A ROC curve for DD in third trimester SPE cases was generated (Fig. 5) and analyzed. DD cut-off that might predict SPE was 1.65 μg/ml at the time of diagnosis during the third trimester, with a sensitivity and specificity of 94% and 67% respectively (AUC = 0.828). We arrived at this cut-off using the Youden index which is maximum at 0.61. However, we decided to use a higher cut-off of 1.94 μg/ml, which will increase the specificity of the test to 75% and still maintain a decent sensitivity of 80%. This decreases the false positive rate with a good true positive rate.
In this study, we describe for the first time trimester specific reference limit of D-dimer concentration using the bootstrap method of RefVal 4.11 program, which implements the recommendations of International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) on estimation of reference intervals.[22,23] The Bootstrap method is the most substantial advance in reference interval estimation since the recommendations of IFCC (1987). We also computed the reference limit using the 95% prediction interval of logarithmized values (Table 3) since biological parameters tend to follow a log normal distribution rather than the Gaussian distribution. The effect of complicated pregnancies on DD concentration was studied and we attempted to find if DD could predict the occurrence of severe preeclampsia.
Our study, as expected confirmed the well-known fact that DD concentration increases gradually with advancing gestation.[7,8] Previous studies in this field used the non-parametric simple rank based estimates for the reference interval (2.5th–97.5th percentile), or have shown that higher cut-off value during pregnancy would increase the specificity while maintaining the sensitivity for the diagnosis of VTE.[10,25] The upper bound of the reference limit estimated was 0.80 μg/ml, 2.7 μg/ml and 5.25 μg/ml for the 1st, 2nd, and 3rd trimester respectively with bootstrap method and 0.78 μg/ml, 1.57 μg/ml, and 4.17 μg/ml using 95% PI of logarithmized values. The trimester specific reference interval could have advantage for use in clinical practice.
According to linear regression model, 41.9% of the variance in DD across the 3 trimesters could be explained by the period of gestation alone. The predicted increase was of (0.006 × POG) when POG is in days. These results agree with previous studies. The data was collected according to strict predefined inclusion and exclusion criteria to exclude any irrelevant causes of increased DD and avoid selection bias. It is also important to recall that the reference interval should be assay specific as the results are substantially different.[3,25] We utilized the PATHFAST D-dimer assay for all the subjects. During the study episode we encountered 3 confirmed cases of DVT, all of which were complicated with other pathologies known to increase DD. Therefore, additional studies are required to investigate the magnitude of rise in DD in confirmed cases of DVT.
For complicated pregnancies, diagnosis was made on admission as per the current guidelines to exclude any diagnostic biases. As the hypercoagulable state is most exacerbated during the third trimester, we included complicated pregnancies during this trimester only. The distribution of DD in the complicated pregnancy group was statistically different for SPE compared to other categories and for PROM compared to PPROM. PE is a condition, which is associated with noticeable exacerbation of the hypercoagulable state compared to normal pregnancy.[26,27] Even though the pathogenesis of PE is not fully elucidated, activation of inflammatory cytokines and coagulation pathways plays a central role. It is documented that in PE and SPE, the endothelial dysfunction leads to a rise in tPA and PAI-1,[14,28] with a net result of hypercoagulability and dampened fibrinolysis in preeclampsia. However, as a manifestation of placental insufficiency, PAI-2 which has a local role in placental functioning during pregnancy is decreased in SPE and early onset PE (<34 weeks), with a lower PAI-1:PAI-2 ratio.[12,28] Thus, in SPE, lower PAI-2 concentration up-regulates the fibrinolytic system and this corroborates the higher circulating DD concentration seen.
The mean DD in the third trimester was 1.63 ± 1.05 μg/ml, 1.27 ± 0.58 μg/ml, and 2.59 ± 1.37 μg/ml for normal, PE and SPE respectively. In SPE, DD concentration is significantly higher despite the fact that the mean gestational age was smaller compared to the other 2 categories (32.4 ± 3.3); however, the reason for lower mean DD in preeclampsia is unclear. This could probably be explained by the fact that PAI-2 concentration is lower only in SPE, that is, more fibrin clot cleavage and this depends upon the degree of placental insufficiency. In their met-analysis in 2012, Pinheiro et al included 7 studies comparing DD in PE and normal pregnancies, out of which 5 studies found slightly higher DD than in the control, whereas the 2 others have shown no difference. The elevated DD concentration in SPE is the result of exaggerated hypercoagulable state and ongoing fibrinolysis despite elevated PAI-1 concentration.
With the exceptional distribution of DD in SPE subjects, we decided to generate a ROC curve to evaluate if DD might predict SPE for that matter. Area under the curve (AUC) was 0.828, thus revealing good relationship between DD and SPE. We used the Youden index, which is maximum at 0.61, with a sensitivity and specificity of 94% and 67% respectively. The cut-off value of DD that might predict SPE was 1.65 μg/ml at the time of diagnosis during the third trimester. We decided to use a higher cut-off value of 1.94 μg/ml to increase the specificity to 75%. This higher cut-off will decrease the false positive rate but the sensitivity decreases to 80%. Pinheiro et al in their study in 2014 had an AUC of 0.938 for DD and 0.873 for PAI-1 in SPE, suggesting both the role of DD and PAI-1 in predicting SPE.
Conflictingly, the DD concentration shows no statistical difference between third trimester normal and GDM category (1.74 ± 0.97). A prior study by Bellart et al have demonstrated that DD is higher in GDM throughout the 3 trimesters with statistical difference only in the third trimester. They also found that GDM is associated with higher TAT complex, and lower protein C and S compared to normal pregnancy. Gorar et al in 2016 reported higher PAI-1 and lower tPA in GDM compared to control groups, but the results were not significant statistically. Overall, the results of prior studies are quite ambiguous to interpret. Our finding is in accordance to a study done by Pöyhönen-Alho et al in 2012.
It has been suggested that hypercoagulation is involved in the etiology of preterm labor and PPROM. Thrombin a serine protease is formed by the cleavage of prothrombin (coagulation factor II) during the clotting process, and in turn, thrombin converts soluble fibrinogen into insoluble fibrin. Based on studies that have demonstrated that thrombin stimulates uterine smooth muscle contractions both in pregnant and non-pregnant animal uterus, it has been suggested that thrombin plays a role in the pathophysiology of preterm birth in patients with intrauterine bleeding.[17,19] This was supported by the fact that pregnant women with preterm births and PPROM had significantly higher concentration of thrombin-antithrombin (TAT) complex which is an indicator of coagulation activation compared to term pregnancy. A. Keren-Politansky et al in 2014 found that PT and aPTT were significantly reduced in pregnant women with premature uterine contractions who delivered preterm compared to those delivering at term, furthermore women with PPROM had lower PT than those delivering preterm. They also found that patients with preterm labor (including PPROM) had no statistical difference in the mean DD concentration compared to term labor. In our study, subjects with PROM (1.93 ± 0.90) had similar distribution of DD as normal pregnancy; however, PPROM (1.31 ± 0.46) had statistically lower DD distribution compared to PROM. This is possibly due to the lower gestational age of PPROM patients at presentation (33.0 ± 2.81 for PPROM vs 38.8 ± 1.0, P < .005 for PROM).
There are certain limitations to this study. Despite showing that a cut-off value for DD of 1.94 μg/ml might predict severe preeclampsia before its occurrence, the diagnostic accuracy is uncertain due to small sample size. Furthermore, as the aim of this paper was generation of a screening test for severe preeclampsia, the design did not include DD concentration prior to the disease onset but only at its first recognition. Therefore, validation of these findings is required in larger cohort.
For the interpretation of an observed value, definition of an appropriate reference interval is mandatory. The trimester specific reference interval that we derived greatly increases the specificity of the test, however it needs further verification. The role of DD in pregnancy is not limited to exclude VTE as it may have additional roles for diagnosing pregnancy complications in which the hemostatic system contributes to pathophysiology. From our data we could find the obvious relationship between DD concentration and severe preeclampsia, but additional comprehensive studies are needed to shed light upon its diagnostic role.
We thank all participants recruited for this study. We would also like to thank Yang Chengwu, Research Associate Professor, Department of Epidemiology and Health Promotion, College of Dentistry, New York University for his statistical expertise which was invaluable during the analysis and interpretation of the data.
Conceptualization: Uttam Baboolall, Haiyi Liu, Xun Gong, Dongrui Deng, Fuyuan Qiao.
Data curation: Uttam Baboolall, Ying Zha, Haiyi Liu.
Project administration: Uttam Baboolall, Haiyi Liu, Ying Zha, Xun Gong, Dongrui Deng, Fuyuan Qiao.
Formal analysis: Uttam Baboolall, Ying Zha, Haiyi Liu, Xun Gong, Dongrui Deng, Fuyuan Qiao.
Software: Uttam Baboolall
Validation: Uttam Baboolall, Haiyi Liu.
Investigation: Uttam Baboolall, Haiyi Liu.
Writing – original draft: Uttam Baboolall, Haiyi Liu.
Writing – review & editing: Uttam Baboolall, Haiyi Liu, Ying Zha.
. Riley RS, Gilbert AR, Dalton JB, et al. Widely used types and clinical applications of D-dimer
assay. Lab Med 2016;47:90–102.
. Haase C, Joergensen M, Ellervik C, et al. Age- and sex-dependent reference intervals for D-dimer
: Evidence for a marked increase by age. Thromb Res 2013;132:676–80.
. Khalafallah AA, Morse M, Al-Barzan AM, et al. D-Dimer
levels at different stages of pregnancy in Australian women: A single centre study using two different immunoturbidimetric assays. Thromb Res 2012;130:e171–7.
. Kawaguchi S, Yamada T, Takeda M, et al. Changes in d-dimer
levels in pregnant women according to gestational week. Pregnancy Hypertens 2013;3:172–7.
. Szecsi PB, Jørgensen M, Klajnbard A, et al. Haemostatic reference intervals in pregnancy. Thromb Haemost 2010;103:718–27.
. Réger B, Péterfalvi Á, Litter I, et al. Challenges in the evaluation of D-dimer
and fibrinogen levels in pregnant women. Thromb Res 2013;131:e183–7.
. Hedengran KK, Andersen MR, Stender S, et al. Large D-dimer
fluctuation in normal pregnancy: a longitudinal cohort study of 4,117 samples from 714 healthy Danish women. Obstet Gynecol Int 2016;2016.
. Murphy N, Broadhurst DI, Khashan AS, et al. Gestation-specific D-dimer
reference ranges: A cross-sectional study. BJOG An Int J Obstet Gynaecol 2015;122:395–400.
. Kline JA, Williams GW, Hernandez-Nino J. D-Dimer
concentrations in normal pregnancy: new dignostic threshold are need. Clin Chem 2005;51:825–9.
. Kovac M, Mikovic Z, Rakicevic L, et al. The use of D-dimer
with new cutoff can be useful in diagnosis of venous thromboembolism in pregnancy. Eur J Obstet Gynecol Reprod Biol 2010;148:27–30.
. Pinheiro MB, Gomes KB, Dusse LMS. Fibrinolytic system in preeclampsia
. Clin Chim Acta 2013;416:67–71.
. Moiz B. A review of hemostasis in normal pregnancy and puerperium. Natl J Heal Sci 2017;2:123–7.
. Dusse LM, Rios DRA, Pinheiro MB, et al. Pre-eclampsia: Relationship between coagulation, fibrinolysis
and inflammation. Clin Chim Acta 2011;412:17–21.
. Jacobsen AF, Skjeldestad FE, Sandset PM. Incidence and risk patterns of venous thromboembolism in pregnancy and puerperium-a register-based case-control study. Am J Obstet Gynecol 2008;198:1–7.
. Phillippe M, Elovitz M, Saunders T. Thrombin-stimulated uterine contractions in the pregnant and nonpregnant rat. J Soc Gynecol Investig 2001;8:260–5.
. Chaiworapongsa T, Espinoza J, Yoshimatsu J, et al. Activation of coagulation system in preterm labor and preterm premature rupture of membranes. J Matern Neonatal Med 2002;11:368–73.
. Keren-Politansky A, Breizman T, Brenner B, et al. The coagulation profile of preterm delivery. Thromb Res 2014;133:585–9.
. Erez O, Espinoza J, Chaiworapongsa T, et al. A link between a hemostatic disorder and preterm PROM: a role for tissue factor and tissue factor pathway inhibitor. J fetal neonatal Med 2010;21:732–44.
. Gorar S, Alioglu B, Ademoglu E, et al. Is there a tendency for thrombosis in gestational diabetes mellitus? J Lab Physicians 2016;8:101–5.
. Jacobsen AF, Skjeldestad FE, Sandset PM. Ante- and postnatal risk factors of venous thrombosis: a hospital-based case-control study. J Thromb Haemost 2008;6:905–12.
. Linnet K. Nonparametric estimation of reference intervals by simple and bootstrap-based procedures. Clin Chem 2000;46(6 Pt 1):867–9.
. Solberg HE. The IFCC recommendation on estimation of reference intervals. The RefVal program. Clin Chem Lab Med 2004;42:710–4.
. Farrokh Habibzadeh, Parham Habibzadeh MY. On determining the most appropriate test cut-off value: the case of tests with continuous results. Biochem Medica 2016;26:297–307.
. Chan WS, Lee A, Spencer FA, et al. D-dimer
testing in pregnant patients: Towards determining the next “level” in the diagnosis of deep vein thrombosis. J Thromb Haemost 2010;8:1004–11.
. Sarah A. Hale, Burton Sobel, Anna Benvenuto, et al. Coagulation and Fibrinolytic system ptotein profiles in women with normal pregnancies and pregnancies complicated by hypertension. Pregnancy Hypertens 2012;152–7.
. Pinheiro MB, Carvalho MG, Martins-Filho OA, et al. Severe preeclampsia
: Are hemostatic and inflammatory parameters associated? Clin Chim Acta 2014;427:65–70.
. Khadijah Ismail S, Higgins JR. Hemostasis in pre-eclampsia. Semin Thromb Hemost 2011;37:111–7.
. de Barros Pinheiro M, Junqueira DR, Coelho FF, et al. D-dimer
: systematic review and meta-analysis. Clinica Chimica Acta 2012;414:166–70.
. Bellart J, Gilabert R, Fontcuberta J, et al. Coagulation and fibrinolysis
parameters in normal pregnancy and in gestational diabetes. Am J Perinatol 1998;15:479–86.
. Pöyhönen-Alho M, Joutsi-Korhonen L, Lassila R, et al. Alterations of sympathetic nervous system, coagulation and platelet function in gestational diabetes. Blood Coagul Fibrinolysis
Keywords:Copyright © 2019 The Authors. Published by Wolters Kluwer Health, Inc. All rights reserved.
D-dimer; fibrinolysis; preeclampsia; severe preeclampsia