Tracheal resection and reconstruction have undergone considerable changes in the last 60 years. Before the 1950s, tracheal resections were limited to less than 2 cm of trachea. In 1957, however, Barclay et al1 demonstrated that a segment larger than 2 cm could be removed and a primary anastomosis could be used for repair. For the next several decades, many investigators showed that complex tracheal resections could be performed with up to 50% of the adult trachea resected.2–4 Tracheal replacement for long-segment (>50%) removal, however, still has significant limitations.5 Our patient illustrates a technique of reconstruction using AlloDerm (Lifecell, Branchburg, NJ USA) as the conduit for reconstruction of a long-segment of tracheal defect.
The patient is a 38-year-old white woman who underwent a thyroid lobectomy in 2011 for a benign goiter. Two months later, she developed dysphagia and burning in the esophagus and underwent a laryngoscopy, which showed findings concerning for gastroesophageal reflux disease A barium esophagram showed a delay in the transit of the barium through the cervical esophagus and a computed tomography (CT) scan of the neck (Video 1, http://links.lww.com/INNOV/A115) showed an esophageal/tracheal, mass and she was referred to thoracic surgery.
The patient had yet to be seen by thoracic surgery in her region when she presented to our emergency department with complaints of a cough with eating in early 2012. We reviewed her imaging and then preformed a bronchoscopy and esophagogastroduodenoscopy (EGD), which showed an abnormal esophagus with ulcerations and a mass-like lesion 2 cm distal to the upper esophageal sphincter that extended for 4 cm with two areas concerning for a fistula. Starting five rings below the vocal cords, the trachea was abnormal in appearance for approximately 4 cm. Biopsies came back atypical cells but not diagnostic for malignancy.
Two days later, she underwent an EGD with biopsy and a percutaneous endoscopic gastrostomy placement. Biopsies again came back atypical but not malignant. A positron emission tomography (PET) scan then showed significant uptake in the paratracheal/esophageal lesion; evidence of aspiration pneumonitis and the tracheoesophageal (TE) fistula was clearly evident (Fig. 1). The scan also showed right vocal cord paralysis.
Because of the aspiration pneumonitis seen on the PET scan, a bronchoscopy with a tracheal stent was performed, which showed the fistula to be approximately 2 cm in size and the abnormal tissue to be 2 cm proximal and distal to the fistula. The stent was placed starting from 3 cm distal to the vocal cords down to 1 cm proximal to the carina. Biopsy was again nondiagnostic for malignancy. The following day, she underwent a mediastinoscopy; no obvious mass was seen, just dense tissue. Biopsies from this procedure also came back atypical but nondiagnostic for malignancy. She was sent home and recovered from her pneumonia. An outpatient repeat CT results showed the stent to be in place, disease only in the area of the esophagus/trachea, and no concern for metastatic disease (Video 2, http://links.lww.com/INNOV/A114).
Approximately 1 month after presenting to our ED, she underwent a repeat rigid bronchoscopy with stent removal and EGD, which showed the esophagus to have abnormal tissue beginning at 18 cm from the incisors with the upper esophageal sphincter being at 15 cm. The fistula began at 19 cm and extended to 21 cm. There was abnormal mucosa down to approximately 24 cm from the incisors, and the rest of the esophagus was normal. On the tracheal side, the trachea was normal from the vocal cords for approximately 5 cm. There was a 2 cm fistula with 1 to 2 cm of abnormal trachea both proximal and distal to this area. The distal 3 cm of tracheal appeared normal. Again, biopsies at this time again came back atypical but not diagnostic for malignancy. All biopsy material was sent to an outside institution for an additional opinion and was again nondiagnostic for malignancy.
One month later, we performed a diagnostic right video-assisted thoracoscopic surgery and a right thoracotomy using a double-lumen endotracheal tube, during which we obtained a diagnosis of classical Hodgkin lymphoma. To get adequate tissue for diagnosis, the fistula had to be taken down. As this was done, portions of the trachea and esophagus liquefied, as did the right recurrent laryngeal nerve.
We were left with a large defect in both the esophagus and the trachea and proceeded to take town the rest of the TE fistula. Once this was completed, 7 cm of the esophagus had been removed. There was also an 8-cm portion of the posterior trachea and a 5-cm section of the right trachea wall missing (Fig. 2). The anterior wall and the left tracheal wall were still intact.
We resected the remainder of the esophagus, stapled off the stomach, and closed the esophageal hiatus. We brought out a cervical esophagostomy, which was quite difficult because the tissue was abnormal and would not mobilize well. The tracheal defect was too large for a patch repair, so we reconstructed her trachea with a tube of AlloDerm. The tube was 6 cm in circumference × 10 cm in length. We sutured the AlloDerm into a tube with 3–0 vicryl. Next, we connected the tube to the distal normal trachea at the level of the carina and then to the proximal trachea in the neck with 3–0 vicryl.
A latissimus dorsi muscle flap was used to cover the repair in the chest and a sternocleidomastoid muscle flap was used to cover the repair in the neck. The mediastinum was widely drained, and she went to the intensive care unit intubated. She underwent two separate bronchoscopies in the immediate postoperative period; the second one showed a defect in the proximal repair with a cavity in the paratracheal space.
The patient underwent a re-exploration that showed that the graft had slipped off the trachea. We subsequently repositioned the graft into the neck around normal tracheal and resutured it in place with 3–0 vicryl. Attempts to extubate the patient failed, and she underwent placement of a tracheostomy tube through the AlloDerm where it overlaid the normal anterior tracheal wall. Her postoperative course was challenging and required nine bronchoscopies, with two of these being dilations (Fig. 3A). For this hospitalization, her length of stay was 67 days.
The patient was staged for her lymphoma and found to have limited disease. She was started on standard chemotherapy for lymphoma, was readmitted multiple times for pneumonia, and underwent multiple bronchoscopies (Figs. 3B, C). She completed her chemotherapy, and her follow-up CT and PET scans showed no evidence of the disease. She desired reconstruction, was evaluated by otolaryngology surgery, and was found to have bilateral vocal cord paralysis. They recommended a laryngectomy.
One year after tracheal reconstruction, she underwent a laryngectomy with colon interposition esophageal reconstruction. Though still difficult, this postoperative course was much easier: she only required four bronchoscopies and her length of stay was just 25 days. Soon after, she underwent unsuccessful placement of a TE speech prosthesis.
In the last 3 years, multiple follow-up CT scan results have shown no evidence of recurrence. She has, however, had issues with stricture of the proximal and distal colonic anastomosis, as well as the tracheal/AlloDerm conduit. In total, she has required 10 additional bronchoscopies (Fig. 3D) and EGDs with dilation since her last surgical reconstruction in 2013. Currently, 3½ years out, she is eating well and living her normal life.
Management of a long-segment defect is still challenging for the thoracic surgeon, because the trachea is a complex conduit. In 1950, Belsey6 stated that the ideal conduit for tracheal replacement would be rigid in the anterior to posterior axis while still being flexible and mobile in the longitudinal axis and would ideally have an inner surface of ciliated respiratory epithelium.
Numerous techniques to replace the trachea have been described, including rigid scaffolding with different structures such as rib cartilage, silicone, mesh, and steel.7 Stents have also been used but are not typical in long-segment defects. These replacement options, however, have flaws that have limited their success. In our case, we did not anticipate needing these reconstruction options, because we were expecting to be able to do a patch repair of the trachea. Once in the operating room, it became apparent that a patch was not possible because of the large defect.
A newer technique, tissue re-engineering seems to be a promising option. It involves construction of a conduit to replace the trachea with a tissue re-engineered tracheal transplant.5,7 Although tissue re-engineering is now available, it is not widely accessible and was not an option at the time of our patient's reconstruction. Moreover, because our patient is currently doing well, tracheal transplantation would not be indicated at this time.
AlloDerm is a donated human dermis that is processed to keep the matrix as a scaffold and then support regeneration of normal tissue by the patient. Replacement with an AlloDerm tube worked well in this patient. Although AlloDerm does not provide any structural support, her replacement was successful; this is likely due to her having a scaffold of cartilaginous trachea in the anterior and left lateral location to serve as the structural support for the AlloDerm conduit. She has had the expected complication of conduit stricture but seems to tolerate the dilations well. She is disease free from her lymphoma and seems to have a good quality of life.