Natural History of Hepatic and Extrahepatic Hepatitis C Virus Diseases and Impact of Interferon-Free HCV Therapy

Abstract

The hepatitis C virus (HCV) infects 71.1 million persons and causes 400,000 deaths annually worldwide. HCV mostly infects the liver, causing acute and chronic necroinflammatory damage, which may progress toward cirrhosis and hepatocellular carcinoma. In addition, HCV has been associated with several extrahepatic manifestations. The advent of safe and effective direct-acting antivirals (DAAs) has made the dream of eliminating this public health scourge feasible in the medium term. Prospective studies using DAA-based regimens have shown the benefit of HCV clearance in terms of both liver- and non-liver-related mortality.

Hepatitis C virus (HCV) infection poses a major global health threat. In 2015, HCV was estimated to infect ∼71.1 million persons worldwide and to cause approximately 400,000 deaths annually, two-thirds attributable to decompensated cirrhosis, and one-third to hepatocellular carcinoma (HCC) (www.who.int/hepatitis/publications/global-hepatitis-report2017/en). The figures provided by the Polaris Observatory (H Razavi, pers. comm.) for 2016 are only slightly different; with a viremic pool, at the beginning of 2016, at 71.1 million infected persons, an annual input of 1,498,000 acute infections of which 270,000 were spontaneously cured, whereas, as output, there were a total of 1,615,000 patients cured by antiviral treatment, 342,000 liver-related and 794,000 non-liver-related deaths (Fig. 1).

Figure 1.

Inputs and outputs of the global hepatitis C virus (HCV)-infected pools, at the start of 2016. According to the models by the Polaris Observatory, during 2016, there were an estimated 1,498,000 new HCV infections of which 270,000 spontaneously cleared, whereas 1,228,000 persisted, thus contributing to the pool. At the same time, an estimated 1,700,000 were treated with antivirals, of which 95% achieved sustained virologic response (SVR), corresponding to 1,615,000 cleared infections. In addition, 342,000 died of liver-related deaths associated with their HCV infection, whereas another 794,000 people infected with HCV died of non-liver-related causes (an unknown of the latter causes included extrahepatic disorders caused by HCV; see text for further discussion). (Figure created from data provided by H Razavi [pers. comm.] and cdafound.org/polaris.)

The real figures are, however, unknown because of the fact that a high proportion of deaths attributable to HCV are not reported as such, with hepatitis C not even being mentioned in the death certificate (Mahajan et al. 2014; Keiser et al. 2018). Another reason suggesting that the global health burden caused by HCV is underestimated is the fact that many infected patients may also die of disorders that are non-liver-related and yet are caused by HCV. Indeed, HCV may contribute to the pathogenesis of many extrahepatic disorders, potentially leading to death, although a precise estimate of the fraction of mortality attributable to HCV is difficult to obtain. These figures may be relevant, and yet were not included in the tally reported by the WHO Global Hepatitis Report (WHO 2017).

The incidence of new HCV infections remains high; there was an estimated 1.75 million cases in 2015, particularly in resource-poor countries mostly attributable to unsafe injections from both IV drug use and in some health care settings. In these countries, the number of new infections may largely outpace the number of patients cured of their hepatitis C after antiviral therapy (Hill et al. 2017). Coupled with the relentless progression of hepatitis C toward end-stage liver disease, it is likely that in these areas the burden of HCV will increase in the coming years, jeopardizing the WHO goal of eliminating HCV as a public health threat by the year 2030. This article deals with the natural history of hepatic and extrahepatic diseases associated with HCV infection, and with the benefits of HCV infection cure with direct-acting antivirals (DAAs).

HEPATIC DISEASES: ACUTE HEPATITIS C

HCV mostly infects hepatocytes. Early after acute infection, serum levels of HCV increase rapidly to reach a plateau. Although HCV infection is not cytolytic, it triggers innate and adaptive immune responses. The cellular immune response is the most prominent histological hallmark of hepatitis C, mediating the killing of infected and uninfected hepatocytes, in synergy with an array of cytotoxic cytokines released by the local inflammatory infiltrate. The biochemical correlate of this response is the elevation of transaminases witnessing hepatocyte lysis.

A robust, broad HCV-specific CD8⁺ T-cell response is reputed to play a pivotal role in eradicating the infection (Rehermann and Thimme 2019). On the other hand, HCV is capable to mount complex strategies of evasion leading to persistence in most patients. On average, only 25% of acutely infected patients clear the virus (Grebely et al. 2014). Although a symptomatic infection is an important predictor of viral clearance (Gerlach et al. 2003), factors like female sex and HCV genotype may play a role (Grebely et al. 2014). In addition, host genetic factors contribute prominently to the fate of primary HCV infection (for review, see Gauthiez et al. 2017). Genome-wide association studies (GWAS) have consistently identified polymorphisms in IFNL3/IFNL4 as associated with HCV clearance (Ge et al. 2009). Variants in MHC and KIR genes (Suppiah et al. 2011) and the GPR158 (Vergara et al. 2019) additively predict clearance, whereas variants in other genes involved in the immune response, such as IFNG-AS1 and STAT1, seem relevant only among patients of European ancestry (Huang et al. 2017). Of particular interest seem recent data on the role of some ERAP1 polymorphisms, because these may alter the CD8⁺ T-cell repertoire (Kemming et al. 2019).

Clinically, most cases of acute hepatitis C are symptomless or characterized by nonspecific symptoms, like fatigue or abdominal discomfort, and may proceed unnoticed. Severe hepatitis C, requiring hospitalization, is infrequent. The role of HCV in acute liver failure is controversial but may be relevant when primary HCV infection occurs in patients with a preexisting, chronic liver damage, such as chronic hepatitis B (Féray et al. 1993; Liang et al. 1993). So, the clinical impact of acute hepatitis C is essentially related to the high risk of chronicity.

HEPATIC DISEASES: CHRONIC HEPATITIS C

Chronic hepatitis C is a progressing liver disease, which may evolve to cirrhosis, usually after decades from infection. Cirrhosis caused by HCV may then progress to decompensation, liver failure and/or HCC at a rate of 2%–6% per year. These complications account for the majority of liver-related morbidity and mortality associated with HCV.

HCV induces fibrosis via two major mechanisms. First, fibrosis proceeds via a typical wound-healing process that accompanies the inflammatory infiltrate elicited by the infection (Bataller and Brenner 2005). This secretes soluble factors that activate hepatic stellate cells to secrete collagen and indeed, the single most important factor affecting the rate of fibrosis progression is the extent of intrahepatic inflammation (Leandro et al. 2006). Accordingly, patients with persistently normal alanine transaminase (ALT) have little inflammatory infiltrate and, on average, a slow liver disease progression (Mathurin et al. 1998; Zeuzem et al. 2006). Because a small minority of these may have advanced liver disease, it has been postulated that a second mechanism may account for fibrosis deposition, that is, the direct secretion of profibrogenic mediators—for example, the transforming growth factor β1—by infected hepatocytes (Schulze-Krebs et al. 2005). This may then explain the frequent observation, at liver histology, of extensive fibrosis with a remarkable paucity of infiltrate.

The variability of liver disease progression in chronic hepatitis C patients is significant. The fibrosis progression rate (FPR) is modified by several factors, which may act alone or in combination, albeit to a different extent (Rüeger et al. 2015), but the risk of HCC development is also affected by diverse factors. As far as viral factors are concerned, although the baseline serum HCV RNA level is not associated with prognosis, HCV genotype 3 is associated with accelerated FPR (Bochud et al. 2009; Probst et al. 2011). Patients with cirrhosis and HCV genotype 3 are at increased risk of end-stage liver disease (McMahon et al. 2010) and HCC (Kanwal et al. 2014).

However, more relevant liver disease modifiers are host and environmental factors. A study assessing the FPR in African-Americans versus Caucasian-Americans failed to show a significant difference (Terrault et al. 2008). A large Veterans Administration (VA) study, however, showed that patients of Hispanic ancestry were at the highest risk of having cirrhosis (adjusted hazard ratio [HR] = 1.28, 95% confidence interval [CI] = 1.21–1.37) and HCC (1.61, 95% CI = 1.44–1.80), whereas African-Americans had a lower risk of progression to both cirrhosis and HCC compared with non-Hispanic whites, after adjusting for HCV genotype, HCV treatment, diabetes, body mass index (BMI), and other factors (El-Serag et al. 2014). It is likely that genetics explains only in part the variance in FPR attributed to ethnicity, and that lifestyle and access to care may have some impact as well.

Although the male sex is also associated with accelerated FPR (Poynard et al. 1997) and increased incidence of HCC, after menopause, the FPR risk also increases in women, via unclear mechanisms (Di Martino et al. 2004). Women with history of multiple pregnancies and those receiving estrogen-replacement therapy are partially protected against accelerated fibrosis progression (Di Martino et al. 2004).

Several candidate gene studies have suggested a role for numerous host gene polymorphisms, alone or in combination, leading to the elaboration of scores predictive of liver disease progression. Among the variants found to affect fibrosis progression, some lie in proximity of IFNL3/IFNL4 and, interestingly, are the same that are associated with an increased viral clearance. However, only a few polymorphisms have been confirmed by more robust GWAS. A multicohort study (Patin et al. 2012) identified single-nucleotide polymorphisms (SNPs) in RNF7 (a gene encoding for an antioxidant protecting against apoptosis), MERTK and TULP1 (two factors involved in the phagocytosis of apoptotic cells by macrophages), as associated with accelerated FPR. Additionally, several genetic variations within the MHC region are also predictive of cirrhosis (Urabe et al. 2013). A more recent meta-analysis of four independent cohorts, focusing on the attributable fraction of risk of accelerated FPR, identified the variants of TULP1, PNPLA3, MERTK and within the MHC region as being significantly and independently associated to FPR, with the SNPs in MERTK being the most relevant in terms of attributable fraction (14.4%) (Rüeger et al. 2015). The prognostic role of SNPs in MERTK was confirmed by a longitudinal study (Jiménez-Sousa et al. 2018). On the other hand, PNPLA3 variants are associated with an increased risk of HCC (Valenti et al. 2011).

Age at infection is, by and large, the most important feature affecting prognosis (Poynard et al. 1997; Minola et al. 2002), contributing to at least one-third of the FPR (Rüeger et al. 2015). However, this should not underestimate the unquestionable impact on both fibrosis progression and HCC development played by modifiable factors, especially the excessive alcohol consumption (Poynard et al. 1997; Tagger et al. 1999) and the metabolic syndrome (Veldt et al. 2008; Bugianesi et al. 2012). Also, tobacco and cannabis smoking may contribute to fibrosis progression, although the data is scantier (Hézode et al. 2005; Mallat et al. 2008).

Finally, coinfections with hepatitis B virus (HBV) and the human immunodeficiency virus (HIV) are of major clinical relevance, because of the frequent overlapping of transmission routes. HBV coinfection increases the risk of incident HCC in cirrhosis patients (Tagger et al. 1999), whereas FPR is accelerated in patients coinfected with HIV (Thein et al. 2008; Konerman et al. 2014), an outcome only partially prevented by antiretroviral therapy (Deng et al. 2009). A low CD4 count is also a risk factor for HCC (Ioannou et al. 2013).

HCV-ASSOCIATED EXTRAHEPATIC DISEASES

HCV infection is associated with several extrahepatic manifestations (EHMs), although the level of causality has not been definitely proven for all of them (Table 1; Lodi et al. 2010; Negro et al. 2015; Younossi et al. 2016a; Balakrishnan et al. 2017; Ramos-Casals et al. 2017; Sayiner et al. 2017; Segna and Dufour 2017; Cacoub et al. 2018a). These EHMs add significantly to the HCV-related morbidity and mortality. As many as 38%–76% of patients with chronic hepatitis C develop at least one EHM (Himoto and Masaki 2012). Regrettably, the health burden caused by HCV-induced EHMs has not received the same attention as its liver-related outcomes, probably in part because of the difficulty in estimating the mortality risk fraction directly attributable to HCV. Many EHMs largely occur and progress independently of HCV, such as B-cell lymphoma or type 2 diabetes, and only recently the relative contribution of the virus has been appreciated owing to studies reporting on the benefits of antiviral therapy.

Table 1.

Extrahepatic manifestations reported in hepatitis C virus (HCV)-infected patients

Manifestation	Strength of association
Mixed cryoglobulinemia	++
Chronic kidney disease	++
B-cell lymphoma	++
Insulin resistance/type 2 diabetes	++
Major adverse cardiovascular events (ischemic)	++
Fatigue	++
Cognitive impairment	++
Porphyria cutanea tarda	++
Autoimmune thyroiditis	++
Sjögren-like sialadenitis	+
Depression	+
Rheumatoid arthritis	+
Lichen planus
Skin	+
Oral	++
Extrahepatocellular cancers
Intrahepatic cholangiocarcinoma	++
Oral squamous cell carcinoma	+
Pancreatic carcinoma	+
Renal carcinoma	+
Papillary thyroid carcinoma	?
Extrahepatic cholangiocarcinoma	?
Lung carcinoma	?
Arthralgia	?
Interstitial lung disease	?
Idiopathic thrombocytopenic purpura	?
Psoriasis	?

(++) proven causality, (+) debated causality, (?) association reported but requiring further study (see text for references).

Specific recommendations on the management of EHMs in HCV-infected persons exist (Ramos-Casals et al. 2017). Although the current treatment indication is universal in case of detectable HCV RNA (European Association for the Study of the Liver 2018), the presence of severe symptoms associated with EHMs are one among many reasons for prioritizing antiviral therapy.

The causal link between EHMs and HCV has been established based on epidemiological and clinical data (Lee et al. 2012; Negro et al. 2015; Younossi et al. 2016a), whereas treatment-induced viral clearance is associated with clinical improvement of many EHMs (Cacoub et al. 2018a). Furthermore, the availability of potent interferon (IFN)-free regimens has made possible the treatment of patients in whom IFN was contraindicated (e.g., with EHMs related to autoimmunity), hence providing the opportunity to verify the causal association between infection and disease.

The pathogenesis of HCV-related EHMs proceeds via diverse mechanisms, not necessarily involving the replication of HCV at extrahepatic sites, reported only in selected cases, although data have not been independently validated (Zignego et al. 2007; Fletcher and McKeating 2012).

Extrahepatic Manifestations Related to Chronic Immune Stimulation

Immunostimulation, autoimmune, or cytokine-mediated mechanisms have been advocated (Negro et al. 2015). Mixed cryoglobulinemia (type II or III) is historically the first EHM to have been found associated with HCV infection (Pascual et al. 1990). Type II cryoglobulinemia is rather specific of HCV, because up to 95% of patients have HCV infection (Sene et al. 2004). It is a lymphoproliferative disorder characterized by the production and extravascular deposition of immune complexes formed by monoclonal and polyclonal immunoglobulins. According to a meta-analysis, on average, ∼30% of hepatitis C patients have cryoglobulinemia, with an 11-fold increased risk versus controls (Younossi et al. 2016a), although only ∼5% of patients may present with the classical triad of symptoms (fatigue, arthralgia, and palpable purpura). Complexes tend to precipitate in small-to-medium-sized blood vessels, with a leukocytoclastic vasculitis as the typical histopathologic hallmark. Several organs, primarily the skin but also the central nervous system (CNS), gut, and peripheral nerves, can be affected with clinical consequences encompassing, respectively, atypical strokes, digestive bleeding, and sensitive neuropathy.

Deposition of immune complexes may affect the kidneys, resulting in glomerulonephritis, mostly of the membranoproliferative type, that may lead to kidney failure. There is a 26% risk increase of developing chronic kidney disease in HCV-infected versus -uninfected persons, although not all of them are attributable to cryoglobulinemia (Younossi et al. 2016a). In any case, according to the recently revised KDIGO (Kidney Disease: Improving Global Outcomes) guidelines, all patients with chronic kidney disease should be screened for HCV and, conversely, all patients with hepatitis C should undergo an assessment for the presence of kidney damage (Kidney International Supplements 2018). Another ominous complication is the occurrence of several subtypes of B-cell non-Hodgkin's lymphoma (Pozzato et al. 1994). The risk of lymphoma is 60% higher in patients with HCV infection compared with uninfected controls (Younossi et al. 2016a).

Insulin Resistance and Diabetes

HCV is also associated with a higher risk of developing type 2 diabetes (Bugianesi et al. 2012). This proceeds via the establishment of peripheral insulin resistance and affects primarily patients who are already at risk for glucose metabolism alterations, in whom diabetes may develop at least one decade earlier than in uninfected persons (Mehta et al. 2003). In a recent longitudinal study, HCV is an independent predictor of diabetes even after correction for transaminase elevations, suggesting that glucose metabolism disturbances are partly independent of liver inflammation (Lerat et al. 2017; Lin et al. 2017). Chronic hepatitis C with type 2 diabetes are at a higher risk of decompensated cirrhosis (Elkrief et al. 2014) and HCC (Veldt et al. 2008).

Cardiovascular Disorders

Among the other prominent EHMs associated with HCV, cardiovascular disorders merit a special mention as a significant cause of mortality. The risk of cardiovascular disease in HCV is 20% higher than among uninfected persons, whereas, more specifically, that of stroke is 35% higher (Younossi et al. 2016a). Patients with HCV infection tend to develop surrogate markers of cardiovascular mortality (intima-media thickness and carotid plaques) at an earlier age than uninfected persons, especially if presenting with advanced liver fibrosis (Petta et al. 2018).

Neurological Disorders

HCV infection has been associated with neuropsychiatric manifestations (Yarlott et al. 2017). Indirect mechanisms of CNS dysfunction encompass the hepatic encephalopathy—a nonspecific condition affecting patients with advanced liver disease and portal hypertension—and cerebritis, reported in occasional patients with cryoglobulinemia (Cacoub et al. 2005). However, a wealth of neurological and psychiatric alterations has been reported in hepatitis C patients even in the absence of advanced liver damage or cryoglobulinemia. The most prominent is fatigue (Goh et al. 1999; Poynard et al. 2002), the severity of which bears no correlation with the liver disease activity or common parameters of autoimmunity. In a recent meta-analysis (Younossi et al. 2016a), the pooled mean fatigue score was 3.90, that is, 1.60 points higher than score reported by healthy controls, suggesting a significant fatigue linked to HCV. The pathogenesis is unclear. Although a study has pinpointed an association with some inflammatory markers (Huckans et al. 2014), fatigue is a multifactorial condition, potentially affected by several social and psychiatric factors unrelated to HCV. Thus, it is impossible to establish a priori whether fatigue in any given patient is specifically linked to the viral infection, or what the effect of a successful antiviral therapy will be.

Depression is another prominent condition reported in chronic hepatitis C patients (Fontana et al. 2002), and has been a formidable impediment to the widespread use of IFN-α-based therapies before the advent of DAAs (EASL 2018). Based on a meta-analysis, the risk of developing depression is about twofold in HCV-infected versus uninfected controls (Younossi et al. 2016a). Although depression may be associated with drug abuse and thus precedes HCV infection, an indirect viral effect mediated by the host immune activation has been suggested (Pawlowski et al. 2014).

A third complex of neuropsychiatric symptoms associated with HCV are collectively referred to as cognitive dysfunction (Forton et al. 2001). The role of viral infection in its pathogenesis has been disputed in the past, owing to the fact that chronic hepatitis C patients may present with several confounding factors that may result in cognitive impairment per se (e.g., drug abuse or psychiatric conditions). However, recent data from clinical trials assessing cognitive dysfunction longitudinally, that is, before and after successful antiviral therapy, has allowed confirming a viral pathogenesis for many such symptoms (see below). Dysfunctions observed in chronic hepatitis C patients include diminished concentration and attention span, impaired memory, poor speed of processing, and learning efficiency (Yarlott et al. 2017), whereas motor performance and visuospatial functions seem relatively preserved, suggesting some specific mechanisms of damage (Weissenborn et al. 2004). Sophisticated imaging studies have analyzed the neuroanatomical basis of cognitive dysfunctions linked to HCV, although the small study size and the heterogeneity of the assessment protocols limit the robustness of findings. The current evidence suggests that HCV may not only induce brain inflammation, but, more specifically, reduce brain metabolism and interfere with dopaminergic signaling (Yarlott et al. 2017).

Fatigue, depression, and cognitive impairment largely account for the reduction in health-related quality of life (HRQOL) universally reported by chronic hepatitis C patients, a finding that challenges the dogma of an asymptomatic condition usually attributed to chronic hepatitis. A large body of evidence exists confirming the reduced HRQOL in chronic hepatitis C, irrespective of the assessment instrument. In particular, demographic features and study design do not affect HRQOL scores as do most clinical features with the exception of cirrhosis, which reduces the physical component score (Younossi et al. 2016a). The mental health domains are those who score the worst (in particular, social functioning), compared with physical health domains (Younossi et al. 2016a). Anxiety, stigmatization, and especially labeling following diagnosis may play a role (Rodger et al. 1999), although this has been later disputed by a study conducted in blood donors unaware of their HCV serostatus (Strauss et al. 2014). The recent clinical trials assessing HRQOL before and after antiviral therapy have contributed significantly to clarify these issues (see below).

ADVENT OF DAAs AND THEIR IMPACT ON HCV-RELATED MORBIDITY AND MORTALITY

DAAs are not only very efficacious, but are also characterized by a high level of safety. Contraindications and safety issues are rare. An important consequence is the ease of use also among patients with severe liver dysfunction (with the notable exception of protease inhibitors) or extrahepatic comorbidities. This has allowed the unprecedented opportunity to reduce HCV-related morbidity and mortality in subgroups of patients at high risk of fatal complications and in whom IFN-α was contraindicated. I will outline some of the most important observations related to the impact of DAAs on the risk of liver decompensation, HCC, EHMs, and mortality.

SVR AND DECOMPENSATED CIRRHOSIS

The efficacy of DAAs in patients with decompensated cirrhosis became evident because of early studies using a combination of sofosbuvir, ledipasvir, and ribavirin (SOLAR-1 study) (Charlton et al. 2015). In the following phase III study (SOLAR-2), patients were randomized to the same combination given for 12 or 24 weeks. Among those with genotype 1 (the vast majority of enrolled), SVR was achieved by 20/23 (87%) child B patients with 12 weeks of treatment and 22/23 (96%) child B patients with 24 weeks of treatment, 17/20 (85%) of child C patients (12 weeks of treatment), and 18 (78%) of 23 child C patients (24 weeks of treatment) (Manns et al. 2016). These regimens were also sufficiently well tolerated, although the ribavirin dose had to be adjusted frequently. The latest combination of sofosbuvir and velpatasvir was tested in a further clinical trial (ASTRAL-4). Here, 267 patients with child B cirrhosis were randomized to receive sofosbuvir/velpatasvir for 12 weeks without or with ribavirin, or sofosbuvir/velpatasvir for 24 weeks (Curry et al. 2015). The SVR rates were, respectively, 83%, 94%, or 86%. Among the 250 patients with Child-Pugh-Turcotte with data available at baseline and 12 weeks after the end of therapy, about one-half improved their score. The MELD (model for end-stage liver disease) score improved in 114/223 (51%) of those with a score <15 at baseline, and in 22/27 (81%) of those who had a score of 15 or higher.

These data were confirmed by real-world studies (Table 2), which also analyzed in more detail the clinical outcomes. The UK Expanded Access Program showed a successful eradication of HCV in 381/467 (81.6%) patients (409 with decompensated cirrhosis at enrollment) (Foster et al. 2016). Most importantly, in the follow-up (FU) study, SVR was associated to a reduction in incident decompensation events, from 28% among untreated patients to 18% in the first 6 months from DAA treatment start to 7% during months 6–15 (Cheung et al. 2016). Overall, during the first 6 months following SVR, 60% of patients reported an improved liver function, whereas in 17% this remained unchanged. Approximately one-fourth of patients saw their MELD score worsen. The improvement of liver function and MELD score in patients with decompensated liver disease waiting for liver transplantation may lead to patients’ delisting. A European study showed that after achieving SVR, one out of five candidates were removed from the liver transplant waiting list because of improved liver function (Belli et al. 2016). A more recent reassessment of the ELITA (European Liver and Intestine Association) cohort, based on a longer FU, showed that the proportion of delisted patients may be as high as 30.9% (Perricone et al. 2018). Another recent study from Spain confirmed these data, reporting that about one-quarter of patients with decompensated cirrhosis were delisted because of an improved MELD score (Pascasio et al. 2017). Indeed, the extended use of DAAs in cirrhotic patients, even before decompensation episodes have occurred, has led to a significant decrease of the proportion of patients with HCV infection listed for liver transplantation, both in the United States, with a 30% decrease since the introduction of DAAs (Flemming et al. 2017), and in several European series (Crespo et al. 2018; Ferrarese et al. 2018; Vaziri et al. 2019).

Table 2.

Clinical benefits of DAA-induced SVR in decompensated HCV-related cirrhosis

Investigators (year)	n	Regimen	SVR (%)	Major clinical outcomes
Clinical trials
Charlton et al. (2015) SOLAR-1	102a	SOF/LDV + RBV, 12 wk	45/52 (86.5)	In CPT-B, CPT improved in 64% and MELD in 53%
		SOF/LDV + RBV, 24 wk	44/50 (88.0)	In CPT-C, CPT improved in 70% and MELD in 50%
Manns et al. (2016) SOLAR-2	107a	SOF/LDV + RBV, 12 wk	39/47 (83.0)	In CPT-B, CPT improved in 70% and MELD in 65%
		SOF/LDV + RBV, 24 wk	43/50 (86.0)	In CPT-C, CPT improved in 87% and MELD in 83%
Curry et al. (2015) ASTRAL-4	267	SOF/VEL, 12 wk	75/90 (83.3)	117/250 (47%) improved their CPT score at 12 wk after the end of therapy; MELD improved in 114/223 (51%) if <15 at baseline, and in 22/27 (81%) if ≥15
		SOF/VEL + RBV, 12 wk	82/87 (94.3)
		SOF/VEL, 24 wk	77/90 (85.6)
Real-world experience
Cheung et al. (2016)	406	SOF/DCV, 12 wk	7/11 (63.4)	Decrease in decompensation events in months 6–15 (7%) versus months 0–6 (18%) or untreated (28%); no reduction in HCC incidence
		SOF/DCV + RBV, 12 wk	111/149 (74.5)
		SOF/LDV, 12 wk	12/18 (66.7)
		SOF/LDV + RBV, 12 wk	187/228 (82.0)
Belli et al. (2016)	103	SOF/RBV, 24–48 wk	47/52 (90.4)	33.3% inactivated from LT waiting list and 19.2% delisted after 60 wk from treatment start (predictors: MELD, delta MELD after 12 wk, delta albumin after 12 wk)
		SOF + 2nd DAA, 12–24 wk	50/51 (98.0)
Pascasio et al. (2017)	171	Various	142/171 (83.0)	29/122 (24%) of patients without HCC were delisted after a median of 50 (IQR 39–63) wk from treatment start (0/6 patients with MELD >20)
Fernández-Carrillo et al. (2017)	144	Various	112/144 (78)	97% of patients with MELD 18 were alive 36 wk from treatment start versus 68% of those with MELD ≥18
Perricone et al. (2018)	142	Various	NA	44/142 (30.9%) were delisted after 2 years from treatment start (four relisted and one died of HCC)

(DAA) direct-acting antiviral, (SVR) sustained virologic response, (HCV) hepatitis C virus, (SOF) sofosbuvir, (DCV) daclatasvir, (RBV) ribavirin, (LDV) ledipasvir, (LT) liver transplantation, (CPT) Child-Pugh-Turcotte, (MELD) model of end-stage liver disease.

^aExcluding patients treated after LT.

The DAA-induced clinical benefit in patients with decompensated cirrhosis seems to be limited to patients with a baseline MELD score <18 (Fernández-Carrillo et al. 2017) or <20 (Belli et al. 2016; Pascasio et al. 2017). Those with higher scores were less likely to be removed from the waiting list because of insufficiently improved liver parameters. A decrease in MELD score may not necessarily be sufficient to delist patients or, even worse, it may lead to delisting without being accompanied by an improved quality of life or a reduced risk of potentially lethal complications, such as rapidly progressing HCC (Perricone et al. 2018). Relisting may also occur (Perricone et al. 2018). Thus, in some cases, patients may no longer qualify for liver transplant, while still being at risk of complications and even death, an untoward situation that has be referred to as “MELD purgatory.” To avoid this, guidelines have suggested to treat patients with DAAs only if the MELD score is below 18–20, and to transplant first and treat later if the MELD score is higher (EASL 2018), although these rules should be modulated based on local trends and availability of organ donors. For example, patients with a MELD score ≥18–20 and an expected waiting time before transplantation higher than 6 months may be treated without delay.

SVR AND HEPATOCELLULAR CARCINOMA

Several meta-analyses have shown that IFN-induced SVR is consistently associated with a significantly reduced risk of developing HCC, including among patients with previous HCC ablative therapy (Ikeda et al. 1999; Morgan et al. 2013; Moon et al. 2015). Because IFN was limited to patients without advanced liver damage, the introduction of safe and effective DAAs allowed the evaluation of the benefit of viral clearance on HCC incidence in advanced cirrhosis. A first manuscript (Reig et al. 2016) opened a fierce debate challenging this optimistic outlook, befuddling expectations. In a retrospective cohort study of 58 patients treated first for HCC and then with DAAs, these investigators reported a staggering HCC recurrence rate of 27.6% after a median FU of 5.7 months (Reig et al. 2016). Despite some criticism (e.g., the study was underpowered, the design was retrospective, and the limited FU was calculated using the time from DAA initiation instead of the time from HCC treatment), these results were supported by another study from Italy, with a 29% HCC recurrence rate after SVR (Conti et al. 2016). Here, in addition, the incidence rate of de novo HCC was 3.16%, not dissimilar from historical rates in untreated patients. Other studies soon followed, in which the incidence of HCC in patients with SVR was relatively high (Cardoso et al. 2016; Kozbial et al. 2018), whereas the HCC recurrence rate in patients who had undergone ablative therapy was variable (Kolly et al. 2017; Nagata et al. 2017; Shimizu et al. 2017; Hassany et al. 2018; Persico et al. 2018b; Nishibatake Kinoshita et al. 2019). In some studies, the HCC pattern of recurrence also seemed unusually aggressive in terms of BCLC (Barcelona Clinic for Liver Cancer) staging, vascular invasion, and short-term mortality, suggesting not only a potential lack of benefit of HCV clearance on the incidence of de novo HCC, but also an increased risk of rapidly progressing HCC after ablative therapy. Overall, it seems as if IFN-based regimens provide an additional benefit, particularly in terms of protection against incidence of de novo liver cancer, arguing in favor of an immunomodulatory effect on top of a mere antiviral effect, such as that afforded by DAAs (Bielen et al. 2017).

These data were not confirmed by all studies (Bielen et al. 2017; Ioannou et al. 2017; Virlogeux et al. 2017). An important (although retrospective) study on 62,354 patients from the VA (Ioannou et al. 2017) showed that DAA-induced SVR was associated with a significant reduction in de novo HCC risk, both in cirrhotic and noncirrhotic patients. The analysis was adjusted for many confounders, encompassing all major risk factors for HCC, including age, sex, alcohol abuse, and diabetes. Most importantly, these adjustments allowed fair comparisons between patients’ populations treated with IFN-based regimens versus those having been treated with DAAs, known for having more advanced liver disease. Thus, SVR was associated with a significantly decreased risk of HCC irrespective of the type of antiviral regimens, with a risk reduction of 71% (patients treated with DAAs only), 52% (when treated with DAAs plus IFN), and 68% (after treatment with IFN-based regimens but without DAAs).

An FU study also from the VA confirmed the benefit of DAAs, although among patients with SVR and cirrhosis, it remained clear that the absolute risk of HCC remained somehow high (1.82/100 person-years), requiring patients’ surveillance even after viral clearance (Kanwal et al. 2017). Another large study from Spain (Calleja et al. 2017) reported a lower incidence of de novo HCC (0.93%), a figure still far from zero.

Thus, the questions to be addressed are (1) whether the risk of de novo HCC is indeed reduced by successful treatment with DAAs and to what extent, which may influence the need (or lack thereof) for a strict surveillance after SVR, and (2) whether treatment with DAAs in patients having received ablative therapy for HCC may be protected from recurrence, and if not, whether the recurrence rate may be higher and characterized by the appearance of more aggressive tumors, potentially impacting survival.

These concerns were analyzed in a meta-analysis (Waziry et al. 2017) of 26 studies on HCC occurrence and 17 on HCC recurrence. The meta-regression adjusting for study FU and age, DAA therapy was not associated with higher HCC occurrence (risk ratio [RR] 0.68, 95% CI 0.18–2.55; P = 0.55) or recurrence (RR 0.62, 95% CI 0.11–3.45; P = 0.56), suggesting that no evidence could be put forth supporting a differential HCC occurrence or recurrence risk following SVR from DAA and IFN-based regimens. These investigators rightly pointed out how, at baseline, DAA-treated patients had a higher HCC risk profile than IFN-treated patients, that is, they were older, had more advanced cirrhosis, and presented more frequently with esophageal varices and low platelets, among other features. Furthermore, in most studies (i.e., 19 out of 24), the timing of FU for HCC detection was calculated from the end of antiviral treatment. A consequence of this was that HCC cases undiagnosed at baseline were more likely to be diagnosed as de novo HCC whenever the antiviral treatment duration was short (for DAAs, ranging from 12 to 24 weeks), compared with long (a typical IFN-based schedule lasting up to 48 weeks). Thus, a higher HCC risk together with its progressive decrease with time (i.e., a typical cohort effect) corresponds to a higher HCC risk profile at the start of DAA therapy, and may be explained by the diagnosis of HCC early after DAA therapy ends. Adjustment for age and FU time annulled the difference between DAA- and IFN-based regimens. The investigators concluded that there is no evidence that DAA therapy is associated with HCC development, and that clearance of HCV following DAA therapy in patients with cirrhosis (63% in subgroup analysis) should afford the same benefit (i.e., risk reduction), as an IFN-based therapy (77% in a historical meta-analysis) (Morgan et al. 2013). Following studies corroborated this view regarding the incidence of de novo HCC, that is, once differences in severity (e.g., lower platelet counts, lower albumin, previous episodes of decompensation, prior treatment failure, or presence of esophageal varices) or age are accounted for, the benefit of DAAs is evident. This was shown by both retrospective (Innes et al. 2018; Li et al. 2018) and prospective studies (Calvaruso et al. 2018; Nahon et al. 2018; Lleo et al. 2019; Pinero et al. 2019).

The French prospective CirVir cohort (Nahon et al. 2018), enrolling 1270 patients with compensated, biopsy-proven HCV-associated cirrhosis since 2006, merits a mention. Here, the treatment effect of DAAs from the time until HCC was evaluated by constructing a robust time-dependent Cox regression model was weighted by the inverse probability of treatment and censoring (IPTCW). Patients receiving DAAs were older, with more frequently diabetes, portal hypertension, or severe liver dysfunction. Although the crude 3-year cumulative incidence of HCC was 5.9% in the DAA group and 3.1% in the IFN-cleared group, after Cox analyses weighted by IPTCW, there was no statistically significant increase in risk of HCC associated with DAA use (HR 0.89, 95% CI 0.46–1.73). Furthermore, patients treated with DAAs were also less likely to undergo strict surveillance. The investigators concluded that the apparent increase in HCC incidence observed in patients with cirrhosis treated with DAAs compared with patients who achieved SVR following an IFN therapy can be explained by patient characteristics (age, diabetes, reduced liver function) and lower screening intensity. The pattern of aggressiveness of HCC was not increased at the time of diagnosis in DAA-treated patients. An additional prospective study on 3917 patients from Italy with advanced fibrosis showed a yearly incidence of HCC comparable or slightly smaller to a historical untreated control group, at least during the first year of FU, but that this risk would further decline thereafter, suggesting that early de novo HCC may be because of small, undetected HCC predating DAA therapy, and that the DAA beneficial effect would only be evident at a later stage (Romano et al. 2018). The importance of the length of FU was also suggested by another prospective study from Japan on 401 patients without history of HCC, followed for 12.7 months from SVR and analyzed by a sensitive magnetic resonance imaging (Toyoda et al. 2019). This study failed to show any difference in the incidence of hypervascularization or the new emergence of non-hypervascular hypointense nodule between SVR and untreated patients, matched for propensity scores. The investigators concluded that SVR did not promote HCC (Toyoda et al. 2019). Possibly, the apparent lack of suppression may have been related to the short FU after SVR, as in the Italian cohort.

Finally, a vast prospective cohort study in adult patients with chronic hepatitis C enrolled at 32 expert centers across France (HEPATHER cohort) (Carrat et al. 2019) analyzed 7344 patients treated with DAAs and followed for a median of 33.4 months. Patients with decompensated liver disease were excluded (Figs. 2 and 3). Another 2551 patients remained untreated. In the adjusted multivariable analysis, compared with untreated patients (adjustments were made, among other factors, for age, sex, BMI, geographical origin, infection route, fibrosis stage, being naive of HCV therapy, HCV genotype, alcohol consumption, diabetes, arterial hypertension), SVR after DAAs was definitely associated with a decreased incidence of HCC.

Figure 2.

Global survival, survival free from hepatocellular carcinoma, and survival free from decompensated cirrhosis, according to exposure to direct-acting antivirals in all patients analyzed in the study from Carrat et al. (2019). The upper panels show the unadjusted survival curves, whereas the lower panels show the multivariable-adjusted survival curves estimated with a time-dependent Cox proportional hazards model. (HR) hazard ratio. (Figure from The Lancet, Carrat et al., Vol. 393, pp. 1453–1464, © 2019; reprinted, with permission, from Elsevier.)

Figure 3.

Global survival, survival free from hepatocellular carcinoma, and survival free from decompensated cirrhosis, according to exposure to direct-acting antivirals in patients with cirrhosis analyzed in the study by Carrat et al. (2019). The upper panels show the unadjusted survival curves, whereas the lower panels show the multivariable-adjusted survival curves estimated with a time-dependent Cox proportional hazards model. (HR) hazard ratio. (Figure from The Lancet, Carrat et al., Vol. 393, pp. 1453–1464, © 2019; reprinted, with permission, from Elsevier.)

The above studies demonstrated a reduced risk of de novo HCC (or at least the lack of increased risk, in some studies with short post-SVR FU) in patients with cirrhosis treated with DAAs. However, because a residual risk has been shown in all studies, this raises the unresolved question of how to conduct appropriate surveillance after SVR. It has been suggested that patients with portal hypertension (esophageal varices, lower platelet count) may be a candidate for stricter surveillance, whereas those fulfilling the extended Baveno criteria (i.e., platelets >110 G/L and liver stiffness measurement <25 kPa) may have a relaxed surveillance (Lleo et al. 2019), although the precise modalities remain to be established. In the CirVir cohort (Nahon et al. 2018), before weighting for IPTCW, an increased risk of HCC was observed in patients who were older and had diabetes, indicating which patients may benefit from closer surveillance, but the modalities of this are unclear.

Properly adjusted analyses on patients with prior HCC therapy showed that the risk of HCC recurrence and its aggressiveness do not seem to be increased by DAA therapy (Guarino et al. 2018; Huang et al. 2018a,b; Mashiba et al. 2018; Ide et al. 2019), with a single important exception from Egypt (El Kassas et al. 2018). Clearly, there is increasing evidence that weighting methodologies are necessary to avoid biased interpretations, and when applied properly there is no evidence that DAAs increase the risk of HCC, either de novo or recurrent after ablative or palliative therapy. In the latter cases, whether treatment with antivirals should be started early or after HCC cure has been ascertained remains possibly the only open question.

EFFECT OF SVR ON EXTRAHEPATIC MANIFESTATIONS

A recent meta-analysis (Cacoub et al. 2018a), supported by a rigorous assessment of the quality of published evidence, reported that SVR is associated with a ∼50% reduction in HCV-related extrahepatic mortality. This not only adds to the evidence in favor of universal antiviral treatment, but provides a strong evidence supporting the causal link between HCV infection and EHMs. The issue has been debated for a long time, mostly because of the fact that studies assessing the effect of antiviral therapy on EHMs were rarely randomized, but also in part because of the inconsistency of case definition, or the scarcity of data altogether, both in terms of number of patients and of duration of post-SVR FU.

Mixed Cryoglobulinemia, Kidney Disease, and B-Cell Lymphoma

Early data were obtained with IFN-based regimens, which posed the known safety issues and contraindications, preventing their expanded use to more severe cases, especially if presenting with severe kidney dysfunction. Moreover, some manifestations associated with cryoglobulin deposition seemed to be in part irreversible. An important prospective study analyzed the effect of a treatment with pegylated IFN-α and ribavirin in 253 chronic hepatitis C patients with mixed cryoglobulinemia (121 symptomatic) and 158 controls (Gragnani et al. 2015). Although this study was primarily aimed at assessing whether the presence of cryoglobulins may independently decrease the response to the above combination, more importantly, its results confirmed that SVR was invariably associated with a clinical response. On the other hand, the symptom improvement was not universal, because 21.1% of patients still reported weakness at the end of FU, 18% still complained of sicca syndrome and 13.1% of peripheral neuropathy (Gragnani et al. 2015). It is possible that in these cases the long-standing deposition of cryoprecipitate in small and medium vessels may have caused irreversible damage, with particular regard to vasa nervorum. This limitation of antiviral treatment was frequently observed also in recent studies using DAAs. In the prospective international multicenter cohort study of 148 patients with symptomatic HCV-related cryoglobulinemia vasculitis (The VASCUVALDIC3 study) (Cacoub et al. 2019), all treated with different DAA regimens, a complete response was observed in 106 (73%), a partial in 33 (22.6%), and no response in only seven patients (4.8%). At the end of FU (lasting a median of 15.3 months), cryoglobulins disappeared in slightly above one-half of patients, despite a 97% SVR rate. Factors predicting lack of complete response included severity of damage and peripheral neuropathy. Although more long-term FU studies are needed, a large-scale benefit on the long-term consequences of the cryoglobulinemia syndrome, especially kidney damage and B-cell lymphoma is expected to the same extent as reported in patients treated with IFN-based regimens (Mahale et al. 2018). However, early treatment is to be advocated, that is, before irreversible end organ damage is constituted, because the size of the clinical benefit seems to be inversely correlated with the time from diagnosis to antiviral therapy initiation (Mahale et al. 2018). Another limitation seems to be represented by the fact that B-cell clones persist in a “dormant” state long after SVR has been reached, and that may reactivate leading to recurring cryoglobulinemia syndrome (Visentini et al. 2019).

The limited beneficial effect of antiviral-induced viral clearance is particularly serious in patients with stage 4 and 5 chronic kidney disease. Treatment with a 12-week course of a combination of grazoprevir and elbasvir, or glecaprevir and pibrentasvir failed to result in improved renal function in all dialysis patients (Roth et al. 2015; Gane et al. 2017). However, in the first study only two patients had cryoglobulinemia, whereas in the second trial no specific information was provided.

B-cell lymphoma can be induced into remission by IFN therapy (Peveling-Oberhag et al. 2016; Cacoub et al. 2018a). In the DAA era, data is still scarce, but the pattern seems to be consistent with historical reports. An observational study on 46 patients with indolent B-cell non-Hodgkin lymphoma (37 had marginal zone type) or chronic lymphocytic leukemia treated with different DAA regimens (Arcaini et al. 2016). SVR was reached in 98% of cases. A lymphoproliferative disease response was observed in 67% of patients and was complete in 26%. The 1-year progression-free and overall survival (after a median FU of 8 months) was 75% and 98%, respectively.

The response seems favorable also in diffuse large B-cell lymphoma, where DAAs were given together with rituximab-containing chemotherapy (Persico et al. 2018a). The overall survival at 1 year was similar to a historical control group not receiving antivirals. However, the disease-free survival was better in patients treated with DAAs. Importantly, there were no toxicity issues deriving from the combinations of the antivirals and chemotherapy agents. In another study (Merli et al. 2018) on 47 consecutive patients with diffuse large B-cell lymphoma treated with DAAs concurrently or subsequently to chemotherapy (25% with cirrhosis), a complete tumor response was observed in 46 cases. After a median FU of 2.8 years, two patients had died, corresponding to a 2-year overall survival of 97.4%. Liver toxicity was more frequent in patients having received chemotherapy first; the investigators argued for a concomitant administration.

Insulin Resistance and Type 2 Diabetes

SVR leads to a decreased level of insulin resistance and a reduced risk of incident diabetes and its complications, such as end-stage renal disease, ischemic stroke, and retinopathy (Lee et al. 2012; Cacoub et al. 2018; Li et al. 2019). The effect afforded by SVR induced by IFN seems confirmed by the experience with DAAs (Li et al. 2019). Glucose metabolic alterations are overall improved after SVR induced by DAAs. The beneficial effects encompass the fasting glucose level (Ciancio et al. 2018; Drazilova et al. 2018; Weidner et al. 2018), a reduced rate of impaired fasting plasma glucose (Weidner et al. 2018), and reduced hemoglobin A1C levels (Ikeda et al. 2017; Ciancio et al. 2018). Some diabetic patients may even reduce the dose of antidiabetic medicines, if not suspending them altogether (Dawood et al. 2017; Ciancio et al. 2018), although it is unclear whether this benefit will be long-lasting. These benefits are independent of other factors such as presence of cirrhosis or BMI.

Cardiovascular Disorders

As far as cardiovascular outcomes are concerned, DAA-induced SVR seems to impact some surrogate markers of cardiovascular outcomes. In a series of 182 consecutive patients treated with DAAs (all SVR), intima-media thickness decreased significantly at FU, together with a reduction of the proportion of patients with carotid thickening (from 43% to 17%) (Petta et al. 2018). No changes were observed for carotid plaques. Stratification of patients for cardiovascular risk factors and liver disease severity did not affect the results. The investigators concluded that clearance of HCV improves carotid atherosclerosis, although they were cautious regarding the long-term clinical impact of these data.

Neurocognitive Impairment

The advent of IFN-α-free treatment regimens has allowed confirming the role of HCV in neuropsychiatric manifestations, which had for a long time been a major obstacle to IFN-α therapy. In a large study encompassing data collected in 1952 on patients enrolled in three landmark clinical trials, assessing the efficacy of sofosbuvir and ledipasvir with or without ribavirin (Younossi et al. 2015), HRQOL improved as quickly 2 weeks after initiation of treatment, in parallel with the inhibition of viral replication. For most patient-reported outcomes (PROs) measured by the SF-36 and the FACIT-F instruments, an improvement bigger than the minimal clinically important difference was observed in about one-half of treated subjects at the end of therapy. The persistence of HRQOL improvement after the end of therapy was definitely proven by the large placebo-controlled clinical trial using a combination of sofosbuvir and velpatasvir (Younossi et al. 2016b). PROs improvements persisted up to 24 weeks after the end of active treatment, although they were not noted among those who received placebo. The multivariable analysis, including clinical and demographic factors, also showed that treatment-emergent modifications of PROs were independently associated with the antiviral therapy, at least for some summary scores. Fatigue, depression, and anxiety all improved with successful therapy. Similar results were reported for other antiviral regimens (Younossi et al. 2016c; Back et al. 2019). It is important to mention that these data on the amelioration of depression symptoms are at odds with a recent meta-analysis assessing data on IFN-α-treated patients (Cacoub et al. 2018a), possibly because of the extensive difference among the treated populations. On the other hand, data on the effects of DAAs on brain metabolism are scanty. A magnetic resonance spectroscopy analysis of patients receiving sofosbuvir and ledipasvir showed the normalization of cerebral N-acetyl aspartate in patients in whom HCV was suppressed, suggesting a virally induced neuronal dysfunction (Alsop et al. 2014).

Non-Liver-Related Mortality

Data on non-liver-related mortality is already available. The retrospective analysis carried on the large ERCHIVES, U.S. VA database (Butt et al. 2019) identified 4436 chronic hepatitis C patients who had been treated with pegylated IFN-α and ribavirin and 12,667 treated with DAAs. Treated patients were matched for age, race, sex, and baseline values with untreated control patients. Patients with cardiovascular events at enrollment were excluded. The incidence of cardiovascular disorders was the lowest among those treated with DAAs, that is, 16.3 per 1000 patient-years, approximately one-half as many as observed in untreated controls, and about one-third less than in patients receiving pegylated IFN-α and ribavirin. Overall, SVR was associated with a 13% decreased risk of incident cardiovascular disorders, especially if treated with DAA.

Three prospective cohort studies also support these observations. First, the above-mentioned prospective French cohort (Carrat et al. 2019) showed that, after appropriate and thorough adjustments, SVR after DAAs was associated with a decrease in all-cause mortality, including both liver- and non-liver-related mortality. Importantly, because adjustments had been made for markers of liver failure and comorbidities, and a sensitivity analysis excluding the first 12 months of FU confirmed the overall results, the investigators could reasonably rule out a reverse causality, that is, the risk that patients may have been withheld DAAs if they presented with a significant risk of death from any cause other than liver-related.

The French ANRS/CirVir prospective study of 1323 patients with compensated, histologically proven cirrhosis (enrolled between 2006 and 2012) and treated with different regimens (including DAAs) analyzed SVR versus non-SVR patients matched for propensity scores and followed for a median of 58.2 months (Nahon et al. 2017). SVR patients had a reduced incidence of HCC (−71%), decompensation events (−74%) and cardiovascular events (−58%). The decrease in mortality was also significant (−73%), for both liver- and non-liver-related causes.

Finally, in a prospective cohort study of 878 patients with histologically proven cirrhosis without history of decompensation (Cacoub et al. 2018b) treated with a variety of regimens (including DAAs) and followed for a median of 57.5 months, 62 had incident major cardiovascular events. The multivariate analysis showed that SVR was an independent predictor of decreased risk of events (HR = 0.35, 95% CI 0.09–0.97; P = .044). Asian ethnic origin, arterial hypertension, smoking, and low serum albumin level were predictors of events. In addition, the 5-year survival was 60.1% in patients with cardiovascular events versus 87.5% in patients without (P < .001).

Evidence showing the specific effect of DAAs on other less frequent EHMs is scanty. Small sample size, retrospective study design, and lack of control of untreated arms characterize most cases series. Because DAA regimens are now used without restrictions in elimination strategies, data on less frequent EHMs will become common with time.

CONCLUSIONS

Several prospective studies using DAAs have now shown the benefit of viral clearance induced by these agents in terms of hard clinical outcomes, that is, liver- and non-liver-related mortality. The benefit is clear when compared with properly matched, untreated cohorts. With regard to EHMs, data suggest that viral clearance should be reached earlier than later. Thus, this adds to the current evidence supporting universal treatment, irrespective of disease stage. Considering also the benefit deriving from improved quality of life, reduced risk of onward transmission, and other social and economic gains, the overall social return from the resources invested in treating all HCV-infected patients (i.e., irrespective of their clinical profile) is unquestionable, and fully supports the WHO goal of eliminating HCV as a public health threat by the year 2030.