24 - 352 Chronic Hepatitis

352 Chronic Hepatitis

Esperance A. K. Schaefer,

Raymond T. Chung, Jules L. Dienstag

Chronic Hepatitis Chronic hepatitis represents a series of liver disorders of varying causes and severity in which hepatic inflammation and necrosis continue for at least 6 months. Milder forms are nonprogressive or only slowly progressive, while more severe forms may be associated with scarring and architectural reorganization, which, when advanced, lead ulti­ mately to cirrhosis. Several broad categories of chronic hepatitis have been recognized. These include chronic viral hepatitis, drug-or toxininduced chronic hepatitis (Chap. 351), alcohol-associated liver disease (Chap. 353), metabolic dysfunction-related chronic hepatitis (Chap 354), and autoimmune chronic hepatitis. In many cases, clinical and laboratory features are insufficient to allow assignment into one of these distinct categories; these “idiopathic” cases are also believed to represent autoimmune chronic hepatitis. Moreover, because of the high prevalence of steatotic liver diseases (metabolic dysfunction-associated steatotic liver disease and alcohol-associated liver disease), patients with more than one, overlapping type of hepatitis are encountered not infrequently (e.g., viral and steatotic liver injury, alcohol-related and nonalcohol-related steatotic liver injury). Finally, clinical and labora­ tory features of chronic hepatitis are observed occasionally in patients with other hereditary/metabolic disorders, such as Wilson disease (copper overload) and α1 antitrypsin deficiency (Chaps. 355 and 427). Although all types of chronic hepatitis share certain clinical, laboratory, and histopathologic features, chronic viral and chronic autoimmune hepatitis are sufficiently distinct to merit separate discussions. For dis­ cussion of acute hepatitis, see Chap. 350. CLASSIFICATION OF CHRONIC HEPATITIS Common to all forms of chronic hepatitis are histopathologic distinc­ tions based on localization and extent of liver injury. These vary from the milder forms, previously labeled chronic persistent hepatitis and chronic lobular hepatitis, to the more severe form, formerly called chronic active hepatitis. When first defined, these designations were believed to have prognostic implications, which were not corroborated by subsequent observations. Categorization of chronic hepatitis based primarily on histopathologic features has been replaced by a more informative classification based on a combination of clinical, serologic, and histologic variables. Classification of chronic hepatitis is based on (1) its cause; (2) its histologic activity, or grade; and (3) its degree of progression based on level of fibrosis, or stage. Thus, neither clinical features alone nor histologic features—requiring liver biopsy or non­ invasive markers of fibrosis—alone are sufficient to characterize and distinguish among the several categories of chronic hepatitis. ■ ■CLASSIFICATION BY CAUSE Clinical and serologic features allow the establishment of a diagnosis of chronic viral hepatitis, caused by hepatitis B, hepatitis B plus D, or hepatitis C; autoimmune hepatitis, including several subcategories, I and II, based on serologic distinctions; drug-associated chronic hepatitis; and a category of unknown cause, or cryptogenic chronic hepatitis (Table 352-1). These are addressed in more detail below. ■ ■CLASSIFICATION BY GRADE Grade, a histologic assessment of necroinflammatory activity, is based on examination of the liver biopsy. An assessment of important histo­ logic features includes the degree of periportal necrosis and the disrup­ tion of the limiting plate of periportal hepatocytes by inflammatory cells (so-called piecemeal necrosis or interface hepatitis); the degree of confluent necrosis that links or forms bridges between vascular structures—between portal tract and portal tract or even more impor­ tant bridges between portal tract and central vein—referred to as bridging necrosis; the degree of hepatocyte degeneration and focal necrosis

TABLE 352-1  Clinical and Laboratory Features of Chronic Hepatitis TYPE OF HEPATITIS DIAGNOSTIC TEST(S) THERAPY Chronic hepatitis B HBsAg, IgG anti-HBc, HBeAg, HBV DNA IFN-α, PEG IFN-α Oral agents: First-line: entecavir, tenofovir  Chronic hepatitis C Anti-HCV, HCV RNA Sofosbuvir-ledipasvir, glecaprevirpibrentasvir, sofosbuvirvelpatasvir, elbasvir-grazoprevir, sofosbuvir-velpatasvir-voxilaprevira Chronic hepatitis D Anti-HDV, HDV RNA, HBsAg, IgG anti-HBc IFN-α, PEG IFN-αc, bulevirtideb Autoimmune hepatitis ANA (homogeneous), anti-SMA, anti-SLA, anti-LKM1 (±) Hyperglobulinemia First line: prednisone, azathioprine Second line: mycophenolate mofetil, tacrolimus, rituximab See www.hcvguidelines.org. aDetailed treatment recommendations can be found at www.hcvguidelines.org. bCurrently approved by European Medicines Agency, under review at the U.S. Food and Drug Administration. Abbreviations: HBc, hepatitis B core; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; HCV, hepatitis C virus; HDV, hepatitis D virus; IFN-α, interferon α; IgG, immunoglobulin G; LKM, liver-kidney microsome; PEG IFN-α, pegylated interferon α; SLA, soluble liver antigen; SMA, smooth muscle antibody. within the lobule; and the degree of portal inflammation. Several scor­ ing systems for viral hepatitis that take these histologic features into account have been devised, and the most popular are the histologic activity index (HAI), used commonly in the United States, and the METAVIR score, used in Europe (Table 352-2). Based on the presence and degree of these features of histologic activity, chronic hepatitis can be graded as mild, moderate, or severe. CHAPTER 352 ■ ■CLASSIFICATION BY STAGE The stage of chronic hepatitis, which reflects the level of progression of the disease, is based on the degree of hepatic fibrosis. When fibro­ sis is so extensive that fibrous septa surround parenchymal nodules and alter the normal architecture of the liver lobule, the histologic lesion is defined as cirrhosis. Staging is based on the degree of fibrosis as categorized on a numerical scale 0−6 (HAI) or 0−4 (METAVIR) (Table 352-2). Several noninvasive approaches have been introduced to provide approximations of hepatic histologic stage, including serum biomarkers of fibrosis; fibrosis scores such as FIB-4, a validated algo­ rithm based on such routine lab tests as aspartate and alanine amino­ transferase (AST and ALT) levels and platelet counts (PLT) (age [years] × AST divided by PLT ×); the Enhanced Liver Fibrosis (ELF) score, a proprietary score (applied more frequently in steatotic liver disease than viral liver disease) based on a combination of three nonroutine blood tests (hyaluronic acid, procollagen III N-terminal peptide, and tissue inhibitor of metalloproteinase 1); and imaging determinations of liver elasticity/stiffness. Chronic Hepatitis CHRONIC VIRAL HEPATITIS Both of the enterically transmitted forms of viral hepatitis, hepatitis A and E, are self-limited and do not cause chronic hepatitis (rare reports notwithstanding in which acute hepatitis A serves as a trigger for the onset of autoimmune hepatitis in genetically susceptible patients or in which hepatitis E [Chap. 350] can cause chronic liver disease in immunosuppressed hosts, for example, after liver transplantation). In contrast, the entire clinicopathologic spectrum of chronic hepatitis occurs in patients with chronic viral hepatitis B and C as well as in patients with chronic hepatitis D superimposed on chronic hepatitis B. ■ ■CHRONIC HEPATITIS B The likelihood of chronicity after acute hepatitis B varies as a function of age. Infection at birth is associated with clinically silent acute infec­ tion but a 90% chance of chronic infection, whereas infection in young adulthood in immunocompetent persons is typically associated with clinically apparent acute hepatitis but a risk of chronicity of only ~1%. Most cases of chronic hepatitis B among adults, however, are recognized

TABLE 352-2  Histologic Grading and Staging of Chronic Hepatitis   HISTOLOGIC ACTIVITY INDEX (HAI)a METAVIRb HISTOLOGIC FEATURE SEVERITY SCORE SEVERITY SCORE Necroinflammatory Activity (grade) Periportal necrosis, including piecemeal necrosis and/or bridging necrosis (BN) None Mild Mild/moderate Moderate Severe

None Mild Moderate Severe Bridging necrosis Intralobular necrosis

Confluent —None —Focal —Zone 3 some —Zone 3 most —Zone 3 + BN few —Zone 3 + BN multiple —Panacinar/multiacinar

None or mild Moderate Severe Focal —None —≤1 focus/10× field —2–4 foci/10× field —5–10 foci/10× field —>10 foci/10× field

Portal inflammation None Mild Moderate Moderate/marked Marked Total

0–18           A0–A3c PART 10 Disorders of the Gastrointestinal System Fibrosis (stage) None Portal fibrosis—some Portal fibrosis—most Bridging fibrosis—few Bridging fibrosis—many Incomplete cirrhosis Cirrhosis               Total

F0 F1 F1 F2 F3 F4 F4

aIshak K, Baptista A, Bianchi L, et al: Histologic grading and staging of chronic hepatitis. J Hepatol 22:696, 1995. bBedossa P, Poynard T, French METAVIR Cooperative Study Group: An algorithm for grading activity in chronic hepatitis C. Hepatology 24:289, 1996. cNecroinflammatory grade: A0 = none; A1 = mild; A2 = moderate; A3 = severe. in patients who never had a recognized episode of clinically apparent acute viral hepatitis. The degree of liver injury (grade) in patients with chronic hepatitis B is variable, ranging from none in inactive carriers to mild to moderate to severe. Among adults with chronic hepatitis B, histologic features are of prognostic importance. In one long-term study of 379 patients with untreated chronic hepatitis B in the United States, investigators found a 5-year survival rate of 97% for patients with mild chronic hepatitis, 86% for patients with moderate to severe chronic hepa­ titis, and only 55% for patients with chronic hepatitis and cirrhosis. The 15-year survival in these cohorts was 77%, 66%, and 40%, respectively. On the other hand, more recent observations do not allow us to be so sanguine about the prognosis in patients with mild chronic hepatitis; among such patients followed for 1−13 years, progression to more severe chronic hepatitis and cirrhosis has been observed in more than a quarter of cases. Co-infection with HIV has long been appreciated to accelerate the progression of hepatitis B; early studies documented a nearly 20-fold increase in liver-related mortality between persons with hepatitis B virus (HBV)-HIV co-infection compared to those with HBV infection alone. Similarly, coexisting steatohepatitis has also been found to accelerate disease progression; in one study, patients with HBV infection and steatohepatitis were found to have both higher-grade inflammation and more advanced-stage fibrosis.

More important to consider than histology alone in patients with chronic hepatitis B is the degree of HBV replication. As reviewed in Chap. 350, chronic HBV infection can occur in the presence or absence of serum hepatitis B e antigen (HBeAg), and generally, for both HBeAg-reactive and HBeAg-negative chronic hepati­ tis B, the level of HBV DNA correlates with the level of liver injury and risk of progression. In otherwise healthy adults with HBeAg-positive chronic hepatitis B, two phases have been recognized based on the relative level of HBV replication. The relatively replicative phase is characterized by the presence in the serum of HBeAg and HBV DNA levels well in excess of 103−104 IU/mL, sometimes exceeding 109 IU/mL; by the presence in the liver of detectable intrahepatocyte nucleocapsid antigens (primarily hepatitis B core antigen [HBcAg]); by high infectivity; and by accompanying liver injury (see below for an important exception during the early decades of life). In contrast, the relatively nonreplicative phase is characterized by the absence of the conventional serum marker of HBV replication (HBeAg), the appear­ ance of anti-HBe, levels of HBV DNA below a threshold of ~103 IU/mL, the absence of intrahepatocytic HBcAg, limited infectivity, and minimal liver injury. Patients in the relatively replicative phase tend to have more severe chronic hepatitis and more rapid progression, whereas those in the relatively nonreplicative phase tend to have minimal or mild chronic hepatitis or to be inactive hep­ atitis B carriers. Thus, the transition between HBeAgpositive chronic hepatitis B to HBeAg-negative disease is of clinical importance, signaling a slowing of disease progression. In the past, the likelihood of spontaneous HBeAg loss was estimated to be ~10% per year; how­ ever, a 2020 systematic review of 26 studies involving 7550 persons with HBeAg-positive chronic hepatitis B revealed a spontaneous HBeAg clearance rate of 6.46 per 100 person-years. Such spontaneous HBeAg loss was found to occur more frequently with older age and in European populations but less frequently in patients with genotype C.

Yes No

Distinctions in HBV replication and in histologic cat­ egory, however, do not always coincide. In patients with HBeAg-positive chronic HBV infection, especially when acquired at birth or in early childhood, as recognized commonly in Asian countries, a dichotomy is common between very high levels of HBV replication during the early decades of life (when the level of apparent host immunologic tolerance of HBV is relatively high) and negligible levels of liver injury; during this phase of chronic hepatitis B (labeled by some as “HBeAg-positive chronic infection” or “immune tolerant,” see Chap. 350, “Complications and Sequelae”), the level of viral replica­ tion does not correlate with liver injury or late complications. Yet despite the relatively immediate, apparently benign nature of liver disease for many decades in this population, in the middle decades, activation of liver injury emerges as what appears to be the relative tolerance of the host to HBV declines, and these patients with childhood-acquired HBV infection are ultimately at increased risk later in life for cirrhosis, hepa­ tocellular carcinoma (HCC) (Chap. 87), and liver-related death; the link between high-level HBV replication and these late liver complications has been demonstrated convincingly in, and confined mostly to, persons in their middle decades, especially age ≥40. A discussion of the patho­ genesis of liver injury in patients with chronic hepatitis B appears in Chap. 350. HBeAg-negative chronic hepatitis B (i.e., chronic HBV infection with active virus replication and readily detectable HBV DNA but without HBeAg [anti-HBe-reactive]) is more common than HBeAg-positive chronic hepatitis B in Mediterranean and European countries and in Asia. Compared to patients with HBeAg-reactive chronic hepa­ titis B, patients with HBeAg-negative chronic hepatitis B have HBV

DNA levels several orders of magnitude lower (usually no more than 105−106 IU/mL) than those observed in the HBeAg-reactive subset. Most such HBeAg-negative cases represent precore or core-promoter mutations acquired late in the natural history of mostly early-lifeonset disease; these mutations prevent translation of HBeAg from the precore component of the HBV genome (precore mutants) or are characterized by downregulated transcription of precore mRNA (corepromoter mutants; Chap. 350). Although their levels of HBV DNA tend to be lower than among patients with HBeAg-reactive chronic hepatitis B, patients with HBeAg-negative chronic hepatitis B can have progressive liver injury (complicated by cirrhosis and HCC) and experience episodic reactivation of liver disease reflected in fluctuat­ ing levels of aminotransferase activity (“flares”). The biochemical and histologic activity of HBeAg-negative disease tends to correlate closely with levels of HBV replication. Worth reiterating, the level of HBV replication is the most important risk factor for the ultimate devel­ opment of cirrhosis and HCC in both HBeAg-reactive and HBeAgnegative patients. Although levels of HBV DNA are lower and more readily suppressed by therapy to undetectable levels in HBeAg-negative (compared to HBeAg-reactive) chronic hepatitis B, achieving sustained responses that permit discontinuation of antiviral therapy is less likely in HBeAg-negative patients (see below). Inactive carriers are patients with circulating hepatitis B surface antigen (HBsAg), normal serum aminotransferase levels, minimal or no histologic evidence of liver injury, undetectable HBeAg, and levels of HBV DNA that are either undetectable or present at a threshold of ≤103 IU/mL. This serologic profile occurs not only in inactive carriers but also in patients with HBeAg-negative chronic hepatitis B during periods of relative inactiv­ ity; distinguishing between the two requires sequential biochemical and virologic monitoring over many months. The spectrum of clinical features of chronic hepatitis B is broad, ranging from asymptomatic infection to debilitating disease or even end-stage, fatal hepatic failure. As noted above, the onset of the dis­ ease tends to be insidious in most patients, apart from the very few in whom chronic disease follows failure of resolution of clinically appar­ ent acute hepatitis B. The clinical and laboratory features associated with progression from acute to chronic hepatitis B are discussed in Chap. 350. Fatigue is a common symptom, and persistent or intermittent jaundice is a common feature in severe or advanced cases. Intermittent deepening of jaundice and recurrence of malaise and anorexia, as well as worsening fatigue, are reminiscent of acute hepatitis; such exacer­ bations may occur spontaneously, often coinciding with evidence of virologic reactivation; may lead to progressive liver injury; and, when superimposed on well-established cirrhosis, may cause hepatic decom­ pensation. Complications of cirrhosis occur in end-stage chronic hepatitis and include ascites, edema, bleeding gastroesophageal varices, hepatic encephalopathy, coagulopathy, and hypersplenism. Occasion­ ally, these complications bring the patient to initial clinical attention. Extrahepatic complications of chronic hepatitis B, similar to those seen during the prodromal phase of acute hepatitis B, are associated with tissue deposition of circulating hepatitis B antigen–antibody immune complexes. These include arthralgias and arthritis, which are common, and the rarer purpuric cutaneous lesions (leukocytoclastic vasculi­ tis), immune-complex glomerulonephritis, and generalized vasculitis (polyarteritis nodosa) (Chap. 375). Laboratory features of chronic hepatitis B do not always distinguish adequately between histologically mild and severe hepatitis. Amino­ transferase elevations tend to be modest for chronic hepatitis B but may fluctuate in the range of 100 to 1000 units. As is true for acute viral hepatitis B, ALT tends to be more elevated than AST; however, once cirrhosis is established, AST tends to exceed ALT. Levels of alkaline phosphatase activity tend to be normal or only marginally elevated. In severe cases, moderate elevations in serum bilirubin (51.3−171 μmol/L [3−10 mg/dL]) occur. Hypoalbuminemia and prolongation of the prothrombin time occur in severe or end-stage cases. Hyperglobulin­ emia and detectable circulating autoantibodies are distinctly absent in chronic hepatitis B (in contrast to autoimmune hepatitis). Viral markers of chronic HBV infection are discussed in Chap. 350.

TREATMENT Chronic Hepatitis B Although progression to cirrhosis is more likely in severe than in mild or moderate chronic hepatitis B, all forms of chronic hepatitis B can be progressive, and progression occurs primarily in patients with active HBV replication. Moreover, in populations of patients with chronic hepatitis B who are at risk for HCC (Chap. 87), the risk is highest for those with continued, high-level HBV replica­ tion and lower for persons in whom initially high-level HBV DNA falls spontaneously over time. Therefore, management of chronic hepatitis B is directed at suppressing the level of virus replication. Although clinical trials tend to focus on clinical endpoints achieved over 1−2 years (e.g., suppression of HBV DNA to undetectable levels, loss of HBeAg/HBsAg, improvement in histology, normal­ ization of ALT), these short-term gains translate into reductions in the risk of clinical progression, hepatic decompensation, HCC, liver transplantation, and death; regression of cirrhosis and of esophageal varices has been documented to follow long-term phar­ macologic suppression of HBV replication. In addition, restoration of impaired HBV-specific T-cell function has been shown following successful suppression of HBV replication with antiviral therapy. To date, eight drugs have been approved for treatment of chronic hepatitis B: injectable interferon (IFN) α and pegylated interferon (long-acting IFN bound to polyethylene glycol, PEG [PEG IFN]) and the oral agents lamivudine, adefovir dipivoxil, entecavir, telbi­ vudine, tenofovir disoproxil fumarate (TDF), and tenofovir alafen­ amide (TAF). CHAPTER 352 Antiviral therapy for hepatitis B has evolved rapidly since the mid-1990s, as has the sensitivity of tests for HBV DNA. When IFN and the first oral antiviral lamivudine were evaluated in clinical trials, HBV DNA was measured by insensitive hybridization assays with detection thresholds of 105−106 virions/mL; when subsequent treatments were studied in clinical trials, HBV DNA was measured by sensitive amplification assays (polymerase chain reaction [PCR]) with detection thresholds of 101−103 viral copies/mL or IU/mL. Recognition of these distinctions is helpful when comparing results of clinical trials that established the efficacy of these therapies (reviewed below). Of the eight approved treatments, PEG IFN, entecavir, and the two tenofovir preparations (TDF and TAF) are recommended as first-line agents, and generally, the oral agents are favored over injectable PEG IFN. Chronic Hepatitis HISTORICAL, EARLY-GENERATION ANTIVIRAL AGENTS Interferon  IFN-α was the first approved therapy (1992) for chronic hepatitis B. Although it is no longer used to treat hepatitis B, standard IFN is important historically, having provided important lessons about antiviral therapy in general. For immunocompetent adults with HBeAg-reactive chronic hepatitis B (who tend to have high-level HBV DNA [>105−106 virions/mL] and histologic evi­ dence of chronic hepatitis on liver biopsy), a 16-week course of subcutaneous IFN, 5 million units daily or 10 million units thrice weekly, resulted in a loss of HBeAg and hybridization-detectable HBV DNA (i.e., a reduction to levels below 105−106 virions/mL) in ~30% of patients, with a concomitant improvement in liver histol­ ogy. Seroconversion from HBeAg to anti-HBe occurred in ~20%, and, in early trials, ~8% lost HBsAg. Successful IFN therapy and seroconversion were often accompanied by an acute hepatitis-like elevation in aminotransferase activity, postulated to result from enhanced cytolytic T-cell clearance of HBV-infected hepatocytes. Relapse after successful therapy was rare (1 or 2%). Responsive­ ness to IFN was higher in patients with low-level HBV DNA and substantial ALT elevations. Therapy with IFN was not effective in immunosuppressed persons, persons with neonatal acquisition of infection and minimal-to-mild ALT elevations, or patients with decompensated chronic hepatitis B (in whom such therapy was actually detrimental, sometimes precipitating decompensation, often associated with severe adverse effects). After HBeAg loss

during IFN therapy, 80% experienced eventual loss of HBsAg and ALT normalization over the ensuing decade. In addition, improved long-term and complication-free survival as well as a reduction in the frequency of HCC were documented among IFN respond­ ers, supporting the conclusion that successful antiviral therapy improves the natural history of chronic hepatitis B.

Brief-duration IFN therapy in patients with HBeAg-negative chronic hepatitis B was disappointing, suppressing HBV replication transiently during therapy but resulting only rarely in sustained antiviral responses. Complications of IFN therapy include systemic “flu-like” symp­ toms; marrow suppression; emotional lability (irritability, depres­ sion, anxiety); autoimmune reactions (especially autoimmune thyroiditis); and miscellaneous side effects such as alopecia, rashes, diarrhea, and numbness and tingling of the extremities. With the TABLE 352-3  Comparison of Pegylated Interferon (PEG IFN), Lamivudine, Adefovir, Entecavir, Telbivudine, and Tenofovir Therapy for Chronic Hepatitis Ba FEATURE PEG IFNb LAMIVUDINE ADEFOVIR ENTECAVIR TELBIVUDINE TENOFOVIR (TDF) TENOFOVIR (TAF) Route of administration Subcutaneous injection (180 μg/ week) Oral (100 mg/d) Oral (10 mg/d) Oral (0.5 mg/d) Oral (600 mg/d) Oral (300 mg/d) Oral 25 mg/d) Status First-line No longer preferred No longer preferred Duration of therapyc 48–52 weeks ≥52 weeks ≥48 weeks ≥48 weeks ≥52 weeks ≥48 weeks 48 weeks Tolerability Poorly tolerated Well tolerated Well tolerated; creatinine monitoring recommended PART 10 Disorders of the Gastrointestinal System HBeAg seroconversion   1 yr Rx    >1 yr Rx   18–20%  NA   16–21%  up to 50% at 5 yrs   12%  43% at 3 yrsd Log10 HBV DNA reduction (mean copies/mL)   HBeAg-reactive   HBeAg-negative      4.5 4.1      5.5 4.4–4.7      Median 3.5–5 Median 3.5–3.9 HBV DNA PCR negative (at then current PCR sensitivitya) at end of yr 1   HBeAg-reactive   HBeAg-negative        10–25% 63%        36–44% 60–73%        13–21% 48–77% ALT normalization at end of yr 1   HBeAg-reactive   HBeAg-negative     39% 34–38%     41–75% 62–79%     48–61% 48–77% HBsAg loss, yr 1   >yr 1 3–4% 12% 5 yr after 1 yr of Rx ≤1% No data 0% 5% at yr 5 Histologic improvement (≥2 point reduction in HAI) at yr 1   HBeAg-reactive    HBeAg-negative       38% 6 months after 48% 6 months after       49–62%  61–66%       53–68%  64% Viral resistance None 15–30% at 1 yr 70% at 5 yrs None at 1 yr 29% at 5 yrs Pregnancy category C Cf C C B B B aGenerally, these comparisons are based on data on each drug tested individually versus placebo in registration clinical trials; with rare exception, these comparisons are not based on head-to-head testing of these drugs. In addition, the sensitivity of HBV DNA assays increased in sensitivity over the two decades between the introduction of the earliest and latest of these approved drugs. Therefore, relative advantages and disadvantages should be interpreted cautiously. bAlthough standard interferon α administered daily or three times a week is approved as therapy for chronic hepatitis B, it has been supplanted by PEG IFN, which is administered once a week and is more effective. Standard interferon has no advantages over PEG IFN. cDuration of therapy in clinical efficacy trials; use in clinical practice may vary. dBecause of a computer-generated randomization error that resulted in misallocation of drug versus placebo during the second year of clinical trial treatment, the frequency of HBeAg seroconversion beyond the first year is an estimate (Kaplan-Meier analysis) based on the small subset in whom adefovir was administered correctly. e7% during a year of therapy (43% at year 4) in lamivudine-resistant patients. fDespite its category C designation, lamivudine has an extensive pregnancy safety record in women with HIV/AIDS. Abbreviations: ALT, alanine aminotransferase; HAI, histologic activity index; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; NA, not applicable; PEG IFN, pegylated interferon; PCR, polymerase chain reaction; Rx, therapy; yr, year; TAF, tenofovir alafenamide; TDF, tenofovir disoproxil fumarate.

possible exception of autoimmune thyroiditis, all these side effects are reversible upon dose lowering or cessation of therapy. Although no longer competitive with the newer generation of antivirals, IFN did represent the first successful antiviral approach, set a standard against which to measure efficacy of subsequent drugs, and demonstrated the benefit of antiviral therapy on the natural history of chronic hepatitis B. Standard IFN has been sup­ planted by long-acting PEG IFN (see below). Lamivudine  The first of the nucleoside analogues to be approved (in 1998) for hepatitis B, the dideoxynucleoside lamivudine inhibits reverse transcriptase activity of both HIV and HBV and is an effec­ tive agent for chronic hepatitis B; however, it is now superseded by newer, more potent, less resistance-prone agents. For a summary of its virologic, serologic, biochemical, and histologic efficacy, as well as its resistance profile, please refer to Table 352-3. In clinical First-line No longer preferred, withdrawn First-line First-line Well tolerated Well tolerated Well tolerated; creatinine monitoring recommended Well tolerated   21%  31% at 2 yrs 44% at 6 yrs   22%  30% at 2 yrs   21%  40% at 5 yrs   10% (14% HBeAg loss) 18% at yr 2 (HBeAg loss 22%)      6.9 5.0      6.4 5.2      6.2 4.6 Not reported in clinical trials, likely same as TDF        67% (91% at 4 yrs) 90%        60% 88%        76% 93%        64% 94%     68% 78%     77% 74%     68% 76%     72% 83% 2% 6% at yr 6 <1% No data 3% 10% at yr 5 1% 1%       72%  70%       65%  67%       74%  72% Not included in clinical trials ≤1% at 1 yre Up to 5% at yr 1 Up to 22% at yr 2 0% at yr 1 0% through yr 8 0% at yr 1 0% through yr 2 1.2% at 6 yrse

trials, lamivudine therapy at daily doses of 100 mg for 48−52 weeks suppressed HBV DNA, as measured by sensitive PCR amplifica­ tion assays, by a median of ~5.5 log10 copies/mL in HBeAg-positive chronic hepatitis B and ~4.5 log10 copies/mL in HBeAg-negative chronic hepatitis B (baseline HBV DNA levels are lower in HBeAgnegative than in HBeAg-positive chronic hepatitis B) and to undetectable levels in ~40% and ~70%, respectively. Lamivudine was shown to improve histology, retard hepatic fibrosis, prevent progression to cirrhosis, and, in patients with advanced fibrosis, reduce decompensation. Lamivudine was also effective in patients unlikely to respond to IFN (e.g., those with high-level HBV DNA) or in prior IFN nonresponders. In HBeAg-positive chronic hepatitis B, lamivudine-associated HBeAg seroconversion occurred in ~20%, as was true for IFN therapy. Also similar to IFN therapy, patients with near-normal ALT activity tended not to experience HBeAg responses (despite suppression of HBV DNA), while patients with ALT levels ≥5× the upper limit of normal could expect 1-year HBeAg seroconversion rates of 50−60%. Generally, HBeAg seroconversions were confined to patients who achieved suppression of HBV DNA to <104 copies/ mL (equivalent to ~103 IU/mL). Lamivudine-associated HBeAg responses were accompanied by a delayed posttreatment HBsAg seroconversion rate comparable to that seen after IFN. Among Western patients who experienced HBeAg responses during a yearlong course of therapy and in whom the response was sustained for 4−6 months after cessation of therapy, the response was durable thereafter in the vast majority (>80%); therefore, the achievement of an HBeAg response represented a viable stopping point in therapy for HBeAg-positive patients. Reduced durability, however, was reported in Asian patients; therefore, to support the durability of HBeAg responses, a period of consolidation therapy of ≥6 months in Western patients and ≥1 year in Asian patients was recommended after HBeAg seroconversion (see treatment guidelines below; a full 12-month consolidation period is recommended currently for treatment extension after oral agent–induced HBeAg seroconver­ sion). Close posttreatment monitoring was recommended to iden­ tify HBV reactivation promptly and to resume therapy. If HBeAg was unaffected by lamivudine therapy, lamivudine was continued until an HBeAg response occurred, but long-term therapy was required to suppress HBV replication and, in turn, limit liver injury; HBeAg seroconversions increased to a level of 50% after 5 years of therapy. After a cumulative course of 3 years of lamivudine therapy, necroinflammatory activity was reduced in the majority of patients, and even cirrhosis was shown to regress to precirrhotic stages in as many as three-quarters of patients. Losses of HBsAg were few during the first year of lamivudine therapy, and this observation had been cited as an advantage of IFNbased therapy over lamivudine therapy; however, in head-to-head comparisons between standard IFN and lamivudine monotherapy, HBsAg losses were rare in both groups. Trials in which lamivudine and IFN were administered in combination failed to show a benefit of combination therapy over lamivudine monotherapy for either treatment-naïve patients or prior IFN nonresponders. Patients with HBeAg-negative chronic hepatitis B (i.e., in those with precore or core-promoter HBV mutations and who lack HBeAg) cannot achieve an HBeAg response to nucleoside ana­ logue therapy—a stopping point in HBeAg-reactive patients; almost invariably, when therapy was discontinued, reactivation was the rule. Therefore, these patients required long-term lamivudine therapy. Clinical and laboratory side effects of lamivudine were negligible and indistinguishable from those observed in placebo recipients; however, lamivudine doses were reduced in patients with reduced creatinine clearance. During lamivudine therapy, transient ALT elevations, resembling those seen during IFN therapy and dur­ ing spontaneous HBeAg-to-anti-HBe seroconversions, occurred in one-fourth of patients. These ALT elevations may result from restored cytolytic T-cell activation permitted by suppression of HBV replication. Similar ALT elevations, however, occurred at an

identical frequency in placebo recipients; however, ALT elevations associated temporally with HBeAg seroconversion in clinical trials were confined to lamivudine-treated patients. When therapy was stopped after a year, two- to threefold ALT elevations occurred in 20−30% of lamivudine-treated patients, representing renewed liver-cell injury as HBV replication returned. Although these post­ treatment flares were almost always transient and mild, rare severe exacerbations, especially in cirrhotic patients, were observed, man­ dating close and careful clinical and virologic monitoring after discontinuation of treatment. Many authorities cautioned against discontinuing therapy in patients with cirrhosis, in whom posttreat­ ment flares could precipitate decompensation.

Long-term monotherapy with lamivudine was associated with methionine-to-valine (M204V) or methionine-to-isoleucine (M204I) mutations, primarily at amino acid 204 in the tyrosinemethionine-aspartate-aspartate (YMDD) motif of the C domain of HBV DNA polymerase, analogous to mutations that occur in HIVinfected patients treated with this drug. During a year of therapy, YMDD mutations occurred in 15−30% of patients; the frequency increased with each year of therapy, reaching 70% at year 5. Ulti­ mately, patients with YMDD mutants experienced degradation of clinical, biochemical, and histologic responses; therefore, if treat­ ment was begun with lamivudine monotherapy, the emergence of lamivudine resistance, reflected clinically by a breakthrough from suppressed levels of HBV DNA and ALT, was managed by adding another antiviral to which YMDD variants are sensitive (e.g., adefo­ vir, tenofovir; see below). Currently, lamivudine has been eclipsed by more potent antivi­ rals that have superior resistance profiles (see below); it is no longer recommended as first-line therapy. Still, as the first successful oral antiviral agent for use in hepatitis B, lamivudine provided proof of principle that polymerase inhibitors can achieve virologic, sero­ logic, biochemical, and histologic benefits, including retardation and reversal of fibrosis and even of cirrhosis. CHAPTER 352 Chronic Hepatitis Cirrhosis  In addition, lamivudine was shown to be effective in the treatment of patients with decompensated hepatitis B (for whom IFN is contraindicated), in some of whom decompensa­ tion can be reversed. Moreover, among patients with cirrhosis or advanced fibrosis, lamivudine was shown to be effective in reducing the risk of progression to hepatic decompensation and, based on subsequent population studies, the risk of HCC. In the half decade following the introduction in the United States of lamivudine therapy for hepatitis B, referral of patients with HBV-associated end-stage liver disease for liver transplantation fell by ~30%, sup­ porting further the beneficial impact of oral antiviral therapy on the natural history of chronic hepatitis B. HIV Co-infection  Because lamivudine monotherapy in persons with HIV infection can result universally in the rapid emergence of YMDD variants, testing for HIV infection was recommended for all patients with chronic hepatitis B prior to lamivudine therapy; if HIV infection was identified, lamivudine monotherapy at the HBV daily dose of 100 mg was contraindicated. These patients require treatment for both HIV and HBV with an HIV drug regimen that includes or is supplemented by at least two drugs active against HBV; antiretroviral therapy (ART) often contains two drugs with antiviral activity against HBV (e.g., tenofovir and emtricitabine), but if lamivudine was part of the regimen, the 300-mg daily dose was required (Chap. 208). The safety of lamivudine during preg­ nancy has not been established; however, the drug is not terato­ genic in rodents and has been used safely in pregnant women with HIV infection and with HBV infection. As shown for subsequent nucleoside analogues, administration of lamivudine during the last months of pregnancy to mothers with high-level hepatitis B viremia reduced the likelihood of perinatal transmission of hepatitis B. Adefovir Dipivoxil  At an oral daily dose of 10 mg, the acyclic nucleotide analogue adefovir dipivoxil, the prodrug of adefovir (approved for hepatitis B in 2002), reduces HBV DNA by ~3.5−4 log10

copies/mL; i.e., it is less potent than lamivudine or any of the newer antiviral agents. For a summary of its virologic, serologic, biochem­ ical, and histologic efficacy, as well as its resistance profile, please refer to Table 352-3. Like IFN and lamivudine, adefovir dipivoxil is more likely to achieve an HBeAg response in patients with high baseline ALT; HBeAg responses to it are highly durable and can be relied upon as a treatment stopping point, after a period of consoli­ dation therapy; and biochemical, serologic, and virologic outcomes improve over time with continued therapy.

In HBeAg-negative chronic hepatitis B, as was true for lamivudine, because HBeAg responses—a potential stopping point—cannot be achieved, reactivation is the rule when adefovir therapy is dis­ continued, and indefinite, long-term therapy is required. Reported attempts to stop adefovir after 5 years were followed by a period of maintained suppression of HBV DNA and ALT; however, most such patients had persistent hepatitis B viremia, and most HBeAgnegative patients were treated indefinitely unless HBsAg loss, albeit very rare, was achieved. Adefovir was less prone to resistance than lamivudine, and thus the primary contribution of adefovir, its effectiveness in lamivudineresistant, YMDD-mutant HBV, led to its adoption for lamivudineresistant hepatitis B. When lamivudine resistance occurred, adding adefovir (i.e., maintaining lamivudine to preempt the emergence of adefovir resistance) was superior to switching to adefovir. An advantage of adefovir was its relatively favorable resistance profile; however, it was not as potent as the other approved oral agents, it did not suppress HBV DNA as rapidly or as uniformly as the others, it was the least likely of all agents to result in HBeAg seroconversion, and 20−50% of patients failed to suppress HBV DNA by 2 log10 (“primary nonresponders”). For these reasons, adefovir, which has been supplanted in both treatment-naïve and lamivudine-resistant patients by the more potent, less resistanceprone tenofovir (see below), is no longer recommended as first-line therapy. PART 10 Disorders of the Gastrointestinal System Telbivudine  Telbivudine, a cytosine analogue (approved in 2006), is similar in efficacy to entecavir (see below) but slightly less potent in suppressing HBV DNA (median 6.4 log10 reduction in HBeAgreactive disease and a 5.2 log10 reduction in HBeAg-negative disease). In its registration trial, telbivudine at an oral daily dose of 600 mg suppressed HBV DNA to <300 copies/mL in 60% of HBeAg-positive and 88% of HBeAg-negative patients, reduced ALT to normal in 77% of HBeAg-positive and 74% of HBeAg-negative patients, and improved histology in 65% of HBeAg-positive and 67% of HBeAg-negative patients. Although resistance to telbivu­ dine (M204I, not M204V, mutations) was less frequent than resis­ tance to lamivudine at the end of 1 year, resistance mutations after 2 years of treatment occurred in up to 22%. Generally well tolerated, telbivudine was associated with a low frequency of asymptom­ atic creatine kinase elevations and with a very low frequency of myopathy and peripheral neuropathy; frequency of administration had to be reduced for patients with impaired creatinine clearance. Its excellent potency notwithstanding, the inferior resistance and safety profile of telbivudine limited its appeal; telbivudine is neither recommended as first-line therapy nor used widely. CURRENT FIRST-LINE TREATMENT OPTIONS PEGYLATED IFN  After long-acting PEG IFN was shown to be effective in the treatment of hepatitis C (see below), this more con­ venient IFN preparation was evaluated in the treatment of chronic hepatitis B. Once-a-week PEG IFN is more effective than the more frequently administered, standard IFN, and several large-scale trials of PEG IFN versus oral lamivudine were conducted in patients with chronic hepatitis B. In HBeAg-reactive chronic hepatitis B, two large-scale studies were done. In one study, PEG IFN-α2b (100 μg weekly for 32 weeks, then 50 μg weekly for another 20 weeks for a total of 52 weeks) was evaluated against a comparison arm of combina­ tion PEG IFN with oral lamivudine in 307 subjects. The other

study involved PEG IFN-α2a (180 μg weekly for 48 weeks) in 814 primarily Asian patients, three-fourths of whom had ALT ≥2× the upper limit of normal, with comparison arms of lamivudine monotherapy and combination PEG IFN plus lamivudine. At the end of therapy (48−52 weeks) in the PEG IFN monotherapy arms, HBeAg loss occurred in ~30%, HBeAg seroconversion in 22−27%, undetectable HBV DNA (<400 copies/mL by PCR) in 10−25%, and normal ALT in 34−39%, and a mean reduction in HBV DNA of 2 log10 copies/mL (PEG IFN-α2b) to 4.5 log10 copies/mL (PEG IFN-α2a) was seen. Six months after completing PEG IFN mono­ therapy in these trials, HBeAg losses were present in ~35%, HBeAg seroconversion in ~30%, undetectable HBV DNA in 7−14%, and normal ALT in 32−41%, and the mean reduction in HBV DNA was 2−2.4 log10 copies/mL. Although the combination of PEG IFN and lamivudine was superior at the end of therapy in one or more serologic, virologic, or biochemical outcomes, neither the combina­ tion arm (in both studies) nor the lamivudine monotherapy arm (in the PEG IFN-α2a trial) demonstrated any benefit compared to the PEG IFN monotherapy arms 6 months after therapy. Moreover, HBsAg seroconversion occurred in 3−7% of PEG IFN recipients (with or without lamivudine); some of these seroconversions were identified by the end of therapy, but many were identified during the posttreatment follow-up period. The likelihood of HBeAg loss in PEG IFN–treated HBeAg-reactive patients was associated with HBV genotype A > B > C > D (shown for PEG IFN-α2b but not for PEG IFN-α2a). PEG IFN-α2a was approved in the United States for hepatitis B in 2005; PEG IFN-α2b, which is not approved for hepa­ titis B in the United States, is used in other countries. Based on these results, some authorities concluded that PEG IFN monotherapy should be the first-line therapy of choice in HBeAgreactive chronic hepatitis B; however, this conclusion has been challenged. Although a finite, 1-year course of PEG IFN results in a higher rate of sustained response (6 months after treatment) than is achieved with oral nucleoside/nucleotide analogue therapy, the comparison is confounded by the fact that oral agents are not discontinued at the end of 1 year. Instead, taken orally and free of side effects, therapy with oral agents is extended indefinitely or until after the occurrence of an HBeAg response. The rate of HBeAg responses after 2 years of oral-agent nucleoside analogue therapy is at least as high as, if not higher than, that achieved with PEG IFN after 1 year; favoring oral agents is the absence of injections, difficult-to-tolerate side effects, and laboratory monitoring as well as lower direct and indirect medical care costs and inconvenience. HBsAg loss with PEG IFN therapy occurs in such a small propor­ tion of patients that subjecting everyone to PEG IFN for the mar­ ginal gain of HBsAg responses during or immediately after therapy in such a very small minority is questionable. Moreover, with the newer, more potent nucleoside analogues, the frequency of HBsAg loss during the first year of therapy nearly equals that of PEG IFN. Of course, resistance is not an issue during PEG IFN therapy, but the risk of resistance is much lower with new agents (≤1% up to 3−8 years in previously treatment-naïve, entecavir-treated patients and 0% in tenofovir-treated patients; see below). Finally, the level of HBV DNA inhibition that can be achieved with the newer agents, and even with lamivudine, exceeds that achieved with PEG IFN, in some cases by several orders of magnitude. In HBeAg-negative chronic hepatitis B, the role of PEG IFN is uncertain. One trial of PEG IFN-α2a (180 μg weekly for 48 weeks vs comparison arms of lamivudine monotherapy and of combination therapy) in 564 patients showed that PEG IFN monotherapy resulted at the end of therapy in suppression of HBV DNA by a mean of 4.1 log10 copies/mL, undetectable HBV DNA (<400 copies/mL by PCR) in 63%, normal ALT in 38%, and loss of HBsAg in 4%. Although lamivudine monotherapy and combination lamivudine−PEG IFN therapy were both superior to PEG IFN at the end of therapy, no advantage of lamivudine monotherapy or combination therapy was apparent over PEG IFN monotherapy 6 months after therapy— suppression of HBV DNA by a mean of 2.3 log10 copies/mL, unde­ tectable HBV DNA in 19%, and normal ALT in 59%. In patients

involved in this trial followed for up to 5 years, among the twothirds followed who had been treated initially with PEG IFN, 17% maintained HBV DNA suppression to <400 copies/mL, but ALT remained normal in only 22%; HBsAg loss increased gradually to 12%. As was the case for standard IFN therapy in HBeAg-negative patients, only a small proportion maintained responsiveness after completion of PEG IFN therapy, raising questions about the rela­ tive value of a finite period of PEG IFN versus a longer course with a potent, low-resistance oral nucleoside analogue in these patients. Moreover, the value of PEG IFN for HBeAg-negative chronic hepatitis B has not been confirmed. In the only other controlled clinical trial of PEG IFN for HBeAg-negative chronic hepatitis B, the hepatitis C regimen of PEG IFN plus ribavirin was compared to PEG IFN monotherapy. In this trial, HBV DNA suppression (<400 copies/mL) occurred in only 7.5% of the two groups combined, and no study subject lost HBsAg. Reductions in quantitative HBsAg levels have been shown to correlate with and to be predictive of responsiveness to PEG IFN in chronic hepatitis B. If HBsAg levels fail to fall within the first 12–24 weeks or to reach <20,000 IU/mL by week 24, PEG IFN therapy is unlikely to be effective and should be discontinued. (Similar observations of HBsAg levels in oral agent–treated patients are of interest but of limited clinical relevance, given the very high likelihood of virologic responses during such therapy.) While PEG IFN remains one of the recommended first-line agents for hepatitis B, subsequent-generation, injection-free, very-well-tolerated, highbarrier-to-resistance, oral agents are used much more widely. Entecavir  Entecavir, an oral cyclopentyl guanosine analogue polymerase inhibitor (approved in 2005), is a highly potent HBV antiviral and is just as well tolerated as lamivudine. In HBeAg-positive patients, one 709-subject clinical trial compared oral entecavir, 0.5 mg daily, with lamivudine, 100 mg daily. At 48 weeks, entecavir was superior to lamivudine in suppression of HBV DNA (mean 6.9 vs 5.5 log10 copies/mL), percentage with undetectable HBV DNA (<300 copies/mL by PCR; 67 vs 36%), histologic improve­ ment (≥2-point improvement in necroinflammatory HAI score; 72 vs 62%), and normal ALT (68 vs 60%). The two treatments were indistinguishable in percentage with HBeAg loss (22 vs 20%) and seroconversion (21 vs 18%). Among patients treated with entecavir for 96 weeks, HBV DNA was undetectable cumulatively in 80% (vs 39% for lamivudine), and HBeAg seroconversions had occurred in 31% (vs 26% for lamivudine). After 3–6 years of entecavir, HBeAg seroconversions were observed in 39–44% and HBsAg loss in 5–6%. Similarly, in a 638-subject clinical trial among HBeAg-negative patients, at week 48, oral entecavir, 0.5 mg daily, was superior to lamivudine, 100 mg daily, in suppression of HBV DNA (mean 5.0 vs 4.5 log10 copies/mL) and in percentage with undetectable HBV DNA (90 vs 72%), histologic improvement (70 vs 61%), and normal ALT (78 vs 71%). No resistance mutations were encountered in pre­ viously treatment-naïve, entecavir-treated patients during 96 weeks of therapy, and in a cohort of subjects treated for up to 6 years, resis­ tance emerged in only 1.2%. Entecavir-induced HBeAg seroconver­ sions are as durable as those achieved with other antivirals. Its high barrier to resistance coupled with its high potency renders entecavir a first-line drug for patients with chronic hepatitis B. Entecavir is also effective against lamivudine-resistant HBV infection. In a trial of 286 lamivudine-resistant patients, enteca­ vir, at a higher daily dose of 1 mg, was superior to lamivudine, as measured at week 48, in achieving suppression of HBV DNA (mean 5.1 vs 0.48 log10 copies/mL), undetectable HBV DNA (72 vs 19%), normal ALT (61 vs 15%), HBeAg loss (10 vs 3%), and HBeAg seroconversion (8 vs 3%). In this population of lamivudineexperienced patients, however, entecavir resistance emerged in 7% at 48 weeks. Although entecavir resistance requires both a YMDD mutation and a second mutation at one of several other sites (e.g., T184A, S202G/I, or M250V), resistance to entecavir in lamivudineresistant chronic hepatitis B was reported to increase progressively to 43% at 4 years and 57% at 6 years; therefore, entecavir is not as

attractive a choice (and is not recommended, despite its approval for this indication) as adefovir was or as tenofovir is for patients with lamivudine-resistant hepatitis B.

In clinical trials, entecavir had an excellent safety profile. In addition, on-treatment and posttreatment ALT flares are relatively uncommon and relatively mild in entecavir-treated patients. Doses should be reduced for patients with reduced creatinine clearance. Entecavir does have low-level antiviral activity against HIV and, therefore, cannot be used as monotherapy to treat HBV infection in HIV/HBV co-infected persons. Tenofovir  Tenofovir disoproxil fumarate (TDF), an acyclic nucle­ otide analogue and potent antiretroviral agent used to treat HIV infection (approved for hepatitis B in 2008), is similar to adefovir but more potent in suppressing HBV DNA and inducing HBeAg responses; it is highly active against both wild-type and lamivudineresistant HBV and active in patients whose response to adefovir is slow and/or limited. At an oral once-daily dose of 300 mg for 48 weeks, tenofovir suppressed HBV DNA by 6.2 log10 (to undetect­ able levels [<400 copies/mL] in 76%) in HBeAg-positive patients and by 4.6 log10 (to undetectable levels in 93%) in HBeAg-negative patients; reduced ALT to normal in 68% of HBeAg-positive and 76% of HBeAg-negative patients; and improved histology in 74% of HBeAg-positive and 72% of HBeAg-negative patients. In HBeAgpositive patients, HBeAg seroconversions occurred in 21% by the end of year 1, 27% by year 2, 34% by year 3, and 40% by year 5 of tenofovir treatment; HBsAg loss occurred in 3% by the end of year 1, 6% at year 2, and 10% by year 5. After 5 years of tenofovir therapy, 87% of patients experienced histologic improvement, including reduction in fibrosis score (51%) and regression of cir­ rhosis (74%). The 5-year safety (negligible renal toxicity, in 1%, and mild reduction in bone density, in ~0.5%) and resistance profiles (none recorded through 8 years) of tenofovir are very favorable as well; therefore, tenofovir has supplanted adefovir both as first-line therapy for chronic hepatitis B and as rescue therapy for lamivu­ dine-resistant chronic hepatitis B. Studies of tenofovir and entecavir reviewed in 2015 showed no difference in long-term risks of renal and bone toxicity; however, among patients treated with tenofovir, instances of acute renal failure and of low blood phosphate levels have been reported. Thus, in patients receiving tenofovir, monitor­ ing bone density is not recommended, but periodic (at least annual) monitoring for renal injury is recommended (serum creatinine and phosphate, urine glucose and protein). Frequency of tenofovir administration should be reduced for patients with impaired cre­ atinine clearance. CHAPTER 352 Chronic Hepatitis Tenofovir alafenamide (TAF), a second-generation tenofovir (approved for hepatitis B in 2016), is a prodrug of tenofovir that requires activation to tenofovir in hepatocytes. This targeted delivery to hepatocytes allows a lower dose to suffice and reduces systemic exposure by 90%, thereby minimizing TDF-associated proximal tubular renal injury, its associated phosphate wasting, and the potential consequent loss of bone mineral density. The dose of TAF is 25 mg, which is equivalent in antiviral potency to 300 mg of TDF; both formulations have the same high barrier to resis­ tance, and clinical resistance has not been encountered. Random­ ized, controlled, double-blind, phase 3 noninferiority trials, one in HBeAg-positive patients and the other in HBeAg-negative patients, provided the safety and efficacy data to support TAF approval. In 873 HBeAg-positive patients treated for 48 weeks, TAF versus TDF achieved (1) HBV DNA reductions to <29 IU/mL in 64% versus 67%; (2) ALT normalization in 72% versus 67% (an unex­ plained TAF biochemical advantage confirmed in other trials); (3) HBeAg loss in 14% versus 12%; (4) HBeAg seroconversion in 10% versus 8%; and (5) a negligible loss of HBsAg in 1% versus 0.3%. Compared to TDF, TAF was associated with reduced impairment of renal function (median reduction in estimated glomerular filtration rate of –0.6 mL/min for TAF vs –5.4 mL/min for TDF) and of bone density (in hip measurements, mean reduction of –0.10% for TAF vs –1.72% for TDF; adjusted difference, 1.62%).

In the parallel trial among 426 HBeAg-negative patients treated for 48 weeks, reductions in HBV DNA to <29 IU/mL occurred in 94% versus 93% of individuals treated with TAF versus TDF, respec­ tively; normalization of ALT occurred in 83% versus 75%, but no HBsAg loss occurred in either group. Similar TAF advantages in maintaining renal function and bone density were reported: reduc­ tion in median estimated glomerular filtration rate (–1.8 mL/min for TAF vs –4.8 mL/min for TDF) and in median bone density (in hip measurements, mean reduction of –0.29% for TAF vs –2.16% for TDF; adjusted percentage difference, 1.87%).

At week 96, TAF and TDF HBV DNA and ALT reductions (including the TAF advantage observed at 48 weeks) were main­ tained. In the original TDF group, when TDF was switched to TAF after week 96, all differences observed during the first 96 weeks (in normalization of ALT and reductions in renal function and bone density) had resolved at week 120. Resistance did not emerge to either TAF or TDF throughout the trial. Based on these trial outcomes, TAF joined the list of recom­ mended first-line antiviral agents for chronic hepatitis B. This drug is recommended over TDF by the American Association for the Study of Liver Diseases (AASLD) and the European Association for the Study of the Liver (EASL) for patients with reduced renal function (creatinine clearance <50 mL/min), reduced bone density, and risk factors for renal injury (including, according to EASL guidelines, decompensated cirrhosis, creatinine clearance <60 mL/ min, poorly controlled hypertension or diabetes, proteinuria, active glomerulonephritis, concomitant nephrotoxic medications, or solid-organ transplantation); the EASL recommendation extends to persons >60 years, who are at increased risk of TDF nephrotoxicity. In patients with creatinine clearances <15 mL/min, neither TDF nor TAF is recommended. PART 10 Disorders of the Gastrointestinal System A comparison of antiviral therapies for chronic hepatitis B appears in Table 352-3; their relative potencies in suppressing HBV DNA are shown in Fig. 352-1. Combination Therapies  Adding a nucleos(t)ide inhibitor to PEG IFN does not appear to improve the frequency of HBeAg response or other treatment endpoints. Although the combination of lamivu­ dine and PEG IFN suppresses HBV DNA more profoundly during therapy than does monotherapy with either drug alone (and is much less likely to be associated with lamivudine resistance), this combination used for a year is no better than a year of PEG IFN

–1 –2 Log10 HBV DNA –3 –3.5 –4 –4.5 –5 –5.5 –6 –6.2 –6.4 –6.9 –7 ADV PEG IFN LAM TDF TBV ETV FIGURE 352-1  Relative potency of antiviral drugs for hepatitis B, as reflected by median log10 hepatitis B virus (HBV) DNA reduction in HBeAg-positive chronic hepatitis B. These data are from individual reports of large, randomized controlled registration trials that were the basis for approval of the drugs. In most instances, these data do not represent direct comparisons among the drugs, because study populations were different, baseline patient variables were not always uniform, and the sensitivity and dynamic range of the HBV DNA assays used in the trials varied. ADV, adefovir dipivoxil; ETV, entecavir; LAM, lamivudine; PEG IFN, pegylated interferon α2a; TBV, telbivudine; TDF, tenofovir disoproxil fumarate. Because of potency and a high barrier to resistance, ETV and tenofovir (either TDF or the second-generation tenofovir alafenamide) are recommended as first-line therapy. While PEG IFN remains a first-line agent, the oral agents developed earlier, LAM, ADV, and TBV, are no longer preferred agents.

in achieving sustained responses. To date, combinations of oral nucleoside/nucleotide agents have not achieved an enhancement in virologic, serologic, or biochemical efficacy over that achieved by the more potent of the combined drugs given individually. In a 2-year trial of combination entecavir and tenofovir versus entecavir monotherapy, for a small subgroup of patients with very high HBV DNA levels (≥108 IU/mL), a reduction in HBV DNA to <50 IU/mL was higher in the combination group (79 vs 62%); however, no dif­ ferences in HBeAg responses or any other endpoint were observed between the combination-therapy and monotherapy groups, even in the high-HBV DNA subgroup. In the setting of resistant virus, combinations of nucleos(t)ide inhibitors can be effective. For resistance to lamivudine or adefo­ vir, adding a second, non-cross-resistant agent was previously the chosen approach. Whereas, initially, in clinical studies of adefovir as rescue therapy for lamivudine resistance, adding adefovir to lamivudine (combination therapy) was considered a better strategy than replacing lamivudine with adefovir monotherapy (to minimize ALT flares and to avoid adefovir resistance), newer antiviral thera­ pies have altered the paradigm. According to current treatment recommendations of the AASLD and the EASL, switching from the resistant drug to the new drug is preferred. Because the current generation of antivirals is so potent and has such a high barrier to resistance, monotherapy with the rescue drug (e.g., tenofovir for lamivudine resistance) is as effective (as demonstrated in observa­ tional reports for up to 5 years) in maintaining viral suppression without the emergence of resistance as combination therapy with the resistant drug and the rescue drug. Generally, in patients treated initially with entecavir and tenofovir preparations, antiviral drug resistance is no longer encountered. For currently rare patients who already have acquired multidrug resistance (to both nucleo­ side analogues [lamivudine, entecavir, telbivudine] and nucleotide analogues [adefovir, tenofovir]), treatment with a combination of entecavir and tenofovir has been shown to be highly effective in suppressing HBV DNA and overcoming drug resistance. In addition to the eight approved antiviral drugs for hepatitis B, emtricitabine, a fluorinated cytosine analogue very similar to lami­ vudine in structure, efficacy, and resistance profile, offers no advan­ tage over lamivudine. A combination of emtricitabine and tenofovir is approved for the treatment of HIV infection and is an appealing combination therapy for hepatitis B, especially for lamivudineresistant disease; however, neither emtricitabine nor the combina­ tion is approved for hepatitis B. Worth considering, however, in patients with chronic hepatitis B who are at high risk of human immunodeficiency virus (HIV) exposure, the combination of TDF and emtricitabine (FTC) may have a unique but complicated role. Although clinical trial data are limited, combined TDF/FTC is used often as preexposure prophylaxis (PrEP) to prevent HIV infec­ tion, with a 94% reduction in new HIV infection. Thus, in patients with chronic hepatitis B who require treatment for their hepatitis B but who also desire PrEP, TDF/FTC may be an appealing treat­ ment option. Adherence is essential, however, because virologic and clinical reactivation of hepatitis B is common if TDF/FTC is discontinued. Novel Antivirals and Strategies  The current generation of welltolerated oral antivirals has been very successful in the management of chronic hepatitis B and in altering its natural history favorably; however, most patients require long-duration, usually indefi­ nite, therapy. Therefore, the quest continues for novel antiviral approaches that can achieve “cure” but are finite in duration. The cure of HBV infection is defined as either “functional cure,” with loss of HBsAg and undetectable DNA (that are durable for at least 6 months after treatment), with or without the emergence of antiHBs, or as “complete (sterilizing) cure,” in which covalently closed circular DNA (cccDNA) is eradicated as well. Even an approach to achieving “functional cure” with finite-duration therapy, which is elusive in patients treated with nucleoside analogues, would be wel­ come. Currently, innovative approaches being investigated focus on

viral-targeting strategies or immunomodulatory strategies. Current virologic approaches in clinical testing include HBV entry blockers, therapies to block viral RNA translation, and HBV capsid assembly modulators. Bulevirtide, an entry inhibitor blocking the receptor for HBV on cell membranes, sodium taurocholate co-transporting polypeptide (NTCP), that decreases new cccDNA production, has been approved in the European Union for treatment of HBV/HDV co-infection (see HDV, below) but has not met with success in treat­ ing HBV mono-infection. Several therapies targeting viral RNA are under investigation. These include RNA interference (RNAi) therapies and antisense oligonucleotides (ASO). Experimental RNAi therapies tend to reduce HBsAg by ~2 log10 IU/mL during therapy, but ultimately, HBsAg reappears over time after therapy. An early signal of efficacy was reported for the ASO bepirovirsen, which achieved functional cure in ~10% of patients treated with six monthly injections (as monotherapy or in combination with nucleoside analogues); however, the durability of this response was degraded over ensuing follow-up, yielding a very small residual rate of functional cure. Drugs to inhibit capsid assembly have the potential to modulate cccDNA synthesis and contribute to func­ tional cure; however, while this class of drugs reduces HBV DNA by ~2–3 log10 IU/mL, HBsAg levels remain unaffected. Moreover, at least one experimental drug in each of these categories of virologic inhibitors had to be abandoned for toxicity. Immunomodulators being studied have included toll-like receptor agonists (TLR7, TLR8), immune checkpoint inhibitors (programmed cell death 1 [PD-1] and programmed cell death ligand 1 [PD-L1] blockade), T-cell vaccines, and broadly neutralizing antibody approaches. While some immunomodulatory strategies have shown promise, such as one small phase 2 clinical trial in which treatment with the TLR8 agonist selgantolimod resulted in HBsAg loss in a small num­ ber, most immunomodulatory approaches have been abandoned for lack of efficacy or for toxicity (e.g., upregulation of RNA sen­ sor retinoic acid-inducible gene I [RIG-I]). Even after more than a decade of early clinical trials, none of these treatments has yet been TABLE 352-4  Recommendations for Treatment of Chronic Hepatitis Ba HBeAg STATUS CLINICAL HBV DNA (IU/mL) ALT RECOMMENDATION HBeAg-reactive b

2 × 104 ≤2 × ULNc                   Chronic hepatitis Cirrhosis: Cirrhosis compensated  2 × 104d  2 × ULNd  2 × 103 < or > ULN ULN < or > ULN < or > ULN <2 × 103 Cirrhosis decompensated Detectable Undetectable HBeAg-negative b ≤2 × 103 ≤ULN 1 to >2 × ULNd Chronic hepatitis       2 × 103            Chronic hepatitis Cirrhosis: Cirrhosis compensated  2 × 103  2 × ULNd  2 × 103 < or > ULN ULN < or > ULN < or > ULN <2 × 103 Cirrhosis decompensated Detectable Undetectable aBased on practice guidelines of the American Association for the Study of Liver Diseases (AASLD). Except as indicated in footnotes, these guidelines are similar to those issued by the European Association for the Study of the Liver (EASL). bLiver disease tends to be mild or inactive clinically; most such patients do not undergo liver biopsy. cAccording to the EASL guidelines, treat if HBV DNA is >2 × 103 IU/mL and ALT >ULN. dOne of the potent oral drugs with a high barrier to resistance (entecavir or tenofovir) or PEG IFN can be used as first-line therapy (see text for further detail). eBecause HBeAg seroconversion is not an option, the goal of therapy is to suppress HBV DNA and maintain a normal ALT. PEG IFN is administered by subcutaneous injection weekly for a year; caution is warranted in relying on a 6-month posttreatment interval to define a sustained response, because the majority of such responses are lost thereafter. Oral agents, entecavir or tenofovir, are administered daily, usually indefinitely or until, as very rarely occurs, virologic and biochemical responses are accompanied by HBsAg seroconversion. fFor older patients and those with advanced fibrosis, consider lowering the HBV DNA threshold to >2 × 103 IU/mL. Abbreviations: ALT, alanine aminotransferase; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; HBV, hepatitis B virus; PEG IFN, pegylated interferon; ULN, upper limit of normal.

shown to “cure” hepatitis B, and none is likely to be competitive, unless it can be shown to go beyond current antivirals in achieving recovery (sustained HBsAg loss) from HBV infection. Whether combinations of immune modulators and direct antiviral agents will improve prospects for curative therapies remains to be seen. TREATMENT RECOMMENDATIONS Several learned societies and groups of expert physicians have issued treatment recommendations for patients with chronic hepa­ titis B; the most authoritative and updated are those of the AASLD and the EASL. Although the recommendations differ slightly, a con­ sensus has emerged on most of the important points (Table 352-4). No treatment is recommended or available for inactive “nonrepli­ cative” hepatitis B carriers (undetectable HBeAg with normal ALT and HBV DNA ≤103 IU/mL documented serially over time). Nor is treatment recommended currently for HBeAg-positive chronic hep­ atitis B with normal ALT (labeled by some as reflecting a relatively “immune-tolerant” phase but more accurately a period of dissocia­ tion between high-level HBV replication and a paradoxical paucity of inflammatory liver injury; see Chap. 350, “Complications and Sequelae”); however, such patients merit careful monitoring with HBV DNA and ALT at least every 6 months. Conversely, treatment is universally recommended for any patient with compensated cirrhosis and detectable HBV DNA; in decompensated cirrhosis, treatment should be carefully considered. Treatment considerations otherwise are based on the presence or absence of HBeAg, HBV DNA levels and presence or absence of active necroinflammatory activity (as based on ALT or liver biopsy).

CHAPTER 352 HBeAg Positive  In patients with detectable HBeAg and HBV DNA levels >2 × 104 IU/mL, treatment is recommended by the AASLD for those with ALT levels >2× the upper limit of normal. (The EASL recommends treatment in HBeAg-positive patients for HBV DNA levels >2 × 103 IU/mL and ALT above the upper limit of normal.) For HBeAg-positive patients with ALT ≤2× the upper limit of normal, in whom sustained responses are not likely Chronic Hepatitis No treatment; monitor, except in patients >40, with family history of cirrhosis or hepatocellular carcinoma, with extrahepatic manifestations, with a history of previous treatment, and/or with liver biopsy (or noninvasive fibrosis determination) evidence for moderate to severe inflammation or fibrosis Treatd  Treatd with oral agents, not PEG IFN Treatd with oral agents, not PEG IFN; refer for liver transplantation Observe; refer for liver transplantation Inactive carrier; treatment not necessary No treatment; monitor, except in patients >40, with family history of cirrhosis or hepatocellular carcinoma, with extrahepatic manifestations, with a history of previous treatment, and/or with liver biopsy (or noninvasive fibrosis determination) evidence for moderate to severe inflammation or fibrosis Treate,f  Treatd with oral agents, not PEG IFN Treatment suggestedf Treatd with oral agentsg, not PEG IFN; refer for liver transplantation Observe; refer for liver transplantation

and who would require multiyear therapy, antiviral therapy is not recommended currently. As noted above, these untreated patients (as well as patients with HBV DNA between 2 × 103 and 2 × 104 IU/mL) merit continued monitoring with HBV DNA and ALT every 3–6 months. Even in these groups (in the “gray zone”), however, therapy would be considered for patients >40 years of age, with extrahepatic manifestations of HBV infection, or with a family history of cirrhosis or HCC; patients whose liver biopsy or noninvasive testing shows moderate to severe necroinflammatory activity or fibrosis; or patients with a history of previous treatment. In HBeAg-positive patients, treatment may be discontinued after HBeAg loss. In a 2016 systematic review of 1716 patients involved in 25 clinical trials, responses after oral-agent therapy were found to be durable. Among patients with HBeAg-reactive chronic hepa­ titis B, the pooled rates of durable HBeAg seroconversions main­ tained after cessation of nucleoside/nucleotide analogue therapy (including all the oral agents) were 92 and 88% at posttreatment months 12 and 24, respectively, unaffected by the duration of postHBeAg-response consolidation therapy (>6 months in all studies evaluated); the pooled rate of durable biochemical remission after therapy in this population was 76%. Thus, per current AASLD recommendations, antiviral treatment with oral agents can be stopped after HBeAg loss/seroconversion in noncirrhotics, and the suggested period of consolidation therapy is 12 months with close monitoring for recurrent viremia (monthly × 6, then every 3 months for the rest of a year) after cessation of therapy.

HBeAg Negative  For patients with HBeAg-negative chronic hepatitis B, antiviral therapy is recommended if ALT is >2× the upper limit of normal (above the upper limit of normal according to EASL) and HBV DNA is >2 × 103 IU/mL. If HBV DNA is >2 × 103 IU/mL and ALT is 1 to >2× the upper limit of normal, the same considerations apply as for HBeAg-positive patients with border­ line ALT levels (the “gray zone”) for patients >40 years of age, with extrahepatic manifestations of HBV infection, or with a family his­ tory of cirrhosis or HCC; patients whose liver biopsy or noninvasive testing shows moderate to severe necroinflammatory activity or fibrosis; or patients with a history of previous treatment. This group also merits continued careful monitoring with HBV DNA and ALT every 3–6 months. For patients with HBeAg-negative chronic hepa­ titis, the current recommendation with oral agents is for indefinite therapy; stopping therapy in this group can be considered after HBsAg loss, which remains exceedingly rare. PART 10 Disorders of the Gastrointestinal System The potential for stopping antiviral therapy in noncirrhotic HBeAg-negative patients after protracted (≥2–5 years) antiviral therapy has been the subject of several studies After such pro­ longed courses of entecavir or tenofovir, in one small, 57-patient study (DARING-B), 18-month virologic relapse rates (HBV DNA

2000 IU/mL) were high (in 72%); however, 25% of study subjects underwent HBsAg loss after 18 months of follow-up. In a similar study (FINITE), virologic relapse rates were high, but 62% did not meet criteria for retreatment, and 19% lost HBsAg. In contrast, in a study among Asian patients, only ~30% had sustained responses for which resumption of therapy was not introduced, and HBsAg responses were negligible. In the only randomized, controlled trial of stopping therapy versus continuing therapy in HBeAg-negative patients after prolonged antiviral therapy (Toronto STOP study), only 33% had sustained responses after cessation of therapy, and HBsAg loss occurred with equal, small frequencies in both the stop-treatment group (4%) and the continue-treatment group (5%). Most recently, in a large 1552-subject, international cohort study (RETRACT B), treatment cessation in HBeAg-negative patients without cirrhosis was followed by virologic relapse rates of 68.9% at 12 months and 83.4% at 48 months. Relapse requiring retreat­ ment occurred in 29.8% of subjects at 12 months and 54.7% at 48 months. Similar to prior studies, HBsAg loss was observed in 3.2% of subjects at 12 months and increased to 13% at 48 months. Predictors of HBsAg loss were identified, which may offer guidance to determine which patients may benefit from stopping nucleos(t)

ide analogue therapy. In this study, the most important predictor of HBsAg loss was quantitative HBsAg level at the time therapy was discontinued (at posttreatment year 4, in Caucasian patients with HBsAg <1000 IU/mL, 41% had lost HBsAg, and in Asian patients with HBsAg <100 IU/mL, 33% had lost HBsAg). At pres­ ent, however, quantitative HBsAg determination is not routinely available in clinical care. Generally, then, although HBsAg loss can be achieved in a small fraction, and although a subgroup may not require reintroduction of therapy in the short run, enthusiasm for this approach is limited, and for most HBeAg-negative patients, recommendations support indefinite treatment, unless they experi­ ence HBsAg loss. For patients with compensated cirrhosis, because antiviral ther­ apy has been shown to retard clinical progression, treatment is recommended regardless of HBeAg status and ALT as long as HBV DNA is detectable. In addition, stopping therapy in patients with cirrhosis comes with the risk of decompensation. For patients with decompensated cirrhosis, treatment is recommended regardless of serologic and biochemical status, as long as HBV DNA is detect­ able. In addition, patients with decompensated cirrhosis should be evaluated as candidates for liver transplantation. Cirrhotics should be treated indefinitely (see considerations for stopping antiviral therapy in noncirrhotics, above). Choice of Treatment Agent  Among the eight available drugs for hepatitis B, PEG IFN has supplanted standard IFN, entecavir has supplanted lamivudine, and tenofovir has supplanted adefovir (Table 352-5). PEG IFN, entecavir, or tenofovir (TDF or TAF) is recommended as first-line therapy (Table 352-3). PEG IFN requires finite-duration therapy, achieves the highest rate of HBeAg TABLE 352-5  Pegylated Interferon Versus Oral Nucleoside Analogues for the Treatment of Chronic Hepatitis B   PEG IFN NUCLEOSIDE ANALOGUES Administration Weekly injection Daily, orally Tolerability Poorly tolerated, intensive monitoring Well tolerated, limited monitoring Duration of therapy Finite, 48 weeks ≥1 year, indefinite in most patients Maximum mean HBV DNA suppression 4.5 log10 6.9 log10 Effective in high-level HBV DNA (≥109 IU/mL) No Yes HBeAg seroconversion   During 1 year of therapy   During >1 year of therapy   ~30% Not applicable   ~20% 30% (year 2) to up to 50% (year 5) HBeAg-negative posttreatment HBV DNA suppression 17% at 5 years 7% at 4 years (lamivudine) HBsAg loss   During 1 year of therapy   During >1 year of therapy     After 1 year of therapy–

HBeAg-negative   3–4% Not applicable 12% at 5 years   0–3% 3–8% at 5 years of therapy 3.5% at 5 years Antiviral resistance None Lamivudine: ~30% year 1,

~70% year 5 Adefovir: 0% year 1, ~30% year 5 Telbivudine: up to 4% year 1, 22% year 2 Entecavir: ≤1.2% through year 6 Tenofovir: 0% through year 8 Use in cirrhosis, transplantation, immunosuppressed No Yes Cost, 1 year of therapy ++++

• to ++ Abbreviations: HBV, hepatitis B virus; HBeAg, hepatitis B e antigen; HBsAg, hepatitis B surface antigen; PEG IFN, pegylated interferon.

responses after a year of therapy, and does not support viral muta­ tions, but it requires subcutaneous injections and is associated with inconvenience, more intensive clinical and laboratory monitor­ ing, and intolerability. Oral agents do not require injections or cumbersome laboratory monitoring, are very well tolerated, lead to improved histology in 50−90% of patients, suppress HBV DNA more profoundly than PEG IFN, and are effective even in patients who fail to respond to IFN-based therapy. Although oral agents are less likely to result in HBeAg responses during the first year of therapy, as compared to PEG IFN, treatment with oral agents is usually extended beyond the first year and, by the end of the second year (of the current-generation, potent, high-barrier-to-resistance agents entecavir and tenofovir), yields HBeAg responses (and even HBsAg responses) comparable in frequency to those achieved after 1 year of PEG IFN (and without the associated side effects). Choice of agent should be based on clinical features, any history of prior treatment, and patient and physician preference; availability, rela­ tive cost, and insurance coverage often also play a role. As noted above, some physicians prefer to begin with PEG IFN, while most physicians and patients prefer oral agents as first-line therapy. Between the two first-line oral agents, entecavir and tenofovir, data to support the choice of one compared to the other are limited. Several observational studies have suggested that TDF is superior to entecavir in reducing the risk of HCC. Such studies, however, sophisticated statistical analyses notwithstanding, are subject to confounding influences that could favor TDF (e.g., physician pref­ erence for TDF in milder disease and for entecavir in more severe disease, including in older patients and those with impaired renal function). In addition, while several studies confirm a differential effect of TDF on long-term HCC risk, many others do not, and, in most cases, adjusting for such factors as age, sex, country, albumin, platelets, AFP, cirrhosis, and diabetes, eliminated the benefit of TDF over entecavir. Therefore, currently, the preponderance of data are insufficient to support this benefit of TDF over entecavir. Between TDF and TAF, TAF is preferred in patients with creatinine clearance <50 mL/min, reduced bone density, and risk factors for renal injury (e.g., decompensated cirrhosis, creatinine clearance <60 mL/min, poorly controlled hypertension or diabetes, proteinuria, active glo­ merulonephritis, concomitant nephrotoxic medications, or solidorgan transplantation). As noted above, the EASL recommendation for TAF extends to persons >60 years, who are at increased risk of TDF nephrotoxicity. POPULATIONS WITH UNIQUE TREATMENT CONSIDERATIONS Pregnancy  Substantial experience with lamivudine during preg­ nancy (see above) has identified no teratogenicity; although widely used during pregnancy, lamivudine remains classified as pregnancy category C. Although IFNs do not appear to cause congenital anomalies, these have antiproliferative properties and should be avoided during pregnancy. Adefovir during pregnancy has not been associated with birth defects; however, the risk of spontaneous abortion may be increased, and adefovir is categorized as pregnancy category C. Data on the safety of entecavir during pregnancy have not been published (pregnancy category C). Sufficient data in animals and limited data in humans suggest that telbivudine and tenofovir (both pregnancy category B) can be used safely during pregnancy; however, telbivudine is not an acceptable first-line drug. In general, then, except for lamivudine and tenofovir, and until additional data become available, the other antivirals for hepatitis B should be avoided or used with extreme caution during pregnancy. Tenofovir is the current drug of choice in pregnancy. Children  For children aged 2 to <18 years old with HBeAgpositive hepatitis B (most children will be HBeAg-positive; children with HBeAg-negative chronic hepatitis B have not been well rep­ resented in treatment studies), treatment is recommended if HBV DNA is detectable and ALT levels are elevated, but not if ALT levels are normal. Each of the available first-line drugs is approved for dif­ ferent childhood age groups (PEG IFN α2a age ≥5 years [approved

for hepatitis C, not B, but can be used in hepatitis B]; entecavir and tenofovir age ≥2 years). Package inserts should be consulted for childhood doses.

HIV Co-infection  Patients with HBV-HIV co-infection can have progressive HBV-associated liver disease and, occasionally, a severe exacerbation of hepatitis B resulting from immunologic recon­ stitution following antiretroviral therapy. Tenofovir and the com­ bination of tenofovir and emtricitabine in one pill are approved therapies for HIV and represent excellent choices for treating HBV infection in HBV-HIV co-infected patients. Generally, even for HBV-HIV co-infected patients who do not yet meet treatment criteria for HIV infection, treating for both HBV and HIV is recom­ mended. In HIV-HBV co-infection, TAF, because of its better safety profile, is preferable to TDF. Cirrhosis  For patients with compensated or decompensated cir­ rhosis, PEG IFN should not be used; the emergence of such lifethreatening side effects as sepsis can result in further deterioration and loss of antiviral effectiveness. Therefore, in this patient subset, therapy with a very favorable resistance profile (e.g., entecavir or tenofovir) is the preferred treatment. For patients with end-stage chronic hepatitis B who undergo liver transplantation, reinfec­ tion of the new liver is almost universal in the absence of antiviral therapy. The majority of patients become high-level viremic carriers with minimal liver injury. Before the availability of antiviral therapy, an unpredictable proportion experienced severe hepatitis B-related liver injury, sometimes a fulminant-like hepatitis and sometimes a rapid recapitulation of the original severe chronic hepatitis B (Chap. 350). Currently, however, prevention of recurrent hepati­ tis B after liver transplantation has been achieved definitively by combining short-term (5–7 days) intravenous hepatitis B immune globulin (HBIG) with lifelong low-resistance oral entecavir or TDF or TAF (Chap. 356); in some patients, especially those with a low risk for recurrence, the newer, more potent, and less resistanceprone oral agents may be used instead of HBIG for posttransplanta­ tion therapy. For patients at high risk for recurrence and progressive disease (e.g., patients with HDV-HBV or HIV-HBV co-infections) as well as for nonadherent patients, lifelong combination HBIG-oral agent therapy should be considered. For patients receiving livers from anti-HBc-positive donors, lifelong oral-agent therapy is rec­ ommended (without HBIG). CHAPTER 352 Chronic Hepatitis Immunosuppression  Patients with chronic hepatitis B who undergo cytotoxic chemotherapy for treatment of malignancies as well as patients treated with immunosuppressive, anticytokine, or anti–tumor necrosis factor (TNF) therapies experience enhanced HBV replication and viral expression on hepatocyte membranes during chemotherapy coupled with suppression of cellular immu­ nity. When chemotherapy is withdrawn, such patients are at risk for reactivation of hepatitis B, often severe and occasionally fatal. Such rebound reactivation represents restoration of cytolytic T-cell func­ tion against a target organ enriched in HBV expression. Preemptive treatment with the first of the oral HBV antivirals, lamivudine, prior to the initiation of chemotherapy was shown to reduce the risk of such reactivation substantially; treating after reactivation has occurred is less effective. The newer, more potent oral antiviral agents, entecavir and tenofovir, which are even more effective in preventing hepatitis B reactivation and with a lower risk of antiviral drug resistance, are preferred. The risk of HBV reactivation and liver injury varies from highest (e.g., B cell–depleting agents, anthracycline derivatives, moder­ ate-/high-dose corticosteroids for ≥4 weeks) to moderate (e.g., TNF-α inhibitors, cytokine or integrin inhibitors, tyrosine kinase inhibitors, low-dose corticosteroids for ≥4 weeks) to lowest (e.g., immunosuppressive agents like methotrexate and azathioprine, intraarticular corticosteroids, any dose of corticosteroids for ≤1 week). Most authorities (e.g., Centers for Disease Control and Prevention, AASLD, American Gastroenterological Association, EASL) recommend HBsAg and anti-HBc (± anti-HBs) screen­ ing of all patients prior to initiation of chemotherapy or other

long-term immunosuppressive regimens. Preemptive antiviral pro­ phylaxis is recommended for any patient with detectable HBsAg, given the high risk of reactivation (which can occur in ~25% of cases but potentially in up to 68%). While, generally, reactivation in HBsAg-negative but anti-HBc-reactive patients is rare, in this population, careful monitoring is recommended, except in the set­ ting of B cell–depleting therapies, which are associated with a high risk for reactivation. Thus, antiviral prophylaxis is recommended in HBsAg-negative/anti-HBc-positive patients undergoing B-cell depletion therapy or stem cell transplantation.

The optimal duration of antiviral therapy after completion of chemotherapy is not known, but a suggested approach is 6 months (12 months for B cell–depleting agents) for inactive hepatitis B car­ riers and longer-duration therapy in patients with baseline HBV DNA levels >2 × 103 IU/mL, until standard clinical endpoints are met (Table 352-4). ■ ■CHRONIC HEPATITIS D (DELTA HEPATITIS) Chronic hepatitis D virus (HDV) infection is most commonly observed in patients who were exposed to HDV while already harboring chronic HBV infection (so-called superinfection). Chronic HDV infection may also follow acute co-infection with HBV and HDV but at a rate no higher than the rate of chronicity of acute hepatitis B. That is, although HDV co-infection can increase the severity of acute hepatitis B, HDV does not increase the likelihood of progression to chronic hepatitis B. When, however, HDV superinfection occurs in a person who is already chron­ ically infected with HBV, long-term HDV infection is the rule, and a worsening of the liver disease is the expected consequence. Except for severity, chronic hepatitis B plus D has similar clinical and laboratory features to those seen in chronic hepatitis B alone. Relatively severe and progressive chronic hepatitis, with or without cirrhosis, is the rule, and mild chronic hepatitis is the exception. Occasionally, however, mild hepatitis or even, rarely, inactive carriage occurs in patients with chronic hepatitis B plus D, and the disease may become indolent after several years of infection. The clinical and laboratory features of chronic HDV infection are summarized in Chap. 350. PART 10 Disorders of the Gastrointestinal System TREATMENT Chronic Hepatitis D Management is not well defined, and the host cellular RNA poly­ merase upon which HDV replication depends cannot be targeted by conventional antiviral agents. Treatment with type I interferons is the only approved treatment for chronic HDV infection, although, per expert opinion, treating underlying HBV infection is suggested as well if the HBV DNA level is >2 × 103 IU/mL. Preliminary experimental trials of IFN-α suggested that conventional doses and durations of therapy lower levels of HDV RNA and aminotransfer­ ase activity only transiently during treatment but have no impact on the natural history of the disease. In contrast, high-dose IFN-α (9 million units three times a week) for 12 months was reported to be associated with a sustained loss of HDV replication and clinical improvement in up to 50% of patients. Furthermore, a substantial improvement in 12-year survival after liver transplantation was observed in patients with hepatitis D who received high-dose IFN-α (86 vs 31% with standard-dose IFN-α). A suggested approach to therapy has been high-dose, long-term IFN for at least a year and, in responders, extension of therapy until HDV RNA and HBsAg clearance; however, extension of therapy to a second year provided no advantage, and sustained responses after completion of therapy have been rare. Compared to standard IFN-α, PEG IFN has been shown to be more effective in chronic hepatitis D but still of limited therapeutic value; after 48 weeks of therapy, durable undetectable HDV RNA for 24 posttreatment weeks has been reported in a quarter to just over a half of patients. Disappointingly, loss of viro­ logic responses (reappearance of HDV RNA) was observed during long-term (median 4.5 years) monitoring in a majority of 24-weekposttreatment responders, with durable HDV RNA suppression to

undetectable in only 12%. None of the nucleoside analogue anti­ viral agents for hepatitis B are effective in hepatitis D, and adding ribavirin or oral nucleoside agents (lamivudine and, more recently, adefovir and tenofovir) to PEG IFN was not found to be more effec­ tive than PEG IFN monotherapy. While recommended, 12 months of PEG IFN therapy is far from satisfactory. The lackluster response to type I interferon treatment with or without nucleoside agents highlights the need for novel therapies to target HDV infection. The approach farthest along in clinical testing is the entry inhibi­ tor bulevirtide, a lipopeptide that blocks the NTCP receptor, pre­ venting hepatocyte entry of both HBV and HDV. In phase 2 studies of subcutaneous bulevirtide alone or combined with PEG IFN for varying treatment durations, a significant reduction occurred in HDV RNA in >71% of subjects, with a more pronounced effect in combination therapy. Similarly, a phase 2 study of bulevirtide combined with tenofovir (vs tenofovir monotherapy) for 24 weeks demonstrated significant reductions in HDV RNA at the end of treatment (50–77% vs 4% with tenofovir alone), but rebound of viremia was universal at follow-up week 48. Most recently, in an ongoing phase 3 clinical trial, ALT normalized in 51–56% of patients receiving bulevirtide after 48 weeks of treatment, and 45–48% of patients had a ≥2 log10 reduction in or undetectable HDV RNA; this treatment effect was maintained though 96 weeks of treatment. In 2020, bulevirtide was authorized conditionally by the European Medicines Agency for the treatment of HDV with or without PEG IFN or nucleoside analogues. Review by the U.S. Food and Drug Administration (FDA) is anticipated. Preliminary trials have also been performed with lonafarnib, a prenylation inhibitor, which is combined with ritonavir to improve bioavailability. Prenylation, the posttranslational covalent addition of the prenyl lipid farnesyl to large HDV antigen, is required for this HDV protein to interact and form secreted viral particles with HBsAg. Phase 2 studies of lonafarnib/ritonavir have demonstrated significant, on-treatment reductions in HDV RNA, and in one study, the addition of IFNα increased the likelihood of achieving a viral response to 89%. Additional experimental approaches to the treatment of hepatitis D include nucleic acid polymer therapy, often combined with PEG IFNα. Nucleic acid polymer therapy is thought to inhibit HBsAg release and has been administered alone or with PEG IFN; to date, these studies have been done in one Eastern European site. While promising reductions in HDV RNA and HBsAg have been reported, these studies have been plagued by adverse effects, and confirmatory studies have not materialized. In patients with end-stage liver disease secondary to chronic hepatitis D, liver transplantation has been effective. If hepatitis D recurs in the new liver without the expression of hepatitis B (an unusual serologic profile in immunocompetent persons but com­ mon in transplant patients), liver injury is limited. In fact, the outcome of transplantation for chronic hepatitis D is superior to that for chronic hepatitis B; in such patients, combination HBIG and nucleoside analogue therapy for hepatitis B is indicated (Chap. 356). ■ ■CHRONIC HEPATITIS C Regardless of the epidemiologic mode of acquisition of hepatitis C virus (HCV) infection, chronic hepatitis follows acute hepatitis C in 50−70% of cases; chronic infection is common even in those with a return to normal in aminotransferase levels after acute hepatitis C, adding up to an 85% likelihood of chronic HCV infection after acute hepatitis C. Few clues had emerged to explain host differences associ­ ated with chronic infection until recently, when variation in a single nucleotide polymorphism (SNP) on chromosome 19, IL28B (which codes for IFN-λ3, now renamed IFNL3), was identified that distin­ guished between responders and nonresponders to IFN-based antiviral therapy (see below). The same variants correlated with spontaneous resolution after acute infection: 53% in genotype C/C, 30% in geno­ type C/T, but only 23% in genotype T/T. The association with HCV clearance after acute infection is even stronger when IL28B (IFNL3)

haplotype is combined with haplotype G/G of an SNP near human leukocyte antigen (HLA) class II DBQ1*03:01. In patients with chronic hepatitis C followed for 20 years, progres­ sion to cirrhosis occurs in ~20−25%. Such is the case even for patients with relatively clinically mild chronic hepatitis, including those without symptoms, with only modest elevations of aminotransferase activity, and with mild chronic hepatitis on liver biopsy. Even in cohorts of well-

compensated patients with chronic hepatitis C referred for clinical research trials (no complications of chronic liver disease and with nor­ mal hepatic synthetic function), the prevalence of cirrhosis may be as high as 50%. Most cases of hepatitis C are identified initially in asymp­ tomatic patients who have no history of acute hepatitis C (e.g., those dis­ covered while attempting to donate blood, while undergoing lab testing as part of an application for life insurance, or as a result of routine labo­ ratory tests). The source of HCV infection in many of these cases is not defined, although a long-forgotten percutaneous exposure (e.g., injection drug use) in the remote past can be elicited in a substantial proportion and probably accounts for most infections; most of these infections were acquired in the 1960s and 1970s among persons in the 1945–1965 birth cohort (Chap. 350), coming to clinical attention decades later. Approximately one-third of patients with chronic hepatitis C have normal or near-normal aminotransferase activity; although one-third to one-half of these patients have chronic hepatitis on liver biopsy, the grade of liver injury and stage of fibrosis tend to be mild in the vast majority. In some cases, more severe liver injury has been reported— even, rarely, cirrhosis, most likely the result of previous histologic activ­ ity. Among patients with persistent normal aminotransferase activity sustained over ≥5−10 years, histologic progression has been shown to be rare; however, approximately one-fourth of patients with normal aminotransferase activity experience subsequent aminotransferase elevations, and histologic injury can be progressive once abnormal biochemical activity resumes. Therefore, continued clinical monitor­ ing and antiviral therapy are indicated, even for patients with normal aminotransferase activity. Despite this substantial rate of progression of chronic hepatitis C and even though liver failure can result from end-stage chronic hepati­ tis C, the long-term prognosis over 1–2 decades for chronic hepatitis C in most patients is relatively benign. Mortality over 10−20 years among patients with transfusion-associated chronic hepatitis C has been shown not to differ from mortality in a matched population of trans­ fused patients in whom hepatitis C did not develop. Although death in the hepatitis group is more likely to result from liver failure and although hepatic decompensation may occur in ~15% of such patients over the course of a decade, the majority (almost 60%) of patients remain asymptomatic and well compensated, with no clinical sequelae of chronic liver disease. Overall, chronic hepatitis C tends to be very slowly and insidiously progressive, if at all, in most patients, whereas in approximately one-fourth of cases, chronic hepatitis C will progress eventually to end-stage cirrhosis. In fact, because HCV infection is so prevalent, and because a proportion of patients progress inexorably to end-stage liver disease, hepatitis C was the most frequent indication for liver transplantation (Chap. 356) in the era prior to the availability of direct-acting antiviral (DAA) therapy (see below). In the United States, hepatitis C accounts for up to 40% of all chronic liver disease; as of 2007, mortality caused by hepatitis C surpassed that associated with HIV/AIDS, and as of 2012, reported deaths caused by hepatitis C surpassed those associated with all other notifiable infectious diseases (HIV, tuberculosis, hepatitis B, and 57 other infectious diseases). More­ over, because the prevalence of HCV infection is so much higher in the “baby boomer” cohort born between 1945 and 1965, three-quarters of the mortality associated with hepatitis C occurs in this age cohort. In a 2010 modeling study based on the prevalence of HCV infec­ tion and its known natural history, liver disease mortality associated with hepatitis C was projected to peak in the year 2022; however, the observed HCV-related mortality in the United States peaked in 2013 and has fallen since, paralleling the introduction in 2013 of highly effective DAA agents (see below). The declining death rates related to HCV infection following the wide use of DAAs have been observed across most of the United States.

Progression of liver disease in patients with chronic hepatitis C has been reported to be more likely in patients with older age, longer dura­ tion of infection, advanced histologic stage and grade, more complex HCV quasispecies diversity, increased hepatic iron, concomitant other liver disorders (alcoholic liver disease, chronic hepatitis B, hemochro­ matosis, α1 antitrypsin deficiency, and steatohepatitis), HIV infection, and obesity. Among these variables, however, duration of infection appears to be one of the most important, and some of the others prob­ ably reflect disease duration to some extent (e.g., quasispecies diversity, hepatic iron accumulation). No other epidemiologic or clinical features of chronic hepatitis C (e.g., severity of acute hepatitis, level of amino­ transferase activity, level of HCV RNA, presence or absence of jaundice during acute hepatitis) are predictive of eventual outcome. Despite the relatively benign nature of chronic hepatitis C over time in many patients, cirrhosis following chronic hepatitis C has been associated with the late development, after several decades, of HCC (Chap. 87); the annual rate of HCC in cirrhotic patients with hepatitis C is 1–4%, occurring primarily in patients who have had HCV infection for 30 years or more.

Referral bias may account for the more severe outcomes described in cohorts of patients reported from tertiary care centers (20-year progres­ sion of ≥20%) versus the more benign outcomes in cohorts of patients monitored from initial blood-product-associated acute hepatitis or identified in community settings (20-year progression of only 4−7%). Still unexplained, however, are the wide ranges in reported progression to cirrhosis, from 2% over 17 years (eventually 19% over 36 years) in a population of Irish women with hepatitis C infection acquired from contaminated anti-D immune globulin to 30% over ≤11 years in recipi­ ents of contaminated intravenous immune globulin. CHAPTER 352 Perhaps the best prognostic indicator in chronic hepatitis C is liver histology; the rate of hepatic fibrosis may be slow, moderate, or rapid. Patients with mild necrosis and inflammation as well as those with limited fibrosis have an excellent prognosis and limited progression to cirrhosis. In contrast, among patients with moderate to severe necro­ inflammatory activity or fibrosis, including septal or bridging fibrosis, progression to cirrhosis is highly likely over the course of 10−20 years. Among patients with compensated cirrhosis associated with hepatitis C, the 10-year survival rate is close to 80%; mortality occurs at a rate of 2−6% per year; decompensation at a rate of 4−5% per year; and, as noted above, HCC at a rate of 1−4% per year. Estimates of the natural history of chronic hepatitis C have been made, based on data available on the prevalence of HCV infection in the U.S. population and on the rate of disease progression. Weighted primarily by the concentration of chronic hepatitis C in the baby boomer generation, the peak prevalence was estimated to have occurred in 2015. The calculated frequency of cirrhosis in U.S. patients with hepatitis C was 5% in 1990 and 25% in 2010 and was projected to be 37% in 2020. A discussion of the patho­ genesis of liver injury in patients with chronic hepatitis C appears in Chap. 350. Chronic Hepatitis Clinical features of chronic hepatitis C are similar to those described above for chronic hepatitis B. Generally, fatigue is the most common symptom; jaundice is rare. Immune complex–mediated extrahepatic complications of chronic hepatitis C are less common than in chronic hepatitis B (despite the fact that assays for immune complexes are often positive in patients with chronic hepatitis C), with the exception of essential mixed cryoglobulinemia (Chap. 350), which is linked to cutaneous vasculitis and membranoproliferative glomerulonephritis as well as lymphoproliferative disorders such as B-cell lymphoma and unexplained monoclonal gammopathy. In addition, chronic hepatitis C has been associated with extrahepatic complications unre­ lated to immune-complex injury. These include Sjögren’s syndrome, lichen planus, porphyria cutanea tarda, renal injury, type 2 diabetes mellitus, and the metabolic syndrome (including insulin resistance and steatohepatitis). In addition, a link has been observed between HCV infection and cardiovascular/cerebrovascular disease, rheumatologic/ immunologic disorders, mental health and cognitive disorders, and nonliver malignancies. Laboratory features of chronic hepatitis C are similar to those in patients with chronic hepatitis B, but aminotransferase levels tend to

fluctuate more (the characteristic episodic pattern of aminotransferase activity) and to be lower, especially in patients with long-standing dis­ ease. An interesting and occasionally confusing finding in patients with chronic hepatitis C is the presence of autoantibodies. Rarely, patients with autoimmune hepatitis (see below) and hyperglobulinemia have false-positive immunoassays for anti-HCV. On the other hand, some patients with serologically confirmable chronic hepatitis C have circu­ lating anti-LKM. These antibodies are anti-LKM1, as seen in patients with autoimmune hepatitis type 2 (see below), and are directed against a 33-amino-acid sequence of cytochrome P450 IID6.

Histopathologic features of chronic hepatitis C, especially those that distinguish hepatitis C from hepatitis B, are described in Chap. 350. TREATMENT Chronic Hepatitis C Therapy for chronic hepatitis C has evolved substantially in the 30 years since IFN-α was introduced for this indication in 1991. The therapeutic armamentarium grew to include PEG IFN with ribavirin and, then, in 2011, the introduction of the first protease inhibitors, telaprevir and boceprevir, used in combination with PEG IFN and ribavirin in patients with HCV genotype 1. The field of antiviral therapy for hepatitis C was transformed beginning in 2013, with the approval of the first nucleoside analogue polymerase inhibitor sofosbuvir. Although several of these combination regi­ mens have been supplanted by better, later-generation drugs, as of 2020, five all-oral, highly effective (>95%), low-resistance, pangeno­ typic, well-tolerated, short-duration (primarily 8–12 weeks) combi­ nation regimens of DAA drugs are recommended. The remarkable historical evolution of antiviral therapy for hepatitis C is instructive. THE INTERFERON ERA (1991–2011) IFN-based therapy has been supplanted by DAA agents introduced in the second decade of the twenty-first century; however, many important lessons about antiviral therapy for chronic hepatitis C were learned from the experience with IFN-based treatment, and many of the limitations of—and disparities in responsiveness to—IFN-based therapy have been overcome by current-generation DAA treatments. Mechanistically, HCV proteins inhibit several steps of the JAK-STAT signal transduction pathway, and by acti­ vation of JAK-STAT signaling, exogenous IFN culminates in and restores intracellular expression of IFN-stimulated genes and their protein products that have antiviral properties. PART 10 Disorders of the Gastrointestinal System When first approved, subcutaneous IFN-α three times a week for 6 months achieved a sustained virologic response (SVR; defined then as a reduction of HCV RNA to PCR-undetectable levels ≥24 weeks after completion of therapy) in <10%. Doubling Structural proteins Nonstructural proteins E1 E2 NS2 p7 NS3 C 4A NS5A NS5B Target NS3-4A protease inhibitors NS5A inhibitors NS5B inhibitors Function Block N3-4A cleavage Inhibit NS5A protein, essential for replication Suffix -previr -asvir -buvir Drugs Glecaprevir, voxilaprevir Velpatasvir, ledipasvir, pibrentasvir Sofosbuvir FIGURE 352-2  Hepatitis C virus (HCV) viral genome and drug targets. The HCV viral genome encodes three structural proteins (c-nucleocapsid and e1-2 envelope proteins) and seven nonstructural proteins. Current, first-line direct-acting antivirals have three viral targets: (1) the NS3/4A viral protease; (2) the NS5A protein, which is a critical component of the viral replication apparatus; and (3) NS5B RNA-dependent RNA polymerase.

the duration of therapy increased the SVR rate to ~20%, and addi­ tion to the regimen of daily ribavirin (ineffective when used alone), an oral guanosine nucleoside, increased the SVR rate further to 40% by reducing the likelihood of virologic relapse after comple­ tion of treatment. Although its mechanism of action remains poorly understood, ribavirin retains a limited role in supporting DAA agents in several subgroups of otherwise refractory patients (see below). Beginning in 2011, for the treatment of hepatitis C, standard IFNs were supplanted by PEG IFNs, which have substantially longer half-lives, permitting administration once (rather than three times) a week. Treatment with the combination of PEG IFN and ribavirin increased SVR rates to 55% overall—to >40% in genotypes 1 and 4, requiring 48 weeks of therapy, and to >80% in genotypes 2 and 3, requiring only 24 weeks of therapy—and resulted in histologic improvement in approximately three-fourths of patients. After initiation of IFN treatment, ALT levels fell precipitously, and up to 90% of virologic responses were achieved within the first 12 weeks of therapy. Failure to achieve an early virologic response (EVR), a ≥2 log10 reduction in HCV RNA by week 12, predicted failure to experience a subsequent SVR. Similarly, patients in whom HCV RNA became undetectable within 4 weeks (i.e., who achieve a rapid virologic response [RVR]) had a very high likelihood of achieving an SVR (Fig. 352-2). Most relapses occurred within the first 12 weeks after treatment, and absence of HCV RNA 12 weeks after completion of therapy has become the current standard for SVR (SVR12); relapses are very rare 6 months to a year after SVR and almost unheard of after 2 years. Of documented durability decades after successful therapy, an SVR to antiviral therapy for chronic hepatitis C is tantamount to a cure and is followed by marked improvements in liver disease outcomes (see below). Patient variables that correlated with IFN-based SVRs included favorable genotype (genotypes 2 and 3 as opposed to genotypes 1 and 4); low baseline HCV RNA level (<800,000 IU/mL); histologi­ cally mild hepatitis and minimal fibrosis, especially absence of cir­ rhosis; immunocompetence; age <40; female gender; and absence of comorbid metabolic disease. High levels of HCV RNA, more histologically advanced liver disease, and high HCV quasispecies diversity all went hand in hand with advanced duration of infection and reduced IFN responsiveness. Ironically, patients whose disease was least likely to progress were the ones most likely to respond to IFN and vice versa. As described above in the discussion of spontaneous recovery from acute hepatitis C, IFN gene variants discovered in genomewide association studies were shown to have a substantial impact on IFN responsiveness of patients with genotype 1 to antiviral therapy. In studies of patients treated with PEG IFN and ribavirin, variants RNA-dependent RNA polymerase Phosphoprotein NS3/4a viral protease Block the RNAdependent RNA polymerase

of the IL28B (now renamed IFNL3) SNP that code for IFN-λ3 (a type III IFN, the receptors for which are more discretely distributed than IFN-α receptors and more concentrated in hepatocytes) cor­ related significantly with responsiveness. Homozygotes for the C allele at this locus (C/C) achieved SVRs of ~80%, heterozygotes (C/T) SVRs of ~35%, and homozygotes for the T allele (T/T) SVRs of ~25%. Side effects of IFN therapy are described in the section on treatment of chronic hepatitis B (see above). Besides ribavirinassociated nasal and chest congestion, pruritus, and precipitation of gout, the most pronounced ribavirin side effect is hemolysis, often requiring dose reduction or addition of erythropoietin therapy (not shown, however, to increase the likelihood of an SVR); therefore, close monitoring of blood counts is crucial, and ribavirin should be avoided in patients with anemia, hemoglobinopathies, coronary artery disease or cerebrovascular disease, or renal insufficiency (the drug is excreted renally) and in pregnancy (the drug is teratogenic, mandating contraception during, and for several months after, ther­ apy in women of child-bearing age [because of their antiprolifera­ tive properties, IFNs also are contraindicated during pregnancy]). For most of the decade prior to 2011, when protease inhibitors were introduced for HCV genotype 1 (see below), the standard of care was a combination of PEG IFN plus ribavirin (unless riba­ virin was contraindicated) for all HCV genotypes. Even after the introduction of protease inhibitors for genotypes 1 and 4, however, PEG IFN–ribavirin remained the standard of care for patients with genotypes 2 and 3 until late 2013. Responsiveness to IFN-ribavirin– based therapy was diminished in immunocompromised patients and in patients with HIV-HCV co-infection and contraindicated in patients with decompensated liver disease or end-stage renal disease. The cumbersome nature of IFN-ribavirin–based therapy (injections, complicated laboratory monitoring, side effects and poor tolerability, modest efficacy, variables and patient subsets associated with poor responsiveness, tailored therapy, futility rules, etc.) was supplanted eventually (in 2016) by DAAs for all genotypes (see below). Most of the variables associated with poor responsiveness to IFN-based ther­ apy became irrelevant, and difficult-to-treat patient subpopulations began to experience responses to DAAs that were indistinguishable from responses in standard patients (see below). Persons with chronic HCV infection suffer increased liverrelated mortality, all-cause mortality, and multiple extrahepatic disorders (see above). On the other hand, successful antiviral therapy of chronic hepatitis C resulting in an SVR was shown to improve survival (and to reduce the need for liver transplantation); to lower the risk of liver failure, liver-related death, and all-cause mortality; to slow the progression of chronic hepatitis C; to reverse fibrosis and even cirrhosis; and to improve such HCV-associated extrahepatic disorders as type 2 diabetes and renal disease. Whereas the 10- and 20-year survival in the absence of an SVR is reduced in cirrhotic patients with chronic hepatitis C, survival at these inter­ vals after an SVR has been found to be indistinguishable from that of the general population. In cirrhotic patients (and in those with advanced fibrosis), although successful treatment reduces mortality and liver failure (three- to fourfold 10-year reduction) and reduces the need for liver transplantation and the likelihood of HCC (14fold 10-year reduction), the risk of liver-related death and HCC persists, albeit at a much reduced level, necessitating continued clinical monitoring and cancer surveillance after SVR in patients with advanced fibrosis and cirrhosis. On the other hand, in the absence of an SVR, IFN-based therapy does not reduce the risk of HCC. Fortunately, PEG IFN–ribavirin nonresponders can now be retreated with DAAs and experience SVR rates comparable to those in treatment-naïve persons (see below). FIRST-GENERATION PROTEASE INHIBITORS (2011–2013) The HCV RNA genome encodes a single polyprotein, which is cleaved during and after translation by host and viral-encoded proteases (Fig. 352-2). One protease involved in the cleavage of the viral polyprotein is an NS3/4A viral protein that has serine protease

activity. Telaprevir and boceprevir were first-generation serine pro­ tease inhibitors that target the genotype 1 NS3/4A. In 2011, telapre­ vir and boceprevir used in combination with PEG IFN and ribavirin were approved by the FDA as the first oral DAA agents for the treatment of hepatitis C genotype 1 (not other genotypes) in adults with stable liver disease. Although now replaced by more effective, all-oral regimens, these first-in-class agents represented a break­ through in the treatment of chronic hepatitis C and established milestones against which subsequent therapies could be measured.

Because selection of resistant variants developed rapidly during monotherapy with telaprevir and boceprevir, these drugs had to be used in combination with PEG IFN and ribavirin. Ribavirin, in particular, appeared to reduce relapse rates significantly in protease-inhibitor-based regimens, such that those who could not take or were intolerant to ribavirin were unlikely to benefit from the addition of these agents. Telaprevir and boceprevir regimens consisted of periods of triple therapy (protease inhibitor plus PEG IFN plus ribavirin) and periods of dual therapy (PEG IFN plus riba­ virin). Cumbersome periods of “run-in” therapy or late-treatment consolidation therapy with PEG IFN-ribavirin, reliance on early virologic responses to determine duration of therapy (“responseguided therapy”), and futility rules (for stopping therapy based on poor early responsiveness) rendered first-generation DAA treat­ ment regimens complicated and unwieldy. For patients with HCV genotype 1, protease inhibitors improved the frequency of RVRs and SVRs significantly as compared to PEG IFN plus ribavirin alone. In treatment-naïve patients, telaprevirbased SVRs were achieved in up to 79% of patients who received 12 weeks of triple therapy followed by 12–36 weeks of dual therapy, with higher rates of response (83–92%) observed in those with undetectable RNA at weeks 4 and 12. In studies with boceprevir in treatment-naïve patients, SVRs occurred in 59–66% of patients, and among those with undetectable HCV RNA at 8 weeks, the SVR rate increased to 86–88%. Adding to the complexity of treatment with these protease inhibitors were absolute stopping rules for futility, that is, absence of HCV RNA reductions at critical treat­ ment milestones, which were shown to be invariably predictive of nonresponse. CHAPTER 352 Chronic Hepatitis The first-generation protease inhibitors were successful as well among patients in whom prior PEG IFN plus ribavirin had failed, even in “null responders,” who had never experienced even a 2 log10 reduction in HCV RNA while on treatment. In this hardest-to-treat group, telaprevir-based therapy achieved an SVR in 29–33% and boceprevir achieved an SVR in 30–40%. In a substantial proportion of protease inhibitor nonresponders, resistance-associated substitutions (RASs; previously referred to as resistance-associated variants [RAVs]) could be identified, but these variants were not archived, and wild-type HCV reemerged in almost all cases within 1.5–2 years. SVRs to these protease inhibitors were highest in prior relapsers and treatment-naïve patients (with SVR rates of >75%) and lowest in cirrhotic prior null responders, for whom no benefit accrued over PEG IFN–ribavirin treatment. Both protease inhibitors had substantial toxicities. Telaprevir was associated with a severe, generalized (trunk and extremities), often confluent, maculopapular, pruritic rash in ~6% of treated patients (that required careful dermatologic monitoring in all patients and systemic corticosteroid therapy in the most severely affected). Other common side effects included pruritus, rectal burning, nau­ sea, diarrhea, fatigue, dysgeusia (altered or unpleasant taste), and anemia, which required close monitoring, could be relatively refrac­ tory, and occasionally required transfusion and even hospitalization (especially in cirrhotic prior nonresponders). Anemia occurred in half of boceprevir-treated patients. The other most common side effects of boceprevir were fatigue, nausea, headache, dysgeusia, dry mouth, vomiting, and diarrhea. Both drugs came with an inconveniently high pill burden and had to be administered every 8 h with food (telaprevir with a 20-g fat meal). Use of protease inhibitors was further complicated by numerous drug-drug interactions. As telaprevir and boceprevir are

both eliminated by and inhibit CYP3A4, these agents could not be administered with other medications that induce CYP3A4 or are dependent on CYP3A4 for elimination. Care had to be taken to examine for any potential interactions between these protease inhibitors and other medications the patient was taking, and a convenient website became available to check for such drug-drug interactions (www.hep-druginteractions.org). This website remains updated and of important clinical utility for checking drug-drug interactions for current HCV DAA regimens.

Despite the improvement in SVRs with protease-inhibitor-based regimens for genotype 1 compared to PEG IFN–ribavirin (e.g., in treatment-naïve patients 66–79% vs 38–44%), triple-drug protease inhibitor therapy was hampered by amplified intolerability, the complexity of response-guided regimens and futility stopping rules, the inconvenience of thrice-daily dosing with meals and a high pill burden, the need for PEG IFN injections and ribavirin with all their intolerability, and multiple drug-drug interactions. Moreover, side effects appeared to be more severe and burdensome once these drugs entered practice, especially in cirrhotic nonresponders, in whom studies reported from Europe showed serious adverse events in up to 45% and deaths in up to 3%. All these issues, as well as rapidly accelerating progress on next-generation and all-oral DAA TABLE 352-6  Indications and Recommendations for Antiviral Therapy of Chronic Hepatitis Ca Standard Indications for Therapy All patients with chronic HCV infection (detectable HCV RNA, with or without elevated ALT) except for those with short life expectancies owing to comorbid conditions. Any stage of fibrosis; pretreatment biopsy is no longer embraced and has been supplanted by noninvasive measures of fibrosis, e.g., imaging to determine liver elasticity. Retreatment Recommended Relapsers, partial responders, or nonresponders after a previous course of interferon-based therapy or prior direct-acting antiviral therapy Antiviral Therapy Not Recommended Pregnancy: No sizable clinical studies of directacting antivirals during pregnancy are available. Ribavirin is contraindicated during pregnancy; therefore, any regimen including ribavirin should not be used. Sofosbuvir and sofosbuvir plus ledipasvir are classified as pregnancy category B, but the other direct-acting antivirals do not have a pregnancy classification. Therefore, these therapies are not indicated routinely in pregnancy and should be used, with caution, only if the benefit of treatment outweighs the potential for fetal risk. Therapeutic Regimens (based on AASLD-IDSA recommendations, www.hcvguidelines.org)b   Treatment-Naïve Previously Treated with DAA Simplified treatment regimenb All genotypes Sofosbuvir-velpatasvir × 12 weeks Glecaprevir-pibrentasvir × 8 weeks   PART 10 Disorders of the Gastrointestinal System Cirrhosis – compensatedc Genotypes 1–6 Glecaprevir-pibrentasvir × 8 weeks  Genotypes 1, 2, 4, 5, or 6 Sofosbuvir-velpatasvir × 12 weeks Cirrhosis – decompensated Ribavirin-eligible Genotypes 1–6 Sofosbuvir-velpatasvir with weight-based ribavirind × 12 weeks  Genotypes 1,4, 5, or 6 Ledipasvir-sofosbuvir with ribavirine × 12 weeks  Ribavirin-ineligible Genotypes 1–6 Sofosbuvir-velpatasvir × 24 weeks  Genotypes 1, 4, 5, or 6 Ledipasvir-sofosbuvir × 24 weeks  FEATURES ASSOCIATED WITH REDUCED RESPONSIVENESS TO DIRECT-ACTING ANTIVIRAL COMBINATION THERAPY Genotype and subtype (genotype 1a less responsive than genotype 1b for several drugs) Treatment experience Advanced fibrosis (bridging fibrosis, cirrhosis) Reduced adherence aRapidly evolving new recommendations continue to be issued; for up-to-date treatment recommendations, please see www.hcvguidelines.org. bExcludes: cirrhosis, prior treatment, pregnancy, hepatitis B surface antigen positive, hepatocellular carcinoma (HCC), or liver transplant. No adjustment required for end-stage renal disease. cChild-Pugh class A cirrhosis only; excludes: prior treatment, pregnancy, hepatitis B surface antigen positive, end-stage renal disease, HCC, or liver transplant. After treatment, ongoing surveillance for HCC is required. dFor Child-Pugh class C, initial low dose of ribavirin 600 mg is recommended with increase as tolerated. eInitial dose of ribavirin 600 mg with increase as tolerated. eNot recommended for prior NS3/4 protease inhibitor failure or genotype 3. fIn compensated cirrhosis, the addition of weight-based ribavirin is recommended. gConsider extending treatment to 24 weeks if high-risk features (e.g., genotype 3, cirrhosis, or failure following sofosbuvir-glecaprevir-pibrentasvir). Abbreviations: ALT, alanine aminotransferase; HCV, hepatitis C virus.

therapy (see below), conspired to temper enthusiasm for these new antivirals; after a brief stint as recommended therapy (2011–2013), these drugs became obsolete and are no longer recommended or available. DIRECT-ACTING ANTIVIRAL COMBINATIONS OF SECONDGENERATION PROTEASE INHIBITORS, FIRST-GENERATION POLYMERASE INHIBITORS, AND FIRST-GENERATION NS5A INHIBITORS (2014–2015) Since late 2013, the number of new antiviral agents for hepatitis C has expanded substantially, and currently, PEG IFN–based treatments have been supplanted by five remaining therapeutic regimens, which are all oral, IFN-free, highly efficacious (>95% SVR), and well toler­ ated, with high barriers to resistance, simple dosing, low pill burdens, treatment durations as brief as 8–12 weeks, and pangenotypic efficacy (Table 352-6). These drugs are distributed among three classes of DAAs: NS3/4 protease inhibitors (which cleave the single HCV poly­ protein into constituent structural and nonstructural proteins [drug name ending in “-previr”]), NS5B nucleoside and nonnucleoside polymerase inhibitors (which interfere with the RNA-dependent RNA polymerase [a replicase] involved in synthesis of viral RNA [drug name ending in “-buvir”]), and NS5A inhibitors (which interfere Sofosbuvir failure All genotypes Sofosbuvir-velpatasvir-voxilaprevir × 12 weeks Genotypes 1, 2, 4, 5, or 6 (alternative) Glecaprevir-pibrentasvir × 16 weekse Glecaprevir-pibrentasvir failure All genotypes Glecaprevir-pibrentasvir plus sofosbuvir and weightbased ribavirin × 16 weeks Sofosbuvir-velpatasvir-voxilaprevir × 12 weeksf  Multiple direct-acting antiviral failures All genotypes Glecaprevir-pibrentasvir plus sofosbuvir and weightbased ribavirin × 16 weeksg  Sofosbuvir-velpatasvir-voxilaprevir plus weight-based ribavirin × 24 weeks  Cirrhosis – decompensated All genotypes Sofosbuvir-velpatasvir plus weight-based ribavirin × 24 weeks  Genotypes 1, 4, 5, or 6 Ledipasvir-sofosbuvir plus ribavirine × 24 weeks DECOMPENSATED CIRRHOSIS Any protease-inhibitor-containing regimen is CONTRAINDICATED in decompensated cirrhosis (glecaprevir, grazoprevir, voxilaprevir).

with a membrane-associated phosphoprotein essential to the HCV RNA replication complex [drug name ending in “-asvir”]). The first of the new DAA agents (approved in November 2013) was simeprevir, a second-generation protease inhibitor for geno­ type 1, followed shortly thereafter (December 2013) by sofosbuvir, a pangenotypic nucleoside polymerase inhibitor. For genotype 1, both of these agents had to be combined with PEG IFN and ribavirin; for genotypes 2 and 3, sofosbuvir was administered with ribavirin, without PEG IFN; however, these treatment regimens have been supplanted by combinations of all-oral, IFN-free DAAs, and ribavirin is rarely needed and retained only for very limited indications. Simeprevir  When simeprevir was used with PEG IFN, its efficacy (genotype 1b > 1a) was similar to that of first-generation prote­ ase inhibitors but required only once-a-day dosing without the complexity of response-guided therapy. Similar to first-generation protease inhibitors, however, simeprevir was hampered by many drug-drug interactions, restriction to genotype 1 HCV, side effects, and limited efficacy in patients with a highly prevalent HCV NS3 polymorphism. Little about simeprevir supported its adoption in combination with PEG IFN and ribavirin. On the other hand, the combination of simeprevir (150 mg) along with sofosbuvir (400 mg) for 12 weeks was found to be effective in treatment-naïve (97% SVR12) or treatment-experienced (95% SVR12) patients without cir­ rhosis and in treatment-naïve (88% SVR12) or treatment-refractory (79% SVR12) patients with cirrhosis. While this all-oral, simeprevirsofosbuvir regimen was a landmark in the treatment of chronic hepatitis C, simeprevir was eclipsed by more effective DAAs and is no longer in clinical use. Sofosbuvir  Sofosbuvir, which inhibits the NS5B RNA-dependent RNA polymerase, was the first non–protease inhibitor DAA to be approved and had a paradigm-shifting excellent profile—high potency, high barrier to resistance, pangenotypic activity, very well tolerated with limited adverse effects (most commonly mild fatigue, insomnia, headache, and nausea), once-daily oral admin­ istration, and relative freedom from major drug-drug interactions. Sofosbuvir has efficacy in all genotypes (1–6); in treatment-naïve subjects and prior nonresponders to PEG IFN–based and proteaseinhibitor-based therapy; with PEG IFN–ribavirin or in IFN-free regimens; in combination with ribavirin or with NS5A inhibitors (see below); and for treatment periods as brief as 8–12 weeks. Cur­ rently, sofosbuvir is used in combination with one of two NS5A inhibitors and is a component of three of the five currently recom­ mended DAA regimens (Table 352-6). Paritaprevir-Ritonavir, Ombitasvir, and Dasabuvir  The combina­ tion of ritonavir (100 mg)–boosted paritaprevir (150 mg), a prote­ ase inhibitor; ombitasvir (25 mg), an NS5A inhibitor; and dasabuvir (250 mg), a nonnucleoside NS5B polymerase inhibitor, with or without weight-based ribavirin (total of five drugs), was approved in December 2014 for genotypes 1 and 4. In clinical trials, this com­ bination achieved SVR12 rates of 87–100% in treatment-naïve and treatment-experienced patients with genotype 1; without ribavirin, this combination in genotype 1a was ~7% less responsive than in genotype 1b. Therefore, in treatment-naïve patients with genotype 1a, this combination was administered with ribavirin for 12 weeks in the absence of cirrhosis (95–97% SVR12) or for 24 weeks in the presence of compensated cirrhosis (94% SVR12), whereas in patients with genotype 1b, the combination did not require ribavirin, and the duration of therapy was 12 weeks for both noncirrhotic and cir­ rhotic patients (99–100% SVR12). In July 2016, the FDA approved a long-acting formulation of dasabuvir, allowing once-a-day instead of twice-a-day treatment; for genotype 1a, twice-daily ribavirin dosing remained. This combination was well tolerated with generally mild side effects. Hyperbilirubinemia (primarily unconjugated) and eleva­ tions in ALT levels could occur but resolved during or shortly after treatment. Because of occasional hyperbilirubinemia and potential hepatotoxicity (FDA warning letter issued October 2015

regarding hepatic failure/decompensation reported in treated cir­ rhotic patients), this combination (and all subsequently introduced protease-inhibitor-containing combinations) was contraindicated in patients with decompensated cirrhosis, and treated cirrhotic patients had be monitored closely for decompensation.

Compared to superior regimens that came later (e.g., sofosbuvirledipasvir, see below), this regimen had the disadvantage of requir­ ing twice-a-day ribavirin therapy for genotype 1a and of being contraindicated in decompensated cirrhosis; however, it had the advantage of offering a 12-week, ribavirin-free regimen for prior null responders with cirrhosis and providing an option for patients with renal failure. Based on regimen simplicity and superior­ ity, subsequent-generation, ribavirin-free combination DAAs have supplanted paritaprevir-ritonavir, ombitasvir, and dasabuvir; this regimen is no longer recommended by the AASLD; however, it is retained in EASL recommendations as an alternative regimen for genotype 1b only. Sofosbuvir and Daclatasvir  Daclatasvir, an NS5A inhibitor, along with the polymerase inhibitor sofosbuvir, was approved by the FDA in July 2015 for genotype 3 and in February 2016 for genotype 1. At the time of its approval for genotype 3, daclatasvir filled a need inadequately met by other available combination DAAs; however, eventually, recommendation of this combination regimen was extended to genotypes 1–4 in the United States and to all genotypes (1–6) in Europe. Daclatasvir, a 60-mg tablet, and sofosbuvir, a sepa­ rate 400-mg tablet, were taken once a day for 12–24 weeks. In clinical trials among treatment-naïve or treatment-experienced patients, SVR12 rates for 12 weeks of daclatasvir plus sofosbuvir were 98% for genotype 1 (comparable results in genotypes 1a and 1b), 92% for genotype 2, and 89% for genotype 3. In patients with cirrhosis, SVR12 was achieved in 93% with Child-Pugh class A and B but in only 56% with class C decompensated cirrhosis. For patients with genotype 3 and cirrhosis, the combination was effective in treatment-naïve patients (94% SVR12) but less so in prior nonre­ sponders (69% SVR12). CHAPTER 352 Chronic Hepatitis Like other sofosbuvir–NS5A inhibitor combinations, daclatasvir plus sofosbuvir was well tolerated (mild fatigue, headache, nausea, or diarrhea in 5–14%) but could cause severe bradycardia when administered with amiodarone (contraindicated). Responsiveness to daclatasvir-containing drug combination therapy was reduced in cirrhotic patients with genotype 1a and in both cirrhotic and noncirrhotic patients with genotype 3 who had baseline daclatasvirassociated NS5A RASs. As new combination DAAs were introduced, however, dacla­ tasvir-sofosbuvir was less competitive and no longer filled a niche. This regimen has been supplanted by better, later-generation com­ bination DAAs and is no longer recommended. CURRENT FIRST-LINE COMBINATIONS OF DIRECT-ACTING ANTIVIRALS Sofosbuvir-Ledipasvir  The DAA combination that has had an early and dominant role in the treatment of hepatitis C is sofosbuvir (400 mg) plus the NS5A inhibitor ledipasvir (90 mg) in a once-aday, fixed-dose, single pill, approved in October 2014 for genotype 1 and in November 2015 for genotypes 4, 5, and 6. Phase 3 trials were conducted in treatment-naïve noncirrhotic patients, in treat­ ment-naïve cirrhotic and noncirrhotic patients, and in treatmentexperienced cirrhotic and noncirrhotic patients treated for 8, 12, or 24 weeks, both with and without ribavirin. In treatment-naïve noncirrhotics, an SVR12 was achieved in 97–99% of subjects, and no benefit was observed by extending therapy from 12 to 24 weeks or by adding ribavirin. Moreover, for treatment-naïve, noncirrhotic patients with baseline HCV RNA <6 × 106 IU/mL, a treatment dura­ tion of 8 weeks was as effective as one of 12 weeks (94–95% SVR12), which may be a consideration for a proportion of patients. In cir­ rhotic patients, SVR12 was achieved in 97–100% of treatment-naïve subjects (no advantage of extending therapy from 12 to 24 weeks or of adding ribavirin); however, for cirrhotic prior nonresponders to IFN-based therapy, 12 weeks of therapy was inferior (86% SVR12)

to 24 weeks of therapy (100% SVR12). This combination, which is equally effective in patients with HIV-HCV co-infection and in African-American patients, has been shown to be highly effective in patients with decompensated cirrhosis and in patients with hepa­ titis C after liver transplantation and after kidney transplantation. Initially, sofosbuvir-ledipasvir was not recommended in patients with advanced renal failure; however, subsequently, the safety and efficacy of sofosbuvir-ledipasvir in patients with advanced renal failure were established, and the combination was approved for this indication (November 2019). All sofosbuvir-containing regimens can be associated with severe bradycardia in patients taking the antiarrhythmic agent amiodarone, especially along with beta block­ ers; sofosbuvir-containing combinations are contraindicated with amiodarone. Drug-drug interactions are few, but P-glycoprotein inducers, such as St. John’s wort and rifampin, and proton pump gastric acid inhibitors, such as omeprazole, may reduce sofosbuvirledipasvir concentrations. Generally, responsiveness to sofosbuvirledipasvir is not reduced in patients with baseline RASs to these agents, with the exception of treatment-experienced patients who have baseline NS5A RASs (see Table 352-6).

Elbasvir-Grazoprevir  Elbasvir (50 mg), an NS5A inhibitor, com­ bined in a single, fixed-dose pill with grazoprevir (100 mg), an NS3/4 protease inhibitor, was approved in January 2016 as a oncea-day (with or without food) treatment for genotypes 1 and 4. In clinical trials, a 12-week course was effective in treatment-naïve and treatment-experienced patients without cirrhosis or with compensated cirrhosis. In treatment-naïve patients, this combina­ tion yielded an SVR12 in 92% of patients with genotype 1a, 99% with genotype 1b, and 100% with genotype 4 (very small numbers, however); 10 patients with genotype 6 were included, but only 80% achieved SVR12. Cirrhotic and noncirrhotic patients had compa­ rable rates of SVR12, 97% and 94%, respectively. For this drug com­ bination, however, ~11% of patients with genotype 1a harbor NS5A polymorphisms, which may include RASs, at baseline. If present, these NS5A RASs reduce efficacy of elbasvir-grazoprevir (unlike baseline RASs to most of the other combination DAA regimens described above and below) from 99% to 58% in treatment-naïve patients. Therefore, all patients with genotype 1a required base­ line RAS testing prior to starting this regimen; when these RASs were present, treatment extension to 16 weeks and the addition of weight-based ribavirin were documented to bring SVR12 rates up to expected levels of close to 100%. In treatment-experienced patients, both extending treatment to 16 weeks and adding ribavirin were studied; however, generally, in the absence of baseline NS5A RASs, SVR12 rates were not increased over those without ribavirin for 12 weeks (94–97%). Among nonresponders to prior protease inhib­ itor therapy, even in the absence of baseline NS5A RASs, ribavirin had to be added to a 12-week regimen; in the presence of baseline NS5A RASs, treatment was extended to 16 weeks and ribavirin added. For genotype 1b, NS5A RASs are not an issue, and the only subgroup requiring modification of a 12-week course of therapy were prior nonresponders to protease inhibitor regimens, for whom ribavirin was added. For genotype 4, the recommended regimen for all prior nonresponders (whether to PEG IFN–ribavirin or protease inhibitor regimens) was 16 weeks of elbasvir-grazoprevir plus ribavirin. Now that simpler, improved combination regimens are available, for patients with NS5A RASs, extending the duration of elbasvir-grazoprevir and adding ribavirin have been abandoned (Table 352-6); however, elbasvir-grazoprevir is one of the currently recommended DAA combinations (Table 352-6). PART 10 Disorders of the Gastrointestinal System This combination is just as effective in patients with HIVHCV co-infection and in patients with advanced renal failure (including those requiring hemodialysis), but like all proteaseinhibitor-including DAA combinations, it is contraindicated in decompensated cirrhosis. In this vein, like other protease inhibitor regimens, elbasvir-grazoprevir can be associated with aminotrans­ ferase elevations and potential hepatotoxicity; because these drugs are excreted by the liver, plasma drug concentrations may become

elevated substantially in the presence of impaired hepatic function. Therefore, all treated patients should have ALT screening periodi­ cally during therapy, and the drug should be stopped for elevations exceeding 10-fold or for elevations of conjugated bilirubin, alkaline phosphatase, or prothrombin time. Elbasvir-grazoprevir is well tolerated, with only low levels of mild adverse effects (fatigue, headache, or nausea in 5–11%) seen just as frequently in placebo recipients. Both elbasvir and grazopre­ vir are substrates for CYP3A and are subject to multiple potential drug-drug interactions. Therefore, this combination should not be used with potent CYP3A inducers; conversely, CYP3A and OATP1B1 inhibitors can lead to untoward elevations of plasma elbasvir-grazoprevir concentrations. Checking for potential drugdrug interactions is advisable prior to initiating therapy (www

.hep-druginteractions.org). Compared to other available regimens for genotypes 1 and 4, elbasvir-grazoprevir has the disadvantage/inconvenience of requir­ ing baseline NS5A RAS testing but the advantages of a comparable regimen for cirrhotic and noncirrhotic patients, for treatment-naïve and treatment-experienced patients, and for patients with normal renal function and with renal failure. As noted above, its limitations and complexities notwithstanding, elbasvir-grazoprevir remains one of the recommended DAA combinations (albeit not one of the “simplified” treatment algorithms, see below). Sofosbuvir-Velpatasvir  The combination in a single, fixed-dose pill of velpatasvir (100 mg), a highly potent, pangenotypic NS5A inhibitor, and the polymerase inhibitor sofosbuvir (400 mg) was approved in June 2016 for genotypes 1–6 in treatment-naïve and treatment-experienced noncirrhotic and cirrhotic patients. In August 2017, approval was extended to include patients with HCVHIV co-infection. Ribavirin is not required, including in patients with genotypes 2 and 3, except in patients with decompensated cirrhosis. In a series of clinical trials, this combination for 12 weeks in the absence of ribavirin was shown to yield a 99% SVR12 (range 97–100%) in genotypes 1, 2, 4, 5, and 6 and 95% in genotype 3. Baseline NS5A RASs had no impact on responsiveness. Prior to the availability of this drug combination, patients with genotype 3, especially those with cirrhosis and prior null response to other therapies, proved to be the most refractory subset of patients. In treatment-naïve patients with genotype 3, 12 weeks of sofosbuvir-velpatasvir (95% SVR12) was superior to 24 weeks of sofosbuvir plus ribavirin (80% SVR12). In patients with genotype 3, the combination of sofosbuvir-velpatasvir for 12 weeks was comparable in noncirrhotic (97% SVR12) and cirrhotic patients (91% SVR12) and in treatment-naïve (97% SVR12) and treatmentexperienced (90% SVR12) patients and was superior in all these categories to 24 weeks of sofosbuvir plus ribavirin (87, 66, 86, and 63%, respectively). In cirrhotic null responder patients, most avail­ able IFN-free regimens for genotype 3 (including daclatasvir plus sofosbuvir, which had been approved specifically for this genotype) achieved SVR12 rates in the range of ~60–75%, while the combina­ tion of PEG IFN, ribavirin, and sofosbuvir could boost SVR12 to the mid-80% range. For treatment-experienced patients with genotype 3, sofosbuvir-velpatasvir in noncirrhotic and cirrhotic patients had similarly high efficacy (91 and 89% SVR12, respectively); this was the highest recorded SVR12 for genotype 3 cirrhotic null respond­ ers treated with IFN-free DAA regimens. Finally, in patients with genotypes 1–4 and 6 and with decompensated, class B cirrhosis (55% treatment-experienced), sofosbuvir-velpatasvir plus ribavirin for 12 weeks yielded an SVR12 in 94%; this result was better than sofosbuvir-velpatasvir without ribavirin for 12 weeks (83% SVR12) or 24 weeks (86% SVR12). Like other all-oral DAAs, sofosbuvir-velpatasvir was very well tolerated; in noncirrhotic and compensated cirrhotic patients, mild headache and fatigue were seen in >10% (this occurred in a compara­ ble proportion of placebo recipients), and in patients with decompen­ sated cirrhosis, mild fatigue, headache, nausea, insomnia, diarrhea,

and anemia (ribavirin was part of the regimen) were seen in >10%. Like other sofosbuvir-containing regimens, sofosbuvir-velpatasvir should not be administered along with amiodarone (potential serious bradycardia); in addition, P-glycoprotein inducers and moderate-topotent CYP3A inducers can reduce plasma levels of sofosbuvir and/ or velpatasvir. Checking for drug-drug interactions prior to therapy is advisable (www.hep-druginteractions.org). Baseline RASs do not influence responsiveness to this combination. Sofosbuvir-velpatasvir is one of the currently recommended DAA combinations for hepatitis C (Table 352-6). Because it is so simple and broadly effective across patient subgroups, sofosbuvir-velpatasvir is one of the two combination DAA regimens recommended by the AASLD and EASL as a preferred, simplified treatment algorithm (Table 352-6). Sofosbuvir-Velpatasvir-Voxilaprevir  Approved in July 2017, the pangenotypic, high-barrier-to-resistance protease inhibitor voxila­ previr (100 mg) added to the polymerase inhibitor–NS5A inhibitor combination of sofosbuvir-velpatasvir yields a very well tolerated triple-drug combination with ~97% SVR12 across all HCV geno­ types and patient subgroups. These include the small percentage of patients with genotype 1 and genotype 3 refractory to previously approved DAA combinations as well as noncirrhotic/cirrhotic, treatment-naïve/treatment-experienced groups, including those who had or who had not received prior NS5A treatment. Efficacy was independent of the number of prior DAA drug classes received, and no effects of baseline NS5A RASs were noted. The potential for abbreviated (8-week) treatment with this triple combination was explored in a clinical trial involving treatmentnaïve patients; however, the shortened duration was inferior to a full 12-week course. The side effect profile for sofosbuvir-velpatasvirvoxilaprevir was similar to that in the placebo arm of clinical trial patients and included mild and uncommon headache, fatigue, nausea, and diarrhea. Because other DAA regimens are so effective in most patients with chronic hepatitis C, recommendations for sofosbuvir-velpatasvirvoxilaprevir are limited to a small subset of otherwise refractory patients: for treatment-naïve cirrhotic patients with genotype 3 and baseline NS5A velpatasvir RAS Y93H, for treatment-naïve (according to AASLD, not EASL) or IFN-ribavirin–experienced noncirrhotic or cirrhotic patients with genotype 3 (Table 352-6), and for patients with or without compensated cirrhosis and prior, failed NS5A inhibitor–containing therapy (consult www. hcvguidelines.org). This triple-drug combination, like all sofosbuvir-containing combinations, is contraindicated in patients taking amiodarone and, like all protease inhibitor–containing combinations, in patients with decompensated cirrhosis. Concomitant omeprazole, 20 mg, can be taken with this sofosbuvir-containing regimen. Prior to initi­ ating therapy, checking for drug-drug interactions is recommended. Glecaprevir-Pibrentasvir  A regimen of 8 weeks of this single-pill, fixed-dose combination of the protease inhibitor glecaprevir (300 mg) and NS5A inhibitor pibrentasvir (120 mg), two pangenotypic, highpotency DAAs with high barriers to resistance (approved in August 2017), achieves SVR12 in close to 100% of treatment-naïve patients with all genotypes, with or without cirrhosis: SVR12 of ~99% for genotypes 1, 2, and 4–6 and of 95–98% for genotype 3. Extended treatment for 12 weeks did not increase efficacy. In trials among treatment-experienced patients, treatment with 12 weeks of this DAA combination was just as effective as 16 weeks for all geno­ types except genotype 3; however, with increasing numbers of prior treatment courses, SVR12 rates fell—100% for patients treated with a protease inhibitor only, 88% for patients treated with an NS5A inhibitor only, and 79% for patients treated previously with both a protease inhibitor and an NS5A inhibitor. Similarly, baseline RASs reduced SVR12 rates—from 100% without RASs (or with RASs limited to those reflecting protease inhibitor resistance) to 89% for baseline NS5A RASs. For retreatment of patients with prior glecaprevir-pibrentasvir failure, 16 weeks of glecaprevir-pibrentasvir plus sofosbuvir are

recommended (alternatively, sofosbuvir-velpatasvir-voxilaprevir for 12 weeks [plus ribavirin in cirrhotic patients]). Glecaprevirpibrentasvir for 16 weeks is recommended as well after failure to respond to the triple-drug combination of sofosbuvir-velpatasvirvoxilaprevir (see below). For retreatment of patients with sofosbuvir-velpatasvir-voxilaprevir failure, 16 weeks of glecap­ revir-pibrentasvir plus ribavirin is recommended, as is a repeat course of sofosbuvir-velpatasvir-voxilaprevir plus ribavirin for 24 weeks. For navigating the complexities of treating DAArefractory patients, checking up-to-date guidelines (www.hcvguid­ lines.org) and referral to a specialized center are advisable.

As is the case for any DAA combination containing a protease inhibitor, glecaprevir-pibrentasvir is contraindicated in decom­ pensated cirrhosis; it has been shown to achieve an SVR12 in 98% of patients with stage 4 or 5 renal disease (in treatment-naïve or experienced, cirrhotic or noncirrhotic patients) and is a preferred treatment for patients with severe renal impairment. This DAA combination should be taken with food, and drug-drug interactions should be considered prior to initiating treatment. Because it is so simple and broadly effective across patient subgroups (8 weeks for all noncirrhotic treatment-naïve patients [including with HIV co-infection]; 12 weeks for all treatment-experienced cirrhotics [including with HIV co-infection] and treatment-naïve cirrhotics with genotype 3 [except treatment-experienced cirrhotic or noncirrhotic genotype 3 {16 weeks}]), glecaprevir-pibrentasvir is one of the two combination DAA regimens recommended by the AASLD and EASL as a preferred, simplified treatment algorithm (Table 352-6). Emerging data on the impact of DAAs on the natural history of chronic hepatitis C indicated that, as was documented for IFNbased therapy, successful DAA therapy is associated with a gradual reduction in fibrosis progression and a regression of advanced fibrosis (cirrhosis), improvement in survival among patients with decompensated cirrhosis, a reduction in HCC, and a decline in the number of patients with hepatitis C being referred for liver trans­ plantation. Indeed, early and broad implementation of DAAs in the treatment of chronic hepatitis C has been instrumental in enabling the decline of HCV-related deaths since 2013. Modeling estimates suggest that application of DAA therapies has the potential to result in a 50–70% reduction in the hepatitis C–associated disease burden by 2050. CHAPTER 352 Chronic Hepatitis TREATMENT RECOMMENDATIONS Because the pace of new drug development and approval has been so rapid, the AASLD and the Infectious Diseases Society of America (IDSA) have been providing a consensus of updated treatment recommendations for patients with hepatitis C; these recommenda­ tions, which continue to be revised regularly based on new data, are available online at www.hcvguidelines.org and should be consulted before initiating therapy. The EASL issues similar (but not identical) treatment recommendations annually for hepatitis C (www.easl.eu), most recently in November 2020. Prior to therapy, all patients should undergo pretreatment evalu­ ation, which includes HCV RNA, HCV genotype, renal function, hepatitis B status (testing should include HBsAg and anti-HBc), HIV status, and fibrosis stage. Despite the presence of pangenotypic regimens, HCV genotype should still be determined, because the genotype may contribute to decisions about which treatment regi­ mens are indicated and can guide insurance coverage (Table 352-6). Fibrosis assessment is an especially important element of the pretreatment evaluation: the absence or presence of cirrhosis or advanced fibrosis determines the treatment options from which to select, including the antiviral agents to be used, the duration of ther­ apy, and the now rare need for ribavirin (Table 352-6). In the past, a pretreatment liver biopsy was relied upon to assess histologic grade and stage as well as to identify such histologic factors as steatosis, which can influence adversely responsiveness to therapy. As therapy has improved for patients with a broad range of histologic severity, and as noninvasive measures of the stage of fibrosis (e.g., assess­ ment of liver stiffness by imaging, FIB-4 score [see above]) have

advanced in accuracy and popularity, noninvasive approaches have supplanted histology in almost most cases. As noted above, if cir­ rhosis or advanced fibrosis is present prior to therapy, the risk of HCC, although reduced substantially by successful therapy, is not eliminated, and twice-yearly posttreatment imaging for HCC surveillance remains indicated even after an SVR. In patients with low-level fibrosis at baseline, achievement of SVR allows cessation of such surveillance.

Several reports have appeared describing hepatitis B reactivation, often severe, during and after DAA therapy in patients co-infected with HCV and HBV who were not being treated for their HBV infections. Therefore, screening for HBV infection is recommended prior to initiating DAA therapy for hepatitis C (which should have been done to determine HBV immunity status as a prelude to recommended hepatitis B vaccination in patients with chronic hepatitis C), and therapy for HBV infection (for those meeting HBV treatment criteria, see above) should be initiated prior to or simultaneously with HCV therapy. Monitoring of serum HCV RNA levels before, during, and after treatment is crucial in assessing the response to therapy; moreover, the baseline level may contribute to determining the duration of therapy (e.g., in noncirrhotic patients with genotype 1 and HCV RNA <6 × 106 IU/mL, 8 [instead of the usual 12] weeks of sofosbuvir-ledi­ pasvir may be a consideration). The goal of treatment is to eradicate HCV RNA during therapy and to document that the virus remains undetectable for at least 12 weeks after completion of therapy (SVR12). Most recent guidelines support a simplified treatment algorithm for patients with HCV infection, leveraging the well tolerated and highly effective pangenotypic regimens. Patients who are eligible for this simplified algorithm must not have any of the following clinical characteristics: prior treatment failure, cirrhosis, presence of HBsAg, pregnancy, HCC, or history of liver transplantation. In the absence of any of these features, the patient may be treated with either sofosbuvir/velpatasvir for 12 weeks or glecaprevir/pibrentas­ vir for 8 weeks. PART 10 Disorders of the Gastrointestinal System INDICATIONS FOR ANTIVIRAL THERAPY Patients with chronic hepatitis C who have detectable HCV RNA in serum, whether or not aminotransferase levels are increased, and chronic hepatitis of any grade and stage are candidates for antiviral therapy with DAA agents. The only exception would be patients with short life expectancies, for whom treating hepatitis C

would have no influence on longevity. Certainly, for patients with advanced liver disease, early treatment merits a high priority. Although patients with persistently normal aminotransferase activ­ ity tend to progress histologically very slowly or not at all, they respond to antiviral therapy just as well as do patients with elevated aminotransferase levels; therefore, such patients are candidates for antiviral therapy. As noted above, antiviral therapy has been shown to improve survival and complication-free survival and to slow progression of and to reverse fibrosis. Patients who have relapsed after, or failed to respond to, a course of IFN-based (currently no longer encountered commonly) or DAA agent–based therapy are candidates for retreatment with a DAA therapy regimen (Table 352-6). For patients who have failed to respond to a DAA combination, options include increasing the duration of therapy with the failed regimen, adding ribavirin, or changing the drug class (e.g., after failed protease and polymerase inhibitors, switching to an NS5A-containing combination). In the presence of cirrhosis or a need for urgent retreatment, patients who have failed protease inhibitor plus polymerase inhibitor com­ bination therapy or who have failed an NS5A combination are candidates for RAS testing and tailored therapy based on such resistance testing. If reliable RAS testing is not available, adding ribavirin or extending the duration of therapy are options. For prior nonresponders to IFN-based therapy, NS5A inhibitor–containing regimens are highly effective; however, reduced responsiveness can be encountered, especially in cirrhotic patients. The simplest current approach for patients who have not responded to prior

DAA therapy containing sofosbuvir is to re-treat with sofosbuvirvelpatasvir-voxilaprevir for 12 weeks or glecaprevir-pibrentasvir for 16 weeks. Additional details for treatment of such patient subgroups can be found at www.hcvguidelines.org; in the small proportion of patients in whom initial DAA courses do not achieve SVR12, referral to an experienced treatment center is advisable. Worth reiterating, protease inhibitors are contraindicated for patients with decompensated cirrhosis, and sofosbuvir-containing regimens are not recommended for patients taking amiodarone (especially with beta blockers) for treatment of cardiac arrhythmias. While sofosbuvir-containing DAA combinations were not recommended initially for patients with advanced renal failure, subsequent studies demonstrated safety and efficacy in this subgroup, and sofosbuvircontaining DAA combinations are now approved for advanced renal failure. Persons with acute hepatitis C are also candidates for antiviral therapy (Chap. 350) with the same pangenotypic combination DAA agents and the same duration of treatment approved for chronic hepatitis C. For example, in a trial of a shortened, 6-week course of sofosbuvir-velpatasvir for acute hepatitis C, a 6-week course was less effective than a full 12-week course (81.7 vs 90.5%, respectively). Delaying the initiation of therapy in acute hepatitis C to allow for spontaneous recovery is no longer recommended. While acute hepatitis C should be treated, postexposure prophylaxis after HCV exposure is not recommended currently In patients with biochemically and histologically mild chronic hepatitis C, the rate of progression is slow; however, such patients respond just as well to antiviral therapy as those with elevated aminotransferase levels and more histologically severe hepatitis. Because of the high cost of DAA treatments, in the past, a higher priority was assigned to patients with advanced fibrosis/cirrhosis; however, this controversial approach was relied upon by some medical insurers and pharmacy benefit management organizations to withhold therapy from patients with low-level fibrosis. Unfor­ tunately, delaying therapy until fibrosis becomes advanced misses the opportunity to prevent all the dire consequences of chronic hepatitis C (liver failure, death/transplantation, HCC), which can be reduced, but not eliminated completely, once advanced fibrosis is established. Therefore, therapy for patients with mild disease is justified as well as cost-effective. TREATMENT OF UNIQUE POPULATIONS Decompensated Cirrhosis  Patients with compensated cirrhosis can respond to therapy, and their likelihood of a sustained response with DAAs is comparable to that in the absence of cirrhosis. Patients with decompensated cirrhosis, who were not candidates for IFN-based antiviral therapy, respond well to non–proteaseinhibitor-containing regimens, and specifically sofosbuvir com­ bined with an NS5A inhibitor, velpatasvir or ledipasvir, with or without ribavirin. As noted above, protease-inhibitor-containing combinations have been associated with potential hepatotoxicity and hepatic decompensation and are contraindicated in this patient subset. Treatment decisions are informed by whether the patient is eligible for ribavirin: in ribavirin-eligible patients, ribavirin is added to a 12-week course of sofosbuvir-NS5A therapy; however, in cases of ribavirin ineligibility, the duration of therapy should be extended to 24 weeks. In cases of prior failure to respond to sofosbuvirNS5A therapy, the sofosbuvir-NS5A regimen should be repeated but supplemented with ribavirin and extended to 24 weeks (www. hcvguidelines.org). Also, patients with decompensated cirrhosis should be referred to a liver transplantation center. DAAs are highly effective not only for patients with end-stage liver disease awaiting liver transplantation but also for patients with recurrent hepatitis C after liver transplantation. Ideally, patients should be treated prior to liver transplantation. Given the effec­ tiveness of treatment immediately after transplantation (either for donor or recipient HCV infection), however, the utility of treat­ ment may be limited if the patient has a high Model for EndStage Liver Disease (MELD) score and may soon receive a donor

liver for transplantation. Therefore, advocacy has been expressed for postponing DAA therapy in patients with high-MELD-score (≥18–20), HCV-associated, end-stage liver disease until after liver transplantation; for patients with MELD scores <18–20, pretrans­ plantation DAA therapy is advised. Still, the decision whether to treat pretransplantation or posttransplantation should be individu­ alized thoughtfully for each patient, based on such factors as MELD score, time anticipated prior to availability of a donor organ, rela­ tive clinical stability, and comorbidities (Chap. 356). Because DAA therapy is so effective, most transplantation centers are accepting organs from HCV-infected donors, transplanting them into HCV-uninfected recipients, and treating recipients with sofosbuvirvelpatasvir for 12 weeks or glecaprevir-pibrentasvir for 8 weeks after transplantation—with excellent results. Extrahepatic Manifestations  The cutaneous and renal vasculitis of HCV-associated essential mixed cryoglobulinemia (Chap. 350) may respond to antiviral therapy, but sustained responses were rare after discontinuation of therapy in the IFN era, and prolonged, potentially indefinite, therapy was recommended. Now that more effective DAAs are available, a 12-week course of sofosbuvir-based combination therapy has been shown to yield an SVR12 rate exceed­ ing 80% in cryoglobulinemic vasculitis. Anecdotal reports suggest that IFN-based antiviral therapy may be effective in porphyria cutanea tarda or lichen planus associated with hepatitis C; whether the more appealing DAAs are effective in these groups remains to be documented. HIV Co-infection  In patients with HCV/HIV co-infection, hepa­ titis C is more progressive and severe than in HCV mono-infected patients. Although patients with HCV/HIV co-infection responded less well to IFN-based antiviral therapy for hepatitis C, they respond as well as patients with HCV infection alone to DAA combination regimens and, generally, should be treated with the same DAA combinations and for the same duration as those for patients (in the absence or presence of cirrhosis) with HCV mono-infection. For patients with HCV/HIV co-infection, an abbreviated, 8-week course of sofosbuvir-ledipasvir for low-level HCV RNA is not recommended, and a full 12 weeks should be given; similarly, for patients with genotype 4, a 12-week course of glecaprevir-pibren­ tasvir is recommended instead of an 8-week course for treatmentnaïve or -experienced patients with or without cirrhosis (Table 352-6). In HCV/HIV-infected patients, ribavirin can potentiate the toxicity of didanosine (e.g., lactic acidosis) and the lipoatrophy of stavu­ dine, and zidovudine can exacerbate ribavirin-associated hemolytic anemia; therefore, these drug combinations should be avoided. In HCV/HIV co-infected persons, the list of potential drug-drug interactions is extensive and should be consulted carefully before beginning DAA treatment (www.hcvguidelines.org). Substance Use Disorder  Patients with a history of injection drug use and alcoholism can be treated successfully for chronic hepati­ tis C, preferably in conjunction with drug and alcohol treatment programs. Moreover, because persons who use injection drugs (PWID), as a source of transmission to others, account dispro­ portionately for perpetuating the spread of HCV infection in the population, the impact of treating people who are actively using injection drugs is amplified by reducing such transmission. In a randomized, controlled clinical trial, treatment with sofosbuvir and velpatasvir for 12 weeks was highly successful in PWID, with SVR12 rates of 74–76%. Based on this study, and others, active injection drug use is no longer considered a contraindication to treatment, and indeed, mechanisms to facilitate care of PWID are being sought to address the U.S. and global burden of HCV infection and move toward the World Health Organization goal of HCV elimination by 2030 (see below). Renal Impairment  The approved oral DAA combinations are effective in patients with mild-modest renal impairment and require no dose adjustments. For patients with severe renal impairment (creatinine clearances <30 mL/min), including those undergoing

hemodialysis, recommended combinations are 12 weeks of elbas­ vir-grazoprevir for genotypes 1 and 4 or 12 weeks of glecaprevirpibrentasvir for all genotypes. Both in severe renal impairment and after kidney transplantation, levels of SVR12 in patients treated with these oral DAA combinations have approached 100%. Initially, in patients with severe renal impairment, sofosbuvir-containing com­ binations were not recommended. Subsequently, however, based on efficacy and safety in a series of clinical trials, sofosbuvir-containing regimens were approved by the FDA in November 2019 for patients with severe renal impairment.

Pregnancy  No large clinical studies of the use of DAAs during pregnancy are available. One small study demonstrated the safety and efficacy of sofosbuvir and ledipasvir given for 12 weeks at week 23–24 of gestation. Ribavirin is contraindicated during preg­ nancy; therefore, any regimen including ribavirin should not be used. Sofosbuvir; sofosbuvir-ledipasvir; and paritaprevir-ritonavir, ombitasvir, and dasabuvir are classified as pregnancy category B; the other DAAs do not have a pregnancy classification. Therefore, these therapies are not indicated routinely in pregnancy and should be used, with caution, only if the benefit of treatment is compelling and justified compared to the potential for fetal risk. Currently, screening all pregnant women for HCV infection is recommended. Breast feeding is not contraindicated in women with HCV infection (unless the mother has a break in the integrity of the nipples or is co-infected with HIV). CHOOSING AMONG AVAILABLE TREATMENT OPTIONS Choosing among the number of all-oral DAA combinations approved since 2013 was daunting to treating clinicians. Cur­ rently, however, the number of recommended DAA combina­ tions has narrowed to a very manageable few. The most popular of the regimens have been fixed-dose, single-pill, pangenotypic combinations. For simplicity, two “one-size-fits-all” pangenotypic regimens—sofosbuvir-velpatasvir and glecaprevir-pibrentasvir— can be used, for 8–12 weeks in almost all treatment-naïve, noncir­ rhotic and cirrhotic patients, including those with advanced renal failure and HCV-HIV co-infection. Applicability of the triple-drug combination sofosbuvir-velpatasvir-voxilaprevir is quite limited in treatment-naïve patients, reserved primarily for cirrhotic patients with genotype 3. As noted above, protease-inhibitor-containing DAA regimens (elbasvir-grazoprevir, glecaprevir-pibrentasvir, and sofosbuvir-velpatasvir-voxilaprevir) are contraindicated in decom­ pensated cirrhosis. CHAPTER 352 Chronic Hepatitis APPROACHES TO STEAMLINING DAA TREATMENT The World Health Organization has set an elimination goal of a 90% reduction in the global prevalence of HCV infection by 2030, and the Viral Hepatitis National Strategic Plan includes a goal of an 80% reduction in the U.S. prevalence of HCV infection by 2030; however, at the current pace of identification and treatment, this ambitious goal is unlikely to be met. To address this shortfall, efforts have been devoted to improving the frequency of screening and diagnosis and to streamlining the process of enrolling persons with HCV infection into treatment programs. Now that DAA regimens have been simplified so dramatically, in-person specialist care is usually not needed (except in complex patients such as those with advanced fibrosis, comorbid conditions, and prior nonresponse to DAA treatment). Programs with an impressive record of success in streamlining the treatment of persons with hepatitis C include those based on telehealth, one-stop screening and treatment initia­ tion, minimal monitoring (dispensing the full course of pills at the beginning of therapy, relying on a limited number of in-person or even all-remote patient contacts during therapy, and involving no HCV RNA or liver test monitoring during treatment and not until 12 weeks after completion of treatment), and collaborative participation on the treatment team of clinicians experienced in managing substance use disorder. The expansion of such programs is anticipated eagerly.

AUTOIMMUNE HEPATITIS

■ ■DEFINITION Autoimmune hepatitis (AIH) is a chronic disorder characterized by continuing hepatocellular necrosis and inflammation, usually with fibrosis, which can progress to cirrhosis and liver failure. The presenta­ tion of AIH varies, from asymptomatic liver enzyme abnormalities to severe or even, in its most severe form, marked hepatocellular drop­ out, multilobular collapse, and clinical features of fulminant hepatitis with high mortality. Based on contemporary estimates of the natural history of AIH, the 10-year survival is 80−98% for treated and 67% for untreated patients. The prominence of extrahepatic features of autoimmunity and seroimmunologic abnormalities in this disorder supports an autoimmune process in its pathogenesis; this concept is reflected in the prior labels lupoid and plasma cell hepatitis. Similar to other autoimmune diseases, AIH has a strong female predominance, with women accounting for ~75% of all cases. The age distribution is bimodal, with one peak in the early decades (age <30) and the second in middle to late years (often >60). Autoantibodies and other typical features of autoimmunity, however, do not occur in all cases; among the broader categories of “idiopathic” or cryptogenic chronic hepati­ tis, many, perhaps the majority, are probably autoimmune in origin. Cases in which hepatotropic viruses, metabolic/genetic derangements (including nonalcoholic fatty liver disease), and hepatotoxic drugs have been excluded represent a spectrum of heterogeneous liver disorders of unknown cause, a proportion of which are most likely AIH. ■ ■IMMUNOPATHOGENESIS The weight of evidence suggests that the progressive liver injury in patients with AIH is the result of a cell-mediated immunologic attack directed against liver cells in the setting of a loss of, or failed, immunologic tolerance for self-liver antigens. In all likelihood, pre­ disposition to autoimmunity is inherited, whereas the liver specific­ ity of this injury is triggered by environmental (e.g., chemical, drug [e.g., minocycline], or viral) factors. For example, patients have been described in whom apparently self-limited cases of acute hepatitis A, B, or C led to AIH, presumably because of genetic susceptibility or predisposition. More recently, the emergence of AIH has been reported following—presumably triggered by—COVID-19 and after COVID-19 vaccination, postulated mechanistically to result from the vigorous immune response and loss of immune tolerance in suscep­ tible persons. PART 10 Disorders of the Gastrointestinal System Evidence to support an autoimmune pathogenesis in this type of hepatitis includes the following: (1) in the liver, the histopathologic lesions are composed predominantly of cytotoxic T cells and plasma cells; (2) circulating autoantibodies (nuclear, smooth muscle, and others, see below), rheumatoid factor, and hyperglobulinemia are common; (3) other autoimmune disorders—such as autoimmune thyroiditis, rheumatoid arthritis, autoimmune hemolytic anemia, ulcerative colitis, membranoproliferative glomerulonephritis, juvenile diabetes mellitus, vitiligo, celiac disease, and Sjögren’s syndrome— occur with increased frequency in patients who have AIH and in their relatives; (4) histocompatibility haplotypes associated with autoim­ mune diseases, such as HLA-B1, B8, DR3, and DR4 as well as extended haplotype DRB10301 and DRB10401 alleles, are common in patients with AIH; and (5) this type of chronic hepatitis is responsive to gluco­ corticoid/immunosuppressive therapy, effective in a variety of autoim­ mune disorders. Loss of immune tolerance as well as cytotoxic cellular immune responses appear to be important in the pathogenesis of AIH. In vitro studies have suggested that in patients with this disorder, CD4+ T lymphocytes are capable of becoming sensitized to hepatocyte mem­ brane proteins and of destroying liver cells. Molecular mimicry by cross-reacting antigens that contain epitopes similar to liver antigens is postulated to activate these T cells, which infiltrate, and result in injury to, the liver. Abnormalities of immunoregulatory control over cytotoxic lymphocytes (impaired regulatory CD4+CD25+ T-cell influ­ ences) may play a role as well. Studies of genetic predisposition to AIH demonstrate that certain haplotypes are associated with the disorder,

as enumerated above, as are polymorphisms in cytotoxic T lymphocyte antigens (CTLA-4) and tumor necrosis factor α (TNFA*2). The precise triggering factors, genetic influences, and cytotoxic and immunoregu­ latory mechanisms involved in this type of liver injury remain incom­ pletely defined. Intriguing clues into the pathogenesis of AIH come from the obser­ vation that circulating autoantibodies are prevalent in patients with this disorder. Among the autoantibodies described in these patients are antibodies to nuclei (so-called antinuclear antibodies [ANAs], primarily in a homogeneous pattern) and smooth muscle (so-called anti-smooth-muscle antibodies [ASMAs], directed at actin, vimen­ tin, and skeletin), antibodies to F-actin and anti-LKM (see below), antibodies to soluble liver antigen (directed against a uracil-guanineadenine transfer RNA suppressor protein), antibodies to α-actinin, and antibodies to the liver-specific asialoglycoprotein receptor (or “hepatic lectin”) and other hepatocyte membrane proteins. Although some of these provide helpful diagnostic markers, their involvement in the pathogenesis of AIH has not been established. Humoral immune mechanisms have been shown to play a role in the extrahepatic manifestations of autoimmune and idiopathic hepatitis. Arthralgias, arthritis, cutaneous vasculitis, and glomerulo­ nephritis occurring in patients with AIH appear to be mediated by the deposition of circulating immune complexes in affected tissue vessels, followed by complement activation, inflammation, and tissue injury. While specific viral antigen-antibody complexes can be identified in acute and chronic viral hepatitis, the nature of the immune complexes in AIH has not been defined. ■ ■CLINICAL FEATURES Many of the clinical features of AIH are similar to those described for chronic viral hepatitis. The onset of disease may be insidious or abrupt; the disease may present initially like, and be confused with, acute viral hepatitis. The most common symptom, occurring in ~85% of patients, is fatigue. In severe disease, which accounts for 20–25% of cases, jaundice may be present as well. In approximately a quarter of patients, the diagnosis is made in the absence of symptoms, based on abnormal liver laboratory tests. A subset of patients with AIH has distinct features. Such patients are predominantly young to middleaged women with marked hyperglobulinemia and high-titer circulat­ ing ANAs. This is the group with positive lupus erythematosus (LE) preparations (initially labeled lupoid hepatitis) in whom other autoim­ mune features are common. Fatigue, malaise, anorexia, amenorrhea, acne, arthralgias, and jaundice are common. Occasionally, arthritis, maculopapular eruptions (including cutaneous vasculitis), erythema nodosum, colitis, pleuritis, pericarditis, anemia, azotemia, and sicca syndrome (keratoconjunctivitis, xerostomia) occur. In some patients, complications of cirrhosis, such as ascites and edema (associated with portal hypertension and hypoalbuminemia), encephalopathy, hyper­ splenism, coagulopathy, or variceal bleeding may bring the patient to initial medical attention. The course of AIH may be variable. In patients with mild dis­ ease or limited histologic lesions (e.g., interface hepatitis, labeled in the past as piecemeal necrosis [inflammation and erosion of the limiting place of periportal hepatocytes] without bridging necrosis and fibrosis), progression to cirrhosis may be limited, but, even in this subset, treatment and careful monitoring remain warranted, because, if untreated, up to half can progress to cirrhosis over 15 years. Also, asymptomatic patients, if left untreated, have a lower 10-year overall survival. In North America, cirrhosis at presentation is more common in African Americans than in whites. In those with severe symptomatic AIH (aminotransferase levels >10 times normal, marked hyperglobulinemia, “aggressive” histologic lesions—bridging necrosis or multilobular collapse, cirrhosis), the 6-month mortality without therapy may be as high as 40%. Such severe disease accounts for only 20% of cases; the natural history of milder disease is variable, often accentuated by spontaneous remissions and exacerbations. In a 10-year (2006–2016) national Dutch study, mortality in patients with AIH was higher than that of the general population only in patients with cirrhosis; for patients without cirrhosis, survival was

comparable to that of the general population. Especially poor prog­ nostic signs include the presence histologically of multilobular col­ lapse at the time of initial presentation and failure of serum bilirubin to improve after 2 weeks of therapy. Death may result from hepatic failure, hepatic coma, other complications of cirrhosis (e.g., variceal hemorrhage), and intercurrent infection. In patients with established cirrhosis, HCC may be a late complication (Chap. 87) but occurs less frequently than in cirrhosis associated with viral hepatitis. Laboratory features of AIH are similar to those seen in chronic viral hepatitis. Liver biochemical tests are invariably abnormal but may not correlate with the clinical severity or histopathologic features in indi­ vidual cases. Many patients with AIH have normal serum bilirubin, alkaline phosphatase, and globulin levels with only minimal amino­ transferase elevations. Most commonly, serum AST and ALT levels are increased and fluctuate in the range of 100−1000 units. In severe cases, the serum bilirubin level is moderately elevated (51−171 μmol/L [3−10 mg/dL]). Hypoalbuminemia occurs in patients with very active or advanced disease. Serum alkaline phosphatase levels may be moder­ ately elevated or near normal. In a small proportion of patients, marked elevations of alkaline phosphatase activity occur; in such patients, clinical and laboratory features overlap with those of primary biliary cholangitis (Chap. 355). The prothrombin time is often prolonged, particularly late in the disease or during active phases. Polyclonal hypergammaglobulinemia (>2.5 g/dL) is common in AIH, as is the presence of rheumatoid factor. As noted above, circulat­ ing autoantibodies are also prevalent, most characteristically ANAs in a homogeneous staining pattern. Smooth-muscle antibodies are detected in 63% of cases and, when combined with a positive ANA, have high specificity. Of important note, low-titer ANA or ASMA is observed often in other liver diseases, including metabolic dysfunction–associated steatotic liver disease (in one series, 20% of patients had a positive ANA and 14% had a positive ASMA) and viral hepatitis. Studies of autoan­ tibodies in AIH have led to the recognition of two categories of AIH. Type I AIH is the classic syndrome prevalent in North America and northern Europe occurring in young women, associated with marked hyperglobulinemia, lupoid features, circulating ANAs, and HLA-DR3 or HLA-DR4 (especially B8-DRB1*03). Also associated with type I AIH are autoantibodies against F-actin and atypical perinuclear antineutro­ philic cytoplasmic antibodies (pANCA). Included in the spectrum of type I AIH is a subset of patients who lack ANA and anti-LKM1 but who have circulating antibodies to soluble liver antigen (SLA). Most of these patients are women and have clinical features similar to, or perhaps more severe than, those of other patients with type I AIH. The presence of SLA autoantibodies has been reported to be associated with a worse prognosis. Type II AIH, often seen in children, more common in Mediterranean populations, and linked to HLA-DRB1 and HLA-DQB1 haplotypes, is associated not with ANA but with anti-LKM. Actually, anti-LKMs rep­ resent a heterogeneous group of antibodies. In type II AIH, the antibody is anti-LKM1, directed against cytochrome P450 2D6. This is the same anti-LKM seen in some patients with chronic hepatitis C. Anti-LKM2 is seen in drug-induced hepatitis, and anti-LKM3 (directed against uridine diphosphate glucuronyltransferases) is seen in patients with chronic hepatitis D but has also been observed in type II AIH. Another autoan­ tibody observed in type II AIH is directed against liver cytosol formimi­ notransferase cyclodeaminase (anti-liver cytosol 1 [LC-1]). Liver biopsy is imperative to make the diagnosis of AIH. Abnormali­ ties are similar to those described for chronic viral hepatitis, but classic features of AIH include inflammation involving not only the portal tracts but extending beyond the portal tracts (interface hepatitis), the presence of plasma cells in the inflammatory infiltrate, and central vein perivenulitis. Necroinflammatory activity characterizes the lobular parenchyma, and evidence of hepatocellular regeneration is reflected by “rosette” formation, the occurrence of thickened liver cell plates, and regenerative “pseudolobules.” Septal fibrosis, bridging fibrosis, and cir­ rhosis are frequent. In patients with early AIH presenting as an acutehepatitis-like illness, lobular and centrilobular (as opposed to the more common periportal) necrosis has been reported. Bile duct injury and granulomas are uncommon; however, a subgroup of patients with AIH

has histologic, biochemical, and serologic features overlapping those of primary biliary cholangitis (Chap. 355).

■ ■DIAGNOSTIC CRITERIA An international group has suggested a set of criteria for establishing a diagnosis of AIH. Exclusion of liver disease caused by genetic dis­ orders, viral hepatitis, drug hepatotoxicity, and alcohol is linked with such inclusive diagnostic criteria as hyperglobulinemia, autoantibod­ ies, and characteristic histologic features. This international group has also suggested a comprehensive diagnostic scoring system that, rarely required for typical cases, may be helpful when typical features are not present. Factors that weigh in favor of the diagnosis include female gender; predominant aminotransferase elevation; presence and level of globulin elevation; presence of nuclear, smooth-muscle, LKM1, and other autoantibodies; concurrent other autoimmune diseases; charac­ teristic histologic features (interface hepatitis, plasma cells, rosettes); HLA-DR3 or DR4 markers; and response to treatment (see below). A more simplified, more specific scoring system relies on four variables: presence of autoantibodies (ANA, ASMA, anti-SLA or LKM), elevated serum IgG level, typical or compatible histologic features, and absence of viral hepatitis markers. Elevated aminotransferase levels are also required for the diagnosis of AIH. Weighing against the diagnosis are predominant alkaline phosphatase elevation, mitochondrial antibod­ ies, markers of viral hepatitis, history of hepatotoxic drugs or excessive alcohol, histologic evidence of bile duct injury, or such atypical histo­ logic features as fatty infiltration, iron overload, and viral inclusions. ■ ■DIFFERENTIAL DIAGNOSIS Early during the course of chronic hepatitis, AIH may resemble typical acute viral hepatitis or drug-induced liver injury (Chap. 350). Exclu­ sion of acute viral hepatitis with serologic testing is critical. In adoles­ cence, Wilson disease (Chaps. 355 and 427) may present with features of chronic hepatitis long before neurologic manifestations become apparent and before the formation of Kayser-Fleischer rings (copper deposition in Descemet’s membrane in the periphery of the cornea). In this age group, serum ceruloplasmin and serum and urinary copper determinations plus measurement of liver copper levels establish the correct diagnosis. Cryptogenic cirrhosis and primary biliary cholan­ gitis (Chap. 355) share clinical features with AIH, and both alcoholassociated hepatitis (Chap. 353) and metabolic dysfunction–associated steatohepatitis (MASH; formerly nonalcoholic steatohepatitis [NASH], Chap. 354) may present with many features common to AIH; his­ toric, biochemical, serologic, and histologic assessments are usually sufficient to allow these entities to be distinguished from AIH. The distinction between autoimmune and chronic viral hepatitis may not be straightforward, especially when viral antibodies occur in patients with autoimmune disease or when autoantibodies occur in patients with viral disease. Furthermore, the presence of extrahepatic features such as arthritis, cutaneous vasculitis, or pleuritis—not to mention the presence of circulating autoantibodies—may cause confusion with rheumatologic disorders such as rheumatoid arthritis and systemic LE. The existence of clinical and biochemical features of progressive necro­ inflammatory liver disease distinguishes chronic hepatitis from these other disorders, which are not associated with severe liver disease. Rarely, hepatic venous outflow obstruction (Budd-Chiari syndrome) may present with features suggestive of AIH, but painful hepatomegaly, ascites, and vascular imaging provide distinguishing diagnostic clues. Other diagnostic considerations would include celiac disease and isch­ emic liver disease, which would be readily distinguishable by clinical and laboratory features from AIH. CHAPTER 352 Chronic Hepatitis In patients treated with immune checkpoint inhibitors for malig­ nancy, the liver may be one of the autoimmune targets of therapy; the syndrome of immune checkpoint inhibitor (ICI) hepatitis resembles AIH in clinical features (without the female preponderance, how­ ever) and response to glucocorticoid-based treatment; however, patients with ICI are less likely to have ANA, which, if present, is low-titer and unlikely to have plasma cell inflammatory infiltrates histologically. Similar to de novo AIH, ICI hepatitis encompasses a spectrum of severity, and some cases may be refractory to standard

therapeutics. Finally, occasionally, features of AIH overlap with features of autoimmune biliary disorders such as primary biliary cholangitis, primary sclerosing cholangitis (Chaps. 355 and 357), or, even more rarely, mitochondrial antibody-negative autoimmune cholangitis. Such overlap syndromes are difficult to categorize, and often response to therapy may be the distinguishing factor that estab­ lishes the diagnosis.

TREATMENT Autoimmune Hepatitis The mainstay of management in AIH is glucocorticoid therapy, which is highly effective for the induction of remission. Several controlled clinical trials have documented that such therapy leads to symptomatic, clinical, biochemical, and histologic improvement as well as increased survival. A therapeutic response to glucocorti­ coids can be expected in up to 80% of patients, and recent consensus opinion defines response as a >50% reduction in ALT by 4 weeks of therapy; however, a dramatic reduction within days is uncom­ mon after the initiation of corticosteroids. Although some advocate the use of prednisolone (the hepatic metabolite of prednisone), prednisone is just as effective and is favored by most authorities. A popular regimen in the United States relies on an initiation dose of 60 mg/d. This high dose is tapered successively over the course of a month down to a goal of 10–20 mg/d. In the European literature, successful treatment for AIH has been achieved with a prednisone starting dose of 20 mg. For patients with a mild presentation, budesonide, while inferior to prednisone in AIH, is an alternative first-line corticosteroid with limited systemic effects related to its extensive hepatic first-pass metabolism that achieves similar rates of biochemical response in patients with aminotransferase levels <2× the upper limit of normal. PART 10 Disorders of the Gastrointestinal System After an early response to corticosteroids is observed, a steroidsparing agent should be initiated to maintain remission. Azathio­ prine, considered first-line treatment for nonicteric or severe AIH, should be introduced 7–14 days after corticosteroids are started and a response is observed. Testing for thiopurine S-methyltrans­ ferase activity is recommended during this period prior to initiation of azathioprine. Dosing of azathioprine may be between 50 and 150 mg, and some experts titrate the dose upward after starting at 50 mg. Once on maintenance azathioprine (typically 1–2 mg/kg per day), prednisone is tapered further to a maintenance dose of 5–10 mg. If the patient achieves full biochemical remission, an attempt to taper prednisone completely is reasonable. The advantage of this combination approach is a reduction, over the span of an 18-month course of therapy, in serious, life-threat­ ening complications of steroid therapy (e.g., cushingoid features, hypertension, diabetes, osteoporosis) from 66% down to <20%. Improvement of fatigue, anorexia, malaise, and jaundice tends to occur within days to several weeks; biochemical improvement occurs over the course of several weeks to months, with a fall in serum bili­ rubin and globulin levels and an increase in serum albumin. Serum aminotransferase levels usually drop promptly, but improvements in AST and ALT alone do not appear to be reliable markers of recovery in individual patients; histologic improvement, characterized by a decrease in mononuclear infiltration and in hepatocellular necrosis, may be delayed for 6−24 months. Still, if interpreted cautiously, ami­ notransferase levels, and immunoglobulin levels in some, are valuable indicators of relative disease activity, and although recommended, many authorities do not advocate for serial liver biopsies to assess therapeutic success or to guide decisions to alter or stop therapy. Therapy should continue for at least 24 months after biochemical remission. After tapering and cessation of therapy, the likelihood of relapse is at least 50%, and in one multicenter study, the rate of relapse was 81% after 3 years. In fact, most (95% of) patients will require treatment at maintenance doses indefinitely. Patients who relapse after completing one course of therapy almost always relapse after subsequent courses; because each relapse increases the risk of

cirrhosis, death, and need for liver transplantation, ideally, patients who relapse once should be treated indefinitely. If azathioprine is not tolerated because of adverse effects (such as marrow suppression, fatigue, or gastrointestinal symptoms), both 6-mercaptopurine (6MP) and mycophenolate mofetil (MMF) are considered second-line alternative agents; if azathioprine is ineffec­ tive, MMF (1.5–2.0 g/d in divided doses) is the next drug of choice. While MMF has yet to be compared to azathioprine in a random­ ized controlled clinical trial, in a recent prospective study, MMF was at least as effective, if not more so, than azathioprine in achiev­ ing biochemical remission and may be better tolerated. A downside of MMF that merits consideration is the risk of teratogenicity. In cases of AIH refractory to standard treatment (azathioprine, 6-MP, or MMF, with or without prednisone) and in whom coex­ isting other causes for liver disease have been excluded, several additional treatments are options. Tacrolimus has been shown to be effective. Small studies of tacrolimus have shown a decline in ami­ notransferase levels in up to 80% after 3 months of therapy; how­ ever, tacrolimus requires careful trough monitoring and comes with a risk of nephrotoxicity. Treatment with monoclonal anti-CD20 (rituximab), described initially in a small pediatric case series, is another promising option that has been used with increasing frequency in refractory AIH. In one cohort study, rituximab given to patients with refractory AIH (1000-mg IV infusion given twice, 2 weeks apart) achieved remission rates of 95% at 1 year and 71% at 2 years and permitted substantial corticosteroid dose reduction. When medical therapy fails, including in severe acute AIH refractory to high-dose corticosteroids or when AIH progresses to cirrhosis and end-stage liver disease, liver transplantation is the only recourse (Chap. 356); in patients with severe AIH, failure of the bilirubin to improve after 2 weeks of therapy should prompt early consideration of the patient for liver transplantation. Recur­ rence of AIH in the new liver occurs rarely in most experiences but in as many as 35−40% of cases in others; nonetheless, 5-year patient and graft survival exceed 80%. Like all patients with chronic liver disease, patients with AIH should be vaccinated against hepatitis A and B, ideally before immunosuppressive therapy is begun, if practical. Similarly, age- and gender-appropriate vaccinations, such as human papillomavi­ rus, varicella-zoster virus, pneumococcal, and COVID-19, should be administered prior to immunosuppressive therapy. Patients with AIH and cirrhosis should be screened for HCC with ultrasound at 6-month intervals and may merit screening for varices with upper gastrointestinal endoscopy, based on the severity of liver disease (Chap. 355). ■ ■FURTHER READING AASLD/IDSA HCV Guidance Panel: Hepatitis C guidance 2019 update: American Association for the Study of Liver Diseases– Infectious Diseases Society of America recommendations for test­ ing, managing, and treating hepatitis C virus infection. Hepatology 71:686, 2020. Updated regularly and available at http://www.hcvguide­ lines.org. Accessed April 20, 2020. Asselah T, Rizzetto M: Hepatitis D virus infection. N Engl J Med 389:58, 2023. Bhattacharya D et al: Hepatitis C guidance 2023 update: American Association for the Study of Liver Diseases-Infectious Diseases Soci­ ety Recommendations for testing, managing, and treating hepatitis C virus infection. Clin Infect Dis 2023. Buti M et al: Tenofovir alafenamide versus tenofovir disproxil fuma­ rate for the treatment of HBeAg-negative chronic hepatitis B virus infection: A randomized, double-blind, phase 3 non-inferiority trial. Lance Gastroenterol Hepatol 1:196, 2017. Carrat F et al: Clinical outcomes in patients with chronic hepatitis C after direct-acting antiviral treatments: A prospective cohort study. Lancet 393:1453, 2019. Chan HLY et al: Tenofovir alafenamide versus tenofovir disproxil fumarate for the treatment of HBeAg-positive chronic hepatitis B