Risk factors for bleeding hepatocellular adenoma in a United States cohort.

BACKGROUND & AIMS: Known risk factors for hepatocellular adenoma (HCA) bleeding are size >5 cm, growth rate, visible vascularity, exophytic lesions, ß-catenin and Sonic Hedgehog activated HCAs. Most studies are based on European cohorts. The objective of this study is to identify additional risk factors for HCA bleeding in a US cohort. METHODS: Retrospective chart review was performed on patients diagnosed with HCA on magnetic resonance imaging (n = 184) at an academic tertiary institution. Clinical, pathological, and imaging data were collected. Primary outcomes measured were HCA bleeding and malignancy. Statistical analysis was performed with SAS 9.4 using Chi-Square, Fisher's exact test, sample t test, non-parametric Wilcoxon test, and logistic regression. RESULTS: After excluding patients whose pathology showed focal nodular hyperplasia and non-adenoma lesions, follow-up data were available for 167 patients. 16% experienced microscopic or macroscopic bleeding and 1.2% had malignancy. HCA size predicted bleeding (P < .0001) and no patients with lesion size <1.8 cm bled. In unadjusted analysis, hepatic adenomatosis (≥10 lesions) trended towards 2.8-fold increased risk of bleeding. Of patients with a single lesion that bled, 77% bled from a lesion >5 cm. In patients with multiple HCAs that bled, 50% bled from lesions <5 cm. In patients with multiple adenomas, size (P = .001) independently predicted bleeding and hepatic steatosis trended towards increased risk of bleeding (P = .05). CONCLUSIONS: In a large US cohort, size predicted increased risk of HCA bleeding while hepatic adenomatosis trended towards increased risk of bleeding. In patients with multiple HCAs, size predicted bleeding and hepatic steatosis trended toward increased risk of bleeding.

Regional Variations in Diagnostic Radiology Residency Stipends Are Accentuated When Correlated With Cost of Living.

OBJECTIVES: To determine and compare the relative value of diagnostic radiology resident stipends when adjusted for regional cost of living. METHODS: The ACGME database was queried for a list of accredited allopathic diagnostic radiology residency programs for academic year 2017-2018. Stipend information for R1 positions (post-graduate year 2) was identified through each program's website. Data was grouped and analyzed by city, state, and geographical region. Stipends were then correlated with the 2017 annual region-specific average cost of living index (COLI). RESULTS: There were 194 programs identified, of which 118 (60.8%) were analyzed after exclusions for lack of stipend or corresponding COLI data. The average annual stipend was $57,161±$4,242 (range, $49,547-$72,000). The COLI-adjusted value was $51,357±$9,927 (range, $26,915-$68,827). The average difference between stipend and cost-ofliving adjusted value was -$5,804±$12,610 (range, -$40,953-$10,958), corresponding to an average -9.1% stipend value loss (range, -58.1%-21.7%, P = .00000004). Value loss in large metropolitan cities was as high as 53.8%. DISCUSSION: Financial issues are one of the biggest challenges faced by trainees and have been correlated with increased stress as well as poor academic performance. The pressures of debt can also play a significant role in ultimate subspecialty career choice. Discrepancies between resident stipends, which are not adjusted based on the local COLI, are exacerbated by large regional variations in cost of living. Residency applicants should not discount regional cost of living when deciding where to train, and training programs should consider cost of living when setting stipend levels for their trainees.

Extrahepatic Vascular Parasitization by a Hepatocellular Carcinoma.

Hepatocellular carcinoma is the most common primary hepatic malignancy. For patients not amenable to surgical treatment, transarterial chemoembolization is a viable therapeutic alternative. Extrahepatic collateral arterial supply to the tumor may occur in a variety of scenarios and timely detection of this phenomenon is of fundamental importance to achieve optimal outcomes and response to treatment. This report presents a case of hepatocellular carcinoma that was supplied mainly by a parasitized right phrenic artery and was only successfully treated once this was identified. Further discussion of extrahepatic collateral arterial supply is also presented.

Comparison of Cone-Beam Tomography and Cross-Sectional Imaging for Volumetric and Dosimetric Calculations in Resin Yttrium-90 Radioembolization.

PURPOSE: To compare the use of cone-beam computed tomography versus contrast-enhanced computed tomography (CT) and magnetic resonance imaging (MRI) in the calculation of liver volume and planned dose for yttrium-90 radioembolization. MATERIALS AND METHODS: The study retrospectively assessed 47 consecutive patients who underwent resin Y-90 radioembolization consecutively over a 2-year period at a single center. Volume calculation software was used to determine perfused lobar liver volumes from cone-beam CT (CBCT) images obtained during mapping angiography. CBCT-derived volumes were compared with perfused lobar volume derived from contrast-enhanced CT and MRI. Nominal activities as determined by the SIR-Spheres Microspheres Activity Calculator were similarly calculated and compared using both CBCT and conventionally acquired volumes. RESULTS: A total of 82 hepatic lobes were assessed in 47 patients. The mean percentage difference between combined CT-MRI- and CBCT-derived calculated lobar volumes was 25.3% (p = 0.994). The mean percentage difference in calculated dose between the two methods was 21.8 ± 24.6% (p = 0.42). Combined left and right lobar CT-derived dose difference was less than 10% in 22 lobes, between 10 and 25% in 20 lobes, between 25 and 50% in 13 lobes and greater than 50% in 5 lobes. Combined left and right lobar MRI-derived dose difference was less than 10% in 11 lobes, between 10 and 25% in 7 lobes, between 25 and 50% in 2 lobes and greater than 50% in 1 lobe. CONCLUSIONS: Although volume measurements derived from CT/MRI did not differ significantly from those derived from CBCT, variability between the two methods led to large and unexpected differences in calculated dose.