Direct-Acting Antiviral Therapy for HCV Infection is Associated with Increased Survival in Patients With a History of Hepatocellular Carcinoma

Abstract

Background & Aims

There is controversy over benefits of direct-acting antiviral (DAA) therapy for hepatitis C virus (HCV) infection for patients with a history of hepatocellular carcinoma (HCC). We performed a multi-center cohort study to compare overall survival between patients with HCV infection treated with DAAs vs patients who did not receive DAA treatment for their HCV infection after complete response to prior HCC therapy.

Methods

We conducted a retrospective cohort study of patients with HCV-related HCC who achieved a complete response to resection, local ablation, trans-arterial chemo- or radioembolization, or radiation therapy, from January 2013 through December 2017 at 31 healthcare systems throughout the United States and Canada. We used Cox proportional hazards regression to determine the association between receipt of DAA therapy, modeled as a time-varying covariate, and all-cause mortality, accounting for informative censoring and confounding using inverse probability weighting.

Results

Of 797 patients with HCV-related HCC, 383 patients (48.1%) received DAA therapy and 414 patients (51.9%) did not receive treatment for their HCV infection after complete response to prior HCC therapy. Among DAA-treated patients, 43 deaths occurred during 941 person-years of follow up, compared with 103 deaths during 526.6 person-years of follow up among patients who did not receive DAA therapy (crude rate ratio, 0.23; 95% CI, 0.16–0.33). In inverse probability-weighted analyses, DAA therapy was associated with a significant reduction in risk of death (hazard ratio [HR], 0.54; 95% CI, 0.33–0.90). This association differed by sustained virologic response (SVR) to DAA therapy; risk of death was reduced in patients with SVR to DAA therapy (HR, 0.29; 95% CI, 0.18–0.47) but not in patients without an SVR (HR, 1.13; 95% CI, 0.55–2.33).

Conclusions

In an analysis of nearly 800 patients with complete response to HCC treatment, DAA therapy was associated with a significant reduction in risk of death.

LAY SUMMARY

Treatment of hepatitis C virus with direct-acting antiviral agents in patients who have been successfully treated for liver cancer increases overall survival.

Graphical Abstract

BACKGROUND

Hepatocellular carcinoma is the third leading cause of cancer-related death worldwide and a leading cause of death in patients with compensated cirrhosis.(1) Given the majority of HCC cases occur in the setting of chronic liver disease, prognosis depends on both tumor burden and the degree of liver dysfunction. Most patients with HCC are diagnosed at late stages, with limited treatment options and a median survival of approximately one year.(2) Those diagnosed with HCC at an early stage are eligible for curative treatments, resulting in 5-year survival approaching 70%; however, outside of liver transplantation, these therapies are limited by a high risk of recurrence and a continued risk of hepatic decompensation.(3–6)

Chronic hepatitis C virus (HCV) infection remains the most common cause of HCC in North America and Europe.(6) Several studies have reported a benefit of direct acting antiviral (DAA) therapy in reducing incident HCC and mortality among HCV-infected individuals(7–9), but the impact of these therapies on prognosis in patients who have had a complete response to HCC treatment is less clear. Interferon (IFN)-based therapy was associated with reduced risk of recurrence and mortality in patients with a history of HCC; however, there is uncertainty whether this was mediated by a direct IFN-mediated effect that would not translate to DAA-based therapy.(10) In fact, some early observational studies suggested a higher than expected risk of HCC recurrence after DAA therapy.(11) Subsequent studies have produced conflicting data, with most reporting no significant difference in recurrence between DAA-treated and untreated patients, including our multi-center North American study, in which we found no significant difference in overall recurrence, early recurrence, or recurrence aggressiveness.(9,12–16)

To date, most studies have focused on the association between DAA therapy and HCC recurrence, with fewer examining the association between DAA therapy and overall survival in patients with a history of HCC. Notably, data from ITA.LI.CA group demonstrated that hepatic decompensation may more strongly influence mortality in HCV-infected patients with a history of complete response to prior HCC therapy than HCC recurrence.(17) Therefore, DAA therapy may improve overall survival in patients with a history of HCC, independent of recurrence risk. The aim of this multi-center study was to compare overall survival between DAA-treated and untreated patients in a large cohort of North American patients with complete response to prior HCC-directed therapy.

METHODS

Study population

As previously described,(15) we conducted a multi-center retrospective cohort study of adult patients with HCV-related HCC who achieved HCC complete response between January 2013 and December 2017. In brief, patients were identified from 31 health systems throughout the United States and Canada. HCC diagnosis was based on characteristic radiologic appearance or histology per AASLD criteria.(18) Patients were required to have liver-localized tumor burden at presentation, and those with extrahepatic disease were excluded. We included patients who achieved complete HCC response by surgical resection, local ablative therapies, transarterial chemoembolization (TACE) or bland embolization, transarterial radioembolization (TARE) or stereotactic body radiation therapy (SBRT). Complete response to HCC treatment was defined by mRECIST criteria, i.e. disappearance of arterial enhancement from all HCC lesions on contrast-enhanced cross-sectional imaging. Patients with complete response after liver transplantation or systemic therapy were excluded.

We excluded patients who received IFN-based therapy or initiated DAA therapy prior to HCC complete response. We also excluded patients with unknown HCC response, e.g. lack of contrast-enhanced imaging after HCC treatment, and patients who died within 30 days of complete response. The study was approved by Institutional review boards at each study site.

Covariates

Demographic and clinical characteristics

We used a standardized data collection form to collect patients’ demographic and clinical characteristics from electronic medical records at time of HCC diagnosis, including age, sex, race/ethnicity, presence of hepatitis B (HBV) and HIV co-infection, platelet count, AST, ALT, HCV viral load, HCV genotype, Eastern Cooperative Oncology Group (ECOG) performance status, and alpha fetoprotein (AFP). Degree of liver dysfunction was assessed by Child Pugh and MELD scores. Tumor burden, as determined by radiologists’ interpretation of imaging at each site, was categorized as very early stage (single tumor < 2 cm), early stage (single tumor < 5 cm or 2–3 tumors with each < 3 cm in maximum diameter), or intermediate stage (beyond early stage but without extra-hepatic spread). We also collected the type of treatment leading to complete response (surgical resection, ablation, TACE, other).

DAA treatment

We considered DAA treatment to be time-varying, allowing patients to switch between unexposed and exposed status. Specifically, we assigned follow-up time for patients before DAA initiation as unexposed time and time after DAA initiation as exposed time. Patients who initiated DAA therapy after liver transplantation only contributed unexposed time. For DAA-treated patients, we extracted information on treatment regimen, time from HCC complete response to DAA initiation, and outcome (i.e., sustained virological response [SVR]).

Statistical analyses

Primary analysis

We used Chi-square and Student’s t tests to characterize the study population by DAA treatment (DAA-treated vs. untreated). Because we allowed treatment to be time-varying, for this purpose, we defined DAA-treated patients as patients who initiated DAA at any time during follow-up, regardless of SVR.

For both DAA-treated and untreated patients, we defined the index date as the date of complete response to HCC treatment. We followed each patient from complete response to date of death, liver transplantation, or last clinic visit. We classified causes of death into four mutually exclusive categories: liver-related (e.g. acute on chronic liver failure, hepatorenal syndrome, spontaneous bacterial peritonitis); HCC-related (e.g. recurrent advanced stage HCC); non-liver/non-HCC related; and unknown (e.g. died elsewhere). The primary outcome was death from all causes, and we estimated the rate of all-cause mortality as the number of deaths divided by person-years of follow-up. We also calculated the crude mortality rate ratio as the ratio between the mortality rate among DAA-treated patients and that among untreated patients.

We then used a Cox proportional hazards regression model with inverse probability (IP) weights to examine overall survival between DAA-treated and untreated patients. Using this approach, each patient is weighted to create a pseudo-population in which: 1) DAA therapy is not associated with baseline covariates such that confounding is eliminated; and 2) censoring by liver transplantation is not associated with DAA therapy or covariates such that selection bias due to censoring is eliminated.(19,20) We used separate logistic regression models to estimate the probability of treatment (i.e., initiating DAA therapy) and for the probability of not being censored (i.e., not receiving liver transplant), including the following covariates: age, sex, Child Pugh class, tumor burden, AFP level, treatment leading to complete response, and time since complete response (censoring weight only). The resulting IP weight is the product of an estimated time-fixed IP treatment weight and an estimated time-varying IP censoring weight.

Sensitivity analysis

To assess the robustness of DAA treatment as a time-varying exposure, we performed a sensitivity analysis using a 180-day landmark. In this restricted subgroup, patients who initiated DAA within 180 days of complete response were categorized as DAA-treated, and those who did not initiate before 180 days were categorized as untreated. Patients who died, had HCC recurrence, received liver transplantation, or had their last clinic visit before the landmark were excluded. These restrictions avoid immortal time bias by synchronizing the start of follow-up for treated and untreated patients.(21,22) Survival by treatment group was then compared using Cox proportional hazards models with IP weights.

Secondary analyses

We hypothesized the association between DAA treatment and all-cause mortality differed among patients who achieved and did not achieve SVR. Therefore, we compared overall survival of DAA-treated patients with and without SVR to untreated patients. We also conducted secondary analyses stratified by tumor burden (within vs. beyond Milan criteria), treatment leading to HCC complete response (resection vs. ablation vs. TACE/SBRT), and HCC recurrence.

All tests were two-sided and performed at the 5% significance level. Statistical analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC USA).

RESULTS

Patient Characteristics

Selection of the study population is illustrated in Figure 1. Of 852 eligible HCV-infected HCC patients who achieved complete response to HCC-directed therapy between January 2013 and December 2017, we excluded 47 patients who initiated DAA therapy and 8 patients who died within 30 days of HCC complete response. There were 797 remaining patients included in the final analysis, of whom 383 initiated DAA therapy and 414 were untreated.

Figure 1.

Study cohort inclusion and exclusion diagram

Patient demographics are described in Table 1. Median age of patients was 61.5 years. Nearly three-fourths of patients were male and the cohort was racially diverse. At HCC diagnosis, nearly three-fourths of patients had a unifocal HCC and over 80% were within Milan Criteria. One-third of patients achieved HCC complete response from local ablation and 14% from surgical resection, while approximately half achieved complete response from locoregional therapies. A higher proportion of DAA-treated patients were older, non-Hispanic white, had compensated cirrhosis with less portal hypertension, achieved HCC complete response by resection than untreated patients; however, the two groups had similar tumor burden at diagnosis, including a similar proportion of patients presenting within Milan Criteria.

Table 1.

Patient characteristics, stratified by DAA treatment status

Variable*	DAA-treated (n=383)	DAA-untreated (n=414)	p-value
Age at time of complete response (years)	62.1 (58.6 – 66.1)	61.2 (57.0 – 65.0)	0.01
Gender (% male)	276 (72.1)	320 (77.3)	0.09
Race/ethnicity			0.02
Non-Hispanic White	187 (48.8)	159 (38.4)
Hispanic White	58 (15.1)	66 (15.9)
Black	75 (19.6)	92 (22.2)
Other	18 (4.7)	22 (5.3)
Missing	45 (11.8)	75 (18.1)
Number of HCC nodules at diagnosis			0.01
One	291 (77.2)	270 (67.8)
Two	65 (17.2)	88 (22.1)
Three	17 (4.5)	25 (6.3)
Four	4 (1.1)	15 (3.7)
Maximum HCC diameter (cm) at diagnosis	2.4 (1.7 – 3.3)	2.6 (2.0 – 3.7)	0.42
HCC within Milan Criteria at diagnosis	313 (82.6)	332 (80.8)	0.51
AFP at time of HCC diagnosis (ng/mL)	18.5 (7.3 – 67)	18.3 (7.9 – 68)	0.29
Treatment leading to complete response			< 0.001
Resection	81 (21.2)	33 (8.0)
Local ablation	138 (36.0)	130 (31.4)
TACE	136 (35.5)	222 (53.8)
TARE/SBRT/other	28 (7.3)	28 (6.8)
Number of HCC therapies required to achieve complete response			0.13
One
Two	220 (60.8)	205 (55.3)
Three or more	92 (25.4)	95 (25.6)
	50 (13.8)	71 (19.1)
Child Pugh class at complete response			< 0.001
Child Pugh A	235 (61.4)	199 (48.1)
Child Pugh B	127 (33.2)	160 (38.7)
Child Pugh C	21 (5.5)	55 (13.3)
Presence of ascites	98 (26.0)	151 (37.8)	<0.001
Presence of hepatic encephalopathy	56 (14.9)	89 (22.3)	0.008
Platelet count at complete response	109 (75 – 158)	87 (60 – 133)	< 0.001
Bilirubin at complete response (mg/dL)	1.0 (0.6 – 1.6)	1.3 (0.7 – 2.1)	0.18
HCV genotype			< 0.001
Genotype 1	293 (77.9)	331 (69.6)
Genotype 2	27 (7.2)	17 (3.9)
Genotype 3	43 (11.4)	70 (14.9)
Genotype 4 – 6	11 (2.9)	17 (2.9)
Viral co-infection			0.38
Hepatitis B	8 (2.2)	5 (1.3)
HIV	9 (2.5)	6 (1.6)
DAA regimen**
Sofosbuvir/ledipasvir	237 (63.0)
Sofosbuvir	56 (14.9)
Simeprevir/sofosbuvir	32 (8.5)	N/A
Sofosbuvir/velpatasvir	18 (4.8)
Daclatasvir/sofosbuvir	16 (4.3)
	7 (1.9)
Ombitasvir/partiprevir/ritonavir/dasabuvir	3 (0.8)
Elbasavir/grazoprevir	7 (1.9)
Other
DAA regimen duration
Less than 12 weeks	15 (4.0)
12 weeks	197 (53.1)	N/A
>12 weeks but < 24 weeks	26 (7.0)
24 weeks	126 (34.0)
Greater than 24 weeks	7 (1.9)
Time from HCC complete response to DAA
Less than 3 months	76 (20.2)
>3 – 6 months	81 (21.5)	N/A
>6 – 12 months	96 (25.5)
>12 – 24 months	92 (24.4)
Greater than 24 months	32 (8.5)

^*Continuous data presented as median (IQR)

^**All DAA regimens are with or without ribavirin

AFP – alpha fetoprotein; DAA – direct acting antiviral; HCC – hepatocellular carcinoma; HCV – hepatitis C virus; SBRT – stereotactic body radiation therapy; TACE – transarterial chemoembolization; TARE – transarterial radioembolization

Median time from HCC diagnosis to HCC treatment was 2.7 (IQR 1.4 – 5.6) months, and median time from HCC treatment to complete response was 1.6 (IQR 1.1 – 3.2) months. Median time from HCC complete response to DAA initiation was 7.7 (IQR 3.6 – 14.1) months, with 20.2% treated within 3 months, 21.5% between 3–6 months, 25.5% between 6–12 months, and 32.9% more than 12 months after HCC complete response. SVR was documented in 79.4% of DAA-treated patients, while 11.5% had treatment-failure and 9.1% did not have documented assessment of SVR.

Association between DAA Therapy and Overall Survival

Among DAA-treated patients, 43 deaths occurred during 941.0 person-years of follow-up compared to 103 deaths during 526.6 person-years of follow-up among untreated patients, yielding a crude rate ratio of 0.23 (95%CI 0.16 – 0.33). The 1- and 2-year risk of mortality for DAA-treated patients was 5.5% and 11.8%, compared to 10.0% and 23.5% among untreated patients. The median time from HCC complete response to death among DAA-treated patients was 25.7 (IQR 19.4 – 33.9) months compared to 11.5 (IQR 7.1 – 20.2) months for untreated patients. The proportion of deaths that were liver-related was significantly lower in DAA-treated than untreated patients (16.3% vs. 34.0%, p=0.03), whereas the proportion of HCC-related deaths was similar between the two groups (30.2% vs. 29.1%, p=0.89).

In the primary analysis, DAA therapy was associated with reduced mortality in crude (HR 0.37, 95%CI 0.26 – 0.54) and weighted (HR 0.54, 95%CI 0.33 – 0.90) models (Table 2). In the sensitivity analysis using a 180-day landmark, DAA therapy was associated with reduced mortality (weighted HR 0.57, 95%CI 0.30 – 1.09); however, few patients were classified as DAA-treated (n=133), and the effect estimate did not reach statistical significance.

Table 2.

Associations between DAA treatment and mortality in HCV-infected patients with HCC

	Patients	Deaths	Person-years	Mortality rate (per 100 person years)	Crude rate ratio	Adjusted HR*	Sample size in adjusted HR
DAA treatment as time-varying exposure†
Overall
Untreated	414	103	526.63	19.56 (16.05, 23.62)			401
DAA-treated	383	43	941.01	4.57 (3.35, 6.09)	0.23 (0.16, 0.33)	0.54 (0.33, 0.90)	339
Within Milan
Untreated	332	83	427.95	19.35 (15.55, 23.91)			323
DAA-treated	313	38	777.54	4.89 (3.51, 6.63)	0.25 (0.17, 0.37)	0.56 (0.32, 0.98)	280
Beyond Milan
Untreated	79	18	96.22	18.71 (11.48, 28.93)			78
DAA-treated	66	5	155.00	3.23 (1.22, 7.07)	0.17 (0.06, 0.46)	0.53 (0.21, 1.39)	59
Surgical resection
Untreated	33	9	58.91	15.28 (7.55, 27.89)			30
DAA-treated	81	5	235.89	2.12 (0.80, 4.65)	0.14 (0.05, 0.41)	0.21 (0.06, 0.76)	73
Local ablation
Untreated	130	42	174.59	24.06 (17.58, 32.19)			123
DAA-treated	138	13	326.02	3.99 (2.23, 6.63)	0.17 (0.09, 0.31)	0.26 (0.13, 0.51)	116
TACE
Untreated	222	43	264.59	16.25 (11.92, 21.67)			220
DAA-treated	136	23	312.42	7.36 (4.79, 10.86)	0.45 (0.27, 0.75)	0.99 (0.47, 2.08)	129
Child Pugh A
Untreated	199	45	301.80	14.91 (11.02, 19.76)			191
DAA-treated	235	17	600.47	2.83 (1.71, 4.43)	0.19 (0.11, 0.33)	0.25 (0.14, 0.48)	206
Child Pugh B
Untreated	160	44	182.43	24.12 (17.76, 32.07)			155
DAA-treated	127	18	303.71	5.93 (3.64, 9.17)	0.25 (0.14, 0.43)	0.43 (0.24, 0.79)	116
Child Pugh C
Untreated	55	14	42.40	33.02 (18.91, 53.92)			55
DAA-treated	44	11	36.83	3.62 (1.92, 6.27)	0.90 (0.41, 1.99)	1.92 (0.56, 6.51)	17
Recurrence
Untreated	205	62	326.40	19.00 (14.70, 24.18)			196
DAA-treated	209	39	535.88	7.28 (5.25, 9.84)	0.38 (0.26, 0.57)	0.86 (0.49, 1.52)	184
No recurrence
Untreated	199	40	197.62	20.24 (14.67, 27.27)			195
DAA-treated	174	4	405.13	0.99 (0.33, 2.35)	0.05 (0.02, 0.14)	0.09 (0.02, 0.29)	155
No prior recurrence
Untreated	400	95	508.91	18.67 (15.19, 22.71)			387
DAA-treated	372	43	917.13	4.69 (3.44, 6.25)	0.25 (0.18, 0.36)	0.57 (0.34, 0.94)	329
Using 180-day landmark‡
Overall
Untreated	423	73	714.16	10.22 (8.07, 12.78)			394
DAA-treated	133	11	204.14	5.39 (2.86, 9.33)	0.53 (0.28, 0.99)	0.57 (0.30, 1.09)	115

^*Adjusted for age, sex, Child Pugh class, HCC tumor stage, AFP level, and type of HCC treatment

^†Exposure to DAAs modeled as a time-dependent covariate

^‡DAA group includes patients who initiated DAA therapy within 180 days from HCC complete response. Patients who died, developed recurrence, received liver transplantation, or had their last clinic visit within 180 days were excluded.

DAA – direct acting antiviral; HCC – hepatocellular carcinoma; SVR – sustained virological response; TACE – transarterial chemoembolization

Subgroup analyses are detailed in Table 2. DAA therapy was associated with improved survival among patients with Child Pugh A or B cirrhosis; assessing mortality benefit in those with Child C cirrhosis was limited by small sample size. We also observed a consistent benefit of DAA therapy among patients who initially had tumor burden within and beyond Milan Criteria. The reduction in mortality with DAA therapy was greater in patients who received resection or ablation than in those who received TACE, (interaction p-value=0.03). Results were consistent in the subgroup of patients without prior history of HCC recurrence. However as expected, there was a significantly greater benefit of DAA therapy in patients who remained recurrence-free during the study period (HR 0.09, 95%CI 0.02 – 0.29) compared to those who experienced HCC recurrence (HR 0.86, 95%CI 0.49 – 1.52) (interaction p-value<0.001).

Association between DAA Therapy and Overall Survival, Stratified by SVR Status

We expected the association between DAA therapy and mortality to differ by SVR and examined DAA-treated patients with and without SVR in greater detail. Characteristics of DAA-treated patients, stratified by SVR status, are detailed in Supplemental Table 1. Patients with SVR were more likely to be female and non-Hispanic white, but they had a similar number of HCC nodules, maximum tumor diameter, proportion within Milan Criteria at diagnosis, as well as type and number of treatments leading to HCC complete response. We also found no significant difference in the degree of liver dysfunction at complete response, HCV genotype, type and duration of DAA regimen, and timing of treatment from HCC complete response. Although a higher proportion of patients with SVR subsequently underwent liver transplantation, this difference did not reach statistical significance (22.7% vs. 13.6%, p=0.17).

Among DAA-treated patients, patients with SVR had significantly reduced mortality from time of DAA initiation compared to non-SVR patients (HR 0.27, 95%CI 0.13 – 0.57). In a sensitivity analysis in which patients with missing SVR status were assumed to be non-SVR, SVR continued to be associated with reduced mortality (HR 0.47, 95%CI 0.35 – 0.64). We next compared both groups of DAA-treated patients (with and without SVR) to untreated patients. DAA-treated patients who achieved SVR had significantly reduced mortality compared to untreated patients (HR 0.29, 95%CI 0.18 – 0.47); however, a mortality benefit was not observed among DAA-treated patients who did not achieve SVR compared to untreated patients (HR 1.13, 95%CI 0.55–2.33).

DISCUSSION

There has been extensive debate about the potential benefit of DAA therapy in patients with a history of HCC, primarily related to concerns about the risk of HCC recurrence. While recent data have suggested that DAA therapy after complete HCC response does not appear to increase recurrence, to the best of our knowledge our study represents the first to demonstrate a survival benefit of DAA therapy in this patient population.

Although HCC treatments such as surgical resection and local ablative therapies are potentially curative, they are limited by high risk of recurrence and 5-year mortality approaches 50%. Prior attempts to identify adjuvant therapies to reduce HCC recurrence, such as the STORM Trial evaluating sorafenib after resection or ablation, largely failed.(23) Cabbibo and colleagues recently demonstrated that hepatic decompensation was a stronger driver of mortality than HCC recurrence in patients with complete response to HCC therapy, suggesting interventions that reduce portal hypertension and other manifestations of advanced liver disease could improve survival.(17) Albeit of limited sample size, results from the ANRS CirVir prospective cohort study suggested a potential benefit of DAA-associated SVR to reduce hepatic decompensation in patients with HCC.(24) Hepatic decompensation was one of the most common causes of death in patients without SVR but was not observed in any of the four patients with SVR. Similarly, Bruno and colleagues found that over one-third of HCC patients with untreated HCV infection died of hepatic decompensation compared to 0 of 19 patients who had achieved SVR prior to HCC diagnosis.(25) However, these studies both evaluated the benefit of SVR prior to HCC diagnosis, where there may be a longer time frame to experience fibrosis regression and improvement in portal hypertension.(26,27) Our study adds to this literature and reinforces that DAA therapy may also reduce mortality in patients with a history of HCC.

Although DAA therapy was associated with a consistent survival benefit across most patient subgroups, this association was mitigated in those who were treated with TACE as well as those with HCC recurrence, likely due to a higher risk of HCC mortality. Although DAA therapy can help improve or stabilize liver function in these patients, the high risk of recurrence and persistent HCC-related mortality highlights a continued need for adjuvant therapy after HCC complete response. In the absence of proven adjuvant agents, close surveillance after complete response may increase early tumor detection and minimize HCC-related mortality.(28,29) As previously reported, over two-thirds of patients with recurrence in our study were detected at an early stage (BCLC stage 0/A), and over 50% of those with known treatment response had an objective response.(15) It is important to note that although the association between DAA therapy and survival was not statistically significant in these subgroups, the samples sizes were too small to make definitive conclusions, and the point estimates still suggested a potential benefit. Further studies are needed in these patients to determine if DAA therapy may be of potential benefit.

Although we observed a benefit of DAA therapy in patients with a history of HCC, the optimal timing of DAA therapy in relation to HCC treatment remains unclear. Should patients initiate DAA therapy concurrent with HCC treatment, immediately thereafter, or is it best to wait for documented complete response? Although some studies suggest a lower proportion of patients with active HCC achieve SVR, it is unclear if this concern should drive decisions of timing.(9,30–32) If the potential benefit of DAA therapy is truly related to preventing liver dysfunction, earlier treatment may be better. Extending this further is the question if DAA therapy would have similar benefits among patients with active intermediate-stage or even advanced stage HCC.(33) Further studies in these patient populations to address these important clinical dilemmas are clearly needed.

It is possible that the association between DAA therapy and reduced mortality may be driven by confounding, with DAA therapy being selectively used in patients with better prognosis. We attempted to adjust for measured confounders and selection bias with IP-weighting, and we mitigated the effect of immortal time bias by using time-varying and landmark analyses. Further, we observed no association between DAA therapy and survival among those who were treated but did not achieve SVR. We acknowledge still the potential for residual confounding. For example, DAA-treated patients had less portal hypertension than untreated patients, which could partly explain results.

Results from our study must be interpreted in light of other potential limitations. First, we used imaging interpretation through routine clinical care instead of centralized imaging review, which could have affected classification of HCC complete response.(34) However, all included health systems were academic centers with GI-trained radiologists, and most centers use multidisciplinary tumor boards.(35,36) Importantly, reduction in mortality were most evidence in patients undergoing surgical resection or ablation which would be less prone to misclassification of complete response. The weaker association between DAA therapy and reduced mortality in patients undergoing TACE may have been partly related to unrecognized microscopic residual disease. Second, approximately 9% of patients were lost to follow-up, with a higher proportion in the DAA-untreated group. Finally, we did not have data on incident development of hepatic decompensation during study follow-up and were unable to demonstrate whether the survival benefit we observed was definitively related to preservation of liver function. Ongoing multi-center prospective cohort studies with longer follow-up will hopefully address some of these limitations; however these are years away from reporting. These limitations were felt to be outweighed by the study’s notable strengths, including its multi-center design, large cohort of patients, rigorous statistical analysis plan, and inclusion of a contemporary untreated comparison group.

In summary, DAA therapy was associated with a significant reduction in mortality among HCV-infected patients with documented complete response to HCC therapy. Overall, our results suggest use of DAA therapies is likely beneficial in HCV-infected patients with a history of HCC.

Figure 2.

Overall survival, stratified by receipt of direct acting antiviral hepatitis C therapy

What you need to know.

Background & Context

Direct acting antiviral (DAA) treatment is associated with increased survival in patients with cirrhosis, but little is known about its effect in patients with a history of hepatocellular carcinoma (HCC).

New Findings

DAA therapy for patients who had a complete response to HCC therapy is associated with increased overall survival, as long as the patients had a sustained virologic response to DAA treatment, which reduces liver-related mortality.

Limitations

This was a retrospective cohort study with possible selection bias and residual confounding.

Impact

Patients with HCV-related HCC who have a confirmed complete response should be considered for DAA treatment, which can increase survival.

Abbreviations

AASLD

American Association for the Study of Liver Diseases

AFP

alpha fetoprotein

BCLC

Barcelona Clinic Liver Cancer

DAA

direct acting antiviral

ECOG

Eastern Cooperative Oncology Group

HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

IFN

interferon

NASH

nonalcoholic steatohepatitis

SBRT

stereotactic body radiation therapy

SVR

sustained virological response

TACE

transarterial chemoembolization

TARE

transarterial radioembolization