Volume 54, Issue 3

Background: Geriatric reference intervals (RIs) are not commonly available and are rarely used. It is difficult to select a reference population from a cohort with a high degree of morbidity. Also important are the statistical approaches used to determine health-associated reference values. It is the aim of this study to examine the statistical methods used in the calculation of geriatric RIs. Methods: A search was conducted on EMBASE and Medline for articles between January 1989 and January 2014. Studies were selected if they: 1) were English primary articles; 2) performed a clinical chemistry test on a blood fraction; 3) had a population sub-group consisting of individuals ≥65 years of age; and 4) calculated a RI for the subgroup ≥65 years of age. Results: There were 64 articles identified, of which 78.1% described the RI calculation method used. RI calculation was performed by non-parametric (21.9%), parametric (42.2%), robust (3.1%), or other (17.2%) methods. Outlier detection (SD, Grubb’s test, Tukey’s fence, Dixon) was infrequently used and although most studies performed partitioning, only 57.8% tested the statistical significance of the partitions. Few studies (17.2%) reported confidence intervals for the RI estimates. Overall, only 14.1% of studies provided RI estimates which followed the CLSI guideline EP28-A3c. Conclusions: Statistical methods for RI calculation and partitioning varied considerably between studies and many failed to provide adequate descriptions of these methods. Challenges in analyses arose from insufficient sample sizes and heterogeneity in the elderly population. Geriatric RIs, although present in the literature, may not be properly calculated and should be carefully considered before applying them for clinical care.

High-throughput technologies such as next-generation genomics, transcriptomics and proteomics are capable of generating massive amounts of data quickly, and at relatively low costs. It is tempting to use this data for various medical applications including preclinical disease detection and for prediction of disease predisposition. Pilot projects, initiated by various research groups and Google, are currently underway, but results with not be available for a few years. We here summarize some possible difficulties with these approaches, by using examples from already tried cancer and other screening programs. Population screening, especially with multiparametric algorithms, will identify at least some false positive parameters and screening programs will identify abnormal results in otherwise healthy individuals. Whole genome sequencing will identify genetic changes of unknown significance and may not predict accurately future disease predisposition if the disease is also influenced by environmental factors. In screening programs, if the disease is rare, the positive predictive value of the test will be low, even if the test has excellent sensitivity and specificity. False positive results may require invasive procedures to delineate. Furthermore, screening programs are not effective if the cancer grows quickly, and will identify indolent forms of the disease with slow-growing tumors. It has also been recently shown that for some cancers, more intensive and radical treatments do not usually lead to better clinical outcomes. We conclude that new omics testing technologies should avoid overdiagnosis and overtreatment and need to be evaluated for overall clinical benefit before introduction to the clinic.

Background: α -Thalassemia, caused by mutations in the α-globin genes, is one of the most common monogenic inherited disorders in the world. However, non-deletional α -thalassemia mutations remain undetected in routine clinical testing due to the lack of a suitable method. In this study, a closed- and single-tube assay for the detection of six common non-deletional α -thalassemia mutations in the HBA 2 gene was developed based on multicolor melting curve analysis. Methods: The assay consisted of one pair of primers specific for the HBA 2 gene and four dual-labeled, self-quenched probes targeting six non-deletional α -thalassemia mutations. The sensitivity, reproducibility, and accuracy of the method were validated via 700 genomic DNA samples. Results: The assay had a reproducibility of 100%, could detect gDNA of different genotype as low as 1 ng per reaction, and had an overall accuracy of 100% when compared with RDB analysis and Sanger sequencing. Conclusions: The developed assay is rapid, robust, and cost-effective while maintaining high sensitivity, specificity, and throughput.

Background: Quality control (QC) is an essential part of clinical biochemistry to ensure that laboratory test results are reliable and correct. Those tests without a defined reference method constitute a special challenge, as is the case with bone-specific alkaline phosphatase (BAP). Methods and Results: The present study reports an example where a shift in a BAP assay was detected by use of a patient pool and supported by a retrospective calculation of “patient mean”, while the external QC and specific assay control material were unaffected by the shift. Conclusions: Patient pools and the use of patient means remain a useful and inexpensive procedure for internal QC.

Background: Long transportation times of samples can occur due to centralization of laboratories, and also in, for instance epidemiological multicenter studies with one core laboratory. Unsatisfactory glycolysis inhibition is known to threaten the correct measurements of glucose concentration in patient samples. In former studies Terumo Glycaemia tubes proved to be superior to other anticoagulant systems for time periods of up to 24 h. We investigated long-term stability of glucose concentration in Terumo Glycaemia tubes for up to 96 h at room temperature and imitated transport conditions by continuous sample shaking. Methods: Human venous blood samples were collected from 40 healthy blood donors using Terumo Glycaemia tubes. Immediately after sampling, tubes were mixed according to the manufactures recommendations. To simulate transportation conditions samples were placed on a shaker for the entire study period and maintained at room temperature. Samples were (re)centrifuged at 0, 24, 36, 48, 72 and 96 h prior to measuring glucose concentration. The glucose concentration at 0 h was used as baseline for evaluation of long-term stability. Results: The recovery of glucose was 100% throughout the study, including the 96-h measurements. Deviations of single glucose measurements were within the imprecision of the measurement procedure. Conclusions: Terumo Glycaemia tubes can effectively stabilize glucose in whole blood samples kept at room temperature on a shaker during a 96-h time period. Therefore, we consider Terumo Glycaemia tubes as a suitable glucose stabilizing tube for long intervals between sample collection and glucose quantification.

Background: Delayed sample processing can affect accurate glucose measurement. Our aim was to investigate the stability of glucose in samples collected in serum, sodium fluoride/potassium oxalate (NaF/KOx) and Glucomedics tubes processed according to different controlled pre-centrifugation delays (up to 180 min after venipuncture) in order to simulate prolonged sample transport between venipuncture and centrifugation. Methods: Samples were collected from healthy volunteers (n=80) into either serum or NaF/KOx and Glucomedics tubes. Glucose concentration was measured in samples centrifuged immediately after venipuncture and compared with tubes processed with a delay of 60, 120 and 180 min prior to centrifugation. Differences between baseline and respective delayed centrifugation glucose value for each tube type were tested using the paired t-test. Mean bias calculated for each tube type and delay protocol was compared to recommended quality specifications for glucose (2.2%). Results: Glucose concentrations measured in all three delayed tube types were lower in comparison to respective baseline glucose concentrations measured in immediately processed tube (p<0.001). The highest decrease in glucose was observed in serum tubes in all specified time points (p<0.001), while glucose was most stable in Glucomedics tubes (p<0.001). The decrease in glucose observed for serum and NaF/KOx tubes was clinically significant at all specified time points while the bias for Glucomedics tubes did not exceed the criteria even with a centrifugation delay of 180 min. Conclusions: Glucose stability in un-centrifuged Glucomedics tubes is much superior to serum and NaF/KOx tubes. Glucomedics tubes can be left un-centrifuged for up to 3 h without affecting glucose concentration.

Background: Endogenous interferences are an important source of biased laboratory results. Hemolysis, lipemia and icteria are the main source of endogenous interference in laboratory medicine. Accreditation according to ISO 15189 improves the overall quality of the laboratory procedures. The aim of our study was i) to assess the level of knowledge of Croatian medical biochemists about the proper detection and management of hemolysis, lipemia and icteria; and ii) to identify possible differences in the level of knowledge respective to the laboratory accreditation status. Methods: An on-line self-report survey was carried out by the Working Group for Preanalytical Phase of the Croatian Society of Medical Biochemistry and Laboratory Medicine during April to May 2015. Survey included 14 statements (Q1–Q14) about procedures for samples with interferences and participants were asked to assess the degree of agreement with the statement using a 4-point Likert scale. Results: The lowest level of knowledge was observed for statements Q10 (dealing with icteric sample; 40.9% participants agreed with the correct procedure), Q12 (allowable error for interference; 47.2%) and Q11 (dealing with lipemic sample; 60.1%). Almost all participants (97.4%) agreed that laboratories in Croatia should have a harmonized protocol for management of samples with interferences. Participants from accredited laboratories showed higher knowledge of hemolysis detection (p=0.031), rejection of hemolyzed sample (p<0.001), management of icteric samples (p=0.038) and allowable error for interferences (p=0.040). Conclusions: Croatian laboratories have a good knowledge of the proper detection and management of hemolyzed, icteric and lipemic samples. Accreditation is associated with higher knowledge about management of samples with interferences.

Background: In recent years, fatty acid ethyl esters (FAEEs) and ethyl glucuronide (EtG) in meconium emerged as reliable, direct biological markers for establishing gestational ethanol exposure. We investigated whether EtG in maternal hair measured during the three trimesters of pregnancy correlated with EtG and FAEEs in neonatal meconium. Methods: In a prospective sample of 80 mother-infant dyads from Barcelona (Spain), we measured EtG and FAEE in maternal hair segments and meconium samples using a validated UHPLC-MS/MS method. Results: Fifty-eight (72.5%) women had EtG concentrations in the hair shafts >7 pg/mg in one or more pregnancy trimesters, and EtG and FAEEs in meconium samples were documented in 50 and 24 of their neonates, respectively. The best significant correlations (p<0.0001) were found between EtG concentration in the proximal 0–3 and 3–6 hair shaft segments corresponding to the last two pregnancy trimesters and EtG in neonatal meconium (ρ=0.609 and ρ=0.577, respectively). Using the combination of EtG in meconium ≥30 ng/g and a median of EtG >11 pg/mg in maternal hair during the second and third trimesters of pregnancy, prenatal ethanol exposure could be predicted with a sensitivity of 85.7% and specificity of 73.7%. Conclusions: This study provides evidence of proven fetal exposure to ethanol during the second and third trimesters of pregnancy by linking detection of ethanol biomarkers (EtG) in maternal hair segments and EtG in neonatal meconium.

Background: Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is routinely used for analysis of immunosuppressive drugs. This study investigated whether replacing analog internal standards (ANISs) with isotopically labeled internal standards (ILISs) has an impact on the performance of a LC-MS/MS method for the quantification of tacrolimus (TAC), sirolimus (SIR), ciclosporin A (CsA) and everolimus (EVE) in whole blood. Methods: Following hemolysis, protein precipitation, and extraction with either ANISs (ascomycin, desmethoxy-rapamycin, CsD), or ILISs (TAC- 13 C,D 2 ; SIR- 13 C,D 3 ; CsA-D 12 ; EVE-D 4 ), samples were centrifuged and injected into a LC-MS/MS device equipped with a C18 reversed phase column. The effect of the two ISs on the linearity, precision, accuracy, trueness, matrix effects, and carryover was investigated by using the same patient-, proficiency testing-, and quality control samples. Statistical analysis of agreement between results includes a standard random effects model and Passing-Bablok regression. Results: Within-day imprecision was <10%, between-day <8%, and trueness 91%–110% for all the analytes with both ISs. No carryover or matrix effects were observed. The median accuracy was −2.1% for CsA, 9.1% for EVE, 12.2% for SIR, and −1.2% for TAC with the ILISs; and −2% for CsA, 9.8% for EVE, 11.4% for SIR, and 0.2% for TAC with the ANISs. Results of patient and proficiency testing samples were not statistically different. Conclusions : Although ILISs are generally considered superior to ANISs, they may not be always essential. When optimizing a LC-MS/MS method other factors must be also considered.

Background: Deregulation of synaptic plasticity and oxidative stress are reported to play a crucial role in the pathogenesis of schizophrenia. Matrix metalloproteinase-9 (MMP-9) is an extracellular protease involved in regulation of synaptic plasticity. Malondialdehyde (MDA) is a marker of lipid peroxidation which is elevated in schizophrenia. Earlier studies have reported polymorphism of MMP-9 and its association with schizophrenia. The present study was designed to assess the serum levels of MMP-9, MDA and total antioxidant status (TAS) and their association in schizophrenia. Methods: A total of 40 cases and 40 controls were included in the study. Serum MMP-9, MDA and TAS were estimated in all the subjects. Disease severity was assessed using Positive and Negative Syndrome Scale (PANSS). Results: MMP-9 and MDA were significantly increased and TAS were significantly reduced in schizophrenia cases compared to controls. MMP-9 was positively correlated with MDA (r=0.353, p=0.025) and negatively correlated with TAS (r=−0.461, p=0.003). TAS was significantly correlated with total (r=0.322, p=0.043) and negative symptom scores (r=0.336, p=0.034). Higher MMP-9 levels were associated with previous exposure to antipsychotics (p=0.032). Conclusions: MMP-9 and oxidative stress were increased and correlate well with each other in schizophrenia cases. Though total oxidant status showed positive association with disease severity, MMP-9 and MDA were not associated with the severity of the disease.

Background: Fibrinogen-based clot firmness is reported as the maximum amplitude (MA) when using the citrated functional fibrinogen (CFF) assay in thrombelastography (TEG), and as the maximum clot firmness (MCF) together with several clot amplitude parameters when using the FIBTEM assay in thromboelastometry (ROTEM). Concern is currently being raised that these two tests have different platelet inhibiting performance and consequently provide different values. This is relevant for the clinical setting of fibrinogen replacement. We aim herein to compare the parameters of these two fibrinogen-based clot quality tests and their correlation with the plasma fibrinogen level as determined by the Clauss method. Methods: In total 261 whole blood samples taken from 163 clinical routine surgical patients were analyzed with TEG 5000 and ROTEM tests, and correlation with Clauss fibrinogen level was assessed. Results : Using TEG, the overall fibrin-based clot firmness measured in the CFF assay was significantly higher than the MCF measured by FIBTEM assay. Both assays showed significantly positive correlations with the fibrinogen levels measured using the Clauss method. However, individual values of Clauss fibrinogen concentration corresponded with different values for the two viscoelastometric tests; e.g. within the range of 1.9–2.1 g/L Clauss fibrinogen the median of CFF MA was 16.3 mm whereas FIBTEM MCF was 12.0 mm. Conclusions : We showed herein by measurements of citrated whole blood samples from surgical patients that CFF MA values were different from FIBTEM MCF values measured in the same sample. Awareness that these whole blood assays provide different clot amplitude results is mandatory, particularly if they are being considered as tools for guiding fibrinogen supplementation. Thromboembolic side effects caused by a potentially too high fibrinogen substitution must also kept in mind in this context.

Background: Immunoglobulin κ and λ free light chains (FLC) are important serum biomarkers for diagnosing and monitoring plasma cell dyscrasias (via the κ:λ FLC ratio), and assessing immune competence and activation status (via ∑FLC). FLCs are produced, in excess of heavy chains, from healthy plasma cells during immunoglobulin production, but unlike intact immunoglobulins that are cleared by cellular catabolism over a number of weeks, FLC are rapidly cleared from the bloodstream by the renal glomerulus with a half-life of 3 (κ FLC)–6 (λ FLC) hours. Marathon running has been shown to acutely and transiently decrease renal function, however, the impact of prolonged aerobic exercise on FLC levels remains unknown. Methods: We measured serum FLC levels in 60 runners before, and immediately after, the 2010 Eindhoven Marathon. Results: A significant increase (p<0.01) in κ FLC levels was observed after the marathon, and κ FLC correlated positively with serum creatinine levels. No changes were observed for λ FLC, and thus, there were subtle elevations in the ∑FLC and FLC ratio in some participants. Indeed, we found that 13% of participants had an abnormally increased FLC ratio upon completion of the marathon; a phenomenon previously observed in renal diseases. Conclusions: Abnormal FLC ratios observed after exercise reflected an increase in serum κ FLC levels, which may be due to acute and transient reductions in renal function during exercise, though we also observed an increase in serum IgG and IgA and thus cannot exclude exercise-induced immune stimulation or immunoglobulin redistribution.

Background: Loss of circadian rhythms and reduced concentrations of endogenous melatonin are common in critically ill patients. After exogenous administration, supra-physiological concentrations in serum are only ephemeral, which may explain the absence of significant therapeutic effect on sleep. The aim of this study is to describe the pharmacokinetics of enteral melatonin in critically ill patients administered in a novel regimen aiming to simulate endogenous release. Methods: Thirteen patients in the recovery phase of critical illness were randomised to receive enteral melatonin or placebo. In the melatonin group, a total of 6 mg was administered as solution through their feeding tube, commencing with a 3 mg loading dose at 9 pm and six subsequent 0.5 mg doses hourly. The placebo was administered using a similar regimen. Serial blood samples were taken and measured using a validated chromatographic method. The concentration-time data for serum melatonin concentrations were described using non-linear mixed-effects modelling. Results: The observed concentrations in the melatonin patients were significantly higher than that observed in the placebo patients. The concentrations in the patients administered melatonin were also higher than endogenous melatonin concentrations previously reported in non-critically ill patients. The patients administered melatonin had a mean clearance, volume of distribution and absorption rate constant of melatonin was 55.2 L/h, 767 L and 0.76 h –1 , respectively. Conclusions: Exogenous administration of melatonin with a loading dose of 3 mg followed by an hourly dose of 0.5 mg demonstrates good oral bioavailability and results in supra-physiological and sustained concentrations of serum melatonin during 12 h overnight.

Background: Evaluation of appropriateness of laboratory tests on the basis of individual requests remains a serious problem as the clinical question is usually not reported with the test order. This study explored the comparison of the rate of tumor marker orders with cancer prevalence as a putative indicator of inappropriateness. Methods: Tumor marker orders (2011 and 2012) were obtained from the Ministry of Health and cancer prevalence from the Italian Association of Cancer Registries. The rate of tumor marker orders was matched with demographic data and tumor prevalence and examined by using the confidence interval approach. Region-to-region and year-to-year variations were also examined. Focus was placed on CEA, CA125, CA19.9 and CA15.3. Results: Tumor markers ordered in Italy were 13,207,289 in 2012 (221.3/1000 individuals). Given an estimated prevalence of 2,243,953 cancer cases, 7.04 tumor markers appear to be requested for each prevalent case of epithelial cancer per year. The rate of requests of CEA, CA125, CA19.9 and CA15.3 (in aggregate 5,834,167 requests in 2012, 44.2% of total) from the first and the last ranked region (96 and 244/1000 individuals) are significantly different (p<0.01). Region-to-region differences do not correspond to any known variation of prevalence in the different regions. Conclusions: The developed approach provides a proxy indicator of inappropriateness showing that tumor markers are overused in Italy and their ordering pattern is not related to tumor prevalence. The model is suitable to be validated in other laboratory tests used in diseases whose prevalence is known.

Background: PCA3 and ERG are mRNA-based prostate cancer (PCa) specific biomarkers that can be detected in urine. However, urine is a complex substrate that can be separated in several fractions. In this study we compared the levels of PCa-specific biomarkers ( PCA3 and ERG ) and KLK3 as prostate-specific reference gene in three urine substrates–whole urine, urinary sediment (cell pellet) and exosomes–and evaluated the influence of performing a digital rectal examination (DRE) prior to urine sampling. Methods: First-voided urine samples were prospectively obtained before and after DRE from 29 men undergoing prostate biopsies. The urine was separated in whole urine, cell pellet and exosomes and the biomarker levels were measured with RT-qPCR. Results: PCa was identified in 52% (15/29) of men. In several samples the mRNA levels were below the analytical limit of detection (BDL). The biomarker levels were highest in whole urine and significantly higher after DRE in all substrates. In PCa patients higher levels of PCA3 and ERG were found in all urine substrates after DRE compared to non-PCa patients. Conclusions: This is the first study in which urinary PCa-specific biomarker levels were compared directly in three separate urine fractions. These results suggest that whole urine could be the urine substrate of choice for PCa-diagnostics based on analytical sensitivity, which is reflected directly in the high informative rate. Moreover, the significant positive effect of performing a DRE prior to urine sampling is confirmed. These findings could be of influence in the development of PCa-diagnostic urine tests.

Background: Quantification of human immunodeficiency virus type 1 (HIV-1) RNA in plasma has become the standard of care in the management of HIV-infected patients. There are several commercially available assays that have been implemented for the detection of HIV-1 RNA in plasma. Here, the new Hologic Aptima ® HIV-1 Quant Dx assay (Aptima HIV) was compared to the Roche COBAS ® TaqMan ® HIV-1 Test v2.0 for use with the High Pure System (HPS/CTM). Methods: The performance characteristics of the assays were assessed using commercially available HIV reference panels, dilution of the WHO 3rd International HIV-1 RNA International Standard (WHO-IS) and plasma from clinical specimens. Assay performance was determined by linear regression, Deming correlation analysis and Bland-Altman analysis. Results: Testing of HIV-1 reference panels revealed excellent agreement. The 61 clinical specimens quantified in both assays were linearly associated and strongly correlated. Conclusions: The Aptima HIV assay offers performance comparable to that of the HPS/CTM assay and, as it is run on a fully automated platform, a significantly improved workflow.

Background: The chronic stage of human immunodeficiency virus (HIV) infection, although clinically asymptomatic, is characterized by activation of the immune system and persistent inflammation. Procalcitonin (PCT) has been studied in HIV infection as a marker of bacterial infection. Our aim was to assess the effect of persistent immune activation on PCT levels in asymptomatic treatment naïve HIV infected subjects. Methods: This was a cross-sectional study of 68 asymptomatic antiretroviral therapy-naive HIV infected participants and 42 uninfected controls. Stored serum samples were used to measure: PCT, interleukin-6 (IL-6), lipopolysaccharide binding protein (LBP), high sensitivity C-reactive protein (hsCRP), immunoglobulin G (IgG) and albumin. PCT was correlated with markers of: disease progression (CD4 count and viral load), immune activation (CD 38 on CD8+ T cells, IgG and LBP), inflammation (IL-6, hsCRP and albumin). Results: IL-6, IgG and CD8/38 were all significantly increased while albumin and CD4 counts were significantly lower in the HIV infected group. PCT levels were not significantly different between the two groups. There was no significant difference in LBP and hsCRP; however, their levels were increased in both groups. PCT correlated only with LBP (p=0.0001). IL-6 and LBP correlated positively with hsCRP and IgG. Albumin correlated inversely with IL-6 and viral load. Only IgG and CD8/38 correlated inversely with CD4 counts. Conclusions: We demonstrated the activation of the innate (raised LBP), humoral (raised IgG) and cellular immune systems (increased CD8/38 T cells). Despite a state of persistent inflammation, PCT levels are not elevated in asymptomatic untreated HIV infection.

Background: Gestational diabetes (GDM) is usually diagnosed late in pregnancy, precluding early preventive interventions. This study aims to develop a predictive model based on clinical factors and selected biochemical markers for the early risk assessment of GDM. Methods: Based on a prospective cohort of 7929 pregnant women from the Quebec City metropolitan area, a nested case-control study was performed including 264 women who developed GDM. Each woman who developed GDM was matched with two women with normal glycemic profile. Risk prediction models for GDM and GDM requiring insulin therapy were developed using multivariable logistic regression analyses, based on clinical characteristics and the measurement of three clinically validated biomarkers: glycated hemoglobin (HbA 1c ), sex hormone binding globulin (SHBG) and high-sensitivity C-reactive protein (hsCRP) measured between 14 and 17 weeks of gestation. Results: HbA 1c and hsCRP were higher and SHBG was lower in women who developed GDM (p<0.001). The selected model for the prediction of GDM, based on HbA 1c , SHBG, BMI, past history of GDM, family history of diabetes and soft drink intake before pregnancy yielded an area under the ROC curve (AUC) of 0.79 (0.75–0.83). For the prediction of GDM requiring insulin therapy, the selected model including the same six variables yielded an AUC of 0.88 (0.84–0.92) and a sensitivity of 68.9% at a false-positive rate of 10%. Conclusions: A simple model based on clinical characteristics and biomarkers available early in pregnancy could allow the identification of women at risk of developing GDM, especially GDM requiring insulin therapy.