Study objective: To determine the feasibility of therapeutic drug monitoring for adjusting low serum antimycobacterial concentrations in patients with both tuberculosis and advanced human immunodeficiency virus (HIV).
Design: Retrospective cohort study.
Data source: De-identified dataset from a tuberculosis clinic.
Patients: Twenty-one patients (median age 38 yrs, range 25-68 yrs) with advanced HIV infection (CD4(+) cell count < 100 cells/mm(3)) who received treatment for active tuberculosis between March 2002 and September 2007.
Measurements and main results: We evaluated data based on the practices performed at the tuberculosis clinic. After the daily doses of isoniazid and rifamycins (rifampin or rifabutin) were ingested, serum concentrations were obtained at 2 hours for isoniazid and rifampin, at 3 hours for rifabutin, and, when possible, at 6 hours for all three drugs to detect delayed absorption. Antimycobacterial drug concentrations were compared with published reference levels, and dosages were adjusted to achieve desired concentrations. Costs of monitoring were recorded for all patients. Of the 21 patients, 18 (86%) had low serum concentrations of at least one drug 2 hours after ingestion: 2 (10%) had low isoniazid concentrations, 5 (24%) had low rifamycin concentrations, and 11 (52%) had low serum concentrations of both drugs. The median number of dosage adjustments to attain normal concentrations was 1 (range 0-4 adjustments). The median cost/patient for therapeutic drug monitoring was $619 (range $230-1948). The median final doses to achieve normal concentrations were isoniazid 600 mg/day (range 300-1500 mg/day), rifampin 1050 mg/day (range 600-1200 mg/day), and rifabutin 300 mg (range 150-450 mg) 3 times/week. No patient demonstrated any adverse effects attributed to these higher doses.
Conclusion: Low serum concentrations of antituberculous drugs, which suggest malabsorption, are common among patients with advanced HIV who also have tuberculosis but can be overcome with higher doses. Therapeutic drug monitoring may be an effective tool to optimize therapy, but needs further study.