Article Text

Download PDFPDF

What’s new for injection drug users with HIV infection?
  1. S M Clarke1,
  2. F M Mulcahy2
  1. 1Department of Infectious Diseases, Boston Medical Center, 1 Boston Medical Center Place, Dowling 3N, Room 3211, Boston, MA 02118, USA
  2. 2GUIDE Clinic, St James’s Hospital, James’s Street, Dublin 8, Ireland
  1. Correspondence to:
 Susan Clarke;

Statistics from

Difficulties of providing long term medications to this unique group of patients

Injection drug use represents the risk factor for human immunodeficiency virus (HIV) infection for the minority of worldwide HIV disease.1 In contrast, injection drug use in developed countries accounts for between 30% and 60% of the HIV infected population.1 Much of the published HIV literature focuses on large randomised clinical trials of highly active antiretroviral therapy (HAART) regimens, potential adverse reactions from HAART, and the basic science of HIV infection. There is generally a paucity of research focusing exclusively on injecting drug users (IDUs). A literature review was undertaken to review published data from 1999 to 2001, focusing on articles that directly relate to IDUs and HIV infection. Most of the derived data relate to the following areas:

  1. Increasing epidemics of HIV infection in IDUs in developing countries

  2. Improving access to care for IDUs

  3. Co-infection with hepatitis C virus

  4. Pharmacokinetic interactions between antiretrovirals and methadone.


Until recently, the concept of IDU was essentially a Western world phenomenon, concentrated in regions of the United States and Western Europe where HIV infection is mainly subtype B disease. In developing countries and countries in political transition where HIV infection was previously vastly under-reported, and IDU was not a significant problem, it is only now that real epidemics of HIV infection among IDUs are occurring and being accurately reported.1 Molecular methods and phylogenetic analysis are helping to trace and follow the epidemics.

In many parts of Asia, the HIV epidemic has spread extensively since the 1980s, with multiple, genetically divergent subtypes, but with a predominance of subtype A.2 In China, a similar epidemic has been reported with subtype C predominating. Lai et al describe data from Guang Xi province in China, where there has been a threefold increase in the incidence of HIV infection in IDUs between 1998 and 1999.3 The authors directly relate this to the rapid adoption of injecting drug use among younger and male heroin users in the area. They also identified needle sharing activity as particularly common among this group of IDUs. There is also evidence of an HIV epidemic affecting IDUs in Ho Chi Minh City, Vietnam.4 In a series of three cross sectional studies of IDUs in Ho Chi Minh City between 1995 and 1998, the prevalence rate of HIV infection was 44% and 38.5% for IDUs “on the street” and “in the rehabilitation centre” respectively. The authors also identified significant rates of needle sharing and use of “shooting galleries” within this population. An explosive epidemic of HIV-1 has also been documented among IDUs in the former Soviet Union republics, with the Russian Federation currently experiencing the fastest growing HIV epidemic in the world. The two largest epidemics have occurred in the Moscow and Irkutsk regions in the Russian Federation, where 44% of all HIV infections in Russia have been identified.5 Here, subtype A predominates. Similar data from the republic of Maldova describe a predominance of subtype A disease, while the Romanian epidemic among IDUs is mainly subtype F.6 There is also evidence for rapid spread of HIV infection in IDUs in Belarus, Kazakhstan, Ukraine, India, and Malaysia.7–9 Factors responsible for the rapid spread in epidemic proportions among IDUs in these countries, centres around the concept of the establishment of a “risk environment.” Key factors within this environment include rapid diffusion within the cohort, population migration and mixing, economic transition and decline, increasing unemployment and impoverishment, declines in public health revenue, and political, ideological, and cultural transition. This highlights the urgent need for interventions targeting these areas. Without such focused intervention, the prevalence of HIV infection among IDUs will reach levels comparable to that of the heterosexual risk groups of African nations.

In developed countries, where the HIV epidemic has progressed as least one decade ahead of such countries in transition, the question must be asked as to whether the epidemic of HIV infection in IDUs is now under control.

After the original epidemic of HIV infection in IDUs was identified, many countries acted promptly with huge resources focused towards drug treatment services, methadone maintenance programmes, and needle exchange services. Initial results of the benefits from these efforts have yielded conflicting results.

In Amsterdam, after the introduction of intensive prevention and treatment measures, there was a significant reduction in risk behaviour among IDUs between 1986 and 1991.10 However, while the rate of behavioural change significantly reduced between 1991 and 1993, there has been no substantial change in risk behaviour or HIV seroprevalence since that time. Vancouver introduced the largest needle exchange programme (NEP) and street nurse programme in North America in 1988.11 While the seroprevalence of HIV infection remained stable initially, there has been a rapid increase in seroprevalence since 1994. Strathdee et al have also demonstrated that 40% of both HIV seropositive and seronegative IDUs persist with needle sharing and lending.12 Data from Dublin also demonstrate a recent increase in the incidence of HIV infection in IDUs; however, the actual increase is difficult to estimate.13

Other centres have had more success with their programmes for HIV prevention in IDUs. Des Jarlais et al have described five cities (Glasgow, Lund, Sydney, Tacoma, Toronto) able to maintain a low seroprevalence of HIV, all centres having common prevention strategies—implementation of prevention activities when seroprevalence of HIV was still low, provision of sterile injecting equipment, and community outreach programmes to IDUs.14 The same authors have outlined data from New York City, where the seroprevalence of HIV infection remains stable with a rate of 56% between 1984 and 1992.15 They also described data that those patients participating in NEPs are three times less likely to be HIV positive than those not participating. More recent data from the same group describe a decline in the epidemic among IDUs in New York City. The authors describe a declining phase (current incidence rate of one per 100 person years) characterised by low incidence and declining prevalence, and the authors suggest that very large seroprevalence HIV epidemics may be reversed.16 In the same period they also demonstrate a significant reduction in needle sharing activity from 46% in 1990 to 24% in 1997. Some other countries have also reported stabilisation of their HIV incidence data in IDUs, with Italy describing stable incidence rates between 1993 and 1999.17 Cities with most success in averting or reversing HIV epidemics among IDUs appear to have introduced intervention methods that emphasise rapid reorientation towards “user friendly” and “low threshold,” community based approaches, in conjunction with public policies and funding supportive of such interventions.


The current standard of care for the treatment of HIV infection involves combination antiretroviral therapy with nucleoside reverse transcriptase inhibitors (NRTIs), nucleotide reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitors (NNRTIs), and protease inhibitors.18 Published data show that morbidity and mortality related to HIV infection have significantly reduced since the introduction of such therapy.19 Unfortunately, many of the large clinical trials assessing the efficacy of such therapy either exclude IDUs from the study population, or have only included IDUs from developed countries with presumed subtype B disease. It is therefore not realistic to equate the results of such trials with potential viral efficacy outcomes in all IDUs.

Previous studies have demonstrated that IDUs are less likely than other risk groups to receive antiretroviral therapy (ART), increasing to threefold less likely if they are not enrolled in a drug treatment programme.20 Both French and Irish studies have confirmed these findings.21,22 Strathdee et al describe only 40% of their ART eligible IDUs to be receiving ART, and most of these were only receiving dual therapy, despite the widespread availability of ART in their unit.20 That study was reflective of the period 1996–7. More recently the same group in Vancouver demonstrated an increase in uptake of ART by IDUs, with 60% of IDUs eligible for ART receiving it, but with still 30% of treatment eligible IDUs not receiving therapy.23 Many factors have been associated with poor acceptance of ART, including continued injecting drug use and non-enrolment in a drug treatment clinic.

Owing to the absence of large published randomised controlled clinical trials of IDUs and ART, we are left with small cohort studies of the efficacy of ART in this patient population. From the data that are available, there have been conflicting results. Palepu et al have provided some recent data on enhanced uptake of ART among IDUs in Baltimore between 1996 and 1998.24 Not only do they demonstrate increased uptake of HAART by IDUs (pre-1997 40%, post-1997 89.2%), but they also demonstrate an improved outcome to therapy. Data from this group also suggest that IDU, as a risk factor, was not significantly associated with the time to achieving viral suppression compared with non IDUs. In contrast, Poundstone et al have also recently published results on differences in disease progression in the era of HAART.25 They found that for IDUs, disease free survival increased from 16% to 34% post-HAART, compared with non-IDUs improving from 65% to 135%. Therefore, while disease free survival is increased in IDUs, there is significant disparity in the response to therapy for the different risk groups.

Individualising HAART by using novel methods of drug dispensing, utilising once daily dosing when appropriate, increasing accessibility to drug treatment programmes, and an awareness of interactions between methadone and HAART may improve the overall outcome for IDUs

Evidently the delivery of ART to IDUs is fraught with difficulties, as a result of which methods of optimising access to care and adherence with therapy have been explored. This not only includes novel mechanisms of delivering care, but also through enhanced knowledge of the pharmacokinetic issues that surround methadone maintenance therapy and ART. The resolution of the latter can only provide enhanced patient care and both initial and extended acceptance of ART. While there are numerous antiretroviral agents available for use in combination antiretroviral regimens, the optimal regimen remains to be defined. As for any chronic medical condition, drugs must be chosen that will have minimal immediate and long term adverse events, and that have minimal impact on the patients’ quality of life. A once daily preparation is therefore preferable, with ease of use, and a degree of forgiveness from drugs with long half lives. Of the currently available antiretroviral drugs, only four are licensed as once daily dosing—ddI, NVP, EFV, and tenofovir, while favourable pharmacokinetic data suggest that d4T, 3TC and, possibly lopinavir/ritonavir and other boosted protease inhibitor combinations may also be appropriate for once daily dosing. Such therapeutic options are prime candidates for programmes of directly observed therapy (DOT).

St James’s Hospital in Dublin first described the concept of delivering HAART to IDUs at methadone maintenance clinics. Clarke et al describe a cohort of 39 patients receiving HAART as DOT through their drug treatment clinic.26 At 48 weeks of follow up 58% of a mixed cohort of ART naive and experienced patients had achieved maximal viral suppression (<50 copies/ml). Stenzel et al have also studied a similar cohort of 37 patients in Vancouver who received ART as DOT over a 12 month period.27 DOT was not, however, provided at their methadone maintenance clinic, but provided by an outreach worker in the patients’ home environment. Fifty six per cent of their original cohort were still receiving treatment at 12 months, with a mean reduction in viral load of 1.28 copies/ml. Both the Vancouver and Dublin studies depend on patients taking their evening medications by self administration. To eliminate the nocturnal element of patient self medication, Fischl et al have studied the efficacy of DOT in a correctional facility where DOT is provided for all doses.28 Initial reports from DOT programmes of antiretroviral therapy in the Department of Corrections in the United States, have shown excellent outcomes. Fischl compares the efficacy of combination antiretroviral therapy in two groups of antiretroviral naive patients: patients receiving ART as DOT in the Department of Corrections (50 patients), and patients self administering therapy (SAT) attending the Clinical Research Unit (50 patients). They demonstrated a statistically significant difference in efficacy outcome between the two cohorts after 80 weeks of follow up, with 95% of the DOT group compared with 75% of the SAT group below the limit of detection.

Another study confirms the potential strength of ART when taken with complete compliance. Altice et al looked at the administration of ART within a correctional institute in the Connecticut Department of Corrections.29 They found the acceptance of ART (80%) and adherence (84%) to HAART among this group of prisoners (n = 205) to be high. Interestingly, they noted no difference in adherence among those receiving DOT (82%) or SAT (85%). But, while the study has just been reported, it is worth noting that the study was performed in 1996 when 79% of the patients were receiving only monotherapy or dual nucleoside reverse transcriptase inhibitor therapy.


The rate of hepatitis C virus (HCV) co-infection in HIV infected IDUs has been reported as between 60% and 90%.30 Since the introduction of HAART, treatment of HCV is now seen as being a priority, with newer therapeutic options available for HCV. Essentially, the real concern for IDUs is that the reduced morbidity and mortality associated with HAART as graphically illustrated by Palella et al,19 will be inversely related to morbidity and mortality from concurrent HCV infection, and not protease inhibitor (PI) therapy (fig 1).

Figure 1

A hypothetical reciprocal relation between the increased use of antiretroviral drugs, and hepatitis C related morbidity and mortality. (Adapted from Palella et al19.)

The most promising approach to the treatment of HCV has been the introduction of combination therapy with interferon alfa-2b (IFN) and ribavarin. While this therapy has proved efficacious in the HIV negative population, the efficacy of such therapy in HIV co-infected patients is as yet unclear.31 Landau et al describe the long term efficacy of combination therapy with IFN and ribavarin in patients with severe chronic hepatitis and HIV infection.32 Those with genotype 3a disease had an improved response to therapy with end of treatment response and sustained viral suppression (0.002, 0.003), but 29% of patients discontinued treatment early due to significant adverse drug reactions. The overall response rate seen in this cohort was 20%. Nasti et al have shown combination HCV therapy to be well tolerated, with 31% of patients achieving sustained viral suppression.33 Initial problems associated with therapy for HCV, encountered especially with IDUs, include depressive or psychiatric symptoms and issues around self administering subcutaneous injection. Many of these initial problems may be reduced by the introduction of the pegylated form of interferon with once weekly dosing. Studies differ on the efficacy of pegylated interferon with ribavarin treatment in HIV negative patients, but response rates as high as 88% have been seen with genotype 2 and 3, as opposed to response rates of 48% in patients with genotype 1 disease.34 Other issues yet to be resolved include an in vitro demonstrated interaction between ribavarin and the NRTIs.35 Patient studies are currently under way to determine the pharmacokinetic and clinical significance of this interaction.

Des Jarlais recently produced an article entitled “Hepatitis C among drug users: Déjà vu all over again,” suggesting that the delayed response to the introduction of widespread IDU prevention measures throughout the United States (and indeed Europe), will result ultimately in further rapid spread of the infection and limit treatment options for patients.36 He challenges policy makers to learn from mistakes that have been made, and are still being made during the HIV/AIDS pandemic.

On a more positive note, in a study from Germany, approximately 30% of IDUs were identified as co-infected with the flavivirus GB virus C (also known as hepatitis G-V), a virus not yet known to cause disease. Fortunately, it was determined that co-infection with HGV is associated with a reduced mortality rate from HIV infection, and a slower progression to AIDS. This survival benefit remained present even in the years post-HAART.37 Tillmann et al also demonstrated that HGV viraemia was a more significant predictor of improved HIV disease outcome compared to those patients with evidence of previous HGV infection but no active viraemia.


One of the primary concerns for both physicians prescribing HAART to IDUs, and for the IDUs themselves, is whether antiretrovirals interact with methadone. Initial experience, both international medical and “on the street,” was that antiretroviral therapy would reduce the effect of methadone, with the consequent emergence of uncomfortable symptoms of opioid withdrawal. The clinical prescribing of methadone is not an exact science, with a poorly defined dose-response curve. Studying the interactions between other drugs and methadone has therefore proved to be difficult. The pharmacokinetic interaction between methadone and the NRTIs has been described; however, studies are limited by the intracellular phosphorylation of the NRTIs, with interpretation of results often difficult.38 Methadone is primarily metabolised by CYP 3A4, and the NNRTIs and PIs are potential substrates for this pathway.39 A series of four clinical and pharmacokinetic studies have evaluated the interaction between methadone and four commonly prescribed antiretrovirals—nevirapine, efavirenz, nelfinavir, and lopinavir.40–43 Pharmacokinetic data, the appearance of methadone withdrawal symptoms, and the need for additional methadone dosing were documented for all patients.

The results of the pharmacokinetic studies demonstrated a highly significant reduction in methadone AUC (0–24 hours) of 60%, 50%, 37%, and 36% after the initiation of efavirenz, nevirapine nelfinavir, lopinavir/ritonavir respectively. Clinical data demonstrated that all patients who received nevirapine or efavirenz, 20% of patients who received nelfinavir, and none of the patients who received lopinavir/ritonavir complained of symptoms of opioid withdrawal after days 7–10 of therapy. The mean increase in methadone dose prescribed was 10–20%. The reasons for this discrepancy in pharmacokinetic and clinical data for some patients are unclear, and may be related to plasma protein binding, p-glycoprotein, or differing enantiomer effects. A recent study by Gerber et al of the clinical and pharmacological effects of ritonavir and saquinavir on methadone metabolism has provided new data of the effect of plasma protein binding on methadone.44 They demonstrated a significant discordant effect on R and S isomers with a 40% reduction in S-methadone compared to a 32% reduction in R-methadone. Unlike other studies of methadone metabolism, they also corrected for changes in plasma protein binding associated with ritonavir/saquinavir therapy and, interestingly, demonstrated the actual reduction in free methadone isomers to be 24.6% for S-methadone and 19.6% for R-methadone. Further research is required in this area to specifically clarify the pharmacodynamics of this interaction between antiretrovirals and methadone.

Even by utilising programmes of DOT, and anticipating interactions between antiretrovirals and methadone, we are still failing to meet the requirements of many patients from this risk group, There remains a cohort of patients who fulfil standard criteria to commence HAART, but fail to access the available services to obtain therapy. Patient friendly and patient convenient methods are already used in many branches of medicine to improve both the uptake of available facilities and to enhance patient care. The “one stop shop” approach, as provided by national breast screening campaign, and multidisciplinary diabetic clinics, obviates the need for numerous clinic appointments, and ultimately enhances the care of their target populations. Equally, for IDUs, it is essential that we provide “patient friendly” facilities to further improve their access to care. One of the significant practical problems seen in the Dublin cohort is the lack of proximity between the drug treatment clinics and the HIV clinic. Traditionally, drug treatment clinics are situated off hospital grounds, away from the general medical services. A potential way to improve this would be to develop a unit within the hospital setting, where patients could simultaneously access drug treatment programmes and HIV assessment clinics.

The paradigm of HIV treatment has shifted from simply treating HIV infected patients to optimising patient care by individualising patient care. The real challenge for all of us involved in the care of HIV infected patients is the treatment of those patients marginalised in society—that is, IDUs, people of colour, refugees, and asylum seekers. Achieving this demonstrates an ability to truly individualise therapy for a patient with particular needs. Perhaps it is no longer appropriate to define treatment success in terms of non-selected patient cohorts but rather on our ability to individualise therapy to those HIV infected patients from marginalised groups within our society.

Difficulties of providing long term medications to this unique group of patients


Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.