Introduction

The introduction of interferon (IFN)-free regimens based on direct-acting antivirals (DAAs) for the treatment of chronic hepatitis C virus (HCV) infection has offered hope of eliminating the disease as a major public health threat by 2030. This was the goal set by the World Health Organization (WHO) in 2016, defined as a 90% reduction in new HCV infections and a 65% decrease in mortality from chronic hepatitis C (CHC) and its complications.1 The most recent achievement of the revolution in the treatment of HCV infection was the registration of highly potent pangenotypic options. These regimens offer a very good safety profile and high efficacy exceeding 95%, regardless of HCV genotype.2

The first registered DAA option with proven activity against all HCV genotypes was a combination of sofosbuvir (SOF), an inhibitor of viral nonstructural protein (NS) 5B, and velpatasvir (VEL), an inhibitor of NS5A HCV.3-5 It is worth noting that this regimen can be used in patients with decompensated liver cirrhosis.6 Another pangenotypic option, which is a combination of glecaprevir (GLE), an HCV protease inhibitor, and pibrentasvir (PIB), an NS5A blocker, is approved for use in patients without cirrhosis or those with compensated cirrhosis.7 The advantage of the GLE/PIB regimen is that it is registered in patients with severe renal insufficiency.8

Despite the overall excellent efficacy of DAA pangenotypic options documented in clinical trials and many real-world experience (RWE) studies, including those carried out in Poland, there is still a small percentage of patients who do not respond to the therapy.5,6,9-11 Previously published Polish studies, including that by Pabian et al,11 did not include the extremely difficult-to-treat subpopulation of patients who failed to respond to pangenotypic regimens simply because such patients appeared only recently.

The only regimen registered in patients after previous ineffective DAA therapy containing the NS5A inhibitors that are included in both the abovementioned pangenotypic options is a triple combination of SOF, VEL, and voxilaprevir (VOX), an HCV protease inhibitor.12 However, due to limited availability of this rescue option in routine clinical practice, other therapeutic regimens are also used in the patients who have failed prior pangenotypic treatment, including another triple combination of SOF+GLE/PIB or an even more frequent alternative, also pangenotypic regimen, which is extended previous therapy or addition of ribavirin (RBV).13,14 Unfortunately, up to now, RWE studies of pangenotypic therapy in pangenotypic failures were limited and included small numbers of patients. The aim of the current analysis, conducted in the RWE setting, was to evaluate the effectiveness of pangenotypic regimens in patients from the Polish population in whom prior treatment with a pangenotypic option was ineffective.

Patients and methods

The patients included in this study were selected from the EpiTer-2 database, which is the ongoing retrospective, multicenter, national, real-world register evaluating DAA-based antiviral treatment in individuals with chronic hepatitis C treated in 22 Polish hepatology centers. The database includes 15 123 patients treated for hepatitis C between 2015 and 2022. The current analysis included 4345 patients treated with all-oral pangenotypic regimens as the first-line therapy or retherapy after genotype-specific DAAs (PAN group), and 48 patients treated with pangenotypic regimens after pangenotypic failure (PAP group) between 2018 and 2021 (Table 1).

Table 1. Characteristics and comparison of patients receiving either 1 or 2 pangenotypic regimens

Parameter

PAP (n = 48)

PAN (n = 4345)

P value

Male sex

38 (79.2)

2304 (53)

<⁠0.001

Age, y

51.1 (40.7–58.7)

46 (37–59)

0.18

BMI, kg/m2

27.5 (25.7–30.1)

25.7 (22.9–28.7)

<⁠0.001

Genotype

1a

3 (6.3)

261 (6)

0.2

1b

16 (33.3)

2527 (58.2)

<⁠0.001

3

28 (58.3)

1187 (27.3)

<⁠0.001

4

1 (2.1)

262 (6)

0.01

Fibrosis

F0

1 (2.1)

129 (3)

0.2

F1

13 (27.1)

1955 (45.4)

0.004

F2

11 (22.9)

825 (19.2)

0.12

F3

1 (2.1)

476 (11.1)

0.02

F4

22 (45.8)

921 (21.4)

<⁠0.001

Liver stiffness, kPa

10.2 (6.9–23.6)

7.4 (5.5–12)

<⁠0.001

Past decompensation

4 (8.3)

97 (2.2)

0.02

MELD

7 (6–9)

7 (6–8)

0.09

CP>A

2 (4.2)

128 (3)

0.94

HCC history

2 (4.2)

48 (1.1)

0.19

OLTx history

0

6 (0.1)

0.09

HIV coinfection

8 (16.7)

453 (10.4)

0.24

HBV coinfection

8 (16.7)

616 (14.2)

0.78

HCV RNA × 106, IU/ml

1.2 (0.3–3.7)

1 (0.3–2.6)

0.65

Data are shown as number (percentage) of patients or median (interquartile range).

Abbreviations: BMI, body mass index; CP, Child-Pugh score; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; MELD, Model for End-Stage Liver Disease; OLTx, orthotopic liver transplantation

The selection of the antiviral regimen, including retherapy, was made by the treating physicians based on current national recommendations15-17 and the reimbursement policy by the National Health Fund (NHF). DAA doses and treatment duration were consistent with the Summary of Product Characteristics. The patients provided their informed consent before the start of the treatment according to the NHF requirements. Clinical and laboratory data were collected retrospectively and submitted via a web platform operated by Tiba Ltd. (Wrocław, Poland) following the National General Data Protection Regulation in Poland.

The information captured at baseline included demographic and clinical data, such as age, sex, body mass index (BMI), comorbidities and concomitant medications, the severity of liver disease, hepatitis B virus (HBV) and HIV coinfections, and the history of previous antiviral treatment. The degree of liver disease was evaluated noninvasively by transient elastography or shear wave elastography, or histologically by liver biopsy. The results were presented with reference to fibrosis stage F0-4 according to the METAVIR score, using the recommendations of the European Association for the Study of the Liver (EASL) with 13 kPa as a threshold to define liver cirrhosis.18 The patients with cirrhosis were scored on the Child–Pugh (CP) scale and Model of End-Stage Liver Disease (MELD).

The laboratory parameters analyzed included the activity of serum alanine transaminase (ALT), the concentrations of bilirubin, albumin, creatinine, and hemoglobin, white blood cell and platelet counts, and HCV viral load. The HCV RNA was assessed by real-time polymerase chain reaction assays at baseline, at the end of the treatment, and at least 12 weeks after the therapy completion. The intention-to-treat group (ITT) included the patients who received at least 1 dose of an antiviral drug, and the per-protocol (PP) group was established by excluding the patients due to a nonvirologic failure.

Statistical analysis

The results were expressed as median (interquartile range [IQR]) or number (percentage). A P value below 0.05 was considered significant. The significance of differences was calculated by the χ2 or the Fisher exact test for nominal variables, and by the Mann–Whitney test and the Kruskal–Wallis analysis of variance for continuous variables. For nominal variables with more than 2 categories the P values were corrected for multiple testing with the Bonferroni test. Univariable comparisons were calculated using the GraphPad Prism 5.1 software (GraphPad Software, Inc., La Jolla, California, United States).

Results

Characteristics of the study population

A total of 48 patients not responding to the pangenotypic treatment received retherapy with another pangenotypic regimen (Table 1). The patients treated twice with the pangenotypic regimens (PAP group) were predominantly men (79% vs 53%; P <⁠0.001), had higher BMI (27.5 [IQR, 25.7–30.1] vs 25.7 [IQR, 22.9–28.7] kg/m2; P <⁠0.001), were more often infected with HCV genotype 3 (58% vs 27%; P <⁠0.001), and more frequently had liver cirrhosis (46% vs 21%; P <⁠0.001) than the patients who received 1 course of the pangenotypic treatment (PAN group). On the other hand, the frequency of coinfections with HBV, HIV, the history of hepatocellular carcinoma and the liver cirrhosis stage expressed by CP and MELD scales as well as baseline HCV viral loads were comparable.

Importantly, after selection of the commonly accepted the hardest-to-treat population4,6,9,10 in the PAP and PAN groups, which comprised men with cirrhosis infected with HCV genotype 3, only BMI was slightly higher in the PAP cohort (29.7 [IQR, 27.4–34.0] vs 27.8 [IQR, 25.4–30.8] kg/m2; P = 0.06) (Table 2). Furthermore, the distribution of the first pangenotypic regimens was comparable in the PAP and PAN groups, with baseline VEL/SOF schedules in 50% and 41% of the patients, and GLE/PIB in 50% vs 59% of the patients, respectively. Detailed characteristics of baseline DAA regimens are shown in Table 3.

Table 2. Characteristics and comparison of difficult-to-treat subpopulation of men with cirrhosis infected with genotype 3 and treated either with 1 or 2 pangenotypic regimens

Parameter

PAP (n = 14)

PAN (n = 209)

P value

Age, y

55.5 (50.7–58.2)

51 (44–58)

0.16

BMI, kg/m2

29.7 (27.4–34)

27.8 (25.4–30.8)

0.06

Liver stiffness, kPa

21.1 (15.4–35)

23.6 (16.8–36.4)

0.74

Past decompensation

1 (7.1)

15 (7.2)

>0.99

MELD

8 (7–9.2)

8 (7–10)

0.63

CP>A,

1 (7.1)

12 (5.7)

0.58

HCC history

1 (7.1)

10 (4.8)

0.51

OLTx history

0

0

–

HIV coinfection

2 (14.3)

16 (7.7)

0.31

HBV coinfection

2 (14.3)

37 (17.7)

>0.99

HCV RNA × 106, IU/ml

1 (0.4–2.7)

0.9 (0.2–2.1)

0.52

Data are shown as number (percentage) of patients or median (interquartile range).

Abbreviations: see Table 1

Table 3. Regimens administered to the patients with double and single pangenotypic therapy

Regimen

PAP (n = 48)

PAN (n = 4345)

Previous therapy

Retreatment

VS12a

13 (27.1)

5 (10.4)

1607 (37)

VS24

1 (2.1)

7 (14.6)

8 (0.2)

VSR12

5 (10.4)

2 (4.2)

101 (2.3)

VSR24

5 (10.4)

10 (20.8)

56 (1.3)

VVS8

0

1 (2.1)

0

VVS12

0

6 (12.5)

4 (0.1)

Total VS based

24 (50)

31 (64.6)

1776 (40.9)

GP8

17 (35.4)

2 (4.2)

2202 (50.7)

GP12

3 (6.3)

9 (18.7)

313 (7.2)

GP16

4 (8.3)

3 (6.2)

50 (1.2)

GPSR12

0

1 (2.1)

4 (0.1)

GPSR16

0

2 (4.2)

0

Total GP based

24 (50)

17 (35.4)

2569 (59.1)

Data are shown as number (percentage) of patients.

a The numbers after the regimen abbreviation denote the therapy duration in weeks.

Abbreviations: GP, glecaprevir / pibrentasvir; GPSR, glecaprevir / pibrentasvir / sofosbuvir / ribavirin; VS, velpatasvir / sofosbuvir; VSR, velpatasvir / sofosbuvir / ribavirin; VVS, velpatasvir / voxilaprevir / sofosbuvir

The effectiveness of the second pangenotypic regimen after the failure of the first pangenotypic therapy

The regimens used for retherapy in the cases of failed pangenotypic therapy were more often based on VEL/SOF than on GLE/PIB (65% vs 35%), and generally lasted longer. Short-term (12 weeks) VEL/SOF therapy was applied in only 10% of cases, while 8-week long GLE/PIB treatment was administered in 4% of the patients. Of 48 patients in whom the pangenotypic therapy failed, 4 did not respond to the second pangenotypic therapy either, while 1 patient was lost to the follow-up. The efficacy of the double pangenotypic therapy in our cohort was 89.6% in ITT analysis, and 91.5% in PP analysis. Clinical characteristics of the 4 patients in whom the double pangenotypic therapy failed showed a distinctive pattern and are presented in Table 4. All of them were men aged over 50 years, with liver cirrhosis and high liver stiffness (range, 17.5–37.9 kPa) caused by HCV genotype 3 infection, 2 of them suffered from obesity, and 2 from overweight. Moreover, they were treated with different pangenotypic regimens as a second therapy, depicted with red lines in Figure 1. All these patients responded initially to the therapy, but relapsed after the end of the treatment (Figure 1).

Table 4. Characteristics of patients who failed the double pangenotypic therapy

Characteristics

Patients

1

2

3

4

Age

58

56

54

59

Sex

Male

Male

Male

Male

BMI, kg/m2

33.21

36.3

29.07

29.14

Genotype

3

3

3

3

Retreatment regimen

VSR12

VSR24

GP12

VS24

Type of current failure

Relapse

Relapse

Relapse

Relapse

Previous regimen

GP12

GP16

VS12

GP16

Type of previous failure

Relapse

Relapse

No response

Relapse

Very first regimen

None

SOF+RBV

SOF+RBV

SOF+PegIFN+RBV

Liver stiffness, kPa

37.9

18.6

31.8

17.5

Metavir estimation

4

4

4

4

History of decompensation

No

No

Yes

No

HCC history

No

No

No

No

HIV coinfection

No

No

No

No

HBV coinfection

No

No

No

No

OLTx history

No

No

No

No

Baseline HCV RNA, IU/ml

1 270 000

1 080 000

8 660 000

4 820 000

ALT, IU/l

36

111

232

44

MELD score

10

8

8

8

CP

A

A

A

A

HCV RNA at the EOT

Undetectable

Undetectable

Undetectable

Undetectable

Abbreviations: ALT, alanine aminotransferase; EOT, end of therapy; PegINF, pegylated interferon; RBV, ribavirin; SOF, sofosbuvir; others, see Tables 1 and 3

Figure 1. Selection of a rescue pangenotypic regimen after failure of previous pangenotypic therapy. Red lines correspond to the failure of the rescue therapy.

Abbreviations: see Table 3

Importantly, treatment effectiveness calculated in both ITT and PP analyses in the PAP patients was not statistically inferior to patients with the PAN regimen (Table 5). In detail, sustained virologic response (SVR) rates for the PAP group were 89.6% (ITT) and 91.5% (PP) vs 93.7% (ITT; P = 0.39) and 97.6% (PP; P = 0.17) in the PAN group (Figure 2). Similarly, no difference in the treatment effectiveness between the PAP and PAN groups was noted for the subgroups of patients (men, genotype 3, genotype 4) in the ITT analyses. Only in the PP analyses, SVR was statistically inferior in the PAP vs PAN group for men, patients infected with genotype 3, and those with liver cirrhosis (Table 6). Importantly enough, in the majority of the PAP group, SVR (ITT or PP) exceeded 80%, being lower only when all of the negative factors for response were present, that is, genotype 3, fibrosis score 4, and male sex (SVR, 71.4% in ITT and PP).

Table 5. Sustained virologic response in patients with double and single pangenotypic therapy depending on the type of the administered regimen

Parameter

ITT

PP

PAP

PAN

P value

PAP

PAN

P value

GP-based regimen

16/17 (94.1)

2451/2569 (95.4)

0.74

16/17 (94.1)

2451/2495 (98.2)

0.72

VS-based regimen

27/31 (87.1)

1619/1776 (91.2)

0.35

27/30 (90)

1619/1677 (96.5)

0.09

P value

0.64

<⁠0.001

>0.99

<⁠0.001

Dual regimens

33/38 (86.8)

4062/4337 (93.7)

0.17

33/37 (89.2)

4062/4164 (97.6)

0.007

Triple regimensa

10/10 (100)

8/8 (100)

–

10/10 (100)

8/8 (100)

–

P value

0.57

0.99

0.56

0.49

Data are shown as number of patients with sustained virologic response / total number of patients (percentage).

a Among 10 PAP patients, 5 were infected with genotype 3, 7 were cirrhotic, and 9 were men, and of the 8 PAN patients, all were genotype 1a or 1b infected, none had cirrhosis, and 5 were men.

Abbreviations: see Table 3 and Figure 2

Figure 2. Sustained viral response in patients retreated after pangenotypic failure with another pangenotypic regimen (PAP) vs those treated for the first time with a pangenotypic regimen (PAN)

a P <⁠0.05

Abbreviations: GT3, genotype 3; F, fibrosis; ITT, intention-to-treat; Male/GT3/F4, men, hepatitis C virus genotype 3, and fibrosis 4; PP, per-protocol (lost to follow-up excluded); SVR, sustained virologic response

Table 6. Sustained virologic response in patients with double and single pangenotypic therapy depending on the factor affecting efficacy

Parameter

ITT

PP

PAP

PAN

P value

PAP

PAN

P value

All

43/48 (89.6)

4070/4345 (93.7)

0.39

43/47 (91.5)

4070/4223 (97.6)

0.17

Women

10/10 (100)

1950/2041 (95.5)

0.93

10/10 (100)

1950/1980 (98.5)

>0.99

Men

34/38 (89.5)

2120/2304 (92.0)

0.79

34/38 (89.5)

2120/2192 (96.7)

0.04

G3

23/28 (82.1)

1081/1187 (91.1)

0.17

23/27 (85.2)

1081/1130 (95.7)

0.03

F4

18/22 (81.8)

831/921 (90.2)

0.26

18/22 (81.8)

831/881 (94.3)

0.04

F4, HCV GT3, men

10/14 (71.4)

174/209 (83.3)

0.27

10/14 (71.4)

174/199 (87.4)

0.09

Data are shown as number of patients with sustained virologic response / total number of patients (percentage).

Abbreviations: see Figures 2 and 3

The influence of a specific regimen on the effectiveness of treatment in the patients in whom the pangenotypic therapy failed

The selection of a retherapy regimen after the first pangenotypic therapy had failed did not seem to affect SVR. In both analyses, GLE/PIB and VEL/SOF-based therapies showed comparable effectiveness (94.1% vs 87.1; P = 0.64 for ITT, and 94.1% vs 90%; P >0.99 for PP). Interestingly enough, this was in contrast with the patients from the PAN group in whom GLE/PIB-based regimens showed higher efficacy in both ITT (95.4% vs 91.2%; P <⁠0.001) and PP (98.2% vs 96.5%; P <⁠0.001) analyses (Table 5). This could be a result of a selection preference for VEL/SOF in the patients with more advanced liver disease. Accordingly, the selection of the last regimen, that is, GLE/PIB vs VEL/SOF, showed comparable efficacy in the patients with 1 or 2 failures of the previous therapy.

Discussion

The very high overall success rate of DAA therapy, exceeding 95%, has made the problem of nonresponse to treatment in CHC marginal. However, recent observations show that there are still patients with a lower chance of therapeutic success.19-21 Effective retreatment options for the patients who did not respond to DAA therapy are a key prerequisite for achieving the WHO’s goal of HCV elimination. This retrospective analysis is a part of our national observational study that demonstrated evolution of the Polish HCV-infected population.22 We documented an SVR of 97.6% in the RWE population treated with a single course of pangenotypic regimens, and a 91.5% success rate among those receiving double pangenotypic treatment. The population of 48 patients who failed the first pangenotypic regimen comprised mostly men (79%) and patients infected with genotype 3 (58%); almost half of them were diagnosed with liver cirrhosis. All these features were present in 14 patients accounting for 30% of those nonresponding to the pangenotypic options, and it was from this group that all 4 patients who failed to respond to the retreatment with pangenotypic regimen came from. Contrary to that, such patient characteristics were found in only 5% of the group treated with a single course of pangenotypic therapy.

Genotype 3-infected individuals with liver cirrhosis constitute a specific last line of HCV defense, even in the DAA era.23-25 At the beginning of the IFN-free period, they were considered the most difficult-to-treat patients, because of the suboptimal efficacy of the only DAA regimen available for this population, that is, a combination of SOF and RBV.5,26 The efficacy of antiviral treatment for genotype 3-infected cirrhotic patients has improved significantly with the availability of pangenotypic regimens but still lags behind the results obtained in clinical trials in patients with cirrhosis caused by infection with other HCV genotypes, as confirmed by the current analysis.5,6,27

One half of the 48 patients retreated with pangenotypic regimens received in the first course a SOF/VEL-based option, and the other half a GLE/PIB-based combination. There were 3 nonresponders to SOF/VEL and 1 to GLE/PIB-based regimen among those in whom the second pangenotypic therapy failed. Thus, the retreatment with SOF/VEL-based option resulted in an 87% SVR rate, whereas GLE/PIB containing salvage therapy was effective in 94%, but the difference was not significant.

One hundred percent efficacy was demonstrated with pangenotypic regimens combining 3 viral enzyme inhibitors. Among 10 effective pangenotypic triple regimens, 7 were SOF/VEL/VOX combinations, the only option registered for those who failed NS5A-containing DAA treatment.12 Its high overall 96% efficacy in the patients in whom NS5A-based regimens failed was confirmed in the POLARIS-1 trial.12 Twenty individuals previously treated with SOF/VEL option had a response rate of 90%. A larger group of 47 such patients was analyzed in a real-world study conducted by Belperio et al28 and the efficacy of the regimen was 83%, with the lowest rate for genotype 1 infection. In an RWE study evaluating a cohort of Spanish patients, efficacy in an 8-people group who received SOF/VEL/VOX retherapy after SOF/VEL treatment failure was reduced to 87% due to nonresponse in a single patient, who was a man with genotype 3 infection and liver cirrhosis.29 This is consistent with our findings on the difficulty treating patients with these characteristics. The efficacy of SOF/VEL/VOX retherapy in patients after unsuccessful SOF/VEL treatment in larger RWE populations, including 19 to 31 patients, has reached from 94% to 100%.30-32 It should be noted that the POLARIS-1 clinical trial did not include patients in whom GLE/PIB therapy failed, and only RWE studies are available for this patient population retreated with SOF/VEL/VOX.12 According to their results, the efficacy of this triple pangenotypic option was 91% to 100% in the groups of 10 to 46 patients.32-34 Only Flamm et al30 reported an SVR of 75%, but the GLE/PIB failure population was small, consisting of only 4 patients, 1 of whom did not respond to SOF/VEL/VOX retherapy, thus significantly reducing the efficacy.

While SOF/VEL/VOX combination is recommended as the first-line rescue therapy in SOF/VEL and GLE/PIB nonresponders, another triple pangenotypic regimen used in our analysis with 100% efficacy, GLE/PIB+SOF with RBV addition is listed as a salvage option for GLE/PIB and multiple DAA treatment failures, including SOF/VEL/VOX.35,36 This combination was evaluated in the Magellan-3, phase 3 clinical trial with an overall SVR of 96%.37 All 23 participants of this study had been previously treated with a GLE/PIB regimen, while in our study 1 patient was GLE/PIB and 2 were SOF/VEL treatment-experienced. RWE data assessing GLE/PIB+SOF±RBV retreatment option are scarce. A 77% success rate in 13 SOF/VEL/VOX nonresponders was documented in a combined multicenter analysis involving patients from Germany, Spain, and Italy. Also, a dual-center case series study from the United States reported 6 successfully treated patients infected with genotype 1 or 3, with a history of multiple DAA failures, including SOF/VEL and GLE/PIB regimens; all were diagnosed with liver cirrhosis, 5 scored as A and 1 as B on the CP scale.13

Although triple salvage therapies are recommended for use in the patients after failed treatment with pangenotypic regimens, the limited availability of these options (reimbursement of SOF/VEL/VOX was introduced in Poland only in mid-2021), resulted in most patients in our analysis (79%) being assigned to dual regimens, a SOF/VEL or GLE/PIB combination. The retreatment strategy was based on switching the regimen, or if the same option was used, extending the duration of the therapy and / or RBV addition. The treatment with dual pangenotypic regimens was effective in 89% of the patients who failed a previous course of such an option, as compared with 97.6% efficacy in those who received 1 course of the pangenotypic combination in our analysis. We found significant differences in patient characteristics between these 2 groups, with significantly higher proportion of men, genotype 3-infected patients, and patients with cirrhosis, including those with decompensation among individuals retreated with the pangenotypic regimens.

According to the recommendations issued by the American Association for the Study of the Liver, the GLE/PIB regimen is listed as an alternative option for patients in whom SOF-based therapy has failed.35 For those after SOF/VEL failure, this recommendation is based on the results of the phase 3b randomized study carried out by Lok et al38, with an overall SVR of 93%, which comprised 10 patients after ineffective SOF/VEL therapy. However, it should be noted that GLE/PIB option is not recommended in patients after the failure of the same regimen.35 The EASL guidelines are even more strict, indicating that this regimen in retherapy is not recommended after the failure of any DAA regimen, especially those containing an NS5A inhibitor, due to insufficient barrier to resistance.36 According to current EASL recommendations, the choice of retherapy after failed DAA treatment can be optimized based on resistance analysis. We are aware that the lack of evaluation of resistance-associated substitutions (RASs) before the second pangenotypic treatment is one of the limitations of our study. The selection of RASs, particularly of NS5A inhibitor, may adversely affect the effectiveness of the retherapy, although this effect has not been demonstrated for pangenotypic triple regimens, being an issue in patients in whom a therapy with multiple DAAs has failed.39,40 Other limitations of our study are related to its retrospective nature with the potential bias and data entry errors. However, its main strengths are the collection of data from a real-world heterogeneous patient population from different regions of the country and the large size of the analyzed population treated with pangenotypic regimens.

Conclusions

The study documented the excellent effectiveness of pangenotypic regimens and demonstrated that most patients in whom these options were ineffective could be successfully retreated with another pangenotypic therapy. The best retreatment strategy is a triple pangenotypic regimen, especially in the patients with unfavorable response factors, such as HCV genotype 3 infection, cirrhosis, and male sex.