To the Editor:
Severe COVID19 patients require mechanical ventilation (MV) in up to 15.2% of cases [1]. Postintubation (PITS), post-tracheostomy (PTTS) tracheal stenosis and tracheoesophageal (TE) fistulas are tracheal complications of MV. The incidence of tracheal complications of MV was 1 in 200 000 patients annually before the COVID19 pandemic [2]. The most important risk factors for these complications include prolonged MV, reintubation and excessive tracheal tube cuff pressure (>30 cm H2O), all of which were common in mechanically ventilated COVID19 patients [3, 4]. We experienced a significant increase in referrals to our interventional pulmonology (IP) unit at a tertiary teaching hospital for PITS and PTTS treatment after the most severe COVID19 surges and thus conducted this retrospective study with prospective follow up aiming to describe this unique cohort of patients.
All patients referred to our interventional pulmonology (IP) unit for evaluation and treatment of PITS, PTTS or TE fistulas complicating MV for COVID19 from March 2021 to June 2022 were included. Clinical and radiological data were collected from hospital records of both our and referring hospitals. Patients were followed up with flexible bronchoscopy (FB) in 4-week intervals and additional diagnostic imaging as indicated. Assessment of the severity and grading of stenosis was done with computed tomography (CT) and FB. Based on the Myer-Cotton and McCaffrey classification systems and key features such as length, shape, location and cartilage involvement, stenoses were divided in two groups - simple and complex [5, 6]. After assessment, patients were treated either endoscopically or surgically. Endoscopic treatment was performed under general anaesthesia using rigid bronchoscopy (RB). First, tissue incision was made at 3, 9 and 12 óclock using an endobronchial electrocauter. Mechanical dilatation followed with Rusch Wiruthan® (Willy Rusch GmbH, Kernen, Germany) or CRE™ Pulmonary (Boston Scientific, Natick, MA, USA) dilatation catheters and rigid bronchoscopes with a diameter of 1.1 and 1.4 cm. Surgical treatment included cricotracheal or tracheotracheal resection and end-to-end anastomosis. Microsoft Excel (Microsoft, Redmond, WA, USA) was used to tabulate data, calculate frequencies, percentages and perform statistical analysis. The study was approved by the ethics committee of University Hospital Centre Zagreb (class 8.1-22/21-2; number 02/013 AG).
A total of 24 patients were evaluated and 21 of them were treated during the study period. The median duration of follow up was 8 (IQR 6–9) months. Patient characteristics, ICU treatment strategies and treatment data are shown in table 1. Most patients (18/24, 11/14 and 5/7 of total, RB and surgical patients, respectively) were referred from a single hospital, after treatment in the same ICU. The time between extubation and clinical presentation varied widely, from 5 to 110 days. The mean diameter of simple PITS/PTTS was 5.6±1.5 mm. Tracheoesophageal fistula developed in 2 patients with complex PITS/PTTS. 14 patients (58.3%) had simple and 10 (41.7%) complex PITS/PTTS. RB was used as the initial treatment in 14 patients which mostly had simple PITS/PTTS (85.7%), while surgery was the initial approach for 7 patients with mostly complex PITS/PTTS (60%). All 14 patients treated with RB achieved complete recanalisation without any procedure related complication. Despite good initial success, restenosis occurred in 93.9% patients after a median of 32 (IQR 22.75–61.25) days, requiring a total of 32 RBs which established airway patency in 10 (71.4%) patients, while 4 (28.6%) patients ultimately had to be surgically treated due to severe restenosis. Three patients (42.8%) initially treated surgically needed endoscopic interventions due to post-surgery restenosis. Successful recanalisation was achieved in all of them. Aside from the high rate of restenosis, no other major complications occurred during or after RB. A total of 39 therapeutic RB and 111 diagnostic FB were performed during the study period.
Patient characteristics, ICU treatment strategies and treatment data
Our IP department is the Croatian national centre for RB which admits patients with tracheal disease from the whole country. In the years preceding the pandemic (2017–2019) we treated an average of 12 PITS/PTTS patients annually. This number almost doubled after the COVID19 pandemic, with 24 referrals during the 15-month study period. In concurrence with previous studies, we observed a disproportionate number of obese patients, with an even higher incidence of patients with DM (41.7% versus 17.4%) [7]. Obesity is a well described risk factor for PITS/PTTS in pre-pandemic studies and also complicates the course of acute COVID19 with a higher rate of MV and airway management issues among these patients [3, 8]. DM could play an important role in the development of tracheal complications due to the changes in the microvasculature and impaired wound healing [3, 9, 10]. The median and maximum duration of MV, reintubation and tracheostomy rates were similar to other cohorts and higher than in non COVID19 patients [7, 8, 11]. The wide time window between extubation, clinical presentation and diagnosis was also almost identical to previous reports [7]. Interestingly, more patients in our study presented with simple stenosis than in previous reports where complex stenosis was more common [7, 12]. Complex stenosis was significantly more common among PTTS patients.
We used the same management strategy for both simple and complex stenosis after COVID19 as in non COVID19 patients. However, the treatment course of PITS/PTTS after COVID19 was complicated with an extremely high rate of restenosis requiring a multimodal approach. Initial RB was preferred for patients with simple stenosis, while complex stenosis was treated surgically whenever possible. The reported success rate of endoscopic procedures of 40–82% depends on the length and complexity of stenosis. Surgical procedures have a success rate as high as 95% [13]. Despite the high rate of restenosis in our cohort (93.9%), we ultimately had a high treatment success rate of 71.4% among the 14 patients initially treated with RB, after an average of 2.3 endoscopic procedures per patient. We avoided stent placement as per usual institutional protocol due to potential complications, similarly to the protocol reported by Freitas et al. who treated simple PITS with even >4 dilatations in some cases [12]. We observed a substantial rate of restenosis after surgery (42.8%) requiring endoscopic management of restenosis. This high failure rate could be attributed to the complexity of stenosis and impaired healing after surgery, which is most likely a consequence of chronic inflammation, low tissue quality after prolonged critical illness and patient comorbidities such as DM [9, 10, 14].
The most intriguing finding is that the majority of patients (18/24) were referred from a single hospital, after treatment in the same ICU, from an area that experienced the most severe COVID19 surge in Croatia. The extremely high number of patients that required ICU treatment presented unique organisational challenges. Physician and nurse shortages led to fatigue and stress which could have contributed to suboptimal ICU care. This highlights the fact that PITS and PTTS are, despite some contributing intrinsic patient characteristics, in essence iatrogenic diseases preventable by meticulous ICU care.
In conclusion, RB was successfully used for post COVID19 PITS and PTTS as the primary treatment and pre- and post-surgery, which emphasizes the importance of multidisciplinary management. Given the high restenosis rate and complicated treatment course, care should be given to prevent tracheal complications with maintenance of high ICU quality standards. Finally, delayed accurate diagnosis emphasizes the importance of awareness of tracheal complications among mechanically ventilated COVID19 survivors, since respiratory symptoms could be misattributed to parenchymal lung disease or neuromuscular disease in long COVID19 syndrome.
Footnotes
Conflict of interest: Summary conflict of interest statements: the authors have no conflicts of interest to declare.
- Received August 1, 2022.
- Accepted December 27, 2022.
- Copyright ©The authors 2023
This version is distributed under the terms of the Creative Commons Attribution Non-Commercial Licence 4.0. For commercial reproduction rights and permissions contact permissions{at}ersnet.org