Imaging characteristics of hypervascular focal nodular hyperplasia-like lesions in patients with chronic alcoholic liver disease

Abstract

BACKGROUND

Focal nodular hyperplasia (FNH)-like lesions are hyperplastic formations in patients with micronodular cirrhosis and a history of alcohol abuse. Although pathologically similar to hepatocellular carcinoma (HCC) lesions, they are benign. As such, it is important to develop methods to distinguish between FNH-like lesions and HCC.

AIM

To evaluate diagnostically differential radiological findings between FNH-like lesions and HCC.

METHODS

We studied pathologically confirmed FNH-like lesions in 13 patients with alcoholic cirrhosis [10 men and 3 women; mean age: 54.5 ± 12.5 (33-72) years] who were negative for hepatitis-B surface antigen and hepatitis-C virus antibody and underwent dynamic computed tomography (CT) and magnetic resonance imaging (MRI), including superparamagnetic iron oxide (SPIO) and/or gadoxetic acid-enhanced MRI. Seven patients also underwent angiography-assisted CT.

RESULTS

The evaluated lesion features included arterial enhancement pattern, washout appearance (low density compared with that of surrounding liver parenchyma), signal intensity on T1-weighted image (T1WI) and T2-weighted image (T2WI), central scar presence, chemical shift on in- and out-of-phase images, and uptake pattern on gadoxetic acid-enhanced MRI hepatobiliary phase and SPIO-enhanced MRI. Eleven patients had multiple small lesions (< 1.5 cm). Radiological features of FNH-like lesions included hypervascularity despite small lesions, lack of “corona-like” enhancement in the late phase on CT during hepatic angiography (CTHA), high-intensity on T1WI, slightly high- or iso-intensity on T2WI, no signal decrease in out-of-phase images, and complete SPIO uptake or incomplete/partial uptake of gadoxetic acid. Pathologically, similar to HCC, FNH-like lesions showed many unpaired arteries and sinusoidal capillarization.

CONCLUSION

Overall, the present study showed that FNH-like lesions have unique radiological findings useful for differential diagnosis. Specifically, SPIO- and/or gadoxetic acid-enhanced MRI and CTHA features might facilitate differential diagnosis of FNH-like lesions and HCC.

Key Words: Focal nodular hyperplasia; Alcoholic liver disease; Hepatocellular carcinoma; Magnetic resonance imaging; Liver

Core Tip: Two enhancement patterns were observed for the hepatobiliary phase on gadoxetic acid-enhanced magnetic resonance imaging (MRI): Heterogeneous hyperintense (43%) and ring-like enhancement (57%), and all lesions exhibited a marked homogeneous uptake pattern on superparamagnetic iron oxide-enhanced MRI in patients with focal nodular hyperplasia-like lesions. This finding is of clinical relevance because it is useful for the differential diagnosis of hypervascular liver nodules in patients with chronic liver disease.

INTRODUCTION

Focal nodular hyperplasia (FNH)-like lesions were initially reported by Gluud et al[1] as hyperplastic formations in patients with micronodular cirrhosis and a history of alcohol abuse. These lesions, now widely identified in patients with alcoholic liver cirrhosis, are histopathologically similar to FNH in non-cirrhotic livers[2-7]. However, FNH-like lesions further exhibit unique vascular characteristics, including hypervascularity and sinusoidal capillarization[7], which are also common in hepatocellular carcinoma (HCC). Conversely, unlike HCC, FNH-like lesions are generally considered benign and non-progressive, particularly following alcohol abstinence. Despite these findings, FNH-like lesions are not yet classified in major diagnostic guidelines, such as those of the International Working Party[8] or the World Health Organization (WHO)[9], presenting challenges in diagnosis and clinical management. Clinically, the differentiation between FNH-like lesions and HCC is essential as the treatment approach differs significantly. While HCC commonly requires aggressive treatment options, including surgery, ablation, or systemic therapies, FNH-like lesions may not require any interventions, making an accurate diagnosis crucial for appropriate patient care. The current literature on FNH-like lesions primarily addresses their pathology, and there are relatively few studies on their imaging characteristics, making non-invasive diagnosis challenging.

This study aimed to address these gaps by examining the radiological features of FNH-like lesions in alcoholic liver cirrhosis using dynamic computed tomography (CT), superparamagnetic iron oxide (SPIO), and gadoxetic acid-enhanced magnetic resonance imaging (MRI). By identifying specific imaging criteria, we aimed to support non-invasive differentiation of FNH-like lesions from early-stage HCC, potentially reducing the need for invasive biopsy and providing clearer diagnostic pathways for clinicians.

MATERIALS AND METHODS

Patients

This retrospective study was approved by the institutional ethics committee of Kindai University, Faculty of Medicine (ethical identification No. 27-067). All patients provided informed consent for the use of their CT and MRI, as well as their biopsy specimens. The study protocol conformed to the ethics guidelines of the 2002 Declaration of Helsinki.

This study enrolled 13 patients with pathologically confirmed FNH-like lesions who underwent dynamic CT, unenhanced MRI, and SPIO- and/or gadoxetic acid-enhanced MRI between 2000 and 2020. In all the patients, liver cirrhosis was determined through clinical examination and blood chemistry tests (for aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, bilirubin, albumin, and globulin). None of the patients included in this study had comorbidities such as diabetes, hypertension, or metabolic syndrome, as verified by assessment of their medical histories and clinical examinations. Further, each patient had at least one FNH-like nodule pathologically diagnosed through biopsy, with the remaining lesions being diagnosed based on clinical findings indicating no decrease in size after abstinence from alcohol during the 1-year follow-up.

Imaging technique

Plain and three-phase contrast-enhanced CT (CE-CT) scans were obtained in a craniocaudal direction using Sensation 64 CT (Siemens, Erlangen, Germany) or Aquilion multi 64 (Canon Medical System, Otawara, Japan) scanners. Routine scanning was conducted at a 5-mm section thickness and 5-mm scan increment. Subsequently, the scans were reconstructed with a 2-mm thickness using a routine abdominal algorithm. An injector system was used to intravenously administer 100-150 mL non-ionic contrast media (Iopamiron^®; 300/370 mgI/mL; Bayer Schering Pharma, Berlin, Germany) at a rate of 3–3.5 mL/s.

All MRIs were performed on a 1.5-T system using the Magnetom Avanto (Siemens Healthcare, Erlangen, Germany) or Signa HDxt (GE Healthcare, Waukesha, WI, United States) scanner. Baseline MRI were acquired using a respiratory-triggered T2-weighted turbo spin-echo (TSE) sequence, breath-hold T2*-weighted gradient-echo (GRE) imaging with steady-state precession (FISP) sequence, and breath-hold T1-weighted GRE sequence. All images were obtained in the transaxial plane, using a phased-array multi-coil. For all sequences, a 7-mm slice thickness was used, with a 10% intersection gap and a field of view of 35-40 cm, depending on the liver size. SPIO-enhanced MRI comprised the respiratory-triggered T2-TSE sequence, breath-hold T2*-weighted FISP sequence, and breath-hold T1-weighted GRE sequence, with the same parameters as those used in baseline MRI. Fercarbotran (Resovist^®; Bayer Yakuhin, Osaka, Japan) at a dose of 8.0 μmol iron per kg (body weight) was manually injected as a rapid bolus through a filter with 5-μm pore size, which was immediately followed by a 10 mL saline solution flush. Imaging was then performed after 7 minutes. Gadoxetic acid-enhanced MRI comprised dynamic images constructed using fat-suppressed T1-weighted GRE images obtained before (pre-contrast) as well as 14-30 seconds (arterial phase), 70 seconds, and 3 minutes after intravenous administration of 0.025 mmol of gadoxetic acid (Primovist^®, Bayer Schering Pharma, Berlin, Germany) per kg (body weight) at a rate of 2.0 mL/s, followed by a 20 mL saline flush. Hepatobiliary phase (HBP) images were obtained 20 minuntes after gadoxetic acid injection.

All CT during arterial portography and double-phase CT during hepatic angiography (CTHA) examinations were performed using a unified angiography and CT system (Multistar Plus/Somatom Plus 4 Volume Zoom; Siemens, Erlangen, Germany). These images were subsequently obtained in a craniocaudal direction with 5-mm section thickness and 5-mm reconstruction intervals. For double-phase CTHA, first-phase data acquisition was started 5 seconds after second-phase data acquisition, which was performed 21-26 seconds after initiation of a transcatheter hepatic arterial injection of a fixed dose of 20 mL iohexol (Omnipaque 300, 300 mgI/mL; Daiichi-Sankyo Parma, Tokyo, Japan) diluted with 33% sterile water (iodine, 100 mg iodine/mL) at a rate of 2 mL/s.

Pathological analysis

All patients underwent ultrasound-guided core needle biopsy, following which the liver samples were fixed in 10% neutral-buffered formalin, routinely processed for paraffin embedding, sectioned (4 µm), and stained with hematoxylin and eosin or Masson trichrome (Muto Pure Chemicals, Tokyo, Japan) according to standard procedures. Immunohistochemical analysis for OATP8 and CD68 antigens were performed on formalin-fixed, paraffin-embedded specimens extracted from seven patients. The stained sections were evaluated under a microscope. The expressions of OATP8 and CD68 antigen were typically assessed by determining the intensity and localization of staining within hepatocytes. The intensity was graded semi-quantitatively (as either none, weak, moderate, or strong), and the localization was confirmed as predominantly cytoplasmic and/or membrane-associated, depending on the known expression pattern of OATP8. The percentage of positively stained cells was also recorded.

Image analysis

Two radiologists with > 10 years of experience evaluated the following lesion characteristics: Number and size (largest axial diameter) of each lesion; dynamic enhancement pattern on dynamic CT and CTHA (1 and 2) for qualitative imaging analysis; and other MRI features (3-8) as follows: (1) Arterial enhancement pattern of the dynamic CT and/or first phase of the CTHA; (2) Washout appearance (defined as low density compared with the density of the surrounding liver parenchyma) on the portal venous phase (PVP) of the dynamic CT and/or second phase of the CTHA; (3) Signal intensity on T1-weighted image (T1WI); (4) Signal intensity on T2-weighted image (T2WI); (5) Central scar (defined as a central T2 hyperintensity and/or T1 hypointensity representing fibrotic tissue); (6) Chemical shift (defined as a drop in signal intensity on out-of-phase images compared with that on in-phase images); (7) Uptake pattern on the HBP of the gadoxetic acid-enhanced MRI; and (8) Uptake pattern on the SPIO-enhanced MRI. Arterial enhancement was classified as homogeneous (peripherally or entirely) or heterogeneous. Signal intensity on T1WIs was classified as hypointensity, iso- or hyperintensity, or heterogeneous signal intensity. The signal intensity on T2WIs was classified as iso-intensity or hyperintensity. Uptake on the HBP of gadoxetic acid- and SPIO-enhanced MRI was classified as homogeneous (peripherally or entirely) or heterogeneous. In case of interobserver discordance, the radiologists re-evaluated the images and reached a consensus. Both radiologists were informed about the location of each lesion; however, they were blinded to the clinical information and final diagnosis.

RESULTS

This study included 13 patients [10 men and 3 women; mean age: 54.5 ± 12.5 (33-72) years] with chronic liver failure due to alcoholic liver disease. Table 1 presents a summary of the clinical findings. All patients with FNH-like lesions were pathologically diagnosed using fine-needle biopsy. The levels of alpha-fetoprotein (AFP; ng/mL) and prothrombin induced by vitamin K deficiency or antagonist-II (mAU/mL) were normal in all patients with FNH-like lesions, except for one patient with HCC who showed elevated AFP (ng/mL) levels. Among the 10 patients with FNH-like lesions, 2 (20%) showed new lesions following abstinence from alcohol. The size of the FNH-like lesions remained stable or decreased in eight (80%) patients.

Table 1 Summary of clinical findings.

Patient No.	Age (years)	Sex	Background of liver failure	Child Pugh grade	AFP	PIVKA-II	Diagnosis	Lesion growth at imaging follow up
1	41	M	Alcoholic	A	6	23	FNB	N/A
2	33	F	Alcoholic	A	7	23	FNB	No growth
3	34	F	Alcoholic	B	4	10	FNB	Increased number
4	60	M	Alcoholic	B	9	20	FNB	No growth
5	58	M	Alcoholic	A	7	40	FNB	No growth
6	57	M	Alcoholic	B	6	19	FNB	N/A
7	53	M	Alcoholic	B	231	33	FNB	No growth
8	65	M	Alcoholic	A	10	18	FNB	Increased number
9	63	M	Alcoholic	A	6	12	FNB	No growth
10	48	F	Alcoholic	B	11	18	FNB	No growth
11	55	M	Alcoholic	B	5	16	FNB	N/A
12	70	M	Alcoholic	A	5	22	FNB	No growth
13	72	F	Alcoholic	A	4	15	FNB	No growth

¹Case 7 was only the patient with hepatocellular carcinomas.

PIVKA-II: Prothrombin induced by vitamin K deficiency or antagonist-II; FNB: Fine needle biopsy; AFP: Alpha-fetoprotein; N/A: Not available.

Table 2 presents the CT findings of the FNH-like lesions. Among the 13 patients with 26 FNH-like lesions, seven had a single lesion, while six had multiple lesions (two, one, and three patients had two, three, and over four lesions, respectively). The lesion size ranged from 10 mm to 30 mm (mean ± SD, 14.0 ± 5.5 mm). All 26 FNH-like lesions exhibited marked hypervascularity during the arterial phase on CE-CT and CTHA (seven patients) as well as washout appearance during the portal venous and equilibrium phases on CE-CT. Pathological examination revealed numerous unpaired arteries and sinusoidal capillarization in all patients, which were similar to those in HCC.

Table 2 Summary of tumor characteristics, dynamic patterns assessed using computed tomography and computed tomography hepatic arteriography, and pathological findings.

Patient No.	No. of nodules	Size	Arterial enhancement pattern	Washout appearance	Modality	Central scar	Unpaired arteries and capillarization	Liver cell atypia	Scar-like fibrosis
1	2	15	Hyper	No	CECT/CTHA	-	+	-	-
2	1	15	Hyper	No	CECT/CTHA	-	+	-	-
3	> 4	10-15	Hyper	No	CECT/CTHA	-	+	-	-
4	> 4	10-30	Hyper	No	CECT/CTHA	+	+	-	+
5	1	10	Hyper	No	CECT	-	+	-	-
6	3	10-30	Hyper	No	CECT/CTHA	-	+	-/+	-
7	1	15	Hyper	No	CECT/CTHA	+	+	-	+
8	1	10	Hyper	No	CECT	-	+	-	-
9	1	10	Hyper	No	CECT	-	+	-	-
10	> 4	10-15	Hyper	No	CECT/CTHA	-	+	-	-
11	2	15	Hyper	No	CECT	-	+	-	-
12	1	20	Hyper	No	CECT	-	+	-	-
13	1	14	Hyper	No	CECT	+	+	-	-

CT: Computed tomography; CECT: Contrast-enhanced computed tomography; CTHA: Computed tomography hepatic arteriography.

Table 3 presents the MRI findings of the FNH-like lesions. Among the 13 patients with FNH-like lesions, 11 (85%) showed high or iso-high intensity on T1WI, seven (54%) showed iso-high intensity on T2WI, six (46%) showed low intensity, three (23%) showed high-intensity central scars, and 10 (77%) showed no central scars on T2WI. Chemical shift artefacts were observed in all 13 patients with FNH-like lesions (Figures 1 and 2). In the six patients who underwent gadoxetic acid-enhanced MRI (Figures 3 and 4), all lesions exhibited marked homogeneous enhancement during the arterial phase. Two enhancement patterns were observed on the HBP: Heterogeneous hyperintense (n = 3, 43%) and ring-like enhancement (n = 4, 57%). In the eight patients who underwent SPIO-enhanced MRI, all lesions exhibited a marked homogeneous uptake pattern on post-SPIO-enhanced MRI (Figures 1 and 3).

Figure 1 A 58-year-old man with alcoholic liver cirrhosis (patient No. 5). A and B: T1-weighted gradient-echo (GRE) magnetic resonance (MR) (A: In-phase; B: Opposed-phase) image shows a small and high-signal-intensity nodule (arrow) in the lateral segment; C: T2-weighted fast spin-echo MR imaging (MRI) shows iso-intensity nodule (arrow); D and E: Gadolinium-enhanced T1-weighted GRE MRI obtained during the arterial- and portal venous phase reveals a nodule with arterial phase hyperenhancement (arrow) and without washout (arrow); F: The superparamagnetic iron oxide (SPIO)-enhanced T2*-weighted GRE MRI shows the lesion as a low-signal intensity nodule (arrow) with SPIO uptake compared with the surrounding liver parenchyma.

Figure 2 Histopathological features of the non-nodular and nodular portions. A: Non-nodular portion showing chronic hepatitis with pericellular fibrosis and diffuse capillarization of sinusoids, likely due to alcoholic injury. Mild inflammatory activity and moderate fibrosis are observed (hematoxylin and eosin staining, × 100); B: Nodular portion displaying a hyperplastic nodule with a mild increase in cell density and pericellular fibrosis. Unpaired arteries (arrow) are present in the nodule. (Hematoxylin and eosin staining, × 100); C: Non-nodular portion highlighting pericellular fibrosis and sinusoidal capillarization (Masson trichrome staining, × 100); D: Nodular portion illustrating pericellular fibrosis and unpaired arteries (arrow) within the hyperplastic nodule (Masson trichrome staining, × 100).

Figure 3 A 65-year-old man with alcoholic liver cirrhosis (patient No. 8). A and B: T1-weighted gradient-echo (GRE) magnetic resonance (MR) (A: In-phase; B: Opposed-phase) image shows a small and high-signal-intensity nodule (arrow) in segment 7 of the liver; C: Pre-contrast T1-weighted GRE MR imaging (MRI) shows iso-slightly high-intensity nodule (arrow); D and E: Gadoxetic acid-enhanced T1-weighted GRE MRI obtained during the arterial-and portal venous phase reveal a nodule with arterial phase hyperenhancement (arrow) and without washout (arrow); F: The hepatobiliary phase of the gadoxetic acid-enhanced T1-weighted GRE MRI shows homogeneous high-intense uptake (arrow); G: T2*-weighted GRE MRI shows iso-intensity nodule (arrow); H: The superparamagnetic iron oxide (SPIO)-enhanced T2*-weighted GRE MRI shows the lesion as a low-signal intensity nodule (arrow) with SPIO uptake compared with the surrounding liver parenchyma.

Figure 4 Histopathological and Immunohistochemical features of the non-nodular and nodular portions (A, C: Non-nodular, B, D, E: Nodular). A and B: Cell density, iron deposits, and sinusoidal capillarization in the surrounding liver tissue and the focal nodular hyperplasia (FNH)-like nodule. The FNH-like nodule shows increased cell density (A: hematoxylin and eosin staining × 400); C and D: OATP8 (red-orange color) is expressed on the cellular membrane of hepatocytes at the sinusoidal side. The expression of OATP8 is nearly absent in the nodule (B: × 400); however, it is diffusely found in the surrounding tissue (C: × 400). OATP8 is immunohistochemically detected using anti-OATP8; E: Immunohistochemical staining of CD68 antigen showing Kupffer cells. The staining of nodule CD68 antigen shows diffuse Kupffer cell infiltration.

Table 3 Summary of magnetic resonance imaging characteristics and findings.

Patient No.	T1-weighted images	T2-weighted images	Central scar on magnetic resonance imaging	Chemical shift	Uptake pattern on the HBP	Uptake pattern on the SPIO
1	High	Iso-high	No	No	N/A	Homogenous
2	High	Iso-high	No	No	N/A	N/A
3	High	Low	No	No	N/A	Homogenous
4	Iso-high	Low	Yes	No	N/A	Homogenous
5	High	Iso-high	No	No	Ring	Homogenous
6	Low-high	Low	No	No	N/A	Homogenous
7	High	Iso	Yes	No	N/A	Homogenous
8	High	Low	No	No	Ring/hetero	Homogenous
9	High	Iso-high	No	No	N/A	Homogenous
10	High	Iso-high	No	No	Hetero	N/A
11	High	Low	No	No	Ring	N/A
12	Iso	Low	No	No	Hetero	N/A
13	Iso-high	Iso-high	Yes	No	Ring	N/A

SPIO: Superparamagnetic iron oxide; N/A: Not available.

DISCUSSION

We studied pathologically confirmed FNH-like lesions in 13 patients with alcoholic cirrhosis. The radiological features of FNH-like lesions included hypervascularity, despite small lesion sizes, lack of a “corona-like” enhancement in the late phase on CTHA, high-intensity on T1WI, slightly high- or iso-intensity on T2WI, no signal decrease in out-of-phase images, and complete SPIO uptake or incomplete/partial uptake of gadoxetic acid. There have been several recent reports of FNH-like lesions in cirrhotic livers[1-6]. Although there is currently no established definition for FNH-like lesions/nodules, FNH is known to develop in normal livers. Accordingly, FNH-like lesions refer to similar nodules observed in patients with chronic liver disease[7]. These lesions have been extensively reported in alcoholic liver disease[8-10]. Most of these hyperplastic lesions are highly vascular and relatively small; additionally, they are difficult to differentiate from HCC lesions, especially early-stage HCC lesions[7-12]. We included 13 patients with alcoholic liver disease and hyperplastic liver lesions who were negative for hepatitis virus markers. Here, we discuss the imaging findings with respect to reported imaging characteristics of such lesions[13-20].

In our cases, except for a normal hepatic background and absence of a central scar, the pathological findings were consistent with those of FNH; accordingly, the patients were diagnosed with FNH-like lesions, based on the original text in the WHO classification (version 5)[9]. There has been no previous report on the radiographic findings of FNH-like lesions based on our exact definition since there remains no established definition. There were no differences in imaging diagnoses according to age or gender, as in previous reports. However, considering the similarities of the histopathology and imaging findings with those of FNH, it is highly likely that FNH-like lesions have been previously described as atypical FNH based on imaging findings. Approximately 80% of FNH cases are reported to be atypical on imaging[21]. Typical FNH on plain CT is characterized by hypo- to iso-attenuation, with homogeneous attenuation in early-phase contrast enhancement, and isoattenuation with the surrounding liver from the PVP to the equilibrium phase. On MRI, these lesions present as iso- to moderately hypointense on T1WIs, iso- to somewhat hyperintense on T2WIs, and iso- to somewhat hyperintense or ring-like hyperintense in the hepatocyte phase after gadoxetic acid-enhanced-MRI[13-22]. The central scar is generally hypointense on plain CT and early-phase contrast enhancement CT, with delayed enhancement into the equilibrium phase. Additionally, the central scar is hypointense on T1WIs and hyperintense on T2WIs; further, the region, including the surrounding area, has low uptake of gadoxetic acid during the hepatocyte phase. However, a central scar is only observed in approximately 50% of cases[23].

Single-level dynamic CTHA has indicated that blood flows from a dilated feeding artery inside the central scar; moreover, blood within the lesion directly flows into the hepatic vein via the fibrous septum and dilated veins at the margin of the lesion[24]. Grazioli et al[22] did not report the imaging characteristics of six nodules with hypointense enhancement; as such, whether there was a complete lack of enhancement currently remains unclear. Notably, the enhancement was relatively lower than that of the surrounding liver; moreover, the heterogeneity was similar to that in our cases; however, their nodules exhibited overall hypointense enhancement. In addition, similar to our cases, the nodule lacked a central scar. Furthermore, although there is heterogeneity in the signal intensity in the hepatocyte phase of typical FNH, which ranges from iso- to hyperintense enhancement, the mechanisms underlying these findings remain unclear. Moreover, the cause of the relatively low enhancement in the uptake area within the nodule remains unclear.

This study has several limitations. First, this retrospective study had inevitable selection bias. Additionally, all patients with FNH-like lesions were pathologically confirmed using fine-needle biopsy specimens rather than surgical specimens. Finally, there was no standard follow-up imaging of the FNH-like lesions; further, we did not describe the frequency of this phenomenon. Further large-scale multicenter studies using surgical specimens are warranted.

CONCLUSION

In conclusion, the present study showed that FNH-like lesions in chronic alcoholic liver disease demonstrate distinct imaging features, including small lesion size, hypervascularity, high signal on T1WIs, and characteristic uptake patterns on SPIO or gadoxetic acid-enhanced MRI. These features support a non-invasive differentiation from HCC, thereby reducing the need for invasive diagnostics and aiding appropriate clinical management. Overall, our findings provide valuable imaging criteria that can improve diagnostic accuracy and patient outcomes.