April 2017

OBJECTIVE: To examine early performance on an eye surgery simulator and its relationship to subsequent live surgical performance in a single large residency program. DESIGN: Retrospective study. SETTING: Massachusetts Eye and Ear, Harvard Medical School, Department of Ophthalmology. METHODS: In a retrospective study, we compared performance of 30 first-year ophthalmology residents on an eye surgery simulator to their surgical skills as third-year residents. Variables collected from the eye surgery simulator included scores on the following modules of the simulator (Eyesi, VRmagic, Mannheim, Germany): antitremor training level 1, bimanual training level 1, capsulorhexis level 1 (configured), forceps training level 1, and navigation training level 1. Subsequent surgical performance was assessed using the total number of phacoemulsification cataract surgery cases for each resident, as well as the number performed as surgeon during residency and scores on global rating assessment of skills in intraocular surgery (GRASIS) scales during the third year of residency. Spearman correlation coefficients were calculated between each of the simulator performance and subsequent surgical performance variables. We also compared variables in a small group of residents who needed extra help in learning cataract surgery to the other residents in the study. MAIN OUTCOME MEASURES: Relationships between Eyesi scores early in residency and surgical performance measures in the final year of residency. RESULTS: A total of 30 residents had Eyesi data from their first year of residency and had already graduated so that all subsequent surgical performance data were available. There was a significant correlation between capsulorhexis task score on the simulator and total surgeries (r = 0.745, p = 0.008). There was a significant correlation between antitremor training level 1 (r = 0.554, p = 0.040), and forceps training level 1 (r = 0.622, p = 0.023) with primary surgery numbers. There was a significant correlation between forceps training level 1 (r = 0.811, p = 0.002), and navigation training level 1 (r = 0.692, p =0.013) with total GRASIS score. There was a significant inverse correlation between total GRASIS score and residents in need of extra help (r = -0.358, p =0.003). CONCLUSION: Module scores on an eye surgery simulator early in residency may predict a resident׳s future performance in the operating room. These scores may allow early identification of residents in need of supplemental training in cataract surgery.

Transposons can be used to easily generate and label the location of mutations throughout bacterial and other genomes. Transposon insertion mutants may be screened for a phenotype as individual isolates, or by selection applied to a pool of thousands of mutants. Identifying the location of a transposon insertion is critical for connecting phenotype to the genetic lesion. In this unit, we present an easy and detailed approach for mapping transposon insertion sites using arbitrarily-primed PCR (AP-PCR). Two rounds of PCR are used to (1) amplify DNA spanning the transposon insertion junction, and (2) increase the specific yield of transposon insertion junction fragments for sequence analysis. The resulting sequence is mapped to a bacterial genome to identify the site of transposon insertion. In this protocol, AP-PCR as it is routinely used to map sites of transposon insertion within Staphylococcus aureus, is used to illustrate the principle. Guidelines are provided for adapting this protocol for mapping insertions in other bacterial genomes. Mapping transposon insertions using this method is typically achieved in 2 to 3 days if starting from a culture of the transposon insertion mutant. © 2017 by John Wiley & Sons, Inc.

The use of viral vectors for inner ear gene therapy is receiving increased attention for treatment of genetic hearing disorders. Most animal studies to date have injected viral suspensions into neonatal ears, via the round window membrane. Achieving transduction of hair cells, or sensory neurons, throughout the cochlea has proven difficult, and no studies have been able to efficiently transduce sensory cells in adult ears while maintaining normal cochlear function. Here, we show, for the first time, successful transduction of all inner hair cells and the majority of outer hair cells in an adult cochlea via virus injection into the posterior semicircular canal. We used a "designer" AAV, AAV2/Anc80L65, in which the main capsid proteins approximate the ancestral sequence state of AAV1, 2, 8, and 9. Our injections also transduced ~10% of spiral ganglion cells and a much larger fraction of their satellite cells. In the vestibular sensory epithelia, the virus transduced large numbers of hair cells and virtually all the supporting cells, along with close to half of the vestibular ganglion cells. We conclude that this viral vector and this delivery route hold great promise for gene therapy applications in both cochlear and vestibular sense organs.

Antigen-presenting cells (APCs) play an important role in transplant rejection and tolerance. In high-risk corneal transplantation, where the graft bed is inflamed and vascularized, immature APCs in the donor corneal stroma quickly mature and migrate to lymphoid tissues to sensitize host T cells. In this study, using a mouse model of corneal transplantation, we investigated whether enrichment of tolerogenic APCs (tolAPCs) in donor corneas can enhance graft survival in corneal allograft recipients with inflamed graft beds. Treatment of donor corneas with interleukin-10 (IL-10) and transforming growth factor-β1 (TGFβ1) altered the phenotype and function of tissue-residing APCs. Transplantation of these tolAPC-enriched corneas decreased frequencies of interferon gamma (IFNγ)(+) effector T cells (Teffs), as well as allosensitization in the hosts, diminished graft infiltration of CD45(+) and CD4(+) cells, and significantly improved corneal allograft survival compared to saline-injected controls. These data provide a novel approach for tolAPC-based immunotherapy in transplantation by direct cytokine conditioning of the donor tissue.

PURPOSE: To determine whether botulinum toxin is as effective as strabismus surgery in the treatment of acute-onset comitant esotropia in children. DESIGN: Retrospective, nonrandomized, comparative clinical study. METHODS: Setting: Tertiary care pediatric hospital. STUDY POPULATION: Forty-nine children with acute-onset comitant esotropia. INTERVENTION: Treatment with either botulinum toxin ("chemodenervation group") or standard incisional strabismus surgery ("surgery group"). MAIN OUTCOME MEASURE: Success rate at 6 months (total horizontal deviation of 10 prism diopters or less and evidence of binocular single vision). RESULTS: There were 16 patients in the chemodenervation group and 33 patients in the surgery group. The success rate was not significantly different at 6 months (81% vs 61%, P = .20) or at 18 months (67% vs 58%, P = .74). The median angle of deviation and median stereoacuity were not significantly different at 6 or 18 months. The chemodenervation procedure was not inferior to incisional strabismus surgery at 6 months. The duration of general anesthesia (5 vs 71 min, P < .001) and time in the post-anesthesia care unit (37 vs 93 min, P < .001) were significantly shorter in the chemodenervation group. Botulinum toxin injection payment averaged $874 per procedure compared with $2783 for strabismus surgery. CONCLUSIONS: Botulinum toxin is at least as effective as surgery in the treatment of acute-onset comitant esotropia at 6 months while reducing the duration of general anesthesia and healthcare costs.

PURPOSE: To compare diagnostic performance and structure-function correlations of multifocal electroretinogram (mfERG), full-field flash ERG (ff-ERG) photopic negative response (PhNR) and transient pattern-reversal ERG (PERG) in a non-human primate (NHP) model of experimental glaucoma (EG). METHODS: At baseline and after induction of chronic unilateral IOP elevation, 43 NHP had alternating weekly recordings of retinal nerve fiber layer thickness (RNFLT) by spectral domain OCT (Spectralis) and retinal function by mfERG (7F slow-sequence stimulus, VERIS), ff-ERG (red 0.42 log cd-s/m(2) flashes on blue 30 scotopic cd/m(2) background, LKC UTAS-E3000), and PERG (0.8° checks, 99% contrast, 100 cd/m(2) mean, 5 reversals/s, VERIS). All NHP were followed at least until HRT-confirmed optic nerve head posterior deformation, most to later stages. mfERG responses were filtered into low- and high-frequency components (LFC, HFC, >75 Hz). Peak-to-trough amplitudes of LFC features (N1, P1, N2) and HFC RMS amplitudes were measured and ratios calculated for HFC:P1 and N2:P1. ff-ERG parameters included A-wave (at 10 ms), B-wave (trough-to-peak) and PhNR (baseline-to-trough) amplitudes as well as PhNR:B-wave ratio. PERG parameters included P50 and N95 amplitudes as well as N95:P50 ratio and N95 slope. Diagnostic performance of retinal function parameters was compared using the area under the receiver operating characteristic curve (A-ROC) to discriminate between EG and control eyes. Correlations to RNFLT were compared using Steiger's test. RESULTS: Study duration was 15 ± 8 months. At final follow-up, structural damage in EG eyes measured by RNFLT ranged from 9% above baseline (BL) to 58% below BL; 29/43 EG eyes (67%) and 0/43 of the fellow control eyes exhibited significant (>7%) loss of RNFLT from BL. Using raw parameter values, the largest A-ROC findings for mfERG were: HFC (0.82) and HFC:P1 (0.90); for ff-ERG: PhNR (0.90) and PhNR:B-wave (0.88) and for PERG: P50 (0.64) and N95 (0.61). A-ROC increased when data were expressed as % change from BL, but the pattern of results persisted. At 95% specificity, the diagnostic sensitivity of mfERG HFC:P1 ratio was best, followed by PhNR and PERG. The correlation to RNFLT was stronger for mfERG HFC (R = 0.65) than for PhNR (R = 0.59) or PERG N95 (R = 0.36), (p = 0.20, p = 0.0006, respectively). The PhNR flagged a few EG eyes at the final time point that had not been flagged by mfERG HFC or PERG. CONCLUSIONS: Diagnostic performance and structure-function correlation were strongest for mfERG HFC as compared with ff-ERG PhNR or PERG in NHP EG.

Whole-brain networks derived from diffusion tensor imaging (DTI) data require the identification of seed and target regions of interest (ROIs) to assess connectivity patterns. This study investigated how initiating tracts from gray matter (GM) or white matter (WM) seed ROIs impacts (1) structural networks constructed from DTI data from healthy elderly (control) and individuals with Alzheimer's disease (AD) and (2) between-group comparisons using these networks. DTI datasets were obtained from the Alzheimer's disease Neuroimaging Initiative database. Deterministic tractography was used to build two whole-brain networks for each subject; one in which tracts were initiated from WM ROIs and another in which they were initiated from GM ROIs. With respect to the first goal, in both groups, WM-seeded networks had approximately 400 more connections and stronger connections (as measured by number of streamlines per connection) than GM-seeded networks, but shared 94% of the connections found in the GM-seed networks. With respect to the second goal, between-group comparisons revealed a stronger subnetwork (as measured by number of streamlines per connection) in controls compared to AD using both WM-seeded and GM-seeded networks. The comparison using WM-seeded networks produced a larger (i.e., a greater number of connections) and more significant subnetwork in controls versus AD. Global, local, and nodal efficiency were greater in controls compared to AD, and between-group comparisons of these measures using WM-seeded networks had larger effect sizes than those using GM-seeded networks. These findings affirm that seed location significantly affects the ability to detect between-group differences in structural networks.

Posner-Schlossman syndrome (PSS) shares some clinical features with uveitis and open angle glaucoma. Cytokines and autoantibodies have been associated with uveitis and open angle glaucoma. However, the role of serum cytokines and autoantibodies in the pathogenesis of PSS remains unknown. This study aimed to evaluate the associations of type 1 T helper (Th1) and Th17 related cytokines and autoantibodies with PSS. Peripheral blood serum samples were collected from 81 patients with PSS and 97 gender- and age-matched healthy blood donors. Th1 and Th17 related cytokines, including interleukin-1β (IL-1β), IL-12, tumor necrosis factor-α (TNF-α), interferon- γ (IFN-γ), IL-6 and IL-17, and glucose-6-phosphate isomerase (GPI) were determined by double antibody sandwich ELISA. Anti-nuclear antibody (ANA), anti-keratin antibody (AKA) and anti-neutrophil cytoplasmic antibody (ANCA) were detected by indirect immunofluorescence assay. Anti-cardiolipin antibody (ACA)-IgG, ACA-IgM, ACA-IgA, anti-double stranded DNA (anti-dsDNA) and anti-cyclic citrullinated peptide antibody (anti-CCP) were detected by indirect ELISA. Serum levels of IL-1β, IL-12 and IL-6 in PSS patients were significantly lower than those in controls (P < 0.003), and these associations survived the Bonferroni correction (Pc < 0.018). There was no significant difference in serum levels of TNF-α, IFN-γ and IL-17 between the PSS and control groups (Pc > 0.12). Positive rate of serum anti-dsDNA in PSS patients was significantly higher than that in the control group (P = 0.002, Pc = 0.018), while positive rates of serum ANA, AKA, ANCA, ACA-IgG, ACA-IgM, ACA-IgA, GPI and anti-CCP in the PSS group were not significantly different from those in the control group (Pc > 0.09). These results suggest that anti-dsDNA may contribute to the pathogenesis of PSS, while Th1 and Th17 related cytokines and other autoantibodies may not be major contributors to PSS.