Analysis of Risk Factors for Febrile Neutropenia in Patients with Small-Cell Lung Cancer Receiving Carboplatin plus Etoposide Therapy

Introduction Febrile neutropenia (FN) is an oncologic emergency requiring immediate empiric antibiotic therapy. Although carboplatin plus etoposide combination chemotherapy is associated with a relatively high frequency of FN, the risk factors are unclear. Hence, this retrospective study aimed to identify predictive markers of carboplatin/etoposide-induced FN. Methods We conducted a retrospective cohort analysis of patients with previously untreated small-cell lung cancer who received combination chemotherapy with carboplatin (area under the concentration curve: 5 mg/mL·min, day 1) and etoposide (80 or 100 mg/m2, days 1-3) between July 2007 and June 2022. FN was assessed during the 21 days after initiation of carboplatin and etoposide therapy according to the Japanese Society of Medical Oncology's definition. Fisher's exact test for categorical variables and Mann-Whitney U-test for continuous variables were used to compare the two groups. Statistical significance was set at p values <0.05. Explanatory variables with p values <0.05 in the univariate analysis were included in the multivariate logistic regression analysis. Results Among the 176 eligible patients, the incidence of FN during the first cycle of chemotherapy was 25.0% (44/176). Multivariate analysis revealed that co-administration of proton pump inhibitors (PPIs) or potassium-competitive acid blockers (PCABs) and body mass index (BMI) were significantly associated with FN (p=0.0035 and 0.0011, respectively). Patients with both co-administration of PPIs or PCABs and a BMI ≤22.509 kg/m2 presented with significantly higher frequencies of FN compared with the other patients (13/24 [54.2%] vs. 31/152 [20.4%] patients; odds ratio: 4.56, 95% confidence interval: 1.70-12.48; p=0.00147). Conclusion Patients who received carboplatin plus etoposide for SCLC with co-administration of PPIs or PCABs and a BMI ≤22.509 kg/m2 more frequently present with FN than those without the two factors.

A Patient Safety Champion Program for Interprofessional Health Care Educators: Implementation and Outcomes

INTRODUCTION

Health care educators are challenged with helping clinicians develop competencies beyond their foundational training. In health care systems where continuing professional development is not integral to practice, clinicians may have few opportunities. We describe the design, implementation, and evaluation of a professional development program in patient safety for Japanese clinical educators to acquire simulation instructional skills and become Patient Safety Champions at their organizations.

METHODS

Mixed methods were used in a longitudinal pre/post study design. The Kirkpatrick evaluation model was used to evaluate outcomes of a workshop, overall program, on-site training experiences, and impact as Patient Safety Champions. Self-assessment data on skills and knowledge of patient safety, simulation instructional methods, interprofessional collaboration, and leadership were collected and analyzed.

RESULTS

Eighty-nine percent of participants facilitated on-site patient safety training within 6 months of workshop completion. Skills and knowledge improvement were observed immediately postworkshop in four categories: patient safety, simulation instructional methods, interprofessional collaboration and communication, and leadership as a patient safety champion. Skills and knowledge increased at 6 months after facilitation of on-site safety training. Program mean satisfaction scores ranged from 84% to 92%. Mean Patient Safety Champion in-facility evaluations were 4.2 to 4.7 on a 5-point scale.

DISCUSSION

High levels of knowledge, skill retention, and behavior change are attributed to goal setting, outcome-oriented pedagogy, and reflective sessions. The Patient Safety Champion model and experiential learning approach gave Japanese clinical educators in medicine, nursing, and pharmacy an opportunity to learn from each other in simulations reflecting the practice environment.

Pharmacokinetics of Neoadjuvant Axitinib Influenced the Efficacy in Patients With Advanced Renal Cell Carcinoma

Although axitinib shows a good objective response rate and acceptable tolerability for advanced renal cell carcinoma, substantial differences in drug concentrations among individuals have hampered the reliable administration of the drug in a neoadjuvant setting. This study aimed to evaluate the relationship between axitinib pharmacokinetics and clinical efficacy in patients with advanced renal cell carcinoma treated in a neoadjuvant setting. We retrospectively reviewed 16 patients who underwent neoadjuvant axitinib treatment from prospective phase 2 study cohorts treated with axitinib and assessed whether the drug concentration was associated with clinical efficacy for primary tumors of advanced metastatic/oligometastatic clear cell renal cell carcinoma. Axitinib was administered orally at a starting dose of 5 mg twice daily for 2 months in principle before the operation, and the axitinib pharmacokinetics were examined. Best response, reduction rate, adverse events (AEs), and surgical complication were assessed. Four patients (25.0%) showed a partial response, and 12 (75.0%) had stable disease, with a mean reduction rate of 22.8%. No progressive disease was noted, and 9 of the 16 patinets (56.3%) showed downstaging. The trough level of axitinib significantly correlated with the objective response rate (P = .0052) and best tumor reduction (P = .0128). All AEs could be safely managed until termination of the dosing period. With respect to perioperative complications, grade 2 anemia was observed. Neoadjuvant axitinib treatment showed acceptable antitumor activity and safety profile for advanced renal cell carcinoma. The pharmacokinetics of neoadjuvant axitinib influenced the efficacy in patients with advanced renal cell carcinoma.

Development of hemiacetal esterified levofloxacin to prevent chelation with metal-containing drugs

OBJECTIVES

To avoid the chelate formation between levofloxacin (LVFX) and aluminium hydroxide in gastrointestinal tract, an ethoxycarbonyl 1-ethyl hemiacetal ester of levofloxacin (LVFX-EHE) was synthesised as a prodrug.

METHODS

The effects of aluminium hydroxide on the bioavailability of LVFX following oral administration of LVFX-EHE were investigated in rats. Furthermore, the effects of aluminium hydroxide on small intestinal absorption of LVFX and LVFX-EHE when subjected to a hydrolysis experiment using in situ everted gut sac were investigated, and the minimal inhibitory concentrations (MICs) of LVFX and LVFX-EHE for various intestinal bacteria were measured.

KEY FINDINGS

When LVFX-EHE was co-administered with and without aluminium hydroxide, the AUC_{0-4 h} values of LVFX hydrolysed from LVFX-EHE were similar to that of LVFX alone. In everted gut sac experiments, LVFX-EHE was efficiently absorbed even in the presence of aluminium ions after 1 h of incubation, whereas the absorption of LVFX decreased significantly in the presence of aluminium ions. MIC values of LVFX-EHE were far higher than LVFX.

CONCLUSIONS

This study suggests the benefit of ethoxycarbonyl 1-ethyl hemiacetal esterification of the carboxyl group of new quinolone as a prodrug which is able to avoid chelate formation.

[Trial of the integrated cross-field pharmaceutical education in the first year of faculty of pharmacy]

The six-year pharmacist education course has begun, and now first-year students receive clinical training. Interdisciplinary problem-solving capabilities covering chemistry, biology, molecular biology, pharmacology, pathology, and pharmacokinetics are necessary for new pharmacists. However, the conventional pharmaceutical science education was so separate from other fields that education for interdisciplinary cooperative capability was insufficient. This was especially true of elemental science courses, because they are not directly connected with clinical knowledge, and there is a problem of low student interest in those courses. As a result, students acquired only recall-level knowledge in clinical courses and their problem-solving capabilities in clinical treatment and drug development deteriorated. Therefore we offered a trial lecture aimed to help students recognize the important relationship between elemental science courses and clinical courses and increase their motivation to enroll in these courses. Specifically, the trial lecture covered cancer therapy, in reference to mechanisms of carcinogenesis, epidemiology, physiology of cancer, anticancer drugs with explanations of the mechanism of action of carcinogens, anticancer drugs, and molecular-targeted drugs from the viewpoints of organic chemistry and biochemistry by a specialized teacher. This paper reports on this experimental lecture with evaluations from students.

Extracranial carotid plaque is increasing in Japanese ischemic stroke patients

OBJECTIVE

The objectives of this study were to investigate the prevalence of extracranial carotid plaque and the association between risk factors and carotid plaque in Japanese patients with ischemic stroke.

METHODS

We consecutively recruited patients with ischemic stroke admitted to our hospital from January 2000 to September 2002. Neurologic signs and a brain magnetic resonance imaging diagnosed ischemic stroke. All subjects underwent a carotid ultrasonography. Multiple logistic regression analysis was used to determine the risk factors that independently contributed to the presence of carotid plaques.

RESULTS

Carotid plaques were identified in 76.2% of the patients and bilateral plaques were found in 58.2%. These lesions were more frequent in comparison with previous Japanese reports. The risk factors that independently contributed to the presence of extracranial carotid plaques were hypertension, age, smoking and past history of ischemic stroke.

CONCLUSIONS

This study demonstrates that extracranial carotid plaque is increasing in Japanese patients with ischemic stroke.

Delayed administration of JTP-2942, a novel thyrotropin-releasing hormone analogue, improves cerebral blood flow and metabolism in rat postischaemic brain

1. The aim of the present study was to examine the central nervous system action of JTP-2942, a novel thyrotropin-releasing hormone (TRH) analogue, from the point of view of cerebral blood flow (CBF) and metabolism in the postischaemic brain. 2. Left middle cerebral artery ischaemia was induced for 90 min followed by reperfusion. 3. Animals were separated into four groups: (i) low-dose (0.003 mg/kg) JTP-2942; (ii) high-dose (0.03 mg/kg) JTP-2942; (iii) cystidine diphosphate choline (500 mg/kg); and (iv) saline. The test drug or saline was administered intravenously 1 week after ischaemia. 4. Local CBF and local cerebral glucose utilization were measured autoradiographically, adjacent sections were stained with haematoxylin-eosin and infarction size was measured. 5. JTP-2942 ameliorated the reduction of local CBF and glucose utilization except in the ischaemic core. In particular, the higher dose (0.03 mg/kg) of JTP-2942 significantly increased local CBF and glucose utilization not only in peri-infarcted areas, but also in distal and contralateral areas. 6. These results suggest that JTP-2942 treatment may be beneficial for improving cerebral circulation and metabolism in the postischaemic brain.

[Photostimulation that induces flattening of photopic electroretinograms--first report. An apparatus for slope photostimulation]

PURPOSE

We developed a new photostimulation system for electroretinograms (ERGs). This apparatus is capable of varying the transients during stimulus on and off (up-slope and down-slope times, respectively) as well as the stimulus and background intensities. We applied this system to evaluate the photopic ERG in human eyes.

CASES AND METHODS

Seven normal volunteers and a 43-year-old patient with pre-proliferative diabetic retinopathy served as test subjects. We recorded ERGs with relatively long intervals of up-slope and down-slope times for photostimulation.

RESULTS

As the up-slope and down-slope times were increased, the amplitude was reduced and the peak latency was prolonged for both the on and off responses. When the normal subjects' retinas were stimulated with a 341-msec up-slope time and a 34.1-msec down-slope time, the ERG waves showed complete flattening in both on and off responses. The ERGs of a patient with pre-proliferative diabetic retinopathy showed much longer peak latency than that of the normal volunteers as the up-slope time was prolonged. When the patient's retina was stimulated for a prolonged up-slope time, the average of the amplitude of ERG waves was similar to that of the normal volunteers.

CONCLUSION

These results suggest that this new type of photostimulation might potentially uncover some novel aspect of clinical ERGs and might be useful for testing retinal functions of patients with diabetic retinopathy.

SPARC from corneal epithelial cells modulates collagen contraction by keratocytes

PURPOSE

Contraction of the scar tissue during corneal wound healing changes the shape of the cornea and corneal refraction. In a previous study, it was found that corneal epithelial cells secrete the factor that stimulates collagen gel contraction by keratocytes in vitro. The purpose of the present study was to purify and identify the contraction-stimulating factor derived from corneal epithelial cells.

METHODS

The cultured medium of rabbit corneal epithelial cells was collected and used as an epithelial cell-conditioned medium (ECCM). Subcultured rabbit keratocytes were embedded in a collagen gel, and collagen gel contraction was investigated. The contraction-stimulating factor in the ECCM was purified through acetone precipitation, affinity chromatography (heparin Sepharose), gel filtration, and reversed-phase chromatography. The amino acid sequence of a contraction-stimulating factor was analyzed.

RESULTS

Collagen gel contraction by keratocytes was enhanced by the addition of ECCM in a dose-dependent manner. The amino acids sequence of the contraction-stimulating factor was homologous to a 32-kDa glycoprotein, a secreted protein that is acidic and rich in cysteine (SPARC). Western blot analysis confirmed that SPARC was contained in the ECCM. Collagen gel contraction by keratocytes was enhanced by the addition of purified SPARC in a dose-dependent manner. SPARC was found in the basal layer of the migrating epithelium and activated keratocytes adjacent to the wound 3 days and 1 week after perforating injury in rabbit corneas.

CONCLUSIONS

Epithelial cells secrete SPARC, which modulates the contraction of scar tissue in the corneal stroma.

[Basic and clinical aspects of modern perimetry]

Clinical perimetry has made a remarkable advance in the last 40 years and many new automated perimeters and sophisticated techniques are now available. We developed a Central Field Screener, O-II, an automated perimeter, ATS-85, and a central critical fusion frequency (CFF)-meter. First, reappraisal of the clinical significance of manual kinetic perimetry was made in comparison with automated light threshold perimetry in the diagnosis of retinitis pigmentosa and chiasmal tumors. We concluded that manual kinetic perimetry was the basis of clinical perimetry and superior to automated static light threshold perimetry in some cases. Secondly, automated static flicker perimetry was performed with Octopus 1-2-3 and a software package developed by us. We found that automated static flicker perimetry was far superior to automated static light threshold perimetry in the detection of early changes of glaucomatous field defects, especially in cases of immature senile cataract. We also confirmed that measurement of CFF was essential in the diagnosis of optic neuritis and diseases of chiasmal syndrome. Dissociation of central vision and central CFF was observed in many cases of optic neuritis and inverse dissociation of central vision and central CFF was seen in some cases of Leber's disease.

The effect of corneal epithelial cells on the collagen gel contraction by keratocytes

To understand the mechanism of the contraction of cicatricial tissues after corneal stromal wounding, we investigated the effect of corneal epithelial cells on the collagen gel contraction by keratocytes. Subcultured rabbit keratocytes were embedded in type I collagen gel and cultured. Rabbit corneal epithelial cells were cultured separately and the culture medium was collected and used as an epithelial cell-conditioned medium (ECCM). The collagen gel contraction by keratocytes was determined by measuring the diameters of the collagen disks once a day for 5 days. In the presence of fetal calf serum, the diameter of the collagen gel decreased in linear proportion to the number of keratocytes, but in inverse proportion to the concentration of collagen. The addition of the ECCM enhanced the collagen gel contraction by keratocytes in a dose-dependent manner. This stimulatory activity was found in the ECCM obtained from 3- to 7-day cultures of epithelial cells. These findings suggest that corneal epithelial cells secrete factor(s) that stimulate the collagen gel contraction by keratocytes.

Collagenolytic activity of keratocytes cultured in a collagen matrix

To study the mechanism of collagen degradation by keratocytes, we developed the new in vitro model in which keratocytes were cultured three-dimensionally in a collagen matrix. Subcultured rabbit keratocytes were embedded in a type I collagen matrix and cultured in serum-free medium. Collagenolytic activity of the cells was determined by measuring the amount of hydroxyproline released into the medium from degraded collagen. Activities of collagenase in the medium were also measured, using collagen labeled with fluorescein isothiocyanate as a substrate. The presence of plasminogen was required for collagen degradation by keratocytes. In the presence of plasminogen, the amount of collagen degradation depended on both the cultivation period and the number of cells. The addition of interleukin-1 (IL-1) stimulated the collagen degradation in a dose-dependent manner. This stimulatory effect of IL-1 was completely inhibited by the addition of IL-1 receptor antagonist (IL-1ra). Collagenase activity in the medium was stimulated by the addition of IL-1, and IL-1ra antagonized this stimulatory effect. These findings indicate that our present model may be useful for investigating the mechanism of collagen degradation by keratocytes.

[Measurement of regional cerebral blood flow and glucose utilization in rat brain under chronic hypoperfusion conditions following bilateral carotid artery occlusion. Analyzed by autoradiographical methods]

OBJECTIVE

Although the pathology of chronic hypoperfusion in rat following bilateral carotid artery occlusion has been documented, long term changes in cerebral blood flow and metabolism have not been reported. In this study regional cerebral blood flow (rCBF) and regional glucose utilization (rCGU) were analyzed by autoradiographical methods, pathological observation of the brains was also conducted.

METHODS

Male Wistar rats aged 12 weeks were anesthetized and the bilateral carotid arteries were occluded. Physiological parameters, ABG, MABP, and rectal temperature were measured before and through occlusion. After 2 days, 1, 4, and 8 weeks and controls of rCBF (Sakurada) and rCGU. (Sokoloff) were measured (n = 6). Evaluated regions included frontal cortex (Fcor), parietal cortex (Pcor), temporal cortex (Tcor), occipital cortex (Ocor), genu corpus callosum (gCC), corpus callosum (CC), splenium corpus callosum (sCC), caudate putamen (CPu), globus pallidus (GP), internal capsule (IC), thalamus (Thal), hippocampus CA 1 (CA 1), hypothalamus (HypoTh), amygdal (Amygd), and substantia nigra (SNR). In separate animals (n = 3) tissue sections were stained using Klüver-Barrera (KB) and Hematoxylin-Eosin (HE), and pathological changes were observed.

RESULTS

After 2 days the rCBF values were significantly reduced to 33-58% of control values in the Fcor. Pcor, Tcor, Ocor, gCC, CC, sCC, CPu, GP, JC, and Amygd. The reductions were observed from the 2nd day to the 1st week. From the 1st week to the 4th week values began to recover to control levels. However, after 4 weeks they were still significantly reduced in the Ocor, gCC, CC, sCC, GP, IC, and SNR (51-63%). After 8 weeks, the rCBF values in the areas except white matter, CPu, GP, and Ocor, etc., recovered to approximately 90% of control levels. However, in the Ocor, gCC, CC, sCC, CPu, GP, and IC, they were still 70-89% of control levels. After 2 days the rCGU values were reduced to 56-95% (except Amygd) in the measured regions, although not as reduced as the rCBF levels. From the 2nd day to the 1st week, the rCGU values were reduced further and after 1 week the rCGU values were significantly reduced to 39-69% in the Fcor, Pcor, Tcor, Ocor, gCC, CC, sCC, CPu, GP, IC (Lt), Thal (Rt), Amygd, and SNR From the 1st week to the 4th week, the values began to recover to control levels. After 4 weeks, rCGU improved to approximately 90-100% of controls, and remained at that level through 8 weeks occlusion. Rarefaction of the myelinated fibers was observed in the white matter from the 1st week to the 4th week in the KB stained sections, while little change in cortex was observed throughout 1st to 8th week.

CONCLUSION

In this chronic hypoperfusion model rCBF remained depressed after 8 weeks in the Ocor, white matter, and basal ganglia, and rarefaction of the white matter was observed. These results indicate that this model is suitable for the study of chronic cerebral hypoperfusion.

Detection of interleukin-1 beta in the tear fluid of patients with corneal disease with or without conjunctival involvement

To investigate the role of Interleukin-1 (IL-1) in the pathobiology of the cornea, we measured IL-1 beta concentration in tear fluid samples from patients with corneal disease. Twenty patients with unilateral corneal disease were included in the study. Tear fluid samples were collected during the active stages of the disease and following resolution. The fellow (unaffected) eyes served as controls. The concentration of IL-1 beta in the tear fluid samples was measured using a sandwich ELISA method. IL-1 beta was detected in tear fluid from five eyes (three eyes with chemical burns, one with a Pseudomonas aeruginosa corneal ulcer, and one with a peripheral corneal ulcer) at concentrations between 29 and 218 pg/mL. IL-1 beta was not detected in tear fluid from the remaining 15 affected eyes, nor from the control eyes. The detection of IL-1 beta in tear fluid correlated with limbal conjunctival involvement, but did not correlate with the type of disease, size of epithelial defect, or degree of stromal involvement. IL-1 beta in the tear fluid may be one of the factors modifying the complex inflammatory process of the anterior ocular surface.

Entactin modulates the attachment of rabbit corneal epithelial cells

PURPOSE

To understand the biological activity of entactin, a component of the basement membrane of the corneal epithelium, we investigated the ability of rabbit corneal epithelial cells to attach to an entactin matrix and the effect of entactin on the cells' attachment to other corneal basement proteins.

METHODS

Multiwell plastic plates were coated with bovine serum albumin (BSA), alone or with BSA and entactin, laminin, fibronectin or collagen type IV. Cultured rabbit corneal epithelial cells were seeded on the plates. After incubation (usually 90 min), the cells were fixed and stained with 1% crystal violet. The number of attached cells was counted under a light microscope.

RESULTS

The numbers of attached cells increased in proportion to both the incubation period and the concentration of entactin coated. Furthermore, the number of cells attached to the entactin-coated plate was greater than the number attached to the BSA-coated plate for each incubation period (30 to 120 min). Likewise, when laminin-coated plates were treated with entactin, the number of the attached cells increased in proportion to the concentration of entactin. However, entactin did not affect the cellular attachment of fibronectin or type IV collagen. Cellular attachment to entactin was partially inhibited by the cells' preincubation with the synthetic peptide (GRGDSP).

CONCLUSIONS

The present results showed that cultured corneal epithelial cells adhere to entactin and that entactin stimulated the attachment of these cells to the laminin matrix. These findings suggest that entactin plays a specific role in maintaining the normal integrity of the corneal epithelium.

Ocular fundus changes in malignant or precocious stroke-prone spontaneously hypertensive rats after administration of antihypertensive drugs

1. Malignant or precocious stroke-prone spontaneously hypertensive rats (M-SHRSP) showed hypertensive ocular fundus changes with severe hypertension, but various anti-hypertensive drugs given over a proper period improved funduscopic findings. 2. We treated a M-SHRSP with SQ29,852 (an angiotensin converting enzyme inhibitor, ACEI) or manidipine (a calcium antagonist) and observed hypertensive vascular changes in the fundus. 3. The M-SHRSP treated with anti-hypertensive drugs lived longer and the hypertensive fundic changes improved in both groups. But there were some differences of histochemical staining reaction on the endothelial cell-surface and mucopolysaccharides accumulations between the treated group of ACEI and the group treated with the calcium antagonist. 4. In the treated group with calcium antagonist, the anionic ion functions of the endothelial cell-surface were impaired and organic retinal or choroidal damages appeared to have deteriorated.

[The effect of corneal epithelial cells on the collagen gel contraction by keratocytes]

To understand the mechanism of corneal stromal wound contraction, we investigated the effect of corneal epithelial cells on the collagen gel contraction by keratocytes. Subcultured rabbit keratocytes embedded in type I collagen gel were cultured. Rabbit corneal epithelial cells were also cultured and the cultured medium was collected and used as an epithelial cell conditioned medium (ECCM). The collagen gel contraction by keratocytes was estimated by measuring the diameter of the collagen disc once a day for 5 days. The diameter of the collagen gel decreased in proportion to the number of keratocytes in the presence of fetal calf serum (FCS), but it decreased in inverse proportion to the concentration of collagen. The collagen gel contraction by keratocytes was enhanced by the addition of ECCM in a dose dependent manner. This stimulatory activity was found in ECCM obtained from 3 to 7 day cultivation of corneal epithelial cells. These findings demonstrated that keratocytes contracted collagen gel in the presence of FCS. It also appeared that corneal epithelial cells secrete a factor or factors that stimulate the collagen gel contraction by keratocytes.

Up-regulation of integrin alpha 5 beta 1 expression by interleukin-6 in rabbit corneal epithelial cells

Interleukin-6 (IL-6) has been shown to promote the attachment of rabbit corneal epithelial cells to fibronectin-coated substratum and ex vivo migration of the cells on the corneal stroma. To examine whether IL-6 promotes cell attachment through up-regulation of expression of integrin alpha 5 beta 1, i.e., the major cell surface fibronectin receptor, we quantified the levels of both alpha 5 and beta 1 subunit transcripts by reverse transcription-polymerase chain reaction in cultured rabbit corneal epithelial cells pretreated with various concentrations of IL-6. The levels of both alpha 5 and beta 1 mRNAs were dose-dependently elevated by IL-6, attaining 1.5- and 1.8-fold increases, respectively, at 10 ng/ml. The stimulatory effect of IL-6 was transient; the levels of both subunit mRNAs reached a maximum 1 h after the addition of IL-6 and returned to the basal levels after 6 h. The IL-6-induced up-regulation of integrin alpha 5 and beta 1 mRNAs was also confirmed by Northern blot analysis. These results indicate that the increased attachment of corneal epithelial cells to fibronectin and enhanced ex vivo migration on corneal stroma by IL-6 is, at least in part, due to the temporal up-regulation of integrin alpha 5 beta 1 expression in corneal epithelial cells.

Binding of hyaluronan to plasma fibronectin increases the attachment of corneal epithelial cells to a fibronectin matrix

We wished to determine whether hyaluronan would affect the attachment of epithelial cells to extracellular matrix proteins. Multiwell tissue culture plates were coated with human plasma fibronectin, laminin, or collagen type IV (0.01-10.0 micrograms/ml). Single-cell suspensions of rabbit corneal epithelial cells were placed in the wells, and after 45 minutes incubation the cells adhering to the matrix proteins were stained and counted. Cells attached to all three types of proteins. Preincubation of the matrix proteins with hyaluronan (0.1-1.0 mg/ml) significantly increased the number of cells attached to the fibronectin matrix, but it did not increase the numbers of cells attached to laminin or collagen type IV. Hyaluronidase inhibited this stimulatory effect. Glycosaminoglcyans other than hyaluronan (chondroitin sulfate, keratan sulfate, or heparan sulfate) failed to increase the numbers of attached cells. Treatment of the fibronectin matrix with monoclonal antibodies against the cell-binding domain of fibronectin (FN12-8 or FN30-8, 0.03-0.3 mg/ml, for 1 hour), before or after hyaluronan treatment, significantly decreased the numbers of attached cells. Monoclonal antibody against the fibrin- and heparin-binding domain at the N-terminal (FN9-1), however, significantly decreased the number of attached cells only when this antibody treatment preceded the hyaluronan treatment. Preincubation of the cells with hyaluronan had no effect; preincubation with GRGDSP (1 mg/ml), a synthetic peptide that blocks the cell surface receptor for fibronectin, significantly decreased cell attachment whether the fibronectin matrix was treated with hyaluronan or not. Further studies demonstrated that monoclonal antibody against the fibrin- and heparin-binding domain at the N-terminal of plasma fibronectin prevented radiolabeled hyaluronan from binding to fibronectin; likewise, the isolated N-terminal fragment, coupled with Sepharose 4B, bound to hyaluronan in columns. We conclude that hyaluronan binds to a fibrin- and heparin-binding domain at the N-terminal of plasma fibronectin and facilitates the attachment of epithelial cells.

Combined effects of hyaluronan and fibronectin on corneal epithelial wound closure of rabbit in vivo

We investigated the combined effects of hyaluronan and fibronectin on corneal epithelial wound closure in vivo. When fibronectin eye drops (1 mg/ml) were administered for the first 4 hours after chemical cautery and hyaluronan eye drops (1 mg/ml) were administered for following 32 hours, the mean healing rate was significantly higher than in control eyes treated with phosphate buffered saline instead of fibronectin or with initial fibronectin alone. These results demonstrated that hyaluronan and fibronectin have a synergistic effect, with fibronectin pretreatment augmenting hyaluronan-stimulated corneal epithelial wound closure.

[The influence of target blurring and simulated opacity of the ocular media on automated perimetric thresholds]

The influence of target blurring and simulated opacity of the ocular media on the sensitivity of the central visual field in quantitative static perimetry was studied using a SARGON program of the automated perimeter Octopus 201. Ten eyes of 10 normal subjects aged from 21 to 35 were examined with target sizes 1, 3, and 5; 4 spherical plus lenses of +1.0D, +2.0D, +3.0D, and +5.0D; 5 kinds of occlusion diffusers which corresponded to the visual acuities of 1.0, 0.8, 0.6, 0.4, and 0.1; and 4 neutral density filters of 0.1, 0.3, 0.6, and 1.0 log units. The studies suggested that there was little difference between the sensitivity for target size 1 and that for target size 3 in spite of slight blurring of the target and simulated opacity of the ocular media. We concluded that target size 1 was as useful as target size 3 in quantitative static perimetry.

Effects of antimicrobials on corneal epithelial migration

The slowed healing rates observed by some investigators may be caused by vehicles or preservatives in the antimicrobials preparations tested. To determine whether antimicrobials directly inhibit corneal epithelial wound healing, we cultured blocks of the rabbit cornea in media containing various concentrations of antibiotics or antimicrobials (at 1, 10, or 100 micrograms/ml); after 24 hours, we measured the distance of epithelium that had migrated down the side of each block. The higher concentrations of fluoroquinolones (ofloxacin; 74 +/- 5.8% of control at 100 micrograms/ml, p < 0.05, ciprofloxacin; 4.4 +/- 1.5% of control at 100 micrograms/ml, p < 0.01, or norfloxacin; 71 +/- 7.0% at 10 mu g/ml, p < 0.01, and 1.5 +/- 0.4% of control at 100 mu g/ml, p < 0.01) and the highest concentrations of peptides (polymyxin B; 64 +/- 3.0% of control at 100 micrograms/ml, p < 0.01, or colistin; 67 +/- 5.7% of control at 100 micrograms/ml, p < 0.01) or fosfomycin (79 +/- 6.2% of control at 100 micrograms/ml, p < 0.05) had an inhibitory effect on corneal epithelial migration. Among aminoglycosides tested, sisomicin (85 +/- 10.0% of control, not significant), dibekacin (76 +/- 11.6% of control, p < 0.05) and streptomycin (77 +/- 9.4% of control, not significant) were inhibitory at 100 micrograms/ml, but tobramycin had no effect. Penicillins (aspoxicillin, sulbenicillin or ampicillin), cephalosporins (cefmenoxime or cefminox), oxytetracycline, erythromycin and chloramphenicol did not affect epithelial migration at all. These results demonstrate that some antimicrobials are inhibitory at high concentrations, but penicillins, cephalosporins, oxytetracycline, erythromycin or chloramphenicol has no inhibitory effect on corneal epithelial migration.

Role of urokinase type plasminogen activator (u-PA) in corneal epithelial migration

The role of plasminogen activator (PA) in the migration of corneal reepithelialization was studied. Rabbit corneal blocks were cultured, and both the extent of epithelial migration over the exposed corneal stroma and the activity of PA released into the culture media were measured. A significant, direct correlation between epithelial migration and PA activity in the medium was observed, even when the migration was stimulated by fibronectin or EGF, or was inhibited by cytochalasin B or cycloheximide. Zymography confirmed that the PA released into the culture medium was of the urokinase type (u-PA). Immunohistochemical studies showed that u-PA and plasmin(ogen) were present at the leading edge of the migrating epithelium. Studies of corneal cell cultures indicated that epithelial cells rather than endothelial cells or fibroblasts were the source of the u-PA. The addition of antihuman u-PA IgG or protease inhibitors retarded the migration of the corneal epithelium in a dose-dependent manner, indicating that u-PA activity is essential for the migration of the corneal epithelium. These findings suggest that the migration of corneal epithelial cells requires not only cell attachment to the extracellular matrix through the fibronectin but also degradation of the fibronectin by the release of cellular u-PA.

Interleukin 6 promotes epithelial migration by a fibronectin-dependent mechanism

We investigated the effect of interleukin 6 (IL-6) on the migration of rabbit corneal epithelium in vitro and on the attachment of dissociated corneal epithelial cells to a fibronectin matrix. When corneal blocks were cultured with IL-6 for 24 hours, the length of the path of epithelial migration over exposed corneal stroma increased significantly (p less than 0.005 at the concentration of 10 ng/ml) in proportion to the concentrations of IL-6 (0.1-10.0 ng/ml). The addition of antiserum against fibronectin or of GRGDSP abolished the stimulatory effect of IL-6 on epithelial migration. When corneal epithelial cells were cultured with various concentrations of IL-6, suspended, and plated on wells coated with fibronectin (10 micrograms/ml), the number of cells attached to the wells increased in a dose-dependent manner. The presence of antibody against fibronectin or of GRGDSP during the attachment assay decreased the number of cells attached to the fibronectin matrix, regardless of the fact that the cells had been cultured with IL-6 or not. IL-6 stimulated the attachment of corneal epithelial cells to collagen type IV and to laminin matrices. However, the presence of GRGDSP did not affect the cell attachment to collagen type IV and to laminin. These findings strongly indicate that IL-6 stimulates epithelial migration in the cornea by a fibronectin-dependent mechanism, presumably the increased expression of fibronectin receptors.

Interleukin 6 facilitates corneal epithelial wound closure in vivo

Rapid corneal epithelial wound healing is essential to the maintenance of clear visual acuity. The cytokine interleukin 6 is thought to participate in the wound-healing process. We investigated the effect of interleukin 6 eye drops on the rate of corneal epithelial wound closure in rabbits in vivo. Recombinant human interleukin 6 in phosphate-buffered saline at concentrations of 0.1, 0.3, or 1 mg/L was administered immediately after the epithelium was débrided with the n-heptyl alcohol treatment and 2, 4, 6, 8, 10, 18, 20, 22, 24, 26, and 28 hours after débridement. The eyes were stained with fluorescein and photographed at baseline and at 7, 10, 18, 21, 24, 27, 30, and 43 hours after débridement. The rate of wound healing was calculated with a linear regression analysis based on the areas of the epithelial defects, which were recorded from hour 10 to hour 30. The mean (+/- SD) healing rate in the control group, which received phosphate-buffered saline alone, was 1.03 +/- 0.15 mm2/h. The administration of interleukin 6 at concentrations of 0.1, 0.3, or 1.0 mg/L increased the healing rate significantly (P less than .001) to 1.27, 1.39, or 1.44 mm2/h, respectively. Our results indicate that the administration of interleukin 6 might have clinical applications in the treatment of persistent corneal epithelial defects.

Coordinated appearance of beta 1 integrins and fibronectin during corneal wound healing

Fibronectin, which appears at wound sites, serves as a temporary matrix for the epithelial migration involved in healing. Cellular responses to fibronectin have not yet been elucidated. We visualized beta 1 integrins, fibronectin, laminin, and collagen type IV by immunofluorescence techniques at various intervals after single, nonpenetrating incision in the corneas of rats to investigate the chronologic changes in the localization of these proteins during corneal epithelial wound healing. In unwounded corneas beta 1 integrins were found on basal cells of the corneal epithelium and fibronectin was found in the corneal subepithelial region, at Descemet's membrane, and in streaks between collagen fibers of the stroma. Both laminin and collagen type IV were found in the subepithelial region and Descemet's membrane. Immediately after the incision, fibronectin was visible on the surface of the V-shaped defect of the stroma; epithelial cell expressed beta 1 integrins began to migrate over the defect and to fill it. When healing was almost completed, beta 1 integrin staining of epithelial cells diminished, except for in basal cells. Staining of fibronectin diminished at the interface between the new epithelium and the stroma. Laminin and collagen type IV were not seen between the migrating epithelial cells and the underlying stroma until epithelial cells completely covered the defect. Appearance and disappearance of beta 1 integrins and fibronectin are thus well coordinated during corneal epithelial wound healing process, suggesting an active involvement of beta 1 integrins and fibronectin.

Transforming growth factor-beta modulates effects of epidermal growth factor on corneal epithelial cells

In order to understand the mechanisms that bring about maintenance and restoration of the integrity of corneal epithelium, we investigated independent and combined effects of transforming growth factor-beta (TGF-beta) and epidermal growth factor (EGF) on rabbit corneal epithelial cells in cell and organ culture. Specifically, we determined whether incubation with these factors influenced 1) cellular proliferation, 2) ability of cells to attach to a fibronectin matrix, and 3) the rate of epithelial migration over corneal stroma. Incubation with TGF-beta caused a dose-related decrease in the incorporation of 3H-thymidine by the epithelial cells. EGF increased 3H-thymidine incorporation, but this effect was antagonized by the addition of TGF-beta into the incubation medium. Incubation with EGF increased the numbers of cells that attached to a fibronectin matrix. TGF-beta itself did not affect the number of attached cells but, again, it antagonized the stimulatory effect of EGF. Similarly, when corneal blocks were cultured with EGF, epithelial migration increased in a dose-related manner. TGF-beta itself did not affect epithelial migration at any of the concentrations tested (0.1-10 ng/ml), but it antagonized EGF-stimulated epithelial migration. These findings suggest that the proliferation and the migration of corneal epithelial cells are regulated by different mechanisms, and that TGF-beta serves as a modulator of the effects of EGF.

Epidermal growth factor stimulates corneal epithelial cell attachment to fibronectin through a fibronectin receptor system

Epidermal growth factor (EGF) stimulates corneal epithelial migration in vivo and in vitro. Antibody against fibronectin inhibits this effect in vitro, suggesting that a fibronectin-dependent mechanism in involved. To elucidate the action of EGF, we placed rabbit corneal epithelial cells, preincubated in the absence or presence of EGF (10 ng/ml), into wells coated with fibronectin. After 45 minutes of incubation, the numbers of cells attached to the wells were counted. Preincubation with EGF for 6 hr was not effective, but preincubation for 9 hr significantly increased the numbers of cells attached to the wells. These numbers were not increased further by additional preincubation. When concentrations of EGF were reduced, numbers of attached cells decreased proportionally, but remained significantly higher than the numbers obtained with cells not exposed to EGF. The EGF-stimulated attachment to the fibronectin matrix was inhibited in a dose-dependent fashion by antifibronectin IgG and by GRGDSP, a synthetic peptide that mimics the amino acid sequence of the cell-binding domain of fibronectin. The authors conclude that a fibronectin/fibronectin receptor system mediates EGF-induced stimulation of cellular attachment. These findings suggest that EGF may increase the expression of fibronectin receptors.

Hyaluronan stimulates corneal epithelial migration

Hyaluronan (hyaluronic acid), well-known for its viscoelastic properties, is also recognized as a biological signal to cells. Using organ cultures of the rabbit cornea, we investigated the effects of hyaluronan on the migration of corneal epithelium. The addition of hyaluronan to the culture medium increased the length of the path of the corneal epithelial layer in a dose-dependent fashion. Other glycosaminoglycans (chondroitin, chondroitin sulphate, keratan sulphate and heparan sulphate) were also tried, but only hyaluronan exhibited a stimulatory effect on corneal epithelial migration. The effects of hyaluronan and fibronectin or epidermal growth factor (EGF) were additive; the addition of antisera against fibronectin or against EGF did not alter the stimulatory effect of hyaluronan. These results demonstrate that hyaluronan stimulates corneal epithelial migration by mechanism(s) different from those of fibronectin and EGF.

Long-term storage of endocytosed latex beads in keratocytes in vivo

Endocytosis by keratocytes (corneal fibroblasts) is an important part of the host defense system. To investigate the long-term fate of endocytosed materials, we injected polystyrene latex beads into the corneal stroma of four rabbits. The corneal stroma was observed under a transmission electron microscope 4 and 800 days after the injection. After 4 days, the beads were found not only between the collagen fibers of the stroma, but also in some keratocytes. After 800 days, no extracellular beads were seen, but endocytosed beads remained, surrounded by limiting membranes, in the cytoplasm of keratocytes. These observations demonstrate that keratocytes endocytose latex beads and store them for a long time, isolating these foreign materials from the corneal stroma. These observations suggest that keratocytes, like some other fibroblasts perform a noninflammatory and nonimmunological defense function.

Expression of fibronectin receptor (integrin) in the uterus of rats in relation to the estrous cycle

The expression and localization of fibronectin receptor (integrin), fibronectin, laminin and collagen type IV in the endometrium of the rat uterus during each period of the estrous cycle were investigated by immunofluorescent microscopy. Fibronectin receptor was observed at the epithelial cells of the endometrium and at vascular endothelial cells. At proestrus, when epithelial cells actively migrate, fibronectin receptor was observed at the basal and lateral epithelial cell surfaces. During estrus, fibronectin receptor had begun to disappear, little fibronectin receptor was observed at metestrus or diestrus. No prominent changes in the localization of fibronectin (seen at the vascular endothelial cells and in the stroma) or of laminin and collagen type IV (seen at the muscles and at the basement membranes of the epithelial and vascular endothelial cells) were observed in relation to the estrous cycle. Thus, uterine epithelial cells, like epithelial cells of the healing cornea, increase their expression of fibronectin receptor during active migration, probably facilitating their attachment to stromal fibronectin. This fibronectin-fibronectin receptor mechanism may underlie epithelial repair, whether the defect results from physiological processes or from an insult.

Hard contact lens-induced corneal neovascularization treated by oxygenation

A 39-year-old man suffered from corneal neovascularization, through to result from hypoxia caused by improper use of polymethylmethacrylate (PMMA) hard contact lenses. The condition was successfully treated by oxygenation of the corneas under swimming goggles.

Differential modes of action of fibronectin and epidermal growth factor on rabbit corneal epithelial migration

In order to clarify the roles of fibronectin (FN) and epidermal growth factor (EGF) in corneal wound healing, we cultured blocks of excised rabbit cornea for 24 hours in media containing one of these agents, then measured the length of the path of the epithelial layer that had migrated down the side of the block. Both FN and EGF stimulated epithelial migration significantly in a dose-dependent fashion. Responses to EGF involved a time lag of at least 12 hours before stimulation could be observed, but there was no lag-time for FN-stimulated migration. FN was maximally effective only if it was continuously present. In contrast, exposure to EGF for 6 hours did not stimulate epithelial migration, but exposure for 9 hours resulted in the same stimulatory effects as were observed after 24 hours' continuous exposure. Anti-FN antibody inhibited the FN- and EGF-stimulated migration of corneal epithelium. But anti-EGF antibody inhibited only EGF-stimulated migration and had no effect on FN-stimulated migration. These results indicate that, unlike FN, EGF need not be present, once the epithelial cells have recognized its signal. Furthermore, the stimulatory effect of EGF depended on FN, while that of FN was independent of EGF. The effects of EGF on migration of corneal epithelium may, therefore, be mediated by FN.

Role of actin filaments and microtubules in the spreading of rabbit corneal epithelial cells on the fibronectin matrix

The migration and spreading of the corneal epithelial cells adjacent to a wound is the first step in successful epithelial resurfacing. To understand the role of actin filaments and microtubules of the cytoskeletal system in the spreading of corneal epithelial cells, we plated rabbit corneal epithelial cells on a fibronectin matrix and studied the effects of cytochalasin B, which inhibits actin filaments assembly, and colchicine, which inhibits microtubules assembly, on the ability of individual cells to spread. Changes in the morphology of actin filaments and microtubules were also studied using immunofluorescent microscopy. The area of spread epithelial cells depended on the concentration of fibronectin used to coat the surface. In spread cells, stress fibers of actin filaments were evenly distributed throughout the cytoplasm, whereas microtubules were observed only at the perinuclear region. The presence of cytochalasin B during the cell attachment and spreading decreased the area of the spread cells more than did colchicine. However, once the epithelial cells were spread on a fibronectin matrix, cytochalasin B and colchicine each decreased the cell area only slightly, and to the same extent. These results indicate that formation of actin filaments is more important than formation of microtubules to the spreading of corneal epithelial cells.

A peptide from fibronectin cell-binding domain inhibits attachment of epithelial cells

Fibronectin is thought to contribute to the adhesion of corneal epithelium during epithelial repair, presumably by mediating epithelial attachment to stroma during epithelial migration. To understand the mechanism of the interaction of fibronectin with corneal epithelial cells, the effects of intact plasma fibronectin and of the synthetic peptide GRGDS from the cell binding domain of fibronectin on the attachment of rabbit corneal epithelial cells have been examined. When dissociated epithelial cells were plated on fibronectin-coated culture dishes, the number of cells that attached increased in proportion to the concentration of fibronectin used for coating. Attachment of the cells was inhibited in a dose-dependent manner by the addition of GRGDS to the medium. When the structurally similar control peptide GRGES was added to the medium, no inhibition of epithelial attachment was observed. Our results demonstrate that corneal epithelial cells use an adhesive recognition system for fibronectin related to that characterized in fibroblastic cells. GRGDS thus presumably competes for receptors for fibronectin on epithelial cells of the cornea, resulting in the observed inhibition of attachment. These results identify a mechanism for corneal epithelial cell adhesion, compatible with that identified in other fibronectin-dependent systems.

The network structure of corneal fibroblasts in the rat as revealed by scanning electron microscopy

The overall morphology of corneal fibroblasts in the rat was examined by scanning electron microscopy after removal of stromal connective tissue elements by tryptic digestion and HCl hydrolysis. Fibroblasts thus exposed were flattened and stellate cells with a diameter of about 10 micron and had a number of ramified cytoplasmic processes. The processes of neighboring fibroblasts contacted with each other to form an extensive and continuous network structure parallel to the plane of collagenous lamellae. The cellular network appears to constitute an integrated system which may be involved in the synchronized regulation of the metabolic and physiologic homeostasis for the maintenance of corneal transparency.

Interactions of extracellular collagen and corneal fibroblasts: morphologic and biochemical changes of rabbit corneal cells cultured in a collagen matrix

Corneal fibroblasts, also known as keratocytes are surrounded by an extracellular matrix of collagen in vivo. To understand the physiology and pathology of these corneal fibroblasts, it is important to study their interactions with this extracellular matrix. We cultured rabbit corneal fibroblasts on tissue culture plastic dishes or in a hydrated collagen gel and compared the changes in morphology and mitotic activity. Corneal fibroblasts on plastic dishes were flattened and widely spread, whereas those in collagen gel became spindle-shaped with long processes. Examination with an electron microscope revealed that the corneal fibroblasts in collagen gel formed gap junctions with neighboring cells. Gap junctions were hardly ever observed between corneal fibroblasts cultured on plastic dishes. Corneal fibroblasts cultured in a collagen matrix showed much less incorporation of [3H]thymidine than did corneal fibroblasts cultured on plastic, and this incorporation decreased with increasing concentration of collagen. Our present results suggest that the morphologic and biochemical characteristics of corneal fibroblasts cultured in collagen gel are different from those cultured on plastic.

Dexamethasone inhibition of phagocytosis by corneal keratocytes in culture

To investigate the effect of steroid hormone on phagocytosis by keratocytes, we investigated the phagocytic activities for latex beads by rabbit corneal keratocytes cultured in the presence and absence of dexamethasone. Culturing the corneal keratocytes with dexamethasone inhibited the phagocytic activity for latex beads in a dose-dependent fashion. The inhibitory effects depended on the duration of the incubation period of the cells with dexamethasone.

Uptake of india ink particles and latex beads by corneal fibroblasts

The fate of India ink particles and polystyrene latex beads injected into the corneal stroma of rabbits was studied by the naked eye, light microscopy, and electron microscopy. All the injected ink particles or latex beads were unchanged in shape, size, and number for at least 6 months. India ink particles and latex beads were endocytosed by the corneal fibroblasts within 3-4 days after injection. Numerous ink particles were packed into vacuoles, 0.5-10 micron in diameter, which occupy a large volume of the cytoplasm of the cell body and processes of fibroblasts in and near the injected area. Each latex bead, 0.72 micron in diameter, is usually enclosed in one vesicle, and a large number of vesicles are distributed throughout the cytoplasm. In corneal tissue removed 10 min after injection of India ink and cultured for 3 or 7 days, uptake of many ink particles by the fibroblasts was seen. By this experiment, the contribution of the blood-derived cells was completely excluded, and it is more distinctly shown that the corneal fibroblast has a strong endocytotic activity. The uptake and long-term storage of ink particles and latex beads by the corneal fibroblast are reactions that protect the organ without inflammation from the injury and harm by non-toxic foreign materials.

Fibronectin enhances the phagocytic activity of cultured rabbit keratocytes

To understand corneal stromal wound healing, we investigated the effect of fibronectin (FN) on the phagocytic activity of rabbit corneal keratocytes. The phagocytic activity was measured under a dark-field microscope using polystyrene latex beads coated with either FN, bovine serum albumin, or both. The percentage of phagocytosing cells increased with incubation time. The phagocytic activity was significantly enhanced in a dose-dependent manner by coating beads with FN. Transmission electron microscopic observation demonstrated that latex beads were internalized into the cytoplasm and were surrounded by limiting membranes. These findings indicate that cultured keratocytes have phagocytic activity and that FN shows opsonic effects on the phagocytic activity, suggesting that FN in the injured corneal stroma may play a role in stromal wound repair by opsonic effects.

Electron-microscopic studies on the presence of gap junctions between corneal fibroblasts in rabbits

Corneal fibroblasts, major cellular components of the corneal stroma, are loosely arrayed between collagen lamellae. They play an important role in the metabolic and physiological homeostasis mechanisms by which the cornea is kept transparent. This paper deals with the demonstration of the gap junctions between the corneal fibroblasts of rabbits by transmission electron microscopy of thin sections and of freeze-fracture specimens. Under the transmission electron microscope, the corneal fibroblasts are seen between the lamellae of collagen fibers of the corneal stroma. Their long cytoplasmic processes are in contact with those of neighboring fibroblasts. Typical gap junctions are found between these cytoplasmic processes. In the freeze-fracture images, intramembrane particles with a diameter of 10.3 nm form polygonal aggregates on P faces. These findings suggest that corneal fibroblasts, coupled with each other, might function synchronously through gap junctions responsible for metabolic activities essential for the maintenance of corneal transparency.

Spontaneous persistent epithelial defects after cataract surgery

The natural course of corneal epithelial erosions was studied retrospectively in patients who had undergone cataract extraction. Postsurgical epithelial defects occurred in 41 of 796 eyes (5.2%). No difference was noted in the incidence of corneal epithelial defects in eyes after intracapsular and extracapsular cataract extraction. Of the 41 eyes with postsurgical corneal erosion, 26 (63.4%) showed corneal epithelial defects by the fourth postoperative day. Two types of healing patterns were noticed: rapid healing, in which there was complete epithelial resurfacing within 4 days after the onset of corneal erosion, and prolonged healing, in which the period for complete epithelial resurfacing ranged from 6 days to 34 days. Since fibronectin has been reported to facilitate corneal epithelial migration and adhesion, we administered autologous fibronectin eyedrops in five cases in which corneal epithelial erosion developed after cataract surgery and epithelial defects persisted more than 7 days (except case 2). In all cases receiving fibronectin treatment, complete epithelial resurfacing occurred. In four of five eyes, fibronectin eye drops were effective in epithelial resurfacing within 3-9 days; in one eye, complete epithelial resurfacing took 23 days. The present results indicate that there are two types of epithelial erosion and that fibronectin eye drops might be one of the possible therapeutic approaches for the treatment of persistent epithelial defects. The weak adhesion of epithelial cells underlying basement membrane may be one of the causes of persistent epithelial defects after cataract surgery.