Managing Focal Nodular Hyperplasia of the Liver: Surgery or Minimally-invasive Approaches? A Review of the Preferable Treatment Options

Discussion of Treatment Options

1. Conservative approach. Silent lesions smaller than 4 cm in diameter can simply be observed and surveilled through imaging studies (1-5, 7-21, 24-27). Such an approach is justified by the minimal risk for bleeding, necrosis and malignant degeneration (12). Furthermore, spontaneous regression, although rare, can always occur, especially following withdrawal of birth control pills in females (1-4).

2. Surgery. All lesions becoming symptomatic or developing an increase in tumor size should deserve treatment. Traditionally, open or laparoscopic resection of FNH (associated or not with hepatic artery ligation) has been advocated as the preferred curative treatment in the case of changes in clinic, radiological or clinicoradiological characteristics (1-5, 7-20). However, there are only limited data evaluating the risk–benefit balance following this type of measure: the morbidity rate has been reported to reach 27% and 9% after open and laparoscopic surgery, respectively, with a mortality rate of 2% (1, 5). Some authors warn against such approaches because of these rates after surgery, considering them as serious and unacceptable in the management of a benign disease (12). Related complications include intra- or postoperative bleeding, biliary fistula, ileus, intra-abdominal abscess, liver failure, pleural effusion and deep vein thrombosis (1). Surgery is, however, the recommended procedure in emergency situations characterized by hemoperitoneum due to FNH rupture (21). To date, only one case of recurrent FNH has been described following surgery: a 22-year-old man with a previous history of left-lobe FNH who in fact developed three bilobal lesions 3 years after the first intervention (22). Surgery was the kind of treatment selected also for recurrences. Differently from HCC, no recurrence of FNH has been documented following laparoscopic liver resection thus far (23).

3. Radiofrequency ablation. In 2010, Hedayati and colleagues described the first (and, to date, the sole) case report of RFA of a 2.2 cm symptomatic FNH in a 21-year-old symptomatic woman. The procedure was safe but only partially effective since a small residual was noted at 2-month follow-up computed tomography (2). Hence, longer-term follow-up is essential to ultimately verify the effectiveness of RFA (2).

4. Transarterial embolization. We found only 17 articles dealing with TAE applied to FNH (Table I) (1, 3-5, 8-20). Altogether, 128 patients with FNH patients were effectively treated: TAE was performed at least once per patient. Usual indications were unresectable lesions with/without symptoms, patients unfit for surgery, necessity of sparing normal hepatic parenchyma, multiple FNHs, patient refusal to undergo surgery and pediatric patients (1, 3-5, 8-20). Additionally, TAE has also been used as a form of bridge therapy to surgery in selected cases for size reduction and pain control (5, 16). Finally, TAE can also be performed in emergency settings as a ready method of arresting hemorrhage arising from a ruptured FNH (arteriographic hemostatic TAE has been applied to only five cases as of 2017) (22-25). On the other hand, due to the uncertainty about the natural history of FNH and its correlation with HCC, some authors still prefer surgical resection to this minimally invasive approach (15). Besides these oncological considerations, given the vascular nature of this disease, TAE appears a logical strategy of treatment (1, 15). In fact, FNH is usually fed by a single artery which supplies blood from the center to the periphery of the tumor (1, 11, 15). Commonly, in its course, the feeding artery provides no or little arteriovenous shunting and is unaccompanied by portal vein branches or biliary ducts (50-60% of cases); this kind of one-vessel FNH is extremely amenable to a single administration of TAE (1, 3, 15). In contrast, in the case of multiple small feeding vessels, the necessity for repeat TAE should be anticipated (1, 3, 11, 15). In contrast to surgery, TAE does not require general anesthesia, entails a significantly shorter length of stay (1.4 days versus 9.5 and 1.9 days, respectively, for open, and laparoscopy-assisted surgery) as well as lower morbidity and mortality rates (1). Such a practice is not free from possible complications: main consequences derive from superinfection of embolized tissue and massive discharge of toxic radicals. However, most minor (liver abscess, infection) as well as major adverse events (post-embolization syndrome/tumor lysis syndrome, liver or gallbladder infarction, pneumonia, pleural effusion, renal failure and death) occurred in patients with severe underlying liver diseases, such as portal vein obstruction, cirrhosis, chronic hepatitis and biliary obstruction (3, 12). Out of these, inadequacy of the portal blood flow seems to be the most important predisposing risk factor (7). Post-embolization syndrome usually starts within 24-48 hours of embolization and can last up to 7 days (1). It is characterized by pain at the site of treatment, fever or low-grade temperature, nausea, vomiting, malaise and leukocytosis; its severity depends on the size of the embolized tissue (1). To date, only one case of local recurrence was found in one treated patient at the 54-month follow-up (4).

Currently, what appears still unsolved with TAE is the choice of the best embolization material (1, 3-5, 7-19). In fact, since the first description of TAE in 1979, various agents have been tested and progressively used in the studies in literature: low-diameter polyvinyl alcohol (PVA) particles and ethiodized oil (lipiodol) emulsions have been the most frequently adopted embolic agents (Table I). Hepatic intra-arterial administration of PVA causes the permanent occlusion of the terminal hepatic arterioles, whereas the concomitant injection of lipiodol is thought to reduce the blood flow of the terminal portal venules temporarily and embolize the small abnormal vessels within the hyperplastic lesion (3). However, because FNH relies much less on the portal circulation than other tumor types, the use of lipiodol could be a matter of debate. On the other hand, in addition to helping cause ischemia, there is a second effect using this material:direct tumoricidal activity (3). In fact, for reasons that remain unknown, lipiodol adheres to and is phagocytated by hyperplastic Kupffer cells, causing their lysis (3, 11). More recently, some authors have introduced the employment of bleomycin in combination with other agents (such as PVA and iodized oil); this mechanism adds to the sclerosing effect exerted by bleomycin the embolic action of iodinated oil associated with PVA (4). More precisely, bleomycin is thought to affect vascular endothelium inducing a secondary formation of intraluminal microthrombi, thereby resulting in the destruction of the feeding artery of the lesion (4). Sclerosing cholangitis, interstitial pneumonia and pulmonary fibrosis have been described as the major possible complications related to the use of this agent (18). Of note, whatever materials have been employed, the use of microcatheters seems to permit a more precise deposition of the embolization particles in the small feeding arteries (selective and superselective TAE), minimizing the risk of occluding major arterial branches (12, 18).