Biodegradable tracheal stents: our ten-year experience with adult patients

We present a retrospective review of our collected database of adult patients who underwent BD polydioxanone tracheal stent implantation between September 2013 and December 2022 at the Department of Respiratory Medicine of the 1st Faculty of Medicine in Prague and Thomayer University Hospital. All patients who received BD polydioxanone custom-made stents (ELLA-CS Ltd., Hradec Kralove, Czech Republic) were included in the study.

The indications for stent implantation were functionally significant nonmalignant tracheal stenosis and tracheomalacia. Every patient was reviewed by two interventional pulmonologists and a thoracic surgeon to determine the best approach, and only those who were not candidates for primary surgical treatment underwent stent implantation. Contraindications included the presence of aero-digestive communication and pregnancy. We focused on demonstrating the effectiveness and safety of this approach. Special attention was given to a 60-day period postimplantation, which was considered generally safe by both us and the manufacturer, i.e., the period of anticipated guaranteed mechanical efficiency of the stent. As there is no chance of longer persistence of the biodegradable stent in the airways, a six-month period after stent implantation, or a second implantation if there was more than one stent, was deemed sufficient to evaluate long-term effects. More than 2 stents could be implanted. For practical reasons, mainly due to the initial nature of the narrowing and the related evaluation of stent efficacy, only the first or second stent effects were assessed. When relevant, for the sake of completeness of the overview, essential data on the implantation of other (third and fourth) BD stents and data from patients with longer follow-up are mentioned.

Efficacy was defined as an improvement in tracheal narrowing and relief from symptoms. The monitored phenomena included immediate postprocedure improvement of tracheal narrowing, details related to the procedure, and postoperative outcomes. A correct endoscopic effect was defined as a lumen improvement of at least 2 degrees according to Freitag´s classification [7] lasting 60 days postimplantation with no breach of integrity detected. Cases that did not fulfil this definition were considered stent failures. Clinical effectiveness was related to the need for subsequent bronchoscopy intervention for restenosis. Improvement of symptoms was defined as the reduction or disappearance of dyspnoea, easier coughing, loss of stridor, etc. Weaning from the tracheostomy tube and/or ventilator was assessed separately in patients as appropriate. Safety parameters included periprocedural events and complications, the need for stent fixation, endoscopy-detected situations, and complications, as well as the development of new symptoms, such as dyspnoea, stridor, respiratory failure, coughing up of stent fibres, haemoptysis, worsening of cough, and difficult expectoration. For events requiring intervention, the need for acute treatment, i.e., treatment carried out within 3 days from the first contact with the physician, was evaluated.

The time to detection of the first signs of disintegration of the stent (onset of rapid stent degradation), as well as the time to detection of the loss of the mechanical efficiency of the stent, were recorded. The first observed event was coughing up of the stent fibres experienced by patients or disruption of the fibres or the mesh of the stent being noted during bronchoscopy. The second event was a recurrence of stenosis due to apparent stent degradation or significant loss of stent integrity, as observed via bronchoscopy. Demographic data, type of stenosis (narrowing), aetiology and severity were collected from the hospital information system database. The presence of a dynamic component of stenosis (localized tracheomalacia) or diffuse tracheomalacia was considered a separate factor influencing stent degradation and outcomes.

We used self-expandable, biodegradable, polydioxanone tracheal stents manufactured by ELLA-CS Ltd. (Hradec Kralove, Czech Republic) with the following dimensions (mm) (diameter × length): 18 × 50, 18 × 70, 20 × 50 and 20 × 70. The choice of stent size was guided by the decision of the bronchoscopist. Before implantation, the stents were manually loaded into the dedicated application apparatus, which worked as a pull–back delivery device. The trachea was intubated with a rigid bronchoscope with an internal diameter of 8.5–11.0 mm (Karl Storz, Germany). Patients were placed under total intravenous (i.v.) anaesthesia with jet ventilation. The distance was measured using a flexible bronchoscope and, when possible, under direct visualization using a thin flexible bronchoscope or rigid optics, and the stent was advanced to the desired location. Fluoroscopy was not used. After insertion, the position of the stent was adjusted using rigid forceps, and stent balloon dilatation was performed to compress the mesh of the stent towards the tracheal mucosa and to allow for better stent reopening. If deemed necessary, a thoracic surgeon secured the stents via external (percutaneous) fixation [8].

After insertion, patients were observed in the hospital for at least one day; nebulized salbutamol, ipratropium and ambroxol were administered; and i.v. methylprednisolone (40–80 mg) was administered at the discretion of the physician. The first follow-up bronchoscopy was performed during the first week after the procedure. Patients were equipped with an aerosol nebulizer with a recommendation for regular inhalation of strongly mineralized water (Vincentka) and/or ambroxol solution b.i.d. A short course of alpha-aescin (40 mg orally, t.i.d., two weeks) was recommended to mitigate concomitant oedema. Additional follow-ups were carried out on a monthly or as-needed basis.

Descriptive statistics and graphical presentations were used to summarize the data. Categorical variables were calculated and are expressed as counts and percentages; their relationships were investigated using Fischer´s exact test. For continuous variables, means and standard deviations (SDs) were calculated; t tests were applied for normally distributed variables, while Wilcoxon tests were used for nonnormally distributed variables. For more than two groups, the Kruskal‒Wallis test was used. The associations between variables were measured by the Pearson correlation coefficient. All calculations were performed with the help of the statistical program JMP 16.2.0 (SAS Institute Inc.).

The majority of stent implantations went smoothly and were followed by a substantial increase in lumen capacity. The stents showed good adaptation to the airway anatomy immediately after insertion (Fig. 3). Of the 65 stenting procedures, 8 (12,3%) were accompanied by difficulties mostly related to the rigid technique or the general condition of the patients. The technical difficulty of placement seemed to be somewhat lower than that of silicone stents, which are slightly more complex, or almost comparable to that of SEMSs. Rigid bronchoscopy and general anaesthesia are necessary for tracheal application. Some SEMSs can be introduced with a flexible technique, which is a theoretical advantage [9]. Insufficient reopening due to a stiff stricture in patients is unlikely to be serious and should not be considered a complication, unless it is followed by distal migration of the stent (under the stenosis). In this case, extraction of the stent was subsequently problematic because the stent was completely deformed when it was pulled out with rigid forceps (Fig. 4). This issue appears to be typical of BD stents since they do not have any extraction threads and may deform in a disorganized manner during manipulation. To our knowledge, we are the first to report this complication with BD stents. Although the integrity of the polydioxanone stent and its fibres is pronounced at the beginning, the radial force itself is less than that of SEMSs. Therefore, thorough balloon dilatation or further additional treatment of the lumen before stent insertion is a prerequisite.

External fixation was performed in 21 (31.8%) implantations and, except in two patients, was performed immediately after deployment of the stent. External fixation appears to be a suitable supplement to implantation, especially for stents in the upper trachea and in cases of posttracheotomy stenoses, where there is a greater risk of migration [8, 10, 11]. Although it prolongs the procedure and makes the implantation more difficult, the stent does not have to produce a significant radial expansion force, which does not allow dislocation but does lead to mucosal irritation. The technique we used was similar to that used for permanent stents [8, 12]. Over the years, as we gained experience, we used it less often. The securing stitch was always removed after 2 to 3 weeks, so its impact on the patients' quality of life and aesthetics was minimal.

In the case of BD stents, we can distinguish between early dislocation when fibres of the stent are not incorporated in the mucosa and late dislocation, which involves pieces of mesh and fibres and a period of stent degradation. Early dislocation occurred in 4 patients—4 events (6.5%) out of 62 successfully implanted stents. In two patients, the stents were repositioned and fixed, while the third patient refused this procedure and was satisfied with overnight CPAP after stent removal. In the last patient, the stent moved 10 mm distally, and no intervention was necessary. Although we considered not securing the stents immediately after implantation because of our lack of experience at that time, we did not classify early stent dislocations as failures. The repositioned stents worked well afterwards, and their properties were not affected. In benign stenoses, we encountered a relatively large number of stent migrations, at a frequency of approximately 20% [11]. This incidence has been previously described to be greater, especially in cases of upper tracheal narrowing treated with either silicone stents or SEMSs [10, 13‐16]. Dumon et al. reported a frequency of migrations in 18.6% of all silicone stents inserted for benign tracheal stenosis [17]. However, with a newer type of silicone stent with narrow central and widened distal portions, which was intended for benign tracheal stenosis, there was no stent migration at a mean follow-up time of almost 23 months [18]. The low number of stent migrations recorded in this study is likely because BD stents are uncovered and adapt well to the anatomy of the airways and to the use of external stent fixation.

During stent disintegration, parts of the stent (mesh or fibres) may break off and move distally to the periphery. In two patients, this occurrence was detected during premature degradation. In another patient with a reduced ability to expectorate, the fibre pieces were stuck in the left main bronchus. The authors agree that in children, these events may be dangerous due to their smaller airway dimensions [19].

In classic stents, mucostasis is the most common side effect and is observed in at least one-third of all stent patients [11]. In our cohort, only one case of narrowing due to mucus stagnation was observed in a patient with a strongly impaired expectoration ability (1 of 39; 2.6%). Dumon et al. reported a mucostasis frequency of 5.7% [17] for silicone stents used for tracheal stenosis; Brichet et al. reported mucostasis in 5 of 18 patients stented due to postintubation tracheal stenosis [20]; and Karush and colleagues identified mucus accumulation as the most frequent cause of reintervention in patients with silicone stents placed for benign central airway stenosis (60%, 131/220) [14]. Sputum retention was significantly greater for both noncovered and covered modern SEMSs (15% and 42%, respectively) [21], and a frequency of 15% was reported by Dahlquist et al. for fully covered SEMSs, while Fortin et al. reported a frequency of only 2.5% [15, 16]. Mucostasis can be associated with halitosis resulting from colonization of the stent with bacteria and fungi [11]. None of our patients complained of this condition.

Subjective intolerance was observed in 5% of patients treated with fully covered SEMSs by Fortin et al. [16]. Xiong et al. reported cough and chest pain in 25% and 21% of patients, respectively, who were implanted with SEMSs; lower rates were observed for bare stents [21]. Haemoptysis occurred slightly more often with bare SEMSs than with covered SEMSs: 17% vs. 12%, respectively. In patients with silicone stents, the occurrence of haemoptysis is usually linked to the stent introduction and related trauma of the airways [17]. In our study, pain or discomfort related to the expansion force of the stent was not reported by any patient, while purulent and difficult expectorations were reported by 15% of patients during the first two months postimplantation and in 50% of patients thereafter when they experienced stent degradation. Only one patient reported haemoptysis, which was related to the repeated irritation of the mucosa due to contact with the loose portion of the stent in the unconstricted part of the trachea.

The inflammatory mucosal response is an important determinant of stenting success. In animal studies, necrosis of the epithelium and lamina propria at the point of contact between the stent fibre and the mucosa has been proven. Hyperplasia of goblet cells resulting in hypersecretion was also observed [2]. Valverde et al. [22] showed that successive stenting caused mild inflammatory changes in the tracheal wall and no increase in the collagen matrix or modifications in the cartilage structure. Generally, the response to damage is the formation of granulation tissue. In an animal study conducted by Choi and colleagues, granulation tissue growth occurred mainly beside stent fibres, reaching a maximum at 4 weeks and decreasing in thickness at 8 weeks postimplantation [23]. Rodriges Zapater et al. [24] reported that the overall granulation formation rate was 30% in their study of rabbits. In our study, significant granulation tissue growth with some narrowing of the lumen was detected in 9 out of 39 (23%) patients during the first two months postimplantation and in 16 out of 34 (47%) patients after the first two months. The contribution of this growth to renarrowing was difficult to assess in cases of concomitant stent degradation. In all patients, we exclusively used stents with a diameter of 18 or 20 mm, which could have played a role in the induction of granulations in patients with differently narrowed tracheas (Fig. 5).

Granulation tissue growth occurs to some degree in more than half of all patients; in approximately one-quarter of all patients, it becomes clinically significant [11]. Low complication rates were reported for some covered SEMSs (7.5%) [16]; however, other authors reported an incidence of up to 35% [15]. Granulomas induced by excessive mobility were reported in 17.2% of all silicone stents inserted for benign tracheal stenosis by Dumon and colleagues [17] and in only one of 13 patients treated with a silicone stent with a narrowed central part [18]. For uncovered SEMSs, granulation tissue growth (through the mesh of the stent) is generally more pronounced, making their extraction difficult, so they are not usually the first choice for benign stenoses.

Our stents allowed a subgroup of 14 patients to undergo removal of the tracheostomy tube or weaning from the ventilator. We thus build on the comprehensive work of Noppen et al., who focused on permanent stents to address this issue [25]. Evidence of the suitability of BD stents is supported by experience in both children and adults [5, 6, 19, 26]. We and other authors used a temporary combination of a tracheostomy cannula placed with its end in the stented section of the trachea to secure the airways [5, 6]. We mention this patient group separately to highlight one of the possible indications for BD stents. These patients were in very poor general condition and often had a poor ability to expectorate; therefore, from our point of view, they were unlikely candidates for a permanent stent. The removal of the tracheostomy tube improved their psychological state and enabled further progress in their rehabilitation.

At the onset of rapid degradation, stents quickly lose their mechanical properties and disintegrate. This period begins with the occurrence of a fracture in the first stent fibre and technically ends with the complete loss of the supporting capabilities. Even beyond this point, individual stent fibres can subsequently be found to be embedded in the hyperplastic mucosa. Some of the fibres are coughed up by the patient, and some are absorbed by the airway wall. Novotny et al. suggested [2] that swallowing and ingestion could have occurred in addition to expectoration. In rabbits sacrificed at week 10 after implantation, no stent material was found. Morante-Valverde and colleagues [22] observed complete degradation by the 14th week; Choi et al. [23], who worked with polydioxanone stents produced elsewhere, reported that the stent disappeared by the 12th week. The latter study assessed whether the fragments had entered the lungs, but no fragments were found. Unlike this animal experiment, we had to address dislocation of the bundle of fibres into the left main bronchus in one patient with impaired expectoration. Similar experience in a child was reported by Sztanó and colleagues [19], who considered unpredictable and nonlinear degradation to be a substantial disadvantage. The process of stent disappearance and its consequences in humans have been well described by clinicians [3‐6, 19]. Based on our observations, the predominant elimination pathway seems to be mostly determined by the apposition of the stent to the wall of the airways. The parts of the stent without close contact with the wall degraded first (Fig. 6). Our stents started to disintegrate earlier in patients with localized or extensive tracheomalacia (75.5 (13.9) vs. 82.2 (16.3) days (SD); p < 0.0001), which could be due to repeated mechanical stress on the stent. This observation agrees with the work of Choi et al. [23].

We achieved a “without stent” state in 22 out of the 34 patients. The mean (SD) time to cure was 6 (3) months, and the longest was 14 months. In addition to the fact that the stents themselves were responsible for tracheal remodelling, balloon dilations within and outside stenting procedures could have been beneficial (Fig. 7). The use of a combination of stenting and nonstenting techniques is common in postintubation and posttracheostomy stenoses and leads to permanent improvement in lumen spaciousness [27]. After degradation in our tracheomalacia patients, we did not observe reduced collapsibility of the lumen compared to the original state, which was contrary to our original assumptions [6]. We judged that improvements in the general condition of patients played a vital role. Dutau et al. reported that no recurrence after stent removal varied from 40% when a permanent stent was placed for 6 months to 70% when it remained in place for 18 months [28]. The time required to heal the stenosis thus appears to be longer. Galluccio et al., who used silicone stents in some simple stenoses and all complex stenoses, reported a mean duration of stenting of 18 months; in the case of complex stenoses, the mean duration of treatment (SD) was 20 (3) months [29]. With this method, they were able to cure 69% of complex lesions. Brichet and colleagues were less successful in this regard, but their treatment algorithm favoured stent extraction after 6 months in patients who became candidates for surgical treatment [20]. The BD stents were characterized by very good patient tolerance and safety within the first two months after implantation, but worse symptom control occurred starting in the third month. The safety of the stents in this period depends mainly on care management. We honoured the patient’s preference for surgical treatment, which unfortunately was not initially feasible for any of our patients. However, because of the sufficiently long periods of relief needed for rehabilitation, surgery could be performed later in seven of these patients. With BD stents, there is no need to consider the time of stent extraction due to their limited lifespan. Instead, the physician is faced with the decision of whether to dilate, restent, or switch to another treatment option. In our study, we detected an average of 2.47 therapeutic bronchoscopies per patient in the long-term monitoring group. Galluccio et al. reported a mean of 2.07 procedures in patients with simple stenosis and 3.27 in those with complex stenosis, using silicone stents in the latter cohort [29]. Sun et al. reported the average need for 4 or 5 procedures for successfully and unsuccessfully treated patients using a variety of procedures, including stents [27]. Focusing on third-generation SEMSs, Fortin et al. reported the need for more than two bronchoscopy interventions per patient [16]. Therefore, we believe that biodegradability can reduce treatment time while maintaining the number of necessary interventions at an acceptable level. Table 6 shows the positions of BD stents and permanent stents for various tracheal indications.

Our study had several limitations. This was a retrospective review, and the set of evaluated patients was still small and heterogeneous, including patients with different aetiologies and types of tracheal narrowing. The symptoms assessed were not quantified, and a quality-of-life assessment was not used. Endoscopic evaluation of airway dimensions was limited, and the study was conducted at a single institution. In the analysis of the patient group, the most clinically relevant phenomena and categories from the authors' point of view were considered, with follow-up duration as a covariate.