Abstract

Coronavirus disease 2019 (COVID-19) is still an ongoing pandemic and has already resulted in millions of deaths worldwide. Since it is frequently associated with thrombotic complications, most hospitalised patients receive some form of anticoagulant therapy. Lately, bleeding events in patients with COVID-19 have also been increasingly reported. It is not yet clear whether these are also part of the spectrum of coagulation dysfunction in COVID-19. We report two cases of patients with COVID-19 with a large spontaneous haemorrhage during therapeutic anticoagulant therapy, after a long-term stay in the ICU. We propose several hypotheses for the occurrence of bleeding in COVID-19. Physicians should be aware of this risk, especially when patients are being treated with anticoagulants, because changes in the patient’s condition during the course of the disease could lead to a relative overdose. We advise regular reassessment of anticoagulation dosage in patients with COVID-19.

Introduction

In December 2019, a new coronavirus called Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) emerged in the region of Wuhan, in China. Since then, the virus has been spreading across the whole world, resulting in an ongoing pandemic. The disease caused by this novel virus is known as Coronavirus Disease 2019 (COVID-19), which mostly manifests with symptoms of fever, cough and/or dyspnoea.[1] Studies have described coagulation abnormalities in patients with COVID-19, which seem to be related to the severity of the disease.[2,3] The occurrence of thromboembolism is frequently reported in hospitalised COVID-19 patients. The incidence appears to be higher in severely ill patients, in particular those who are admitted to the ICU. Cumulative incidence rates of up to 31% are documented in ICU patients.[2-4] Therefore, both prophylactic and therapeutic anticoagulants are often prescribed in COVID-19 patients.[5] Much less has been published about the occurrence of bleeding in COVID-19 patients, although studies on this subject have increased lately.[2,6-12] It is not yet clear whether this complication is also part of the spectrum of coagulation dysfunction in COVID-19. We report two cases of patients with COVID-19 with a large spontaneous haemorrhage during therapeutic anticoagulant therapy, after a long-term stay in the ICU. We propose several hypotheses for the occurrence of bleeding events in COVID-19 patients.

Cases

Patient A, a 73-year-old man with a history of atrial fibrillation for which he was taking rivaroxaban, was admitted to the hospital with symptoms of general malaise for more than a week and dyspnoea. On admission, the patient was moderately ill. Apart from mild hypoxaemia, for which he required oxygen therapy, no other significant abnormalities were reported on physical examination. Computed tomography angiography (CTA) of the chest showed widespread bilateral ground glass opacities and dorsobasal consolidations, suspect for COVID-19 (CO-RADS 5, classified as ‘very high suspicion’ and a severity score of 15 out of 25).[13,14] Shortly after, this diagnosis was confirmed by reverse transcription polymerase chain reaction (RT-PCR). Within one day after admission, the patient was transferred to the ICU for intubation because of respiratory insufficiency. Rivaroxaban was then replaced by low-molecular-weight heparin (LMWH) in a therapeutic dose of 10,000 international units (IU) twice daily subcutaneously.
Several CTAs were performed during his stay in the ICU, which showed signs of extensive consolidations and fibrosis, but no pulmonary emboli. Finally, the patient’s condition slowly improved and after 38 days in the ICU he was discharged to the COVID-19 step down ward. Two days later, the patient complained of right-sided chest pain. On physical examination, a painful swelling with a diameter of about 15 cm on the right lateral side of the chest just below the axilla was found. There were no visible signs of inflammation or bleeding. Laboratory tests showed a decrease in haemoglobin, as compared with two days before (table 1). Suddenly, the patient’s condition deteriorated, showing the clinical presentation of hypovolaemic shock. Fluid resuscitation was immediately given, followed by blood products as severe haemorrhage was suspected. Emergency CTA of the chest showed a large haematoma in the right latissimus dorsi muscle with signs of active bleeding originating from the thoracodorsal artery (figure 1). Coil embolisation was successfully performed and the patient was then transferred to the ICU for further stabilisation. In the following days, the patient needed several blood transfusions because of declining haemoglobin levels. Ongoing or new blood loss was excluded by repeated CTAs. Subsequently, he developed severe pneumonia and delirium. Despite administration of antibiotics and oxygen therapy, the patient’s condition deteriorated further and eventually he died one and a half months after initial admission to the hospital.

Patient B, a 63-year-old man with a history of diabetes mellitus type 2, hypertension and chronic mild kidney dysfunction, presented to the emergency department with a ten-day history of general malaise. On physical examination, the patient was found to have a sinus tachycardia, hypoxaemia and severe tachypnoea. The chest radiograph showed bilateral consolidations, suspect for COVID-19. Ceftriaxone and oxygen were administered. Within 24 hours after admission, the patient had to be transferred to the ICU for intubation due to respiratory insufficiency. CTA of the chest demonstrated widespread abnormalities, compatible with COVID-19 (CO-RADS 5 and a severity score of 19 out of 25) and no signs of pulmonary embolism.[13,14] SARS-CoV-2 RNA was detected in a nasopharyngeal swab specimen by RT-PCR. Because of hypercoagulability in COVID-19, prophylactic LMWH 5,000 IU twice daily was started according to the local protocol at that time. The patient was discharged to the COVID-19 step down ward after two weeks, but had to be readmitted to the ICU two days later because of respiratory insufficiency. CTA was suspicious for segmental and subsegmental pulmonary emboli, for which the prophylactic LMWH dosage was increased to a therapeutic dose of 10,000 IU twice daily subcutaneously. After a four-day stay in the ICU, the patient returned to the step down ward again. Unfortunately, repeated respiratory insufficiency led to readmission to the ICU on the same day. CTA demonstrated extensive consolidations and fibrosis. Bronchoalveolar lavage showed a herpes simplex virus type 1. Treatment with acyclovir intravenously for 14 days was then started. In total, the patient stayed in the ICU for approximately six weeks; for about five of these six weeks he was mechanically ventilated. Two days after his third return from the ICU, the patient complained of left-sided chest pain. Physical examination revealed a local swelling, extending to the back. Laboratory studies showed a decrease in haemoglobin (table 1). Apart from a sinus tachycardia, the patient was otherwise haemodynamically stable. The LMWH was discontinued and the haemoglobin level was monitored and remained stable during the following hours. The next day, a CTA was conducted that demonstrated a large haematoma along the dorsolateral thoracic wall with no evidence of active bleeding (figure 1).