1948P Concordance of the pathological diagnosis between local institutional and central judgment in high-grade non-round-cell sarcomas: A supplementary analysis of JCOG1306

JCOG1306 was a multicenter randomized phase II/III trial comparing perioperative chemotherapy with adriamycin plus ifosfamide (AI) and gemcitabine plus docetaxel (GD) for high-grade non-round-cell sarcomas of FNCLCC classification ≥G2. After the histopathological diagnosis was confirmed by each local institution, specialized sarcoma pathologists reviewed each pathological diagnosis. In this study, we examined the concordance proportion of pathological diagnoses between the local institutional judgment (LIJ) and central judgment (CJ) and its impact on the study. A total of 142 patients with proven high-grade non-round-cell sarcoma based on an incisional biopsy were included. We evaluated the concordance between the LIJ and CJ for pathological types and FNCLCC classification. The concordance proportion for pathological types between the LIJ and CJ was 63.4%. Compared to the LIJ (G1: 0%, G2: 59.9%, G3: 40.1%), the CJ (G1: 15.5%, G2: 61.3%, G3: 23.2%) showed more patients diagnosed with G1, and the concordance proportion for the FNCLCC classification was 62.7%. Histological types diagnosed as G1 were myxofibrosarcoma in 15 patients, myxoid liposarcoma in 6 patients, and fibrosarcoma in 1 patient. The proportion of patients in which the LIJ agreed with the CJ for each histological type was higher for synovial sarcoma (100%) and myxofibrosarcoma (88.2%) and lower for undifferentiated/unclassifiable sarcoma (17.6%). The 3-year survival proportions for 120 patients excluding G1 cases (AI: 59, GD: 61) were 89.8% and 74.9%, respectively, and those for all 143 patients (AI: 70, GD: 73) were 91.4% and 79.2%, respectively. The present analysis included 15% of G1 patients who were not originally eligible for this trial. However, this did not significantly affect the results, as the AI and GD groups contained similar numbers of G1 patients. Patients with a background of myxoid components tended to be judged as G1 by the CJ. Although the limitations of assigning a grade using an incisional biopsy should be considered, the CJ is still useful for ensuring the accuracy of the clinical trial for such rare cancers.