Clinical features of infection caused by non-tuberculous mycobacteria: 7 years’ experience

Introduction

Non-tuberculous mycobacteria (NTM) are ubiquitous organisms associated with various infections. The aim of the study was to determine the most relevant clinical characteristics of NTM during the 7-year period.

Methodology

A retrospective study of NTM infections was conducted between January 2009 and December 2016. The American Thoracic Society/Infectious Disease Society of America criteria were used to define cases of pulmonary or an extrapulmonary site.

Results

A total of 85 patients were included in the study. Pulmonary cases predominated 83/85 (98%), while extrapulmonary NTM were present in 2/95 (2%) patients. Overall, ten different NTM species were isolated. The most common organisms were slow-growing mycobacteria (SGM) presented in 70/85 (82.35%) patients. Isolated SGM strains were Mycobacterium avium complex (MAC) in 25/85 (29.41%) patients, M. xenopi in 20/85 (23.53%) patients, M. kansasii in 15/85 (17.65%) patients and M. peregrinum and M. gordonae in 5/85 (5.88%) patients each. Isolated rapid-growing mycobacteria (RGM) strains were M. abscessus in 8/85 (9.41%) patients, M. fortuitum in 4/85 (4.71%) patients and M. chelonae in 3/85 (3.53%) patients. Almost all patients (98%; 83/85) had comorbidities. Among 75 (88.24%) patients who completed follow-up, 59 (69.41%), 10 (11.76%) and 6 (7%), were cured, experienced relapse and died, respectively.

Conclusion

In the present study, pulmonary NTM infections were more frequent compared to extrapulmonary disease forms. SGM were most common isolates with MAC pulmonary disease the most frequently found. Comorbidities have an important role in NTM occurrence. Further investigation should focus on an NTM drug susceptibility testing.

     

The effect of alcohols as the third component on diffusion in mixtures of aromatics and ketones

With laboratory and numerical work, we demonstrate that one of the main diffusion coefficients and the smaller eigenvalue of the Fick diffusion matrix are invariant to the number of methylene groups of the alcohol in ternary mixtures composed of an aromatic (benzene), a ketone (acetone) and one of three different alcohols (methanol, ethanol or 2-propanol). A critical analysis of the relationship between the kinetic and thermodynamic contributions to the diffusion coefficients allows us to explain this intriguing behaviour of this class of mixture. These findings are reflected by the diffusive behaviour of the according binary subsystems. Our approach provides a promising systematic framework for future investigations into the important and challenging problem of transport diffusion in multicomponent liquids.

The Fick diffusion coefficient matrix of three ternary mixtures composed of an aromatic (benzene), a ketone (acetone) and one of three different alcohols (methanol, ethanol or 2-propanol) is investigated with laboratory and numerical work.

Multicomponent diffusion plays a crucial role in various natural and industrial processes involving mass transfer.^1–3 Liquids appearing in nature and technical applications are essentially multicomponent. However, only data on binary diffusion coefficients are relatively abundant because the diffusion behavior of ternary and higher mixtures is much more complex.^4,5 Describing the isothermal–isobaric diffusion of a ternary mixture by Fick’s law requires four different diffusion coefficients that are composition dependent. The presence of cross diffusion coefficients aggravates the interpretation and data processing in experimental work, resulting in large uncertainties.^6,7 Thus, efforts are being made to develop new methods for analysis of multicomponent diffusion explicitly addressing various degrees of complexity.^8–10 Predictive equations for multicomponent diffusion of liquids mostly rely on extensions of the Darken relation,^11–13 which is only valid for ideal mixtures.¹⁴ The underlying physical phenomena in non-ideal mixtures are not well understood and the lack of experimental data impedes the development and verification of new predictive equations.The objective of this study was not only to measure and predict the Fick diffusion coefficient matrix for a series of ternary liquid mixtures, rather, the emphasis lied on understanding common features and whether they can be related to the behavior of the pure components and binary subsystems. Three ternary mixtures that are composed of organic compounds were selected, i.e. an aromatic, a ketone and an alcohol. Throughout, the first two components were benzene (1) and acetone (2) and the third component was one of the alcohols, methanol, ethanol or 2-propanol. For each mixture, nine state points along a composition path with a constant content of benzene, x₁ = 0.33 mol mol⁻¹, were studied under ambient conditions (298.15 K and 0.1 MPa). Seven of the state points were ternary mixtures and two were binary subsystems. To obtain reliable results for the Fick diffusion coefficient matrix, two complementary approaches were used, i.e. experiments and predictive molecular simulations. This combination allows for a critical analysis and leads to a deeper understanding of the underlying phenomena.^14,15The Taylor dispersion technique was utilized for the experiments.^16,17 In this method, a small quantity of mixture with a slightly different composition is injected into a laminar stream. It disperses due to convection and diffusion while flowing through a capillary tube and the refractive index is measured at its end to sample the concentration distribution. We have used the same apparatus as in previous works.^6,7 The Fick diffusion matrix is obtained by fitting working equations to the measured signal, i.e. the Taylor peak. The mathematical model of the Taylor dispersion technique was originally developed on the basis of Fick’s law in the volume reference frame. In a ternary mixture, two molar fluxes J^v_i relative to a volume averaged velocity are related to gradients of molar concentration ∇C_i with four diffusion coefficients D^v_ij. Alternatively, fluxes expressed in the molar reference frame J_i are relative to a molar averaged velocity and the mole fraction gradients ∇x_i act as a driving force1with molar density ρ. The fluxes of all three components are constrained by ΣJ_i = 0. The main diffusion coefficients D₁₁ and D₂₂ relate the flux of one component to its own mole fraction gradient and the cross diffusion coefficients D₁₂ and D₂₁ describe the coupling of the flux of one component with the gradient of the other. The third component does not appear in eqn (1) explicitly, but in general it affects all four diffusion coefficients. The transformation of experimental data from the volume to the molar reference frame (D^v_ij to D_ij) could be done here on the basis of the pure component volumes (see the ESI^†).Equilibrium molecular dynamics (MD) simulations were employed in this work, allowing for examination at the microscopic scale. The underlying molecular models were rigid, non-polarizable force fields of united atom type, consisting of a varying number of Lennard–Jones, point charge, dipole and quadrupole sites (see the ESI^†). Note that the force field parameters were adjusted to pure fluid properties only so that all simulation results for the mixtures are strictly predictive. Diffusion coefficients were sampled with the Green–Kubo formalism, based on integrated correlation functions of net velocities of the contained species.^11,15 Thereby, phenomenological coefficients Δ_ij were obtained, associating the diffusive fluxes with the chemical potential gradients ∇μ_i2with gas constant R and temperature T. Fluxes J_i correspond to the molar reference frame as in eqn (1).The diffusion coefficients from experiment and simulation are related to different driving forces so that the chemical potential gradients have to be transformed to the mole fraction gradients for their comparison.¹⁸ This transformation is contained in the thermodynamic factor matrix Γ3with the activity coefficient of species i being γ_i, which expresses the non-ideality of a mixture with respect to the composition. This relationship shows that the Fick diffusion coefficients are actually the product of two contributions, a kinetic Δ_ij and a thermodynamic Γ_ij. The separate observation of these two contributions promotes understanding of the underlying physical phenomena. In the present study, the thermodynamic factor was calculated using the Wilson excess Gibbs energy (g^E) model, using parameters fitted to experimental vapor–liquid equilibrium data of the binary subsystems (see the ESI^†). This combination of MD simulation results with a g^E model was successfully used in previous work to predict Fick diffusion coefficients, including several binary subsystems of the ternary mixtures studied here.¹⁹The four elements of the Fick diffusion coefficient matrix were determined for the three ternary mixtures, benzene + acetone + methanol/ethanol/2-propanol, for nine different compositions, each at ambient temperature and pressure.Results for the first main element of the diffusion matrix D₁₁, which relates the flux of benzene to its own mole fraction gradient, are shown in Fig. 1(a). The experimental data agree quantitatively with the molecular simulation data. D₁₁ increases with the acetone content in the ternary mixture. Since mixtures with a constant mole fraction of benzene (x₁ = 0.33 mol mol⁻¹) were studied throughout, the left edge of Fig. 1(a) corresponds to the binary limit of benzene + alcohol, while the right edge corresponds to that of benzene + acetone. Analysis of the ternary diffusive fluxes implies the following asymptotic behavior of the diffusion coefficients towards the binary limits:⁷ (i) at the infinite dilution limit, x₂ → 0, the ternary coefficient D₁₁ tends to the binary Fick diffusion coefficient of benzene + alcohol; (ii) at the other limit, x₃ → 0, D₁₁ − D₁₂ = D₂₂ − D₂₁ → D_bin (benzene + acetone) should hold. The present experimental and simulation results for D₁₁ are consistent with these asymptotic limits.Open in a separate windowFig. 1Top: The main Fick diffusion coefficient (molar reference frame) of benzene D₁₁ in the three ternary mixtures benzene (1) + acetone (2) + alcohol (3) at a constant benzene mole fraction x₁ = 0.33 mol mol⁻¹ from experiment (triangles) and MD simulation combined with the Wilson g^E model (circles). Both data sets were sampled at the same compositions, but are slightly shifted in the plot for visibility reasons. The symbols at the edges of this plot are the binary diffusion coefficients of benzene + alcohol (x₂ → 0) and of benzene + acetone (x₃ → 0). Bottom: The binary Fick diffusion coefficient of the subsystems benzene + alcohol and benzene + acetone. Most of the binary experimental data were taken from the literature.^20–27An inspection of Fig. 1(a) provides an unexpected finding: the main element D₁₁ is almost identical for all three mixtures along the examined composition path, i.e. it is independent of the contained type of alcohol. To explain this intriguing behavior of D₁₁, the properties of the pure components are considered first (see

Peripheral bronchial identification on chest CT using unsupervised machine learning

Purpose

To automatically identify small- to medium-diameter bronchial segments distributed throughout the lungs.

Methods

We segment the peripheral pulmonary vascular tree and construct cross-sectional images perpendicular to the lung vasculature. The bronchi running with pulmonary arteries appear as concentric rings, and potential center points that lie within the bronchi are identified by looking for circles (using the circular Hough transform) and rings (using a novel variable ring filter). The number of candidate bronchial center points are further reduced by using agglomerative hierarchical clustering applied to the points represented with 18 features pertaining to their 3D position, orientation and appearance of the surrounding cross-sectional image. Resulting clusters corresponded to bronchial segments. Parameters of the algorithm are varied and applied to two experimental data sets to find the best values for bronchial identification. The optimized algorithm was then applied to a further 21 CT studies obtained using two different CT vendors.

Results

The parameters that result in the most number of true positive bronchial center points with > 95% precision are a tolerance of 0.15 for the hierarchical clustering algorithm and a threshold of 75 HU with 10 spokes for the ring filter. Overall, the performance on all 21 test data sets from CT scans from both vendors demonstrates a mean number of 563 bronchial points detected per CT study, with a mean precision of 96%. The detected points across this group of test data sets are relatively uniformly distributed spatially with respect to spherical coordinates with the origin at the center of the test imaging data sets.

Conclusion

We have constructed a robust algorithm for automatic detection of small- to medium-diameter bronchial segments throughout the lungs using a combination of knowledge-based approaches and unsupervised machine learning. It appears robust over two different CT vendors with similar acquisition parameters.

     

Age-associated accumulation of CMV-specific CD8+ T cells expressing the inhibitory killer cell lectin-like receptor G1 (KLRG1)

Large clonal expansions of peripheral CD8+ T cells carrying receptors for single epitopes of CMV and EBV are common in the elderly and may be associated with an immune risk phenotype predicting mortality. Here we show that the frequency of CD8+ T cells expressing the inhibitory killer cell lectin-like receptor G1 (KLRG1), a marker of cells unable to undergo further clonal expansion, was markedly elevated in CD8+ T cells from old donors. Moreover, tetramer staining revealed that the elevated frequency of CMV-specific CD8+ T cells in the elderly was due to an accumulation of cells bearing this dominant negative receptor. The fraction of CMV-specific T cells able to secrete interferon-gamma after specific antigenic stimulation was significantly lower in the elderly than in the young, although the total number of functional cells was comparable. Therefore, the majority of the clonally expanded virus-specific CD8+ cells in the elderly was dysfunctional. Thus, T cell responses are altered in the aged by an accumulation of replicatively senescent dysfunctional T cells carrying receptors for persistent herpes viruses. The presence of clonal expansions of such virus-specific cells may shrink the available repertoire for other antigens and contribute to the increased incidence of infectious disease in the elderly.     

Myeloablative intensified conditioning regimen with in vivo T-cell depletion (ATG) followed by allografting in patients with advanced multiple myeloma. A phase I/II study of the German Study-group Multiple Myeloma (DSMM)

We investigated toxicity and efficacy of in vivo T-cell depletion with anti-thymocyte globulin (ATG) as part of an intensified myeloablative conditioning regimen followed by allogeneic stem cell transplantation in patients with advanced multiple myeloma. The conditioning regimen consisted of modified total body irradiation, busulfan and cyclophosphamide (n=15) or in the case of prior dose-limiting radiotherapy of busulfan and cyclophosphamide (n=3). The median age was 44 years (range, 29-53) and the median time from diagnosis to transplant was 12 months (range, 6-144). Grade II-IV acute graft-versus-host disease (GvHD) occurred in six patients (35%). Severe grade III/IV GvHD developed in one patient (6%). Three patients died of therapy-related causes (17%). A complete remission (CR) with negative immunofixation after allogeneic transplantation was seen in eight of the evaluable patients (53%). After a median follow-up of 41 months (range, 8-84), the estimated overall survival at 6 years for all patients is 77% (CI 95%: 58-96%). The estimated progression-free survival at 6 years for all patients is 31% (CI 95%: 2-59%) and 46% (CI 95%: 9-83%) for patients with CR. In vivo T-cell depletion with ATG resulted in a low rate of severe GvHD with low treatment-related mortality, and a substantial number of long-term survivors.     

18F‐Fluorodeoxyglucose (FDG)‐PET features of focal nodular hyperplasia (FNH) of the liver

Abstract: Aim: The aim of this paper is to describe the imaging pattern of focal nodular hyperplasia (FNH) by ¹⁸F‐fluorodeoxyglucose (18F‐FDG) positron emission tomography (PET). Methods: Eight consecutive asymptomatic patients with histologic proof of FNH underwent 18F‐FDG PET imaging. The lesions were found incidentally. The 18F‐FDG PET imaging was performed with a dedicated PET tomograph after intravenous injection of 300–370 MBq 18F‐FDG. The 18F‐FDG accumulation in the lesions was (semi)quantified by calculating the standardized uptake value (SUV) and SUV has been corrected for the lean body mass (LBM). Eight patients with liver metastases spread from melanoma (n=2) and colorectal carcinoma (n=6) served as controls. The size of the FNH lesions and of the control group ranged from 2.0 to 8.5 cm (mean 4.83 cm±2.37) and from 1.5 to 6 cm (mean 3.28±1.52), respectively. Results: While in malignant liver lesions the accumulation of 18F‐FDG was significantly increased, all FNH lesions showed normal or even decreased accumulation of 18F‐FDG. In FNH lesions, SUV ranged between 1.5 and 2.6 (mean 2.12±0.38), whereas all liver metastases showed an increased SUV ranging between 6.20 and 16.00 (mean 10.07±3.79). The SUV corrected for LMB (SUV_LBM) was similar to the SUV and ranged between 0.9 and 2.2 (mean 1.81±0.41) for FNH and between 5.9 and 16.3 (mean 9.15±4.03), respectively. Conclusion: In contrast to liver metastases, there is no increased glucose metabolism in FNH in vivo. The imaging feature of FNH by 18F‐FDG‐PET imaging is not specific for FNH; however, it may be helpful to differentiate FNH from liver metastases in cancer patients if radiological methods are not diagnostic.     

Interstitial Capillary in Normal and in Transplanted Kidneys: An Ultrastructural Study

Knowledge about the normal structure and pathology of interstitial capillary is limited. Splitting and multilayering of the basal membrane (BM), as a marker of chronic rejection, has been published in association with transplant glomerulopathy. The authors investigated the ultrastructural features of the interstitial capillary basal membrane in normal (15 biopsies) and in transplanted kidneys (27 biopsies from 21 patients), expressing transplant glomerulopathy (8 biopsies from 6 patients), acute tubulo-interstitial rejection (9 biopsies from 6 patients), and recurrent or de novo glomerulonephritis (10 biopsies from 8 patients). All biopsies were fixed in 1%OsO 4, embedded in Epon, and examined by electron microscope. Measurements of the interstitial capillary BM were made. The BM of interstitial capillary of intact kidney was a homogenous continuous structure, 88 nm in width on average. Thickening with diffuse multilayering of BM was most intensive in patients with transplant glomerulopathy, and much less intensive in patients with acute tubulointerstitial rejection and in patients with recurrent or de novo glomerulonephritis. These findings may provide the first information about the morphology of the normal basal lamina of interstitial capillary and support the diagnostic value of interstitial capillary changes in chronic rejection.     

On hemispheric specialisation and visual field effects in the perception of print: A comment on Jordan,Patching, and Thomas

Introduction Due to structural characteristics of the visual pathways, stimuli that are presented in the right half of the visual field (RVF) are initially projected to the left cerebral hemisphere, while those presented in the left half of the visual field (LVF) are projected to the right cerebral hemisphere. This anatomical feature has frequently been taken to support the notion that the well-documented RVF advantage in recognising printed words is a reflection of functional differences between the two hemispheres; notably that of the dominance of the left hemisphere for processing language. Word stimuli that are sent straight to the left hemisphere are believed to profit from more efficient processing than those sent initially to the right hemisphere, because the latter stimuli must follow a longer and more noisy pathway before reaching the language-dominant hemisphere. In the work by Jordan, Patching, and Thomas (2003) the above notion is further developed to speculate that the point of entry of visual information into the cortex may determine the procedure that will underlie the ensuing word recognition process: "... the left hemisphere can process words by mapping orthographic information in parallel onto lexical entries whereas the right hemisphere has a more rudimentary process, that can only map orthographic information sequentially" (p. 50).