The study included 22 healthy adult female New Zealand rabbits (Oryctolagus cuniculus) weighing 4.61 ± 0.52 kg. The animals were fed ad libitum with standard kibble and housed in individual cages with a grating floor intended for the species, with a light/dark cycle of 12 h.

They were randomly distributed into 3 groups with different survival times (Fig. 1). All the animals in each group were implanted with a BPS, except 1 from each group which acted as a negative control.

Fig. 1.

Distribution of animals in study groups.

(0.12MB).

The study device was a BPS manufactured by ELLA-CS sro, Czech Republic, constructed with a single 3.5 EP polydioxanone woven filament. The stents are self-expandable, measure 8 ⋅ 30 mm and are cylindrical in shape with a radiopaque gold marker at each end. They are supplied separately from the delivery system, which had to be loaded immediately before implantation. Polydioxanone is a biodegradable polymer used since the 1980s in sutures and other prostheses. It causes little tissue reaction and is neither antigenic nor pyrogenic. This material is a polyester that is degraded by hydrolysis (accelerated in acid medium), resulting in degradation products that are eliminated in the urine or metabolized in the citric acid cycle. In suture form, the absorption time is 182–238 days.21,22

The animals underwent a clinical examination prior to the procedure. The implantation was carried out under general anesthesia with intramuscular administration of a combination of medetomidine (Sedator®, Eurovet Animal Health, Netherlands) (0.5 mg/kg) and ketamine (Imalgene 1000®, Merial, Spain) (25 mg/kg). The animals were kept oxygenated with a laryngeal mask and anesthesia was maintained with intramuscular injections of ketamine (0.5 mg/kg) every 20 min. Monitoring continued throughout the procedure.

In order to maintain the laryngeal mask to control the airway and oxygenate and ventilate the animals throughout the procedure, implantations were performed through a mini tracheostomy, under fluoroscopic control. In each intervention the animal was positioned in a supine posture with the neck hyperextended. A small incision was made in the neck and the tissues were dissected until the trachea could be seen. The trachea was punctured immediately caudal to the cricoid ring between two tracheal rings with a peripheral 18 G plastic trocar (Introcan®, B. Braun, Germany). A total of 0.15 ml of 5% lidocaine was administered through the plastic sheath of the trocar (B. Braun, Barcelona, Spain). A 0.035'' Teflon guidewire (StarterTM Guidewire, Boston Scientific, Ireland) impregnated with lidocaine was advanced under fluoroscopic guidance. The access was dilated with a 12 Fr dilator (12 F dilator, Cook Medical, min). The delivery system loaded with the stent (11.8 Fr) was advanced over the guidewire and the stent was released at least 1 cm cranial to the carina. Implantation was performed under fluoroscopic guidance, visualizing the radiomarkers. The same procedure was performed in the control animals, but using an empty delivery system.

Before the procedure, 50 mg/kg of oxytetracycline (Terricina LA, Pfizer, Madrid, Spain) was administered for antibiotic coverage and 0.2 mg/kg meloxicam (Metacam ®, Boehringer Ingelheim, Germany) was given for post-surgical analgesia.

Technical success was defined as an uneventful procedure.

Following implantation, the animals were clinically monitored during the pre-defined survival period (30, 60, and 90 days depending on the group), during which physical status and the presence of respiratory symptoms were assessed. Clinical success was defined as survival to the end of the programed period and no symptoms requiring additional therapy.

After the scheduled survival time, CT studies (GE Healthcare Brivo™ CT325 16 slice) were performed under sedation in all animals (a CT scan was performed in the pilot animal at 30 and 90 days). The studies were performed with a slice thickness of 1 mm at 0.5 mm intervals and multiplanar reconstructions were made.

To facilitate the interpretation of the images, the trachea was divided into 5 sections, as shown in Fig. 2. If the stent had degraded, its position was estimated in relation to its location in the fluoroscopy performed during implantation.

Fig. 2.

Division of the trachea for CT study.

(0.18MB).

The presence of the radiomarkers and images suggestive of granulomas were evaluated, and the thickness of the tracheal wall was measured.

After the CT, the animals were sacrificed while still sedated with pentobarbital sodium (Dolethal®, Vétoquinol, France). Tracheas were extracted for macroscopic and microscopic study.

The samples obtained were fixed in 10% formalin, embedded in paraffin, sliced, and stained with hematoxylin-eosin. They were then divided into 3 portions (cranial, medial and caudal) for histopathological study. The degree of changes was assessed using the scoring system shown in Table 1.23

All statistical data and analyses were processed using SPSS Statistics (IBM SPSS Statistics for Macintosh, version 21.0; IBM Corp., Armonk, NY, USA). An error level of 0.05 was set. Qualitative variables were expressed as frequencies and quantitative variables were described as mean ± standard deviation. Qualitative variables were compared using the likelihood ratio or Fisher's exact test. Before the comparisons, the quantitative variables were tested for normality using the Shapiro-Wilk test. If the data followed a normal distribution, Student's t test was used for independent samples (2 means) or variance analysis, ANOVA (more than 2 means). The Mann–Whitney U test or the Kruskal–Wallis test were used for non-normal distributions. The Wilcoxon test was used for paired samples.

Radiomarkers were detected in all implanted animals who underwent CT at 30 days post-implantation (n = 7). They were present in half (n = 3) of the animals in the D60 group, and had disappeared in the other half (n = 3). In the D90 group, radiomarkers had disappeared in all animals (n = 7) (p < 0.001; Cramér’s V: 0.837) (Fig. 3).

Fig. 3.

CT images, sagittal reconstructions. A) Control. B) D30 (undegraded stent). C) D90 (degraded stent).

(0.05MB).

No stent migration was observed in any CT where radiomarkers were present. In CTs where radiomarkers were not observed in the trachea, they were not observed in the lung either. In 7 CTs (30.4%), images suggestive of granulomas were observed, with preserved tracheal patency in all animals. The location of the possible granulomas was: A (n = 1), B1 (n = 3) and C1 (n = 3). Significant differences were observed in area C1 (p = 0.038; Phi: 0.509) in the thickness of the tracheal wall depending on whether possible granulomas were present (1.37 ± 0.30 mm) or absent (1.02 ± 0.13 mm). No significant differences were found between controls and implanted animals in any of the studied parameters. Tracheal wall thickness results are shown in Fig. 4.

Fig. 4.

Tracheal wall thicknesses in CT study. Graph A: Measurements according to the presence or absence of radiomarkers. Graph B: Measurements according to post-implantation time.

aSignificance by Student's t-test.

bSignificance by Kruskal–Wallis test.

cSignificance by ANOVA.

(0.29MB).

In all cases in which radiomarkers were observed on the CT, the presence of the stent was confirmed in the macroscopic study of the specimen, while in cases in which they were not, no stent remains were found in the anatomical specimen.

The histological study revealed no granulomas that could confirm the images observed on CT. Stent reepithelialization did not occur in any case in which the stent was still present, and only impressions of filaments in the epithelium or epithelial invaginations over the filaments that were incompletely covered were observed.

In the implanted animals, significant differences (p = 0.028) were found in the degree of histological change between area A scores (4.56 ± 2.23 points) and area C1 scores (5.22 ± 2.60 points), but no significant differences with area C2 were observed (4.56 ± 2.31 points).

The total scores are shown in Fig. 5 and the images in Fig. 6.

Fig. 5.

Overall mean scores from histopathological studies. Yellow: Scores depending on if stent was present or degraded. Green: Post-implantation time scores.

Values with the same superscript indicate the absence of statistically significant differences, according to the Mann−Whitney U test for matched values.

* Significance by Kruskal-Wallis test.

(0.11MB).

Fig. 6.

CT Pathology images. A) 30D group B) 60D group (undegraded stent). C) 60D group (degraded stent). D) 30D group E) Control group.

(1.41MB).

Studies performed at 30, 60, and 90 days showed histological changes in the implanted animals. In general, there was a slight increase over time in the thickness of the epithelium, and a change in the cell populations of the epithelium with an increase in the number of caliciform cells and a decrease in ciliated cells, although the cilia generally appeared normal. Both chronic and acute inflammatory foci were also observed, although in the latter the differences are not significant with respect to the control group.

In the groups sacrificed at 30 and 60 days post-implantation, a slight increase in subepithelial thickness was observed, with no significant differences between groups. In cases where the subepithelial thickness was less than 400 μm (less than that of the control animals), the connective tissue appeared denser.

The squamous metaplasia foci observed appeared as small foci surrounded by normal epithelium. The scores obtained for squamous metaplasia and neovascularization were highest in the D30 group and lower without being significantly different in the D60 group, while in the D90 group there were no foci of squamous metaplasia or neovascularization. There are significant differences between the scores of the D30 and D90 groups in both the squamous metaplasia parameter (p = 0.001) and the neovascularization parameter (p = 0.006).

With respect to the sum of the score of all the parameters, a decrease in the scores over time was observed, the mean of the scores of the D90 group being significantly different from those of the D30 group (p = 0.003) and the D60 group (p = 0.003).

With respect to degradation, two situations were found: complete stent degradation and stent persistence. Significant differences in mean total scores were observed between cases where the stent was still present and those where it had already degraded (p = 0.001). The scores for subepithelial thickness, squamous metaplasia, neovascularization, and acute inflammation was 0 in all animals in which the stent had degraded, whereas in animals in which the stent was still present, the subepithelial thickness had increased in 6 of 9 animals (66.67%); the difference with the degraded group was significant (p = 0.004). Furthermore, all animals in which the stent had not degraded showed squamous metaplasia foci (p < 0.001), 8 of 9 animals showed neovascularization foci (88.89%; p < 0.001), and 7 of 9 showed acute inflammatory foci (77.78%; p = 0.002).