麻豆传媒

Full-dose prophylactic anticoagulation held substantial benefit for moderately ill COVID-19 patients regardless of initial D-dimer level, according to interim pooled results from the , , and trials.

For the primary endpoint of organ support-free days to day 21 among the hospitalized patients not initially in the intensive care unit (ICU), the therapeutic-dosing strategy had a proportional odds ratio that met criteria for superiority over typical prophylactic dosing:

• With low D-dimer, 1.57 (95% credible interval [CrI] 1.14-2.19)

• With high D-dimer, 1.53 (95% CrI 1.09-2.17)

• With missing D-dimer, 1.51 (95% CrI 1.06-2.15)

Overall, the rates were about 16% versus 23% in post hoc analysis, according to Ryan Zarychanski, MD, a leader in both the ATTACC and REMAP-CAP trials, of the University of Manitoba in Winnipeg, and colleagues.

For the critically ill COVID-19 patients for whom the trial drew an earlier stop due to futility, the higher dose strategy was actually significantly more risky (OR 0.76, 95% CrI 0.60-0.97).

Release of the results in a cautioned that these interim results were "pre-publication, not from locked databases and not peer reviewed."

Topline results in a press release the prior week hinted at a mortality benefit, with the slide set now defining this component of the primary endpoint as occurring in 5.7% of therapeutic anticoagulation patients versus 7.7% in the usual care group in the non-ICU cohort. In the ICU, things again went the other direction (35.3% vs 32.6%).

For the moderate group, that mortality benefit is "very significant" and on the order of that seen with dexamethasone, commented Nigel Key, MD, director of the University of North Carolina at Chapel Hill's Hemophilia and Thrombosis Center.

By comparison, dexamethasone in the RECOVERY trial reduced mortality by a relative 18% (23.3% vs 26.2%) among those on supplemental oxygen without mechanical ventilation and didn't hold any benefit in those in the hospital who didn't need respiratory support (17.8% vs 14.4%).

No other pharmaceuticals have yet proven a mortality benefit in this setting.

"It's practice changing for sure," Key argued, pointing to the wide availability and relatively low cost of heparin as extra advantages.

While most medical centers have come up with consensus recommendations on antithrombotic therapy, Key's group found wide variability in practices in a survey of 17 major U.S. medical centers. However, almost no one had guessed empirically that ICU patients might actually need de-escalation of therapy rather than more intense anticoagulation, he told 麻豆传媒.

"These results turn existing expectations 180 degrees," he said. "Each center is now going to need to look at their standards of care internally. We have potential to do harm by getting this wrong."

The fact that D-dimer didn't matter suggests a universal strategy without dose escalation based on that biomarker, as some centers are currently doing, commented Behnood Bikdeli, MD, of Brigham and Women's Hospital and Harvard Medical School in Boston.

Major bleeding rates were somewhat higher but didn't seem to be unreasonable, Key noted. The rate was reported as 1.6% versus 0.9% in the prophylactic dosing arm among non-ICU patients, but it and thrombotic event rates are being looked at again after some .

Across the three platform trial arms coordinated to compare therapeutic versus prophylactic dosing, there were 1,772 moderately ill patients and 1,123 severe patients randomized and 1,398 and 895, respectively, with known outcome at the time of interim analysis. These arms of the platform trials were stopped at the interim analysis, although other treatment randomization is ongoing.

Whether to de-escalate dosing for people who become critically ill after being on therapeutic anticoagulation in the general wards is something centers will have to figure out, noted Bikdeli.

The researchers pointed out that their protocol specified continuing therapeutic dosing into the ICU for patients who worsened after initially being randomized in the general wards, but it wasn't clear that the therapeutic dosing arm wouldn't have looked even better if those patients had dropped to prophylactic doses after becoming critically ill.

The fact that timing does seem to matter may hold some mechanistic clues, he noted.

Further details will be more illuminating, such as comparing time from symptom onset to receiving the study drug in the moderate and severe illness groups, Bikdeli noted. If it's the same onset duration from symptoms in both groups but the advantage was largely in those more rapidly progressing, there would be a different plausible explanation, he added.

If the higher dose primarily prevented thrombotic events like subsegmental pulmonary embolism and distal deep vein thrombosis the mortality benefit might be more likely related to tackling microthrombosis, whereas a reduction driven by stroke, type 1 myocardial infarction, and central pulmonary embolism might be more plausibly due to an effect on macrothrombi, he suggested.

"We have been taught over the past year that caution would never do us a disservice," Bikdeli concluded.

Comment