Machine intelligence identifies soluble TNFa as a therapeutic target for spinal cord injury.

Traumatic spinal cord injury (SCI) produces a complex syndrome that is expressed across multiple endpoints ranging from molecular and cellular changes to functional behavioral deficits. Effective therapeutic strategies for CNS injury are therefore likely to manifest multi-factorial effects across a broad range of biological and functional outcome measures. Thus, multivariate analytic approaches are needed to capture the linkage between biological and neurobehavioral outcomes. Injury-induced neuroinflammation (NI) presents a particularly challenging therapeutic target, since NI is involved in both degeneration and repair. Here, we used big-data integration and large-scale analytics to examine a large dataset of preclinical efficacy tests combining five different blinded, fully counter-balanced treatment trials for different acute anti-inflammatory treatments for cervical spinal cord injury in rats. Multi-dimensional discovery, using topological data analysis (TDA) and principal components analysis (PCA) revealed that only one showed consistent multidimensional syndromic benefit: intrathecal application of recombinant soluble TNFα receptor 1 (sTNFR1), which showed an inverse-U dose response efficacy. Using the optimal acute dose, we showed that clinically-relevant 90 min delayed treatment profoundly affected multiple biological indices of NI in the first 48 h after injury, including reduction in pro-inflammatory cytokines and gene expression of a coherent complex of acute inflammatory mediators and receptors. Further, a 90 min delayed bolus dose of sTNFR1 reduced the expression of NI markers in the chronic perilesional spinal cord, and consistently improved neurological function over 6 weeks post SCI. These results provide validation of a novel strategy for precision preclinical drug discovery that is likely to improve translation in the difficult landscape of CNS trauma, and confirm the importance of TNFα signaling as a therapeutic target.

Interplay of strain and race/ethnicity in the innate immune response to M. tuberculosis.

BACKGROUND: The roles of host and pathogen factors in determining innate immune responses to M. tuberculosis are not fully understood. In this study, we examined host macrophage immune responses of 3 race/ethnic groups to 3 genetically and geographically diverse M. tuberculosis lineages. METHODS: Monocyte-derived macrophages from healthy Filipinos, Chinese and non-Hispanic White study participants (approximately 45 individuals/group) were challenged with M. tuberculosis whole cell lysates of clinical strains Beijing HN878 (lineage 2), Manila T31 (lineage 1), CDC1551 (lineage 4), the reference strain H37Rv (lineage 4), as well as with Toll-like receptor 2 agonist lipoteichoic acid (TLR2/LTA) and TLR4 agonist lipopolysaccharide (TLR4/LPS). Following overnight incubation, multiplex assays for nine cytokines: IL-1ß, IL-2, IL-6, IL-8, IL-10, IL-12p70, IFNγ, TNFα, and GM-CSF, were batch applied to supernatants. RESULTS: Filipino macrophages produced less IL-1, IL-6, and more IL-8, compared to macrophages from Chinese and Whites. Race/ethnicity had only subtle effects or no impact on the levels of IL-10, IL-12p70, TNFα and GM-CSF. In response to the Toll-like receptor 2 agonist lipoteichoic acid (TLR2/LTA), Filipino macrophages again had lower IL-1 and IL-6 responses and a higher IL-8 response, compared to Chinese and Whites. The TLR2/LTA-stimulated Filipino macrophages also produced lower amounts of IL-10, TNFα and GM-CSF. Race/ethnicity had no impact on IL-12p70 levels released in response to TLR2/LTA. The responses to TLR4 agonist lipopolysaccharide (TLR4/LPS) were similar to the TLR2/LTA responses, for IL-1, IL-6, IL-8, and IL-10. However, TLR4/LPS triggered the release of less IL-12p70 from Filipino macrophages, and less TNFα from White macrophages. CONCLUSIONS: Both host race/ethnicity and pathogen strain influence the innate immune response. Such variation may have implications for the development of new tools across TB therapeutics, immunodiagnostics and vaccines.

Biomarkers of Tuberculosis Severity and Treatment Effect: A Directed Screen of 70 Host Markers in a Randomized Clinical Trial.

More efficacious treatment regimens are needed for tuberculosis, however, drug development is impeded by a lack of reliable biomarkers of disease severity and of treatment effect. We conducted a directed screen of host biomarkers in participants enrolled in a tuberculosis clinical trial to address this need. Serum samples from 319 protocol-correct, culture-confirmed pulmonary tuberculosis patients treated under direct observation as part of an international, phase 2 trial were screened for 70 markers of infection, inflammation, and metabolism. Biomarker assays were specifically developed for this study and quantified using a novel, multiplexed electrochemiluminescence assay. We evaluated the association of biomarkers with baseline characteristics, as well as with detailed microbiologic data, using Bonferroni-adjusted, linear regression models. Across numerous analyses, seven proteins, SAA1, PCT, IL-1ß, IL-6, CRP, PTX-3 and MMP-8, showed recurring strong associations with markers of baseline disease severity, smear grade and cavitation; were strongly modulated by tuberculosis treatment; and had responses that were greater for patients who culture-converted at 8weeks. With treatment, all proteins decreased, except for osteocalcin, MCP-1 and MCP-4, which significantly increased. Several previously reported putative tuberculosis-associated biomarkers (HOMX1, neopterin, and cathelicidin) were not significantly associated with treatment response. In conclusion, across a geographically diverse and large population of tuberculosis patients enrolled in a clinical trial, several previously reported putative biomarkers were not significantly associated with treatment response, however, seven proteins had recurring strong associations with baseline radiographic and microbiologic measures of disease severity, as well as with early treatment response, deserving additional study.

Molecular characterization of genomic AML1-ETO fusions in childhood leukemia.

T(8;21) AML1(CBFA2)-ETO(MTG8) is the most common chromosomal translocation in acute myeloid leukemia (AML) in both children and adults. We sought to understand the structure and gain insight into the fusion process between AML1 and ETO by sequencing genomic fusions in 17 primary childhood AMLs and two cell lines with t(8;21). Reciprocal translocations were sequenced for seven of the 19 samples. We assumed a null hypothesis that the translocation breakpoints would be evenly distributed along the intronic breakpoint cluster regions. Testing for multimodality via smoothed bootstrap statistical methods suggested, however, the presence of two separate cluster regions within both the AML1 and ETO breakpoint cluster regions. ETObreakpoints were predominantly located in intron 1B in a defined cluster 5' of exon 1A (scan statistic P value = 0.00001). All patients with available RNA expressed an AML1-ETO mRNA fusion between exon 5 of AML1 and exon 2 of ETO. Since the structural restraints for the fusion protein of AML1-ETO exclude exon 1A, we reason that ETO intron 1B harbors a structural feature with propensity for breakage and/or recombination. Chromosomal breakpoints displayed evidence of fusion by a non-homologous end joining process, with microhomologies and nontemplate nucleotides at some fusion junctions. Breakpoints in general displayed similar complexity of duplications, deletions, and insertions to other common pediatric leukemia translocations (TEL-AML1, MLL-AF4, PML-RARA, CBFB-MYH11) that we and others have analyzed.

Serum biomarkers of treatment response within a randomized clinical trial for pulmonary tuberculosis.

RATIONALE: Biomarkers for monitoring response to anti-tuberculosis treatment are needed. We explored immune markers previously published as having predictive capability for 8 week culture status in 39 adults enrolled in a clinical trial in Kampala, Uganda. METHODS: We consecutively selected 20 HIV-negative pulmonary TB subjects with positive cultures, and 19 subjects with negative cultures at the end of intensive phase therapy. At baseline and after 8 weeks, serum was assayed for nine cytokines and soluble cytokine receptors using multiplexed platforms or ELISA. We evaluated their association with week 8 culture status first using single-variable logistic models, then using cross-validated estimates of the C-statistic, a measure of discrimination, of candidate models including 2 or 3 analytes in addition to age. RESULTS: All but one analyte decreased from baseline to week 8 (all p < 0.01). Individual biomarkers were not associated with 8 week culture status. Logistic models including increasing age, higher baseline soluble tumor necrosis factor receptor alpha 1 (sTNF-R1), and higher week 8 C-reactive protein (CRP) concentration classified subjects by culture status with up to 85% accuracy and acceptable discrimination (cross-validated C-statistic 0.76) and calibration (Hosmer-Lemeshow P > 0.2). CONCLUSION: Exploratory post-hoc models including sTNF-R1, CRP, and age, classified 8 week culture status with promising accuracy.

Patterns of weight change and their relation to diet in a cohort of healthy women.

Patterns of weight change for 31,940 non-smoking women aged 30-55 y in 1976 were examined for 8 y of follow-up. Each woman reported her weight every 2 y on questionnaires, and diet was assessed in 1980 with a semiquantitative food frequency questionnaire. Self-reported weight was highly correlated over time, decreasing from r = 0.95 over 2 y to r = 0.89 over 8 y. Weight gain was inversely related to age (r = -0.06). Weight change in a given 2-y interval was inversely related to change in weight over the subsequent 2 y (r = -0.30). Age, relative weight, and prior weight change were stronger predictors of recent weight change than were intake patterns of specific nutrients. Recent prior weight change was the strongest predictor of subsequent weight change. In a population of free-living women, prior weight loss and younger age are far stronger predictors of subsequent weight gain than are the qualitative aspects of diet.

Body fat distribution, blood pressure, and hypertension. A prospective cohort study of men in the normative aging study.

The relation between the abdominal accumulation of body fat, blood pressure, and hypertension was assessed prospectively among 1972 male participants in the Veterans Administration Normative Aging Study. Body mass index (BMI = weight [kg]/height [m]2) and the ratio of abdominal circumference to hip breadth (AC/HB), measured at regular exams, were used as indices of total adiposity and body fat distribution, respectively. Considering blood pressure as a continuous outcome variable (in models that allowed for intraclass correlation), the AC/HB ratio was significantly positively associated with both diastolic and systolic blood pressure (P < 0.001), adjusting for age and BMI. Blood pressure was dichotomized and hypertension risk was assessed using the proportional hazards model, adjusting for age and BMI. Seven hundred cases of hypertension were recorded by study physicians during 35,496 person-years of follow-up. The risk of hypertension increased approximately three-fold (95% confidence interval, 1.7 to 5.2) with a change of one unit in the AC/HB ratio. These estimates were little changed when the effects of smoking and alcohol intake were considered. Thus, the abdominal accumulation of body fat, apart from overall level of adiposity, was associated with both increased blood pressure and an increased risk of hypertension.

Temporal trends in chronic obstructive lung disease case fatality in hospitalized US veterans: 1970-1987.

Vital statistics data have suggested that age-adjusted mortality from chronic obstructive lung diseases (CLD) is increasing. The present investigation has used the US Veterans Administration (VA) hospital computer database to determine whether trends in CLD case fatality follow trends in CLD population mortality. Data for male patients discharged from 172 VA hospitals from 1970 through 1987 were utilized. Patients were included if they had a CLD as a first-listed hospital discharge diagnosis, did not have any of a number of smoking-related cardiovascular or malignant diseases, and were born in the years 1900 through 1939. While crude case fatality for all CLD increased from 5.2 percent in 1970 to 7.4 percent in 1987, age-adjusted case fatality decreased from a peak of 8.5 percent in 1971 to 5.8 percent (95 percent confidence interval, 5.4 to 6.61 percent) in 1987. This trend was seen for all CLD diagnoses, including asthma. Age-specific case fatality decreased for each successive 5-year birth cohort, and age at death remained relatively constant over the last 14 years of the study. The declining case fatality could not be explained by changes in the ICD coding rubrics. The extent to which case fatality has been declining due to improved treatment and/or cohort-related changes in tobacco smoke exposure could not be determined definitively from the data.

Subcortical ischemic vascular dementia: assessment with quantitative MR imaging and 1H MR spectroscopy.

BACKGROUND AND PURPOSE: Subcortical ischemic vascular dementia is associated with cortical hypometabolism and hypoperfusion, and this reduced cortical metabolism or blood flow can be detected with functional imaging such as positron emission tomography. The aim of this study was to characterize, by means of MR imaging and 1H MR spectroscopy, the structural and metabolic brain changes that occur among patients with subcortical ischemic vascular dementia compared with those of elderly control volunteers and patients with Alzheimer's disease. METHODS: Patients with dementia and lacunes (n = 11), cognitive impairment and lacunes (n = 14), and dementia without lacunes (n = 18) and healthy age-matched control volunteers (n = 20) underwent MR imaging and 1H MR spectroscopy. 1H MR spectroscopy data were coanalyzed with coregistered segmented MR images to account for atrophy and tissue composition. RESULTS: Compared with healthy control volunteers, patients with dementia and lacunes had 11.74% lower N-acetylaspartate/creatine ratios (NAA/Cr) (P = .007) and 10.25% lower N-acetylaspartate measurements (NAA) in the cerebral cortex (P = .03). In white matter, patients with dementia and lacunes showed a 10.56% NAA/Cr reduction (P = .01) and a 12.64% NAA reduction (P = .04) compared with control subjects. NAA in the frontal cortex was negatively correlated with the volume of white matter signal hyperintensity among patients with cognitive impairment and lacunes (P = .002). Patients with dementia, but not patients with dementia and lacunes, showed a 10.33% NAA/Cr decrease (P = .02) in the hippocampus compared with healthy control volunteers. CONCLUSION: Patients with dementia and lacunes have reduced NAA and NAA/Cr in both cortical and white matter regions. Cortical changes may result from cortical ischemia/infarction, retrograde or trans-synaptic injury (or both) secondary to subcortical neuronal loss, or concurrent Alzheimer's pathologic abnormalities. Cortical derangement may contribute to dementia among patients with subcortical infarction.

The relationship of nasal disorders to lower respiratory tract symptoms and illness in a random sample of children.

We examined the relationship of nasal disorders, defined as frequent colds and sinus trouble, to lower respiratory tract symptoms in a random population of 718 children aged 4 to 11 years in East Boston, Massachusetts. Frequent colds were significantly associated with maternal smoking (odds ratio (OR) = 3.00; 95% confidence interval (CI) = 1.97, 4.58), and so was sinus trouble (OR = 4.73; 95% CI = 1.78, 12.51). After adjustment for maternal smoking, age and sex, frequent colds (OR = 2.88; 95% CI = 1.87, 4.42) and sinus trouble (OR = 4.95, 95% CI = 1.83, 13.39) remained significant predictors of lower respiratory tract symptoms in separate logistic regressions. If one restricted the cohort to the 513 children who also had personal smoking information and adjusted for this variable as well, the results for colds were unchanged (OR = 2.94; 95% CI = 1.78, 4.84) but the results for sinus trouble were now not statistically significant (OR = 2.30, 95% CI = 0.67, 7.94). We conclude that nasal disorders are associated with lower respiratory tract symptoms in children.

Extending the elements of tree-structured regression.

The usage of tree-structured or recursive partitioning methods has grown steadily in the decade since the appearance of the definitive monograph 'Classification and Regression Trees'. Accompanying this growth have been many methodologic and software extensions that have served to give the tree-structured approach even wider applicability. This overview highlights some of these developments, emphasizing the regression setting. An illustrative example describes how tree-structured regression, modified to handle right-censored, left-truncated survival data with time-dependent covariates, can be used to assess whether rates of progression from HIV to AIDS have changed over time.

Magnetic resonance imaging measurement of relaxation and water diffusion in the human lumbar intervertebral disc under compression in vitro.

STUDY DESIGN: Twelve lumbar intervertebral disc specimens were imaged with magnetic resonance imaging to estimate relaxation constants, T1 and T2, and tissue water diffusion, before and after applying compression. OBJECTIVES: The objectives of the study were to measure T1, T2, and water diffusion for differences with loading state, region of the disc (anulus fibrosus or nucleus pulposus), and grade of degeneration. SUMMARY OF BACKGROUND DATA: Magnetic resonance imaging can be used qualitatively to estimate water content and degeneration of the intervertebral disc. Beyond structural information of images, the relaxation times T1 and T2 may contain information on the changes occurring with degeneration. A modified spin-echo sequence can be used to estimate tissue water diffusion in cartilage and disc specimens with the ability to measure anisotropy. METHODS: Specimens were imaged in a 1.5-Tesla clinical scanner. T1, T2, and water diffusion were estimated from midsagittal images. Magnetic resonance imaging parameters were calculated before and after axial loading. The measured T1, T2, and D (diffusion coefficient) were compared before and after compression, and for the diffusion data, also by direction to consider anisotropy. RESULTS: For the T1 data, a significant difference was found by region, nucleus > anulus, and loading state, loaded > unloaded. For the T2 values, there was a significant difference by region, nucleus > anulus, and Thompson grade. For diffusion, significant differences were found by region, nucleus > anulus, Thompson grade, direction of diffusion, and state of compression, loaded > unloaded. CONCLUSIONS: This study demonstrated that magnetic resonance imaging can be used to measure significant changes in T1, T2, or diffusion in intervertebral disc specimens by region, loading condition, or Thompson grade.

Distinguishing recurrent intra-axial metastatic tumor from radiation necrosis following gamma knife radiosurgery using dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging.

BACKGROUND AND PURPOSE: MR image-guided gamma knife radiosurgery is often used to treat intra-axial metastatic neoplasms. Following treatment, it is often difficult to determine whether a progressively enhancing lesion is due to metastatic tumor recurrence or radiation necrosis. The purpose of our study was to determine whether relative cerebral blood volume (rCBV), relative peak height (rPH), and percentage of signal-intensity recovery (PSR) derived from dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging can distinguish recurrent metastatic tumor from radiation necrosis. MATERIALS AND METHODS: Twenty-seven patients with systemic cancer underwent gamma knife radiosurgery for metastatic lesions of the brain and subsequently developed enlarging regions of enhancement within the radiation field. Subsequent surgical resection or clinicoradiologic follow-up established a diagnosis of recurrent metastatic tumor or radiation necrosis. Perfusion MR imaging datasets were retrospectively reprocessed, and regions of interest were drawn around the entire contrast-enhancing region. The resulting T2* signal-intensity time curves produced rCBV, rPH, and PSR values for each examination. A Welch t test was used to compare imaging values between groups. RESULTS: The mean, minimum, and maximum PSR values were significantly lower (P < .01) in cases of recurrent metastatic tumor. The mean and maximum rCBV and rPH values were significantly higher (P < .02) in the recurrent metastatic tumor group. CONCLUSIONS: The findings of our study suggest that perfusion MR imaging may be used to differentiate recurrent intra-axial metastatic tumor from gamma knife-induced radiation necrosis.

Robust inference for multivariate survival data.

Multivariate survival data arise when an individual records multiple survival events or when individuals recording single survival events are grouped into clusters. In this paper we propose a new method for the analysis of multivariate survival data. The technique is a synthesis of the Poisson regression formulation for univariate censored survival analysis and the generalized estimating equation approach for obtaining valid variance estimates for generalized linear models in the presence of clustering. When the survival data are clustered, combining the methods provides not only valid estimates for the variances of regression parameters but also estimates of the dependence between survival times. The approach entails specifying parametric models for the marginal hazards and a dependence structure, but does not require specification of the joint multivariate survival distribution. Properties of the methodology are investigated by simulation and through an illustrative example.

Trees and tracking.

Epidemiologists have used the term 'tracking' to connote an individual's maintenance of relative rank of some longitudinally measured characteristic over a given time span. To assess the extent to which an attribute tracks we have first to summarize individual growth curves, and second to quantify the notion of maintenance of relative rank, both in the face of random error. A sequence of papers appearing in 1981 provided differing methodologies for appraising tracking. Here we take a different approach to tracking by using regression trees for longitudinal data. The above two concerns are simultaneously addressed in that the procedure identifies subgroups, defined in terms of covariates, within which the collection of growth curves is homogeneous. After reviewing the existing approaches to tracking we describe the tree-structured methodology, and present an illustrative example pertaining to lung function growth in children.

Design effects for binary regression models fitted to dependent data.

Dependent data, such as arise with cluster sampling, typically yield variances of parameter estimates which are larger than would be provided by a simple random sample of the same size. This variance inflation factor is called the design effect of the estimator. Design effects have been derived for cluster sampling designs using simple estimators such as means and proportions, and also for linear regression coefficient estimators. In this paper, we show that a method to derive design effects for linear regression estimators extends to generalized linear models for binary responses. In particular, some simple expressions for design effects in the linear regression model provide accurate approximations for binary regression models such as those based on the logistic, probit and complementary log-log links. We corroborate our findings with two examples and some simulation studies.

Survival trees with time-dependent covariates: application to estimating changes in the incubation period of AIDS.

Tree-structured methods for exploratory data analysis have previously been extended to right-censored survival data. We further extend these methods to allow for truncation and time-dependent covariates. We apply the new methods to a data set on incubation times of acquired immunodeficiency syndrome (AIDS), using calendar time as a time-dependent covariate. Contrary to expectation, we find that rates of progression to AIDS appear to be faster after August 1989 than before.

On a method of mortality analysis incorporating age--year interaction, with application to prostate cancer mortality.

A model used in the analysis of tabular mortality data (or incidence data), in which deaths are classified by age and year of occurrence, is shown to arise from the assumption that effects due to epoch of birth (the cohort effects) act multiplicatively on an underlying hazard function which may change with calendar year. The model disentangles the cohort effects and the effects due to epoch of death (year effects) and age at death (age effects), and incorporates an age--year interaction which has an interpretation in terms of changing the shape of the underlying hazard function. In addition, the derivation of the hazard function suggests meaningful interpretation of certain combinations of the model parameters. As an illustration, application of the model to prostate cancer mortality in England and Wales indicates, among other things, that after allowing the changing cohort effects, the risk for young men has increased relative to that for older men in recent years.

A comparison of estimated proportional hazards models and regression trees.

We present examples of the usage of regression trees for censored response via two real world datasets, one a rheumatoid arthritis survival study and the other a hip replacement study, and draw comparisons with the results of Cox proportional hazards modelling. The two methods pursue different goals. Motivation of the tree techniques is the desire to extract meaningful prognostic groups while the proportional hazards model enables assessment of the impact of risk factors. The methods are thus complementary. For the arthritis study the two techniques corroborate one another, although the flavour of the conclusions derived differ. For the hip replacement study, however, the regression tree approach reveals structure that would not emerge from a routine proportional hazards analysis. We also discuss the treatment of data analytic issues such as the handling of missing values and influence in the presence of non-uniform censoring.

An empirical comparison of multivariable methods for estimating risk of death from coronary heart disease.

BACKGROUND: Logistic regression and, more recently, Cox regression have been the predominant methods for identifying risk factors and developing risk estimation equations for coronary heart disease (CHD). Software for the regression tree method is now available for binary and survival outcomes and thus offers an alternative methodology. This paper compares these four methods for identifying significant risk factors from among a set of candidate factors and for estimating the risk of death from CHD using baseline and mortality follow-up data on 1,701 men participating in the Busselton Health Study. The candidate risk factors were age, body mass index, systolic and diastolic blood pressure, treatment for hypertension, cholesterol and smoking. METHODS: Logistic regression, Cox proportional hazards regression, binary regression tree, and survival regression tree analyses have been applied to data obtained from the same cohort of men for CHD death risk estimation and prediction. The four methods are compared in terms of the variables selected, goodness-of-fit of models, similarity of cross-validated estimated risks for individuals, and ability to discriminate between those who died from CHD and those who did not die from CHD during the follow-up period, including the comparison of Receiver Operating Characteristic (ROC) curves. RESULTS: Although age and a blood pressure variable were selected by all four methods, body mass index was also selected by the regression tree methods and smoking was also selected by Cox regression. There was good, but not excellent, agreement between methods in estimates of risk for individuals, the areas under the ROC curves were 0.66 for the binary tree, 0.72 for logistic regression, 0.71 for the survival tree method and 0.78 for Cox regression. The average differences in estimated risk between those who died from CHD and those who did not die from CHD during the follow-up period were 0.051 for logistic regression, 0.070 for the binary tree method, 0.073 for the survival tree method and 0.088 for Cox regression. CONCLUSION: For a moderately sized cohort typical of many applications of these methods in the literature, the two methods which used the survival outcome performed better than the methods using a binary outcome. Despite selecting some different variables and showing moderate differences in risk estimates for individuals, the two binary approaches were similar in performance. Cox regression appeared to be superior to the survival tree method, but further larger studies of completely separate samples for model development and evaluation of prediction performance are required to confirm this finding.