Accuracy of Histology and Malignancy Grade between Preoperative Biopsy and Surgical Specimens in Primary Retroperitoneal Sarcoma. A Study from the Prospective Retroperitoneal Sarcoma Registry (Resar).

Abstract

This study aimed to prospectively assess the accuracy of preoperative biopsy in primary retroperitoneal sarcoma (RPS) across sarcoma referral centers. Histological subtype and malignancy grade are key for guiding RPS treatment strategies. However, the accuracy of preoperative biopsy remains uncertain. Data on adult patients with primary localized RPS who underwent preoperative biopsy followed by curative-intent surgery (2017-2020) were collected from the Retroperitoneal Sarcoma Registry. The study aimed to assess concordance between biopsy and surgical specimen histology and grade, using Cohen's kappa statistic. Concordance was also analyzed by center volume (high ≥13 vs. low <13 cases/year). Of 894 enrolled patients, histologic concordance was observed in 87.7% of cases (unweighted κ=0.814; 95% CI, 0.773-0.854). Among 172 tumors initially diagnosed as well-differentiated liposarcomas, 44 (25.6%) were reclassified as dedifferentiated liposarcomas. Grade concordance was observed in 232 of 346 cases (76.1%; weighted κ=0.652; 95% CI, 0.589-0.715), with no difference between computed tomography- and ultrasound-guided biopsies. Concordance by tumor grade was 98.9% (grade 1), 62.1% (grade 2), and 40.2% (grade 3). In dedifferentiated liposarcomas, grade concordance was 59.7% (weighted κ=0.385; 95% CI, 0.292-0.479). High-volume centers showed higher concordance for both histology (κ=0.780) and grade (κ=0.680) compared with low-volume centers (κ=0.622 and 0.564, respectively). While preoperative biopsy for RPS provides satisfactory histologic accuracy, tumor grade is frequently underestimated. This diagnostic inaccuracy may impact treatment decisions, particularly regarding preoperative therapies. Incorporating additional diagnostic factors may improve the accuracy of preoperative assessment.