Noninjection drug use: an under‐appreciated risk factor for hepatitis C virus transmission

Hepatitis C virus (HCV) infection is a leading cause of chronic liver disease with an estimated global prevalence of >120 million individuals (1). Of those who are exposed to the virus, an estimated 50–80% will develop chronic infection that can ultimately lead to hepatic fibrosis, hepatocellular carcinoma, and cirrhosis. As the virus is most effectively transmitted via blood, the majority of prevalent infections can be attributable to injection drug use or blood transfusions administered before 1990. Among incident HCV infections in the United States in 2005, the most frequently reported risk factors were injection drug use (50%), multiple sex partners (23%), surgery (14%) and percutaneous injury (10%) (2). Interestingly, no identifiable risk factors were reported by 26% of the participants. In at least some of the HCV-infected individuals without identifiable risk factors, noninjection drug use might be the mechanism of viral acquisition. As 8% of the US population engages in illicit substance use monthly (3), HCV transmission via noninjection drug use may be a common, yet under-appreciated, public health problem. Although most incident HCV infections occur among injection drug users, noninjection drug use is increasingly recognized as an emerging risk factor. Several recent studies have reported that the prevalence of HCV among noninjection drug users (NIDUs), estimated to range from 2 to 35%, is greater than that observed in the general population (4). The wide variability in the prevalence estimates among NIDUs results from differences in the primary aims and in the quality of the data obtained in individual studies. In the majority of these studies, HCV risk factor assessment among NIDUs was a secondary aim as the investigation of HCV transmission among injection drug users (IDUs) was their primary objective. Additional weaknesses of these studies include limited data regarding the likely time of HCV exposure and possible false-negative HCV antibody measurements due to waning serological markers, as HCV antibodies have been shown to disappear in a significant percentage of individuals 20 years after spontaneous resolution of the infection. Additionally, several studies did not use the recombinant immunoblot assay or HCV RNA quantitation to confirm positive serologic results. Misclassification and recall bias is another potential limitation of studies of both noninjection as well as injection drug use, as these studies largely rely on risk factor assessment based upon patient self-report. Misclassification might occur if prior injectors are erroneously classified as NIDUs. Recall bias, the inability to remember episodes of high-risk behavior, may have occurred from the effect of mind-altering substances or patient reluctance to report specific instances. Consequently, the specific behaviors among NIDUs that result in increased HCV seroprevalence and their relative importance toward the establishment of HCV infection remain unclear. How might NIDU behaviors result in HCV transmission? Most likely, two factors are required, exposure to HCV RNA in biological fluids and mucous membrane disruptions permissive for blood–virus interaction. HCV RNA has been detected in saliva from chimpanzees (5) and in up to 52% of humans with chronic HCV infection (6, 7). Injection of saliva derived from an HCV-infected chimpanzee into a second chimpanzee resulted in HCV infection (5). Similar observations have been noted in humans where a skin bite from an HCV-infected individual resulted in productive HCV infection (8). Other body fluids besides saliva also harbor HCV and may be vehicles of viral transmission. For example, HCV RNA has been detected in 59% of gingival crevicular fluid (GCF) specimens, and the detection of HCV RNA in GCF required a minimum plasma level of 100 000 copies/ml (7). In this study, only patients with measurable HCV RNA in GCF had detectable HCV RNA in saliva suggesting that GCF may be a potential viral reservoir. Additionally, secretions from other body compartments may also transit HCV. As the virus has also been detected in nasal sections of an NIDU (9), intranasal cocaine use could lead to HCV infection. Thus, sharing of inhalational equipment contaminated by either intranasal or oral fluids could permit viral transfer among individuals. The other factor required for the development of HCV infection is direct interaction between the virus and the host's blood. Among NIDUs, chronic substance abuse can lead to mucous membrane disruptions. For example, through its vasoconstricting effects as well as irritation from substances with which the drug is diluted (i.e. talc), cocaine can be locally irritating to the thin respiratory epithelium of the nasal cavity resulting in nasal septum perforations of both the cartilaginous and bony tissues. In the oral cavity, persistent cocaine use can result in ischaemic mucosal ulceration, rapid gingival recession, and dental erosions (10). In addition, individuals with chronic substance abuse display high rates of dental abnormalities, including decay and periodontal disease, which could further facilitate interaction between the virus and the host. Consequently, blood or saliva on inhalation equipment in combination with cocaine-induced lesions of oral or nasal mucous membranes is a mechanism by which noninjection drug use could result in productive HCV infection. In this issue of Liver International, Macias et al. (11) assessed risk factors for HCV acquisition in 182 NIDUs. Study participants were recruited from a drug abuse treatment center, and detailed information on previous drug addiction behaviors was available. Subjects underwent a structured interview that included detailed questions on potential routes of exposure, substance abuse patterns, and behaviors that increase the likelihood of HCV transmission. A study physician subsequently examined all subjects for signs of skin perforation such as evidence of recent venipuncture, tattoos and body piercings. Serological testing for human immunodeficiency virus, hepatitis B virus and HCV as well as measurement of HCV RNA levels was subsequently performed on each individual. The authors report an HCV prevalence of 12.6%, and they observed that age ≥34 years, the presence of tattoos, and sharing of crack cocaine inhalation equipment were independently associated with HCV infection in NIDUs. These results, which are corroborated by those of prior studies (12–17) (Table 1), suggest that HCV transmission can occur via shared crack cocaine inhalation equipment. The authors hypothesize that blood (from oral ulcerations) or saliva contamination of inhalational equipment could transmit quantities of virus sufficient to surpass the critical threshold, estimated to be 20 viral copies/ml in chimpanzees (18), necessary for productive infection. Macias et al. also found that tattoos were significantly associated with HCV transmission. Tattooing, especially when performed by a friend or relative, has been reported to be an independent risk factor for HCV transmission among high-risk individuals (12–14, 16, 17) (Table 1). Unprofessional tattooing, especially among incarcerated individuals, can be performed using a paperclip, staple or other sharp objects to break the surface of the skin, which is then pigmented with ink commonly from a ballpoint pen. HCV transmission might occur as a consequence of the same object being used on more than one individual and the fact that these implements are infrequently sterilized between insertions. Besides noninjection drug use and tattooing, other risk factors for HCV include incarceration (13, 19), altercations that result in traumatic injury (20), and sharing the same electric shears among many prisoners (20). How does this study rate in comparison with other studies that have assessed risk factors for HCV acquisition among NIDUs? A recent meta-analysis of 28 studies of NIDUs reported mean and median scores of 7.11 and 7.00, respectively, on an objective quality measurement scale (4). Using these criteria, we calculated a score of 10 for the Macias et al. study. Strengths of the present study are the inclusion of a relatively large sample, the fact that physicians experienced in drug addiction management conducted patient interviews and physical examinations, the fact that a comprehensive ascertainment of prior noninjection drug use behaviors was performed, and the analysis of the data using multivariate techniques. Of note is the fact that study participants did not receive payment for their involvement. The exclusion of IDUs from this study is an additional advantage. Recall bias, especially because patients may have been reluctant to report or did not recall specific high-risk behaviors, is a potential limitation of the study as it is with all studies that rely on patient self-report. This important study by Macias et al. identifies potential modes of HCV transmission in NIDUs, and it underscores the importance of noninjection drug use as a potential mechanism of viral acquisition. The exclusion of IDUs may permit detection of the weaker association between NIDU behaviors and HCV. Several factors concerning the mechanism of viral transmission, such as the relative contribution of blood versus salivary contamination and the minimum viral quantity required for human oral transmission, remain to be investigated. With an enhanced understanding of the factors that promote viral transmission in NIDUs, effective infection control interventions may be implemented.