Original article Analysis of blood flow and glucose metabolism in mammary carcinomas and normal breast: A H 2 15 O PET and 18 F-FDG PET study Michael Hentschel a , Timo Paulus b , Michael Mix a , Ernst Moser a , Egbert U. Nitzsche c and Ingo Brink a Objective To determine parameters of perfusion, distribution coefficient, and glucose metabolism as part of the tumour-specific micromilieu of breast cancer and compare them with corresponding values in normal breast tissue. Methods H 2 15 O PET and 18 F-FDG PET were performed on 10 patients with advanced invasive ductal carcinomas of the breast. Perfusion, distribution coefficient, and glucose metabolism and standardized uptake were quantified and analysed. Results Mean values based on the regions of interest were 59.2 ± 43.9 ml  min –1  100 g –1 (perfusion), 0.58 ± 0.26 ml  g –1 (distribution coefficient), 7.76± 6.10 (standardized uptake), and 5.4 ± 2.5 mg  min –1  100 g –1 (glucose metabolism). The corresponding values for normal breast tissue were 22.1 ± 13.2 ml  min –1  100 g –1 (perfusion), 0.16 ± 0.05 ml  g –1 (distribution coefficient), 0.33 ± 0.07 (standardized uptake), and 0.18 ± 0.08 mg  min –1  100 g –1 (glucose metabolism). For each tumour–normal tissue parameter pair, the mean values were significantly higher in tumours than normal breast tissue. Region-of-interest and pixel-wise correlation analysis revealed a positive association between glucose metabolism and distribution coefficient and glucose metabolism and perfusion for 7/10 tumours investigated. Conclusions H 2 15 O and 18 F-FDG PET were able to differentiate breast cancer and normal breast tissue. The pixel-wise analysis revealed information about the heterogeneity of tumour fine structure in perfusion, distribution coefficient, and glucose metabolism, which may provide important guidelines for improving individual treatment. Nucl Med Commun 28:789–797  c 2007 Lippincott Williams & Wilkins. Nuclear Medicine Communications 2007, 28:789–797 Keywords: PET, H 2 15 O, 18 F-FDG, perfusion, distribution coefficient, metabolic rate of glucose a Division of Nuclear Medicine and PET Center, University Hospital of Freiburg, Germany, b Philips Technologie GmbH Forschungslaboratorien, Aachen, Germany and c Division of Nuclear Medicine and PET Center, Kantonsspital Aarau, Switzerland Correspondence to Dr Michael Hentschel, Division of Nuclear Medicine and PET Center, University Hospital of Freiburg, Hugstetter Str. 55, 79106 Freiburg i. Br., Germany Tel: +49 761 270 3960; fax: +49 761 270 3989; e-mail: michael.hentschel@uniklinik-freiburg.de Received 28 March 2007 Revised 30 May 2007 Accepted 1 June 2007 Introduction Malignant tumours are characterized by a heterogeneous, frequently inadequate, and partly disorganized vascular architecture. The functions associated with such a vascular structure are altered compared to those of normal tissue. Heterogeneities in blood flow are pre- dicted for tumours. In fact, both low and high perfusion values have been reported [1]. Moreover, some tumours have an inadequate oxygen supply (hypoxia). Thus, each tumour has a micromilieu that is characterized by a variety of parameters, including angiogenesis, vasculariza- tion, tissue perfusion, nutrient metabolism, oxygenation, oxygen consumption, pH and bioenergetic status [1–4]. Perfusion is an important parameter by which to characterize individual tumours. Since perfusion is a prerequisite for transporting nutrients and therapeutics and determines their local distribution and concentra- tion, it may affect tumour sensitivity to radiation and chemotherapy [5]. In addition to the microsphere method, which is not approved for use in patients [6], 13 NH 3 and H 2 15 O positron emission tomography (PET) are the ‘gold standards’ for measuring myocardial [7–10] and brain [11] perfusion. There are few perfusion data on human tumours outside of the skull, even when magnetic resonance imaging (MRI) and computed tomography (CT) data are taken into account [12–18]. 15 O-labelled water is well suited for tracing perfusion. It diffuses freely through membranes, and its first-pass extraction is B100%. Thus, H 2 15 O PET enables quantification of tumour blood flow. In contrast to 13 NH 3 and other tracer agents, H 2 15 O perfusion is not affected by metabolic processes. Another important characteristic of some tumours is increased glucose metabolism, which is the basis for imaging with 2-[ 18 F]fluoro-2-deoxy-D-glucose ( 18 F-FDG) PET [19–21]. Furthermore, quantification of 18 F-FDG metabolism may be important in evaluating response to 0143-3636  c 2007 Lippincott Williams & Wilkins Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.