Low-grade myxofibrosarcoma: a clinicopathologic analysis of 49 cases treated at a single institution with simultaneous assessment of the efficacy of 3-tier and 4-tier grading systems.

Grading of myxofibrosarcoma (MFS) is contentious and based on a variety of factors, such as the percentage of myxoid or solid areas, tumor necrosis, mitotic counts, and so on. These factors are often used in combination for different grading schemes, which have not been uniformly evaluated by consistent criteria for patients treated and prospectively followed up at a single institution. Because only a subset of low-grade (LG) MFS will progress to high grade and metastasize after relentless local recurrences, we analyze various histologic parameters and grading methods to identify prognostic predictors of the LGMFS. Forty-nine cases were classified as LGMFS after review, by using > or =30% of myxoid component, but < or =20% of solid areas and only focal, < or =10%, of tumor necrosis as cutoffs, as modified from a 2-tier system that is used in our hospital. These cases were also graded in parallel by French Federation of Cancer Centers Sarcoma Group (FNCLCC) 3-tier and 4-tier grading schemes. The study cohort consisted of 26 men and 23 women, with a median age of 60.5 years. Nineteen cases were superficial, and 30 were deep seated, with the most common site being the lower limb (57%), followed by the upper limb (31%), trunk (8%), and head and neck (4%). The primary tumors ranged from 1.5 to 24 cm in size. Solid areas (5% to 20%) were seen in 23 cases, tumor necrosis (5% to 10%) was observed in 4 cases, and a predominant myxoid area (> or =75%) was noted in 22 cases. Mitotic activity ranged from 0 to 16 (median, 2) per 10 HPF. Comparing FNCLCC 3-tier versus 4-tier grading, respectively, 26 versus 10 tumors were classified as grade I, and 23 versus 39, as grade II, with 16 cases (33%) graded discordantly by 2 schemes. A median follow-up of 55 months in 49 patients (range, 9 to 171 months) revealed local recurrence occurring in 28 patients (57%), 15 and 7 of which developed multiple local recurrences and distant metastases, respectively. There was only 1 case with pulmonary metastasis without a prior local recurrence. Currently, 33 patients are alive with no evidence of disease, 4 are alive with disease, 9 are dead of disease, and 3 are dead of unknown causes. The 5-year recurrence-free survival, metastasis-free survival (MeFS), and disease-specific mortality (DSM) rates were 41%, 90%, and 4.4%, respectively. Size of larger than 5 cm (P = 0.032), tumor necrosis (P = 0.033), and < 75% of myxoid area (P = 0.042) were significant risk factors for DSM; the former two (P = 0.011 for size larger than 5 cm, P = 0.038 for necrosis) were also significantly related to MeFS. Both FNCLCC 3-tier and 4-tier schemes failed to show a significantly better outcome in grade I LGMFS than grade II lesions with respect to all 3 endpoints. In conclusion, our data statistically validated the previous impression that even the blandest LGMFS still carries a recurrent potential that cannot be foreseen by either different grading schemes or other clinicopathologic parameters. However, DSM rate is significantly related to tumor necrosis, large size, and decreased myxoid area. Tumors having necrosis or exceeding 5 cm are at significant risk of metastatic relapse.