Background and aims: Patients with severe Coronavirus Disease 2019 (COVID-19) are prone to secondary bacterial pneumonia. The use of probiotics against oral pathogens might prevent lung colonization and progression to bacterial pneumonia. This study aimed to assess the effect of Streptococcus salivarius K12 combined with Lactobacillus brevis CD2 in preventing secondary bacterial pneumonia in patients with severe COVID-19. Methods: This randomized placebo-controlled phase 2 trial involved 70 patients with severe COVID-19 admitted to the intensive care unit (ICU). Patients were randomly assigned to a 7-day course of oral gel containing Streptococcus salivarius K12 2 billion colony-forming units (CFU) and Lactobacillus brevis CD2 4 billion CFU every 8 hours or placebo, starting in the first ICU day. The primary outcome was bacterial pneumonia, established according to clinical, laboratory, radiological, and microbiological findings, whereas secondary outcomes were ICU stay in days and hospital mortality.

Design This was a randomized, double-blind, placebo-controlled, parallel-group, phase 2 clinical trial.

Setting All participants were studied after admission on ICU at the Clinics Hospital of Passo Fundo (HCPF). This institution is a regional reference for the treatment of patients with COVID-19, with ICU specialized for this condition, serving an estimated population over 500,000 inhabitants.

Participants Adult patients with suspected or confirmed diagnosis of COVID-19 who presented to the HCPF with severe acute respiratory syndrome, requiring ICU admission, were invited to participate. This project followed the rules of the Helsinki Declaration and was carried out after approval by the Ethics Committee of the HCPF and the Research Ethics Committee of the UPF/CONEP. Patients were included after signing the consent form, by the patient or his/her guardian.

ICU management The patients followed the clinical routine practiced in the HCPF ICU. They received measures of non-invasive and invasive ventilatory support, hemodynamic, nutritional, among others, in addition to specific treatment for COVID-19 in severe cases, using drugs at the discretion of the hospital protocol, in addition to broad-spectrum antibiotic therapy in suspected superimposed bacterial infection. Monitoring of the pulmonary condition was performed with routine tests including laboratory, imaging, microbiology.

Probiotic and placebo Patients were treated with a gel containing Streptococcus salivarius K12 (2 billion live bacilli) and Lactobacillus brevis CD2 (4 billion live bacilli) or a placebo. This amount of probiotics (and placebo) was prepared in a sachet containing one gram of gel, by a local compounding pharmacy (NATUPHARMA, probiotics marketed by FRAGON, Netherlands). These probiotics are exempt from registration with The Brazilian Health Regulatory Agency (ANVISA). Each patient received 1 gram of oral gel every 8 hours for 7 days, applied by the researcher. It should be noted that all researchers followed the current care protocols for the prevention of contamination by the new coronavirus.

Sample size The number of patients was estimated according to the outcome "incidence of secondary bacterial pneumonia". Considering a 65% probability of occurrence of this outcome without treatment (placebo), with a reduction to 30% after the use of experimental treatment (probiotics), with a bilateral test for an alpha of 5% and power of 80%, the number number of participants in each group was 30. Thus, the estimated sample size was 60 patients.

Study protocol All research activities were conducted in the ICU bed. After fulfilling the eligibility criteria and signing the consent form, each patient was drawn to receive a probiotic or placebo. The treatments were coded with computer-generated random numbers (groups with two treatments 1:1 = probiotic:placebo) and the list of codes was under the responsibility of the research pharmacist, for disclosure only after data analysis. The codes were packaged in a manila envelope to be removed by a family member of the patient involved in the study, or by the patient himself, or through a drawing site. The gels were identical, such that neither the researchers nor the patients/relatives know what treatment is being offered to a particular patient. At each visit by the researcher to the ICU, the patient was examined for verification of oral and dental status. Afterward, oral hygiene and application of the gel with probiotics or placebo were performed. This procedure is repeated 3 times a day (every 8 hours) for 7 days. Any reactions at the application site (mouth) were monitored and recorded daily by the researcher who applied the treatments. The occurrence of unusual pulmonary events, in quality (symptoms and laboratory/imaging findings) and/or quantity (intensity of manifestations), potentially related to the use of probiotics, was monitored by the responsible pulmonologist. The responsible researchers undertook to notify CONEP within 24 hours if they found serious adverse events. Even if the clinical trial is a phase 2 trial with a small sample, the finding of a serious adverse reaction to the use of oral treatments would result in an immediate interruption of the study. The occurrence of serious adverse events was also observed by an independent data monitoring and safety committee, composed of two physicians with experience in the field (a pulmonologist and an intensivist). The committee performed a daily assessment of patients throughout the study and had autonomy, with the researchers' agreement, to notify CONEP/ANVISA in the event of serious adverse events.

Statistical analysis Descriptive statistics were used with average (or median) and standard deviation (or interquartile range) for quantitative data, and absolute and percentage frequencies for qualitative data. Student's t-test and Fisher's exact test were used for the respective data. Survival analysis was performed. Treatments were maintained blindly between researchers and patients. Data were analyzed with the Graph Pad Prism software. A value of P < 0.05 will be indicative of significance.

Inclusion Criteria:

• patients hospitalized in the ICU with severe acute respiratory syndrome
• progressive deterioration of the ventilatory status in the first 48 hours in the ICU, which may require mechanical ventilation within this period;
• confirmed diagnosis of COVID-19 within 3 days of ICU admission;
• acceptance to participate.

Exclusion Criteria:

• relevant immunodeficiency (decompensated HIV, aggressive immunosuppression, terminal cancer)
• decompensated systemic disease before the onset of COVID-19.

Oral probiotics and placebo were physically identical and were revealed after statistical analysis