Pneumothorax, usually due to prolonged ventilation with positive pressure, is a relatively common complication of COVID-19 pneumonia, affecting up to 5.9% of patients7.
In 2004, Sihoe et al. reported a 1.7% incidence of SP in Severe Acute Respiratory Syndrome (SARS) patients8. In this series, pneumothorax was a late complication of SARS, occurring from 14 to 37 days after initial diagnosis, suggesting that a sustained period of lung inflammation is first required. Four patients were treated with chest drain, and two were managed conservatively. Chest drains were removed from 14 up to 31 days after insertion: permanence of drains was mainly due to prolonged air loss. None of the patients underwent surgical operation because of both severe impairment of patient’s lung function and high anesthetic risk, and because of concern for infection risk to operating room staff8.
As for novel coronavirus disease, other authors have already reported SP as a clinical manifestation for COVID-194,5,9,10; however, none of them was managed with a surgical approach.
Tian and colleagues11 reported of two patients submitted to pulmonary lobectomy for lung cancer who developed fever and respiratory failure after surgery: a postoperative oropharyngeal swab test revealed a Sars-CoV2 infection. A review of preoperative CT scan revealed bilateral GGOs in one of them, while the second developed radiologically detectable interstitial pneumonia on POD 2. This is actually the first report of COVID-19 patients with lung pneumonia submitted to lung surgery.
We report the first two cases of patients with novel coronavirus pneumonia with SP as presenting symptoms who were surgically treated.
Our first approach for both patients was chest drain, because we had no knowledge of whether COVID-19 could favor complications after surgery. However, after recurrence of SP in both patients we had to consider surgical treatment because of the need to avoid possible further respiratory distress in patients with ongoing interstitial pneumonia.
Intraoperative findings were of segmental and subsegmental areas characterized by atelectasis and vascular congestion, which proved to be very frail at manipulation and traction. These areas corresponded to GGO infiltrates at preoperative CT scan (Figs. 1 and 2). In particular, the rupture of the bulla at the apical segment of right lower lobe within an area of important vascular congestion and pulmonary atelectasis was probably the cause of patient 2 hemoptysis before hospital admission (Fig. 1).
Notably, despite regular coagulation parameters and no anticoagulant therapy, both patient had a tendency for bleeding during gentle manipulation, so that we preferred not to perform extensive pleural scarification. This finding may be explained by systemic inflammation, which can impair proper coagulation 12.
Finally, during pulmonary re-ventilation at the end of the procedure, the COVID-19 atelectasis areas did not re-expand as expected, due to loss of parenchymal compliance, despite the use of positive pressure air flows.
At histological specimen evaluation the main finding was an interstitial pneumonia with ongoing reparative processes, associated with vascular changes probably responsible for increased vascular resistance. In particular, intense, diffuse and uniform chronic inflammation was located in the septal interstitium, with prevalence of lymphocytes and plasma cells, as typically found in non specific interstitial pneumonia. Extensive endoalveolar fibroblastic overhangs located in the centrolobular seat were retrieved. Alveolar spaces were filled with rare alveolar macrophages and proteinaceous exudates. Signs of viral infection consisted of nuclear pseudoinclusions, indicating cytopathic effects.
Vascular changes were of multiple and extensive interstitial and endoalveolar blood extravasations, and marked myo-intimal thickening with associated blood stasis, sometimes with microthrombi. Subendothelial infiltration of lymphocytes, suggestive of endotelitis, was also retrieved (Fig. 3).
Lastly, areas of distal emphysema and subpleural bullous formations with hemorrhagic content were also observed.
These findings are comparable to what Tian has described11.
Although other authors have described SP as a clinical manifestation of SARS-Cov2 infection, this is the first report of diagnosed COVID-19 patients who underwent surgical treatment for spontaneous pneumothorax: intraoperative findings of rotten bulla in a COVID area explained why conservative treatment had failed in patient 2, as well as similar findings of dystrophic and fragile parenchyma within COVID areas for patient 1.
Differently from SARS patients8, in these cases pneumothorax was the first symptom of COVID-19 and occurred at the beginning of pneumonia. Moreover, none of them experienced severe respiratory impairment requiring invasive ventilation. This fact may reflect a different mechanism of pneumonia, where lung rupture does not occur as a final stage of prolonged inflammation, but is an immediate consequence of virus infection in selected patients: microvessel inflammation and capillary wall thickening at the early stages of the disease may play a significant role in this process, for which further studies are needed.
Despite understandable concerns for staff safety, all operative room members wore personal protection equipment as recommended by Italy SSI13 and none of them developed COVID-19 infection. At 1 month follow up, all staff workers were COVID-19 IgG-free.