A rabbit model of tracheal collapse for optimal self-expanding metal stents

Abstract

Investigating the characteristics of tracheas can help the understanding of diseases related to the trachea, particularly tracheal collapse (TC) in dogs. This study aimed to compare the mechanical properties of tracheas from New Zealand White (NZW) rabbits and dogs and to introduce a method for inducing a model of TC in the normal trachea. Tracheal samples were obtained from NZW rabbit cadavers (n=5) weighing 3.62–3.92 kg and from dog cadavers (n=5) weighing 2.97–3.28 kg. Three live NZW rabbits weighing 3.5–4.0 kg were used to establish the model. The radial forces of both sample sets were measured using a digital force gauge and statistically compared. Subsequently, TC was surgically induced in three female NZW rabbits by physically weakening their tracheal cartilage under general anesthesia. Their clinical signs were monitored for 3 months, and radiographic examinations were performed monthly for 3 months. The mean radial forces of the two sample sets were comparable (P>0.05). The clinical signs, radiographic examinations, and macroscopic examinations were all comparable to those of dogs with TC. The cadaveric study between the rabbits and dogs demonstrated that the surgically induced rabbit model of TC is an excellent candidate for the experimental study of dogs with TC. This study also provides a reference of tracheal radial force values to enable selection of appropriate mesh types and wire diameters of self-expanding metal stents.

Tracheal collapse (TC) is commonly encountered as a cause of cough and is a progressive airway disease in toy- to small-breed dogs [6, 14]. Although the precise etiology of TC is not fully understood, it is caused by a weakening of the cartilage support in the airway [18]. TC in dogs is classified into traditional and malformation types [16]. Due to the laxity of the trachealis dorsalis muscle, the airway becomes flattened dorsoventrally, followed by chondromalacia and progressive cartilage collapse in traditional TC [17]. Dogs with malformation TC have a trachea with W-shaped cartilage rings, narrowing the airway dorsoventrally, causing static mechanical collapse. As stenosis of the trachea progresses, the amount of air entering the lungs rapidly decreases and clinical symptoms occur, including abnormal breathing sounds, exercise intolerance, and respiratory distress. Chronic inflammation of the tracheal mucosa causes coughing and a vicious cycle of worsening inflammation, which leads to poor quality of life or sudden death. Medical therapy is recommended as initial management, and surgery is indicated when medical management is ineffective. Surgical options include the installation of extraluminal ring prostheses and intraluminal stents. However, since extraluminal prostheses have a risk of laryngeal paralysis and airway obstruction during surgery, intraluminal stenting is preferred because it is a rapid, non-invasive procedure and results in quicker recovery [2, 13, 18]. However, major complications may occur after stent installation. Many attempts have been made by veterinarians to minimize complications of self-expanding metal stents (SEMS) used for dogs with TC because the severity of complications, such as stent fracture, migration, and inflammation, may lead to worsening of clinical signs [11, 12, 17].

However, an animal model of TC has not yet been developed. We selected rabbits to develop an animal model because the rabbit trachea shares characteristics with the canine trachea; in addition, surgical access to the rabbit trachea, which has a long cervical section, is straightforward. The unique property of the trachea is that the cervical region is exposed to atmospheric pressure and the intrathoracic region is exposed to pleural pressure [6]. Dogs with TC in the cervical region, which occurs more commonly, exhibit airway obstruction during inspiration [7, 8]. Rabbits have longer tracheas than dogs, particularly in the cervical region, which makes them an excellent candidate for an animal model [3].

Mechanical properties of SEMS determine their suitability in dog TC cases. We have previously reported on the relevance of radial force (RF) values in SEMS, including strong positive correlations between the RF and wire diameter and those between the axial force (AF) and wire diameter. From a clinical standpoint, a high RF and a low AF are favorable [5]. As the wire diameter of the SEMS increases, the RF and AF values increase significantly. However, if the RF is high, excess pressure exerted on the trachea may damage the tracheal walls and lead to pressure necrosis or tracheal wall perforation [4]. Nevertheless, guidelines on RF values in selecting stent design for SEMS are lacking. Selecting an optimal SEMS may be aided by comparison between RF values for the SEMS and those for the trachea.

TC is a disease unique to dogs. No previous study has proposed an animal model of TC or compared SEMS for TC in vivo prior to their implementation. This study aimed to (i) confirm the validity of a rabbit model (RM) of TC by comparing the mechanical properties of the trachea between rabbits and dogs with traditional TC, (ii) to evaluate whether an RM for TC would be created after physically weakening the rigidity of the tracheal cartilage, and (iii) to provide a reference for the tracheal RF in developing an optimal intraluminal stent as a treatment for TC.

MATERIALS AND METHODS

Ethics approval

This study was approved by the Institutional Animal Care and Use Committee of Konkuk University (IACUC number: KU17026-1) and followed the principles of the Declaration of Helsinki.

Cadaveric tracheal sample preparation in normal rabbits and dogs

Tracheal samples were obtained from five New Zealand White (NZW) rabbit cadavers (weight, 3.62–3.92 kg) purchased from DBL (Chungbuk, Korea) after having been used in a study unrelated to the trachea. A radiographic examination was performed to exclude any underlying diseases related to the trachea before euthanasia. The tracheas were collected within 30 min of euthanasia. General anesthesia was induced with intramuscular administration of ketamine (35 mg/kg; Ketamine 50 inj., Yuhan Co., Seoul, Korea), xylazine (5 mg/kg; Rompun, Bayer, Seoul, Korea), and butorphanol (0.1 mg/kg; Butophan inj., Myungmoon Pharmaceutical, Seoul, Korea), and euthanasia was performed by intravenously injecting KCl (35 mg/kg; Potassium chloride inj., JW Pharmaceutical, Seoul, Korea). Cervical tracheas were harvested through a ventral midline incision of the neck, and thoracic tracheas were approached through median sternotomy. Any soft tissue attached to the trachea was resected for precise RF measurement. Tracheal samples were obtained from five dog cadavers weighing 2.97–3.28 kg, donated for study purposes after euthanasia due to diseases unrelated to the trachea. Radiographic examinations were performed on all dogs prior to sample collection to eliminate any underlying tracheal diseases. The tracheas were collected using the same method used for rabbits. One tracheal sample was obtained from a dog previously diagnosed with TC and donated for research purposes. The donated dog, a castrated male Yorkshire Terrier, was aged 13 years at the time of death. The dog was diagnosed with TC grade 4, had a history of syncope and cyanosis, and weighed 3.1 kg at diagnosis. The cause of death was respiratory distress. This sample was compared with the tracheal samples from the rabbit and dog cadavers.

Radial force measurement

A longitudinal compression test was performed to measure the RF of each sample using a digital force gauge (SH-200^®, Sundoo Instruments, Zhejiang, China). Each trachea was situated between two flat plates (3 cm in width). The plates were compressed toward each other longitudinally up to 50% of the diameter of the trachea, and the RF was measured (Fig. 1). Each tracheal sample was measured five times to obtain a mean value; the same person performed all five measurements for consistency.

Fig. 1.

Longitudinal compression test. Each tracheal sample is situated between two flat plates, and radial force is measured when each plate is compressed longitudinally toward the other by 50% of the tracheal diameter.

Surgical procedure for a rabbit model of tracheal collapse

Three female NZW rabbits (DBL) weighing 3.5–4.0 kg were used in this study. The animals were anesthetized using a combination of intramuscular ketamine (35 mg/kg; Yuhan Co.), xylazine (5 mg/kg; Bayer), and butorphanol (0.1 mg/kg; Myungmoon Pharmaceutical), as described earlier. The animal was positioned in dorsal recumbency, with the neck extended, clipped from the caudal mandibular area to the cranial thorax, and prepared for aseptic surgery. The skin and subcutaneous tissue were incised along the ventral cervical midline, from the larynx to the manubrium. The sternohyoid and sternocephalicus muscles were separated to expose the cervical trachea. First, the RF of the trachea was measured using a digital force gauge, and tracheal cartilage rings 3 cm in length in the distal cervical region near the thoracic inlet were broken down with a No. 10 blade (Fig. 2). Care was taken to avoid penetrating the submucosal layer beneath the cartilage ring. The RF of the collapsed region in the trachea was measured using a digital force gauge to confirm cartilage fragility. Physical and radiographic examinations were performed after the surgery to ensure that there was no emphysema near the surgical site. The sternohyoid and sternocephalicus muscles were sutured with a simple continuous suture (3–0 Maxon; Covidien, Minneapolis, MN, USA), and the subcutaneous tissues and skin were closed. Postoperatively, enrofloxacin (10 mg/kg; Baytril, Bayer, Korea) and butorphanol (0.1 mg/kg; Myungmoon Pharmaceutical) were administered subcutaneously for a week.

Fig. 2.

Creating a rabbit model of tracheal collapse. (A) The trachea is exposed after a ventral midline incision. Subsequently, a length of 3 cm of tracheal cartilage rings in the distal cervical region is measured. (B) Radial force of the normal trachea measured using a digital force gauge prior to cartilage damage. (C) Cartilage rings dissected using a No. 10 blade. (D) Radial force of the damaged cartilage measured.

Follow-up evaluation

All the rabbits were observed daily for 3 months to evaluate their clinical signs related to TC, and any respiratory distress was recorded by using the respiratory distress grading system [10]. Respiratory distress was assessed using the following grading of respiratory distress: 0 =normal, respiration without goose-honking sound; 1 =mild distress, intermittent goose-honking sound observed in excitement; 2 =moderate distress, intermittent goose-honking sound observed at rest; and 3 =severe distress, continuous goose-honking sounds observed [10].

Radiography (TITAN 2000^®; Comed, Seongnam, Korea) was performed before and after the surgery. Follow-up radiography was performed monthly for 3 months. Physical palpation near the surgical site and radiographic examination were performed regularly to confirm that the submucosal layer was not penetrated, as penetration may result in emphysema. Three months after the surgery, the rabbits were euthanized by intravenously injecting KCl (35 mg/kg; JW Pharmaceutical) under anesthesia with intramuscular administration of ketamine (35 mg/kg; Yuhan Co.), xylazine (5 mg/kg; Bayer), and butorphanol (0.1 mg/kg; Myungmoon Pharmaceutical).

Tracheal samples from the cricoid cartilage to the thoracic inlet were successfully extracted from all three RMs after euthanasia. These samples were compared with tracheal samples from the dog previously diagnosed with TC.

Histological findings

Tracheal samples collected from the normal rabbits, RMs, and dog with TC were fixed in 10% neutral buffered formalin for 48 hr, washed for 30 min, and decalcified in a decalcifying solution (Decalcifying Solution-Lite; Sigma-Aldrich, St. Louis, MO, USA) for 6 hr. After washing for 30 min, the tracheal samples were fixed in 10% neutral buffered formalin for 17 hr. The same procedure was repeated the following day. After embedding in paraffin, the specimens were sectioned at 4-µm thickness and stained with hematoxylin and eosin. Thereafter, morphological changes were compared among the trachea samples from a normal rabbit, an RM, and the dog previously diagnosed with TC.

Statistical analyses

Statistical analyses were performed using SPSS Statistics for Windows version 24 (IBM, Armonk, NY, USA). Differences for the clinical grades of respiratory distress between months were assessed using Friedman’s test with post hoc pairwise comparisons using the Wilcoxon signed-rank test. RF values were obtained from the normal tracheas of dogs and rabbits and were compared using the independent t-test. A P-value of <0.05 was considered statistically significant. The correlation between the RF values and weight was examined via simple correlation analysis.

RESULTS

Rabbit model of tracheal collapse

The integrity of the cartilage rings was physically compromised when creating an RM for TC. To ensure suitable cartilage rigidity, the RF must be uniformly maintained below 0.08 kgf. Such a value was obtained in the dog with TC, a value that was difficult to detect. The mean compression force values of the three RMs before and after surgery were 0.3 and 0.07 kgf, respectively. After the cervical cartilage rings were weakened, the collapsed trachea was visualized, particularly during inspiration.

Radial force of tracheal samples

The RF values from the tracheal samples of the rabbits and dogs are summarized in Table 1 along with their weights. The median values were selected from five measurements. The RF of five samples was expressed as the mean ± standard deviation (SD) kgf. The mean tracheal RFs of the five rabbits and five dogs were 0.43 ± 0.004 and 0.43 ± 0.030 kgf, respectively. RF values from the tracheas of the normal dogs and rabbits were normally distributed, and the tracheal RF values were comparable between rabbits and dogs (P>0.05). The dogs’ characteristics are summarized in Table 2.

Table 1. Tracheal radial force measurement and weight of normal rabbits and dogs.

Samples	Weight (kg)	RF (kgf)
Rabbit 1	3.86	0.4
Rabbit 2	3.92	0.46
Rabbit 3	3.76	0.41
Rabbit 4	3.62	0.39
Rabbit 5	3.81	0.48
Mean ± SD	3.79 ± 0.10	0.43 ± 0.04
Dog 1	2.97	0.39
Dog 2	3.05	0.43
Dog 3	3.01	0.41
Dog 4	3.17	0.44
Dog 5	3.28	0.48
Mean ± SD	3.09 ± 0.11	0.43 ± 0.03

RF=radial force.

Table 2. Characteristics of five dog cadavers evaluated for tracheal radial force and compared to five rabbit cadavers.

Samples	Weight (kg)	Breed	Age	Sex
Dog 1	2.97	YT	11 y	M
Dog 2	3.05	Maltese	10 y	MC
Dog 3	3.01	Mix	2 y	F
Dog 4	3.17	Maltese	9 y	MC
Dog 5	3.28	Shi Tzu	11 y	FS

YT=Yorkshire Terrier, M=Male, sexually intact, MC=Male, castrated, F=Female, sexually intact, FS=Female spayed.

Correlations between tracheal radial force and body weight of normal rabbits and dogs

Positive or negative linear correlations between the tracheal RF of rabbits and dogs and their weights were assessed using simple correlation analyses. The tracheal RF of rabbits was not correlated with their weights (P>0.05). There was a strong positive correlation between tracheal RF and weight in dogs (r=0.985, P=0.002, and R²=0.97) (Fig. 3).

Fig. 3.

Linear correlation between tracheal radial force and weight of dogs. A strong positive correlation is revealed (r=0.985, P=0.002, and R²=0.97).

Follow-up evaluation

All RMs survived for 3 months after the surgery. During daily clinical evaluation, all rabbits exhibited respiratory distress ranging from grade 2 to grade 3; no significant differences in clinical grade were observed between the induced models. Monthly clinical grade in the first month was higher than that in the second (P=0.102) and third month (P=0.102), and the grade in the second month was higher than that in the third month (P=0.276), although the differences were not statistically significant.

In all three RMs, the chest radiographs were unremarkable prior to the surgery. After the surgical procedure, the trachea collapsed dorsoventrally in the cervical region, as observed during thoracic radiography (Fig. 4). Tracheal diameters in the cervical region were measured before and after the surgery and are shown as the mean ± SD (mm). The tracheal diameters before and after the surgery were 6.72 ± 0.11 and 2.27 ± 0.12 mm, respectively.

Fig. 4.

Thoracic radiographic examination of rabbit models at full inspiration. (A) Radiographic examination findings obtained before the surgery are unremarkable. (B) A dorsoventrally collapsed trachea is visualized through radiography after the surgery. (C) The magnified image provides better visualization of the collapsed trachea, which is indicated by the yellow arrowhead.

Gross examination of tracheal samples from a rabbit model and a dog with tracheal collapse

Collapsed areas of the trachea were observed in the cervical region of all RM samples. The tracheal sample was flattened dorsoventrally from 1 cm caudal to the cricoid cartilage to the end of the sample. Furthermore, all harvested tracheal samples were palpated gently, and their rigidity was compared. The samples lacked rigidity in the area of surgically induced TC, and the sample from the dog with TC lacked rigidity throughout (Fig. 5).

Fig. 5.

Harvested tracheal samples of a rabbit model and a dog with tracheal collapse. (A) Surgically induced trachea of a rabbit model dorsoventrally collapsed in the proximal cervical region and obliquely collapsed in the middle cervical region (yellow arrohead). (B) The collapse is easily detectible throughout the trachea sample of a dog with tracheal collapse (yellow arrowheads).

Histological findings

Tracheal samples from a normal rabbit, an RM, and the dog with TC were compared after hematoxylin and eosin staining (Fig. 6). Morphological changes were noted in samples from the RM and dog. The tracheal lumens of the RM and dog were significantly smaller than those of the normal rabbit. The histological characteristics of the canine tracheal sample were similar to those of general TC dogs—the tracheal membrane was stretched, the tracheal lumen had collapsed, inflammatory cells were present in the tracheal mucosa, and there was considerable edema (Fig. 6D). In the RM, tracheal mucosal epithelial cells were present, and no edema or inflammation was detected in the tracheal mucosa (Fig. 6E).

Fig. 6.

Microscopic views of the trachea in rabbit models and a dog with tracheal collapse. (A) A morphological change in the lumen is noted in the tracheal sample of a rabbit model. (B) The area inside the tracheal lumen of the dog previously diagnosed with tracheal collapse has decreased. (C) The lumen of a tracheal sample collected from a normal rabbit maintains the morphology of the C-shaped cartilage ring. (D) Canine trachea with tracheal collapse. Narrowed tracheal lumen due to prolapse of the dorsal tracheal membrane, with an increase in loose connective tissue and smooth muscle. A few glands and goblet cells are noted in the tracheal membrane. Inflammatory cell infiltration and obvious edema are noted in the tracheal mucosa. (E) Rabbit model of tracheal collapse. Incised tracheal cartilage and slightly narrowed tracheal lumen are observed in the trachea of the rabbit model. No change in the length of the tracheal membrane is noted.

DISCUSSION

An RM was created for the development of a SEMS suitable for the canine trachea. RF values in dogs and rabbits were similar. TC was surgically induced in rabbits, and the reduced tracheal lumen was confirmed through follow-up tests. Clinical symptoms of RMs were comparable to those of TC dogs, and morphological comparisons between dogs and rabbits were made via radiographic and histopathologic assessments.

In the present study, the tracheal RF values, measured via 50% longitudinal compression, were comparable between rabbits weighing 3.79 ± 0.11 kg and dogs weighing 3.09 ± 0.12 kg, suggesting that the RM is representative of TC in dogs. In addition to the mechanical similarities between the tracheas of rabbits and dogs, using an RM for TC to develop an optimal SEMS has certain advantages. Compared to other laboratory animals, rabbits are easy to raise, harmless, and inexpensive, and thus achieving the required sample size is straightforward. Although TC mainly occurs in toy- to small-breed dogs, typical laboratory dogs are medium-breed dogs such as Beagles. Rabbits and small-breed dogs have a comparable weight and tracheal RF, making rabbits a suitable model of small-breed dogs with TC.

After creating the RM, the tracheal rings in the cervical area were weakened and visualized, particularly during inspiration. These findings were observed in all three RMs. When preparing an RM, care must be taken not to penetrate the submucosal layer underneath the cartilage rings.

In comparing the tracheal RF of rabbits weighing 3.79 ± 0.11 kg and dogs weighing 3.09 ± 0.12, no significant difference was identified. Along with the tracheal RF, the clinical signs of the RMs resembled those of dogs with TC, accompanied by distinctive goose-honking coughs. Goose-honking coughs are caused by negative pressure in the collapsed region during inspiration. In addition, radiographic examination, gross examination, and histological findings of the RM were comparable with those of dogs with TC, as described in previous studies [18]. In a study by Weisse et al., radiographic, endoscopic, and postmortem findings of TC in dogs revealed a ventral deviation of the trachealis muscle, a flaccid trachealis muscle compromising tracheal lumen patency, and flattened tracheal cartilage rings [18]. Similarly, in our study, respiratory failure of grade 2 to 3 was observed with TC, alongside a dorsoventrally collapsed trachea, and biopsy results revealed a decreased tracheal lumen in the RM.

When selecting a SEMS for the treatment of TC, the determination of an adequate RF is critical. If the RF is too high, excessive pressure exerted on the trachea may damage the tracheal walls and result in pressure necrosis or wall perforation [4]. In our study, the tracheal RF values in rabbits weighing 3.79 ± 0.11 kg was 0.41 ± 0.36 kgf, and those of dogs weighing 3.09 ± 0.12 kg was 0.43 ± 0.032 kgf. However, according to our previous study, the mean RF values of a D-type stent with a wire diameter of 0.008 inches (D8) was 1.295 ± 0.046 kgf [9]. Therefore, the RF with respect to the D8 SEMS is approximately 3–4 times greater than that of the trachea. Hence, the RF of the D8 SEMS may be excessive for use in the trachea of toy- to small-breed dogs.

SEMS are commonly used in palliative therapies for malignant airway obstruction [1]. Therefore, SEMS are mostly applied to adult humans. The weight distribution in adult humans is significantly less heterogeneous than that in adult dogs. This study identified a strong positive correlation between RF and weight in five rabbits and five dogs. Further studies are required to verify this observation regarding how the correlation between bodyweight and the RF in dogs can affect the pathological condition or stent suitability. Although TC occurs mostly in toy- to small-breed dogs, a few cases of TC in large-breed dogs have been previously reported [15, 17]. In addition, even among toy- to small-breed dogs, the variation in weight is fairly large. Therefore, further studies are required to ensure a SEMS with an adequate RF for the treatment of TC depending on the animal’s bodyweight, with the exception of brachycephalic dogs. For further study using a RM of TC, it is necessary to identify specific lesions of the disease. For example, canine TC occurs in the cervical, thoracic, or whole trachea, and the four stages are graded depending on the degree of TC [6]. Accurate identification of lesion location and extent may help evaluate stent suitability.

In this study, the validity of an RM was determined by comparing the mechanical properties of tracheal samples from rabbits and dogs and evaluating the clinical, radiographic, and histological characteristics of the RM. This RM of TC can be used to develop an optimal SEMS for dogs with TC. The results from this study may provide a reference for the RF with respect to the trachea, thus narrowing the spectrum of options available to develop an optimal SEMS.

CONFLICTS OF INTEREST

All authors have no potential conflicts of interest.