| |
|
Direct acting antivirals after successful treatment of early hepatocellular carcinoma improve survival in HCV-cirrhotic patients
|
| |
| |
JNL of Hepatology - March 22 2019 - Authors: Giuseppe Cabibbo1,* Ciro Celsa1,* Vincenza Calvaruso1, Salvatore Petta1, Irene Cacciola2, Maria Rita Cannavo3, Salvatore Madonia4, Margherita Rossi1, Bianca Magro1, Francesca Rini1, Marco Distefano5, Licia Larocca6, Tullio Prestileo7, Giuseppe Malizia8, Gaetano Bertino9, Francesco Benanti10, Anna Licata11, Ignazio Scalisi12, Giovanni Mazzola13, Maria Antonietta Di Rosolini14, Giuseppe Alaimo15, Alfonso Averna16, Fabio Cartabellotta17, Nicola Alessi1, Salvatore Guastella1, Maurizio Russello3, Gaetano Scifo5, Giovanni Squadrito2, Giovanni Raimondo2, Franco Trevisani18, Antonio Craxì1, Vito Di Marco1 and Calogero Cammà1 on behalf of Rete Sicilia Selezione Terapia - HCV (RESIST-HCV), and Italian Liver Cancer (ITA.LI.CA.) Group.
Download the PDF here
Highlights
• DAAs improve survival in patients with HCV-related successfully treated early HCC.
• Improvement of survival seems due to reduction in hepatic decompensation.
• DAAs don't impact on HCC recurrence.
Abstract
Background & Aims
The effectiveness of direct-acting antivirals (DAAs) against hepatitis C virus (HCV) after successful treatment of early hepatocellular carcinoma (HCC) has been studied extensively. However, the benefit in terms of overall survival (OS) remains to be conclusively demonstrated. The aim of this study was to assess the impact of DAAs on OS, HCC recurrence, and hepatic decompensation.
Methods
We enrolled prospectively 163 consecutive patients with HCV-related cirrhosis and at first diagnosis of early Barcelona Clinic Liver Cancer (BCLC) 0/A HCC who had achieved a complete radiologic response after curative resection or ablation, subsequently treated with DAAs. DAA-untreated patients from ITA.LI.CA. cohort (n=328) served as controls. After propensity-score matching, outcomes of 102 DAA-treated (DAA group) and 102 DAA-untreated patients (No DAA group) were compared.
Results
In DAA group, 7/102 patients (6.9%) died, HCC recurred in 28/102 patients (27.5%) and hepatic decompensation occurred in 6/102 patients (5.9%), after a mean follow-up of 21.4 months. OS was significantly higher in DAA group compared with No DAA group (hazard ratio [HR]=0.39; 95% confidence Interval [CI]=0.17-0.91, p=0.03). HCC recurrence was not significantly different between DAA and No DAA groups (HR=0.70; 95% CI=0.44-1.13, p=0.15). A significant reduction in the rate of hepatic decompensation was observed in DAA group compared with No DAA group (HR=0.32; 95% CI=0.13-0.84, p=0.02).
In DAA group, sustained virologic response was a significant predictor of overall survival (HR 0.02, 95% CI 0.00-0.19, p<0.001), HCC recurrence (HR 0.25, 95% CI 0.11-0.57, p<0.001) and hepatic decompensation (HR 0.12, 95% CI 0.02-0.38, p=0.02).
Conclusions
In patients with HCV-related cirrhosis and previous successful treatment of early HCC, DAAs significantly improved OS compared with No DAA treatment.
In conclusion, the results observed in patients with HCV-related cirrhosis who underwent DAA therapy after successful treatment of early HCC demonstrated that patients treated with DAAs had improved OS compared to those who had not received DAAs. DAA treatment was found to improve OS through reduction of hepatic decompensation. HCC recurrence risk remains high despite DAA therapy, highlighting the need for new adjuvant strategy for prevention of HCC recurrence. Finally, DAA-induced SVR significantly reduces mortality, hepatic decompensation and HCC recurrence. A meta-analysis of individual patient data, including studies performed in different clinical settings, may prove useful to substantiate the benefit of DAA therapy on long-term outcomes. [40]
The main aim of this prospective multicentre study was to estimate, in patients with compensated cirrhosis who are infected with hepatitis C virus (HCV) and at first diagnosis of early hepatocellular carcinoma (HCC) (without history of HCC recurrence before DAA treatment) who had achieved a complete radiologic response after curative resection or ablation, if DAAs (DAA group) prolong OS compared with an appropriately matched control group of patients who had not received DAAs (No DAA group). As secondary outcomes we considered the impact of DAAs on HCC recurrence and on hepatic decompensation.
A recent meta-analysis showed that, among HCV-infected compensated cirrhotic patients in whom early HCC was successfully treated, who remained unexposed to direct-acting antivirals (DAAs), the 2-year actuarial pooled HCC recurrence rate was 47.0% and the 3-year actuarial pooled survival rate was 79.8%. [4] These data indicate that there is an urgent need for an effective adjuvant strategy given the prior failures of other adjuvant treatments, including sorafenib. [5]
In 2016, an alarm signal was released about a potentially increased risk for early HCC recurrence after DAA therapy, raising concerns about the safety of DAA use in patients with previously treated early HCC. [8, 9] Since then, several prospective studies [10, 11, 12, 13] and two meta-analyses [14, 15] have provided evidence that the risk of HCC recurrence after treatment with DAAs in patients with history of successful treatment of early HCC is similar, if not lower, than that observed in interferon-treated or DAA-unexposed controls. However, field-practice prospective studies that prove the benefit of DAAs on overall survival and hepatic decompensation are lacking, and the longer-term effect of DAAs on mortality remains to be established.
Outcomes
Table 3 showed the outcomes observed during follow-up of DAA group (n=102) and No DAA group (n=102), after propensity-score-matching. During a mean follow-up of 21.4 months (range 1-37 months), 7 of 102 patients (6.9%) in DAA group and 18 of 102 patients (17.7%) in No DAA group died. Actuarial survival rates at 6, 12, 24, and 36 months between DAA and No DAA groups are shown in Table 3 and Figure 1. DAA group had a significantly higher survival rate than No DAA group (Hazard Ratio [HR] 0.39, 95% Confidence Interval [CI] 0.17-0.91, p = 0.03). Similar results were obtained by IPTW (HR 0.42, 95% CI 0.18-1, p=0.05). In order to assess the impact of survivor treatment selection bias in observational studies, four other models were performed using DAA as a time-dependent covariate or starting follow-up from DAA initiation in DAA group and from index time in the No DAA group. In all the robust analyses, the estimate of the DAA treatment effect on overall survival was significant. (see Supplementary Table S2 and Figure S4).
HCC recurred in 28 of 102 patients in DAA group (27.5%) and in 38 of 102 (37.3%) patients in No DAA group. Actuarial 6, 12, 24 and 36 month HCC recurrence rates between DAA and No DAA groups are shown in Table 3 and Figure 2. In DAA group, HCC recurrence rate was lower than in No DAA group, but difference was not statistically significant (HR 0.70, 95% CI 0.44-1.13, p=0.15).
During follow-up, 6 of 102 (5.9%) patients in DAA group and 14 of 102 (13.7%) patients in No DAA group experienced hepatic decompensation. The 6, 12, 24 and 36 month actuarial rates of hepatic decompensation are shown in Table 3 and Figure 3. DAA group showed significantly lower decompensation rate than no DAA group (HR 0.32, 95% CI 0.13-0.84, p=0.02).
Predictors of mortality, HCC recurrence and hepatic decompensation in DAA group
Predictors of outcomes (mortality, HCC recurrence, hepatic decompensation) in the cohort of 102 DAA-treated patients are showed in Table 4. SVR was the only variable associated with a decrease in mortality (HR 0.02, 95% CI 0.00-0.19, p<0.001). It is noteworthy that SVR was independently associated with a decrease in HCC recurrence risk (HR 0.25, 95% CI 0.11-0.57, p<0.001) and hepatic decompensation (HR 0.12, 95% CI 0.02-0.38, p=0.02) by multivariate analyses.
Discussion
This prospective real-world multicentre observational study involved a cohort of patients with HCV-related cirrhosis who had received treatment with DAAs after curative resection or ablation of early HCC. According to methods for comparative effectiveness, we used our observational data to emulate a hypothetical randomized trial by comparing DAA exposed versus DAA unexposed patients. [16, 26] Its results show that patients treated with DAAs had significantly better OS and lower hepatic decompensation than propensity-score-matched controls who did not receive DAAs. To the best of our knowledge, the benefit of DAAs on OS and hepatic decompensation was demonstrated for the first time in our matching-propensity analyses adjusted for demographic, liver function, and cancer-related characteristics. In contrast, the risk of HCC recurrence was not significantly different between patients treated with DAAs and propensity-score-matched controls who did not receive DAAs.
Most investigators and clinicians now accept DAAs as the standard of care, even for patients with advanced liver disease and a history of HCC. [6, 7] As a result, it is not feasible to design RCTs for direct comparison of patients who did versus did not receive DAAs. Therefore, DAA and No DAA groups may be compared only by propensity score methods to correct for potential confounding. Despite the lack of randomization, our findings suggest that DAAs improve OS and hepatic decompensation without affecting HCC recurrence risk, after adjustment by matching-propensity analyses. We underline that DAA-treated patients, compared with controls who did not receive DAAs had more advanced liver disease in terms of higher Child-Pugh score, prevalence of oesophageal varices and tumor burden. Therefore, our results are particularly relevant considering that a full balance between DAA and No DAA groups was obtained after propensity-score-methods.
It is noteworthy that the reported 2-year HCC recurrence actuarial rate of our matched controls who did not receive DAAs is in keeping with the results of previous meta-analysis of DAA-unexposed patients with compensated cirrhosis, [4] providing that our control group is a useful benchmark for comparison of DAA benefit on this outcome. [27] Among the DAA group, the 2-year HCC recurrence rate, although not significantly different compared to No DAA group, still remains high (27%). This finding is consistent with the results of two recent published studies. [28, 29] Multivariate analysis of our cohort of DAA-exposed patients showed that no SVR after DAAs was an independent risk factor for HCC recurrence. So, our study suggests that future models including SVR could identify patients at higher risk for HCC recurrence reliably and thereby maximize the clinical benefit and the cost/effectiveness of imaging follow-up. Lack of data on other potential risk factors for HCC recurrence, such as microscopic vascular invasion, histology grade, and both cancer and patient gene profile, [30] may have affected the accuracy of the results. Incomplete knowledge of molecular characteristics and tumor pathobiology (e.g., neo-angiogenic pathways) may explain why adjuvant therapy in HCC represents an area of high unmet medical need. New markers, such as liver angiopoietin-2, have shown promising results in predicting HCC growth and recurrence, as well as risk for death. [30] A recent report suggested that a DAA-mediated increase in VEGF may act as a trigger in patients with highly activated neo-angiogenic pathways in cirrhotic tissue. [31]
Our data provide evidence that DAAs significantly reduce the risk of hepatic decompensation. The findings presented above strongly support the current practice of DAA treatment, even in patients with advanced liver disease and HCC. Exploring causality multifactorial mechanism of the observed reduction in mortality associated with DAAs is outside the purpose of the present study. We can speculate that DAAs improve OS through long-term preservation of liver function, resulting in greater likelihood that the patient will receive curative treatment, even in the case of HCC recurrence. This result suggests that, although the risk of HCC recurrence remains high, the inefficacy of DAA treatment on HCC recurrence risk could be overcome by way of the proven survival-enhancing benefit with regard to hepatic decompensation. [3] In this line, a recent retrospective cohort study by Huang et al. found that DAA use does not change the risk of HCC recurrence after local-regional therapy among waitlisted patients but rather is associated with reduced risk of waitlist dropout due to tumor progression or death. [7] Although waitlisted patients had more advanced liver disease (half had CP B or C cirrhosis) compared to our cohort of compensated cirrhotic patients, these results add further evidence that DAAs reduce the risk of decompensating events also in patients with advanced liver disease and HCC. These findings align with previous literature showing that DAA therapy is associated with improvement in liver function [12, 32, 33, 34] as well as lower rates of hepatic decompensation. [12, 32] All these evidences lend support to the use of DAA therapy in patients with advanced liver disease and successfully treated HCC. Finally, treatment of HCV infection even in the sickest patients with more advanced BCLC stage remains a controversial topic. [35]
Our study has some limitations. First, a methodological issue of the current study is the potential limitation of the generalizability of its results to different populations and settings, given that the results were obtained in a cohort of real-world, compensated patients with cirrhosis and early HCC enrolled in tertiary care centres, who may be different in terms of age, clinical features and comorbidities from patients treated with DAAs in other settings. In particular, we excluded patients with HBV/HCV or HIV/HCV coinfections, as well as those with comorbidities or those who did not receive potential curative HCC treatments like resection or ablation. It should be also emphasized that our cohort included older patients (mean age 70 years) than those recruited in other studies (ranging from 54 to 58 years). [6, 12, 32, 33, 34]
Second, this nonrandomised study regards the issue of survivor treatment selection bias in observational studies, such as immortal and time-lag biases. [20, 21, 36, 37] The effectiveness of DAAs was also confirmed when DAA treatment was assessed as time-dependent covariate, mitigating misclassified immortal-time bias. Regarding the time-lag bias [20] (i.e. a potential bias due to the fact that patients may change their disease stage during the time-lag between CRR and DAA start), we are confident that this bias did not affect our results because no clinical events occurred during this short interval (median 2.1 months). [38] Moreover, in all the robust analyses performed with and without time-dependent methods or starting follow-up from DAA initiation or from CRR, the estimates of the treatment effect of DAAs on survival were similar and significant. Finally, although matching and IPTW were used in order to handle the issue of lack of randomization, residual confounding from both measured and unmeasured variables could not be definitively rule out. Whatever efforts were made to control for confounding and survivor treatment selection bias, "bias-zero" observational studies of drug effects do not exist, remaining randomized controlled trials the gold standard to evaluate the effectiveness of treatments.
Finally, another limitation of our study is the use of historical, instead of contemporaneous controls. [39] However, differences in HCC management, regarding diagnosis, follow-up and treatment were not observed between DAA-treated and DAA-untreated patients. Moreover, the 3-year survival and 2-year HCC recurrence of our historical controls from the ITA.LI.CA. cohort are in keeping with the benchmark provided by a previous published meta-analysis of DAA-untreated patients, [4] as well as by the placebo arm of the STORM trial [5] (that included patients with successfully treated early HCC), making us confident that our historical controls are representative of the population of DAA-unexposed patients.
In conclusion, the results observed in patients with HCV-related cirrhosis who underwent DAA therapy after successful treatment of early HCC demonstrated that patients treated with DAAs had improved OS compared to those who had not received DAAs. DAA treatment was found to improve OS through reduction of hepatic decompensation. HCC recurrence risk remains high despite DAA therapy, highlighting the need for new adjuvant strategy for prevention of HCC recurrence. Finally, DAA-induced SVR significantly reduces mortality, hepatic decompensation and HCC recurrence. A meta-analysis of individual patient data, including studies performed in different clinical settings, may prove useful to substantiate the benefit of DAA therapy on long-term outcomes. [40]
| |
| |
| |
|
|
|
