Thrombotic markers in pregnant patients with and without SARS-CoV-2 infection

Abstract

Introduction:

COVID-19 is associated with coagulation abnormalities and increased risk for venous and arterial thrombi. To evaluate D-dimer levels and lupus anticoagulant (LAC) positivity in pregnant individuals with and without SARS-CoV-2 infection.

Materials and Methods:

This was a prospective cohort study of pregnant individuals delivering at a single academic institution from April 2020 to March 2022. Individuals with a positive SARS-CoV-2 result during pregnancy were compared with a convenience sample of those without a positive SARS-CoV-2 result. For individuals with SARS-CoV-2 infection, severity was assessed based on the National Institutes of Health classification system. The primary outcome was D-dimer level measured during delivery admission. The secondary outcomes were LAC positivity and thromboembolic events. Outcomes were compared between individuals with and without a positive SARS-CoV-2 result, and further by disease severity.

Results:

Of 98 participants, 77 (78.6%) were SARS-CoV-2 positive during pregnancy. Among individuals with SARS-CoV-2 infection, severity was asymptomatic in 20 (26.0%), mild in 13 (16.9%), moderate in 4 (5.2%), severe in 38 (49.4%), and critical in 2 (2.6%). The D-dimer concentration at delivery did not significantly differ between those with a SARS-CoV-2 positive result compared to those without (mean 2.03 μg/mL [95% CI 1.72–2.40] vs 2.37 μg/mL [95% CI 1.65–3.40]; p=0.43). Three individuals (4%) with SARS-CoV-2 infection and none (0%) without infection were LAC positive (p=0.59). There were no clinically apparent thromboses in either group. D-dimer concentrations and LAC positive results did not differ by COVID-19 severity.

Conclusions:

Thrombotic markers did not differ in pregnant individuals by SARS-CoV-2 infection; however, high rates of LAC positivity were detected.

Introduction

Individuals with Coronavirus disease 2019 (COVID-19) are at increased risk for both venous and arterial macro- and micro-thrombi.^1–4 Coagulation abnormalities and inflammatory changes resulting from SARS-CoV-2 infection are attributed as mechanistic pathways for increased rates of thromboembolism, although the pathophysiologic process is not fully understood.^5–7 As a result, prophylactic and therapeutic anticoagulation are in widespread use for hospitalized patients with SARS-CoV-2 infection.^8–11

Changes in thrombotic markers, including elevated D-dimer levels and presence of antiphospholipid antibodies, among individuals with SARS-CoV-2 infection have been identified.^12,13 Higher D-dimer levels are associated with severe COVID-19 and mortality.^14,15 Case series describe 11–91% of individuals hospitalized with SARS-CoV-2 infection as positive for lupus anticoagulant (LAC).^16–18 Although antiphospholipid antibodies may transiently appear during acute viral infections, persistence of these antibodies was identified in some studies, raising concerns about long-term health implications.^19–21 The clinical utility of routine testing for thrombotic markers in individuals with COVID-19 to inform risk of thrombosis and need for anticoagulation remains uncertain.²²

Thromboembolic risk related to COVID-19 extends to pregnancy and the puerperium.^23–25 Pregnant and postpartum individuals with SARS-CoV-2 infection are at increased risk for thromboembolism, as compared to those without SARS-CoV-2 infection.²⁶ The risk for thromboembolism in the obstetric population is further associated with COVID-19 severity.²⁷ It is unknown whether similar changes in thrombotic markers seen in non-pregnant individuals with SARS-CoV-2 infection occur, and potentially contribute to thromboembolism, in pregnant individuals. LAC positivity has additional relevance in obstetric populations based on the known association with thrombotic risk and poor pregnancy sequelae (e.g., stillbirth, miscarriage, fetal growth restriction).^28–29 The prothrombotic physiology of pregnancy and postpartum make understanding changes in the thrombotic profile during SARS-CoV-2 infection in this population clinically relevant.^28,30 Therefore, we aimed to evaluate D-dimer levels and LAC positivity in pregnant individuals with and without SARS-CoV-2 infection.

Methods

This was a prospective cohort study of pregnant individuals aged 18 or older delivering at a single academic institution between April 2020 to March 2022. Individuals were excluded if unable to provide consent in English or Spanish. There were no additional exclusion criteria. Eligible individuals were approached upon admission to Labor and Delivery and provided written, informed consent.

The exposure was SARS-CoV-2 infection during pregnancy. Participants were considered exposed if they had any of the following: positive SARS-CoV-2 result on Labor and Delivery universal screening or patient self-report of a positive SARS-CoV-2 test at any time during the pregnancy. During the COVID-19 pandemic, universal SARS-CoV-2 screening was implemented on Labor and Delivery as part of routine clinical care for patient and clinician safety. All patients admitted to the hospital underwent screening via rapid antigen testing and polymerase chain reaction (PCR) testing. Self-report of a positive SARS-CoV-2 test during pregnancy was confirmed with medical record abstraction when possible. Participants were considered SARS-CoV-2 uninfected if their Labor and Delivery universal screening was negative, they had no self-report of SARS-CoV-2 infection, and they had no prior SARS-CoV-2 positive results by medical record abstraction. All individuals with SARS-CoV-2 positive results were approached for enrollment, while convenience sampling was used to enroll the control group of participants without a positive SARS-CoV-2 result.

Participants with SARS-CoV-2 infection were further classified as having an active infection if SARS-CoV-2 positive result was during or within 10 days prior to delivery admission, or as having a recovered infection if SARS-CoV-2 positive result was greater than 10 days prior to delivery admission. For individuals with SARS-CoV-2 infection during pregnancy, COVID-19 severity was assessed based on the National Institutes of Health (NIH) classification as asymptomatic, mild, moderate, severe or critical (Appendix A).³¹ Severity classification was informed by patient reported symptoms and medical record abstraction of COVID-19 clinical status for those admitted during infection (e.g., oxygen requirements, imaging results).

The primary outcome was D-dimer level (μg/mL) measured during delivery admission. The secondary outcome was LAC positivity during delivery admission. Blood was collected after enrollment on Labor and Delivery during the delivery hospitalization. Assays for D-dimer concentration and LAC were performed through ARUP Laboratories. LAC testing was by a series of assays following the International Society on Thrombosis and Haemostasis (ISTH) guidelines with final interpretation as positive or negative.³² Individuals with SARS-CoV-2 infection peri-delivery received heparin-based pharmacologic prophylaxis postpartum. Blood samples for the study were obtained prior to thromboprophylaxis initiation.

Additional secondary outcomes included LAC positivity at 5 to 12 weeks postpartum and clinically diagnosed thromboembolism within 6 weeks postpartum. Antiphospholipid antibodies may be present transiently. Therefore, an additional blood sample was collected at 5 to 12 weeks postpartum to reassess LAC positivity. Thromboembolism was defined as pulmonary embolism or deep vein thrombosis by standard clinical imaging studies (e.g., computed tomographic angiography, lower extremity Doppler).

Trained perinatal research staff abstracted data from medical records including demographics, medical and surgical history, obstetric history, pregnancy course, COVID-19 vaccination status, SARS-CoV-2 test results, clinical laboratory studies, and COVID-19 clinical course. COVID-19 vaccines were clinically available beginning in December 2020. For SARS-CoV-2 positive individuals, risk factors for disease and quality and severity of reported symptoms were assessed through a patient survey. The electronic survey was adapted from the Centers for Disease Control and Prevention (CDC) case report form (Appendix B).

Baseline characteristics were summarized and compared between participants with and without positive SARS-CoV-2 result during pregnancy using χ² or Fisher exact tests for categorical variables, and student’s t test for continuous variables. Due to a right-skewed distribution, D-dimer was analyzed on the log-scale, with summary statistics reported as geometric mean with 95% confidence interval. In secondary analysis, outcomes were further compared between those with active versus recovered SARS-CoV-2 infection at delivery admission, and by disease severity.

Statistical significance was defined as p<0.05. This study was approved by the Institutional Review Board of the University of Utah (#00139802). Study data were managed using the Research Electronic Data Capture (REDCap) system hosted at University of Utah.³³ All analyses were performed using SAS version 9.4 and graphics created using GraphPad Prism version 9.0. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines were followed.³⁴

Results

Of 98 individuals meeting inclusion criteria and consenting to participate, 77 (78.6%) were SARS-CoV-2 infected. Sixty-nine patients (89.6%) classified as SARS-CoV-2 infected had confirmatory positive laboratory test results from medical record abstraction or Labor and Delivery universal screening.

Of individuals with SARS-CoV-2 infection, 41 (53.2%) were designated with an active infection at delivery admission while 36 (46.8%) were designated as having a recovered infection at delivery admission. The median time from SARS-CoV-2 positive result to delivery was 9 days (IQR 1–148 days). For those with active infection at delivery, the median time from SARS-CoV-2 positive result to delivery was 1 day (IQR 0–6 days), while for those classified with a resolved infection at delivery, the median was 151 days (IQR 12–207 days). Among participants with SARS-CoV-2 infection, severity was classified as asymptomatic in 20 (26.0%), mild in 13 (16.9%), moderate in 4 (5.2%), severe in 38 (49.4%), and critical in 2 (2.6%) (Figure 1).

Figure 1.

Study population

Overall, 67 (68.4%) participants had received at least one COVID-19 vaccination. Individuals with SARS-CoV-2 infection were more likely to be younger and have public health insurance (Table 1). Individuals without SARS-CoV-2 infection had a lower mean hemoglobin (12.14 ± 1.10 g/dL vs 12.81 ± 1.16 g/dL; p=0.02) but other clinically available laboratory studies did not differ by SARS-CoV-2 infection (Table 2).

Table 1.

Baseline characteristics between individuals with and without SARS-CoV-2 infection

Characteristic	SARS-CoV-2 Positive N=77	SARS-CoV-2 Negative N=21	P-value
Age (years)	29.12 ± 5.43	32.48 ± 5.17	0.014
Weight (kilograms) (mean, 95% CI)	86.49 (82.62–90.55)	89.23 (81.02–98.27)	0.551
BMI Category			0.974
Normal	7 (9.1)	1 (4.8)
Overweight	24 (31.2)	6 (28.6)
Class I obesity	23 (29.9)	7 (33.3)
Class II obesity	12 (15.6)	3 (14.3)
Class III obesity	11 (14.3)	4 (19.0)
Hispanic ethnicity*	16 (20.8)	1 (4.8)	0.11
Race*
American Indian/Alaskan	1 (1.3)	0 (0.0)	0.99
Asian	3 (3.9)	0 (0.0)	0.596
Caucasian/White	56 (72.7)	20 (95.2)	0.037
Pacific Islander/Hawaiian	3 (3.9)	0 (0.0)	0.596
Unknown	8 (10.4)	1 (4.8)	0.679
Insurance
Self-pay	2 (2.6)	0 (0.0)	0.99
Private	54 (70.1)	20 (95.2)	0.02
Public	21 (27.3)	1 (4.8)	0.037
Nulliparity	27 (35.5)	7 (33.3)	0.99
Multifetal gestation	0 (0)	1 (4.8)	0.21
Mode of delivery			0.771
Cesarean	17 (22.1)	4 (19.0)
Spontaneous vaginal delivery	56 (72.7)	17 (81.0)
Operative vaginal delivery	3 (3.9)	0 (0)
Preterm delivery (<37 weeks)	7 (9.2)	0 (0)	0.159
Medical comorbidity
Chronic hypertension	2 (2.6)	2 (9.5)	0.2
Hypertensive disorders of pregnancy	15 (19.5)	4 (19.0)	0.99
Depression or anxiety	28 (36.4)	13 (61.9)	0.047
Renal disease	0 (0.0)	0 (0.0)	–
Heart disease	0 (0.0)	0 (0.0)	–
Gestational diabetes	9 (11.7)	2 (9.5)	0.99
Fetal growth restriction	2 (2.6)	1 (4.8)	0.99
Fetal demise	1 (1.3)	0 (0.0)	0.99
Assisted reproductive technology	5 (6.5)	6 (28.6)	0.02
COVID-19 vaccination	49 (63.6)	18 (85.7)	0.154
Low dose aspirin use	9 (11.7)	6 (28.6)	0.084

BMI, body mass index.

Data are mean ± standard deviation, or n(%), unless otherwise noted.

^*Self-reported by participant.

Table 2.

Laboratory studies between individuals with and without SARS-CoV-2 infection

Laboratory Study	SARS-CoV-2 Positive N=77	SARS-CoV-2 Negative N=21	p-value
WBC count (/L)	9.50 ± 2.80	9.41 ± 1.92	0.868
Hemoglobin (grams/dL)	12.81 ± 1.16	12.14 ± 1.10	0.021
Hematocrit (%)	38.52 ± 2.98	37.15 ± 2.73	0.055
Hemoglobin/Hematocrit ratio	0.33 ± 0.01	0.33 ± 0.01	0.052
Platelet count (/mL)	230.47 ± 69.66	235.65 ± 75.64	0.784
Creatinine (mg/dL)*	0.57 ± 0.11	0.55 ± 0.13	0.684

Mean ± standard deviation

^*Missing values for 55 in SARS-CoV-2 positive and 13 of SARS-CoV-2 negative

The D-dimer concentration did not significantly differ at delivery between those with a SARS-CoV-2 positive test compared to those without a positive test (mean 2.03 μg/mL, 95% CI 1.72–2.40 vs 2.37 μg/mL, 95% CI 1.65–3.40; p=0.43; Table 3; Figure 2A). Three participants (4%) with SARS-CoV-2 infection and none (0%) without SARS-CoV-2 infection were LAC positive at delivery (p=0.59). From the three LAC positive participants at delivery, one remained positive postpartum, one converted to LAC negative postpartum, and one participant did not have a longitudinal blood sample. There were no thromboses in either group.

Table 3.

D-dimer and lupus anticoagulant positivity by SARS-CoV-2 infection

Outcome	SARS-CoV-2 Positive N=77	SARS-CoV-2 Negative N=21	p-value
D-dimer µg/mL	2.03 (1.72–2.40)	2.37 (1.65–3.40)	0.428
Lupus anticoagulant positive	3 (4.0)	0 (0.0)	0.592

Data as mean (95% confidence interval) or n(%)

Figure 2.

D-dimer concentration by SARS-CoV-2 infection status

(A) D-dimer concentration (μg/mL) for SARS-CoV-2 positive and negative individuals

(B) D-dimer concentration (μg/mL) by SARS-CoV-2 active and recovered infection at delivery

(C) D-dimer concentration (μg/mL) by COVID-19 severity

In secondary analyses, D-dimer concentration did not differ between those with active SARS-CoV-2 infection (geometric mean 2.0 μg/mL, 95% CI 1.6–2.5) versus recovered SARS-CoV-2 infection (geometric mean 2.1 μg/mL, 95% CI 1.6–2.6) at delivery admission (p=0.69; Figure 2B). LAC positivity also did not differ by SARS-CoV-2 active versus recovered infection (n=2/41 (4.9%) vs n=1/36 (2.8%), p=0.79). By COVID-19 severity, D-dimer concentrations were 2.1 μg/mL (95% CI 1.7–2.8) for asymptomatic to mild, 2.0 μg/mL (95% CI 1.6–2.5) for moderate to severe, and 1.3 μg/mL (95% CI 0–38.0) for critical (p=0.63; Figure 2C). LAC positive results did not differ by COVID-19 severity (p=0.31).

Discussion

In this single center prospective cohort study, D-dimer concentration did not differ between individuals with and without SARS-CoV-2 infection during pregnancy. The LAC positive rate among those with SARS-CoV-2 infection was 4% but did not significantly differ from those without SARS-CoV-2 infection. Further, D-dimer concentration and LAC positivity did not differ by active versus recovered SARS-CoV-2 infection, nor by COVID-19 severity. In the two LAC positive patients at delivery with longitudinal data available, LAC positivity persisted into the postpartum period in one. There were no clinical thromboses reported in any of the participants.

D-dimer has become an important thrombotic marker in non-obstetric patients with SARS-CoV-2 infection. In a case series of individuals with laboratory-confirmed SARS-CoV-2 infection in China, elevated D-dimer levels (≥0.5 μg/mL) were identified in nearly half (n=260/560, 46.4%) of patients.² Subsequent studies found higher D-dimer levels in individuals with SARS-CoV-2 infection were associated with increased risk for thrombi, intensive care unit (ICU) admission, and mortality.^{14,15,35–37} Using D-dimer to inform thrombotic risk, three open-label randomized controlled trials compared therapeutic to prophylactic dose heparin in hospitalized patients with SARS-CoV-2 and elevated D-dimer levels, finding therapeutic heparin resulted in a reduction in mortality, thromboembolism, and cardiorespiratory organ support.^38–40 National guidelines now recommend therapeutic dose heparin for non-pregnant individuals hospitalized with SARS-CoV-2 infection requiring oxygen with elevated D-dimer levels.¹¹

However, data on the association between SARS-CoV-2 infection in pregnant individuals and D-dimer levels are limited and mixed, and similar national guidelines for heparin-based prevention of thrombosis informed by D-dimer concentration are not in place.⁴¹ The baseline peri-partum reference range for D-dimer is higher (0.5 – 1.45 μg/mL) than for non-pregnant individuals reflecting the known physiologic prothrombotic state of pregnancy.^42–44 In a case series of 61 pregnant individuals with SARS-CoV-2 infection, D-dimer levels were elevated among those with symptoms (n=17/32, 53.1%) compared to those without symptoms (n=4/29, 13.8%) (p<0.001).⁴¹ There was not a SARS-CoV-2 negative comparison. In another single center retrospective cohort study of 227 pregnant and postpartum individuals with SARS-CoV-2 infection, Lombardi et al identified D-dimer concentrations above normal across four timepoints during infection but D-dimer concentration was not associated with oxygen supplementation.⁴⁵ We found D-dimer levels did not differ by SARS-CoV-2 infection status, nor COVID-19 severity. We hypothesize that the baseline increase in D-dimer concentration in pregnancy limits the clinical utility of this thrombotic marker in those further infected with SARS-CoV-2. Findings may also reflect less virulent SARS-CoV-2 variants, or COVID-19 immunization protection.

Similar to D-dimer, case reports early in the COVID-19 pandemic described the presence of antiphospholipid antibodies among individuals with SARS-CoV-2 who experienced arterial and venous thromboses.^12,46 Case series identified 11.1% to 91% of non-pregnant adults with SARS-CoV-2 infection as LAC positive.^16–18,21 In a single center retrospective cohort study of individuals with SARS-CoV-2, there was a significantly higher rate of thromboses among LAC positive individuals (n=19/30, 63.3%) compared with LAC negative individuals (n=13/38, 34.2%; p=0.03).²¹ Unlike D-dimer, however, LAC positivity has not been sufficiently predictive of thrombotic risk to inform COVID-19 management guidelines.^22,47

Studies to date on LAC positivity in pregnant individuals with SARS-CoV-2 infection are limited. In a single center prospective cohort study of 151 individuals admitted to Labor and Delivery with SARS-CoV-2 infection (symptomatic or asymptomatic) tested for antiphospholipid antibodies (LAC, anti-beta2glycoprotein I, and anti-cardiolipin), Gozzoli et al found 16 patients had at least one positive antibody of which nine individuals (6.0%) were LAC positive.⁴⁸ Those with a positive antiphospholipid antibody had a higher rate of preeclampsia compared with those without a positive antiphospholipid antibody (25% vs 5%; p=0.004). There was a single thrombosis identified in an antiphospholipid antibody negative patient.⁴⁸ Again, there was no uninfected comparison group. Our results are similar with a 4% LAC positivity rate among individuals with SARS-CoV-2 infection and no thromboses. Although the LAC positivity rate did not significantly differ by SARS-CoV-2 infection, the rate of LAC positivity among those with SARS-CoV-2 infection (4%) was higher than anticipated. Prior literature estimates an upper normal rate of 1–1.4% of LAC positivity among uncomplicated pregnancies.^49,50 These findings warrant further study in a larger population.

Antiphospholipid antibodies may transiently appear during acute viral infections as described for other viruses including hepatitis and human immunodeficiency virus (HIV).^19,46 Antiphospholipid antibodies in those with SARS-CoV-2 infection are theorized to result from an inflammatory cytokine response pathway.²⁰ Persistence of antibodies may suggest the virus hastens progression to antibody development in predisposed individuals and helps to identify those at risk for pathogenic sequelae. In three LAC positive patients with 12-week follow up testing, Gozzoli et al found all were LAC negative.⁴⁸ We identified LAC positivity persisted in one of two patients with follow-up data. It is unknown if persistent LAC positivity in this population has clinical implications for subsequent pregnancies or thrombotic risk.

This study has several limitations. While larger than prior studies addressing thrombotic markers in pregnant individuals with SARS-CoV-2 infection, the sample size is nonetheless modest and findings should be considered exploratory. A convenience sampling approach was used for the SARS-CoV-2 uninfected group. Sampling occurred across various SARS-CoV-2 variants but there was insufficient sampling in each period to analyze data by variant (e.g., Delta, Omicron). There is a possibility that those included in the SARS-CoV-2 uninfected group could have experienced an asymptomatic SARS-CoV-2 infection. The population studied was heterogenous in medical comorbidities and pre-existing risk factors for thrombotic marker elevation. We had insufficient sample size to explore time-dose relationships and inclusion of individuals with recovered infections may bias results toward the null secondary to diminished thrombotic risk further from the time of infection.

This study also has strengths. Unlike prior studies, we included a comparison group of individuals without a positive SARS-CoV-2 result. The study spanned time periods when different SARS-CoV-2 variants predominated increasing generalizability of findings. Detailed medical record abstraction was used to obtain confirmation of SARS-CoV-2 testing and disease course, as well as demographics and medical history.

In conclusion, D-dimer concentration and LAC positivity did not differ by SARS-CoV-2 infection in pregnant individuals in this single center cohort study. While thrombotic markers have a useful role in informing thrombotic and morbidity risk in the non-obstetric population with SARS-CoV-2 infection, their utility to inform care in obstetric patients with SARS-CoV-2 infection remains limited. With the known morbidity associated with LAC positivity in obstetric patients, additional study of LAC positivity and persistence post-acute SARS-CoV-2 infection is warranted.