Biomarker-guided personalised emergency medicine for all - hope for another hype?

Polymorbid patients, diverse diagnostic and therapeutic options, more complex hospital structures, financial incentives, benchmarking, as well as perceptional and societal changes put pressure on medical doctors, specifically if medical errors surface. This is particularly true for the emergency department setting, where patients face delayed or erroneous initial diagnostic or therapeutic measures and costly hospital stays due to sub-optimal triage. A "biomarker" is any laboratory tool with the potential better to detect and characterise diseases, to simplify complex clinical algorithms and to improve clinical problem solving in routine care. They must be embedded in clinical algorithms to complement and not replace basic medical skills. Unselected ordering of laboratory tests and shortcomings in test performance and interpretation contribute to diagnostic errors. Test results may be ambiguous with false positive or false negative results and generate unnecessary harm and costs. Laboratory tests should only be ordered, if results have clinical consequences. In studies, we must move beyond the observational reporting and meta-analysing of diagnostic accuracies for biomarkers. Instead, specific cut-off ranges should be proposed and intervention studies conducted to prove outcome relevant impacts on patient care. The focus of this review is to exemplify the appropriate use of selected laboratory tests in the emergency setting for which randomised-controlled intervention studies have proven clinical benefit. Herein, we focus on initial patient triage and allocation of treatment opportunities in patients with cardiorespiratory diseases in the emergency department. The following five biomarkers will be discussed: proadrenomedullin for prognostic triage assessment and site-of-care decisions, cardiac troponin for acute myocardial infarction, natriuretic peptides for acute heart failure, D-dimers for venous thromboembolism, C-reactive protein as a marker of inflammation, and procalcitonin for antibiotic stewardship in infections of the respiratory tract and sepsis. For these markers we provide an overview on physiopathology, historical evolution of evidence, strengths and limitations for a rational implementation into clinical algorithms. We critically discuss results from key intervention trials that led to their use in clinical routine and potential future indications. The rational for the use of all these biomarkers, first, tackle diagnostic ambiguity and consecutive defensive medicine, second, delayed and sub-optimal therapeutic decisions, and third, prognostic uncertainty with misguided triage and site-of-care decisions all contributing to the waste of our limited health care resources. A multifaceted approach for a more targeted management of medical patients from emergency admission to discharge including biomarkers, will translate into better resource use, shorter length of hospital stay, reduced overall costs, improved patients satisfaction and outcomes in terms of mortality and re-hospitalisation. Hopefully, the concepts outlined in this review will help the reader to improve their diagnostic skills and become more parsimonious laboratory test requesters.

Assessing the improved discriminatory power of a new biomarker in prognostic models

Although the area under the receiver operating characteristic (AUC) is the most popular measure of the performance of prediction models, it has limitations, especially when it is used to evaluate the added discrimination of a new biomarker in the model. Pencina et al. (2008) proposed two indices, the net reclassification improvement (NRI) and integrated discrimination improvement (IDI), to supplement the improvement in the AUC (IAUC). Their NRI and IDI are based on binary outcomes in case-control settings, which do not involve time-to-event outcome. However, many disease outcomes are time-dependent and the onset time can be censored. Measuring discrimination potential of a prognostic marker without considering time to event can lead to biased estimates. In this dissertation, we have extended the NRI and IDI to survival analysis settings and derived the corresponding sample estimators and asymptotic tests. Simulation studies were conducted to compare the performance of the time-dependent NRI and IDI with Pencina’s NRI and IDI. For illustration, we have applied the proposed method to a breast cancer study.^ Key words: Prognostic model, Discrimination, Time-dependent NRI and IDI ^

Biomarker circadian rhythm profiles in critically ill mechanically ventilated patients

Objective: To explore the natural trajectory of core body temperature (CBT) and cortisol (CORT) circadian rhythms in mechanically ventilated intensive care unit (MV ICU) patients. ^ Design: Prospective, observational, time-series pilot study. ^ Setting: Medical-surgical and pulmonary ICUs in a tertiary care hospital. ^ Sample: Nine (F = 3, M = 6) adults who were mechanically ventilated within 12 hrs of ICU admission with mean ± SD age of 65.2 ± 14 years old. ^ Measurements: Core body temperature and environmental measures of light, sound, temperature, and relative humidity were logged in 1-min intervals. Hourly urine specimens and 2-hr interval blood specimens were collected for up to 7 consecutive days for CORT assay. Mechanical ventilation days, ICU length of stay, and ICU mortality were documented. Acute Physiology and Chronic Health Evaluation (APACHE) II scores were computed for each study day. The data of each biologic and environmental variable were analyzed using single cosinor analysis of 24-hr serial segments. One patient did not complete the study because mortality occurred within 8 hrs of enrollment. Nine ICU patients completed the study in 1.6 to 7.0 days. ^ Results: No normal circadian rhythm pattern was found when the cosinor-derived parameters of amplitude (one-half the peak-trough variability) and acrophase (peak time) were compared with cosinor-derived parameter reference ranges of healthy, diurnally active humans, although 83% of patient-day CBT segments showed statistically significant (p ≤ .05) and biologically meaningful (R2≥ 0.30) 24-hr rhythms with abnormal cosinor parameters. Cosinor parameters of the environmental temporal profiles showed 27% of light, 76% of ambient temperature, and 78% of relative humidity serial segments had a significant and meaningful 24-hr diurnal pattern. Average daily light intensity varied from 34 to 187 lx with a maximum light exposure of 1877 lx. No sound measurement segment had a statistically significant cosine pattern, and numerous 1-minute interval peaks ≥ 60 dB occurred around the clock. Average daily ambient temperature and relative humidity varied from 19 to 24°C and from 25% to 61%, respectively. There was no statistically significant association between CBT or clinical outcomes and cosinor-derived parameters of the environmental variables. Circadian rhythms of urine and plasma CORT were deferred for later analysis. ^ Conclusions: The natural trajectory of the CBT circadian rhythm in MV ICU patients demonstrated persistent cosinor parameter alteration, even when a significant and meaningful 24-hr rhythm was present. The ICU environmental measures showed erratic light and sound exposures. Room temperature and relative humidity data produced the highest rate of significant and meaningful diurnal 24-hr patterns. Additional research is needed to clarify relations among the CBT biomarker of the circadian clock and environmental variables of MV ICU patients. ^

INTEGRATIVE BIOMARKER IDENTIFICATION AND CLASSIFICATION USING HIGH THROUGHPUT ASSAYS

It is well accepted that tumorigenesis is a multi-step procedure involving aberrant functioning of genes regulating cell proliferation, differentiation, apoptosis, genome stability, angiogenesis and motility. To obtain a full understanding of tumorigenesis, it is necessary to collect information on all aspects of cell activity. Recent advances in high throughput technologies allow biologists to generate massive amounts of data, more than might have been imagined decades ago. These advances have made it possible to launch comprehensive projects such as (TCGA) and (ICGC) which systematically characterize the molecular fingerprints of cancer cells using gene expression, methylation, copy number, microRNA and SNP microarrays as well as next generation sequencing assays interrogating somatic mutation, insertion, deletion, translocation and structural rearrangements. Given the massive amount of data, a major challenge is to integrate information from multiple sources and formulate testable hypotheses. This thesis focuses on developing methodologies for integrative analyses of genomic assays profiled on the same set of samples. We have developed several novel methods for integrative biomarker identification and cancer classification. We introduce a regression-based approach to identify biomarkers predictive to therapy response or survival by integrating multiple assays including gene expression, methylation and copy number data through penalized regression. To identify key cancer-specific genes accounting for multiple mechanisms of regulation, we have developed the integIRTy software that provides robust and reliable inferences about gene alteration by automatically adjusting for sample heterogeneity as well as technical artifacts using Item Response Theory. To cope with the increasing need for accurate cancer diagnosis and individualized therapy, we have developed a robust and powerful algorithm called SIBER to systematically identify bimodally expressed genes using next generation RNAseq data. We have shown that prediction models built from these bimodal genes have the same accuracy as models built from all genes. Further, prediction models with dichotomized gene expression measurements based on their bimodal shapes still perform well. The effectiveness of outcome prediction using discretized signals paves the road for more accurate and interpretable cancer classification by integrating signals from multiple sources.

A relationship between aerosol transport and compound-specific d13C land plant biomarker and pollen records

We examined near-surface, late Holocene deep-sea sediments at nine sites on a north-south transect from the Congo Fan (4°S) to the Cape Basin (30°S) along the Southwest African continental margin. Contents, distribution patterns and molecular stable carbon isotope signatures of long-chain n-alkanes (C27-C33) and n-alkanols (C22-C32) are indicators of land plant vegetation of different biosynthetic types, which can be correlated with concentrations and distributions of pollen taxa in the same sediments. Calculated clusters of wind trajectories and satellite Aerosol Index imagery afford information on the source areas for the lipids and pollen on land and their transport pathways to the ocean sites. This multidisciplinary approach on an almost continental scale provides clear evidence of latitudinal differences in lipid and pollen composition paralleling the major phytogeographic zonations on the adjacent continent. Dust and smoke aerosols are mainly derived from the western and central South African hinterland dominated by deserts, semi-deserts and savannah regions rich in C4 and CAM plants. The northern sites (Congo Fan area and northern Angola Basin), which get most of their terrestrial material from the Congo Basin and the Angolan highlands, may also receive some material from the Chad region. Very little aerosol from the African continent is transported to the most southerly sites in the Cape Basin. As can be expected from the present position of the phytogeographic zones, the carbon isotopic signatures of the n-alkanes and n-alkanols both become isotopically more enriched in 13C from north to south. The results of the study suggest that this combination of pollen data and compound-specific isotope geochemical proxies can be effectively applied in the reconstruction of past continental phytogeographic developments.

Biomarker proxies on sediment cores in the central Arctic Ocean

Although the permanently to seasonally ice-covered Arctic Ocean is a unique and sensitive component in the Earth's climate system, the knowledge of its long-term climate history remains very limited due to the restricted number of pre-Quaternary sedimentary records. During Polarstern Expedition PS87/2014, we discovered multiple submarine landslides over a distance of >350 km along Lomonosov Ridge. Removal of younger sediments from steep headwalls has led to exhumation of Miocene to early Quaternary sediments close to the seafloor, allowing the retrieval of such old sediments with gravity cores. Multi-proxy biomarker analyses of these gravity cores reveal for the first time that the late Miocene central Arctic Ocean was relatively warm (4-7°C) and ice-free during summer, whereas sea ice occurred during spring and autumn/winter. A comparison of our proxy data with Miocene climate simulations seems to favour relatively high late Miocene atmospheric CO2 concentrations. These new findings from the Arctic region provide new benchmarks for groundtruthing global climate reconstructions and modeling.

Sea surface temperature reconstruction from biomarker in sediments off Portugal

High resolution reconstructions of sea surface temperature (Uk'37-SST), coccolithophore associations and continental input (total organic carbon, higher plant n-alkanes, n-alkan-1-ols) in core D13882 from the shallow Tagus mud patch are compared to SST records from deep-sea core MD03-2699 and other western Iberian Margin cores. Results reveal millennial-scale climate variability over the last deglaciation, in particular during the LGIT. In the Iberian margin, Heinrich event 1 (H1) and the Younger Dryas (YD) represent two extreme episodes of cold sea surface condition separated by a marine warm phase that coincides with the Bølling-Allerød interval (B-A) on the neighboring continent. Following the YD event, an abrupt sea surface warming marks the beginning of the Holocene in this region. SSTs recorded in core D13882 changed, however, faster than those at deep-sea site MD03-2699 and at the other available palaeoclimate sequences from the region. While the SST values from most deep-sea cores reflect the latitudinal gradient detected in the Iberian Peninsula atmospheric temperature proxies during H1 and the B-A, the Tagus mud patch (core D13882) experienced colder SSTs during both events. This is most certainly related to a supplementary input of cold freshwater from the continent to the Tagus mud patch, a hypothesis supported by the high contents of terrigenous biomarkers and total organic carbon as well as by the dominance of tetra-unsaturated alkenone (C37:4) observed at this site. The comparison of all western Iberia SST records suggests that the SST increase that characterizes the B-A event in this region started 1000 yr before meltwater pulse 1A (mwp-1A) and reached its maximum values during or slightly after this episode of substantial sea-level rise. In contrast, during the YD/ Holocene transition, the sharp SST rise in the Tagus mud patch is synchronous with meltwater pulse IB. The decrease of continental input to the mud patch conflrms a sea level rise in the region. Thus, the synchronism between the maximum warming in the mid-latitudes off the western Iberian margin, the adjacent landmasses and Greenland indicates that mwp-lB and the associated sea-level rise probably initiated in the Northern Hemisphere rather than in the South.

Specific microbial biomarker signatures from Congo fan sediments of ODP site 175-1075, Congo soils, wetland and estuary sediments

Methane (CH4) is a strong greenhouse gas known to have perturbed global climate in the past, especially when released in large quantities over short time periods from continental or marine sources. It is therefore crucial to understand and, if possible, quantify the individual and combined response of these variable methane sources to natural climate variability. However, past changes in the stability of greenhouse gas reservoirs remain uncertain and poorly constrained by geological evidence. Here, we present a record from the Congo fan of a highly specific bacteriohopanepolyol (BHP) biomarker for aerobic methane oxidation (AMO), 35-aminobacteriohopane-30,31,32,33,34-pentol (aminopentol), that identifies discrete periods of increased AMO as far back as 1.2 Ma. Fluctuations in the concentration of aminopentol, and other 35-aminoBHPs, follow a pattern that correlates with late Quaternary glacial-interglacial climate cycles, with highest concentrations during warm periods. We discuss possible sources of aminopentol, and the methane consumed by the precursor methanotrophs, within the context of the Congo River setting, including supply of methane oxidation markers from terrestrial watersheds and/or marine sources (gas hydrate and/or deep subsurface gas reservoir). Compound-specific carbon isotope values of -30 per mil to -40 per mil for BHPs in ODP 1075 and strong similarities between the BHP signature of the core and surface sediments from the Congo estuary and floodplain wetlands from the interior of the Congo River Basin, support a methanotrophic and likely terrigenous origin of the 35-aminoBHPs found in the fan sediments. This new evidence supports a causal connection between marine sediment BHP records of tropical deep sea fans and wetland settings in the feeding river catchments, and thus tropical continental hydrology. Further research is needed to better constrain the different sources and pathways of methane emission. However, this study identifies the large potential of aminoBHPs, in particular aminopentol, to trace and, once better calibrated and understood, quantify past methane sources and fluxes from terrestrial and potentially also marine sources.