MRI of Pulmonary Nodules: Closing the Gap on CT

HomeRadiologyVol. 302, No. 3 PreviousNext Reviews and CommentaryFree AccessEditorialMRI of Pulmonary Nodules: Closing the Gap on CTMark O. Wielpütz Mark O. Wielpütz Author AffiliationsFrom the Translational Lung Research Center (TLRC), German Center for Lung Research (DZL), University of Heidelberg, Im Neuenheimer Feld 156, 69120 Heidelberg, Germany; Department of Diagnostic and Interventional Radiology, University of Heidelberg, Heidelberg, Germany; and Department of Diagnostic and Interventional Radiology with Nuclear Medicine, Thoraxklinik at University of Heidelberg, Heidelberg, Germany.Address correspondence to the author (e-mail: [email protected].).Mark O. Wielpütz Published Online:Nov 30 2021https://doi.org/10.1148/radiol.212516MoreSectionsPDF ToolsAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookXLinked In See also the article by Ohno et al in this issue.Dr Mark O. Wielpütz is a professor in the Department of Diagnostic and Interventional Radiology at the Heidelberg University Medical Center. His research interest is focused on quantitative chest MRI and CT in chronic lung diseases. As a member of the German Lung Research Center, he is a principal investigator for imaging trials in cystic fibrosis lung disease.Download as PowerPointThe field of pulmonary MRI is rapidly emerging, yet it had a tough start. The physically disadvantageous conditions inside the lung had long made it look impossible to achieve diagnostic images with MRI. Low proton density and fast signal decay resulted in rather dull images of the lungs. Also, why should one seek out alternatives to CT when CT excels with great image quality and steadily declining radiation exposure? Still, sequence improvements in MRI technology also benefited pulmonary MRI, and research interest increased.The main arguments for lung MRI could be summarized as follows. First, MRI lacks radiation exposure, which, even with the advent of low-dose CT, still holds true for pediatric and chronic lung diseases with long-term observation required. Second, functional techniques allow for repeated assessments of processes like ventilation and perfusion with high spatial and temporal resolution, even without the use of contrast material (1). Third, pulmonary diseases, such as infections or nodules, increase the number of protons per voxel and displace airspace, thus improving conditions for structural MRI and allowing for a good sensitivity for these "plus pathologies." The aforementioned advantages led to the robust implementation of lung MRI for cystic fibrosis. Fourth, growing interest in whole-body MRI for patients with cancer as a "one-stop shop" examination, and in hybrid imaging with PET/MRI, demands for diagnostic accuracy in all parts of the body imaged, including the lungs—for example, in the context of lung cancer staging (2). Finally, lung cancer screening will develop into a major factor in radiation exposure to certain populations. In the interest of noninvasiveness, MRI scans as a potential alternative may even be easier to evaluate because nodules stand out against the signal-free normal lung. In addition, it may be more cost-effective due to fewer false-positive findings (3).During an assessment, high isotropic imaging resolution is desirable to detect and characterize a pulmonary nodule. Size (diameter, volume) and certain visual criteria, such as relative contribution of solid and subsolid components, calcifications, and nodule-lung interaction at its circumference (spiculation, bubble-like or cystic necrosis, lobulation), constitute the most important criteria for malignancy risk at imaging. These criteria are systematically applied in various evaluation systems used to determine nodule management for incidental nodules (eg, Fleischner guidelines) and lung cancer screening (eg, Lung CT Screening Reporting and Data System [Lung-RADS]).In a 2003 study using ex vivo porcine lung explants, Biederer et al (4) demonstrated a sensitivity of 84%–88% for artificial pulmonary nodules smaller than 5 mm using gradient-echo sequences at 1.5 T. But then, initial efforts to detect nodules in a lung cancer screening population using combinations of conventional MRI sequences with low-dose CT as the standard of reference brought sobering results: The sensitivity of MRI was shown to be only 48%, with 88% specificity for lesions with an average long-axis diameter of 15 mm. Within the group of confirmed lung cancers, however, sensitivity was at least 78%. This reassuring finding indicated that MRI may be more sensitive to malignant nodules, which in part also results from the fact that extensively calcified nodules are naturally a blind spot in MRI (5). Other work brought somewhat better results with conventional gradient-echo MRI sequences. Still, the clinical impact of pulmonary MRI for nodules remained low.The most recent developments in MRI sequences for the lung made ultrashort echo time (UTE) measurements available on state-of-the-art MRI systems. UTE MRI requires MRI hardware with very short times to switch between transmit and receive modes; such systems are now more widely available. Simply speaking, the use of UTEs, when the time between MRI pulse excitation and readout is kept below 200 µsec, may forestall the extremely short T2* of lung tissue, thereby avoiding its fast signal decay. Furthermore, gating techniques have been developed to compensate for respiratory displacement during the relatively long acquisition on the order of several minutes in the lung (6). This very appealing approach of UTE MRI has first shown impressive advantages in displaying intrapulmonary airways in cystic fibrosis (7).In the detection of pulmonary nodules, Ohno et al (8) demonstrated similar sensitivity and specificity to those of standard-dose and low-dose CT in a nonscreening population with pulmonary nodules using UTE on a 3-T system in the order of 93.0% sensitivity and 99.6% specificity. In subsequent work (9), the group demonstrated that the intermethod agreement for morphologic assessment of solid, part-solid, and entirely subsolid pulmonary nodules using UTE compared with CT as the standard of reference was within the same range as interreader agreement with CT. Thus, for the first time, classification of pulmonary nodules using MRI with respect to established malignancy criteria (developed using CT) has become possible. Of note, MRI systematically underestimated nodule diameters (solid and subsolid) by approximately 1–2 mm (9).In this issue of Radiology, Ohno et al (10) extend their previous experience and translate MRI-derived nodule characteristics into a decision on management. A total of 205 patients undergoing lung cancer screening were prospectively enrolled. Participants underwent UTE MRI at 3 T in conjunction with standard-dose and low-dose CT. As a great strength of the study, the consensus of two radiologists not otherwise involved with the study data set on the standard-dose CT scans was used as the reference standard.Another advantage of this study is that the authors did not hesitate to include subsolid nodules, an entity of great prevalence in screening populations but which is often excluded from recent studies. Because these nodules are more difficult to evaluate visually and computationally, there is lower interreader agreement in the assessment of diameter or volume. Of the 1037 nodules evaluated in the study by Ohno et al, a substantial number were part-solid (n = 132) and purely ground-glass (n = 182) nodules. The inclusion of subsolid nodules resulted in lower reader agreement and sensitivity for UTE MRI compared with low-dose CT. We can conclude that ground-glass components are more difficult to detect with UTE MRI than with CT. Still, the intermethod agreement for Lung-RADS classification was almost perfect, even for UTE MRI versus standard-dose CT, with a κ value of 0.92. The κ value for agreement between low-dose versus standard-dose CT was 0.97. In other words, intermethod agreement was not different from interreader agreement using a given single modality.Combining the results of the current study and prior work on this topic, I conclude that nodule management based on radiation-free UTE MRI at 3 T appears to be similar (noninferior) to that of low-dose CT in a lung cancer screening setting. Of course, I would like to see these results validated in a larger multicenter trial.Unfortunately, UTE sequences are not always available; sequence performance may vary by vendor. In addition, postprocessing of MRI data is far less developed compared with CT data, so computer-aided detection, segmentation and volumetry, and malignancy estimation (radiomics) are not available for MRI. However, it seems very likely that UTE sequences will shortly find their way into MRI protocols for the lung for structural lung assessment, not only for nodules but also for inflammatory and airway disease.Disclosures of conflicts of interest: M.O.W. Grants from Boehringer Ingelheim Pharma, Vertex Pharmaceuticals.References1. Wielpütz MO. Making Contrast Material Obsolete: Functional Lung Imaging with MRI. Radiology 2020;296(1):200–201. Link, Google Scholar2. Ohno Y, Koyama H, Yoshikawa T, et al. Three-way comparison of whole-body MR, coregistered whole-body FDG PET/MR, and integrated whole-body FDG PET/CT imaging: TNM and stage assessment capability for non–small cell lung cancer patients. Radiology 2015;275(3):849–861. Link, Google Scholar3. Allen BD, Schiebler ML, Sommer G, et al. Cost-effectiveness of lung MRI in lung cancer screening. Eur Radiol 2020;30(3):1738–1746. Crossref, Medline, Google Scholar4. Biederer J, Schoene A, Freitag S, Reuter M, Heller M. Simulated pulmonary nodules implanted in a dedicated porcine chest phantom: sensitivity of MR imaging for detection. Radiology 2003;227(2):475–483. Link, Google Scholar5. Sommer G, Tremper J, Koenigkam-Santos M, et al. Lung nodule detection in a high-risk population: comparison of magnetic resonance imaging and low-dose computed tomography. Eur J Radiol 2014;83(3):600–605. Crossref, Medline, Google Scholar6. Wielpütz MO, Triphan SMF, Ohno Y, Jobst BJ, Kauczor HU. Outracing Lung Signal Decay - Potential of Ultrashort Echo Time MRI. Rofo 2019;191(5):415–423. Crossref, Medline, Google Scholar7. Dournes G, Grodzki D, Macey J, et al. Quiet Submillimeter MR Imaging of the Lung Is Feasible with a PETRA Sequence at 1.5 T. Radiology 2015;276(1):258–265. Link, Google Scholar8. Ohno Y, Koyama H, Yoshikawa T, et al. Standard-, Reduced-, and No-Dose Thin-Section Radiologic Examinations: Comparison of Capability for Nodule Detection and Nodule Type Assessment in Patients Suspected of Having Pulmonary Nodules. Radiology 2017;284(2):562–573. Link, Google Scholar9. Wielpütz MO, Lee HY, Koyama H, et al. Morphologic Characterization of Pulmonary Nodules With Ultrashort TE MRI at 3T. AJR Am J Roentgenol 2018;210(6):1216–1225. Crossref, Medline, Google Scholar10. Ohno Y, Takenaka D, Yoshikawa T, et al. Efficacy of ultrashort echo time pulmonary MRI for lung nodule detection and Lung-RADS classification. Radiology 2022;302(3):697–706. Link, Google ScholarArticle HistoryReceived: Oct 5 2021Revision requested: Oct 14 2021Revision received: Oct 17 2021Accepted: Oct 21 2021Published online: Nov 30 2021Published in print: Mar 2022 FiguresReferencesRelatedDetailsCited ByMRI Compared with Low-Dose CT for Incidental Lung Nodule Detection in COPD: A Multicenter TrialQian Li, Lin Zhu, Oyunbileg von Stackelberg, Simon M. F. Triphan, Jürgen Biederer, Oliver Weinheimer, Monika Eichinger, Claus F. Vogelmeier, Rudolf A. Jörres, Hans-Ulrich Kauczor, Claus P. Heußel, Bertram J. Jobst, Mark O. Wielpütz, Stefan AndreasRobert BalsJürgen BehrKathrin KahnertThomas BahmerBurkhard BewigRalf EwertBeate StubbeJoachim H. FickerManfred WagnerChristian GrohéMatthias HeldWolfgang GesierichFelix HerthMichael KreuterFanziska TrudzinskiAnne-Marie KirstenHenrik WatzRembert KoczullaJuliane KronsbeinCornelia Kropf-SanchenAntonia Sassman-SchwedaMichael PfeiferWinfried J. RanderathWerner SeegerMichael StudnickaChristian TaubeHartmut TimmermannPeter AlterBernd SchmeckClaus VogelmeierTobias WelteHubert Wirtz27 April 2023 | Radiology: Cardiothoracic Imaging, Vol. 5, No. 2MRI versus CT in Detecting Pulmonary NodulesArosh S. Perera Molligoda Arachchige, 21 June 2022 | Radiology, Vol. 304, No. 3Lung Nodules Missed in Initial Staging of Breast Cancer Patients in PET/MRI—Clinically Relevant?KaiJannusch, Nils MartinBruckmann, Charlotte JohannaGeuting, JannaMorawitz, FredericDietzel, ChristophRischpler, KenHerrmann, Ann-KathrinBittner, OliverHoffmann, SvjetlanaMohrmann, Harald H.Quick, LaleUmutlu, GeraldAntoch, JulianKirchner2022 | Cancers, Vol. 14, No. 14Recommended Articles Subsolid Lung Nodules: Potential for OverdiagnosisRadiology2019Volume: 293Issue: 2pp. 449-450Radiomics and Computerized Analysis of CT Images: Looking ForwardRadiology: Cardiothoracic Imaging2020Volume: 2Issue: 6Disease Severity Scoring for COVID-19: A Welcome Semiquantitative Role for Chest RadiographyRadiology2021Volume: 302Issue: 2pp. 470-472Use of MRI to Measure Bronchial Inflammation in Cystic FibrosisRadiology2019Volume: 294Issue: 1pp. 197-198Lung-RADS: Pushing the LimitsRadioGraphics2017Volume: 37Issue: 7pp. 1975-1993See More RSNA Education Exhibits Role Of Radiology In Addressing The Challenge Of Lung Cancer After Lung Transplantation.Digital Posters2021Beyond Idiopathic In Pulmonary Fibrosis: Genetics, Immunology And MoreDigital Posters2021Lung Cancer Screening Beyond the Nodules: Smoking-related Lung DiseaseDigital Posters2020 RSNA Case Collection Pulmonary manifestations of cystic fibrosisRSNA Case Collection2020Diffuse idiopathic pulmonary neuroendocrine cell hyperplasiaRSNA Case Collection2020Bleomycin induced lung toxicityRSNA Case Collection2021 Vol. 302, No. 3 Metrics Altmetric Score PDF download