Search

Background: Multiparametric magnetic resonance imaging (mpMRI) undoubtedly affects the diagnosis and treatment of localized prostate cancer (CaP). However, clinicians need a better understanding of its accuracy and limitations in detecting individual CaP foci to optimize management., Objective: To determine the per-lesion detection rate for CaP foci by mpMRI and identify predictors of tumor detection., Design, Setting, and Participants: We carried out a retrospective analysis of a prospectively managed database correlating lesion-specific results from mpMRI co-registered with whole-mount pathology (WMP) prostatectomy specimens from June 2010 to February 2018. Participants include 588 consecutive patients with biopsy-proven CaP undergoing 3-T mpMRI before radical prostatectomy at a single tertiary institution., Outcome Measurements and Statistical Analysis: We measured mpMRI sensitivity in detecting individual CaP and clinically significant (any Gleason score ≥7) CaP foci and predictors of tumor detection using multivariate analysis., Results and Limitations: The final analysis included 1213 pathologically confirmed tumor foci in 588 patients with primarily intermediate- (75%) or high-risk (12%) CaP. mpMRI detected 45% of all lesions (95% confidence interval [CI] 42-47%), including 65% of clinically significant lesions (95% CI 61-69%) and nearly 80% of high-grade tumors. Some 74% and 31% of missed solitary and multifocal tumors, respectively, were clinically significant. The majority of missed lesions were small (61.1% ≤1cm); 28.3% were between 1 and 2cm, and 10.4% were >2cm. mpMRI missed at least one clinically significant focus in 34% of patients overall, and in 45% of men with multifocal lesions. On multivariate analysis, smaller, low-grade, multifocal, nonindex tumors with lower prostate-specific antigen density were more likely to be missed. Limitations include selection bias in a prostatectomy cohort, lack of specificity data, an imperfect co-registration process, and uncertain clinical significance for undetected lesions., Conclusions: mpMRI detects less than half of all and less than two-thirds of clinically significant CaP foci. The moderate per-lesion sensitivity and significant proportion of men with undetected tumor foci demonstrate the current limitations of mpMRI., Patient Summary: Magnetic resonance imaging of the prostate before surgical removal for prostate cancer finds less than half of all individual prostate cancer tumors. Large, solitary, aggressive tumors are more likely to be visualized on imaging., (Published by Elsevier B.V.)

Objective: The objectives of our study were to compare MR elastography (MRE) based on gradient-recalled echo (GRE) imaging with spin-echo echo-planar imaging (SEEPI) and rapid fractional (RF)-GRE MRE sequences at 3 T in terms of liver stiffness (LS) and image quality and to evaluate the effect of liver R2* on image quality., Materials and Methods: Eighty-one patients underwent 3-T liver MRE with GRE, SE-EPI, and RF-GRE sequences performed in variable order in this study. LS and ROI areas on the LS 95% confidence maps were compared among the three sequences. The relationship between liver R2* and ROI area was investigated., Results: There was no significant difference in mean LS among the three sequences (p = 0.49). Mean ROI area was significantly larger for RF-GRE (18,213 ± 9292 [SD] mm 2 ) than for GRE (13,196 ± 8149 mm 2 ) and SE-EPI (12,896 ± 8656 mm 2 ) (p < 0.0001). Liver R2* was significantly higher among patients with one or more failed sequences (mean ± SD, 116 ± 76 s -1 ) than for patients with no failed sequences (59 ± 26 s -1 ) (p = 0.001). Technical failure rates were 10% (8/81), 4% (3/81), and 2% (2/81) for GRE, SE-EPI, and RF-GRE, respectively. Among patients with iron overload (R2* ≥ 100 s -1 ), there was a trend toward larger ROI area for SE-EPI (p = 0.09)., Conclusion: SE-EPI-and RF-GRE-based MRE sequences provide equivalent measures of LS compared with GRE-based MRE, and both have lower technical failure rates. The RF-GRE sequence yielded the largest measurable area of LS. Among patients with iron overload, there was a trend toward larger measurable area of LS for the SE-EPI sequence.

Purpose: We present a method for generating a T2 MR-based probabilistic model of tumor occurrence in the prostate to guide the selection of anatomical sites for targeted biopsies and serve as a diagnostic tool to aid radiological evaluation of prostate cancer., Materials and Methods: In our study, the prostate and any radiological findings within were segmented retrospectively on 3D T2-weighted MR images of 266 subjects who underwent radical prostatectomy. Subsequent histopathological analysis determined both the ground truth and the Gleason grade of the tumors. A randomly chosen subset of 19 subjects was used to generate a multi-subject-derived prostate template. Subsequently, a cascading registration algorithm involving both affine and non-rigid B-spline transforms was used to register the prostate of every subject to the template. Corresponding transformation of radiological findings yielded a population-based probabilistic model of tumor occurrence. The quality of our probabilistic model building approach was statistically evaluated by measuring the proportion of correct placements of tumors in the prostate template, i.e., the number of tumors that maintained their anatomical location within the prostate after their transformation into the prostate template space., Results: Probabilistic model built with tumors deemed clinically significant demonstrated a heterogeneous distribution of tumors, with higher likelihood of tumor occurrence at the mid-gland anterior transition zone and the base-to-mid-gland posterior peripheral zones. Of 250 MR lesions analyzed, 248 maintained their original anatomical location with respect to the prostate zones after transformation to the prostate., Conclusion: We present a robust method for generating a probabilistic model of tumor occurrence in the prostate that could aid clinical decision making, such as selection of anatomical sites for MR-guided prostate biopsies.

Objective: The objective of our study was to determine the clinical and MRI characteristics of clinically significant prostate cancer (PCA) (Gleason score ≥ 3 + 4) in men with Prostate Imaging Reporting and Data System version 2 (PI-RADSv2) category 3 transition zone (TZ) lesions., Materials and Methods: From 2014 to 2016, 865 men underwent prostate MRI and MRI/ultrasound (US) fusion biopsy (FB). A subset of 90 FB-naïve men with 96 PI-RADSv2 category 3 TZ lesions was identified. Patients were imaged at 3 T using a body coil. Images were assigned a PI-RADSv2 category by an experienced radiologist. Using clinical data and imaging features, we performed univariate and multivariate analyses to identify predictors of clinically significant PCA., Results: The mean patient age was 66 years, and the mean prostate-specific antigen density (PSAD) was 0.13 ng/mL 2 . PCA was detected in 34 of 96 (35%) lesions, 14 of which (15%) harbored clinically significant PCA. In univariate analysis, DWI score, prostate volume, and PSAD were significant predictors (p < 0.05) of clinically significant PCA with a suggested significance for apparent diffusion coefficient (ADC) and prostate-specific antigen value (p < 0.10). On multivariate analysis, PSAD and lesion ADC were the most important covariates. The combination of both PSAD of 0.15 ng/mL 2 or greater and an ADC value of less than 1000 mm 2 /s yielded an AUC of 0.91 for clinically significant PCA (p < 0.001). If FB had been restricted to these criteria, only 10 of 90 men would have undergone biopsy, resulting in diagnosis of clinically significant PCA in 60% with eight men (9%) misdiagnosed (false-negative)., Conclusion: The yield of FB in men with PI-RADSv2 category 3 TZ lesions for clinically significant PCA is 15% but significantly improves to 60% (AUC > 0.9) among men with PSAD of 0.15 ng/mL 2 or greater and lesion ADC value of less than 1000 mm 2 /s.

Purpose To determine the diagnostic yield of in-bore 3-T magnetic resonance (MR) imaging-guided prostate biopsy and stratify performance according to Prostate Imaging Reporting and Data System (PI-RADS) versions 1 and 2. Materials and Methods This study was HIPAA compliant and institution review board approved. In-bore 3-T MR-guided prostate biopsy was performed in 134 targets in 106 men who (a) had not previously undergone prostate biopsy, (b) had prior negative biopsy findings with increased prostate-specific antigen (PSA) level, or (c) had a prior history of prostate cancer with increasing PSA level. Clinical, diagnostic 3-T MR imaging was performed with in-bore guided prostate biopsy, and pathology data were collected. The diagnostic yields of MR-guided biopsy per patient and target were analyzed, and differences between biopsy targets with negative and positive findings were determined. Results of logistic regression and areas under the curve were compared between PI-RADS versions 1 and 2. Results Prostate cancer was detected in 63 of 106 patients (59.4%) and in 72 of 134 targets (53.7%) with 3-T MR imaging. Forty-nine of 72 targets (68.0%) had clinically significant cancer (Gleason score ≥ 7). One complication occurred (urosepsis, 0.9%). Patients who had positive target findings had lower apparent diffusion coefficient values (875 × 10 -6 mm 2 /sec vs 1111 × 10 -6 mm 2 /sec, respectively; P < .01), smaller prostate volume (47.2 cm 3 vs 75.4 cm 3 , respectively; P < .01), higher PSA density (0.16 vs 0.10, respectively; P < .01), and higher proportion of PI-RADS version 2 category 3-5 scores when compared with patients with negative target findings. MR targets with PI-RADS version 2 category 2, 3, 4, and 5 scores had a positive diagnostic yield of three of 23 (13.0%), six of 31 (19.4%), 39 of 50 (78.0%), and 24 of 29 (82.8%) targets, respectively. No differences were detected in areas under the curve for PI-RADS version 2 versus 1. Conclusion In-bore 3-T MR-guided biopsy is safe and effective for prostate cancer diagnosis when stratified according to PI-RADS versions 1 and 2. © RSNA, 2016.

Purpose: To retrospectively determine whether hepatobiliary phase (HBP) sequence outperforms unenhanced T1-weighted imaging (uT1wI) in distinguishing the ablation margin (AM) from hepatocellular carcinoma (HCC) 24 h after thermoablation., Material and Methods: Ninety-one patients [mean age, 65.7 years; 68 M/23F] with 138 HCCs (>6 months follow-up) underwent pre- and postablation gadoxetate disodium-enhanced MRI. AM showed a hyperintense middle zone (MZ) surrounding central hypo- or hyperintense HCCs on uT1wI, and an intermediate-intense MZ encompassing central hypo- or hyperintense HCCs during HBP. The visible AM was defined as persistent MZ around HCCs, which were demarcated from MZ, or peripherally band encompassing MZ, which were not demarcated from HCC. The indefinite AM was defined as no demarcating HCCs from MZ. The ability to distinguish AM from HCC was classified as visible or indefinite on axial (ax)-uT1wI, ax-HBP, coronal (cor)-HBP, and combined all images. To investigate the AM visibility during HBP, significance of differences upon comparison of ax-uT1wI with combined images was analyzed. Preablation liver-tumor contrast ratio (LTCR) on ax-uT1wI and ax-HBP sequence is compared between the visible and indefinite AM., Results: The McNemar test demonstrated a significant increase (p < 0.05) in visible AM from ax-uT1wI (60), to ax-HBP (70), cor-HBP (79), and combined images (83). TLCR with visible AM was significantly higher than that with indefinite AM on ax-uT1wI (0.4 vs. 0.2, p = 0.001) and ax-HBP sequence (0.9 vs. 0.6, p = 0.004)., Conclusions: HBP sequence might have higher feasibility to distinguish AM from tumor than ax-uT1wI. The TLCR value in visible AM was higher than that in indefinite AM on both ax-uT1wI and ax-HBP sequences.

Purpose: To determine which clinical variables and gadoxetic acid disodium (Gd-EOB-DTPA)-enhanced magnetic resonance (MR) imaging features are associated with histologically proved hepatocellular adenoma (HCA) genotypic subtypes., Materials and Methods: In this institutional review board-approved and Health Insurance Portability and Accountability Act-compliant study, clinical information and MR images of 49 histologically proved HCAs from January 2002 to December 2013 (21 patients; mean age, 39 years; age range, 15-59 years) were retrospectively reviewed by two radiologists. Qualitative and quantitative imaging features, including the signal intensity ratio relative to liver in each phase, were studied. HCA tissues were stained with subtype-specific markers and subclassified by a pathologist. Clinical and imaging data were correlated with pathologic findings and compared by using Fisher exact or t test, with a Bonferroni correction for multiple comparisons., Results: Forty-nine HCAs were subclassified into 14 inflammatory, 20 hepatocyte nuclear factor (HNF)-1α-mutated, one β-catenin-activated, and 14 unclassified lesions. Intralesional steatosis was exclusively seen in HNF-1α-mutated lesions. Marked hyperintensity on T2-weighted images was seen in 12 of 14 (86%) inflammatory lesions compared with four of 21 (19%) HNF-1α-mutated, seven of 14 (50%) unclassified, and zero of one (0%) β-catenin-activated lesion. Two large lesions (one β-catenin-activated and one unclassified) transformed into hepatocellular carcinomas and were the only lesions to enhance with marked heterogeneity. In the hepatobiliary phase, all HCA subtypes were hypoenhancing compared with surrounding liver parenchyma, and they reached their nadir signal intensity by 10 minutes after the administration of contrast material before plateauing. HNF-1α-mutated lesions had the lowest lesion signal intensity ratio of 0.47 ± 0.09, compared with 0.73 ± 0.18 for inflammatory lesions (P = .0004), 0.82 for the β-catenin-activated lesion, and 0.73 ± 0.06 for the unclassified lesion (P = .00002)., Conclusion: In this study, all HCA subtypes were hypoenhancing at Gd-EOB-DTPA-enhanced MR imaging in the hepatobiliary phase and reached their nadir signal intensity at 10 minutes. HNF-1α-mutated lesions could be distinguished from other subtypes by having the lowest lesion signal intensity ratio.

Background & Aims: In patients with cirrhosis, hepatocellular carcinoma (HCC) is detected by ultrasound (US), computed tomography (CT), or magnetic resonance imaging (MRI); US is recommended for screening and surveillance. We performed a retrospective analysis of the abilities of these cross-sectional imaging modalities to detect HCC., Methods: We analyzed data from 638 consecutive adult patients with cirrhosis who received liver transplants within 6 months of imaging at a tertiary care institution. Imaging reports and serum alpha-fetoprotein levels were compared with results from pathology analysis of explants as the reference standard. Sensitivities of US, CT, and MRI were calculated overall and in defined size categories. False-positive imaging results and patient-based specificities were evaluated., Results: Of the 638 patients, 225 (35%) had HCC, confirmed by pathology analysis of liver explants. In 23 cases, the lesions were infiltrative or extensively multifocal. In the remaining 202 explants (337 numerable, discrete nodules), respective lesion-based sensitivities of US, CT, and MRI were 46%, 65%, and 72% overall and 21%, 40%, and 47% for small (<2 cm) HCC. The sensitivity of US increased with the availability of CT or MRI data (P = .049); sensitivity values were 62% and 85% for lesions 2-4 and ≥ 4 cm, respectively. Patient-based specificities of US, CT, and MRI were 96%, 96%, and 87%, respectively., Conclusions: US, CT, and MRI did not detect small HCC lesions with high levels of sensitivity, although CT and MRI provide substantial improvements over unenhanced US in patients with cirrhosis who received liver transplants., (Copyright © 2011 AGA Institute. Published by Elsevier Inc. All rights reserved.)

Objectives: To analyze all cases of nephrogenic systemic fibrosis (NSF) at our institution and to compare them with controls., Methods: After the institutional review board approval, 13 biopsy-proven NSF cases were identified. Ten cases had complete records and were compared in a case-control format with 10 age- and sex-matched, dialysis-dependent controls. Analyzed risk factors included single and cumulative gadolinium dose, medication and transplant history, and serum electrolytes at the time of gadolinium exposure., Results: There were 1.9% of dialysis-dependent, gadolinium-exposed patients who developed NSF. There was no difference in gadolinium dose, transplant history, or serum electrolytes. Seven of 10 cases and 3 of 10 controls were treated with erythropoietin (P = 0.13). At the time of NSF diagnosis, 7 of 10 cases were on immunosuppressive therapy. Two of 7 cases developed NSF only after immunosuppressive therapy was initiated. Two of 10 controls were on immunosuppressive therapy (P = 0.06)., Conclusions: All cases of NSF occurred in dialysis-dependent, gadolinium-exposed patients. Associations between immunosuppressive and erythropoietin therapies and NSF need further investigation.

The purpose of this work is to describe our experience with single-session percutaneous ethanol injection (PEI) under the guidance of 0.2-T open MRI for hepatocellular carcinomas (HCC) that were not suitable for ablation under ultrasound (US) or computed tomography (CT) guidance. None of the lesions (N = 7) were detectable on US. MRI was chosen over CT as the guidance modality because the nodules were located in the hepatic dome (N = 4) or were invisible on noncontrast CT (N = 3). All of the nodules were targeted successfully, and apparently complete tumor necrosis was achieved in six nodules (86%). During a follow-up of one to 41 months, only one patient developed local recurrence four months after PEI. MR-guided PEI is feasible and effective for treating HCC when other imaging guidance methods are not appropriate., ((c) 2005 Wiley-Liss, Inc.)

Objective: The objective of our study was to compare a breath-hold fat-suppressed fast-recovery fast spin-echo (FSE) T2-weighted sequence with a respiratory-triggered fat-suppressed FSE T2-weighted sequence to assess the effect on image quality and lesion detection and characterization in clinical hepatic imaging., Materials and Methods: Both the breath-hold fat-suppressed fast-recovery FSE and respiratory-triggered fat-suppressed FSE T2-weighted sequences were acquired in 46 patients. Two radiologists, blinded to clinical data, independently evaluated randomized images from both sequences. Qualitatively, images were graded on a 5-point scale for five different characteristics. The number and location of lesions were recorded. The confidence of detection and the confidence of characterization (solid vs nonsolid) were graded on a 5-point scale. A consensus review using radiology, clinical, and pathology data served as the standard. Receiver operating characteristic (ROC) curve analysis (area under the ROC curve [A(z)]) was used to compare each reviewer's interpretation against the consensus interpretation. Quantitative analysis was performed by calculating the liver signal-to-noise ratio (SNR), liver-to-spleen contrast-to-noise ratio (CNR), and lesion-to-liver CNR. Both one- and two-tailed Student's t tests were used to check for significance., Results: Qualitatively, both reviewers graded the breath-hold fat-suppressed fast-recovery FSE T2-weighted sequence better than the respiratory-triggered fat-suppressed FSE T2-weighted sequence on all five characteristics (p < 0.005). Of 78 lesions detected, 29 were characterized as solid; 47, nonsolid; and two, indeterminate. On ROC analysis, there were no significant differences between the breath-hold fat-suppressed fast-recovery FSE and respiratory-triggered fat-suppressed FSE T2-weighted sequences in lesion detection (A(z) reviewer 1, 0.77 and 0.83, respectively, [p = 0.12]; A(z) reviewer 2, 0.84 and 0.80, respectively [p = 0.12]) or in lesion characterization (A(z) reviewer 1, 0.86 and 0.92, respectively [p = 0.33]; A(z) reviewer 2, 0.90 and 0.91, respectively [p = 0.79]). Quantitatively, liver SNRs, spleen CNRs, and lesion CNRs (solid and nonsolid lesions) were significantly better on the breath-hold fat-suppressed fast-recovery FSE T2-weighted images than on the respiratory-triggered fat-suppressed FSE T2-weighted images (p < 0.005)., Conclusion: Breath-hold fat-suppressed fast-recovery FSE T2-weighted images were of better quality than respiratory-triggered fat-suppressed FSE T2-weighted images, and lesion detection and characterization were comparable.

Purpose: To compare the performance of dynamic gadolinium-enhanced and ferumoxides-enhanced MRI in the detection and characterization of hepatic lesions, on 1.5-T and 0.2-T magnets, Materials and Methods: In 41 patients (23 men, 18 women), 52 hepatic MR examinations were performed and retrospectively analyzed; 39 and 13 examinations were performed on 1.5-T and 0.2-T magnets, respectively. A total of 33 of 41 patients had known malignancies, and 31 of 33 patients had biopsy of at least one lesion. First, a combination of unenhanced T2-weighted sequences and gradient-echo T1-weighted sequences were performed. Then, dynamic gadolinium-enhanced (0.1 mmol/kg) T1 GRE sequences were obtained, followed by intravenous drip infusion of ferumoxides (10 micromol/kg). The T2-weighted sequences were then repeated. The unenhanced and gadolinium-enhanced images (the Gd set) were reviewed separately from the unenhanced and ferumoxides-enhanced images (the ferumoxides set) by two abdominal imagers. The reviewers were blinded to clinical history and reviewed the individual studies in each set randomly. Each detected lesion was scored on a five-point scale for characterization scores: nonsolid (1 or 2), indeterminate (3), or solid (4 or 5). A consensus review was then performed correlating all available pathology, imaging, clinical findings, and follow-up to act as a gold standard. Receiver-operating-characteristic (ROC) curves were generated and both area-under-the-curve (Az values) and sensitivity values were calculated. Significance of Az and sensitivity differences was assessed using standard Z-test and chi-square., Results: Of 270 lesions detected by consensus, 211 were on 1.5-T and 59 were on 0.2-T scanners. The accuracy (Az values) of lesion detection overall, of both readers, was greater for the ferumoxides set than for the Gd set (reader 1: 0.95 vs. 0.89 (P < 0.05); reader 2: 0.91 vs. 0.78 (P < 0.05)). Az values for both readers were greater on the ferumoxides set for both the 1.5-T scans and the 0.2-T scans. Out of 270 lesions detected, 231 were characterized by consensus review as solid (185) or nonsolid (46). There was a significant improvement in lesion characterization for both readers on the GD set compared with the ferumoxides set on both high and low field scanners (Az reader 1: 0.99 vs. 0.96 (P < 0.05); Az reader 2: 0.99 vs. 0.95 (P < 0.05)), respectively., Conclusion: At both 1.5-T and 0.2-T, ferumoxides-enhanced sequences were better for lesion detection, while gadolinium-enhanced sequences were better for lesion characterization, respectively., (Copyright 2004 Wiley-Liss, Inc.)

Purpose: To assess the utility of magnetic resonance cholangiopancreatography (MRCP) in preoperative mapping of biliary anatomy in adult-to-adult living related liver transplant (LRLT) donors., Materials and Methods: From 57 potential donors with preoperative MRCP, 27 cases (16 men, 11 women, age range 22-51 years, mean 37.2 years) underwent right lobe resection and had intraoperative cholangiography (IOC) for comparison. The MRCP and IOC reports were retrospectively reviewed in all 27 cases. The MRCP was performed on 1.5 Tesla MR magnets using breath-hold heavily T2-weighted sequences in axial/coronal thin sections, and variable-thickness rotating slabs. The accuracy of preoperative MRCP for biliary mapping in potential LRLT donors was analyzed compared to the IOC findings., Results: Of 27 donors, 26 (96.3%) had MRCP which showed adequate information of central intrahepatic biliary anatomy. Of these, 19 had normal bifurcation confirmed by IOC, and single biliary anastomosis was created in the recipient at transplantation. MRCP correctly predicted 17 of 19 normal cases (sensitivity for normals: 89.5%). In seven donors with variant biliary anatomy, two separate biliary anastomoses were performed in the recipient. MRCP correctly predicted five of seven variants (sensitivity for variants: 71.4%). Overall, MRCP had an accuracy 84.6% (22/26)., Conclusions: MRCP has potential in the preoperative assessment of nondilated bile ducts in LRLT donors, however further improvements are desired to increase its quality and accuracy., (Copyright 2004 Wiley-Liss, Inc.)

Objectives: The purpose of this study was to evaluate the 10-year overall survival and local tumor progression (LTP) of percutaneous radiofrequency ablation (RFA) for single nodular hepatocellular carcinoma (HCC) < 3 cm using a large longitudinal hospital registry and clinical factors associated with overall survival and LTP.Methods: A total of 467 newly diagnosed patients with single nodular HCC < 3 cm who underwent RFA as first-line therapy between January 2008 to December 2016 were analyzed. Overall survival and LTP were estimated using the Kaplan-Meier method. Cox regression and competing risks Cox regression analysis were performed to identify prognostic factors for overall survival and LTP, respectively.Results: The 5- and 10-year overall survival rates after RFA were 83.7% and 74.2%, respectively. LTP (hazard ratio (HR), 2.03; 95% confidence interval (CI), 1.19-3.47) was one of the important factors for overall survival after RFA. The 5- and 10-year LTP rates after RFA were 20.4% and 25.1%, respectively. Periportal location (subdistribution HR, 2.29; 95% CI, 1.25-4.21), subphrenic location (2.25, 1.34-3.86), size ≥ 1.5-< 2.0 cm (1.88, 1.05-3.39), and size ≥ 2.0 cm (2.10, 1.14-3.86) were independent factors for LTP.Conclusion: Ten-year therapeutic outcomes of percutaneous RFA as first-line therapy were excellent for single HCC < 3 cm. LTP was an important prognostic factor for overall survival after RFA. Periportal and subphrenic location of HCCs and tumor size were predictors for the development of LTP after RFA.Key Points: • Updated 10-year survival outcome of percutaneous radiofrequency ablation as first-line therapy for single hepatocellular carcinoma < 3 cm was higher than previously reported. • Local tumor progression was an important prognostic factor for overall survival after percutaneous radiofrequency ablation. • Periportal and subphrenic location of hepatocellular carcinomas and tumor size were predictors for the development of local tumor progression after percutaneous radiofrequency ablation. [ABSTRACT FROM AUTHOR]