While several published Moroccan studies exist in the scientific literature on hemodialysis, haemophiliacs and chronic HCV-infected patients (Sekkak et al.; Benouda et al.; Benani et al.; Benjelloun et al.) [10, 24–26], the IDU group in the Maghreb region has remained a largely under-studied cohort to date. Due to the paucity of published data, this study sought to characterize this elusive and high-risk group, as well as to elucidate key difference between it and other chronically HCV-infected patients in Morocco.
The average age of IDUs (n = 154) was found to be 38.5 ± 8.7 years and the median 40 years. The youngest and oldest subjects were 19 and 58 years old respectively, thereby our sample had an age range of 39 years. Only a small fraction of subjects were reportedly in a conjugal relationship (married, 14.5%), while the overwhelming majority (85.5%) were either single, divorced, widowed or separated. A large part of the subjects were male, with a male/female sex ratio of 7.78 (n = 167). This gender asymmetry may be explained in two ways: either that the majority of the injecting drug users in Morocco are indeed male and this is a predominantly male-associated drug habit, or that females hesitated to report their addiction because of social and cultural reasons.
In contrast, the chronically infected patients used as a comparator population had an average age of 59.8 ± 13.8 years (n = 151), with a median of 60 years old, indicating that his cohort was somewhat older than the IDUs. The age range was slightly larger as well (18 to 65 years, range of 47 years), as we speculate this is due the fact that there were older subjects present in the chronically infected group. Another demographic difference observed was the male to female sex ratio (51/100 = 0.51), as the majority of the chronic carriers were women. This skewedness did not bias the mean age of all patients, as women (n = 100) had an average age of 59.1 ± 13.4 years, and men (n = 51) an average of 60.12 ± 14.7 years respectively.
Slightly more than half of our IDU cohort was found to be anti-HCV positive (89/148, 60%), allowing us to classify them into and confirm the ‘highly seroprevalent’ stigma associated with IDUs worldwide. Previously reported studies in various parts of the world disclosed lower HCV-positive proportions in their local IDUs: Lebanon (50%) and Australia (45.8%) (Mahfoud et al.; Kynotch Aitken et al.) [1, 15], while other studies reported higher HCV-positive proportions: San Francisco Bay Area, China, and Greece at 80.2%, 66.7% and 73% respectively (Gigi; Ucellini et al.; Bao & Liu) [27–29].
Age and seropositive status were shown to be positively correlated (Baha et al.; Benouda et al.; Mathei et al.) [9, 10, 30], in that older subjects (as a group) were more likely to be seropositive as compared to younger subjects (n =132). Due to our sufficient sub-sample size and wide range of ages, we were able to repeat the analysis by binning our IDU cohort into 7 age groups and calculating the average seropositivity for each group. Indeed, we report a moderately positive correlation between age and average seropositivity, confirming that the younger subjects were less likely to be anti-HCV positive (R =0.726) (Figure 1). As previously mentioned, Morocco implemented a national anti-HCV blood-screening practice in 1994 (Baha et al.) , and this is thought to have substantially contributed to the reduction of new HCV transmissions among younger generations. In addition, we also believe that the younger IDUs lower seropositivity can be explained by reduced duration of exposure to the risk associated with injecting drug use.
Already identified and well-studied risk factors associated with HCV transmission are incarceration, tattooing, needle-sharing, dental therapy, and a history of surgery (Baha et al.; Rice et al.) [9, 31]. In terms of route of drug administration, only a small proportion (8/101 = 0.08%) of IDUs reported smoking as their habitual route of drug administration, as opposed to the more popular intravenous route. Five out of the eight users who smoked as their primary method for taking the drug were found to be anti-HCV positive, however, we note that the majority of the IDUs who smoked heroine were not subsequently screened for HCV-positive status. Further studies need to be done in with a sufficiently large sample size in order to establish route of drug administration as a bone fida risk factor, one which has implications for risk of infection.
The other major risk factor our study sought to address was the sharing of needles, with the majority of subjects reportedly doing so (86/101, 85.15%). This habit reflects the cultural aspect of the Moroccan and the larger Arab society, in the sense that these collectivist cultures are known for their sharing of food and goods as a symbol of bonding, brotherhood, and hospitality. One might also speculate that needle-sharing may be due to financial reasons during hard economic times, as the partage of one syringe between multiple users is less costly. The transmission of HCV is not only associated with contaminated syringes, but also with the drug preparation equipment, such cookers and filters, and the water in which the heroine is dissolved (Doerrbecker et al.) . Not surprisingly, the act of needle-sharing was associated with a higher HCV-positive status among our cohort, and the χ2 test yielded significant results (χ2 = 43.22, p = 4.87e-11), leading us to conclude that the drug users under study who actively shared their needles were more likely to be HCV-positive as compared to those that did not.
We next sought to characterize possible serological and genotypic differences between IDUs and other chronically-infected patients. The average viral load (IU/mL) was not significantly different between IDUs and chronically-infected patients (t = 0.159, df = 109.46, p = 0.4369). Viral load is a factor routinely taken into account by practitioners when designing custom therapies for their patients. Pre-treatment viral load level has implications for treatment and is inversely related to treatment response, such that high viral load is often associated with a poor response to therapy (Amjad et al.) .
Perhaps the most striking difference between our IDU cohort and the chronically infected patients was their genotype profile (χ2 = 151.1, df = 5, p = 2.2e-16). The majority of IDUs harboured the genotypes 1a (60.8%), 3a (25.6%), while other chronically HCV mono-infected patients at the Pasteur Institute of Morocco in Casablanca primarily exhibited genotypes 1b and 2a/2c respectively (Table 1). A similar IDU profile has been documented in anti-HCV-positive subjects in other countries, some of which are in direct contact with Morocco and hugging the Mediterranean basin, such as Spain, Turkey, Croatia, and France (Payan et al.; Yildiz et al.; Vince et al.; Lopez-Labrador et al.) [17, 21, 34, 35]. The Moroccan IDU profile does not match with neighbouring country Algeria or central Africa, which exhibit 1b, 2a/2c, and 4 as the predominant genotypes respectively (Rouabhia et al.; Xu et al.) [36, 37], however we note that these cited studies were carried out in chronically infected and hemodialysis patients. One can speculate as to the multiple channels of heroin trafficking taking place across the Moroccan borders and the interactions and relationships that contribute to the distinct Moroccan IDU profile. Morocco shares a border with two neighbouring countries: Algeria on its Western front and Mauritania to the South. In the North, one simply crosses the Gibraltar strait to enter Spain and the European Union. Bordering on the East is the Atlantic Ocean, however direct access to the Spanish Canary Islands is possible via direct flights to Gran Canaria and Tenerife, two very popular touristic destinations. While the scientific literature lacks data on Mauritania, Algeria has an anti-HCV positive prevalence estimated at 0.02% (Lavancy; Rouabhia et al.) [36, 38]. Interestingly, one Spanish study on both IDUs and chronically infected hemodialysis patients revealed that the majority of injecting drug users were genotype 1a (48.5%) (Garcia et al.) . Finally, two drug user cohorts studied in Lisbon predominantly exhibited the genotypes 1a and 3a, resulting in this country harbouring a profile most similar to the local Moroccan drug users (Almeida Calando et al.) . Collectively, these findings are clues and provide insight into the possible origin of local circulating genotypes among the drug users of Morocco. Available published data on HCV genotypes from IDUs living in countries close in geographic proximity to Morocco are displayed in Table 2.
Genotyping was performed on 59 out of the 154 IDUs who had age demographic data available, and patients were grouped into 7 age intervals (Table 3). At first glance, it seems that genotype 1a is found in all age groups, while genotypes 3a and 1b were present mostly in younger or older IDUs respectively. These differing proportions for genotypes 3a and 1b could be explained by a gradual shift in genotype profile over time and with older age groups having been exposed to different ‘sources’ of HCV as compared to younger generations. Genotype 1b has been found in the general population and among high-risk groups such as hemodialysis and blood transfusion patients (Benani et al.) , effectively classifying this genotype as ‘transfusion-related’. Since implementation of the national blood screening program in 1994 one would expect that younger subjects would have had a reduced risk of contracting this genotype, effectively explaining its absence in patients aged less than 30 years old. The low presence of genotype 4 across all age groups mirrors the lack of genotype 4 in the general Moroccan population. Authors acknowledge the insufficient number of subjects in some age intervals, and further studies with more patients need to be done in order to accurately assess the uniformity of genotypes in heroin addicts of all ages.
We next sought to determine if our IDU cohort had a higher viremia compared between the different genotypes. ANOVA analysis of genotype and average viremia (IU/mL)) was significant (F = 2.98, p = 0.0412, n = 49), indicating that individuals infected with genotype 1a had a higher viral load titer as compared to genotypes 1b, 3a, and 4a respectively (Figure 2). Our results conflict with a previous study that did not observe a higher viremia associated with genotype 1a across all age groups (Schijman et al.) .
Further complicating the picture is the co-infection with Hepatitis B virus (HBV) and Human Immunodeficiency virus (HIV-1), which ultimately alters the natural course of the disease. It is now well-established that a substantially higher viral load is present in HIV-1 co-infected individuals as compared to HCV mono-infected patients (Matthews-Greer et al.; Cribier et al.) [41, 42]. HBV/HCV dual-infection results in cross-suppression of viral replication, as well as rapid progression of liver disease and cirrhosis, while HIV-1/HBV co-infected individuals often exhibit immune suppression and higher viral RNA loads (Sulkowsky; Filippini) [43, 44], mirroring the situation in HIV-1/HCV co-infected patients. A possible next step is to address the proportion of HCV-positive Moroccan IDUs co-infected with either HIV-1 or HBV.
It is widely recognized in the medical community that different genotypes respond varyingly to antiviral therapy and that no standard of care (SOC) treatment exists for all 6 genotypes (Petta & Craxi; Grassi & Aghemo; Esmat et al.) [45–47], with the ‘gold standard’ treatment currently being pegylated interferon-α (pegINFα) in combination with ribavirin (RBV) for 24 to 48 weeks. The addition of protease inhibitors Boceprevir and Teleprevir implemented in triple therapy significantly increases response rates and sustained virology response (SVR), however this comes at the price of additional side effects and the risk of therapy discontinuation (Chopra et al.) . Studies are at odds in terms of the efficacy of antiviral treatment based on genotype: genotype 1a has been shown to be negatively correlated to SVR (Petta & Craxi) , while another study reported a better response rate to dual antiviral therapy in mono-infected patients using Boceprevir and Telaprevir (Pelicelli et al.) . Generally speaking, genotype 1b associated with a more aggressive disease course, and patients are more likely to exhibit liver cirrhosis and decompensated liver disease (Zein) . Better response rates are obtained in genotype 2 and genotype 3 patients (about 80%), while genotype 1-infected patients remain the hardest to treat overall, reaching an SVR of only about 40% (Jazwinski & Muir) . With respect to genotype 3 infected individuals, the above mentioned SOC treatment was tailored to 24 weeks and was shown to be effective in terms of SVR, and further shortened to 12-14 weeks for those patients exhibiting rapid virologic response (RVR) (Andriulli et al.; Zeuzem et al.) [52, 53]. In light of these findings, the treatment implications for the Moroccan IDU cohort are somewhat positive, seeing that the majority are infected with either genotypes 1a or 3a. However, the medical community must first tackle the issue of treatment refusal before effective therapy can be implemented, as IDUs have been known to decline medical help (Mehta et al.) . For it to be successfully resolved, the issue must be approached on multiple fronts: reaching out to drug addicts as they are often marginalized from mainstream society; ensuring that IDUs are aware of various treatment options at their disposal; undertaking a cost-benefit analysis for each patient in order to choose the best course of action and a tailored therapy.