Anti-Xa assay monitoring for thromboprophylaxis in critical care patients with Covid-19 pneumonia

Infection with SARS-CoV-2 is associated with an inflammatory response and prothrombotic state. The incidence of venous thromboembolism is significant, with acute pulmonary embolus identified in 30% of screened patients. ¹ D-dimer levels correlate with disease severity, ² and are particularly raised in non-survivors. ³ Current guidelines advocate use of prophylactic anticoagulation in all patients presenting with SARS-CoV-2 pneumonia at standard dose, with some suggesting an intermediate dose in high-risk patients or where biomarkers such as D-dimer are significantly raised.^4,5

In our tertiary cardiothoracic hospital, our role has been to support neighbouring units by receiving ventilated patients with SARS-CoV-2 and continuing management. In the face of an increasing body of anecdotal evidence, we monitored effectiveness of thromboprophylaxis with enoxaparin via assessment of anti-Xa levels. Patients required higher doses of enoxaparin to achieve a prophylactic effect than we expected, and so a higher dose regime was adopted by our unit.

Data are presented from the most recently admitted 10 patients receiving enhanced dose thromboprophylaxis from admission. Initial enoxaparin dose was chosen according to presenting D-dimer and calculated ideal body weight (D-dimer > 3000 ng/mL, 1.5 mg/kg OD; D-dimer 1000–3000 ng/mL, 1 mg/kg OD; D-dimer <1000 ng/mL, standard BD prophylactic dose). Anti-Xa levels were measured 4 h post the third enoxaparin dose, and where dose changes were required following the third dose again. The target anti-Xa level was 0.2–0.4 IU/mL. Where the level was within the target range, it was repeated every 48–72 h. Where levels were outside of the target range, the dose was titrated by 10 mg or 10% of the total dose for extremes of weight, and the level was re-assessed 4 hours post the third dose. In addition to routine daily analysis of full blood count and clotting profile, D-dimer levels were assessed regularly. This service evaluation was registered within our institution, and in view of its observational nature requirement for ethical approval was waived.

Levels of anti-Xa for a given dose of enoxaparin were lower than expected. To achieve a prophylactic anti-Xa level the median dose of enoxaparin required was 1.8 mg/kg IBW (range 1.0–2.0 mg/kg IBW). Mean admission SOFA score was 11.5. Higher doses of enoxaparin were required to achieve a prophylactic effect with a rise in SOFA score, as demonstrated by a fall in anti-Xa level as SOFA score increased (rho −0.49, p = 0.01). As patients clinically improved they achieved higher anti-Xa levels, allowing gradual dose reduction. This was a small cohort of patients, but no bleeding complications were reported. Eight patients survived to ITU discharge and two died.

Use of anti-Xa monitoring to guide thromboprophylaxis has been discouraged as existing evidence does not suggest an associated reduction in thrombotic or bleeding events. While applicable to the wider critical care population, patients with SARS-CoV-2 infection demonstrate a significantly higher incidence of thrombotic complications, despite standard prophylaxis. Thromboprophylaxis in this patient group is essential but must be titrated according to risk and severity of disease. Monitoring of anti-Xa level may provide a means to do this. Without an individualised approach, suboptimal dosing may impact morbidity and mortality.