Clinical considerations in study designs that use cotinine as a biomarker

Subjects enrolled in studies are not always screened for routine habits such as smoking. Personal history is not always reliable and therefore an objective biomarker is necessary to screen for smokers. The objectives of this article were to review the metabolism of nicotine and other metabolic considerations associated with smoking; to review some of the routine methods used to assess exposure to nicotine-containing products; to revisit cotinine breakpoints utilized to distinguish smokers from non-smokers during screening for clinical trials; to assess the utility of screening questions regarding smoking practices; and to recommend standards for clinical pharmacology studies. The results indicated that cotinine levels serve as a useful biomarker of tobacco exposure; racial issues may be clinically relevant in determining smoking status; cessation of smoking should occur at least 14 days prior to the start of the study; adverse effects from nicotine withdrawal such as craving, hunger and weight gain may persist for more than 6 months; potential metabolic interactions via cytochrome P2A6 and P1A2 need to be considered when designing a study; and the use of a single calibrator as a breakpoint is acceptable if a categorical outcome such as 'smoker' versus 'non-smoker' is desired. Nicotine from food products is not expected to impact assay sensitivity or to be clinically relevant; a serum cotinine concentration of 10 ng ml^-1 be employed as a breakpoint for non-smokers versus smokers; other non-invasive alternatives are collection of urine, saliva, or hair (with suggested breakpoints of 200 ng ml^-1, 5 ng ml^-1 and 0.3 ng mg^-1, respectively; screening questions be accompanied by testing for cotinine; and the inclusion of smokers in studies should be considered once the impact of smoking on the targeted population is understood.     

Clinical considerations in study designs that use cotinine as a biomarker

Subjects enrolled in studies are not always screened for routine habits such as smoking. Personal history is not always reliable and therefore an objective biomarker is necessary to screen for smokers. The objectives of this article were to review the metabolism of nicotine and other metabolic considerations associated with smoking; to review some of the routine methods used to assess exposure to nicotine-containing products; to revisit cotinine breakpoints utilized to distinguish smokers from non-smokers during screening for clinical trials; to assess the utility of screening questions regarding smoking practices; and to recommend standards for clinical pharmacology studies. The results indicated that cotinine levels serve as a useful biomarker of tobacco exposure; racial issues may be clinically relevant in determining smoking status; cessation of smoking should occur at least 14 days prior to the start of the study; adverse effects from nicotine withdrawal such as craving, hunger and weight gain may persist for more than 6 months; potential metabolic interactions via cytochrome P2A6 and P1A2 need to be considered when designing a study; and the use of a single calibrator as a breakpoint is acceptable if a categorical outcome such as 'smoker' versus 'non-smoker' is desired. Nicotine from food products is not expected to impact assay sensitivity or to be clinically relevant; a serum cotinine concentration of 10 ng ml^?1 be employed as a breakpoint for non-smokers versus smokers; other non-invasive alternatives are collection of urine, saliva, or hair (with suggested breakpoints of 200 ng ml^?1, 5 ng ml^?1 and 0.3 ng mg^?1, respectively; screening questions be accompanied by testing for cotinine; and the inclusion of smokers in studies should be considered once the impact of smoking on the targeted population is understood.     

Quantification of cotinine in dried blood spots as a biomarker of exposure to tobacco smoke

Objective: We present an ultra-sensitive, minimally-invasive method for quantifying cotinine in dried blood spot (DBS) samples as a biomarker of exposure to tobacco smoke that can be collected using a simple heel or finger prick to obtain blood samples.

Methods: Cotinine levels were measured in matched plasma and reconstituted DBS samples from smokers and nonsmokers to evaluate assay parameters. In addition, we applied this new method to finger-prick DBS samples that were collected from infants, children and young adults ages 1–21 to estimate exposure to tobacco smoke. Partitioning of cotinine across red blood cells and haematocrit effects were investigated.

Results: Cotinine levels measured in matched plasma and reconstituted DBS samples from smokers and nonsmokers were found to be highly correlated (R²=0.94), with 100% sensitivity and 94% specificity to differentiate reported smokers from nonsmokers. With this method, the LOQ is <0.25?ng/mL using a single 3.2?mm punch of a DBS, and haematocrit effects are negligible.

Conclusions: This sensitive, high-throughput and minimally-invasive method for quantifying cotinine in DBS samples provides a simple and cost effective means for estimating exposure to tobacco smoke in population based studies, and has particular advantages in studies involving infants and children.     

Adaptive enrichment designs with a continuous biomarker

A popular design for clinical trials assessing targeted therapies is the two-stage adaptive enrichment design with recruitment in stage 2 limited to a biomarker-defined subgroup chosen based on data from stage 1. The data-dependent selection leads to statistical challenges if data from both stages are used to draw inference on treatment effects in the selected subgroup. If subgroups considered are nested, as when defined by a continuous biomarker, treatment effect estimates in different subgroups follow the same distribution as estimates in a group-sequential trial. This result is used to obtain tests controlling the familywise type I error rate (FWER) for six simple subgroup selection rules, one of which also controls the FWER for any selection rule. Two approaches are proposed: one based on multivariate normal distributions suitable if the number of possible subgroups, k, is small, and one based on Brownian motion approximations suitable for large k. The methods, applicable in the wide range of settings with asymptotically normal test statistics, are illustrated using survival data from a breast cancer trial.     

The use of PIB-PET as a dual pathological and functional biomarker in AD

Amyloid imaging with positron emission tomography (PET) is presently used in Alzheimer's disease (AD) research. In this study we investigated the possibility to use early frames (ePIB) of the PIB scans as a rough index of CBF by comparing normalised early PIB values with cerebral glucose metabolism (rCMRglc). PIB-PET and FDG-PET were performed in 37 AD patients, 21 subjects with mild cognitive impairment (MCI) and 6 healthy controls (HC). The patients were divided based on their PIB retention (amyloid load) as either PIB positive (PIB+) or PIB negative (PIB−). Data of the unidirectional influx K₁ from a subset of the subjects including 7 AD patients and 3 HC was used for correlative analysis. Data was analysed using regions of interest (ROI) analysis. A strong, positive correlation was observed across brain regions between K₁ and ePIB (r = 0.70; p ≤ 0.001). The ePIB values were significantly lower in the posterior cingulate (p ≤ 0.001) and the parietal cortices (p = 0.002) in PIB+ subjects compared to PIB−, although the group difference were stronger for rCMRglc in cortical areas (p ≤ 0.001). Strong positive correlations between ePIB and rCMRglc were observed in all cortical regions analysed, especially in the posterior cingulate and parietal cortices (p ≤ 0.001). A single dynamic PIB-PET scan may provide information about pathological and functional changes (amyloidosis and impaired blood flow). This might be important for diagnosis of AD, enrichment of patients in clinical trials and evaluation of treatment effects. This article is part of a Special Issue entitled: Imaging Brain Aging and Neurodegenerative disease.     

Nitrogen isotopic fractionation as a biomarker for nitrogen use efficiency in ruminants: a meta-analysis

Animal proteins are naturally ¹⁵N enriched relative to the diet and the extent of this difference (Δ¹⁵N_animal-diet or N isotopic fractionation) has been correlated to N use efficiency (NUE; N gain or milk N yield/N intake) in some recent ruminant studies. The present study used meta-analysis to investigate whether Δ¹⁵N_animal-diet can be used as a predictor of NUE across a range of dietary conditions, particularly at the level of between-animal variation. An additional objective was to identify variables related to N partitioning explaining the link between NUE and Δ¹⁵N_animal-diet. Individual values from eight publications reporting both NUE and Δ¹⁵N_animal-diet for domestic ruminants were used to create a database comprising 11 experimental studies, 41 treatments and individual animal values for NUE (n=226) and Δ¹⁵N_animal-diet (n=291). Data were analyzed by mixed-effect regression analysis taking into account experimental factors as random effects on both the intercept and slope of the model. Diets were characterized according to the INRA feeding system in terms of N utilization at the rumen, digestive and metabolic levels. These variables were used in a partial least squares regression analysis to predict separately NUE and Δ¹⁵N_animal-diet variation, with the objective of identifying common variables linking NUE and Δ¹⁵N_animal-diet. For individuals reared under similar conditions (within-study) and at the same time (within-period), the variance of NUE and Δ¹⁵N_animal-diet not explained by dietary treatments (i.e. between-animal variation plus experimental error) was 35% and 55%, respectively. Mixed-effect regression analysis conducted with treatment means showed that Δ¹⁵N_animal-diet was significantly and negatively correlated to NUE variation across diets (NUE=0.415 −0.055×Δ¹⁵N_animal-diet). When using individual values and taking into account the random effects of study, period and diet, the relationship was also significant (NUE=0.358 −0.035×Δ¹⁵N_animal-diet). However, there may be a biased prediction for animals close to zero, or in negative, N balance. When using a novel statistical approach, attempting to regress between-animal variation in NUE on between-animal variation in Δ¹⁵N_animal-diet (without the influence of experimental factors), the negative relationship was still significant, highlighting the ability of Δ¹⁵N_animal-diet to capture individual variability. Among the studied variables related to N utilization, those concerning N efficiency use at the metabolic level contributed most to predict both Δ¹⁵N_animal-diet and NUE variation, with rumen fermentation and digestion contributing to a lesser extent. This study confirmed that on average Δ¹⁵N_animal-diet can predict NUE variation across diets and across individuals reared under similar conditions.     

Urinary desmosine as a biomarker in acute lung injury.

Acute lung injury (ALI) is a complex disorder associated with an acute inflammatory response thought to contribute to tissue injury. Desmosine, a cross-linking amino acid present in elastin, is released during matrix degradation and cleared by the kidney. Results from animal models and human disease studies have suggested that ALI is associated with the release of desmosine, resulting in increased urinary desmosine. A radioimmunoassay was used to monitor urinary desmosine levels over 10 days in ten patients with ALI. The concentration of desmosine was measured with and without acid hydrolysis. Baseline urinary desmosine was increased in two of ten patients. The concentration of desmosine at baseline did not appear to be related to age, gender, neutrophil elastase (NE)/alpha(1)-antiprotease complex concentration or P(a)O(2)/F(i)O(2) ratio. No meaningful changes in desmosine levels were noted after removal from mechanical ventilation. Baseline desmosine concentrations did not appear to correlate with the risk of death. The limited sensitivity, predictive correlations and dynamic modulation would suggest that urine desmosine has a limited role as a biomarker for ALI. Hydrolysis of urine samples appears necessary for optimal measurement of urine desmosine.     

Decline in serum antibodies to methyltetrahydrophthalic anhydride after cessation of exposure.Implications for use as a biomarker

T he association between exposure intensity and serum levels of immunoglobulins E and G against low molecular weight compounds was evaluated. The decay of levels of specific IgE and IgG antibodies was studied after cessation of exposure in workers exposed to the inhalant allergen methyltetrahydrophthalic anhydride in a plant using epoxy resins. Sera have been collected in workers for 18-84 (mean value 54) months after cessation of exposure. Specific IgE and IgG was assessed by RAST and ELISA, respectively. The mean of individual half-times for IgE ( N = 10) and IgG ( N = 8) was 0 9 (range 0 1-1 8) and 0 4 (range 0 2-0 6) years, respectively, after total avoidance of exposure. Corresponding decreases of IgE and IgG were also observed after reduction, but not total elimination, of exposure. No correlation was seen between biologic halftimes of specific IgE and total IgE, atopy, smoking habits or gender. The results indicate that the levels of specific antibodies in sensitized individuals reflect long term exposure, and may persist for years after the end of exposure.     

Clinical utility of decorin in follicular fluid as a biomarker of oocyte potential

This study investigated the concentration of decorin (DCN) in mature follicular fluid and the existence in the granulosa cells. It also investigated whether DCN is useful as a biomarker for outcomes of assisted reproductive technology (ART). A retrospective cohort study was performed involving 130 oocytes of 88 patients treated with ART because of unexplained infertility. The concentration of DCN in the follicular fluid (F-DCN) was 39.26 ng/ml (median value); it was higher than that in serum. F-DCN of the oocytes fertilized by intracytoplasmic sperm injection (ICSI) was significantly lower than that of oocytes that were not fertilized (33.24 ng/ml vs 40.18 ng/ml; P = 0.043). When a cut-off level of 34.5 ng/ml was set according to the receiver-operating characteristic curve, the fertilization rate of the oocytes from the follicles in which F-DCN was lower than the cut-off level tended to be good compared to that of the oocytes with F-DCN higher than the cut-off level (P = 0.052). DCN is less likely to be produced by the granulosa cells (GCs), because it was not detected in GCs by immunostaining and Western blot analysis. F-DCN has a possibility to be a biomarker indicating the quality of oocytes collected from the corresponding follicle.     

DNA-protein crosslinks as a biomarker of exposure to solar radiation: a preliminary study in brick-kiln workers.

In India, fired clay bricks are produced in small-scale factories. There are 60, 000 active brick kilns, providing employment to nearly 12 million people in different suboccupations. This industry is largely non-mechanized and operates from November to June. Almost all the workers are exposed to direct sunlight for 8-10 h a day. Cellular DNA-protein crosslinks (DPCs) are the biologically active nucleoprotein complexes formed between DNA and proteins. Ultraviolet light and gamma-rays, and other suspected carcinogens in humans, induce DPC formation in blood cells. DPCs have therefore been identified as a biomarker for monitoring exposure to these hazardous agents. Here we report steady-state levels of DPCs in human peripheral lymphocytes from 46 brick-kiln workers exposed occupationally for 8-10 h a day to solar radiation in brickfields and 25 unexposed controls. A significant increase (p <0.05) in DPC content and DPC coefficients in peripheral lymphocytes was observed in the brick-kiln workers compared with the controls. The data suggest that the DPC content of lymphocytes could be a possible biomarker of exposure to solar radiation. However, further work is necessary to confirm this.     

The use of fluorescence correlation spectroscopy (FCS) as an alternative biomarker detection technique: a preliminary study

Biomarkers are essential part of daily medical practice. Currently, biomarkers are being used both for diagnostic and prognostic purposes. There are many approaches e.g. ELISA by which biomarker levels are detected from patient samples. However, all these approaches are laborious, time consuming and expensive. There is therefore a general need for exploring new technique which can overcome these drawbacks. Here, we present a preliminary study for detection of serum biomarkers by fluorescence correlation spectroscopy (FCS) based diagnostic technique. FCS is a technique basically used for spatial and temporal analysis of molecular interactions of extremely low-concentration biomolecules in solution. FCS is able to measure diffusion time of the fluorescent molecules passing through the open detection volume and it can also measure the average number of fluorescent molecules passing through the detection volume. Because diffusion speed is correlated with shape and molecular mass of the fluorescent molecule, this property makes it possible to study the complex formation between a small fluorescently labelled and a large unlabelled molecule. In this preliminary study, we utilize this FCS property for detection of serum biomarker. Further studies on various pathological serum samples are warranted to explore further aspects of this technique.     

Sputum neutrophils as a biomarker in COPD: findings from the ECLIPSE study

Introduction

The percentage of neutrophils in sputum are increased in COPD patients, and may therefore be a biomarker of airway inflammation. We studied the relationships between sputum neutrophils and FEV₁, health status, exacerbation rates, systemic inflammation and emphysema, and long term variability at 1 year.

Methods

Sputum samples were obtained from 488 COPD patients within the ECLIPSE cohort. 359 samples were obtained at baseline, and 297 after 1 year. 168 subjects provided samples at both visits. Serum interleukin-6 (IL-6), IL-8, surfactant protein D and C-reactive protein levels were measured by immunoassays. Low-dose CT scans evaluated emphysema.

Results

Sputum neutrophil % increased with GOLD stage. There was a weak association between % sputum neutrophils and FEV₁ % predicted (univariate r² = 0.025 and 0.094 at baseline and year 1 respectively, p < 0.05 after multivariate regression). Similar weak but significant associations were observed between neutrophil % and health status measured using the St Georges Respiratory Questionairre. There were no associations between neutrophils and exacerbation rates or emphysema. Associations between sputum neutrophils and systemic biomarkers were non-significant or similarly weak. The mean change over 1 year in neutrophil % was an increase of 3.5%.

Conclusions

Sputum neutrophil measurements in COPD are associated weakly with FEV₁ % predicted and health status. Sputum neutrophil measurements were dissociated from exacerbation rates, emphysema and systemic inflammation.     

Statistical considerations of optimal study design for human plasma proteomics and biomarker discovery

A mass spectrometry-based plasma biomarker discovery workflow was developed to facilitate biomarker discovery. Plasma from either healthy volunteers or patients with pancreatic cancer was 8-plex iTRAQ labeled, fractionated by 2-dimensional reversed phase chromatography and subjected to MALDI ToF/ToF mass spectrometry. Data were processed using a q-value based statistical approach to maximize protein quantification and identification. Technical (between duplicate samples) and biological variance (between and within individuals) were calculated and power analysis was thereby enabled. An a priori power analysis was carried out using samples from healthy volunteers to define sample sizes required for robust biomarker identification. The result was subsequently validated with a post hoc power analysis using a real clinical setting involving pancreatic cancer patients. This demonstrated that six samples per group (e.g., pre- vs post-treatment) may provide sufficient statistical power for most proteins with changes>2 fold. A reference standard allowed direct comparison of protein expression changes between multiple experiments. Analysis of patient plasma prior to treatment identified 29 proteins with significant changes within individual patient. Changes in Peroxiredoxin II levels were confirmed by Western blot. This q-value based statistical approach in combination with reference standard samples can be applied with confidence in the design and execution of clinical studies for predictive, prognostic, and/or pharmacodynamic biomarker discovery. The power analysis provides information required prior to study initiation.     

Urinary TWEAK as a biomarker of lupus nephritis: a multicenter cohort study

Introduction

TNF-like weak inducer of apoptosis (TWEAK) has been implicated as a mediator of chronic inflammatory processes via prolonged activation of the NF-κB pathway in several tissues, including the kidney. Evidence for the importance of TWEAK in the pathogenesis of lupus nephritis (LN) has been recently introduced. Thus, TWEAK levels may serve as an indication of LN presence and activity.

Methods

Multicenter cohorts of systemic lupus erythematosus (SLE) patients and controls were recruited for cross-sectional and longitudinal analysis of urinary TWEAK (uTWEAK) and/or serum TWEAK (sTWEAK) levels as potential biomarkers of LN. The performance of TWEAK as a biomarker for nephritis was compared with routinely used laboratory tests in lupus patients, including anti-double stranded DNA antibodies and levels of C3 and C4.

Results

uTWEAK levels were significantly higher in LN patients than in non-LN SLE patients and other disease control groups (P = 0.039). Furthermore, uTWEAK was better at distinguishing between LN and non-LN SLE patients than anti-DNA antibodies and complement levels, while high uTWEAK levels predicted LN in SLE patients with an odds ratio of 7.36 (95% confidence interval = 2.25 to 24.07; P = 0.001). uTWEAK levels peaked during LN flares, and were significantly higher during the flare than at 4 and 6 months prior to or following the flare event. A linear mixed-effects model showed a significant association between uTWEAK levels in SLE patients and their disease activity over time (P = 0.008). sTWEAK levels, however, were not found to correlate with the presence of LN or the degree of nephritis activity.

Conclusions

High uTWEAK levels are indicative of LN, as opposed to non-LN SLE and other healthy and disease control populations, and reflect renal disease activity in longitudinal follow-up. Thus, our study further supports a role for TWEAK in the pathogenesis of LN, and provides strong evidence for uTWEAK as a candidate clinical biomarker for LN.     

Confidence regions for treatment effects in subgroups in biomarker stratified designs

Subgroup analysis has important applications in the analysis of controlled clinical trials. Sometimes the result of the overall group fails to demonstrate that the new treatment is better than the control therapy, but for a subgroup of patients, the treatment benefit may exist; or sometimes, the new treatment is better for the overall group but not for a subgroup. Hence we are interested in constructing a simultaneous confidence interval for the difference of the treatment effects in a subgroup and the overall group. Subgroups are usually formed on the basis of a predictive biomarker such as age, sex, or some genetic marker. While, for example, age can be detected precisely, it is often only possible to detect the biomarker status with a certain probability. Because patients detected with a positive or negative biomarker may not be truly biomarker positive or negative, responses in the subgroups depend on the treatment therapy as well as on the sensitivity and specificity of the assay used in detecting the biomarkers. In this work, we show how (approximate) simultaneous confidence intervals and confidence ellipsoid for the treatment effects in subgroups can be found for biomarker stratified clinical trials using a normal framework with normally distributed or binary data. We show that these intervals maintain the nominal confidence level via simulations.