A Comparison of Nine Machine Learning Mutagenicity Models and Their Application for Predicting Pyrrolizidine Alkaloids

Random forest, support vector machine, logistic regression, neural networks and k-nearest neighbor (lazar) algorithms, were applied to a new Salmonella mutagenicity dataset with 8,290 unique chemical structures utilizing MolPrint2D and Chemistry Development Kit (CDK) descriptors. Crossvalidation accuracies of all investigated models ranged from 80 to 85% which is comparable with the interlaboratory variability of the Salmonella mutagenicity assay. Pyrrolizidine alkaloid predictions showed a clear distinction between chemical groups, where otonecines had the highest proportion of positive mutagenicity predictions and monoesters the lowest.

The Interlaboratory Performance in Measurement of Dabigatran and Rivaroxaban: Results of the College of American Pathologists External Quality Assessment Program

Context.— Assessing direct oral anticoagulant (DOAC) drug levels by reliable laboratory assays is necessary in a number of clinical scenarios. Objective.— To evaluate the performance of DOAC-specific assays for various concentrations of dabigatran and rivaroxaban, assess the interlaboratory variability in measurement of these DOACs, and investigate the responsiveness of the routine clotting assays to various concentrations of these oral anticoagulants. Design.— College of American Pathologists proficiency testing survey data from 2013 to 2016 were summarized and analyzed. Results.— For dabigatran, the interlaboratory coefficient of variation (CV) of ecarin chromogenic assay was broad (ranging from 7.5% to 29.1%, 6.3% to 15.5%, and 6.8% to 9.0% for 100-ng/mL, 200-ng/mL, and 400-ng/mL targeted drug concentrations, respectively). The CV for diluted thrombin time for dabigatran was better overall (ranging from 11.6% to 17.2%, 9.3% to 12.3, and 7.1% to 11.2% for 100 ng/mL, 200 ng/mL, and 400 ng/mL, respectively). The rivaroxaban-calibrated anti-Xa assay CVs also showed variability (ranging from 11.5% to 22.2%, 7.2% to 10.9%, and 6.4% to 8.1% for 50-ng/mL, 200-ng/mL, and 400-ng/mL targeted drug concentrations, respectively). The prothrombin time (PT) and activated partial thromboplastin time (aPTT) showed variable dose and reagent-dependent responsiveness to DOACs: PT was more responsive to rivaroxaban and aPTT to dabigatran. The undiluted thrombin time showed maximum prolongation across all 3 dabigatran concentrations, making it too sensitive for drug-level monitoring, but supporting its use as a qualitative screening assay. Conclusions.— DOAC-specific assays performed reasonably well. While PT and aPTT cannot be used safely to determine DOAC degree of anticoagulation, a normal thrombin time excludes the presence of dabigatran.

Harmonized procedures lead to comparable quantification of total oxylipins across laboratories

Oxylipins are potent lipid mediators involved in a variety of physiological processes. Their profiling has the potential to provide a wealth of information regarding human health and disease and is a promising technology for translation into clinical applications. However, results generated by independent groups are rarely comparable, which increases the need for the implementation of internationally agreed upon protocols. We performed an interlaboratory comparison for the MS-based quantitative analysis of total oxylipins. Five independent laboratories assessed the technical variability and comparability of 133 oxylipins using a harmonized and standardized protocol, common biological materials (i.e., seven quality control plasmas), standard calibration series, and analytical methods. The quantitative analysis was based on a standard calibration series with isotopically labeled internal standards. Using the standardized protocol, the technical variance was within ±15% for 73% of oxylipins; however, most epoxy fatty acids were identified as critical analytes due to high variabilities in concentrations. The comparability of concentrations determined by the laboratories was examined using consensus value estimates and unsupervised/supervised multivariate analysis (i.e., principal component analysis and partial least squares discriminant analysis). Interlaboratory variability was limited and did not interfere with our ability to distinguish the different plasmas. Moreover, all laboratories were able to identify similar differences between plasmas. In summary, we show that by using a standardized protocol for sample preparation, low technical variability can be achieved. Harmonization of all oxylipin extraction and analysis steps led to reliable, reproducible, and comparable oxylipin concentrations in independent laboratories, allowing the generation of biologically meaningful oxylipin patterns.

Multicentre study on the consistency of PD-L1 immunohistochemistry as predictive test for immunotherapy in non-small cell lung cancer

AimsInvestigate the impact of interlaboratory- and interobserver variability of immunohistochemistry on the assessment of programmed death ligand 1 (PD-L1) in non-small cell lung cancer (NSCLC).MethodsTwo tissue microarrays (TMAs) were constructed from 50 (TMA-A) and 51 (TMA-B) resected NSCLC cases, and distributed among eight centres. Immunostaining for PD-L1 was performed using Agilent’s 22C3 pharmDx Assay (pharmDx) and/or a 22C3 laboratory developed test (LDT). The interlaboratory variability of staining- and interobserver variability of scoring for PD-L1 were assessed in selected critical samples (samples at the cut-off of positivity) and non-critical samples. Also, PD-L1 epitope deterioration in time in stored unstained slides was analysed. Krippendorff’s alpha values (0=maximal, 1=no variability) were calculated as measure for variability.ResultsFor interlaboratory variability of immunostaining, the percentage of PD-L1 positive cases among centres ranged 40%–51% (1% cut-off) and 23%–30% (50% cut-off). Alpha values at 1% cut-off were 0.88 (pharmDx) and 0.87 (LDT) and at 50% cut-off 0.82 (pharmDx) and 0.95 (LDT). Interobserver variability of scoring resulted in PD-L1 positive cases ranging 29%–55% (1% cut-off) and 14%–30% (50% cut-off) among pathologists. Alpha values were at 1% cut-off 0.83 (TMA-A) and 0.66 (TMA-B), and at 50% cut-off 0.77 (TMA-A) and 0.78 (TMA-B). Interlaboratory variability of staining was higher (p<0.001) in critical samples than in non-critical samples at 50% cut-off. Furthermore, PD-L1 epitope deterioration in unstained slides was observed after 12 weeks.ConclusionsThe results provide insight in factors contributing to variability of immunohistochemical assessment of PD-L1, and contribute to more reliable predictive testing for PD-L1.

Interobserver, intraobserver, and interlaboratory variability in reporting pT4a colon cancer

Clinical significance of the pT4 category in colon cancer is increasing with several therapeutic implications. The aim of this study was to evaluate variability in diagnosing pT4a colon cancer. Twelve pathologists classified 66 preselected scanned Hematoxylin/Eosin-stained slides with tumor cells at a distance of 25–1500 μm (n = 22), 0–25 μm (n = 22), or on (n = 22) the peritoneal surface. Inter- and intraobserver variability were calculated using Kappa statistics. For interlaboratory variability, pathology reports of pT3 and pT4a colon cancer were extracted from the Dutch Pathology Registry between 2012 and 2015. The proportion of pT4a (pT4a/(pT3+pT4a)) was compared between 33 laboratories. Potential risk of understaging was assessed by determining the average number of blocks taken from pT3 and pT4a N0-2M0 tumors with metachronous peritoneal metastasis. Interobserver variability among 12 pathologists was 0.50 (95%CI 0.41–0.60; moderate agreement). Intraobserver variability (8 pathologists) was 0.71 (substantial agreement). A total of 7745 reports with pT3 or pT4aN0-2M0 colon cancer from 33 laboratories were included for interlaboratory analysis. Median percentage of pT4a was 15.5% (range 3.2–24.6%). After adjustment for case mix, 8 labs diagnosed pT4a significantly less or more frequently than the median lab. Metachronous peritoneal metastases were histologically verified in 170 of 6629 pT3 and in 129 of 1116 pT4a tumors, with a mean number of blocks of 4.03(SD 1.51) and 4.78 (SD 1.76) taken from the primary tumors, respectively (p < 0.001). A substantial variability in diagnosing pT4a colon cancer exists, both at pathologist and laboratory level. Diagnosis of pT4a stage appears to be challenging and there is a need for standardizing assessment of this pathological entity.

The Brazilian Reproducibility Initiative: a systematic assessment of Brazilian biomedical science

With concerns over research reproducibility on the rise, systematic replications of published science have become an important tool to estimate the replicability of findings in specific areas. Nevertheless, such initiatives are still uncommon in biomedical science, and have never been performed at a national level. The Brazilian Reproducibility Initiative is a multicenter, systematic effort to assess the reproducibility of the country’s biomedical research by replicating between 60 and 100 experiments from Brazilian life sciences articles. The project will focus on a set of common laboratory methods, performing each experiment in multiple institutions across the country, with the reproducibility of published findings analyzed in the light of interlaboratory variability. The results, due in 2021, will allow us not only to estimate the reproducibility of Brazilian biomedical science, but also to investigate if there are aspects of the published literature that can be used to predict it.