Results for two primary endpoints
The curves for the primary outcome of prevention had already separated by day 3 and continued to widen over the 30 days in which outcomes were compared. At the end of 30 days, 11.2% of the dapagliflozin group and 13.8% of the placebo group had an event. By hazard ratio, dapagliflozin was linked to 20% nonsignificant relative protection from events (hazard ratio, 0.80; 95% confidence interval, 0.58-1.10).
The trend (P = .168) for the primary endpoint for prevention was reflected in the individual components. For dapagliflozin related to placebo, there were generally similar or greater reductions in new or worsening organ failure (HR, 0.80), cardiac decompensation (HR, 0.81), respiratory decompensation (HR, 0.85), and kidney decompensation (HR, 0.65). None were statistically significant, but the confidence intervals were tight with the upper end never exceeding 1.20.
Moreover, the relative risk reduction for all-cause mortality moved in the same direction (HR, 0.77; 95% CI, 0.52-1.16).
In the hierarchical composite endpoint of recovery, there was no significant difference in the time to discharge, but again many recovery metrics numerically favored dapagliflozin with an overall difference producing a statistical trend (P = .14) similar to organ failure events and death.
In safety analyses, dapagliflozin consistently outperformed placebo across a broad array of safety measure, including any severe adverse event (65% vs. 82%), any adverse event with an outcome of death (32% vs. 48%), discontinuation caused by an adverse event (44% vs. 55%), and acute kidney injury (21% vs. 34%).
Data could fuel related studies
According to Ana Barac, MD, PhD, director of the cardio-oncology program in the Medstar Heart and Vascular Institute, Washington, these data are “thought provoking.” Although this was a negative trial, she said that it generates an “exciting hypothesis” about the potential of SGLT2 inhibitors to provide organ protection. She called for studies to pursue this path of research.
More immediately, Dr. Barac agreed that these data argue against stopping SGLT2 inhibitors in patients admitted to a hospital for COVID-19 infection.
“These data show that these drugs are not going to lead to harm, but they might lead to benefit,” she said.
For James Januzzi, MD, a cardiologist at Massachusetts General Hospital and professor of medicine at Harvard Medical School, both in Boston, DARE-19 was perhaps most impressive because of its rigorous design and execution in the midst of a pandemic.
Over the past year, “the medical literature was flooded with grossly underpowered, poorly designed, single-center studies” yielding results that have been hard to interpret, Dr. Januzzi said. Despite the fact that this study failed to confirm its hypothesis, he said the investigators deserve praise for the quality of the work.
Dr. Januzzi also believes the study is not without clinically relevant findings, particularly the fact that dapagliflozin was associated with a lower rate of adverse events than placebo. This, at least, provides reassurance about the safety of this drug in the setting of COVID-19 infection.
Dr. Kosiborod reported financial relationships with more than 10 pharmaceutical companies, including AstraZeneca, which provided funding for DARE-19. Dr. Barac reported financial relationships with Bristol-Myers Squibb and CTI BioPharma. Dr. Januzzi reported financial relationships with Boehringer Ingelheim, GE Healthcare, Johnson & Johnson, Merck, Novartis, Pfizer, and Roche.