Prognostic impact of tumor perfusion in MR-imaging studies in Ewing tumors.

BACKGROUND: Intratumoral hypoxia is associated with poor prognosis in various solid tumors. Severe and long-lasting hypoxia results in necrosis. The presence of necrosis therefore might also be correlated with unfavorable outcome. This has been demonstrated for head and neck cancers, gliomas and adult soft tissue sarcomas. We have investigated whether or not the patterns of contrast enhancement and the presence of visible necrosis on pretreatment MR images has prognostic impact in Ewing tumors. PATIENTS AND METHODS: From December 1993 though March 1997, 79 patients with Ewing tumors were prospectively analyzed for the presence and amount of necrosis in their tumors. The median age was 12 years (range 4-30 years). The median follow-up at the time of this analyses was 3 years. All patients were treated according to the multicentric EICESS-92 protocol with multi-agent chemotherapy (VACA or VAIA or EVAIA) and local therapy (radiotherapy with 50-60 Gy or surgery or surgery with pre- or postoperative irradiation with 45-50 Gy). For the measurement of necrosis, gadolinium contrast-enhanced T1-weighted MR images were used. Necrosis was defined as non-perfused areas in the tumor and the necrotic volume was expressed as percentage of total volume.
RESULTS: Out of 79 tumors, 10 (13%) showed no necrosis, 30 (38%) had 1-25% necrosis, 21 (27%) had 26-50% necrosis and the remaining 18 (23%) more than 50% necrosis. There was a correlation between tumor size and necrosis (p = 0.001): the median tumor volume increased with amount of necrosis (47 cm3 in non-necrotic tumors, 59 cm3 vs 280 cm3 vs 284 cm3 for tumors with 1-25% vs 26-50% vs > 50% necrosis). Tumor site (central location vs proximal extremities vs distal extremities) had no impact on necrosis (p = 0.71). 23 out of 79 patients had metastases (M1) at the time of diagnosis. The frequency of metastatic spread increased with the amount of necrosis: 1/10 (10%) patients with non-necrotic tumors had metastases vs 7/30 (23%) vs 6/21 (28%) vs 9/18 (50%) for tumors with 1-25% vs 26-50% vs > 50% necrosis. "Unfavorable" metastatic spread (bone or multiple metastases) was only noted in patients with high amount of necrosis (> 25% necrosis) whereas even large tumors did not show unfavorable metastases if they contained no or only small amounts of necrosis. All patients with non-necrotic tumors survived event-free during the observation period. Patients with necrotic tumors had a 3-year event-free survival of 55% (p = 0.06 vs tumors without necrosis).
CONCLUSIONS: The presence of non-perfused (presumably necrotic) areas on pretreatment contrast-enhanced MR-images is associated with an increased risk of metastases, especially an unfavorable pattern of metastatic spread at diagnosis. This observation may be explained by a more aggressive biological behavior of hypoxic tumor cells. The small group of patients with non-necrotic tumors (13%) had an excellent prognosis suggesting that the absence of necrosis might be helpful in identifying a very favorable prognostic subgroup in Ewing tumors.