The impact of direct-acting antiviral agents on liver and kidney transplant costs and outcomes

1 INTRODUCTION

Direct-acting antivirals (DAAs) have dramatically altered the care of patients with chronic hepatitis C virus (HCV) infection.1-6 For the last 2 decades, chronic HCV infection was the leading indication for liver transplant (LT) in the United States. Treatment of HCV infection prior to 2014 consisted primarily of interferon- and ribavirin-based treatment regimens, which had limited efficacy, were associated with debilitating side effects, and, generally, were contraindicated in patients with decompensated cirrhosis or, in the case of ribavirin, advanced chronic kidney disease (CKD).7, 8 DAAs reduce morbidity and result in sustained virological response 12 weeks after completing treatment (SVR12) rates >94% for most genotypes in both compensated and decompensated patients.4, 9, 10 DAAs have also been safely used in the post LT setting to prevent recurrent inflammation and fibrosis, which was universal in the absence of effective pretransplant treatment. Historically, recurrence of HCV in the liver graft resulted in cirrhosis in 20%-30% of LT recipients and fibrosing cholestatic hepatitis C in 2%-9% within the first 12 months.11, 12 DAAs have been shown to achieve SVR12 after LT in multiple clinical trials, although the impact of DAAs on longer-term LT and KT outcomes has not been reported.12

Among patients with end-stage renal disease (ESRD), HCV prevalence is 5 times greater than in the general population.13, 14 Historically, HCV-infected ESRD patients on dialysis were 60% more likely to die than their noninfected counterparts. Prior to DAAs, interferon-based regimens had low levels of efficacy and high rates of intolerability in this population and were generally contraindicated in HCV-infected KT recipients because of unacceptable rates of rejection and allograft dysfunction. Because most HCV-infected KT recipients in the pre-DAA era were therefore either untreated or intolerant to pretransplant interferon/ribavirin, these patients experienced higher rates of graft loss and death than noninfected patients.15 In contrast, DAAs are both safe and well tolerated in patients with advanced CKD, with SVR12 rates from 95% to 98%.16, 17 DAA treatment in KT recipients successfully eradicates the virus without negatively affecting graft function in clinical series.18, 19

While DAA treatment is revolutionary, access to it has been hampered nationally by its high cost.20-22 Initial regimens resulted in total healthcare expenditures exceeding $100 000 USD (United States dollars) per treatment course. As more DAAs have entered the market and competition has increased, costs have diminished, but remain in excess of $25 000 per course, depending on specific agent. Consequently, major private payers developed preauthorization processes that initially restricted use to patients with defined clinical conditions such as demonstrated hepatic fibrosis, cirrhosis, or advanced kidney disease, despite evidence suggesting clinical benefits even in patients without advanced liver disease.23 While long-term economic analyses suggest that DAAs are cost effective, few are cost saving despite reducing the need for transplant.24-26 Furthermore, the cost savings are accrued far in the future when many patients have changed health insurance, diminishing enthusiasm for broader treatment of patients without qualifying conditions.

While clinical trial experience with DAAs in the posttransplant LT and KT populations have demonstrated high SVR rates, no large-scale, population-based assessment of access to DAA treatment, effects on longer-term transplant outcomes, or increases in the cost of treatment has been conducted. Using a unique data set linking pharmacy claims data and transplant registry outcomes, we developed a national cohort of HCV-positive transplant recipients with sufficient power to assess DAA use in LT and KT recipients before and after introduction of DAAs, characteristics associated with posttransplant HCV treatment, and the independent impact of these medications on patient and graft outcomes.

2 METHODS

3 RESULTS

Between 2007 and 2016, 58 509 LTs were performed; pharmacy claims data for at least 1 year after transplant were available for 41 690 (71%) of these (Table 1). Among patients with claims, 15 671 (38%) were HCV donor negative and recipient positive (D-/R+), 1468 (3.5%) were D+/R+, and 47 (0.1%) were recorded as D+/R-. In the same period, 157 873 KTs were performed; pharmacy claims data were available for 121 800 (71%). In this population, 3681 (3.0%) were D-/R+, 1738 (1.4%) were D+/R+, and 294 (0.2%) were D+/R-.

3.1 Use of HCV treatment

Overall, 12.9% of patients undergoing LT received HCV medications within 3 years after transplant. Among serologic subgroups, treatment prevalence varied from 0.75% of HCV D-/R- patients to 35.2% of D+/R+ patients (Figure 1A). In the pre-DAA era, 4.0% of HCV D+/R+ and 5.5% of D-/R+ received HCV treatment within 1 year. Post-DAA, 13.5% of D+/R+ patients and 4.4% of D-/R+ patients received treatment within 1 year. By 3 years, 17.0% D+/R+ and 17.1% D-/R+ patients received treatment in the pre-DAA era. Post-DAA, 15.3% of D+/R+ and 5.5% of D-/R+ recipients had received treatment at the end of follow-up (2.3 years and 2.6 years, respectively). Clinical factors associated with use of HCV treatments pre-DAA include male sex (adjusted hazard ratio [aHR] _0.830.91_1.00), diabetes (aHR _0.800.89_0.98), hypertension (aHR _1.031.13_1.25), higher model for end-stage liver disease (MELD) score (15-30, aHR _1.011.10_1.20; >30, aHR _1.001.18_1.38 vs MELD<15), and black race (aHR _0.720.81_0.90). Post-DAA, donor HCV+ status (aHR _1.661.95_2.29), higher MELD score (15-30, aHR _1.231.42_1.65; >30, aHR _1.431.81_2.28), and black race (aHR _0.650.79_0.96) were associated with the likelihood of HCV treatment (Table S2A).

The impact of payer on LT recipient access differed by DAA era. Pre-DAA, HCV+ LT recipients with private insurance were equally likely to be treated with HCV medications (Figure 2A). After adjustment for donor and recipient characteristics, HCV+ recipients with private insurance were somewhat more likely to be treated (aHR _1.011.10_1.19). However, in the post-DAA era, both D+/R+ and D-/R+ recipients with private insurance were significantly less likely to receive HCV treatment (Figure 2B). Nearly 20% of D+R+ recipients with public insurance, compared with 11% of recipients with private insurance (P < .0001), received treatment. After adjustment for other donor and recipient characteristics, D+/R+ recipients with private insurance were 45% less likely than publicly insured recipients to receive HCV treatment (aHR _0.430.55_0.71, P < .0001). Among D-/R+ recipients, 6.2% with public insurance received DAA treatment, compared with 5.0% with private insurance (aHR _0.740.84_0.96, P = .01).

As expected, HCV treatment was substantially less common in KT recipients; overall, 1.4% received HCV treatments within 3 years of transplant. Utilization was highest, at 22.8%, for D+R+ recipients; 11.3% of D-/R+ recipients received treatment (Figure 1B). Pre-DAA, 2.8% of D-/R+ and 5.5% of D+R+ recipients received treatment with HCV medications within 3 years. Post-DAA, 3.9% of D-/R+ and 12.9% of D+/R+ recipients received treatment (P < .0001). Differences by payer in the post-DAA era were modest. Among D+R+ recipients, 13.5% with public insurance received treatment compared with 11.5% with private insurance (P = .35). Among D-/R+ recipients, 4.3% with public insurance received treatment, compared with 2.9% with private insurance (P = .05). After adjustment for donor and recipient characteristics, HCV treatment in R+ D+ KT patients was more likely than D- patients before and after DAA introduction (Pre-DAA: aHR _1.432.02 _2.84 vs Post DAA aHR: _2.312.93 _3.71). In the post-DAA era, other patients whose race was not black or white and older patients were more likely to be treated (Table S2b).

3.2 Impact of HCV treatment on patient and graft survival

LT outcomes among all HCV+ recipients have improved in the post-DAA era (Figure 3A). Among all HCV+ LT recipients, after adjustment for donor and recipient characteristics, HCV requiring treatment pre-DAA was associated with a significantly higher risk of posttransplant mortality (aHR _1.471.68_1.91, P < .0001) and all-cause graft failure (aHR _1.641.85_2.10, P < .0001) (Figure 4A). In contrast, there was not increased risk of death or graft failure post-DAA (aHR for death: _0.740.94_1.19, P = .61; aHR for graft failure: _0.740.94_1.19, P = .62) (Figure 4; Table S3A). In D-/R+ recipients, HCV treatment pre-DAA was not associated with an increased risk of death (P = .96) or graft failure (P = .40). Among D+/R+ post-DAA, the reduction in the risk of death (aHR _0.340.67_1.32, P = .25) and graft failure (aHR _0.320.64_1.26, P = 0.20) within 3 years of transplant was not statistically significant (Table S3B).

Among HCV+ KT recipients, there was no significant improvement in unadjusted patient survival post-DAA compared with pre-DAA (Figure 3B). In the adjusted analysis, pre-DAA HCV treatment was associated with a non–statistically significant increased risk of death (aHR_0.67 1.80_4.87, P = .25) and graft failure (aHR _0.711.61_3.62, P = .24) (Figure 4B; Table S4A). There was a nonsignificant protective effect of treatment for KT recipients post-DAA for mortality (aHR _0.370.66_1.18, P = .16) and graft failure (aHR _0.490.79_1.27, P = .33) (Table S3; Table S4A). Among D+R+ KT recipients, DAA treatment was associated with nonsignificant reduction in graft failure (aHR _0.270.54 _1.07, P = .08) and a statistically significant lower risk of death after KT (_0.190.43_0.95, P = .04). (Table S4B)

4 DISCUSSION

Introduction of DAAs has markedly improved care for LT and KT recipients with chronic HCV infection. Among LT recipients, our data support findings from smaller clinical studies that suggest that DAAs in the posttransplant setting improve posttransplant outcomes. In the pre-DAA era, HCV reinfection after LT was universal and treatment was generally reserved only for recipients with early aggressive recurrence resulting in fibrosing cholestatic hepatitis C or other complications. Treatment was often ineffective, resulting in accelerated graft loss and death, as confirmed in this analysis. DAAs, by contrast, are well tolerated and recommended for all actively infected patients.7, 9, 28 Based on national data and early follow-up, there is a consistent pattern of improved outcomes in LT recipients with HCV treated with DAAs. In KT recipients, treatment of HCV is less common than in LT recipients. However, a similar protective effect is noted in D+/R+ KT recipients, who have reduced mortality and, likely, graft loss if they receive posttransplant DAAs. Access to these expensive medications, however, appears limited, as <20% of D+/R+ LT and KT recipients receive them, and DAA treatment rates appear to be even lower among privately insured patients.

The availability of effective posttransplant HCV treatment allows LT and KT recipients to time treatment to achieve the greatest benefit.24, 29 Pretransplant patients who are precirrhotic or who have well-compensated disease may benefit from early HCV treatment with stabilization or regression of chronic liver disease, potentially avoiding LT entirely. Ahmed et al recently reported a Markov analysis comparing delayed or immediate HCV treatment among patients waiting for LT.30 The benefit of HCV treatment varied according to clinical condition. Among patients with decompensated liver disease, pre-LT treatment was associated with improved survival (9.3 vs 8.7 quality-adjusted life-years) but higher costs ($304 800 USD vs $283 789 USD). Results were sensitive to MELD score at evaluation, such that pre-LT treatment was of less benefit to patients with greater decompensation. Decompensated patients may be further disadvantaged by pre-LT treatment, because the availability of HCV-infected donor organs has historically been significantly greater, allowing earlier transplant. Among patients with stable liver disease who undergo transplant due to hepatocellular carcinoma, immediate HCV treatment was associated with a gain of 11.5 quality-adjusted life-years vs 10.4 for delayed treatment; however, healthcare expenditures were increased by $82 000 USD per patient. Our data suggest that HCV treatment after transplant is no longer associated with a decrement in graft or patient survival, and may in fact be protective. In light of the organ shortage, reserving treatment of patients with decompensated cirrhosis until after LT may be clinically and economically beneficial, provided they have access to DAAs after transplant.

Dialysis patients infected with HCV are frequently encouraged to seek DAA treatment before transplant despite demonstrated efficacy of DAA in the posttransplant setting.13, 31 In recipients with available, compatible live donors, this strategy can be justified as it allows viral clearance prior to transplant, thereby avoiding potential posttransplant drug-drug interactions and mitigating risk of any early HCV-related complications (for example, new-onset diabetes). However, patients who are waiting for deceased donor organs may delay HCV treatment until after transplant to allow greater access to HCV+ donor organs, which is associated with a marked reduction in expected waiting times.32 Importantly, these data are the first to suggest improved patient and allograft survival among HCV+ KT patients who are treated with DAAs early posttransplant, yet <15% of patients who receive a HCV D+ organ received timely DAA treatment.

The cost of HCV therapy with DAAs increased markedly compared with interferon and ribavirin. The cost of care is further increased for HCV patients with co-existing organ dysfunction. Patients with CKD or ESRD who require HCV treatment incur fourfold higher per member per month costs than HCV patients without ESRD ($5481 USD vs $1922 USD, P < .001). Despite these high costs, DAA treatment in the majority of patients has been found to be generally cost effective, with some regimens characterized as cost saving in the nontransplant population.33, 34 While early treatment regimens, such as those identified in this data set, were very expensive, competition from newly released agents has dramatically reduced the cost of treatment in an effort to increase market share. Treatment costs have also been reduced after shorter durations of care and have been demonstrated to be equally effective. A recent meta-analysis assessing the cost effectiveness of pretransplant treatment revealed that 71% of analyses found second-generation DAAs to be cost saving and 22% cost effective, while only 7% were not cost effective. Further savings may be expected through reduced graft loss and need for retransplantation (LT) and HCV-related kidney disease. Because use of kidneys from HCV-infected donors can markedly reduce wait times for transplant for HCV-infected kidney candidates, the cost-saving realized by a shorter dialysis burden in these patients must be accounted for as well, especially in regions associated with lengthy waiting times.35 It is therefore unclear why access to treatment with DAAs should be limited for transplant recipients, who are even more likely to benefit from treatment than dialysis patients.24, 36

This analysis has several key limitations. First, we lack information regarding viral load and genotype in both the pre- and posttransplant setting. Therefore, it is impossible to determine definitively which patients were actively infected at the time of transplant. This is particularly important in the current era, as HCV seropositivity in the absence of a positive NAT is consistent with virological cure, either related to prior administration of efficacious therapy or spontaneous clearance. This may account for the lower rate of utilization of HCV treatment in the post-DAA era among D-/R+ recipients. Second, the price of HCV medications has fallen since 2016 as new medications have been developed and marketed. This competition has resulted in somewhat lower costs for treatment regimens than reported in this analysis. Therefore, the reduced treatment access for privately insured patients that we found may have improved, as insurance companies start to develop new policies relating to DAAs. Third, despite our assembling the largest cohort reported of treatment in HCV+ recipients, the number of presumptively viremic patients (D+/R+) is still limited in this national study. This may limit inferences about the impact of treatment and comparisons of treatments. Additional data with longer periods of follow-up may provide important insight into the benefits of these treatments. Finally, among KT patients, the severity of HCV-related liver disease, an important outcome determinant, was unknown. However, in this large-scale analysis, it is unlikely to have differed between eras.

In conclusion, HCV treatment patterns have changed with the introduction of highly effective DAAs. Fewer HCV D-/R+ LT patients are treated in the posttransplant setting, because many may have been treated prior to transplant, while rates have increased for D+/R+ patients. In contrast to the reduced survival observed in LT patients treated in the pre-DAA era, HCV treatment with DAAs was not associated with poor outcomes in D+/R+ HCV+ LT recipients. Future studies capturing larger samples may demonstrate improved survival following DAA treatment. In KT patients, HCV treatment remains rare, but is more common in the DAA era and appears to improve outcomes in HCV+ KT recipients. Finally, DAAs were associated with a 10-fold increase in the cost of HCV treatment after LT and KT compared with medications in the pre-DAA era. These data suggest that patients with private insurance have reduced access to DAAs, which may result in higher rates of graft failure and death in patients who receive HCV+ donor organs or remain viremic at the time of transplant. Further study is needed to determine the optimal time and treatment strategy for transplant candidates and recipients infected with HCV.

DISCLOSURE

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.