Skip main navigation
Close Drawer MenuOpen Drawer Menu
Search
Quick Search in Journals
Quick Search anywhere
Advanced Search
Menu
  • Information
    • Previous
    Next
    Original ResearchFree Access

    Long-term Evolution of Hepatocellular Adenomas at MRI Follow-up

    Published Online:Mar 17 2020https://doi.org/10.1148/radiol.2020191790

    Abstract

    Background

    Hepatocellular adenomas (HCAs) are rare benign liver tumors. Guidelines recommend continued surveillance of patients diagnosed with HCAs, but these guidelines are mainly based on small studies or expert opinion.

    Purpose

    To analyze the long-term evolution of HCAs, including solitary and multiple lesions, and to identify predictive features of progression with MRI.

    Materials and Methods

    In a retrospective study, patients diagnosed with pathologically proven solitary or multiple HCAs between January 2004 and December 2015 were included; β-catenin–mutated HCAs and HCAs with foci of malignancy were considered to be at risk for progression. MRI examinations were analyzed, and tumor evolution was evaluated by using Response Evaluation Criteria in Solid Tumors, version 1.1. Student t, Mann-Whitney, χ2, Fisher exact, and McNemar tests were used, as appropriate.

    Results

    In total, 118 patients (mean age, 40 years ± 10 [standard deviation]; 108 women) were evaluated, including 41 with a solitary HCA (mean age, 40 years ± 14; 36 women) and 77 with multiple HCAs (mean age, 40 years ± 10; 72 women). At a median follow-up of 5 years, 37 of 41 (90%) patients with a solitary HCA and 55 of 77 (71%) patients with multiple HCAs showed stable or regressive disease. After resection of solitary HCAs, new lesions appeared in only two of 29 (7%) patients, both of whom had HCAs at risk of progression. In patients with multiple HCAs, hepatocyte nuclear factor 1α–inactivated HCAs showed a higher rate of progression compared with inflammatory HCAs (11 of 26 [42%] vs seven of 37 [19%], P = .04) despite lower use (28 of 32 patients [88%] vs 45 of 45 patients [100%]; P = .03) and shorter duration (mean, 12.0 years ± 7.5 vs 19.2 years ± 9.2; P = .001) of oral contraceptive intake.

    Conclusion

    Long-term MRI follow-up showed that 78% of hepatocellular adenomas had long-term stability or regression. After resection of solitary hepatocellular adenomas, new lesions occurred only in hepatocellular adenomas at risk of progression. Patients with multiple hepatocellular adenomas were more likely to show progressive disease, with hepatic nuclear factor 1α–inactivated hepatocellular adenomas being the most common subtype showing progression.

    © RSNA, 2020

    Online supplemental material is available for this article.

    Download as PowerPoint

    Summary

    MRI follow-up may be discontinued in patients with a solitary hepatocellular adenoma after resection unless worrisome features are detected at pathologic assessment; conversely, patients with multiple hepatocellular adenomas should undergo continuous follow-up regardless of surgery.

    Key Results

    • ■ Long-term MRI follow-up showed that 71% (55 of 77) of patients with multiple hepatocellular adenomas (HCAs) and 90% (37 of 41) of patients with solitary HCAs had stable or regressive disease.

    • ■ After resection of solitary HCAs, only 7% (two of 29) of the patients had new lesions after a median follow-up of 5 years.

    • ■ Progression rate in patients with multiple HCAs who underwent surgery but had at least one lesion left in place was similar to that in patients who did not undergo surgery (31% [17 of 54] vs 25% [five of 20], respectively).

    • ■ Hepatic nuclear factor 1α–inactivated HCAs showed a higher rate of progression and growth from less than 50 mm to 50 mm or larger compared with inflammatory HCAs (progression rate of approximately 40% [11 of 26] vs 20% [seven of 37]), despite lower use and shorter duration of oral contraceptive intake.

    Introduction

    Hepatocellular adenomas (HCAs) are rare benign liver tumors that occur in 0.001%–0.004% of the population and typically affect young women (10:1 female-to-male ratio) (1–3). These tumors may be complicated by hemorrhage or malignant transformation in up to 27% and 4% of affected patients, respectively (4,5). Tumor size increase is considered one of the risk factors for complications (4,5). It is known that HCAs may progress in size and number over time, potentially placing patients at risk for complications. Therefore, European Association for the Study of the Liver (EASL) guidelines recommend MRI surveillance every 6 months for the 1st year after diagnosis of HCA and annually thereafter, as well as lesion resection for tumors that continue to grow or that are larger than 50 mm (1).

    The time points for lesion follow-up according to EASL guidelines are mainly based on the findings of small studies or on expert opinion rather than solid evidence. Indeed, it is known that progression of HCAs may occur in up to 19% of patients (6–16) and that size changes could correlate with different patient- and lesion-related variables (14–16). However, prior studies were limited by small numbers of patients, relatively short overall follow-up periods, lack of HCA subtyping, and, above all, lack of consistent use of MRI for lesion follow-up. MRI is preferable to CT in EASL guidelines for HCA imaging follow-up, not only because of the lack of exposure to ionizing radiation, but also because of its high soft-tissue resolution and its multiparametric nature, both of which may result in higher diagnostic accuracy for HCA detection, characterization, and subtyping (17). To date, there have been few longitudinal analytical studies of long-term MRI follow-up of HCAs. An evidence-based understanding of the natural history of different subtypes of HCAs at MRI may help to inform patient-tailored recommendations, which in turn may result in better disease management and outcome.

    The primary aim of this study was to analyze the long-term evolution of different subtypes of HCAs. Secondary aims included determining if there were differences in natural history between solitary and multiple lesions and identifying predictive features of progression.

    Materials and Methods

    This retrospective single-center cohort study was approved by the institutional review board of Beaujon Hospital, a tertiary referral center for liver diseases, and a waiver of written informed consent was obtained. For patients who participated in the telephone interview, oral consent was given.

    Study Cohort

    Figure 1 shows the patient accrual flowchart, which was based on Strengthening the Reporting of Observational Studies in Epidemiology guidelines (18). We retrospectively searched the departmental database at Beaujon Hospital for consecutive patients diagnosed with pathologically proven (ie, after biopsy or surgery) solitary or multiple HCAs between January 2004 and December 2015. Patients were excluded if (a) preoperative or follow-up MRI scans were not available (this category included patients in whom the overall follow-up period was less than 6 months); (b) hepatic malignancies (eg, hepatocellular carcinoma, metastases) were present, which could potentially influence the natural evolution of HCAs; (c) they had undergone local-regional therapy (ie, transarterial embolization, radiofrequency ablation) or transplantation; or (d) the MRI protocol was inadequate. The MRI protocol was considered inadequate if the (a) fat-saturated T2-weighted sequence, (b) dual-phase sequence, or (c) fat-saturated T1-weighted sequence before or after intravenous contrast agent administration was not performed or was nondiagnostic.

    Figure 1:

    Figure 1: Flowchart shows study enrollment based on recommended Standards for Reporting of Diagnostic Accuracy criteria and the reference standard.

    Download as PowerPointOpen in Image Viewer

    Clinical Data

    We reviewed the electronic records of the included patients and assessed demographic information, notably age and sex, for each patient. Then we performed a telephone interview to collect information regarding the following patient-related factors: symptoms at diagnosis, height and weight at diagnosis, diabetes, hypercholesterolemia, arterial hypertension, use of oral contraceptives (including age at initial intake and interruption), age at menarche, onset of menopause, and history of pregnancy. Height and weight were used to calculate the body mass index (15). Time on oral contraceptives before diagnosis was calculated as the difference between age at interruption and age at first intake. As per the standard of care, all women with an HCA diagnosis had stopped using oral contraceptives.

    Reference Standard

    All specimens were reviewed by an experienced hepatobiliary pathologist (V.P., 25 years of experience) who was blinded to clinical information, imaging test results, and the original pathologist’s report. The following information was assessed for each patient: (a) aspect of nontumoral liver (fibrosis and steatosis, including the percentage of steatosis, and the presence of microadenomas in the resected specimen [ie, microscopic foci of adenomas smaller than 1 cm undetected on preoperative radiologic images]); (b) subtype of HCA according to the updated classification published by Nault et al (4), including inflammatory, hepatocyte nuclear factor 1α (HNF-1α)–inactivated, β-catenin–mutated (in exon 3 or exons 7–8), Sonic the Hedgehog, and unclassified HCAs (Appendix E1 [online]); (c) micro- or macroscopic hemorrhage within the lesion; and (d) malignant foci within the lesion. Malignancy within the lesion was diagnosed when the pathologic examination revealed foci of hepatocellular carcinoma inside the adenomatous proliferation. Pathologic proof of HCA was obtained with biopsy in 32 patients and with surgery in 86. At pathologic evaluation, lesion characteristics (ie, subtype, hemorrhage, or malignancy within the lesion) were available for all 41 patients with a solitary HCA and for 129 lesions in the 77 patients with multiple HCAs; characteristics of nontumoral liver were available in 95 patients, including 38 with a solitary HCA and 57 with multiple HCAs. MRI findings were used for subtype classification of HCAs without pathologic proof in patients with multiple HCAs (55.9%, 166 of 297) and of newly developed HCAs after surgery. Specifically, MRI sequences used to assign a subtype to HCAs were T2-weighted, in- and opposed-phase multiphasic contrast-enhanced sequences during late arterial, portal venous, and delayed phases according to the literature (19,20). MRI signal features that enabled identification of HCA subtypes at imaging are described in Appendix E1 (online). HCAs were defined as at risk of progression in the presence of β-catenin mutation in exon 3 or in the presence of foci of malignancy in resected tumors.

    MRI Analysis

    By using the institutions’ picture archiving and communication system (Carestream PACS, version 12.1.5.6009; Kodak, Rochester, NY), MRI examinations were reviewed according to Response Evaluation Criteria in Solid Tumors, version 1.1, by two radiologists in consensus (F.V., V.V.; 6 years and 32 years of experience in liver imaging, respectively) who were blinded to the original radiology report but not to the pathologic findings regarding the lesions (21). Time points for MRI follow-up had been determined at the discretion of the treating physician or were based on patients’ symptoms. Further information is provided in Appendix E1 (online). In patients with multiple HCAs, in addition to the surgically resected HCAs (n = 88) evaluated at baseline MRI, a maximum of three target lesions per subtype per patient for a total of 209 lesions were analyzed for lesion diameter and subtype at baseline and follow-up MRI.

    Statistical Analysis

    Clinical, pathologic, and MRI data were summarized in a spreadsheet. Categorical variables were summarized as percentages; continuous variables were summarized as means ± standard deviations or as medians and interquartile ranges (IQRs). For continuous variables, differences were evaluated with the Student t or Mann-Whitney test, as appropriate. For categorical variables, differences were evaluated by using the χ2, Fisher exact, or McNemar test, as appropriate. Median follow-up times were compared by using the Kruskal-Wallis test.

    First, we analyzed multiple patient and lesion characteristics, including sex, age, and lesion subtype, in the overall study cohort. Then, we divided all patients into two groups according to the number of lesions: those with a solitary HCA and those with multiple HCAs. We analyzed and compared clinical features, lesion size at baseline and follow-up MRI examinations, and pathologic characteristics of the lesions and nontumoral liver parenchyma between the two groups. Of note, except for demographics, other clinical variables were available only in the 95 (80.5%) of 118 patients who participated in the telephone interview, and the predictive role was analyzed within these patients only (further information on missing data appears in Appendix E1 [online]).

    Long-term evolution of patients with HCAs was analyzed first on a per-patient basis (solitary HCA group vs multiple HCAs group) and then on a per-lesion basis, and these analyses were aimed at identifying clinical, pathologic, or imaging variables predictive of HCA progression (patients with progressive disease vs patients without progressive disease), as further detailed in Appendix E1 (online) (22–24).

    Statistical analyses were performed with statistical software (SPSS, version 20.0; SPSS, Chicago, Ill) or by using the R computing platform (www.r-project.org). All P values were two tailed. P < .05 indicated a significant difference.

    Results

    Study Cohort: Clinical and Pathologic Data

    Figure 1 shows the patient accrual flowchart. A total of 145 patients were excluded for the following reasons: (a) preoperative or follow-up MR images were not available (category included patients in whom the overall follow-up period was shorter than the minimum of 6 months) (n = 107), (b) hepatic malignancies (eg, hepatocellular carcinoma, metastases) were present (n = 13), (c) patients had been treated with local-regional therapies (ie, transarterial embolization, radiofrequency ablation) or transplantation and pretreatment MRI follow-up of 6 months or longer was lacking (n = 22), or (d) MRI protocol was inadequate (n = 3).

    The final study cohort consisted of 118 patients (mean age, 40 years ± 10; 10 men, 108 women), including 41 patients with a solitary HCA (mean age, 40 years ± 14; five men, 36 women) and 77 patients with multiple HCAs (mean age, 40 years ± 10; five men, 72 women). Overall, 338 HCAs (mean diameter at diagnosis, 3.8 cm ± 3.3) were analyzed in this study at baseline MRI, including 172 (50.9%) with pathologic proof (41 lesions in patients with a solitary HCA, 131 lesions in patients with multiple HCAs) and 166 (49.1%) without pathologic confirmation. The telephone interview could be performed in 95 (80.5%) of 118 patients; therefore, some of the clinical information regarding these 95 patients was available and was analyzed. The characteristics of the study cohort are summarized in Table 1. The only significant differences in clinical and imaging characteristics between cohorts with a solitary HCA and those with multiple HCAs were the duration of oral contraception intake before diagnosis of HCA, which was lower in patients with a solitary HCA (mean, 12.4 years ± 7.8 vs 17.7 years ± 9.5, respectively; P = .003), and lesion sizes at diagnosis and follow-up, which were larger in patients with a solitary HCA (mean diameter at diagnosis, 6.1 cm ± 3.7 vs 3.5 cm ± 3.1, respectively; P < .001; mean diameter at follow-up, 5.9 cm ± 3.6 vs 2.2 cm ± 2.5, respectively; P < .001).

    Table 1: Characteristics of Study Population including Clinical and Pathologic Data

    Table 1:

    Inflammatory, HNF-1α–inactivated, and β-catenin–mutated exon 3 or exon 7–8 HCAs were encountered in 64 of 118 (54.2%), 46 of 118 (39.0%), and seven of 118 (5.9%) patients, respectively (Table E1 [online]). Eight of 118 (6.8%) patients had two different subtypes of HCAs—including four patients with inflammatory and HNF-1α–inactivated HCAs, two patients with inflammatory and β-catenin–mutated HCAs, one patient with inflammatory and unclassified HCAs, and one patient with HNF-1α–inactivated and β-catenin–mutated HCAs (Table E2 [online]).

    Intratumoral hemorrhage and malignancy were encountered at pathologic evaluation in 29.7% (35 of 118) and 5.9% (seven of 118) of the patients, respectively. Foci of malignancy developed in β-catenin–mutated HCAs in five of seven (71.4%) patients with β-catenin mutation—four with β-catenin mutation in exon 3 and one with β-catenin mutation in exons 7–8—and in two of 64 (3.1%) patients with inflammatory HCAs. None of the HNF-1α–inactivated HCAs in our cohort showed malignancy at pathologic assessment at baseline.

    In a per-lesion analysis, of the 44 HCAs with microscopic or macroscopic hemorrhage detected at pathologic evaluation, 26 of 90 (8.9%) were inflammatory HCAs, eight of 60 (13.3%) were HNF-1α–inactivated HCAs, five of six (83.3%) were Sonic the Hedgehog HCAs, four of six (66.7%) were β-catenin–mutated exon 3 HCAs, and one of two (50%) was a β-catenin–mutated exon 7–8 HCA. None (0%) of the six unclassified HCAs showed micro- or macroscopic hemorrhage at pathologic evaluation.

    Solitary and Multiple Adenomas: Long-term Evolution

    Median follow-up, calculated from baseline MRI to last available follow-up MRI, of the entire study cohort was 5 years (IQR, 3.0–7.5 years), and it was not different between the solitary and multiple HCA cohorts (5.1 years [IQR, 3.4–8.3 years] vs 4.9 years [IQR, 3.0–7.3 years], respectively; P = .62).

    Thirty-seven (90.2%) of 41 patients with a solitary HCA and 55 (71.4%) of 77 patients with multiple HCAs showed stable or regressive disease (Figs 2, 3; Table 2). Consequently, 22 (28.6%) of 77 patients with multiple HCAs showed progressive disease, including 12 (54.5%) with at least five lesions at baseline and one (4.5%) with malignancy at pathologic evaluation. Overall, new lesions were identified at follow-up in two (4.9%) of 41 patients with a solitary adenoma and 12 (15.6%) of 77 patients with multiple HCAs (P = .09).

    Figure 2:

    Figure 2: Flowchart shows long-term evolution of histologically proven solitary hepatocellular adenomas by pathologic proof. * Patient 1, inflammatory hepatocellular adenoma (HCA) with signs of malignancy; patient 2, β-catenin–activated HCA. † Patient 1, nuclear factor 1α–mutated HCA; patient 2, inflammatory HCA and history of tamoxifen intake for breast cancer.

    Download as PowerPointOpen in Image Viewer
    Figure 3:

    Figure 3: Flowchart shows long-term evolution of histologically proven multiple hepatocellular adenomas by pathologic proof.

    Download as PowerPointOpen in Image Viewer

    Table 2: Follow-up Per-Patient Analysis of Solitary and Multiple Adenomas

    Table 2:

    In the cohort of patients with a solitary lesion, inflammatory HCAs, HNF-1α–inactivated HCAs, and β-catenin–mutated exon 3 HCAs showed progression in two of 20 (10%), one of 15 (6.7%), and one of three (33.3%) patients, respectively. Of the 29 patients with resected solitary adenoma, only two (6.9%) showed progressive disease, and both of them had at-risk HCAs (one β-catenin–mutated in exon 3 and one inflammatory HCA with foci of malignancy) (Fig 2). Conversely, of the 12 patients with a solitary HCA not resected, only two (16.7%) had progression because of lesion growth. In the cohort of patients with multiple HCAs of a single subtype, only inflammatory HCAs and HNF-1α–inactivated HCAs showed progression in seven of 37 (18.9%) patients and 11 of 26 (42.3%) patients, respectively. Representative images of long-term follow-up of inflammatory HCAs and HNF-1α–inactivated HCAs are shown in Figures 4 and 5, respectively. Only one (33.3%) of the three patients with at least one β-catenin–mutated HCA showed progressive disease during long-term follow-up (Table E2 [online]). The long-term evolution in men (10 of 118 patients) with HCAs included progressive disease in four of the 10 patients (40%) (Appendix E1 [online]).

    Figure 4:

    Figure 4: Representative images show size changes in inflammatory hepatocellular adenomas (HCAs). Outcome in patients with only inflammatory HCAs in our study is also represented by numbers and percentages on the left. Top row: Axial contrast-enhanced MRI scans in a 30-year-old woman with HCAs, A, at baseline and, B, at 4-year follow-up show complete regression of inflammatory HCAs (arrows). Bottom row: Axial contrast-enhanced MRI scans in a 41-year-old woman with HCAs, C, at baseline and, D, at 6-year follow-up show substantial progression of inflammatory HCA (arrow). PD = progressive disease, RD = regressive disease, SD = stable disease.

    Download as PowerPointOpen in Image Viewer
    Figure 5:

    Figure 5: Representative images show size changes in hepatocyte nuclear factor 1α (HNF-1α)–inactivated hepatocellular adenomas (HCAs). Outcome in patients with only HNF-1α–inactivated HCAs in our study is also represented in numbers and percentages on the left. Top row: Axial contrast-enhanced MRI scans in a 47-year-old woman with HCAs, A, at baseline and, B, at 12-year follow-up show complete regression of HNF-1α–inactivated HCA (arrow). Bottom row: Axial contrast-enhanced MRI scans in a 42-year-old woman with HCAs, C, at baseline and, D, at 5-year follow-up show substantial progression of HNF-1α–inactivated HCA (arrow).

    Download as PowerPointOpen in Image Viewer

    Figure 6 summarizes long-term follow-up of HCAs according to lesion management; new lesions were observed in only two (6.2%) of 32 patients who underwent complete resection of HCAs, whereas progressive disease was demonstrated in 17 (31.5%) of 54 and seven (21.8%) of 32 patients who underwent partial resection or no resection, respectively. Specifically, of the 77 patients with multiple HCAs (Fig 3), three (3.9%) underwent complete resection of HCAs, and none of these three patients had new lesions at follow-up. The remaining 74 patients with multiple HCAs included 54 patients with incomplete resection and 20 patients with no surgery at all, and progressive disease was detected in 17 (31.5%) of 54 patients (ie, eight with lesion growth only, seven with new lesions, and two with lesion growth and new lesions) and five (25%) of 20 patients (ie, two with lesion growth, one with a new lesion, and two with both lesion growth and a new lesion), respectively.

    Figure 6:

    Figure 6: Flowchart shows long-term evolution of histologically proven hepatocellular adenomas according to management.

    Download as PowerPointOpen in Image Viewer

    Overall, seven HCAs initially smaller than 50 mm in five patients had progressed to a diameter of 50 mm or greater at overall follow-up. These seven lesions that progressed to a diameter of 50 mm or greater included one lesion in a patient with a solitary nonresected HNF-1α–inactivated HCA, three lesions in one patient with multiple HNF-1α–inactivated HCAs, two HNF-1α–inactivated HCAs in two patients, and one inflammatory HCA in the remaining patient. None of the patients had malignant transformation of HCAs during follow-up, including HCAs that increased in size. None of the patients experienced clinically relevant bleeding of their HCAs.

    Predictive Variables of Progression

    Neither surgery nor the presence of β-catenin–mutated subtype had an impact on progression (P = .99 and P = .67, respectively) (Table 3), whereas lower body mass index at diagnosis, symptoms at diagnosis, and presence of multiple HCAs were associated with higher probability of progressive disease at univariate (P = .001, P = .04, and P = .02, respectively) and multivariate (P = .006, P = .01, and P = .05, respectively) analyses. However, the number of lesions at diagnosis in patients with multiple HCAs was not associated with progressive disease (P = .54).

    Table 3: Patient and Lesion Variables in Progressive Disease in All Patients with Adenomas at Per-Patient Univariable Analysis

    Table 3:

    Overall, inflammatory HCAs showed a significantly greater reduction in size compared with HNF-1α–inactivated HCAs (Fig 7, Table 4, Fig E1 [online]), whereas baseline diameter was not significantly correlated with size changes (r = −0.01). Finally, in a per-lesion analysis with use of a cut-off value of 20% for defining progression of each lesion, 23 (27%) of 85 HNF-1α–inactivated HCAs and 13 (13%) of 103 inflammatory HCAs showed progression, and none of the HCAs within the remaining subtypes showed progressive disease. The risk of progression of HNF-1α–inactivated HCAs was greater—although marginally—compared with the remaining subtypes (non–HNF-1α–inactivated HCAs vs HNF-1α–inactivated HCAs: odds ratio, 0.20; P = .05). Of note, none of the 11 patients with unclassified, Sonic the Hedgehog, or β-catenin–mutated exon 7–8 HCAs showed progressive disease. In patients with multiple HCAs of a unique subtype only, HNF-1α–inactivated HCAs showed a higher rate of progression compared with inflammatory HCAs (11 of 26 [42.3%] vs seven of 37 [18.9%], P = .04) (Table 5) during a similar median follow-up period (5.5 years [IQR, 3.8–7.7 years] vs 5.25 years [IQR, 2.9–7.2 years], respectively; P = .17) and lower use (28 of 32 [87.5%] vs 45 of 45 [100%], P = .03) and shorter duration (mean, 12.0 years ± 7.5 vs 19.2 years ± 9.2; P = .001) of oral contraceptive intake.

    Figure 7:

    Figure 7: Boxplots show distribution of change in lesion diameter in percentage (baseline MRI examination at diagnosis to last available MRI examination) by subtype, mean, median, and first and third quartiles of changes in percentages by subtype. HCA = hepatocellular adenoma, HNF1α = hepatocyte nuclear factor 1α.

    Download as PowerPointOpen in Image Viewer

    Table 4: Estimated Effects by Subtype

    Table 4:

    Table 5: Per-Patient Analysis of Type of Progression in Patients with Multiple Inflammatory or HNF-1α–inactivated HCAs of Unique Subtype Showing Progressive Disease at Imaging Follow-up

    Table 5:

    Discussion

    Current European Association for the Study of the Liver (EASL) guidelines recommend continued surveillance with MRI for patients diagnosed with hepatocellular adenomas (HCAs) (1). However, to date, few studies have provided evidence regarding the natural history of HCAs at MRI. The studies that do exist were limited by short or midterm follow-up; use of combined diagnostic criteria, including imaging criteria, pathologic criteria, or both, and lack of use of MRI for follow-up (8–16). Our study, which analyzed the MRI long-term course of HCAs in pathologically proven and subtyped solitary and multiple HCAs, demonstrated that 90% of patients with a solitary HCA and 71% of patients with multiple HCAs showed stability or regression in the size of HCA after oral contraception withdrawal at a median MRI follow-up of 5 years. We also demonstrated that multiplicity of lesions and hepatocyte nuclear factor 1α (HNF-1α)–inactivated subtype were significantly (P = .02 and P = .05, respectively) associated with progression. In addition, surgery does not significantly change progression rate in patients with multiple HCAs (P = .18).

    The proportion of patients with progressive disease in the literature varies from 0% to 19% (6–16). Shao et al (25) have reported stability or regression in size in 95% of HCAs at an average follow-up of 43 months, which is slightly different compared with our rates of regression or stability (range, 71%–90%). However, these differences may be explained by the different definition of significant reduction in size (at least 20% decrease in size, compared with at least 30% decrease in size in our study), the inclusion of more than three lesions per patient (in our study, we analyzed no more than three lesions in each patient to prevent clustering bias), and the use of a per-lesion analysis instead of a per-patient analysis. Our results demonstrated new lesions in only two of 29 (6.9%) patients with a resected solitary HCA, and both of these patients had at-risk HCAs at baseline pathologic assessment, including HCA with presence of β-catenin mutation in exon 3 or foci of malignancy in resected tumor. Conversely, patients with solitary resected HCA without such worrisome features at pathologic assessment did not show new lesions after a median follow-up of 5.1 years. Therefore, our data may suggest that after resection of solitary HCAs, follow-up should be maintained in β-catenin–mutated HCAs or in the presence of foci of malignancy within the resected tumor. Discontinuation of follow-up may be discussed otherwise. One prior study by Karkar et al (12) assessed the tumor course in a cohort of 22 patients with resected solitary HCA and found new lesions in only two patients, whose resected HCA exceeded 10 cm. In our study, the rate of progression of nonresected solitary HCAs was 16% because of tumor size increase in all the patients showing progression; therefore, these patients need to undergo follow-up according to EASL guidelines.

    Our results show that HNF-1α–inactivated HCAs have a higher probability of progression compared with inflammatory HCAs (42% vs 19%, P = .04). These results agree with those of a recent study by Klompenhouwer et al (26) that demonstrated that inflammatory HCAs are more likely to regress during 1 and 2 years of follow-up. This may be potentially explained by different levels of estrogen dependency of different HCA subtypes (4,27). Unfortunately, as recently pointed out by Haring et al (15), to our knowledge, no study has included any subtype analysis of expression of androgen and estrogen receptors. Overall, 29% of multiple HCAs showed progression in our study, with a similar percentage in the biopsy and surgery groups (25% vs 29%, P = .18). Surprisingly, none of the patient- and lesion-related variables that have been previously reported to be risk factors for HCA appearance and progression (eg, body mass index) (14–16,28) were associated with tumor progression in our study. As a result, and in line with EASL guidelines (1), patients with multiple HCAs should be monitored regardless of prior surgery or patient- or lesion-related variables.

    Our study had several limitations. First, we had a very low number of unclassified, Sonic the Hedgehog, or β-catenin–mutated exon 7–8 HCAs (n = 11 of 118). Second, besides demographics, other clinical variables were analyzed in only the 95 (80.5%) of 118 patients who were reached for the telephone interview, thereby potentially skewing our results. However, a telephone interview is considered an effective method of data collection (29). Third, owing to the retrospective nature of our study, the lack of serial MRI examinations did not allow for analysis of the time to progression or valid assessment of tumor volume doubling times and therefore the identification of the optimal interval for follow-up MRI. Fourth, because of our stringent inclusion criteria, we excluded 145 patients, which could lead to selection bias. However, our aim was to fill the current need for robust data regarding long-term evolution of path-proven HCAs in which MRI was used as the reference standard. We also acknowledge that we did not collect data regarding changes in MRI signal features over time—including occurrence of hypersignal at T1-weighted imaging during follow-up that could correspond to hemorrhage—because this was beyond the scope of the study. However, no clinically relevant bleeding was identified during patient follow-up. Similar findings have been described in prior literature (4,5), including one study of MRI that was performed by our team (19). Finally, Response Evaluation Criteria in Solid Tumors, version 1.1, assessment of outcome as progressive, stable, or regressive disease is not validated for benign liver diseases. However, Response Evaluation Criteria in Solid Tumors, version 1.1, is the most known, widely adopted, and reproducible system in the oncologic setting, and it has been used in a very recent study (15).

    In conclusion, our data suggest that in patients with a solitary hepatocellular adenoma, surveillance may be potentially discontinued after resection, except in β-catenin–mutated hepatocellular adenomas or with foci of malignancy within the resected tumor. Our data also show that patients with multiple hepatocellular adenomas are more likely to have progressive disease regardless of surgery, with hepatocyte nuclear factor 1α–inactivated hepatocellular adenomas being the most common subtype showing progression. The lower exposure to oral contraceptives in patients with hepatocyte nuclear factor 1α–inactivated hepatocellular adenomas compared with inflammatory ones suggests the presence of other possible influencing factors for the development and progression of these lesions. Further studies analyzing changes in MRI signal features over time and other possible influencing factors of progression in patients with multiple hepatocellular adenomas are needed.

    Disclosures of Conflicts of Interest: F.V. disclosed no relevant relationships. M.R. disclosed no relevant relationships. M.D.B. disclosed no relevant relationships. F.C. disclosed no relevant relationships. K.R.C. disclosed no relevant relationships. S.D. disclosed no relevant relationships. O.S. disclosed no relevant relationships. D.V. disclosed no relevant relationships. J.Z. disclosed no relevant relationships. V.P. disclosed no relevant relationships. V.V. Activities related to the present article: disclosed no relevant relationships. Activities not related to the present article: institution received grants from HECAM, RIDHO, and H2020 FORCE; gave lectures for Supersonic, Canon, and Guerbet. Other relationships: disclosed no relevant relationships.

    Author Contributions

    Author contributions: Guarantors of integrity of entire study, F.V., O.S., V.V.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; approval of final version of submitted manuscript, all authors; agrees to ensure any questions related to the work are appropriately resolved, all authors; literature research, F.V., M.R., S.D., D.V., J.Z., V.P., V.V.; clinical studies, F.V., M.R., F.C., S.D., O.S., D.V., J.Z., V.P., V.V.; statistical analysis, M.R., K.R.C., J.Z., V.P.; and manuscript editing, F.V., M.R., M.D.B., F.C., S.D., O.S., D.V., J.Z., V.P., V.V.

    References

    • 1. European Association for the Study of the Liver (EASL). EASL Clinical Practice Guidelines on the management of benign liver tumours. J Hepatol 2016;65(2):386–398. Crossref, Medline, Google Scholar
    • 2. Bonder A, Afdhal N. Evaluation of liver lesions. Clin Liver Dis 2012;16(2):271–283. Crossref, Medline, Google Scholar
    • 3. Karhunen PJ. Benign hepatic tumours and tumour like conditions in men. J Clin Pathol 1986;39(2):183–188. Crossref, Medline, Google Scholar
    • 4. Nault JC, Couchy G, Balabaud C, et al. Molecular Classification of Hepatocellular Adenoma Associates With Risk Factors, Bleeding, and Malignant Transformation. Gastroenterology 2017;152(4):880–894.e6. Crossref, Medline, Google Scholar
    • 5. van Aalten SM, de Man RA, IJzermans JNM, Terkivatan T. Systematic review of haemorrhage and rupture of hepatocellular adenomas. Br J Surg 2012;99(7):911–916. Crossref, Medline, Google Scholar
    • 6. Bioulac-Sage P, Laumonier H, Couchy G, et al. Hepatocellular adenoma management and phenotypic classification: the Bordeaux experience. Hepatology 2009;50(2):481–489. Crossref, Medline, Google Scholar
    • 7. Herman P, Coelho FF, Perini MV, Lupinacci RM, D’Albuquerque LA, Cecconello I. Hepatocellular adenoma: an excellent indication for laparoscopic liver resection. HPB 2012;14(6):390–395. Crossref, Medline, Google Scholar
    • 8. de’Angelis N, Memeo R, Calderaro J, et al. Open and laparoscopic resection of hepatocellular adenoma: trends over 23 years at a specialist hepatobiliary unit. HPB 2014;16(9):783–788. Crossref, Medline, Google Scholar
    • 9. Chun YS, Parker RJ, Inampudi S, et al. Imaging Surveillance of Hypervascular Liver Lesions in Non-Cirrhotic Patients. J Gastrointest Surg 2016;20(3):564–567. Crossref, Medline, Google Scholar
    • 10. Dokmak S, Paradis V, Vilgrain V, et al. A single-center surgical experience of 122 patients with single and multiple hepatocellular adenomas. Gastroenterology 2009;137(5):1698–1705. Crossref, Medline, Google Scholar
    • 11. Shafizadeh N, Genrich G, Ferrell L, Kakar S. Hepatocellular adenomas in a large community population, 2000 to 2010: reclassification per current World Health Organization classification and results of long-term follow-up. Hum Pathol 2014;45(5):976–983. Crossref, Medline, Google Scholar
    • 12. Karkar AM, Tang LH, Kashikar ND, et al. Management of hepatocellular adenoma: comparison of resection, embolization and observation. HPB 2013;15(3):235–243. Crossref, Medline, Google Scholar
    • 13. Klompenhouwer AJ, Bröker MEE, Thomeer MGJ, Gaspersz MP, de Man RA, IJzermans JNM. Retrospective study on timing of resection of hepatocellular adenoma. Br J Surg 2017;104(12):1695–1703. Crossref, Medline, Google Scholar
    • 14. Bunchorntavakul C, Bahirwani R, Drazek D, et al. Clinical features and natural history of hepatocellular adenomas: the impact of obesity. Aliment Pharmacol Ther 2011;34(6):664–674. Crossref, Medline, Google Scholar
    • 15. Haring MPD, Gouw ASH, de Haas RJ, Cuperus FJC, de Jong KP, de Meijer VE. The effect of oral contraceptive pill cessation on hepatocellular adenoma diameter: A retrospective cohort study. Liver Int 2019;39(5):905–913. Crossref, Medline, Google Scholar
    • 16. Klompenhouwer AJ, Sprengers D, Willemssen FEJA, Gaspersz MP, Ijzermans JNM, De Man RA. Evidence of good prognosis of hepatocellular adenoma in post-menopausal women. J Hepatol 2016;65(6):1163–1170. Crossref, Medline, Google Scholar
    • 17. Zulfiqar M, Sirlin CB, Yoneda N, et al. Hepatocellular adenomas: understanding the pathomolecular lexicon, MRI features, terminology, and pitfalls to inform a standardized approach. J Magn Reson Imaging 2019 Aug 16 [Epub ahead of print]. Crossref, Medline, Google Scholar
    • 18. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg 2014;12(12):1495–1499. Crossref, Medline, Google Scholar
    • 19. Ronot M, Bahrami S, Calderaro J, et al. Hepatocellular adenomas: accuracy of magnetic resonance imaging and liver biopsy in subtype classification. Hepatology 2011;53(4):1182–1191. Crossref, Medline, Google Scholar
    • 20. Laumonier H, Bioulac-Sage P, Laurent C, Zucman-Rossi J, Balabaud C, Trillaud H. Hepatocellular adenomas: magnetic resonance imaging features as a function of molecular pathological classification. Hepatology 2008;48(3):808–818. Crossref, Medline, Google Scholar
    • 21. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45(2):228–247. Crossref, Medline, Google Scholar
    • 22. Fitzmaurice G, Laird N, Ware J. Applied Longitudinal Analysis. 2nd ed. Hoboken, NJ: Wiley, 2011. Crossref, Google Scholar
    • 23. Pinheiro J, Bates D, DebRoy S, Sarkar D. nlme: Linear and Nonlinear Mixed Effects Models, R package 2019, version 3.1-141. https://CRAN.R-project.org/package=nlme. Accessed October 27, 2019. Google Scholar
    • 24. Bates D, Machler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. J Stat Softw 2015;67(1):1–48. Crossref, Google Scholar
    • 25. Shao N, Pandey A, Ghasabeh MA, et al. Long-term follow-up of hepatic adenoma and adenomatosis: analysis of size change on imaging with histopathological correlation. Clin Radiol 2018;73(11):958–965. Crossref, Medline, Google Scholar
    • 26. Klompenhouwer AJ, Alblas M, van Rosmalen BV, et al. Development and validation of a model to predict regression of large size hepatocellular adenoma. Am J Gastroenterol 2019;114(8):1292–1298. Crossref, Medline, Google Scholar
    • 27. Nault JC, Paradis V, Cherqui D, Vilgrain V, Zucman-Rossi J. Molecular classification of hepatocellular adenoma in clinical practice. J Hepatol 2017;67(5):1074–1083. Crossref, Medline, Google Scholar
    • 28. Bröker MEE, Gaspersz MP, Klompenhouwer AJ, et al. Inflammatory and multiple hepatocellular adenoma are associated with a higher BMI. Eur J Gastroenterol Hepatol 2017;29(10):1183–1188. Crossref, Medline, Google Scholar
    • 29. Musselwhite K, Cuff L, McGregor L, King KM. The telephone interview is an effective method of data collection in clinical nursing research: a discussion paper. Int J Nurs Stud 2007;44(6):1064–1070. Crossref, Medline, Google Scholar

    Article History

    Received: Aug 10 2019
    Revision requested: Sept 24 2019
    Revision received: Dec 29 2019
    Accepted: Jan 10 2020
    Published online: Mar 17 2020
    Published in print: May 2020