Nuclear grading of myxoid liposarcoma robustly predicts patient outcome: a study of 280 patients

Why clinicians should know about it

• Picked for Pathology and Forensic Medicine (paper of the day, 17 October 2026): Nuclear grading predicts outcome, improves prognostic reproducibility

Abstract

Myxoid liposarcoma (MLPS) exhibits variable morphology and clinical behavior. Current grading systems, which rely on the proportion of round cell areas, were developed for resections, rather than core biopsies. Additionally, criteria for round cell and "transitional" areas are subjective/poorly reproducible. It has recently been suggested that nuclear grading may be superior to traditional grading in predicting the behavior of MLPS. We compared nuclear and cellularity-based grading in a large, multinational MLPS cohort. 280 MLPS (1990-2025) were reviewed; nuclear grade was assigned using a slight modification of the Kawaguchi et al. system. Disease‑specific survival (DSS), metastasis‑free survival (MFS), and disease‑free survival (DFS) were estimated and the performance of both grading systems was compared. Tumors occurred in 152 males/128 females (median 45 years of age; range 15-91 years) and predominantly involved proximal deep soft tissues. Initial diagnoses were made with needle core (149 cases) and open biopsies (131 cases). DDIT3 rearrangement was documented in 195 tumors. Nuclear grades were 1 (71 cases), 2 (121 cases) and 3 (88 cases); grades 1 and 2 were considered "low-grade" and grade 3 "high-grade". Both systems assessed 192 (69%) and 88 (31%) tumors to be low and high-grade, respectively, although there was not perfect overlap. Clinical follow-up (280 patients; median 68 months, range: 3-355 months) showed 33 local recurrences, 96 distant metastases and 63 cancer-specific deaths. Tumors with low nuclear grade demonstrated significantly better DSS, MFS and DFS; those with low conventional grade did not. Identical results were seen with tumors diagnosed by needle biopsy (149 cases). The Harrell's C-index demonstrated nuclear grading to be superior in the prediction of DSS, MFS and DFS. We conclude that nuclear grading demonstrates superior prognostic performance over conventional grading, including in tumors diagnosed by needle biopsy. These results argue for more widespread adoption of nuclear grading in MLPS prognostication.

Abstract as published, via PubMed.