Hepatotoxicity with Vismodegib: An MD Anderson Cancer Center and Research on Adverse Drug Events and Reports Project

We ascertained that hepatic events associated with vismodegib in FAERS produced a safety signal. This association was previously described in the study by Ventarola and Silverstein [16]. We identified an increasing incidence over the extended study period. The first case was reported prior to FDA approval in 2011, and there has been an upward trend in cases reported to FAERs over the last 4 years (Fig. 1). It has been reported that a median of 7 years post-marketing elapses before newly detected SAEs are announced by pharmaceutical suppliers and/or the FDA [17], but this report of vismodegib-associated hepatotoxicity is being disseminated within 4 years of approval. The importance of systematic interrogation and analyses with large databases such as FAERS is made evident in that, in two NCI-designated comprehensive cancer centers, initial interrogation showed no detectable reports of hepatotoxicity. Such a finding is not unexpected, given the small number of treated patients (n = 15) in two of the NCI-designated 41 comprehensive cancer centers. Since hepatotoxicity may progress to be severe and/or life threatening, further exploration of this important association is essential. According to the NCI, only about 15% of US cancer patients are diagnosed and treated at the nation’s major academic-based cancer centers, and the large remainder may not be easily tracked for outcomes [18]. Moreover, since FAERS case reports are de-identified, it is not practical to ascertain, and/or contact the initial reporter for, further clinical information and outcomes. Moreover, FAERS reports of vismodegib-associated hepatotoxicity typically do not provide detail on related comorbidities such as prior liver abnormalities or alcohol abuse.

Some possible mechanisms of drug-related hepatotoxicity include disruption of the cell membrane and drug binding to cell protein, leading to cell death; cell apoptosis; inhibition of cell mitochondria function; cellular pathway inhibition; and abnormal bile flow leading to cholestasis and jaundice. Moreover, obesity and malnutrition, pregnancy, history of drug reactions, and pre-existing liver disease are factors that lead to susceptibility to hepatotoxicity [19]. Liver abnormalities may also include etiological factors such as nonalcoholic steatohepatitis (NASH), medication use, alcohol use, and chronic liver disease, among others [20].

In published RCTs, vismodegib resulted in elevated alanine aminotransferase levels compared with the placebo arm [21]. SAEs (CTCAE grade 3–4) were seen in two participants in an ovarian cancer RCT [9]. Hedgehog signaling has generally been considered inactive in healthy adult hepatocytes. Paracrine connections with hepatic epithelial cells are important for embryonic development, cell proliferation, and recovery from chronic liver damage [22‐24]. Vismodegib functions as a small-molecule inhibitor of the hedgehog pathway by binding to smoothened (SMO) receptors in cells. Therefore, it may be reasonable to hypothesize that the damage caused by NASH creates an idiosyncratic trait within patients, placing them at a higher risk for experiencing a liver-based AE if they are treated with vismodegib as the drug attempts to regulate a highly damaged system.

Concomitant medication use, summarized in FAERS for 14 of the serious cases, may prove to be quite important, as four of the reports in FAERS indicated concomitant use of acetaminophen, a common over-the-counter medication (Table 1). Acetaminophen may lead to acute hepatotoxicity or necrosis, especially with overdose; acetaminophen poisoning accounts for nearly 50% of all cases of acute liver failure in the USA [25]. Ash and Jolly [26] reported a case of acute liver injury related to a possible drug–drug interaction between vismodegib, aspirin, and naproxen. Thus, concomitant drug use with vismodegib may contribute to drug–drug interactions that may play a major role in the development of liver toxicity.

The press announcement at the time of the original FDA approval on 30 January 2012 and the DailyMed website, on which full prescribing information documents are posted by the US National Library of Medicine, both specified that safety and effectiveness were not established for patients with hepatic impairment who receive vismodegib [27‐29]. LiverTox currently lists vismodegib as a potentially hepatotoxic medication [30]. Health Canada has a statement similar to that of the FDA in relation to liver impairment but with an arrangement whereby the sponsor agrees to submit additional safety studies, specifically to include results from a renal/hepatic impairment study by March 2015 [31]. The European Medicines Agency (EMA) reported that data on moderate and severe liver impairment were too limited to enable conclusions about liver impairment, but nevertheless posted a table including abnormal liver enzymes as a common side effect in 1–10% of patients [32]. The Australian Government’s Therapeutic Goods Administration released a statement citing study SHH4489g, which attributed three cases of hepatotoxicity to vismodegib, cases resulted in hospitalizations, but no deaths were reported [33]. The agency recommended vismodegib be avoided in patients with severe hepatic impairment [33]. As such, although the FDA labeling (full prescribing information) mentions that medications that are liver metabolized should be avoided, it provides no information on drug interaction risks for vismodegib [27, 28]. Especially in light of healthcare system reform, busy health practitioners must access and utilize ever-increasing amounts of information in the prescribing of medications and patient management, with a continuous need for new and updated post-marketing risk information and newly identified clinically significant drug interactions [34].

In contrast with other drug regulatory agencies, the FDA has not yet advanced newly evolving risk information regarding vismodegib and hepatotoxicity, although it has issued recommendations for studies of vismodegib use in individuals with renal and liver dysfunction [29]. Patients with mild liver impairment should be carefully monitored if vismodegib is considered for therapy. We also recommend that potentially hepatotoxic agents such as acetaminophen and alcohol be avoided during vismodegib therapy.

Limits of this study may include under-reporting and potential reporting bias because of the voluntary nature of the FAERS database and the impracticability of contacting AE reporters. As such, the ability to establish causality of hepatotoxicity associated with vismodegib based on FAERS data is limited, albeit significant signal detection values were found.