Feed

Background/Objectives: Metastasizing pleomorphic adenoma (MPA) is a rare salivary gland tumor characterized by benign pleomorphic adenoma morphology despite regional or distant metastatic behavior. Because most evidence derives from case reports and small series, this systematic review and pooled case-report analysis aimed to summarize published patient-level data and to explore clinicopathologic features, metastatic distribution, treatment, and survival without inferring causal treatment effects. Methods: PubMed, Scopus, Google Scholar, and reference lists were searched for English-language reports published up to 1 August 2025. Eligible reports described primary pleomorphic adenoma and metastatic disease reported as MPA with benign morphology; carcinoma ex pleomorphic adenoma and other malignant pleomorphic adenoma variants were excluded. The primary pooled analysis was restricted to 122 published case-report patients. Institutional cases are presented separately as illustrative cases and were not included in the primary pooled statistical analyses. Unknown, not reported, NA, and unclear values were treated as missing, and denominators are reported. Because only 16 deaths were available, survival analyses were exploratory and no multivariable Cox model was fitted. No formal risk-of-bias or certainty-of-evidence assessment was performed. Results: In the 122 published patients, sex was available for 121 patients and 71/121 (58.7%) were female. Primary tumors most frequently arose in the parotid gland (92/122, 75.4%), followed by the submandibular gland (15/122, 12.3%) and palate/soft palate (10/122, 8.2%). The median interval between primary pleomorphic adenoma and MPA was 12 years (IQR 7-21; range 0-69; n = 118). Metastatic site was reported in 120 patients and was non-mutually exclusive: bone/skeleton and lymph nodes/neck were each reported in 42/120 patients (35.0%), followed by lung/pulmonary metastases in 31/120 (25.8%). Treatment information was available for 98/122 patients. MPA-directed surgery was performed in 86/98 patients with available treatment information (87.75%). The non-surgical cohort received radiotherapy, chemotherapy, palliative treatment, treatment refusal, observation, or other non-surgical strategies. A total of 72 patients had complete covariate data for the reported Cox analyses, including 16 deaths. Estimated 1- and 5-year overall survival were 89.97% and 66.2%, respectively. Surgery was associated with improved overall survival in exploratory Kaplan-Meier comparison (log-rank p = 0.008) and univariable Cox analysis (HR 0.119, 95% CI 0.018-0.810, p = 0.030), but this association should not be interpreted as causal. Conclusions: MPA remains difficult to predict and the available evidence is limited by case-report design, missing data, publication bias, and heterogeneous follow-up. Surgical management may be considered in selected patients when technically feasible and clinically appropriate, but the observed survival association is exploratory and may be influenced by selection bias and disease characteristics. Multicenter registries, standardized reporting, centralized pathology review, molecular characterization, and longer follow-up are needed.