Daria Pašalić
Department of Medical Chemistry, Biochemistry and Clinical Chemistry
Zagreb University School of Medicine
Šalata ul 2.
10 000 Zagreb, Croatia
Phone +385 (1) 4590 205; +385 (1) 4566 940
E-mail: dariapasalic [at] gmail [dot] com

Useful links

A – Laboratory diagnostics and monitoring of malignant diseases


Assessment of tumor marker ProGRP concentration in patients with small lung cancer

Jozo Ćorić1, Radivoj Jadrić2, Elma Kučukalić1, Berina Hasanefendić1, Mirsad Panjeta1

1Department of Clinical Chemistry, Clinical Center of Sarajevo University, Sarajevo, Bosnia and Herzegovina

2Department of Biochemistry, School of Medicine, Sarajevo, Bosnia and Herzegovina


Introduction: Gastin-releasing peptide (GRP) is a specific and secreted product of small cell lung carcinoma (SCLC) cells. The assay ProGRP is used to aid in the differential diagnosis in lung cancer in conjunction with other tumor markers like determination serum concentrations neuron-specific enolase (NSE).

Materials and Methods: ProGRP and NSE were measured by electrochemiluminescence immunoassay (ECLIA) on Roche-Cobas e601 analyzer in sera obtained from 25 healthy subjects and 32 patients with histological proven SCLC. Median age of SCLC patients was 62 year old (range: 42-69) and included 19 females and 13 males. Analytical assessment of ProGRP determination comprised within-run and between-run imprecision. For quality control we use PreciControl ProGRP 1 and 2.

Results: Median values for ProGRP were significantly higher in patients with SCLC to the control group (30.3 vs. 141.2 pg/mL, P<0,001). Serum ProGRP levels were elevated in 78% (25/32) patients. We found similar results with NSE. Within-run imprecision on the commercially controls for PreciControl ProGRP 1 is 4.6% and 3.6% for PreciControl ProGRP 2; between-day imprecision for PreciControl ProGRP 1 is 7.5% and 8.2% for PreciControl ProGRP 2 respectively.

Conclusion: Serum ProGRP level were specifically elevated in SCLC patients. The presented results of the analytical evaluation methods for the determination of ProGRP on the Roche-Cobas e601 analyzer showed an acceptable accuracy and precision.

e-mail: coricjozo [at] hotmail [dot] com 



Correlation between procalcitonin and C-reactive protein values on first postoperative day and patient’s recovery period after malignant bowel disease surgery

Sanja Dobrijević1, Ljiljana Mayer1, Danko Velimir Vrdoljak2, Mihaela Gaće1, Zvjezdana Špacir Prskalo1, Petra Pozaić1, Iva Kirac2

1Medical Biochemistry Laboratory, Institute for Tumors, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

2Department of Surgical Oncology, Institute for Tumors, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: Procalcitonin (PCT) and C-reactive protein (CRP) are inflammatory markers with different synthesis and degradation dynamics. Our aim was to determine number of cases where CRP does not follow PCT dynamics, and investigate if PCT values on first postoperative day after malignant bowel disease surgery correlates with patient’s recovery period.

Subjects and Methods: In 1-year period 181 patients were operated due to malignant bowel disease. Data about length of postoperative hospital stay, PCT and CRP concentration in serum from 1st to 6th postoperative days were collected retrospectively. Patients whose inflammatory parameters were not measured on first postoperative day were excluded. Group included 67 patients (61% men) with median age 64 (30-80). PCT was determined on Elecsys 2010 (Roche Diagnostics GmbH, Germany), and CRP on Architect c4000 (Abbott Laboratories, USA). MedCalc Version (Mariakerke, Belgium) was used for statistical analysis. P<0.05 was considered statistically significant.

Results: Significant correlation between recovery period and PCT values (r=0.31; P=0.011) and CRP values on first postoperative day (r=0.29; P=0.019) was observed. On the second postoperative day, decrease in PCT concentration and increase of CRP value were noted in 51% subjects. There was no statistically significant difference (P=0.096) in PCT values in the group of patients who spent ≤10 days in hosiptal after surgery of those who spent >10 days. PCT values higher than 1.16 ng/mL determined on first postoperative day indicate that postoperative hospital stay will be longer than 10 days with sensitivity of 45.2% (95%CI 27.3-64.0) and specificity of 77.8% (95%CI 60.8-89.9).

Conclusion: Correlation between recovery period of patients in the hospital after malignant bowel disease surgery and inflammatory parameters on first postoperative day was statistically significant. In more than half of patients decrease of PCT values can be seen one day before decrease of CRP values.

e-mail:saana [dot] smail [at] gmail [dot] com


The diagnostic potential of human epididymis protein 4 (HE4) and ROMA index in women with ovarian cancer

Zvjezdana Špacir Prskalo1, Ljiljana Mayer1, Mihaela Gaće1, Sanja Dobrijević1, Mario Puljiz2, Ilija Alvir2, Ivica Mamić2, Damir Danolić2

1Medical Biochemical Laboratory, Institute for Tumors, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

2Department of Surgical Oncology, Department of Gynecology Oncology Surgery, Institute for Tumors, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: Ovarian cancer which in Croatia is fifth in newly diagnosed cancer in women; 440 newly diagnosed die 301 women. In recent years, the arguments are that the HE4, either by itself or in combination with CA 125 as part of the ROMA calculations far more reliable marker (no increases in endometriosis, cysts and non-gynecological malignancies). First time in Croatia examined the utility of HE4 and ROMA calculating the classification of ovarian tumors of other gynecological malignancies.

Subjects and Methods: The study included 63 patients of Institute for tumors with preoperative findings demonstrated tumor mass in the pelvis of unknown etiology. In 20/63 women subsequent histopathologic diagnosis was confirmed ovarian cancer. HE4 and CA125 measurement were the Elecsys 2010 (Roche). The subjects were classified according to menopausal status, to the ROMA calculation formula is used.

Results: Measured concentrations of HE4 in a group of women with ovarian cancer 131.7 (76.6-639.9) mg/mL was significantly higher than other gynecological malignancies 52.6 (48.1-62.2) mg/mL; P<0.001. ROC analysis revealed 64.7% sensitivity and 90.5% specificity with cutt-of 95.62 mg/mL; AUC of 80.3% for the HE4 and 70.6% sensitivity and 88.9% specificity with the cutt-of 34.1 mg/mL; AUC of 82.2% for the calculation of the ROMA in distinguishing ovarian cancer from other cancers.

Conclusion: The results confirm literature data of the usability of HE4 and ROMA calculation. Confirmed the relevance and usefulness of preoperative findings clinician operator to be doing just about ovarian cancer. Incomparably smaller number of false positive results in comparison with CA125 makes HE4 almost ideal organ specific tumor marker. Calculation HE4 with CA125 within ROMA calculation increases the sensitivity with only a slight reduction in specificity. Gynecologists and oncologists in HE4 recognize the potential for monitoring therapy and progress of the disease.

e-mail: zspacir [at] net [dot] hr       



Diagnostic value of HE4 tumor marker in gynecological oncology

Dragana Šegulja, Danica Matišić, Ivana Lapić, Dunja Rogić

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Ovarian cancer is the second most common malignancy and the fifth cause of cancer death in women. Five-year survival in epithelial ovarian cancer detected in FIGO stage I is 90%. Today, <30% ovarian Ca are diagnosed in FIGO stage I or II. CA125, recently the only tumor marker in diagnostics of ovarian Ca, has insufficient diagnostic specificity and sensitivity. Several different proteins are investigated to improve diagnostics of ovarian Ca. Promising results have been obtained for secretory glycoprotein HE4. The aim of this study was to examine the diagnostic value of HE4 in patients referred to UHC Zagreb for determination of CA125.

Subjects and Methods: HE4 and CA125 markers were determined in 80 patients with median age 60 (21-85) by electrochemiluminescent method on Cobas 6000cee (ECLIA,Roche Diagnostics). Results were analyzed using statistical program MedCalc.

Results: Patients were divided into two groups: those with benign diseases (N=28): endometriosis, ovarian cysts, benign ovarian neoplasm, leiomyoma, and patients with malignant gynecological diseases (N=52): malignant ovarian, uterus, cervix and breast neoplasm. The median value of CA125 was 64.5 KIU/L (7.09-8277), while the median value of HE4 was 117.4 pmol/L (11.2-1999). Although the area under the ROC curve for HE4 (AUC 0.790) was larger than the area under the ROC curve for CA125 (AUC 0.739), the difference in AUCs was not statistically significant (P=0.410). The CA125 sensitivity of 92% and specificity of 43% was obtained by cut-off 35 KIU/L while for HE4 sensitivity was 53% and specificity 82% by cut-off 140 pmol/L.

Conclusion: Results of this study are consistent with current literature data. Since CA125 has been shown on studied population as a marker of greater sensitivity and HE4 of greater specificity in the diagnosis of gynecological malignancies, particularly ovarian cancer, optimal diagnostic value would be achieved by simultaneously determining the both studied markers.

e-mail:ivana [dot] lapic [at] hotmail [dot] com           


Serum free light chains as new prognostic biomarker in chronic lymphocytic leukemia

Blaženka Dobrošević, Mirjana Fijačko, Jasna Pavela, Barbara Vuković, Vatroslav Šerić

Department Of Clinical Laboratory Diagnostics, Clinical Hospital Centre Osijek, Osijek, Croatia


Introduction: Chronic lymphocytic leukemia B-cell ( B-CLL) is the most common form of leukemia among elderly. Male-to-female ratio is 2:1 with a median age of 60 years. It has unpredictable clinical course, median survival varies from 2-20 years. Because of this, continuous effort has been put into finding prognostic biomarkers useful for predicting survival. Serum β2-microglobulin, cell markers CD38 and ZAP-70 are established independent prognostic biomarkers in B-CLL. Recent data has shown that elevated levels of serum free light chains (sFLC) were shown to be adverse prognostic factor in B-CLL. The goal of our research is to compare the concentrations of sFLC with a concentrations of a well known prognostic biomarker, serum β2-microglobulin, among patients with B-CLL.
Subjects and Methods: We studied retrospectively 39 patients (28 male and 11 female, median age 69.03±11.1) with newly diagnosed B-CLL. All patients had typical immunophenotype, coexpress CD5, CD19, CD20 and CD23, restricted kappa or lambda chain, determined by flow cytometry. All patients had retrospectively determined concentrations of sFLC kappa and lambda and concentration of β2-microglobulin by nephelometry (Siemens Healthcare Diagnostics, Marburg, Germany). By calculation we have determined the total concentration of sFLC (kappa+lambda). We used the statistic Spearman’s rho test to determine the correlation between total sFLC and β2-microglobulin.

Results: Based on recent literature articles, the value of total sFLC above 60 mg/L (cut off ) is a independent prognostic biomarker. Value of serum β2-microglobulin ranged from 1.79-13.70 g/L (mean 5.04). Value of total sFLC ranged from 13.4-520.2 mg/L (mean 111.8). Total concentration of sFLC was higher than 60 mg/L in 46% of patients. The total concentration of sFLC correlated significantly with the serum β2-microglobulin (P=0.01; r=0.54).

Conclusion: Our results match the ones from previously published literature articles. The concentration of total sFLC is a good, routinely available biomarker of survival.

e-mail: dobrosevic [dot] blazenka [at] kbo [dot] hr


B – Statistical analysis of laboratory data


Results of the recommended laboratory criteria appliance for determining diabetes in pregnancy

Bojana Kranjčec1, Jadranka Šanjug2

1Medical Biochemistry Laboratory, General Hospital Zabok and Hospital of Croatian Veterans, Zabok, Croatia

2Department of Gynecology and Obstetrics, General Hospital Zabok and Hospital of Croatian Veterans, Zabok, Croatia


Introduction: The comparison of oGTT test in pregnant women according to WHO criteria from 2006, with the Standard laboratory procedure for the diagnosis of diabetes in pregnancy which is in compliance with the recommendations of the IADPSG and which is recommended by the Commission for Professional Issues CCMB.

Subjects and Methods: For the first group of women the oGTT test was performed, measuring venous plasma glucose concentration before loading them with 75 g of glucose, and after 120 minutes. For the second group of pregnant women the oGTT test was performed which measured glucose concentration before loading with 75 g of glucose, after 60 and 120 minutes. The concentration of glucose was determined by BC enzymatic UV test (BC AU680), with supervising preanalytical, analytical and postanalytical conditions.

Results: Using WHO criteria for diagnosis of diabetes in pregnancy with the limit value for concentrations of fasting glucose of 7 mmol/L, and then after 120 minutes of 7.8 mmol/L, from the first group of 259 pregnant women, 21 (8.1%) who had one of these glucose concentrations equal or increased were singled out. By applying the Standard laboratory procedure for the diagnosis of diabetes in pregnancy with the limit values ​​for fasting glucose of ≥5.1 mmol/L, after 60 minutes of ≥10.0 mmol/L, and after 120 minutes of ≥8.5 mmol/L, in the second group of 285 pregnant women, 73 (25.6%) which had value of the concentration of glucose equal or increased were singled out. The difference in proportions test demonstrated a statistically significant difference for the diagnosis of diabetes between these two groups.

Conclusion: The following Standard laboratory procedure for the diagnosis of diabetes in pregnancy as recommended by IADPSG and CCMB, in relation to the earlier used WHO recommendations, a greater incidence of diabetes in pregnancy, which is in line with the findings in the literature was detected.

e-mail:bojanakranjcec [at] gmail [dot] com  


B02 (Oral presentation)

The difference of verification protocols for precision assessment applied to routine coagulation tests

Sanda Jelisavac Ćosić1, Désirée Coen Herak2, Vanja Radišić Biljak3

1Department of Oncology and Radiotherapy, University Hospital Centre Zagreb, Zagreb, Croatia

2Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

3Institute of Clinical Chemistry and Laboratory Medicine, Merkur University Hospital, Zagreb, Croatia


Introduction: The aim of the study was to compare two verification protocols for precision study applied to routine coagulation tests: duplicate measurement during ten consecutive days (Protocol 1) and triplicate measurement during five consecutive days (EP-15). EP-15 was introduced as a highly efficient, time- and money-saving protocol. We hypothesized that there is no difference in results obtained with two protocols.

Materials and Methods: Simultaneous validation of coagulation analyzers Instrumentation Laboratory ACL TOP 500 and ACL TOP 300 (Bedford, USA) was performed using original HemosIL reagents for PT (PT RecombiPlasTin 2G), APTT (SynthASil), fibrinogen (Fibrinogen-C), AT (Liquid Antithrombin) and D-dimer (D-Dimer HS 500) and commercial control plasma samples, Normal and Low Abnormal Control Assayed for PT, APTT, fibrinogen and AT, as well as, Low and High Control for D-dimer analysis. Coefficients of variation (CVs) were calculated for each protocol. CV’s proposed by the manufacturer were considered acceptable.

Results: No difference in CVs for APTT and D-Dimer, performed with two protocols, was observed, and the manufacturer criteria were fully achieved. For the remaining analysis increased CVs with the EP-15 protocol were noticed on both analyzers, compared to Protocol 1. CVs ranged from 23% for fibrinogen to 241% for AT on ACL TOP 500 and 26% for fibrinogen to 85% for AT on ACL TOP 300. Manufacturer’s criteria were achieved for fibrinogen on ACL TOP 300 for both concentration levels but not for pathological concentration on ACL TOP 500 according to EP-15 protocol. For PT on ACL TOP 500 both protocols for both concentrations obtained manufacturer’s criteria while CVs (2.9% and 3.6%) obtained with the verification protocol EP-15 on ACL TOP 300, were slightly above the manufacturer’s criteria (2.2% and 3.1%).

Conclusion:  Results showed the difference between the applied verification protocols for some analysis that has to be considered during the precision assessment.

e-mail: jelisavaccosic [dot] sanda [at] gmail [dot] com 



QConnect run control result monitoring with QConnect software in blood donors NAT screening

Ivana Babić, Margareta Maslović, Jasna Bingulac-Popović, Vesna Đogić, Nina Juraković-Lončar, Melita Balija, Irena Jukić

Croatian Institute of Transfusion Medicine, Zagreb, Croatia


Introduction: Nucleic acid testing (NAT) screening procedure for blood born viruses in Croatian blood donors includes routine usage of external run controls to monitor daily testing variation. QConnect software through EDCNet enables real-time graphing and analysis of controls data using robust peer group data. Evaluation of NAT run controls results with EDCNet enables noticing deviating result/trend in performance of testing depending on lot reagents, run control (HBV-DNA, HCV-RNA or HIV-1 RNA positive) and instrument used .

Materials and Methods: Blood donors NAT screening in Croatia is performed on individual donation by using a Procleix Ultrio Plus test on Tigris instruments (Grifols, Spain). The test simultaneously detects HBV-DNA, HCV-RNA and HIV-1 RNA based on transcription mediated amplification (TMA). We use QConnect HBVDNA+, HCVRNA+ and HIV-1 RNA+ control (Life Technologies LTD, UK) to monitor daily performances of three Tigris instruments. NAT negative result for positive run control is causing invalidation of the run. From June, 2014 we started with daily data input for run controls lot, test results-s/co values, reagent lot and used instrument, in EDCNet (NRL, Australia). The data are being statistically processed and displayed as Levey-Jenning chart and Mean/Scatter report with mean, ±2SD values for our peer group (eight laboratories) and our laboratory in particular.

Results: Mean values of our results are very similar to those of our peer group. QConnect HIVRNA+ has highest ±2SD value (12.06±1.64) and HCVRNA+ lowest (9.33±0.82). Approximately 4-5% of our results are outliers that do not show any trend.

Conclusion: Through EDCNet we are able to be part of global control chart of our peer group and to evaluate our results in broader context. By enrolment of more users in our peer group we will be able to obtain even more accurate data on our routine work control.

e-mail: ivana [dot] babic [at] hztm [dot] hr



Comparison of quality of laboratory processes for Troponin I determination

Adriana Unić, Lovorka Đerek, Nevenka Stančin

Clinical Department of Laboratory Diagnostics, University Hospital Dubrava, Zagreb, Croatia


Introduction: Laboratory test results have major impact on clinical decision-making, which makes the quality of all laboratory processes extremely important. Therefore, it is necessary to objectively evaluate each phase of overall laboratory process. Six Sigma quality system enables the comparison of quality indicators and information on unacceptable processes which require further improvement. The aim of the study was to assess and compare the quality of individual process for determination of Troponin I (TnIDx, Beckman Coulter Inc.) in the Clinical Department of Laboratory Diagnostics, UH Dubrava in the six months period.

Materials and Methods: To assess the quality of preanalytical phase the number of hemolyzed samples was observed. To assess the quality of analytical phase the internal quality control precision (three concentration levels) and bias (from the external quality control program) data were analyzed. Desirable specifications based on biological variability were used as an acceptance criteria (TEa<27.91%). To assess the quality of postanalytical phase Turnaround time (TAT) was used. The acceptance criteria for issuing Troponin I results is 60 minutes. Sigma values were calculated using calculators available on

Results: The percentage of unacceptable hemolytic samples for Troponin I determination was 3.7%, Sigma value 3.3. In that period, Sigma values for analytical performance on the three concentration levels of troponin I 0.3 g/L, 1.9 g/L and 5.8 g/L were 1.8, 3.0 and 5.3 respectively. 12.71% results of Troponin I were issued outside the established criteria for TAT, Sigma value 2.7.

Conclusion: Although all laboratory processes require improvement and carefully designed strategy of quality control, postanalytical phase had the lowest sigma value and thus requires additional improvement of its quality. Generally, laboratories are mainly focused on improving the quality of preanalytical and postanalytical processes but the quality of the analytical process, despite the significant progress, still leaves room for improvement.

e-mail: adriana [dot] unic [at] gmail [dot] com 



An estimation of IgA anti-tTG cut-off specific for celiac disease screening of diabetic children

Merica Aralica, Jasminka Matica

Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia


Introduction: Diabetic children are prone to celiac disease (CD) due to autoimmune feature of both diseases. According to current clinical guidelines IgA anti tissue transglutaminase antibody (IgA anti-tTG) is recommended screening test for CD in every suspected or high risk case. In routine laboratory practice single IgA anti-tTG cut-off (producer provided) is reported for all screened individuals regardless their clinical status. In cohort of diabetic children we tested such IgA anti-tTG cut-off to estimate its utility for this specific group of pediatric patients.

Subjects and Methods: From 2002-2014, total of 149 diabetic children (ages <18 years) were screened by IgA anti-tTG using Enzyme Linked Immunoassay (Euroimmun, Germany); cut-off 20 RU/ml (all genders and ages). Children had previously diagnosed diabetes mellitus type I and no one had total serum IgA deficiency. All children with positive results underwent the bowel biopsy if they missed to turn out negative in period of nine months after initial positive screening result and regular gluten containing diet. Statistical analysis was done in MedCalc (Mariakerke, Belgium).

Results: IgA anti-TTG was positive in 14 diabetic children following positive biopsy results. One child had negative IgA anti-tTG but positive biopsy finding. Among all CD free diabetic children (N=134) false positive results of IgA anti-tTG had 15 of them. ROC curve analysis showed a cut-off of 34.5 RU/ml in screened cohort; an area under the ROC curve of 0.934 with 95%CI 0.882-0.968 and P<0.05. At cut-off of 34.5 RU/ml IgA anti-tTG sensitivity and specificity are 93.3% (95%CI 68.1-99.8) and 91.0% (95%CI 84.9-95.3).

Conclusion: In screened cohort, ROC analysis proposed higher IgA anti-tTG cut-off with high sensitivity and specificity. Providing specific IgA anti-tTG cut-off for diabetic children may improve its diagnostic accuracy and clinical utility in sense of avoiding retesting patients with initial low positive screening results.

e-mail: merica [dot] aralica [at] gmail [dot] com



Macroprolactin – to screen or not to screen?

Adriana Bokulić , Ivana Zec, Valentina Vidranski, Željka Bukovec Megla, Iva Petek Tarnik, Iva Petek

Department of Oncology and Nuclear Medicine, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

Introduction: Monomeric prolactin is the most common form of circulating prolactin in healthy individuals, but forms with higher molecular mass are also present, such as big prolactin and big-big prolactin or macroprolactin. Macroprolactin is an important source of immunoassay interference and often leads to misdiagnosis and mismanagement of hyperprolactinemic patients. Precipitation with polyethylene glycol (PEG) is widely used to detect the presence of macroprolactin. Our aim was to assess macroprolactin incidence in hyperprolactinemic adults and to decide on its screening policy.

Subjects and Methods: From August 2014 to February 2015, all adult patients’ samples with prolactin levels above corresponding upper reference limit (male 381 mIU/L, female 496 mIU/L) were tested for macroprolactin (N=380, median age 37, min 19, max 84). Pre-PEG and post-PEG prolactin measurements were performed with electrochemiluminescence method on Cobas e601 analyzer (Roche, Mannheim, Germany). Manufacturer’s recommended procedure for PEG participation was used.

Post-PEG results within their corresponding reference range were considered as falsely hyperprolactinemic (male 63-245 mIU/L, female 75-381 mIU/L). Chi-square test was used to test for possible sex-dependent difference in incidence (a=0.05)

Results: In 343 (90%) samples, post-PEG results were above post-PEG reference range (57 male and 286 female). In 37 (10%) samples, post-PEG results fell within post-PEG reference range (5 male, 32 female) and they could be classified as falsely hyperprolactinemic. There is no sex-dependent difference in incidence (P=0.802).

Conclusion: Results show that 10% of hyperprolactinemic patients are falsely classified which could lead to unnecessary investigation, incorrect diagnosis and inappropriate treatment. As a result of this study, our laboratory routinely screens all hyperprolactinemic samples for the presence of macroprolactin, using gender appropriate upper limit as a cut-off. Screening excludes requests repeated within 6 months, unless they were, as a result of previous testing, classified as falsely hyperprolactinemic.

e-mail: adriana [dot] bokulic [at] gmail [dot] com


B07 (Usmeno izlaganje)

Izračun slobodnog testosterona: usporedba dvije formule

Tihana Pavošević, Iva Lukić, Sanja Mandić, Vesna Horvat,
Vatroslav Šerić

Odjel za kliničku laboratorijsku dijagnostiku, Klinički bolnički centar Osijek, Osijek, Hrvatska


Uvod: Testosteron je steroidni hormon koji je ~97% u cirkulaciji specifično vezan za hormon koji veže spolne hormone (SHBG) i nespecifično za albumine, a samo mali dio je nevezan (slobodan). Slobodni testosteron (fT) biološki je aktivan oblik testosterona, a može se mjeriti direktnom metodom ili indirektno računskim metodama. Cilj ovog rada je usporediti dvije različite formule za izračun fT: (1) pomoću testosterona i SHBG-a i (2) pomoću testosterona, SHBG-a i albumina.

Materijali i metode: Analizirani su uzorci 172 ispitanika (58 muškaraca i 114 žena) kojima je zatražena analiza fT u Odjelu za kliničku laboratorijsku dijagnostiku KBC-a Osijek. U uzorcima seruma su izmjereni ukupni testosteron i SHBG imunokemijskom (CMIA) metodom na analizatoru Architect i1000SR, te albumini fotometrijskom metodom na analizatoru Olympus AU680. Pomoću navedenih parametara izračunat je postotak fT dvijema različitim metodama: (1) iz testosterona i SHBG-a i (2) iz testosterona, SHBG-a i albumina. Ispitanici su podjeljeni u skupine prema spolu, te je napravljena usporedba dobivenih rezultata Passing-Bablock regresijskom analizom u MedCalc programu, verzija (MedCalc Software, Mariakerke, Belgija).

Rezultati: U skupini muškaraca prvom formulom bilo je 43% pacjenata čije su vrijednost fT bile iznad referentnog intervala, dok je u skupini žena takvih bilo 14%. Drugim izračunom u skupini muškaraca taj broj bio je znatno manji i iznosio je 10%, a u skupini žena ostao je isti. Passing-Bablock regresijskom analizom dobiveno je da u skupini žena nema razlike između ova dva načina izračuna (y=0,0000(0,0000-0,0000)+1,0000(1,0000-10000)x), dok je u skupini muškaraca utvrđeno da razlika postoji, te da je ona konstantna i neproporcionalna (y=-1,5000(-2,4500-(-1,0600))+2,0000(1,8000-2,5000)x).

Zaključak: Premda se u rutinskoj praksi uglavnom koristi izračun fT pomoću testosterona i SHBG-a, ispitanicima s vrijednostima fT izvan referentnog intervala trebalo bi izračunat fT formulom koja koristi i albumine, naročito kada su u pitanju muškarci.

e-adresa: tihanapavos [at] gmail [dot] com


C - Harmonization in laboratory medicine


Insight in antinuclear antibodies (ANA) determination in Croatia – survey results of Croatian Society of Medical Biochemistry and Laboratory Medicine (CSMBLM) working group for guidelines in laboratory diagnosis of autoimmune diseases

Lovorka Đerek1, Andrea Tešija Kuna2, Ana Kozmar3, Vedrana Drvar4

1Clinical Department for Laboratory Diagnostics, University Hospital Dubrava, Zagreb, Croatia

2University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

3Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

4Clinical Department for Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia


Introduction: The increase in the prevalence of autoimmune diseases is followed by an increase in humoral immunodiagnostics in routine laboratory work. The objective was to gain insight into a number and type of Croatian laboratories that are performing humoral diagnostics, methodology and diagnostic algorithms that are used in ANA determination.

Materials and Methods: A survey was created and published using SurveyMonkey application. It was sent on 26/9/2014 to 88 laboratories in secondary, tertiary health care and private institutions, and closed on 22/04/2015.

Results: Of 80 laboratories that answered the questionnaire 33 perform humoral immunodiagnostics: 16 in general/county hospitals, 13 in clinical hospitals/clinical hospital centers, 13 in private institutions, 5 in specialized hospitals. ANA is determined in 16/33 laboratories. ANA screening is performed using indirect immunofluorescence (IIF)/Hep-2 in 7/16, fluoro enzyme immunoassay (FEIA) in 5/16, line immunoassay (LIA) in 1/16 and enzyme immunoassay (ELISA) in 3/16 laboratories. Anti-dsDNA antibodies are determined in 15 laboratories using: FEIA (6/15), ELISA (5/15), multiplex immunoassay (MIA) (3/15) and IIF (1/15) methods. Antibodies to extractable nuclear antigens (ENA) are determined in 13 laboratories using: ELISA 4/13, FEIA 4/13, MIA 3/13, LIA 1/13 and FEIA+LIA 1/13. Antibodies to centromere (CENP), histones and nucleosomes are determined sporadically. For ANA/IIF 5/7 laboratories determine titer and type of fluorescence. If the ANA screening test is positive: 6/14 laboratories do not determine specificity, 2/14 laboratories determine every ordered antibody, 4/14 determines antibody specificity depending on ANA titer and 2/14 depending on the type of fluorescence. When another screening method is used (beside IIF), 5/10 laboratories report covered antigens.

Conclusion: Results indicate the need of creating recommendations and algorithms in order to harmonize the approach to laboratory diagnostics of systemic autoimmune diseases in Croatia.

e-mail: vedranadrvar [at] gmail [dot] com


D – Markers of kidney and urinary tract diseases


Frequency of microhematuria in men older than 35

Zorana Todorić1, Suzi Žlabravec2, Dunja Turner2, Mirjana Sikirica2

1Polyclinic LabPlus, Split, Croatia

2Polyclinic LabPlus, Zagreb, Croatia


Introduction: The aim of the project was to estimate the frequency of microhematuria in men older than 35 in the polyclinic LabPlus.

Subjects and Methods: One hundred samples of fresh urine (men older than 35) were used to determine number of erythrocytes (Erc) by automated system LabUMat/UriSed (77 Elektronika Kft, Budapest, Hungary). System works by a refractometry principle (dipstick) and manual microscopy (sediment) comparatively to light microscopy of a wet sample which is supravitally coloured urine sediment, prepared in accordance with recommendations of the European group for urine analysis.

Results: 23/100 patients had positive results in erythrocytes/hemoglobin (Erc/Hb) on a dipstick. 3/23 results on a dipstick had 10 Erc/Hb/µL and negative in the high power field (HPF); 4/23 had a dipstick results 10-50 Erc/Hb/µL and 2-5 Erc/HPF with presence of calcium oxalate; 3/23 had 50-300 Erc/Hb/µL and 1-4 Erc/HPF with creatine kinase activity and proteinuria (<0.3 g/L). 4/23 patients had a dipstick result 10-50 Erc/Hb/µL and 1–3 Erc/HPF with PSA 2-4 µg/L. 4/30 patients had 10 Erc/Hb/µL and 1-5 Erc/HPF with leukocyturia. 9/23 patients had 10–50 Erc/Hb/µL and 1-3 Erc/HPF without any other pathological results.

Conclusion: Microhematuria is almost always accidental and often neglected result. It can be a consequence of various benign but also malignant diseases. The first indicator of neoplasma urothelium is isolated microhematuria. According to American Urological Association (AUA) more than 3 Erc/HPF is first indicator for further analysis and it is necessary to confirm positive results in three random samples. If microhematuria is still present, it is necessary to find the origins of the erythrocytes. Dysmorphic erythrocytes need further nephrological and native erythrocytes need urological analysis. Results show that 10% of patients with microhematuria demand further analysis and treatment.

e-mail: zorana [dot] todoric [at] gmail [dot] com


E – Molecular diagnostics


E01 (Oral presentation)

The influence of CYP3A4/5 and CYP2D6 polymorphisms on serum concentrations of risperidone and 9-hydroxyrisperidone in patients using long-acting injectable risperidone

Lana Ganoci1, Mila Lovrić1, Maja Živković2, Marina Šagud3, Nada Božina1

1Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

2Psychiatric Hospital Vrapče, Zagreb, Croatia

3Department of Psychiatry, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Risperidone (RIS) is metabolized to its active metabolite 9-hydroxyrisperidone (9-OHRIS), mainly by the enzyme CYP2D6 and, to a lesser extent by CYP3A4/5. Its antipsychotic effect is assumed to be related to the active moiety of RIS and 9-OHRIS. The aim of this study is to investigate the role of genetic variations of CYP2D6 *dupl, *3, *4, *5, *6, *41, CYP3A4*22 and CYP3A5*3 on serum concentrations and metabolic ratio of RIS and 9-OHRIS.

Subjects and Methods: 70 patients with schizophrenia treated with long-acting risperidone (RIS-LAI) (dose range 25-75 mg) were genotyped, and their serum steady-state concentrations of RIS and 9-OH RIS were measured on 5th and 14th day. Analysis for CYP2D6 *dupl and allele *5 was performed by PCR, while real-time PCR analysis by TaqMan® assays was performed for genotyping of CYP2D6 alleles *3, *4, *6, *41, CYP3A4*22 and CYP3A5*3. Concentrations of RIS and 9-OH RIS were measured by HPLC-DAD.

Results: The genotype distribution for CYP2D6 was: 34 extensive metabolizers (EM), 23 intermediate metabolizers (IM), 5 poor metabolizers (PM), 3 ultrarapid metabolizers (UM) and 5 duplications of variant alleles. The median active moiety concentrations were on 5th day 76.4 nmol/L (95%CI=56.8-102.2), on 14th day 42.3 nmol/L (95%CI=34.1-52.4). The active moiety concentrations on 14th day were significantly different according to CYP2D6 genotype (Kruskall-Wallis test, P=0.042). Genotyping of CYP2D6*41 and metabolic ratio RIS/RIS-OH revealed 3 more patients to be PM and 3 IM. No significant influence of CYP3A4*22 and CYP3A5*3 variants was found.

Conclusion: The CYP2D6 genotypes had a strong influence on the steady-state serum levels of RIS and 9-OHRIS. Genotyping of CYP2D6*41 is important for detection of additional CYP2D6 PM and IM. The active moiety concentrations for UM on 5th and 14th day were bellow recommended therapeutic range. CYP3A4*22 and CYP3A5*3 did not show influence on the steady-state serum drug concentrations.

e-mail: lana [dot] pejnovic [at] gmail [dot] com


Genotyping of interleukin 28 b polymorphysm rs12979860 and prediction of viral response to therapy in chronic hepatitis C patients

Jasna Bingulac-Popović1, Nadia Komparić2, Ivana Furčić3, Vesna Đogić1, Ivana Babić1, Jadranka Žgrablić-Cetina2, Ivana Babić2, Irena Hrstić2, Nina Juraković-Lončar1, Melita Balija1, Irena Jukić1

1Croatian Institute of Transfusion Medicine, Zagreb, Croatia
2General Hospital Pula, Pula, Croatia
3Institute for Anthropological Research, Zagreb, Croatia


Introduction: Genetic polymorphisms in the interleukin 28B (IL28B) gene has been found to predict spontaneous clearance or successful treatment of HCV infection. Patients with CC genotype of rs12979860 IL28B SNPs are more likely to eliminate HCV than those with TT or CT genotype. Patients with CC genotype were twice as likely to achieve an SVR compared with patients with other genotypes. Aim of the study was to determine the virological response to therapy in patients with chronic HCV infection due to the IL28B genotype polymorphism.
Subjects and Methods: The patients group from General Hospital Pula are divided into two groups: 1- on dual therapy with pegylated interferon alfa-2a + ribavirin (N=9) and 2- on triple therapy with NS3/4A protease inhibitors (N=13). After isolation of the genomic DNA, IL28B genotype was determined using “in house” validated RT-PCR on an ABI 7500 real time PCR system. Viral load and genotype were determined prior to treatment using commercial kits on the COBAS AmpliPrep/COBAS TaqMan analyzer.
Results: In the group of the patients on dual therapy with good respond to therapy, 5/9 were CC, 3 CT and 1 TT genotype. Of the 13 patients on triple therapy, 3 of them were CC, 8 CT and 2 TT genotypes as well. 5/13 patients with viral genotype 1; subtypes 1a and 1b were excluded from therapy due to poor response (1 CC, 4 CT). Overall, 5/13 patients were non-responders, of which 1/5 IL28B was CC genotype (20%) and the remaining 4/5 (80%) patients were non-CC genotype.
Conclusion: Preliminary results show slightly better response to therapy in chronic HCV patients carriers of CC IL28B genotype polymorphism. The study continues with including more patients to the triple therapy in Croatia, which will enable significant statistical comparison.

e-mail: jasna [dot] bingulac [dot] popovic [at] hztm [dot] hr

F - Automatization of procedures and processes in laboratory medicine


Malaria detection on DXH800 Beckman Coulter® hematology analyzer in non-endemic area

Sanja Kozić Dokmanović1, Valentina Vidranski2, Vedrana Jukić1, Zdravka Čulig1, Martina Kramar1, Adrijana Dorotić1, Renata Laškaj1, Anita Klasić3

1Department for biochemistry and hematology, University hospital for infectious diseases “Dr Fran Mihaljevic”, Zagreb, Croatia

2Department of Oncology and Nuclear Medicine, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

3Special Hospital for Medical Rehabilitation Krapinske Toplice, Krapinske Toplice, Croatia

Introduction: The prevalence of malaria infection in Croatia is low and usually associated with traveling to endemic areas. A microscopic examination of blood smear is the gold standard for malaria parasite detection, but time consuming and requiring a well trained laboratory personnel. Hematology analyzers that use VCS (V-Volume, C-Conductivity, S-Scatter) technology, as DxH800 Beckman Coulter®, may detect lymphocytes and monocytes volume changes caused by malaria infection. In 2006, Briggs C. et al introduced the «malaria factor», a number calculated from lymphocyte and monocyte volume SD. During malaria infection, it should be greater than 3.7. Additional parameters as platelet count <150×10/L, eosinophil percentage <0.15%, presence of additional peak at threshold of WBC histogram, SD volume of monocytes >23.2 and a mean volume of monocytes >180 improve specificity.

Materials and Methods: We retrospectively collected results of CBC analysis for 12 patients with proved malaria infection.

Results: Malaria factor >3.7 was found in 10/12 patients (median 5.9, IQR: 5.1-7.0), platelet count <150×109/L was found in 8/12 patients (median 99, IQR: 62-171×109/L), eosinophil percentage <0.15% was found in 4/12 patients (median 0.2, IQR: 0.1-1.2), presence of additional peak at threshold of WBC histogram was found in 9/12 patients, SD volume of monocytes >23.2 was found in 9/12 patients (median 26.5, IQR: 23.3-28.9), mean volume of monocytes >180 was found in 9/12 patients (median 191, IQR: 187.5-196.5).

Conclusion: We showed that DxH800 Beckman Coulter® can detect the most changes in CBC caused by malaria. Data are generated during routine CBC analysis with no additional cost and possibility to create a flag to warn the operator for possible malaria infection. However, each positive result has to be confirmed by microscopic examination.

e-mail: sanja [dot] kozic1 [at] gmail [dot] com

F02 (Oral presentation)

Capillary electrophoresis for routine determination of HbA1c in patients with diabetes: Short analytical verification and preliminary clinical results

Sandra Božičević, Vanja Radišić Biljak, Maja Krhač,
Marijana Vučić Lovrenčić

Department of Medical Biochemistry and Laboratory Medicine, Merkur University Hospital, Zagreb, Croatia


Introduction: HbA1c is a crucial laboratory test for clinical management of diabetes and the risk assessment for the development of micro- and macrovascular complications. Expanding indications and limitations of analytical methods represent a continuous challenge in meeting the rigorous requirements of analytical quality for reliable clinical use. Capillary electrophoresis, one of the two reference methods for HbA1c, recently available for routine laboratory work, certainly enables a qualitative shift in analysis of HbA1c. The aim of this preliminary study is to determine the analytical features and possible clinical benefits of HbA1c results measured by capillary electrophoresis compared to immunoassays.

Materials and Methods: Analytical features of the system for measuring concentration of HbA1c by capillary electrophoresis (MiniCap Flex Piercing, Sebia, France) are verified in accordance with CLSI EP15-A2 protocol. Comparison of the results of HbA1c measured by capillary electrophoresis and accredited Immunoassay on the automatic analyzer (Tina-quant HbA1c Gen 2, Cobas Integra 400+, Roche Diagnostics, USA) was performed on samples of diabetic patients (N=78) within the clinically relevant concentration range of HbA1c.

Results: The total imprecision of HbA1c measured by capillary electrophoresis was 1.02/1.62% (HbA1c=5.0%/30.6 mmol/mol) and 1.24/1.55% (HbA1c=8.0%/64.3 mmol/mol). Bland-Altman analysis showed the average deviation of the measured values of HbA1c by capillary electrophoresis in relation to immunoassays -0.24%/- 2.8 mmol/mol (95%CI: (-0.16)-(-0.32)%/(-3.8)-(-1.7) mmol/mol), with continued significant (P<0.001) increase in bias. Passing-Bablok analysis confirmed the systematic and proportional difference between the methods: HbA1c(%)capillary electrophoresis=0.8375+0.8750×HbA1c(%)immunochemistry; intercept A=0.85; slope B=0.87; HbA1c(mmol/mol)capillary electrophoresis=5.44+0.88×HbA1c(mmol/mol)immunochemistry; intercept A=5.44; slope B=0.88.

Conclusion: Capillary electrophoresis meets the rigorous criteria of the analytical quality for reliable clinical use of HbA1c. The concentration dependence of the observed difference in the values of HbA1c is probably a consequence of enhanced specificity of capillary electrophoresis compared to immunochemistry methods. Potential clinical implications of the observed differences need to be brought in future research.

e-mail: vanja [dot] radisic [at] gmail [dot] com

F03 (Oral presentation)

Evaluation of rules in autovalidation algorithm

Vladimira Rimac1, Krešimir Kuleš2, Željka Vogrinc1, Dunja Rogić1

1Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

2IN2 d.o.o, Zagreb, Croatia


Introduction: Autovalidation is a part of the laboratory information system whereby laboratory results are released without manual human intervention through defined rules and algorithms. Autovalidation rules may include: analytical measurement range, interference indices (hemolysis, icterus, lipemia), critical values or delta checks. Rules must be set for laboratory testing and patient population for which autovalidation will be applied. The aims of the study were to examine the credibility of the rules in the algorithm for autovalidation and determine the percentage of samples that were autovalidated in comparison to the total number of samples included in validation, as well as rules that stop autovalidation.

Materials and Methods: The validation results included 9805 samples of biochemical tests analyzed in the Department of Laboratory Diagnostics, University Hospital Centre Zagreb from May to July 2014. Validation was performed in such a way that, before starting the system of autovalidation, the results were reviewed by a medical biochemist who recorded samples that should meet the rules set in the algorithm. Statistical analysis was made using the software MedCalc (version

Results: 78.3% (7677) of the samples included in validation were autovalidated. The highest percentage of samples (54.9%) was not autovalidated due to set rules for analytical measurement ranges, while the lowest percentage of non-validated samples was recorded in the rules for interference indices for icterus (0.6%). The correspondence of autovalidation and the person who carried it out was 99.5%, i.e. the mismatch is observed only in 0.5% of samples (38). An analysis of Χ2-test data showed no statistically significant difference (P=0.523) in the number of samples that were autovalidated (7677) in relation to those validated by medical biochemist (7639).

Conclusion: The analysis of results showed that set rules in the algorithm are credible, and that system autovalidation can be implemented in routine laboratory use.

e-mail: kutnjakvl [at] gmail [dot] com


OptiScanner 5000: an intermittent glucose monitoring in septic patients

Alessandra Barassi1, Michele Umbrello2, Valentina Salice3, Paolo Spanu2, Paolo Formenti2, Luca Massaccesi4, Giancarlo Goi4, Clara Anna Linda Damele5, Angela Leone5, Rossana Stefanelli5, Gaetano Iapichino2,3, Gian Vico Melzi d’Eril1

1Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milano, Italy

2Unità Operativa di Anestesia e Rianimazione, Azienda Ospedaliera San Paolo - Polo Universitario, Milano, Italy

3Dipartimento di Fisiopatologia Medico-Chirurgica e dei Trapianti, Università degli Studi di Milano, Milano, Italy

4Dipartimento di Scienze Biomediche, Chirurgiche ed Odontoiatriche, Università degli Studi di Milano, Milano, Italy

5Laboratorio di Analisi, Ospedale San Paolo, Milano, Italy


Introduction: The optimal level and modality of glucose control in critically ill patients is still debated. A protocolized approach and the use of nearly-continuous technologies are recommended to manage hyperglycemia, hypoglycemia and glycemic variability. We recently proposed a pato-physiology-based glucose control protocol which takes into account patient glucose/carbohydrate intake and insulin resistance. Aim of the present investigation was to assess the performance of our protocol with an automated intermittent plasma glucose monitoring device (OptiScanner™ 5000).

Materials and Methods: OptiScanner™ was used in 6 septic patients, providing glucose measurement every 15 minutes from a side-port of an indwelling central venous catheter. Target level of glucose was 4.4-8.3 mmol/L. Insulin infusion and kcal with nutritional support were also recorded.

Results: Six septic patients were studied for 319 h (1277 measurements); 58 (45-65) hours for each patient (measurements/patient: 231 (172-265)). Blood glucose was at target for 93 (90-98)% of study time. Mean plasma glucose was 7.0±0.6 mmol/L. Only 3 hypoglycemic episodes (4.3, 4.3, 3.8 mmol/L) were recorded. Glucose variability was limited: plasma glucose coefficient of variation was 11.7±4.0% and plasma glucose standard deviation was 0.8±0.3 mmol/L.

Conclusion: The local glucose control protocol achieved satisfactory glucose control in septic patients along with a high degree of safeness. Automated intermittent plasma glucose monitoring seemed useful to assess the performance of the protocol.

e-mail: alessandra [dot] barassi [at] unimi [dot] it

G – Accreditation, organization, quality and laboratory management


Quality indicators in clinical laboratory hematology

Dragica Ferenec Ružić, Biserka Getaldić, Sandra Margetić, Ivana Vuga, Nada Vrkić

Department of Laboratory Hematology and Coagulation, University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: Defining and continuous monitoring of quality indicators is obligatory for laboratories accredited according to the HRN EN 15189. Measurement of indicators should include different phases of laboratory work in which, according to the specificity of diagnostic area, selected nonconformities are appropriate for quality self-assessment, when counted and evaluated enable process improvement. The aim was self-assessment of the quality in the hematology laboratory, with integrated organization of emergency and routine samples, by monitoring three key nonconformities within one year: 1. number of clotted samples, 2. number of repeated measurements due to analytical errors, 3. number of emergency CBC samples, which exceeded TAT.

Materials and Methods: Nonconformities of pre-analytical, analytical and post-analytical phase were recorded by the Department of Laboratory Hematology and Coagulation. CBC was measured on hematology analyzers DxH (Beckman Coulter) and Cell-Dyn Sapphire (Abbott). Data on the selected indicators of the working process were obtained by exploring a nonconformities database in Microsoft Access.

Results: In one year period (1.1.2014.-31.12.2014.) 124602 CBC reports were issued. Records show that 918 repeated sampling were requested (0.7%), of which the majority was clotted samples 770 (0.6%). Measurement was repeated for 1091 samples (0.9%) out of which 509 (46%) for checking differential leucocyte count. 94974 (76.2%) of the total number of issued CBC reports were emergency requirements, of which 685 samples (0.7%) monthly exceeded TAT, while the annual average TAT value for CBC was 17.8 minutes.

Conclusion: According to the quality indicators, our results show a good proportion of acceptable samples considering that 15654 of analyzed samples were capillary samples. Initial analyses of the hematology analyzers have eligible criteria for efficiency and average time for issuing CBC emergency report. The improvement will be established on the analysis of the monthly dynamics of these indicators and will speed up the management of recorded errors.

e-mail: dfruzic [at] gmail [dot] com 


G02 (Oral presentation)

Identification of laboratory processes with major influence on laboratory turnaround time

Vesna Šupak Smolčić1,2, Dragana Antončić1, Ivana Vladilo1, Eliza Bašić1, Diana Jamaković1, Dubravka Lenac1, Jasna Marčelja1, Snježana Salopek1, Elizabeta Fišić1, Lidija Bilić-Zulle1,2

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia


Introduction: Time reduction for issuing laboratory results accompanied with quality improvement presents a major challenge in laboratory work. We aimed to identify processes with major influence on turnaround time (TAT).

Materials and Methods: From January till March 2015, we manually recorded duration of several processes in laboratory routine work for 1432 samples arrived in the Department of laboratory diagnostics, CHC Rijeka. Sample reception was divided in 5 intervals: R1 (7:00-8:00h), R2 (8:00-9:00h), R3 (9:00-10:00h), R4 (10:00-11:00h), R5 (11:00-12:00h). Duration (minutes) of the following processes was recorded: from sample reception to centrifugation (D1), from centrifugation to manual barcoding (D2), from barcoding to placing the sample into analyser (D3), time on analyser (D4), from the end of analysis to archiving (D5) and overall process duration (TAT) (D6). Difference of process duration regarding different time of receipt were tested using Kruskal-Wallis test (level of significance P<0.05).

Results: The majority of samples were received in R3 (25%). Average duration (minutes) of each process (median, interquartile range) was as follows: D1=25(20-30); D2=15(15-20); D3=30(15-50); D4=20(15-24); D5=12(10-20); D6=115(100-135). Sample reception (D1) lasts equally for intervals R1 to R5 (P=0,122). Process D2 is the fastest, 15(10-15), for samples received in R4 (P=0,004). Duration of D3 and D4 is significantly longer for samples received during R1 (D3=40(20-59); D4=20(14-25)) and R2 (D3=30(15-50); D4=20(15-24)) when compared with other intervals (both P<0.001). Process D5 is longer for R1, R2 and R3 in comparison to R4 and R5 (P<0.001). TAT (D6) is the longest for samples received in R1, 125(105-142), following R2, 115(95-130), while in other intervals remains the same (P<0.001).

Conclusion: From 9:00-10:00 am laboratory sample reception is burdened the most. Samples received from 7:00-9:00 am are processed the slowest. Critical points are manual barcoding and time for placing the sample in the analyser. Personnel management optimization or automatization is needed for TAT reduction.

e-mail: vesnasupak [at] gmail [dot] com


Number of requests required but not registered (missed tests) – preanalitycal quality indicator

Vesna Šupak Smolčić1,2, Dragana Antončić1, Lidija Bilić-Zulle1,2

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia


Introduction: Quality indicators are objective way for accessing the quality of laboratory work. One of the preanalytical quality indicators is the percentage of tests that are required but not primarily requested i.e. requests for additional analysis of the samples already received and analyzed in the laboratory. We aimed to establish this quality indicator and to identify departments which frequently make additional requests.

Materials and Methods: Added test requests were recorded during 2014 in the Department of laboratory diagnostics (location Rijeka), Clinical Hospital Centre Rijeka. Data was analyzed according to Clinical departments and Clinics which requested additional analysis. Quality indicator was defined as percentage of requests for additional analysis of the samples already analyzed out of total number of requests for each Clinical department and Clinic in 2014. Sigma value was calculated using web calculator ( Sigma >3 was considered as acceptable.

Results: In 2014 we received a total of 107016 requests for laboratory analysis from all Clinical departments and Clinics in CHC Rijeka, location Rijeka. The total number of requests for additional analysis was 1356 (1.3%) corresponding to a sigma value of 3.8. Departments that make requests for additional testing most frequently are: Emergency department (1113/16187, 6.9%, sigma 3.0); Clinic for Infectious Diseases (31/3451, 0.9%, 3.9), Department of Gastroenterology (62/11618, 0.5%, sigma 4.1), Clinic for Psychiatry (5/958, 0.5%, sigma 4.1), Clinic for Surgery (14/4882, 0.3%, sigma 4.3), Department of Hematology (41/12859, 0.3%, sigma 4.3), Clinic for Gynecology (35/11752, 0.3%, sigma 4.3). Clinical departments and Clinics with sigma >4.5 are not shown.

Conclusion: Despite the seemingly low percentage of additional requests for laboratory analysis, sigma values from 3.0 to 4.5 for the seven Clinical departments and Clinics leaves significant room for improvement.

e-mail: vesnasupak [at] gmail [dot] com



Verification of immunoassay for determination of anti-Mullerian hormone

Danijela Županić, Renat Mujagić, Lorena Honović

Department of Laboratory Diagnostics, General Hospital Pula, Pula, Croatia


Introduction: Anti-Mullerian hormone (AMH) is produced by the ovaries and testes. Its role is to make a reserve of oocytes and regulate the transition of follicles into a mature phase in women and the normal development of the reproductive organs in men. The aim of this paper is to investigate the acceptability of AMH reagent (Roche Diagnostics GmbH, Mannheim, Germany) for the application in immunochemical analyzer Cobas e411 to ensure reliable results of the laboratory test. Verification of immunoassay was performed using the CLSI/NCCLS procedure EP 15-A2.

Materials and Methods: We evaluated the following parameters: Precision in which we determined repeatability, interprecision and overall laboratory precision and systematic error (deviation from the expected values). We used Roche control samples (PC-AMH 1 and PC-AMH 2).

Results: In the evaluation of precision the following results were obtained: For control sample level 1 (declared mean value is 7.07 pmol/L), mean value of 7.3 pmol/L, value for repeatability (Sr=0.07, CV%=1.0%), interprecision (Sb=0.08, CV%=1.1%) and overall laboratory precision (Sl=0.1, CV%=1.4%). For control sample level 2 (declared mean value is 38.6 pmol/L), mean value of 39.5 pmol/L, value for repeatability (Sr=0.7, CV%=1.8%), interprecision (Sb=0.53, CV%=1.3%) and overall laboratory precision (Sl=0.78, CV%=2.0%). The resulting systematic error for control sample level 1 is 0.184 (2.6%), while for level 2 it is 0.559 (1.4%).

Conclusion: Based on the results of this study, we can conclude that tested Roche reagent AMH is accetable and we recommend its use in immunochemical analyzer Cobas e411.

e-mail: danijela [dot] zupanic [at] pu [dot] t-com [dot] hr



Molecular viral diagnostics - external control results 2010 - 2014

Jasna Bingulac-Popović, Vesna Đogić, Ivana Babić, Tomislav Vuk, Dorotea Šarlija, Nina Juraković- Lončar, Melita Balija, Irena Jukić

Croatian Institute of Transfusion Medicine, Zagreb, Croatia


Introduction: External Quality Assessment were applied in the Molecular Diagnostics department (MDD) of CITM regularly since 1997 for molecular diagnostics of viral hepatitis B and C, and HIV-1. The aim is to present the results of external control from 2010-2014.

Materials and Methods: MDD participates in the Quality Control for Molecular Diagnostics (QCMD, Scotland) for viral loads of HCV, HBV, HIV-1 and HBV, HCV genotypes. Annually were tested 4 panel samples for viral load and 1 for genotyping. The viral load is determined by RT-PCR commercial kits on the COBAS AmpliPrepTaqMan docking station (Roche Diagnostics, USA) and genotyping using PCR-SSOP on the AutoLIPa (Siemens, Germany). In the study were used QCMD individual and group reports.

Results: Acceptance criteria are ±0.5 log10 (95% CI) or to 2SD for quantitative results, while for qualitative criteria is 90-100% results accuracy. QCMD calculates panel scores as an indicator of (un)satisfactory results compared to the target value. Score 0 signs highly satisfactory and 3 highly unsatisfactory result. For HCV genotype determination in five years the panel score was 0 with 100% accurate testing, for HBV and HIV-1 load panel scores ranged from 0-1. For HCV load tests one positive sample was negative (score 2) 2011; for 1 sample log titre differed by 1.2 (0.5 log is allowed; score 1) 2012; one panel had a score 2 (2SD) 2013, and for the same method panel score was 0 2014. For HBV genotype twice deviations has occurred; the sample is contaminated in a series of 8 samples; and HBV genotype H is wrong interpreted as mixed D+F (panel score 1).

Conclusion: All MDD employees analyze the results of EQA, discuss the possible critical points of the process, and improve decontaminating measures which resulted in excellent results of external quality assessment of molecular viral diagnostics.

e-mail: jasna [dot] bingulac-popovic [at] hztm [dot] hr



Comparison of two immunoassays for TSH, FT4 and FT3

Jelena Omazić1, Sanja Mandić1,2, Vesna Horvat1,2, Vatroslav Šerić1,2

1Department of Clinical Laboratory Diagnostics, Clinical Hospital Centre Osijek, Osijek, Croatia

2University J.J. Strossmayer Medical School Osijek, Osijek, Croatia


Introduction: The thyroid-stimulating-hormone (TSH), thyroxin (FT4) and triiodothyronine (FT3) are ones of the most required tests. The aim of this study was to examine comparability of results from Unicel Dxl600 (Beckman Coulter, Tokyo, Japan) and Architect i1000 (Abbott, Wiesbaden, Germany) analyzers.

Materials and Methods: The verification of methods was done on control samples BioRad Lypocheck Immunoassay Plus Control (Biorad Laboratories, Marnes-la-Coquette, France) (level1 and level2) for all three hormones on Unicel and Architect analyzers. Control samples were tested every day in duplicate during 10 days for imprecision in series and day-to-day imprecision. Results are shown as coefficient of variation (CV) and compared with Westgard rules and with CV provided by the manufacturer. Methods comparison was conducted using 50 patient samples. Statistical analysis was performed using MedCalc software (MedCalc, Mariakerke, Belgium). Correlation of results is show trough Spearman correlation coefficient. Passing-Bablok regression was also used, including the Cusum test for linearity. The level of significance was set at P<0.05.

Results: The measured values of control samples were within the range recommended by the manufacturer. Results of the imprecision in series and day-to-day imprecision were in the range of desirable specifications provided by manufacturer. The measured values of TSH on both analyzers, were within desirable specification according to Westgard (I=9.7%), while for FT4 (I=2.9%) and FT3 (I=4%), the measured values were not within desirable specification for neither analyzers. Passing-Bablok regression showed that there was proportional measurement error between analyzers for TSH (y=0.001857+0.9087x) and FT3 (y=0.2439+0.8403x), and constant and proportional measurement error between analyzers for FT4 (y=4.0083+0.8183x).

Conclusion: Data gathered through verification of method was within range recommended by the mabufacturer. Statistical analysis showed that is impossible to compare results from two different analyzers. That proves that for monitoring of patients, the thyroid hormones need to be determined always on one analyzer.

e-mail: jelena [dot] omazic [at] gmail [dot] com


G07 (Oral presentation)

Does the supervision of the Croatian Accreditation Agency (CAA) lead to more effective reporting of nonconformities in the laboratory?

Domagoj Marijančević, Adriana Unić

Clinical Department of Laboratory Diagnostics, University Hospital Dubrava, Zagreb, Croatia


Introduction: Nonconformity is any deviation from the requirements of the ISO 15189. Accredited laboratories commit to develop the most appropriate way to manage nonconformities. The aim of this study was to assess the impact of conformity assessment bodies first audit on the raise of awareness about the importance of reporting nonconformities in the medical laboratories.

Materials and Methods: Using Clinical Department for Laboratory Diagnostics network where the documentation of the QMS is distributed, the data on nonconformities before and after supervision of the CAA was retrospectively obtained. Nonconformities are then further categorized according to the grounds for their registration.

Results: During multiyear establishment of the QMS in the laboratory we reported only four nonconformities. During the year following the first CAA audit we reported eleven nonconformities in total. Before supervision of the CAA, nonconformities were mostly reported on the basis of the internal audits findings (3/4) rarely by normal application procedure of nonconformities (1/4). After supervision of the CAA, we experienced noticeable increase in the number of initiated nonconformities on the basis of the internal audits findings (6/11) and in particular we noticed the increase in standard reporting of nonconformities (5/11).

Conclusion: Considering that the first CAA audit the number of reported nonconformities was noticeably increased in all categories, we are convinced that external supervision particularly stimulates the development of awareness about the importance of reporting nonconformities. In addition, through practical assessment in the laboratory, external assessors further train the staff about the need to report all nonconformities promoting an atmosphere in which there will be no deliberate or accidental concealment of nonconformities. Ultimately, the effective management of nonconformities leads to the necessary corrective and preventive actions that form the basis for a continuous improvement process and enhancement of all segments of the quality system in the medical laboratories.

e-mail: dmarijan [at] kbd [dot] hr



Performance estimation of biochemical methods validated on analyzer Beckman Coulter AU5800 using normalized MEDx

Anđela Žic1, Vesna Šupak Smolčić1,2, Zvonka Bačić1, Lidija Bilić-Zulle1,2

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia

Introduction: Prior to the implementation of a new analytical system in routine laboratory work it is required to perform analytical validation. Method validation includes measuring inaccuracy (bias) and imprecision (CV), measures of systematic and random error, essential data for calculation of the total error (TE). TE represents the measure of quality i.e. overall error that may occur during measurement. Aim of the study was to show all validated methods on normalized MEDx graph, simple tool for fast assessment of methods performance.

Materials and Methods: The process of validation of 32 analytical methods on the analyzer Beckman Coulter AU5800 (Beckman Coulter, USA) has been conducted during July 2014 in the Department of Laboratory Diagnostics, CHC Rijeka. Each method is evaluated according to the recommended quality criteria from the database Minchinela et al. ( /biodatabase1.htm). TE is calculated according to the equation TE=Bias+σCV where σ indicates desired confidence level (σ=2-6). Expressing CV and bias as the percentage of the TE enables to display all operating points in the same normalized MEDx graph where each method is presented by its coordinates CV and bias and defined according to the σ value for method performance classification.

Results: According to the given criteria, performance of the validated analytical methods is classified into the following categories: excellent (6σ): CK, triglycerides, GGT, ferritin, CRP; good (5σ): amylase, direct bilirubin, IgM; border (4σ): total bilirubin, IgA, ALT, cholesterol; poor (3σ): glucose, AST, LDH, urea, iron; unacceptable (2σ): IgG, uric acid, ALP; points outside the graph: creatinine, total protein, albumin, Mg, P, Ca, Na, K, Cl, HDL, C3, C4.

Conclusion: The results of analytical validation for many parameters of the new system are not satisfactory according to six sigma. Normalized MEDx graph enables us to easily observe the methods which require more frequent control and more strict quality criteria.

e-mail: angela [dot] zic [at] gmail [dot] com



Application of a flexible scope of accreditation

Ines Alpeza Viman, Anita Herceg, Dunja Rogić

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Department of Laboratory Diagnostics, UHC Zagreb requested the first accreditation assessment to obtain objective evidence for establishing quality management system, and for its capability to implement specific laboratory procedures. Accreditation certificate with precise list of these procedures and accompanying informational labels represents such evidence.

Materials and Methods: 71 laboratory procedure were included in the first accreditation assessment conducted in February 2014. As a laboratory process is dynamic, some laboratory procedures have undergone changes at the time of obtaining certificate of accreditation. Changes were not substantial, but resulted in a new version of documents, not specified in certificate. After a year >70% of laboratory procedures were revised in documentation and practice. Therefore, assessors proposed establishment of flexible scope of accreditation.

Results: Management of flexible scope of accreditation is documented with clearly defined elements of application: Conditions under which such scope is applicable - changes apply only to elements not involving method modification; Limits within which it is applicable – it only applies to already accredited laboratory procedures; Responsibilities in the application - identification and assessment of the need for changes, verification and documentation of changes and approval of the application, estimate of the need for additional training, implementation of training, storage of records etc.; Requirements to be met - publicly available, constantly maintained list of laboratory methods covered by the flexible scope of accreditation; continuous self-assessment of such scope.

Conclusion: Application of flexible scope of accreditation is the legalization of introducing changes to already accredited laboratory procedures under clearly defined conditions. It allows proper documentation tracking of continuously present but not substantial changes. Its application can be considered a sign of confidence in assessed body, as it implies not only independence in introducing changes without additional verification of accreditation body but also appropriate - higher - level of responsibility in their implementation.

e-mail: ialpeza@kbc-zagreb.


Analytical validation of cyclosporine and tacrolimus on Roche Cobas 6000 analyzer

Katarina Čepić, Dijana Dževrnja Viro, Antonija Rodin Kurtović, Nada Bošnjak

Department of Laboratory Diagnostics, Clinical Hospital Centre Split, Split, Croatia


Introduction: Immunossupresive drugs are used to reduce the immune response in organ transplantation and autoimmune disease. Such therapies require lifelong use and nonspecifically suppress the entire immune system, exposing patients to considerably higer risk of adverse effects. Therapeutic drug monitoring (TDM) ensures that concentrations are not too high or too low, thereby reducing the risks of toxicity or rejection, respectively. The aim of this study was to evaluate analytical performance of Roche Cobas 6000 analyzer for TDM of cyclosporine and tacrolimus.

Materials and Methods: For the purpose of analytical validation of cyclosporine and tacrolimus by ECLIA (electrochemiluminescence immunoassay) method on Roche Cobas e6000 analyzer following parameters were determined: within-run and between run imprecision, inaccuracy (bias) using two levels of commercial control materials and method comparison with analytical system Abbott Architect i1000, CMIA (chemiluminiscent microparticle immunoassay) method on human samples (N=40).

Results: Coefficients of variation (CV) for within-run imprecision were 1.06% and 1.90% for cyclosporine and 2.94% and 2.52% for tacrolimus. CV for between run imprecision were 3.85 and 4.8% for cyclosporine and 3.45% and 2.10% for tacrolimus. Inaccuracy (bias) results for cyclosporine were -4.74% and 6.80%, and for tacrolimus -7.10% and 1.49%. Passing-Bablock regression analysis of cyclosporine comparison on two analyzers showed that there was constant deviation between methods (a=21.38, 95%CI 14.44-29.74; b=1.036, 95%CI 0.95-1.15). Bland-Altman plot showed the average deviation between two methods. Passing-Bablock regression analysis of tacrolimus comparison on two analyzers showed statistically significant but clinically insignificant (a=0.73, 95%CI 0.28-1.07; b=1.00, 95%CI 0.9-1.07) deviation between methods.

Conclusion: Analytical validation of cyclosporine and tacrolimus fulfilled all previously established criteria and could be implemented in a routine laboratory work. Data obtained by Passing-Bablock regression analysis showed that results are not completely comparable, especially for cyclosporine which confirms the importance of recommendations on monitoring the values of cyclosporine and tacrolimus on the same analyzer with the same method.

e-mail: katarina [dot] cepic [at] kbsplit [dot] hr



Comparison of methods for creatinine determination on Olympus AU400 and Dimension Xpand analyzers

Valentina Šenjug1, Danijela Polak-Erceg1, Anita Klasić1, Lada Surjan2

1Department for Boichemistry and Hematology, Hospital for Medical Rehabilitation Krapinske Toplice, Krapinske Toplice, Croatia

2Department for Boichemistry and Hematology, General Hospital Šibenik, Šibenik, Croatia


Introduction: Standardization of creatinine determination methods is of high value for estimation of glomerular filtration rate (eGFR) since eGFR is the main clinical use of creatinine. The aim was to compare compensated Jaffe method with uncompensated Jaffe method, compensated Jaffe method with enzymatic method and two compensated Jaffe methods on different analyzers used in laboratory.

Subjects and Methods: Study included 47 patients, serum creatinine values 26-1061 μmol/L. The samples were measured with four different methods on two different analyzers: enzymatic method (A), compensated Jaffe (B) and uncompensated Jaffe method (C) on Olympus AU400 analyzer and compensated Jaffe method on Siemens Dimension Xpand analyzer (D). The results were analyzed using Passing-Bablok regression analysis.

Results: The obtained results were: y(B) =-0.23+1.03x(A), 95%CI for intercept (-1.70-1.16), for slope (1.02-1.05); y(D)=5.88+0.98x(B), 95%CI for intercept (4.00-7.67), for slope (0.96-1.00); y(B)=-18.71+1.07x(C), 95%CI for intercept (-19.84-(-16.43)), for slope (1.04-1.08). The results indicate good method agreement for method B and method A with no significant difference in linearity (P=0.11) and constant difference between method B and method D, without significant difference between linearity (P=0.11). There is a constant difference between method B and method C, just as expected because of the setting of the methods.

Conclusion: There is a good comparability between compensated and enzymatic method on Olympus AU400. However, we use compensated Jaffe method in routine work in our laboratory because the price is lower, while enzymatic method is used only in creatinine determination in pediatric population. The comparison of uncompensated and compensated method showed constant difference due to mathematical correction of the compensated method (18 μmol/L), which makes these methods statistically incomparable. Constant difference between the two compensated methods on Olympus AU400 and Dimension Xpand is within the allowable deviation defined by the CSMBLM in the external quality assessment criteria.

e-mail: valentina [dot] senjug [at] sbkt [dot] hr


G12 (Oral presentation)

Immunosupressants therapeutic drug monitoring: immunoassay vs. liquid chromatography-tandem mass spectrometry (LC-MS/MS)

Andrea Radeljak1, Tajana Filipec-Kanižaj2, Matea Zorić1, Sonja Perkov1, Zlata Flegar-Meštrić1

1Department of Medical Biochemistry and Laboratory medicine, Merkur University Hospital, Zagreb

2Department of Internal Medicine, Merkur University Hospital, Zagreb, Croatia


Introduction: Immunosuppressants therapeutic drug monitoring (TDM) is a crucial requirement for the prevention of adverse drug reaction and graft rejection due to incorrect dosage after organ transplantation. The aim of this study was to compare concentrations of immunosuppressants tacrolimus and cyclosporin A determined by immunoassay and liquid chromatography-tandem mass spectrometry (LC-MS/MS) throughout longitudinal monitoring of transplanted patient.

Subjects and Methods: Tacrolimus (N=216) and cyclosporin A (N=163) concentrations were measured by chemiluminescent microparticle immunoassay (CMIA, Abbott Architect i1000SR), accredited according to HRN EN ISO 15189, and online SPE LC-MS/MS (Shimadzu UPLC NEXERA X2 - LCMS-8040) in whole blood samples collected 12 hours postdose. The obtained results were within the target values of internal and external quality assurance programme (Referenzinstitut für Bioanalytik, Deuchland, Survey for immunosuppressives). The correlation between methods was assessed using Passing-Bablok linear regression. Further, 15 patient were monitored simultaneously by both methods during period of 3 months by TDM of immunosuppressants in 6 to 22 samples,depending on patient.

Results: A method comparison showed correlation between concentrations measured with immunoassay versus LC-MS/MS according to the equation y=-0.3575(-0.7700-0.08947)+1.2250(1.1579-1.3000)x for tacrolimus and y=-6.4789(-12.2273–2.0000)+1.5634(1.4773-1.6515)x for cyclosporin A, respectively. Median of bias was 17.1%(13.5%-20.3%) for tacrolimus and 50.4%(45.9%-54.9%) for cyclosporin A. Longitudinal monitoring of 15 patients showed that the degree of overestimation by immunoassay varied from -10.9% to 71.1% (tacrolimus) and -17.9% to 133.8% (cyclosporin A).

Conclusion: Results of our study are in agreement with literature data - immunoassay overestimates drugs concentration due to nonspecific cross-reaction with metabolites, consequently leding to under-dosage of the drug and increasing the risk of organ rejection. The degree of overestimation is individual. Therefore, longitudinal monitoring of immunosuppressants is necessary, prefereably using LC-MS/MS as a gold standard in TDM owing to it substantial specificity,sensitivity and low cost of analysis.

e-mail: radeljak [dot] andrea [at] gmail [dot] com


G13 (Oral presentation)

Comparison of two ELISA methods for measuring the concentration of fecal calprotectin in the pediatric population

Irena Linarić, Jasna Obuljen

Department of Medical Biochemistry and Hematology, Department of Laboratory Diagnostics, Children’s hospital Zagreb, Zagreb, Croatia


Introduction: Calprotectin is non-invasive marker of intestinal inflammation, clinically significant in the diagnosis and monitoring of chronic inflammatory bowel disease (IBD). The aim of this study was to examine the correlation and diagnostic accuracy of two ELISA methods for measuring the concentration of fecal calprotectin in the pediatric population.

Materials and Methods: The study included 68 children (age 4-17 years) with clinical suspicion of IBD. The concentration of fecal calprotectin was determined by two ELISA methods: Calprest (Eurospital, Italy) and Calprotectin (Bühlmann Laboratories AG, Germany). Diagnostic criteria for IBD was positive finding of colonoscopy.

Results: Passing-Bablok regression was used for method comparison of two calprotectin ELISA methods. Regression analysis demonstrated no constant difference (A=0), but significant proportional difference (B≠1); regression equation y=-2.2860+0.5882x with 95% confidence interval (95%CI) for the section (A) -17.1360-2.3520 that includes the value 0 (A=0) and the 95%CI for the slope (B) from 0.5388-0.6425, which does not include the value 1 (B≠1). The receiver operating characteristic (ROC) analysis showed excellent diagnostic accuracy of both methods, with the area under the curve (AUC) 0.973(95%CI 0.901-0.997) for Calprest method and 0.982 (95%CI 0.915-0.999) for Calprotectin method. Comparison of AUC showed that the diagnostic accuracy of the two methods are not significantly different (P=0.218). ROC analysis showed optimal cut-off values for Calprest and Calprotectin methods of 188 mg/kg (sensitivity 90.7%, specificity 96%) and 383 mg/kg (sensitivity 90.7%, specificity 100%), respectively.

Conclusion: Comparative testing of Calprest and Calprotectin ELISA methods for the determination of fecal calprotectin showed no constant significant difference, but significant proportional differences. These methods are not comparable and cannot be simultaneously used in the diagnosis and monitoring of patients. ROC analysis showed excellent and equal diagnostic accuracy of both methods.

e-mail:irenalinaric16 [at] gmail [dot] com



Comparison of monoclonal and polyclonal ELISA method for determination of the
fecal elastase in stool in the pediatric population

Jasna Obuljen, Irena Linarić, Ana Milčić

Department of Medical Biochemistry and Hematology, Department of Laboratory Diagnostics, Children’s hospital Zagreb, Zagreb, Croatia


Introduction: Determination of fecal elastase (FE) in stool is clinically significant in diagnosis of pancreatic insufficiency of various etiologies, including cystic fibrosis, Shwachman syndrome, chronic pancreatitis, cholelithiasis, and malnutrition caused by dysfunction of pancreas. The aim of this study was to compare two ELISA methods for quantification of FE in pediatric population.

Materials and Methods: The study included 55 children (aged 2 months to17 years) with clinical suspicion of pancreatic insufficiency. The concentration of FE was determined with ScheBo stool pancreatic elastase (ScheBo Biotech AG, Germany) ELISA (monoclonal antibody) and Bioserv pancreatic elastase (Bioserv Diagnostics GmbH, Germany) ELISA (polyclonal antibody). The measuring range for both methods is 15-500 μg/g and the concentration of FE <200 μg/g is the cut-off value indicating the insufficiency of the exocrine pancreas.

Results: Histogram distribution data obtained by analyzing the results from both methods showed that regression analysis could not be applied to compare these methods. Analysis of compatibility of results was made by comparison of results to the cut-off value (200 μg/g). Results indicated satisfactory correlation (39/55, 0.71) of the results of both ELISA methods within the reference range (>200 μg/g). In the area of reduced FE values (<200 μg/g) results were matched in 3/16: ScheBo method detected 3 children with reduced values, whereas Bioserv method detected 16 children, therefore methods were inconsistent in 13/16 children.

Conclusion: Compatibility trial of monoclonal Schebo and polyclonal Bioserv method showed unsatisfactory comparability of results in the area of ​​reduced values that are clinically significant in the diagnosis of pancreatic insufficiency. Further testing is necessary to determine the possible reasons for discrepancies of these methods. According to the literature data, polyclonal method shows higher sensitivity than monoclonal method, which could partially explain the discrepancy found in the area of ​​low values in this study.

e-mail: jasna [dot] obuljen [at] xnet [dot] hr



Verification of fully automated method
for measuring Anti Mullerian hormone in serum on Roche Cobas e601 analyzer

Daniela Šupe-Domić, Lada Stanišić, Antonija Rodin-Kurtović, Nada Bilopavlović, Branka Knežević

Department of Laboratory Diagnostics, Clinical Hospital Centre Split, Split, Croatia


Introduction: Measurement of Anti Mullerian hormone (AMH) is mainly used for assessment of ovarian reserve and prediction of response to controlled stimulation which demands timely results reporting. In our laboratory AMH was measured monthly with modified AMH Gen II (Beckman Coulter, Czech Republic) ELISA method on ElisysDuo (Human, Germany) analyzer. The aim of this study was verification of first fully automated electrochemiluminescence immunoassay (ECLIA) for AMH determination on Roche Cobas e601 (Roche, Germany) analyzer and comparison with routine AMH Gen II ELISA method.

Materials and Methods: Verification included following parameters: within-run precision (repeatability), between-run precision and comparison with routine ELISA method, the same one used by manufacturer in method validation process. For precision determination patient serum samples with AMH concentrations (Sample 1=6.38 pmol/L, Sample 2=21.67 pmol/L) close to concentrations of human serums at which manufacturer claims precision (5.31 pmol/L and 18.20 pmol/L), were analyzed in three replicates over five days. Method comparison study was done using 40 serum samples (concentration range 0,92 – 38,6 pmol/L) and Passing-Bablok and Bland-Altman analysis of data was performed using MedCalc(Mariakerke, Belgium) program.

Results: Coefficients of variation were 0.65% and 0.65% for repeatability (manufacturer’s claim: 1.1% and 1.2%); 0.81% and 1.8% for between-run precision (manufacturer’s claim: 4.0% and 3.3%). Passing-Bablok regression analysis yielded following equation: y(Roche)=0.39(95%CI (-1.31)-0.83)+0.77(95%CI 0.67-0.94)x. Bland-Altman plot showed mean difference of 2.5 pmol/L ((95%CI (-4.3)-9.4) with the largest and also clinically significant differences at 10-20 pmol/L concentration range where the cut-off limit that separates patients with decreased and optimal fertility lies.

Conclusion: Precision verification of AMH ECLIA obtained results that fully satisfy manufacturer specifications. Differences obtained in comparison study confirmed previously published data about poor method agreement. Because of full automation, daily results reporting and great precision AMH ECLIA method was introduced in our routine practice.

e-mail: antonija [dot] kurtovic [at] gmail [dot] com



Comparison of within-laboratory precision, trueness and a total error of measurement procedures for CMIA and ECLIA methods in cyclosporine measurement

Marina Njire Bratičević, Antonija Perović, Diana Ljubimir

Department of Biochemical and Hematological Laboratory Diagnostics, General Hospital Dubrovnik, Dubrovnik, Croatia


Introduction: In the laboratory, verification of measurement procedures opens the possibility to compare different methods. The aim of this study was to compare within-laboratory precision (CVSl), trueness (BIAS) and total error (TE) for CMIA (Chemiluminescent microparticle immunoassay) and ECLIA (Electrochemiluminescence immunoassay) method used for cyclosporine measurement.

Materials and Methods: According to CLSI guideline protocol EP15-A2, verification of measurement procedures for cyclosporine measurement was conducted for CMIA method on Architect i2000SR (Abbott Laboratories, Illinois, USA) and for ECLIA method on Elecsys 2010 (Roche Diagnostics GmbH, Mannheim, Germany). For both measurement procedures calibration was performed on the first day. Cyclosporine concentration was measured in quality control materials in two concentration levels (L1 and L3) in triplicate for 5 days. Used control material for CMIA-Abbott method was MultiChem WBT (Technopath, Ireland) target values 92.3 ng/mL and 905 ng/mL and for ECLIA-Roche method PreciControl ISD (Roche Diagnostics GmbH, Germany) target values 91.5 ng/mL and 1170 ng/mL. Pretreatment of all used samples was performed using the same automatic pipette. Resulting values for precision were compared against the manufacturer’s criteria.

Results: Precision of measurement procedures CMIA-Abbott and ECLIA-Roche method meet manufacturer’s set criteria. Obtained values for CMIA-Abbott vs. ECLIA-Roche method for within-laboratory precision (CVSl), trueness (BIAS) and total error (TE) were as following: CVSl (L1) 13.87% vs. 2.91% and CVSl (L3) 9.49% vs. 2.23%, BIAS (L1) -8.86% vs. 1.57% and BIAS (L3) -14.91% vs 1.37%, TE (L1) 36.04% vs. 7.27% and TE (L3) 33.51% vs. 5.74%.

Conclusion: With the advantage of easier and faster sample pretreatment, measurement procedure for ECLIA-Roche method showed better accuracy, better trueness according to target values assigned to control materials and lower total error.

e-mail: marinanjire [at] yahoo [dot] com


G17 (Oral presentation)

Analytical verification of Combur10Test® M strips and urine analyzer Cobas u411

Alen Vrtarić, Nora Nikolac, Marijana Miler, Ana-Maria Šimundić

University Department of Chemistry, University Hospital Centre Sestre Milosrdnice, Zagreb, Croatia


Introduction: Our aim was to: a) verify urinary test strips for urine chemistry analysis Combur10Test® M (Roche Diagnostics GmbH, Mannheim) on analyzer Cobas u411; b) asses the comparability of Cobas u411 and Miditron Junior II (both Roche Diagnostics, GmbH, Mannheim).

Materials and Methods: Analytical verification was performed according to CLSI guideline EP12-A. Repetability was tested on 10 consecutively repeated measurements on negative control (BIORAD, Liquicheck Urinalysis Control Level 1) and in two patient samples with pathological values. Precision from day-to-day was assesed on control samples (Level 1 and 2) in duplicate for 10 days. Comparability was evaluated comparing 87 patient›s samples on analyzers Cobas U411 and Miditron Junior II. Accuracy for glucose, bilirubin and proteins was tested by comparing the results of test strips with values measured by reference method on analyzer Architect c8000 (Abbott, IL, USA). Precision was estimated by the degree of agreement of repeated measurements, accuracy by the degree of agreement with result of reference method. Comparability was expressed with kappa coefficient with 95% confidence interval. Acceptance criteria were: precision: 90% agreement of repeated measurements, pH: acceptable bias ±1 pH unit, specific gravity: acceptable bias ± 0.005 and comparability: kappa coefficient ≥0.6. MedCalc (v12.7.2.0, Ostend, Belgium) was used for statistical analysis.

Results: Precision was acceptable for all parameters, except for leukocytes in one patient›s sample (agreement for only 8/10 measurements). Comparability between analyzers was satisfactory, except for bilirubin (kappa=0.370 (0.112 to 0.628)). Accuracy for bilirubin was satisfactory (agreement 10/10). The observed differences in proteins (8/10) and glucose (6/10) were not clinically significant.

Conclusion: Analytical performance of Combur10Test® M on Cobas u411 analyzer are within acceptance limits. Observed bias for glucose and proteins are not clinically significant and are caused by differences in methodology. Analyzers are comparable, except for bilirubin measurement.

e-mail: alenvrtaric [at] gmail [dot] com



Verification of the method for determining human chorionic gonadotropin on Cobas e411 analyzer

Petra Filipi, Monika Doželenčić, Ivana Zec, Marijana Miler, Nora Nikolac, Ana-Maria Šimundić

University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: The aim of this study was to check the reproducibility, hemolysis interference and carryover for human chorionic gonadotropin (hCG) test on Cobas e411 analyzer (Roche, Mannheim, Germany).

Materials and Methods: The reproducibility was examined using control samples (Preci Control Universal 1 and 2) for five days in triplicate. Manufacturers acceptability criteria were used: CV (PCU1)=2.1%; CV (PCU2)=4.0%. The impact of hemolysis was tested at three hCG concentration levels according to CLSI protocol EP7-A2. Plasma hemolysate was prepared by osmotic shock. Hemolytic samples were prepared by adding hemolysate in increasing concentrations to reach free hemoglobin concentrations of 0 (no hemolysate), 2.5; 5.0; 7.5 and 10.0 g/L. To assess the impact of hemolysis, RCPA (Royal College of Pathologist of Australasia) acceptability criteria were used: ±1 IU/L for hCG <10.0 IU/L; ±10% for hCG >10.0 IU/L. Effect of carryover was examined by analyzing the low concentration sample in triplicate, immediately after the sample with high concentration in duplicate, according to the IUPAC (International Union of Pure and Applied Chemistry) protocol. All samples were analyzed on Cobas e411 with original reagent. Statistical analysis was performed using Microsoft Excel.

Results: For lower concentration level, reproducibility was out of the acceptable criteria (2.8% for PCU1 and 1.5% for PCU2). In examination of hemolysis interference, the maximum biases in the concentration levels of 181 and 3076 mU/mL were 6.1% and -7.5%. In samples with low hCG concentration (7.9 mU/mL), bias was greater than 1 IU/L in the sample with Hb>5 g/L. We did not observe false positive results in the carryover experiment.

Conclusion: The method for hCG determination has satisfactory reproducibility. Impact of hemolysis was clinically significant only in the samples with low analyte concentrations in hemoglobin levels above 5 g/L. There is no influence of carryover from samples with the high concentration.

e-mail: petrafilipi6 [at] gmail [dot] com


G19 (Oral presentation)

Prealbumin method verification on Beckman Coulter AU680 analyzer

Snježana Hrabrić Vlah1, Davor Valenčić1, Lidija Bilić-Zulle1,2

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia


Introduction: In clinical practice prealbumin is considered a good indicator of malnutrition as well as a negative acute phase reactant. The aim of this study was to evaluate reliability of prealbumin method on Beckman Coulter AU680 analyzer (Beckman Coulter, Brea, California, USA) before use in routine practice in the laboratory and examine the comparability of results with the Cobas c501 as a reference analyzer (Roche Diagnostics, Mannheim, Germany).

Materials and Methods: We conducted a short verification of prealbumin analytical method on Beckman Coulter AU680 analyzer. Prealbumin concentrations were determined for 5 days in triplicate at two concentration levels (total of 30 measurements). It was used a commercial control samples ITA Control Sera L2 and L3. Verification includes: within-run imprecision (Cwd), between-run imprecisions (CVbd) and bias (10 days, ITA Control Sera L2 and L3). Comparability of the methods were tested by measuring the patient samples (N=20) in a wide concentration range. Beckman Coulter AU680 was investigated analyzer, Cobas c501 was a reference analyzer.

Results: Results of the within-run imprecision for Beckman Coulter AU680 analyzer were: Cwd=1.05% (ITAL2) and Cwd=2.12% (ITAL3). Results of between-run imprecision were: CVbd=3.7% (ITAL2) and CVbd=5.44% (ITAL3). Bias was ITAL2=-0.7% and ITAL3=-4.1% for AU680 analyzer. The results of prealbumin concentrations were compared by Passing-Bablok regression analysis. Results of the regression analysis showed good compatibility (y=0.0230435+1.043478x, 95%CI 1.00-1.167 for the slope, 95%CI (-0.009)-0.030 for the intercept of the regression line, P=0.14). Correlation coefficient (r=0.95) was satisfactory between AU680 and reference analyzer.

Conclusion: Desirable analytical specification for within-run imprecision is 10.9%, between-day imprecision is 19.1% and bias is 5.5%, who derived from biological variation ( Prealbumin analytical method verification on AU680 analyzer shows acceptable imprecisions, bias, good compatibility and can be used in routine practice.

e-mail: snjezana [dot] hrabricvlah [at] gmail [dot] com



Verification of serum bicarbonate assay on Beckman Coulter AU680 analyzer

Snježana Hrabrić Vlah1, Lidija Bilić-Zulle1,2

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia


Introduction: Bicarbonate measurements are used in clinical practice as an indicator disorders associated with changes in body acid-base balance. The aim of this study was to evaluate reliability of bicarbonate analytical method on BC AU680 analyzer (Beckman Coulter, Brea, California, USA) before using in routine practice. We wanted to compare serum bicarbonate concentration measured on BC AU680 analyzer and results of whole blood bicarbonate concentration measured by potentiometric method on Rapid ABS 348 (Bayer Corporation, USA) as reference analyzer.

Materials and Methods: We performed short verification of bicarbonate analytical method according to CLSI guideline EP15-A2. We used commercial control samples BioRad Liquid Assayed in two concentration levels. Bicarbonate concentrations were measured for 5 days in triplicate. Verification includes: within-run imprecision (Cwd), between-run imprecisions (CVbd) and bias (10 days, BioRad L1 and L2). Comparability was examined by measuring bicarbonate concentration in the serum samples and in the whole venous blood (N=20).

Results: Results of the within-run imprecision for Beckman Coulter AU680 analyzer were: Cwd=3.44% (BioRad L1), Cwd=2.65% (BioRad L2). Results of between-run imprecision were: CVbd=3.2% (BioRad L1), CVbd=3.18% (BioRad L2). Bias was 0.48% (BioRad L1), -4.81% (BioRad L2) for BC AU680. The results of bicarbonate concentrations were compared by Passing-Bablok regression analysis. Results of the regression analysis were y=-2.176042+1.104167x, 95%CI 0.8364-1.4651 for the slope, 95%CI (-10.5023)-4.0600 for the intercept, P=0.98). Correlation coefficient was r=0.84.

Conclusion: Desirable analytical specification for within-run imprecision is 4%, between-day imprecision is 4.8% and bias is 1.56%, derived from biological variation ( Bicarbonate analytical method verification on the BC AU680 showed acceptable imprecision. Bias was departed from desirable values; the results were slightly lower than the target value in control samples BioRad L2. Results of bicarbonate concentration showed good compatibility between these two methods.

e-mail: snjezana [dot] hrabricvlah [at] gmail [dot] com



Comparison of transcutaneous and photometric methods for the determination of total bilirubin in healthy newborns with clinically manifest jaundice

Sonja Perkov1, Marko Vukasović2, Jelena Starčić1, Andrea Radeljak1

1Department of Medical Biochemistry and Laboratory Medicine, Merkur University Hospital, Zagreb, Croatia

2Department of Gynecology and Obstetrics, Merkur University Hospital, Zagreb, Croatia


Introduction: Determination of total serum bilirubin (TSB) is one of the most frequently requested laboratory test in newborns and the gold standard for assessment of neonatal hyperbilirubinemia.The aim of this study was to evaluate the comparability of the transcutaneous bilirubin (TcB) and TSB results for clinical application of noninvasive methods to supervise neonatal hyperbilirubinemia.

Subjects and Methods: The study included 66 pairs of results of the 33 newborns gestation age ≥35 weeks during the first week after birth, without phototherapy or exchange transfusion. The concentration of TSB was determined by photometric method with 3,5-dichlorophenyldiazonium tetraflouoro borate, accredited according to HRN EN ISO 15189. The level of TcB was determined as the difference in the optical density of light in the blue and green range of wavelengths at three measuring points forehead- sternum-forehead by Draeger JM-103 (Drager Medical Inc., Telford, PA) bilirubinometer. Method comparability was assessed using the Passing-Bablok regression and Bland-Altman analysis.

Results: Passing-Bablok analysis of the values obtained (with corresponding 95% confidence intervals) y-axis intercept 16.9091 (-33.8378-53.6455), slope 0.9091 (0.7455-1.351), indicating that there is no statistically significant constant or proportional error between two methods.The Bland-Altman analysis demonstrated that transcutaneous bilirubin measurements tended to underestimate the value of TSB: mean difference from the corresponding 95% confidence intervals between photometric and transcutaneous methods was 1.4(-2.0-9.3)±11.4%. Although 90% of results within ±1.96 SD, range difference of -20.9 (-25.7-(-16.1)) to 23.7% (18.9-28.5)% outside the desirable biological criteria for accuracy (8.95%).

Conclusion: Our results indicate that there is statistically significant correlation of transcutaneous and photometric methods in measuring total bilirubin in term neonates, suggesting a potential clinical application of this noninvasive methods in conjunction with other clinical indicators to assist in the control of neonatal hyperbilirubinemia. On the other hand, a wide range of difference indicates a possible clinically significant deviations, which is important to know when interpreting the results.

e-mail: radeljak [dot] andrea [at] gmail [dot] com



Validation of high performance liquid chromatography method for measuring of antifungal drug voriconazole in serum

Mila Lovrić, Nada Božina, Dunja Rogić

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Voriconazole (VRCZ) is a triazole antifungal agent with a broad spectrum for invasive fungal infection. VRCZ is mainly metabolized to the N-oxide predominantly by cytochrome P450: CYP 3A4 and CYP 2C19. Because of genetic polymorphisms of the cytochrome CYP2C19, voriconazole’s nonlinear and saturable pharmacokinetics and numerous drugdrug interactions standard doses of the drug do not produce predictable trough concentrations and therefore therapeutic drug monitoring is highly recommended. The aim of this study was to develop and validate a high performance liquid chromatography method for voriconazole therapeutic drug monitoring.

Materials and Methods: Serum sample (250 μL) and chloramfenikol as internal standard (25 μL) pretreatment was based on liquid extraction with ethyl acetate/hexane at alkaline pH. Organic layer was then evaporated and residue desoved in 200 μL mobile phase. The separation was obtained on C18, 250 x 4.6 mm I.D. 5mm column (Macherey-Nagel) by isocratic reversed-phase chromatography. Compounds were eluated with phosphate buffer/acetonitrile/methanol mixture (pH 4.6) as the mobile phase, at flow rate of 1.5 ml/min. Analytes were monitored at diode array detector at 254 nm.

Results: All calibration curves showed good linearity (r>0.998) through the range of 0.1-10 mg/L. The within-run precision in two concentration levels was 2.45 and 5.44%. Between-run precisions were 7.56% and 6.11%. Accuracy, expressed as percent error, was ranged from 97.3-106.8 %. The limit of quantitation was 0.038 mg/L and limit of detection 0.013 mg/L. The method has been applied to monitor voriconazole in serum of patients.

Conclusion: Based on analytical parameters linearity, precision, accurancy, limit of detection and quantitation presented method is suitable for therapeutic drug monitoring and pharmacokinetic studies. The method shows good specificity with other prescribed drugs. Voriconazole has a high degree of interpatient variability and some data are suggestive that therapeutic drug monitoring may improve voriconazole efficacy and toxicity.


e-mail: mila [dot] lovric [at] kbc-zagreb [dot] hr


H – Laboratory hematology and


Testing of primary haemostasis at patients with polycythaemia rubra vera and secondary erythrocytosis by PFA 100 analyzer

Branka Pauković Sekulić1, Leida Tandara1, Anuška Trlaja2

1Department of Laboratory Diagnostics, Clinical Hospital Centre Split, Split, Croatia

2Clinical Centre of Transfusion Medicine, Clinical Hospital Centre Split, Split, Croatia


Introduction: RBC mass is elevated in polycythemia rubra vera (PRV) and in secondary erythrocytosis (SE), associated with hypoxia or increased erythropoietin production. Both diseases have been accompanied by thrombosis due to increased viscosity of blood. Low-dose aspirin are applied in the treatment of blood hyperviscosity. Bleeding risk at PRV is rare and mostly affects patients with high platelets count. It manifests itself like skin and gastrointestinal bleeding. At SE bleedings do not appear. Aim of examination was to estimate the platelet function by determining closure time (CT) by PFA 100 analyzer (Siemens Healthcare Diagnostics) and compare them with referent intervals.

Subjects and Methods: 36 patients were examined: 18 with PRV (7 W and 11 M; 55-87 years) and 18 with SE (1 W and 17 M; 32-79 years). For all patients PCR for the mutation V617F in the JAK-2 gene and CBC counts on the Advia 2120 (Siemens) have been performed. CT was determined in the presence of the collagen and adrenaline (CEPI–CT) as well as collagen and ADP (CADP–CT) in samples of citrated whole blood (0.105 M Na-citrate) and were compared with referential intervals: (CEPI-CT:85-165 s); (CADP-CT:71-118 s). Statistical analysis was made by MedCalc 14.8.1 software. Mann-Whitney U-test (P<0.05) was used for comparison of the results.

Results: Values of the CT were significantly higher in the patients with PRV (CEPI-CT: median 285,5 s; 95%CI:168.6-300 s; P=0.020; CADP-CT: median 128 s; 95%CI:122.6-167 s; P=0.0025) than in the patients with the SE (CEPI-CT: median 148 s; 95%CI:120.8-224.7 s; CADP-CT: median 100.5 s; 95%CI:87-111.8 s).

Conclusion: The comparison of the results confirmed the clinical utility of this test in the assessment of platelets function in the patients with PRV at the appearance of bleeding because of the disease or therapy. Results at SE accord with results from the literature.

e-mail: branka [dot] paukovic [dot] sekulic [at] gmail [dot] com


H02 (Oral presentation)

Are platelet indices valuable in assessment of bone marrow thrombopoietic activity in neonates with sudden decrease in platelet count?

Ana Mlinarić, Gordana Fressl Juroš, Ivana Rako, Dunja Rogić

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Immature platelet fraction (IPF), mean platelet volume (MPV), platelet large cell ratio (P-LCR) and platelet distribution width (PDW) are platelet indices (PIs) measured as research parameters in complete blood cell analysis. We investigated if PIs could be useful in assessing bone marrow thrombopoietic activity in neonates following sudden decrease in platelet count (PC) and tested the hypothesis that PIs increase with adequate bone marrow thrombopoietic response.

Subjects and Methods: 32 neonates admitted to neonatal intensive care unit were monitored during 2 months. Blood was sampled in K3-EDTA tubes and analysed on Sysmex XE-5000 analyser. Based on platelet biological variability criteria, PC decrease greater than 20% was considered significant. In neonates with observed PC decrease, we tested PIs before and after this event for differences, using paired samples t-test. Otherwise, in neonates without PC decrease, we tested PIs of two consecutive samples for differences. The P<0.01 was considered statistically significant.

Results: We observed significant PC decrease in 16 neonates and found statistically significant increase in IPF (3.1 vs 5.2; P=0.004), MPV (10.6 vs 11.2; P=0.001) and P-LCR (29.7 vs 34.4; P<0.001) values before and after PC decrease, respectively. In 16 neonates without significant PC decrease, we found no statistically significant difference in IPF (P=0.189), MPV (P=0.021) and P-LCR (P=0.020) values. There was no significant difference in PDW values in both neonate groups.

Conclusion: Increased MPV, P-LCR and IPF indicate platelets were larger and more immature in neonates following significant PC decrease. Neonates without PC decrease exhibited no significant difference in PIs, indicating no additional activation of bone marrow was present. These findings suggest IPF, MPV and P-LCR are useful in assessing bone marrow thrombopoietic response following sudden decrease in PC.

e-mail: ana [dot] mlinaric [at] yahoo [dot] com



Analytical estimation of Sclavo Diagnostics coagulation reagents on semi-automatic and automated coagulometers

Vladimira Rimac1, Désirée Coen Herak1, Sanela Šimić Vojak2, Vanja Radišić Biljak3, Dunja Rogić1

1Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

2Department of Laboratory Diagnostics, General hospital Požega, Požega, Croatia

3Department of Medical Biochemistry and Laboratory Medicine, Merkur University Hospital, Zagreb, Croatia


Introduction: According to good laboratory practice, before applying new reagents in routine work, an analytical assessment is necessary. The aim of this study was the verification of Sclavo Diagnostics International SrI (Sovicille, Italy) coagulation reagents for determination of prothrombin time (PT), activated partial thromboplastin time (aPTT), functional concentration of fibrinogen (FBG) and the activity of factor V (FV).

Materials and Methods: The verification of Sclavo reagent PT Kit (PT), aPTT S kit (APTT), Fibrinogen Kit (FBG) and Factor V Kit (FV) was performed on automated coagulometers BCS XP/BCT (Siemens Healthcare Diagnostics, Germany) and semi-automatic coagulometer KC-4A (Amelung, Germany). Within-run imprecision (N=10) was examined with plasma pools in the normal (N levels) and pathological ranges (level P) and between-run imprecision (10 days in duplicate) with control plasma Sclavo Normal and Abnormal Control Plasma kit. For method comparison 30 samples of patient plasma were analyzed after analysis on reference analyzers BCS XP/BCT using Siemens Healthcare Diagnostics reagents. Inaccuracy results were expressed as mean, standard deviation and coefficient of variation (CV). Passing-Bablok regression analysis was used to compare results with reference reagents.

Results: Within-run imprecision CVs were <5% for all tests. For between-run imprecision the smallest CVs (<2%) were obtained on both coagulometers for the PV (s) in both concentration levels. The highest CVs (<12%) were obtained for the FV (level N). The correlation coefficients obtained with reference reagents ranged from 0.789 (FV) to 0.967 (FBG) on the KC-4A, and from 0.870 (FV, BCT) to 0.978 (PT-INR, BCS XP). By comparing the results obtained on coagulometers KC-4A and BCS XP/BCT using Sclavo Diagnostics reagents, correlation coefficients ranged from 0.904 for the FV to 0.983 for the APTT.

Conclusion: The tested Sclavo Diagnostics reagents showed satisfactory analytical properties and are suitable for use in routine work on semi-automatic and automatic coagulometers.

e-mail:kutnjakvl [at] gmail [dot] com



Red blood count of the term newborns

Ksenija Paradinović, Blaženka Dobrošević, Tara Rolić, Marija Milić, Vatroslav Šerić

Department of Clinical Laboratory Diagnostics, Clinical Hospital Centre Osijek, Osijek, Croatia


Introduction: Pediatric population, the most sensitive age group for the sample, is not fully covered by the reference intervals. We do not have our own reference intervals for this population, so it is recommended to use reference intervals in the literature, where the main criterion is compliance of revised analytical methods with recommended methods of the Croatian Chamber of Medical Biochemists.

The goal is to determine the concentration of red blood cells parameters of term newborns, and compare them with the recommended reference intervals.

Materials and Methods: Samples of capillary whole blood of term newborns (347 samples) at the age of one to five days. The concentration and the calculated parameters of red blood cells in the blood counter Sysmex XE2100. The calculation of the average values ​​obtained and comparison with the Harmonization reference interval.

Results: Newborns in the first week of life have an average value of red cells = 5.82×1012/L, Hb = 210 g/L, hematocrit = 0.592 L/L, MCV = 101.9 fL, MCH = 37.6 pg, MCHC = 359 g/L.

Conclusion: The values ​​of red blood cells of newborns in Osijek and the surrounding area partly show increased value of the proposed reference range in age from 1-14 days. The proposed range is a period of significant physiological changes in the concentration of red blood cells. It would be useful to create own reference intervals and to divide range.

e-mail:ksenija [dot] paradinovic [at] gmail [dot] com



Validation of Coagulation Systems STA Compact Max and STA-R Evolution

Ivana Baršić1, Snježana Semenski2, Vesna Šupak Smolčić3, Dragana Antončić3, Darka Stošić4, Désirée Coen Herak1, Dunja Rogić1

1Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

2Medical Biochemistry Laboratory, General Hospital Zabok and Hospital of Croatian Veterans, Zabok, Croatia

3Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

4Biomedica dijagnostika d.o.o., Zagreb, Croatia


Introduction: The aim of this study was to perform the validation of coagulation systems STA Compact Max and STA-R Evolution (Diagnostica Stago, France) with emphasis on method performance for routine and special coagulation analyses and to compare them with routine coagulation systems.

Materials and Methods: Validation was performed at the same time on both analyzers according to Stago protocol System Method Validation PQ for STA® System for PT, APTT, D-dimer, antithrombin, PC, FVII, FVIII and free PS:Ag. The study included the determination of within-run and between-run imprecision using commercial control samples and patients plasma pools with normal and pathological values, inaccuracy, carry-over (APTT), method comparisons with routine analyzers BCS XP/BCT and miniVidas using patient plasma samples (N=33-127), linearity (D-dimer, free PS:Ag) and verification of reference interval (³20 samples from healthy individuals). Results were statistically analyzed with MedCalc software (version and estimated according to specification criteria from Stago quality documents SAV 4012 and 4013.

Results: For all analytes, CVs met the required specifications for within-run imprecision except for FVIII on STA-R Evolution analyzer (10.1% for normal and 8.6% for pathological values). Although higher than expected CVs for D-dimer were obtained (STA-R Evolution: 15.1%, STA Compact Max: 7.9%), SDs (0.02-0.07 mg/L FEU) were within required specifications. Between-run CVs for all analytes fulfilled specific requirements. Passing-Bablok regression analysis yielded correlation coefficients ranging from 0.811 (APTT) to 0.988 (PC) on STA-R Evolution and from 0.827 (APTV) to 0.988 (PC) on STA Compact Max. The linearity stated by manufacturer for D-dimer and free PS:Ag was confirmed. Assessment of reference intervals met the defined criteria confirming that they are applicable for routine work.

Conclusion: The obtained results from validation study show optimal analytical performance of both coagulation systems and therefore can be implemented in routine practice.

e-mail: ibarsic [at] kbc-zagreb [dot] hr


I – Preanalytical and postanalytical phase of laboratory work


Abnormal flotation of separator gel in blood test tubes in three hemodialysis patients with intradialytic hypotension

Meltem Demir1, Sebahat Ozdem2

1Biochemistry Laboratory, Medicalpark Hospital, Antalya, Turkey

2Department of Medical Biochemistry, Akdeniz University Medical Faculty, Antalya, Turkey.


Introduction: Blood sampling tubes containing separator gel are widely used in many clinical biochemistry laboratories since they provide an easy and effective way of separating blood cells and serum. We present 3 cases with abnormal flotation of separator gel in test tubes after centrifugation.

Material and Methods: The blood sample obtained in Hemodialysis Unit of Medicalpark Hospital from a 60 years old male patient with end stage renal disease (ESRD) following completion of hemodialysis (HD) was drained into BD SST II Advance Tube (Becton Dickinson, NJ, USA) containing separator gel. Following completion of coagulation, sample was centrifuged at 3000 rpm for 10 min.

Results: Following placement of test tube in autoanalyzer, a preanalytical error caused discontinuation of the measurements and delay in delivery of lab results. A careful analysis revealed bending and obstruction of the probe of autoanalyzer due to stuck in separator gel that was abnormally floated on to the top of tube. One week and 2 months after the first case, abnormal floating of separator gel was observed in post-HD blood samples of two other patients with ESRD. Intradialytic hypotension and high protein load (>162.8 g/L) of post-HD blood samples were the common findings in all 3 cases. Raising the protein content of blood samples from healthy volunteers above 161.6 g/L by adding bovine albumin caused abnormal flotation of separator gel similar to those we observed in our cases.

Conclusion: Our findings suggested that increased protein load probably caused by intradialytic hypotension due to dehydration induced the observed abnormal flotation of separator gel in our cases. Present findings emphasize the importance of visual control of test tubes for abnormal flotation of separator gel in blood samples with high protein content especially in the presence of accompanying volume loss to avoid unfavorable effects on test results and laboratory costs.

e-mail: meldemir52 [at] gmail [dot] com


I02 (Oral presentation)

Ionized calcium in serum – the influence of time periods from sampling to centrifugation, and from centrifugation until analysis

Antonija Perović, Marina Njire Bratičević, Diana Ljubimir

Department of Biochemical and Hematological Laboratory Diagnostics, General Hospital Dubrovnik, Dubrovnik, Croatia


Introduction: After verification of serum as sample equivalent to arterial blood for ionized calcium (iCa) measurement, aim of our study was to evaluate the influence of time periods from sampling to centrifugation, and from centrifugation until analysis on the iCa in serum, in order to assess the urgency of analysis due to possible sample instability.

Subjects and Methods: The influence of time periods from sampling to centrifugation was examined on 10 subjects; 3 tubes of blood were collected and centrifuged 15 (Tube 1), 30 (Tube 2) and 60 minutes (Tube 3) after sampling, and analyzed within 10 minutes. The influence of time periods from centrifugation until analysis was examined on 15 subjects; 3 tubes of blood were collected, centrifuged after 30 minutes of the sampling and analyzed at the following timing: 0-10 (Tube A), 30-40 (Tube B) and 90-100 minutes (Tube C). Venous blood samples were collected in ‘’gel separator’’ serum tubes, all tubes were completely filled, centrifuged according to the manufacturer’s instructions, without temperature adjustment and left at room temperature until analysis. iCa was measured by potentiometric method on the RapidLab 348 analyzer (Siemens, Suffolk, UK). The obtained values were compared according to RiliBÄK’s criteria (acceptable deviation < 7.5%).

Results: iCa deviations in Tube 2 and Tube 3 in relation to the values in Tube 1 were less than established criteria; medium bias (range); 1.25% (-3.1-3.13) for Tube 2; 1.12% (-1.55-3.91) for Tube 3. iCa deviations in Tube B and Tube C in relation to the values in Tube A were less than established criteria; medium bias (range); 2.38% (-4.58-2.34) for Tube B; 2.79% (-4.72-4.80) for Tube C.

Conclusion: Different time periods from sampling to centrifugation (up to 60 minutes) and from centrifugation until analysis (up to 90 minutes) did not influence the stability of the iCa in the serum.

e-mail: perovic [dot] antonija [at] gmail [dot] com



Effect of hemolysis on serum protein electrophoresis

Božica Sokolić, Adrijana Dorotić, Renata Laškaj, Gabrijela Smoljkić, Marija Mijakić

Department of Medical Biochemistry, Hematology and Coagulation, University Hospital for Infectious Diseases “Dr. Fran Mihaljević“, Zagreb, Croatia


Introduction: According to manufacturers’ recommendations, hemolyzed samples are not suitable for serum protein electrophoresis (SPE) because they increase alpha2- and beta-globulin fraction. The aim of this study was to estimate the influence of hemolysis on the values of SPE.

Materials and Methods: During the period of 2 months, the total of 34 hemolyzed and non-hemolyzed serum samples were collected. LIH reagent (Beckman Coulter, USA) was used to estimate the degree of hemolysis. According to the concentration of free haemoglobin derived from LIH test, all samples with the concentration of free haemoglobin >0.50 g/L were described as hemolyzed.

Total serum proteins were measured using a photometric test on a Beckman Coulter AU 680 analyzer (Beckman Coulter, USA). SPE was performed using Hydragel 30 Protein(e) on a Hydrasis 2SCAN (Sebia, USA) in both, hemolyzed and non-hemolyzed samples. Comparison of the values of proteins (g/L), alpha-2 (%) and beta-globulin fraction (%) was performed using MedCalc 12.5.0. statistical software (Ostend, Belgium). P<0.005 was set as the treshold of significance.

Results: Kolmogorov-Smirnov test showed that the values of proteins and alpha-2-globulin fraction follow normal distribution and that beta-globulin fraction does not distribute normally and is consequently tested using non-parametric tests. Paired t-test showed that there is a statistically significant difference between non-hemolyzed and hemolyzed samples for the values of alpha-2-globulin fraction (P<0.001). Other parameters did not show statistically significant difference. Bland-Altman method showed that there is a mean difference between the values of alpha-2-globulin fraction between non-hemolyzed and hemolyzed samples of -14.2% (±1.96 SD -49.8% to 21.3%) which can be considered as clinically significant difference when compared to desirable bias specification based on biological variation.

Conclusion: According to our results, hemolyzed serum samples are not suitable samples for SPE because hemolysis can lead to statistically and clinically higher values of the alpha-2-globulin fraction.

e-mail: bsokolic [at] bfm [dot] hr


I05 (Oral presentation)

Vacuette® Glucomedics tubes and prevalence of gestational diabetes

Marija Kocijančić, Jelena Čargonja, Alma Delić Knežević

Medical Biochemistry Laboratory of Primorsko-Goranska County Health Care Rijeka, Rijeka, Croatia


Introduction: An accurate plasma glucose measurement is essential for correct diagnosis of diabetes, especially for gestational diabetes mellitus (GDM). We hypothesized that glucose concentration in Glucomedics tubes should be slightly elevated in regard to the reference tubes due to the immediate additive activity. The aims of this study were to ascertain whether the results from Glucomedics tubes compatible with sodium fluoride tubes results and potential impact on the diagnosis of gestational diabetes on basis change of positive results.

Subjects and Methods: Venous blood samples were collected from 97 pregnant women who were admitted to the laboratory in the morning for GDM screening. Diagnostic test for GDM screening included three points 2h oral glucose tolerance test (OGTT) with 75 g glucose load. Blood was taken three times (0h, 1h, 2h) directly into two different plasma vacuum tubes: Tube I (reference): BD Vacutainer® NaF/K2oxalate (lot 4007104, Becton, Dickinson and Company, Plymouth, UK) and Tube II: Vacuette® Glucomedics Na2EDTA/NaF/Nacitrate-citric acid buffer (lot A140411U, Greiner Bio-one, Kremsmunster, Austria). Glucose determination was performed on Beckman AU 2700 using hexokinase method. The Wilcoxon and McNemar’s tests were applied to evaluate the difference of the new blood collection tube. Results were analyzed using Passing-Bablok and Bland-Altman plots for association and differences. The level of significance was set as P<0.05.

Results: Plasma glucose concentrations were statistically significant higher in Glucomedics tubes compared to reference NaF/KOx tubes for all three measurements (P<0.001). Bland-Altman difference plots showed high mean bias (8.9 to 9.7%), but Passing-Bablok regression analysis showed good association among the measured values. Significant difference in change of positive results was shown for Glucomedics tubes (49.3%, P<0.001).

Conclusion: Glucomedics tubes have advantages for the most accurate estimation of glucose measurements, but additional improvements are needed due to dilution factor.

e-mail: marija [dot] percan [at] gmail [dot] com




Comparison of glucose concentration in venous plasma and in serum

Bojana Kranjčec, Snježana Semenski, Pavao Tomašković

Medical Biochemistry Laboratory, General Hospital Zabok and Hospital of Croatian Veterans, Zabok, Croatia


Introduction: The venous plasma is the recommended sample for determining glucose. It can take up to several hours between taking a sample in the physician’s office and transporting it to the laboratory. Comparison was made between glucose concentration in serum and venous plasma.

Materials and Methods: Blood samples were collected in serum tubes without anticoagulant – Vacuette Z Serum Sep Clot Activator, Greiner Bio-One and in a test tube containing several inhibitors of glycolysis (EDTA, NaF, citric acid and Na citrate) - Vacuette Glucomedics, Greiner Bio-One. They were analyzed after it has been more than one hour from the time of sampling. Concentration of glucose was determined by enzymatic UV test on BC AU680.

Results: A total of 279 serum samples and 279 samples of venous plasma concentration in range of 3.7 to 15.8 mmol/L glucose were analyzed. Average time from sampling till analysis was 4.3h (1 to 6h). Pearson correlation coefficient r=0.9935 with P<0.05, indicate a strong correlation between the concentration of glucose. Passing Bablok analysis indicates the existence of a constant (+4.722) and proportional differences (-7.6%) between the measured values, but with different sign was obtained. Bland-Altman analysis showed that more than 95% of the results were within ±1.96SD. Differences between glucose concentrations in the area of ​​higher concentration are greater (it is lower in venous plasma compared to serum).

Conclusion: No statistical significant differences where found in glucose concentration if blood sampling is done via serum tubes or in the tube with glycolytic inhibitor. In the higher concentration range glucose is lower in the test tube containing a glycolytic inhibitor than in serum specimens which is not in accordance with the manufacturer’s declarations regarding to stability of samples. It is possible to obtain blood sample for the glucose determination in serum tubes only if the analysis is made within six hours of sampling.

e-mail: bojanakranjcec [at] gmail [dot] com


I07 (Oral presentation)

Data analysis of surrogate markers for detection of K-EDTA contamination

Lora Dukić, Nora Nikolac, Ana-Maria Šimundić

University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: There is ongoing debate among laboratory professionals on one of the recommended steps in phlebotomy procedure – order of draw. Many contradictory data on K-ethylenediaminetetra acetic acid (K-EDTA) contamination of samples as consequence of non-compliance to order of draw are published. Analytes commonly used as surrogate markers of K-EDTA contamination are potassium (K), calcium (Ca) and magnesium (Mg). Aim of our study was to assess frequency of K-EDTA contamination on samples analyzed in our laboratory during one-year period by data analysis of surrogate markers.

Materials and Methods: We extracted data from our Laboratory Information System (LIS) for one–year period (from April of 2014 to April of 2015). Creatinine, potassium, calcium and magnesium results were analyzed for patient samples collected at hospital wards and outpatient samples. Greiner Bio One collection devices were used for collection of samples. Samples were analyzed on Abbott Architect c8000 biochemistry analyzer. Creatinine concentration >150 μmol/L was defined as cut-off for kidney function disorder, while concentration of K ≥5.5 mmol/L was considered as hyperkalemia. Cut-off for hypocalcaemia was Ca <2.14 mmol/L and for hypomagnesaemia it was Mg <0.65 mmol/L. Data were analyzed using Microsoft office excel 2003 program.

Results: Data from 50435 samples were analyzed. Out of total number (N=50435), there were found 812 (1.6%) samples with hyperkalemia. For those samples (N=391, 0.8%) having creatinine concentration <150 μmol/L, pseudohyperkalemia was suspected. In that sample group, we found N=31 (0.06%) samples having hypocalcaemia, while hypomagnesaemia was not found.

Conclusion: Based on this retrospective data analysis on surrogate markers we conclude that K-EDTA contamination of samples is not preanalytical error in our laboratory.

e-mail: lora [dot] dukic [at] gmail [dot] com



The influence of inadequate sample volume on complete blood count results

Dragana Antončić1, Doris Ožanić1, Vesna Šupak Smolčić1,2,
Lidija Bilić -Zulle

1Clinical Department of Laboratory Diagnostics, Clinical Hospital Centre Rijeka, Rijeka, Croatia

2Department of Medical Informatics, Rijeka University School of Medicine, Rijeka, Croatia


Introduction: The inadequate ratio of blood and liquid anticoagulant due to insufficient blood volume can lead to erroneous results. The aim of this study was to investigate the influence of insufficient sample volume on complete blood count (CBC) results.

Materials and Methods: The study was conducted in the Department of laboratory diagnostics, CHC Rijeka from December 2014 to March 2015. All the samples were taken in the BD Vacutainer® (Becton Dickinson and Company, Franklin Lakes, NJ, USA) with liquid K3EDTA anticoagulant (volume=3mL). CBC was measured on hematology analyzer ADVIA 2120i (Siemens, Dublin, Ireland). Each insufficient sample (<3 mL) was re-sampled to correct volume (N=57). Paired t-test was used for comparison of erythrocytes, hemoglobin, hematocrit, MCV, platelets and Wilcoxon paired test for leukocytes. Significance level was set to P<0.05. Clinically significant difference was evaluated according to reference change value (RCV).

Results: There is a statistically significant difference for all the parameters except for MCV and leukocytes. Mean values of the first and second sample (X±SD) and associated P values are: erythrocytes (3.81±0.71×1012/L vs. 3.88±0.71×1012/L; P<0.001), hemoglobin (110±21g/L vs. 112±20g/L; P<0.001); hematocrit (0.331±0,065L/L vs. 0.340±0,066L/L; P=0.002), MCV (87,8±6,5fL vs. 87.6±6,5fL 6,5fL; P=0.287), platelets (213±162×109/L vs. 219±162×109/L; P=0.019) and leukocytes (median (range)): (7.0 (4.6-10.0)×109/L vs. 7.6 (4.5-11.0)×109/L, P=0.360). There was no clinically significant difference for any parameter. RCV values are calculated according to analytical coefficients of variation (CVa) of internal quality control (December 2014 - March 2015). The measured bias (%)/RCV/CVa (%) are: erythrocytes 2.0/9.2/0.9; hemoglobin 1.9/8.4/1.0; hematocrit 3.0/8.3/1.3; MCV -0.2/4.3/0.7; platelets 2.3/25.8/1.9; leukocytes 2.0/32.0/1.8.

Conclusion: There is no clinically significant difference between samples despite statistically significant differences for most parameters. However, results of the study should be interpreted with caution because clinical decision limits can not always be interpreted according to the biological variability especially in therapeutic procedures.

e-mail: dragana [dot] antocic [at] gmail [dot] com



I09 (Oral presentation)

Stability of lactate in a plasma sample - aliquot or NaF?

Ivana Rako, Ana Mlinarić, Gordana Fressl Juroš, Dunja Rogić

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Glycolysis continues after phlebotomy and results in decreased glucose and increased lactate plasma levels. In order to prevent changes in lactate concentration, NaF is the most commonly used stabilizer. In UHC Zagreb, blood for lactate determination is collected in Li-heparin tube, centrifuged within 15 minutes of sampling, aliquoted and stored at +4°C until delivery to laboratory. The goal was to find out if there are statistically and/or clinically significant differences in the concentration of lactate measured in an aliquot of heparinized plasma (AHP), heparinized plasma which is not separated from cells (HP), and plasma from blood collected in NaF/K3EDTA tube (NaFP) recommended for lactate determination.

Materials and Methods: In 52 patients with suspected hypoxia, blood was collected in Li-heparin and NaF/K3EDTA tubes. Lactate was measured on Roche Cobas C501 analyser in all samples. A statistically significant difference was tested using the Wilcoxon’s rank sum test (P<0.05). Clinically significant difference was tested using Passing-Bablok regression analysis.

Results: Lactate values were between 0.34 and 5.66 mmol/L (median AHP 1.17; 95%CI 1.0887-1.2575; HP 1.38; 95%CI 1.2800-1.5875; NaFP 1.01; 95%CI 0.8725-1.1300). Wilcoxon’s rank sum test results have shown that there is a statistically significant difference in the concentration of lactate measured in AHP and HP compared to NaFP (P<0.0001). Passing-Bablok regression analysis for values in AHP compared to NaFP has shown that there is a constant difference in lactate concentration (95%CI 0.018-0.163; y=0.1005+1.0342x), which is not clinically significant. Also, there was a proportional difference between values ​​in the HP compared to NaFP (95%CI 1.157-1.420; y=0.0744+1.280x), which is unacceptable in view of the narrow reference interval and large intraindividual variation of 27%.

Conclusion: We can conclude that AHP can be used as a sample to determine lactate levels, if pre-analytical requirements are met, because it shows an acceptable deviation of results in relation to the recommended blood sampling.

e-mail: irako [at] kbc-zagreb [dot] hr 



Risk assessment for manual handling of hemolyzed samples on patient safety

Marijana Miler1, Nora Nikolac1, Ana Helena Lukšić1, Ana Bakliža2,
Lora Dukić
1, Ana-Maria Šimundić1

1University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

2Department of Laboratory Diagnostics, Psychiatric Hospital “Sveti Ivan”, Zagreb, Croatia


Introduction: Manual handling of hemolyzed samples is not standardized and is error prone. This could lead to significant impact on clinical decision and patient safety. We aimed to assess the risk for patient safety caused by laboratory errors due to manual handling of hemolyzed samples.

Materials and Methods: All laboratory samples (N=3185) were visually assessed for hemolysis by one person in the period of one week. In this retrospective analysis 25 emergency tests were included. Hemolysis was assessed by comparison with a color chart and all samples with concentration of free hemoglobin >0.5 g/L were considered hemolyzed. Data were retrospectively obtained from laboratory information system. Correct way of handling the results (results reported or sample rejected) was determined for each test. Risk assessment was done according to ISO 14971 standard with 5 risk categories (S1=minimum impact, S2=additional sampling, S3=delayed diagnosis, S4=inappropriate therapy, S5=possible fatal outcome) and frequencies of occurrence (O1<10%, O2=10-20%, O3=21-50%, O4=51-75%, O5=76-100%). Critical categories were defined as those with the highest combination of risk and occurrence rate. For these tests some immediate corrective measures are necessary to improve the patient safety.

Results: In the period of one week emergency laboratory has received 2518 test requests in 495 hemolyzed samples (15.5%). Even 780 results (31%) were incorrectly reported accounting for 4.8% of the total test volume. The majority (67%) of erroneously reported results was made available to physicians in spite of being hemolysed. Even 33% of samples were unnecessarily rejected. Tests in critical category with the highest combination of risk and occurrence rate were troponin T (S5, O5), potassium (S5, O4) and total bilirubin (S4, O4).

Conclusion: Manual handling with hemolyzed samples may lead to risk of errors in reporting results for troponin T, potassium and total bilirubin which may impact clinical decision and patient outcome.

e-mail: marijana [dot] miler [at] gmail [dot] com



J – Other


Diagnostic value and presence of IgA and IgG Helicobacter pylori antibodies in blood serum from symptomatic patients

Jordan Petrov

PZU PAVLINA, Skopje, Macedonia



Introduction: Infection with Helicobacter pylori is one of the most common infections in the world. There are lots of methods for diagnosing H. pylori infection. Beside endoscopy, Rapid urease test, Breath test, Fecal antigen test is measuring H. pylori antibodies in blood serum. The aim of this study was to investigate presence of Helicobacter pylori antibodies IgG and IgA in patients with symptoms and already diagnosed patients with Helicobacter pylori infection.

Subjects and Methods: In our study were included 64 patients, 35 of them with already diagnosed Helicobacter pylori infection. We use ELISA kits H. pylori IgA and H. pylori IgG from Monobind USA for measuring serum levels.

Results: All patients prior diagnosed have higher values for H. pylori IgG. 40 patients or 64.5% have higher values for IgG antibodies and 10 patients or 16% have higher values for IgA antibodies respectively. From the patients IgA positive 4 have IgG and IgA antibodies positive and other 6 were IgG negative.
No one of the prior diagnosed patients was positive H. pylori IgA antibodies. The results of H. pylori IgA antibodies very well correlated with prior other studies.

Conclusion: Individuals with H. pylori infection have predominant IgG immune response but is also known that there is small group seronegative patients who are IgG negative and IgA positive. We recommend testing IgA with IgG antibody test for more accurate diagnosis. IgA positive result may have very big clinical significance especially when IgG serology is negative. 

e-mail: genolabdooel [at] yahoo [dot] com



Abdominal obesity may be the independent predictor of elevated serum cystatin C levels in overweight/obese postmenopausal women

Aleksandra Klisic1, Nebojsa Kavaric1, Milovan Jovanovic1, Najdana Gligorovic-Barhanovic2, Dragica Bozovic2, Marija Matic3

1Primary Health Care Center, Podgorica, Montenegro

2Center of Clinical-Laboratory Diagnostics, Clinical Center of Montenegro, Podgorica, Montenegro

3Institute of Medical and Clinical Biochemistry, Faculty of Medicine, University of Belgrade, Serbia


Introduction: Previous studies have reported the association of cystatin C with postmenopause but it is not clear whether this association is due to increased weight gain or menopause per se. Therefore, we aimed to evaluate the association between cystatin C and sex hormones (total estradiol, total testosterone, follicle-stimulating hormone (FSH), luteinizing hormone (LH)), sex hormone-binding globulin (SHBG) and anthropometric parameters in overweight/obese postmenopausal women with normal kidney function.

Subjects and Methods: A total of 100 overweight/obese postmenopausal women (mean age 56.7±4.8 years), and 50 age-matched normal weight controls were included in this cross-sectional study. Serum cystatin C levels, creatinine, total estradiol and testosterone, FSH, LH, and SHBG were determined. Anthropometric parameters (body weight, body height and waist circumference (WC)) were measured. Body mass index (BMI) and estimated glomerular filtration rate (eGFR) were calculated. A correlation analysis by Pearson’s (r) correlation coefficient was used to determine the relationships between serum cystatin C level and other variables. Multiple regression analysis was performed to identify independent factors affecting serum cystatin C and to estimate the final predictors of its variability. A P value of <0.05 was considered as statistically significant.

Results: Overweight/obese postmenopausal women displayed higher cystatin C levels than normal weight group (P<0.001). Serum cystatin C correlated with age (r=0.308, P<0.001), BMI (r=0.391, P<0.001), WC (r=0.440, P<0.001), FSH (r=-0.209, P=0.038), SHBG (r=-0.254, P=0.011), creatinine (r=0.361, P<0.001), and eGFR (r=-0.406, P<0.001), but did not correlate with either total estradiol, or with total testosterone levels. After stepwise multiple linear regression analysis, only waist circumference (Beta=0.349, P<0.001) and eGFR (Beta=-0.377, P<0.001) were independently associated with cystatin C (R2=0.318, P<0.001).

Conclusion: Abdominal obesity, but not sex steroids, could independently predict higher serum cystatin C levels in overweight/obese postmenopausal women.

e-mail: aleksanadranklisic [at] gmail [dot] com 



Levels of human erythrocyte’s glycohydrolases are associated to oxidative stress in patients with prosthetic-joint-associated infection

Gian Vico Melzi d’Eril1, Luca Massaccesi2, Giancarlo Goi2, Daniela Erba3, Carlo Luca Romanò4, Angela Leone5, Clara Anna Linda Damele5, Rossana Stefanelli5, Alessandra Barassi1, Lorenzo Drago6,7

1Department of Health’s Sciences, University of Milan, Milan, Italy

2Department of Biomedical, Surgical and Dental Sciences, University of Milan, Milan, Italy

3DeFENS-Department of Food, Environmental and Nutrition Sciences, University of Milan, Milan, Italy

4C.R.I.O. Unit, IRCCS Galeazzi, Milan, Italy

5Laboratory of Clinical Chemistry, San Paolo Hospital, Milan, Italy

6Laboratory of Clinical Chemistry and Microbiology, IRCCS Galeazzi, Milan, Italy

7Department of Biomedical Sciences for Health, University of Milan, Milan Italy


Introduction: Primary arthroplasties of total hip and total knee improve the quality of life. However, they may fail, necessitating revision or resection arthroplasty. Infection is the most serious complication, occurring in 0.8-1.9% of knee arthroplasties and 0.3-1.7% of hip arthroplasties. A central role in the pathophysiologic “vicious cycle” of inflammation, deeply related to infection, is played by oxidative stress (OS) due to an over production of ROS. Recently has been showed that the degree of proteins O-GlcNAcylation may influence the stress response pathway; cellular levels of O-GlcNAc, regulated by O-GlcNAc transferase (OGT) and O-β-N-Acetyl-Glucosaminidase (OGA), are considered as OS sensor. Moreover, OS induces modifications of the physicochemical properties of erythrocyte (RBC) plasma membranes and of the enzyme content of the same membranes. Due to their role in signaling early membrane alterations in OS related pathologies, several plasma membrane and cytosolic glycohydrolases of human RBC have been proposed as new markers of cellular OS.

Materials and Methods: To evaluate the possible association between RBC glycohydrolases and OS, thiobarbituric acid reactive substances (TBARS), plasma hydroperoxides (ROS) and antioxidant total defenses (by lag-time method), cytosolic OGA, membrane Hexosaminidase (Hex), β-D-Glucuronidase (GCR), and α-D-Glucosidase (αGLU) (by fluorimetric assay) have been studied in 11 patients with Prosthetic-Joint-associated Infection (PJI) and in 30 matched controls.

Results: In PJI subjects plasma membrane Hex (P<0.05), GCR and αGLU activities (P<0.001) were significantly higher as well as TBARS and ROS (P<0.001), while lag-time values (P<0.01) and OGA (P<0.05) activities were significantly lower.

Conclusion: Our data confirm the strong OS in PJI, the role of considered enzymes as sensitive OS biomarkers and suggest their possible use to monitor conditions of PJI patients under therapies in order to manage and/or prevent the debridement and retention of the prosthesis.

e-mail: gianlodovico [dot] melzi [at] unimi [dot] it



J04 (Oral presentation)

Serum total bilirubin in overweight/obese postmenopausal women with metabolic syndrome

Aleksandra Klisic, Nebojsa Kavaric, Milovan Jovanovic

Primary Health Care Center, Podgorica, Montenegro



Introduction: Metabolic syndrome (MS) is associated with increased oxidative stress. Recent findings indicate that total bilirubin (TB) has important role in antioxidant defense. Therefore, we aimed to determine serum TB level in overweight/obese postmenopausal women with MS, as well as to investigate its potential association with MS components.

Subjects and Methods: A total of 100 overweight/obese postmenopausal women, mean age 56.7±4.8 years (49 without MS and 51 with MS) without diabetes, thyroid dysfunction or cardiovascular disease, and who were not using hormonal therapy or any other medication, participated in this cross-sectional study. MS was defined using International Diabetes Federation criteria. Biochemical parameters including serum TB, fasting glycemia, total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol and triglycerides were determined. Anthropometric parameters (body weight, body height and waist circumference), and blood pressure were measured. Body mass index (BMI) was calculated. A correlation analysis by Spearman’s (ρ) correlation coefficient was used to determine the relationships between serum TB level and other variables. A P value of <0.05 was considered as statistically significant.

Results: Women with MS displayed lower serum TB level as compared with the group without MS (7.90±2.81 vs. 9.53±4.68 µmol/L, respectively, P=0.047). Serum TB inversely correlated with total cholesterol (ρ=-0.201, P<0.05) LDL-cholesterol (ρ=-0.235, P<0.05), triglycerides (ρ=-0.283, P<0.01), but positively correlated with HDL-cholesterol (ρ=0.201, P<0.05) in all women. There were no significant correlations between TB and fasting glycemia, blood pressure or anthropometric indices.

Conclusion: Decreased antioxidant defense as measured by TB was observed in overweight/obese postmenopausal women with MS. Moreover, the association of TB with lipid profile in overweight/obese women suggests that TB may be a useful marker in MS.

e-mail: aleksandranklisic [at] gmail [dot] com 



Cardiotonic steroids selectively inhibit L-arginine uptake by the y+L transporter in human endothelial cells

Marietha J. Nel, Geoff P. Candy

Department of Surgery, Medical School, University of the Witwatersrand, Johannesburg, South Africa



Introduction: Nitric oxide (NO) plays many physiological roles, of which importantly, is the vasodilation of the vasculature in order to regulate blood pressure. NO is synthesized from its precursor, L-arginine, by the nitric oxide synthase (NOS) enzyme. Therefore, NO synthesis is limited by the rate of L-arginine uptake by the vascular endothelial cells. The transport of L-arginine into the endothelial cells is by means of two main cell membrane transporters namely: the y+L (high affinity/low capacity) and the y+ (low affinity/high capacity) transporters. Indeed, arginine uptake by the y+L transporter in particular, is involved in NO production. Endogenous cardiotonic steroids affect the efficiency of the Na+/K+-ATPase pump and have therefore been implicated in the etiology of hypertension since the 1970s. The cardiotonic steroid, Ouabain which was characterized as such in the 1980s, inhibits the Na+/K+-ATPase pump, but the mechanism by which this pump inhibition affects blood pressure has not yet been fully elucidated. Aim of this study was to establish whether and to what extent Ouabain influences the uptake of L-arginine by the y+L and the y+ endothelial cell membrane transporters.

Materials and Methods: Confluent human umbilical cord vein endothelial cells were depleted of arginine overnight. The initial rate of [3H]-L-arginine uptake was measured over 30 seconds in the presence/absence of Ouabain, with and without N-ethylmaleimide, an inhibitor of the y+ transporter.

Results: At a concentration of 10 µM Ouabain inhibited the y+L transport of L-arginine by 34%, while 300 µM Ouabain inhibited L-arginine transport by this transporter by 88%. In contrast, the y+ membrane transporter was not affected by Ouabain.

Conclusion: These data indicate that indeed, Ouabain and thus possibly other cardiotonic steroids such as Digoxin, do affect the uptake of L-arginine into endothelial cells. However, whether this is a direct effect on L-arginine uptake, independent of the Na+/K+-ATPase pump activity, remains to be determined.

e-mail: marietha [dot] nel [at] wits [dot] ac [dot] za 



Small nonfunctional parathyroidal cyst: diagnostic and therapeutic challenges

Ljubica Fuštar Preradović1, Milena Tadijanović Krnjaić2, Davorin Đanić3, Branka Maljković2, Ana Đanić Hadžibegović3, Marina Josipović2

1Department for Pathology and Cytology, General Hospital „Dr Josip Benčević“, Slavonski Brod, Croatia

2Department for Laboratory Diagnostics, General Hospital „Dr Josip Benčević“, Slavonski Brod, Croatia

3Department for Otorhinolaryngology, General Hospital „Dr Josip Benčević“, Slavonski Brod, Croatia


Introduction: Parathyroid cysts (PCs) account for less than 1% of all parathyroid lesions and are most commonly located along thyroid lobes, rarely at ectopic sites. PCs are important because they can pose a differential diagnostic challenge against other cystic formations of the neck. PCs can be functional (elevated serum parathyroid hormone level) and nonfunctional.

Subjects and Methods: Four cases of nonfunctional PCs are presented. All four female patients underwent physical examination and ultrasonography of the neck with ultrasound-guided fine-needle aspiration biopsy (UG-FNA). The material thus obtained was stained by the standard May-Grunwald-Giemsa method. Parathyroid hormone level was determined in aspirate and serum (ECMIA-Cobas e411), along with serum levels of total calcium, inorganic phosphates.

Results: In two asymptomatic patients, remission occurred after initial aspiration biopsy; one patient had compression syndrome with vocal cord paresis that required surgical treatment; and one patient had cyst recurrence that was surgical removed.

Conclusion: Cystic neck masses can pose a major differential diagnostic problem considering different approach, treatment method, and preoperative and postoperative follow up. Surgical treatment is necessary in case of functional and large nonfunctional PCs (due to compression syndrome), whereas individualized therapeutic approach is used in case of small nonfunctional PCs. UG-FNA, cytological analysis and determination of parathyroid hormone level in aspirate and serum are crucial for making an accurate diagnosis.

e-mail: mtadijanovic [at] gmail [dot] com 



Intrathecal chitotriosidase activity as a potential biomarker of disease progression in patients with multiple sclerosis

Ivana Lapić, Ljiljana Zaninović, Željka Vogrinc, Milica Trbojević-Čepe

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: Multiple sclerosis (MS), as an immune-mediated inflammatory disease, is characterized by the activation of resident macrophages in the central nervous system. Chitotriosidase activity is considered to be a biochemical marker of macrophage activation. The aim of our study was to assess the correlation of intrathecal chitotriosidase activity with the extent of disability symptoms in patients suffering from MS.

Subjects and Methods: The study included 170 MS patients classified into two groups according to the severity of neurological disability. Chitotriosidase activity in cerebrospinal fluid and serum samples was determined by measuring the enzymatic hydrolysis product 4-methylumbelliferone fluorimetrically at specified excitation (365 nm) and emission (450 nm) wavelengths. Chitotriosidase ratio (CSF Chit/Serum Chit) was used as the indicator of intrathecal rather than peripheral activity. Statistical analysis was performed using MedCalc and non-parametric Mann-Whitney test was used to assess differences between groups.

Results: The median of chitotriosidase activity ratio was 0.10 (95%CI 0.09-0.11) in the group of patients with minimal disability or abnormal neurological signs and 1.18 (95%CI 0.95-1.29) in patients experiencing moderate or severe disability, showing statistically significant difference (P<0.001). Our results show that a higher chitotriosidase activity ratio is associated with disability accumulation.

Conclusion: Our study underlines the importance of the innate immune response in the pathogenesis of multiple sclerosis. A significant increase of intrathecal chitotriosidase activity was found in severely disabled patients with limited physical and cognitive functions. Therefore, chitotriosidase activity ratio could be used as a potential biomarker of disease progression in patients with multiple sclerosis. Further investigations with detailed clinical data should be performed in order to establish optimal discrimination criteria.


e-mail: ivana [dot] lapic [at] hotmail [dot] com



Case report of high values of troponin-I using high sensitive troponin-I test on patients in acute conditions

Nena Peran, Lada Surjan, Dijana Pamuković Jaram, Željana Reškov, Davorka Sladić

Laboratory of Biochemistry and Hematology, General Hospital Šibenik, Šibenik, Croatia


Introduction: Troponin-I(TnI) is a sensitive marker of myocardial injury which plays great role in AIM diagnosis. Although elevated hsTnI values indicate myocardial injury, they don’t indicate the mechanism of injury. Our goal was to show all cases of elevated TnI values in patients admitted through ER, in 20 days period, using new ARCHITECT STAT High Sensitive Troponin-I assay.

Materials and Methods: This study included 67 patients (33 women, age median 83 years and 34 men, age median 78 years) with elevated hsTnI. 25 patients were diagnosed with AIM. 42 patients (22 women, age median 86 years; 20 men, age median 80.5 years) didn’t show signs of acute coronary syndrome. ARCHITECT STAT High Sensitive Troponin-I assay was used for TnI quantification on Architect i1000 analyser.

Results: Patients diagnosed with AIM showed a significant rise/fall of hsTnI levels, within 3-12 hours. In 10 patients the first mesured value was >10×99th percentile. In 3 patients only one hsTnI level was over 99th percentile. 39 patients, who didn’t show signs of AIM, didn’t have a significant rise/fall in hsTnI within 3-14 hours. 2 patients showed a significant rise, and 1 showed a significant fall. In 10 patients the first mesurement was >10×99th percentile.

Conclusion: High sensitive hsTnI is very sensitive biomarker of myocardial injury, which is often, but not always, conected with AIM. It can detect very low TnI levels. It can also detect TnI levels >99th percentile of an healthy population, within 3 h after the onset of chest pain. The rise and/or fall of TnI levels is the evidence for AIM and chronic myocardial injury differentiation. hsTnI levels can be significantly elevated, even similar to AIM values. If these values don’t change significantly and there are no clinical signs of ischemic injury, other diagnoses, conected with myocardial injury, should be considered.


e-mail:nena [dot] peran1 [at] gmail [dot] com



Serological tests in selective IgA deficiency: gluten-free diet performance indicator in celiac disease - a case report

Mirjana Fijačko1, Igor Marjanac2, Jasna Pavela1, Blaženka Dobrošević1, Silvija Pušeljić2, Vatroslav Šerić1

1Department of Clinical Laboratory Diagnostics, Clinical Hospital Centre Osijek, Osijek, Croatia

2Pediatric Clinic, Clinical Hospital Centre Osijek, Osijek, Croatia


Introduction: Performance evaluation of treatment of celiac disease in children on gluten free diet is based on the improvement of the clinical picture and the normalization of serology: IgA antibodies against tissue transglutaminase (anti-tTG IgA) and IgG antibodies against deaminated gliadin peptide (anti-dGp IgG). Recent studies have shown that serological tests do not correlate with the recovery of the duodenal mucosa. In children with selective IgA deficiency (IgA <0.07 g/L), a serological test based on IgA-specific antibodies are negative. For those cases diagnostic procedure includes test based on IgG-specific antibodies. However, serological tests based on IgG-specific antibodies (anti-tTG, anti-dGp) can be elevated despite histological recovery. Intention of this case report is to show whether serologic tests for celiac disease show recovery duodenal mucosa in gluten free diet, carried out for at least 12 months.

Subjects and Methods: 10 year old girl was diagnosed insulin-dependent diabetes and celiac disease. Since then receives insulin therapy with adequate diabetic and gluten-free diet. IgA was measured immunonephelometrically on SIEMENS ProSpec Nephelometer. Serological tests were performed on the Luminex 200 analyzer with multiplex technology on commercial BMD kits.

Results: In accordance with the low concentration of IgA (IgA <0.066 g/L, reference value 0.91-2.55) titer of anti-tTG IgA and anti-dGp IgA at the time of diagnosis was low. However, titers of anti-tTG IgG (118 AU/ml) and anti-dGp IgG (52 AU/ml) was high (reference value <15). After two years gluten-free diet, titer of anti-tTG IgA and anti-dGp IgA was low and the titer of anti-tTG IgG (86 AU/ml) and anti-dGp IgG (40 AU/ml) was high, although lower compared to the previous value.

Conclusion: The serological markers of celiac disease do not show the recovery of the duodenal mucosa in the gluten free diet. A definite answer would give biopsy, assuming that the diet is properly implemented.

e-mail: fijacko [dot] mirjana [at] kbo [dot] hr


The significance of high sensitivity cardiac troponin I in coronary and non-coronary disease

Andrea Radeljak1, Ingrid Prkačin2, Dean Mileta2, Ivona Herceg1, Jelena Starčić1, Sonja Perkov1

1Department of Medical Biochemistry and Laboratory Medicine, Merkur University Hospital, Zagreb, Croatia

2Department of Internal Medicine, Merkur University Hospital, Zagreb, Croatia


Introduction: Introducing high-sensitivity cardiac troponin I (hsTnI) in clinical use it has been found that elevated levels, above the upper limit of the reference interval(RI) (men <34.2 ng/L; women <15.6 ng/L), occur in coronary and non-coronary disease. The aim of this study was to determine the importance of hsTnI levels regarding the diagnosis and clinical condition and determine the limit of quantification for the hsTnI on the analyzer Abbott Architect i1000SR.

Materials and Methods: Retrospective analysis included 54 patients: 19 with coronary heart disease (10 with angina pectoris, 9 with acute transmural myocardial infarct), and 35 with non-coronary disease for which initial measurement and control measurement of the hsTnI level was conducted. Concentration of hsTnI was determined by a chemiluminescent microparticle immunoassay (CMIA). The limit of quantitation (LoQ) was determined according to CLSI protocol EP17-A.

Results: In the group with coronary heart disease the maximum concentration of hsTnI was measured in patients with acute myocardial infarct (186134.0 ng/L) and in the group of non-coronary disease in patients with peripheral atherosclerotic disease (4822.8 ng/L). For the following diagnoses the obtained concentration levels, presented as median and range (ng/L) for initial/control test were: acute transmural myocardial infarct (N=9): 269.9(19.6-4055.8)/903.8(35.8-186134.0); angina pectoris (N=10): 97.8(10.5-1477.5)/82.7(7.2-1363.3); atrial fibrillation waveform (N=12): 20.3(5.2-282.2)/21.6(4.5-271.6); peripheral atherosclerotic disease (N=5): 427.9(3.0-4822.8)/2081.6(3.3-4695.7); diabetes mellitus (N=7): 22.3(1.3-73.3)/20.1(1.9-80.4); chronic renal disease (N=7): 118.9(49.9-163.9)/83.9(25.9-163.0); essential hypertension (N=4): 12.25(3.3-16.7)/14.45(4.9-33.1). LoQ was verified at the concentration of 4.71 ng/L with CV% 8.41%.

Conclusion: Quantification of very low concentrations hsTnI increases the possibility of more frequent and earlier detection of acute coronary disease with mandatory monitoring the dynamics of change in relation to the initial value. The obtained preliminary results show elevated levels of concentration, regarding to the upper limit of RI depending on gender, in coronary and non-coronary disease.Therefore, hsTnI level always must be interpreted considering the overall clinical condition.

e-mail: radeljak [dot] andrea [at] gmail [dot] com



Relationship between serum chitotriosidase activity and LDL phenotype in patients with developing metabolic syndrome

Martina Horvat, Željka Vogrinc, Milica Trbojević-Čepe

Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia



Introduction: Central obesity, increased body mass index, hypertension, impaired glucose tolerance and dyslipidemia are factors of metabolic syndrome which are associated with increased risk of atherosclerosis. In patients with metabolic syndrome, prevalence of atherogenic, small dense low-density lipoprotein (LDL) particles (LDL phenotype B) is found more frequently than in healthy individuals. LDL phenotype B is additional cardiovascular risk factor while phenotype A is typical for healthy individuals. Besides vascular and metabolic components, immune system also plays important role in atherosclerotic lesions. It has been found that chitotriosidase secreted by macrophages has increased activity in atherosclerotic lesions. The aim of this study was to investigate serum chitotriosidase activity in subjects with different LDL phenotypes. Our hypothesis was that, among subjects with developing metabolic syndrome, those with LDL phenotype B would present higher chitotriosidase activity.

Subjects and Methods: LDL phenotype and chitotriosidase activity were determined in sera of 91 subjects (22-81 years) with at least two factors of metabolic syndrome. LDL phenotype was assessed by horizontal electrophoresis in non-denaturing gradient polyacrylamide gel. Chitotriosidase activity was determined by measuring enzymatic hydrolysis product 4-methylumbelliferone fluorimetrically. Differences in chitotriosidase activities between subjects with LDL phenotype A and B were tested using non-parametric Mann-Whitney test. P<0.05 was considered statistically significant.

Results: Serum chitotriosidase activity median in subjects with phenotype A (N=64) was 118 mU/mL (IQR=82) and in subjects with phenotype B (N=27) it was 179 mU/mL (IQR=161). Serum chitotriosidase activities were statistically higher in LDL phenotype B in comparison to phenotype A (P=0.03).

Conclusion: Significantly higher serum chitotriosidase activity detected in subjects with LDL phenotype B in comparison to LDL phenotype A indicates possible increased macrophage activation and more developed atherosclerotic changes. Therefore, assessment of chitotriosidase could be a useful marker for estimation of atherosclerotic risk.

e-mail: martina [dot] horvat001 [at] gmail [dot] com




MacroCK-type 1 isoenzyme - a case report

Marina Pavić1, Lara Milevoj Kopčinović1, Dragana Šegulja2,

1University Department of Chemistry, University Hospital Centre Sestre Milosrdnice, Zagreb, Croatia

2Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia


Introduction: A rare cause of elevated creatine kinase (CK) activity is the presence of macroenzymes (macroCK). MacroCK complicates laboratory results’ interpretation because elevated enzyme activity does not correlate with clinical presentation. Our aim was to report a case of macroCK-type 1 isoenzyme.

Subjects and Methods: A female patient (58 years) was admitted to the internal clinic with chest discomfort and dizziness. The patients’ history revealed high blood pressure, high fasting and postprandial glucose in the last few months and dry cough present for the last few days. Symptoms of gastroesophageal reflux disease were also present. An ECG and venous blood sampling were performed for CK, aspartate-aminotransferase (AST), alanine-aminotransferase (ALT), creatine kinase MB isoenzyme (CK-MB), cholesterol, triglycerides, troponin I (TnI), hemoglobin A1c (HbA1c) and thyrotropin (TSH) determination. CK, AST, ALT were determined using recommended methods; CK-MB, triglycerides and cholesterol using dry-chemistry methods; HbA1c, TnI and TSH using immunoassays. MacroCK was determined by polyethylene glycol (PEG) precipitation followed by agarose electrophoresis.

Results: Upon admission, the patients’ laboratory results were: elevated CK=264 U/L, CK-MB=86 U/L and CK-MB/CK=32%; mildly elevated AST=32 U/L and ALT=57 U/L; elevated HbA1c=54 mmol/mol, triglycerides=2.0 mmol/L, cholesterol=7.8 mmol/L and TSH=4.63 mIU/L. TnI was negative (<10 ng/L) and the ECG normal. Repeated laboratory results after 42 days revealed increased enzyme activities (CK=363, CK-MB=185, CK-MB/CK=51%). After PEG precipitation, residual CK activity of 50.2% was indicative of macroCK presence. MacroCK-type 1 isoenzyme (21.1%) was confirmed by electrophoresis.

Conclusion: Elevated CK, CK-MB, and ratio of CK-MB/CK greater than 25% with negative TnI, normal ECG findings and non-specific symptoms indicate the possible presence of macroCK. Early macroCK identification is important for correct diagnosis, it contributes to avoiding unnecessary diagnostic work-up and additional treatment costs. Since current data on macroCK diagnostic significance are scarce, it is essential to monitor and document this laboratory finding.

e-mail: marina [dot] pavic [at] kbcsm [dot] hr



Proadrenomedullin levels in the early phase of septic shock

Alessandra Barassi1, Raffaele Pezzilli2, Luca Massaccesi3, Giancarlo Goi3, Angela Leone4, Clara Anna Linda Damele4, Rossana Stefanelli4, Gian Vico Melzi d’Eril1

1Department of Health’s Sciences, University of Milan, Milan, Italy

2Pancreas Unit, Department of Digestive System, Sant’Orsola-Malpighi Hospital, Bologna, Italy

3Department of Biomedical, Surgical and Dental Sciences, University of Milan, Milan, Italy

4Lab of Clinical Chemistry, San Paolo Hospital, Milan, Italy


Introduction: It has been shown that proadrenomedullin is a good marker of the severity of septic shock. However, there are no data on the early changes in serum proadrenomedullin concentrations in patients with this pathology.

Materials and Methods: Twenty-one patients with septic shock and 21 healthy subjects studied as controls. Serum concentrations of proadrenomedullin, procalcitonin, ferritin, CRP and IL-6 were determined in all subjects at the initial observation. Patients with septic shock were also studied after 24 and 48 hours.

Results: The concentrations of the acute phase proteins were significantly higher in patients with septic shock than in the control subjects during the entire study period (P<0.001). Only procalcitonin significantly decreased on the third day of observation with respect to both the first day (P=0.002) and the second day (P=0.006). Proadrenomedullin (P=0.017) and IL-6 (P=0.001) showed an AUC significantly different from the null hypothesis in differentiating the patients who survived and those who did not. The sensitivity and specificity of proadrenomedullin in the assessment of death were 71.4% and 72.7%, respectively, while IL-6 had a sensitivity of 92.9% and a specificity of 60.6%.

Conclusion: Proadrenomedullin is a reliable prognostic marker in patients with shock; further studies on a more consistent number of septic patients will definitively assess whether proadrenomedullin may replace the current prognostic markers in critically ill patients with shock due to sepsis.

e-mail: alessandra [dot] barassi [at] unimi [dot] it



Human erythrocyte membrane–bound glycohydrolases activity is affected by Plasmodium falciparum products

Gian Vico Melzi d’Eril1, Luca Massaccesi2, Giancarlo Goi2, Nadia Papini3, Silvia Parapini4, Donatella Taramelli4, Angela Leone5, Clara Anna Linda Damele5, Rossana Stefanelli5, Nicoletta Basilico2, Alessandra Barassi1

1Department of Health’s Sciences, University of Milan, Milan, Italy

2Department of Biomedical, Surgical and Dental Sciences, University of Milan, Milan, Italy

3Department of Medical Biotechnology and Translational Medicine, University of Milan, Milan, Italy

 4Department of Pharmacological and Biomolecular Sciences, University of Milan, Milan, Italy

5Lab of Clinical Chemistry, San Paolo Hospital, Milan, Italy



Introduction: Plasmodium falciparum (Pf) malaria causes about 600,000 deaths each year because of complications including cerebral malaria and severe anemia. Anemia is due to parasite induced hemolysis, diserythropoiesis and reduced deformability, accelerated senescence and removal of uninfected red blood cells (RBC). Modifications of RBC membrane’s properties can be ascribed to increased oxidative stress caused by parasite heme products such as Fe(III)-protoporphyrin IX (hematin) or hemozoin, present at high concentration in the plasma of malaria patients. Several plasma membrane-bound glycohydrolases of human RBC were recognized to have a role in signaling early membrane alterations both in pathologies suggesting their use as markers of cellular oxidative stress. The present work was aimed to investigate the ability of malaria parasite products to alter RBC membrane by affecting glycohydrolases activity.

Materials and Methods: RBC from human donors were incubated for 24 hours with presence of supernatants, derived from Pf cultures, containing heme products released during parasite growth. Then RBC were treated with different concentrations of hematin (20-10-5 µg/ml) or hemozoin (10-5-2.5 µg/mL) purified from Pf cultures. Hexosaminidase, β-D-Glucuronidase, α-D-Glucosidase and acidic Sialidase activities were evaluated by fluorimetric assays; membrane fluidity by fluorescence anisotropy method.

Results: A decrease in the activity of the tested enzymes was observed after incubation of RBC with culture supernatants from different Pf strains. Moreover, a significant dose dependent inhibition was observed after treatment with either hematin or hemozoin. At the highest dose used, hematin induced a decrease in the activity of the enzymes between 65% and 55%. Likewise a decrease of RBC membrane fluidity was observed following treatment with hematin or hemozoin.

Conclusion: The decrease of glycohydrolases activity and the increase in membrane rigidity are in agreement with accelerated RBC senescence and the parasite heme products seem to be the main contributors to these membranes damages.

e-mail: gianlodovic [dot] melzi [at] unimi [dot] it 



Evaluation of plasma myeloperoxidase in patients with erectile dysfunction of arteriogenic and non-arteriogenic origin

Alessandra Barassi1, Elena Dozio2, Monica Gioia Marazzi2, Elena Vianello2, Giovanni Maria Colpi3, Umberto Solimene2,4, Clara Anna Linda Damele5, Angela Leone5, Rossana Stefanelli5, Massimiliano Marco Corsi Romanelli2,6, Gian Vico Melzi d’Eril1

1Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milano, Italy

2Dipartimento di Scienze Biomediche per la Salute, Cattedra di Patologia Clinica, Università degli Studi di Milano, Milano, Italy

3ISES – Istituto per la Sterilità e la Sessualità, Milano, Italy

4Centro di Ricerche in Bioclimatologia Medica, Biotecnologie e Medicine Naturali, Università degli Studi di Milano, Milano, Italy

5Laboratorio di Analisi, Ospedale San Paolo, Milano, Italy

6Servizio di Medicina di Laboratorio1-Patologia Clinica, Dipartimento dei Servizi Sanitari di Diagnosi e Cura–Medicina di Laboratorio, IRCCS Policlinico San Donato, Milano, Italy


Introduction: Endothelial dysfunction and the disruption of the nitric oxide-cyclic guanosine monophosphate (cGMP) pathway have been considered the early mechanisms for the development of erectile dysfunction (ED). Myeloperoxidase (MPO), a heme-containing enzyme mainly released by activated neutrophils and monocytes, may contribute to endothelial dysfunction by promoting oxidation of different substrates and thus may play a role in ED. MPO level and its correlation with different plasma biomarkers of endothelial dysfunction were studied in patient with ED of arteriogenic (A-ED) and non-arteriogenic (NA-ED) to assess potential differences between the two ED subgroups.

Materials and Methods: Diagnosis of ED was based on the International Index of Erectile Function Score. Its etiology was classified with penile echo-color Doppler at baseline and after intracavernous injection of prostaglandin E1. MPO, soluble(s) cGMP, sICAM-1, sVCAM-1 and sP-Selectin were measured by enzyme-linked immunosorbent assay.

Results: MPO concentration in A-ED was significantly higher compared to control subjects and NA-ED patients (P<0.05). Plasmatic cGMP level resulted lower both in A-ED and in NA-ED patients, whereas no difference has been observed between the two ED groups. sICAM-1 concentration resulted higher in A-ED compared both to controls and NA-ED. sVCAM-1 level was the same in controls, A-ED and NA-ED patients. sP-Selectin concentration resulted higher both in A-ED and in NA-ED patients than in controls, whereas no difference has been observed between the two ED groups. Correlation analysis indicated a positive correlation between plasmatic MPO, sICAM-1 and sP-Selectin levels.

Conclusion: MPO may represent an important link between oxidation, inflammation and cardiovascular diseases and may also represent a potential marker to distinguish between the two subgroups of ED patients. Moreover, in ED subjects circulating cGMP may reflect the local signaling dysfunction. The use cGMP as a potential marker for monitoring the disease needs further investigation.

e-mail: alessandra [dot] barassi [at] unimi [dot] it



Early retinol-binding protein levels are associated with growth changes in infants born to diabetic mothers

Gian Vico Melzi d’Eril1, Gaia Francescato2, Massimo Agosti2, Luca Massaccesi3, Giancarlo Goi3, Clara Anna Linda Damele4, Angela Leone4, Rossana Stefanelli4, Carlo Agostoni5, Alessandra Barassi1

1Department of Health’s Sciences, University of Milan, Milan, Italy

2Neonatology and NICU, Maternal and Child Department, Filippo Del Ponte Hospital, Varese, Italy

3Department of Biomedical, Surgical and Dental Sciences, University of Milan, Milan, Italy

4Lab of Clinical Chemistry, San Paolo Hospital, Milan, Italy

5Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy


Introduction: Biochemical predictors of infants’ growth changes are not available. We tested whether retinol-binding protein (RBP), docosahexaenoic acid and insulin (I) measured within 72h from birth are associated with growth changes in infants born to mothers with gestational diabetes mellitus (GDM).

Materials and Methods: Fifty six children, 32 born to diabetic mothers treated with insulin (GDM-I) and 24 born to diabetic mothers treated with diet (GDM-D), were evaluated at 0, 1, 3, 6 and 12 months of life.

Results: At multivariable regression performed using generalized estimating equations, early RBP levels and maternal body mass index were associated to average weight changes and early RBP and insulin levels to average length changes, respectively. There was no difference between GDM-I and GDM-D infants.

Conclusion: This exploratory study suggests that early RBP levels may be a predictor of growth changes.

e-mail: gianlodovico [dot] melzi [at] unimi [dot] it



J17 (Oral presentation)

Monoclonal gammophaty as an accidental finding on CSF isoelectric focusing for oligoclonal IgG used as part of the diagnostic workup for multiple sclerosis

Ines Vukasović1, Andrea Tešija Kuna1, Vanja Bašić Kes2,
Dubravka Čaržavec
3, Nada Vrkić1,

1University Department of Chemistry, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

2University Department of Neurology, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia

3University Department of Internal diseases, University Hospital Centre Sestre milosrdnice, Zagreb, Croatia


Introduction: We aimed to present three cases of monoclonal gammopathy (MG) unexpectedly found on CSF isoelectric focusing for oligoclonal IgG (OIgG) carried out as a part of multiple sclerosis (MS) diagnostic workup.

Subjects and Methods: In the period from March 1st 2014 to March 1st 2015, isoelectric focusing (IEF) on agarose gel followed by immunofixation for IgG of paired liquor and serum samples from 240 patients presented with symptoms corresponding to MS were performed using Sebia semiautomated electrophoresis system. Serum samples of patients for whom type 5 pattern (monoclonal IgG bands in CSF and serum) was found on IEF were further subjected to immunofixation (Sebia) and IgG, IgA, IgM quantification (immunoturbidimetry, Architect-c8000, Abbott). In patients with positive OIgG, antibody specificity index (ASI) for neurothropic viruses (Morbilli, Rubella, Varicella Zoster, Herpes simplex) were assesed.

Results: Out of 240 patients (F/M=166/74; age 40.2±15.3), type 5 pattern on paired serum/CSF IEF was found in three patients (F/M=2/1), in two of them in addition with type 3 pattern (indicating intrathecal IgG synthesis). Monoclonal IgG-kappa was found in two and monoclonal IgG-lambda in one patient. In patients with 3 and 5 OIgG patterns, demyelinization on magnetic resonance imaging (MRI) together with antibody specificity index (ASI) for neurothropic viruses supported the MS diagnosis in coexistence with multiple myeloma (MM) in one and monoclonal gammopathy of undetermined significance (MGUS) in other patient. In third patient MGUS was diagnosed and MS excluded.

Conclusion: Coexistence of demyelinization disease and MG (mainly MGUS) has been reported as extremely rare with controverses about reporting of type 5 OIgG pattern with intention to avoid anxiety and unnecessary examination. Our data stressed out the importance of performing the additional examinations for distinguishing MM and MGUS in order to apply therapy immediately or just follow up the patient.

e-mail: ines [dot] vukasovic [at] gmail [dot] com



Unbelievably critical value – case report

Gordana Tkalec, Lidija Ivković, Bojana Kranjčec

Medical Biochemistry Laboratory, General Hospital Zabok and Hospital of Croatian Veterans, Zabok, Croatia


Introduction: According to HKMB guidelines it is necessary to send an immediate report to the doctor who is responsible for the patient if there is a critical value in laboratories analysis that is life threatening for the patients. For glucose critical levels in serum are below 2.5 and higher than 27.8 mmol/L. When we acquired a result of 101.4 mmol/L, our first thought was to assign it to preanalytical error by contamination of the sample with i.v. fluid.

Subjects and Methods: Patient (56y) was admitted to emergency room in General Hospital Zabok in severe condition. Laboratory analyses as ordered were: complete blood count, pH analyses, gas analyses, prothrombin time, glucose, creatinine, BUN, total bilirubin, potassium, sodium, AST, ALT, GGT, AP, CK, CK-MB, Troponin I, D-dimer, C-reactive protein, ethyl alcohol and urine analysis. Patient was transferred to the intensive care unit where he was closely monitored by analysing blood glucose levels, BUN, creatinine, pH and gas analysis. After a few days the patient was transferred to another hospital because the need for dialysis.

Results: Laboratory results were pointing to hyperosmolar coma, hemoconcentration and impairment of the kidney function. First laboratory results were reported to the doctor in charge as a value of glucose greater than 40 mmol/L and after dilution and repeated analysis an accurate value was reported. Consecutive analysis in the next hours and days had the results of 93.0 mmol/L, 74.9 mmol/L, 59.3 mmol/L, 53.3 mmol/L, 41.4 mmol/L, 30.9 mmol/L, 18.4 mmol/L, 12.5 mmol/L and 4.7 mmol/L.

Conclusion: From the values acquired in that period of time it is clear that the first result of 101.4 mmol/L was accurate although we had doubts about the already mentioned preanalytical error. The doctor we were communicating with wrote in the patients charts “lab reported glucose level above 40 mmol/L”.

e-mail: laboratorij [dot] voditelj [at] bolnica-zabok [dot] hr



Significance of measuring amino acid concentrations on tandem mass spectrometer in diagnosing and follow-up treatment of argininosuccinic aciduria

Ana Škaričić1, Marija Zekušić1, Ksenija Fumić1, Karmen Bilić1,
Dunja Rogić
1, Johannes Häberle2, Danijela Petković-Ramadža3, Vladimir Sarnavka3, Ivo Barić3,4

1Department of Laboratory Diagnostics, University Hospital Centre Zagreb, Zagreb, Croatia

2Division of Metabolism, University Children’s Hospital Zürich, Zürich, Switzerland

3Department of Pediatrics, University Hospital Centre Zagreb, Zagreb, Croatia

4University of Zagreb School of Medicine, Zagreb, Croatia


Introduction: Argininosuccinic aciduria (ASA) is an autosomal recessive aminoacidopathy caused by deficiency of the urea cycle enzyme argininosuccinate lyase (ASL) that cleaves the argininosuccinic acid to arginine and fumarate. The disease presents in infancy with extreme hyperammonemia and clinical symptoms including lethargy, somnolence, appetite loss, vomiting, tachypnea, respiratory alkalosis and progressive encephalopathy. Since hyperammonemia is a condition that requires urgent therapeutic approach, it is necessary for timely diagnosis to identify the deficient enzyme of the urea cycle by measuring the concentration of amino acids in plasma and urine.

Subjects and Methods: A male infant with hyperammonemia of unknown etiology. Amino acids were quantified by tandem mass spectrometry coupled with high performance liquid chromatography (API 3200, AB Sciex; UPLC Nexera, Shimadzu) using aTRAQ™ reagent (AB Sciex). Method allows simultaneous quantification of 45 amino acids in plasma using internal standards for each amino acid. All coding regions of the ASL gene were analyzed in collaborative international center University Children’s Hospital in Zürich, Switzerland.

Results: Besides extreme hyperammonemia (1047 µmol/L, N 24-48 µmol/L), plasma amino acid analysis showed high levels of citrulline (415 µmol/L, N 9-35 µmol/L) and argininosuccinate (1308 µmol/L), which is not present in healthy individuals. Arginine concentration was borderline low (18 µmol/L, N 6-130 µmol/L). Based on these findings, the diagnosis of argininosuccinic aciduria was established. In accordance with the diagnosis, molecular genetic investigation identified non-sense mutation c.1128C>A p.(Tyr376*) in the ASL gene in a homozygous state that has not been described so far.
Conclusion: Measurement of the concentration of amino acids using tandem mass spectrometry is necessary in all conditions with hyperammonemia of unknown etiology, particularly in the differential diagnosis of urea cycle disorders. It is also of great importance to continuously measure ammonia and amino acid levels in plasma and urine as part of treatment monitoring in patients with confirmed diagnoses.

e-mail: askaricic [at] yahoo [dot] com 



Optimization of the method for determination of the catalytic activity of ecto-ATPase in human serum

Anita Somborac Bačura1, Karmela Barišić1, Monika Janković1,
Diana Salem1
, Ognjen Čulić2, Tihana Žanić Grubišić1

1Department of Medical Biochemistry and Hematology, Faculty of Pharmacy and Biochemistry, University of Zagreb, Zagreb, Croatia
2Faculty of Pharmacy and Biochemistry, University of Zagreb, Zagreb, Croatia


Introduction: Ecto-ATPases are transmembrane enzymes involved in hydrolysis of extracellular ATP and ADP. Extracellular ATP and other nucleotides and nucleosides are important signaling molecules that regulate a variety of effects in almost all tissues: smooth muscles contractility, vascular tone, immune response, nociception, neurotransmission, embryonic development, hemostasis. In addition to the membrane bound form of the enzyme, soluble active forms of ecto-ATPase may be found in body fluids such as serum. Dysregulation of ecto-ATPase participates in pathogenesis of various diseases (degenerative neurological response, acute inflammation, tumor growth and metastasis). Therefore, selective inhibition and regulation of ecto-ATPase is a new scientific field currently being explored with great interest. The aim of this study was to optimize conditions for determination of the catalytic activity of ecto-ATPase in human serum.
Materials and Methods: The catalytic activity of ecto-ATPase in human serum was measured using the ATP as a substrate. The amount of released inorganic phosphate was measured using the spectrophotometric method, with ammonium heptamolybdate and ascorbic acid as a reducing agent, at 620 nm. ‘’Pool’’ serum was used for optimization of reaction conditions, and catalytic activity of ecto-ATPase was determined in 20 healthy individuals. Statistical analysis was performed using the SigmaStat program.
Results: The satisfactory enzyme activity at 37°C was obtained with 50 mmol/L HEPES-Tris buffer, pH 7.2, containing 5 mmol/L CaCl2 and 3 mmol/L ATP, whereas the reaction was stopped with 3 mol/L trichloroacetic acid. Within-run imprecision was determined using the ‘’pool’’ serum (N=10; 89.2±19.2 nmol/ml/hr; CV=21.5%). The catalytic activity of ecto-ATPase measured in 20 healthy subjects was 63.7 (29.6-100.1) nmol/ml/hr.
Conclusion: In this work conditions for determination of the catalytic activity of ecto-ATPase in human serum were optimized, which is the basis for further research of ecto-ATPase in pathogenesis of various diseases.

e-mail: kbarisic [at] pharma [dot] hr



The detection of macroprolactin by polyethylen glycol precipitation method

Iva Lukić, Tihana Pavošević, Vesna Horvat, Sanja Mandić,
Vatroslav Šerić

Department of Clinical Laboratory Diagnostics, Clinical Hospital Centre Osijek, Osijek, Croatia



Introduction: Prolactin exists in human blood in several molecular forms and can form a macromolecular complex. Macroprolactin is a product of monomeric prolactin and imunoglobulin G association. While macroprolactin retains immunoreactivity to a certain extent in all prolactin immunoassays, it has almost no biological activity in vivo and therefore its presence is considered clinically irrelevant. The aim of this study was to perform screening procedures, using polyethylen glycol (PEG) precipitation method, to ensure routine detection of macroprolactin.

Materials and Methods: In a five month period we collected 29 serum samples with elevated prolactin concentration (≥700 mIU/L). To evaluate the prolactin concentrations and rate of recovery in the supernatant, the PEG precipitation test according to the method proposed by Olukoga and Kane (1999) was performed on all serum samples. Initial prolactin and post-PEG prolactin concentrations were measured using CMIA (Architect i1000SR, Abbott). In accordance with most literature data, samples in this study were considered positive for macroprolactinemia if recovery rate of prolactin was <40%. Recoveries of >65% were classified as predominantly monomeric and values between 40% and 65% were classified as ambiguous.

Results: Using the 40% limit for PEG precipitation, hyperprolactinemia was found to be attributed to macroprolactin in 11 (38%) of 29 hyperprolactinemic patients. Significantly increased concentrations of prolactin (defined as ≥700 mIU/L) were found in 15 (52%) cases. Furthermore, 3 samples showed PEG recoveries between 40-65% and should be classified as indeterminate and be subject to gel-filtration chromatography for a definitive diagnosis. Serum prolactin levels ranged from 700-1900 mIU/L.

Conclusion: Being one of the major causes of apparent hyperprolactinemia, screening for macroprolactin should be included in the routine management of all hyperprolactinemic patients to prevent misdiagnosis, amplification of laboratory testing and inappropriate treatment. PEG precipitation enables easy identification of macroprolactin in routine clinical practice.

e-mail: lukiciva84 [at] gmail [dot] com




Report of two recently discovered cases of cystic fibrosis; the value of the simultaneous use of diagnostic tests: chloride in sweat and fecal elastase

Irena Linarić1, Jasna Obuljen1, Oleg Jadrešin2

1Department of Medical Biochemistry and Hematology, Department of Laboratory Diagnostics, Children’s Hospital Zagreb, Zagreb, Croatia
2Department of Gastroenterology and Nutrition, Department of Pediatrics, Children’s Hospital Zagreb, Zagreb, Croatia



Introduction: Cystic fibrosis is the most common lethal inherited disease in the Caucasians which affects the secretory epithelial cells of organs, or exocrine glands. Clinical presentation characterizes chronic obstructive pulmonary disease, and in the majority of patients (85%) chronic pancreatic insufficiency. The aim is to show the importance of the determination of chloride (Cl) in sweat and concentration of fecal elastase (FE), unavoidable tests for the detection of cystic fibrosis.

Subjects and Methods: Case 1, a child five months old, received due to failure to weight, on an artificial diet, had a 5-7 mushy bowel movements a day, with no signs of fat in the stool, by previous diseases registered respiratory catarrh. Case 2, child aged two hours, the surgery was indicated due to abdominal distension and bowel perforation in doubt; postoperatively in good clinical condition with less progression of weight - a history of meconium ileus with peritonitis, nasopharyngeal aspirate collected. In both cases, in a short period of time, the concentration of Cl in sweat by pilocarpine iontophoresis method and the concentration of FE, ELISA (Schebo, Germany), have been done repeatedly.

Results: Case 1: anemia, hypoalbuminemia, without hyponatremia; Cl in sweat 1,2=111.7, 101.8 mmol/L; FE1,2,3=<15 µg/g; genotyping: ΔF508 mutation. Case 2: mild anemia, hypoalbuminemia, without hyponatremia, Pseudomonas aeruginosa isolated from nasopharyngeal aspirate; Cl in sweat 1,2=103.6, 93.7 mmol/L; FE1,2=<15 µg/g, genotyping in progress.

Conclusion: In both cases cystic fibrosis was suspected; clinical and laboratory findings justified urgent requests for analyses: the determination of chloride in sweat and concentration of fecal elastase. These analyses are fast, reproducible and relatively inexpensive and are an important step in the early diagnosis of cystic fibrosis, without necessarily genotyping. Such an approach helps to quickly care for sick children, because early diagnosis allows proper multidisciplinary interventions that positively affect the general course of the disease.

e-mail: irenalinaric16 [at] gmail [dot] com 




Speakers Bureau: a new educational resource offered by European Federation of Clinical Chemistry and Laboratory Medicine (EFLM)

Andjelo Beletić1, Elizabeta Topić2, Sverre Sandberg3, Mauro Panteghini4

1Center for Medical Biochemistry, Clinical Centre of Serbia, Belgrade, Serbia; EFLM Working Group Congresses and Postgraduate Education-Young Scientist Member

2Faculty of Pharmacy and Biochemistry, University of Zagreb, Zagreb, Croatia; EFLM Committee Education and Training-Chair

3Norwegian Quality Improvement of Primary Care Laboratories, Noklus Department of Global Public Health and Primary Care,University of Bergen, Bergen, Norway; EFLM President Elect

4Department of Biomedical and Clinical Sciences “Luigi Sacco”, University of Milano, Milano, Italy; EFLM President


Introduction: Considering the professional and scientific leadership of European Federation of Clinical Chemistry and Laboratory Medicine (EFLM), it was assumed that a pool of scientists serving as speakers for educational events organized by national EFLM member societies might be selected from EFLM Committees (C) and Working Groups (WG).

Subjects and Methods: With goal to establish the EFLM Speakers Bureau and identify eligible lecturers and topics, in 2013 the Chairs and members of EFLM C and WG have been invited to give their availability and propose topics, accompanied with information regarding language(s) used for presentation. Proposals were evaluated and selected by chairs of EFLM Committee for Education and Training and Committee for Science and finally endorsed by EFLM Executive Board (EB).

Results: In total 207 topics, grouped into 29 areas, were selected. Areas with the highest number of topics (in parentheses) were Preanalytics (15) and Cardiac Markers (15), followed by Accreditation (14), Patient Focussed Laboratory Medicine (14) and Laboratory Management (12). Topics were proposed by 40 EFLM officers from 19 EFLM member societies. All lectures were offered to be presented in English, while some of them in other languages too (i.e., French, Dutch, Croatian, Danish). EFLM EB endorsed the establishment of Speakers Bureau in June 2014.

Conclusion: Number of available speakers and areas covered by the proposed topics indicate that resources offered through the EFLM Speakers Bureau ( may represent a suitable contribution to support educational events organized by EFLM member.


e-mail: andjelo [dot] beletic78 [at] gmail [dot] com