Progress in the elimination of hepatitis C virus infection: A population-based cohort study in Spain

Background The World Health Organization set targets to eliminate hepatitis C virus (HCV) infection through detection and treatment of all cases by 2030. This study aimed to describe the progress and difficulties in the elimination of HCV infection in Navarra, Spain. Methods Using electronic healthcare databases, we performed a population-based prospective cohort study to describe changes in the prevalence of diagnosed active HCV infection at the beginning of 2015 and the end of 2017, the rate of new diagnoses and the rate of post-treatment viral clearance (PTVC) during this period. Results At the beginning of 2015 there were 1503 patients diagnosed with positive HCV-RNA, 2.4 per 1000 inhabitants, and at the end of 2017 the prevalence had decreased by 47%. In the study period, 333 (18 per 100,000 person-years) new positive HCV-RNA cases were detected, but only 76 (23%; 4.2 per 100,000 person-years) did not have anti-HCV antibodies previously detected. Prevalent cases and new diagnoses of active infection were more frequent in men, people born in 1950–1979, HIV-infected patients and in those with lower income levels. Among patients with HCV-RNA, 984 achieved PTVC (22.7 per 100 person-years). PTVC was less frequent in patients born before 1940, in immigrants and in patients with lower income levels. Conclusions The prevalence of diagnosed active HCV infection has dropped by almost half over three years, because the number of patients with PTVC was much higher than the number of new diagnoses. Interventions specifically targeted at population groups with less favourable trends may be necessary.

In 2016 the World Health Organization set targets to eliminate HCV infection as a major public health threat by 2030 [12]. These targets should be translated into operational objectives in geographical areas and population groups, including the detection of undiagnosed HCVinfections, diagnosis of active infections among HCV positive patients and antiviral treatment of active infections [13,14]. In 2015, the National Health System in Spain launched a plan to progressively provide access of all HCV-infected patients to the DAAs regimens [15], without treatment restrictions since early 2017. In Spain a high proportion of active HCV infections have already been diagnosed. Some infections may have an incomplete diagnosis, if the detection of anti-HCV has not been followed by determination of the HCV-RNA, making it difficult to establish whether active infection continues or has cleared spontaneously. Finally, some active infections may remain undiagnosed [16].
Good practices on aspect related to HCV-elimination have been described [17,18], but comprehensive studies evaluating the progress towards the HCV-elimination are necessary. In order to monitor the process of eliminating HCV infection in Spain, we quantified the changes in the prevalence of diagnosed active HCV infection, the rate of new diagnoses and the rate of post-treatment viral clearance (PTVC) according to socio-demographic characteristics of the population during the first three years when DAAs were widely used. The rate of PTVC was considered as a summarized indicator of the access, adherence and effectiveness of antiviral treatment.

Study population and design
The study population included people with stable residence in the Navarra region, Spain, and covered by the Regional Health Service (about 615,000 inhabitants). Using electronic healthcare databases we performed a prospective cohort study to describe and compare the prevalence of HCV-RNA positive diagnosed cases at the beginning of 2015 and at the end of 2017. We also monitored new HCV-RNA diagnoses and the PTVC during this follow-up period.
4% of the population was annually tested for anti-HCV in Navarra with an increased screening effort over time. Some of these tests were done in persons with suspected HCV infection, and others were done within screening protocols (risk exposure history, prison inmates, pregnant women, blood donors, etc.). The HCV diagnoses included serology for anti-HCV (ARCHI-TECT, Abbott Laboratories, Wiesbaden, Germany), and RT-PCR for HCV-RNA (Cobas HCV, Roche Diagnostics, Mannheim, Germany) in all patients with a positive anti-HCV result.
A prevalent active HCV infection was defined as HCV-RNA detection in the most recent test previous to 1st January 2015. Patients testing positive for HCV-RNA for the first time between 1st January 2015 and 31st December 2017 were considered new diagnoses of active HCV infection regardless of when they could have acquired the HCV infection. For some analyses we considered those patients with a first positive result for both anti-HCV and HCV-RNA during the study period.
The information on HIV infection was obtained from the regional enhanced surveillance system [19]. From the databases of hospital pharmacies we obtained the HCV treatments administered up to the end of 2017. The guidelines of the European Association for the Study of the Liver panel and the Spanish health authorities were followed in the treatment of patients [20]. Since 2015, oral DAAs have been widely used and replaced interferon containing treatments. In HCV-RNA positive patients, viral clearance was defined as undetectable HCV-RNA in the last test during the study period. We defined PTVC as undetectable HCV-RNA in all tests performed after the antiviral therapy, and considered it the endpoint indicator of a process that summarizes early access, good adherence and high effectiveness of the DAAs therapies among the patients who initially had an active HCV infection.
From administrative databases we obtained the sex, birth year, origin country (native or immigrant), municipality of residence and income level. Place of residence in municipalities with more than 10,000 inhabitants was considered urban and all others were considered rural. The income level was established in the categories used by the Spanish Health System: low (no incomes or great dependence), lower-middle (<18,000 €/year) and high income level (>18,000 €/year).

Statistical analysis
The prevalence of HCV-RNA positive tests in the population was calculated at the beginning of 2015 and the end of 2017, and the percentage of change between the two estimates was obtained. The chi-square test was used to compare percentages. Multivariate logistic regression models were used to identify differences in the prevalence, in terms of adjusted odds ratios (ORs) with their 95% confidence intervals (CIs).
Among the initially susceptible population we calculated the rate of new HCV-RNA detection between 2015 and 2017. We also calculated the rate of new diagnoses of positive anti-HCV and HCV-RNA. Among the population with a diagnosis of active HCV infection at the beginning or during the follow-up, we obtained the incidence rate of PTVC during the period 2015-2017. The factors associated with the rate of new diagnoses of active HCV infection and with the rate of PTVC were analysed using Poisson regression models to estimate the adjusted rate ratio (RR) and its 95% CI. Multivariate analyses were adjusted by sex, birth decade, origin country, place of residence, income level, HIV infection and year of diagnosis or PTVC.

Ethical considerations
The study protocol was approved by the Navarra Ethical Committee for Clinical Research. The data were obtained from epidemiological surveillance and electronic healthcare databases and were analyzed anonymously.

Prevalence of diagnosed active HCV infection
On first January 2015 there were 1503 patients with a HCV-RNA positive result in their last determination (2.4 per 1000 inhabitants). The prevalence was almost double in men (3.1) than in women (1.6; p<0.001). The highest prevalence was observed in people born in the 1960s (8.4 per 1000 inhabitants), which included more than half of all cases (52%). Birth cohorts born in the 1950s and 1970s also exceeded 2 cases per 1000 inhabitants, and only in birth cohorts born after 1980 the prevalence was below 1 per 1000 inhabitants. About 78% of cases had lower-middle or low income level, and they showed the highest prevalence, 2.5 and 11.1 per 1000 inhabitants, respectively. Among people infected with HIV the prevalence of active HCV infection was 30% and among people with active HCV infection the prevalence of coinfection reached 19% (Table 1).
In the multivariate analysis, the prevalence of positive HCV-RNA in 2015 was lower in women than in men (OR = 0.6, 95% CI 0.5-0.6), and was lower in immigrants than in the native population (OR = 0.5, 95% CI 0.4-0.6), while the factors associated with higher prevalence were HIV infection (OR = 89.0, 95% CI 75.3-105.1), birth before 1990, urban residence (OR = 1.4, 95% CI 1.3-1.6) and lower-middle (OR = 2.1, 95% CI 1.8-2.4) and low income level (OR = 8.2, 95% CI 6.9-9.8) ( Table 2). From the beginning of 2015 until the end of 2017, active HCV infection was detected in 257 patients who had been diagnosed with anti-HCV in previous years, 76 persons were newly diagnosed with anti-HCV and HCV-RNA, and 126 patients previously diagnosed with active HCV infection arrived in the region during the study period. On the other hand, 57 active HCV-infected patients left the region, 65 died, 69 became spontaneously HCV-RNA negative, and 984 had completed the treatment and became HCV-RNA negative (Fig 1).
As a result of that, the prevalence of diagnosed HCV-RNA declined by 47%, from 2.4 to 1.3 cases per 1000 inhabitants (p<0.001). Decreases in the prevalence of HCV-RNA were observed among almost all population groups, but with different magnitude. The decline in males was 50% and in females 39%. The prevalence decreased around 50% in the cohorts born between 1940 and 1969, but non-statistically significant declines were observed in cohorts born before 1940 and after 1980. The decline in the prevalence of active HCV infection was more pronounced in the native population (50% vs. 24%), in HIV-coinfected patients (65% vs. 44%), and in the high income level category (Table 1).

New diagnoses of active HCV infection
From 2015 to 2017 a total of 333 patients were detected with active HCV infection for the first time (18.5 per 100,000 person-years); the majority of them (n = 257; 77%) had tested positive for anti-HCV antibodies in previous years, but had not been tested for HCV-RNA. In general, the predictors of new diagnoses of active HCV infection were the same as those described for the prevalence of active HCV infection; furthermore, the rate of new diagnoses of HCV-RNA presented a statistically significant decline in 2017 as compared to 2015 (RR = 0.7, 95% CI 0.6-1.0) ( Table 3 Table 4). 95% CI 0.1-1.0). PTVC was less frequent in immigrants than in the native population (RR = 0.7, 95% CI 0.6-0.9), and among the population with lower-middle (RR = 0.8, 95% CI 0.7-0.9) or low income level (RR = 0.7, 95% CI 0.5-0.8). The rate of PTVC increased to 29.7 per 100 person-years in 2017 (RR = 1.5, 95% CI 1.3-1.8) ( Table 5).

Discussion
In 2015, when the use of the second generation of DAAs was expanded in Spain, the prevalence of diagnosed active HCV infection in Navarra was high and showed important sociodemographic differences. In the following three years this prevalence has decreased to near half, thanks to the large number of patients with PTVC. An important reduction in the Progress to HCV elimination in Spain prevalence was observed in all population groups, although the socio-demographic inequalities have not disappeared. The rate of diagnoses of active HCV infection was 18 per 100,000 person-years, mainly as result of the first HCV-RNA testing in patients diagnosed with anti-HCV years ago. The decline in this rate of diagnoses of active infection in 2017 suggests a progressive reduction in the number of patients with positive anti-HCV who were pending HCV-RNA testing. During the period 2015-2017 a considerable proportion of the population was tested for anti-HCV, but only 4.2 new diagnoses of anti-HCV with active infection per 100,000 person-years were detected. This is consistent with the moderate rate of undiagnosed HCV-infections that has been described in Navarra [16].
Males and people born between 1950 and 1979 presented the highest prevalence rates of active HCV infection by both, the beginning and the end of the study period, and they also showed the highest rates of new diagnoses of active infection. This distribution is consistent with the pattern of the epidemic of injecting heroin consumption in Spain, which was responsible for many of these infections [21,22]. It is also consistent with the recently reported pattern of transmission among men who have sex with men [23], and is not very different from the distribution of HCV infection described in other countries [24,25]. The prevalence of active infection was also high among the people born before 1950, who accounts for most cases of iatrogenic HCV transmission [3]. The rate of PTVC was significantly lower among people born before 1940. Although since early 2017 there is no restriction in the DAAs treatments in Spain, advanced age and the presence of comorbidities may reduce the priority of treatment in practice; therefore, some interventions aimed at improving the recruitment of these populations to be treated may be necessary.
A recent meta-analysis reported no major differences in the prevalence of anti-HCV among native and immigrant populations in Spain [26]. Our results describe a lower Progress to HCV elimination in Spain prevalence and rate of new diagnoses of active HCV infection in the immigrant population after adjustment for other socio-demographic characteristics. The high transmission rate during the past century in Spain may explain this finding [2]. In immigrants, a worse rate of PTVC was observed, which indicates late request of medical care, since all the study population had free health care. The higher prevalence of active HCV infection in urban residents is consistent with other studies, and may be due to the higher frequency of risk behaviour for HCV transmission in cities [23]. Independently of other variables, lower income levels were associated with a higher prevalence and rate of new HCV-RNA diagnoses, as well as a lower rate of PTVC. Similar findings were described in the interferon era and the progressive reduction of differences during the DAAs era [27], and other studies have found that HIV-coinfected patients with lower income were less likely to initiate treatment [28,29]. In our study the prevalence of active HCV infection declined less in the lower income level categories so that the pre-existing socio-demographic inequalities in the prevalence have remained. The association between income level and active HCV infection may have several explanations: some risk exposures for HCV infection, especially injecting drug use, were more frequent in the low socio-economic level; the manifestations of HCV chronic infection may cause occupational difficulties and a resulting decline in income level; mental problems or drug addiction may favour the HCV infection as well as a decline in income level; and lastly, income level may be associated with late request for medical care and worse compliance with treatment, even in countries where HCV treatment is free of charge for the patient [29,30].
HIV infection was associated with a much higher prevalence and rate of new diagnoses of active HCV infection, which is consistent with results reported by other authors [23]. However, no differences were observed in the rate of PTVC by HIV status, which may be due to the specific priority of HIV-infected patients for antiviral treatments [15].
Several limitations may affect these results. Undiagnosed active HCV infections would be the main limitation; however, in spite of the high number of analysed persons only 76 new diagnoses of active infection were detected in three years, which suggests that the number of undiagnosed infections is probably low [16]. The fact that some patients with positive anti-HCV had not been tested for HCV-RNA may have resulted in an underestimate of the prevalence of active HCV infections; however these cases decreased during the study period. We analysed the factors associated with PTVC, but spontaneous clearance was also observed in few patients [31]. We considered PTVC the change from detectable to undetectable HCV-RNA in all available tests after the treatment as a summarized indicator of access, adherence and effectiveness of antiviral treatment. This concept does not exactly match the definition of sustained virological response usually used in clinical trials or clinical studies as a treatment effectiveness outcome [32].
The situation described is rapidly changing, thus some results may have short-term validity and updates may be necessary. Since interventions have been introduced in Navarra to improve the hepatitis C control program, the observed differences may decline in subsequent years. Other factors, such as differential mortality and migrations may affect the prevalence of active HCV infection. The high mobility of some population groups makes them difficult to study, although the inputs and outputs of the population partially offset each other. Although the epidemiology of HCV infection and the introduction of DAAs have been relatively homogeneous in Spain, we can not rule out some geographical differences.

Conclusions
In a comprehensive population-based approach, a favourable trend in the prevalence of active HCV infection has been observed in Spain, with a reduction by nearly half after three years of use of DAAs. This may be explained by the fact that the rate of viral clearance has been much higher than the rate of new diagnoses. Even in a country with universal healthcare and HCV treatment free of charge for the patient, we found that lower income levels were associated with a higher prevalence, higher rate of new diagnosis of active HCV infection and a lower frequency of PTVC. Immigrants and older population also showed a lower rate of PTVC. These findings suggest the need of interventions specifically targeted at population groups with less favourable trends.