The role of perineural invasion in parotid malignancy outcomes: A systematic review and meta-analysis

Purpose: At present, perineural invasion is used as a histologic indicator of aggressive salivary gland disease. In other head and neck malignancies, perineural invasion impacts staging of cancer lesions and therefore affects treatment options. Objective: To compare survival outcomes in primary parotid malignancies with and without perineural invasion. A systematic review pooled data from the scientific literature in patients with any primary parotid malignancy to investigate the prognosis of those with perineural invasion. Data Sources: PubMed (Medline), Scopus and Cochrane databases were queried from inception to July 2020 without any initial search constraints. Additional publications were included from review of pertinent articles. Study Selection: Our inclusion criteria included primary parotid cancers with reported perineural invasion on survival outcomes. Exclusion criteria were non-English language text, non-human studies, reviews, textbooks, abstracts, case reports and case series. Two authors independently reviewed articles for inclusion. Of the initial 465 records, 83 articles were reviewed in full to yield a final collection of 14 studies. Data Extraction and Synthesis: PRISMA-p guidelines were used in the reporting of our studies. A MOOSE Checklist was also used. MINORS criteria were applied to assess risk of bias. Random-effects models were used to estimate pooled effect sizes. No institutional review board review was needed for our study. Main Outcomes and Measures: Primary study outcomes were set prior to data collection and included overall survival (OS), disease-free survival (DFS), disease-specific survival (DSS) and distant-metastasis-free survival (DMFS) in patients with and without perineural invasion. Results: Fourteen studies contributed to this meta-analysis. Compared to patients without perineural invasion, the pooled rate of mortality (HR = 3.64), time to recurrence (HR = 3.56), disease-specific mortality (HR = 2.77) and distant metastasis (HR = 3.84) was significantly higher for patients with PNI (all p <.001). Controlling for perineural invasion status, no moderator was associated with these survival outcomes (all p >.05). Given the clinical severity of perineural invasion, few studies were null as shown in a panel of publication bias plots. Conclusion: Perineural invasion portends a poor survival outcome in patients with parotid malignancies.