Reticular Pseudodrusen: Interreader Agreement of Evaluation on OCT Imaging in Age-Related Macular Degeneration

Purpose

To determine the interreader agreement for reticular pseudodrusen (RPD) assessment on combined infrared reflectance (IR) and OCT imaging in the early stages of age-related macular degeneration across a range of different criteria to define their presence.

Design

Interreader agreement study.

Participants

Twelve readers from 6 reading centers.

Methods

All readers evaluated 100 eyes from individuals with bilateral large drusen for the following: (1) the presence of RPD across a range of different criteria and (2) the number of Stage 2 or 3 RPD lesions (from 0 to ≥ 5 lesions) on an entire OCT volume scan and on a selected OCT B-scan. Supportive information was available from the corresponding IR image.

Main Outcome Measures

Interreader agreement, as assessed by Gwet's first-order agreement coefficient (AC₁).

Results

When evaluating an entire OCT volume scan, there was substantial interreader agreement for the presence of any RPD, any or ≥ 5 Stage 2 or 3 lesions, and ≥ 5 definite lesions on en face IR images corresponding to Stage 2 or 3 lesions (AC₁ = 0.60-0.72). On selected OCT B-scans, there was also moderate-to-substantial agreement for the presence of any RPD, any or ≥ 5 Stage 2 or 3 lesions (AC₁ = 0.58-0.65) and increasing levels of agreement with increasing RPD stage (AC₁ = 0.08, 0.56, 0.78, and 0.99 for the presence of any Stage 1, 2, 3, and 4 lesions, respectively). There was substantial agreement regarding the number of Stage 2 or 3 lesions on an entire OCT volume scan (AC₁ = 0.68), but only fair agreement for this evaluation on selected B-scans (AC₁ = 0.30).

Conclusions

There was generally substantial or near-substantial-but not near-perfect-agreement for assessing the presence of RPD on entire OCT volume scans or selected B-scans across a range of differing RPD criteria. These findings underscore how interreader variability would likely contribute to the variability of findings related to the clinical associations of RPD. The low levels of agreement for assessing RPD number on OCT B-scans underscore the likely challenges of quantifying RPD extent with manual grading.

Financial Disclosures

Proprietary or commercial disclosure may be found after the references.

Deep learning-based classification and segmentation of retinal cavitations on optical coherence tomography images of macular telangiectasia type 2

AIM

To develop a fully automatic algorithm to segment retinal cavitations on optical coherence tomography (OCT) images of macular telangiectasia type 2 (MacTel2).

METHODS

The dataset consisted of 99 eyes from 67 participants enrolled in an international, multicentre, phase 2 MacTel2 clinical trial (NCT01949324). Each eye was imaged with spectral-domain OCT at three time points over 2 years. Retinal cavitations were manually segmented by a trained Reader and the retinal cavitation volume was calculated. Two convolutional neural networks (CNNs) were developed that operated in sequential stages. In the first stage, CNN1 classified whether a B-scan contained any retinal cavitations. In the second stage, CNN2 segmented the retinal cavitations in a B-scan. We evaluated the performance of the proposed method against alternative methods using several performance metrics and manual segmentations as the gold standard.

RESULTS

The proposed method was computationally efficient and accurately classified and segmented retinal cavitations on OCT images, with a sensitivity of 0.94, specificity of 0.80 and average Dice similarity coefficient of 0.94±0.07 across all time points. The proposed method produced measurements that were highly correlated with the manual measurements of retinal cavitation volume and change in retinal cavitation volume over time.

CONCLUSION

The proposed method will be useful to help clinicians quantify retinal cavitations, assess changes over time and further investigate the clinical significance of these early structural changes observed in MacTel2.

Retinal cavitations in macular telangiectasia type 2 (MacTel): longitudinal structure-function correlations

BACKGROUND/AIMS

To quantify retinal cavitation size over time in macular telangiectasia type 2 (MacTel) and to correlate changes with visual acuity and area of ellipsoid zone loss.

METHODS

Optical coherence tomography (OCT) macula volume scans from sham eyes included in a prospective, phase II clinical trial of human ciliary neutrophic factor for MacTel at baseline, 1 year and 2 years of follow-up were analysed. Cavitations were segmented by two independent readers. Total cavitation volume was compared with area of ellipsoid zone loss and best-corrected visual acuity (BCVA).

RESULTS

Fifty-one eyes from 51 unique patients (mean age 62 years, range 45-79 years) were included. Intraclass correlation between readers for cavitation volume was excellent (>0.99). Average cavitation volume was 0.0109 mm³, 0.0113 mm³ and 0.0124 mm³ at baseline, 1 year and 2 years, respectively. The average rate of cavitation volume change was +0.0039 mm³/year. 10 eyes (20%) had a significant change in cavitation volume during the study (3 decreased, 7 increased). Eyes with increased cavitation volume had worse BCVA compared with eyes with no change/decreased cavitation volume (71.5 vs 76.1 ETDRS letters, respectively). Cavitation volume was negatively correlated to BCVA (r=-0.37) but not to area of ellipsoid zone loss. Cavitation volume was negatively predictive of BCVA in both univariate and multivariate mixed-effects modelling with ellipsoid zone loss.

CONCLUSIONS

Retinal cavitations and their rate of change in MacTel can be reliably quantified using OCT. Cavitations are negatively correlated with visual acuity and may be a useful OCT-based biomarker for disease progression and visual function in MacTel.

Effect of clinically relevant culture conditions on antimicrobial susceptibility of Chlamydia trachomatis

Infection of polarized human endometrial-gland epithelial cells obtained at hysterectomy with Chlamydia trachomatis serovar E may provide a relevant in vitro model for studies of pharmacokinetics in genital chlamydial infections. The minimal bactericidal concentration of azithromycin against C. trachomatis was lower in this model than in studies with nonpolarized cells (0.125 and 0.5 mg/L, respectively). Polarized cells also internalized more azithromycin over 24 hours. The results indicate that the eradication of chlamydial infections may be difficult to prove by antigen detection methods: the persistence of chlamydial envelope material within the intracellular vacuoles of azithromycin-treated cells may lead to a false-positive diagnosis of persisting chlamydial infection.

In-vitro activity of azithromycin on Chlamydia trachomatis infected, polarized human endometrial epithelial cells

The in-vitro activity of azithromycin on Chlamydia trachomatis infected human endometrial epithelial cells, both primary and transformed cells growing in a polarized and non-polarized orientation, was analyzed. Addition of azithromycin two hours after adsorption inoculation with continued exposure until 72 h gave an MIC90 and MBC90 of 0.063 and 0.5 mg/L, respectively. In addition, the MBC results were more pronounced in infected cells growing in a polarized orientation. Numerous small fluorescent 'spots' (presumed small abnormal inclusions) were visible in the infected cells exposed to MIC concentrations of azithromycin. Immuno-transmission electron microscopy examination revealed intracellular inclusions filled with chlamydial envelope ghosts. Since standard diagnostic antigen detection methods use anti-envelope antibodies, the aberrant envelope-filled inclusions might be interpreted as viable inclusions by fluorescent microscopy and result in high false positive readings. To simulate treatment of an infected patient, azithromycin was added at 18 h to infected cells containing many reticulate bodies and exposure continued for 54 h after which killing of chlamydiae was seen. The use of polarized human cells may offer a more relevant in-vitro model system for examining the efficacy of antimicrobial action.

Examination of chlamydial glycolipid with monoclonal antibodies: cellular distribution and epitope binding

A chlamydial glycolipid antigen (GLXA) is shed into the medium of C. trachomatis-infected cell cultures. This study screened monoclonal antibodies (mAb), prepared in different laboratories by immunization with embryonated egg propagated elementary bodies (EB), for their ability to bind with infected cells and to react with purified GLXA isolated from supernatants of infected McCoy cells. The fluorescent antibody (FA) staining pattern exhibited by a number of mAb indicated that they bound antigen present within the inclusion and at the inner membrane surface of infected cells; the observed pattern differs significantly from the distribution seen when anti-lipopolysaccharide (LPS) (mAb) were used. The staining pattern observed by immunofluorescence was confirmed and extended by ultrastructure studies of immunogold-labelled, infected human endometrial gland epithelial cells (HEGEC) and a human endometrial carcinoma-derived cell line (RL95-2). Additionally, the immunoelectron microscope studies revealed binding within the inclusion and on reticulate bodies, within the cell cytoplasm and at the surface of infected cells. The specificity of the reactive mAb, examined by molecular shift chromatography and isolated, affinity-purified GLXA, indicated that two mAb of the IgG isotype recognized an antigen which had been purified from tissue culture supernatants by affinity chromatography using an IgM mAb. The results suggest that GLXA is an important determinant whose role and function during in vitro and in vivo infections deserves further analyses.

Recombinant Escherichia coli clones expressing Chlamydia trachomatis gene products attach to human endometrial epithelial cells

To identify Chlamydia trachomatis genes involved in attachment to host cells, a chlamydial genomic library was screened on the basis of binding characteristics by two methods. In the whole-cell screen, individual recombinant Escherichia coli clones were assayed for adherence to eukaryotic cells. In the membrane-binding screen, each recombinant colony of E. coli was treated with CHCl3 and assayed for binding to purified, 3-[(3-cholamidopropyl)-dimethyl-ammonio]-1-propanesulfonate (CHAPS)-solubilized, 35S-labeled eukaryotic membrane material. Initial screening with McCoy cells was refined by using HEC-1B cells, a human endometrial epithelial cell line, which discriminate among recombinants adhering to McCoy cells. Some recombinants demonstrate significantly greater adherence to HEC-1B cells than to McCoy cells and appear, by transmission electron microscopy, to associate with electron-dense areas of the epithelial cell plasma membrane, resembling coated pits. Recombinants positive by one or both screening methods were examined by Southern and Western (immunoblot) analyses, which revealed the presence of chlamydial sequences inserted in the plasmids and the expression of novel 18-, 28-, and approximately 82 kDa, and perhaps of 18 Maxicell analysis of selected recombinants confirmed that the proteins of 28 and approximately 82 kDa, and perhaps of 18 kDa, are plasmid encoded. Antiserum generated against the recombinant approximately 82-kDa protein reacted in Western analysis with a similar-sized protein from C. trachomatis serovar E elementary bodies (EB) and reticulate bodies, serovar L2 EB, and C. psittaci EB. E. coli JM109(pPBW58) contains a 6.7-kb plasmid insert which encodes proteins of all three sizes. Under a number of different conditions in the whole-cell attachment assay--i.e., at 4 degrees C, in Ca(2+)- and Mg(2+)-free medium, in the presence of trypsin or dextran sulfate, and with rabbit aortic endothelial cells--the binding specificity of JM109(pPBW58) parallels that of C. trachomatis EB. Finally, the adherence phenotype of E. coli JM109(pPBW58) correlates directly with the presence of the recombinant plasmid; the phenotype is lost concurrently with loss of the recombinant plasmid, and the into E. coli JM109. The role of the 18-, 28-, and approximately 82-kDa proteins in mediating attachment, whether they act in concert as a complex or individually, has yet to be determined.

Entry of genital Chlamydia trachomatis into polarized human epithelial cells

To study the initial invasion process(es) of genital chlamydiae, a model system consisting of hormonally maintained primary cultures of human endometrial gland epithelial cells (HEGEC), grown in a polarized orientation on collagen-coated filters, was utilized. After Chlamydia trachomatis inoculation of the apical surface of polarized HEGEC, chlamydiae were readily visualized, by transmission electron microscopy, in coated pits and coated vesicles. This was true for HEGEC maintained in physiologic concentrations of estrogen (proliferative phase) and of estrogen plus progesterone (secretory phase), despite the finding that association of chlamydiae with secretory-phase HEGEC is significantly reduced (P = 0.025; A.S. Maslow, C.H. Davis, J. Choong, and P.B. Wyrick, Am. J. Obstet. Gynecol. 159:1006-1014, 1988). In contrast, chlamydiae were rarely observed in the clathrin-associated structures if the HEGEC were cultured on plastic surfaces. The same pattern of coated pit versus noncoated pit entry was reproducible in HeLa cells. The quantity of coated pits associated with isolated membrane sheets derived from HeLa cells, grown on poly-L-lysine-coated cover slips in medium containing the female hormones, was not significantly different as monitored by radiolabeling studies and by laser scanning microscopy. These data suggest that culture conditions which mimic in vivo cellular organization may enhance entry into coated pits for some obligate intracellular pathogens.

Estrogen enhances attachment of Chlamydia trachomatis to human endometrial epithelial cells in vitro

Ultraviolet light-inactivated elementary bodies of Chlamydia trachomatis serovar E were fluorescently tagged with rhodamine isothiocyanate (5 micrograms/ml) and added to primary cultures of human endometrial gland epithelial cells. The elementary bodies, at a multiplicity of infection of 600:1, were allowed to adsorb to the cell monolayers for 1 hour at 35 degrees C in an atmosphere of 5% carbon dioxide. The monolayers were disaggregated by trypsinization and the individual cells were processed in the fluorescent activated cell sorter for chlamydial attachment. This method of analysis revealed attachment of C. trachomatis to approximately 50% of human endometrial gland epithelial cells. Addition of estrogen (10(-10) mol/L) to the culture medium enhanced chlamydial attachment to human endometrial gland epithelial cells to approximately 80% (p less than or equal to 0.005), and progesterone in combination with estrogen reduced chlamydial attachment in a dose-dependent fashion: 1 ng/ml progesterone, approximately 50%; 5 ng/ml, about 30%; 10 ng/ml, about 18%, respectively (p less than or equal to 0.025).

Ultrastructural study of endocytosis of Chlamydia trachomatis by McCoy cells

The entry of Chlamydia trachomatis into McCoy cells (fibroblasts) was studied by transmission electron microscopy. On adsorption of elementary bodies (EBs) to host cells at 37 degrees C, the EBs were bound primarily to preexisting cell-surface microvilli. They were also observed in coated pits located at the bases of the microvilli and along smooth surfaces of the host cells and were internalized within coated vesicles at this temperature. Postembedding immunogold labeling on Lowicryl thin sections with anti-clathrin antibody as the primary reagent revealed the gold marker localized in pits and vesicles containing chlamydiae. Some EBs were present in smooth-surfaced invaginations at or near the bases of microvilli and in vesicles devoid of distinguishable coat material. A similar entry process was observed with centrifugation-assisted inoculation of EBs onto the McCoy cells. Individual EBs were initially internalized into tightly bound endocytic vesicles. However, within 1 to 3 h postinfection, multiple C. trachomatis EBs were observed in large, loosely bound vesicles. Evidence suggests that vesicles containing C. trachomatis may have fused with one another early in the infectious process. These results indicate that chlamydiae can exploit the specific process of adsorptive endocytosis for entry into host cells and for translocation to a given intracellular destination, which may be different for each species.

Comparison of scores obtained on the PPVT and the PPVT-R

The purpose of this study was to compare age equivalent values obtained from the PPVT and PPVT-R. The 80 subjects, ages 3:6 to 4:6 years old, were presented one form each of the original and revised forms of the test. Results indicate the subjects achieved age equivalent scores significantly closer to their chronological ages on the PPVT-R.