THROMBOPHILIA : WOMEN-SPECIFIC REFERENCE�RANGES CAN PREVENT MISDIAGNOSIS IN WOMEN ��BY CAROLINE S.B. VEEN,1 MARC F. DURIAN,1 MARIEKE J.H.A. KRUIP,1 MUSTAFA AHMADI,1�SIZWE M. PETRONIA,1 SJEF G. VAN ASTEN,1 WILLY VISSER,2 AND MONIEK P.M. DE MAAT1*�

Sufiza Bt Jamaludin

Introduction

• Thrombophilia is a state where inherited or acquired abnormalities of the hemostatic system are present that predispose a patient to thrombosis.
• Inherited thrombophilia results from deficiencies of

a)anticoagulant factors b)antithrombin(AT)

c)protein C (PC) d)protein S (PS)

• Acquired thrombophilia is often caused by the antiphospholipid syndrome, a diagnosis that is defined by a combination of clinical and laboratory criteria.
• Clinical criteria are the presence of vascular thrombosis or pregnancy complications.
• Laboratory criteria are the presence of lupus anticoagulant (LAC) and/or anticardiolipin (aCL) and/or beta 2-glycoprotein (β2GP) antibodies of at least 12 weeks apart (3).

Cont…

• During a normal pregnancy, major changes in hemostasis are seen.

​

• Briefly, pregnancy is associated with increased concentrations of most coagulation factors, a decrease in the concentrations of some of the natural anticoagulants, and impaired fibrinolytic activity, which together induce a thrombophilic state, especially in the last trimester (4, 5).

​

• In 40% to 72% of preeclamptic women, the presence of at least one thrombophilic factor after delivery has been reported.

Cont…

• To determine whether a woman who experiences a thrombotic event during pregnancy has a congenital or acquired hemostatic abnormality, analysis of hemostatic factors is usually performed after delivery (12) when the hemostatic abnormalities related to pregnancy are expected to be normalized.

​

• Also, the immunoglobulin concentration is higher in women than in men, and antibody production in response to primary and secondary antigen stimulations seems to be more pronounced in women (19).

​

Cont…

• Diagnosing thrombophilia in young postpartum women remains a challenge, and the use of universal reference ranges may be misleading in these women (15, 16, 21).

​

• Therefore, sex specific reference ranges are recommended (16,20, 21).

​

• We hypothesize that this will affect the analysis of thrombophilia after pregnancy complications.

​

• Therefore, the aim of our study was to investigate the effect of women-specific reference ranges (WRRs) on the interpretation of hemostatic variables in postpartum women.

​

Methods

a) Participants

• To obtain WRRs, we recruited healthy women (group H) for low-risk pregnancies 3 months after an uncomplicated pregnancy [i.e., no preeclampsia (PE), hemolysis, elevated liver enzymes, low platelet count (HELLP) syndrome, or intrauterine growth retardation] from January 2009 to February 2011.
• To investigate the outcome of using WRRs in a group for whom these reference ranges can have clinical relevance, we included PE patients are routinely investigated for thrombophilia 3 months postpartum.
• PE patients (group P) were included in this study (Fig. 1).
• These women were retrospectively evaluated.

​

​

Cont…

b) Blood collection

• Blood was drawn in tubes containing 0.106mol/L citrate (1 part of citrate to 9 parts of blood) as the anticoagulant and serum tubes (BD Biosciences).
• Citrated samples were centrifuged 2 times at 2000g for 10 min at 4 °C followed by centrifuging at 14000 g for 10 min at 4 °C; the plasma and serum were stored at −80 °C until analysis.
• Testing for thrombophilia is performed 3 months after delivery, and therefore we collected blood samples from the healthy women 3 months postpartum(group H3).

Cont…

• In the healthy women, a second sample was taken at least 6 months after delivery (group H6) when hemostatic markers are no longer influenced by pregnancy (2, 14, 15, 17).

• In the healthy women, blood was drawn solely for research purposes.

​

• In the PE women, blood was drawn 3 months postpartum as part of clinical care after a complicated pregnancy.

​

• At the time of blood sampling, the healthy women provided information about the use of oral contraceptives (OCs) and medication using a questionnaire.

​

Cont…

C) Assays

• Prothrombin time (PT; Thromborel S)
• activated partial thromboplastin time lupus (aPTTL)
• lupus anticoagulant with diluted Russell viper venom time (La-DRVVT)
• protein C activity (PC) (Siemens Healthcare Products);
• antithrombin (AT-III; chromogenix, ILC)
• protein S activity (PS; Roche Diagnostica)
• activated protein C resistance (APC-resistance; Chromogenix) were performed on Sysmex CA-1500 (Siemens HealthCare Diagnostics).
• In-house ELISA assays were performed for the detection of anticardiolipin (ACL)-IgG and IgM and for β 2-glycoprotein I (β-2GP1)-IgG and-IgM antibodies.

​

​

Cont…

d) Reference ranges

• Routine reference ranges (RRRs) used in our center are based on 40 healthy male blood donors with the lower limit mean −2SD and the upper limit mean +2SD or ±3SD for lupus tests.

​

• WRRs were calculated with the Reference Value Advisor Software (v2.1), which closely follows the Clinical and Laboratory Standards Institute (CLSI) guideline (22, 23).

​

• For any series of data, Reference Value Advisor calculates and reports 5 reference intervals based on assumptions about data distribution.

​

Cont…

• For the outliers, the software uses Dixon–Reed and Tukey tests, which are recommended by International Federation of Clinical Chemistry and Laboratory Medicine (IFCC)-CLSI.

​

• In this study, the non transformed robust data were used (22).

​

• We based the WRRs for APC-resistance ratio on women without factor V Leiden (FVL), defined as women with an APC-resistance ratio above 0.8 and/or women negative as carriers of FVL.

​

Cont…

e) Statistics

• Continuous data were analyzed with the Student t-test (in the case of a normal distribution) or the Mann–Whitney U-test (in the case of a skewed distribution).

​

• Categorical data were analyzed using the McNemar test for paired data and the Pearson chi-square test for unpaired data.

​

• All tests were 2-tailed and groups were considered statistically significant if P < 0.05.

​

• Analyses were carried out using IBM SPSS Statistics Data Editor, version 21.

​

Results

• 61 women with an uncomplicated pregnancy were included (group H).

​

• In all 61 women, blood was collected 3 months postpartum (range 83–128 days) (group H3).

​

• The median age at time of delivery was 32.2 years (range 18.2–39.1).

​

• In 55 of the 61 (90%) women, a second blood sample was collected at least 6 months postpartum (range181–280 days) (group H6).

​

• The WRRs were determined based on this last measurement (Table1).

​

• Three and 6 months after delivery, 18% and 27% of the women used oral contraceptives, respectively.

​

• Seven healthy women used over-the counter medication, such as pain medication and vitamin supplements, at time of inclusion.

​

Cont…

Cont…

Women-specific reference ranges

• The WRR for AT was 0.69–1.37 U/mL, compared with an RRR of 0.80–1.20 U/mL. For PC activity, the lower limit of the WRR was 0.75 U/mL compared with the lower limit of the RRR of 0.70 U/mL.
• In contrast, for PS activity, the lower limit of the WRR was 0.57 U/mL compared with the lower limit of the RRR of 0.70 U/mL.
• The women-specific cutoff value of the APC resistance ratio was calculated to be 0.60, whereas the cutoff value of the RRR used was 0.80.
• The upper limit of the WRR for APTT-Lupus was 42 s vs 39 s for the RRR; for the DRVVT ratio, the upper limit of the WRR was 1.26 vs 1.20 for the RRR; for aCL IgM antibody, 50 U/mL was the upper limit of the WRR vs 23 U/mL for the RRR; and for β2GP-1IgG and IgM antibody, 45 U/mL and 40U/mL, respectively, were the upper limits of the WRRs vs 30 U/mL and 15 U/mL, respectively, for the RRR.

Cont…

Cont…

• In contrast, the aCL IgG antibody upper limit of the WRR was 13 U/mL compared with 21U/mL for the RRR (Table 2).

​

• In group H3, 48% of women had normal results when using routine reference ranges compared with 89% when using WRRs (P <0.05) (Table 3).

​

• The most pronounced abnormalities were found regarding PS activity (7 vs 0 abnormal values) and APC-resistance ratio (26 vs 2 abnormal values) (Table 4).

Cont..

Oral contraceptive use and women-specific reference ranges

• In our study population, women that used oral contraceptives had significantly decreased antithrombin activity 6 months after pregnancy.
• We did not find a significant difference in the other anticoagulant factors between women who used oral contraceptives and women who did not use oral contraceptives.

Preeclampsia and thrombophilia

• 197 women who suffered from PE during their last pregnancy were included (Group P). Blood was collected 3 months postpartum (range 83–142 days) as part of routine care. The median age of the PE women at time of delivery was 31.3 years (19.2– 44.7) (Table1).

Cont…

Cont…

• When using RRRs, 26% of the women in group P showed no abnormalities in anticoagulant factors and/or lupus anticoagulant tests vs up to 66% when WRRs were used (P < 0.05) (Table 3).

​

• The most pronounced abnormalities were found regarding PS activity (32 vs 4 abnormal values), APC resistance ratio (88 vs 16 abnormal values), and DRVVT ratio (54 vs 30 abnormal values) (Table 4).

​

Discussion

• The main result of our study is that over 40% of women, based on laboratory results, would be falsely classified as having thrombophilia using RRRs.
• In healthy women, when using routine hemostatic reference ranges based on healthy male or mixed-population blood donors, significantly fewer women had normal results (48%) compared with using WRRs (89%).
• To give an example of a patient group for whom the use of these WRRs can be clinically relevant, we also evaluated anticoagulant factors measured in a group of women with pregnancies complicated by preeclampsia, who are routinely evaluated for the presence of thrombophilia.
• In this group, significantly more normal results were seen when using WRRs (66%) compared with those when using routine hemostatic reference ranges (26%).

​

Cont…

• This finding indicates that WRRs should be used to interpret hemostatic variables in this group as false classification can have major consequences for future pregnancies and other prothrombotic situations throughout life.

​

• Our finding that reference ranges are different in men and women is in line with the literature as previously reported data consistently show that sex has an influence on coagulation factors (16, 18,21, 24, 25).

​

• With this present study, we concur with the suggestion of Lowe et al. (21) that sex specific reference values at least should be considered in the diagnosis of congenital thrombophilias as a significant difference was seen in our study group.

​

Cont…

• There are, however, several arguments not to have reference ranges for each variable that can have an effect on these reference ranges.
• First, some differences in reference ranges influenced by a variable are small and thus can be considered clinically irrelevant.
• Second, the IFCC recommends estimating reference ranges on at least 120 subjects (22).
• Enrolling a sufficient number of reference individuals for each variable is difficult, costly, and time-consuming.
• Third, taking into account all the variables that can have an influence on reference ranges (e.g., age, dietary habits, ethnicity, environment, use of medication, pregnancy and menopause, muscle mass, and body mass index), it is impossible to calculate reference ranges for each of these different variables and implement them into clinical practice.

Cont…

• Our study had several limitations and strengths.

​

• To determine the WRRs, we included 55 healthy women with uncomplicated pregnancies.

​

• Guidelines recommend to use samples of 120 individuals for determination of reference ranges.

​

• However, it is not always possible to obtain the suggested number of 120 individuals of a specific group to define reference ranges (23).

​

• Therefore, the revised CLSI guideline introduced determination of reference ranges from smaller reference samples based on a robust method, preferably after transformation of the data to a distribution that is closer to gaussian or normal(22, 23, 26).

​

​

Conclusion

​

• The use of WRR should allow a more accurate definition of true congenital thrombophilia and prevent misclassification of thrombophilia, which can have aggravating clinical consequences.

​

​

​

Thank you