explanation for our negative findings. Although a higher dose might have been effective, the concentration of -3 fatty acids that was administered was similar to the formulation evaluated in the Belluzzi et al Study. Although we agree with Dr Aberra that a well-balanced diet “cannot hurt,” we would like to underscore our main conclusion that -3 supplementation is ineffective for the prevention of relapse and that patients would be better served by taking maintenance therapies of known efficacy rather than using the -3 supplements that are widely available in health food stores. BRIAN G. FEAGAN, MD QUANTITATIVE DYNAMIC CONTRAST- ENHANCED MAGNETIC RESONANCE IMAGING FOR ASSESSING TREATMENT RESPONSE IN HEPATOCELLULAR CARCINOMA Liang PC, Ch’ang HJ, Hsu C, et al. (Department of Medical Imaging, National Taiwan University Hospital, Taipei, Taiwan). Dynamic MRI signals in the second week of radiotherapy relate to treatment outcomes of hepatocellular carcinoma: a preliminary result. Liver Int 2007;27:516 –528. Angiogenesis is a process represented by the formation of new blood vessels from the existing vascular bed. Although this is a common mechanism associated with normal tissue repair and regeneration, the presence of disordered, extensive neoangiogenesis is also a distinctive feature of neoplasms (Annu Rev Med 2006;57:1–18). In clinical practice, contrast-enhanced imaging is used for determining the presence and extent of malignancy based on the detection of hypervascularity. After an interven- tion (local or systemic), a reduction in tumor size and/or a significant qualitative reduction in arterial perfusion is often defined as a satisfactory response (Nat Rev Cancer 2006;6:409 – 414). In the absence of overall survival or progression-free survival as an endpoint, however, there remains an inability to provide a quantitative assessment of treatment response as a way of risk stratifying an individual’s prognosis (AJR Am J Roentgenol 2004;183: 713–719). Dynamic contrast-enhanced magnetic resonance imag- ing (DCE-MRI) has been shown to detect the later phases of angiogenesis characterized by increased perfusion and new vessel formation (Top Magn Reson Imaging 2001; 12:301–308). DCE-MRI is sensitive to differences in blood volume and vascular permeability that can be as- sociated with tumor angiogenesis. The technique in- volves delayed acquisition of T1-weighted images after contrast administration. For human studies, the contrast agent is generally a low-molecular-weight, gadolinium- based compound. Because extravasation of contrast agent occurs from the vascular space to the interstitial space (providing information about blood volume and microvascular permeability), the accumulation of con- trast agent within the interstitium results in a signal increase on T1-weighted MRI. A subsequent washout effect can be observed if the vascular permeability is high. Signal intensity changes in proportion to the contrast agent concentration in the volume element of measure- ment, or voxel (J Clin Oncol 2006;24:3293–3298). Evaluation of human tumor angiogenesis using DCE- MRI is based on the observation that angiogenesis can increase the perfusion and permeability of tumors (J Magn Reson Imaging 1999;10:254 –259; Adv Drug Deliv Rev 2000;41:91–110; Clin Radiol 2001;56:607– 620). Be- cause no gold standard is available to directly verify pharmacodynamic measurements by DCE-MRI, investi- gations to date have compared DCE-MRI with clinical outcomes or existing prognostic markers such as mi- crovessel density count and vascular endothelial growth factor (VEGF) tissue expression (J Clin Oncol 2006;24: 3293–3298; J Magn Reson Imaging 1999;10:254 –259; Clin Radiol 2001;56:607– 620). Multiple reports regard- ing the evaluation of various kinds of malignancies using DCE-MRI have been published. Despite inherent differ- ences in tumor etiology and data analytic approach, there is a general correlation between stage and prognosis with enhancement parameters (Adv Drug Deliv Rev 2000;41: 91–110; Top Magn Reson Imaging 2001;12:301–308; J Clin Oncol 2006;24:3293–3298). Hepatocellular carcinoma (HCC) is the 5th most com- mon cancer worldwide and the 3rd most common cause of cancer mortality (Gastroenterology 2007;132:2557–2576). For patients with cirrhosis from hepatitis B and C infec- tion, it has become the major cause of death as opposed to complications from portal hypertension (Mayo Clin Proc 2005;80:1501–1508). Given the known hypervascu- larity of this tumor and diagnostic validity of radiologic criteria (Hepatology 2005;42:1208 –1236), there has also been evidence documenting the occurrence of angiogen- esis with development of HCC (Hepatology 1998;28:68 – 77; Arch Pathol Lab Med 2000;124:1061–1065; Gut 2001; 48:87–96; Ann Surg Oncol 2003;10:355–362; Br J Surg 2004;10:1354 –1360; Hepatology 2007;46:446 – 455). Studies to date have examined the role of DCE-MRI in assessing patients with HCC. Among 30 patients with histopathologically proven HCC undergoing DCE-MRI, both microvessel density count and VEGF tissue expres- sion were higher in hypervascular tumors as compared with iso- and hypoenhancing lesions on arterial phase images. These observations suggest that contrast-en- hancement patterns on DCE-MRI are likely to be influ- enced by angiogenesis (Acta Radiol 2005;46:353–388). DCE-MRI has also been investigated for its ability to measure the effects of antiangiogenic therapy in HCC. Preliminary data suggest a reduction in tumor perfusion defined by semiquantitative parameters from DCE-MRI September 2008 SELECTED SUMMARIES 1007