Recognition and management of respiratory coinfection and secondary bacterial pneumonia in patients with COVID-19

RESPIRATORY INFECTION AND COVID-19

Even as severe acute respiratory syndrome coronavirus type-2 (SARS-CoV-2), the etiological agent of coronavirus disease 2019 (COVID-19), spreads across the globe, the pathophysiology of the disease remains incompletely understood. Respiratory infection caused by more than one viral pathogen (viral coinfection) or both viral and bacterial pathogens (combined viral and bacterial pneumonia) have been well-described. Secondary bacterial pneumonia can follow the initial phase of viral respiratory infection or occur during the recovery phase.¹ Data on SARS-CoV-2 is limited, but thus far, the overall incidence of viral co-infection has varied widely from 0% to 15% in different case series,^2–7 and combined viral and bacterial pneumonia rates appear to be low.^3,8–10 There is also a dearth of data on the predisposing factors and causative organisms.

Combined viral and bacterial pneumonia and secondary bacterial pneumonia by Staphylococcus aureus and other common community-acquired pneumonia pathogens have been best studied in seasonal² and pandemic^2,3 influenza and contribute significantly to morbidity and mortality. Secondary bacterial pneumonia with severe acute respiratory syndrome virus (SARS-CoV) infection occurred as ventilator-associated pneumonia in 25% of patients at a single center of which methicillin-resistant Staphylococcus aureus was the causative organism in 47% of cases, although there was significant concern for cross-transmission.⁷

Because no obvious pattern or guidelines exist for viral coinfection, combined viral and bacterial pneumonia, or secondary bacterial pneumonia in the context of SARS-CoV-2, the following commentary is based on existing clinical data and experience with similar viruses such as influenza and SARS-CoV. With what we know so far, the approach in the context of COVID-19 would, ideally, take into consideration the overall presentation as well as the trajectory of illness.

All patients presenting with symptoms of respiratory infection should be tested for influenza with polymerase chain reaction (PCR)-based assay in addition to SARS-CoV-2. Polymerase chain reaction assays can be performed for other respiratory viruses if available.

Regardless of disease severity, all patients with influenza A/B viral coinfection should be treated with oseltamivir or an alternative agent.¹¹ Empiric treatment for influenza viral coinfection can be considered while waiting for test results if an obvious exposure or risk factor is present. If viral coinfection with another respiratory virus such as respiratory syncytial virus is identified, treatment options are limited and effective only in specific scenarios such as immunosuppression and/or hypogammaglobulinemia.^12,13 Infectious disease consultation is strongly recommended to determine the benefits of such treatment in light of the potential risk for exacerbating COVID-19–related organ failure and the potential adverse effects of the medication(s).

Recognizing combined viral and bacterial pneumonia or secondary bacterial pneumonia with COVID-19 requires a high index of suspicion. Some characteristics of bacterial infection may still be identifiable despite a significant overlap of viral and bacterial symptomatology (Table 1). Neutrophilic leukocytosis is the hallmark of bacterial pneumonia, whereas COVID-19 patients typically present with a normal white blood cell count with lymphopenia.^5,8,14,15

Procalcitonin is neither sensitive nor specific in differentiating the etiology of community-acquired pneumonia.¹¹ However, several reported series of COVID-19 have consistently reported normal (low) procalcitonin levels in isolated SARS-CoV-2 infection leading to its widespread, albeit unvalidated, use to “rule out” combined viral and bacterial pneumonia, although the exact cut-off remains to be determined. This observation highlights the need to consider all variables in the context of the clinical scenario.

In patients with mild- to-moderate respiratory failure consistent with the presentation of COVID-19 and without obvious signs of bacterial infection, the likelihood of combined viral and bacterial pneumonia is low, and antibiotics can be safely held off. In this case, gradually worsening respiratory failure within the first week of presentation is more likely to be from progression of COVID-19 than from a new superimposed secondary bacterial pneumonia. This includes patients who are started on non-invasive forms of supplemental oxygen support and then ultimately require invasive mechanical ventilation.

In the absence of supporting evidence of bacterial pneumonia, antibiotics should not be initiated despite progression of respiratory distress. If, however, a patient was to develop new or acutely worsening respiratory failure, sepsis, or both after an initial phase of consistent improvement (considered to be days), then nosocomial acquisition of secondary bacterial infection is likely unless proven otherwise—either secondary bacterial pneumonia in the form of hospital-acquired pneumonia and/or at another extrapulmonary site.

While COVID-19 by itself can cause acute respiratory decompensation, data regarding secondary bacterial pneumonia playing a role in such decompensation is limited; therefore, guideline-driven empiric antibiotic use may be reasonable until this secondary infection is ruled out. Supportive evidence for secondary bacterial pneumonia include one or more of the following: new or recrudescent fever, new onset or change in the character of sputum, new leukocytosis and/or new neutrophilia, new relevant imaging findings, and new or increasing oxygen requirements. It is also important to consider all other sources of hospital-acquired infections in these patients such as indwelling central venous catheters or urinary tract catheters and treat accordingly.

For a critically ill patient admitted with severe respiratory failure, empiric treatment for all possible etiologies upfront is essential. This is especially important because procalcitonin levels can be falsely elevated in patients with multi-organ failure,^29,30 and imaging studies may be limited in differentiating bilateral infiltrates of acute respiratory distress syndrome from obscured consolidation of bacterial infection.

Empiric therapy for community-acquired pneumonia should be based on the Infectious Diseases Society of America/American Thoracic Society guidelines, host risk factors, and prior microbiological data.¹⁹ Respiratory samples (tracheal aspirate in mechanically ventilated patients is preferable to sputum) and blood cultures should be sent for all patients, ideally prior to starting antibiotics. Streptococcus pneumoniae urine antigen should be tested in all patients presenting with severe community-acquired pneumonia. Legionella pneumophila urine antigen and Mycoplasma pneumoniae IgM and IgG antibodies can be sent based on clinical context and epidemiology.

In the absence of signs of bacterial pneumonia, a positive respiratory culture can represent colonization, especially in those with prior pneumonia with the same organism or altered airway anatomy. Laboratory markers, radiologic features (see Table 1 and above) as well as quantitative and/or semi-quantitative culture methods can help in making this distinction.¹⁹

Secondary bacterial pneumonia in a patient on invasive mechanical ventilation has a similar presentation to that of hospital-acquired pneumonia but warrants the aggressive use of empiric broad-spectrum antibiotics with coverage for methicillin-resistant Staphylococcus aureus, Pseudomonas aeruginosa, and possibly other multi-drug resistant organisms in accordance with the Infectious Diseases Society of America/American Thoracic Society guidelines.¹⁹ It is also important to consider the side effects of antibiotics and institutional antibiograms.

Patients with ventilator-associated tracheobronchitis often lack the classical signs of secondary bacterial pneumonia, may have increased secretions and low-grade fevers, and can be difficult to wean from ventilatory support. The evidence to support antibacterial therapy for this clinical entity is limited and warrants a judicious case-based analysis.

The duration of antibacterial therapy is generally 5 to 7 days for community-acquired pneumonia³¹ and 7 days for hospital-acquired pneumonia and ventilator-associated pneumonia¹⁹ in the absence of complications. Consider shortening the duration if patients demonstrate signs of clinical stabilization, especially if adverse effects are seen. Checking the procalcitonin level at presentation will help in the de-escalation of antibiotics based on the trend of procalcitonin levels in 24 to 48 hours.³² If a microbiological source is not identified within 48 hours of testing and the procalcitonin level is < 0.5 μg/L and/ or decreases by ≥ 80% from peak concentration, it is reasonable to discontinue all antibiotics.¹⁹

The use of interleukin 6 (IL-6) inhibitors such as tocilizumab for COVID-19–related cytokine activation syndrome presents a unique challenge because it suppresses common signs of sepsis. The risk of serious bacterial infections has been consistently reported to be higher with tocilizumab use for rheumatological diseases.^33–36 C-reactive protein (CRP) and other acute phase reactants including white blood cell count may be unreliable acute phase reactants and may not rise in response to a secondary bacterial infection after tocilizumab use.^34,37,38 Exactly how long this effect lasts with 1 or 2 doses is unclear. Procalcitonin may be less affected by IL-6 inhibitors,^21–23 but the data to differentiate bacteria from viral pneumonia in this context is limited and should be further evaluated.

Lastly, invasive pulmonary aspergillosis has been described in critically ill patients with seasonal^39,40 as well as pandemic⁴¹ influenza and is associated with high morbidity and mortality rates. Invasive pulmonary aspergillosis was also reported on post-mortem lung pathology in 2 of 20 patients with SARS-CoV-1 infection.⁴² This complication should be considered in high-risk individuals such as those with immune compromising conditions, precedent or concomitant influenza viral co-infection, clinical deterioration despite appropriate antibiotics, and those with positive fungal markers such as galactomannan on cul ture. If invasive pulmonary aspergillosis is suspected, treatment with broad antifungal agents such as voriconazole should be initiated promptly in consultation with infectious disease colleagues.