Granulocyte-colony stimulating factor use in alcohol-associated hepatitis: is it time to promote liver regeneration?

Severe alcohol-associated hepatitis (sAH) is a life-threatening condition characterized by an abrupt onset of jaundice, malaise, decompensated liver disease, coagulopathy, and a Model of End-stage Liver Disease (MELD) score > 20 or a modified Maddrey’s discriminant function (mDF) ≥32 (1). sAH is often associated with bacterial infections and multiorgan failure, reaching a mortality of around 40–50% at 6 months (2,3). To date, scarce specific therapeutic agents are available to treat sAH in clinical practice (4). Currently, corticosteroids represent the first line of pharmacological therapy and are recommended by clinical guidelines (5–7). However, evidence supporting the use of corticosteroids is controversial, and some centers do not use them routinely. For example, the Steroids or Pentoxifylline for Alcoholic Hepatitis (STOPAH) study, a randomized clinical trial including 1,103 individuals with sAH in the United Kingdom, evidenced a non-significant reduction in mortality at 28 days in those who underwent corticosteroids but no pentoxifylline (PTX) (2). Further studies, including in individuals with a more severe liver-related condition, have shown that corticosteroids could decrease 30-day mortality in sAH, but this benefit is lost at 90-day mortality (3). In particular, the maximum benefit from corticosteroid use was observed in individuals with a MELD score between 25–39. Corticosteroid therapy is based on 40 mg of oral prednisolone per day, which is extended by 4 weeks in those with initial response according to the Lille model (8,9). Unfortunately, infections are common in patients with sAH, which limits the wide use of corticosteroids for sAH in clinical practice (10). Additionally, corticosteroids do not address the impaired hepatic regeneration associated with sAH.

Although corticosteroids are recommended, sAH-related mortality remains high, with approximately 35% mortality within 6 months (11). Beyond the use of corticosteroids for sAH, there are only a few potential therapeutic options that may be considered for sAH. One decade ago, a landmark study conducted in France and Belgium demonstrated that early liver transplantation is a therapeutic option for those patients who do not respond to medical therapy, reaching a survival comparable to other indications of liver transplantation (12). Since then, multiple centers have included sAH as an indication for early liver transplantation regardless of length of alcohol abstinence, constituting the fastest-growing indication for liver transplantation in the United States (13) and changing paradigms in the treatment of alcohol use disorder in patients with advanced liver disease (14). Unfortunately, liver transplantation is an exceptional therapy in most regions worldwide and several patients could not be eligible for early liver transplantation. Therefore, the development of novel drugs is essential to increase the therapeutic armamentarium of options for sAH.

In this issue of the American Journal of Drug and Alcohol Abuse, Fangfang Duan et al. report the results of a network meta-analysis to compare different treatments for sAH. They included 31 randomized controlled trials, including 19 treatment regimens. The primary outcome was 1-month mortality, while other outcomes were medium-term survival, long-term survival, and complications (i.e., infections, gastrointestinal bleeding, and hepatorenal syndrome [HRS]). They identified that granulocyte-colony stimulating factor (G-CSF) plus PTX or corticosteroids compared to placebo were significantly associated with lower 1-month mortality in sAH. The combination of G-CSF plus PTX also demonstrated an improved 3-month survival and a reduced incidence of infections.

In particular, G-CSF mobilizes hematopoietic stem and immune cells and represents a potential therapeutic agent in sAH. In animal models, the G-CSF accelerates and enhances hepatocyte proliferation in partially hepatectomized rats (15). In animal models of hepatitis, G-CSF promotes hematopoietic stem cell mobilization, induces liver regeneration, and improves survival (16). Also, a proof of concept clinical study in moderate and severe AH patients evidenced that G-CSF administration mobilized CD34+ stem cells, increased circulating hepatic growth factor, and induced proliferation of hepatic progenitor cells in liver-biopsy specimens at 7 days after administration (17). Further studies were conducted in patients with acute-on-chronic liver failure (ACLF). In particular, a study including sAH patients (57.4%) evidenced that G-CSF use could improve survival at 2 months, decreasing the risk of HRS, hepatic encephalopathy, and sepsis (18). In terms of implementation in clinical practice, G-CSF and PTX are widely available worldwide, are easy to use, and have a lower cost compared to other therapies (including liver transplantation).

Some emerging evidence on G-CSF supported its use in clinical practice. However, a recent European randomized controlled trial including 176 patients with ACLF showed that G-CSF did not improve 90-day transplant-free survival, liver function scores, or occurrence of infections; these results were confirmed in an additional analysis focused specifically on the alcohol-related ACLF subgroup (19). Another multicenter open-label randomized trial of pegfilgrastim including 34 patients with sAH did not demonstrate a survival benefit at 90 days, nor did it show differences in acute kidney injury (AKI), HRS, hepatic encephalopathy, or infections (20). In addition, a meta-analysis including seven studies in ACLF patients (84.8% had AH) demonstrated a significant benefit of G-CSF use in Asian individuals but in European individuals, probably due to a high heterogeneity and selection bias (21). Thus, this contradictory evidence has raised concerns about the potential benefit of G-CSF administration in sAH.

Among the 31 enrolled studies included in this network meta-analysis, 9 had a low bias, 12 showed a high bias, and 10 had a bias. Also, 21 of them were published before 2013 (twelve between 1971–1998). Therefore, although heterogeneity was low in several analyses, supportive management has undoubtedly evolved in sAH during the last five decades. The definition of severity is also a critical issue, since a recent global study including 3,380 sAH patients demonstrated a benefit of corticosteroid use mainly in patients with a more severe condition (MELD score between 25–39) (3). Unfortunately, although most clinicians and studies use mDF to define severity, the mDF has lower diagnostic accuracy in sAH (22). Similarly, most studies perform a clinical diagnosis of AH over a liver biopsy. Although the National Institute on Alcohol Abuse and Alcoholism (NIAAA) consensus criteria have facilitated the management of these patients, the clinical diagnosis has a lower sensitivity and specificity compared to a liver biopsy (63% and 78%, respectively) (23). Alcohol abstinence and return to drinking is another important issue, especially in assessing long-term mortality. For example, a study including 398 French patients with sAH demonstrated that alcohol consumption (≥30 g/day) was associated with mortality after 6 months, observing a dose-dependent relationship between alcohol intake and mortality in the long term (24). Due to these relevant differences in diagnosis and assessment of severity, studies included in this network meta-analysis should be carefully considered. Being infections one of the leading complications and cause of mortality, the use of G-CSF could also make it difficult to interpret the white blood cell (WBC) count, and consequently, the clinical suspicion of infections. For example, in the phase II study using pegfilgrastim, 9 (50%) patients had a WBC count > 30,000/mm³ on day 8 after one single dose (20). Also, one of the largest trials reported 7 serious G-CSF-related adverse reactions in 4 patients, one related to a WBC count above the threshold of 70,000/mm³. In the case of PTX, the STOPAH trial did not evidence a higher risk of adverse events compared to the placebo. However, after its negative results, the use of PTX has been abandoned in the routine management of sAH.

We consider that the results that emerged from this network meta-analysis by Fangfang Duan et al. are interesting and promote more research in this field (34). It is also intriguing that, in addition to the potential role of G-CSF in sAH, initial animal studies have provided evidence for a potential role of G-CSF as a novel therapeutic target for addictive disorders (25,26); while speculative at this stage, this notion is intriguing given that patients with sAH have a dual pathology represented by the alcohol-associated liver disease and the underlying alcohol addiction. However, as summarized above, the literature on the potential role of G-CSF in patients with sAH remains inconclusive at this stage. Given the overall potential risks and benefits based on the knowledge gained so far, further evidence is necessary to support the wide use of G-CSF plus PTX in sAH. In particular, larger trials are required to determine the specific G-CSF agent, the dosage, and the duration in combination with PTX. Finally, it is critical to mention that supportive therapies, including nutritional support, the prevention of AKI and infections, and the promotion of alcohol abstinence (27), along with the need to develop integrated models of addiction and hepatology care (28,29) are the most important therapies and goals in the field, but also still the most neglected in clinical practice (30). At the same time, public health policies and strategies to decrease alcohol use and early detect individuals with alcohol-related health consequences could also contribute to prevent mortality due to sAH (31–33).