Background. The prognostic significance of the status of surgical marg
ins remains unclear in patients with cancer of the larynx. Methods. To
further evaluate this issue, a cohort of 159 surgically treated patie
nts with advanced (stage III, IV) squamous carcinoma of the larynx who
were part of a prospective multi-institutional clinical trial (VA Lar
yngeal Cancer Study Group) were studied. Results. All patients underwe
nt total laryngectomy (n 144) or supraglottic laryngectomy (n = 15) fo
llowed by radiotherapy. Surgical margins were assessed intraoperativel
y and reported on final pathology as either clear (n = 134), tumor wit
hin 5-10 mm of the surgical margin (n = 6), less than 5 mm from the su
rgical margin (n = 8), or involved (n = 11). Postoperative complicatio
ns, especially fistula rates, were significantly associated with margi
n status (p = .04 and p = .004, respectively). No significant differen
ces were detected in patterns of recurrence (p = .584) or time to recu
rrence (p = .712) according to margin status. Patients with close or i
nvolved margins received significantly higher doses of radiation (p =
.0017), as prescribed in the protocol. Patient survival percentages we
re 50% for clear margins, 57% for close margins, and 27% for involved
margins (p = .286). When grouped by margins >5 mm or <5 mm, patients w
ith wider margins tended to have prolonged disease-free survival (p =
.1413) and overall survival (p = .1879). Conclusions, Despite the smal
l number of patients with histologically proven close margins (16%) in
this trial, the data suggest that high-dose postoperative radiation d
oes not eliminate the benefit of obtaining widely clear surgical margi
ns. Better methods of assessing and obtaining wide surgical margins ma
y translate into added survival benefit and lower complication rates.
(C) 1996 John Wiley & Sons, Inc.