Developing a diagnostic model to differentiate the well-differentiated lipomatous tumors based on clinicopathological...

This study evaluated the diagnostic value of clinicopathological features and immunohistochemical markers for distinguishing lipomas from atypical lipomatous tumors/well-differentiated liposarcomas (ALT/WDLPS). An integrated diagnostic model for ALT/WDLPS was developed to guide diagnosis, treatment planning, and prognosis. This retrospective analysis included 216 patients with lipomatous tumors diagnosed between February 2018 and December 2024, including lipomas (n = 149), spindle cell lipomas (n = 3), ALTs/WDLPs (n = 62), and WDLPS with a low-grade de-differentiated component (n = 2). Immunohistochemical data for MDM2, CDK4, and p16 were available for 131 patients. MDM2 amplification was significantly more frequent in patients ≥ 55 years and in tumors of the lower limbs (the thigh) and retroperitoneum (p = 0.000). Larger tumor size and multiplicity were also associated with MDM2 amplification (p < 0.05). Immunohistochemistry sensitivities for ALT/WDLPS vs. lipomas: 65% (MDM2), 100% (CDK4), and 80.4% (p16); combined, the specificity was 100% and sensitivity 85.6%. The diagnostic model achieved 93.3% sensitivity and 72.2% specificity. Scores < 0.219 indicated a higher likelihood of lipoma, while scores > 0.652 indicated a higher likelihood of liposarcoma. Age ≥ 55 years, lower extremity/retroperitoneal location, tumor diameter ≥ 9.9 cm, and positive markers were independent risk factors. This model provides an effective tool for ALT/WDLPS identification.