An experimental model of retinal pigment epithelial and neurosensory serous detachment

PURPOSE

The authors document an animal model of retinal pigment epithelial (RPE) detachments.

METHODS

N-ethylmaleimide (NEM) and sodium iodate (sulfhydryl-alkylating agents) were injected into the vitreous of Dutch pigmented rabbits. Subretinal fluid samples were withdrawn 0.5 and 3 hours after NEM injections and analyzed for albumin content by gel electrophoresis. Selected eyes were examined histologically.

RESULTS

Injections of 9.2 micrograms NEM caused wrinkling of the retina within 15 minutes, at which time fluorescein angiography revealed multifocal RPE detachments. At this stage, histologic sections showed moderate choroidal edema and RPE detachment. Elevated retinal detachments developed over the next 15 to 25 minutes. Histologic sections showed massive choroidal edema and RPE fragmentation. The albumin concentration of subretinal fluid rose from 60% of serum level at 30 minutes after NEM to 80% after 3 hours. Sodium iodate did not cause choroidal edema or RPE detachment.

CONCLUSION

In rabbits, intravitreal NEM causes the rapid appearance of choroidal edema and RPE detachment, followed by serous retinal detachment. The initial effect probably is caused by an alteration of choroidal vascular permeability. The relationship of these effects to sulfhydryl alkylation is unclear because sodium iodate failed to produce RPE detachments. N-ethylmaleimide effects may model aspects of clinical RPE and serous retinal detachments.

Cytochalasin D reversibly weakens retinal adhesiveness

This study asks whether retinal adhesiveness is affected by cytochalasin D, a drug that is known to alter the apical morphology of the retinal pigment epithelium (RPE). Cytochalasin D was injected intravitreally in Dutch rabbits and retinal adhesiveness measured 0.5 to 72 h later by two methods: in vitro peeling of the retina from retinal pigment epithelium to observe the amount of adherent pigment, and in vivo measurement of the pressure needed to achieve retinal separation. Electroretinograms were recorded, and RPE apical morphology was examined by scanning electron microscopy. The injection of 60 microM cytochalasin D caused in vitro retinal adhesiveness to fall within 3 h to 10% of normal although the electroretinogram (a, b, and c-waves) remained normal. Smaller doses of cytochalasin D had a lesser effect. The RPE apical surface at 3 h showed large bullet-like microvilli, swollen cone sheaths, and an absence of filamentous microvilli. The severity of these changes was dose-related. At 72 h after cytochalasin D, retinal adhesiveness had largely recovered, and RPE apical morphology appeared normal again. Thus, cytochalasin D weakens retinal adhesiveness acutely but reversibly, and both the initial effect and recovery correlate with changes in RPE microvillar morphology. This suggests that actin microfilaments may be involved in mechanisms of retinal adhesion.

The interphotoreceptor matrix mediates primate retinal adhesion

OBJECTIVE

To assess the participation of cone matrix sheaths, which are domains of the cone photoreceptor-associated interphotoreceptor matrix that extend from the neural retina to the surface of the retinal pigment epithelium (RPE), in retinal adhesion.

METHODS

Monkey and human retinas were partially peeled from the RPE, and the tissues were examined by lectin histochemistry to determine the effects of physical separation on the cone matrix sheath.

RESULTS

A firm attachment of cone matrix sheaths to both the RPE and the neural retina that was strong enough to cause detachment of sheets of RPE cells from Bruch's membrane or tearing of the sheaths as a result of retinal peeling was observed. Cone matrix sheaths can stretch considerably and contract following tearing. Their integrity was compromised rapidly after the first postmortem minute.

CONCLUSION

Cone matrix sheath glycoconjugates are likely to play a major role in mediating retinal adhesion by forming a molecular bridge between the neural retina and the RPE.

The metabolic dependency of retinal adhesion in rabbit and primate

OBJECTIVE

To determine the oxygen and glucose dependency of retinal adhesion in primate and rabbit.

METHODS

Experiments were performed on Dutch rabbits and monkeys. Retinal adhesiveness was measured by peeling the retina from the retinal pigment epithelium in vitro, under different conditions of PO2 and glucose supply, and by observing the amount of adherent pigment. In vivo ischemia was produced by raising the intraocular pressure.

RESULTS

Retinal adhesion failed quickly at low oxygen tensions, but a well-oxygenated solution preserved strong retinal adhesion in vitro for 15 to 20 minutes in rabbit tissue and up to 50 minutes in primate tissue. Ischemic adhesive failure was reversible on raising the PO2. Glucose levels did not affect adhesiveness. Ischemia in vivo for more than 1 minute caused rabbit retina to lose its adhesiveness.

CONCLUSIONS

Retinal adhesion is continually and reversibly dependent on oxygenation, and probably on aerobic metabolism. Primate tissue is more resistant to metabolic adhesive failure than is rabbit tissue, but the metabolic requirements appear qualitatively similar.

The rate and source of albumin entry into saline-filled experimental retinal detachments

PURPOSE

To investigate the rate and source of albumin entry into experimental nonrhegmatogenous detachments.

METHODS

Detachments were made in Dutch rabbits by injecting Hanks' balanced salt solution into the subretinal space through a micropipette. Subretinal fluid was withdrawn 0 to 4 hours later through a similar micropipette and analyzed for osmolality and albumin content (by gel electrophoresis). Sodium iodate was injected intravenously in some rabbits to damage the retinal pigment epithelium (RPE). In some rabbits fluorescein isothiocyanate albumin (FITC-albumin) was injected intravitreally or intravenously to measure its entry into the subretinal fluid by fluorophotometry. Results from 4 to 8 eyes were averaged for each data point.

RESULTS

The albumin concentration and total amount of albumin in the subretinal fluid increased steadily over 4 hours in retinal detachments initially filled with Hanks' solution. Pretreating rabbits with sodium iodate injection resulted in a 50-fold increase in the rate of albumin entry, although the levels were still low relative to those of serum. Intravitreal FITC-albumin entered the subretinal fluid at a rate independent of sodium iodate damage, but intravenous FITC-albumin only entered the subretinal space after RPE damage. Subretinal fluid osmolality remained within the range of 291 to 294 mOsm/kg, irrespective of sodium iodate damage or differences in the rate of fluid absorption.

CONCLUSIONS

These results indicate that albumin can diffuse into the rabbit subretinal space from both vitreous and bloodstream, although entry from serum requires damage to the RPE. Subretinal fluid appears to be transported actively (control eyes) or passively (iodate-damaged eyes) out of the subretinal space, despite albumin entry and without major osmolar shifts.

Conditions necessary for the formation of serous detachment. Experimental evidence from the cat

OBJECTIVE

To evaluate experimentally the conditions necessary for the formation of serous detachments.

METHODS

Selective injury to the retinal pigment epithelium (RPE) and choriocapillaris was produced in cats using weak laser burns and intense diffuse light, with or without photosensitization with rose bengal. The fundi were analyzed by observation, fluorescein angiography, and histologic examination.

RESULTS

Injuring the RPE alone did not cause detachments. Focal injury to the RPE and choriocapillaris caused moderate detachment only when an area surrounding the focal leakage site had been subjected to light damage. Diffuse injury to the RPE and choriocapillaris caused broader detachments.

CONCLUSIONS

Three conditions are necessary for serous detachments to form: (1) a source of fluid pressure, (2) a defect in the blood-retinal barrier (entry site), and (3) an area of impaired fluid transport beyond the site of leakage (if fluid is to accumulate over a broader area).

Retinal adhesiveness in the monkey

PURPOSE

To determine factors that influence retinal adhesion in the primate and compare these with previous data from the rabbit.

METHODS

Retinal adhesiveness was studied in monkey eyes immediately after enucleation. The retina was peeled manually from the retinal pigment epithelium, and the amount of pigment that remained adherent to the retina was used as an index of adhesiveness.

RESULTS

The rate of post mortem failure of retinal adhesiveness in the monkey was less than in the rabbit under similar conditions. However, as in the rabbit, adhesiveness was sensitive to temperature, pH, and the concentrations of calcium and magnesium, and subretinal injections of neuraminidase weakened adhesion beyond the injection sites.

CONCLUSIONS

Mechanisms of retinal adhesion are similar in primates and rabbits.

Retinal adhesiveness in surgically enucleated human eyes

PURPOSE

Retinal adhesion in rabbits and monkeys has been shown to involve the interphotoreceptor matrix and depend on environmental factors, such as temperature, pH, and calcium concentration. This study was designed to see whether these same factors modulate retinal adhesiveness in humans.

METHODS

Four surgically enucleated human eyes were studied. Strips of eyecup were placed in experimental baths shortly after enucleation to evaluate environmental conditions. Retinal adhesive strength was quantified by the amount of retinal pigment epithelial pigment adherent to retina peeled from strips of eyecup. Cone and rod sheaths of the interphotoreceptor matrix were studied by microscopic examination of samples stained with the lectins peanut and wheat germ agglutinin.

RESULTS

Retinal adhesiveness fell to 0 approximately 40 minutes after enucleation. Adhesive strength was maintained by cold temperature and was weakened by low pH or low calcium in the bath. Cone and rod matrix sheaths became stretched (indicating a bond) as outer segments and retinal pigment epithelium were peeled apart.

CONCLUSION

Retinal adhesion in humans appears to involve similar mechanisms as in rabbits and monkeys.

Oxygen dependency of retinal adhesion

PURPOSE

Normal retina is firmly attached to the retinal pigment epithelium, but the force of this adhesion drops precipitously within the first 2-3 min after enucleation. The purpose was to study metabolic factors that might be relevant to this postmortem failure of adhesion.

METHODS

Dutch rabbit retina was manually peeled from the retinal pigment epithelium on strips of enucleated eyecup within a 37 degrees C bath. Retinal adhesiveness was measured by observing the amount of retinal pigment epithelium that remained adherent to the retina.

RESULTS

Autologous whole blood in place of salt solution retarded the decrease in adhesiveness. A solution of hemoglobin alone was similarly effective, whereas methemoglobin solution failed to help the persistence of retinal adhesion. Bubbling oxygen into the salt solution and circulating it to avoid oxygen depletion at the tissue boundary also proved effective at sustaining retinal adhesiveness. Eyes made ischemic in vivo for 5 min or longer, by elevating intraocular pressure, showed virtually no retinal adhesion when enucleated immediately thereafter. However, eyes made ischemic for 10 min, but allowed to regain circulation for 5 min before enucleation, showed a return of retinal adhesiveness to 80% of normal.

CONCLUSIONS

Oxidative metabolism is critical to the maintenance of retinal adhesiveness, and the effects of oxygen deprivation on adhesion are reversible within a certain time period.

Nuclear morphometry and its prognostic significance in laryngeal squamous cell carcinomas

The five parameters including nuclear area (NA), nuclear perimeter (NP), nuclear diameter (ND), axis ratio and nuclear roundness were evaluated with MIAS200 Image Analysis System (Sichuan University, Chengdu) in 45 patients with laryngeal primary squamous cell carcinoma, which were confirmed pathologically and operated on from January 1977 to October 1987. The results showed that NA, NP and ND are important prognostic indicators. The survival rate in patients with large nuclei (NA greater than 73 microns2, NP greater than 32.5 microns, ND greater than 12 microns) was significantly lower than that in those with small nuclei (NA less than 73 microns2, NP less than 32.5 microns, ND less than 12 microns) (P less than 0.0025). These findings strongly suggest that patients with laryngeal squamous cell carcinoma with large nuclei should be treated more aggressively.

Induction of serous retinal detachment in rabbit eyes by pigment epithelial and choriocapillary injury

Serous retinal detachments have been produced experimentally by occluding the choroidal circulation. We sought a more realistic model by adjusting rose bengal photosensitization to injure the retinal pigment epithelium and choriocapillaris in rabbits without causing vascular occlusion. We accomplished this by either using low doses of rose bengal or cooling the animal to reduce ocular temperature by 4 degrees C. Detachments typically appeared within 1 day after light exposure, centered over the area of light damage. Fluorescein angiography showed no capillary occlusion, but there was leakage of dye from areas of light damage. Light microscopy of light-damaged areas showed swelling or disruption of retinal pigment epithelial cells but a patent choriocapillaris. These data suggest that a combination of retinal pigment epithelial and choriocapillary damage may be necessary for serous detachment to occur.

Recovery of retinal adhesion after enzymatic perturbation of the interphotoreceptor matrix

Previous investigations established that focal subretinal injections of neuraminidase, chondroitinase, and hyaluronidase in the rabbit lead to a diffuse loss of retinal adhesiveness beyond the site of injection. This loss of adhesiveness, measured by peeling of the retina immediately after enucleation, correlates with changes in the interphotoreceptor matrix (IPM), as monitored by lectin histochemistry. In this study, rabbits were evaluated during recovery of retinal adhesiveness after subretinal injections of neuraminidase and chondroitinase. Adhesion recovered steadily 5-20 days after chondroitinase injection. After administration of neuraminidase, adhesion remained low for approximately 14 days but recovered to normal by 20 days. The recovery of adhesiveness correlated closely with reestablishment of the normal distribution of peanut agglutinin-binding glycoconjugates in the IPM, one group of molecules thought to participate in retinal adhesion. Electroretinography and light microscopy showed no abnormalities in the retina or retinal pigment epithelium after recovery. These results suggest that IPM glycoconjugates participate in maintaining retinal adhesion.

[The expression of nuclear estrogen receptor and its relation to cytoplasmic estrogen receptor in breast cancer]

Sixty specimens of breast cancer were assayed with ER-monoclonal antibody by immunocytochemical staining. Twenty-nine (48.33%) were nuclear estrogen receptor positive (ERn+). The number of ERn+ cancer cells decreased in the following sequence: mucinous carcinoma, invasive lobular carcinoma, invasive ductal carcinoma, and papillary adenocarcinoma. Two apocrine carcinomas were ERn-. ERn+ rate was higher in patients over 60 years of age (P less than 0.05). The amount of ERn+ cells was much greater in cancer cells than in the surrounding benign epithelial cells. This phenomenon may indicate that malignant cells are more hormone-dependent than benign cells. The results of immunocytochemical staining and steroid binding assay were compared. By immunocytochemical staining, twenty-four of fifty-seven cases (42.11%) were ERn+ and ERc+. Sixteen cases (28.07%) were ERn- and ERc-. This study showed that in a number of breast cancers ERs were positive in both cytoplasms and nuclei and the concordant rate was 70.18%. In the remaining cases 13 (22.81%), ERs were positive in the cytoplasm, and in 4 cases (7.02%) positive in nuclei only. Additionally, fifty-two out of 60 cases were assayed by 3H-estradiol and 3H-R5020 by means of steroid binding assay. Twenty-seven cases of them (51.92%) showed ERc+ and PRc+, and seventeen cases (32.69%) were ERc- and PRc-. Their correspondent rate was 84.61%.

DNA content and estrogen receptors in primary carcinoma of the breast

DNA content and estrogen-receptor status were studied in 54 consecutive patients with primary breast carcinoma. Estrogen-receptor determinations were performed by immunohistochemical assay on frozen sections with a monoclonal antibody against the estrogen-receptor molecule and by biochemical analysis with a dextran-coated charcoal method. Nuclear DNA content was measured by flow cytometry performed on formalin-fixed, paraffin-embedded sections. Seventy-two percent of tumours were positive for estrogen receptors by immunohistochemical assay and 67% by biochemical assay. Comparison of the qualitative results of immunohistochemical and biochemical estrogen-receptor determinations revealed a strong correlation between the two assays, with agreement in 90% of the cases (p less than 0.001). Regression analysis showed only a weak relationship between the quantitative results of the two assays. DNA analysis was performed in 51 cases, and 54% demonstrated aneuploid stemlines by flow cytometry. An association was demonstrated between aneuploidy and low levels of estrogen receptor. The association was highly significant with the immunohistochemical assay but not with the biochemical assay. The authors' results suggest that immunohistologic determinations of estrogen receptor status may better reflect the biologic features of the tumour cells. However, improved standardization in reporting the results is necessary if the test is to have widespread use.

[Influences of open thoracoscopy on human body: experience in 65 cases]

The influence of open thoracoscopy with fiber-bronchoscope on human body was studied in 65 patients with pleural disease. The items observed included blood pressure, Pulse rate, respiration rate, blood gas analysis and clinical manifestations. It was found that open thoracoscopy has little influence on human body and the patients tolerated it well. No severe complication was found and operative procedures might be avoided popularization of this technique is therefore suggested.

[The relation of deoxyribonucleic acid contents and nuclear estrogen receptors in breast cancers]

Flow cytometry was applied to the analysis of DNA contents in 59 cases of breast cancer, among which 28 cases were diploidy cancers, and 31 were aneuploid. DNA index of these cases was found to be ranging from 0.45 to 3.35, including S-phase fraction (SPF) from 2.46% to 78.87%, and proliferation index (PI) from 17.06 to 87.04. In 60 cases of breast carcinoma, nuclear estrogen receptors (ERn) were assayed by immunocytochemistry with ER-monoclonal antibody. 29 of them were ERn-, and 31 were ERn+. The DNA contents and ERn levels, both detected in each of 57 cases, were found to be correlated to some extent. In aneuploidy cancer group, there were more ERn- cases than ERn+ cases (P less than 0.05). The mean value of PI was higher in the ERn- group than in the ERn+ group, and the corresponding ERn- cases were more frequent with increase of PI value. Among those cases in which SPF was over 40%, there were more ERn- cases than ERn+ cases. The relations between ERn levels and DNA contents may represent the degree of differentiation of breast cancers and may help to predict prognosis.

Systemic mannitol increases retinal adhesiveness measured in vitro

Pigmented rabbits were given mannitol intravenously, and at various times thereafter blood osmolality was measured and eyes were enucleated to measure retinal adhesiveness to the retinal pigment epithelium. Both blood osmolality and retinal adhesion increased in proportion to the dose of mannitol. We found a measurable increase in retinal adhesion as early as 10 minutes and as late as 4 hours after mannitol injection. We suspect that mannitol dehydrates the subretinal space and thereby enhances the adhesive properties of the interphotoreceptor matrix or tightens the interdigitation between photoreceptors and pigment epithelial cells. Our data suggest that mannitol could have clinical application in the management of retinal detachments.

Retinal adhesiveness is weakened by enzymatic modification of the interphotoreceptor matrix in vivo

The role of interphotoreceptor matrix (IPM) constituents in mediating adhesion between the retina and retinal pigment epithelium (RPE) was investigated by injecting specific enzymes into rabbit eyes either intravitreally or subretinally. Retinal adhesiveness was measured by peeling the retina from the pigment epithelium 1-3 days later and observing the amount of adherent pigment. Effects of enzymes on the IPM were monitored by observation of peanut agglutinin (PNA) binding to cone matrix sheaths; retinal and RPE toxicity was excluded by electroretinography and histology. Three enzymes that degrade glycosaminoglycans or saccharides known to be constituents of the IPM (chondroitinase ABC, neuraminidase, and testicular hyaluronidase) both weakened adhesion and altered PNA binding, although the effects on the cone matrix sheaths were different for each enzyme. An enzyme specific for hyaluronic acid (Streptomyces-derived hyaluronidase), which has not been identified as a major IPM constituent, had no effect on either adhesion or PNA binding. The authors conclude that IPM-associated glycoconjugates participate in retinal adhesion, although their precise composition, interaction with IPM components, and relationship to other mechanisms of adhesion remain to be determined.

[Sequelae in 4122 cases of cesarean section]

From 1986-1987, there were altogether 28,048 deliveries in our hospital, and there were 4,122 cases of Cesarean section, a rate of 14.69%. Early postoperative complications occurred in 522 cases (12.66%). These were namely: postpartum hemorrhage 242 cases (5.87%), postoperative morbidity 151 cases (3.66%), bladder injury during operation 6 cases (0.15%), urinary fistula 1 case, transient hematuria after operation 39 cases (0.95%). 348 Cesarean section cases were followed-up 1-3 years after operation and 44 of them were found to suffer from late complications (12.64%). These included chronic overiosalpingitis 23 cases (6.6%), parametritis 9 cases (2.58%), paracystitis 6 cases (1.7%), abdominal incision scar endometriosis 3 cases (0.86%), ectopic pregnancy 3 cases (0.86%).

[Clinical and pathologic comparison of adult respiratory distress syndrome and infant respiratory distress syndrome]

The authors compared the clinical and pathological findings between adult respiratory distress syndrome (ARDS), and infant respiratory distress syndrome (IRDS). In ARDS, the most common causes were injury, infection, shock and acidosis. The clinical course was longer. The weight of the lungs increased markedly, the hyaline membrane formation in the alveoli was late in the clinical course, and the degree of edema in the interstitium of the lungs and microthrombosis within the blood vessels was more serious. The pathogenesis of ARDS was related to the activation of the complements and neutrophils by inflammation in which proteinase, oxygen radical, thromboxane, leukotriene and prostaglandin were released. Thus the endothelial cells of the blood vessels and capillary-alveoli membrane were damaged by these mediators. On the other hand, the main contributory factors of IRDS were suffocation of premature fetus by various reasons in the uterus and aspiration of meconium during delivery by the infant. The clinical course was shorter, alveolar hemorrhage and collapse were severe and hyaline membrane in alveoli was formed in early stage of the clinical course. Insufficiency of surfactant in premature fetus, damage of the surfactant system by hypoxia, aspiration of foreign materials and defect of the epithelial cells of infant were the pathogenic factors of IRDS, they resulted in increase of permeability of fluid and, as a result, led to pulmonary edema and atelectasis.

The enhancement of retinal adhesiveness by ouabain appears to involve cellular edema

Our previous work showed that ouabain increases retinal adhesion to the retinal pigment epithelium, but the mechanism of this effect has been unclear. In this study, we have found that the ouabain-induced increase in retinal adhesiveness was blocked reversibly by removing sodium from the medium or by exposure to hyperosmotic solution. Hyposmotic solution slowed down the postmortem failure of adhesion. Tissue edema could not be documented histologically, but retinal tissue water content increased by 2.9% after exposure to ouabain. These data suggest that the ouabain effect involves cellular edema, which may enhance adhesion by tightening the interdigitation between photoreceptor outer segments and retinal pigment epithelial microvilli.

Reversibility of retinal adhesion in the rabbit

We have studied retinal adhesion in the rabbit using a new technique for quantifying adhesiveness by the observation of how much pigment remains adherent to the photoreceptors after separation. Previous in vitro work has shown that retinal adhesiveness falls within 3 min after enucleation, is weakened by low pH or the removal of calcium and magnesium, and is increased by cold temperature. This report shows that the effects of low pH, low calcium/magnesium and temperature are all rapidly reversible. Thus, restoring normal pH or normal calcium/magnesium will restore adhesiveness to control levels, and adhesion may be repeatedly weakened or strengthened by the adjustment of temperature or the ionic environment. The actions of pH and calcium/magnesium are in part additive, while cold temperature can maintain adhesion even in low pH or low calcium/magnesium. These findings suggest that irreversible processes such as metabolic or enzymatic decay are not primarily responsible for the loss of retinal adhesion which occurs so rapidly after enucleation or death. They also suggest that retinal adhesion is a multifactorial process.