Abstract
We report the clinical outcomes following implementation of initial COVID-19 treatment guidelines in Thailand. A composite poor outcome was defined as death, ICU admission, requiring intubation, or high-flow oxygen. 744 COVID-19 patients (48.8% male) were included, median (IQR) age was 37 (27-48) years [8.4% >60 years] and 21.4% had pneumonia at admission. Admission < 4 days from symptom onset had a reduced risk of poor outcome. In a subgroup analysis, favipiravir use reduced the risk of a poor outcome for patients admitted < 4 days from symptom onset (OR 0.320 (0.152-0.662), P=0.003). Thai guidelines now include favipiravir to treat all symptomatic COVID-19 patients.
Keywords : COVID-19, Favipiravir, Thailand
Main text: Thailand is one of the few countries that has been able to successfully control the COVID-19 outbreak. As of January 10, 2021, there were 10,294 confirmed cases of COVID-19 in Thailand, resulting in 69 deaths. Thailand encountered its first outbreak in early March 2020 originating from attendees of a famous Thai Boxing competition in Bangkok which rapidly spread to individuals in surrounding bars and night clubs resulting in more than 2,500 cases. Due to this rapid surge in cases, the Ministry of Public Health held meetings with multidisciplinary experts, including infectious diseases specialists, pulmonologists, intensivists, and policymakers, to develop clinical practice guidelines to guide clinicians using the scientific evidence available at that time. The first version of the COVID-19 guidelines was posted on March 21, 2020 and recommended (i) No antiviral treatment for asymptomatic patients and (ii) Combination antiviral treatment directed against different viral targets for symptomatic cases, with the regimen dependent on the presence or absence of pneumonia COVID-19 illness. For patients with non-pneumonia COVID-19 illness but with specific risk factors, e.g. age >60 years or <5 years, BMI>35 kg/m2, immunocompromised, diabetes, chronic conditions of lungs, kidney, liver, cardiovascular or cerebrovascular diseases, or lymphocytes <1,000 cells/mm3, the recommended treatment was chloroquine (CQ) (500 mg, twice daily) or hydroxychloroquine (HCQ) (600 mg, twice daily on the first day, then 400 mg twice daily) plus a boosted protease inhibitor (bPI) (lopinavir/ritonavir 400/100 mg, twice daily or darunavir/ritonavir 600/100 mg, twice daily). A triple drug combination of CQ or HCQ, a bPI, and favipiravir (FPV), an oral RNA dependent RNA polymerase inhibitor (60 mg/kg/day twice daily on the first day, then 20 mg/kg/day twice daily afterwards) was recommend for patients with pneumonia COVID-19 illness. The use of FPV for only the most severe cases was due to the limited availability at that time. The duration of treatment was 5 days but could be extend to 10 days in cases of severe pneumonia. Azithromycin was recommended as adjunctive treatment for all symptomatic cases, particularly for severely ill patients. All COVID-19 patients were hospitalized regardless of symptoms. Our aim was to describe the clinical outcomes following antiviral treatment according to these initial guidelines and identify factors associated with a poor clinical outcome.
Nine large hospitals in Bangkok were invited to submit data on their COVID-19 patients using standardized case record forms. Data collected from March to June, 2020, included demographics, initial diagnosis, antiviral and antibiotic treatment, clinical outcome, including the need for oxygen therapy (intubated, high-flow, or lower oxygen support), intensive care unit (ICU) admission, and death. A composite poor outcome was defined as death, ICU admission, requiring intubation, or requiring high-flow oxygen. If treatment was indicated, antiviral treatment was initiated on the day of admission; however, favipiravir initiation may be delayed for a day in some hospitals due to approval and drug delivery process. Factors associated with the composite poor outcome were assessed using multivariate analysis. The study was approved by the local Ethics Committee.
A total of 744 COVID-19 patients (48.8% male) were included in this analysis with a median (interquartile range) age of 37 (27-48) years [8.4% >60 years], 5.6% were obese and 16.5% had underlying conditions. At hospital admission, 28.4% were asymptomatic, 50.2% had non-pneumonia illnesses, of which 22.7% had risk factors, and 21.4% had pneumonia (Figure 1). 637 (85.6%) cases received antiviral treatment in compliance with the guidelines. Overall, there were 8 (1.1%) deaths, 66 (8.9%) required ICU admission, 17 (2.3%) required intubation, and 19 (2.6%) required high-flow oxygen. Seventy (9.4%) of patients had at least one of these outcomes (composite poor outcomes).
In univariate analysis among symptomatic patients, factors associated with a higher risk of a composite poor outcome were: being male (odd ration [OR] 1.89 (95% confidential interval [CI] 1.12-3.20), P=0.018), >60 years-old (OR 4.42 (2.36-8.27), P<0.001), obesity (OR 2.79 (1.32-5.88), P=0.007) , underlying risk condition (OR 4.14 (2.45-6.99), P<0.001), pneumonia at admission (OR 11.51 (6.32-20.94), P<0.001), receiving an initial regimen containing FPV (OR 10.65 (5.62-20.17), P<0.001) and azithromycin use (OR 4.52 (2.68-7.62), P<0.001). Admission to the hospital within 4 days from symptom onset significantly reduced the risk of a composite poor outcome (OR 0.30 (0.17-0.52), P<0.001), and this remained significant among patients with pneumonia at presentation (OR 0.34 (0.16-0.72), P=0.005), but not among patients with non-pneumonia illness (OR 0.65 (0.24-1.77), P=0.397). The longer time to hospital admission significantly increased the risk of poor outcome (OR 1.10 (1.04-1.16), P=0.001), i.e. the risk of a composite poor outcome increased 10% each day from symptom onset until hospital admission.
In a multivariate analysis of symptomatic patients, factors associated with a composite poor outcome were an underlying risk conditions, receiving an initial regimen containing FPV, and azithromycin use; while admission within 4 days of onset of symptoms was associated with a reduced risk of a poor composite outcome (Table 1).
A triple drug antiviral regimen of CQ/HCQ + bPI + FPV was associated with a composite poor outcome but this was likely due to its use only in the sickest patients. However, in a subgroup analysis, regimens with FPV reduced the risk of a poor outcome for patients admitted< 4 days from symptom onset, compare to after 4 days, even after adjusting for confounding factors (OR 0.320 (0.152-0.662), P=0.003) (Table 2). This benefit was not observed for patients receiving regimens using CQ/HCQ plus bPI without FPV.
Large studies have reported no benefit of lopinavir1,2and hydroxychloroquine3-5and we also observed no clear benefit of this dual treatment combination. Recent data from the open-label randomized SOLIDARITY study,6 led by WHO, found that neither remdesivir, hydroxychloroquine, lopinavir, nor interferon beta-1a treatment reduce overall mortality. No benefit of remdesivir in terms of all-cause mortality was also reported in the ACTT-1 trial,7 but remdesivir did significantly decrease the time to recovery (10 vs 15 days, RR 1·29, 95% CI 1·12–1·49), with a clear benefit for patients starting treatment within 10 days of symptom onset. The approval of remdesivir for COVID-19 treatment is encouraging but widespread access is expected to be limited in most resource-limited settings, including Thailand. Favipiravir has a similar antiviral mechanism to remdesivir and can easily be scaled up in Asia due to its lower cost and simpler oral administration. Studies have shown favipiravir to be effective in viral clearance and improve clinical outcomes when compared to lopinavir/ritonavir8 or with arbidol9, or without antiviral treatment.10 Initiation of FPV ‘early’ compared to ‘late’ (6 days later) was found to reduce the time to defervescence in Japanese adults admitted with COVID-19 who were asymptomatic or mildly ill. Data from our study, while observational, indicates that early treatment (<4 days from symptom onset) with a combination of antivirals containing favipiravir reduces the risk of a poor outcome. These data build on the evidence that early antiviral treatment could be beneficial while later antiviral treatment after the process of hyperinflammatory has kicked in, and viral load11waning, is less effective, and instead a glucocorticoid would help reduced mortality.12
The initial Thai guidelines were based on available data at that time with many of the evidence based on in-vitro data. Considering the recent literature, along with these local data, Thai guidelines were recently revised to stop using CQ/HCQ plus bPI, and recommend using favipiravir early in the course of illness for symptomatic patients independent of risk factors. It is conceivable that early treatment with FPV per initial guidelines could have contributed towards the relatively low mortality observed in our setting. In the second wave in Thailand, favipiravir plays an important role for the treatment of COVID-19.