The outbreak of Coronavirus disease 2019 (COVID-19) has caused a high morbidity and mortality and an unprecedented global mobilization. Most cases present with none or only mild symptoms, but severe presentations requiring intensive care and mechanical ventilation are observed. As the pandemic progresses, COVID-19 appears to be a lethal complex systemic disease with pneumonia and coagulation disorders as a primum movens leading to a multiple organ dysfunction syndrome [1]. Moreover, awareness of uncommon presentations and complications increases. Review of the current literature reveals spontaneous pneumothorax as a known complication of COVID-19 infection [2],[3], although with an undetermined prevalence and long-term risk. In this report, we discuss a case of pneumothorax complicating COVID-19 pneumonia to raise awareness toward this complication as a potential cause of acute decompensation and poor outcome.

A 68-year-old patient was admitted to our unit for the management of a COVID-19 viral pneumonia with negative nasopharyngeal RT-PCR (real-time reverse transcriptase polymerase chain reaction) testing, positive IgG serology, and typical extensive lesions on chest computed tomography (CT) scan (Figure 1). His past medical history was significant for a 20-year history of chronic smoking, weaned 10 years ago. He presented to the emergency room with impaired general condition and worsening shortness of breath for 7 days. His vitals upon admission were: heart rate 124 beats per minute, respiratory rate 28 cycles per minute, oxygen saturation of 70% on room air, blood pressure 120/80 mmHg, Glasgow score of 15, temperature 36 °C, and blood glucose of 1.4 g/dL. Laboratory assessment showed white blood cell count at 18.44 × 10³/μL (normal range 4–10 × 10³/μL) with lymphocyte count at 0.81 × 10³/μL (normal range 0.9–5.2 × 10³/μL), platelets at 684 × 10³/μL (normal range 150–400 × 10³/μL), C-reactive protein (CRP) at 180 mg/L (normal range 0–5 mg/L), lactate dehydrogenase (LDH) of 578 U/L (normal range 0–248 U/L), and ferritin of 967 μg/L (normal range 20–300 μg/L). The patient was placed on non-invasive positive pressure respiratory support based on the alternance of a bi-level positive airway pressure (BiPAP) mode through a Helmet and high-flow nasal cannula (HFNC) combined with prone positioning. The pressure support was adjusted to maintain a tidal volume of 6–8 mL/kg. The positive end-expiratory pressure (PEEP), fraction of inspired oxygen (FiO₂), and oxygen flow rate were adjusted (gradually increased or decreased in increments of respectively 2 cmH₂O, 5% and 5 L/min) to maintain blood oxygen saturation (SpO₂) ≥90% or PaO₂≥60 mmHg and a respiratory rate at around 18 cycles/min. Medical treatment following local guidelines included hydroxychloroquine, azithromycin, corticotherapy, heparin, aspirin, glycemic control, vitamin therapy, and nutritional support. He received ceftazidime and amikacine to treat a secondarily superimposed bacterial pneumonia. He did not require any other invasive procedures or central lines. The patient’s clinical and biological condition progressively improved and he was weaned from non-invasive mechanical ventilation on day 12. He was placed on nasal cannula 5 L/min, which improved his oxygen saturation from 90% to 96%. His respiratory rate was around 16 cycles per minute. On day 18, he developed a sudden onset of dyspnea and oxygen saturation drop to 68%. He was conscious; his heart rate was of 140 beats per minute and blood pressure of 90/60 mmHg. He was placed on high concentration mask 15 L/min. Rapid clinical assessment noted an asymmetrical thoracic amplification. A point of care ultrasound was performed and revealed no pleural sliding with a “barcode sign” at the M-mode evaluation on the right lung while partially dismissing concerns for severe pulmonary embolism. Bedsides chest X-ray confirmed a massive right pneumothorax (Figure 2A). A chest tube was placed under close monitoring and vascular filling. This resulted in immediate respiratory and hemodynamic improvement with clinical and radiological evidence of lung re-expansion (Figure 2B). Despite lung expansion after drainage, patient’s oxygen requirements increased and high flow oxygen therapy (HFNC) was administrated. A chest angioscan was performed and ruled out a proximal pulmonary embolism. Complete regression of pneumothorax was observed in the following radiographic and ultrasound controls. However, the patient presented progressive respiratory deterioration requiring intubation, sedation, and protective mechanical ventilation (low volume and limited pressure). Despite support measures and intra-hospital management, multiple organ failure (respiratory then renal then hemodynamic) occurred and he died on day 33 after admission.

Approximately 1% of patients admitted with COVID-19 develop pneumothorax [4],[5],[6]. This incidence may be higher (roughly 2%) in COVID-19 patients requiring Intensive Care Unit (ICU) admission [7]. The significance and frequency of COVID-19 and pneumothorax association remains unclear and challenging. The known imaging features of initial CT in COVID-19 cases include bilateral, multilobar ground glass opacities with a peripheral or posterior distribution (or both), mainly in the lower lobes. Septal thickening, bronchiectasis, pleural thickening, subpleural involvement consolidation, and crazy paving patterns are mainly seen in the later stages of the disease. Pneumothorax is one of the uncommon but possible findings seen with disease progression [8]. Spontaneous pneumothorax refers to the presence of air in the pleural space with no evidence of precipitating factor (trauma or invasive procedure). An underlying pulmonary disease is the primary risk factor for the development of secondary spontaneous pneumothorax [9]. Current understanding of histopathology in COVID-19 is restricted to case reports and small series. The physiopathology basis of air-leak disease in COVID-19 patients and its clinical importance have yet to be elucidated. Meanwhile, underlying mechanisms of spontaneous pneumothorax in COVID-19 patients are thought to be multiple: lung parenchyma structural changes leading to alveolar tears (mainly cystic and fibrotic changes) and the increase in intrathoracic pressure resulting from prolonged coughing and/or mechanical ventilation [4],[10]. COVID-19 pneumonia, through a cytokine storm and a dysregulated immune response, can cause severe lung injury with a diffuse alveolar damage leading to alveoli rupturing [11]. In this case, air leak occurred on the side of maximum lung damage. COVID-19 induced thrombosis and microangiopathy can also lead to an ischemic breakdown of the alveolar wall secondary to intravascular fibrin micro-thrombi [12],[13]. Moreover, the COVID-19 viral entry into target cells through angiotensin converting enzyme-2 (ACE-2) receptors may cause surfactant-producing type II pneumocytes injury, leading to a surfactant production dysregulation and an impaired lung compliance [14]. Our patient received non-invasive ventilation (NIV) (BiPAP) and HFNC, an intermediate level of support between conventional oxygen delivery and NIV. Though the risk of barotrauma and subsequent complications are low with NIV and practically uncommon with HFNC, such cases have been reported [15],[16]. End-expiratory pressure provided by positive pressure ventilation increases the pressure gradient between alveoli and the interstitial space and may cause alveolar rupture with extension of air into the pleura. These reports may partially confound the role of COVID-19 induced injury in the onset of pneumothorax but the vast majority of patients on NIV do not experience this complication, based on our daily practice and literature evidence. In addition, air leak cases have been reported in some series prior to mechanical ventilation [4],[6],[17],[18],[19]. This highlights that barotrauma is not the disease and may suggest that COVID-19 makes normally low-risk patients at higher risk of pneumothorax, through a combination of diffuse alveolar damage and decreased lung compliance. Severity of respiratory symptoms may also be a potential trigger. The increase in respiratory effort to compensate ventilation/perfusion mismatches oxygenation and the frequent cough may increase the intra-alveolar pressure and contribute to the cystic lesion rupture and pneumothorax formation [17].

Our patient presented with elevated inflammatory markers and the CT scan findings were consistent with the hyper-inflammatory lung injury and the progression of COVID-19 disease. Even though he did not present with evidence of cysts or bullae on initial CT, the delayed time frame to onset of pneumothorax could have given rise to pneumothorax predisposing pulmonary changes. In fact, cystic lesions may require weeks to develop as a consequence of initial inflammatory aggression and resulting increased respiratory effort [2]. Thus, cases of pneumothorax in COVID 19 patients without cysts in initial CT may be explained by both the lower CT resolution to identify small lesions or their development during the follow-up [20]. COVID-19-induced pneumothorax can occur at any point of patient care course: upon admission revealing the disease [21], during hospitalization, following recovery and even after discharge from hospital [2]. This case introduces the consideration of tension pneumothorax as a cause of acute deterioration in patients with underlying COVID-19 pneumonia. Examination, clinical acumen, and bedside diagnostic techniques resulted in an expedited diagnosis of a pneumothorax and both early and adequate management while excluding differential diagnosis of severe pulmonary embolism and avoiding potential fatal positive pressure ventilation. The benefits of rapid bedside diagnostic techniques, such as point of care ultrasound, are increasingly being acknowledged in COVID-19 management [22],[23] but should be considered as a continuum of the clinical examination. In some cases, auscultation accuracy may be limited because of isolation clothes. The fact that patients' chests are no longer routinely auscultated because of exposure concerns or the lack of single-use stethoscopes may delay such life-threatening complication management, especially upon admission in an emergency setting [21]. When drainage is necessary, keeping the chest tube head pointed away from medical staff faces and negative pressure suction (−20 cm/H₂O) of pneumothorax to avoid air leakage are recommended to decrease the risk of iatrogenic infection [24]. The onset of pneumothorax in COVID-19 pneumonia course may be associated with more severity and worse outcome [4],[25],[26]. In a recent study, barotrauma in patients with COVID-19 infection and requiring invasive mechanical ventilation occurred at higher than expected rates and was associated with longer hospital stay and death [27]. In this case, our patient was apparently stabilizing before the sudden onset of pneumothorax, which triggered the course of respiratory deterioration. Moreover, progressive respiratory deterioration was noted despite lung expansion after drainage. We believe that the occurrence of pneumothorax in this case is consistent with evolutive parenchyma lesions and inflammatory exacerbation as assessed by the follow-up inflammatory markers and imaging. A proximal embolism on pulmonary imaging was ruled out but the existence of micro-thrombi worsening the shunt may be responsible for increased oxygen requirements, despite antithrombotic and corticoid therapies administration. Furthermore, no biological or microbiological evidence of surinfection was noted. This raises two questions regarding the unpredictable course of COVID-19 pneumonia and the prognostic value of the COVID-19/pneumothorax association. Future studies are needed to further characterize such prognostic implications.

During this pandemic crisis, emergent but treatable complications, such as pneumothorax, should not be neglected. Our case highlights the increased inflammatory response as a potential underlying mechanism of pneumothorax in COVID-19 patients, which may be associated with a poor prognosis. But further research is needed.