Surgical Factors Affecting Regionally Metastatic Cutaneous Squamous Cell Carcinoma to the Parotid Gland

Presentation: P782
Topic: Skin Cancers
Type: Poster
Date:
Session:
Authors: Vishwanatha M Rao1; Allen L Feng, MD2; Daniel G Deschler, MD1; Derrick T Lin, MD1; Jeremy D Richmon, MD1; Mark A Varvares, MD1; Kevin S Emerick, MD1
Institution(s): 1Department of Otolaryngology – Head and Neck Surgery, Massachusetts Eye and Ear; 2Department of Otolaryngology – Head and Neck Surgery, University of Michigan


Introduction:

Survival from regionally metastatic cutaneous squamous cell carcinoma (cSCC) is generally considered poor.  Regional metastasis to the parotid gland from cSCC is a common location for head and neck cSCC metastasis. Understanding the impact of surgical treatment on this disease is important as treatment paradigms shift with the introduction and evolution of immune checkpoint inhibitors. Herein, we aim to describe surgical features as well as the clinical and pathologic tumor features that impact disease control and survival.


Methods:

A retrospective series of patients undergoing parotidectomy and neck dissection for regionally metastatic cSCC to the parotid between January 2008 to December 2018 was identified at a single tertiary academic center. Patient demographics, tumor information, and surgical factors were gathered including extent of parotidectomy, nodal yield, and margin status. Overall survival (OS), disease-specific survival (DSS), and disease-free survival (DFS) were collected over a 3-year period. Outcomes were represented using Kaplan-Meier curves. Multivariate analysis was completed using Cox proportional hazard models with backwards elimination (removal p<0.10). All statistics were performed using STATA 13.0 (StataCorp, College Station, TX).


Results:

147 patients were included in the study. The mean age was 74.9 years (range, 43-94 years),  87.8% were male. On average, patients had 26.8 nodes (SD, 15.8 nodes) resected from their lymphadenectomy. An average of 4.0 nodes (SD, 4.3 nodes) were removed from the parotidectomy specimen with no significant difference seen between superficial or total parotidectomies with or without nerve sacrifice (p=0.700). The median number of positive nodes was one (range, 1-16 nodes). Key features include: 55.8% had perineural invasion (PNI), 31.3% had lymphovascular invasion (LVI), and 65.3% had extracapsular extension. Positive margins were seen in the parotid specimen in 29.1% of cases, while 35.5% had close margins and 35.5% had negative margins. 14.3% of patients were immunocompromised, 6.8% related to solid organ transplant and 7.5% related to lymphoproliferative disorders (7.5%). 53.5% of patients received adjuvant radiation alone, 27.8% received adjuvant chemoradiation, and 18.8% received no adjuvant therapy. 32.7% of patients experienced recurrence, of which 10.2% had local recurrences, 16.3% had regional recurrence, and 6.1% had distant metastases. At 3 years, OS was 69.3%, DSS was 81.8% and DFS was 61.1%. On multivariate analysis, immune status (OS HR3.30, p=0.009; DSS HR5.51, p=0.002; DFS HR2.31, p=0.017) and LVI (OS HR2.35, p=0.028; DSS HR4.85, p=0.002; DFS HR2.72, p=0.001) were the only factors predictive for OS, DSS, and DFS. Margin status (OS HR2.13, p=0.002; DSS HR2.31, p=0.009), and >18 resected nodes (OS HR0.26, p=0.001; DSS HR0.33, p=0.019) were predictive of OS and DSS. Age (HR1.06, p=0.012) and >1 positive node (HR2.66, p=0.012) was predictive of OS, while adjuvant therapy (HR0.44, p=0.015) was predictive of DSS.


Conclusions:

In patients with metastatic cSCC to the parotid, immunocompromise and LVI are associated with worse OS, DSS, and DFS, while positive margins are associated with worse OS and DSS. In these patients, those who have >18 nodes resected have improved OS and DSS, whereas the extent of parotidectomy has no effect on OS, DSS, DFS, or nodal yield.