Focal nodular hyperplasia after systemic chemotherapy: Pathological features of a series of 15 cases

Chemotherapy, particularly oxaliplatin, has been associated with the development of focal nodular hyperplasia (FNH). Imaging diagnosis of FNH is well standardized, but it can be misdiagnosed as liver metastasis. The aim of this study was to describe the pathological features of FNH occurring after systemic chemotherapy.


| INTRODUC TI ON
Focal nodular hyperplasia is a benign, frequent, non-neoplastic lesion, which usually occurs in a normal background liver or more rarely in a context of vascular liver disease. [1][2][3][4] Focal nodular hyperplasia (FNH) is thought to be a hyperplastic response to a localized vascular abnormality. 5,6 Its pathogenesis implies dysregulated expression of angiopoietin 1 and 2, leading to the development of dystrophic vessels, seen in typical FNH. 7,8 This local increased blood flow hyperperfuses the parenchyma, leading to focal hyperplasia. 6 In most cases, the diagnosis of FNH is incidental on imaging. On magnetic resonance imaging (MRI), FNH is a non-encapsulated, homogeneous nodule. The central scar present in two-thirds of cases is typically more intense on T1-weighted imaging and shows delayed enhancement 9,10 Because FNH presents a typical imaging appearance, liver biopsy is not required in most cases. 11,12 If performed, biopsy of the presumed FNH has excellent diagnostic performance in typical cases. 12 However, in atypical cases (lacking the central scar and some histopathological criteria), the diagnosis of FNH could be challenging, not only on imaging but also on biopsy. 13 Because FNH cases are usually asymptomatic, with few complications, neither surgical resection nor follow-up is required. Therefore, an accurate diagnosis of FNH is mandatory. 14 FNH developing after systemic chemotherapy was first reported in children with malignant tumours treated with high-dose chemotherapy or undergoing chemotherapy before haematopoietic stem cell transplantation as well as more recently in adults with colorectal cancer after an oxaliplatin-based regimen. [15][16][17][18][19][20] Observations are increasing of chemotherapeutic damage to the non-tumoral liver, particularly associated with oxaliplatin treatment, the main drug in colorectal cancer chemotherapeutic protocols. 21,22 Reported liver injuries induced by oxaliplatin include sinusoidal obstruction syndrome (SOS) and nodular regenerative hyperplasia (NRH). 21,23 In a few cases, such vascular changes can be so severe as to induce liver failure and increase morbidity after hepatectomy. 24,25 SOS and NRH observed in patients receiving oxaliplatin may be because of a toxic effect of the drug on sinusoidal endothelial cells leading to disrupted sinusoidal wall integrity. 23 Oxaliplatin-induced vascular liver injury is well described, 21 but the association with FNH has been reported in only a limited number of cases including case reports (3 cases) and a series of 14 cases diagnosed by MRI. [18][19][20] The aim of the study was to report a series of surgically resected FNH cases occurring de novo after chemotherapy for liver metastasis and to describe their pathological features.

| Patients
From our pathological files for 1990-2021, we retrospectively retrieved data for 15 patients with liver metastasis treated with systemic chemotherapy in whom liver lesions diagnosed as FNH on surgical resection developed during follow-up. As required by French legislation, written consent was obtained for all patients.
Clinical data collected included age at surgery, sex, type and origin of primary cancer, history of chemotherapeutic treatment.

| Pathological assessment
All archival slides (originally prepared from formalin-fixed and paraffin-embedded samples) were reviewed, including slides of FNH and non-tumoral liver located at a distance from the FNH.
The morphological analysis was based on haematoxylin eosin and safran (HES) and reticulin staining. An immunohistochemical study with glutamine synthetase (GS; glutamine synthetase-6, 1:500, BD Biosciences) was performed for all FNH cases. Macroscopic features of FNH (number, size, aspect) were recorded according to the initial pathological report.
Two pathologists (ABr and ABe) systematically reviewed slides.
Pathological criteria evaluated on FNH are in Table 1. A GS immunostaining pattern was also recorded. In FNH, strong cytoplasmic staining of broad anastomosing groups of hepatocytes was defined as a characteristic GS geographical map-like expression pattern.
The following features were evaluated in the parenchymal liver located at a distance from the FNH or liver metastasis: SOS, intensity of sinusoidal dilatation, NRH, atrophy, steatosis and fibrosis. Liver atrophy was defined as the reduced size of the lobule illustrated by portal tracts that become closer together. The fibrosis stage was evaluated according to Kleiner scoring and steatosis according to the SAF score system. 26,27

| Clinical characteristics
Clinical characteristics of patients are summarized in Table 2. The mean age of the 15 patients (13 women) was 56 years (range 44-78).

Lay summary
We report a series of 15
In 60% (9/15) of cases, the GS immunostaining had the typical map-like expression pattern. Three (20%) cases showed a heterogeneous pattern, two (13%) cases moderate diffuse staining and one (7%) case no staining. In these cases, lesional hepatocytes exhibited neither SAA nor aberrant nuclear expression of β-catenin.
Five (33%) cases exhibited steatosis, and in all of these cases, steatosis was also present in FNH. Among these five cases, two patients were overweight, one had high blood pressure and one had metabolic syndrome. No steato-hepatitis features were present. Seven (47%) cases exhibited parenchymal atrophy. The fibrosis stage was F0 in six (40%) cases, F1 in four (27%) cases, F2 in four (27%) cases and F3 in one (7%) case.

| D ISCUSS I ON
We report 15 cases of FNH occurring after chemotherapy, 11 of these after oxaliplatin-based chemotherapy. To our knowledge, Another relevant finding is the high rate of intralesional steatosis (53%) in this context, which was frequently associated with steatosis in background liver (62%) and with at least one risk factor for metabolic syndrome. Indeed, in the literature, the rate of intralesional steatosis varies greatly (16.7%-52%). 1 reported. 23 Moreover, FNH has been also reported in children after high-dose chemotherapy for cancer or haematopoietic stem cell transplantation. [15][16][17][34][35][36][37] In these studies, the variety of cancers affecting patients is wide, and the occurrence of FNH is more likely linked to the nature of treatment received rather than the type of cancer.
Otherwise, the prime cause of FNH is thought to result from hyperarterialization as a result of portal vein injury, which leads to arterial proliferation and arteriovenous shunts. 6,38,39 An association between FNH and the presence of SOS may lead to the development, in some cases, of NRH. 17,18 The high rate of parenchymal lesions we observed, including SOS and NRH, could support this hypothesis, even though a control series is needed to confirm it.
The main limitations of our series include the monocentric retrospective study, which prevented us from reviewing imaging data in all cases and then performing a radiopathological correlation.
In conclusion, our series highlights that FNH occurring during the follow-up of patients after systemic chemotherapy (mostly based on oxaliplatin) is atypical on morphology and develops on a damaged background liver with SOS and NRH. Considering the wide use of oxaliplatin for colorectal carcinoma, a potential diagnosis of FNH should be systematically discussed in the follow-up of patients and not be misinterpreted as metastasis.

CO N FLI C T O F I NTE R E S T S
The authors do not have any disclosures to report.