Pharyngocutaneous Fistulas in Advanced Cancer: Closure With
Musculocutaneous or Muscle Flaps
Riley Rees, MD, Asher Gary, MD, FL Bruce Shack, MD, and Perry F. Harris, DDS, MD, Nashville, Tennessee
Over a century has passed since Billroth reported the first pharyngocutaneous fistula after a partial laryngectomy for head and neck cancer [I]. This troublesome complication of head and neck opera- tions continues to prolong hospital stay, delay ra- diotherapy, and compromise postoperative nutri- tion. Despite improvements in anesthesia, operative technique, and nutrition, pharyngocutaneous fistu- las occur in 4 to 38 percent of patients [2-51. Many investigators have suggested that infections, previ- ous radiotherapy, previous tracheostomy, or accom- panying radical neck dissection increase the inci- dence of these fistulas [3,4,6,7]. In the past, when fistulas did not close spontaneously, surgery was often unsuccessful [5,8,9].
The purpose of this study is to clearly demon- strate the use of musculocutaneous and muscle flaps for closure of pharyngocutaneous fistulas, even in cases of advanced cancer. Herein, we report our successful operative experience with these flaps for closing pharyngocutaneous fistulas in patients with extensive disease, previous radiotherapy, and, in some cases, involved microscopic operative mar- gins.
Material and Methods
At the Nashville Veterans Administration Medical Center, a review of over 200 operations for intraoral can- cer during a 6 year period revealed a postoperative fistula rate of 6 percent (19 patients). In these patients, a postop- erative pharyngocutaneous fistula developed which re- quired operative treatment. Sixteen fistulas were closed with muscle or musculocutaneous flaps, and three fistulas were closed primarily with local tissues and were excluded from this study. The patients were elderly (mean age 59 f 7 years) and had advanced disease. Tumor resection was palliative in eight patients who had involved operative margins as noted on permanent histologic sections. Pri- mary resections included laryngectomy in 14 patients, radical neck dissection in 13 patients, mandibulectomy in
From the Department of Plastic Surgery, Vanderbilt University, The Nash- ville Veterans Administration Medical Center and Baptist Medical Center, Nashville, Tennessee. This work was supported in part by a grant from the Veterans Administration, Nashville. Tennessee.
Requests for reprints shouM be addressed to Riley Rees, MD, Depart- ment of Plastic Surgery, S-2221 Medical Center North, Vanderbilt Universi- ty, School of Medicine, Nashville, Tennessee 37232.
Presented at the 33rd Annual Meeting of the Society of Head and Neck Surgeons, London, England. April 25-30, 1987.
Volume 154, October 1987
6 patients, and partial glossectomy in 3 patients. Parallel neck incisions were utilized for all resections. Fourteen patients were treated with radiotherapy (greater than 5,000 rads) for their head and neck cancers. Two patients had preoperative radiotherapy, and in one patient, thera- peutic radiotherapy failed. Five patients received previ- ous radiotherapy for other head and neck cancers after ablative operation, and seven patients received radiother- apy after fistula closure. All radiation fields included both the entire neck and the primary lesion site. Two patients received chemotherapy.
Prior to operative closure with myocutaneous flaps, all fistulas were treated in a standard way; that is, with early local wound care, pressure dressings, and antibiotics. Ag- gressive nutritional support of more than 5,000 calories per day was instituted immediately when fistulas were encountered.
Once the muscle or musculocutaneous flap was select- ed, the fistula was approached through either the inferior or superior limb of the McFee incision. We thereby avoid- ed additional scars in infected or irradiated wounds. Sub- platysmal skin flaps were elevated and the fistula debrid- ed into the hypopharynx with thorough resection of all necrotic tissue. The muscle or musculocutaneous flaps were then tunneled into the operative field overlying the fistula site, and the skin defect was closed with meshed skin grafts or local skin. Suction drains were strategically placed, and nothing was given by mouth for 10 days.
We were able to identify factors critical to the closure of pharyngocutaneous fistulas in patients with advanced disease. Because patients had exten- sive disease or previous radiotherapy, our treatment was directed at single-stage reconstruction to avoid further complications. Seventeen musculocutane- ous or muscle flaps were used to close 16 pharyngo- cutaneous fistulas in 16 patients. In the six patients in whom it was recorded, the average fistula size was 2 f 1.3 cm. All fistulas occurred by 9 days after operation and required an average of 39 f 10 days of treatment before they were operatively closed.
Flap selection for fistula closure was determined by the location of the fistula in the neck and catego- rized into four groups for analysis (Figure 1). Type I fistulas were peristomal (one patient), type II fistu- las were within the lower McFee incision (three
Rem et al
-e- TYPE III
T$E% Figure 1. Classiflcatkm of pharyngocutaneous fistuias based on anatomic iocati~. Type I fistuias were peristomai( top left), type ii fistuias were within the inferior McFae incision (top right), tyfm lb fistuias were in the midneck (bottom left), and type IV fistuias were in the superior McFea inciskm (bottom, right).
patients), type III fistulas occurred in the midneck (five patients), and type IV fistulas occurred within the upper McFee incision (seven patients). Presum- ably, type IV fistulas were the most common be- cause of the weakness of a trifurcated mucosal clo- sure in the superior aspect of the pharyngeal closure.
The flaps used in this series were Pectoralis mus- culocutaneous flaps (seven patients), pectoralis ma- jor muscle flaps (four patients), sternocleidomas- toid flaps (four patients), and latissimus dorsi flaps (two patients) were used in this series. The sterno- cleidomastoid muscle flap was useful only with type III and IV fistulas because the cephalic-based mus- cle could be easily rotated into that area. The latissi- mus dorsi flap was used in patients with extensive resections or when previous radiotherapy was given. The large skin defects were closed primarily when this flap was used. Pectoralis major muscle and musculocutaneous flaps were effective for all types of fistulas and were used extensively with type I and II fistulas. In our most recent patients, they were used only as muscle flaps with skin grafts (Figure 2) or local skin advancement. A latissimus dorsi mus- culocutaneous flap was reserved for patients in whom additional bulk and skin were needed.
Our success rate with fistula closure was 88 per-
Figure 2. Closure with a pectoralis major muscle flap and skin graft. A, the fistuia defect. B, generous resection of necrotic tissue excised from the pharynx. C, inciston for the pectoralis muscle flap around the tunnel into the neck. D, the skin graft in place ciosing tfw type I fistuia 7 year after operation.
The American Journal ot Surgery
Phaiyngocutaneous Fistulas in Advanced Cancer
cent using muscle and musculocutaneous flaps. Al- though many operative debridements (7 patients) and multiple skin grafts (11 patients) were required to close the fistulas, only 1 patient died with persis- tent fistula who also had a recurrence present. One patient required a second pectoralis major musculo- cutaneous flap to close the fistula. There were no carotid ruptures in this series because of good soft tissue coverage. Recurrent fistulas did not devlop in any patient after closure with musculocutaneous or muscle flaps. The mean survival rate was a surpris- ing 50 percent at 2 years and follow-up was 100 percent.
In this study, we have analyzed the factors neces- sary for successful use of musculocutaneous and muscle flaps for closure of pharyngocutaneous fistu- las, which is a concept first reported by Vieta et al [IO]. Our success rate of 88 percent in 16 patients reflects the extreme durability and versatility of these flaps when used for head and neck reconstruc- tion. These results were achieved despite advanced disease, radiation-damaged tissues, involved micro- scopic margins, and extensive resection. The deci- sion as to which muscle to use in closure of the fistulas was based on fistula location and availabili- ty of regional muscle groups-not prior radiothera- py, fistula size, or disease stage.
Prior to the advent of muscle flaps, traditional treatment of pharyngocutaneous fistulas had forced surgeons to carefully examine factors such as previ- ous radiotherapy, availability of local skin flaps, and disease stage [3-5,8,9]. Musculocutaneous and mus- cle flaps avoid many of these problems because they are vascularized by a single dominant pedicle and are generally outside the irradiated field. With the use of such flaps, which have their own blood supply and muscle bulk, operative closure is accomplished and patients who have a poor prognosis do not spend their remaining days undergoing multistaged reconstruction. Our ability to achieve fistula closure within a mean of 37 days is better than the 90 days using more traditional types of reconstruction as previously reported [5,9].
Continuation of radiotherapy and chemotherapy are positively influenced by rapid closure of pharyn- gocutaneous fistulas. In this study, we were able to administer radiotherapy earlier. Closure of fistulas using this technique will decrease morbidity and may enhance survival by reducing local recurrence. In addition, postoperative delays caused by fistulas
prevent placement of a voice prosthesis or training with an electrolarynx. Certainly, early closure of a pharyngocutaneous fistula will enhance the pa- tients quality of life.
This study has focused on a solid concept in the management of pharyngocutaneous fistulas using musculocutaneous or muscle flaps for patients with extensive head and neck cancer. Its obvious advan- tage is single-stage reconstruction with rapid wound healing.
Two hundred cases of head and neck cancer were reviewed and 16 pharyngocutaneous fistulas identi- fied, for an incidence of 6 percent. The fistulas were closed with pectoralis major muscle flaps in four patients, pectoralis musculocutaneous flaps in sev- en patients, sternocleidomastoid muscle flaps in four patients, and latissimus dorsi flaps in two pa- tients. Four types of fistulas were identified, and flap selection was determined by fistula location. Successful closure was obtained in 15 patients (88 percent), although one patient died from recurrence with a persistent fistula.
Acknowledgment: We thank J.B. Lynch, MD, Arnold Malcolm, MD, and Robert Ossoff, MD of Vanderbilt Uni- versity for their help and advice, and Tom Ebers for his technical support.
1. Thorwald J. Imperial tragedy. In: The triumph of surgery. Winston R, Winston C, trans. New York: Pantheon Books, 1960.
2. Cummings CW, Chung CK, Johnson J, et al. Complications of laryngectomy and neck dissection following planned pre- operative radiotherapy. Ann Dtoi Rhino1 Laryngol 1977; 66: 745-50.
3. Soil AM, Sessions DG, Ogura JH. Complications following surgery for cancer of the larynx and hypopharynx. Cancer 7977; 39: 624-37.
4. Lavelle RT, Man AR. The etiology of postiatyngectomy phar- yngocutaneous fistuiae. J Laryngol Dtoi 7972; 66: 765-93.
5. Dedo DD, Aionso WA, Ogura JH. incidence, predisposing factors, and outcome of pharyngocutaneous fistulas com- plicating head and neck cancer surgery. Ann Otolaryngol 1975; 64: 833-40.
6. Briant TDR. Pharyngeai fistula and wound infection following laryngectomy. Laryngoscope 7975; 85: 829-34.
7. Horgan EC, Dedo HH. Prevention of major and minor fistulae after laryngectomy. Laryngoscope 1979; 88: 250-60.
8. Chandler JR. Treatment of pharyngeai fistuias after total laryn- gectomy. Laryngoscope 7979; 89: 671-3.
9. Ballinger C. Classification of pharyngostomes: a guide to closure. Plast Reconstr Surg 1977; 47: 54-60.
70. Vieta CJ, Quintana RA, Obregon MA. Plastic repair of pharyn- geal fistuias. Arch Otolaryngol 7 968; 87: 78-8 7.
Volume 154, October 1907 383