Secondary Logo

Journal Logo

Clinical Science

Daily Low-Dose Subcutaneous Interleukin-2 Added To Single- or Dual-Nucleoside Therapy in HIV Infection Does Not Protect Against CD4+ T-Cell Decline or Improve Other Indices of Immune Function: Results of a Randomized Controlled Clinical Trial (ACTG 248)

Vogler, Mary A. MD*; Teppler, Hedy MD; Gelman, Rebecca PhD; Valentine, Fred MD*; Lederman, Michael M. MD§§; Pomerantz, Roger J. MD; Pollard, Richard B. MD**; Cherng, Deborah Weng MS; Gonzalez, Charles J. MD*; Squires, Kathleen E. MD§; Frank, Ian MD‡‡; Mildvan, Donna MD††; Mahon, Laura F. BS¶¶; Schock, Barbara MS***the AIDS Clinical Trials Group 248 Study Team

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: May 1st, 2004 - Volume 36 - Issue 1 - p 576-587


Interleukin-2 (IL-2) is a cytokine produced by activated T lymphocytes that has immunomodulatory action and is under intense scrutiny as interest in immunologic therapy for HIV disease has reemerged. 1–3 IL-2 has been shown to be effective at increasing the CD4+ T-cell count when used as adjunctive therapy in the treatment of chronic HIV infection. Early trials of intermittent (5 days every 8 weeks) high-dose intravenous (IV) IL-2 administered with oral nucleoside therapy demonstrated progressive and sustained increases in CD4+ T-cell count with each subsequent cycle. 4,5 The ability of IL-2 to successively increase CD4+ T-cell counts was attributed to the use of an intermittent schedule of administration, but despite an intermittent schedule, significant treatment-associated morbidity was also observed, and substantial dose reductions were required. Immunologic benefit was greatest in those with higher CD4+ T-cell counts and lower viral burden at study entry, whereas adverse effects were significantly more frequent and severe in patients with lower CD4+ T-cell counts. Toxicities associated with higher doses of IL-2 and the lack of data demonstrating clinical efficacy have hindered the widespread application of IL-2 in the treatment of HIV disease. Since the publication of the original trials of IL-2, new potent antiretroviral therapy has changed the outcome for HIV disease in the developed world, yet well-tolerated and potent therapies to enhance immunologic function remain elusive. Clinical endpoint evaluations of intermittently administered IL-2 are in progress.

The present study was done to evaluate the potential for a substantially lower subcutaneous dose of IL-2 administered daily in conjunction with standard (at the time of the trial) nucleoside therapy to prevent deterioration of immune function in subjects with relatively normal CD4+ T-cell counts. The primary endpoint of the study was prevention of CD4+ T-cell loss over time. Secondary endpoints included evaluation of toxicity, safety, viral load changes, serologic responsiveness to new or boosting doses of vaccines, skin test reactivity to recall antigens, and laboratory measures of specific immunologic responses.


Study Design

The study was designed as a randomized, controlled, open-label comparison of antiretroviral therapy (ART) alone vs. ART plus daily low-dose IL-2. Low-dose (1 million IU) recombinant human IL-2 was administered daily by subcutaneous injection in conjunction with standard-of-care ART. At the time this study was initiated, generally, single or dual nucleosides were standard ART.

Patients were randomly allocated between arms for a 6-month (24-week) period (step 1) and then were to be crossed over to the alternate treatment arm (step 2). This report only includes a comparison of subjects through step 1, the original 6 months of randomized treatment with ART alone vs. ART + IL-2. During this period, no changes in ART drugs were allowed. This report does not include an analysis of the crossover data, because patients were allowed to change ART regimens during an optional 2-month period between the end of step 1 and the beginning of step 2; however, during the crossover period for most of the enrolled patients, 3-drug therapy including a protease inhibitor became the standard of care, so comparisons between steps 1 and 2 were difficult to interpret.

Adults ≥18 years of age with confirmed HIV infection by positive serology, HIV antigen, or viral culture were eligible if they met the following criteria: a screening CD4+ T-cell count between 300 and 700 cells/mm3; no history of AIDS-defining clinical conditions by the 1993 Centers for Disease Control case definition (except limited cutaneous Kaposi sarcoma); and no IL-2 treatment within the prior 3 months. Subjects were required to be receiving stable, Food and Drug Administration–approved or parallel-track ART using the same drug or combination of drugs for at least 2 consecutive months prior to study entry. Choice of ART regimens was at the discretion of the patient’s prescribing physician. Patients must have had no active opportunistic infection, no immunomodulating drugs including systemic cytotoxic chemotherapy or systemic corticosteroids within 4 weeks prior to study entry, no active cardiac disease, no untreated thyroid disease, no systemic malignancies, no asthma, and no autoimmune disease. Pregnant or breast-feeding women were also excluded. Data on nadir CD4+ T-cell counts prior to screening were not collected.

All patients provided written informed consent as approved through the institutional review board at each participating center.

Subjects were stratified by type of ART (single vs. dual nucleoside), duration of ART (2–6 vs. >6 months) at time of study entry, and whether or not they participated in an immunologic substudy [ACTG 819, which measured lymphocyte responses by lymphocyte proliferation assay (LPA), and by natural killer (NK) cell and lymphokine-activated killer (LAK) cell activities] for a total of 2 × 2 × 2 = 8 strata.

Patients were randomly assigned to receive either continued ART alone or open-label treatment with 1 million IU daily of IL-2 self-administered subcutaneously in 0.2-mL volume at rotating skin sites in combination with continued ART for 24 weeks. Subjects were trained to inject themselves and were provided with prefilled syringes containing IL-2 at the appropriate dose. Dose reductions of IL-2 were allowed for any toxicity the subjects found intolerable and were allowed in 2 levels: first to 500,000 IU once daily and then to 250,000 IU once daily. ART was not supplied by the study as patients were expected to be on their choice of stable ART therapy at study entry and were not allowed to change ART regimen during the study period.

At the time of randomization it was determined whether the patient was eligible to receive any of 4 vaccines: tetanus toxoid (Td), meningococcal polysaccharide (Menomune-A/C/Y/W-135), influenza (Fluzone 1995–1996 season; Connaught Laboratories, Swiftwater, PA), or polyvalent pneumococcal (Pneumovax 23; Merck Research Laboratories, West Point, PA). Patients were eligible if they had not been recently immunized and had no history of allergy to vaccine preservatives or components. Vaccines were administered to eligible patients in both arms of the study 4 weeks after enrollment, with follow-up serologies drawn done at weeks 8 and 24, with the exception that pneumococcal vaccine was given between weeks 8 and 20 to patients who had not received it within the prior year, and follow-up serologies were drawn done 4 weeks later.

Patients had skin tests placed for measurement of delayed-type hypersensitivity (DTH) responses at baseline and at week 16 with a panel of 0.1-mL intradermal injections of each of the following skin test antigens: Tuberculin PPD (Connaught Laboratories), 5TU; Candida albicans skin test antigen (CASTA) (Greer Laboratories), at 2 doses, 1 and 10 CU; tetanus toxoid fluid, adsorbed (TT) (Connaught), 0.1 mL of 1:10 vol/vol dilution; and mumps skin test antigen (MSTA) (Connaught), 4 CFU. Skin test reactions were read at 48 hours, using a ballpoint pen technique. The pretreatment skin tests were read before the first dose of IL-2 was given.

Flow cytometry was used to measure CD4+ T cells once at screening for initial eligibility, twice at preentry, and once at weeks 2, 4, 8, 16, and 24. The screening CD4+ T-cell count was required to be between 300 and 700 cells/mm3 but was not used in the analysis. The 2 preentry CD4+ T-cell counts were averaged to obtain a baseline value. Serum drawn at entry and at various follow-up time points was frozen and stored for subsequent measurement of Td, meningococcal, influenza, and pneumococcal serologies, and for measurement of soluble tumor necrosis factor receptor II (sTNF-RII), β2-microglobulin, and neopterin. Plasma samples were frozen and stored for HIV-1 RNA by bDNA, and then all were measured at the same time.

Patients were also offered enrollment in an intensive immunologic substudy to evaluate cellular immune responses, including NK and LAK cell cytotoxicity, as well as lymphocyte proliferative responses to recall antigens and to HIV-specific antigens. Separate written informed consent was obtained from participants for the substudy.

Statistical Analyses

CD4+ T Cells and Viral Loads

The study was designed primarily to examine effects of subcutaneous low-dose IL-2 on the prevention of CD4+ T-cell count decline, on virologic safety, and on clinical safety and tolerability. It was planned to enroll at least 52 patients per arm to have 80% power to detect a treatment difference if the true percentages of patients with a <25% drop in CD4+ cell count were 31% on ART alone and 9% on ART + IL-2. (This calculation assumes that 10% of patients might not be used in the analysis, either because they never got study treatment or because they never had a follow-up CD4+ T-cell count.) One of the protocol-specified secondary analyses was a longitudinal analysis of slope of CD4+ T-cell counts during step 1; the protocol was designed to have 80% power to detect between treatment arms a true difference in slopes at least as big as 79 cells/mm3 /y. Based on the earlier version of the bDNA assay, the study had sufficient power to detect a difference between treatment arms in median change in viral load of 1 log.


The protocol specified that an interim analysis of toxicity be done after 54 patients had been followed for a minimum of 6 months and again at the end of the study; grade 3 (severe) or worse toxicity was compared between arms using a 2-sided Fisher exact test not adjusting for the interim analysis or for multiple types of toxicities. With the planned sample size, each toxicity comparison at the end of the study had 80% power to detect differences between study arms if the true percents of grade 3 or worse toxicity were as different as 10% and 31%.

Comparisons of categorical variables (e.g., percent of patients who had at least a 25% decrease in CD4+ T cells or at least a 25% increase in CD4+ T cells or a <25% change) all used a Fisher exact test. Power calculations for these tests were based on the exact binomial calculations. Comparisons of time to event (i.e., time to first dose reduction or discontinuation, to first drop of CD4 count by at least 25%) all used the log rank text. Comparisons of continuous variables (i.e., changes at various follow-up times in CD4 count, log10 plasma HIV RNA, β2-microglobulin, neopterin, sTNF-RII, vaccine serologies, lymphocyte proliferation responses, LAK and NK cytotoxicity) used the Wilcoxon rank sum test. Power calculations for these tests were based the t test and on the fact that the asymptotic relative efficiency of the Wilcoxon test to the t test is no worse than 0.86 for any distribution. Comparison of treatments with respect to variance of change of CD4+ T-cell counts used an F test and comparisons of slopes of change of CD4+ T-cell counts used a random effects linear model.


Baseline Data

Between October 1995 and June 1997, a total of 115 patients were enrolled at 14 sites. Of these, 58 were randomly assigned to receive ART alone, and 57 were randomly assigned to receive ART + IL-2. One of the ART-alone patients was declared ineligible before randomization because he did not want to stop regular allergy injections, but he was randomized in error. His data are excluded from all analyses. There were no large differences in baseline characteristics between treatment arms (Table 1).

Patient Characteristics at Baseline Percents Are Used for Categories; Numbers in Parentheses

Treatment Administration

Sixty-one percent of the patients had ART consisting of combinations of drugs, and 39% had ART consisting of a single drug. The most common types of ART were zidovudine (ZDV) (21%), ZDV + zalcitabine (ddC) (15%), ZDV + lamivudine (3TC) (12%), and ZDV + 3TC + a protease inhibitor (11%); other less frequent types of ART were didanosine (ddI) alone, ddC alone, stavudine (d4T) alone, ZDV + ddI, and d4T + 3TC.

Of the patients randomly allocated to receive continued ART alone, 54 of 57 (95%) completed 24 weeks on treatment. Three ART-alone patients stopped treatment between 12 and 22 weeks after randomization (1 was lost to follow-up, 1 patient refused to continue, and 1 left to change ART drugs). Of the patients randomly assigned to receive ART + IL-2, 48 of 57 (84%) completed 24 weeks on treatment. Nine ART + IL-2 patients stopped treatment between 6 and 18 weeks after randomization (1 because of physician request, 1 to change ART drugs, 1 due to toxicity, and 6 patients refused to continue IL-2 therapy).

All patients are included in the analysis, but CD4+ T-cell counts and other laboratory assay results were not obtained after a patient stopped treatment (so in the following sections, “last CD4+ T-cell count” should be interpreted as the last CD4+ T-cell count while the patient was still on step 1 treatment). Treatment discontinuation seemed to be unrelated to CD4+ T-cell count at the time of treatment discontinuation. (For the 9 ART + IL-2 patients, the last CD4 count on therapy was an increase over baseline for 4 patients and a decrease from baseline for 5 patients. This 50/50 split between increases and decreases held up at each time point for which these patients had data. Of the 3 ART-alone patients who discontinued treatment early, only 2 had baseline CD4+ T-cell counts recorded; the last CD4 count in these 2 represented an increase but at earlier time points CD4+ T-cell counts were decreased from baseline values.)

Doses Reduced or Missed

Dose reduction or discontinuation of either ART or IL-2 was allowed in the protocol on both treatment arms (as described in “Methods”). Those patients randomly assigned to ART + IL-2 who then discontinued their ART were required to discontinue IL-2 for the same period; however, patients who discontinued IL-2 or reduced their IL-2 dose were not required to interrupt or decrease the doses of their ART.

Five patients had a first level reduction of IL-2 dose to 500,000 IU once daily, and 1 patient had a second level reduction to 250,000 IU once daily. The level of adherence to IL-2 injections was good, with 6 patients missing 1 day of the 24 weeks of IL-2, 1 patient missing 3 days, and 1 patient missing 5 days. Ten other patients missed ≥8 days (range 8–35 days) of IL-2 because of toxicity in the first 24 weeks of treatment; almost all had additional days of drug after week 24 to make up for missed doses. Adherence to ART therapy was also good. One patient on ART alone had ART doses reduced because the patient requested it and one on ART + IL-2 had ART doses reduced because of nausea; one patient on ART alone missed 14 days of ART because of patient request and one on ART + IL-2 missed 9 days of treatment while hospitalized. No significant difference between treatment arms was seen in time to first dose reduction or discontinuation of either IL-2 or ART.

Changes in CD4+ T-Cell Counts and Percents

The mean changes in CD4+ T-cell count from baseline by study week and treatment arm are shown in Figure 1 and Table 2. A statistically significant treatment difference in change in CD4+ T-cell counts from baseline was seen at week 4 but was not seen at any other time point. Table 3 also shows summary statistics and tests of treatment differences in change of CD4+ percents; a significant difference was seen at weeks 4, 8, 16, and 24 but not at last follow-up. The tests for treatment differences assume equal variances, but in fact the variances were significantly different (e.g., for change from baseline to last CD4 assay, ratio of variances 1.8 with P = 0.04 for CD4+ T-cell counts and ratio of variances 2.6 with P = 0.0007 for CD4+ percents.). This increased splay in CD4+ T-cell count in the IL-2–treated group is also illustrated in the histograms in Figure 2.

CD4+ T Cells in Counts and Percents, log10 Plasma HIV RNA by bDNA
Serologic Responses to Vaccination
Mean and 95% confidence interval (CI) of mean for CD4 count change from baseline.
Time to first drop of CD4 count by at least 25%.

Because the primary objective of this study was to examine the effect of IL-2 on preserving CD4+ T-cell count in this relatively immunologically healthy group of patients, rather than to examine the effect of IL-2 on increasing CD4+ T-cell counts, the protocol-specified primary analysis was based on differences by treatment arm in the number of patients who ever had a CD4+ T-cell count <75% of baseline (therefore equal to a ≥25% drop) over the first 24 weeks. Three patients (2 on ART alone and 1 on ART + IL-2) did not have an entry or preentry CD4+ T-cell count and therefore are omitted from this analysis. Of the 55 patients on ART alone with baseline and follow-up values, 15 (27%) had a drop of ≥25% in their CD4+ T-cell count at some time over the 24 weeks. Of the 56 patients on ART + IL-2, 23 (41%) had a CD4+ T-cell count drop of ≥25% at some time over the first 24 weeks of treatment; this difference was surprising but not statistically significant (P = 0.16). Additionally, the difference in time to first drop in CD4+ T-cell count by ≥25% is not statistically significant (P = 0.11, Fig. 3).

Time to first increase of CD4 count by at least 25% for first step.

However, a parallel analysis of the number of patients who ever had a CD4+ T-cell count increase over the first 24 weeks (using the definition of at least a 25% increase over baseline since this is analogous to the protocol-defined decrease specified in the primary analysis) revealed differences that were statistically significant. On ART alone, 18 of 55 patients (33%) had a ≥25% increase in CD4+ T-cell count at some time during the first 24 weeks, whereas 39 of 56 of patients (70%) on ART + IL-2 had a CD4+ T-cell count increase of ≥25% over the same period (P = 0.0001). The time to first increase of ≥25% in CD4+ T-cell count was also significantly different between treatment arms (P = 0.001, Fig. 4).

Histograms of change from baseline to last CD4 count.

To compare the changes in both directions in CD4+ T-cell numbers in the 2 treatment groups, we analyzed overall change from baseline to the last CD4+ T-cell count observed, assigning patients into 1 of 3 categories: ≥25% decrease, <25% change in CD4+ T-cell count, and ≥25% increase. For ART + IL-2, the percents of patients in these 3 categories were 18%, 48%, and 34%, respectively. For ART alone, the percents of patients in these 3 categories were 7%, 80%, and 13%, respectively; the difference was significant (P = 0.03).

Stratifying patients by single vs. combination nucleoside-based treatment at entry and then evaluating for CD4+ T-cell count change revealed a similar pattern of greater splay in the CD4+ T-cell count in the ART + IL-2–treated groups (both more increases and more decreases in CD4+ T-cell count) regardless of the type of nucleoside treatment, except in the stratum of patients who had been treated with nucleoside-based combination regimens for only 2–6 months at study entry. An equal distribution of CD4+ T-cell count decreases and increases in patients on both arms of the study was seen in that stratum alone.

Based on the first 24 weeks of the study, the estimated annual slope of CD4+ T-cell count (based on a random effects analysis) for ART-only patients was an increase of 1 cell/mm3 vs. an estimated annual slope of 57 cells/mm3 for ART + IL-2-treated patients. Again, however, this difference was not significant, even when corrected for initial CD4+ T-cell count (P = 0.18). This analysis assumes the between-patient variances of slope in CD4+ T-cell counts are equal in the 2 treatment groups, but as previously discussed, the variance of change in CD4+ T-cell counts in patients treated with ART+IL-2 was about twice that for the patients treated with ART alone, so the slope would also have different variances.

Changes in Plasma HIV-1 RNA

The interim analysis of plasma HIV RNA in the first 20 patients was based on a bDNA assay with lower level of detection of 10,000 copies/mm3 and showed no significant treatment difference. The final analysis of plasma HIV RNA was based on a bDNA assay with lower level of detection of 50 copies/mm3. The differences between treatment arms for changes in plasma HIV RNA from baseline over the study period were not significant at any time point. Table 2 shows mean and median change in log10 plasma HIV RNA by bDNA: for each arm at each follow-up time; the mean change was never >0.07 (corresponding to a 17% increase or 15% decrease in viral load). Although there were no significant differences between the treatments in viral load change, ART + IL-2 was always associated with a more beneficial direction of viral load change (bigger decrease or smaller increase) than ART alone.

Serologic Responses to Vaccine

Comparing serologic titers after immunization in the 2 groups assessed the effect of low-dose IL-2 on enhancing responses to immunization in these immunocompromised patients. Patients eligible for vaccination (see “Methods”) had baseline prevaccination serologies compared with 4 weeks postvaccination and end-of-study serologies. Depending on vaccine type, serologic results were available for analysis in 47–98% of those vaccinated. The protocol did not define serologic response, but for purposes of the analysis, response was defined as a 4-fold rise in titer. Only 8 patients on ART only and 10 on ART + IL-2 had pneumococcal serologies, and none of them had a response; the results of the other serologies are given in Table 3. There were no statistically significant differences between the treatment arms in the proportions of responders to any vaccine antigen except for 1 influenza strain (for which ART alone was associated with a larger response percent). However, the power to detect reasonable treatment differences in these proportions (e.g., 40% vs. 60%) was quite low (6% for pneumococcal, 18% for tetanus and diphtheria, 12% for influenza, and 32% for meningococcal serologies).

Skin Test Responses

Improvements in DTH responses to recall antigens after treatment with IL-2 would be expected with improvement in CD4+ T-cell numbers, function, or both. DTH response was therefore measured in response to PPD, tetanus, mumps, and 2 doses of Candida at study entry and week 16. Only 5 patients ever had a reaction to PPD, so results were not analyzed for this antigen. Depending on which skin test antigen was analyzed, between 4 and 32 patients per arm had negative responses at baseline and also had a skin test result recorded at 16 weeks. Between 2 and 13% of patients with initial negative responses had positive responses at 16 weeks. For tetanus and mumps, only 10 and 8% of patients, respectively (overall), had an increase in induration ≥5 mm between baseline and week 16. For the lower dose of Candida, 30% of ART-only patients and 21% of ART + IL-2 patients had an increase of induration ≥5 mm; for the higher dose of Candida, 26% of ART-only patients and 23% of ART + IL-2 patients had an increase of induration ≥5 mm. There were no significant differences discerned in the frequency of or changes in skin test responses between treatment arms. However, the power of these comparisons was again very low (e.g., in the comparison of conversion rates from a baseline negative skin test response to positive skin test response at 16 weeks, the power to detect a difference between arms that have true conversion rates of 10% and 30% was 41% for tetanus, 30% for mumps, and 6% for Candida).

Serum Activation Markers

Serum activation markers at baseline, prior to initiating IL-2, were similar in the 2 randomized groups, with means of 2.06 and 1.98 mg/L for β2-microglobulin, 9.62 and 8.96 nml/L for neopterin and 3.01 and 3.02 ng/mL for sTNF-RII (for the ART and ART + IL-2 arms, respectively). After starting IL-2, for all 3 activation markers measured, the median and mean changes from baseline were downward at most time points in the ART-only group, and upward in the ART + IL-2 group at all time points (weeks 2, 4, 8, 16, and 24). Using the 2-sided Wilcoxon rank sum test, there were significant differences between the groups at each time point. For all time points, values were P ≤ 0.01 for β2-microglobulin, P &le 0.0009 for neopterin, and P ≤ 0.0001 for sTNF-RII.

Lymphoproliferative Assays

Of the 56 patients entered on the intensive immunologic substudy, 27 patients had baseline lymphoproliferative assays to tetanus and Candida performed. Eighteen patients had baseline lymphoproliferative assays performed to 2 HIV envelope proteins and 9 patients to an HIV p24 antigen labeled Chi p25; fewer had follow-up assays performed. Results were analyzed both by the difference in counts per minute (CPM) in the presence of antigen or medium alone (delta CPM) and by stimulation index (SI). Over the duration of the study there was little change in response to any antigen for either treatment arm. However, the power to detect reasonable differences (e.g., a 3-fold median increase in SI on one arm and a total lack of change in SI on the other arm) was small (53% for tetanus or Candida and 34% for the HIV envelope proteins if standard deviations were about the same size as means).

Natural Killer Cell Assays

Measurements of NK and LAK cell activities were performed on 16 patients per arm at 12 weeks and 13 patients per arm at 24 weeks. There were no significant differences by treatment arm in the difference between follow-up and baseline results of these assays using effector to target ratios from 6.25 to 50 for assays done with peripheral blood mononuclear cells (PBMCs) alone or with PBMCs + IL-2 (although the power to detect an increase of 1000 cells in one arm and 500 cells in the other arm if standard deviations were equal to means was only 38% at 12 weeks and 32% at 24 weeks).


An interim safety analysis performed using data received on 54 patients followed for 6 months revealed no significant differences between treatment arms in the occurrence of grade 3 or worse toxicity, and there was no protocol interruption or modification. By the end of the study at 24 weeks, only 2 grade 4 toxicities had occurred: one grade 4 hypertriglyceridemia in the ART + IL-2 group and one case of agitation in the same group. Table 4 shows percent of patients with grades 2 and 3 toxicities by treatment arm; some of these are associated with ART therapies rather than with IL-2. In comparing grade 2 or worse toxicity, the ART + IL-2 group had significantly more toxicity only for the category “general body,” and that was primarily due to more grade 2 discomfort and fatigue.

Toxicities: Percents of Patients

Clinical Events and Deaths

New clinical diagnoses seen in ≥5 patients while on study included lymphadenopathy, sinusitis or rhinitis, upper respiratory infection or “strep” throat, bronchitis, oral hairy leukoplakia, and herpes simplex. There were no significant differences between the treatment arms in the occurrence of these events. One patient died at week 38 due to sepsis, with invasive aspergillosis and HIV disease progression as contributing causes. He had been assigned to the ART-only treatment arm.


In this prospective, randomized, controlled, multicenter trial, the addition of low-dose subcutaneous daily IL-2 to ART treatment comprising largely single- or dual-nucleoside therapy for 6 months showed no significant benefit in preserving CD4+ T-cell counts in a population of relatively immunologically healthy patients. Although the study was primarily designed to evaluate only preservation of baseline CD4+ T-cell counts, we observed that a significantly larger proportion of patients treated with this schedule of IL-2 had a >25% increase from baseline in CD4+ T-cell counts than did patients receiving ART therapy alone. However, the frequency of individuals experiencing a decrease in CD4 + T-cell counts by ≥25% was nonsignificantly greater in recipients of IL-2. Importantly, the variation in change in CD4+ T cell counts over time was significantly greater in the patients randomly assigned to the IL-2 arm.

ART therapy in these patients was primarily single or dual nucleoside based, which suppressed viral replication to <50 copies/mL in only 11% of patients as assessed by bDNA assay at baseline. Median plasma HIV RNA on ART therapy obtained at baseline by bDNA was 3.63 log10 in the ART-alone arm and 3.41 log10 in the ART + IL-2 arm. Plasma HIV RNA was unchanged by the addition of IL-2, and no baseline characteristic, including plasma HIV RNA at baseline, predicted a >25% drop in CD4+ T-cell counts over the study period.

Although our study addresses some aspects of IL-2 therapy in HIV disease, several questions remain, including the types of immunologic responses generated at different doses and its utility in different patient populations. In patients with lower CD4+ T-cell counts, it is unclear whether IL-2 boosts CD4+ T-cell counts primarily by expansion of the pool of preexisting CD4+ T cells, 6 by increasing thymic generation of naive cells, 7 or by selective expansion and prolonged survival of a unique subset of naive CD4+CD25+ T cells. 8 In addition, increases in CD4+ T-cell numbers in these patients may not restore a diverse repertoire of T-cell clones once their distribution has been perturbed.

Efforts to maximize immunologic activity and to minimize toxicity led to trials of lower doses given daily 9,10 and to trials of subcutaneous or intradermal administration. 11–13 Indeed, some investigators have proposed that lower-dose IL-2 administered daily is more physiologic since lower concentrations of IL-2 will bind only to T cells with high-affinity receptors and not to NK cells with lower-affinity receptors. 14 Lower doses may also limit toxicities associated with activation of cytokine cascades and limit the extent of rebound cytopenias. It has been argued therefore that “less is more” in IL-2 dosing. 15,16 A recently published randomized trial of highly active antiretroviral therapy (HAART) + daily subcutaneous IL-2 at a starting dose of 1.2 mIU/m2 /d in patients with CD4+ T-cell counts <300 cells/mm3 reported significantly greater increases in the number of CD4+ T cells, in the proportion of naive cells, and in the number of NK cells in the IL-2–treated group at weeks 4 and 8 compared with these indices in HAART-alone–treated controls, but this difference was not sustained at weeks 16 and 26 of the study. 9 There were also substantial changes in antiviral therapies during the trial period. Higher IL-2 doses and route and frequency of dosing were addressed in a randomized, controlled clinical trial in France of IL-2 added to background dual-nucleoside therapy, which showed comparable significant improvements in CD4+ T-cell counts and proliferative responses in the intermittent IV and in the high-dose daily subcutaneous groups. 17

Our study does not address the question of whether this schedule of daily low-dose IL-2 administration would benefit patients on potent ART therapy who have substantial virologic suppression with viral loads below the limit of detection. Higher doses of IL-2 administered intermittently in conjunction with daily ART therapy have shown benefit in increasing CD4+ T-cell counts in subjects whose plasma HIV RNA is below the limit of detection. 18,19 For example, Katlama et al. 20 analyzed CD4+ T-cell area under the curve minus baseline in patients with very low CD4+ T-cell counts who were treated with HAART alone vs. HAART + IL-2 administered as 4 cycles of 9 million IU of IL-2 daily for 5 days every 6 weeks and found significant and sustained increases in CD4+ T-cell area under the curve out to 80 weeks.

The population of patients participating in this study was quite healthy from an immunologic standpoint and by current guidelines might not begin ART therapy. “Adjunctive” therapy to further increase CD4+ T-cell counts might not be necessary, but administration of IL-2 might delay the decline in CD4+ T-cell counts as disease progresses and possibly prolong the time before HAART is initiated.

We were not able to detect an effect of this dose of IL-2 on other immunologic endpoints such as skin test responsiveness, serologic responses to the administration of vaccines, lymphoproliferative responses, and assays of LAK and NK cell activity. In the setting of incomplete control of HIV replication, administration of daily low-dose IL-2 in this study provides very little evidence of immune enhancement. Since responsiveness to immunization is often impaired in HIV infection, immune-based strategies to enhance vaccine responses would have important clinical implications. The utility of low-dose daily or twice weekly subcutaneous or intradermal IL-2 as adjunctive therapy to restore or preserve immunologic function in HIV infection was studied previously in a few small, uncontrolled trials. In these studies, enhanced skin test reactivity suggested an immunologic benefit. However, a recent paper suggests that high-dose intermittent IL-2 administered to patients on HAART does not show evidence of enhancing vaccination responses when compared with responses seen in persons receiving HAART alone, despite inducing significant increases in CD4+ T-cell number. 21

In conclusion, this regimen of daily low-dose subcutaneous IL-2 was well tolerated. The observation that IL-2 was associated with both a more frequent increase and a more frequent decrease in CD4+ T-cell numbers than seen in subjects receiving nucleosides alone is difficult to explain. Nonetheless, a substantial proportion of the patients treated with adjunctive low-dose IL-2 compared with those not treated with IL-2 experienced an increase of ≥25% in CD4+ T-cell count from baseline. Our conclusions are supported to some degree by other studies, in which study protocols monitored CD4+ T-cell counts as IL-2 doses were escalated and showed evidence for a dose-response association between IL-2 and CD4+ T-cell counts. 17–19,22 Phase 3 clinical endpoint trials of higher-dose intermittent IL-2 therapy are in progress and the results of these studies should provide a definitive answer about the clinical benefit of the admittedly more toxic but possibly more active regimen of IL-2 administration for the treatment of HIV disease.


Without the patients who so willingly provided blood samples, learned injection techniques, and returned for frequent follow-up visits, this study could not have been done and the authors thank them for their participation.

Critical concept review, manuscript review, and study support were provided by Anne-Marie Duliege, Chiron Corp. Influenza serologies were performed by Alexander Klimov, PhD, Influenza Branch, Centers for Disease Control, Atlanta, GA. Meningococcal serologies were performed by Mike Bybel, Aventis Pharmaceuticals. Pneumococcal serologies were performed by Daniel Musher, Baylor College of Medicine, Veterans Affairs Medical Center, Houston, TX. β2-Microglobulin, neopterin, and sTNF-RII assays were performed by John Fahey (University of California, Los Angeles, School of Medicine). HIV-1 bDNA assays were performed and funded by Chiron Corp. IL-2 (aldesleukin, rhIL-2, Proleukin) was provided by Chiron Corp. Tetanus and diphtheria toxoids, adsorbed (for adult use) (Td), Meningococcal Polysaccharide Vaccine (Menomune-A/C/Y/W-135), and Influenza Virus Vaccine (Fluzone, 1995–1996 season) were provided by Connaught Laboratories. Pneumococcal vaccine, polyvalent (Pneumovax 23) was provided by Merck Research Laboratories.


1. Blankson J, Siliciano R. Interleukin-2 treatment for HIV infection. JAMA. 2000;284:263–268.
2. Mitsuyasu RT. The potential role of interleukin-2 in HIV. AIDS. 2001;suppl 2:S28–S35.
3. Mitsuyasu RT. Immune therapy: non-highly active antiretroviral therapy management of human immunodeficiency virus-infected patients. J Infect Dis. 2002;185(suppl 2):S115–S122.
4. Kovacs J, Baseler M, Dewar R, et al. Increases in CD4 T lymphocytes with intermittent courses of interleukin-2 in patients with human immunodeficiency virus infection: a preliminary study. N Engl J Med. 1995;332:567–575.
5. Kovacs J, Vogel S, Albert J, et al. Controlled trial of interleukin-2 infusions in patients infected with the human immunodeficiency virus. N Engl J Med. 1996;335:1350–1356.
6. Connors M, Kovacs JA, Krevat S, et al. HIV infection induces changes in CD4+ T-cell phenotype and depletions within the CD4+ T-cell repertoire that are not immediately restored by antiviral or immune-based therapies. Nat Med. 1997;3:533.
7. Marchetti G, Meroni L, Varchetta S, et al. Low-dose prolonged intermittent interleukin-2 adjuvant therapy: results of a randomized trial among human immunodeficiency virus-positive patients with advanced immune impairment. J Infect Dis. 2002;606–616.
8. Sereti I, Martinez-Wilson H, Metcalf JA, et al. Long-term effects of intermittent interleukin-2 therapy in patients with HIV infection: characterization of a novel subset of CD4+/CD25+ T cells. Blood. 2002;100:2159–2167.
9. Bartlett JA, Berend C, Petroni GR, et al. Coadministration of zidovudine and interleukin-2 increases absolute CD4 cells in subjects with Walter Reed stage 2 human immunodeficiency virus infection: results of ACTG protocol 042. J Infect Dis. 1998;178:1170–1173.
10. Lalezari JP, Beal JA, Ruane PJ, et al. Low-dose daily subcutaneous interleukin-2 in combination with highly active antiretroviral therapy in HIV+ patients: a randomized controlled trial. HIV Clin Trials. 2000;1:1–5.
11. Jacobson E, Pilaro F, Smith KA. Rational interleukin 2 therapy for HIV positive individuals: daily low doses enhance immune function without toxicity. Proc Natl Acad Sci U S A. 1996;93:10405–10410.
12. Teppler H, Kaplan G, Smith KA, et al. Efficacy of low doses of the polyethylene glycol derivative of interleukin-2 in modulating the immune response of patients with human immunodeficiency virus type 1 infection. J Infect Dis. 1993;167:291.
13. Teppler H, Kaplan G, Smith KA, et al. Prolonged immunostimulatory effect of low-dose polyethylene glycol interleukin-2 in patients with human immunodeficiency virus type1 infection. J Exp Med. 1993;177:483.
14. Smith KA. Interleukin-2: Inception, impact and implications. Science. 1988;240:1169–1176.
15. Smith KA. Lowest dose interleukin-2 immunotherapy. Blood. 1993;81:1414.
16. Smith KA. Low-dose daily interleukin-2 immunotherapy: accelerating immune restoration and expanding HIV-specific T-cell immunity without toxicity. AIDS. 2001;15:(suppl 2):S28–S35.
17. Levy Y, Capitant C, Houhou S, et al. Comparison of subcutaneous and intravenous interleukin-2 in asymptomatic HIV-1 infection: a randomized controlled trial. Lancet. 1999;353:1923–1929.
18. Mitsuyasu R, Pollard R, Gelman R, et al. Prospective randomized controlled phase II study of highly active antiretroviral therapy (HAART) with continuous IV (CIV) or subcutaneous (SC) interleukin-2 (IL-2) in HIV infected patients with CD4+ counts of 50–350 cells/mm3: ACTG 328 final results at 84 weeks. Paper presented at: 8th Conference on Retroviruses and Opportunistic Infections (CROI); 2001. Abstract 17.
19. Arno A, Rui L Juan M, et al. Efficacy of low-dose subcutaneous interleukin-2 to treat advanced human immunodeficiency virus type 1 in persons with <250 CD4 T cells and undetectable plasma virus load. J Infect Dis. 1999;180:56–60.
20. Katlama C, Carcelain G, Duvivier C, et al. Interleukin-2 accelerates CD4 cell reconstitution in HIV-infected patients with severe immunosuppression despite highly active antretroviral therapy: the ILSTIM study—ANRS 082. AIDS. 2002;16:2027–2034.
21. Valdez H, Mitsuyasu R, Landay A, et al. Interleukin-2 increases CD4+ lymphocyte numbers but does not enhance responses to immunization: results of A5046s. J Infect Dis. 2003;187:320–325.
22. Davey R, Murphy R, Graziano F, et al. Immunologic and virologic effects of subcutaneous interleukin-2 in combination with antiretroviral therapy. JAMA. 2000;284:183–189.


ACTG 248 Contributors

Dawn Bell, BSc (NIAID, NIH)

Margarita Vasquez, RN, Victoria Rosenwald, RN, MPH (NYU/Bellevue)

Holly Ingelfinger-Lopez, RN, and Paul Stockdill, RPh (SUNY-Buffalo, Rochester)

David Mushatt, MD (Tulane University)

Michael F. Giordano, MD (Cornell University)

Michael Chance, RN (Case Western Reserve University)

Mitchell Goldman, MD, Kristin Todd, RN, MSN, CCRC (Indiana University)

Charles van der Horst, MD, and Barbara Longmire, RN (UNC) (University of North Carolina)

Virginia A. Waite, BSN (University of Colorado Health Sciences Center, Denver)

Janice Jacovini, RN, and Chris Helker, RN (University of Pennsylvania, Philadelphia)

Janet Pientka and Elizabeth McCann (Thomas Jefferson University Hospital, University of Pennsylvania, Philadelphia)

Cyndi Frank (Yale University)

Author Contributions

Study concept and design: Dr. Teppler, Dr. Pomerantz, Dr. Gelman

Acquisition of data: all participating ACTG sites

Drafting of manuscript: Dr. Vogler, Dr. Gelman

Critical revision of manuscript for important intellectual content: Dr. Lederman, Dr. Valentine, Dr. Teppler, and Dr. Gonzalez

Statistical expertise: Dr. Gelman, Ms. Cherng

Obtained funding: ACTG

Administrative, technical, or material support: Ms. Mahon, Ms. Schock, Ms. Bell

Study supervision: Dr. Teppler, Dr. Vogler, Dr. Pomerantz, Dr. Pollard


HIV infection; interleukin-2; IL-2; immunotherapy; immune function; CD4

© 2004 Lippincott Williams & Wilkins, Inc.