MRI characterization of residual mediastinal masses in Hodgkin’s disease: long-term follow-up Ernesto Di Cesare a, *, Gabriella Cerone a , Riccardo Maurizi Enrici b , Vincenzo Tombolini a , Paola Anselmo c , Carlo Masciocchi a a Department of Radiology, University of L’Aquila, L’Aquila, Italy b Department of Radiology, University of Rome, La Sapienza, Rome, Italy c Department of Haematology, University of Rome, La Sapienza, Rome, Italy Received 5 March 2003; received in revised form 14 August 2003; accepted 15 August 2003 Abstract Our purpose was to evaluate the role of MRI in distinguishing fibrous from active residual masses in treated Hodgkin’s disease. Forty patients with residual mediastinal mass larger than 1.5 cm underwent MRI 1, 3, 6, and 12 months after the end of cycles of prescribed chemotherapy or combined chemoradiotherapy. The MRI examinations were performed on a 0.5 and a 1.5 T systems, using T 1 before and after gadolinium injection and T 2 -weighted sequences. Each time the residual mass was evaluated in size and signal intensity on spin echo (SE) T 2 -weighted images and on SE T 1 -weighted images after contrast medium. Low signal intensity and low contrast enhancement were considered signs of inactive residues; homogeneous high signal intensity and high contrast enhancement were indicative of active residual disease; heterogeneous signal intensity and heterogeneous contrast enhancement were indicative of partial remission or necrotic/inflam- matory phenomena. MR showed high diagnostic accuracy in the evaluation of Hodgkin’s mediastinal residues after treatment, if performed at least 6 months after the end of therapy, reaching the highest sensitivity and specificity values at 12 month follow-up (considering the three parameters—T 2 signal intensity, contrast-enhancement, and size—all together). If we consider the single parameters individually, we can observe that size variation remains the more valuable parameter to predict or to exclude a relapse. MR diagnostic accuracy at the 6-month follow-up was lower due to the higher incidence of inhomogeneous pattern. The accuracy of MR performed at 1 and at 3 months after the end of therapy was not satisfying. This represents a clinical problem because the most important clinical decisions have to be taken just in this early post-treatment phase. © 2004 Elsevier Inc. All rights reserved. Keywords: Hodgkin disease; MR; Lymphoma; Magnetic resonance; Tissue characterization 1. Introduction Magnetic resonance (MR) has been indicated for the diagnosis and staging of Hodgkin’s disease [1–3] because of its multiplanar imaging potentialities and its intrinsic con- trast resolution. At present time, with the exception of some body dis- tricts, such as supraclavicular region and pulmonary ila, the diagnostic potentialities of computed tomography (CT) are similar to those of MR and an absolutely comparable accu- racy has been reported in literature [4,5]. In addition, MRI is more expensive than CT and less diffuse. On the basis of these considerations, CT can be considered as the most appropriate technique for staging Hodgkin’s disease. Nev- ertheless, some authors [6 – 8] demonstrated that MR is helpful in the follow-up after therapy due to its ability in distinguishing active tissue from residual fibrosis. The incomplete remission is one of the most important problems encountered in the treatment of Hodgkin’s dis- ease: in some cases even a large residue can be formed by fibrous tissue; in other cases instead a residual active tissue can be observed [9 –12]. The volumetric monitoring performed by CT has some limitations because the regression pattern varies in the dif- ferent patients depending on initial size, site, histology, and method of treatment of the tumor. Similarly, the CT atten- uation values may be similar for active and fibrous tissue [13,14]. Even biopsy and surgery can result inappropriately * Corresponding author. Tel.: +39-0862-414258; fax: +39-0862- 311277. E-mail address: ernesto.dicesare@cc.univaq.it (E. Di Cesare). Magnetic Resonance Imaging 22 (2004) 31–38 0730-725X/04/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.mri.2003.08.002