Association between nonrandom X-chromosome inactivation and BRCA1 mutation in germline DNA of patients with ovarian cancer.

BACKGROUND: Most human female cells contain two X chromosomes, only one of which is active. The process of X-chromosome inactivation, which occurs early in development, is usually random, producing tissues with equal mixtures of cells having active X chromosomes of either maternal or paternal origin. However, nonrandom inactivation may occur in a subset of females. If a tumor suppressor gene were located on the X chromosome and if females with a germline mutation in one copy of that suppressor gene experienced nonrandom X-chromosome inactivation, then some or all of the tissues of such women might lack the wild-type suppressor gene function. This scenario could represent a previously unrecognized mechanism for development of hereditary cancers. We investigated whether such a mechanism might contribute to the development of hereditary ovarian cancers. METHODS: Patterns of X-chromosome inactivation were determined by means of polymerase chain reaction amplification of the CAG-nucleotide repeat of the androgen receptor (AR) gene after methylation-sensitive restriction endonuclease digestion of blood mononuclear cell DNA from patients with invasive (n = 213) or borderline (n = 44) ovarian cancer and control subjects without a personal or family history of cancer (n = 50). BRCA1 gene status was determined by means of single-strand conformational polymorphism analysis and DNA sequencing. All statistical tests were two-sided. RESULTS AND CONCLUSIONS: Among individuals informative for the AR locus, nonrandom X-chromosome inactivation was found in the DNA of 53% of those with invasive cancer versus 28% of those with borderline cancer (P = .005) and 33% of healthy control subjects (P = .016). Nonrandom X-chromosome inactivation can be a heritable trait. Nine of 11 AR-informative carriers of germline BRCA1 mutations demonstrated nonrandom X-chromosome inactivation (.0002 < P < .008, for simultaneous occurrence of both). IMPLICATIONS: Nonrandom X-chromosome inactivation may be a predisposing factor for the development of invasive, but not borderline, ovarian cancer.

Cyclosporin A reverses chemoresistance in patients with gynecologic malignancies.

Multidrug resistance is a major obstacle in successful systemic therapy of gynecologic malignancies. The objectives of this study are to evaluate the activity of cyclosporin A used to overcome drug resistance in a variety of gynecologic malignancies. Forty women (29 with ovarian cancer, 7 with uterine cancer, 3 with cervical cancer, and 1 with choriocarcinoma) were treated with cyclosporin A, 4 mg/kg intravenously, 6 hours before and 18 hours after the specific chemotherapeutic agent, to which the tumor had developed drug resistance. All patients had shown resistance to the chemotherapy agent used in combination with cyclosporin A. All patients had been heavily pretreated (mean, 2.8 previous chemotherapy regimens). Overall, among 38 available patients with gynecologic malignancies, a 29% objective response rate was observed. Twenty-six (65%) of all patients received three or more cycles of cyclosporin A. There was a 25% response rate for patients with ovarian cancer patients and 50% for those with uterine cancer. There were no responses among the three patients with cervical cancer, and the patient with choriocarcinoma had a complete response. All patients were evaluable for toxicity. Leukopenia and nausea were the most common toxic reactions, but in most cases they were transient, and only three patients required a treatment delay. The most common grade 3 or 4 toxicity was thrombocytopenia, which was observed in 22% of the patients. Cyclosporin A is well tolerated and has significant potential for reversal of chemoresistance in heavily pretreated patients with ovarian and uterine malignancies.

Colorectal cancer complicating pregnancy.

Colorectal carcinoma complicating pregnancy is rare but associated with a high maternal mortality rate. Stage for stage the survival data are the same for pregnant patients and nonpregnant controls, however diagnosis is often delayed due to pregnancy-associated gastrointestinal symptoms masking cancer symptoms. A high degree of suspicion, especially in high-risk patients, is the key to early detection and improved prognosis. Digital rectal exams, occult fecal blood tests, and flexible sigmoidoscopy or colonoscopy are performed as indicated. Treatment primarily consists of en-bloc resection of the malignancy and regional lymph nodes; timing of colorectal surgery and delivery are based upon clinical presentation, fetal age, and maternal desires.

Field cancerization: why late "recurrent" ovarian cancer is not recurrent.

OBJECTIVE: Late "recurrence" of ovarian cancer may result from either regrowth of dormant tumor cells or from development of a new cancer caused by the phenomenon of field cancerization. Clinically, some recurrent ovarian cancers show the same therapeutic sensitivities to chemotherapy and surgery as did the primary disease, whereas others are refractory to all therapy. We hypothesize that recurrent ovarian cancers are distinguishable on the basis of a molecular genetic fingerprint and that some are actually new primary cancers of the peritoneum rather than recurrent ovarian cancer. STUDY DESIGN: We constructed molecular genetic fingerprints of 13 paired primary and late recurrent ovarian cancers to study their clonal relationships. The tumor pairs were analyzed for p53 mutations and allelotypes, patterns of X-chromosome inactivation, loss of heterozygosity, and microsatellite instability at 12 different loci on 6 different chromosomes. Techniques used included single-strand conformational polymorphism mutation screening and polymerase chain reaction-based sequence analysis of the p53 locus, restriction digestion of the androgen receptor locus to determine X-chromosome inactivation, and polyacrylamide gel electrophoresis of highly polymorphic dinucleotide, trinucleotide, and tetranucleotide repeats. RESULTS: The average age at initial diagnosis for this cohort was 54.7 years (range 45.3 to 65.5). Mean interval to recurrence was 42.7 months (range 28 to 62). Molecular fingerprints were characterized for 4 to 8 informative loci per tumor pair. The fingerprints of 10 (77%) differed significantly, strongly suggesting that a second primary cancer had developed. The remaining 3 tumor pairs demonstrated identical allelotypes consistent with regrowth of dormant tumor cells. CONCLUSION: Our results are consistent with the "field cancerization" hypothesis of ovarian carcinogenesis but could also be explained by a polyclonal tumor origin, which contrasts with the currently accepted monoclonal theory of ovarian carcinogenesis. Late development of a new primary cancer may herald the proband as a member of a familial cancer phenotype. These studies provide a molecular genetic rationale that both explains and prognosticates the clinical course of recurrent ovarian cancer.

Stage IIIC ovarian leiomyosarcoma in a premenopausal woman with multiple recurrences: prolonged survival with surgical therapy.

A case of Stage IIIC primary ovarian leiomyosarcoma in a premenopausal woman with multiple recurrences alive and well 7 years after diagnosis is presented. In addition to the typical light microscopic, immunohistochemical, and electron microscopic features of ovarian leiomyosarcoma, the tumor was progesterone receptor positive. This is 28th report of primary ovarian leiomyosarcoma and the first report of progesterone receptor in this tumor. This is the longest reported survival in a woman with this disease.

Ovarian cancer genomic instability correlates with p53 frameshift mutations.

We hypothesize that genomic instability plays an important role in causing specific types of p53 mutations in ovarian cancer. To test this hypothesis, 78 tumors were analyzed for p53 mutations with SSCP analysis of the entire open reading frame. At the same time, alterations in 10 microsatellite loci including di-, tri-, and tetranucleotide repeats were evaluated. Fourteen (26%) of all mutations were insertion/deletion mutations. All insertion/deletion mutations were associated with one of the following features: runs of purines or pyrimidines, repeats of short sequences, or palindromes. There was a strong association of generalized microsatellite instability with p53 in contrast to tumors with other types of mutations or wild-type p53 (P = 0.007). These characteristic p53 mutations appear to be caused by generalized genomic instability rather than to be the direct cause of genomic instability. These findings suggest the existence of additional novel DNA repair genes important to the carcinogenic process.

An abundance of p53 null mutations in ovarian carcinoma.

Ovarian cancers from 64 midwestern US women were screened for p53 dysfunction both by immunohistochemical staining (IHCS) and single strand conformation polymorphism (SSCP) analysis of the entire open reading frame (ORF). Forty SSCP abnormalities in 39 tumors included nine deletion, one insertion, two splice junction, two nonsense, one silent and 25 missense mutations were confirmed by direct genomic sequencing. Eight of the insertion/deletion defects may have occurred due to slippage during the course of DNA replication. This observation suggests that genomic instability may play an important role in ovarian carcinogenesis. Fifteen percent of the mutations encountered were located outside exons 5-9 and four of these were null. The sensitivity of IHCS was 96% for missense mutations but only 14% for null mutations. This contrasted with 100% sensitivity of the SSCP screening methodology. The 21% overall incidence of null mutations in the present study far exceeds the reported 6.8% incidence in the world literature (P=0.0003). Explanations for this difference include: (1) our complete analysis of the entire ORF of the p53 gene; (2) the tendency of others to rely upon IHCS to screen tumors prior to mutation analysis; and (3) environmental or endogenous genetic influences.

p53 allelotypes and enhanced detection of allelic loss in ovarian cancer: lack of correlation with familial and clinical factors.

To develop a mapping strategy for improved detection of p53 allelic loss associated with ovarian carcinoma, we utilized multiple intragenic polymorphisms and single-strand conformation polymorphism (SSCP) analysis. p53 allelotype distributions were defined for 80 ovarian cancer patients from germ-line DNA. All polymorphic sites studied had polymorphism information content (PIC) rates of greater than 0.25. Tumor loss of heterozygosity (LOH) was determined from informative polymorphisms and migratory shifts on SSCP screening of exons 5-9. Of the four polymorphisms analyzed, the intron 1 (alu) was the most informative (PIC = 0.66). A novel allele of 110 base pairs was found in 4.4% of our ovarian cancer cohort at this site. The intron 3 (16-base pair repeat) and intron 6 (MSP1) polymorphisms were in relative equilibrium; thus, we chose to use only the intron 3 polymorphic site in the mapping strategy. Family cancer history did not influence the allelotype distribution frequencies. Overall, 56 of 61 tumors (91.8%) were informative for allelic loss, and LOH was observed in 66.1 %. A reduction to homozygosity at the p53 locus did not correlate with familial or clinical factors. These observations are consistent with the multiple mechanisms by which p53 dysfunction can occur.

p53 gene mutation analysis and antisense-mediated growth inhibition of human ovarian carcinoma cell lines.

Targeting dysfunctional gene expression in the cancer cell with gene-specific therapeutics requires knowledge of the structure and expression of the designated gene. Because of the prevalence of p53 dysfunction in epithelial ovarian carcinoma, modulation of the expression of this tumor suppressor gene is an attractive target for gene therapy. We sequenced the p53 gene and analyzed its expression in 10 ovarian cancer cell lines. Only five cell line mutations were encountered, three associated with a loss of heterozygosity. Thus, neither p53 mutation nor allelic loss is required for ovarian carcinogenesis or propagation of ovarian cancer cell lines in vitro. SSCP screening, but not immunohistochemical staining, correlated with results of direct genomic sequencing. All p53 immunohistochemical-negative cell lines differed from that reported by another laboratory, underscoring the importance of the knowledge of target gene expression in a given cell line in a given laboratory. We designed pilot studies of antisense oligodeoxynucleotides directed against the p53 gene based on our sequence data. Differential growth inhibition of the A2780-CP-20 cell line (mutant p53 protein), but not of the OVCAR-3 cell line (wild-type p53 protein) confirmed the potential usefulness of this strategy.

Absence of significant germ line p53 mutations in ovarian cancer patients.

Germ line mutations of the p53 gene have been described in the Li-Fraumeni Cancer Family Syndrome and occur in patients with multifocal gliomas, particularly those with a history of a metachronous cancer or a family history of cancer. p53 dysfunction is often associated with ovarian cancer. Patients with ovarian carcinoma frequently develop synchronous or metachronous cancers and may have a family history of this or related cancers. Thus, we hypothesized that germ line p53 mutations might be associated with a significant proportion of ovarian cancers. Germ line DNA isolated from peripheral leukocytes of 73 patients with ovarian carcinoma was screened for p53 sequence abnormalities utilizing single-strand conformation polymorphism analysis and direct PCR sequencing techniques. As many as 40% of this cohort of ovarian cancer patients from 67 families may represent familial phenotypes. Synchronous and metachronous cancers occurred in 19% of the cohort. Only two intron-based polymorphisms were found. Neither has been previously reported. One of these, in intron 6, occurred in three unrelated patients all of whom had a history of metachronous breast cancer. A polymorphism in intron 10 occurred in a patient with synchronous endometrial cancer. No classic germ line mutations of p53 were found.

The use of quantified maximum entropy methods for optimising information from electron paramagnetic resonance spectroscopy.

A quantified maximum entropy method is applied to the optimisation of analytical information from EPR spectra of free radicals. Statistically meaningful errors are produced for the positions and intensities of all spectral peaks and considerable improvements in sensitivity compared with conventional spectral enhancement procedures are obtained with measurements of the intensities of spectra of known radicals.

Isolation and preliminary characterization of an ovarian carcinoma cell line from a patient with familial ovarian cancer.

An epithelial ovarian cancer cell line is established from a patient with recurrent familial ovarian cancer. Two of the patient's sisters and her mother have also had ovarian cancer. The histological resemblance of the cell line to the patient's Stage IV, Grade 3 papillary serous ovarian primary cancer is striking. The cell line does not secrete CA125 and is estrogen and progesterone receptor negative. Overexpression of the p53 tumor suppressor gene but not the HER-2/neu oncogene was detected by immunohistochemical analysis. An unusual chemosensitivity to cisplatin, doxorubicin, etoposide, and taxol is demonstrated, suggesting that a chemosensitivity mechanism might explain prolonged survival of some patients with familial ovarian cancers. This truly unique cell line should prove invaluable in the further evaluation of molecular genetic changes associated with familial ovarian cancers.

Isotope computed tomography using cone-beam geometry: a comparison of two reconstruction algorithms.

A CT scanner has been constructed specifically to determine the three-dimensional distribution of bone mineral in the medullary cavities of the radius, ulna and femur. A source of x-rays (153Gd) and a multiwire proportional counter (MWPC) are mounted at opposite ends of a diameter of an annular mounting. The limb is placed on the axis of rotation of the annulus and a series of two-dimensional transmission projections are obtained at equal angular spacings over 360 degrees. The distribution of bone mineral is reconstructed from the projections either by the method of maximum entropy (ME) or by convolution and back projection (CBP). These two methods have been evaluated by reconstructing a single slice of a phantom, representing the forearm, from projections simulated by computer. With a clinically acceptable exposure time, the mean medullary densities of the ulna and radius were determined with systematic errors of less than 3.5% (ME) and 11% (CBP), although for the latter method of reconstruction the systematic error was reduced to less than 2% by increasing the number of views. The mean medullary densities of the ulna and radius were determined with precisions better than 2.5% (ME) and 3.5% (CBP).

Estimation of the relative avidity of 19S IgM rheumatoid factor secreted by rheumatoid synovial cells for human IgG subclasses.

19S IgM rheumatoid factors (RF) may play an important role in sustaining inflammation in rheumatoid arthritis (RA). As yet, no unique antigenic specificity for RF in RA has been identified. Because the synovium is central to the pathogenic changes in RA, RF produced therein might be pathogenically more important than serum RF. Therefore, we examined the reactivity and relative avidity of 19S IgM-RF in serum and rheumatoid synovial cells (RSC) from 20 patients with seropositive RA. Reactivities were determined by competitive inhibition of serum RF hemolytic activity and RSC RF-plaque-forming cells (PFC) by added soluble antigen, i.e., monomeric human IgG subclasses. Estimation of relative avidities of RSC RF for human IgG subclasses was done by calculation of fractional RF expression in the RSC RF-PFC assay following inhibition by IgG subclasses. RSC RF had greatest reactivity with IgG3 and IgG1, some reactivity with IgG2, and the least reactivity with IgG4. Serum RF reacted most with IgG1 and IgG2, reacted some with IgG4, but reacted poorly with IgG3. The antigenic determinants with which RSC RF reacted were common to many IgG3 molecules. The highest relative avidity of RSC RF was for IgG3. These observations indicate a selective deficiency of serum RF compared with RSC RF and suggest an important pathogenic role for these qualitatively different RSC RF molecules for in situ RF immune complex-mediated inflammation in RA synovial tissue.