Epidemiology and Outcomes of Hepatitis C Infection in Elderly US Veterans

H. B. El-Serag; J. Kramer; Z. Duan; F. Kanwal

Disclosures

J Viral Hepat. 2016;23(9):687-696. 

In This Article

Results

Study Cohort

The study cohort included 161,744 patients with HCV who met our study criteria (Fig. 1). They were followed for 8 32 369 PY, with a mean of 5.1 years (SD, 2.5 years). Their mean age was 52.5 years (SD 7.5 years); 36.8% were in the younger age group (20–49 years), 57.6% were in the middle-age group (50–64 years), and 5.6% were in the oldest group (65–85 years). Most were men (97.1%), 47.8% were white, and 29.3% were African Americans. Approximately 11.5% had diabetes, 27.3% had a BMI ≥30, 51.6% had a diagnosis of alcohol abuse, and 4.1% had HIV co-infection. A total of 15.7% received antiviral treatment, and 5.1% achieved an SVR12.

Figure 1.

Study flow diagram of patient selection for study cohort.

There were significant demographic and clinical differences among the age groups (Table 1). Elderly patients (65–85 years) were disproportionately white and included more men than the 2 younger groups. Elderly patients were also less likely to be obese (BMI≥30) and had a much lower prevalence of alcohol-use disorder, HIV and/or hepatitis B virus coinfection; however, they were more likely to have diabetes and a high Deyo comorbidity index than younger patients with CHC. Elderly patients were much less likely to have HCV genotype 1 infection than younger patients yet were less likely to receive HCV treatment and have an SVR12.

Association Between Age Groups and Risk of Incident Cirrhosis

After we excluded prevalent cases of cirrhosis, there were 1 49 339 patients in the study cohort, with an overall follow-up of 7 45 787 PY. A total of 13 877 (8.96%) patients developed cirrhosis, for an incidence rate of 18.6 per 1000 PY. The incidence rates of cirrhosis were slightly higher in the 50–64 age group than in the younger and older age groups (Table 2).

Older patients (50–64 and 65–85 years old) had a significantly higher risk of cirrhosis than younger patients (20–49 years old) (log-rank test P value <0.0001) (Fig. 2). In univariate Cox proportional hazard models, older patients aged 50–64 and 65–85 years were 1.30 (95% CI = 1.25–1.35) and 1.01 (95% CI = 0.93–1.10), respectively, times more likely to develop cirrhosis than 20- to 49-year-old patients. After adjusting for prespecified demographic, clinical and treatment factors, we found that the risk for cirrhosis was significantly elevated in both groups of older patients (HR for 50–64 years old=1.22 [1.17–1.27] and 65–85 years old=1.14 [1.00–1.29]) compared with younger patients (Table 3).

Figure 2.

Cumulative hazard of cirrhosis starting 12 months following HCV diagnosis stratified by 3 age groups.

Association Between Age Groups and Risk of Incident HCC

After we excluded prevalent cases of HCC, there were 1 60 971 patients in the study cohort. Of these, 3750 (2.33%) patients developed HCC over 8 25 802 PY of follow-up, for an incidence rate of 4.54 per 1000 PY. The incidence rates for HCC were considerably higher among older versus younger patients with CHC (Table 2).

Older patients (50–64 and 65–85 years) were significantly more likely to develop HCC than younger patients (20–49 years old) (log-rank test P value <0.0001) (Fig. 3). In univariate Cox proportional hazard models, older patients aged 50–64 and 65–85 years were 2.46 (95% CI = 2.28–2.66) and 3.71 (95% CI = 3.28–4.20), respectively, times more likely to develop HCC than 20- to 49-year-old patients. Although slightly attenuated, these associations remained after adjusting for prespecified demographic, clinical and treatment factors (HR for 50–64 years old=1.97 [95% CI = 1.81–2.16] and 65–85 years old=2.44 [95% CI = 1.99–2.99] vs 20–49 years) (Table 3).

Figure 3.

Cumulative hazard of hepatocellular carcinoma starting 12 months following HCV diagnosis stratified by 3 age groups.

Association Between Age Groups and Risk of Mortality

As expected, as seen in Table 2 and the cumulative hazard curves, older patients (50–64 and 65–85 years) were significantly more likely to die than younger patients (20–49 years) (log-rank test P value <0.0001) (Fig. 4). In addition, elderly patients were more likely to die than younger patients in both unadjusted and adjusted analyses (adjusted HR for 50–64 years old=1.39 [1.35–1.42] and 65–85 years old= 2.09 [1.98–2.22] vs 20–49 years old) (Table 3).

Figure 4.

Cumulative hazard of death starting 12 months following HCV diagnosis stratified by 3 age groups.

Association Between Antiviral Treatment and CHC Outcomes

Elderly patients received antiviral treatment less frequently than younger patients (3.8% vs 14.8% and 19.1%, P < 0.0001). In a multivariate logistic regression model adjusting for age-related differences, older patients were still less likely to receive treatment (OR for 50–64 vs 20- to 49-year-olds=0.88 (0.85–0.91) and OR for 65–85 vs 20–49 = 0.38 [0.33–0.44]). However, there was no statistical difference (P = 0.12) in SVR12 among those who were treated when stratified by age group 33.2%, 32.1% and 33.5% for 20- to 49-year-olds, 50- to 64-year-olds and 65- to 85-year-olds, respectively. Overall, treatment resulting in SVR12 (compared with no treatment) was associated with a reduction in the risk of all 3 outcomes (Table 3), but there was a small but statistically significant elevated risk of HCC associated with treatment failure compared to no treatment (adjusted HR = 1.17, 95% CI 1.06–1.28). In addition, treatment resulting in SVR12 was associated with a significant reduction in the risk of incident cirrhosis compared with no treatment receipt among patients in the 2 younger age groups. However, this was not the case in the oldest age group (adjusted HR: 1.14 (0.63–2.08) (Table 4). We found a similar association between treatment and HCC; treatment resulting in SVR12 was associated with a significant reduction in HCC risk in all 3 age groups compared with no treatment, although these associations were significant only for the 2 younger age groups but not the oldest group. Finally, SVR12 was strongly associated with a reduced risk of overall mortality. This survival benefit was consistent across all age groups [adjusted HRs for SVR12 vs no treatment were 0.34 (95% CI = 0.30–0.38), 0.32 (95% CI = 0.29–0.35) and 0.36 (95% CI = 0.24–0.53) for the 20–49, 50–64 and 65–85 year groups, respectively]. When we limited the analysis to those who received treatment, SVR12 compared to treatment receipt with no SVR12 was significantly associated with a reduced risk of developing cirrhosis (HR = 0.34; 95%CI: 0.18–0.66). In addition, the effect of SVR12 on the other outcomes was stronger in all 3 age groups; however, the association of SVR12 with HCC in the older age group was still not significant (Table 4).

Secondary Analyses of 65–74 and 75 and 85 Years Age Groups

We evaluated CHC patients 65–74 years (n = 6422) and 75–85 years (n = 2648). These two age groups had prevalent cirrhosis in 7.2% and 5.8% and HCC in 1.1% and 1.3%, respectively. The incidence rates of cirrhosis in the 2 age groups were 17.7 and 10.7 per 1000 PY, respectively, and those of HCC were 9.4 and 5.8 per 1000 PY, respectively. In the multivariable Cox PH, the risk of HCC was considerably elevated in these two age groups compared to those 20–40 years of age (Table 5). Treatment with or without SVR was not associated with cirrhosis, HCC or death; however, the number of events was very low to allow the calculation of precise estimates.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....