The hypotheses of this study are:

1. Ondansetron will show a decrease in cocaine use from baseline in individuals with HIV who are cocaine using.
2. Ondansetron will show a decrease in cravings from baseline in individuals with HIV who are cocaine using.

After informed consent and screening, HIV infected individuals who are cocaine dependent and qualify for the study will be offered ondansetron 4mg BID for six weeks in an open label format 4mg BID has been found to have efficacy compared to placebo. At screening and then at each visit, they will be asked to provide urine and a drug of abuse screen will be conducted to assess for cocaine. They will be asked to detail their recent cocaine use in the last month and then will be given a visual analog scale to assess their craving for cocaine. They will be asked to return weekly for 6 weeks to receive a week's supply of ondansetron and to give a urine sample that will test for cocaine. They will fill out a time line follow back for the past week and asked to assess their craving for cocaine on a visual analog scale.

In 1996, the use of protease inhibitors and triple therapy known as HAART (highly active antiretroviral therapies) became widespread in the United States for the treatment of HIV. This changed the disease process from a relentless and progressive one to a chronic one, with the resulting need to focus on issues related to adherence. Because of unique aspects of viral resistance with HIV, an adherence rate of 95% or higher is required to minimize or halt the progression of the disease (DeMasi et al., 2001; Gross, Bilker, Friedman, & Strom, 2001). Cocaine use and dependence has been found to have deleterious effects on HIV infected individuals (Fiala et al., 2005) Cocaine use hastens disease progression, increases viral loads and decreases CD4 counts(Arnsten et al., 2002; Baldwin, Roth, & Tashkin, 1998; Hurwitz, ; Kapadia, Vlahov, Donahoe, & Friedland, 2005; Lucas et al., 2006; Roth et al., 2002; Roth, Whittaker, Choi, Tashkin, & Baldwin, 2005) Cocaine use correlates with increased "no show" clinic visits and decreased or no-adherence to HIV medications. (Hinkin et al., 2007; Ingersoll, 2004; Palepu, Horton, Tibbetts, Meli, & Samet, 2004; Sharpe, Lee, Nakashima, Elam-Evans, & Fleming, 2004) There appear to be multiple reasons for non-adherence in this substance abusing population. Factors such as forgetting and running out of medications have been implicated with substances such as crack cocaine and heroin. (Ingersoll, 2004; Kerr et al., 2004)

A medication that decreases or ceases cocaine use potentially improves adherence at clinic appointments as well as adherence to medication. In addition, it might indirectly decrease the progression of the disease. Ondansetron, a serotonin type 3 receptor antagonist (5HT3), at 4mg twice a day has been shown to have a greater rate of improvement in percentage of participants with a cocaine-free week compared to placebo in cocaine dependent non HIV infected treatment seeking individuals. (Johnson, B.A. 2006)

Ondansetron is FDA approved for chemotherapy induced nausea and vomiting, postoperative nausea and vomiting and radiation induced nausea and vomiting. Ondansetron has shown efficacy for nausea and vomiting in HIV infected individuals for palliation, as well as for diarrhea due to cryptosporidium infection.(Currow, Coughlan, Fardell, & Cooney, 1997; Gompels et al., 1993; Schworer, Hartmann, & Ramadori, 1994) Until recently, ondansetron's cost was prohibitive. However, as a generic drug it has become more affordable. Among the benefits of this drug is the qualitative benefits for HIV infected individuals with nausea and diarrhea as well as cocaine dependence. HIV medications can cause nausea (M. O. Johnson, Stallworth, & Neilands, 2003; M. O. Johnson et al., 2005; O'Brien, Clark, Besch, Myers, & Kissinger, 2003; Reynolds & Neidig, 2002) and thus adherence might be improved with this medication in cocaine using individuals. Ondansetron is well tolerated in HIV infected individuals, and according to micromedex there are no known drug interactions with HIV medications (Gompels et al., 1993)

Ondansetron has a mild adverse events profile. In studies with cocaine dependent individuals who were not HIV infected, our group found that Ondansetron had fewer side effects than placebo (B. A. Johnson et al., 2006). Also, in that same study, Ondansetron recipients attended more sessions than those of placebo. Both of these factors make this drug an appealing option for cocaine dependent HIV infected individuals.

Inclusion Criteria:

• HIV infected individuals who are cocaine dependent based on the Diagnostic and Statistical Manual IV-R
• 18-64 years of age
• Ability to read and write in English
• Seeking treatment for cocaine dependence
• Drug of dependence screen or benzoylecgonine urine specimen (metabolite of cocaine) positive within the 2 weeks prior to study medication administration
• If female, non-pregnant or breast feeding and willing to use acceptable form of contraception including oral contraceptives, hormonal (levonorgestrel) or surgical implants or barrier plus spermicide
• Liver function tests and Chemistries from CARECAST or by blood draw within the last 3 months that must show no disease of the kidney or liver that could result in altered metabolism or excretion of the study agent. AST and ALT can be no greater than twice the upper limit of normal
• An EKG that shows no clinically significant abnormalities including but not limited to bundle branch blocks, bradycardia with heart rate less than 50, tachycardia with heart rate greater than 105

Exclusion Criteria:

• Non-English speaking (As these individuals are filling out rating forms, the criteria for exclusion is directly related to ability to read and process information in English.)
• Inability to process and sign informed consent
• Pregnant or nursing or unwilling to use contraception if female
• Restrictions on use of other drugs or treatments: The following medications which are established or theoretically have a drug-drug interaction include: Apomorphine (established),Mesoridazine (theoretical), Pimozide (theoretical), Thioridazine (theoretical), Acecainide (theoretical), Amiodarone (theoretical), Arsenic Trioxide (theoretical), Azimilide (theoretical), Bretylium (theoretical), Dofetilide (theoretical), Droperidol (theoretical), Enflurane (theoretical), Halothane (theoretical), Ibutilide (theoretical), Isoflurane (theoretical), Isradipine (theoretical), Sematilide (theoretical), Sotalol (theoretical)
• History of neuroleptic malignant syndrome
• Allergy to ondansetron
• Clinically significant cardiovascular abnormality (EKG) or history of arrhythmias