Predictors of Occult Nodal Involvement in Clinically Node Negative Primary Parotid Carcinoma: An Updated NCDB Analysis
J. Lorenz,Y. Liu,S. Goyal,J.S. Remick,S.F. Rudra,S. Hanasoge,D.S. Yu,N.C. Schmitt,C. Steuer,K.R. Magliocca,W.A. Stokes
DOI: https://doi.org/10.1016/j.ijrobp.2024.01.053
IF: 8.013
2024-04-01
International Journal of Radiation Oncology*Biology*Physics
Abstract:Purpose/Objective(s) Primary Parotid Carcinomas (PPC) comprise a wide array of histologies, each with varying propensity for local invasion and cervical nodal involvement. Upfront surgical excision of the primary tumor constitutes the standard of care for PPC, but there exists little consensus regarding management of the clinically negative neck (cN0). We conducted a clinicopathologic correlative analysis of the National Cancer Database (NCDB) to identify predictors of occult nodal involvement in PPC to guide elective neck management. Materials/Methods The NCDB was queried for adults diagnosed from 2010-2020 with cN0M0 PPC who underwent adequate surgical resection (defined as subtotal to radical parotidectomy with >/=18 nodes examined) without pre-surgical treatment. Pathologic T-classification, tumor size, histology, grade, lymphovascular space invasion (LVSI), gender, and age at diagnosis were specified a priori as variables of interest. The primary outcome, occult nodal positivity (i.e. cN0 but pN+), was assessed by univariable (UVA) and multivariable (MVA) logistic regression, depicted as odds ratios (OR) and corresponding 95% confidence intervals (95%CI) with a prespecified threshold for statistical significance of p<0.05. Recursive partitioning analysis (RPA) was then performed to hierarchically categorize patients according to occult nodal disease risk. Results From an initial 51,663 patients with PPC 1,899 met inclusion criteria and were included for analysis, of whom 1,047 (55.1%) had occult nodal involvement. By UVA, the incidence of occult nodal positivity was significantly higher with male sex; age 51 and older; high grade; pT2 status or greater; histology of adenocarcinoma, salivary duct carcinoma, or carcinoma ex-PA; tumor size >2cm; and LVSI. On MVA, tumor size, age, and gender were no longer significantly associated with occult nodal involvement; however, LVSI (OR 8.3, 95%CI 6.3-10.8), carcinoma ex-pleomorphic adenoma (OR 3.6, 95%CI 1.7-7.7), adenocarcinoma (OR 3.5, 95%CI 2.2-5.5), salivary duct carcinoma (OR 3.62, 95%CI 2.2-6.1), pT3 (OR 1.9, 95%CI 1.1-3.3), pT4 (OR 3.1, 95%CI 1.8-5.4), and high grade (OR 2.7, 95%CI 2.0-3.5) remained significant. On RPA evaluating LVSI, histology, pT, and grade, LVSI was the strongest predictor of occult nodal involvement (87% if LVSI+ vs 32% if LVSI–). Conclusion LVSI appears to supersede tumor histology, grade, size, and pT status as the strongest risk factor for occult nodal involvement in PPC. To our knowledge, this is the largest series to evaluate this pathologic feature. Further research into this heretofore underappreciated risk factor and its implications for management of the cN0 neck is warranted.
oncology,radiology, nuclear medicine & medical imaging