CLINICAL INVESTIGATION Head and Neck AN EXPLORATORY STUDY INTO THE ROLE OF DYNAMIC CONTRAST-ENHANCED MAGNETIC RESONANCE IMAGING OR PERFUSION COMPUTED TOMOGRAPHY FOR DETECTION OF INTRATUMORAL HYPOXIA IN HEAD-AND-NECK CANCER KATE NEWBOLD, M.D.,* ISABEL CASTELLANO,PH.D., y ELIZABETH CHARLES-EDWARDS, M.SC., z DOROTHY MEARS, D.C.R.,* ASLAM SOHAIB, F.R.C.R.,* MARTIN LEACH,PH.D., z PETER RHYS-EVANS, F.R.C.S., x PETER CLARKE, F.R.C.S., x CYRIL FISHER, M.D., x KEVIN HARRINGTON,PH.D., yx AND CHRISTOPHER NUTTING, M.D. x * The Royal Marsden Hospital, Sutton, United Kingdom; y Institute of Cancer Research, London, United Kingdom; z Institute of Cancer Research, Sutton, Surrey, United Kingdom; and x The Royal Marsden Hospital, London, United Kingdom Purpose: Hypoxia in patients with head-and-neck cancer (HNC) is well established and known to cause radiation resistance and treatment failure in the management of HNC. This study examines the role of parameters derived from dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and perfusion computed tomography (CT) as surrogate markers of intratumoral hypoxia, defined by using the exogenous marker of hypoxia pimonida- zole and the endogenous marker carbonic anhydrase 9 (CA9). Methods and Materials: Patients with HNC underwent preoperative DCE-MRI, perfusion CT, and pimonidazole infusion. Imaging parameters were correlated with pimonidazole and CA9 staining. The strength of correlations was tested by using a two-tailed Spearman’s rank correlation coefficient. Results: Twenty-three regions of interest were analyzed from the 7 patients who completed the DCE-MRI studies. A number of statistically significant correlations were seen between DCE-MRI parameters (volume transfer between blood plasma and extracellular extravascular space [EES], volume of EES, rate constant between EES and blood plasma, time at arrival of contrast inflow, time to peak, average gradient, and time to onset) and areas with a pimonidazole score of 4. In the case of CA9 staining, only a weak correlation was shown with wash-in rate. There were no significant correlations between perfusion CT parameters and pimonidazole staining or CA9 expression. Conclusion: Intratumoral hypoxia in patients with HNC may be predicted by using DCE-MRI; however, perfusion CT requires further investigation. Ó 2009 Elsevier Inc. Head and neck cancer, Hypoxia, DCE MRI, perfusion CT. INTRODUCTION Hypoxia in patients with head-and-neck cancer (HNC) is well established (1–3) and known to cause radiation resis- tance and treatment failure (4). With therapeutic strategies to modify or exploit tumor hypoxia, there is a need for reli- able methods for detection that distinguish areas of clinically relevant hypoxia and are simple, noninvasive, and nontoxic. Tissue perfusion is important for oxygenation. Chaotic vessel formation with incompetent arteriovenous shunts is associated with tumor angiogenesis. Tumor arterioles are relatively more hypoxic than in normal tissue (5) and carry plasma without red blood cells (6). Flow through arteriove- nous shunts within a tumor vascular bed has been estimated at up to 30% of total tumor blood flow (BF), thereby bypass- ing the tumor and not delivering oxygen (7). This abnormal vasculature could be a surrogate marker of hypoxia. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and perfusion computed tomography (CT) characterize tissue microvasculature and function either directly from qualitative descriptions of changes in intensity arising from the passage of contrast agent or more Reprint requests to: Kate Newbold, M.D., The Royal Marsden Hospital, Downs Road, Sutton, Surrey SM2 5PT, United Kingdom. Tel: (00) 44-208-661-3638; Fax: (00) 44-208-915-6719; E-mail: kate.newbold@rmh.nhs.uk Dr. Newbold was funded by a Royal College of Radiologists/ BUPA Research Fellowship and The Royal Marsden Hospital Gen- eral Clinical Research Fund. Conflict of interest: none. Acknowledgments—We thank Dr. Frances Daley and Professor Peter Wardman of The Gray Institute for help with immunohisto- chemical staining of specimens; Dr. Jan Zavada of the Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, and Drs. Silvia Pastorekova and Jaro Pastorek of the Institute of Virology of the Slovak Academy of Sciences as the source of the CA9 antibody. Received March 27, 2008, and in revised form July 21, 2008. Accepted for publication July 22, 2008. 29 Int. J. Radiation Oncology Biol. Phys., Vol. 74, No. 1, pp. 29–37, 2009 Copyright Ó 2009 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/09/$–see front matter doi:10.1016/j.ijrobp.2008.07.039