A Limited Immunohistochemical Panel Can Subtype Hepatocellular Adenomas for Routine Practice

CRP Versus SAA for Identification of Inflammatory Hepatic Adenomas

Subtyping hepatic adenomas is important for patient management due to differing complication risks. Immunohistochemical staining with C-reactive protein (CRP) and serum amyloid-A (SAA) is widely accepted as a surrogate for molecular classification to identify inflammatory hepatocellular adenomas. Limited data, however, has been published on how these 2 stains compare for sensitivity. We conducted a large, multicenter, retrospective study to examine the sensitivity and staining characteristics of CRP and SAA in inflammatory hepatic adenomas, with focal nodular hyperplasia (FNHs) as a control group. Inflammatory adenomas were identified in 133 patients (average age 37 years, 109 were female). In all, 69.9% of cases were resection specimens and 90.2% of all cases showed positive staining for both CRP and SAA; 10 (7.5%) were positive for CRP only and 3 (2.3%) were positive for SAA only. CRP was more sensitive than SAA (97.74% vs. 92.48%, P -value = 0.0961) and showed more extensive and intense staining, with a significantly higher modified H-score ( P <0.001). Focal nodular hyperplasia can also show positive CRP and SAA staining but with a lower modified H-score ( P <0.0001). Based on beta-catenin and glutamine synthetase staining, 26 of inflammatory adenomas also had beta-catenin activation (19.5%). All 3 cases with positive SAA and negative CRP staining were beta-catenin activated. In contrast, the proportion of cases that were CRP positive and SAA negative was similar regardless of beta-catenin activation. The data affirms the strategy of using both CRP and SAA immunostains for hepatic adenoma subtyping and raises the awareness of the highly variable nature of SAA staining characteristics.

Subtyping of hepatocellular adenoma: a machine learning-based approach

Subtyping of hepatocellular adenoma (HCA) is an important task in practice as different subtypes may have different clinical outcomes and management algorithms. Definitive subtyping is currently dependent on immunohistochemical and molecular testing. The association between some morphologic/clinical features and HCA subtypes has been reported; however, the predictive performance of these features has been controversial. In this study, we attempted machine learning based methods to select an efficient and parsimonious set of morphologic/clinical features for differentiating a HCA subtype from the others, and then assessed the performance of the selected features in identifying the correct subtypes. We first examined 50 liver HCA resection specimens collected at the University of Washington and Kobe University/Kings College London, including HNF1α-mutated HCA (H-HCA) (n = 16), inflammatory HCA (I-HCA) (n = 20), beta-catenin activated HCA (β-HCA) (n = 8), and unclassified HCA (U-HCA) (n = 6). Twenty-six morphologic/clinical features were assessed. We used LASSO (least absolute shrinkage and selection operator) to select key features that could differentiate a subtype from the others. We further performed SVM (support vector machine) analysis to assess the performance (sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy) of the selected features in HCA subtyping in an independent cohort of liver resection samples (n = 20) collected at the University of Wisconsin-Madison. With some overlap, different combinations of morphologic/clinical features were selected for each subtype. Based on SVM analysis, the selected features classified HCA into correct subtypes with an overall accuracy of at least 80%. Our findings are useful for initial diagnosis and subtyping of HCA, especially in clinical settings without access to immunohistochemical and molecular assays.

Applying criteria for hepatocellular neoplasm of uncertain malignant potential reclassifies more than half of hepatocellular adenomas

OBJECTIVES

Hepatocellular adenoma (HCA) is a benign well-differentiated hepatocellular neoplasm that can be difficult to distinguish from well-differentiated hepatocellular carcinoma (HCC). The term "well-differentiated hepatocellular neoplasm of uncertain malignant potential" (HUMP) has been proposed for neoplasms resembling HCAs, but arising in atypical clinical situations (in females over 50 years old or under 15, in males, with anabolic steroid use, or in some congenital conditions), and/or with atypical pathological features (focal cytological/architectural atypia, β-catenin activation, or focal reticulin loss) insufficient for an unequivocal diagnosis of HCC.

METHODS

This study evaluated HUMP criteria on 42 previously diagnosed HCAs from 33 patients.

RESULTS

Twenty-six (62%) masses from 21 patients were classified as HUMPs. Eleven (42%) had focal cytological atypia, and two (8%) had focal architectural atypia. Four (15%) showed focal reticulin loss. Five (19%) showed evidence of β-catenin activation. Four (12%) HUMP patients were male.

CONCLUSIONS

In this series, HUMP did not correlate with an increased rate of synchronous or metachronous HCC compared to HCA. Clinical colleagues may not accept such a high rate of tumors placed in a category of "uncertain malignant potential". Additional study is warranted to refine criteria for designating well-differentiated hepatocellular neoplasms as of uncertain malignant potential.

Relevance of morphological features for hepatocellular adenoma classification in pathology practice

Background

Gene mutations correlated with histological and immunohistochemical phenotypes of hepatocellular adenoma were recently identified. Based on these findings, four adenoma subtypes were distinguished. We classify hepatocellular adenoma (HCA) into subtypes based on histologic and immunohistochemical findings and verify the contribution of histological features in pathology practice.

Methods

Thirty hepatocellular adenomas were classified in subtypes. Sinusoidal dilatation, ductular reaction, pseudoportal tracts, pseudoglands, steatosis, inflammatory infiltrate and cellular atypia were analyzed, as well as liver fatty acid binding protein, β catenin, serum amyloid A, glutamine synthetase, and C-reactive protein antibodies.

Results

Histologically, eleven adenomas were classified as HNF1A inactivated (HHCA), five were β-catenin-activated (bHCA) and fourteen were inflammatory adenoma (IHCA). Steatosis was found in all HHCA and was predominantly severe. Sinusoidal dilatation and inflammatory infiltrate were present in all IHCA. Ductular reaction, pseudoportal tracts and cellular atypia were observed in 71.4, 85.7 and 42.8%, respectively. Pseudoglands were present in 60% and cellular atypia in 80% of bHCA. According to immunohistochemistry, 11 were HHCA; 1 was bHCA; 17 were IHCA, among which 5 were β-catenin activated IHCA; and 1 was unclassified UHCA (UHCA). Superior concordance between the histological and immunohistochemical classifications was found for HHCA (К = 0.854) and IHCA (К = 0.657).

Conclusion

Approximately 90% of adenomas may be diagnosed by subgroup based only on morphological features. When aberrant β catenin nuclear staining is not found, glutamine synthetase positivity is useful for diagnosis, although supplementary molecular analysis may be necessary.