Implications of Hepatitis C Viremia vs. Antibody Alone on Transmission among Male Injecting Drug Users in Three Afghan Cities

Abstract

Objectives:

To assess differences between injecting drug users (IDUs) with HCV viremia compared to IDUs with HCV antibody (Ab) or no evidence of prior infection in three Afghan cities.

Methods:

IDUs in Hirat, Jalalabad, and Mazar-i-Sharif completed questionnaires and rapid testing for blood-borne infection including HCV Ab. HCV Ab was confirmed with recombinant immunoblot assay (RIBA); RIBA-positive specimens underwent reverse transcriptase polymerase chain reaction (RT-PCR) for HCV. Risk behaviors associated with viremia were assessed with site-controlled ordinal regression analysis.

Results:

Of 609 participants, 221 (36.6%) had confirmed HCV Ab, among whom 127 (57.5%) were viremic. HCV viremia prevalence did not differ by site (range: 41.7% – 59.1%; p=0.52). Among all IDUs, HCV was independently associated with HIV co-infection (AOR=7.16, 95% CI: 4.41 – 11.64), prior addiction treatment (AOR=1.95, 95% CI: 1.57 – 2.42), ever aspirating and re-injecting blood (AOR=1.62, 95% CI: 1.18 – 2.23), prior incarceration (AOR=1.60, 95% CI: 1.04 – 2.45), sharing injecting equipment in the last 6 months (AOR=1.35, 95% CI: 1.02 – 1.80), and aspirating and re-injecting blood (AOR=1.62, 95% CI: 1.18 – 1.99) in age and site-controlled ordinal regression analysis.

Conclusion:

HCV viremia was present in many participants with prior HCV infection and was associated with some injecting risk behaviors, indicating substantial risk for transmission. Current harm reduction programming should urgently improve HCV awareness and prevention among IDUs in Afghanistan.

Introduction:

Hepatitis C virus (HCV) is a serious health threat for injecting drug users (IDUs), among whom risky practices such as sharing syringes, other injecting equipment, or drug preparations may transmit infection.¹ HCV-associated morbidity among IDUs may manifest as hepatic failure and cirrhosis from chronic infection; this process may be accelerated and amplified by co-infection with HIV or hepatitis B virus (HBV).¹ Acute HCV is symptomatic in 10–15% of cases; those who are asymptomatic in the acute phase of infection are more likely (85 – 95% vs. 48 – 75%) to progress to chronic infection.²

Injecting drug use is a primary mode of transmission for HCV in developed countries. In developing countries, HCV transmission more commonly results from transfusion and unsafe medical practices.^1,2 However, HCV among IDUs in developing countries has been acknowledged as a burgeoning health issue, particularly in settings where HCV transmission is considered a harbinger of HIV epidemics among this population.^3–5 High HCV prevalence has been recorded among IDUs in many Asian settings, including Pakistan, India, and Iran, several of which are also high HCV burden (general population prevalence>2.9%) countries.^1,6–9 HCV prevalence among IDUs in Afghanistan ranges from 36.6% to 49.1% in two urban settings and may increase due to risk through both injection-related and other potential exposures.^10,11 Increasing numbers of IDUs in Afghanistan and the common practice of sharing of needles, syringes, injecting equipment and aspirating and re-injecting one’s blood (khoon bozee, literally “playing with blood”) may also contribute to increasing prevalence of HCV among Afghan IDUs.^10,12

Longitudinal studies indicate that some IDUs previously infected with HCV are able to clear subsequent infections; this association persisted after adjustment for risky behaviors.^13,14 However, the risky behaviors that increase likelihood of re-infection also increase the likelihood of HCV transmission to other IDUs due to presence of viremia in re-infected individuals. IDUs with measurable viremia may differ significantly with respect to risky behaviors than those with either evidence of past infection or no evidence of prior infection and are of special interest due to their ability to transmit infection to others. IDUs with viremia may also differ by biologic factors; however, it is important to assess the relative association of behavioral factors as these may be addressed through prevention programming. Our previous cross-sectional study in Kabul was conducted prior to availability of nucleic acid testing in Afghanistan, so no information could be obtained regarding prevalence and traits associated with viremia. The data presented in this manuscript result from the follow-on assessment conducted between 2006 and 2008 in three other Afghan urban centers.

Recent data from a longitudinal study among IDUs in Kabul reveal high HCV incidence in a setting with many new injectors.¹⁵ Assessment of whether IDUs with detectable viremia are distinct from those with antibody alone may provide important information to prevention programs in other Afghan cities to avert the HCV epidemic currently underway in Kabul. The purpose of this analysis was to determine whether the sub-group of IDUs with HCV viremia differs significantly from IDUs with either no evidence of prior infection or HCV antibody (HCV Ab) and to determine whether behaviors associated with HCV infection are more likely among IDU with HCV viremia in three Afghan cities.

Methods:

Setting

This study was conducted among injecting drug users (IDUs) in Hirat, Mazar-i-Sharif, and Jalalabad, the largest cities in their respective regions of Afghanistan. At the time of this study, private and public detoxification programs were operating in all cities; there was one harm reduction program operating in Hirat city with on-site needle exchange. The other two cities did not have functioning needle and syringe programs (NSPs) and no city had programming offering either opioid substitution therapy or hepatitis B vaccine to IDUs at the time of the study.

Study Design and Participants

This cross-sectional study enrolled participants between September, 2006 and January, 2008 through the Ministry of Public Health-affiliated Voluntary Counseling and Testing centers (VCT), public and private harm reduction outreach programs (a few of which had VCT services on offer within their organizations), and the International Rescue Committee (IRC) offices in each location. Eligible participants were those reporting injecting drugs (confirmed through physical evidence of recent injection) within the past six months, aged 18 years or greater, and able to provide written informed consent. Participants answered a series of questions to ensure comprehension of the study risks and benefits and requirements of participation; participants unable to sign their name provided a fingerprint. Approval was obtained from the institutional review boards of the University of California, San Diego, the Walter Reed Army Institute of Research, the U.S. Naval Medical Research Unit 3 in Cairo, Egypt, and the Ministry of Public Health of the Islamic Republic of Afghanistan.

Measurement of variables and outcomes of interest

The outcomes of interest for this analysis are determination of whether IDUs with HCV viremia differ significantly from those with prior HCV infection and whether the association between HCV infection and various risky behaviors strengthens in a continuum from seronegative to viremic status. The questionnaire instrument assessed sociodemographic, travel, incarceration and medical histories, and drug use and sexual behaviors. Drug use behaviors of interest included sharing needles/syringes ever and in the last six months, sharing injecting “works” (e.g. cookers, cotton) ever and in the last six months, duration of injecting, injecting while incarcerated, and aspirating and re-injecting blood. Iatrogenic routes of blood-borne infection transmission were also assessed, such as receipt of therapeutic injections from both medical and non-medical providers and both provision and receipt of transfused blood.

Procedures

Potential participants were recruited by experienced outreach workers affiliated with local harm reduction programs in each city. Trained study staff obtained informed consent, administered the study questionnaire, and performed rapid testing and counseling in a private setting. Laboratory methods for other pathogens have been reported in a related publication.¹⁶ Briefly, whole blood rapid testing for HCV Ab was performed with Standard Diagnostics HCV Ab (Standard Diagnostics, Kyonggi-do, Korea). The manufacturer reports 100% sensitivity and 94.1% specificity by WHO evaluation.¹⁷ Serum specimens were obtained from participants with positive rapid tests; HCV Ab was confirmed with RIBA (RIBA 3.0 SIA®, Chiron Corporation, Emeryville, CA, USA). Presence of hepatitis C virus was assessed with qualitative reverse transcriptase polymerase chain reaction (RT-PCR). Extraction was performed using Qi-Amp viral RNA mini kit (Qiagen Incorporated, Germantown, Maryland, US) according to manufacturer’s instructions. Nested RT-PCR was performed as previously described with minor modifications of cycle parameters and master mix components.¹⁸ Both RIBA and PCR were performed at the Afghan Public Health Institute laboratory in Kabul, with the exception of HBV PCR performed at NAMRU-3 laboratory in Cairo, Egypt.

Confirmatory results were available after two weeks; participants were provided with follow-up appointments at time of enrollment. All participants received post-test counseling, risk reduction counseling, a small non-monetary gift of hygiene items (e.g. razor, soap) of U.S.$4 value, condoms, and sterile syringes, with referrals for detoxification programs upon request. No data were recorded on those declining or ineligible for study entry. Enrollment was conducted for twelve months in all sites.

Statistical Analysis

Descriptive statistics for viremic individuals were generated, with site differences assessed with Chi-square test. Risky behaviors, such as sharing syringes and injecting works, were reported by frequency; these variables were dichotomized into ever/never in the last six months.

Univariable logistic regression was performed to identify potential associations between HCV viremia and select demographic, biological, and individual risk behavior variables, controlled by enrollment site, of all confirmed HCV Ab positive participants. Variables were entered into a site and age-adjusted multivariable model if they were associated with HCV viremia at the 10% level in univariable analysis or were considered of epidemiologic significance.

Ordinal logistic regression was performed to identify associations between various HCV infection states and select demographic and risk behavior variables, adjusting for site and age. The reference group was defined as participants with no evidence of prior HCV infection (antibody-negative by rapid test or by RIBA) assigned the value of 0, with confirmed HCV Ab-positive participants assigned the value of one, and those with HCV Ab and detectable viremia assigned the value of two. A multivariable model was generated to identify factors independently associated with HCV infection states using the Wald test to determine which variables were retained at p<0.05. The proportional odds assumption of ordinal regression was tested using the Brant test.

Results:

Of 623 IDUs recruited in the three Afghan cities (Hirat=340, Jalalabad=96, and Mazar-i-Sharif=187), eight female IDUs were excluded as the low number of female participants precluded accurate description and adjustment by sex. As stated in the Methods, only participants with confirmed HCV Ab by RIBA were included in subsequent analysis. Of 631 total participants, 630 had HCV rapid testing performed. Of 229 participants with reactive HCV rapid tests with serum available for RIBA, 223 were RIBA-reactive, two were RIBA-indeterminate, and four were negative (positive predictive value=97.4%). A further two male IDUs with reactive RIBA tests did not have sufficient serum available for PCR testing and were excluded from this analysis.

The overall prevalence of RIBA-confirmed HCV Ab was 36.6% (n=223/609), with significant variance by site (p<0.001). Of 221 participants with HCV Ab and sufficient available specimen for PCR testing, 127 (57.5%) had detectable viremia. There was no significant difference in prevalence of viremia among IDUs with positive HCV Ab by site, with 41.7% in Jalalabad, 59.1% in Mazar-i-Sharif, and 59.6% in Hirat. However, there were significant site differences among viremic individuals by age, sharing syringes/needles or injecting paraphernalia, having lived outside the country in the last 10 years, and having previous addiction treatment (Table 1). Further, age alone was positively associated with viremia (OR=1.04, 95% CI: 1.00–1.08). Due to these differences by site and age, all further analyses were controlled for these factors. All detected HIV cases were positive for HCV Ab and were enrolled in Hirat. Reported therapeutic injections in the last six months (13.4%, n=17), therapeutic injections by a non-medical provider (3.1%, n=4) or either blood donation (7.1%, n=9) or transfusion at any time (4.7%, n=6) were rare among viremic IDUs.

Table 1.

Demographic and Behavioral Characteristics of Male Injecting Drug Users with HCV Viremia in Three Afghan Cities. (N=127)

Feature	Hirat	Jalalabad	Mazar-i-Sharif	p-value
HCV viremia; % of HCV Ab+ IDU that were viremic/site; N=221	59.6% (96)	41.7% (5)	59.1% (26)	0.520
	Mean (SD)	Mean (SD)	Mean (SD)
Age	30.2 (7.3)	27.2 (6.3)	31.5 (9.1)	0.006
Age Initiated Injecting	27.3 (7.0)	24.2 (4.3)	26.2 (7.9)	0.736
	% (N)	% (N)	% (N)
Civil Status: Ever Married	46.9% (45)	33.3% (4)	42.4% (11)	0.299
Education: ≤6 years	35.4% (34)	60.0% (3)	30.8% (8)	0.457
Lived/Worked outside Afghanistan in last decade	96.8% (92)	60.0% (3)	76.9% (20)	<0.001
Ever in Jail: Yes	76.8% (73)	40.0% (2)	56.0% (14)	0.177
Injected in Jail*: Yes	16.9% (12)	0 (0)	25.0% (4)	0.767
Shared Syringes in Last 6 Months: Yes	26.1% (24)	80.0% (4)	19.2% (5)	0.018
Shared Injecting Works Last 6 Months: Yes	38.5% (37)	0 (0)	19.2% (5)	0.049
Aspirating & re-injecting blood: Yes	78.1% (75)	80.0% (4)	92.3% (24)	0.261
Prior Addiction Treatment: Yes	76.0% (73)	40.0% (2)	26.9% (7)	<0.001

^*Denominator is number ever having been incarcerated/ site.

Injecting behaviors and other risk markers were compared between IDUs with HCV viremia and those with HCV Ab only (Table 2). HIV co-infection was positively independently associated with HCV viremia, but HBV co-infection or sharing syringes or needles in the last six months were negatively associated in age and site-controlled multivariable regression.

Table 2.

Factors independently associated with HCV viremia, among HCV antibody positive injection drug users in three Afghan cities, controlling for site and age (N=221).

Variable	AOR, (95% CI)
HIV co-infection	1.47, (1.29 – 1.66)
HBV co-infection	0.48, (0.43 – 0.51)
Share syringes in last 6 months	0.40, (0.29 – 0.55)

AOR= Adjusted Odds Ratio

CI= Confidence Interval

N= Number

In age and site-controlled ordinal logistic regression analysis, the spectrum (range: 0–2) of HCV disease from antibody to viremia was independently associated with HIV co-infection, prior incarceration, sharing injecting equipment in the last six months, khoon bozee, and prior addiction treatment were independently associated with HCV infection, while a negative association between HCV infection and ever receiving therapeutic injection from a non-medical provider persisted (Table 3).

Table 3.

Variables independently associated with HCV viremia among male injecting drug users in three cities of Afghanistan in ordinal logistic regression controlled for age and site (N=609).

Variable	AOR, (95% CI)
HIV Co-infection	7.16, (4.41–11.64)
Prior Addiction Treatment	1.95, (1.57–2.42)
Ever Aspirate & Re-inject Blood	1.62, (1.18–2.23)
Ever Been Incarcerated	1.60, (1.04–2.45)
Shared Injecting Equipment Last 6 Months	1.35, (1.02–1.80)
Ever Received Injection from Non-Medical Provider	0.30, (0.10–0.89)

AOR= Adjusted Odds Ratio

CI= Confidence Interval

N= Number

Discussion:

Circulating virus was present in at least 40% of participants with HCV Ab and did not vary significantly by site, ensuring substantial risk of HCV transmission within injecting networks inclusive of these IDUs. We acknowledge that this figure likely underestimates viremia as sustained viremia without detectable antibody levels has been noted among IDUs in longitudinal studies in the Netherlands and the United States.^19–21 Though not statistically significant, IDUs in Jalalabad were less likely to have detectable viremia. One possible reason for this observation may be the social organization of injectors in Jalalabad, where many participants lived outside the city center, initiated injecting in small networks affiliated with the opium processing industry, and only came to the city for addiction treatment or trade reasons (personal communication, Dr. M. Nasiry, assistant study manager, Jalalabad).

In comparisons between viremic IDUs and IDUs with HCV Ab alone, odds of viremia increased with age. Viremia was less likely for older IDUs on opioid replacement therapy in Australia, suggesting that the observed association in Afghanistan may be behavioral rather than physiologic.²² One hypothesis for this observation may be that injecting networks tend to consist of individuals of similar age, seen as peers. HIV co-infection was positively associated with HCV viremia, similar to findings among IDUs in several urban centers in the United States.^23,24 Grebely et al. found that IDUs with HIV co-infection have higher rates of HCV clearance in a longitudinal study; however, incident HCV re-infection was twice as high among HIV-infected individuals.¹⁴ The clearance rate among those with HCV re-infection was 29% (4/14); none of the 6 HIV-infected participants re-infected with HCV cleared their viremia.¹⁴ Sharing needles/syringes in the last 6 months was negatively associated with viremia, which is surprising as this group would likely be at greater risk for recent infection or re-infection, as noted among the Canadian cohort.¹⁴ The negative association between detectable viremia and HBV viremia may be explained by observed higher HCV clearance rates in the presence of chronic HBV infection, similar to findings among hemophiliac and female patient populations.^25,26

Several behaviors or conditions were progressively more associated with presence of both viremia and HCV Ab. The strongest association was between HIV co-infection and HCV, likely reflective of both infections resulting from some of the same risky behaviors, as noted in China, Iran, and India.^6,27,28 Prior incarceration was associated with increasing odds of HCV, potentially reflecting sharing of injecting equipment and mixing of injecting networks in prison settings, previously detected in a variety of settings.^6,29,30 Similarly, previous addiction treatment may reflect mixing of injecting networks during or following the course of treatment, as participants in this study reflect active IDUs who have relapsed.

Sharing injecting “works” (e.g. cookers, cotton) in the last 6 months was associated with HCV spectrum but not with HCV viremia among those with HCV infection. This association was unexpected, as recent risk behaviors would seem more likely to be linked to viremia present in recent infection/re-infection.^31,32 We did not assess whether ever sharing works was associated with HCV viremia, as recent behaviors were deemed more relevant. Sharing of needles and syringes and potentially injecting works may be under-reported or may have ceased due to these messages. Participants enrolled in Hirat, most of whom had been refugees in Iran and were exposed to harm reduction messaging in that setting, had a higher prevalence of HCV Ab. Though analyses were controlled for site, under-reporting of sharing, a stigmatized behavior, in Hirat may influence results. Aspirating and re-injecting blood was independently associated with chronic HCV infection and may serve as a proxy for risky sharing behaviors, as hypothesized for IDUs in Pakistan.⁹

Receipt of a therapeutic injection from a non-medical provider was positively associated with presence of HCV antibody among IDUs in Kabul, which was the rationale for including this variable in our analysis.¹⁰ This variable was negatively associated with HCV in our ordinal analysis and, though this relationship was statistically significant, there were relatively few individuals in any group reporting prior injections from non-medical providers. We believe these low numbers indicate a spurious relationship. Of note, reported receipt of a therapeutic injection ever or in the last six months was similarly rare. Other potential markers of iatrogenic transmission will be considered in future studies.

Due to cost limitations, we did not obtain serum samples or perform PCR for participants with negative HCV Ab rapid tests. Therefore, we are unable to comment on circulating virus among participants with HCV infection without detectable antibody, such as those with acute (<12 weeks) infection, immune compromise, or delayed/absent humoral response.^2,19 Further, the number of participants with viremia may have been further underestimated based on the threshold for detecting virus with the qualitative PCR assay used. This assay may not have been sensitive enough to detect those participants who were chronically infected but had low circulating viral levels.¹⁹ The rapid tests used cite a sensitivity of 100%, indicating that true positives were unlikely to have been missed in the initial screening.¹⁷ However, though the manufacturer’s fact sheet states the evaluation was performed by the World Health Organization (WHO), no WHO source could be identified to confirm the findings. The performance characteristics of the rapid test may have also been affected by setting. Additionally, risky injecting practices were based on self report and may have been under-reported. We did not map injecting networks and so cannot comment on risk behaviors within specific networks as related to HCV prevalence. Last, we did not assess HCV genotype, a factor known to affect both viral load and probability of natural clearance.³³ Future genotype assessment is planned and will provide important insight in treatment recommendations.

High prevalence of HCV viremia has been detected among IDUs previously exposed to HCV in three cities in Afghanistan. This suggests ample opportunity for HCV transmission among IDUs through shared injecting equipment and potentially to the population at large through shared medical equipment. IDUs in Afghanistan are vulnerable to this infection and efforts must be focused on scaling up harm reduction programs to limit further HCV transmission in this population. Currently, harm reduction programs are being established or expanded in Hirat, Jalalabad and Mazar-i-Sharif, indicating the feasibility of harm reduction programs in Afghanistan, despite conservative culture, political instability, and armed conflict. Encouragingly, the Kabul harm reduction programs have established a harm reduction kit that is provided to IDUs in the field and includes sterile saline water bottles, cotton gauze, and alcohol prep pads in addition to unused syringes. We recommend that, as voluntary HIV counseling and testing is scaled up with harm reduction efforts, counseling and testing for HCV and sterile injecting paraphernalia also be provided to this vulnerable group to prevent further expansion of this prevalent infection and associated morbidity and mortality.