Hepatitis C Treatment Response Kinetics and Impact of Baseline Predictors
M. Lindh; B. Arnholm; A. Eilard; M. Färkkilä; K. Hellstrand; M. Lagging; N. Langeland; K. Mørch; S. Nilsson; C. Pedersen; M. R. Buhl; T. Wahlberg; R. Wejstål; J. Westin; G. Norkrans
Authors and Disclosures
Posted: 07/11/2011; J Viral Hepat. 2011;18(6):400-407. © 2011 Blackwell Publishing
Abstract and Introduction
Abstract
The optimal duration of treatment for hepatitis C virus (HCV) infections is highly variable but critical for achieving cure (sustained virological response, SVR). We prospectively investigated the impact of age, fibrosis, baseline viraemia and genotype on the early viral kinetics and treatment outcome. Patients treated with peginterferon alfa-2a and ribavirin in standard dosing were included: 49 with genotype 1 treated for 48 weeks and 139 with genotype 2 or 3 treated for 24 weeks. The reduced SVR rates in patients older than 45 years, with severe liver fibrosis or pretreatment viraemia above 400 000 IU/mL were strongly associated with slower second phase declines of HCV RNA. Genotype 2/3 infections responded more rapidly than genotype 1, reaching week 4 negativity (RVR) in 59%vs 22%. We conclude that baseline response predictors such as age, fibrosis and viral load were well reflected by the early viral kinetics as assessed by repeated HCV RNA quantifications. The kinetic patterns and the high relapse rate in genotype 2/3 patients without RVR suggest that this group might benefit from treatment durations longer than 24 weeks.
Introduction
Treatment of chronic hepatitis C with pegylated interferon and ribavirin cures the infection in 30–90% of the patients depending on baseline parameters.[1–3] The most important response predictor is the viral genotype, and patients with the poorly responsive genotypes 1 or 4 are recommended longer treatment duration than patients with genotypes 2 or 3. Patients with low baseline levels of hepatitis C virus (HCV) RNA, age below 40–50 and absence of significant fibrosis are more likely to achieve a sustained virological response (SVR), but it is unclear how this knowledge should influence the therapeutic regimen.
It has long been known that the viral kinetics, both the initial first phase decline[4] and the decline over the following weeks[5–10] predict the outcome of treatment. It has been difficult, though, to translate this knowledge into clinically useful guidelines, with the exception of the decision to stop treatment if HCV RNA has not decreased by at least 2 logs after 12 weeks.[11] However, there is a trend towards more flexible response-guided treatment (RGT) regimens, in which the duration of treatment is directed by the virological response, but as yet these regimens are based mainly on loss of HCV RNA after 4 weeks of therapy. Thus, it has recently been recognized that genotype 1 infections with undetected HCV RNA after 4 weeks of treatment (rapid virological response, RVR) and low baseline viraemia may be treated for 24 weeks instead of 48 weeks,[12–15] and that selected patients with genotype 2 or 3 infections may be treated for 12–16 weeks,[16–19] in particular, those who reach HCV RNA levels below 1000 IU/mL after 7 days of treatment.[20] It has also been shown that prolonged treatment duration increases the likelihood of SVR for patients with genotype 1 who show slow virological responses.[14,21–23] It is however uncertain to what extent treatment guidelines based on RGT should take into consideration also baseline factors, or if these factors can be presumed to be part of the viral kinetics.
In two recent multicentre studies, we investigated how the early virological response was related to outcome of treatment of genotype 1 or genotype 2/3 infection.[9,20] In this study, we have further analysed merged data from these two trials with the aim of clarifying how the baseline factors age, fibrosis, HCV RNA level and genotype are related to viral kinetics and outcome. Our results indicate that all these factors influence the viral kinetics, and that it might be possible to use an algorithm for RGT that is independent of baseline factors
Materials and Methods
Patients
All patients were treated with a combination of 180 μg of pegylated interferon alfa-2a once weekly and oral ribavirin, given to genotype 2/3 patients at a fixed dose, or to genotype 1 patients at the dose 1000 or 1200 mg daily, depending on whether body weight at inclusion was below or above 75 kg. The KinG1 trial[9] included 53 patients with genotype 1, who were treated for 48 weeks (unless the HCV RNA testing after 12 weeks showed <2 log reduction, or if HCV RNA was detected after 24 weeks of treatment). In this study, we included the 49 per protocol patients from that study, i.e. patients who had taken at least 80% of the planned medication (dose and time) and who were sampled sufficiently for analysing the virological response.
The NORDynamIC trial[20] included 382 patients with genotype 2 or 3, who were randomized to either 12 or 24 weeks. From that study, we included in this study, all the 139 patients in the 24-week arm, who met the same per protocol criteria as for genotype 1. Both studies were approved by the regional ethics committee and the patients gave written informed consents to participate.
HCV RNA
Hepatitis C virus RNA levels before, during and after treatment were determined by the Cobas Taqman assay (Roche Diagnostics, Branchburg, NJ, USA), which has a lower detection limit of 15 IU/mL. All samples were analysed at the Clinical Virology Laboratory at the Sahlgrenska University Hospital, Gothenburg, Sweden. In the genotype 1 patients, HCV RNA levels were analysed at baseline, after 4, 7, 14, 21, 28 days, and 8, 12, 16 and 24 weeks of treatment, while the genotype 2/3 study patients levels were analysed at baseline, after 4, 7, 28 days, and 8 and 12 weeks of treatment. The second phase slopes (weekly decline) were obtained from linear regression analysis including all HCV RNA–positive results from day 4 to week 12 in the KinG1 study. The slope in the NORDynamIC study was assessed as follows: If HCV RNA was detectable at week 4 (n = 57), all the three values (day 4, 7 and 28) were included in linear regression. If HCV RNA was negative at week 4, all three time points were used with the value for day 28 set at 0.5 log IU/mL (n = 23), unless the slope based on only day 4 and 7 was steeper (n = 56). In the latter cases, the slope was based on values from day 4 and 7 only. Three cases were excluded from slope analysis because the baseline level was very low (less then 1500 IU/mL). SVR was defined as a negative HCV RNA result by Cobas Taqman 24 weeks after end of treatment.
Fibrosis Score
Liver biopsies were histologically scored according to the Ludwig-Batts system[24] in the KinG1 trial, and the Ishak system[25] for the NORDynamIC patients.
Statistics
Continuous data were compared by Mann–Whitney ranks sum test and group data by Fischer's exact test. Multivariate analysis was performed by logistic regression with baseline HCV RNA (above or below 400 000 IU/mL), age (above or below 45), fibrosis (mild or severe) and second phase decline (above or below 0.35 log/week for genotype 1, above or below 1.0 log/week for genotype 2/3) as independent variables and SVR as dependent variable, and also by multiple regression using continuous values for baseline HCV RNA and age as dependent variables. The cut-off 0.35 log/week was chosen for genotype 1 because it distinguished SVR and non-SVR well. The cut-off 1.0 log/week was the median slope value for genotype 2/3, and was chosen because there was no natural cut-off.
Results
Baseline HCV RNA Levels, Age and Fibrosis
The relations between HCV RNA levels, age and fibrosis at baseline were analysed in the KinG1 study subjects, in whom the viral kinetics had been closely monitored. The impact on SVR from these baseline predictors was closely related to the rate of the second phase decline in the different groups. As shown in Fig. 1, the poorer response in patients older than 45 years or with severe fibrosis was caused by larger proportions with slow viral kinetics. Patients with pretreatment HCV RNA levels below 400 000 IU/mL showed rapid declines in HCV RNA, with 8/10 becoming HCV RNA negative within 2 or 3 weeks. In comparison, patients with higher baseline viraemia levels showed a wide range of virological responses, with a strong association between the second phase viral decline and SVR.
(Click To Enlarge)
Figure 1.
Individual response patterns in HCV genotype 1 patients with respect to age, fibrosis and HCV RNA level at baseline. Green plot, SVR, blue plot, non-SVR. Dotted line, lower limit of detection
Table 1 ummarizes the impact of the second phase slope on SVR rates. The favourable outcome in patients less then 45 years (73% SVR) was a result of a second phase decline more then 0.35 log IU/week in the 11 patients with SVR. Similarly, of the 31 patients with fibrosis scores 0–2, SVR was achieved in 23 patients (74%), of whom 22 had declines steeper than 0.35 log IU/week. Finally, all the 10 patients with baseline HCV RNA <400 000 IU/mL achieved SVR, and they all showed declines steeper than 0.35 log IU/mL.
When investigated at the individual level, it was obvious that subjects with favourable baseline status (age less then;45, HCV RNA less then;400 000 IU/mL and/or fibrosis score 0–2) did not achieve SVR unless they demonstrated a sufficient second phase viral decline (Fig. 1). Conversely, some patients with unfavourable baseline status achieved SVR if the decline was rapid. Thus, among the 34 patient older than 45 years, all the 19 who had a second phase slope steeper than 0.35 achieved SVR. Similarly, among 13 patients with fibrosis scores 3–4, all the five patients with slopes steeper than 0.35 log IU/mL achieved SVR, including three who were older than 45 years and had HCV RNA levels above 400 000 IU/mL.
When SVR was related to the four parameters age, baseline HCV RNA levels, fibrosis score and second phase decline in multivariate analysis, only the second phase decline showed a significant association with SVR.
Genotype
To investigate if also the strongest baseline factor, genotype, might mediate impact on the viral kinetics, we compared the early viral kinetics from patients with genotype 1 and 2/3. In the genotype 2/3 trial, HCV RNA levels were not analysed at day 14 or 21 when many patients might have reached undetectable levels. Still, this comparison clearly showed that genotype 2/3 infections responded more rapidly than genotype 1. In particular, the reduction in HCV RNA level during the first week was more pronounced in genotypes 2/3 than genotype 1. Also the second phase decline rates in genotype 2/3 patients were steeper than in genotype 1 patients. As shown in Fig. 2, patients who reached RVR (and SVR) showed rapid declines in HCV RNA. The decline rates for genotype 2/3 patients who achieved SVR without first reaching RVR were not significantly faster than the declines in genotype 2/3 patients with relapse (after 24 weeks of treatment) or genotype 1 patients without RVR who achieved SVR (after 48 weeks of treatment). Interestingly, the latter two groups had almost identical patterns of viral decline. Patients with genotype 1 who did not achieve SVR showed markedly slower response profiles
(Click To Enlarge)
Figure 2.
Viral kinetics (median values) in HCV genotype 1 patients (n = 49) treated for 48 weeks and 2/3 patients (n = 139) treated for 24 weeks with different treatment outcomes. Dotted line, lower limit of detection
In Table 2, the genotype differences in viral kinetics are presented in detail. In particular, the HCV RNA reduction from baseline during the first week was greater in genotypes 2/3 than in genotype 1 (median 2.49 vs 1.01 log IU/week), presumably reflecting a greater inhibition of viral replication (antiviral efficiency). As a result, patients with genotypes 2/3, when compared to genotype 1, had lower HCV RNA levels (median 3.58 vs 5.50 log IU/mL) after the first treatment week, i.e. before the second phase decline that is presumed to represent eradication of HCV-infected hepatocytes. There was a significant association between the reduction of HCV RNA after 1 week and the subsequent second phase decline rate for both genotype 1 and 2/3 (P less then; 0.0001 for both groups).
When comparing kinetics in patients with genotypes 1 and 2/3 with or without SVR one must keep in mind that treatment duration was different (48 and 24 weeks, respectively). Still, some observations were interesting. Genotype 2/3 patients with RVR had greater first week reductions (2.97 vs 2.41 log IU/mL) than genotype 1, but the proportion of RVR patients that had levels below 1000 IU/mL after 1 week was similar (50%vs 46%). In the subgroup without RVR, on the other hand, the genotypes differed: In genotype 1, those that achieved SVR had much better kinetics than the relapsers (first week decline 1.11 vs 0.65 log IU/mL, second slope 0.74 vs 0.27).
Genotype 2/3 patients with SVR differed less from those with relapse (first week decline 1.88 vs 1.73 log IU/mL, second slope 0.83 vs 0.80), but relapsers had significantly higher HCV RNA levels after 1 week (P = 0.0064) and 4 weeks (P = 0.05) than SVR patients (Fig. 2). As shown in Table 3, RVR was more common in genotype 2/3 than in genotype 1 patients (59%vs 22%). Non-response (HCV RNA never becoming undetectable) was observed in 32% of genotype 1 patients without RVR, but was rare (5%) in genotype 2/3 patients without RVR, among whom relapse instead was common (28%). In patients without RVR, the rate of non-SVR was 47% in genotype 1 and 33% in genotypes 2/3.
The impact of baseline factors on treatment outcome was similar for genotype 2/3 and genotype 1 (Table 4). A baseline viraemia below 400 000 IU/mL was highly favourable (100% SVR), but was present in only a minority of the patients (16% for genotype 1, 25% for genotype 2/3). An age below 45 was also favourable (94% SVR vs 75% for genotype 2/3, P = 0.002; 73%vs 59% for genotype 1, P > 0.1). However, for the large fraction of patients with mild fibrosis and high baseline viraemia levels (50% in both genotype groups), the impact of age below 45 on SVR was lower: 78%vs 68% for genotype 1, 89%vs 81% for genotype 2/3.
Discussion
The virological response patterns during treatment of hepatitis C are highly variable, and more flexible treatment strategies are required.[8,11,26] An important step in this direction was to shorten treatment with peginterferon and ribavirin to 24 weeks for rapid responders (RVR; negative HCV RNA at week 4)[12–15] and extend treatment to 72 weeks for slow or 'partial early' responders (pEVR; with detectable HCV RNA at week 12).[21–23] Similarly, for genotype 2 and 3 infections, it has been proposed that treatment may be abbreviated from 24 to 12–16 weeks in patients with very rapid response.[16,19,20,27] It is however not well known whether these strategies should consider baseline factors that are associated with the likelihood of SVR.
The results from this study show that the impact of baseline factors like age, fibrosis and baseline viral load on SVR is closely related to the viral kinetics during treatment. The observation suggests that treatment of genotype 1 infection might be directed by the early kinetics independently of age, baseline HCV RNA level or fibrosis stage. Our findings largely agree with the observations by Ferenci et al. and Mangia et al., who found that SVR was not influenced by age or fibrosis in genotype 1 patients with RVR.[12,14] On the other hand, several studies indicate that a low baseline HCV RNA (below 400 000–800 000 IU/mL) is important for the chance of achieving a SVR for genotype 1 patients with RVR.[12,15,28,29] Our observations suggest that baseline HCV RNA might not be important for SVR if the early response is categorized by the second phase decline rather than by a negative HCV RNA test result at week 4. Recent mathematical modelling data indicate that higher antiviral efficiencies (first week reductions) are required for patients with a large fraction of infected hepatocytes and/or a high baseline HCV RNA level.[30] However, it is not clear if these factors are interrelated or how they should be considered in response-guided treatment regimens.
We attempted to evaluate whether genotype, which is the strongest baseline predictor of response, may also be considered a component of the viral kinetics. This analysis was weaker, because we compared patients from two different trials, with different sampling protocols. Still, it was clear that the favourable response in patients infected with genotype 2 or 3 was visible in the early kinetics. Thus, in agreement with previous findings,[26,31,32] there was a greater decrease in HCV RNA during the first week: median 2.49 (IQR 1.74) log reduction in genotype 2/3 vs 1.01 (IQR 1.57) log in genotype 1, corresponding to a higher antiviral efficiency.[26,32] Overall, the second phase decline was also steeper for genotype 2/3: median 1.06 (IQR 0.82) vs 0.53 (IQR 0.87) log/week. For both genotype 1 and 2/3, patients with a greater reduction in HCV RNA after 1 week also had steeper second phase declines, in agreement with a recent report.[33]
As a result of a faster viral clearance, an RVR was achieved in 59% of patients with genotypes 2/3 when compared to 22% in genotype 1. In both groups, 100% of the RVR patients achieved SVR. In non-RVR patients with genotype 1, 47% of patients did not reach SVR, 67% of them being non-responders. In non-RVR patients with genotype 2/3, 33% did not reach SVR, but only 16% of them were non-responders. These results suggest that in genotype 1 patients without RVR, a smaller proportion will reach EVR but relapse following 48 weeks of treatment, and might therefore benefit from extended treatment. A greater impact on cost efficiency would be obtained by an early identification of non-responders, which may be obtained from slope analysis earlier than at week 12. In genotype 2/3 patients without RVR, the majority of those not achieving SVR were relapsers and might have benefited from treatment extended beyond 24 weeks, an issue that is not much discussed, except for as re-treatment of patients with non-SVR.[34,35]
This idea was also supported by our finding that genotype 2/3 relapsers (after 24 weeks of treatment) showed viral kinetics similar to those in genotype 1 patients reaching SVR without RVR (after 48 weeks of treatment). On the other hand, 67% of genotype 2/3 patients without RVR did reach SVR after 24 weeks of treatment. Unfortunately, despite the fact that HCV RNA levels differed significantly after 7 and 28 days, we could not reliably distinguish these groups (genotype 2/3 relapsers from SVR without RVR) by means of kinetics or baseline data (age, fibrosis). Still, our observations indicate that extended treatment would be of value for many non-RVR genotype 2/3 patients to reduce the risk of relapse. In such cases, extension to 36 weeks treatment may be sufficient, considering that almost 60% reached SVR after 24 weeks of treatment, but this requires further study.
Although we believe that treatment duration might be guided by the viral kinetics alone, it is clear that baseline factors strongly influence the chance of treatment success. In particular, baseline HCV RNA below 400 000 IU/mL indicates a very high chance of SVR, and cirrhosis a very low chance. Knowledge about baseline factors is therefore important for the decision to treat. However, the majority of the patients neither have cirrhosis nor low baseline viraemia, and for them, knowledge about baseline factors is of limited value for decision-making or for the choice of treatment duration. Moreover, we found that several patients with unfavourable baseline factors displayed rapid declines in viraemia. This suggests that shorter treatment may be sufficient also for older patients with high HCV RNA levels and pronounced fibrosis if only the early kinetics are favourable.
In summary, our data imply that the main predictors of treatment response were all reflected in the viral kinetics. It therefore might be possible to individualize treatment on the basis of the early viral kinetics independently on these baseline factors, at least if the response is assessed by repeated HCV RNA quantifications.
Abstract and Introduction Materials and Methods Results
Discussion References
No comments:
Post a Comment