Total laryngectomy is a curative-intent surgery for advanced laryngeal cancer. Resection to negative surgical margins is the primary goal and confers an improved prognosis.
To determine if a significant change in the rate of positive margins has occurred in recent years and to evaluate patient factors related to these changes.
Patients from the National Cancer Database (NCDB) with T3 or T4 primary hypopharynx or larynx squamous cell carcinoma undergoing curative-intent total laryngectomy (TL) with or without pharyngectomy from 2004-2016 were analyzed. The primary outcome was positive surgical margin rate. Secondary outcomes included factors associated with positive margins and overall survival.
A total of 6,916 patients that underwent TL were included in the database. The median age was 61 (IQR 54-69). Patients were predominantly male (81.0%), white (73.9%), and publicly insured (64.8% Medicare/Medicaid).
While multivariable logistic regression initially suggested that odds of positive margins decreased yearly (OR 0.94, CI95 0.89-0.99), this effect fell out after excluding 2010 data as an outlier (17.3% vs. 12.6% remaining years; OR 0.97, CI95 0.91-1.03). Nor was there a change in margin rate over time as assessed by linear regression (slope 0.02%, CI95 -0.27-0.30); Figure 1). However, the observed rate of patients presenting with cT4 disease increased by 0.8% yearly (CI95 0.5-1.2) by linear regression (54.5% in 2004 increasing to 63.6% in 2016; Figure 2). Similarly, the observed rate of patients requiring pharyngectomy in addition to total laryngectomy increased by 0.5% yearly (CI95 0.19-0.84; 19.1% in 2004 increasing to 24.8% in 2016; Figure 3). There was also increased time from diagnosis to definitive surgery [median 22 days (IQR 11-35 days) in 2004 to 36 days (IQR 20-56) in 2016; 1.0 day increase/year (CI95 0.7-1.3 days)]. More patients were treated in academic centers over time [64.6% in 2004 to 77.6% in 2016; OR 1.05/year, CI95 1.02-1.06).
On multivariable logistic regression, factors associated with positive margins included having cT4 disease (OR 1.31, CI95 1.05-1.62), poorly differentiated grade (OR 1.91, CI95 1.12-3.27), lymphovascular invasion (OR 2.1, CI95 1.70-2.63), extranodal extension (OR 2.00, CI95 1.59-2.51), and requiring pharyngectomy (OR 1.42, CI95 1.13-1.79). Treatment at a middle or high-volume academic center (OR 0.61, CI95 0.44-0.83, and OR 0.63, CI95 0.46-0.86) was associated with negative margins.
Five-year overall survival was 40.9% (CI95 39.5-42.3) and has not changed over time in multivariable Cox proportional hazards regression (HR 0.99/year CI 95 0.96–1.03; Figure 4). Factors correlated with worse overall survival included Charlson-Deyo comorbidity score³3 (HR 2.43,CI95 1.88-3.14), pathologic stage IV (HR 1.36, CI95 1.12-3.27), lymphovascular invasion (HR 1.31, CI95 1.17-1.48), extranodal extension (HR 1.62, CI95 1.43-1.85), positive margins (HR 1.44, CI95 1.24-1.66), and delay in post-operative radiation (HR 1.22, CI95 1.04-1.43, Figure 4).
Positive margin rates and overall survival for total laryngectomy have not changed, although patients are presenting with more advanced disease and increased time to surgery. Treatment in academic centers increased over time and treatment in middle and high-volume centers was associated with decreased risk of positive surgical margins.