On the basis of the limited solubility of dapsone at a pH of 7.0 in vitro [4], it is widely believed that antacids contained in the didanosine preparation decrease the bioavailability of dapsone. This in turn leads to recurrence of P. carinii pneumonia [4, 5]. Although the antacids in the didanosine tablet formulation dramatically reduce the absorption of other concomitantly administered agents such as ciprofloxacin [6], no clinical pharmacokinetic data support a didanosine-dapsone interaction. We did two studies to determine 1) whether the currently available didanosine tablet decreases the absorption of dapsone and 2) whether the magnesium-aluminum antacids or other excipients in the tablet can produce a decreased absorption.

Methods

Top Methods Statistical Analysis Results Discussion Author & Article Info References

Participants

Six healthy men (group 1) and six men seropositive for the human immunodeficiency virus (HIV) (group 2) participated in two separate, randomized, two-period, two-treatment, crossover studies. The ethics committee of the Ottawa General Hospital, Ottawa, Ontario, approved the study, and all participants provided written informed consent. Each study contained a 21-day washout period between treatments. The mean (±SD) age and weight were 29 ± 5 years and 75.8 ± 11.4 kg, respectively, in group 1 and 41 ± 7 years and 70.5 ± 6.5 kg, respectively, in group 2. The median CD4 lymphocyte count in group 2 was 176 cells/mm³ (range, 116 to 408 cells/mm³). Healthy volunteers were excluded if they had an abnormal laboratory test result or hypersensitivity to any drug or if they used any other medications. Patients who were positive for HIV were excluded if they met one of the following criteria: hypersensitivity reactions to sulfonamides; current therapy with antacids, histamine-2 antagonists, or enzyme-inducing agents; evidence by history or physical examination of gastrointestinal malfunction; any liver function test result more than five times normal; a hemoglobin level of less than 100 g/L; and a neutrophil count of less than 10⁹/L.

Study Protocol

In the first phase of the healthy volunteer study, participants received 100 mg of dapsone (Avlosulfan, Wyeth-Ayerst Canada, Inc., Montreal, Quebec, Canada) with 180 mL of tap water after an overnight fast. In the second phase, participants received two didanosine placebo tablets (Bristol-Myers Squibb Co., Syracuse, New York), which contained 15.6 mmol of magnesium and a slightly higher amount of aluminum (proprietary information), every 12 hours for four doses. The placebo tablet was identical to the marketed tablet except that it contained no didanosine. After an overnight fast, the participants ingested the third didanosine placebo dose. Each tablet was chewed for approximately 30 seconds and swallowed. The participants then rinsed their mouths with 60 mL of water. Five minutes later, participants received 100 mg of dapsone with 120 mL of tap water. Participants were allowed to eat 4 hours after dapsone ingestion and received their fourth and final dose of didanosine placebo about 12 hours after the dapsone dose. The protocol was exactly the same for the HIV-infected patients, except that patients received two 100-mg didanosine tablets instead of placebo and took their dapsone tablet with the 27th of 28 doses of didanosine. The 28th and final dose of didanosine was administered approximately 12 hours after dapsone ingestion.

In both phases of the studies, venous blood samples were collected before dapsone ingestion and 0.5, 1, 2, 3, 4, 5, 6, 8, 24, 48, and 72 hours after the doses were ingested. In the HIV-positive patients, additional samples were obtained at 1.5 and 10 hours. Plasma was harvested immediately [7] and stored at –80 °C until analysis was done (within 4 weeks of harvesting).

Drug Analysis

Plasma dapsone concentrations were measured by high-performance liquid chromatography [8]. We constructed standard curves each day over the concentration range of 0.1 to 3 µg/mL in plasma and fitted data by weighted linear regression analysis. Coefficients of determination were greater than 0.996 for all curves. Predicted high-performance liquid chromatography concentrations of all quality-control samples at three concentrations—0.25, 1.3, and 2.5 µg/mL—were less than 15% from the nominal value. Within- and between-batch coefficients of variation were less than 8% at each concentration.

Pharmacokinetic Analysis

We analyzed the plasma dapsone concentration-versus-time data by noncompartmental methods [9]; these data included the highest observed concentration (C_max), the time to C_max (t_max), the area under the plasma concentration-versus-time curve (AUC), and the plasma terminal disposition half-life (t₁/2,z).

Statistical Analysis

Top Methods Statistical Analysis Results Discussion Author & Article Info References

We separately analyzed data from each group. Differences in the mean pharmacokinetic variables of dapsone between treatments were evaluated by the analysis of variance appropriate for a crossover study [10]. Analysis of variance summary statistics were based on least-squares geometric means. Point estimates for differences in treatment means were expressed as ratios of the geometric means of the coadministered treatment to the means of the dapsone-alone treatment. In the analysis of variance model, the effects of treatment and period were tested by the mean-square residual, and effects of sequence were tested by the participant-within-sequence mean-square term. Calculations of intraparticipant coefficient of variation and power have been previously described [11]. We compared median t_max values for the two treatments using the nonparametric Wilcoxon rank-sum test appropriate for a two-period, crossover design [12]. The difference between treatment means or medians was considered to be significant at P < 0.05.

Results

Top Methods Statistical Analysis Results Discussion Author & Article Info References

Mean pharmacokinetic data for dapsone after each treatment in groups 1 and 2 are summarized in (Table 1 and Table 2), respectively. In both groups, the pharmacokinetic variables determined when dapsone was administered with the didanosine placebo or didanosine did not statistically significantly differ (P > 0.10) from those determined for dapsone alone. The intraparticipant coefficient of variation was low (< 20%) for each pharmacokinetic variable. For the AUC and t₁/2,z variables in group 1, a difference of less than 15% between treatment means was seen; 90% CIs for point estimates ranged from 73% to 140%. The C_max values increased by 21% when dapsone was administered with the placebo tablets. For the same variables in group 2, a difference of less than 10% was seen; 90% CIs for point estimates ranged from 77% to 129%. Both studies had power of at least 80% at the 5% significance level to detect a decrease of more than 33% in C_max, AUC, and t₁/2,z values relative to the reference dapsone alone.

Discussion

Top Methods Statistical Analysis Results Discussion Author & Article Info References

Our results corroborate earlier findings with other alkalinizing agents and dapsone. Breen and colleagues [15] evaluated the effect of multiple doses of a magnesium-aluminum antacid on dapsone absorption in healthy volunteers and found no differences in any pharmacokinetic variables between the two treatments. In addition, Mirochnick and coworkers [16] showed that magnesium-aluminum antacids actually enhanced the absorption of dapsone when the latter was given as a new liquid preparation [16]. Coleman and colleagues [17] evaluated the effect of cimetidine, 1200 mg daily for 3 days, on inhibition of the metabolism of dapsone. In addition to finding that cimetidine slowed the metabolic clearance of dapsone, these investigators also showed that cimetidine coadministration did not decrease the apparent rate of absorption and peak serum concentrations of dapsone. Because cimetidine administration usually increases the stomach pH to 4.0 to 7.0, a decrease in dapsone levels might have been expected. In a recent study, P. carinii pneumonia recurred in only 2 of 13 HIV-infected patients receiving dapsone and alkalinizing agents such as didanosine preparations, antacids, and histamine-2 antagonists [18]. Although these small numbers preclude firm conclusions, they further support the improbability of a drug interaction between dapsone and alkalinizing agents. However, our results cannot be extrapolated to proton-pump inhibitors and histamine-2 antagonists.

The discrepancy between the results of these pharmacokinetic studies and the apparent observations of dapsone failure is puzzling. Dapsone is more soluble in vitro, and the rates of dissolution and disintegration are faster at a pH of 1.0 than at a pH of 6.0 to 7.0 because the drug is ionized at a pH of less than 2.0. (Gallicano K. Unpublished data). However, at a pH greater than 6.0 to 7.0, disintegration and dissolution may be increased in the stomach compared with the controlled in vitro conditions because of uncontrolled mixing effects occurring in the stomach. This may explain why Breen and colleagues [15] found that absorption of dapsone was adequate when the stomach pH was about 6.0. Another possible explanation is that at a pH of 7.0, dapsone dissolution decreases and gastric membrane permeability to the nonionized drug increases; for the ionized drug, the opposite changes occur at a pH of 1.0. The net result is no significant change in the absorption of dapsone in the different gastric pH environments.

Our studies in HIV-infected patients and healthy volunteers show that the antacids, other excipients, and didanosine in the currently marketed didanosine chewable tablet did not decrease plasma concentrations of dapsone. Therefore, a drug interaction and increased gastric pH from antacids are unlikely explanations for the apparent failure of dapsone to prevent P. carinii pneumonia when administered concurrently with didanosine preparations.

Drs. Garber and Cameron: Division of Infectious Diseases, Ottawa General Hospital, 501 Smyth Road, Ottawa K1H 8L6, Ontario, Canada.

Ms. Oliveras: Department of Biochemistry, Ottawa General Hospital, 501 Smyth Road, Ottawa K1H 8L6, Ontario, Canada.

Dr. Gallicano: Bureau of Drug Research, Banting Building, Postal Locator 2202C1, Tunney's Pasture, Ottawa K1A 0L2, Ontario, Canada.