Manifestations of Hepatic Focal Nodular Hyperplasia in Hepatobiliary Phase of Enhancement by Hepatocyte-specific Contrast Agent: Comparison With Pathologic Findings
Jiangfeng Xu, Liqing Zhang, Jianxun Li, Xiaojian Cui, Qingping Jiang, Yan ZhangResearch2021 · DOI
This study explores the diagnostic value and signal characteristics of hepatic focal nodular hyperplasia (FNH) in the hepatobiliary phase (HBP) of gadolinium-ethoxybenzyl-diethylenetriamine-pentaacetic acid (Gd-EOB-DTPA) enhanced MRI. By analyzing 43 pathologically proven lesions, the researchers correlated imaging manifestations with tissue-level findings. While most FNH lesions appear bright (iso-high intensity) in the HBP, a subset of "non-classic" lesions present with low-intensity signals, often due to high fat content (steatosis) or specific genetic markers like β-catenin.
Introduction
Focal Nodular Hyperplasia (FNH) is the second most common benign liver tumor. Distinguishing it from malignant tumors like hepatocellular carcinoma (HCC) is vital to avoid unnecessary surgical intervention. Radiologists rely on Gd-EOB-DTPA (Primovist), a contrast agent taken up by functioning hepatocytes. Typically, FNH appears bright during the HBP because it is composed of hyperplastic hepatocytes. However, "non-classic" FNH can appear dark, mimicking malignancy and posing a significant diagnostic challenge.
Methods
The researchers retrospectively analyzed 43 FNH lesions from 39 patients who underwent Gd-EOB-DTPA enhanced MRI between 2016 and 2019. Imaging data were categorized by signal intensity (homogenous/heterogeneous and bright/dark). These findings were then compared against pathological gold standards, including classifications of "classic" vs. "non-classic" FNH and immunohistochemistry for markers such as β-catenin and CK7/CK19.
Key Results and Statistical Significance
The study found that 88.37% (38/43) of lesions showed the expected iso-high intensity signal. However, 11.63% (5/43) presented as low-intensity (dark) signals. Crucially, every single low-intensity lesion was pathologically classified as non-classic FNH.
Table showing the significant correlation between HBP signal intensity and FNH pathological classification (P = 0.003).
As shown in the data, classic FNH lesions never presented with low-intensity signals in this cohort, whereas non-classic FNH accounted for all atypical imaging presentations. This statistical difference ($P = 0.003$) underscores the relationship between tissue architecture and MRI appearance.
Pathological Drivers of Atypical Signals
The researchers identified several key factors that "kill" the expected bright signal in FNH:
Steatosis (Fatty Change): When more than 40% of the hyperplastic hepatocytes contained fat, contrast uptake was inhibited, leading to a dark appearance.
Fibrous Scars: The presence of fibrous tissue was responsible for the heterogeneous appearance of classic FNH. Lesions lacking these scars appeared homogenous.
Genetic Markers: In rare cases, β-catenin (nucleus +) mutations were associated with low-intensity signals, likely due to altered expression of the transporters that pull contrast into the cells.
Detailed breakdown of the pathological features found in non-classic FNH lesions.
While Gd-EOB-DTPA MRI remains a primary tool for diagnosing FNH, this study highlights that a dark lesion in the hepatobiliary phase is not a definitive indicator of malignancy. Radiologists and clinicians should consider non-classic FNH as a possibility, particularly when pathological features like high steatosis or a lack of typical fibrous scarring are present. Further multi-center studies are needed to refine the diagnostic criteria for these "chameleon" lesions.
Summary updated
Abstract
Abstract Background: To explore the signals and diagnostic value of hepatic focal nodular hyperplasia (FNH) in the hepatobiliary phase of gadolinium-ethoxybenzyl-diethylenetriamine-pentaacetic acid (Gd-EOB-DTPA) enhanced MRI. Methods: Imaging data of 43 pathologically proven FNH lesions from 39 patients who underwent Gd-EOB-DTPA enhanced MRI scanning at our hospital between January 2016 and June 2019 were retrospectively analyzed. The signal characteristics in the hepatobiliary phase were analyzed and compared with the pathologic findings. Results: According to the characteristics of signals in the hepatobiliary phase, the signals were classified as follows: homogenous iso-high intensity signals in 20.93% (9/43) lesions, heterogeneous iso-high intensity signals in 67.44% (29/43) lesions, homogenous low-intensity signals in 4.65% (2/43) lesions, and heterogeneous low-intensity signals in 6.98% (3/43) lesions. Two patients were with multiple lesions, where one was with 2 lesions of heterogeneous high-intensity signals, and the other with 3 lesions of heterogeneous low-intensity signals. Pathologic findings were as follows: the slices of the 38 lesions with high-intensity signals in a hepatobiliary phase were with hyperplasic hepatocytes, inflammatory cell infiltration, and malformed blood vessels. Twenty-nine of the lesions were with fiber tissues of different degrees and were classified as classic type. The remaining 9 lesions were without fibrous scars and were classified as non-classic type. The other 5 of the 43 lesions were non-classic FNH with no evident fibrous tissues, while 4 of them were with >40% steatosis in the hyperplasic hepatocytes; the immunohistochemistry showed CK7(-)/CK19(-) in 1 lesion and β-catenin (nucleus +) in another lesion. Comparisons of pathologic with imaging findings were as follows: twenty-nine lesions were with heterogeneous iso-high intensity signals, of which the slices showed evident fibrous tissues of different degrees, and the slices of 9 lesions with homogenous iso-high intensity signals in the hepatobiliary phase showed no fibrous tissues. Three lesions with heterogeneous low-intensity signals in the hepatobiliary phase showed about 80% mixed steatosis in hyperplasic hepatocytes. The other two lesions both showed homogeneous low-intensity signals in the hepatobiliary phase, where 1 lesion was with >40% macrovesicular steatosis and CK7/CK19 (-), while the other only showed β-catenin (nucleus +) by immunohistochemistry. Conclusions: The signals of FNH in the hepatobiliary phase showed various characteristics, where the signal differences were mainly associated with the number of hyperplastic hepatocytes in lesions, presence of steatosis, fibrous scars, and conditions of small bile ducts, and potentially associated with β-catenin (nucleus+). Low-intensity signals were relatively rare for FNH, thus representing a relatively major challenge for diagnosing this type FNH.
BibTeX
@inproceedings{xu2021manifestations,
author = {Jiangfeng Xu and Liqing Zhang and Jianxun Li and Xiaojian Cui and Qingping Jiang and Yan Zhang},
title = {{Manifestations of Hepatic Focal Nodular Hyperplasia in Hepatobiliary Phase of Enhancement by Hepatocyte-specific Contrast Agent: Comparison With Pathologic Findings}},
booktitle = {Research},
year = {2021},
doi = {10.21203/rs.3.rs-870839/v1},
}