170 AJR:201, July 2013 and the National Cancer Institute grading sys- tems [2]. Tumor necrosis has been repeated- ly shown to be an important prognostic factor in soft-tissue sarcomas with respect to overall and metastasis-free survival [3]. New prognos- tic models (e.g., the SING [size, invasion, ne- crosis, growth] model) have been proposed that include necrosis as one of the main variables [4]. The problem with necrosis on pathology as a prognostic variable is significant interob- server variation in pathologic assessment of a biopsy or surgical specimen, because no more than 60–70% concordance is obtained between (even specialized) observers [5–7]. Fluorine-18-FDG PET/CT can be used as a noninvasive technique to reliably assess for the presence of necrosis in sarcomas [8]. The use of necrosis, diagnosed from metabolic data determined by FDG PET/CT, as an inde- Necrosis on FDG PET/CT Correlates With Prognosis and Mortality in Sarcomas Rajan Rakheja 1 William Makis 2 Rima Tulbah 1 Sonia Skamene 1 Christina Holcroft 3 Ayoub Nahal 4 Robert Turcotte 5 Marc Hickeson 1 Rakheja R, Makis W, Tulbah R, et al. 1 Department of Nuclear Medicine, Royal Victoria Hospital, 687 Pine Ave, Montreal, QC H3A 1A1, Canada. Address correspondence to R. Rakheja (rajan.rakheja@gmail.com). 2 Department of Nuclear Medicine, Brandon Regional Health Centre, Brandon, MB, Canada. 3 Department of Epidemiology, Jewish General Hospital, Montreal, QC, Canada. 4 Department of Pathology, Musculoskeletal Division, Montreal General Hospital, Montreal, QC, Canada. 5 Department of Oncology, McGill University, Montreal, QC, Canada. Nuclear Medicine and Molecular Imaging • Original Research AJR 2013; 201:170–177 0361–803X/13/2011–170 © American Roentgen Ray Society S arcomas are uncommon tumors that arise from the embryonic me- soderm, accounting for 0.7% of all adult malignancies and approxi- mately 6.5% of childhood cancers. They have a high mortality rate and account for 3–4% of an- nual cancer-related deaths. Despite aggressive treatment with surgery and neoadjuvant and adjuvant chemoradiotherapy, the prognosis for most histologic subtypes of sarcomas remains poor, and the survival rate has not changed sig- nificantly over the past several decades [1]. Necrosis is an important part of sarcoma grading and prognostication and features prom- inently in two grading systems recognized by the World Health Organization (WHO)—the French Federation of Cancer Centers Sarco- ma Group (Fédération Nationale des Cen- tres de Lutte Contre le Cancer [FNCLCC]) Keywords: FDG, 18 F-FDG PET/CT, necrosis, prognosis, sarcoma DOI:10.2214/AJR.12.9795 Received August 3, 2012; accepted after revision December 20, 2012. OBJECTIVE. The purpose of this study was to determine if there is an association be- tween necrosis as identified on staging 18 F-FDG PET/CT and overall survival (OS) and pro- gression-free survival (PFS) in patients with sarcoma. MATERIALS AND METHODS. Sixty-six patients with newly diagnosed limb and girdle sarcoma underwent PET/CT at our institution between June 2004 and July 2009 for sarcoma staging before treatment with curative intent. The tumor maximum standardized up- take values (SUV max ), the presence of necrosis, and the volume of necrosis were measured for each primary tumor and correlated with follow-up data. PFS and OS were analyzed using the Kaplan-Meier method. Proportional hazards models were used to estimate hazard ratios. RESULTS. Median patient age was 49 years, and 51.6% of the patients were men. Sar- comas were categorized as soft tissue (69.2%), bone (23.5%), or other (7.3%). Mean follow- up time was 33.3 months. During the follow-up interval, 53% of patients experienced disease progression, and 40.9% died. There was a statistically significant relationship between the presence of necrosis and OS (by log-rank test, p = 0.001), as well as PFS (by log-rank test, p = 0.0001). Twenty-four-month OS was 96%, 65%, and 38% in patients with tumors with ab- sence necrosis, those with presence of necrosis, and with necrosis volume greater than 50%, respectively. Forty-eight-month OS was 81% in patients with absence of necrosis and 41% in patients with presence of necrosis. Twelve-month PFS was 96%, 60%, and 42% in patients with tumors with absence of necrosis, those with presence of necrosis, and those with necro- sis volume greater than 50%, respectively. Twenty-four-month PFS was 83%, 38%, and 22%, respectively, in these groups. CONCLUSION. The presence of necrosis and the volume of necrosis, as identified on the staging FDG PET/CT and after adjusting for SUV max , are strong independent adverse prognostic factors for disease recurrence and death in patients with limb and girdle sarcomas. Rakheja et al. FDG PET/CT of Sarcomas Nuclear Medicine and Molecular Imaging Original Research Downloaded from www.ajronline.org by MCGILL UNIVERSITY LIBRARIES on 12/26/13 from IP address 132.216.1.36. Copyright ARRS. For personal use only; all rights reserved