Squamous cell carcinoma of the supraglottic larynx without clinically detectable lymph node metastases: problem of local relapse and influence of overall treatment time.

From January 1965 until December 1979, 203 patients with squamous cell carcinoma of the supraglottic larynx were treated with curative intent. The mean follow-up time was 10 years. The policy was to try to aim for cure by radiation therapy (RT) only, reserving surgery (S) for radiation therapy failures. For 193 patients the treatment consisted of a first series of radiation therapy to a total dose of 40 Gy; if a good response to radiation therapy was obtained, the treatment was continued to a full course of 60-70 Gy (RT-I, n = 132). Patients with tumors considered to have responded poorly to the first series of radiation therapy but who refused surgery or were found medically unfit for operation, were also carried to a full dose of 60-70 Gy (RT-II, n = 33). Surgery was performed in 33 cases; 23 patients had a laryngectomy because of a poor response to radiation therapy and 10 were treated with surgery upfront because of severe respiratory distress. This paper focuses on the local control and survival in the defined treatment groups. In summary, with advancing T-stage a lower survival and higher local relapse rate was found; that is, a 5-year relapse-free survival (RFS) of 53% and corrected survival (CS) of 83% for T2 tumors vs 39% (RFS) and 52% (CS) for T4 tumors. Age more than 60 was associated with a 2.2 times higher risk of dying due to laryngeal cancer. A lower relapse-free survival (T3,4: 43% vs 61%) but a comparable corrected survival (T3,4: 64% vs 69%) for RT-I patients compared to the surgery treated patients was found, due to salvage of the radiation therapy failures. Although the relapse-free survival of RT-I and RT-II was similar (43% vs 38%), the corrected survival for the RT-II patients was worse (44% vs 69%). No influence of dose (Gy) per se on the local relapse rate was observed; however, a positive association between local relapse rate with overall treatment time was found. Death from intercurrent disease was almost twice as high as might have been expected for the normal Dutch population. More than half of the patients who died of intercurrent disease developed a second primary tumor.