A high proportion of patients with HCV infection achieved SVRs. For patients who did not meet the CT criteria, treatment regimens must be optimized.
Download PDFOur real-world study is representative of monoinfected, non-transplanted patients and the treatment regimens available in Spain in 2015. Because the decision to treat and the choice of treatment were entirely at the discretion of the treating physician and randomization was not possible, this study could not directly compare the effectiveness and safety of the treatment regimens.
In the general cohort, the global efficacy was high (94.6% SVR) and the results were similar to those achieved in the CTs, although almost 60% of the patients had received previous HCV antiviral treatment and more than half had liver cirrhosis.
We found that 0.4% of the subjects who achieved a SVR at week 12 subsequently relapsed at week 24 (did not achieve SVR24), and this percentage was a similar to or even lower than those found in other studies[
16,
17]. Therefore, this finding confirmed previous results in a real-world setting and showed good concordance between SVRs at week 12 and week 24 based on different new AAD-based regimens, including those with shorter durations and/or with drugs with lower barriers to resistance. However, in our opinion, to definitively determine a “cure” in every patient in clinical practice, a SVR must be confirmed at week 24.
Until now, few real-world setting studies have included results that consider the most frequent genotypes (1 to 4). The most significant study is the US retrospective analysis of data from 17487 patients with genotypes 1 to 4 from the Veterans Affairs (VA) National Healthcare System[
18], in which a global SVR of 90.7% was found, which was lower than that in our study. This difference may be linked to early discontinuation of treatment in 4.4% of patients with available SVR data[
18].
In our study, albumin was the only independent predictor of a SVR. Other studies[
14,
18] have also shown that albumin and other variables associated with cirrhosis or worse liver function were related to a lower SVR, thus confirming these findings in a real-world setting and with a wide number of patients and supporting the results of CTs in which patients with a more advanced liver disease have a worse response to treatment.
Most real-world studies reported results in genotype 1 HCV patients[
14,
19,
20]. The SVR rate in our study, which included 362 genotype 1 patients, was 94.5% of the overall genotype 1 patients, which was somewhat higher than previously reported rates (SVRs over 91%), although limited differences were observed among the different DAA combinations, treatment durations and use of RBV. SMV and SOF with or without RBV was the most used treatment in our genotype 1 patients, which was likely because it was the best combination available at the beginning of the study. This treatment was used in 149 of the total genotype 1 patients. Most of these patients had liver cirrhosis and were included in the CT-unmet group because the most severe patients were prioritized. However, these patients achieved a SVR of 93.3%. In other studies with thousands of patients with genotype 1 HCV treated with this regimen, the SVR rates were lower at between 75% and 84%[
14,
15,
21]. The main cause of the differences between our cohort and the others was likely the lower rate of subtype 1a (31.2%) and Q80K variants in our genotype 1 patients. Although these variants were not analyzed in the current study, they appeared in only 2.7% of Spanish genotype 1 patients[
22].
Other treatment combinations also showed high rates of SVR in our study; i.e., 95.0% with SOF/LDV and 94.5% with OBV/PTV/r/DSV. These rates were similar to the 92.9% or 92% SVR rates derived from the first regimen presented in two US VA National Healthcare System studies[
18,
19] and the 94.9% or 95.1% SVR rates achieved with the second regimen in other studies in clinical practice[
18,
20].
In our cohort, only eleven genotype 2 patients were treated, and all of them achieved a SVR regardless of the treatment regimen used. High rates of SVR with the combination SOF + RBV were more similar to those described in Asian CTs[
23] than the SVR of 79.0% or 86.2% achieved in clinical practice in the two VA studies[
14,
18] or the SVR of 88.2% from the recent analysis of 321 genotype 2 HCV infected HCV-TARGET participants[
24]. However, the low number of genotype 2 patients in our study indicate that several of the currently recommended combinations in clinical guidelines, such as SOF and DCV[
25] should be favored because they presented 100% SVR rates in all patients.
Patients with HCV genotype 3 are at a higher risk of liver disease progression and hepatocellular carcinoma development[
26,
27]. However, compared with other HCV genotypes, DAA combinations have lower efficacy against genotype 3 in patients with liver cirrhosis in CTs.
In the current study, the global SVR in patients with genotype 3 HCV infection was 93.3%. In our cohort, 82.2% of patients with this genotype were treated with SOF and DCV, with a global SRV rate of 90.3%-91.9% in patients with liver cirrhosis and 100% without. In others studies in real-world settings, a global SVR of 60%-70% was achieved in genotype 3 infection with SOF plus RBV[
18,
28]. All these studies had remarkably low rates, which was likely related to the use of combinations that are currently not recommended because of their low efficacy[
25].
Patients with HCV genotype 4 infection are poorly represented in pivotal CTs of second-generation DAAs[
25] and in most real-world studies. In the VA study, a SVR of 87.6% with SOF and LDV and 96.4% with OBV and PTV/r was achieved in patients with this genotype[
18]. In the current study, 44 patients who were HCV genotype 4-infected were treated and the SVR rate was 95% (100% with SOF and LDV, 92.3% with OBV and PTV/r and 94.7% with SMV and SOF).
The week 4 response data were available for almost all patients in the current study. We found that 72.9% of patients had an undetectable HCV RNA at week 4, similar to another analysis[
19,
29]. In this last real-world setting study, significant SVR rate reductions of 7.1% to 10.5% according to the addition of RBV or not, respectively, were observed in patients who did not have an undetectable HCV RNA at week 4 compared with those with undetectable HCV RNA at week 4, which was similar to the 6% observed in the current study[
19]. The clinical implications of this finding on treatment decisions, such as potentially adding RBV or extending the treatment duration based on 4 wk of on-treatment HCV RNA, warrants further study.
Despite the real-world nature of our cohort, which included a higher proportion of elderly patients and many patients with liver cirrhosis, the safety and tolerability of all regimens were good. Discontinuation rates were low (< 1%), which is similar to that of CTs, and there were no deaths during treatment or follow up. In Backus et al[
20] higher early discontinuation rates of 5.3% to 15.2% according to the treatment combination were found. In contrast, of the 802 patients in the genotype 1 group from the HCV-TARGET cohort treated with SMV and SOF, the rate of discontinuation for adverse events was only 2%[
15].
In patients from the genotype 1 and genotype 3 groups from the HCV-TARGET cohort, the most commonly reported AEs were fatigue and headache, which is consistent with the results presented here[
15,
28]. However, anemia associated with RBV was less frequent in our study.
Overall, the reported rates of SAEs (2.4%) were similar to those reported in the pivotal CTs and lower than the 5.3% or the 7.3% described in other studies in “real-world”[
15,
28]. Again, in the three studies, the most frequent SAEs were the same decompensating events. However, in the current study, only seven of 262 cirrhotic patients experienced decompensation.
Because the real-world population is heterogeneous, it is important to investigate the treatment outcomes in patients excluded from CTs. Thus, we divided patients into two groups: Patients who met the requirements to take part in a CT and patients who did not meet these requirements. We found that the CT-unmet patients had lower rates of SVR and higher rates of SAEs, liver decompensation and treatment interruptions than the CT-met patients. Thus, in this group of patients, it might be advisable to conduct a more rigorous follow-up investigation to closely monitor tolerability and optimize treatment regimens.
This study has the usual limitations related to its observational, real-world design and electronic data collection. Resistance testing was not performed; thus, we were unable to assess the impact of this factor. The lack of randomization limited the ability to directly compare treatment groups, which is further compounded by the small number of patients in certain subgroups.
In conclusion, our study confirmed the efficacy and safety data reported in CTs in a cohort of patients with genotypes 1-4 and a wide range of basal characteristics, including a high proportion of patients with advanced fibrosis and treatment experience. Our results confirmed and occasionally improved upon the efficacy and safety results reported in other recently published real-world setting studies with a large number of patients[
8,
19], and these results are in sharp contrast to the lower SVR rates reported in certain early real-world studies on interferon-free therapy with second generation DAAs[
14,
15]. Moreover, our results indicate that treatment regimens should be optimized in patients that do not fulfill classical CT inclusion criteria because of their lower rates of SVR and higher rates of SAEs.
