Secondary Logo

Journal Logo

Clinical Science

A Phase-I Study of the Safety, Pharmacokinetics, and Antiviral Activity of Combination Didanosine and Ribavirin in Patients with HIV-1 Disease

Japour, Anthony J.; Lertora, Juan J.*; Meehan, Patricia M.†‡; Erice, Alejo§; Connor, James D.; Griffith, Brigitte P.; Clax, Pamela A.**; Holden-Wiltse, Jeanne†‡; Hussey, S.; Walesky, Mary; Cooney, Elizabeth; Pollard, Richard††; Timpone, Joseph**; McLaren, Colin‡‡; Johanneson, Nils§§; Wood, Kenneth‡‡; Booth, David K.¶¶; Bassiakos, Yannis†‡ Clyde S. Crumpacker, for the AIDS Clinical Trials Group 231 Protocol Team

Author Information
Journal of Acquired Immune Deficiency Syndromes and Human Retrovirology: November 1, 1996 - Volume 13 - Issue 3 - p 235-246
  • Free


Nucleoside analogues require intracellular phosphorylation prior to acting as inhibitors of the human immunodeficiency virus type 1 (HIV-1) reverse transcriptase. 2',3'-Dideoxyinosine (ddI) is an analogue of the naturally occurring purine nucleoside inosine. ddI has potent anti-HIV-1 activity in vitro in human T cells, monocytes, and macrophages (1,2). ddI is converted to its active form, 2',3'-dideoxyadenosine-5'-triphosphate (ddATP), first by conversion of ddI to 2',3'-dideoxyinosine-5'-monophosphate (ddIMP) by cytoplasmic phosphotransferase, followed by conversion to 2',3'-dideoxyadenosine-5'-monophosphate (ddAMP) by adenylosuccinate synthetase and adenylsuccinate lyase, and finally formation of its active form, ddATP, by cellular kinases (3-5)(Fig. 1).

Ribavirin (1-β-D-ribofuranosyl-1H-1,2,4-triazole-3-carboxamide) is a guanosine analogue; it is a broad-spectrum antiviral agent that is converted to its active metabolite, a 5'-phosphate derivative, by cellular enzymes (6). Although oral ribavirin is safe when given to HIV-positive individuals, it has not been shown to have significant antiretroviral activity (7-9). Nevertheless, ribavirin increases the intracellular phosphorylation of ddI by inhibiting inosine monophosphate dehydrogenase (IMPD), an enzyme required for the synthesis of the purine nucleotide guanosine triphosphate (10,11)(Fig. 1). IMPD inhibitors such as ribavirin and tiazofurine block the utilization of IMP for guanidinetriphosphate synthesis, thus yielding higher concentrations of IMP to act as a phosphate donor for the reaction of ddI to ddI-monophosphate (10). This mechanism provides a basis for additive or synergistic inhibition of HIV-1 replication by ribavirin and ddI. Previous in vitro studies have shown that ribavirin enhances the antiviral effects of ddI and other purine nucleosides (12-15). In peripheral blood mononuclear cells (PBMCs) infected with clinical and laboratory HIV-1 isolates, 4.0 μM ribavirin enhanced ddI HIV-1 RNA inhibition by tenfold (16).

Clinical studies have demonstrated the efficacy of ddI therapy for HIV-1 (17-19). Ribavirin is safe and tolerable at 800 mg/day when given to subjects with AIDS, AIDS-related complex, or asymptomatic HIV-1 infection (9,20,21). Clinical trials using ddI and ribavirin individually have not suggested that synergistic or overlapping toxicity should be expected if both drugs are used together. Pancreatitis and peripheral neuropathy are the most severe toxicities associated with ddI therapy (22). Although ribavirin has not been shown to cause any adverse effects on the pancreas or peripheral nervous system, mild anemia has been reported with ribavirin 1200-1600 mg/day (7). Although there is no evidence that ddI and ribavirin share a common metabolic pathway that may suggest a pharmacokinetic interaction regarding plasma levels, a formal evaluation of ddI pharmacokinetics before and after the addition of ribavirin was conducted, given previous reports indicating a relationship between ddI plasma levels and clinical response (23,24).

By shifting intracellular nucleotide pools, the coadministration of ddI and ribavirin, if safe and effective, may increase the antiretroviral effectiveness of ddI either by increasing the magnitude of HIV inhibition or by increasing the durability of the antiretroviral response. Moreover, the combination of ddI to ribavirin offers an alternative regimen for patients unable to tolerate zidovudine (ZDV)-containing regimens. We conducted a phase-I trial to evaluate the pharmacokinetics, safety, and activity of ddI and ribavirin coadministration in patients with HIV-1 disease.


Study Design

Eligible HIV-positive individuals were enrolled into the openlabeled study following written institutional review-board-approved informed consent and were treated with 4 weeks of ddI monotherapy followed by 8 weeks of ddI and ribavirin combination therapy for a total of 12 weeks. Patients who had no adverse effects from the combination regimen and wished to continue were treated for an additional 12 weeks for a total of 24 weeks on study. At the conclusion, patients went off study drugs to their physician-prescribed treatment regimen. The study was designed to have >80% power to detect a twofold increase in the toxicity rate of the ddI-ribavirin combination over ddI monotherapy, assuming a toxicity rate of 35% for ddI monotherapy and using a 5% level of significance.

Study Medication

Didanosine (ddI) was administered as 100-mg and 25-mg tablets at a dose of 200 mg p.o. b.i.d. for patients weighing ≥60 kg (125 mg p.o. b.i.d. for patients weighing <60 kg). The ddI tablets were crushed and dispersed in at least 4 oz (1.2 L) of ice-cold water before the patients drank the solution. Patients were instructed to take the ddI at least 1 h before or 2 h after meals and to take ddI throughout the 12 weeks on the study. Didanosine tablets were supplied by Bristol-Myers Squibb (Wallingford, CT, U.S.A.).

Ribavirin was administered as 200-mg capsules at a dose of 600 mg q.d. taken all at once 6 h after the morning ddI dose. Patients began ribavirin treatment after 4 weeks of ddI monotherapy and continued ribavirin for 8 weeks. Ribavirin capsules were supplied by ICN Pharmaceuticals (Costa Mesa, CA, U.S.A.).

Study Population

Patients with a documented positive serum HIV-1 antibody test result were eligible for study entry if they had an absolute CD4+ lymphocyte count of <500 cells/μl, were at least 18 years of age, had adequate baseline hematology profiles (hemoglobin >11.0 g/dl; absolute neutrophil count ≥1,000/mm3; and platelet count ≥75,000/mm3) and chemistry profiles (creatinine ≤1.5 mg/dl; aspartate aminotransferase/alanine aminotransferase and alkaline phosphatase ≤2.5 times the upper limit of normal; total bilirubin ≤2.5 mg/dl; and amylase ≤1.5 times the upper limit of normal), and had a Karnofsky performance score of ≥60. Patients were excluded if they had ever been treated with ddI, had taken ribavirin within 60 days of study entry, had a history of pancreatitis or peripheral neuropathy, had active liver disease or alcohol abuse, were pregnant women, or had opportunistic infections requiring treatment or malignancies requiring systemic cytotoxic chemotherapy or radiation therapy.

Clinical and Laboratory Evaluation

A complete medical history was recorded and a physical examination was performed within 14 days of study entry; symptom-directed exams were repeated every 4 weeks on study and 4 weeks after completing the drug-dosing period. Routine hematology and chemistry (including amylase) tests were performed at baseline and every 4 weeks. Quantitative PBMC HIV-1 microcultures, screening for syncytium-inducing (SI) HIV-1 phenotype, and CD4+ lymphocyte subsets were performed in real time at two preentry time points within 30 days of study entry, at study entry, and every 4 weeks until 4 weeks after completing the drug-dosing period, using standard ACTG-certified procedures (25,26). The baseline CD4+ values were obtained by taking the median of the three pretreatment CD4+ counts. The viral titer obtained from the HIV-1 quantitative microculture was estimated by the maximum likelihood assuming a Poisson regression model and is expressed in infectious units per million (IUPM) PBMCs (27).

Levels of the immune activation molecules neopterin and β2-microglobulin were assessed in serum at baseline and every 4 weeks. Neopterin levels were determined using a commercial radioimmunoassay (RIA) (IMMUtest Neopterin; Henning Beglin). β2-Microglobulin levels were determined using an automated microparticle enzyme immunoassay system (Imx; Abbott Laboratories, Abbott Park, IL, U.S.A.).

HIV-1 RNA was quantified in plasma at baseline and at weeks 4 and 12 by using the nucleic acid sequence-based amplification (NASBA) system (28,29). The assay was performed as recommended by the manufacturer (Organon Teknica Corporation, Durham, N.C., U.S.A.). Plasma was stored at -70°C throughout the study and assayed for HIV-1 RNA in batch. Briefly, RNA was isolated from 100 μl of each plasma sample as described by Boom et al. (30). The RNA was amplified at 41°C in the presence of three enzymes: reverse transcriptase, RNase H, and T7 RNA polymerase. The amplified fragments were detected by electrochemiluminescence in an automated electrochemiluminescence instrument (NASBA QR system). The quantification was obtained by simultaneous competitive amplification of three different RNA constructs containing 104, 105, and 106 RNA copies, respectively, spiked into the patient samples before extraction. Results are expressed as RNA copies per milliliter.

The emergence of a ddI resistance-conferring HIV-1 reverse-transcriptase mutation (leucine→valine at codon 74) was evaluated at the conclusion of the study on samples from baseline, week 4, and week 12. Nested polymerase chain reaction analysis of DNA derived from quantitative HIV-1 microcultures was employed to detect the mutation as previously described (31).

Dosing and Blood Sampling Procedures

The pharmacokinetic studies of ddI were performed at study week 4 (following 4 weeks of ddI monotherapy), at study week 6 (following 2 weeks of ddI plus ribavirin), and at study week 12 (following 8 weeks of ddI plus ribavirin). The rationale for this procedure was to obtain baseline ddI pharmacokinetics after steady-state ddI levels, to repeat ddI pharmacokinetics after 2 weeks of the combination regimen to detect any major change that might have occurred with early institution of the combined regimen, and to obtain ddI pharmacokinetics during steady-state ribavirin levels (week 12).

Blood samples (5 ml each) were collected for measurements of ddI concentrations in plasma at time zero just before the morning ddI dose. Subsequent samples were taken at 0.25, 0.5, 0.75, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 5.0, and 6.0 h after dosing. Blood was collected by venipuncture into EDTA-containing tubes and held at 4°C. Plasma was separated within 1 h of collection with aliquots frozen at -20°C until assayed.

Analytical Methods

Plasma samples were assayed for ddI and ribavirin by RIA as previously described (32,33). The standard curves are generated by RIA Smart software (Packard Instrument, Meriden, CT, U.S.A.) from a set of standards by calculating the amount of precipitated radioactivity as a percentage of total radioactivity present in a standard and then fitting the data for all the standards to a four-parameter logistic curve. All samples, standards, and controls are analyzed in duplicate; a reportable result is the average of duplicate determinations. The percent coefficient of variation for the means of these duplicates was ≤12.9%.

The ddI method has been validated for serum samples with ddI concentrations from 10 to 2,560 ng/ml. The limit of quantification of the method is 10 ng/ml, and concentrations <10 ng/ml are reported as below the quantification limit. The ribavirin method has been validated for serum samples with ribavirin concentration from the limit of quantification of 0.03-30.0 μM.

Pharmacokinetic Analysis

Plasma concentrations of ddI corresponding to treatment with ddI alone (week 4) and ddI plus ribavirin (weeks 6 and 12) over a 6-h sampling time after oral dosing were plotted on a time (hours)-versus-plasma concentration (μmol/L). Profiles were analyzed using a one-compartment or a two-compartment pharmacokinetic model according to the best fit for each individual data set. The models allow first-order absorption, with or without a lag phase, and first-order elimination (PCNONLIN; SCI Software Statistical Consultants, Lexington, KY, U.S.A.). The pharmacokinetic parameters derived from the program were area under the curve (AUC), first-order elimination rate constant (ke), time to peak concentration (tmax), and peak concentration (Cmax). Plasma elimination half-life (t½) was calculated from the elimination rate constant, according to 0.693/ke. The sample size provides a 90% chance of detecting a 20% difference in the ddI AUC caused by coadministration of ribavirin at a 5% level of significance.

Statistical Analysis

The primary analysis was performed on all of the data collected during the first 12 weeks. Adverse-event data were collected from all patients until 60 days following drug discontinuation. The NIAID standardized toxicity-grading system was used to achieve uniformity in reporting of adverse events among study sites. Using modified World Health Organization criteria adopted by the ACTG, toxicity was graded on a 1-4 scale, with severe and life-threatening events graded as 3 and 4, respectively. The estimated toxicity rate was summarized by an exact binomial confidence interval (34). Secondary analyses were performed on data collected from patients who continued on treatment during the extension phase of the study (to week 24). Activity measures are right-censored or left-censored if the value is above or below the quantifiable range of the assay, respectively. Changes in activity measures and pharmacokinetic parameters over time were assessed by Wilcoxon signed-rank tests. Spearman rank correlation was used to assess the strength of relationships between the pharmacokinetic parameters and activity measures. Since this was a single-arm study, it was not possible to make any direct comparisons between the effect of combination ddI and ribavirin therapy relative to ddI monotherapy. All reported p values are two-tailed.


Patient Accrual, Demographics, and Characteristics

Nineteen patients were enrolled at three AIDS Clinical Trials Units (ACTUs): Beth Israel Hospital, Boston-Harvard ACTU, Minnesota ACTU, and Yale ACTU. Demographics and baseline characteristics of the 19 patients enrolled into the study are summarized in Table 1. Patients ranged from 27 to 53 years of age; 32% of the patients had a history of intravenous drug use, and 89% of patients had prior use of antiretroviral therapy. The median duration of antiretroviral use was 15 months (range, 1 day to 53 months). Of the 17 patients reporting prior antiretroviral use, all had taken ZDV and one had taken dideoxycytidine in addition. The median baseline CD4+ cell count was 167 cells/mm3, and the median baseline HIV-1 RNA was 4.9 log10 copies/ml. Of the patients, 42% had SI variants of HIV-1 at baseline. Two patients entered the study with HIV-1 isolates carrying a mutation (T215Y) in the reverse-transcriptase gene known to be associated with resistance to ZDV.

Adverse Events

The toxicity rate was defined as the percentage of patients who experienced at least one grade-3 or 4 toxicity during the first 12 weeks of treatment. To provide a conservative summary of the toxicity profile, the patient who prematurely quit after 2 weeks of treatment without toxicity was excluded. The remaining 18 patients were included in the toxicity analysis. The observed toxicity rate was 22% (95% confidence interval, 10-48%). Sixteen patients completed the 12 weeks on study. Twelve patients elected to continue into the extension phase of the study and, of these, 10 completed the 24 weeks of treatment. Three individuals did not complete 12 weeks of the study; two patients voluntarily withdrew from study treatment during the ddI-monotherapy phase because of grade-2 diarrhea in one and grade-3 constipation in the other, and one patient was discontinued by the study team because of granulocytopenia.

Three patients experienced grade-3 or 4 toxicity during the first 8 weeks of combination treatment: at week 7, one patient had severe nausea that began with the introduction of ribavirin at week 4, necessitating temporary discontinuation of both study drugs; one patient had an increased serum glutamic-oxaloacetic transaminase level at week 12; and one patient had granulocytopenia, as noted previously. Of the 12 patients who were treated on the extended course of therapy, two developed twofold times the upper limit of normal in lipase and/or amylase. Both of these individuals, however, entered the study with asymptomatic elevations in lipase/amylase at 1.5 times the upper limit of normal. The elevations in lipase/amylase were asymptomatic and resolved with discontinuation of the study drugs. No cases of pancreatitis, peripheral neuropathy, or anemia were observed during the study.


At week 4 of ddI monotherapy, peak ddI levels (Cmax) of 5.0 ± 2.09 μmol/l (mean ± standard deviation) were achieved at a mean of 0.49 ± 0.24 h (tmax). The estimated AUC was 6.73 ± 1.18 μmol·h/L, with a plasma elimination rate constant (ke) of 0.59 ± 0.23 h-1 and a plasma half-life (t½) of 1.38 ± 0.60 h. The main pharmacokinetic parameter of interest, AUC, which reflects the overall exposure to ddI, did not change after 8 weeks of combined therapy with ribavirin (p = 0.64) (Table 2). For the remaining parameters, there was no detectable change in the maximum concentration (p = 0.86); there was a modest increase in the rate of elimination of ddI (p = 0.02) and a suggestion of an increase in the time to the maximum concentration (p = 0.06). The magnitudes were small; the median decline in the elimination half-life was 8 min, and the median increase in time to maximum concentration was <7 min. Figure 2 illustrates the plasma concentration-versus-time curves for ddI at weeks 4, 6, and 12, and indicates that the profiles were essentially unchanged by the presence of ribavirin.

Trough serum ribavirin levels at weeks 6, 8, 12, and 16 were similar (3.56 ± 1.30, 3.8 ± 1.13, 4.58 ± 1.31, and 4.30 ± 1.39 μM, respectively), indicating that a steady state had been achieved and maintained for this drug.


Sixteen patients had available measurements of quantitative HIV-1 culture PBMCs and plasma HIV-1 RNA during the first 12 weeks. Of the 16 patients who had measurements after baseline, one had a right-censored titer measurement and three had left-censored titers in later weeks. A significant decline from baseline in HIV-1 titer as measured by quantitative HIV-1 culture log10 was detected both during the ddI-monotherapy phase (week 4, p < 0.0001) and during the combination-therapy ddI + ribavirin phase (week 12, p = 0.0008). No difference in log10 titer was observed between weeks 4 and 12; the median drop observed was 0.90 log10 at week 4 and 0.92 log10 at week 12. Of the 16 patients, 13 (81%) had at least a -0.5 log10 change in viral titer at week 12.

The median baseline HIV-1 RNA level was 4.9 log10 copies/ml. Figures 3 and 4 illustrate individual and group responses to treatment with combination ddI and ribavirin, respectively. Four patients had RNA levels below the lower level of detection of 1,000 copies/ml at weeks 4 and 12; these values were left-censored at 1,000 copies/ml. The median decline in viral RNA was 0.68 log10 copies/ml at week 4 (p = 0.0006) and 0.67 log10 copies/ml at week 12 (p = 0.0048).

Longitudinal changes in the HIV-1 viral load and the time point at which the L74V mutation emerged for each patient are shown in Fig. 3. Figure 4 illustrates overall changes in plasma HIV-1 RNA and HIV quantitative microculture (cell-associated viremia). Quantitative HIV-1 microculture data were available on all 10 patients who elected to complete 24 weeks of combination treatment; no plasma HIV-1 RNA determinations were performed on these visits. Cell-associated viremia appeared to remain suppressed through week 20, but began to return toward baseline at week 24 (Fig. 4).


CD4+ cell counts were available for 16 patients through week 12. CD4+ cell counts increased during the ddI-monotherapy phase of the study (p = 0.0038) but returned toward baseline during the combination-therapy phase with cell counts that did not significantly differ from baseline at week 12 (p = 0.64). However, the study was not well powered with respect to this measure. The study had only 80% power to detect a 65 CD4+ cell rise from baseline CD4+ cell count with a one-sided α level of 0.05. The median increase was +26 cells/mm3 at week 4 and +11 cells/mm3 at week 12. Among the 10 patients who continued treatment through week 24, CD4+ cell counts remained elevated above baseline throughout the study; the median increase was +10 cells/mm3 at week 24 (Fig. 4).

A decline in β2-microglobulin was observed at week 12 (median change = -0.41 mg/L, n = 13, p = 0.0012) but not at week 4 (median change = -0.13 mg/L, n = 12, p = 0.92). No changes in neopterin were detected at weeks 4 or 12 (data not shown).

Exploratory Virologic and Immunologic Analyses

In addition to the planned analyses of safety and activity measures, the interrelationships between pharmacokinetics and immunologic and virologic activity measures were explored. No relationship between changes in HIV RNA and pharmacokinetic parameters were noted; however, a suggestive relationship between the change in cell-associated viremia log10 titer and plasma ddI AUC was detected at week 4 (p = 0.02, r = -0.60) and week 12 (p = 0.13, r = -0.39); that is, higher plasma ddI AUC correlated with greater suppression in cell-associated HIV-1 virus load on treatment.

Table 3 shows the relationship between the presence of SI HIV-1 variants at baseline and change in viral load and CD4+ cell counts. While the baseline log10 titer was similar for patients entering the study with both non-SI and SI HIV-1 strains, subjects with non-SI HIV-1 at baseline demonstrated larger declines in virus load as measured by log10 titer (p = 0.0047) of HIV-1 RNA and larger increases in CD4+ cell count (p = 0.12) over the first 12 weeks of treatment.

The leucine → valine substitution at codon 74 (L74V) in the HIV-1 reverse-transcriptase gene confers an approximately four- to tenfold decrease in susceptibility to ddI (31). Virus isolates from the 15 patients who completed 12 weeks of the study were screened for the emergence of the L74V resistance-conferring mutation: none of them had the L74V mutation at baseline, but eight developed the mutation at some point during the study. In six of these eight patients, the appearance of the L74V mutation preceded the rebound in virus load (Fig. 3). Table 3 summarizes the changes in cell-associated viremia and CD4+ cell count stratified by the development of the L74V mutation during the study period. Although the baseline virus load was higher among patients who developed the L74V mutation, the response to treatment after 12 weeks of therapy was similar for both groups (Table 3).


In patients with HIV-1 disease, the combination of ddI and ribavirin was safe and well tolerated. Coadministration of ddI and ribavirin was associated with a significant decline in viral load over the 12-week study period; this was a single-arm study, however, and it was not possible to compare the effect of the combination treatment relative to ddI monotherapy. There was no change in the overall exposure to ddI as estimated by the plasma ddI AUC before and after the addition of ribavirin to the therapeutic regimen. Previous pharmacokinetic studies with ribavirin at a dose of 800 mg/day (400 mg b.i.d.) resulted in mean trough serum levels at steady state ranging from 6.6 ± 0.5 to 11.1 ± 2.3 μM (21). The patients in our study received a lower daily dose of ribavirin such that a proportionately lower range of trough levels at steady state was expected. This is consistent with our observation of trough ribavirin levels ranging from 3.56 ± 1.3 and 4.58 ± 1.31 μM.

Exploratory analysis revealed that greater reductions in virus load were seen in patients with greater plasma ddI AUCs. This confirms results from a previous study which has shown that viral suppression (as measured by suppression of HIV p24 antigenemia) is related to both single dose and cumulative exposure to ddI (23).

Adverse events were uncommon during the study. The observed toxicity rate was 22% during the first 12 weeks of treatment, which compares favorably with other studies of ddI monotherapy (17,22). Pancreatitis and peripheral neuropathy are the most common and severe adverse effects of ddI treatment. The mechanisms leading to these toxicities are unknown. Nevertheless, at the outset, the potential for an increased rate of toxicity was recognized, and cases were followed closely for evidence if incipient pancreatitis or peripheral neuropathy. Two patients developed elevations of lipase and/or amylase at study weeks 16 and 17, respectively. Both patients had higher baseline amylase at entry into the study. These patients did not develop clinical pancreatitis; the laboratory abnormalities resolved with discontinuation of the study medications. None of the patients developed peripheral neuropathy during the course of 24 weeks on treatment.

Hematologic parameters were closely monitored during the study. Anemia is common among patients treated with >1,000 mg of ribavirin per day. At the ribavirin dose used in this study, no cases of anemia occurred. One patient entered the study with HIV-1-associated thrombocytopenia that had initially responded to ZDV therapy in the past but it had recurred prior to study entry. The thrombocytopenia neither improved nor worsened on the combination regimen. The most severe hematologic adverse event was one case of neutropenia. Although we cannot be certain that the study medications did not contribute to the adverse event, this occurred in a patient with a preentry diagnosis of granulocytopenia (in the setting of antineutrophil antibodies) who was treated chronically with granulocyte colony-stimulating factor (G-CSF) at study entry and who self-discontinued G-CSF just prior to the episode of neutropenia. The patient's neutropenia resolved with reinstitution of G-CSF and discontinuation of the study medications. No adverse sequelae secondary to the neutropenic episode were observed.

Our phase-I trial had no ddI-monotherapy arm, so we cannot draw any conclusion about the activity of ddI alone compared with the ddI-ribavirin combination. At baseline, our cohort had a median CD4+ cell count and plasma HIV-1 RNA level of 167 and 4.9 log10 copies/mL, respectively; median declines from baseline in log10 RNA were 0.68 log10 copies/ml at week 4 and 0.67 log10 copies/ml at week 12, both of which were highly statistically significant. In addition, cell-associated viremia as determined by HIV-1 quantitative microculture also showed sustained suppression relative to baseline at week 12 of -0.92 log10 virus titer. Although the study did not show significant increases in CD4+ cell count between baseline and week 12, we had only 80% power to detect a 65 CD4+ cell rise from baseline with a one-sided α level of 0.05. Significant declines were observed for virologic measures, but the addition of ribavirin at week 4 did not result in further reductions of viremia in the following weeks on study treatment. Within patients, the longitudinal profiles of plasma HIV-1 RNA and cell-associated viremia were similar, although the magnitude of the decreases in virus load were greater for cell-associated viremia compared with plasma HIV-1 RNA. Data on changes in HIV-1 RNA following initiation of ddI monotherapy are available for one phase-II study (ACTG 143) and a phase-III study (ACTG 116A) (35). Although the data from these studies offer greater insight into the virologic and immunologic responses following ddI monotherapy, results across studies must be interpreted with caution.

ACTG 143 was a phase-II study comparing ddI monotherapy (250 mg p.o. b.i.d.) with combination therapy with ddI plus ZDV. All of the patients in ACTG 143 had CD4+ cell counts ranging 200-500/mm3 and constitute a somewhat healthier population than our study. The median reduction from baseline in log10 titer quantitative HIV-1 culture was ≈1 log at week 12 of ddI monotherapy, and 61% of patients (14 of 23) had a decrease of least 0.5 log IUPM at week 12 (M. Ragni, personal communication). In our study, we found a similar degree of HIV-1 cell-associated viremia suppression, but a greater proportion of the patients on ddI and ribavirin had at least a 0.5-log IUPM decrease at week 12 [13 (81%) of 16].

ACTG 116A was a phase-III study comparing ddI with ZDV after little or no previous ZDV treatment. The baseline median CD4+ cell count and RNA level were 139/μl and 5.0-5.3 log10 copies/ml, respectively. Among those who were treated with ZDV for <16 weeks and randomized to ddI, a decline from baseline of 0.2 log10 copies/ml of HIV-1 RNA was observed at week 4, followed by a rebound to +0.15 log10 copies/ml above baseline values at week 12. Among patients never treated with antiretroviral agents and randomized to ddI, a decrease from baseline of 0.4 log10 copies/ml was observed at week 4 and 0.35 log10 copies/ml at week 12. The virologic response after 4 weeks in our study during the ddI-monotherapy phase indicates potential noncomparability of the study populations. The differences between the HIV-1 RNA data from the ACTG-116A study and our study may be due to differences between the two studies in specimen processing, handling, and storage interval before analysis.

The L74V ddI resistance-conferring reverse-transcriptase mutation emerged in a majority of the patients (53%) completing 12 weeks of treatment. In addition, the L74V mutation appeared before a rise in virus load in most instances. Finally, our results show that patients with non-SI HIV-1 demonstrated greater responses to treatment with larger decreases in virus load and greater increases in CD4+ cell count; the significance of these findings is uncertain because our sample size is limited and pretreatment CD4+ cell counts were higher among patients with non-SI HIV-1.

In summary, combination ddI plus ribavirin in HIV-1-positive patients is safe, well tolerated, without adverse pharmacologic interaction, and associated with significant and sustained declines in virus load over at least 12 weeks of therapy.

Acknowledgment: This work was supported by AI-27659, AI-32766, AI-01101, AI-32782, AI-27661, and AI-95030, and supported in part by the AIDS Clinical Trials Group of the National Institute of Allergy and Infectious Diseases and by the Harvard-Thorndike General Clinical Research Center Unit at Beth Israel Hospital funded by the National Center for Research Resources. The authors express their appreciation to the following: Pat Kasdan and Tom Nevin from the ACTG Operations Office and DAIDS for administrative and logistical support; Karen McLaughlin, Beryl Chapman, Mary Ann Lee, and David Ives at the Beth Israel Hospital; Linda Ecto at Beth Israel Hospital for performance of quantitative HIV-1 cultures and nested polymerase chain reaction studies; Alan Moses, Donna Martin, and Kristine Jordan in the Harvard-Thorndike GCRC; Geraldina Fuentes for assisting with the pharmacokinetics data analysis at Tulane University School of Medicine; Debbie Greenspan at Tulane University School of Medicine; Scott A. Smith for clinical nursing support at the University of Minnesota ACTU; Susan Stehn, Clinical Immunology Research Laboratory at UCLA, for measurement of β2-microglobulin and neopterin levels; Robin B. Garner, Yale University and the VA Connecticut Health Care System, for HIV-1 RNA quantification in plasma; Michael F. Cronin of Organon Teknika Corporation (Durham, NC) for providing support for NASBA kits for HIV-1 RNA assays; and the individuals enrolled in this study, without whom this work could not have been accomplished.

FIG. 1
FIG. 1:
. Phosphorylation of ddI to ddATP. A: Role of IMP in ddIno phosphorylation. 1, Adenosine deaminase; 2, high-Km 5'-nucleotidase (phosphotransferase); 3, adenylosuccinate synthetase; 4, adenylosuccinate lyase; 5, 2'-deoxycytidine kinase; 6, purine nucleoside phosphorylase; 7, hypoxanthine-guanine phosphoribosyl transferase; 8, IMPD; 9, GMP synthetase; 10, xanthine oxidase; 11, pathway for purine biosynthesis de novo. PRPP, 5-phosphoribosyl-1-pyrophosphate; Hx, hypoxanthine. B: Pathways of ddAdo and ddIno metabolism. 1-10, See A. ddAdo and ddIno are also susceptible to nonenzymic hydrolytic cleavage at acid pH. ddRP, 2,3-dideoxyribose-5-phosphate. From Hartman et al. (10) with permission by Molecular Pharmacology.
FIG. 2
FIG. 2:
. Mean concentration (μmol/L)-versus-time curves of ddI in plasma following oral administration of ddI alone for 4 weeks (solid squares) and after coadministration with ribavirin at study week 6 (open circles) and at study week 12 (solid triangles). Data points are ± SEM (n = 17 for weeks 4 and 6; and n = 16 for week 12).
FIG. 3
FIG. 3:
. Individual virologic changes for log10 HIV-1 titer (solid lines) in infectious units per million patient peripheral blood mononuclear cells (PBMCs) and log10 plasma HIV-1 RNA levels (dotted lines) in copies/ml. The vertical dashed line indicates time of last study medication doses. W, wild type at codon 74; and M, Leu → Val mutation at codon 74.
FIG. 4
FIG. 4:
. Median relative changes of CD4+ cell count (cells/mm3), HIV-1 RNA (copies/ml of plasma), and quantitative microculture (QM, in infectious units per million) from baseline at weeks on study treatment. Vertical dashed line separates official study period from the voluntary extension phase; data beyond study week 12 includes only the 10 patients who continued into this portion of the study.


1. Mitsuya H, Broder S. Inhibition of the in vitro infectivity and cytopathic effect of human T-lymphotropic virus type III by 2',3'-dideoxynucleotides. Proc Natl Acad Sci USA 1986;83:1911-5.
2. Perno C-F, Yarchoan R, Cooney D, et al. Inhibition of human immunodeficiency virus replication in fresh and cultured human peripheral blood monocytes/macrophages by azidothymidine and related 2',3'-dideoxynucleosides. J Exp Med 1988;168:1111-25.
3. Ahluwalia G, Cooney D, Mitsuya H, et al. Initial studies on the cellular pharmacology of 2',3'-dideoxyinosine, an inhibitor of HIV infectivity. Biochem Pharmacol 1987;36:3797-800.
4. Jackson JB, Coombs RW, Sannerud K, Rhame FS, Balfour HH. Rapid and sensitive viral culture method for HIV type-1. J Clin Microbiol 1988;26:1416-8.
5. Johnson M, Fridland A. Phosphorylation of 2',3'-dideoxyinosine by cytosolic 5'-nucleotidase of human lymphoid cells. Mol Pharmacol 1989;1989:291-5.
6. Gilbert B, Knight V. Biochemistry and clinical applications of ribavirin. Antimicrob Agents Chemother 1986;30:201-5.
7. Roberts R, Jurica K, Meyer W, Praxton H, Makuch R. A phase 1 study of ribavirin in human immunodeficiency virus-infected patients. J Infect Dis 1990;162:638-42.
8. Roberts RB, Dickinson GM, Heseltine PN, et al. A multicenter clinical trial of oral ribavirin in HIV-infected patients with lymphadenopathy: the Ribavirin-LAS Collaborative Group [see comments]. J Acquir Immune Defic Syndr 1990;3:884-92 [Erratum, J Acquir Immune Defic Syndr 1991;4:450].
9. Crumpacker C, Pearlstein G, Van der Horst C, Valentine F, Spector S, Mills J. A phase one increasing dose trial of ribavirin in patients with AIDS and ARC (ACTG 034) [Abstract]. In: Sixth International Conference on AIDS, San Francisco, CA, 1990. Int AIDS Conf 1990, June 20-23, 6(3):203 (Abst SB 468).
10. Hartman N, Ahluwalia G, Cooney D, et al. Inhibitors of IMP dehydrogenase stimulate the phosphorylation of the anti-HIV nucleosides 2',3'-dideoxyadenosine and 2',3'-didoxyinosine. Mol Pharmacol 1991;40:118-24.
11. Streeter D, Witkowski J, Khare G. Mechanism of action of 1-β-D-ribofuranosyl-1,2,4-triazole-3-carboxamide (virazole), a new broad spectrum antiviral agent. Proc Natl Acad Sci USA 1973;70:1174-8.
12. Baba M, Pauwels R, Balzarini J, Herdewijn P, De Clercq E, Desmyter J. Ribavirin antagonizes inhibitory effect of pyrimidine 2',3'-dideoxynucleosides but enhanced inhibitory effects of purine 2',3'-dideoxynucleosides on replication of human immunodeficiency virus in-vitro. Antimicrob Agents Chemother 1987;31:1613-7.
13. Ahluwalia G, Cooney D, Bondoc L, et al. Inhibitors of IMP dehydrogenase stimulate the phosphorylation of the antiviral nucleoside 2',3'-dideoxyguanosine. Biochem Biophys Res Commun 1990;171:1297-303.
14. Balzarini J, Naesens L, De Clercq E. Potentiating effect of ribavirin on the in-vitro and in-vivo antiretrovirus activities of 2',3'-dideoxyinosine and 2',3'-dideoxy-2,6-diaminopurine riboside. J Acquir Immune Defic Syndr 1990;3:1140-7.
15. Johns DG, Ahluwalia GS, Counet DA, Mitsuya H, Driscoll JS. Enhanced stimulation by ribavirin of the 5' phosphorylation and anti-human immunodeficiency virus activity of purine 2'-β-fluoro-2',3'-dideoxynucleosides. Mol Pharmacol 1993;44:517-23.
16. Japour A, Chatis P, Eigenrauch H, Kim L, Crumpacker C. DDI potentiation by ribavirin. In: Seventh International Conference on AIDS, Florence, Italy, 1991; vol 2. Int AIDS Conf 1991, June 16-21, 7(2):97 (Abst. WA1021).
17. Kahn JO, Lagakos SW, Richman DD, et al. A controlled trial comparing continued zidovudine with didanosine in human immunodeficiency virus infection. N Engl J Med 1992;327:581-7.
18. Abrams DI, Goldman AI, Launer C, et al. A comparative trial of didanosine or zalcitabine after treatment with zidovudine in patients with human immunodeficiency virus infection: the Terry Beirn Community Programs for Clinical Research on AIDS [see comments]. N Engl J Med 1994;330:657-62.
19. Dolin R, Amato DA, Fischl MA, et al. Zidovudine compared with didanosine in patients with advanced HIV type 1 infection and little or no previous experience with zidovudine: AIDS Clinical Trials Group. Arch Intern Med 1995;155:961-74.
20. Crumpacker C, Heagy W, Bubley G, et al. Ribavirin treatment of AIDS and ARC: a phase I study shows transient clinical improvement associated with suppression of HIV and enhanced lymphocyte proliferation. Ann Intern Med 1987;107:664-74.
21. Lertora J, Rege A, Lacour J, et al. Pharmacokinetics and long-term tolerance to ribavirin in asymptomatic patients with human immunodeficiency virus. Clin Pharmacol Ther 1991;50:442-9.
22. Lambert JS, Seidlin M, Reichman RC, et al. 2',3'-Dideoxyinosine (ddI) in patients with the AIDS or ARC. N Engl J Med 1990;322:1333-40.
23. Drusano GL, Yuen GJ, Lambert JS, Seidlin M, Dolin R, Valentine FT. Relationship between dideoxyinosine exposure, CD4 counts, and p24 antigen levels in human immunodeficiency virus infection: a phase I trial. Ann Intern Med 1992;116:562-6.
24. Beltangady M, Knupp CA, Gustafson N, et al. Relation between plasma concentrations of didanosine and markers of antiviral efficacy in adults with AIDS or AIDS-related complex. Clin Infect Dis 1993;16(suppl 1):S26-31.
25. Dimitrov D, Melnick J, Hollinger F. Microculture assay for isolation of human immunodeficiency virus type 1 for titration of infected peripheral blood mononuclear cells. J Clin Microbiol 1990;28:734-7.
26. Japour A, Fiscus S, Arduino J, Mayers D, Reichelderfer P, Kuritzkes D. Standardized microtiter assay for determination of syncytium-inducing phenotypes of clinical human immunodeficiency virus type 1 isolates. J Clin Microbiol 1994;32:2291-4.
27. Myers L, McQuay L, Hollinger F. Dilution assay statistics. J Clin Microbiol 1994;32:732-9.
28. Van Gemen B, Kievets T, Schukkink R, et al. Quantification of HIV-1 RNA in plasma using NASBA during HIV-1 primary infection. J Virol Methods 1993;43:177-88.
29. Van Gemen B, Van Beuningen R, Nabbe M, et al. A one-tube quantitative HIV-1 RNA NASBA nucleic acid amplification using electrochemiluminescent (ECL) labelled probes. J Virol Methods 1994;49:157-68.
30. Boom R, Sol C, Salimans M, Jansen C, Wertheim-Van Dillen P, Van Der Noordaa J. Rapid and simple method for purification of nucleic acids. J Clin Microbiol 1990;28:495-503.
31. St. Clair M, Martin J, Tudor-Williams G, et al. Resistance to ddI and sensitivity to AZT induced by a mutation in HIV-1 reverse transcriptase. Science 1991;253:1557-9.
32. Kalin J, Hill D. Determination of 2',3'-dideoxyadenosine, 2',3'-dideoxyinosine, and 2',3'-dideoxycytidine in biologic samples. J Chromatogr 1988;431:184-91.
33. Austin R, Trefts P, Hintz M, Connor J, Kagnoff M. Sensitive radioimmunoassay for the broad-spectrum antiviral agent ribavirin. Antimicrob Agents Chemother 1983;24:696-701.
34. Weerahandi S. Exact statistical methods for data analysis. New York: Springer-Verlag, 1995.
35. Welles S, Jackson J, Yen-Lieberman B, et al. Prognostic value of plasma HIV-1 RNA levels in patients with advanced HIV-1 disease and with little or no zidovudine therapy. J Infect Dis 1995 (in press).

Didanosine; HIV-1; Phase-I study; Ribavirin

© Lippincott-Raven Publishers.