[Pharyngo-cutaneous fistula after total laryngectomy]

The charts of 319 consecutive patients who underwent total laryngectomy at the Cancer Institute Hospital from 1971 to 1994 were reviewed in order to clarify the relationship between pharyngo-cutaneous fistula formation and age, the dose of pre-operative radiation and radical neck dissection, as well as the need for subsequent surgical repair. The patients did not need to undergo reconstruction by flaps at the time of laryngectomy. Radiation sources were limited to X ray radiotherapy and Cobalt 60. Of the 319 patients 204 (63.9%) underwent neck dissection. Both radical neck dissection and modified radical neck dissection were classified as neck dissection. The chi-square test was used to construct a table of the three parameters age, dose of radiation and neck dissection. With respect to age, the incidence of fistula formation was 13.4% (16 patients of 119) in patients at the age of 59 and below, 5.9% (7/118) in those from 60 to 69, and 8.5% (7/82) in those at 70 years and above. Our analysis reveals that the age at the time of surgery is not a predisposing factor for fistula formation in the three age groups (59 and below, between 60 and 69, and 70 and above). Similarly the need for subsequent surgical repair is also not age-related. With respect to radiation, the incidence of fistula formation was 8.0% (4/50) for patients who received radiotherapy less than 20 Gy, 6.3% (2/32) in those who received between 20 and 40 Gy, 2.6% (2/77) in those who received between 40 and 60 Gy, 13.2% (20/152) in those who received between 60 and 80 Gy and 25.0% (2/8) in those who received over 80 Gy. When the preoperative dose of radiation was divided into three classes, that is, less than 40 Gy, 40 to 60 Gy and over 60 Gy, we observed that the incidence of fistula formation increased significantly in the patients who received over 60 Gy. Surgical repair was also indicated more frequently for those patients who received over 60 Gy than for those who received less than 60 Gy. With respect to neck dissection, the incidence of fistula formation was 12.2% (14/115) for the patients who did not undergo neck dissection or those who underwent only lymphadenectomy, 7.8% (9/115) for the patients who underwent unilateral neck dissection, and 7.9% (7/89) for those who underwent bilateral neck dissection. These data reveal that neck dissection, whether unilateral or bilateral, dose not increase the incidence of fistula formation, nor the need for subsequent surgical repair. Fistulae were present in 30 patients (9.4%) for 24 years, and 14 of these 30 patients did not need subsequent surgery. In these 30 patients with fistulae, we did not find patients with systemic disease such as diabetes mellitus prior to the surgery. When the period of 24 years was divided into 4 periods, the incidence of fistula formation was 19.0% (from 1971 to 1976), 6.9% (from 1977 to 1982), 10.3% (from 1983 to 1989) and 2.6% (from 1989 to 1994), that of the latest period was the lowest with gradual improvement. The average dose of preoperative radiation was 57.7 Gy (from 1971 to 1976), 50.8 Gy (from 1977 to 1982), 39.6 Gy (from 1982 to 1988) and 45.7 Gy (from 1989 to 1994) and reduction in dose of radiation seemed to be one of the reasons for the lower frequency of fistula. Several surgeons performed the operations for different patients, but the procedure of laryngectomy was recently directed by an experienced surgeon. The study also indicates that the risk of fistula formation is reduced not only by the dose of radiation but also by improved surgical skill.