Liver Safety Assessment in Clinical Trials of New Agents for Chronic Hepatitis B

Robert J. Fontana; Mark I. Avigan; Harry L. A. Janssen; Arie Regev; Poonam Mishra; Anuj Gaggar; Nathaniel Brown; Cynthia Wat; Patricia Mendez; Ryan T. Anderson; Bruce Given; Veronica Miller; Maria Beumont


J Viral Hepat. 2020;27(2):96-109. 

In This Article

Serum ALT Flares in CHB Patients Receiving Currently Approved Antiviral Treatment

Four types of serum ALT elevations were noted in clinical trials of IFNs and NrtIs in chronic HBV patients: (a) spontaneous pretreatment flares; (b) early on treatment flares, typically in the first 12 weeks of treatment; (c) later on treatment flares between week 12 and end of treatment; and (d) post-treatment flares. Serum ALT elevations occurring after the screening visit but before the first dose of study drug (during the baseline visit) presumably correspond to the spontaneous flare activity that is part of the natural history of untreated CHB infection. Other situations where elevations in ALT may occur, such as treatment with immunosuppressive agents, pregnancy (mainly post-partum), co-infection with hepatitis C or D viruses, and development of de novo HCC are outside the scope of this review.

Early on Treatment Serum ALT Flares

In the first few months of NrtI or Peg-IFN therapy, some patients with rapid treatment-induced suppression of HBV replication experience transient increases in serum ALT which usually resolve despite continued treatment. These likely reflect a treatment-induced enhancement of immune-mediated cytolysis of HBV-infected liver cells (Table 2). Although close patient monitoring to evaluate alternative causes of the ALT elevations occasionally uncovered an acute intercurrent viral infection or toxin exposure, most of these flares were associated with marked early efficacy responses (Figure 1A). Importantly, these patients typically exhibited stable serum albumin, INR and bilirubin levels during the ALT flare, which helped differentiate these early flare events from severe liver inflammation with compromise of liver function or clinically important DILI. Patients with early ALT flares in the absence of other liver parameter abnormalities and who had persistently suppressed or declining levels of HBV DNA could usually continue study treatment uninterrupted and resume their normal schedule of clinic visits after documentation of stable hepatic functions and persistently declining ALT levels to less than half of their peak 'flare' values. In addition to changes in serum HBV DNA levels, a number of quantitative exploratory viral biomarkers [ie HBV RNA, quantitative HBsAg, HBeAg and HB core-related antigen (HBcrAg)] show promise to help differentiate patients who are experiencing immune reconstitution with cytolysis of HBV-infected hepatocytes from DILI, but further development is needed.[17]

Figure 1.

Types of serum ALT flares in CHB patients receiving approved antiviral treatments. A, A patient with chronic HBV developed a moderate serum ALT flare at month 6 of entecavir therapy that was self-limited and resolved despite continued dosing. B, This patient developed drug resistance to lamivudine with an increase in HBV DNA that preceded the mild ALT flare. The introduction of tenofovir led to a reduction in serum HBV DNA and normalization of serum ALT levels. C, Following completion of a 48-week course of peginterferon, this patient experienced a rise in serum HBV DNA levels that was associated with a severe ALT flare but no loss of HBsAg

Late on Treatment Serum ALT Flares

Serum ALT elevations occurring after 12 weeks of treatment in Peg-IFN and first-generation NrtI studies generally occurred in the setting of imminent HBeAg or HBsAg seroclearance or with the emergence of drug-resistant variants for Nrtl. The majority of Peg-IFN treated patients (>85%) have elevated ALT levels throughout the 48-week treatment period (with normal levels of bilirubin, INR and albumin), although the incidence of higher grades (>5× ULN) of ALT elevation is lower after 12 weeks compared with the initial 12 weeks (Table 2). The increase in viral replication after months of treatment often reflected the development of viral resistance (especially for older NrtI agents with lower barriers to resistance) or patient noncompliance with increasing levels of serum HBV DNA (Figure 1B).

Post-treatment Serum ALT Flares

Potentially severe post-treatment serum ALT flares were noted in studies of Peg-IFN and in the early NrtI studies (Figure 1C).[18,19] Post-treatment flares are attributed to memory T-cell responses to recrudescent HBV replication with rising HBV DNA levels in a subset of the patients who had not achieved HBeAg/HBsAg-seroconversion during treatment.[20] Such events usually present soon after end of treatment but can occur later (ie 24–48 weeks) after treatment discontinuation and may be clinically severe in some instances with occasional progression to liver failure, especially in patients with cirrhosis.[21] Although the frequency and severity of post-therapy flares may vary with HBV genotype as well as the oral agent that is discontinued, these episodes are essentially indistinguishable from the spontaneous flares seen in untreated CHB patients.[22,23] If clinically significant hepatitis develops with confirmed rising HBV DNA levels after treatment withdrawal, it is important to resume HBV suppressive therapy expeditiously. To monitor for such events in clinical trials, a post-treatment monitoring phase of at least 48 weeks is now recommended by the FDA.[24] Early experience with post-treatment flares suggested that severe flares could be averted when antiviral treatment was promptly reinstituted with either the study drug or approved NrtI.[13]

More recent studies have explored the utility of discontinuing NrtIs in noncirrhotic HBeAg-negative patients who have been stably suppressed for prolonged periods of time.[25,26] In a pilot, randomized study of 42 European chronic HBV patients with genotype D infection that had been suppressed on tenofovir for >4 years, 19% (4 of 22) achieved HBsAg loss after tenofovir discontinuation, with a further 43% able to remain off therapy with suppressed HBV DNA and normal ALT at 144 weeks after therapy cessation.[23] However, studies of Asian patients with primarily genotype B and C chronic HBV infection have demonstrated a much lower rate of HBsAg seroclearance and a higher rate of antiviral therapy resumption after NrtI discontinuation.[27–29] Furthermore, whether pretreatment or end of treatment quantitative HBsAg or HBV RNA levels may predict the likelihood of post-therapy seroclearance remains unclear. Nonetheless, most experts advise against NRTI discontinuation in any CHB patient with cirrhosis due to safety concerns until future prospective studies reliably identify patients that are likely to benefit from treatment discontinuation.[30]