Datasets:

rntc

/

ex-cl

like 0

The 188 KJ-RILs and their parental lines were genotyped with the Wheat660K SNP array. SNPs were rejected if they showed minor allele frequency (defined as frequency <0.3) or contained >10% missing data. Markers were binned based on their segregation patterns in the KJ-RIL population using the BIN function in IciMapping 4.1 (http://www.isbreeding.net/) according to Winfield et al.18. Markers that shared their segregation pattern with at least one other marker were retained. One marker was chosen to represent each bin on the basis of the least amount of missing data or, when the percentage of missing data was equal, at random. Markers were tested for significant segregation distortion using a chi-square test. SNPs were sorted into groups using the MAP function in IciMapping 4.1, with the previously mapped 591 loci serving as anchored markers3. A logarithm of the odds (LOD) score of 3.5 and a recombination fraction of 0.3 were used to sort the SNPs with the Kosambi mapping function70. Groups were ordered with the Kosambi mapping function within the JoinMap v. 4.0, using a LOD score ≥3 after preliminary analysis of SNPs with LOD scores ranging from 2 to 10. The long and short arms of each chromosome were identified from the IWGSC wheat survey sequence (http://www.wheatgenome.org/), and groups were orientated to have the short arm above the long arm. MapChart 2.2 (http://www.biometris.nl/uk/Software/MapChart/) was used to draw the genetic map.

study

The QTLs for KNPS were identified using MapQTL 6.0 (based on CIM, https://www.kyazma.nl/index.php/mc.MapQTL/), IciMapping 4.1 (based on ICIM, http://www.isbreeding.net/software/?type=detail&id=18), and QTLNetwork 2.0 (based on MLMA, http://ibi.zju.edu.cn/software/qtlnetwork/). For MapQTL 6.0 and IciMapping 4.1, the average KNPS values of each KJ-RIL in each environment were used for individual environment QTL mapping analysis. For QTLNetwork 2.0, the KNPS values of each KJ-RIL in each replication of the 10 environments were assembled to perform combined QTL analysis across environments to identify QTLs with additive-by-environment (A by E) interaction effects. The overlapping confidence intervals detected with the abovementioned programs were used to predict the candidate genes based on the sequence information of the flanking markers and the IWGSC WGA v0.4 assembly of chromosome 4A (https://urgi.versailles.inra.fr/download/iwgsc/IWGSC-WGA_Sequences/).

study

The SNP flanking sequences mapped in the KJ-RIL map were kindly provided by Professor Jia JZ. We used the Basic Local Alignment Search Tool (BLAST) (ftp://ftp.ncbi.nlm.nih.gov/blast/executables/release/) to align the SNP probes to the IWGSC survey sequences (contigs). All IWGSC survey sequences were downloaded from http://www.wheatgenome.org/. In addition, contig sequences to which the SNPs were best hits were screened in a BLASTN search against the coding sequences (CDSs) of B. distachyon, rice (O. sativa L.), maize (Z. mays L.), and sorghum (S. vulgare L.). All CDSs were downloaded from http://plants.ensembl.org/index.html. An expectation value (E) of 1E−10 was used as the significance threshold. Synteny analyses with common wheat, Brachypodium, rice, maize and sorghum genomes were performed based on the SNP orders in the KJ genetic map and on the corresponding CDSs in the genome sequences of Brachypodium, rice, maize and sorghum genomes.

study

Using the sequences of the markers (including SNPs, PCR-based markers and DArTs), we conducted comparative genomics analysis against the contigs assembled in the chromosome survey sequencing (CSS) project. All of the contig sequences were downloaded from https://wheat-urgi.versailles.inra.fr/Seq-Repository. More recently, the genome assembly of T. aestivum cv. Chinese Spring (CS) has been released (https://urgi.versailles.inra.fr/download/iwgsc/IWGSC-WGA_Sequences/). Based on SNP flanking sequences, we assigned SNPs to this wheat genome assembly.

study

Human Immunodeficiency Virus (HIV) integrase inhibitors (INI) are the most common drug class recommended by the different international guidelines as third agent of the combined antiretroviral therapy (cART). First licensed INI, raltegravir (RAL), is now available for 10 years and determinants of genotypic resistance are well-described for this drug, as well as for the second first-generation INI, elvitegravir (EVG). However, there is still data to provide regarding resistance mechanisms to the most recent, second-generation INI, dolutegravir (DTG).

review

In this manuscript, we have reviewed main in vivo and in vitro knowledge about two integrase resistance-associated mutations: R263K and E157Q. R263K mutation is interesting because this mutation was selected in vivo at failure of a DTG-based regimen and only displayed a moderate increase in phenotypic resistance level . Furthermore, Mark Wainberg and his team conducted several studies regarding in vitro characteristics of R263K integrase mutants. E157Q integrase mutation is of interest too, since it is both polymorphic, with variable prevalence in INI-naïve patients depending on the viral subtype , but also selected at virological failure (VF) of a RAL-based regimen in two case reports and described in a case report of a non virological response to a DTG-based regimen .

review

In the first report of in vitro selection under DTG pressure conducted in primary human cells, the R263K substitution in integrase was identified through culture selection after 20 weeks as a DTG resistance-associated mutation . Furthermore, in these selection experiments, S153Y and S153T substitutions were observed in combination with R263K in one subtype B and in one subtype C virus, respectively . In the study of Quashie et al., site-directed mutagenesis analysis showed that R263K did confer very low-level resistance to DTG (Fold Change (FC) = 1.06), a slight increased for EVG (FC = 1.75) and no change for RAL (FC = 0.63) . More recently, it has been shown that in both single and multiple rounds of HIV-1 infections, bictegravir (BIC) and cabotegravir (CAB), two INIs currently under development, remained active against R263K mutant .

study

To date, in vitro selection experiments failed to select viruses with a high level of resistance to DTG, except when very high concentrations of DTG, until 500 nM, were used. Interestingly, no integrase resistance-associated mutations were detected in this resistant selected virus. However, this work of Malet et al. reported, for the first time, an INI-resistant virus with mutations selected outside integrase gene, located in the 3’ PPT region .

study

In the SAILING randomized trial, which included cART-experienced but INI-naïve patients, selection of INI resistance at time of VF through week 24 was observed in two participants among the DTG arm (n = 354 patients). In both cases the R263K mutation was selected with a plasma viral load at failure comprised between 3 and 4 log10 c/mL . One virus displayed R263K as a single mutation and phenotypic analysis of this clinical isolate showed a FC to DTG and RAL of 1.12 and 0.96, respectively, with a reduced viral replicative capacity equal to 33% . The second virus harbored the V260I mutation added to the R263K, this double-mutant resulted in a FC to DTG and RAL of 1.93 and 1.12, respectively . The single mutant V260I did not confer DTG or RAL FC increase . In addition, R263K site-directed mutant analyses using MT4 cells in a 5-day assay with cell tier glow readout showed a FC of 2.1, 0.8 and 10.6 for DTG, RAL and EVG, respectively . A few additional VF occurred in the SAILING trial after W48 and R263K mutation was detected in one of them . In this latter, VF occurred at week 120 with a viral load of 622 c/mL and R263K was detected added to A49G and S230R integrase mutations. This accumulation of integrase mutations probably resulted from the long duration of replication under treatment, since plasma viral load was above 50 c/mL since week 96. This triple-mutant clinical isolate showed an increased DTG FC of 5.77 and a RAL FC of 2.62, with a very low viral replicative capacity of 12% .

study

Regarding its prevalence, R263K mutation is very rare in cART-naïve patients, in the French epidemiological transmitted drug resistance survey conducted in patients in primary infection with a prevalence of 0.9% (n = 2/233 patients) . In a study based on 92 recently diagnosed, but chronically-infected, cART-naïve patients, no R263K was detected by Sanger sequencing technology and was found in two samples in minority proportion only when using ultra-deep sequencing technology .

study

The study of Quashie et al. showed that the presence of R263K mutation did confer a decreased integration in cell culture without altering reverse transcription step . Further in vitro experiments performed in this study, including biochemical cell-free assays performed with purified integrase enzyme containing R263K mutation, showed a slight decrease in 3′processing and strand transfer activities compared to the wild-type virus. Structural modeling suggested that the R263K mutation affects integrase-DNA interactions and in vitro integrase-DNA binding assays confirmed these data (Figure 1). In the study of Mesplède et al., they performed prolonged infections by transferring culture fluids from infected cells to uninfected cells at weekly intervals and it resulted in a progressive decrease in integrated viral DNA between weeks 2 to 4 of infection. Thus, prolonged infections with R263K mutants led to a progressive decline in integrated HIV-1 DNA over time .

study

Recently, the study of Kessl et al. showed that IN binding to the viral RNA genome is necessary for formation of infectious viral particles. Interestingly, the residues identified in IN to mediate IN-RNA interactions are residues 269–273, very close to the 263 residue . This could participate to the drastic reduction of the viral replicative capacity observed in case of the presence of the R263K mutation.

study

It has been shown that R263K mutants display a low viral replicative capacity . Thus, several studies were conducted to identify potential secondary mutations that would restore viral replicative capacity. First studies reported on the M50I integrase polymorphism, since it was selected in culture secondary to the R263K mutation . The findings showed that M50I polymorphism in combination with R263K increased resistance to DTG in tissue culture and in biochemical assays but did not restore viral replicative capacity of the R263K mutant . Similarly, H51Y integrase mutation also emerged secondary to R263K mutation in DTG selection experiments. So, in vitro characteristics of H51Y single- and H51Y-R263K double-mutants were studied showing that the addition of H51Y to R263K increased phenotypic resistance level to DTG with a FC of 16.5, whereas H51Y alone did not confer resistance to this drug. However, the addition of H51Y is accompanied by dramatic decreases in both enzymatic activity and viral replication. So, the H51Y mutation is also not a compensatory mutation to the R263K .

study

The study of Liang et al. showed that the T66I substitution emerged from a wild-type virus but also from a R263K mutant and from a E138K-R263K double-mutant virus under RAL or EVG pressure . The aim of this study was to assess the effects of the T66I and E138K substitutions, alone and in combination with R263K, on viral replicative capacity and resistance to INI. They showed that the addition of R263K to the T66I substitution did not significantly compromise susceptibility to DTG, while decreasing both viral replicative capacity and strand-transfer activity . The addition of the E138K substitution to the T66I-R263K double-mutant partially compensated for these deficits and resulted in a high level of resistance against EVG but not against DTG or RAL .

study

Interestingly, E157Q mutation is able to partially restore decrease in integrase enzymatic activity caused by the R263K substitution, thereby acting as a possible secondary, compensatory mutation. Furthermore, the E157Q-R263K double-mutant also displayed enhanced DTG resistance by 10-fold compared with lower-level resistance associated with R263K mutation alone. However, to date, the double-mutant E157Q-R263K has not yet been described in vivo at VF of an INI-based regimen .

study

The R263K substitution was reported at VF in a subtype C-infected participant in the SAILING clinical trial . Thus, in vitro characteristics of a R263K site-directed mutant in the subtype C context was assessed showing a significant decrease in strand-transfer activity for the R263K integrase protein to a greater extent than observed in subtype B . The presence of R263K decreased HIV-1 subtype C infectiousness by 70% compared to wild-type, whereas R263K in subtype B only resulted in a 40% decrease . Thus, the R263K substitution appears to be more deleterious in subtype C than in subtype B.

study

In the study of Anstett et al., tissue culture selections with DTG, using viruses resistant to first-generation INIs (RAL and EVG), infectivity and resistance assays were performed. They showed that the presence of the E92Q or N155H resistance mutations was compatible with the emergence of R263K, whereas no R263K selection was observed in presence of G140S-Q148R, E92Q-N155H, G140S, Y143R and Q148R resistance mutations . Thus, some genotypic resistance profiles seem to prevent emergence of the R263K DTG resistance-associated mutation.

study

In the opposite way, in vitro the presence of the R263K delayed the emergence of RAL resistance-associated mutations, whereas the simultaneous presence of either the H51Y or E138K substitutions in combination with R263K somewhat mitigated this inhibitory effect. In contrast, in vitro experiments showed that, in the presence of R263K mutation, resistance selection to EVG appeared earlier than in wild-type virus .

study

The N155H is one of the major resistance pathway to the first-generation INI, and is often associated with secondary resistance mutations. In the study of Anstett et al., L74M, E92Q, T97A, E157Q and G163R resistance mutations were introduced into NL4.3 subtype B HIV-1 vectors harboring N155H and R263K in tandem . They found that the addition of T97A, E157Q or G163R mutation somewhat improved the affinity of the double-mutant N155H-R263K for its target DNA substrate, while the presence of L74M or E92Q had a negative effect on this step. This work showed that the compensatory mutations, that evolve after N155H selection under continued DTG or RAL/EVG pressure, are unable to improve enzyme efficiency, resistance level or viral infectivity in an N155H-R263K background .

study

The study of Singhroy et al. assessed the impact of adding the M184I/V mutation to a R263K-mutated virus on viral replicative capacity. Their results showed that the presence of M184I or M184V with R263K further decreased viral infectiousness and replicative capacity compared to the effects of the individual mutations alone .

study

Integrase E157Q substitution has been described as a polymorphism present in 2.4% of viral sequences obtained from cART-naïve patients in the ARCA Italian database and in 5.0% in the last French transmitted drug resistance survey study . A recent study of the French ANRS AC11 virology network conducted on 8528 integrase sequences from INI-naïve patients showed that the overall prevalence of E157Q polymorphism was 2.7% and its distribution among HIV-1 subtypes was 1.7%, 5.6% and 2.2% in B, CRF02_AG and others non-B subtypes, respectively .

study

E157Q emergence in integrase has also been observed in the case of VF under a RAL-based treatment in two case reports. A case report in an highly cART-experienced patient has shown that mutation E157Q was rapidly selected and archived in intracellular HIV DNA within a short term of eight weeks of low-level replication under a RAL-based treatment . Indeed, in this case report mutation at position 157 of integrase was not present in baseline plasma virus . In a second study based on nine patients experiencing a VF under a RAL-based treatment, E157Q mutation was one of the four different resistance mutations profiles identified at VF .

study

A recent Italian study carried out in vitro phenotypic assays on six clinical samples harbouring E157Q-mutated viruses including three subtype B and three CRF02_AG recombinant, showing they were all susceptible to DTG and to RAL, except one at the limit of the biological cut-off for RAL . The study of Shimura et al. assessed susceptibility to EVG of a E157Q recombinant molecular clone showing a small reduction in EVG susceptibility with a FC of 6.3 .

study

These assays were mainly conducted with a subtype B. In a recent study, we carried out phenotypic assays using E157Q mutant generated in the subtype B context, but also in the CRF02_AG context. In this study, the E157Q site-directed mutants did not show an increased phenotypic resistance level to DTG or RAL both in B and CRF02_AG subtypes contexts . These in vitro phenotypic data are in accordance with the recent findings of the Italian study using similar phenotypic assays . However, we observed a slight increase of FC to EVG at 1.9 and 2.4 in the presence of E157Q in B and CRF02_AG contexts, respectively . Thus, E157Q polymorphism does not seem to impact phenotypic susceptibility to RAL or DTG, in contrast a potential low-level resistance, especially in the context of CRF02_AG recombinant, was observed for EVG.

study

A case report of a non virological response to a DTG-based regimen, in a patient infected with a E157Q-mutated virus, has been described despite adequate DTG plasma concentrations . In vitro characterization of the E157Q-mutated virus issued from this clinical sample has shown that integrase strand-transfer activity was 3-fold greater compared to wild-type virus and that IC50 value was increased, leading to a FC equal to 9 .

clinical case

To date, very few data are available regarding virological response of patients displaying E157Q-mutated virus, only represented by case reports that limit their interpretation. In the study of Ambrosioni et al., only two cART-naïve patients initiated a DTG-based regimen with a very short follow-up of 12 weeks . In the study of Pavkovich et al., only six among the 15 patients initiating an INI-based regimen were in virological success at month 12; however heterogeneous profiles of patients, cART-naïve and cART-experienced, were mixed in the analysis .

study

A recent study of the French ANRS AC11 virology network assessed the virological outcome of 39 INI-naïve patients with E157Q-mutated virus initiating an INI-based regimen . Among them, 15 had a viral load (VL) < 50 c/mL at initiation and virological suppression was maintained during the first-year follow-up in all but two exhibiting a viral blip. Twenty-four patients had a VL > 50 c/mL at time of INI-based regimen initiation. Among them eight were receiving their first-line c ART and the two patients who did not reach VL < 50 c/mL at W24 were receiving EVG-based single-tablet regimen. The 16 remaining patients were cART-experienced in VF with drug-resistant viruses and they displayed several virological outcomes to the INI-based regimen, independently of the genotypic susceptibility score of the cART. Thus, taking into account all the data, in case of E157Q polymorphism, the most recommended INI might be DTG in such patients; EVG should not be considered due to potential low-level resistance, especially in the context of CRF02_AG recombinant and RAL seems adequate with unmounted IC50 but with the caveat that selection of E157Q has been described at VF in two case reports.

study

These findings show that there is still a need for better understanding of resistance mechanisms to INI, specially for DTG. These findings also emphasized the importance of genotypic background in viral evolution under drug pressure and that high level of resistance could result from combined integrase polymorphisms as yet unidentified.

study

The health effects of fetal and child environmental exposure to chemicals are a major concern, and several cohort studies have been conducted worldwide (Ayotte et al. 2003; Govarts et al. 2012; James et al. 2002). Polychlorinated biphenyls (PCBs) are known to affect the endocrine systems in humans. Despite being banned in 2001 by the Stockholm Convention (UNEP 2001), PCBs widely persist in wildlife and humans because of their lipophilic properties, low water solubility, and bioaccumulation in fatty tissues (Safe 1994). PCBs are reported to have numerous adverse human health effects, including reproductive (Leoni et al. 1989), developmental (Jorissen 2007), immunological (Glynn et al. 2008), and neurological (Prince et al. 2006). Food is a major source of exposure to PCBs and dioxins, which are related compounds (Caspersen et al. 2013; Kvalem et al. 2009). It has been reported that the consumption of fish was positively associated with serum levels of total PCBs among Japanese women (Tsukino et al. 2006), suggesting that human PCB contamination levels are strongly affected by eating habits.

review

Self-administered, semi-quantitative food frequency questionnaire (FFQ) has been used to assess the nutrient and food intake of subjects in large-scale epidemiologic studies from several countries, including Japan (Sasaki et al. 2003; Tsugane et al. 2003a; Willett 1994). If groups at high risk for PCB contamination could be identified by an FFQ approach, without chemical analyses, significant cost and time could be saved. Various prediction models for PCB concentrations in humans have been generated using dietary intake (Kvalem et al. 2009; Laden et al. 1999; Xue et al. 2014); however, few studies have addressed this issue in Japanese populations.

study

The Japan Environment and Children’s Study (JECS) is a nationwide birth cohort study of 100,000 parent–child pairs that evaluates the impact of various environmental factors on children’s health and development (Kawamoto et al. 2014). Here, in an adjunct study of JECS, we evaluated the relationship between FFQ responses and serum PCB contamination levels of participants at the Chiba Regional Center.

study

This study was approved by the Ethics Committee of Chiba University, Japan. Data of participants (age, body mass index [BMI], and a number of deliveries) and FFQ were completed by mothers (n = 1716) and fathers (n = 318) recruited from the Chiba Regional Center, which is one of the 15 regional centers of JECS. Age, BMI, and reproductive information were collected from medical record information. Participants who had completed <60% of the FFQ, outlier of energy intake (below 600 kcal/day or above 4000 kcal/day), and not completed the medical record information (age, BMI, and reproductive information) were excluded from the analysis. Finally, the remaining 1477 mothers (86.1%) and 219 fathers (68.9%) were used for all analysis (Table 1).Table 1Subjects’ characteristics, FFQ data in fathers and mothers recruited in Chiba Regional Center of JECSMothers (n = 1477)Fathers (n = 219)MeanStandard deviation0%25%50%75%100%MeanStandard deviation0%25%50%75%100%Age31.05.017.028.031.035.046.032.96.020.029.032.037.064.0BMI21.43.214.619.220.722.739.924.43.916.921.923.826.239.0Height158.25.4140.0155.0158.0162.0178.0171.45.7158.0167.3171.5175.0190.0Weight53.58.937.048.052.057.0106.071.812.347.363.070.078.7110.0A number of deliveries0.90.80.00.01.01.05.0Energy (kcal)18455836851426175821533973220862510561756215825933920Intakes of foods (g/day/kcal)Eggs15.413.10.07.613.319.6194.315.720.40.05.810.718.3145.9Fats and oils6.02.41.24.35.67.226.35.43.90.03.04.56.936.5Fish and shellfish22.115.00.012.420.028.8154.822.018.50.08.017.630.8119.8Fruits89.170.50.039.674.8121.7740.253.271.40.010.333.372.3548.1Meats38.919.50.025.236.149.0143.948.636.80.023.939.760.2241.0Milks138.8126.20.059.6108.6171.2914.594.3163.20.016.445.496.71315Potatoes and starches13.99.50.07.312.118.489.411.011.00.03.68.414.788.5Vegetables106.062.70.064.191.4133.8829.082.178.50.034.457.2110.4492.9Water217.7313.90.022.197.1286.02386174.5322.70.022.368.8215.43585

study

The FFQ used in the JECS is a validated, self-administered diet questionnaire that has been evaluated in previous studies with respect to nutrient factors (Ishihara et al. 2003; Iso et al. 2003; Tsubono et al. 2003; Tsugane et al. 2003b). The FFQ was a self-administered questionnaire comprising 13 pages. The FFQ included food lists (cereals, eggs, fat and oils, fish and shellfish, fruits, meats, mushrooms, milks, potatoes and starches, and vegetables), standard units (volumes; 0.5 for small, 1.0 for medium, and 1.5 for large), portion sizes (e.g., 30 g portion−1), and frequencies (ranging from less than once per month to more than seven times per day). Moreover, eating habits (e.g., frequency of having breakfast, eating out, and speed of eating) were assessed by the questionnaire. The subjects reported their intake frequency and portions of consumption for the past year (Sakurai et al. 2004; Takasuga et al. 2004). Consumption amounts were calculated using the frequencies, standard units, and portion sizes for the various food items of the FFQ. This amount was then energy adjusted using the energy density method (Drewnowski and Specter 2004).

study

During the mother’s pregnancy, blood samples of 30 mL were drawn once from both mothers and fathers. PCB levels were analyzed according to the procedures previously reported (Jotaki et al. 2011; Mori et al. 2014). Approximately 1 g of serum was hydrolyzed with 1 M potassium hydroxide that was diluted in ethanol for 18 h and extracted three times with hexane. The extract was washed three times with distilled water and then dehydrated with anhydrous sodium sulfate. The solution was concentrated by evaporation to approximately 2 mL, eluted through a Florisil column (Florisil PR; GL Sciences, Tokyo, Japan) with 50 mL hexane, evaporated to a final volume of 1 mL, and analyzed by packed-column gas chromatography with an electron capture detector (GC/ECD; GC-17A, Shimadzu, Kyoto, Japan). The PCB concentration was calculated as the sum of the major eight peaks shown after p,p′-DDE on the chromatogram, based on the peaks detected in a Kanechlor mixture standard (Kanechlor 300, 400, 500, and 600 = 1:1:1:1) (JIS K 0093: 2006; Mori et al. 2014). The detection limit was 0.1 ng g−1 wet weight; samples below the lower level of detection were assigned to a value of 0.05 ng g−1 (JIS K 0093: 2006). The total PCB concentrations that were analyzed by packed GC/ECD were strongly correlated (R 2 > 0.996) with values analyzed by GC-HRMS in our previous study (Mori et al. 2014).

study

The relationships between body mass index (BMI), age, FFQ, and PCB concentrations (wet weight) were analyzed by the Bayesian linear regression modeling separately by gender for both parents. The PCB concentration data was assessed for normal distribution with the Shapiro–Wilkes test and was not normally or log-normally distributed but was right-skewed and showed slightly heavy tails. Due to the distribution of PCB concentration, Bayesian linear regression modeling was conducted using a No-U-Turn Sampler variant of the Hamiltonian Markov Chain Monte Carlo (Hoffman and Gelman 2014) approach that was implemented in Stan version 2.8.0 and accessed through the rstan packing (Stan Development Team 2015) of the R program (R Core Team 2014). This approach was chosen because Bayesian linear regression modeling is sufficiently tolerant to handle skewed and heavy-tailed distributions. We used the non-informative N (0, 1002) prior distribution as the prior mean PCB concentration in all models.

study

In the Bayesian linear regression modeling, we ran three chains from random initializations each with 10,000 samples, thinned to every 2 samples, and discarded the first 5000 samples from each as burn-in. We checked convergence by ensuring that R-hat, an estimate of the potential scale reduction of the posterior if sampling were to be infinitely continued, was below 1.1 (Gelman et al. 2014). The reported empirical 95% credible intervals (95% CI) represent the 2.5th to 97.5th percentiles of the highest posterior density interval calculated from the posterior samples.

study

Following this, the predictive accuracy from various models can be compared using the Watanabe–Akaike information criterion (WAIC) (Vehtari and Gelman 2014; Watanabe 2010):\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \mathrm{WAIC}=-2\left(\widehat{lpd}-{\widehat{P}}_{\mathrm{waic}}\right) $$\end{document}WAIC=−2lpd^−P^waicwhere \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ \widehat{lpd} $$\end{document}lpd^ represents the computed log pointwise predictive density and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$ {\widehat{P}}_{\mathrm{waic}} $$\end{document}P^waic is the estimated effective number of parameters (Vehtari and Gelman 2014; Watanabe 2010). The model with the lowest WAIC is selected to achieve a trade-off between model complexities. WAIC was derived based on a singular learning theory as an asymptotically unbiased approximation to the out-of-sample prediction error and is a generalization of Akaike’s information criterion that is applicable to both regular and singular statistical models (Watanabe 2010). The variable selection in the Bayesian linear regression modeling was evaluated based on the stepwise forward selection method by WAIC.

study

The mean ± standard deviation (SD) serum concentrations of PCBs in fathers and mothers were 0.38 ± 0.21 and 0.28 ± 0.16 ng g−1 wet weights, respectively (Table 2). The mean concentrations of PCBs stratified by a number of deliveries in mothers are shown in Table 2. PCB concentrations in the fathers were higher than in the mothers, and among the mothers, serum PCB concentrations were lower as the number of deliveries was greater. PCB levels in this study are comparable or slightly lower than the serum levels of previous cohort studies in Japan, USA, and European countries (Table 3) (Govarts et al. 2012; Inoue et al. 2006; Patterson et al. 2009). Based on the stepwise forward selection results, consumption of fish and eggs was selected along with age and number of delivery for the model of mothers by WAIC (Tables 4 and 6). In mothers, age [β = 0.26, 95% CI (0.23, 0.29)] and fish [β = 0.055, 95% CI (0.029, 0.082)] and egg consumptions [β = 0.025, 95% CI (−0.00027, 0.052)] were positively and a number of deliveries [β = −0.18, 95% CI (−0.21, −0.16)] was negatively associated with PCB concentrations in mothers (Table 6). However, in our analysis, only age [β = 0.25, 95% CI (0.19, 0.32)] was selected for the fathers’ model (Tables 5 and 6).Table 2PCB concentrations (ng/g wet wt.) in serum from fathers and mothersMeanStandard deviationMin25%50%75%MaxFather (n = 219)0.380.210.050.230.320.491.2Mother (n = 1477)0.280.160.050.180.250.351.3Delivery status of mothersA number of deliveries: 0 (n = 550)0.320.170.050.210.280.391.3A number of deliveries: 1 (n = 644)0.260.140.050.170.240.331.3A number of deliveries: 2 (n = 234)0.230.140.050.140.210.290.79A number of deliveries: 3 (n = 37)0.220.150.050.120.190.250.71A number of deliveries: 4 (n = 10)0.150.0820.050.0680.160.220.26A number of deliveries: 5 (n = 2)<0.1 Table 3Comparison of serum concentrations of PCBs (pg g−1 wet wt.) in participants analyzed in this study with previous dataAgeMeanStandard deviationMin25%50%75%MaxReferencesFather (n = 219)32.9 ± 6.00.380.210.050.230.320.491.2This studyMother (n = 1477)31.0 ± 5.00.280.160.050.180.250.351.3This studyJapanese mother (n = 89)*20–29 years (50.6%)30–39 years (44.9%)40–49 years (5.5%)0.360.0620.210.300.451.3Inoue et al. 2006 Mothers from 12 European birth cohorts (n = 7762: CB153)>20% of mothers were <25 years of age at delivery and in two populations >20% were >35 years of age0.180.160.14Govarts et al. 2012 National Health and Nutrition Examination Survey (2003–2004): USA12–19 years (31.2%)20–39 years (24.5%)40–59 years (20.3%)+60 years (24.0%)Male (n = 928)0.835**0.8141.62Patterson et al. 2009 Female (n = 938)0.805**0.8241.65Patterson et al. 2009 *Adjusted by reported average serum lipid level (7.9 g/L) (Longnecker et al. 2003)**Geometric mean Table 4Selection of prediction model based on mothers’ dataset by WAICWAICPCB ~ age−1843.5PCB ~ age + a number of deliveries−2012.9PCB ~ age + a number of deliveries + BMI−2011.4PCB ~ age + a number of deliveries + cereals−2012.2PCB ~ age + a number of deliveries + potatoes and starches−2011.5PCB ~ age + a number of deliveries + vegetables−2013.3PCB ~ age + a number of deliveries + fruits−2011.4PCB ~ age + a number of deliveries + fish and shellfish−2028.5PCB ~ age + a number of deliveries + meats−2011.3PCB ~ age + a number of deliveries + eggs−2014.2PCB ~ age + a number of deliveries + milk−2011.6PCB ~ age + a number of deliveries + fats and oils−2011.5PCB ~ age + a number of deliveries + fish and shellfish + BMI−2027.5PCB ~ age + a number of deliveries + fish and shellfish + cereals−2027.0PCB ~ age + a number of deliveries + fish and shellfish + potatoes and starches−2027.8PCB ~ age + a number of deliveries + fish and shellfish + vegetables−2027.9PCB ~ age + a number of deliveries + fish and shellfish + fruits−2027.0PCB ~ age + a number of deliveries + fish and shellfish + meats−2026.4 PCB ~ age + a number of deliveries + fish and shellfish + eggs −2030.2 PCB ~ age + a number of deliveries + fish and shellfish + milk−2028.3PCB ~ age + a number of deliveries + fish and shellfish + fats and oils−2026.7PCB ~ age + a number of deliveries + fish and shellfish + eggs + BMI−2029.2PCB ~ age + a number of deliveries + fish and shellfish + eggs + cereals−2028.4PCB ~ age + a number of deliveries + fish and shellfish + eggs + potatoes and starches−2029.7PCB ~ age + a number of deliveries + fish and shellfish + eggs + vegetables−2029.4PCB ~ age + a number of deliveries + fish and shellfish + eggs + fruits−2029.1PCB ~ age + a number of deliveries + fish and shellfish + eggs + meats−2028.3PCB ~ age + a number of deliveries + fish and shellfish + eggs + milk−2030.0PCB ~ age + a number of deliveries + fish and shellfish + eggs + fats and oils−2028.4*Italicized model was selected with the smallest WAIC Table 5Selection of prediction model based on fathers’ dataset by WAICWAIC PCB ~ age −172.8 PCB ~ age + BMI−172.4PCB ~ age + cereals−171.0PCB ~ age + potatoes and starches−172.5PCB ~ age + vegetables−169.9PCB ~ age + fruits−171.9PCB ~ age + fish and shellfish−172.0PCB ~ age + meats−170.9PCB ~ age + eggs−171.4PCB ~ age + milks−171.7PCB ~ age + fats and oils−170.7PCB ~ age + potatoes and starches + BMI−172.2PCB ~ age + potatoes and starches + cereals−171.2PCB ~ age + potatoes and starches + vegetables−170.0PCB ~ age + potatoes and starches + fruits−171.0PCB ~ age + potatoes and starches + meats−171.1PCB ~ age + potatoes and starches + fish and shellfish−171.0PCB ~ age + potatoes and starches + eggs−171.3PCB ~ age + potatoes and starches + milk−171.4PCB ~ age + potatoes and starches + fats and oils−171.2*Italicized model was selected with the smallest WAIC Table 6Empirical mean (mean), standard deviation of the mean (SD), and quantiles for each variable in selected prediction model for mothers and fathersMothersFathersCoefficientStandard deviation95% lower95% upperCoefficientStandard deviation95% lower95% upperIntercept−1.320.013−1.34−1.29−1.00.032−1.1−0.93Age0.260.0140.230.290.250.0330.190.32A number of deliveries−0.180.013−0.21−0.16NAFish and shellfish0.0550.0130.0290.082OmittedEggs0.0250.013−0.000270.052Omitted“Omitted” indicates factors eliminated by stepwise procedures NA not analyzed

study

The relationship between PCB concentration and age is commonly assumed to result from the continued exposure to or intake of PCB (Mori et al. 2014; Wolff et al. 2007) and the long half-lives of PCBs in the human body (Ritter et al. 2011). In a former study, our group also reported that delivery experience and lactation affected the PCB concentration level in adults (Mori et al. 2014), indicating women may excrete PCBs through delivery and lactation (Agudo et al. 2009; Kang et al. 1997).

study

Studies have demonstrated that the dietary intake of fish is a major pathway for human exposure to PCBs (Caspersen et al. 2013; Kvalem et al. 2009; Sjodin et al. 2000). Fitzgerald et al. (2004) reported that Mohawk women had high serum levels of PCBs that were associated with consumption of local fish and not with the exposure to other sources, like residential soil, lived sites, and the other foodstuff (Fitzgerald et al. 2004). Furthermore, it was reported that fish and shellfish accounted for 30–84% of the dioxin toxic equivalences from food intake by a Japanese total-diet study (Sasamoto et al. 2006), indicating that dioxin and dioxin-like PCB intake also occurs through fish and shellfish. Sasamoto et al. (2006) reported that meat and egg consumption comprised the second highest contribution to dietary intake of dioxins and dioxin-like PCBs, indicating that egg consumption may be associated with higher serum dioxin and dioxin-like PCB concentrations. The Bureau of Social Welfare and Public Health, Tokyo Metropolitan Government, showed the levels of PCBs in meat and eggs were below the detection limit in a food contamination survey (Bureau of Social Welfare and Public Health, Tokyo Metropolitan Government, 2015); however, continuous monitoring of PCBs in these foodstuffs might be needed.

study

In mother’s models, effect sizes of food intakes were 1 order lower than that of age and delivery experience, indicating the effect of food intake on serum PCB concentrations was weaker than that of age and number of deliveries. However, fish intake was included in the final model in mothers, indicating FFQ might be useful to estimate the PCB levels. In the father’s model, only age is predictive of serum PCB concentrations. Due to lack of information about smoking habits and limitation of sample size of father, development of improved predictive models and detailed investigations are required in future study.

study

Limitations of this study were lack of data about breast-feeding and smoking habit. Several studies have reported that breast-feeding is one of the major elimination pathways of persistent organic pollutants (POPs) (Abraham et al. 1996; Milbrath et al. 2009; Schecter et al. 1996), and after 6 months of breast-feeding, approximately 20% or more of the maternal body burden of these POPs may be transferred to children from mothers (Landrigan et al. 2002; Milbrath et al. 2009; Niessen et al. 1984). These results indicated that the number of deliveries is most likely a surrogate for breast-feeding. It was also reported that the active smoking was negatively affected to levels of PCB and dioxin in serum (Chen et al. 2005; Milbrath et al. 2009) through increased induction of dioxin-degrading enzymes, like cytochrome P450, by activation of the aryl hydrocarbon receptor by polycyclic aromatic hydrocarbons in tobacco smoke (Milbrath et al. 2009; Zevin and Benowitz 1999). Thus, information about smoking status and breast-feeding will be needed to develop more accurate predictive models in any future study.

study

In the current study, we developed a preliminary model to estimate the serum PCB levels from FFQ and other demographic/reproductive data collected from medical records. Finally, in mothers, age and consumption of fish and eggs were positively associated and the number of deliveries was negatively associated with PCB concentration. On the other hand, only age was positively associated with PCBs in fathers in this study.

study

In previous studies, it was shown that age was strongly related to exposure of PCBs, due to intake of PCB and long half-lives of PCBs in the human body. On the other hand, only women might excrete the PCBs through delivery and lactation. Dietary exposure to PCBs, especially in relation to fish consumption, was also reported, indicating out results that supported the previous cohort studies, even in the current Japanese cohort study. While the limitations about the smoking habits and breast-feeding existed, this study shows that the FFQ may help predict PCB levels in serum from women. In the future, development of the more accurate predictive models using information about the smoking status and breast-feeding was needed.

study

Debate about ophthalmology’s place within the undergraduate medical curriculum has been ongoing for nearly a hundred years.1 The explosive growth of scientific discovery has created relentless pressure to include novel subjects in the undergraduate medical curriculum. These changes have led to the erosion of more traditional subjects, including ophthalmology.2-4

other

The earliest reference to an undergraduate ophthalmology curriculum was in 1919, when the Council of British Ophthalmologists issued a report on the ‘Teaching and Examination of Medical Students in Ophthalmology’.1 The undergraduate ophthalmology curriculum described in this report concentrated only on the duration of the ophthalmology teaching and the assessment of knowledge. No reference is made to the content of the curriculum or teaching methods or environment. Despite the brevity of this report, the influence that the British Council of Ophthalmology held resulted in the General Medical Council recommending that at least ten weeks of ophthalmology training was undertaken during the undergraduate course.1 These recommendations are dramatically different to those of the current General Medical Council. With the publication of ‘Tomorrows Doctors’ 5 the UK undergraduate medical education system underwent major changes. More recent guidelines outlining the concept of a core undergraduate medical curriculum make no reference at all to the speciality of Ophthalmology.6 As a result ophthalmology undergraduate rotations have become more vulnerable in the UK.

other

Ophthalmology specialists have been comparatively peripheral to the education and training of doctors, such that ophthalmology has become isolated in the undergraduate medical curriculum.7 A larger emphasis on the learning of clinical skills in general practice rotations, rather than specialty attachments has reduced dedicated ophthalmology teaching.4 Conflict about what the end product of an ophthalmology rotation should achieve: a general practitioner, researcher or specialist has also created further confusion about the content of the curriculum.4 Despite this decline in undergraduate ophthalmology teaching, the ability to perform direct ophthalmoscopy examination is still an expected competency during UK Foundation Year One training.8

other

Between 7-19% of Accident and Emergency presentations have an ophthalmic component.9,10 In addition, it is estimated that 2-5% of all consultations in primary care are for ophthalmic conditions.11,12 These patients may present with sight threatening and potentially life-threatening illnesses. Therefore, it is critically important that junior doctors and general practitioners have the knowledge and skills required to treat and refer ophthalmic patients. This is especially relevant in the UK population where there is a rising prevalence of Type 2 diabetes with a resultant increase in diabetic retinopathy.

other

Several studies have highlighted the lack of confidence reported by clinicians when dealing with eye conditions.13, 14 It was suggested that if undergraduate ophthalmological teaching could be made more effective, the standard of primary ocular care would improve and have beneficial consequences for the population and for the ophthalmologist.2

review

In 2006, the International Task Force on Ophthalmic Education of Medical Students on behalf of The International Council of Ophthalmology (ICO) published the ‘Principles and Guidelines of a Curriculum for Ophthalmic Education of Medical Students’15 to try and address some of these challenges.

other

The ICO suggests that an evidence based ophthalmic curriculum should be included in the core curriculum of all medical schools. The undergraduate curriculum should also contain the basic skills and knowledge needed for satisfactory primary eye care, and ophthalmic instruction should enable students to recognise the ocular manifestations of systemic disease and when to refer to ophthalmology.15 During the report the elements of the curriculum are divided into hours in the curriculum, specific teaching methods, resources, and content. Forty to sixty hours (or 5 to 8 days) duration is recommended. Specific teaching methods such as lectures, clinical placements, case studies and integration with other specialities (e.g. neuroscience, neurology, endocrinology, and geriatric medicine) are also advocated. The suggested content of the curriculum consists of an extensive list of learning objectives. The knowledge topics are broadly divided into twelve categories of “fundamentals and principles of ophthalmology, cornea and external diseases, lens and cataract, neuro-ophthalmology, vitreoretinal disease, glaucoma, paediatric ophthalmology and strabismus, diseases of the eyelid lacrimal system and orbit, ocular manifestation of systemic disease, intraocular tumours, refraction and contact lens, and refractive surgery”.16 Eleven clinical skills are also defined these are “visual acuity, external inspection, pupillary reaction testing, pupillary dilation, ocular motility, direct ophthalmoscopy, intra-ocular pressure measurement, anterior chamber depth assessment, confrontation field testing, upper lid eversion, fluorescein staining of the cornea”.16

other

In the UK, the Royal College of Ophthalmologists (RCOphth) provides a recommended undergraduate curriculum. A brief outline of the recommended ophthalmic diagnostic skills and knowledge that a medical student should attain is described. Knowledge domains include:

other

“Competence and understanding in basic ophthalmic history taking and examination; the ability to differentiate common causes of red eye and understand their management; differentiate common causes of sudden and gradual loss of vision and understand their management; appreciate common ophthalmological manifestations of general medical conditions; understand how visual impairment impacts on activities of daily living; understand the importance of ocular public health issues, both in the UK and internationally”.17

other

Their recommended clinical skills are, visual acuity measurement, pupil assessment, confrontation visual fields, cover test and assessment of extraocular movement, and direct ophthalmoscopy.17 In 2016 the RCOphth updated their undergraduate curriculum in response to our preliminary report.18

other

Several studies have examined what the undergraduate ophthalmology curriculum should contain.19-26 In the last decade, research focus has been on the evaluation of the existing undergraduate ophthalmology curriculum as a method of aiding national curriculum development.27-29 The multiple factors that have to be considered in such curriculum evaluations include the educational environment, educational strategies, learning opportunities, content, learning outcomes, and assessment.30

other

The aim of this research was to investigate the current undergraduate ophthalmology curricula provided by the UK medical schools, compare them with the RCOphth and ICO guidelines, and determine the views of the undergraduate teaching leads in ophthalmology on the future direction of the curriculum. Eight themes on the undergraduate ophthalmology curriculum were explored, based on the framework of Harden’s ten questions of curriculum development. 32

other

This was a cross-sectional web based questionnaire. Participants included were the teaching leads for ophthalmology from each UK medical school. Two UK medical schools offer only pre-clinical courses and therefore do not include ophthalmology in their course. These medical schools were excluded from the study. A total of 31 medical schools were, therefore, invited to take part in the study. The ophthalmology teaching leads representing 29 medical schools responded to the questionnaire.

study

A web-based questionnaire was created using the Bristol Online Survey tool, after ethical approval was granted by the University of Nottingham Faculty of Medicine and Health Sciences, Research Ethics Committee. This questionnaire was modified from a survey designed by Baylis et al28 after permission and copyright was obtained from Informa Healthcare publishing group.

other

The questionnaire explored eight themes of curriculum development. These included, content and learning outcomes of the curriculum, how those learning outcomes are communicated to students, organisation of the curriculum and the teaching methods used, assessment, educational resources, and the educational environment. The opinions of the undergraduate ophthalmology teaching leads on the current and future management of the curriculum were also explored with Likert scale questions and free text responses. The categories of the knowledge and skills were taken from the pre-2016 Royal College of Ophthamologists guideline17 and the International Council of Ophthamology’s guideline on the undergraduate curriculum.15

other

After an initial pilot study the web based survey was distributed to all undergraduate ophthalmology teaching leads via an email link to a Bristol Online Survey webpage. Participants were given two months to respond to the questionnaire. Any respondents that had not replied by the complete by date were followed up and contacted via email to encourage a greater response rate. Responses were collated on the Bristol Online Survey website. Missing data was identified and a few individuals were contacted directly via email to clarify these answers to increase the data reliability.

study

Twenty-seven (96%) of the UK medical schools surveyed had an agreed undergraduate ophthalmology curriculum with twenty-two (76%) of the medical schools reporting defined learning outcomes. An ophthalmologist is involved in the design of the curricula in 93% (n=26) of these medical schools.

other

The clinical ophthalmic skills taught by the UK medical schools also adhered to the RCOphth guidelines33 but failed to encompass all the skills described in the ICO guidelines.15 Only seven of the eleven ICO recommended clinical skills are taught by 75% (n=21) of the UK medical schools, namely visual acuity, pupil reactions, red reflex assessment, direct ophthalmoscopy, confrontation field testing, ocular motility, and the cover test.

other

The undergraduate teaching leads considered knowledge about the different causes of red eye, sudden loss of vision, and assessment of visual acuity to be the most essential knowledge and skills to be included in the undergraduate course. Figures 1 and 2 show the ranking of the knowledge and skills by importance from 1 to 14.

other

Knowledge about intraocular tumours and the assessment of intraocular pressure were considered to be the least essential knowledge and skills in the curriculum. These were also found to be the least widely taught knowledge topic and skill by the UK medical schools (Figure 1 and Figure 2).

other

The organisation of the undergraduate ophthalmology curricula varied across the UK medical schools. Currently, nineteen medical schools provide a stand-alone attachment in ophthalmology of between one and two weeks’ duration. Half of the medical schools provided a one-week attachment, and a further five provided two weeks of ophthalmology teaching. Thirty-one percent (n=9) of the medical schools integrate ophthalmology with other subjects. These subjects included Neurology, Ear, Nose and Throat, Surgery and General Practice. An additional or optional student selected component is provided by twenty-five of the medical schools, these consist of a wide variety of formats. Student selected attachments/electives are widely available at the UK medical schools for students who wish to extend their experience in ophthalmology. These were commonly 4 weeks in duration with opportunities in research and audit, available in the third, fourth and fifth year of the medical course.

other

UK medical schools use various teaching methods in their undergraduate ophthalmology curriculum. The most commonly used teaching method was lectures (96%, n=28), closely followed by self-directed learning (83%, n=24) and small group teaching (82%, n= 24) (Table 1). An increasing number of medical schools (70%, n=21) used E-learning methods.

other

The clinical environments used to provide ophthalmology clinical experience were predominantly outpatient clinics, eye casualty and ophthalmic theaters. When participants were asked to rate whether teaching was given a priority in their departments (curriculum environment), 68% (n=20) agreed and strongly agreed that it was.

other

Ophthalmology is formally assessed by twenty-five medical schools during their curricula. Of these, 62% (n=18) had an ophthalmology practical skills assessment. However, only 55.1% (n=16) of medical schools included ophthalmology in their final written and 58.6% (n=17) in their final clinical skills examinations (Table 2).

other

A variety of resources are available to students during their ophthalmology placements. Eighty-nine percent (n=25) of medical schools provided clinical skills facilities for students to practice on simulated clinical models. Online resources and course handouts are also widely reported. Online learning modules were also used by 78% (n=22) of UK medical schools.

other

The final part of the survey focused on the opinions of the teaching leads on the current and future management of the undergraduate ophthalmology curriculum. Twenty-one respondents knew of the RCOphth undergraduate curriculum and its content. All respondents agreed that having an RCOphth curriculum to follow was vital, important, or useful. Ninety-six percent (n=27) of respondents also agreed that the RCOphth should have resources for teaching. More financial support was considered vital by 62% (n=18) while 78% (n=22) felt that more time for teaching was important or vital. No consensus was reached on the ideal duration of the undergraduate ophthalmology rotation; six respondents suggested 2 weeks, a further six suggested 4 weeks and three suggested 3 weeks. The remaining respondents suggested a range from four days to eight weeks duration. Despite these differences in opinion, all the respondents agreed that their teaching commitment should be recognised in their job plans as programmed activity (PA) sessions. Twenty-eight (28) (96%) of respondents were also prepared to share lectures or teaching resources with other UK medical schools.

other

In this study the ICO’s undergraduate curriculum guidelines15 were adapted to evaluate and compare what content was currently being taught at the UK medical schools. The ICO guidance was chosen as it provided an opportunity to compare the results of the current study to studies that had used the ICO guidance previously as benchmark to evaluate curriculum content both within the UK28, 29 and abroad.27,31

study

This study found that a higher percentage of UK medical schools taught each knowledge topic than had previously been reported.27-29,31 This may be because of a different sample population included in the current survey. However, it may also suggest a greater adherence to the ICO guidelines on the knowledge content of the curriculum is slowly emerging across the UK.

study

In contrast, the ICO recommended clinical skills are not widely taught. Previous UK studies have reported similar findings with less than 30% of medical schools teaching upper lid eversion, anterior chamber depth assessment, and intraocular pressure assessment.28,35 Reasons for this disparity in the taught clinical skills is unclear. Perhaps, the short curriculum duration in some schools does not allow sufficient time to ensure these skills can be taught. The individual educators’ attitudes towards the necessity of certain skills in the curriculum could also have an influence on their inclusion in the curriculum.

other

Boxplot graph showing the ophthalmology undergraduate leads ranking score for the importance of each knowledge topic in the curriculum. 1 representing the most important topic to 14, the least important topic (n=29); Key: OA- Ocular anatomy/physiology; RE- Acute red eye; OE-Ocular emergencies; SLV-Sudden loss of vision; LC-Lens and cataract, G- Glaucoma; MR-Medical retina/ macular degeneration; C- Cornea and external disease; VR- Vitreoretinal diseases; EM- Eye manifestations of systemic disease; N- Neuro-ophthalmology; P- Paediatric ophthalmology; IOT-Intraocular tumours; RCE- Refractive and corrective errors.

other

Opinions on the importance of knowledge topics and clinical skills to be included in the undergraduate curriculum vary. This is reflected in the large box plots in Figures 1 and 2. Previous studies agree that red eye, glaucoma, squint, visual failure, and the eye in systemic disease should be included.2,21,24-26 However, in the study by Spivey in 1971, the eye in systemic disease was not included.20 Clinical skills have also been ranked in previous studies.26 Visual acuity ranked the highest followed by pupillary reflexes, ophthalmoscopy and visual field testing, whilst intraocular pressure assessment was considered to be the least important skill.21 These results, very closely reflect the opinions of the respondents from this current study.

study

In their guidelines, the ICO does not consider being able to test the IOP of a patients’ eye as an essential skill.15 However, it is suggested that the student should know how IOP measurement is done and be given the opportunity to develop this skill, for example during an elective or student selected component in ophthalmology. It seems that uniformity of opinion on the content of the undergraduate ophthalmology curriculum is unlikely to be achieved. The ICO guideline15 provides a compromise on this issue and as such have been widely accepted.27-29,31,35-37

other

The communication of ophthalmic learning objectives has been rarely mentioned in the literature. Previous suggestions included handouts with checklists of conditions.24 This present study has identified the adoption of E-learning platform, such as Moodle, as viable method for communicating learning outcomes to students.

study

The number of medical schools providing standalone attachments in ophthalmology has significantly increased from the 29% previously reported.28 These previous studies suggested much greater levels of integration of ophthalmology teaching with other subjects.27,28 These differences may reflect curriculum organisational change over time in UK undergraduate teaching, and if properly used may increase emphasis on ophthalmology.

other

There is a high variability in the duration of ophthalmology placements across the UK. If we assume a day provides seven hours of ophthalmology teaching, then most students are receiving approximately 49 hours of ophthalmology teaching exposure. This is not dissimilar to the recommended duration of 40-60 hours (5-8 days) by the ICO.15 Previous studies report similar average durations of 7,31 7.6,28 8,29 and 8.9 days.27 The widespread provision of student-selected components may enable enthusiastic students to pursue their interest in ophthalmology and gain further experience and time in the specialty. However, these attachments are only available to a select number of students, and their uptake will greatly depend upon the quality of their initial ophthalmology attachment. This is a view shared by other authors.4,10,38

study

Boxplot graph showing the ophthalmology undergraduate leads ranking score for the importance of each clinical skill in the curriculum. 1 representing the most important clinical skill to 14, the least important clinical skill (n=29); Key: VA-Visual acuity; PI- Pupil reactions; ARR-Assessing red reflex; CT-Cover test; OM- Ocular motility; CFT-Confrontation field testing; DO-Direct ophthalmoscopy; FS- Fluorescein staining of cornea; ULE- Upper lid eversion; AC-Anterior chamber depth assessment; IOP- Intraocular pressure assessment.

other

The positive undergraduate ophthalmology educational environment reported in this study parallels the results of Baylis et al, who reported that most eye departments supported teaching by reducing clinic numbers in order to facilitate education.28 These results imply that UK hospital departments involved in undergraduate ophthalmology education regard teaching as valuable and worthwhile.

study

There remains, however, large variation throughout the UK in the assessment of the undergraduate ophthalmology curriculum. Welch and Eckstein have previously reported a similar level of clinical skills assessment.29 Baylis et al (2011),28 found comparable levels of assessment of ophthalmology knowledge as in other examinations, but fewer schools reporting ophthalmology clinical skills assessments. Only about a third of medical schools require students to pass their ophthalmology assessments in order to complete the year.29 A similar picture is seen in Canada where only 43% of schools count an ophthalmology evaluation towards a students’ assessment record.39 How the curriculum can be most effectively assessed remains largely unstudied.

other

The almost universal adoption of simulation by medical schools to practice ophthalmic clinical skills has occurred recently. Such simulation has been previously adopted in ophthalmology teaching with reports of ophthalmoscopy practice mannequins in Canada in 1998.39 This should be encouraged as it allows students to practice their skills prior to examining patients. However, such simulation should not be considered as adequate replacement to patient examination.

other

This study has shown significant agreement between the respondents about the future direction of the undergraduate ophthalmology curriculum. Closer collaboration between medical schools to improve educational standards and reduce local workload has been proposed previously in Australasia,27 America,40 and Canada.31,39

other

Our results highlight the direction that the undergraduate curriculum could take in the future to survive in the crowded undergraduate medical course. Areas for possible resource sharing could include clinical case vignettes, e-learning modules and a database of exam questions. The findings from this study were shared with the RCOphth Education Committee in 2015. We welcome the launch of a new curriculum by the RCOphth in 2016 “Eyes & Vision Curriculum” for Undergraduate and Foundation Doctors.18

other

The data collected from this questionnaire is limited as only a snap shot of the undergraduate ophthalmology curriculum can be produced from a cross-sectional survey. The sample size of this study was small. However, this sample reflects the individuals responsible for the undergraduate ophthalmology curriculum across the UK. These individuals are therefore best placed to report on the curriculum. With a response rate of 93% (n=29/31), the data collected is likely to be representative.

study

Attempts were made to reduce measurement error. However, misinterpretation of the survey questions by some of the participants may have resulted in variation in the ranking of the importance of knowledge and skills essential to the curriculum. Asking respondents to organise priorities where there are more than five options to rank, has been previously reported as overwhelming.41

other

The final part of the survey asked the respondents their opinions on the curriculum using a rating scale. This type of question allows a relative degree of preference to be documented.41 These types of questions can however be open to interpretation as one participant’s ‘agree’ may be another respondent’s ‘strongly agree’.

other

Improved adherence to the ICO guideline would ensure greater standardisation of ophthalmology teaching across the UK. A formal edict from the RCOphth about the organisation, duration and content of the undergraduate ophthalmology curriculum, would assist the undergraduate teaching leads in ensuring ophthalmology’s continuing presence in the undergraduate medical curriculum. Resource sharing between the medical schools facilitated by the RCOphth, could also improve the effectiveness of undergraduate ophthalmology teaching while reducing the workload of local teaching departments and medical schools.

other

The authors would like to thank Oliver Baylis for sharing his questionnaire proforma; also, Eric Barnes, Daniel Byles, Phillip Murray, and Narciss Okhravi for their advice on the pilot version of the questionnaire. The authors would also like to thank the undergraduate ophthalmology teaching leads for their support in completing the study questionnaire.

other

Echinoderm microtubule-associated protein-like 4- anaplastic lymphoma kinase EML4-ALK gene rearrangements are present in around 5% of non-small cell lung carcinoma (NSCLC) patients . In the majority of cases, they are adenocarcinomas . It is noteworthy that, in some series, more than 17% of lung cancer patients who never smoked have this genomic alteration . Currently, analysis for the EML4-ALK status is mainly performed by immunohistochemistry (IHC) and confirmed by FISH analysis from tissue biopsies . This status can also be detected on cytological samples by immunochemistry (ICC) and/or by FISH analysis . However, molecular biology methods are also used for this genomic alteration detection in tissues or cells, in particular by next generation sequencing (NGS) . The presence of an EML4-ALK rearrangement leads to targeted treatment . Despite a frequent early response, relapse and progression is systematically noted, which can lead to a change in targeted therapy . In most cases, this tumor progression is linked to the emergence of EML4-ALK mutations .

review

The development of liquid biopsies (LB) has changed substantially the care of patients with solid cancers, in particular late stage lung cancer patients . Addition of LB to management of advanced lung cancer can have a strong impact at three different levels: (i) for the initial detection of actionable oncogenic drivers; (ii) for the identification of resistance mutations in patients relapsing on targeted therapies; and (iii) for the monitoring of the response to therapy and the prediction of the clinical outcome . Within the context of targeted therapeutics, the question arises of whether to use LB on a daily basis to evaluate the presence of rearrangements or mutations in ALK and to propose this approach instead of an analysis made from tissue and/or cytological samples.

review

This review will briefly discuss the notion of LB and then analyze the advantages and limitations of the different methods that evaluate the ALK status using blood from patients with advanced stage lung cancer, performed either at the time of diagnosis or during follow-up of the patient on targeted treatments.

review

LB contains several components of interest for research into potential therapeutic targets . Two main components have been studied, circulating tumor cells (CTCs) and plasma . At present, most of the approaches in routine practice use plasma, in particular for activating and resistance mutations detection in EGFR in patients with advanced stage lung cancer.

review

Many direct and indirect methods have been developed to detect and characterize CTCs associated with lung cancer . These different methods have advantages and disadvantages . Different innovative technologies to improve the detection and characterization of CTCs in NSCLC have been developed, including CTC microchips, filtration devices, quantitative reverse-transcription PCR assays, and automated microscopy systems . Direct approaches are based on the isolation of CTCs through their physical characteristics . Thus, in these latter methods, the CTCs are detected through their different size, their loss of deformability, their higher density and their electrical charges . Indirect techniques exploit the molecular properties of CTCs . Among these latter methods, the only FDA approved method for CTC detection, the CellSearch method (Janssen Diagnostics Company, Raritan, NJ, USA), has been FDA approved for CTC detection in metastatic breast, prostate and colon cancer patients. However, the CellSearch method has not been approved for CTC detection in metastatic lung cancer patients by the FDA. It is noteworthy that, since CTCs that have undergone epithelial-mesenchymal transition cannot express epithelial biomarkers, the CellSearch system can miss the detection of a number of CTCs of interest in lung cancer patients . Currently, the results obtained from CTCs characterization do not change the treatment of patients on a daily practice. Thus, to date, studies into the detection and characterization of CTCs are still a matter of translational and clinical research in NSCLC patients. The methods used to isolate free plasma DNA can also isolate circulating tumor DNA, which is present in variable amounts depending on the cancer patients and tumor stage. The volume of whole blood used is generally 10 mL but a larger volume is sometimes necessary for certain analyses, in particular NGS. The analyses are either targeted techniques such as allele specific PCR or droplet digital PCR or NGS techniques using panels of different genes. The sequenom mass array approach can also detect genomic alterations on several genes in blood samples . At present, the detection of activating or resistance mutations in EGFR in plasma samples can be performed in the clinical care of patients with advanced stage lung cancer according to the different practices available.

review

Other components of blood can be used for detection of some therapeutic targets for patients with lung cancer, including free plasma RNA or RNA in a complex with proteins in exosome or in association with platelets. Among these RNA, microRNA are of particular interest. Microparticles in plasma are heterogeneous components associated with exosomes, nucleosomes and particles of larger size. However, currently, none of these elements are used in a routine manner for detection of genomic alterations in lung cancer patients.

review

Several recent studies showed that EML4-ALK rearrangement could be detected within CTCs . In 2012 a study from our group showed very good agreement between the presence of EML4-ALK rearrangements in tissue biopsies and in CTCs . In this study, detection of CTCs was performed with a direct method (ISET) using filtration of blood through a porous filter which retained on its surface CTCs that are larger than circulating hematological cells. The ISET method used in this work has the advantage of being able to characterize the cyto-morphology of the isolated CTCs and the malignant cytological criteria can be used to identify the different populations of cells on the surface of the filter . Moreover, this approach holds the advantage of allowing correlation between the ALK status, determined by ICC and/or FISH and the malignant cytological characteristics which limits potential false positive results. In this context, since it was reported that ICC and FISH analyses gave the same results, the question was to know if ICC can be used for systematic rapid screening of ALK status in isolated CTCs. In fact, ICC results can be easily interpreted by cytopathologists in any pathology laboratory and performed in an identical way to detection of ALK rearrangements using tissues or cells. However, despite the promising results from the initial study, one limitation was that only a small number of patients at a single center were studied . In this regard, a French multi-center study with a large number of patients with a tissue biopsy ALK positive status is now in progress (www.clinicaltrials.gov/ct2/show/NCT02372448). Another limitation of the initial study was that no information existed concerning a possible association between targeted treatments for patients presenting with ALK positive CTCs and the therapeutic response, in particular with respect to the CTC number. Finally, some patients often had less than 100 analyzable CTCs, which make the interpretation of a positive score difficult to take into consideration. A second study using the ISET method also showed a correlation between the presence of an EML4-ALK rearrangement in tissues and CTCs and a correlation with the disappearance of CTCs with a good response to targeted therapy . The analysis with immunofluorescence and confocal microscopy demonstrated that CTCs harboring a unique ALK rearrangement may have a mesenchymal phenotype with the potential to drive metastatic progression of ALK-positive NSCLC . Using the same ISET method for CTCs detection, a recent study looked for the prognostic value of patients with CTCs having ALK-copy number gain (CNG) at baseline and treated with crizotinib . In this study, the dynamic change in the numbers of CTCs with ALK-CNG was a predictive biomarker for crizotinib efficacy in ALK-rearranged NSCLC patients . Another study confirmed the correlation between a positive EML4-ALK status detected in tissues and in CTCs detected by FISH analysis . This latter study showed that persistence of or an increase in the number of ALK positive CTCs on FISH analysis correlated with tumor progression in patients on treatment with crizotinib, in particular with the appearance of a new molecular profile . Finally, a recent work demonstrated that ALK-rearrangement can also be detected in CTCs collected from advanced NSCLC patients by another approach, the NanoVelcro system .

study

Free plasma DNA is routinely used to analyze mutations in EGFR in patients with advanced stage lung cancer. In contrast, plasma RNA is not used routinely, or very rarely used, for detection of EML4-ALK rearrangements. Nonetheless, the detection of this genomic alteration can be performed with plasma using the reverse transcription-polymerase chain reaction (RT-PCR). While in theory this approach can detect an ALK rearrangement in the blood of patients shown to be ALK positive after analysis of a tissue biopsy, it still has some limits. The first hurdle concerns mastering of the pre-analytical delay between the blood sampling and the extraction of the plasma RNA, knowing that the blood RNA degrades rapidly after sampling. The addition of a solution such as Trizol or immediate plasma freezing at low temperature (minimum of −80 °C) allows better conservation of the integrity of the plasma RNA. In practice, this is not always possible to do in clinical routine. The second hurdle is the low detection limit of the EML4-ALK rearrangement by RT-PCR (around 20–25%), which also limits its use in daily practice .

review

Several studies have shown that platelets can sequester tumor RNA by taking up circulating micro-vesicles . These tumor-educated platelets can be isolated and constitute an enriched source of tumor RNA for detection of EML4-ALK rearrangements. A recent study showed that this approach, while being very specific, is more sensitive than those used for plasma RNA . It is noteworthy that the persistence of EML4-ALK rearrangements in platelets of patients treated with crizotinib is associated with tumor progression and poor prognosis . Moreover, the reappearance of this fusion transcript in platelets is a sign of resistance to crizotinib and it occurs several months before progression is detected on radiological examination.

study

In addition to the ALK rearrangements, other fusion transcripts of the genes for ROS proto-oncogene 1 (ROS1), ret-proto-oncogene (RET) or NTRK genes have been identified in 1–2% of patients with advanced stage lung cancer. These genomic alterations are liable to be associated with targeted therapies. In this context, NGS approaches were developed to identify these genomic alterations in purified free nucleic acids in blood samples . Based on the hybrid-capture approach, these techniques not look only for rearrangements but also point mutations, insertions, deletions and variations in the number of copies. At present, the NGS techniques are being developed for use with LB and to thereby allow detection of several targeted therapeutics of interest with a single approach. However, these approaches probably still lack sufficient sensitivity to date and different ring studies will be probably useful before using them in clinical routine practice.

review