By Author
  By Title
  By Keywords

May 1995, Volume 45, Issue 5

Original Article

Pharyngocutaneous Fistula Following Laryngectomy

Iqbal H.U daipurwala  ( Department of Otolaryngology and Cervico-facial Surgery, Dow Medical College and Civil Hospital, Karachi. )
Khalid Iqbal  ( Department of Otolaryngology and Cervico-facial Surgery, Dow Medical College and Civil Hospital, Karachi. )
M. Jalisi  ( Department of Otorhinolaryngology and Cervico-facial Surgery, Dow Medical College and Civil Hospital, Karachi. )


Sixty-eight cases of stage III and IV laryngeal cancer were analysed to detenmne the frequency and management of post- operative pharyngocutaneous fistula (PCF). The overall incidence in our series is 8.9% (6 cases), In five patients PCF healed spontaneously without any surgical intervention while one patient expired due to concomitant other medical illness. The longest time taken by the fistula to heal by conservative treatment was 7 weeks, Retaining the NC tube for a longer period (14-16 days) and adequate and appropriate antibiotic cover reduces the incidence of fistula formation (JPMA 45:130, 1995).


Pharyngocutaneous fistulais a significant cause of post­operative morbidity following total laryngectomy. The inci­dence of this serious complication varies from 14% to 70% depending upon the modality of treatment used in combina­tion with surgery1-6. This study analyses the incidence and management of pharyngocutaneous fistula after radical laryn­geal surgery at our department during the last 7 years (July, 1987-June, 1994).

Patients and Methods

During the period of 7 years (July, 1987 to June, 1994), 68 cases of biopsy proven, stage III and IV, laryngeal cancer patients were subjected to radical surgery at Civil Hospital, Karachi. In all these patients, a thorough clinical, radiological and endoscopic examination was carried out prior to radical laryngeal surgery either alone (54 cases) or followed by radiotherapy (14 cases). Radical neck dissection for nodal metastasis was also carried out along with the primary surgery in 17 patients. Pharyngeal repair was done with 4/0 viciyl suture in a ‘T’ fashion in three layers. Post-operatively all the patients were put on triple antibiotics which included a cephalosponn, gentamycin and metronidazole. Gentamycin continued for 5 days while the other two antibiotics were given for 12-16 days. We routinely put two radivac drains with negative pressure one on each side of neck during laryngeal surgery. Drains were removed one after the other between 4th and 7th post-opera­tive day, with a gap of 1-2 days in between. Post-operatively dressing was done around the neck with crepe bandage for 24-48 hours. All these maneuvers were performed to prevent collection of serumin the neck spaces. Patients were examined between 13th and l5th post- operative days for any pharyngeal leakage by giving sterile water orally. If no leakage was found the patients were then allowed to take liquids orally and NG tube was removed the next day.


There were 55 (80.9%) male and 13 (19.1%) female patients.

Table I shows the age and sex distribution. Majority (64.8%) of the patients were between the ages of 41 and 60 years. Twenty-three (34%) patients had supraglottic lesion while the glottic and transglottic group comprised of 15(22%) and 30 (44%) patients respectively.

Table II shows the topography and ‘T’ status at the time of surgery. Twenty-eight cases had T3 lesion while eight patients had T4 lesion. Up till now both UICC and AJCC system of TNM classification do not include transglottic carcinoma in their classification. Therefore, the tmnsglottic lesions (30 cases) were staged as Tx. Since these lesions were more extensive involving a vertical extension to include more than one topographic site, they were treated with radical laryngeal surgery. Two patients had recurrence after radiotherapy for initial T2 glottic lesion. These were also staged as Tx and were subjected to total laiyngectomy. Out of 68 cases, total laryngectomy was done in 53 cases while in 15 cases extended total latyngectomy was done.

Table III shows the extensions to classical total laiyngectomy. PCF occurred in 6(8.9%) cases. It healed spontaneously by conservative treatment in five patients while one patient expired due to concomitant medical illness. This patient had severe postoperative bleeding which needed exploration and ligation of the bleeders twice, on the same and 1st post-opera­tive day respectively. He also developed severe liverfailure on 3rd postoperative day. The probable cause was chronic liver disease due to hepatitis B, which was aggravated by repeated administration of hepatotoxic anaesthetic gases. This patient developed a large phaiyngocutaneous fistula. Surgical inter­vention was not possible due to his medical illness and thus gastrostomy was done for feeding. Patient succumbed to liver failure. Patients in whom PCF occurred included four male and two female cases. Three patients were above the age of 60 years, while three were between 45 and 60 years (Table IV).

PCF was greater in supraglottic carcinomata where three (13%) patients hadPCFoutof 23 cases whereas intransglottic lesions 3 (10%) out of 30. In 13 patients partial phaiyngec­ tomy was also done (Table III) and PCF occurred in 2(15.3%), while in 56 cases where standard total laryngectomy has been done, PCF occurred in 4 (7.1%) cases. NG tube was not removed in cases with PCF and all the fistulae healed spontaneously by conservative treatment in 3-7 weeks time (Table IV).


A phaiyngocutaneous fistula following total laryngec­tomy is not an uncommon occurrence. It exacethates postop­erative motbidity withprolonged hospitalization. The compli­cation may require additional operative procedures or if untreated may lead to further complications. The incidence of PCF varies worldwide depending upon the modality of treatment used in combination with surgery1-6. We prefer surgical modality either alone or followed by radiotherapy to treat carcinoma of larynx. The incidence figure in our series was 8.9%. There are multiple factors which affect the development of PCF. These include site and size of tumour, surgical technique, suture material used, post-operative care, compro­mised immune status nutrition, pre-operative radiotherapy, chemotherapy and associated surgical procedures including radical neck dissection. Hypothyroidism due to surgical ablation or radiotherapy is also an important cause of delayed wound healing with subsequent fistula formation7. These fistulae are refractory to conservative orsurgical treatmentand substitution therapy results in rapid healing. The advanced stage of the disease and supraglottic tumours extending into the phaiynx, requires more extensive resection of the phaiynx. These are associated with a higher incidence of post-operative PCF formation. Increasing clo­sure tension in the pharynx associated with the increasing width of resected tissue has a direct relation with the development of fistulae8. An adequate pharyngeal lumen can be constructed if there is enough mucosa available that can be sutured without tension over a 18 French feeding tube4. We routinely use two radivac drains with negative pressure for 4-7 days after surgery. The use of radivac drains with negative pressure keeps the skin in close apposition with the phaiyngeal wall. This prevents the collection of blood and semm under the skin and thus prevents infection with subsequent fistula formation. The use of drains with negative pressure is associated with a lower incidence of fistula formation4,9. We also routinely use crepe bandage around the neck for 24-48 hours to fulfill the same purpose. The use of prior and post-operative broad spectrum antibiotics significantly reduces the risk of fistula forma­tion10. Metronidazole has a very important role in preventing this serious complication11,12. The use of metronidazole alongwith gentamycin and cephalosporin in our series had showed reduced postoperative infections and fistula forma­tion. Controversy prevails about the time for extubation of feeding tube. There are proponents who keep the NG tube for a short time and believe that this curtails both mothidity and hospital stay13. Contrary to this some workers keep NG tube for longer duration and believe this, to be instmmental in preventing a fistula4,9. One group does not use NG tube after laryngectomy and starts oral feeding on the 3rd post-operative day with satisfactory results14. Our practice is to give sterile water orally between 13th and 15th post-operative day and look for any leakage. If no leakage is found we allow the patients to take liquids orally and NG is removed on the next day. In case of leakage tube feeding is continued for another 7-10 days and thereafter the same procedure is repeated. In five out of six cases in our series, fistulae healed completely by following this protocol. We advocate retaining the NG tube for a longer period (14-16 days) specially in situation like ours. Prolonged nasogastric intubation carries no additional risks to patients apart from prolonging the hospitalization. Radical neck dissection alongwith the primary surgery is considered to be associated with greater incidence of post- operative phaiyngocutaneous fistula15,16. However, this is contradicted by other workers4,17,18. In our series radical neck dissection was performed in 17 patients and none of them developed fistula, thus no significant correlation was found between radical neck dissection and fistula formation.


1. GaIl, A. M., Sessions, D. C. and Ogura, J. H. Complications following surgery for cancer of the larynx and hypopharynx. Cancer, 1977;39:624-31.
2. Hier, M., Black, M. J. and Lafond, G. Phaiyngocutaneous fistula after total laryngectomy: Incidence, etiology and outcome analysis. J. Otolaryngol., 1993;22(3):164-66.
3. Moses, B. L., Eiaele, D. W. and Jones, B. Radiologic assessment of the early post-operative total laryngectomy patient. Laryngoscope, 1993;103:1 157-60.
4. Gairola, A. and Bahadur, S. Pharyngocutaneous fistula following laryngectomy. Pak. J. Otolaryngol., 1991;7 :86-90.
5. Viani, L., Stell, P. M. and Dalby, J. E. Recurrence after radiotherapy for glottic carcinoma. Cancer, 1991 ;67:577-84.
6. Bresson, K., Hasmussen, H. and Rasmussen, P. Pharyngocutaneous fistula in total layrngectomy patients. J. Laryngol. Otol., 1974;88:835-40.
7. Talmi, Y. P., Finkelstein, Y. and Zohar, Y. Pharyngeal fistula in post-operative hypothyroid patients. Ann. Otol. Rhinol. Laryngol., 1989;98:267-68.
8. Weiss, L. S., Miller, R. H., Bates, M. L. et at. An animal model for pharyngocutaneous fistula Otolaryngol., Head Neck Surg., 1993;108:738-42.
9. De-Jong, P C. and Struben, W.H. Pharyngeal fistula after laryngectomy. J. Laryngol. Otol., 1970;84:897-901.
10. Violaris, N. and Bridger, M. Prophylactic antibiotics and post laryngectomy pharyngocutaneousfistula. J. Laryngol. Otol., 1990;104(3):225.28.
11. Sawyer, R. Clinical implications of metronidazole antibiotic prophylaxis in major head and neck surgical practice. Ear Nose Throat J., 1988 ;67:655-62.
12. Johansen, L. V., Overgaard, J. and Elbrond, 0. Pharyngocutaneous fistulae after laryngectomy: Influence of previous radiotherapy and prophylactic metronida­zole. Cancer, 1988;61 :673-78.
13. Boyce, S. E. and Meyers, A. D. Oral feeding aftertotal laryngectomy. Head Neck Surg., 1989; 11(3):269-73.
14. Aprigliano, F. Use of the nasogastric tube after total laxyngectomy: is it truly necessary?. Ann. Otol. Rhinol. Laryngol., 1990;99:513-14.
15. Lavelle, R. S. and Mao, B. The etiology ofpost-Iaryngectomy fistula. Laryngol. Otol., 1973;86:785-94.
16. Harris, A. and KOmray, R. R. Cost effective management ofpharyngocutaneous fistula following laryngectomy. Ostomy. Wound Management, 1993;39(8):36-44.
17. Briant, T. D. R. Spontaneous pharyngeal fistula and wound infection following laryngectomy. Laryngoacope, 1975;85:829-32.
18. Myers, E. The management of pharyngocutaneous fistula. Arch. Otolaryngol., 1977;95:10-14.

Journal of the Pakistan Medical Association has agreed to receive and publish manuscripts in accordance with the principles of the following committees: