Anastasia Chatzinikolaou

Discrimination indices as screening tests for β-thalassemic trait

The two most frequent microcytic anemias are β-thalassemic trait (β-TT) and iron deficiency anemia (IDA). Several discrimination indices have been proposed to distinguish between these two conditions. These indices are derived from several simple red blood cell indices, like red blood cell (RBC) count, mean cell volume, and RBC distribution width (RDW), as these are provided by electronic cell counters. The purpose of the study is to examine the diagnostic accuracy of six discrimination indices in the differentiation between IDA and β-TT. The six discrimination indices that were examined were as follows: Mentzer Index (MI), Green & King Index (G&K), RDW Index (RDWI), England & Fraser Index (E&F), RDW, and RBC count. We calculated these indices on 373 patients (205 men, 168 women) with β-TT and 120 patients (50 men, 70 women) with IDA, as well as their sensitivity, specificity, positive and negative prognostic value, efficiency, and Youden’s index (YI). G&K shows the highest reliability, followed by E&F, RBC count, MI, and RDWI. On the contrary, RDW completely failed to differentiate between IDA and β-TT. G&K proved to be the most reliable index as it had the highest sensitivity (75.06%), efficiency (80.12%), and YI (70.86%) for the detection of β-TT. These six discrimination indices cannot be relied on for a safe differential diagnosis between β-TT and IDA. They do have high specificity, but their sensitivity for the detection of β-TT is not satisfactory. Consequently, they cannot be used neither as a screening tool for β-TT because they could result in a significant number of false negative results.

Increased values of mean platelet volume and platelet size deviation width may provide a safe positive diagnosis of idiopathic thrombocytopenic purpura.

Introduction: It has been shown recently that platelet indices like mean platelet volume (MPV), platelet size deviation width (PDW) and platelet-to-large-cell ratio (P-LCR) are helpful in the discrimination between hyperdestructive thrombocytopenia like idiopathic thrombocytopenic purpura (ITP) and hypoproductive thrombocytopenia (HT). The aim of the study is to assess the reliability of these indices in the differentiation of ITP from other thrombocytopenias. Methods: We recruited 134 thrombocytopenic patients (69 men, 65 women) who were divided into two groups according to the underlying disease: group I (n = 63) included ITP patients, whereas group II (n = 71) included patients with HT due to myelosuppression secondary to chemotherapy for hematological malignancies. Platelet indices were derived from a Sysmex automated cell counter. Sensitivity, specificity, positive prognostic value, negative prognostic value, efficiency and Youden index were calculated. Results: Concerning MPV and PDW indices, sensitivity, specificity, positive prognostic value, negative prognostic value, efficiency and Youden index were 100% for the diagnosis of ITP. On the contrary, the values for P-LCR were significantly lower. Conclusions: MPV and PDW can be safely relied on for a positive diagnosis of ITP. MPV and PDW were superior to P-LCR.

Evaluation of use of proton pump inhibitors in Greece

INTRODUCTION Proton pump inhibitors (PPIs) are widely used for acid-related gastric diseases. However, several national studies reported increasing use of PPIs for yet unlicensed indications. AIM The aim of our study was to evaluate the extent of PPIs prescription in a Greek tertiary hospital, as well as the adherence to licensed indications according to the Greek National Drug Organization. METHODS We retrospectively studied the discharge letters of 1693 adult patients who were admitted at the First Propedeutic Department of Internal Medicine at AHEPA hospital in Thessaloniki, Greece between July 2005 and December 2006. We studied their discharge letters in order to record all cases in which antisecretory therapy (PPIs or H(2) antagonists) was prescribed, as well as to collect data about indication of PPI treatment and the type of PPI prescribed for each patient. RESULTS PPIs were prescribed in 430 patients (25.4%). In 349 patients, PPIs were prescribed for an improper indication (81.2%), mainly for prophylaxis against medications such as steroids, non-steroidal anti-inflammatory drugs, antiplatelets and warfarin. The most commonly prescribed PPI was omeprazole. CONCLUSIONS PPIs are inappropriately prescribed in Greece. In most cases, physicians prescribe PPIs for unlicensed indications and usually, they do not give specific instructions about the duration of the treatment.

Vitamins and stroke: the homocysteine hypothesis still in doubt.

During the last years, many epidemiologic studies have identified homocysteine as an independent risk factor for cardiovascular diseases like coronary events, stroke, and venous thromboembolism. Supplementation with oral folate and vitamins B6 and B12 (mainly folate) reduce plasma homocysteine levels to a significant degree. Recent clinical trials showed that vitamin supplementation leads to slower progression or even regression of atherosclerotic lesions in the carotid arteries, as confirmed by ultrasonographic measurement of carotid intima media thickness. However, the recent Vitamin Intervention for Stroke Prevention (VISP) study failed to show any clinical effect on stroke prevention. It is unclear if homocysteine-lowering therapy really has a role in the prevention of cardiovascular diseases. Large trials, which are already conducted, will probably give the definitive answer. In this review, we try to keep pace with the data that make the homocysteine hypothesis still doubtful.

An audit of tumour marker utilization in Greece

INTRODUCTION Several international organizations have published guidelines for the correct use of tumour markers in clinical practice. However, there are reports that clinicians do not adhere to these guidelines in clinical practice. The present study constitutes an audit of TM use in a major hospital in Northern Greece. Purpose of our study is to quantify the magnitude of inappropriate TM requests as well as the corresponding financial cost. METHODS We examined retrospectively all TM requests between 10/2006 and 07/2007 in the department of biochemistry of our hospital. The tumour markers included in our study were: CA 19-9, CA 125, CA 15-3, AFP, NSE, CYFRA 21-1 and CEA. RESULTS We found 9782 inappropriate TM orders. For five of them - namely CA 125, AFP, CA 19-9, CYFRA 21-1 and NSE - the proper requests were below 10%. There were 5.6 TM requests per patient. The total cost for inappropriate TM reached 239,748 euro, which corresponds to a monthly cost of 23,974euro. CONCLUSIONS There is considerable inappropriateness in the utilization of TM in Greece which corresponds to significant financial cost. Various measures should be applied in order to increase the cost-effectiveness of TM use.

Detection of CD55- and CD59-deficient granulocytic populations in patients with myelodysplastic syndrome.

Paroxysmal nocturnal hemoglobinuria (PNH) is characterized by absence of CD55 and CD59 from the surface of affected cells. PNH has been associated with myelodysplastic syndromes (MDS). The aim of our study was to estimate the prevalence of the PNH clone in MDS patients by detecting CD55 and CD59 deficiency. We studied 90 MDS patients: 19 patients with RA, 15 with refractory anemia with ringed sideroblasts (RARS), 18 with refractory anemia with excess of blasts (RAEB), 17 with refractory anemia with excess of blasts in transformation (RAEB-t), and 21 with chronic myelomonocytic leukemia (CMML). Twenty healthy individuals were also studied as the control group. We studied the PNH clone on granulocytes of these patients with the aid of flow cytometry. CD55- and CD59-deficient granulocytic populations were detected in 15.5% of MDS patients compared to 2.8% of normal individuals. Among the subgroups of the study, significant difference was present in three cases: (1) between CMML and control, (2) between CMML and RA, and (3) between CMML and RARS. These data indicate a possible association between PNH phenotype and MDS. MDS patients of worse prognosis (CMML) express more strongly the PNH clone compared to those of better prognosis (RA and RARS). Perhaps, the examination of MDS patients for the PNH clone by flow cytometry could provide us with a valuable prognostic tool.

Beta-thalassemic trait cannot be safely differentiated from iron deficiency anemia by red blood cell indices

To the Editor: We read with interest the paper (1) by Beyan et al. concerning the discrimination between b-thalassemic trait (bTT) and iron deficiency anemia (IDA) with the aid of indices based on red blood cell parameters. In particular, they incorporated red blood cell count (RBC), red blood cell distribution width (RDW), Mentzer index, Shine and Lal index, England and Fraser index (E&F), Srivastava index, Green and King index (G&K), RDW index and Ricerca index. This study was conducted simultaneously with our study which focused on a Greek adult population and included six of the aforementioned indices (2). In Beyan’s study, the most reliable index was RBC, which was the third more powerful index in our study. On the other hand, in our study G&K was the most accurate, which ranked second in the study by Beyan et al. Both studies strongly concluded that none of the indices can be relied on for a safe differential diagnosis between b-TT and IDA, as Youden index was only 73.7% and 70.8% respectively. Moreover, sensitivity for diagnosis of b-TT was only 84.8% and 75% respectively, suggesting that these indices cannot even be used as a screening tool for b-TT, as they could result in a significant number of false negative results. The fact that the two studies did not coincide on which one is the most accurate index may be explained by the different sample size which was much larger in our study (111 vs. 493 patients). Another reason may be the fact that Beyan et al. excluded IDA patients with haemoglobin levels lower than 8.7 mg ⁄ dL. The results of these two papers confirm previous work by Demir et al. in a study conducted in a paediatric population (3). In this case, the authors studied 63 children and concluded that no index was reliable in the differential diagnosis between b-TT and IDA. They found that the most reliable index was RBC, for which Youden index was only 82%. However, in a recent study (4) by Alfadhli et al., the authors concluded that E&F is highly reliable in distinguishing b-TT from IDA. In particular, they studied 103 patients and found that in the case of E&F, Youden index was 98.2%. On an attempt to explain the different results compared with Demir’s study, the authors suggested that these may be caused by the different sample size (63 in Demir’s study compared with 103 in Alfadhli’s study), as well as by the different mean age ( 16 yr respectively). However, if we compare the Alfadhli’s study with both Beyan’s and our study, these differences do not exist, as all patients were >16 yr and the sample size was 103, 111 and 493 respectively. That means that there may be a different explanation for the results in Alfadhli’s study, other than the sample size and age of the patients. Thus, we still believe that the aforementioned indices cannot provide a safe diagnostic tool for the discrimination between b-TT and IDA.

The neuroprotective role of erythropoietin in the management of acute ischaemic stroke: from bench to bedside

Recombinant human erythropoietin was produced soon after the discovery of the erythropoietin gene in 1985 and since then, it is used in various clinical conditions such as chronic renal failure. Moreover, experimental studies have shown that erythropoietin exerts neuroprotective action as well. Recently, a clinical trial yielded promising results concerning the use of erythropoietin in stroke management. In this review, we summarize the main data which suggest that recombinant human erythropoietin and its analogues may indeed have a role in stroke treatment.

Evaluation of mean platelet volume in the differential diagnosis of thrombocytopenia

Sir, we read with great interest the article by Numbenjapon et al. (2008) that was recently published in ‘International Journal of Laboratory Hematology’. In this article, the authors evaluated the use of mean platelet volume (MPV) in discriminating hyperdestructive (HD) from hypoproductive (HP) thrombocytopenia. They found that a cut-off MPV value of 8 fl would have a sensitivity of 82.3% and a specificity of 92.5% and concluded that MPV is a reliable diagnostic test to differentiate between these two conditions. A similar work, which was recently published by our group, yielded even more promising results (Ntaios et al., 2008). In particular, we investigated 134 thrombocytopenic patients and found that MPV was absolutely reliable in differentiating the two conditions at cut-off values of 9–10 fl with sensitivity and specificity of 100%. Although the results of both studies indicate that MPV is indeed a useful diagnostic test in thrombocytopenic patients, there is considerable deviation between the two studies concerning the accuracy of this index. There was a certain degree of overlap in MPV values between HD and HP patients in the Numbenjapon study, whereas there was absolutely no overlap in our study. It seems that this difference cannot be related to the MPV values in HP patients, because the mean MPV value of these patients was similar in both studies (7.2 ± 0.6 fl in the Numbenjapon study compared with 7.17 ± 0.54 fl in our study). Therefore, the discrepancy has to be attributed to the MPV values of the HD patients. Indeed, the mean MPV value of the HD patients was only 8.8 ± 1.3 fl in the Numbenjapon study compared with 11.38 ± 0.57 fl in our study. As Numbenjapon et al. mentioned in their study, recent studies demonstrated that HD patients are likely to have higher values of MPV when compared with normal reference values. However, their study did not reproduce this finding, as the mean MPV value in their HD patients was very close to the mean value of 7.9 fl (ranging between 5.6 and 10.9 fl) they used as reference. On the contrary, our mean value was well above our reference range of 7.5–10 fl. To our opinion, the deviation of the MPV mean value between the HD patients of the two studies should be attributed to the different automated cell counter used. Numbenjapon et al. used Coulter STKS, in which platelet parameters are obtained with the electrical impedance method. On the contrary, we used Sysmex XE-2100 (Sysmex Corp., Kobe, Japan), which relies not only on the electrical impedance method but also on the optical fluorescence method (Briggs, Harrison & Machin, 2007). Similar to our study, Sysmex XE-2100 was previously also used by Kaito et al., who found a mean MPV value of 12.2 ± 0.2 fl in HD patients, which is close to the mean value in our study (Kaito et al., 2005). It could be possible that large or giant platelets were excluded from the platelet count with the impedance method in Coulter STKS, thus yielding a lower MPV value compared with Sysmex XE2100, in which optical fluorescence allows their inclusion (Briggs, Harrison & Machin, 2007). This assumption was also suggested in an earlier study by Stanworth et al. (1999)which compared the two methods. To conclude, we believe that MPV is indeed a reliable diagnostic test in the differential diagnosis of thrombocytopenic patients. However, it seems that the level of accuracy depends on the platelet counting method of the automated cell counter.

Homocysteine and carotid intima-media thickness in ischemic stroke patients are not correlated

Introduction Hyperhomocysteinemia has been linked to cardiovascular morbidity and mortality by numerous authors. Whether this association is causal or not remains uncertain. The aim of the study was to investigate the association of hyperhomocysteinemia with the degree of carotid atherosclerosis in stroke patients. Methods We studied 97 Greek patients in our stroke unit who were hospitalized as a result of ischemic stroke between March 2006 and May 2007. The patients were divided into two groups: the first (52 patients) included stroke patients with serum levels of homocysteine below 15 μmol/L, but in the second group (45 patients) serum homocysteine exceeded this value. We measured carotid intima-media thickness (cIMT) in all patients and correlated it with serum homocysteine. Results The mean homocysteine concentration was 11.5 μmol/L in the first group and 21.5 μmol/L in the second group. Carotid IMT was 1.012 mm in the first group, and 1.015 mm in the second group, an insignificant difference. On the contrary, serum folate concentration was 21.3 nmol/L in the first group compared with 16.7 nmol/L in the second group (p < 0.001). VitB12 was 401 pmol/L in the first group and 340 pmol/L in the second group, a statistically significant difference (p < 0.001). Conclusions Serum levels of homocysteine were not correlated with cIMT in ischemic stroke patients. Both folate and vitB12 were decreased in hyperhomocysteinemic ischemic stroke patients.