4.2 – Clinical data
All available evidence on the use of azithromycin in the treatment of COVID-19 was summarized in Table 2.
In March, Gautret et al. showed that the early treatment either with hydroxychloroquine presented superior virological clearance compared to standard of care[8]. Moreover, the addition of azithromycin further improved the activity of hydroxychloroquine alone. However, only 6 patients were treated with hydroxychloroquine and azithromycin.
Based on this study and in vitro data showing synergic activity, some hospitals started to spread the use of this combination. Nevertheless, the International Society of Antimicrobial Chemotherapy raised concerns as they believed that did not meet the society’s expected standard[71].
These authors subsequently expanded the number of included patients evaluating this combination[72,73]. They included those admitted to the infectious disease ward or treated in day-care hospital, so disease presentation was mild. Overall, clinical and viral outcome was positive. On the contrary, Molina et al. challenged these results in sicker patients as this strategy was not associated with any clinical benefit or antiviral activity[74]. In all these studies, unfortunately, a control group was lacking.
Mahevas et al. assessed the efficacy of hydroxychloroquine in 173 hospitalized patients showing no effect in any outcomes[75]. Patients with organ failure, ARDS or ICU at admission and those treated with other experimental therapies (remdesivir, tocilizumab or lopinavir/ritonavir) were excluded. Given that the objective of the study was the evaluation of the efficacy of hydroxychloroquine, the outcomes of azithromycin alone or in combination were not analyzed. Azithromycin was administered in 15 (18 %) patients in the treatment group and 26 (29 %) in the control group. Among those treated with azithromycin alone, 5 (19.2 %) died and 6 (23.1 %) were transferred to the ICU. These patients, however, were not further analyzed nor included in the propensity-score analysis and no data about their baseline and clinical demographics were detailed.
In patients hospitalized at Veterans Health Administration medical centers, Magagnoli et al. demonstrated a higher risk of mortality in hospitalized patients treated with hydroxychloroquine alone after propensity-score adjustment [76]. However, this finding was not observed with combination therapy. The risk of mechanical ventilation was similar among hydroxychloroquine alone (aHR 1.19 [95% CI 0.78-1.82]) and hydroxychloroquine/azithromycin groups (aHR 1.09 [95% CI 0.72-1.66]) when compared to the no-hydroxychloroquine group. The use of other therapies was not assessed and no information about ICU status at admission was reported.
Geleris et al. included 1,085 hospitalized patients in a propensity-score matched analysis in New York[77]. Patients who died or were intubated within 24 hours after presentation were excluded. Azithromycin was used in both groups (59.9 % in the treatment group and 37.2 % in the control group). Other agents as tocilizumab/sarilumab or remdesivir were allowed (data on corticosteroids was not shown). In the multivariate analysis hydroxychloroquine or azithromycin use was not associated with the composite primary endpoint.
Rosenberg et al. showed a trend towards reduced mortality in the azithromycin alone group, after adjusting for multiple factors[78]. Unlike other studies, patients admitted to the ICU were not excluded. In the estimated direct-adjusted model, 21-day mortality was 22.5 % (95% CI 19.7-25.1) in the combination group, 18.9 % (95% CI 14.3-23.2) in the hydroxychloroquine alone group, 10.9 % (95% CI 5.8-15.6) in the azithromycin group and 17.8 % (95% CI, 11.1-23.9) in the control group. When hydroxychloroquine and azithromycin monotherapy groups were compared, no differences were observed in mortality (aHR 1.92 [95% CI 0.99-3.74]), although it was in the limit of significance.
Another pre-print study showed potential benefits of azithromycin alone, but unfortunately was withdrawn.
Recently, Guérin et al. assessed the time to clinical recovery of azithromycin and its combination with hydroxychloroquine compared to standard of care in outpatients[79]. Both treatments accelerated recovery both in the global cohort and after adjusting in a case-control analysis. No significant differences were found when azithromycin monotherapy and combination therapy were compared (P=0.26).
Barbosa et al . evaluated the combination therapy in the need for hospitalization in outpatients[80]. Patients with flu-like symptoms were referred to telemedicine service, where combination therapy was offered. Those who refused to initiate this treatment were considered the control group. The treatment group was associated with a reduction in the need for hospitalization of 3.5 %. Moreover, among those in the treatment group, patients treated before day 7 of symptoms onset required less hospitalization (1.17 % vs. 3.2 %, P<0.001).
To date, 36 clinical trials are recruiting patients to evaluate azithromycin in a wide variety of scenarios (outpatients, combined with hydroxychloroquine or other drugs, ICU…).