Purpose. To determine the expression of differentiation and progenitor cell markers in corneal tissues that previously underwent autologous cultivated oral mucosal epithelial transplantation (COMET). methods. Four eyes from three alkaline-injured patients and one thermally injured patient underwent COMET to promote re-epithelialization or corneal reconstruction. Between 10 and 22 months (mean, 14.2 ± 5.5 months [SD]) after COMET, the corneal tissues were obtained after penetrating keratoplasty (n = 1) or autologous limbal transplantation (n = 3). Immunoconfocal microscopy for keratin (K)3, -12, -4, -13, and -8; connexin (Cx)43; MUC5AC; laminin-5; pan-p63; ABCG2; and p75 was performed in those specimens as well as in the oral mucosa and cultivated oral mucosal epithelial cells (OMECs). results. All four specimens were unanimously positive for K3, -4, and -13 but negative for K8 and MUC5AC, suggesting that the keratinocytes were oral mucosa-derived. However, peripheral K12 staining was positive only in patient 2, suggesting a mixed oral and corneal epithelium in that case. Cx43 staining in the basal epithelium was negative in patients 1, 2, and 3, but was positive in patient 4. Small, compact keratinocytes in the basal epithelium preferentially expressed pan-p63, ABCG2, and p75. Although the staining of pan-p63 and ABCG2 tended to be more than one layer, signal for p75 was consistently localized only to the basal layer. conclusions. The study demonstrated the persistence of transplanted OMECs in human corneas. In addition, small, compact cells in the basal epithelium preferentially expressed the keratinocyte stem/progenitor cell markers, which may be indicative of the engraftment of the progenitor cells after transplantation.

Fungal keratitis (FK) is the most devastating and vision-threatening microbial keratitis, but clinical diagnosis a great challenge. This study aimed to develop and verify a deep learning (DL)-based corneal photograph model for diagnosing FK. Corneal photos of laboratory-confirmed microbial keratitis were consecutively collected from a single referral center. A DL framework with DenseNet architecture was used to automatically recognize FK from the photo. The diagnoses of FK via corneal photograph for comparing DL-based models were made in the Expert and NCS-Oph group through a majority decision of three non-corneal specialty ophthalmologist and three corneal specialists, respectively. The average percentage of sensitivity, specificity, positive predictive value, and negative predictive value was approximately 71, 68, 60, and 78. The sensitivity was higher than that of the NCS-Oph (52%, P < .01), whereas the specificity was lower than that of the NCS-Oph (83%, P < .01). The average accuracy of around 70% was comparable with that of the NCS-Oph. Therefore, the sensitive DL-based diagnostic model is a promising tool for improving first-line medical care at rural area in early identification of FK.

In this study, we aimed to develop a deep learning model for identifying bacterial keratitis (BK) and fungal keratitis (FK) by using slit-lamp images. We retrospectively collected slit-lamp images of patients with culture-proven microbial keratitis between 1 January 2010 and 31 December 2019 from two medical centers in Taiwan. We constructed a deep learning algorithm consisting of a segmentation model for cropping cornea images and a classification model that applies different convolutional neural networks (CNNs) to differentiate between FK and BK. The CNNs included DenseNet121, DenseNet161, DenseNet169, DenseNet201, EfficientNetB3, InceptionV3, ResNet101, and ResNet50. The model performance was evaluated and presented as the area under the curve (AUC) of the receiver operating characteristic curves. A gradient-weighted class activation mapping technique was used to plot the heat map of the model. By using 1330 images from 580 patients, the deep learning algorithm achieved the highest average accuracy of 80.0%. Using different CNNs, the diagnostic accuracy for BK ranged from 79.6% to 95.9%, and that for FK ranged from 26.3% to 65.8%. The CNN of DenseNet161 showed the best model performance, with an AUC of 0.85 for both BK and FK. The heat maps revealed that the model was able to identify the corneal infiltrations. The model showed a better diagnostic accuracy than the previously reported diagnostic performance of both general ophthalmologists and corneal specialists.

Sjögren syndrome (SS) or dry eye disease (DED) is one of the most complicated ocular surface diseases. The goal of this study is to elucidate the relationship of the changes in clinical indices of tear film (TF) homeostasis with respect to tear components to allow for SS-DED monitoring and avoid stably controlled SS-DED patients from re-entering a vicious cycle. This prospective case-control study compared stable SS-DED patients with non-SS-DED control from several aspects, including clinical indices for TF homeostasis, 2 DED diagnostic biomarkers (MMP-9 and lactoferrin), and the proteome of flush tears. Compared with non-SS-DED controls, stably controlled SS-DED subjects had less tear secretion and higher ocular surface inflammation, a higher concentration ratio of tear MMP-9/lactoferrin, a more diverse tear proteome, and lower spectral intensities of lipocalin-1, lacritin, and prolactin-inducible protein among the abundant tear proteins. For stable SS-DED patients, the concentration ratio of tear MMP-9/lactoferrin and the corrected lipocalin-1 signal was positively correlated with ocular inflammation and TF stability, respectively. MMP-9 released from stressed ocular surface epithelium and lipocalin-1 secreted from the energetic lacrimal gland are two tear biomarkers responding well to TF homeostasis. The tear proteomics approach through flush tears is a promising method for monitoring SS-DED patients with a standardized sampling procedure and lactoferrin-corrected analysis.

Tyrosine kinase inhibitors (TKIs) are currently the first-line treatment for non-small cell lung cancer (NSCLC) patients with epidermal growth factor receptor (EGFR) mutations. These patients receive platinum-based chemotherapy as the second-line treatment after they develop resistance to TKIs. Many patients regain sensitivity to the TKIs used in the first-line treatment after the failure of chemotherapy. However, the molecular mechanism for the regain of TKI sensitivity is largely unknown. In this study, we established gefitinib-resistant PC9 and HCC827 cell lines, which did not harbor the EGFR T790M mutation and MET amplification but exhibited the epithelial-mesenchymal transition (EMT) phenotype. Overexpression of EMT inducers, Snail or Slug, in the parental lines promoted their resistance to gefitinib. The gefitinib-resistant cell lines regained their sensitivity to gefitinib and displayed reverse EMT phenotypes after long-term culture in gefitinib-free culture medium. Blockage of reverse EMT by stable expression of Snail or Slug prevented the regain of TKI sensitivity. In conclusion, reverse EMT is one of the major mechanisms for the regain of TKI sensitivity in TKI-resistant NSCLC cells, suggesting that the development of small molecules targeting the EMT process may prolong the efficacy of TKIs in NSCLC patients with EGFR mutations.