Evaluation of protein Z plasma level in beta-thalassemia major patients in Ahvaz city in Iran

Abstract

OBJECTIVES: Thrombotic episodes occurred frequently in beta-thalassemia major (BTM) patients, leading to hypercoagulability of plasma. Protein Z (PZ) is a vitamin-K-dependent anti-coagulation factor that plays a role in the human homeostatic process. The objective of the current study is to investigate the distribution pattern of PZ plasma concentrations between BTM patients and the normal population in Ahvaz city, the center of Khuzestan province, southwest of Iran.

MATERIAL and METHODS: Forty confirmed BTM patients and 40 healthy volunteers were evaluated for complete blood count (CBC) indices and PZ plasma levels. CBC samples were measured using an automated cell counter, and PZ was assayed with an immunoassay method. Statistical analysis was conducted using SPSS software. The ROC curve and binary logistic regression estimated the sensitivity, specificity, and Odd’s ratio for PZ measurement.

RESULTS: The mean±SD of the PZ plasma level in normal individuals was 1.68±0.63 µg/mL, and in BTM patients, it was 1.10±0.52 µg/mL. This shows a significant reduction of PZ in BTM patients statistically (CI= 0.99; p<0.001). Further, the mean±SD of the PZ plasma levels in BTM patients who received washed red blood cells was not significantly different from that of patients undergoing packed red blood cell therapy (CI= 0.95; p = 0.320). The area under the curve (AUC) for PZ was 0.759 (p = 0.00). The cut-off value= 1.4 µg/mL of the PZ plasma level had at least 70% sensitivity and specificity in BTM patients.

DISCUSSION: Several epidemiologic studies have shown thromboembolism episodes in BTM patients. In the current study, PZ was reduced significantly in BTMs.

CONCLUSION: We noticed that BTMs have lower plasma PZ concentration might be predisposed to BTM.

Keywords

major beta-thalassemia / protein Z / thrombosis / immunoassay / anti-coagulation factor

Introduction

The deficiency or absence of thebeta chain of hemoglobin leads to major beta-thalassemia, which isa kind of severe and chronic anemia (Weatherall, 1976). Thalassemia affects various organsand is associated with a considerable increase in morbidity and themortality rate (Cunningham et al.,2004). Beta-thalassemia major is a severe form of thalassemia,and patients are dependent on blood transfusion throughout their lifespans (Steinberg, 2001). Thalassemia is a vastly heterogeneous disease because of variousmutations of the beta globin gene and other disease-associated factors.It is well known that BTM patients undergo an increase in the rateof thrombotic episodes. These episodes include transient ischemia,stroke, and arterial and vascular thrombosis, which in turn resultin an increased rate of hypercoagulability (Michaeli et al., 1992). The relativelyhigh prevalence of co-inherited a-thalassemia and hemoglobinopathies among b-thalassemia carriers reveals the importance of molecularanalysis to diagnose double heterozygous for prenatal diagnostic purposes(Alizadeh et al., 2014). A lack of information about prenatal testing can cause thalassemiamajor births, which can be prevented if we increase awareness of screeningand prenatal genetic diagnosis services (Mendiratta et al., 2017). Various hemostatic disturbancesmediate these episodes, including abnormalities of the red blood cell(RBC) membrane and charge (Kuypersand de Jong, 2004), the presence of activated platelets(Winichagoon et al., 1981), a low or high plasma level of coagulation, and anti-coagulationfactors (Eldor and Rachmilewitz, 2002). It is shown that a decreased level of anti-coagulation factorssuch as anti-thrombin III (AT-III), protein C (PC), and protein S(PS) induces the coagulation process (Schettini et al., 1987). Protein Z (PZ) is a vitamin-K-dependentanti-coagulation factor (Broze andMiletich, 1984) that is necessary for the function ofthe Z-dependent protease inhibitor (ZPI) (Yin et al., 2000). In the presence of protein Z, ZPIinactivates the coagulation factor Xa (Han et al., 2000). In the presenceof Z protein, the inhibitory function of ZPI on the Xa factor increasesby about 1000 times (Han et al., 1999). Published studies proposed that a decrease in the plasma levelof PZ may have an important role in the occurrence of bleeding episodes(Kemkes-Matthes and Matthes, 1995), deep vein thrombosis (DVT) (Kemkes-Mattheset al., 2002), early fetal loss (Gris et al., 2002), and stroke (Vasse et al., 2001).

The current study showed that PZplasma levels are significantly decreased in BTM patients comparedto those of local normal cases. There are some questions about PZ:

1. Does a decrease in the level ofPZ play a role in coagulation induction and the occurrence of thrombosisepisodes?

2. How much of a decrease in thePZ plasma level causes clinical complications?

3. Is PZ necessary as an anti-thrombosisprophylactic agent for thromboembolism prevention in BTM patients?If yes, then at which cut-off point should PZ therapy be started astreatment?

To address these questions, normalPZ ranges should be determined for each local area. In addition, morestudies are necessary among various subpopulations of beta thalassemiaaccording to the presence or absence of thrombosis or splenectomyhistory.

According to what we mentioned ProteinZ existence can decrease bleeding, and decrease in protein Z can beone reason of bleeding in thalassemia patients. Our goal is to runa study to evaluate the protein Z level in thalassemia patients. Alsothere are limited studies on PZ plasma levels in beta thalassemiapatients in Iran. Therefore, we studied and compared the PZ plasmalevels and complete blood count (CBC) indices between cases and normalcontrol individuals to estimate Odd’s ratio and to propose PZas a probable hypercoagulability index in BTM patients.

Material and methods

Study groups

In the current study, 40 beta-thalassemiamajor (BTM) patients (18 males and 22 females) and 40 control individualswere evaluated for CBC indices and plasma levels of PZ. The patient’sage was 17.78±7.76 years, and the control’s age was 18.38±8.75years. All patients and controls who participated in this work completedconsent forms and permitted us to use their samples for analyzingand publishing data.

Patients and samples

Forty BTM cases were selected fromthe research center of thalassemia and hemoglobinopathy hospital ofAhvaz Jundishapur Medical University. Inclusion criteria were chosenaccording to the hemoglobin (Hb), mean corpuscular volume (MCV), meancorpuscular hemoglobin (MCH), Hb-electrophoresis in alkaline pH, familyhistory, and beta-globin gene analysis by molecular technique. Allof these data were present in the patients’ history documents.Exclusion criteria were current fever, infection, or increased serumactivities of alanine amino transferase (ALT) and aspartate aminotransferase(AST).

Forty normal controls were selectedfrom patients referred to the hospital laboratory according to theinclusion criteria Hb>12 g/dL, MCV>80 fl, MCH>30 pg, and a negativehistory of bleeding, thrombosis, cardiovascular, or hepatic problems.Control individuals were matched with BTM patients for age and sex.All BTM patients were transfusion dependent. Some of them receivedwashed red cells and the ot 56+ hers packed cell on a monthly basis.Plasma samples were collected from patients at least 30 days afterthe last transfusion to minimize the transfused blood effect on thePZ and CBC indices. Whole blood samples were collected in EDTA-K2containing tubes. Complete blood counts (CBC) were done using an automatedprocedure (Mindray apparatus). Plasma samples were separated by centrifugationand kept at -70 ºC until theday of the test.

ELISA method and statistical analysis

Plasma PZ concentrations were determinedby using an enzyme-linked immunosorbent assay (ELISA) kit (ZYMUTESTProtein Z, HYPHEN Biomed) and according to the operating manual ofthe kit insert.

First of all, there are numbers ofprinciples as mentioned in the kite the immunoconjugate, which isa polyclonal antibody specific for PZ coupled to horse radish peroxidase(HRP), is introduced into the microwells coated with a polyclonalantibody specific to PZ. Then, the diluted test sample is immediatelyintroduced, and the immunological reaction starts. When present, PZbinds onto the polyclonal-antibody-coated solid phase through oneepitope, and fixes the polyclonal antibody coupled to HRP throughfree epitopes. Following a washing step, the peroxidase substrate,3,3′,5,5′–Tetramethylbenzidine (TMB), in the presenceof hydrogen peroxide (H₂O₂), is introduced, and a blue color develops. When the reaction isstopped with sulfuric acid, a yellow color is obtained. The amountof color developed is directly proportional to the concentration ofhuman PZ in the tested sample.

According to the kite, we used 50microliters (µL) of conjugate anti (h)-PZ-HRP to introduce theanti-(h)-PZ- HRP immunoconjugate in the micro ELISA plate wells. Thenwe added 200 µL of the tested sample introduced immediately,or the tested samples in the corresponding micro ELISA plate well.Next, we mixed them gently on a plate shaker or manually, and incubatedthem for 1 h at room temperature (18–25 ºC). Then, 300µL of wash solution was used five times in a wash procedureusing a washing instrument. Immediately after the washing, we introduced200 µL of TMB/H₂O₂ substrate. We incubated for exactly 5 min at room temperature (18–25ºC). We stopped the color development by introducing 50 µLof 0.45 M sulfuric acid. We waited for 10 min to allow the color tostabilize, and measured the absorbance at 450 nm (A450). We subtractedthe blank value first.

Data analysis was conducted usingSPSS Ver. 20 software for descriptive statistics, a comparison ofmean±SD between groups, correlations, and Odd’s ratioestimations. Furthermore, a receiver operating characteristic (ROC)curve was plotted to estimate a cut-off value of the PZ protein forthe hypercoagulability state in BTM patients.

Results

Odd’s ratio (OR) value forPZ (OR= 5.40; 2.20 to 13.22) suggests that a PZ assessment is valuablefor BTM patients. PZ could be about 2.2- to 13.22-fold lower in BTMpatients than in normal people. Furthermore, a cut-off value of 1.4µg/mL PZ could be considerable in clinical evaluations. A lowervalue indicates a predisposition to hypercoagulability, and highervalues may indicate a normal state of coagulation cascade.

The mean±SD for plasma levelsof PZ were 1.10±0.524 and 1.68±0.63 for BTM and controlindividuals, respectively. There was a significant difference betweenthe two groups statistically for PZ plasma levels (CI= 0.95, p = 0.000) (Fig. 1).

The mean of the PZ plasma levelswas lower in BTM patients who received washed red cell therapy thanpatients who received packed cell therapy. However, the differencewas not statistically significant (CI= 0.95, p = 0.32). Table 1 lists CBC data and PZ plasma levelsin BTM and normal individuals. Furthermore, the table shows comparativeP-vales and Odd’s ratio from binary logistic regression. A highvalue of Odd’s ratio for PZ was obtained, in addition to someCBC indices.

The average platelet count was notsignificantly different in the two studied groups (CI= 0.95, p = 0.365), but the difference was significantfor Hb, hematocrit, and MCV, which were lower in BTM patients (Table1). There was no meaningful correlation between PZ and other hematologicindices for BTM or normal individuals. In addition, the duration oftransfusion therapy did not show a correlation with PZ protein (Table2).

Figure 2 shows the ROC curve of PZ.The area under the curve (AUC) for PZ was 0.759 (p = 0.00). ROC analysis showed that when considering aPZ plasma level of 1.4 µg/mL, the sensitivity and specificityof the test is about 70%. In fact, 28 cases (70% of BTM patients)and 10 control individuals (28% of controls) had PZ plasma levelsthat were less than 1.4 µg/mL. Therefore, this value is proposedas a predictor for the hypercoagulability state in BTM patients.

Discussion

Current therapeutic approaches havethe potentiated life expectancy and survival of BTM patients (Olivieri and Brittenham, 1997). However,these approaches are accompanied with new complications such as thromboembolism,pulmonary embolism, deep vein thrombosis, and portal vein thrombosis(Michaeli et al., 1992; Gillis et al., 1999; Kemkes-Matthes et al., 2002). Several epidemiologic studies have shown thromboembolism episodesin BTM patients (Borgna-Pignatti etal., 1998; Moratelliet al., 1998). Current studies showed that PZ was reducedsignificantly in BTM patients over normal individuals (Schettini et al., 1987; Cappellini et al., 2000; Singer et al., 2006; Hassan et al., 2010). A decreasein PZ may predispose these BTM patients to thromboembolism episodes.

Thus, PZ can be a prophylactic indexfor hypercoagulability states in BTM patients. Hepatic sidrosis couldbe the reason for a reduction in PZ production in BTM patients. Itwas investigated that erythrocytes of splenectomized BTM patientshave more phosphatidyl serine in their outer membranes (Kuypers and de Jong, 2004). Thiscould result in the sequestration of anti-coagulation proteins suchas protein Z by the phosphatidyl serine, and in turn increase thePZ turnover, as reported for proteins C and S (Cappellini et al., 2000; Eldor and Rachmilewitz, 2002).

There are conflicting reports onthe plasma level of PZ and its role in coagulation induction, especiallyin BTM patients. Studies have indicated a significant associationbetween decreased levels of PZ and thromboembolism episodes such asvascular-arterial thrombosis and DVT (Santacroce et al., 2006; Pardos-Gea et al., 2008). However, such findings donot agree with other investigations (Al-Shanqeeti et al., 2005; Elkhateeb et al., 2009). One of the most importantdifferences between these studies is the employment of different cut-offpoints for reporting decreased levels of PZ. An obvious differencein cut-off points could be seen in Elkhateeb et al. (2009) (1.13 µg/mL).

Such differences in the cut-off valuecould cause a bias in reporting Odd’s ratio (OR) and falselyoverestimate reports of the thrombosis rate. These problems may stemfrom the wide reference range intervals that have been reported forPZ plasma concentrations. This phenomenon could be a result of vastintraindividual variability, which in turn arose from gene polymorphismin different investigated populations. As a result, it would be rationalto determine reference ranges of PZ for each population accordingto their ethnicity and polymorphism. Such determinations help to findproper cut-off values to evaluate PZ plasma levels. This result wasconfirmed by Miletich and Broze (1987) but is not in accordance with Heebet al. (2002).

Conclusion

In our study, age and PZ plasma levelsdid not have a significant association. However, we suggested 1.4µg/mL cut-off values in the current study. More investigationshould be performed to find a more precise cut-off value and the benefitsof PZ determination. Then, comprehensive studies with larger samplesizes, in addition to polymorphism determination, are necessary toclarify such discrepancies. We reported that there was no associationbetween most CBC indices, the duration of transfusion therapy, thetype of blood product for transfusion therapy, and decreased levelsof PZ in BTM patients. Significant differences in CBC indices betweenthe case and control categories were not novel findings and were expected.

References

[1] Al-Shanqeeti A, van Hylckama Vlieg  A, Berntorp E,  Rosendaal F R,  Broze G J Jr (2005). Protein Z and protein Z-dependent protease inhibitor.Determinants of levels and risk of venous thrombosis. Thromb Haemost, 93(3): 411–413

[2] Alizadeh S, Bavarsad  M S, Dorgalaleh  A, Khatib Z K,  Dargahi H,  Nassiri N,  Hamid F,  Rahim F,  Jaseb K,  Saki N (2014). Frequency of beta-thalassemia or beta-hemoglobinopathycarriers simultaneously affected with alpha-thalassemia in Iran. Clin Lab, 60(6): 941–949

[3] Borgna-Pignatti C, Rugolotto  S, De Stefano P,  Piga A, Di Gregorio  F, Gamberini M R,  Sabato V,  Melevendi C,  Cappellini M D,  Verlato G (1998). Survival and disease complications in thalassemia major. Ann N Y Acad Sci, 850(1): 227–231

[4] Broze G J Jr,  Miletich J P (1984). Human Protein Z. J Clin Invest, 73(4): 933–938 160;

[5] Cappellini M D,  Robbiolo L,  Bottasso B M,  Coppola R,  Fiorelli G,  Mannucci A P (2000). Venous thromboembolism and hypercoagulability in splenectomized patientswith thalassaemia intermedia. Br J Haematol, 111(2): 467–473

[6] Cunningham M J,  Macklin E A,  Neufeld E J,  Cohen A R, and the Thalassemia Clinical Research Network (2004). Complications of -thalassemia majorin North America. Blood, 104(1): 34–39 

[7] Eldor A, Rachmilewitz  E A (2002). The hypercoagulable state in thalassemia. Blood, 99(1): 36–43160;

[8] Elkhateeb M, Elmenshawy  I, Azzam H,  Selim T,  Abousamra N K,  Elsharawy S (2009). Plasma levels of protein Z and antiphospholipid protein antibodies in patientswith ischemic stroke. Egypt J Neurol PsychiatNeurosurg, 46(1): 11–17

[9] Gillis S, Cappellini  M D, Goldfarb  A, Ciceri L,  Fiorelli G,  Rachmilewitz E A (1999). Pulmonary thromboembolism in thalassemia intermedia patients. Haematologica, 84(10): 959–960

[10] Gris J C, Quéré  I, Dechaud H,  Mercier E,  Pinçon C,  Hoffet M,  Vasse M,  Marès P (2002). High frequency of protein Z deficiency in patients withunexplained early fetal loss. Blood, 99(7): 2606–2608160;

[11] Han X, Fiehler  R, Broze G J Jr (2000). Characterization of the protein Z-dependent protease inhibitor. Blood, 96(9): 3049–3055

[12] Han X, Huang  Z F, Fiehler  R, Broze G J Jr (1999). The protein Z-dependent protease inhibitor is a serpin. Biochemistry, 38(34): 11073–11078 

[13] Hassan T H, Elbehedy  R M, Youssef  D M, Amr  G E (2010). Protein C levels in beta-thalassemiamajor patients in the east Nile delta of Egypt. Hematol Oncol Stem Cell Ther, 3(2): 60–65 

[14] Heeb M J, Paganini-Hill  A, Griffin J H,  Fisher M (2002). Low protein Z levels and risk of ischemic stroke: differences by diabetic status and gender. Blood Cells Mol Dis, 29(2): 139–144

[15] Kemkes-Matthes B, Matthes  K J (1995). Protein Z deficiency: a new cause of bleeding tendency. Thromb Res, 79(1): 49–55

[16] Kemkes-Matthes B, Nees  M, Kühnel G,  Matzdorff A,  Matthes K J(2002). Protein Z influences the prothrombotic phenotype inFactor V Leiden patients. Thromb Res, 106(4-5): 183–185 

[17] Kuypers F A, de Jong  K (2004). The role of phosphatidylserine in recognition and removalof erythrocytes. Cell Mol Biol (Noisy-le-grand), 50(2): 147–158

[18] Mendiratta S L,  Mittal M,  Naaz F, Singh  S, Anand S (2017). Role of thalassemia screening in prevention and control of thalassemia-a5 year experience. Int J Reprod ContraceptObstet Gynecol, 5(9): 3107–3111

[19] Michaeli J, Mittelman  M, Grisaru D,  Rachmilewitz E A (1992). Thromboembolic complications in beta thalassemia major. Acta Haematol, 87(1-2): 71–74

[20] Miletich J P, Broze  G J J Jr (1987). Human plasma protein Z antigen: range in normal subjectsand effect of warfarin therapy. Blood, 69(6): 1580–1586

[21] Moratelli S, De Sanctis  V, Gemmati D,  Serino M L,  Mari R, Gamberini  M R, Scapoli  G L (1998). Thrombotic risk in thalassemic patients. J Pediatr Endocrinol Metab, 11(Suppl 3): 915–921

[22] Olivieri N F, Brittenham  G M (1997). Iron-chelating therapy and the treatment of thalassemia. Blood, 89(3): 739–761 

[23] Pardos-Gea J, Ordi-Ros  J, Serrano S,  Balada E,  Nicolau I,  Vilardell M (2008). Protein Z levels and anti-protein Z antibodies in patients with arterial andvenous thrombosis. Thromb Res, 121(6): 727–734

[24] Santacroce R, Sarno  M, Cappucci F,  Sessa F,  Colaizzo D,  Brancaccio V,  Grandone E,  Margaglione M (2006). Low protein Z levels and risk of occurrence of deep vein thrombosis. J Thromb Haemost, 4(11): 2417–2422

[25] Schettini F, De Mattia  D, Arcamone G,  Sabato V,  Altomare M,  Burattini M G,  Fedele F,  Compagnone A,  Ciavarella G (1987). Coagulation contact phase factors and inhibitors in -thalassemia major children. Pediatr Hematol Oncol, 4(3): 231–236

[26] Singer S T, Kuypers  F A, Styles  L, Vichinsky E P,  Foote D,  Rosenfeld H (2006). Pulmonary hypertension in thalassemia: association with platelet activation and hypercoagulablestate. Am J Hematol, 81(9): 670–675

[27] Steinberg M H (2001). Disorders of hemoglobin: genetics, pathophysiology, and clinical management. J R Soc Med, 94(11): 602–603

[28] Vasse M, Guegan-Massardier  E, Borg J Y,  Woimant F,  Soria C (2001). Frequency of protein Z deficiency in patients with ischaemic stroke. Lancet, 357(9260): 933–934 

[29] Weatherall D J (1976). Molecular pathology of the thalassemia disorders. West J Med, 124(5): 388–402 

[30] Winichagoon P, Fucharoen  S, Wasi P (1981). Increased circulating platelet aggregates in thalassaemia. Southeast Asian J Trop Med Public Health, 12(4): 556–560 

[31] Yin Z F, Huang  Z F, Cui  J, Fiehler R,  Lasky N,  Ginsburg D,  Broze G J Jr (2000). Prothrombotic phenotype of protein Z deficiency. Proc Natl Acad Sci USA, 97(12): 6734–6738 

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag GmbH Germany, part of Springer Nature