| |
Effect of long-cycle structured intermittent versus continuous HAART on quality of life in patients with chronic HIV infection
|
| |
| |
The results of this study suggest that repeated long-cycle SIT may not improve HRQL or reduce symptom distress in patients with chronic HIV infection when compared to continuous HAART over a 10-month-period
AIDS: Volume 20(6) 4 April 2006 p 837-845
Powers, April Ea; Marden, Susan Fa; McConnell, LCDR Rosea; Leidy, Nancy Kb; Campbell, Colleen Ma; Soeken, Karen Lc; Barker, Chrisd; Davey, Richard Te; Dybul, Mark Re
From the aNursing and Patient Care Services, National Institutes of Health, Bethesda, Maryland, USA
bMEDTAP International, Bethesda, Maryland, USA
cUniversity of Maryland, Baltimore, Maryland, USA
dScios Incorporated, Fremont, California, USA
eNational Institute of Allergy and Infectious Disease, National Institutes of Health, Bethesda, Maryland, USA.
Abstract
Objective: To examine the effect of repeated, long-cycle structured intermittent versus continuous HAART on health-related quality of life (HRQL) and symptom distress in patients with chronic HIV infection and plasma HIV RNA of less than 50 copies/ml.
Design: Prospective survey of adult patients (n = 46) enrolled in a randomized clinical trial evaluating intermittent versus continuous HAART on immunological and virologic parameters. Patients (n = 23) randomized to structured intermittent therapy received serial cycles of 4 weeks on/8 weeks off HAART.
Outcome measures: HRQL was measured by the physical and mental health summary scores of the Medical Outcomes Study HIV Health Survey (MOS-HIV). Symptom distress was measured by the Symptom Distress Scale. Patients completed initial questionnaires prior to randomization and at weeks 4, 12, and 40 of the trial via a touch screen computer in an outpatient clinic.
Results: Baseline demographic and clinical characteristics were equivalent in both treatment groups. Although the mental health summary score declined significantly over time for the structured intermittent group, linear mixed modeling ANOVA indicated no significant difference across time for MOS-HIV summary and Symptom Distress Scale scores between the two treatment arms.
Conclusion: In this small sample, repeated long-cycle structured intermittent therapy may not provide HRQL or symptom distress advantage compared to continuous HAART in patients with chronic HIV infection over 10 months of treatment. Further research in a heterogenous chronic HIV population and longer follow-up period is warranted.
Discussion
The results of this study suggest that repeated long-cycle SIT may not improve HRQL or reduce symptom distress in patients with chronic HIV infection when compared to continuous HAART over a 10-month-period. In this small sample, differences in changes in HRQL (PHS, MHS) between the two arms was less than 1 point after the first intermittent cycle of treatment and 2 points or less after repeating two more cycles and a fourth off HAART period. Research has demonstrated that the magnitude of clinically important differences in MOS-HIV summary scores is likely to be within the range of 5-7 points [41,60]. Therefore, even if we had achieved statistical significance, the modest differences in treatment effect on HRQL over time are not within the suggested range for interpretation as clinically meaningful. Differences in changes in symptom distress between the two arms across time were 2 points or less and may also be interpreted as modest. For both groups, mean total symptom distress scores across time remained less than 25 which infers low symptom distress [62].
An unexpected finding of this study was that in patients who received SIT, mental health decreased significantly from baseline after completing the first cycle of off/on HAART and did not improve upon completion of two more intermittent cycles and fourth off HAART period. Rather, patients reported significantly poorer mental health after three cycles of SIT and their fourth period off HAART when compared to baseline and first off HAART period. This trend towards poorer mental health was reflected in decreases in mean MHS scores of approximately 3 and 4 points. Although the magnitude of these changes is not within the range of denoting clinical importance, it is noteworthy that poorer mental health was evidenced despite no significant changes in physical health or symptom distress over time. Thus, perhaps this finding requires further evaluation in larger studies.
In comparison, mental health remained relatively unchanged over a 10-month-period in patients who received continuous HAART. This stability in mental health was accompanied by no significant changes in physical health or symptom distress over time.
One explanation for study findings is that multiple, serial cycles of SIT may be necessary to improve HRQL and reduce symptom distress in patients with chronic HIV infection. SIT required patients to change a familiar medication administration routine. In acclimating to SIT, patients may have experienced discomfort and anxiety as they modified their medication administration schedule and dismissed potential fears that their clinical response to the new regimen may not be optimal. Because learning a new medication administration schedule and gaining confidence in its potential clinical benefits may likely be time dependent, adjusting to SIT may have initially taken additional effort on the part of the patients and in the short term consumed a similar, central role in their life as continuous HAART taken prior to this study.
The need to initially adjust to SIT may explain why patients had a significant decrease in their overall mental health during their first intermittent cycle after reintroduction of HAART. At 12 weeks, patients who received SIT had mental health very similar to patients who received continuous HAART. At 40 weeks, mental health in patients who received SIT, although not significantly different, was slightly better than at 12 weeks suggesting some possible adjustment to SIT and/or reflecting the recent 4-week-period off HAART.
Further support for the notion that an initial adjustment process to SIT may preclude HRQL benefit in the short term is evidenced by the study finding that symptom distress during the first intermittent cycle was slightly lower than baseline levels. Although this finding was not significant, it may suggest that HAART related side effects did not contribute substantially to poorer mental health at 12 weeks. In addition, results of a small study (n = 24) that examined the effect of SIT on HRQL in patients with chronic HIV infection (n = 12) showed that quality of life decreased significantly from baseline upon resumption of HAART for 1 week following the first off-therapy interval [38]. In this study, patients (n = 3) reported negative feelings towards reintroduction of a medication routine into their daily activities after an off HAART interval.
On the basis of study results, a study design allowing for longer follow-up periods beyond 10 months may be important in demonstrating the usefulness of SIT for enhancing HRQL and reducing symptom distress. Because of premature termination of this study, patients received only three serial intermittent cycles of HAART rather than the seven cycles as initially planned. Therefore, the length of the follow-up period must be considered as a limitation when interpreting findings from this study.
Another limitation to consider is that patients participating in this study were research subjects in the MSTI clinical trial. MSTI did not include patients with advanced HIV infection. Patients enrolled in MSTI had CD4 lymphocyte counts > 300 cells/ml, no history of major opportunistic infection, and plasma HIV RNA at entry of < 50 copies/ml. Therefore, the sample of this study was biased towards more physically and possibly psychologically healthy patients with chronic HIV infection. The average MOS-HIV summary scores of patients who participated in this study support this premise. Mean PHS and MHS scores were slightly above 50 indicating patients experienced better physical and mental health than a general HIV/AIDS population in the US [54]. Therefore, the findings of this study are generalizable only to a similar population of patients with chronic HIV infection.
The greater overall health of study participants suggests that they may have been less concerned about the therapeutic benefits of HAART. Consequently, therapeutic benefit may not have been a factor judged by these patients as outweighing the toxicity and administration burdens associated with HAART, even if treatment was provided on a structured intermittent basis. Patients' discordant perceptions of HAART therapeutic benefit versus burden may provide an alternative explanation for the results of this study.
Alternatively, patients who received SIT may have been concerned that treatment interruption would not be of clinical benefit to them and perhaps posed a risk to their health. It is possible that this potential concern in combination with the toxicity and administration burdens associated with HAART contributed to the poorer mental health experienced by patients who received SIT as the number of intermittent cycles increased.
At baseline, patients who were randomly assigned to receive SIT were equivalent on demographic and clinical characteristics to those who were randomly assigned to receive continuous HAART. Although there were no significant group differences on HRQL and symptom distress outcomes at baseline, patients assigned to receive SIT had higher mean HRQL scores and lower symptom distress scores at baseline than those assigned to continuous HAART. This suggests the possibility that other factors not accounted for in this study may have moderated the effect of treatment on HRQL and symptom distress confounding study results. Intervening factors that may not have been equally distributed between the groups and potentially contributed to systematic bias include history of antiretroviral treatment, specific HAART regimen, HAART-related symptom burden, and social support.
To our knowledge, no studies have evaluated the effect of repeated, long-cycle SIT on HRQL and symptom distress. Although direct comparison of study findings cannot be made, our findings contrast with reported study results that SIT enhanced HRQL when compared to continuous HAART [38] and a treatment interruption may overall negatively influence HRQL [39]. Consistent with our findings, results of these studies indicated that HRQL decreased significantly after reintroduction of HAART [38,39]. Our study finding of overall stability in HRQL outcomes in patients receiving continuous HAART is consistent with results of other studies [39,42,48,49]. The low symptom distress experienced by patients in this study was similar to symptom distress levels reported by patients who were HIV positive but asymptomatic and not receiving HAART [63,64].
Conclusion
In this small sample of patients with chronic HIV infection and undetectable plasma HIV RNA, results suggest that repeated long-cycle SIT may not provide HRQL and symptom distress advantage when compared to continuous HAART over 10 months of treatment. Further research is warranted to elucidate the effect of SIT in achieving positive HRQL outcomes using study designs that incorporate heterogenous chronic HIV populations and longer follow-up periods. This study highlights the importance of including patient reported outcomes into study designs as novel treatment interruption approaches to improve adherence and minimize adverse side effects to HAART continue to evolve. As multiple therapies for the treatment of HIV prove effective, those that enhance HRQL will become the regimen of choice [69].
Methods
Patients
A repeated measures HRQL survey was conducted of 52 consecutive patients from an outpatient clinic of the Warren Magnuson Clinical Center of the National Institutes of Health during the period of January 2000 to October 2001. Eligible patients were HIV sero-positive adults who could read English and were enrolled in the randomized, controlled trial, 'Immunologic and Virologic Studies of Intermittent versus Continuous HAART in the Treatment of HIV Disease' (MSTI). In this trial, patients who were receiving at least a three-drug HAART regimen and had plasma HIV RNA of < 50 copies/ml were randomized to continue antiretrovirals without interruption for 22 months or to follow an intermittent schedule of seven serial cycles of 4 weeks off and 8 weeks on HAART. Design and results of this clinical trial are reported elsewhere [25].
This parallel HRQL study to the MSTI trial was approved by the Institutional Review Board of the National Institute of Allergy and Infectious Diseases, National Institutes of Health. All patients provided written informed consent prior to participating. Although this study was designed to assess the effect of SIT on HRQL and symptom distress originally over a 22-month-period, the study was terminated in conjunction with the premature enrollment closure of MSTI [25] and so a shorter observation period (10 months) was chosen.
Measures
Medical Outcomes Study HIV Health Survey Version 2.97 (MOS-HIV)
The MOS-HIV was used to measure HRQL. The MOS-HIV is a 35-item, self-administered questionnaire that consists of eight multi-item subscales (health perceptions, physical function, role function, cognitive function, pain, mental health, energy/fatigue, and health distress) and two single item subscales (social function, quality of life) [53]. Subscale scores are computed by summation of individual item scores and converted to a 0-100-point scale with higher scores indicating better health status. Two summary scores, a physical health summary (PHS) and a mental health summary (MHS), can also be derived by weighting each subscale score with appropriate coefficients from the general HIV/AIDS patient population in the US and then aggregating the weighted scores across subscales. The PHS is weighted toward the pain, health perceptions, physical, role, and social function scales [54,55]. The MHS reflects strongly the mental health, health distress, overall quality of life, cognitive function, and energy/fatigue scales [54,55]. Summary scores are transformed to a standard score with a mean of 50 and standard deviation of 10 with higher PHS and MHS scores indicating better overall physical and mental health status, respectively [54]. The psychometric properties of the MOS-HIV demonstrate that it is a reliable and valid measure of health status in HIV-infected populations across illness stages [53,43,56-58]. The MOS-HIV has shown evidence of responsiveness to differences between treatment arms and clinical change [53,59]. The magnitude of differences in MOS-HIV summary scores that is perceptible to patients as clinically meaningful has been empirically derived using anchor based methods and is estimated in the range of 5-7 points [41,53,54,60].
Symptom Distress Scale
Symptom distress was measured using the Symptom Distress Scale. The 13-item-scale was designed to measure the degree of discomfort associated with the following 11 symptoms as perceived by the patient: nausea, appetite, insomnia, pain, fatigue, bowel pattern, concentration, appearance, outlook, breathing, and cough [61]. For each symptom, the patient rates the degree of discomfort on a 5-point Likert scale. The frequency of occurrence for the symptoms nausea and pain is also rated using a 5-point Likert scale. The 13 items are then summed for a total symptom distress score. Scores range from 13 to 65 with higher scores indicating a greater degree of distress. Author guidelines suggest a score less than 25 indicates low symptom distress, 25-32 indicates moderate distress, and 33 and above indicates severe distress [62]. The Symptom Distress Scale has demonstrated evidence of internal consistency reliability, test-retest reliability, and validity in a wide variety of patients including those with HIV infection [62-64].
Procedures
The MOS-HIV and Symptom Distress Scale were administered via a Fujitsu, touch screen notebook computer with specialized software supplied by Assist Technology (Phoenix, Arizona, USA). The computerized method for questionnaire administration has been associated with a high degree of acceptability by patients [65], evidence of good test-retest reliability [65] and high correlations with paper and pencil method [66].
During their MSTI enrollment visit, patients were invited to participate in this parallel HRQL study by a member of the research team. After providing written informed consent, patients were educated on the use of the touch screen computer using a standard protocol and then completed questionnaires in a private clinic room prior to receiving information of their treatment randomization. At this initial visit, HRQL and symptom distress were assessed with patients in both groups having received HAART on a continuous basis prior to study entry. At subsequent follow-up visits, patients completed questionnaires prior to receiving information on CD4 lymphocyte count and plasma HIV RNA at weeks 4, 12, and 40.
Statistical analyses
Internal consistency reliability of MOS-HIV and Symptom Distress Scale was assessed using Cronbach's formula for coefficient alpha (a) [67] as an estimate of precision, scaled 0-1 with higher values indicating greater reliability. The SIT and continuous groups were compared for baseline equivalence on demographic and clinical characteristics and mean PHS, MHS, and symptom distress total scores using chi-square test of independence or Fisher's exact test for categorical variables and Student's t tests for continuous variables with a two-sided P value < 0.05 considered statistically significant. Analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 11.5 software (SPSS Inc., Chicago, Illinois, USA).
Descriptive statistics were used to summarize scores of the MOS-HIV and total symptom distress score at each observation. Separate repeated measures linear mixed model ANOVA were estimated for each outcome using a compound symmetry variance-covariance structure to test whether there were significant differences between groups across time on PHS, MHS, and total symptom distress scores. Estimates of group differences used the fixed effects model. The 'saturated' version of the mixed model was used [68] with a separate (1 degree of freedom) parameter for each time point and group (8 degrees of freedom). A two-sided P value < 0.05 was considered statistically significant and Statistical Analysis Software (SAS) version 6.12 (SAS Institute, Cary, North Carolina, USA) was used to perform these analyses.
Results
All 52 patients enrolled in MSTI met eligibility criteria and agreed to participate in this parallel HRQL study. One (2%) patient was withdrawn from the study because of inability to read well enough to complete questionnaires at baseline; this patient was not included in the analysis. Five (10%) patients were not included in the analysis because early termination of this study precluded them from returning for their scheduled outpatient visit at week 40 to complete questionnaires. Therefore, 46 (88%) patients (23 per arm) had data sufficient for inclusion in this analysis. Of these, four (8%) patients completed end-of-study questionnaires because they voluntarily withdrew from this study in conjunction with their voluntary withdrawal from MSTI. Three of these patients received SIT and completed end-of-study questionnaires at 23, 32, and 36 weeks, respectively. The other patient received continuous HAART and completed an end-of-study questionnaire at 18 weeks. Data from these questionnaires were included for analysis as the patient's 40 week HRQL and symptom distress assessment.
Baseline demographic and clinical characteristics of sample TOP
Patients (n = 46) were predominantly male (96%), white/not hispanic (83%), with mean age of 43 years, and mean (± SD) total CD4 lymphocyte count of 715 ± 250 cells/ml. The two groups were equivalent on demographic and clinical characteristics although patients who received continuous HAART tended to be younger (Table 1).
MOS-HIV and Symptom Distress Scale reliability estimates and mean scores
High internal consistency reliability was evidenced for the MOS-HIV and Symptom Distress Scale across time-points. Coefficient alpha reliability estimates [67] ranged for: (i) PHS from 0.89 to 0.94; (ii) MHS from 0.92 to 0.97; (iii) MOS-HIV subscales from 0.72 to 0.92 except for role function at baseline (a = 0.60); and (iv) Symptom Distress Scale from 0.85 to 0.91.
Mean MOS-HIV subscale scores across time-points for both groups were lowest in the domains of health perceptions (range, 62-68) and energy/fatigue (range, 56-73), with lower mean energy/fatigue scores for the continuous versus SIT group at each time-point. The highest mean subscale scores were in the domains of physical (range, 84-90), social (range, 84-93) and cognitive function (range, 82-86) across time-points.
The groups did not differ significantly at baseline on mean PHS, MHS, or total symptom distress scores. The SIT group had higher mean PHS and MHS scores than the continuous group at baseline and each observation thereafter. Mean total symptom distress scores were lower in the SIT versus continuous group across observations. The plot of the means for PHS, MHS, and total symptom distress scores illustrates these findings (Fig. 1).
Between and within group comparisons
Linear mixed model ANOVA indicated no significant difference between groups across time for PHS, MHS, and total symptom distress scores (Table 2). For the SIT group, there was a significant change in MHS scores over time. MHS scores decreased significantly from baseline and week 4 to week 12 and 40 (Table 3). PHS and total symptom distress scores did not change significantly over time. For the continuous group, PHS, MHS, and total symptom distress scores did not change significantly over time (Table 4).
Introduction
HIV disease-related morbidity and mortality have improved significantly with the advent of HAART [1-3]. Despite improvements in clinical outcomes, HAART does not eliminate HIV infection [4]. Most individuals with HIV infection manage this chronic illness by taking HAART for prolonged periods of time or for the rest of their lives [5]. This long-term commitment requires maintaining near perfect adherence to HAART to achieve clinical efficacy [6-11] and potentially improve health-related quality of life (HRQL) [12,13], which are the primary goals of therapy [5].
Despite these therapeutic benefits, optimal adherence to HAART is a major challenge [14,15]. Non-adherence is prevalent due to the myriad of HAART related side effects and complicated regimens [16-21]. These burdens often assume a central role in the lives of individuals with HIV infection sometimes leading to feelings of resentment regarding the need for therapy [19].
Structured intermittent therapy (SIT) is a treatment interruption approach that entails pre-specified cycles on and off HAART to provide less total time receiving therapy [22-24]. For individuals with chronic HIV infection who have maintained suppression of HIV replication and adequate CD4 cell counts, SIT may promote adherence to HAART by reducing the associated toxicities and cost of the regimen while preserving clinical efficacy and future therapeutic options [23,24]. Findings from pilot studies and clinical trials suggest SIT may offer alternatives to continuous HAART [22,23,25-33], but concerns remain regarding the use of serial cycles of SIT due to emergence of drug-resistant virus during drug restart [34-36]. Thus, defining the risks and benefits of SIT requires further evaluation [5,37] to include whether SIT may improve patient reported outcomes such as HRQL and symptom distress. Results of a small, prospective study demonstrated that SIT may contribute HRQL benefit when compared to continuous HAART [38]. However, findings of a larger, prospective study indicated that the effect of a treatment interruption may differ depending on the reason for the interruption and may not necessarily influence HRQL positively [39].
Although the impact of SIT on HRQL and symptom distress has not been thoroughly evaluated, results of several studies suggest that a continuous HAART regimen may maintain and enhance HRQL [12,39-50]. A decrease in emotional well-being, however, was demonstrated with the use of HAART over a 2-year-period [51].
Fewer studies have examined symptom distress in patients with HIV infection who have received HAART. Johnson et al. found that low symptom distress may be an important predictor of better HRQL outcomes [52]. In patients with more advanced HIV infection, Brechtl et al. found no improvement in moderate symptom distress with HAART use over time despite enhancement of clinical parameters [49].
The purpose of this study was to examine the effect of repeated long-cycle SIT (serial cycles of 4 weeks off/8 weeks on HAART) on HRQL and symptom distress compared to continuous therapy in individuals with chronic HIV infection and undetectable plasma HIV RNA.
|
|
| |
| |
|
|
|
