce and statistically test a new computerized method to objectively evaluate the extent of tubal damage. DESIGN OF RETROSPECTIVE STUDY: Cooperation of two Departments of Gynecology and Obstetrics (Krankenanstalt Rudolfstiftung, Vienna, Austria and CRES Center, Hôpital Natecia, Lyon, France) with the University of Art and Design, Linz, Austria and University Hospital, Vienna, Austria. MATERIALS AND METHODS: Microsalpingoscopic images from ten female patients, aged between 18 and 45 years with primary infertility, showing stained nuclei in damaged intrafallopian tubal epithelium were provided by Antoine Watrelot, CRES Center, Hôpital Natecia, Lyon, France. These images were evaluated by an experienced medical expert staff examiner and a computerized standard method called cross-correlation and template matching. The obtained numbers of nuclear stainings were statistically evaluated. RESULTS: Computerized evaluation of nuclear staining of damaged intrafallopian epithelial cells in female patients with infertility obtains similar but more reproducible results compared to manual evaluation (p = 0.007). CONCLUSION: Normalized cross-correlation can be used to measure tubal damage diagnosed by in vivo methylene blue dyeing during microsalpingoscopy and might facilitate the decision for in vitro fertilisation in patients with unclear unexplained infertility in further studies.

HER-2 amplification, steroid receptors and epidermal growth factor receptor in primary breast cancer.

Amplification of HER-2 oncogene was analysed in DNAs obtained from 291 primary human mammary carcinomas. 52/291 (18%) were found to contain amplified HER-2 oncogene. Moderate amplification (2- to 5-fold) was noted in 36/291 (12%). Thirteen tumors (4.5%) had a copy number of 5 to 10. A 10- to 20-fold and greater than 20-fold amplification was observed in 2 and 1 patient, respectively. Sample sizes allowed the determination of estrogen receptor (ER) and progesterone receptor (PgR) levels in 253/291 primary breast cancers. HER-2 gene amplification was noted in 14% of ER+ patients and in 28% of ER- patients, respectively (P = 0.02). Similarly a significantly greater number of PgR- primary mammary carcinoma exhibited an amplification of the HER-2 gene compared to PgR+ cases (22% vs. 16%, P = 0.01). Although statistically not significant, tumors with HER-2 gene amplification were found to have lower levels of ER and PgR. No association of HER-2 amplification with the androgen receptor and EGF receptor was observed. Present data combine to suggest that tumor progression is more stringently controlled by the oncogene upon loss of hormone dependency. Differences found in HER-2 amplification between steroid receptor positive and negative tumors could be helpful to define a specific subset of women to whom adjuvant therapy should be directed.

Determination of HER-2/neu amplification and expression in tumor tissue and cultured cells using a simple, phenol free method for nucleic acid isolation.

A rapid, simple and non-toxic procedure for the simultaneous isolation of DNA and RNA from tumor tissue and cells grown in vitro is described. Guanidinium isothiocyanate was used for homogenization of tumor tissue and for cell lysis. Separation of proteins, DNA and RNA was carried out by isopycnic centrifugation in cesium trifluoroacetate. DNA was further purified by salting out residual protein. Nucleic acids prepared by this method from 47 primary human carcinomas and 17 human cell lines were analysed for amplification and expression of the HER-2/neu proto-oncogene. 2- to 10-fold amplification of HER-2/neu was noted in 7/22 mammary carcinomas (32%) and in 4/14 ovarian carcinomas (28%). No amplification of the proto-oncogene was found in 4 laryngeal carcinomas, 1 pharyngeal carcinoma, 2 retrolingual carcinomas, 3 gastric carcinomas and 1 kidney carcinoma. HER-2/neu overexpression was observed in 6/22 of mammary carcinomas (27%) and 7/14 of ovarian carcinomas (50%). No overexpression was found in all other carcinomas studied. Concordance between amplification and overexpression was noted in 3 mammary and 4 ovarian carcinomas, respectively. 3 mammary and 3 ovarian carcinomas showed overexpression without amplification. 5 human mammary carcinoma cell lines showed both amplification and overexpression of HER-2/neu. In two mammary carcinoma cell lines (MDA MB-453 and ZR 75-1) overexpression was noted without amplification of the proto-oncogene. These data combine to suggest that mechanisms other than gene amplification may also lead to overexpression of the HER-2/neu protooncogene in cancer cells.

Serum concentrations of vascular endothelial growth factor in vulvar cancer.

The aim of the present study was to evaluate serum concentrations of vascular endothelial growth factor (VEGF) in patients with vulvar cancer and healthy female controls with respect to correlation of VEGF with clinicopathological parameters and impact on the patients' prognosis. Serum concentrations of VEGF were measured using a commercially available ELISA. Results were correlated to clinical data. Median serum concentrations of VEGF in patients with vulvar cancer (n = 41) and healthy female controls (n = 130) were 260 (range, 33-1216) pg/ml and 216 (range, 0-777) pg/ml, respectively (Mann-Whitney U test, P = 0.048). Serum concentrations of VEGF significantly correlated with tumor stage (Mann-Whitney U test, P = 0.02) but not with histological grade (Mann-Whitney U test, P = 0.2). In a univariate analysis, elevated pretreatment serum concentrations of VEGF were significantly correlated with a shortened disease-free and overall survival (Wilcoxon test, P = 0.03; and Wilcoxon test, P = 0.04, respectively). A multivariate Cox regression model considering tumor stage and serum concentrations of VEGF revealed, however, that serum concentrations of VEGF did not confer additional prognostic information to that already obtained by the established prognosticator tumor stage (multivariate Cox regression model: P = 0.9 and P = 0.8, respectively). Our data indicate that angiogenesis, as reflected by serum concentrations of VEGF, plays a functional role in vulvar carcinogenesis. VEGF seems to be a mediator of vulvar tumor growth but not of tumor cell dedifferentiation. Although associated with impaired disease-free and overall survival, pretreatment serum concentrations of VEGF are not an independent predictor of outcome in patients with vulvar cancer.

Prognostic value of CD44 splice variants in human stage III cervical cancer.

The expression of specific cell adhesion molecule CD44 isoforms (splice variants) has been shown to be associated with poor prognosis in human malignancies, such as breast cancer. We used three different variant exon sequence-specific murine monoclonal antibodies to epitopes encoded by exons v5, v6 or v7-v8 of human variant CD44, to study the expression of CD44 splice variants by immunohistochemistry in human stage III cervical cancer. We investigated 40 pretreatment punch biopsies of cervical cancer FIGO stage III. CD44 splice variants CD44v5, CD44v6 and CD44v7-8 were detected by means of immunohistochemistry in 90%, 55% and 25%, respectively. CD44 epitopes encoded by exon v5 were not correlated with prognosis. Expression of CD44 splice variants containing epitopes encoded by exon v6 were correlated with significantly poorer prognosis (Mantel test, P = 0.008). Five-year survival rates with or without CD44v6 expression were 20% versus 71%, respectively. Expression of CD44v7-8 was also correlated with significantly poorer overall survival (Mantel test, P = 0.02). Expression of CD44 splice variants containing epitopes encoded by exons v7-v8 and especially exon v6 is associated with significantly poorer prognosis in stage III cervical cancer patients.

Serum M3/M21 in cervical cancer patients.

Cytokeratins are polypeptides which constitute a subclass of intermediate filaments in epithelial cells. The serum tumour marker M3/M21 is based on monoclonal antibodies against the epitopes M3 and M21 of cytokeratin 18. In the present study, we measured M3/M21 serum levels in 50 patients with FIGO stage IB-IIB cervical cancer and in 50 control subjects using a two-site radiometric immunoassay directed against soluble fragments of cytokeratin 18. Median serum levels of M3/M21 in patients with cervical cancer and in normal controls were 70.6 U/ml (range 0-397.7) and 6.5 U/ml (range 0-205.2), respectively (Mann-Whitney U-test, P = 0.0001). Median serum levels of M3/M21 prior to therapy and 4 weeks after therapy were 104.2 U/ml (range 24.6-397.7) and 39.3 U/ml (range 0-234.7), respectively (Mann-Whitney U-test, P = 0.004). We found a significant correlation between elevated M3/M21 serum levels and metastatic disease in pelvic lymph nodes (Mann-Whitney U-test, P = 0.002). 24 patients relapsed after complete remission. In these patients, elevated M3/M21 serum levels before the detection of relapse by computed tomography was observed in 13 cases. Considering these preliminary results, further studies with an increased number of patients are justified to clarify the prognostic value and the monitoring abilities of M3/M21 in cervical cancer patients.

Serum interleukin 1 in ovarian cancer patients.

Interleukin-1 (IL-1) is a multifunctional cytokine playing a central role in the immune response and displaying direct cytotoxic activity in vitro. Serum IL-1 alpha and beta levels were measured by enzyme linked immunosorbent assay (ELISA) in 75 ovarian cancer patients, 30 patients with benign ovarian cysts and 50 healthy controls. Both serum IL-1 alpha and IL-1 beta levels were more often elevated in ovarian cancer patients compared with healthy controls (chi-square test, P < 0.001 and P < 0.001, respectively). Mean serum IL-1 alpha and beta levels decreased significantly after surgical intervention (paired t-test, P = 0.0001 and P = 0.0002, respectively). No correlation with histopathological parameters and overall and disease-free survival was found. These preliminary results indicate that serum levels of IL-1 alpha and beta represent a host defence reaction rather than an autonomous tumour cell production.

Expression of cathepsin D in head and neck cancer.

To determine overexpression of cathepsin D in head and neck tumours we examined cytosols from 53 primary tumours, nine cytosols of lymph node metastases and 12 cytosols from adjacent normal tissue. We found a significantly lower concentration in normal tissue compared with tumour cytosol as well as with metastases, even when we compared tumours and corresponding metastases pairwise. In addition, we found a significantly higher concentration of cathepsin D in five lymph node metastases than in the corresponding tumours. We conclude that the reported role of cathepsin D is not restricted to breast cancer but could also be important in head and neck cancer.

HER-2 oncogene amplification and overall survival of breast carcinoma patients.

DNA was extracted from tumour samples of 77 patients with primary breast carcinoma and HER-2 proto-oncogene amplification was assessed. Prognostic indices such as number of positive lymph nodes, tumour size and histological grading were strongly associated with overall survival. No statistically significant correlation between amplification of HER-2 and overall survival was observed. In addition, prognostic indices, HER-2 amplification and disease-free interval was not correlated. Analysis of HER-2 amplification alone is not a useful guide in the management of patients with mammary carcinoma.

Serum CD44 splice variants in cervical cancer patients.

Aberrant expression of the cell adhesion molecule CD44 has been detected in human tumours and has been shown to be associated with metastasis and poor prognosis in human malignancies. We evaluated serum levels of different soluble CD44 molecules (CD44 standard form and CD44 splice variants v5 and v6) in cervical cancer patients stage IB to IIIB. Two-hundred three serum samples were analysed. Serum levels of CD44st and CD44v5 showed no significant correlation with the presence or absence of cervical cancer. The splice variant CD44v6 showed a mean concentration of 227.3 +/- 90.9 (minimum 71.4, maximum 543.9) ng/ml when tumour was present and a mean concentration of 198.7 +/- 135.4 (minimum 67.2, maximum 696.3) ng/ml in cases of complete remission (P-value = 0.0001). However, in this preliminary study the sensitivity/specificity characteristic of CD44v6 was poor.

Serum evaluation of basic fibroblast growth factor in cervical cancer patients.

We present the data of 105 serum samples from 20 patients suffering from cervical cancer. Mean serum levels of basic fibroblast growth factor (bFGF) in patients with or without tumor present were 31.3 +/- 32.1 (minimum 0, maximum 156.7) pg/ml and 4.8 +/- 6.8 (minimum 0, maximum 29.6) pg/ml, respectively (P = 0.0001). bFGF reached a sensitivity of 65.7% at a specificity of 91.5% when applying a cut-off level of 15 pg/ml. Four patients relapsed after complete remission. A continuous increase of bFGF serum levels before the clinical detection of relapse (lead time) was seen in two cases with a mean lead time of 4 months. Preoperative serum levels were not of prognostic value and showed no correlation with pelvic lymph node metastasis. These preliminary results indicate that in cervical cancer patients soluble bFGF may be useful in early detection of primary tumors, recurrences and monitoring of therapy.

Quantitative immunohistochemistry of factor VIII-related antigen in breast carcinoma: a comparison of computer-assisted image analysis with established counting methods.

Microvessel density in the area of the most intense neovascularization in invasive breast carcinoma is reported to be an independent prognostic factor. The established method of enumeration of microvessel density is to count the vessels using an ocular raster (counted microvessel density [CMVD]). The vessels were detected by staining endothelial cells using Factor VIII-related antigen. The aim of the study was to compare the CMVD results with the percentage of factor VIII-related antigen-stained area using computer-assisted image analysis. A true color red-green-blue (RGB) image analyzer based on a morphologically reduced instruction set computer processor was used to evaluate the area of stained endothelial cells. Sixty invasive breast carcinomas were included in the analysis. There was no significant correlation between the CMVD and the percentage of factor VIII-related antigen-stained area (Spearman correlation coefficient = 0.24, confidence interval = 0.02-0.46). Although high CMVD was significantly correlated with poorer recurrence free survival (P = .024), percentage of factor VIII-related antigen-stained area showed no prognostic value. Counted microvessel density and percentage of factor VIII-related antigen-stained area showed a highly significant correlation with vessel invasion (P = .0001 and P = .02, respectively). There was no correlation between CMVD and percentage of factor VIII-related antigen-stained area with other prognostic factors. In contrast to the CMVD within malignant tissue, the percentage of factor VIII-related antigen-stained area is not suitable as an indicator of prognosis in breast cancer patients.

EGF-R and steroid receptors in breast cancer: a comparison with tumor grading, tumor size, lymph node involvement, and age.

In this study epidermal growth factor receptor (EGF-R), estrogen receptor (ER), and progesterone receptor (PR) status was evaluated in 326 primary breast carcinomas. Nineteen percent of samples were EGF-R positive, 63% were positive for ER, and 54% for PR. In 46% of the tumors both ER and PR were positive. These data are presented together, with grading, size of tumor, lymph node involvement, histological subtype, and age. Sixty-nine percent of EGF-R negative tumors were ER-positive and 51% were positive for ER as well as PR. In particular, negative correlation between EGF-R and steroid receptor status was found. A quantitative correlation was also shown. A combination of negative steroid receptor and positive EGF-R was found more often in the population of poorly differentiated tumors. Tumors bigger than 5 cm were related to a positive EGF-R status. No correlation between nodal status and any receptor status was found. Intraductal carcinomas were more often EGF-R positive than infiltrating ductal (NOS) or infiltrating lobular lesions. The age of patients correlated with the concentration of ER only. In our study we reaffirmed the negative correlation between steroid receptor status and the overexpression of EGF-R; furthermore the combination of EGF-R+ and ER- tumors was observed more often in histological high-risk tumors. Patient outcome did not show statistically significant differences concerning the EGF-R status, but was associated with the steroid receptor status.

Serum evaluation of basic FGF in breast cancer patients.

Basic fibroblast growth factor (bFGF) is a potent endothelial cell mitogen found in a variety of normal and tumour tissues and is of prognostic relevance in human malignancies such as renal cell carcinoma and leukaemia. This study presents the data of 104 serum samples of 20 patients suffering from breast cancer. Mean serum levels of bFGF in these patients were 13.9 +/- 17. 1 (min 0, max 56.4) pg/ml and 2.4 +/- 5.9 (min 0, max 24.7) pg/ml, respectively (p = 0.01). Basic FGF reached a sensitivity of 61% at a specificity of 87% when applying a cut-off level of 5 pg/ml. A continuous increase of bFGF serum levels before the clinical detection of relapse (lead time) was seen in 3 out of 8 cases with a mean lead time of 4 months. Preoperative serum levels were not of prognostic value and showed no correlation with axillary lymph node metastasis. These preliminary results indicate that, in breast cancer patients, soluble bFGF may be useful in early detection of primary tumours, recurrences and monitoring of therapy.

Prognostic value of tumour microvessel density in cancer of the uterine cervix stage IB to IIB.

High intensity of tumour angiogenesis has been correlated with an increased potential of metastasis and poor prognosis in human malignancies. We investigated 43 patients with cervical cancer stages IB (n = 13), IIA (n = 8) and IIB (n = 22). All patients were treated by radical hysterectomy and lymphadenectomy. In the tumour specimen blood vessels were highlighted by staining endothelial cells for factor VIII. Microvessels were counted on a 200x field (0.74 mm2) in the most active areas of neovascularisation. The mean microvessel counts per field for stage IB, IIA and IIB tumours were 59.6 +/- 28.1, 56.3 +/- 24.3 and 55.7 +/- 55.6, respectively (p-value = n.s.). We found no significant correlation of microvessel density and established prognostic factors like pelvic lymph node involvement, vascular space invasion and stromal reaction. Patients with tumours showing low microvessel density (< 40 microvessels per field) had a significantly poorer recurrence-free interval (log-rank test: p-value = 0.01).

Tissue polypeptide specific antigen and cancer associated serum antigen in the follow-up of ovarian cancer.

420 clinical and serological examinations prior to surgery and during follow-up were performed in 30 patients suffering from ovarian cancer. The population consisted of three FIGO stage Ia, nine stage Ic, four stage II and fourteen stage III cases. Serous carcinoma of the ovary, mucinous carcinoma and other kinds of ovarian cancer were found in 16, 9 and 5 cases, respectively. The serum levels of the tumor markers tissue polypeptide specific antigen (TPS), cancer associated serum antigen (CASA) and carbohydrate antigen 125 (CA 125) were determined. Cut-off values of 97 U/l, 4 U/ml and 35mU/ml for TPS, CASA and CA 125, respectively, were selected according to the 95% of serum concentrations measured in healthy controls. Sensitivity, specificity, PPV and NPV of CA 125 were 75%/96%/69%/92%, respectively. Sensitivity, specificity, PPV and NPV of TPS were 67%/84%/59%/90%, respectively. CASA showed a sensitivity of 58%, specificity of 96% and a PPV and NPV of 73%/94%, respectively. The combination of TPS and CA125 increased the sensitivity to 81%, reaching a specificity of 82% and a PPV and NPV of 58/96%, respectively. The combination of CASA and CA125 showed a sensitivity, specificity, PPV and NPV of 88/85/65/96%, respectively. Twelve patients developed recurrence of disease after response to primary treatment. TPS, CASA and CA 125 detected recurrent disease in six, two and four cases, respectively. For TPS mean lead time was 4.6 months (range 2-18 months), for CASA 1.7 months (range 1-6 months), and for CA 125 3.5 months (range 1-24 months. As a matter of fact TPS never showed lead time effects in patients without elevated pretherapeutic levels. A combination of all makers showed a mean lead time of 6.72 months. Detection of recurrent disease by CA 125 is improved when CA 125 is used in combination with TPS, especially in those patients with pretherapeutically elevated TPS serum levels.

Prostate specific antigen (PSA) in breast and ovarian cancer.

Prostatic specific antigen (PSA) is commonly used for the diagnosis and monitoring of prostatic adenocarcinoma, and has recently been detected in breast cancer and it is also thought to be produced by ovarian cancer. To examine the prognostic value of PSA tumor expression, we investigated the tumor cytosols of 26 patients with breast cancer and of 16 women with ovarian cancer. We used a chemiluminescence immunoassay for the quantitative determination of PSA (detection limit 0.003 ng/ml). The median PSA level in breast cancer patients was 0.31 ng/mg (minimum 0.003, maximum 4.4). PSA expression was significantly lower in poorly differentiated breast cancers compared to moderately and highly differentiated tumors. Advanced lymph node metastases were also correlated with lower PSA expression. We found no correlation with relapse free or overall survival. Median PSA level in our ovarian cancer patient collective was 0.014 ng/mg (minimum 0.003, maximum 0.046). We found a significant correlation between PSA expression and the estrogen receptor content of the tumor. PSA expression showed no correlation with histological grading, tumor size or other tumor characteristics in ovarian cancer. There was no correlation with relapse free or overall survival in ovarian cancer patients. Our study demonstrates that PSA is expressed in ovarian malignancies but at a lower level compared to breast cancer tissue. PSA was not able to identify a subset of patients with good prognosis in breast and in ovarian cancer patients.

Humoral p53 antibody response is a prognostic parameter in ovarian cancer.

BACKGROUND: Mutant p53 protein may become the target of a tumor-specific humoral and cellular immune response. MATERIAL AND METHODS: We used a specific qualitative p53 antibody ELISA to investigate serum samples of 33 patients with ovarian cancer taken prior to therapy. Additionally, we sought to evaluate whether p53 antibodies are also present in the sera of 17 patients with benign ovarian tumors. RESULTS: p53 antibodies were detected in 36% of serum samples. There was a statistically significant association between p53 serum antibody response and poor overall survival (p < 0.006). No significant associations were found between p53 antibody status and histological type, histological grade, and tumor stage. In 81% of serum samples, no changes from p53 antibody negativity to positivity or vice versa during follow-up were observed. p53 antibodies were also detected in the sera of 18% of patients with benign ovarian tumors. CONCLUSIONS: The results of this preliminary study suggest that a p53 antibody response in patients with ovarian cancer is associated with poor prognosis. A qualitative method of p53 antibody detection cannot be used to monitor the clinical course of ovarian cancer.