Real-world responses to HCV treatment among US veterans match best clinical trial results

Direct-acting antiviral (DAA) treatment is curing people of hepatitis C infection in clinics at similar rates to those seen in clinical trials, and there don’t seem to be major differences between drug regimens, according to results of a large population study presented at the 2016 AASLD Liver Meeting in Boston this weekend.

Clinical trials tend to show the best-case scenario for efficacy of new drugs. Patients in clinical trials are carefully selected from populations who are already attending clinics – by definition, a highly motivated group of patients. Furthermore, patients recruited to clinical trials cost money to recruit and monitor so they will be followed up carefully, and in many cases patients are getting free treatment, maximising the incentive to stay in care. All these factors tend to promote better outcomes in clinical trial populations than in subsequent 'real world' clinic populations.

A review of the Veterans Affairs Cohort – military veterans receiving care through US Veterans Affairs’ (VA) clinics – finds that people receiving DAA treatment in those hospitals are being cured of hepatitis C at similar rates to those seen in clinical trials of the drug combinations in widespread use today. Clinical trials that led to licensing of the DAA combinations recommended for treatment reported sustained virologic response (SVR12) rates above 90% for genotype 1, around 75% for genotype 3, and somewhat lower for each genotype in cases where people had cirrhosis or previous experience of unsuccessful treatment.

The study also found that an 8-week regimen of sofosbuvir/ledipasvir (Harvoni) is just as effective as a 12-week regimen for those who qualify to take it – genotype 1 patients without cirrhosis or previous experience of treatment who have a viral load below 6 million IU/ml.

The study population comprised 17,487 people with hepatitis C virus (HCV) infection who started therapy with DAA agents without interferon between January 2014 and June 2015 at 167 VA medical centres in the United States. Follow-up was to April 2016. The main study outcome was sustained virological response (SVR). This was analysed according to HCV genotype and cirrhosis status.

The study population was 97% male, 52% white, 29% black, 5% Hispanic, and had a high prevalence of cirrhosis (30%) and decompensated cirrhosis (8.3%). Genotype 1 was the predominant form of hepatitis C infection (80%), with 12% having genotype 2, 7% with genotype 3 and 1% with genotype 4.

Regimens used by the participants were:

• Sofosbuvir (SOF)/ribavirin: n = 2986
• Ledipasvir/sofosbuvir (LDV/SOF): n = 11,327
• Paritaprevir/ritonavir/ombitasvir and dasabuvir (PrOD): n = 3174.

The study measured sustained virologic response rates 12 weeks after completion of treatment (SVR12). If SVR12 data were missing, the SVR12 result was imputed from SVR4 measurements (5.7% of cases).

	% treated	SVR 12%
Genotype 1
SOF/LDV	58%	92.%
SOF/LDV/riba	19%	92%
PrOD	6%	94.9%
PrOD/riba	17%	92.5%
Genotype 2
SOF/riba	100%	86.2%
Genotype 3
SOF/riba	57%	77.9%
SOF/LDV/riba	31%	87%
SOF/PEG/riba	11%	70.6%
Genotype 4
SOF/LED/riba	77%	87.6%
PrOD/riba	23%	96.4%

The study also looked at the performance of DAAs in people with cirrhosis. Whereas in people with genotype 1 infection there was little difference in SVR12 between those with or without cirrhosis (90.6 vs 93.6%), virologic response was considerably lower in people with cirrhosis with genotype 2 (77.3 vs 89.1%) and genotype 3 (65.7 vs 81.6%). Response was also lower in people with cirrhosis with genotype 4 (83.9 vs 91.5%), although less markedly so.

A similar pattern was evident when treatment responses were compared in treatment-experienced and previously untreated people. There was no difference in virological response in people with genotype 1, but SVR rates in treatment-experienced people with genotypes 2 or 3 were around 8% lower than in previously untreated people (80.2 vs 88% for genotype 2, and 77.5 vs 69.2% for genotype 3).

The study also looked at treatment outcomes in 1975 people who qualified for an 8-week course of sofosbuvir/ledipasvir – previously untreated people without cirrhosis with viral load below 6 million IU/ml. 95.1% of these people achieved SVR12, compared to 95.8% of those treated with sofosbuvir/ledipasvir for 12 weeks.

Independent predictors of treatment failure included genotype 2 or 3 infection compared to genotype 1 infection, and among those with genotype 2 or 3 infection, treatment experience. Male sex, black or Hispanic ethnicity, diabetes, low platelet count, low serum albumin and elevated bilirubin all independently predicted failure to achieve SVR, as did cirrhosis.