Treatment of unknown primary head and neck squamous cell carcinoma: primary surgery versus primary radiotherapy

Unknown primary head and neck squamous cell carcinoma is a rare condition with poor prognosis compared to tumours with a known primary site. There are no consistent guidelines or strong evidence to guide the management of these tumours. Surgery or radiotherapy are equally common primary treatment modalities for these patients. A systematic review was therefore conducted to assess the effectiveness of primary surgery compared to primary radiotherapy. A pre-defined search strategy was used to search PubMed, Embase and ProQuest databases. Titles and abstracts were screened against inclusion criteria and full texts of potentially relevant studies retrieved to assess final eligibility. These studies underwent critical appraisal by two independent reviewers for assessment of their methodological quality. Five of these studies were included in pooled meta-analysis. Primary outcome measures of interest were overall survival and regional and relapse free survival. Primary emergence, neck recurrence and distant metastasis rates were extracted and analysed to substantiate the primary outcome measures. Following screening of 9376 unique records identified by the search, ten retrospective cohort studies, including a total of 655 participants that analysed data from patient registries, were included in the review. Across the included studies, quality of data synthesis and reporting was poor, especially the stratification of end point survival data, summary statistics, reporting of treatment related toxicities and quality of life measures. Meta-analysis (n=5) revealed no statistically significant difference in overall survival based on the primary treatment modality (HR:0.86, p=0.60) but favoured primary surgery for regional and relapse free survival (HR:0.57, p=0.07). Early stage disease at the time of treatment initiation had improved overall survival, regardless of the treatment modality (HR:0.27, p=0.008). Rate of primary emergence (median = 5%) after five years did not increase when the mucosa was not irradiated in suspected cutaneous cancer patients. However, treatment with neck dissection alone without patient risk stratification increased primary emergence rates as well as neck failure rates. In conclusion, there is no treatment modality dependent difference in overall survival or regional and relapse free survival. Cutaneous origin of unknown primary head and neck cancers need to be considered and treated differently to cancers of occult mucosal origin.Thesis (MClinSc.) -- University of Adelaide, Adelaide Medical School, 201