With interest, we read the recent report of a 73-year-old coronavirus disease 2019 (COVID-19) patient with COVID-19-associated aspergillosis (CAPA), developing massive hemoptysis and subsequent refractory cardiac arrest [1]. Post-mortem analysis revealed a massive gaseous embolism, which the clinicians suspected to be the result of a pneumovascular fistula secondary to CAPA. We thank Mombrun and colleagues [1] for their case presentation and would like to substantiate their hypothesis by showing data on a complication with suspected analogous pathomechanism.

Similar to pneumovascular fistula, pneumopleural fistula (PPF) with associated barotrauma secondary to CAPA seems just as reasonable from a pathophysiologic perspective. Barotrauma constitutes a relevant complication among ventilated COVID-19 patients and appears to occur more frequently in COVID-19-associated acute respiratory distress syndrome (ARDS) than in ARDS of other origin [2]. As data on underlying causes are scarce, we analyzed a propensity-matched cohort of ventilated COVID-19 patients from the University Hospital Cologne (UHC) to identify risk factors for PPF. PPF was defined as imaging-confirmed pneumothorax, pneumomediastinum or extensive subcutaneous emphysema. Treatment was performed according to standardized UHC protocols aiming for a Richmond Agitation-Sedation-Scale of 0 to -1 once prone positioning was terminated without regular use of neuromuscular blocking agents. Matching was performed based on a nearest-neighbor matching procedure without replacement with a threshold for absolute standardized mean differences of 0.1 in R, adjusting for suggested risk factors for barotrauma: age, sex, plateau pressure, and tidal volume per predicted body weight at the time of intubation or admission.

Of 143 COVID-19-ARDS-patients with complete biweekly CAPA-screening out of a total of 289 ventilated COVID-19-ARDS-patients at UCH (PPF prevalence 26%), 35 PPF-patients were matched with 35 non-PPF-patients. Regarding complications, a 37% higher incidence of ventilator-associated pneumonia (VAP) (p < 0.001) was observed in PPF-patients. Possible or probable CAPA, defined by the European Confederation of Medical Mycology (ECMM) and the International Society for Human & Animal Mycology (ISHAM) consensus criteria [3], was diagnosed in 40% of PPF-patients compared to 14% in non-PPF-patients (p = 0.016). The odds ratio (OR) for PPF was examined using multiple logistic regression analysis, including the factors initial Acute Physiology And Chronic Health Evaluation II (APACHE-II) score, VAP, CAPA, and duration of ventilation. Only APACHE-II score (OR 1.14, 95% confidence interval (CI) 1.02–1.31, p = 0.015) and CAPA (OR 5.35, 95% CI 1.11–40.8, p = 0.036) were independent predictors for PPF (Fig. 1).

Fig. 1
figure 1

Clinical images, challenges, and outcome. a Chest‐CT showing pneumothorax, pneumomediastinum and a defect at the left main bronchus (red arrow), verified by bronchoscopy, in a patient with COVID-19-associated pulmonary aspergillosis (CAPA). b Bronchoscopy, showing an ulcerating tracheitis as a manifestation of COVID-19-associated pulmonary aspergillosis (CAPA) (white arrow) with intra-tracheal hemorrhage. c Bronchoscopically extracted blood clot. d Kaplan–Meier estimator for overall survival of a propensity-matched cohort of critically ill COVID-19 ICU patients with and without barotrauma during mechanical ventilation. e ICU survival and incidence of possible or probable COVID-19-associated pulmonary aspergillosis (CAPA), as defined by the 2020 ECMM/ISHAM consensus criteria [3], in a propensity-matched cohort of critically ill COVID-19 ICU patients with and without pneumopleural fistula during mechanical ventilation. Each square represents one COVID-19 ICU patient. ICU Intensive care unit; CAPA COVID-19-associated pulmonary aspergillosis, ECMM European confederation of medical mycology, ISHAM: International society for human and animal mycology

Full size image

A possible correlation between fungal infections and barotrauma has been recently suggested in an observational study, identifying barotrauma in 68% of patients with COVID-19-associated fungal coinfections [4]. The invasive growth of hyphae destroys the integrity of the airways. This effect is amplified by positive airway pressure needed during mechanical ventilation [5]. These changes could not only predispose patients to develop PPF with associated barotrauma, but may also result in pneumovascular fistula by angioinvasion, as suspected by Mombrun and colleagues [1]. Although histopathological correlates are scarce [6], causal relationships between CAPA and both pneumopleural and pneumovascular fistula seem reasonable from pathophysiologic perspective. Besides limitations due to its retrospective and monocentric design, additional unaccounted confounders may exist despite propensity-matching. However, thorough biweekly CAPA-screening and homogeneous matching groups indicate high levels of representativeness and generalizability.

These findings therefore support the suspected theory and shed new light on CAPA-associated complications, emphasizing the importance of diagnostic strategies to allow for early treatment and contributing to the ongoing debate regarding antifungal prophylaxis in COVID-19-associated ARDS.