Temporal bone squamous cell carcinoma: Analyzing prognosis with univariate and multivariate models


  • This study was partly supported by grant No. 60A07-1341/12 (G. Marioni) from the University of Padova, Italy. The authors have no other funding, financial relationships, or conflicts of interest to disclose.



Temporal bone squamous cell carcinoma (SCC) is an uncommon malignancy accounting for less than 0.2% of head and neck cancers. Despite advances in its early diagnosis, skull base microsurgery, radiotherapy, and integrated treatments, prognosis in advanced SCCs remains dismal. The present study aimed to analyze the clinicopathological variables potentially influencing outcome in a series of temporal bone SCCs.

Study Design

The prognosis of 41 patients with temporal bone SCC was assessed retrospectively using univariate and multivariate statistical approaches.

Patients and Methods

Twenty-two women and 19 men consecutively operated for primary temporal bone SCC with a curative intent at a tertiary referral center between 1980 and 2008.


On univariate analysis, cT stage correlated with disease-free survival in months (DFS) (P = 0.037), and pT stage correlated with recurrence rate (P = 0.038), DFS (P = 0.013), and disease-specific survival (DSS) (P = 0.025). Lymph node status (cN0 or pN0 vs. pN+) was associated with DFS (P = 0.025). SCC grading correlated significantly with recurrence rate (P = 0.005), DFS (P = 0.004), and DSS (P = 0.0036). Dura mater involvement was significantly associated with a higher recurrence rate (P = 0.001), a shorter DFS (P = 0.00001), and a lower DSS (P = 0.0001). On multivariate analysis, only dura mater involvement (P = 0.001) and N status (P = 0.012) remained independently prognostic of DFS.


Recurrences occurred despite obtaining block resections according to the tumor's clinical stage and pathologically free margins in all cases. Further analyses are mandatory to investigate hidden microscopic pathways of tumor diffusion, particularly in bone. Multi-institutional protocols are needed to facilitate comparisons between studies and enable meaningful meta-analyses.

Level of Evidence

2b. Laryngoscope, 124:1192–1198, 2014