Therapeutic anticoagulation appears to be beneficial in patients with moderate COVID-19, but not in those with severe disease where it may actually do harm, interim results from three large international trials suggest.

Initial data from ATTACC, REMAP-CAP and ACTIV-4a suggests that in patients who are hospitalised and require oxygen but not ICU organ support, therapeutic anticoagulation has positive effects on morbidity and mortality.

The benefits were seen across D-dimer sub-groups, the interim data report said.

Yet in severe COVID-19 where patients require ICU organ support there is a high probability that therapeutic heparin may be harmful compared with thromboprophylaxis, the early results suggest.

“Even though these are interim results, they are difficult to ignore and are prompting some groups to change practice,” said Beverley Hunt, professor of thrombosis and haemostasis at King’s College London, in an interview with the limbic. “Others are being more cautious and awaiting the full published papers, as will all the major guideline groups.”

An online presentation of the data shows that 2,895 patients had been enrolled in the trials by the 21st January, when the interim analysis was done.

All three trials have halted recruitment for patients with both moderate and severe disease, in the former case due to superiority and the latter due to inferiority. The independent boards overseeing the studies said the data showed therapeutic anticoagulation drugs did not improve outcomes, and said “a potential for harm in this sub-group could not be excluded and increased bleeding was noted with full-dose anticoagulation.”

The primary outcome measure was organ support-free days. Anticoagulation showed superiority across D-dimer groups in moderate disease, but inferiority in severe disease.

While the major bleeding rate in moderately ill patients treated with therapeutic anticoagulation was less than 2%, in severe patients there was a “numeric increase in major bleeding events and mortality.”

Work is ongoing to answer the question of how to manage moderate patients who subsequently become severely ill. The reported interim results have not yet been peer reviewed.

Professor Hunt said there were now numerous papers showing that rates of thromboembolism are high in hospitalised patients with COVID-19 pneumonia, and it was exciting to have some interim results from RCTs on the utility of using therapeutic anticoagulation versus standard of care.

“This combined data from REMAP-CAP, ATTACC and ACTIV-4a trials is an extraordinary international effort combining data from centres across the world,” she told the limbic, calling the divergent results in moderate and severe disease surprising.

Dr Keith Gomez, consultant and associate professor in haematology at the Royal Free London NHS Foundation Trust, said he would not change clinical practice until the full analysis has been published.

“The only thing I would take from this interim data is that D-dimer is not useful for determining anticoagulation dosage in moderate patients,” he said. “What to do with patients whose condition deteriorates is a key question.”

Dr Gomez, who is also President of the British Society of Haemostasis and Thrombosis, added: “The other thing that this demonstrates is the importance of only using modified doses of anticoagulation within a clinical trial, rather than changing treatment protocols without solid evidence.”