Antisense-mediated suppression of human heparanase gene expression inhibits pleural dissemination of human cancer cells.

Heparan sulfate proteoglycans is a major component of the cell surface and extracellular matrix and functions as a barrier against cationic molecules and macromolecules. Heparanase is an endoglucuronidase capable of specifically degrading heparan sulfate, and its activity is associated with the metastatic potential of tumor cells. To inhibit human heparanase expression in human cancer cells, we constructed an adenoviral vector carrying a full-length human heparanase cDNA in an antisense orientation (Ad-AS/hep). Increased heparanase expression in T.Tn human esophageal cancer cells and A549 human lung cancer cells after infection with an adenovirus vector expressing the human heparanase gene (Ad-S/hep) was specifically inhibited by simultaneous infection with Ad-AS/hep in a dose-dependent manner. A modified Boyden chamber assay demonstrated that infection with Ad-AS/hep significantly inhibited in vitro invasion of A549 cells after Ad-S/hep infection. Moreover, intrathoracic administration of Ad-AS/hep reduced the number and size of heparanase-expressing A549 tumors implanted intrathoracically into BALB/c-nu/nu mice. Our results suggest that heparanase contributes to the invasive phenotype of tumor cells, and that antisense-mediated inhibition of heparanase activity may be efficacious in the prevention of pleural dissemination.

Structure of the core regions in lipopolysaccharides from Escherichia coli K12 W2252-11U-, the Ter-15 mutant, and Ter-15 (F'-lac) and Ter-15 (F+) cells.

From Escherichia coli K12 W2252-11U-cells, the Ter-15 mutant, the Ter-15 (F'-lac) and the Ter-15 (F+) cells, lipopolysaccharides were isolated and the primary structure of its core oligosaccharides was elucidated. When the F'-lac episome is transferred to the Ter-15 mutant by conjugation, the structure of the glucose III(1 leads to 3)glucose II(1 leads to 3)glucose I residue and the galactose I(1 leads to 2)-linked to the glucose I residue in the core oligosaccharide from the Ter-15 mutant changes into the structure of the glucose IV(1 leads to 6)glucose III(1 leads to 2)glucose II(1 leads to 3)glucose I residue and the galactose I (1 leads to 6)-linked to the glucose I residue in the core oligosaccharide from the Ter-15 (F'-lac) cells, but the core oligosaccharide in the Ter-15 (F+) cells is the same structure with that of the core oligosaccharide from the Ter-15 mutant when F+ episome is transferred to the Ter-15 mutant. Also, the core oligosaccharide from the Ter-15 (F'-lac) cells shows the same structure with that of the core oligosaccharide from E. coli K12 W2252-11U- cells (the parent cells). As the result, the ability to produce the structure of the core oligosaccharide in E. coli K12 W2252-11U- cells is recovered in the Ter-15 (F'-lac) cells by the dominant expression of lac gene or its containing DNA segment in F'-lac episome.

Specificity of chitosanase from Bacillus pumilus.

Partially (25-35%) N-acetylated chitosan was digested by chitosanase from Bacillus pumilus BN-262, and structures of the products, partially N-acetylated chitooligosaccharides, were analyzed in order to investigate the specificity of the chitosanase. The chitosanase produced glucosamine (GlcN) oligosaccharides abundantly, indicating that the chitosanase splits the beta-1,4-glycosidic linkage of GlcN-GlcN. The chitosanase also produced hetero-oligosaccharides consisting of glucosamine and N-acetyl-D-glucosamine (GlcNAc). Three types of the hetero-oligosaccharides purified by cation-exchange chromatography and HPLC were found to have GlcNAc residue at their reducing end and GlcN residue at their non-reducing end, indicating that the chitosanase can also split the linkage of GlcNAc-GlcN. The determination of the mode of action toward partially N-acetylated chitosan enables a classification of chitosanases according to their specificities and a more precise definition of chitosanases.

Effects of local anesthetics on the Chara plasmalemma.

The effects of lidocaine, tetracaine, procaine and bupivacaine (less than 1000 microM) on the Chara corallina internodal cell were studied. These local anesthetics depolarized the membrane at rest, while they affected the rising phase and the peak level of action potential not appreciably. Instead, they prolonged the time course of the falling phase of action potential as slowly as the repolarization was imperfect, even after enough lapse beyond the refractory period. Consequently, an action potential appeared to enhance the degree of depolarization at rest. Such a depolarization with stimulus/excitation was named use-dependent depolarization, while the depolarization without excitation, the resting one. The order of the potency of the use-dependent depolarization almost coincided with that of the nerve-blocking potency. During depolarization the change in membrane conductance was not simple. However, the conductance-voltage (Gm-Vm) relationship curve in the presence of local anesthetic suggested that depolarization was due to, not only the decrease in the electrogenic H(+)-pump, but also the increase in the diffusion conductance.

Inhibition of hyaluronan synthesis by vesnarinone in cultured human myofibroblasts.

Hyaluronan (HA), which is a major component of the extracellular matrix (ECM), is regulated during myofibroproliferative responses to numerous forms of inflammatory stimuli. It is a key factor involved in cellular migration and adherence. The development of a potent and non-toxic inhibitor of HA synthesis would open up a new avenue for the treatment of fibrocontractive diseases such as pulmonary fibrosis and liver cirrhosis. In this study, the effects of vesnarinone (OPC-8212: 3,4-dihydro-6-[4-(3, 4-dimethoxybenzoyl)-1-piperazinyl]-2(1H)-quinolinone) on the secretion of HA in human myofibroblast cell lines (MRC-5 and LI90 cells, referred to as pulmonary and hepatic myofibroblasts, respectively) were examined. Vesnarinone specifically and dose-dependently inhibited HA secretion by myofibroblasts up-regulated by fetal calf serum (FCS). The treatment of vesnarinone did not modify the phenotype of myofibroblast cells in culture. Vesnarinone also potently inhibited the HA secretion by the two myofibroblast cell lines up-regulated by transforming growth factor-beta1 (TGF-beta1) or tumor necrosis factor-alpha (TNF-alpha). The addition of vesnarinone to myofibroblasts resulted in a significant decrease of HA synthase (HAS) activity, with or without the addition of FCS or either cytokine. These findings suggest that vesnarinone inhibits the secretion of HA in myofibroblasts by specifically suppressing HAS activity, and may therefore prove useful for the treatment of chronic inflammation and tissue fibrosis.

Excessive production of transforming growth-factor beta 1 can play an important role in the development of tumorigenesis by its action for angiogenesis: validity of neutralizing antibodies to block tumor growth.

Angiogenesis is an important part of tumor growth in vivo. We used the transfected Chinese hamster ovary (CHO) cells that overproduced recombinant transforming growth-factor beta 1 (TGF-beta 1) to examine the possible role of this factor in tumor growth and angiogenesis in a nude mouse model. The in-vitro proliferation of TGF-beta 1-transfected CHO cells was unaffected by the treatment of either recombinant TGF-beta 1 or an anti-TGF-beta 1 antibody. The TGF-beta 1-transfected cells grew more rapidly than the parental CHO cells when injected subcutaneously into nude mice. The tumors derived from the TGF-beta 1-transfected cells showed prominent tumor-associated angiogenesis, whereas the parental cells produced tumors without such angiogenesis. In addition, an anti-TGF-beta 1 neutralizing antibody was able to inhibit both growth and angiogenesis in the tumors derived from TGF-beta 1-transfected cells. These findings suggest that the overproduction of TGF-beta 1 by tumor cells can contribute to neovascularization and may help promote tumor development in vivo.

Stimulation of hyaluronan synthesis by tumor necrosis factor-alpha is mediated by the p50/p65 NF-kappa B complex in MRC-5 myofibroblasts.

The lesions of fibrocontractive diseases result from an excessive myofibroproliferative response to numerous forms of inflammatory stimuli, which elicit the net deposition of extracellular matrix (ECM) in the interstitium of the affected tissue. Hyaluronan (HA), reported to be a key player supporting cellular migration and adherence, is a major component of ECM that undergoes dynamic regulation during inflammation. The molecular regulation of HA biosynthesis by inflammatory cytokines on myofibroblasts is not yet completely understood. Here we report the biochemical characteristics of the lung myofibroblast cell line MRC-5, and we demonstrate that the production of HA by this cell line is inducible by the proinflammatory cytokine, tumor necrosis factor-alpha (TNF-alpha), at the message level of HA synthase (HAS). In TNF-alpha-stimulated MRC-5 cells, DNA-binding and competition experiments indicated that the predominant NF-kappa B binding activity detected with nuclear extract-stimulated cells is mediated by the p50/p65 complex. Using antisense oligonucleotides, we confirmed that the TNF-alpha-stimulation of HA synthesis by MRC-5 cells is dependent on the activation of the p50/p65 NF-kappa B complex. These findings indicate that TNF-alpha production within inflamed tissues may enhance the HA synthesis via the transcriptional induction of HAS on myofibroblasts, thereby providing a provisional matrix for supporting cellular migration and adhesion, and that the p50/p65 NF-kappa B complex that plays an important role in the regulation of HA production by TNF-alpha might be an appropriate target for therapeutic compounds to treat tissue fibrosis accompanied by inflammation.

Induction of calponin-h1 by transforming growth factor-beta1 in cultured human ito cells, LI90.

We investigated the effect of transforming growth factor-beta1 (TGF-beta1) on the expression of calponin-h1, alpha-smooth muscle actin (alpha-SMA), and extracellular matrix (ECM) components in a cultured human Ito cell line, LI90. The TGF-beta1 treatment stimulated productions of hyaluronic acid and laminin, and significantly decreased the secretion of hepatocyte growth factor in LI90 cells. The functional characteristics of LI90 cells were compatible with those of human-activated Ito cells that are known as pericyte-like mesenchymal liver cells. TGF-beta1 induced a slight growth-inhibition of LI90 cells. TGF-beta1 enhanced the expressions of both alpha-SMA and calponin-h1 at the protein level, while tumor necrosis factor-alpha and interleukin-1alpha did not affect the expressions of these cytoskeletal proteins on LI90 cells. The addition of TGF-beta1 to LI90 cells resulted in a significant increase of calponin-h1 mRNA levels, but not calponin-h2. These data suggest that the expression of calponin-h1 is controlled at the level of mRNA under the coordinate regulation together with alpha-SMA as the process of perpetuation of activated Ito cells promoted by TGF-beta1. The identification of smooth muscle features promoted by TGF-beta1 support the hypothesis that the activation of Ito cells coincides with their contractile behavior, indicating that these cells may be important in vasoregulation during liver injury and fibrosis.

TNF combined with IFN-alpha accelerates NF-kappaB-mediated apoptosis through enhancement of Fas expression in colon cancer cells.

Immunostaining and EMSA revealed that NF-kappaB was activated strongly by TNF/IFN-alpha compared to TNF alone in a human colon adenocarcinoma cell line, RPMI4788. Although inhibition of activated NF-kappaB, by using an NF-kappaB decoy, reduced cell viability after treatment with TNF only, NF-kappaB decoy resulted in recovery of cell viability after TNF/IFN-alpha treatment. Caspase-3 activity was increased in cells induced by TNF/IFN-alpha, while suppression of caspase-3 activity was observed in cells transfected with NF-kappaB decoy and then treated by TNF/IFN-alpha. On the other hand, Fas expression was strongly enhanced by TNF/IFN-alpha, and inhibition of TNF/IFN-alpha-induced NF-kappaB activation, by using NF-kappaB decoy, decreased Fas expression. Cell viability and caspase-3 activity decreased in cells treated with TNF/IFN-alpha and anti-FasL antibody. Taken together, our findings suggest that activated NF-kappaB induced by the crosstalk between TNF and IFN-alpha is a novel pro-apoptotic signal acting via enhancement of Fas expression.

Therapeutic applications of ribozymes.

The demonstration that RNA can be cleaved by cis or trans ribozymes (catalytic RNAs, RNA enzymes) has potentially important therapeutic implications. Since their discovery in the 1980s, the biochemistry and conserved sequences of ribozymes have been well characterized. Ribozymes are effective modulators of gene expression because of their simple structure, sitespecific cleavage activity, and catalytic potential. The targets of ribozyme-mediated gene modulation have ranged from cancer cells to foreign genes that cause infectious diseases. Additional target sites for ribozymes are in initial phases of development and design. Ribozymes have been targeted against a myriad of genes, including oncogenes (ras, BCR-ABL, c-fos) and drug resistance genes, as well as the human immunodeficiency virus-type I genome. These ribozymes have cleaved the target RNAs in vitro and altered the cellular pathology. Currently, the therapeutic application of ribozymes to human diseases is limited by gene transfer systems. It is anticipated that ribozymes ultimately will play an important role in human gene therapy.

Cytotype control of Drosophila melanogaster P element transposition: genomic position determines maternal repression.

P element transposition in Drosophila is controlled by the cytotype regulatory state: in P cytotype, transposition is repressed, whereas in M cytotype, transposition can occur. P cytotype is determined by a combination of maternally inherited factors and chromosomal P elements in the zygote. Transformant strains containing single elements that encoded the 66-kD P element protein zygotically repressed transposition, but did not display the maternal repression characteristic of P cytotype. Upon mobilization to new genomic positions, some of these repressor elements showed significant maternal repression of transposition in genetic assays, involving a true maternal effect. Thus, the genomic position of repressor elements can determine the maternal vs. zygotic inheritance of P cytotype. Immunoblotting experiments indicate that this genomic position effect does not operate solely by controlling the expression level of the 66-kD repressor protein during oogenesis. Likewise, P element derivatives containing the hsp26 maternal regulator sequence expressed high levels of the 66-kD protein during oogenesis, but showed no detectable maternal repression. These data suggest that the location of a repressor element in the genome may determine maternal inheritance of P cytotype by a mechanism involving more than the overall level of expression of the 66-kD protein in the ovary.

Heparanase expression correlates with malignant potential in human colon cancer.

PURPOSE: Heparanase cleaves carbohydrate chains of heparan sulphate proteoglycans and is an important component of the extracellular matrix. This study was designed to determine the relation between heparanase expression and prognosis of patients with colon cancer. METHODS: The study included 54 patients (35 males and 19 females) who underwent colorectal resection for colorectal cancer between January 1992 and December 1994. Expression of heparanase protein and mRNA were determined and correlated with various clinicopathological parameters. In vitro studies were also performed to examine tumor invasion and to test the effects of heparanase inhibition, and in vivo studies were performed to examine tumor metastasis and prognosis. RESULTS: Heparanase expression was detected in the invasion front of the tumor in 37 of 54 (69%) colon cancer samples, whereas 17 of 54 (31%) tumors were negative. Expression of heparanase was significantly more frequent in tumors of higher TNM stage (P=0.0481), higher Dukes stage (P=0.0411), higher vascular infiltration (P=0.0146), and higher lymph vessel infiltration (P=0.0010). Heparanase expression in colon cancers correlated significantly with poor survival (P=0.0361). Heparanase-transfected colon cancer cells exhibited significant invasion compared with control-transfected colon cancer cells (P=0.001), and the peritoneal dissemination model also showed the malignant potential of heparanase-transfected cells, as assayed by number of nodules (P=0.017) and survival (P=0.0062). Inhibition of heparanase significantly reduced the invasive capacity of cancer cells (P=0.003). CONCLUSIONS: Heparanase is a marker for poor prognosis of patients with colon cancer and could be a suitable target for antitumor therapy in colon cancer.

Suppression of allograft rejection with anti-alphabeta T cell receptor antibody in rat corneal transplantation.

BACKGROUND: Anti-alphabeta T cell receptor monoclonal antibody (R73) has been reported to be a potent immunosuppressant. The suppressive effects of this antibody on allograft rejection after corneal transplantation are unknown. METHODS: Orthotopic rat penetrating keratoplasty was performed using Lewis rats as recipients and Brown Norway and Fisher rats as donors. The treated groups received R73 intraperitoneally until day 12 after the transplantation. In grafted rats with or without R73 treatment, cytokine expression of the aqueous humor, corneal-infiltrating cells, draining lymph nodes, and splenocytes was determined. Delayed-type hypersensitivity (DTH) responses were compared. RESULTS: All allografts in the untreated controls of Fisher-to-Lewis or BN-to-Lewis rat combinations were rejected within 14 days. In contrast, indefinite survival rates of the postoperative R73-treated group increased to 86% in the Fisher-to-Lewis and 23% in the Brown Norway-to-Lewis combinations, respectively. Interferon-y, interleukin (IL)-2 (T helper [Th]1), and IL-10 (Th2), but not IL-4 (Th2), expression of the eye and DTH responses in the control group were suppressed in the R73-treated group. Both IL-2 and IL-10 expression after mixed lymphocyte culture in the R73-treated group were significantly lower than those of the naive and untreated control group. CONCLUSIONS: alphabeta T cell receptor-targeted therapy prevents allograft rejection in rat corneal transplantation as evidenced by suppression of DTH responses. The cytokine profile after R73 treatment was characterized by low interferon-gamma, IL-2, and IL-10, and high IL-4 expression.

Immunohistochemical localization of pregnancy-associated plasma protein A (PAPP-A) in placentae from normal and pre-eclamptic pregnancies.

An immunohistochemical method was used to locate pregnancy-associated plasma protein A (PAPP-A) in the placenta and uterus. In addition to 10 placentae and basal plates from normal pregnancies, ranging in gestational age from 37 to 40 weeks, the following specimens were studied: three uteri obtained by hysterectomy during early pregnancy; and three placentae from patients with severe hypertensive pre-eclampsia. In early gestation, PAPP-A was localized in the villous cytotrophoblastic cell layer and the endometrial glands but was not found in the villous syncytiotrophoblast, the cytotrophoblastic cell columns or the decidual cells. On histochemical examination of placentae from cases of pre-eclampsia with hypertension, an increased number of villous cytotrophoblastic cells and so-called X-cells was observed. The monospecific antiserum to PAPP-A reacted strongly and evenly with the cytoplasm of these cells. The present study strongly suggests that the active production sites of PAPP-A are the villous cytotrophoblastic cells and the X-cells.

Clinical significance of enzyme-linked immunosorbent assay for the detection of circulating anti-BP180 autoantibodies in patients with bullous pemphigoid.

The NC16A domain of the 180-kDa bullous pemphigoid antigen (BP180) is the most immunogenic and, probably, pathogenic region in bullous pemphigoid (BP). In the present study, in order to determine whether serum level of circulating anti-BP180 autoantibodies is a valuable serum marker in BP, the immunoreactivity of sera against the NC16A domain of BP180 was measured using enzyme-linked immunosorbent assay (ELISA) in ten patients with BP. Serum levels of anti-BP180 autoantibodies correlated with the clinical course in BP patients, who received various therapeutic agents. The result suggests that this NC16A-ELISA is a useful method for evaluating the clinical course and efficacy of the therapy in patients with BP.

Results of adjuvant chemotherapy for invasive uroepithelial cancer.

Between June 1982 and July 1990, 55 patients (41 with bladder cancers and 14 with renal pelvic or ureteral cancers) who had undergone radical extirpative surgery and/or node dissection for pathological stage pT2-4 and/or nodal disease received adjuvant chemotherapy consisting of cisplatin alone or in combination with other agents. In all, 26 of the bladder-cancer patients also received preoperative chemotherapy consisting of arterial infusion of cisplatin, mitomycin C, and Adriamycin. Adjuvant chemotherapy was performed according to the following protocol. Between June 1982 and July 1987, 30-50 mg/m2 cisplatin either alone or in combination with Adriamycin and 5-fluorouracil (CAF) was given to 35 patients in an induction and maintenance setting for 1 year. After July 1987, short-course cisplatin (70 mg/m2) or cisplatin, etoposide, and Adriamycin combination chemotherapy (CVA) was given to 20 patients. Of the 55 patients, 38 are alive and show no evidence of disease, three are alive with disease, 13 have died of their disease, and 1 has died of an unrelated cause. The 5-year survival of all patients was 65.1%. The survival of the 20 patients who were treated after July 1987 was better than that of the 35 patients who were treated before June 1987. Local recurrence and/or distant dissemination occurred in 16 patients, 13 of whom died of cancer progression. Nausea and vomiting and anorexia occurred in most patients during the administration of cisplatin. Mild to moderate myelosuppression developed in patients who received CAF or CVA combination chemotherapy. Although adjuvant chemotherapy combined with radical surgery seemed to be effective in cases with a pathological stage of pT3a or less, more intensive pre- or postoperative chemotherapy is needed to improve the poor prognosis of patients with deeply invasive uroepithelial cancer.

The effects of intravesical chemoimmunotherapy with epirubicin and bacillus Calmette-Guérin for prophylaxis of recurrence of superficial bladder cancer: a preliminary report.

The effects of intravesical chemoimmunotherapy with epirubicin and bacillus Calmette-Guérin (BCG) for prophylaxis of recurrence of superficial bladder cancer (pTa, pT1) were investigated in 29 patients aged a median of 70 years between January of 1991 and May of 1993. The patients received intravesical instillation of 40 mg epirubicin immediately after transurethral resection (TUR) of the bladder cancer. At 1 week after TUR, 80 mg Tokyo-strain BCG was instilled into the bladder once a week for 6 weeks. Thereafter, the patients were followed by cystoscopy and urinary cytology at 3-month intervals until recurrence was detected. Of the 29 patients, 28 had no evidence of disease over a mean follow-up period of 20 months. The 1 case of recurrence occurred at 3 months after TUR and that patient died of cancer progression. The simple recurrence rate was 3.5% after therapy. According to the person-years method, the number of recurrent tumors per 100 patient-months was 0.17. The cumulative nonrecurrence rate determined for all cases was 96.5% at 30 months. Adverse reactions, including urinary frequency, urgency, and miction pain, among others, were observed in 27 patients (93%). Only 1 patient was withdrawn from the treatment because of severe bladder-irritation symptoms due to the BCG instillation. The intravesical chemoimmunotherapy with epirubicin and BCG seemed to be effective for prophylaxis of recurrence of superficial bladder cancer.

Adjuvant chemotherapy for invasive bladder cancer.

From June 1982 through December 1985, 25 patients who had undergone radical cystectomy with pelvic node dissection for pathologic stage-pT3 or -pT4 and/or N+ disease received adjuvant chemotherapy involving the injection of cis-platinum alone or in combination with adriamycin and 5-fluorouracil (CAF). Thirteen patients also received preoperative adjuvant chemotherapy involving the infusion of cis-platinum, adriamycin, and mitomycin C into the bilateral internal iliac arteries. Postoperative adjuvant chemotherapy was performed using the following two protocols. Protocol 1 (18 cases) consisted of cis-platinum alone being administered every week for 3 weeks and then every month for 1 year. In protocol 2 (7 cases), cisplatinum, adriamycin, and 5-fluorouracil were administered at 3-week intervals on three occasions and then every month for 1 year. Eighteen patients were still alive with no evidence of disease after an average of 26 months. One patient died as a result of factors unrelated to cancer. Local recurrence and distant metastasis occurred in 6 patients, of whom 3 were still alive for an average of 20.7 months. Three patients died of cancer progression after 9, 19, and 21 months. The survival rate for all 25 patients at 50 months was 77%. Nausea and vomiting occurred in most patients during the administration of cis-platinum. Mild myelosuppression developed in a few patients subjected to protocol 2. Our results indicate that adjuvant chemotherapy consisting of the administration of cis-platinum alone or in combination with other chemotherapeutic agents appears to be effective in patients with invasive bladder cancer.

Release of nitric oxide together with carbon-centered radicals from N-nitrosamines by ultraviolet light irradiation.

Solutions of N-nitrosamines, N-nitrosodimethylamine, N-nitrosodiethylamine, N-nitrosomorpholine and N-nitrosopyrrolidine in phosphate buffer (pH 7.4) were irradiated by ultraviolet (UV) light at room temperature. The N-nitrosamines were extensively degraded due to irradiation for 120 min in a time-dependent fashion as monitored by UV-absorption or high performance liquid chromatographic analysis. Carbon-centered radicals were generated from four N-nitrosamines during the short time irradiation of 10-60 s as monitored by electron spin resonance (ESR) technique using 5,5-dimethyl-1-pyrroline N-oxide and N-tert-butyl-alpha-phenylnitrone as spin traps. Nitric oxide (NO) was generated during the short time irradiation as monitored by ESR technique using cysteine-Fe(II) complex, N-methyl-D-glucamine dithiocarbamate and 2-(4-carboxyphenyl)-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide. Significant amounts of nitrite (4-16%) from four N-nitrosamines and also a significant amount of nitrate (4%) was produced from N-nitrosodimethylamine during the irradiation time of 120 min. Released NO from the N-nitrosamines must be converted into nitrite through intermediary reactive nitrogen oxide species including nitrogen dioxide and dinitrogen trioxide in contact with dissolved oxygen.

Cancer-associated retinopathy in a patient with endometrial cancer.

PURPOSE: We studied a case of unusual retinopathy in a 60-year-old woman. METHOD: The patient developed bilateral retinopathy with iridocyclitis and vitreitis. Corticosteroid therapy could not prevent deterioration of visual acuity and field. The electroretinogram became nonrecordable, and fluorescein angiogram exhibited retinal pigment epithelium window defects, fluorescein stainings to the vascular walls, and narrowed arterioles. RESULTS: The patient underwent hysterectomy and lymphadenectomy for endometrial cancer in the corporis uteri. She was found to possess antibodies against retinal 34-kd protein, leading to a diagnosis of cancer-associated retinopathy. CONCLUSION: Cancer-associated retinopathy may occur in patients with endometrial cancer of the corporis uteri.