Interleukin-2 (IL-2) is a lymphokine with potent immune-modulating effects. In people with HIV infection, intermittent intravenously or subcutaneously administered IL-2 combined with antiretroviral therapy has led to substantial sustained increases in CD4+ cell counts.1-10 IL-2 can cause flu-like symptoms,1-5,11 however, and subcutaneously administered IL-2 also can cause local irritation.3-5 The net impact of the potential clinical benefits and adverse effects of IL-2 on patients' health perception and ability to function is not known. A comprehensive evaluation of IL-2 would include outcome assessment from the patient perspective in addition to conventional clinical and immunologic end points.
We assessed health-related quality of life (HRQOL) outcomes in a randomized clinical trial sponsored by the Adult AIDS Clinical Trials Group (AACTG) of the National Institutes of Health, AACTG Protocol 328. We hypothesized that IL-2 would have short-term adverse effects on patients' HRQOL but would have a neutral or positive long-term impact on HRQOL.
MATERIALS AND METHODS
In brief, patients enrolled under AACTG Protocol 328 had moderately low CD4+ T-cell counts of 50 to 350 × 106/L, had received less than 14 days of prior protease inhibitor therapy, and had not received prior IL-2. Those who completed an initial 12 weeks of a highly active antiretroviral therapy (HAART) regimen (indinavir plus 2 nucleoside analogues) were randomized to (1) continued HAART alone, (2) HAART plus IL-2 subcutaneous injection or (3) HAART plus IL-2 central intravenous injection. The subcutaneous IL-2 dosage was 7.5 MIU administered twice daily for 5 days every 8 weeks. The intravenous IL-2 dosage was 9 MIU administered daily for 5 days every 8 weeks.
Quality of Life Measures
We assessed HRQOL using the ACTG Short Form (SF)-21.12 This instrument, used in ACTG phase III-IV trials,13,14 is closely related to the Medical Outcomes Study (MOS) SF-21,15 for which reliability and validity data are available.16 The questionnaires comprise a brief (21-item) comprehensive assessment of 8 dimensions of HRQOL: physical functioning, pain, role functioning, social functioning, mental health, cognitive functioning, energy/fatigue, and general health perceptions. Dimension scores are calculated on a scale of 0 to 100, where a higher score indicates better health. A summary score is also calculated as the unweighted average of the 8 dimension scores. In addition, the instrument includes a visual analogue scale used to obtain a pseudo-utility health rating.
To explore the baseline distribution of the HRQOL scores, the mean, median, standard deviation, minimum, maximum, percent at floor (score = 0), and percent at ceiling (score = 100) were determined for each scale. The baseline distribution of scale scores among the 3 treatment groups also was examined. Changes from baseline to 16, 28, and 52 weeks (4, 16, and 40 weeks after randomization) were calculated. P values for the differences among the treatment groups were obtained using the nonparametric Kruskal-Wallis test.
To examine the short-term effect of IL-2 administration on HRQOL scores, the change from day 0 to day 5 of treatment cycles 1, 3, and 6 was calculated. P values for the differences between the subcutaneous and intravenous IL-2 groups were obtained from generalized estimation equation (GEE) linear regression17 to account for repeated measures.
The effect of switching treatment and the extent of missing data also were examined. The HRQOL scores of patients who did and did not switch treatment were compared before treatment switches occurred. Differences by treatment received at 52 weeks were calculated, and P values were obtained using the Kruskal-Wallis test. The amount of missing data at weeks 16, 28, and 52 was obtained for each treatment group and evaluated by χ2 and Fisher exact tests. In addition, at each time point, scores were categorized as present or missing, and nonmissing scores from the previous time point were compared using the Kruskal-Wallis test.
All analyses were performed using SAS version 8.1 (SAS Institute, Cary, NC).
One hundred seventy-four patients were randomized to treatment. Of these, 148 had HRQOL scores at baseline and at 1 or more of the follow-up points (47 HAART only, 53 HAART plus subcutaneous IL-2, and 48 HAART plus intravenous IL-2).
The distributions of baseline HRQOL scores are presented in Table 1. For the 8 subscales, the standard deviation (SD) was typically between 20 and 25 points. Few patients reported scores at the floor of individual subscales, but all scales had patients scoring at the ceiling (8.8%-63.4%). Baseline scores by treatment group are presented in Table 2.
There were no statistically significant differences by treatment group in changes from baseline (12 weeks) to 16 weeks, which was the end of the first cycle of treatment (data not shown). Changes from baseline to 28 and 52 weeks are presented in Tables 3 and 4, respectively. At 28 weeks, the subcutaneous IL-2 group had the best mean change score for 6 of the 8 subscales and for the summary scale. The differences were statistically significant for 2 of the subscales (role functioning, P = 0.03; and cognitive functioning, P = 0.03). At 52 weeks, the subcutaneous IL-2 group had the best mean change score for all 8 subscales and for the average of the scales. The differences were statistically significant for 3 subscales and the summary scale (role functioning, P = 0.02; pain, P = 0.05; physical functioning, P = 0.01; and summary scale, P = 0.02).
At 16, 28, and 52 weeks, 11, 18, and 37 patients, respectively, had missing HRQOL scores. The distribution of missing scores did not differ by treatment group at week 16 or 28. At 52 weeks, 21%, 36%, and 17% of patients from the HAART only, HAART plus subcutaneous IL-2, and HAART plus intravenous IL-2 groups, respectively, had missing data (P = 0.07). Scores at 28 weeks did not differ significantly for people with or without data at 52 weeks.
Twenty-nine patients originally assigned to intravenous IL-2 switched to subcutaneous IL-2 by week 52 (27 at 36 weeks and 2 at 44 weeks); this switching of treatment was allowed by protocol. Among the intravenous IL-2 patients, HRQOL scores at 28 weeks were, on average, lower for those who later switched treatment than for those who did not. When scores at 52 weeks were analyzed according to the treatment patients were receiving at that time rather than by assigned treatment, there were no statistically significant treatment differences (data not shown).
Mean midcycle changes in HRQOL, as measured on days 0 and 5 of cycles 1, 3, and 6, were mostly between 5 and 20 points (Table 5). Midcycle changes were statistically significantly worse for the subcutaneous IL-2 group for 4 of the subscales and for the summary scale. The subcutaneous and intravenous IL-2 groups were similar in the percentage of the protocol dose that was actually administered. Differences between the 2 IL-2 treatment groups in midcycle HRQOL changes were attenuated when the cycle 6 scores were analyzed according to treatment received instead of assigned treatment.
Quality of life evaluation was undertaken for this protocol because of concern that IL-2 administration might be immunologically beneficial but have long-term adverse effects on patients' quality of life. The results presented here suggest strongly that IL-2 administration has a significant short-term impact but provide no evidence of a lasting adverse effect on HRQOL. There is even some evidence that IL-2 given subcutaneously has a small positive effect on HRQOL, specifically on physical functioning, role functioning, and pain.
The short-term effects of IL-2 were expected because it is known to cause fever, chills, and malaise during the time therapy is administered.1-5,11 Subcutaneous IL-2 is also known to cause redness and swelling at the injection site.3-5 Therefore, our observation of significant declines in HRQOL during IL-2 administration lends validity to the measurements and to our long-term findings. Furthermore, our findings of possible long-term beneficial effects of IL-2 are supported by the main outcomes of the trial, which were immunologic in nature. Median CD4+ cell count changes at the end of the study (week 84), were +102, +312, and +459 × 106 cells/L, respectively, for the HAART only, HAART plus subcutaneous IL-2, and HAART plus intravenous IL-2 groups (P = 0.001).18 Interestingly, the subcutaneous IL-2 group had the best HRQOL scores overall, but the intravenous group had the best immunologic measures of the 3 treatment groups. HRQOL scores in the intravenous IL-2 group may have been affected by issues of inconvenience and time required for treatment, which were not specifically asked about on our questionnaires.
Our analyses of the amount of missing data showed no difference by treatment group at earlier time points but some evidence of a difference at 52 weeks, with the subcutaneous IL-2 group having the most missing data points. Consequently, some of the treatment effect at 52 weeks could be attributable to differentially missing data. Nevertheless, it is reassuring that the treatment comparison at 28 weeks, when data were not differentially missing, showed similar trends to those observed at 52 weeks.
Our analyses of switches in treatment suggest that the differences observed in the intention-to-treat analysis were not artifacts attributable to patient crossover. All treatment switches occurred at 36 weeks or later; therefore, the 28-week results were not affected. Although the 28-week results showed smaller and fewer significant treatment differences favoring the subcutaneous IL-2 group than did the 52-week results, they are nevertheless supportive of the conclusion that subcutaneous IL-2 did not cause long-term declines in HRQOL. Furthermore, patients who switched from intravenous IL-2 to subcutaneous IL-2 had, on average, lower HRQOL scores at 28 weeks compared with patients who did not switch treatment, and they continued to have lower HRQOL measures after switching treatment. This likely explains why the small advantage of subcutaneous IL-2 administration mostly disappeared when HRQOL measures at 52 weeks were analyzed according to treatment received.
Subcutaneous IL-2 seems to have a greater short-term impact on HRQOL. Average midcycle changes for all scales except the mental health and cognitive functioning scales were frequently in the range of a 15- to 20-point decline in the subcutaneous IL-2 group, representing a large and clinically significant adverse impact. By contrast, the average declines observed in the intravenous IL-2 group were almost always less than 15 points. This most likely is attributable to the added discomfort of local reactions at the site of subcutaneous administration. Although the subcutaneous IL-2 dose was higher than the intravenous dose, no assumptions regarding the effect of the dose difference can be made because of the different routes of administration.
Of note is that the midcycle changes in HRQOL measures were fairly consistent over the 3 cycles during which they were measured. In other words, our data do not suggest that patients adapt to the side effects over time but, rather, that they continue to experience them as a significant detriment to quality of life.
Statistically significant differences between the treatment groups were detected when the difference between the HAART only and HAART plus subcutaneous IL-2 groups was approximately 7 points for a scale with a larger number of items (physical functioning) and approximately 10 points for scales with few items (role functioning and pain). We had limited power to detect small differences, in the 5- to 7-point range, that still might be considered clinically significant.19,20
The ceiling effects we observed are a consequence, at least in part, of the short nature of the questionnaire used and the limited number of items contributing to a given scale. Such ceiling effects may not be of large consequence in trials in which patients are unlikely to get substantially better. In this trial, the ceiling effects may have blunted the ability to see benefits from IL-2 administration and resulted in conservative estimates of such benefits.
The results of the visual analogue scale suggest that it was not as sensitive to changes in HRQOL, a finding that is consistent with those of others.21 The visual analogue scale seemed to be more in line with the subscale results for the short-term midcycle changes than for the long-term changes, suggesting perhaps that individuals' “internal anchoring” for a visual analogue scale may shift over time. Thus, scales that provide anchoring for item responses may provide better long-term responsiveness.
In conclusion, we found no evidence of a long-term adverse impact of IL-2 administration on quality of life, despite the short-term detriments evident during treatment administration cycles. If further research determines that immunologic effects of IL-2 translate into clinical benefit, subcutaneous or intravenous IL-2 may be a reasonable treatment option for patients who are willing to tolerate its short-term side effects.
Study drugs were provided by Chiron Corporation (David Sahner, MD), Bristol-Myers Squibb Company (Steven Schnittman, MD). and Glaxo Wellcome (Diane Goodwin, PharmD, and Alex Rinehart, PhD).
1. Kovacs JA, Baseler M, Dewar RJ, et al. Increases in CD4 T lymphocytes with intermittent courses in interleukin-2
in patients with human immunodeficiency virus infection. N Engl J Med
2. Kovacs JA, Vogel S, Albert JM, et al. Controlled trial of interleukin-2
infusions in patients infected with the human immunodeficiency virus. N Engl J Med
3. Carr A, Emery S, Lloyd A, et al. Outpatient continuous intravenous interleukin-2
or subcutaneous, polyethylene glycol-modified interleukin-2
in human immunodeficiency virus-infected patients: a randomized, controlled, multicenter study. J Infect Dis
4. Hengge UR, Goos M, Esser S, et al. Randomized, controlled phase II trial of subcutaneous interleukin-2
in combination with highly active antiretroviral therapy
(HAART) in HIV
5. Davey RT Jr, Chaitt DG, Albert JM, et al. A randomized trial of high- versus low-dose subcutaneous interleukin-2
outpatient therapy for early human immunodeficiency virus type 1 infection. J Infect Dis
6. Arnó A, Ruiz L, Juan M, et al. Efficacy of low-dose subcutaneous interleukin-2
to treat advanced human immunodeficiency virus type 1 in persons with ≤250/μL CD4 T cells and undetectable plasma virus load. J Infect Dis
7. Losso MH, Belloso WH, Emery S, et al. A randomized, controlled, phase II trial comparing escalating doses of subcutaneous interleukin-2
plus antiretrovirals versus antiretrovirals alone in human immunodeficiency virus-infected patients with CD4+
cell counts ≥350/mm3
. J Infect Dis
8. Levy Y, Capitant C, Houhou S, et al. Comparison of subcutaneous and intravenous interleukin-2
in asymptomatic HIV
-1 infection: a randomised controlled trial. Lancet
9. Abrams DI, Bebchuk JD, Denning ET, et al. Randomized, open-label study of the impact of two doses of subcutaneous recombinant interleukin-2
on viral burden in patients with HIV
-1 infection and CD4+
cell counts of ≥300/mm3
: CPCRA 059. J Acquir Immune Defic Syndr
10. Levy Y, Durier C, Krzysiek R, et al. Effects of interleukin-2
therapy combined with highly active antiretroviral therapy
on immune restoration in HIV
-1 infection: a randomized controlled trial. AIDS
11. Schwartz DH, Skowtron G, Merigan TC. Safety and effects of interleukin-2
plus zidovudine in asymptomatic individuals infected with human immunodeficiency virus. J Acquir Immune Defic Syndr
12. Wu AW. ACTG QOL601-602 Health Survey Manual: Adult AIDS Clinical Trials Group
. 1998. Forms available at: http://www.fstrf.org/qol/aactg/adult_ql.html
13. Coplan PM, Cook JR, Carides GW, et al., for the AIDS
Clinical Trials Group 320 Study Team. Impact of indinavir on the quality of life in patients with advanced HIV
infection treated with zidovudine and lamivudine. Clin Infect Dis
14. Mrus JM, Williams PG, Tsevat J, et al. Gender differences in health-related quality of life
in patients with HIV
. Qual Life Res
15. Wu AW, Hays RD, Kelly S, et al. Applications of the medical outcomes study health-related quality of life
measures in HIV
. Qual Life Res
16. Bozette SA, Hays RD, Berry SH, et al. Derivation and properties of a brief health status assessment instrument for use in HIV
disease. J Acquir Immune Defic Syndr
17. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika
18. Mitsuyasu R, Pollard R, Gelman R, et al. The virologic, immunologic and clinical effects of interleukin-2
with potent antiretroviral therapy in advanced patients with HIV
infection: a randomized controlled clinical trial
-ACTG 328. (In press).
19. Revicki DA, Wu AW, Brown R. Change in clinical status, health status and health utility outcomes in HIV
-infected patients. Med Care
20. Samsa G, Edelman D, Rothman ML, et al. Determining clinically important differences in health status measures: a general approach with illustration to the Health Utilities Index Mark II. Pharmacoeconomics
21. Wu AW, Jacobson DL, Frick K, et al. Validity and responsiveness of the EuroQol as a measure of health-related quality of life
in people enrolled in an AIDS clinical trial
. Qual Life Res
Credit Roster of Participating AACTG Sites and Contributors (in order of the number of patients enrolled)
University of Texas Medical Branch, Galveston: Richard B. Pollard, MD, Michael J. Borucki, MD, and Suzanne Lanier, RN; Duke University Medical Center: Carol Dukes Hamilton, MD, Shelia Tedder, RN, and Kenneth Shipp, RPh; Washington University, St. Louis: Ge-Youl Kim, RN, BSN, and Michael Conklin, RN, CS, RNP; Case Western Reserve University: Kathy Shina, RN, and Ann Conrad, RN; Tulane University: Juan J. L. Lertora, MD, PhD, and David Mushatt, MD; University of California Los Angeles Medical School: Judy Carden, RN, and Mario Guerrero, MD; Beth Israel Medical Center: Donna Mildvan, MD, and Gwen Costantini, FNP; University of North Carolina: Becky Stephenson, MD, and Laurie Frarey; University of Alabama at Birmingham: Michael Kilby, MD, and Kerry Upton, RN; University of Iowa Hospitals and Clinics: Julie Katseres, ARNP, and Jack T. Stapleton, MD; Cornell University: Valery Hughes, FNP, and Roy Gulick, MD; University of Hawaii: Debra Ogata-Arakaki, RN, and Scott Souza, PharmD; University of Southern California: Michael P. Dube, MD, and Hannah Edmondson-Melancon, RN, MPH; University of Washington, Seattle: Sheryl Storey, PA-C, and N. Jeanne Conley, RN; Ohio State University: Susan L. Koletar, MD, and Charlotte Mills, RN; Beth Israel Deaconess-East Campus: Carol Delaney, RN, and Clyde Crumpacker, MD; New York University/Bellevue: Mary Vogler, MD, and Janet Forcht, RN; University of Rochester: Richard Reichman, MD, and Jane Reid, RNC, MS; University of Colorado Health Sciences Center, Denver: Virginia Waite, BSN, and Elizabeth Connick, MD; Stanford University: Sandra Valle, PA-C, and Debbie Slamowitz, RN, BSN, ACRN; University of California, San Francisco: Mark Jacobson, MD; Indiana University Hospital: Kristen Todd, RN, MSN, and Mitchell Goldman, MD.