Most clinical studies investigating predictors of virologic failure of combination antiretroviral therapy (cART) have focused on the role of treatment factors and laboratory measures, with little information from the patient's perspective. Few studies have assessed the importance of patient-reported measures of health. There is evidence that a significant physical and psychological symptom burden exists in HIV clinic populations, despite the apparent success of cART.1-4 Such symptoms may be related to the infection itself, antiretroviral or other drugs, the unique demographics of HIV clinic populations, or the complex social and lifestyle issues faced by HIV-infected individuals. Evidence from the pre-cART5-10 and cART11-18 eras suggests that depression may predict clinical progression and/or mortality among HIV-infected individuals, particularly among patients starting cART.14-16 Depression1,2,19-22 and physical symptoms1 have also been linked to cART nonadherence. However, the extent to which both physical and psychological symptoms affect virologic outcome of current cART regimens and, in particular, the importance of such symptoms among patients who are generally adherent on successful ART, is an issue that has received little attention. Using data from a self-administered questionnaire conducted in a routine HIV clinic, we investigated whether physical and psychological symptoms predicted subsequent virologic rebound, among patients with viral load suppression on cART.
A self-completed questionnaire on health, lifestyle, and treatment issues was conducted in the Royal Free HIV outpatient clinic during April-August 2005.23 All patients aged ≥18 years with sufficient English language and cognitive ability were approached by a researcher and invited to participate. Of 656 eligible patients, 564 (86%) completed the questionnaire, of whom 337 (60% of completers) gave consent to linkage with the clinic database, which provided complete ART history, CD4 counts and viral loads (VLs), and prospective information on these factors up to May 2008. The analysis includes 188 of 337 patients, who were on ART at questionnaire completion (n = 230) and had latest VL (≤6 months previous) ≤ 50 c/mL (n = 190) and had at least one subsequent VL (n = 188).
The questionnaire included the Memorial Symptom Assessment Scale Short Form (MSAS-SF), a standardized inquiry on occurrence of 32 physical and psychological symptoms in the past week, from which validated subscales are derived,24,25 and additional symptom questions. Six symptom measures were used.
Physical Distress Score (MSAS-SF PHYS)
A standard score representing average physical distress from 12 symptoms: pain/lack of energy/lack of appetite/feeling tired/constipation/dry mouth/nausea/vomiting/change in taste/weight loss/feeling bloated/dizziness.25 Each symptom is scored according to how much it distressed, if present: 0.8 (“not at all”); 1.6 (“a little bit”); 2.4 (“somewhat”); 3.2 (“quite a bit”); and 4.0 (“very much”). Symptom not present is scored as 0.
Psychological Distress Score (MSAS-SF PSYC)
A standard score representing average psychological distress from 6 symptoms: worrying/feeling sad/feeling nervous/feeling irritable/difficulty sleeping/difficulty concentrating.25 Each symptom is scored according to frequency of occurrence, if present: 1 (“rarely”); 2 (“occasionally”); 3 (“frequently”); and 4 (“almost constantly”). Symptom not present is scored as 0.
Global Distress Score (MSAS-SF GDI)
A standard score representing average global distress from 10 symptoms: pain/lack of energy/lack of appetite/feeling tired/constipation/dry mouth/worrying/feeling sad/feeling nervous/feeling irritable.25 Responses are scored from 0 to 4 as above.
Total Number of Symptoms
Number of symptoms present (irrespective of frequency/intensity) from 32 (in addition to above includes cough/changes in skin/numbness or tingling in hands or feet/problems urinating/shortness of breath/diarrhea/sweats/mouth sores/problems with sexual interest or activity/itching/difficulty swallowing/hair loss/swelling of arms or legs/feeling “I don't look like myself”).25
The question from Euroqol 5D26 was used. “Anxiety/depression” was recorded if the patient reported they were “moderately anxious or depressed” or “extremely anxious or depressed” (compared with “not anxious or depressed”).
Enquiry was in the MSAS-SF style. “Suicidal thoughts” was recorded if such thoughts had occurred during the past week, irrespective of frequency (“rarely” to “almost constantly” coded as positive, and compared with no suicidal thoughts).4
VL was measured using commercially available assays with a lower limit of 50 c/mL. Virologic rebound was defined as the first VL (1) > 200 c/mL, and (2) > 50 c/mL (irrespective of ART status) post questionnaire completion. These criteria were chosen before analysis and based on a single raised VL to maximize power. However, we also performed a sensitivity analysis defining rebound as 2 consecutive VL > 50 c/mL. Associations between symptom measures and rebound were assessed using Cox proportional hazards regression, considering each symptom measure separately. Measures 1-4 were grouped into tertiles; tests for trend were used to assess significance. For subjects who did not experience the endpoint, follow-up was censored at the date of last VL or death.
The following factors were also considered.
From patient questionnaire: age group (<40/≥40 years); gender/sexuality (homosexual men/heterosexual men/heterosexual women); ethnicity (white/nonwhite); education (university/other); whether born in UK; time since HIV diagnosis (<3 years/≥3 years).
From medical records: history of AIDS; CD4 count at questionnaire (<200/mm3/≥200/mm3).
From medical records: type of cART regimen [nonnucleoside reverse transcriptase inhibitor (NNRTI)-containing regimen/other]; time with VL ≤ 50 c/mL during current suppression episode (<1 year/1-3 years/>3 years); evidence of previous virologic failure (previous VL > 1000 c/mL ≥28 weeks after starting triple-ART); mono/dual ART before triple ART. From patient questionnaire: patient reports having switched ART (having changed ART combination at least once); nonadherence to ART (defined before analysis as missing ≥2 doses in past week, from inquiry ‘how many doses have you missed’; possible responses 0-10).
Characteristics of the 188 patients on ART with VL ≤ 50 c/mL are given in Table 1. Median [interquartile range] scores for MSAS-PHYS, MSAS-PSYC, and MSAS-GDI (potential maximum score: 4) were 0.7 [0.2, 1.2], 1.2 [0.4, 2.0], and 1.0 [0.4, 1.7], respectively. Median [IQR] total number of symptoms (out of 32) was 10.5 [5, 20]. MSAS symptoms with the highest proportion of patients scoring ≥3 were feeling drowsy/tired (25.0%); worrying (25.0%); difficulty sleeping (22.3%); lack of energy (21.3%); feeling irritable (20.2%); problems with sexual interest/activity (19.7%); diarrhea (14.9%); numbness/tingling (14.9%); feeling nervous (14.4%); and feeling sad (14.4%). Almost half (48.4%) of patients reported anxiety/depression, and 27.1% reported suicidal thoughts in the past week; 3.7% had such thoughts frequently or almost constantly. Physical symptoms were strongly associated with psychologic symptoms. For patients in the first, second, and third MSAS-SF PHYS tertile categories, the percentages with an MSAS-SF PSYC score ≥1.73 (upper tertile) were 7.9%, 27.1%, and 60.6%, respectively, whereas the corresponding percentages with anxiety/depression were 28.6%, 52.5%, and 63.6%.
During a median (range) follow-up of 2.2 (0.4, 3.1) years, 22 patients experienced VL > 200 c/mL; 46 patients experienced VL > 50 c/mL. In each case, 7 patients were off ART at rebound. Three of 188 patients died during follow-up. Table 2 (1) shows unadjusted associations between symptom measures and subsequent VL > 200 c/mL. Symptom measures 1-5 were strongly associated with risk of virologic rebound: risk was increased approximately four to 5-fold for the highest versus lowest tertile groups for physical, psychologic, and global MSAS-SF scores, and total number of symptoms, and 3.7 4-fold for anxiety/depression. Measure 6 (suicidal thoughts) had a weaker nonsignificant association with virologic rebound. Other factors with evidence of association with virologic rebound (P ≤ 0.2) in univariable analyses were age group [unadjusted hazard ratios [95% confidence interval (CI)] 0.4 (0.2-1.0) for ≥40 versus <40 years, P = 0.038]; having switched ART [2.8 (0.8-9.5), P = 0.060]; NNRTI-containing regimen [0.5 (0.2, 1.2), P = 0.11]; time with VL ≤ 50 c/mL [0.6 (0.3-1.6) and 0.3 (0.1-1.3) for 1-3 years and ≥3 years, respectively, versus <1 year, P = 0.058 for trend]; subject-reported nonadherence [3.3 (1.2, 9.0), P = 0.030]. The associations between symptom measures and rebound remained, with little attenuation, after adjustment for age group, switching ART, NNRTI-containing regimen, and time with VL < 50 c/mL [Table 2 (1); adjusted A: P = 0.006 to P = 0.050, measures 1-5]. Additional adjustment for nonadherence resulted in some further attenuation, but did not greatly impact on associations [Table 2 (1); adjusted B: P = 0.011 to P = 0.072, measures 1-5].
Table 2 (2) shows the associations between symptoms and risk of subsequent VL > 50 c/mL. Using this definition, all symptom measures (1-6) were associated with rebound risk. Risk was increased 2- to 3-fold for the highest versus lowest tertiles for symptom measures 1-4, almost 2-fold for anxiety/depression, and 2-fold for suicidal thoughts. Adjustment for other factors resulted in little attenuation of associations [Table 2 (2), adjusted A: P = 0.019 to P = 0.053, and adjusted B: P = 0.022 to P = 0.084, for measures 1-6]. The sensitivity analysis defining rebound as 2 consecutive VL > 50 c/mL (18 rebounds) also suggested similar associations with rebound [eg, hazard ratios (95% CI) for anxiety/depression: 4.5 (1.4, 15.2), P = 0.006, “adjusted A” analysis; 3.3 (1.0, 11.5), P = 0.040 “adjusted B” analysis].
Among this subgroup of HIV patients on successful ART, physical and psychological symptoms were common, and were among the strongest predictors of virologic rebound. The associations appeared largely independent of various known risk factors for virologic failure. Our study was conducted in a routine HIV clinic, and subjects who consented to linkage with clinical data did not differ from others in terms of symptoms (eg, 53.1% and 55.5%, respectively, reported anxiety/depression). The main limitation is that confidence intervals are wide due to the low overall risk of rebound. However, associations appeared consistent across different symptom measures and rebound definitions.
Previous studies have suggested that depression is associated with lower chance of achieving viral suppression in patients initiating a cART regimen,22,27-30 but very few studies have investigated its impact on risk of rebound among patients established on successful treatment. In a French study, depression (using the Hospital Anxiety and Depression Scale) was an independent predictor of virologic failure, even after adjustment for nonadherence, among 71 initial cART responders.31 In the Veteran's Affairs HIV Clinics Study, depression (by the Beck Depression Inventory) predicted virologic failure among 81 patients on cART, but not independently of adherence.32 However, in an American study, neither depression nor other clinical or psychosocial factors predicted viral rebound among 194 patients with viral suppression.27 In addition, there is evidence from trials and observational studies that medical and psychological interventions for depression can improve adherence and treatment outcome.22,33 We found no previous reports of physical symptoms and risk of rebound. In this current study, the physical symptom score predicted viral rebound, and was highly correlated with psychological measures, suggesting that physical symptoms may result in depression and anxiety, or be manifestations of psychological distress. For this reason, it is difficult to isolate the effect of purely physical or purely psychological symptoms, or to separate the effect of depression from that of anxiety. Common individual symptoms in each category appeared predictive of rebound [eg, hazard ratios (95% CI) for risk of VL > 50 c/mL: 2.9 (1.0, 8.7) for “worrying”; 3.2 (1.1, 9.6) for “feeling sad”; 3.1 (1.1, 8.3) for “diarrhea”].
The most likely mechanism linking symptoms to virologic rebound is cART nonadherence, including treatment interruption or discontinuation. By selecting only patients who were initially virally suppressed, we aimed to exclude from analysis those with very substantial adherence issues at the time of the questionnaire. Nevertheless, nonadherence (defined here as missing at least 2 cART doses in the past week) was reported by 8% of patients and was strongly associated with viral rebound. However, nonadherence did not substantially explain the associations between symptoms and viral rebound. It is possible that symptom measures capture additional information about nonadherence that is not captured by direct inquiry on missed cART doses. Depression and physical symptoms may be predictive of future nonadherence and treatment interruption/discontinuation, even among patients who are currently doing well on treatment. Symptoms may also result from drug and alcohol use, or concomitant illness and medication-these factors themselves may affect current or future adherence, and cART tolerability and effectiveness. Our questionnaire did not include inquiry on recreational drug use or alcohol intake, and the number of patients infected via intravenous drug use in our clinic is very low (3 of 188 in this analysis), therefore we could not examine the effect of these factors. Alternatively, symptoms may be linked to viral rebound via mechanisms that are independent of nonadherence, such as changes in immune function. Measures of depression and anxiety have been linked to cell-mediated immunity in HIV-infected individuals.34-36
In conclusion, an assessment of virologic failure risk that is based solely on laboratory results, treatment history and adherence may be missing an important dimension-information from the patient's perspective that is not captured by these measures. These results suggest the importance of ongoing clinical focus on physical and psychological symptoms among patients on successful cART, in addition to those starting treatment. Simple symptom inquiry conducted as part of routine clinical care may be valuable to identify patients at risk of future treatment failure, and provide opportunity not only for assessing adherence but also for appropriate medical or psychological interventions to address physical symptoms and psychological distress.
The authors would like to thank Amanda Jayakody and the research nurses at the Royal Free HIV Clinic, Sally Norwood for assistance with data management, and all survey respondents. This research was assisted with an unrestricted educational grant from GlaxoSmithKline, with input from the Adherence Strategy Group.
1. Ammassari A, Murri R, Pezzotti P, et al. Self-reported symptoms and medication side effects influence adherence to highly active antiretroviral therapy in persons with HIV infection. J Acquir Immun Defic Syndr
. 2001; 28:445-449.
2. Starace F, Ammassari A, Trotta MP, et al, for the AdICoNA and the NeuroICoNA Study Groups. Depression is a risk factor for suboptimal adherence to highly active antiretroviral therapy. J Acquir Immun Def Syndr
3. Willard S, Holzemer WL, Wantland DJ, et al. Does asymptomatic mean without symptoms for those living with HIV infection? AIDS Care
4. Sherr L, Lampe F, Fisher M, et al. Suicidal ideation in UK HIV clinic attenders. AIDS
5. Lyketsos CG, Hoover DR, Guccione M, et al. Depressive symptoms as predictors of medical outcomes in HIV infection. Multicenter AIDS Cohort Study. JAMA
6. Burack JH, Barrett DC, Stall RD, et al. Depressive symptoms and CD4 lymphocyte decline among HIV-infected men. JAMA
7. Page-Shafer K, Delorenze GN, Satariano WA, et al. Comorbidity and survival in HIV-infected men in the San Francisco Men's Health Study. Ann Epidemiol
8. Mayne TJ, Vittinghoff E, Chesney MA, et al. Depressive affect and survival among gay and bisexual men infected with HIV. Arch Intern Med
9. Leserman J, Jackson ED, Petitto JM, et al. Progression to AIDS: the effects of stress, depressive symptoms and social support. Psychosom Med
10. Ickovics JR, Hamburger ME, Vlahov D, et al, for the HIV Epidemiology Research Study Group. Mortality, CD4 cell count decline and depressive symptoms among HIV-seropositive women. JAMA
11. Kilbourne AM, Justice AC, Rollman BL, et al. Clinical importance of HIV and depressive symptoms among veterans with HIV infection. J Gen Intern Med
12. Cook JA, Grey D, Burke J, et al. Depressive symptoms and AIDS-related mortality among a multisite cohort of HIV-positive women. Am J Public Health
13. Ironson G, O'Cleirigh C, Fletcher MA, et al. Psychosocial factor predict CD4 and viral load change in men and women with human immunodeficiency virus in era of highly active antiretroviral treatment. Psychosom Med
14. Bouhnik A, Preau M, Vincent E, et al, and the MANIF 2000 Study Group. Depression and clinical progression in HIV-infected drug users treated with highly active antiretroviral therapy. Antiviral Ther
15. Lima VD, Geller J, Bangsberg DR, et al. The effect of adherence on the association between depressive symptoms and mortality among HIV-infected individuals first initiating HAART. AIDS
16. Villes V, Spire B, Lewden C, et al. The effect of depressive symptoms at ART initiation on HIV clinical progression and mortality: implications in clinical practice. Antiviral Ther
17. Hartzell JD, Janke IE, Weintrob AC. Impact of depression on HIV outcomes in the HAART era. J Antimicrob Chem
18. Mugavero MJ, Wells Pence B, Whetten K, et al. Predictor of AIDS-related morbidity and mortality in a southern U.S. cohort. AIDS Pat Care STD
19. Nieuwkerk P, Gisolf E, Sprangers M, et al, for the Prometheus Study Group. Adherence over 48 weeks in an antiretroviral clinical trial: variable within patients, affected by toxicities and independently predictive of virological response. Antiviral Ther
20. Li X, Margolick JB, Conover CS, et al. Interruption and discontinuation of highly active antiretroviral therapy in the Multicentre AIDS Cohort Study. J Acquir Immune Defic Syndr
21. Vranceanu AM, Safren SA, Lu M, et al. The relationship of post-traumatic stress disorder and depression to antiretroviral medication adherence in persons with HIV. Aids Pat Care STDS
22. Horberg MA, Silverberg MJ, Hurley LB, et al. Effects of depression and selective serotonin reuptake inhibitor use on adherence to highly active antiretroviral therapy and on clinical outcomes in HIV-infected patients. J Acquir Immun Defic Syndr
23. Sherr L, Lampe F, Norwood S, et al. Successive switching of antiretroviral therapy is associated with high psychological and physical burden. Int J STD AIDS
24. Vogl D, Rosenfeld B, Breitbart W, et al. Symptom prevalence, characteristics and distress in AIDS outpatients. J Pain Symptom Manage
25. Chang VT, Hwang SS, Feuerman M, et al. The Memorial Symptom Assessment Scale Short Form (MSAS-SF). Cancer
26. EuroQol Group. EuroQol-a new facility for the measurement of health-related quality of life. Health Policy
27. Wells Pence B, Ostermann J, Kumar V, et al. The influence of psychosocial characteristics and race/ethnicity on the use, duration, and success of antiretroviral therapy. J Acquir Immune Defic Syndr
28. Wells Pence B, Miller W, Gayness BN, et al. Psychiatric illness and virologic response in patients initiating highly active antiretroviral therapy. J Acquir Immun Defic Syndr
29. Anastos K, Schneider MF, Gange SJ. The association of race, sociodemographic and behavioural characteristics with response to highly active antiretroviral therapy in women. J Acquir Immune Defic Syndr
30. Barfod TS, Gerstoft J, Rodkjaer L, et al. Patients' answers to simple questions about treatment satisfaction and adherence and depression are associated with failure of HAART: a cross-sectional survey. AIDS Patient Care
31. Parenti JJ, Massari V, Descamps D, et al. Predictors of virologic failure and resistance in HIV-infected patients treated with nevirapine- or efavirenz-based antiretroviral therapy. Clin Infect Dis
32. Patterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
33. Antoni MH, Carrico AW, Duran RE, et al. Randomized clinical trial of cognitive behavioural stress management on human immunodeficiency virus viral load in gay men treated with highly active antiretroviral therapy. Psychosom Med
34. Leserman J, Petitto JM, Perkins DO, et al. Severe stress, depressive symptoms, and changes in lymphocyte subsets in human immunodeficiency virus-infected men. A 2-year follow-up study. Arch Gen Psychiatry
35. Evans DL, Ten Have TR, Douglas SD, et al. Association of depression with viral load, CD8 T lymphocytes and natural killer cells in women with HIV infection. Am J Psychiatry
36. Greeson JM, Hurwitz BE, Llabre MM, et al. Psychological distress, killer lymphocytes and disease severity in HIV/AIDS. Brain Behav Immun