Standardized Evaluation of Small Renal Masses Using the MRI Clear Cell Likelihood Score

HomeRadiologyVol. 303, No. 3 PreviousNext Reviews and CommentaryFree AccessEditorialStandardized Evaluation of Small Renal Masses Using the MRI Clear Cell Likelihood ScoreAchille Mileto , Theodora A. PotretzkeAchille Mileto , Theodora A. PotretzkeAuthor AffiliationsFrom the Department of Radiology, Mayo Clinic, 200 First St SW, Rochester, MN 55905.Address correspondence to A.M. (e-mail: [email protected]).Achille Mileto Theodora A. PotretzkePublished Online:Mar 15 2022https://doi.org/10.1148/radiol.220054MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In See also the article by Schieda et al in this issue.Achille Mileto, MD, is an abdominal radiologist and senior associate consultant in the division of abdominal radiology at the Mayo Clinic (Rochester, Minn). His main clinical interests include the imaging evaluation of gastrointestinal, hepatobiliary and genitourinary diseases. His translational interests revolve around the development and clinical implementation of advanced multimodality imaging techniques to improve lesion detection and treatment response evaluation in abdominal imaging and oncology.Download as PowerPointOpen in Image Viewer Theodora A. Potretzke, MD, is an assistant professor of radiology at the Mayo Clinic (Rochester, Minn) working in the US and abdominal imaging divisions. Her areas of subspecialization are diagnostic genitourinary imaging and US-guided intervention. Clinical and research interests include renal tumor diagnosis and pathologic analysis, diagnosis of upper tract urothelial tumors, cystic renal disease, periprocedural considerations in image-guided biopsy, and clinical implementation of artificial intelligence algorithms in radiology.Download as PowerPointOpen in Image Viewer Approximately 70% of renal cell carcinomas (RCCs) are diagnosed incidentally in asymptomatic patients in the context of cross-sectional imaging performed for unrelated reasons (1). Consequently, most newly detected renal masses are small (≤4 cm) and categorized as small renal masses (stage T1a). This has resulted in so-called stage migration, wherein a larger proportion of new patients with RCC present with stage I disease as compared with earlier data. This stage migration has led to a reconsideration of treatment counseling. Currently, both the American Urologic Association and the American Society of Clinical Oncology include active surveillance as an initial management option for small renal masses (2,3). This is in part predicated on the knowledge that approximately 15%–20% of small renal masses are benign. While the remainder are malignant, only a minority of these will show the aggressive histologic features most associated with disease progression, metastasis, and death (dedifferentiation, histologic necrosis, sarcomatoid phenotype). For example, based on a large recent series of resected renal tumors, of those measuring 4 cm or less, 86% were malignant but only 23% were classified as aggressive, with the remainder being classified as indolent malignant or benign (4). Counseling on active surveillance also considers patient factors (eg, age and comorbidities), as well as risks associated with treatment and patient preference.Multiphasic contrast-enhanced CT and MRI provide important information for patient counseling and treatment decision making. The most well-established imaging feature associated with the risk of malignancy and aggressiveness is tumor size (4–6). Anatomic tumor complexity is also important to determine the feasibility of treatment approaches and the risk of complications, whether that be for partial nephrectomy or percutaneous ablation. In those who are candidates for active surveillance, other features associated with more aggressive malignant histology would be useful to identify. In some, but not all, centers, a renal mass biopsy is pursued in active surveillance of renal masses so histopathology can be factored into risk assessment. At other centers, if active surveillance is planned regardless of histology, then biopsy is deferred. In those instances, growth rate in follow-up is relied on to reflect tumor behavior. Still, given that histologic subtype is an independent predictor of outcomes for patients with RCC (7,8), recognition of imaging features associated with the more commonly aggressive clear cell RCC histology (as opposed to papillary or chromophobe) may be clinically useful.A multidisciplinary team of abdominal radiologists and urologists from the University of Texas Southwestern (Dallas) developed a standardized algorithm to identify clear cell RCC using MRI, referred to as the clear cell likelihood score (ccLS) (9). The ccLS designates the likelihood that a solid renal mass is a clear cell RCC using a five-point Likert scale (1, very unlikely; 2, unlikely; 3, intermediate likelihood [indeterminant findings]; 4, likely; and 5, very likely). The ccLS algorithm has undergone a few updates since its initial inception. The current version number is 2.0. The ccLS workflow provides a standardized analytic technique for solid renal mass evaluation that closely reflects the routine viewing approach of a subspecialty abdominal or genitourinary radiologist, including (a) eligibility criteria, ensuring the absence of macroscopic fat and at least mild (defined as 25%) contrast enhancement; and (b) major criteria, assessing signal on T2-weighted MRI scans, corticomedullary contrast-enhancement degree, and presence of intralesion microscopic fat. These steps are then followed by assessment for less-appreciated ancillary features, such as restricted diffusion, segmental contrast enhancement inversion, and arterial-to-delayed contrast enhancement ratio. This workflow has notable potential for offering a standardized approach across different practices. In particular, it can aid radiologists whose subspecialty is not in abdominal imaging when they are assessing small solid renal masses with MRI.Performance of the ccLS algorithm has been evaluated with reported sensitivity and specificity of 78%–89% and 58%–80%, respectively, with positive and negative predictive values of about 85% (9). Performance of the ccLS algorithm, particularly its positive predictive value, has been reported to increase with higher ccLS scores, ranging from 5% for ccLS 1 lesions to 93% for ccLS 5 lesions (9). This improved performance with higher ccLS scores likely reflects the beneficial effect of greater tumor dimension (ie, more readily apparent tumoral features) on ccLS category assignment. Knowledge of the performance of the ccLS algorithm outside its institution of origin, however, has been limited until now.In this issue of Radiology, a multi-institutional group of experts in the field of uroradiology across five different institutions in Canada and the United States has validated the performance of the ccLS algorithm in a retrospective review of 241 patients with 250 small solid renal masses (10). This study by Schieda et al provides a powerful substratum for the adoption of the ccLS algorithm in the clinical practice of radiology, with sensitivity, specificity, and positive and negative predictive values of 75%, 78%, 76%, and 88%, respectively (10). It is notable that the algorithm’s negative predictive value of a ccLS of 2 or less (the ability to identify tumors that are not clear cell RCC if ccLS ≤ 2) is better than the positive predictive value of a ccLS of 4 or higher. This finding could be interpreted to reflect the more recognizable features of papillary RCC, which were the majority of masses with a ccLS of 2 or less (63%) (10). Still, this multi-institutional work substantiates the prior observations of a gradual increase of positive predictive values of the ccLS algorithm with higher ccLS scores.Another encouraging finding from this work is the moderate interobserver agreement (κ = 0.58) among 10 radiologists with varying levels of experience (range, 5–30 years). This result is in line with that of other radiologic reporting systems currently used in clinical practice, such as the Prostate Imaging Reporting and Data System, or PI-RADS. Of note, the favorable interradiologist agreement obtained with minimal prior exposure of the different site participants to the ccLS algorithm suggests that this system could be used across different radiology practices and by radiologists with varying levels of training or clinical experience.The use of the ccLS algorithm for solid renal mass evaluation may have implications for the daily clinical practice of radiology. To cite a few potential benefits of this algorithm, the ccLS can promote standardized renal mass MRI scan evaluation and may contribute additional information for counseling discussions between clinicians and patients when active surveillance is being considered. In particular, the stronger consideration of active surveillance for small solid renal masses categorized as ccLS 1 or ccLS 2 may enable us to avoid overtreatment of lesions that are rarely clear cell RCCs while also minimizing the number of potential aggressive renal cancers managed with watchful waiting that lack a final pathology-based diagnosis. Conversely, since approximately 90% of solid renal masses with a ccLS of 4 or 5 are estimated to be malignant (of these, about 85% are deemed to represent clear cell RCCs), curative treatment options could be selected with more confidence for lesions within these categories, if appropriate. A ccLS 3 assignment carries a high degree of diagnostic uncertainty, suggesting that these lesions are “indeterminate,” and renal mass biopsy may be considered more often in some practices if it would impact management decisions.Areas of potential intricacy include the risk of equating ccLS 1 or 2 with indolent behavior, as non–clear cell histology can be aggressive or, in contrast, equating ccLS 4 or 5 with aggressive behavior when some clear cell RCCs are indolent and may be appropriate for active surveillance. It also remains to be determined whether the ccLS can outperform other variables that already perform well in risk stratification for small renal masses, such as size and growth rate. Additionally, requiring MRI for renal mass evaluation may represent a major change in the paradigm of renal mass imaging considering the limited access to MRI at some locations and as multiphasic CT often suffices to provide the required clinical information.Some future directions may be considered for the ccLS algorithm. One future direction may be to account for the diagnostic challenge often posed by oncocytomas and oncocytic-spectrum lesions (including hybrid oncocytic-chromophobe neoplasms), which often do not show any specific diagnostic feature and are indistinguishable from RCCs, particularly the chromophobe subtype. It is probable that these lesions are the culprit of many ccLS 4 or 5 false-positive findings. A feature that may help separate oncocytomas from chromophobe RCC is restricted diffusion that is seen more often in the latter. Inclusion within the ccLS algorithm of a specific recommendation for additional imaging with technetium 99m sestamibi SPECT/CT could be considered whenever a lesion shows imaging findings compatible with oncocytomas but does not restrict diffusion. This strategy may help further stratify patients according to risk.Another potential future direction for the ccLS algorithm is identification of other aggressive features, such as the sarcomatoid phenotype—an RCC feature not designated as a distinct histotype but that can be observed with any RCC subtype. The sarcomatoid phenotype carries the most dismal prognosis due to the high propensity for metastasis and resistance to systemic targeted therapies. High-throughput extraction of mineable image features based on radiomics techniques and computational augmentation enabled by machine learning may expand the MRI armamentarium, opening a new dimension into earlier identification of epigenetic signatures of tumor biology and somatic evolution of these aggressive neoplasms. In addition, future studies are likely required to better determine whether and to what extent MRI settings affect use and performance of the ccLS algorithm. For instance, the impact that varying spoiled gradient-echo flip angles, differences in diffusion-weighted b values, or contrast agent bolus timing have on tumoral features. Also, more experience is needed to characterize the impact that radiologists’ learning curve, standardization of the lexicon, and criteria have on ccLS performance.In conclusion, the multi-institutional external validation of the ccLS algorithm represents a considerable step toward the standardized use of MRI in the evaluation of solid renal masses. Like other radiologic reporting systems used in clinical practice, the clinical implementation of the ccLS algorithm may render a powerful framework with which to assess small solid renal masses using MRI. Clinical implementation of the ccLS algorithm has the potential to yield standardized diagnostic evaluation of solid renal masses across various imaging facilities, aid management, and support informed decision making.Disclosures of conflicts of interest: A.M. Consulting fees from Bayer Healthcare. T.A.P. No relevant relationships.References1. Nguyen MM, Gill IS, Ellison LM . The evolving presentation of renal carcinoma in the United States: trends from the Surveillance, Epidemiology, and End Results program. J Urol 2006;176(6 Pt 1):2397–2400.discussion 2400. Crossref, Medline, Google Scholar2. Campbell S, Uzzo RG, Allaf ME, et al . Renal Mass and Localized Renal Cancer: AUA Guideline. J Urol 2017;198(3):520–529. Crossref, Medline, Google Scholar3. Finelli A, Ismaila N, Bro B, et al . Management of Small Renal Masses: American Society of Clinical Oncology Clinical Practice Guideline. J Clin Oncol 2017;35(6):668–68.[Published correction appears in J Clin Oncol 2017;35(10):1141.]. Crossref, Medline, Google Scholar4. Bhindi B, Thompson RH, Lohse CM, et al . The Probability of Aggressive Versus Indolent Histology Based on Renal Tumor Size: Implications for Surveillance and Treatment. Eur Urol 2018;74(4):489–497. Crossref, Medline, Google Scholar5. Frank I, Blute ML, Cheville JC, Lohse CM, Weaver AL, Zincke H . Solid renal tumors: an analysis of pathological features related to tumor size. J Urol 2003;170(6 Pt 1):2217–2220. Crossref, Medline, Google Scholar6. Thompson RH, Kurta JM, Kaag M, et al . Tumor size is associated with malignant potential in renal cell carcinoma cases. J Urol 2009;181(5):2033–2036. Crossref, Medline, Google Scholar7. Lohse CM, Gupta S, Cheville JC . Outcome prediction for patients with renal cell carcinoma. Semin Diagn Pathol 2015;32(2):172–183. Crossref, Medline, Google Scholar8. Leibovich BC, Lohse CM, Crispen PL, et al . Histological subtype is an independent predictor of outcome for patients with renal cell carcinoma. J Urol 2010;183(4):1309–1315. Crossref, Medline, Google Scholar9. Pedrosa I, Cadeddu JA . How We Do It: Managing the Indeterminate Renal Mass with the MRI Clear Cell Likelihood Score. Radiology 2022;302(2):256–269. Link, Google Scholar10. Schieda N, Davenport MS, Silverman SG, et al . Multicenter Evaluation of Multiparametric MRI Clear Cell Likelihood Scores in Solid Indeterminate Small Renal Masses. Radiology 2022;303(3):590–599. Link, Google ScholarArticle HistoryReceived: Jan 7 2022Revision requested: Jan 13 2022Revision received: Jan 13 2022Accepted: Jan 19 2022Published online: Mar 15 2022Published in print: June 2022 FiguresReferencesRelatedDetailsCited BySystematic review and meta-analysis of multiparametric MRI clear cell likelihood scores for classification of small renal massesJunTian, FeixiangTeng, HongtaoXu, DongliangZhang, YinxiuChi, HuZhang2022 | Frontiers in Oncology, Vol. 12Accompanying This ArticleMulticenter Evaluation of Multiparametric MRI Clear Cell Likelihood Scores in Solid Indeterminate Small Renal MassesMar 15 2022RadiologyRecommended Articles Active Surveillance of Renal Masses: The Role of RadiologyRadiology2021Volume: 302Issue: 1pp. 11-24Multicenter Evaluation of Multiparametric MRI Clear Cell Likelihood Scores in Solid Indeterminate Small Renal MassesRadiology2022Volume: 303Issue: 3pp. 590-599Case 297: Mucinous Adenocarcinoma of the ProstateRadiology2021Volume: 302Issue: 1pp. 238-242Invited Commentary on “Differentiation of Solid Renal Tumors with Multiparametric MR Imaging”RadioGraphics2017Volume: 37Issue: 7pp. 2042-2044Using Radiomics for Risk Stratification: Where We Need to GoRadiology2021Volume: 302Issue: 2pp. 435-437See More RSNA Education Exhibits Imaging Features of Cystic Renal Lesions in the Era of the 2016 WHO ClassificationDigital Posters2019Complex Cystic and Solid Masses of the Breast: Friend or Foe?Digital Posters2019What’s New in Classification, Pathology, Imaging Findings, and Management of Renal Cell Carcinomas? 2022 Update for RadiologistsDigital Posters2022 RSNA Case Collection Papillary Renal Cell CarcinomaRSNA Case Collection2020HepatoblastomaRSNA Case Collection2021Primary Dural Lymphoma RSNA Case Collection2022 Vol. 303, No. 3 Metrics Altmetric Score PDF download