Accuracy of cytological grading of spindle-cell sarcomas.

Spindle-cell sarcomas constitute an important component of soft-tissue sarcomas, where accurate grading may be more important than histologic subtype in order to plan treatment strategies and determine prognosis. To evaluate the applicability and accuracy of grading spindle-cell sarcomas on fine-needle aspiration cytology (FNAC) smears, 54 cases of histologically documented spindle-cell sarcomas, consisting of synovial sarcomas (20 cases), neurofibrosarcomas (12 cases), leiomyosarcomas (9 cases), dermatofibrosarcoma protuberans (DFSP; 6 cases), fibrosarcomas (3 cases), hemangiopericytomas (2 cases), and spindle-cell sarcomas, unclassified (2 cases), were graded according to a three-tier system proposed earlier for FNAC smears, while the histological sections were graded using the French Federation of National Cancer Centers (FNCLCC) grading system. The cytological grading was correlated with the histological grade. There was an overall cytologic and histologic concordance in 40/54 (74%) cases, and concordance in 9/13 (69%) grade I, 19/25 (76%) grade II, and 12/16 (75%) grade III cases. Analysis of grading of individual sarcomas revealed a concordance in 92% of neurofibrosarcomas, 78% of leiomyosarcomas, 70% of synovial sarcomas, 67% of dermatofibrosarcoma protuberans, 67% of fibrosarcomas, 50% of hemangiopericytomas, and 50% of cases of malignant mesenchymal tumors, spindle-cell type unclassified. Major noncorrelation was seen in 10/54 (18.5%) cases. Minor noncorrelation was seen in 4/54 (7.4%) cases. Thus it is possible to accurately predict the grade in 74% of cases of spindle-cell sarcomas. The cytological and histological concordance was better (75%) in high-grade (grades II and III) as compared to grade I sarcomas (69%). Sampling errors due to morphologic heterogeneity in sarcomas may cause noncorrelation in a few cases.