Compliance with antiretroviral therapy -- the physician's compliance in prescribing the optimum appropriate regimen and monitoring it and the patient's compliance in taking the medication as prescribed -- is the major determinant of drug treatment success. One way to ensure patient compliance is DOT, in which the ingestion of every drug dose is actually witnessed.

The World Health Organization DOT strategy for the treatment of tuberculosis (TB) advocates DOT with a short-course drug regimen.^[5] Cohort studies with historical controls receiving self-administered antituberculous therapy have shown improved cure rates with DOT in a number of centers. In the United Kingdom, DOT is recommended for patients who are unlikely to adhere, including the homeless, alcohol or drug abusers, those with multidrug resistance, or those with a history of nonadherence to antituberculous therapy.^[6] In the United States, however, the CDC recommends that DOT be considered for all TB patients.^[7] Given the short- and long-term similarities between the courses of HIV infection and TB and the association between inadequate adherence to therapy and the development of resistant strains, HIV-infected patients who are unlikely to adhere to HAART also should be considered for DOT.

Initial reports from DOT programs of antiretroviral therapy in the US Department of Corrections have shown excellent outcomes. In a retrospective study, Fischl and colleagues^[8] compared the efficacy of combination antiretroviral therapy in 2 groups of antiretroviral-naive patients: those receiving antiretroviral therapy as DOT in a correctional facility (50 patients) and those self-administering therapy and attending a clinical research unit (50 patients). They demonstrated a statistically significant difference in virologic efficacy between the 2 cohorts after 80 weeks of follow-up: 95% of the DOT group had viral loads below the limit of detection (less than 50 copies/mL) versus 75% of the self-administering group.

The program of DOT linking methadone maintenance therapy with HAART has proved very successful in the GUIDE Clinic in Dublin, where 39 patients began therapy between September and December 1998. DOT has been initially effective in this group, with data comparing favorably with those of other HAART trials of IDUs. A significant limitation of this study is that it was observational and did not include a randomized control population. (Following the promising preliminary data from this study, an additional prospective randomized study has been planned to further define the role of DOT in such patients.)

In our cohort of patients receiving DOT, 58% had viral loads below the limit of detection at 52 weeks, with a mean CD4⁺ cell count increase of 43/µL (P < .001). These data compare favorably with data reported by Haberl and colleagues^[9] from the Frankfurt HIV cohort. They describe once-daily therapy with ddI, 3TC, and NVP for a similar mixed cohort of IDUs. At 40 weeks, only 53% of the original 70 patients were still enrolled in therapy, and 10% of this group had HIV RNA levels below the limit of detection (less than 500 copies/mL). Considering that drug treatment clinics report 1-year success rates of 53% to 62% for opioid detoxification,^[10] the overall efficacy data indicate a moderate success rate, but there is considerable room for improvement.

More recently, Stenzel and colleagues^[11] reported the efficacy results of a program of DOT in a cohort of IDUs in Providence, RI. Unlike the Dublin study, in which patients received their HAART with their methadone therapy, patients from this cohort received morning doses of HAART in their homes from an outreach worker, independent of their methadone dose. Studying a cohort of patients similar to the Dublin cohort, these researchers found that 49% of patients were still enrolled after a mean follow-up period of 10 months, compared with a 69% follow-up rate in the Dublin study. They also reported less impressive efficacy data, with a mean reduction in HIV RNA of 1.53 log₁₀ copies/mL at 10 months and with none of the patients achieving complete viral suppression. These 2 small studies suggest that DOT is a useful option for providing HAART to IDUs, particularly in conjunction with methadone maintenance therapy.

In our study, medications were prescribed as once-daily dosing, when possible, to facilitate DOT and achieve optimal adherence. The 16 patients treated with regimens containing ddI received this drug as a single morning dose before their methadone, in accordance with the current licensing for this drug.^[12,13] Thirteen patients received regimens that included both ZDV and 3TC, which permitted the administration of an available combination preparation (Combivir) with a q12h dosing schedule. In 9 other patients, 3TC was given as a single daily dose of 300 mg, administered with their methadone. Although unlicensed for once-daily dosing, 3TC has been shown to be pharmacologically adequate when given once a day.^[14]

With regard to nonnucleoside once-daily dosing, 10 patients were given regimens containing a once-daily dose of NVP. Research has shown that this drug can be given effectively on a once-daily basis.^[15-17]

EFV is licensed for once-daily dosing, and it is recommended that it be taken at night because of the risk of CNS symptoms: 52% of patients receiving EFV report such symptoms,^[18] including dizziness, impaired concentration, somnolence, abnormal dreams, and insomnia, which usually begin on the first or second day of therapy and generally resolve with repeated dosing within 2 to 4 weeks. Dosing at bedtime improves symptoms of diurnal dizziness. All of our patients who are treated with regimens containing EFV, including those in the present study, take their EFV dose in the morning before methadone and are counseled regarding the risk of dizziness. To date, 4 patients have reported minor EFV-related CNS symptoms, with 1 patient discontinuing treatment against medical advice because of the sudden onset of severe anxiety and agitation.

There is a potentially significant interaction between methadone and the nonnucleoside analogue drugs. The nonnucleoside analogues are significant inducers of cytochrome P-450, the primary system through which methadone is metabolized.^[15,18]

In our study, 5 of the 10 patients receiving NVP required an increase in methadone dose (20% to 50%), and 4 patients receiving EFV required an increased dose of methadone as a result of the onset of narcotic withdrawal symptoms. The exact interactions of NVP and EFV with methadone have been published.^[19,20] A complete discussion of the pharmacokinetic interactions between antiretroviral therapies and methadone is beyond the scope of this article, but some of the more pertinent interactions are summarized in Table 2.^[19-48]

A confounding factor for 4 of the patients receiving EFV was the early onset (week 1) of neurologic symptoms, which were initially perceived by the patient as withdrawal symptoms. However, because of the prolonged period of hepatic enzyme induction, during which the drug concentration does not reach maximum levels for at least 10 to 14 days,^[49] neurologic symptoms in the first week of therapy cannot be attributed to methadone withdrawal.

Two patients discontinued treatment as a result of development of hepatotoxicity. Both were hepatitis C RNA-positive. In the first patient, who was receiving a regimen containing d4T, ddI, and EFV, significant elevation in all liver function parameters developed after 7 weeks (with an aspartate aminotransferase [AST] level 6 times the normal level). Once treatment was discontinued, liver function returned to normal within 1 month. The second patient acquired hepatitis B and hepatitis C in the 6-month period before commencing HAART with ZDV, 3TC, and NVP. All liver markers became abnormal (AST level 5 times the normal level) within 6 weeks of commencing therapy, returning to normal within 1 month of discontinuing all medications. The deterioration in liver function could be due to an immune reconstitution phenomenon as previously described.^[50] It could also be due to the fact that all of the prescribed medications, especially EFV and NVP, have been associated with hepatotoxicity.^[15,18] The rapid onset of deterioration and the equally rapid reversal when medications were stopped make a drug-induced reaction the more likely cause.

The combination of directly observed methadone and antiretroviral therapy, which we have termed "DAART" (directly administered antiretroviral therapy), is an attractive solution for patients for whom HAART may otherwise have been inaccessible. The paradigm of HIV treatment has shifted from simply treating HIV-infected patients to optimizing patient care by individualizing care. The real challenge for all involved in the care of HIV-infected patients is the treatment of those patients who are marginalized in our society, such as IDUs, persons of color, refugees, and asylum seekers. Achieving this demonstrates an ability to truly individualize therapy for a patient with particular needs. Perhaps it is no longer appropriate to define treatment success in terms of randomized patient cohorts but rather by our ability to individualize therapy for those persons from marginalized groups within our society who are infected with HIV.