Mussini, Cristina MD*; Touloumi, Giota PhD†; Bakoyannis, Giorgos PhD†; Sabin, Caroline PhD‡; Castagna, Antonella MD§; Sighinolfi, Laura MD¶; Erikson, Lars E MD∥; Bratt, Goran MD‖; Borghi, Vanni MD*; Lazzarin, Adriano MD§; Cossarizza, Andrea MD**; Esposito, Roberto MD*
It is well known that the introduction and widespread use of highly active antiretroviral therapy (HAART) has modified the natural history of HIV infection.1 However, HAART is likely to be required lifelong and, as we have learnt from other disease areas, it may be difficult for patients to maintain good adherence to treatment over such a long period of time.2 This is particularly relevant for HIV infection, where >95% adherence to therapy is required to prevent the development of resistance.3 To deal with the fatigue and the eventual onset of side effects that may result from such long-term use of treatment, a minority of patients choose to discontinue treatment for varying periods of time, either with or without the support of their physicians.
Several observational and randomized studies, including our own, have investigated whether treatment interruptions (TI) are associated with an increased risk of clinical progression.4-9 However, almost all of these previous studies have focused on the maximum length of time that a patient can safely spend off treatment and/or on the clinical events that develop during any periods off treatment. Moreover, to evaluate the potential risk of clinical progression, most studies have examined the rate of decrease of CD4+ T lymphocytes during the TI and predictive factors for this. These have included the nadir CD4 count, the level of immune restoration experienced on treatment and the degree to which the patient has maintained good virologic suppression in the past.5,8,10 Fewer studies have focused on responses once therapy is restarted.11 The aim of this present study, therefore, was to evaluate the effect of 1 cycle of TI on immune reconstitution after HAART reintroduction, with a particular emphasis on the magnitude and determinants of CD4 recovery at this time.
PATIENTS AND METHODS
The study population derives from a multicenter retrospective study of patients followed in the following centers: Clinic of Infectious Diseases, Modena, Italy; Clinic of Infectious Diseases, San Raffaele Hospital, Milan, Italy; Division of Infectious Diseases, S. Anna Hospital, Ferrara, Italy; and Venhälsan, Stockholm South General Hospital, Stockholm, Sweden.
Patients were included in the present analysis if they had been on HAART for at least 6 months and had discontinued it for any reason (regardless of their plasma viral load at the time), if they had a CD4 count above 500 cells/μL and had reinitiated HAART after a period of at least 4 weeks. Patients were required to have at least one CD4 measurement available during each of the following 4 periods: (1) the 6-month period before TI (TI baseline value); (2) during TI; (3) during TI and within the 6-month period before HAART resumption; and (4) after HAART resumption. In some cases, these measurements could be the same value.
Data sets were requested in a common format, and included information on patient demographics (date of birth, sex, risk group), clinical events (dates and type of all AIDS events, date of death, date of HIV diagnosis), laboratory measurements at diagnosis and over follow-up (CD4 counts and percentages, CD8 counts and percentages, HIV-RNA levels, hemoglobin), and information on antiretroviral use (dates of starting and stopping all antiretroviral drugs).
Changes in CD4 During TI and After Treatment Resumption
As resumption of HAART is likely to have an impact on an individual's future CD4 decline, a competing risks approach was taken for the analyses of CD4 loss during TI. Thus, the cumulative incidence of treatment resumption or of experiencing a drop in CD4 count to <350 cells/μL during TI were modeled using the Fine and Gray proportional hazards model for the subdistribution of a competing risk.12 For these analyses, the baseline was the start of TI; a CD4 threshold of <350 cells/μL was chosen as treatment resumption would generally be recommended at this point.13 For these models we report ratios of subdistribution hazards (RHS). Factors included in these models were sex, age, risk group, year of HAART initiation, duration of pre-TI HAART and of TI, CD4, CD8, and HIV-RNA at HAART initiation, at TI and at treatment resumption, whether the type of HAART at treatment resumption (PI, NNRTI, or NRTI based) was the same as that immediately before TI, and virologic response to pre-TI HAART (confirmed decrease in plasma HIV-RNA level to <500 copies/mL or not).
Virologic and Immunologic Response After HAART Resumption
Virologic response was defined as a confirmed reduction in HIV-RNA <500 copies/mL. As sustained responders we defined those who had at least one additional HIV-RNA measurement available after initial virologic response and had controlled viremia (ie, HIV-RNA <500 copies/mL) during subsequent follow-up. Single blips in HIV-RNA between 500 and 1000 copies/mL were allowed. Virologic relapse was defined as a confirmed increase in HIV-RNA >500 copies/mL after an initial virologic response. Immunologic response after HAART resumption was considered in 3 ways: first, as an increase in CD4 count to pre-TI levels, second as an increase in CD4 counts of >200 cells/μL from the last (during TI) CD4 count before treatment resumption, and third by estimating CD4 changes after treatment resumption.
Factors associated with immunological response (CD4 increase of >200 cells/μL) were identified using Cox proportional hazards regression, with baseline defined as the date of treatment resumption. Patient follow-up was right censored at the last clinic visit or date of death had the patient not experienced an immunological response at that time. Analyses were stratified by center to allow for differences in baseline hazards. The factors considered were similar to those in the previous analysis, with the exception that rates of CD4 and CD8 change pre-TI and during TI were additionally included. These rates of change were estimated through linear mixed models after transforming all CD4 and CD8 measurements onto the square root scale to normalize the corresponding distributions and to stabilize their variances.14,15
CD4 changes after treatment resumption were estimated using a piece-wise linear mixed model allowing for subject-specific deviations in baseline values. For this model, all CD4 counts were transformed onto the square root scale to normalize the distributions and stabilize variances. The slope was allowed to change at 3 months with this time point being chosen on the basis of both graphical and analytical methods. Several factors, including age, sex, risk group, and type of HAART at therapy resumption, were explored as potential prognostic factors of both initial (during the first 3 months) and last (after the first 3 months) slope of CD4.
All analyses were carried out in Stata SE 8.2 (Stata Statistical Software: Release 8.3; Stata Corp., College Station, TX).
Characteristics of the Study Population
In total, 183 subjects were included in the analysis after having being on HAART for a median interquartile range (IQR) time of 3.13 (1.66, 4.67) years, followed by a TI of a median (IQR) duration of 5.52 (2.64, 14.88) months. Of them, 134 were men (73.2%) and the median age at TI was 43 years (IQR 39-49). Risk factors for HIV infection were intravenous drug use 53 (29.0%); sex between men 68 (37.2%); and sex between men and women 62 (33.8%). Of the 183 subjects, 159 (86.9%) had never experienced an AIDS-defining illness, whereas 17 (9.3%) had experienced AIDS in the past (one of whom experienced a subsequent AIDS event during TI). The remaining 7 patients (4.4%) developed a first AIDS event during TI. Immunologic, virologic, and therapeutic data for all patients are shown in Table 1. Fifty-five of the 180 subjects with recorded HIV-RNA at TI had HIVRNA > 500 copies/mL and 37 (34.5%) of them restarted the same as the pre-TI regimen at treatment restart. Among those with HIVRNA ≤ 500 copies/mL at TI, 69 (65.1%) restarted the same as the pre-TI regimen.
Time From Start of TI to HAART Resumption or CD4 < 350 cells/μL
During TI, 80 of the 183 (43.7%) subjects experienced a drop in CD4 cell count to <350 cells/μL whereas the remaining 103 (56.3%) subjects resumed therapy while their CD4 count remained >350 cells/μL. The estimated cumulative incidence [95% confidence interval (CI)] of experiencing a drop in CD4 cell count to <350 cells/μL was 19.1% (13.8%, 25.1%), 29.0% (22.6%, 35.6%), and 33.9% (27.1%, 40.7%) at 6, 12, and 18 months after TI, respectively. The corresponding figures for treatment resumption were 40.4% (33.3%, 47.4%), 44.8% (37.5%, 51.8%), and 51.4% (43.9%, 58.3%), respectively.
Results from multivariate analysis of the factors associated with a drop in CD4 count or treatment resumption suggested that higher CD4 counts, either pre-HAART or pre-TI, were associated with a lower probability of a fall in CD4 during TI [RHS for 100 cell/μL higher pre-HAART CD4: 0.75 (95% CI 0.63, 0.90), P = 0.001; RHS for 100 cell/μL higher pre-TI CD4: 0.82 (0.73, 0.92); P = 0.001]. Subjects with AIDS tended to reinitiate HAART sooner compared with AIDS-free individuals [3.70 (1.76, 7.78), P = 0.001]. Note that as the pre-HAART and pre-TI CD4 counts are highly correlated (Pearson r = 0.78, P < 0.001), the independent roles of the 2 factors cannot be discriminated in the model.
Virologic and Immunological Response After Resumption of HAART
The median (IQR) duration of follow-up after treatment resumption was 1.36 (0.89, 2.59) years. During this time, 155 of the 170 subjects with HIV-RNA >500 copies/mL at HAART resumption (91.2%) experienced an initial virologic response over a median (95% CI) time of 2.56 (2.14, 2.99) months after HAART resumption (Fig. 1). Factors associated with initial virologic response were HIV-RNA at TI and at HAART resumption, CD4 at HAART resumption and virologic response to pre-TI HAART (Table 2). Of the 168 subjects with initial virologic response and at least one subsequent HIV-RNA measurement, 23 (14.47%) had a subsequent virologic relapse, the majority of whom (18/23) had restarted the same regimen as they were receiving pre-TI. The median (IQR) time from initial response to relapse was 0.50 (0, 4.54) years, whereas for the 136 sustained responders the median (IQR) time with undetectable viral load was 1.12 (0.13, 5.34) years.
In all but 19 subjects, CD4 counts were substantially reduced during TI. In total, over a median TI duration of 5.5 months, the median (IQR) CD4 loss was 335 (167-512) cells/μL. Among those with reduced CD4 counts at treatment restart, 66 (40.2%) subjects experienced an increase in CD4 to pre-TI levels. The estimated cumulative probability (95% CI) of experiencing an increase in CD4 to pre-TI levels at 3, 6, 12, 18, and 24 months was 11.6% (7.6%, 17.6%), 23.6% (17.7%, 30.9%), 31.3% (24.6%, 39.3%), 38.3% (30.7%, 47.1%), and 45.5% (36.7%, 55.4%), respectively (Fig. 1).
In total, 125 (68.3%) subjects experienced an immunological response (an increase in CD4 count from baseline of >200 cells/μL) with the median (95% CI) time to immunological response being 8 (5.7, 11.6) months. The cumulative probabilities (95% CI) of gaining at least 200 CD4 cells at 3, 6, 12, 18, and 24 months after treatment resumption were 24.1% (18.5%, 30.9%), 43.1% (36.2%, 50.6%), 58.1% (50.8%, 65.6%), 68.3% (60.7%, 75.7%), and 76.0% (68.1%, 83.1%), respectively (Fig. 1).
Patients who initiated their first HAART regimen when they were ART naive, thus not exposed to mono or dual therapies before their first HAART regimen, were more likely to experience an immunologic response compared with those with pre-HAART ART experience, although this effect was marginally nonsignificant [hazard ratio 1.39 (0.94, 2.06), P = 0.10). Patients with higher pre-TI CD4 counts were more likely to experience an immunologic response (HR per 100 cells/μL higher: 1.25 (1.16, 1.34), P < 0.001]. However, after adjusting for the pre-TI CD4 count, subjects who reinitiated HAART with higher CD4 counts at the time of treatment resumption (ie, those with smaller CD4 losses during TI) were less likely to experience an immunologic response [HR per 100 cells/μL higher: 0.88 (0.79, 0.98), P = 0.02]. Subjects who initiated a PI-based HAART regimen after TI were more likely to experience an immunological response than those who initiated a NNRTI or triple NRTI-based regimen [2.51 (1.54, 4.10), P < 0.001]. At the same time, patients who initiated the same class of therapy as they were receiving immediately before TI were less likely to experience an immunological response [0.60 (0.41, 0.87), P = 0.007] than those who initiated a regimen based on a different drug class.
CD4 Trends After Treatment Resumption
Observed cross-sectional medians (IQR) before and after treatment resumption are shown in Figure 2A, whereas modeled values (with 95% CI) are shown in Figure 2B. The CD4 increase after treatment resumption was biphasic with the slope being steeper during the first 3 months after resumption (Fig. 2B) than subsequently. The median (95% CI) CD4 increase (on the square root scale) per month during the first 3 months after treatment resumption was 1.08 (0.91, 1.25). In contrast, the increase after the first 3 months was relatively slow and only marginally different from 0 [mean (95% CI) rate of increase: 0.32/yr (−0.05, 0.70) on the square root scale]. At 3, 6, 12, 18, and 24 months after treatment resumption, the estimated median (95% CI) CD4 increases were 149 (125, 173), 153 (130, 177), 161 (136, 186), 170 (140, 199), and 178 (141, 214) cells/μL, respectively.
Factors found to affect initial and subsequent rates of CD4 change after resumption of HAART are shown in Table 3. Subjects who had experienced a virologic response (viral load <500 copies/mL) on their pre-TI HAART regimen tended to have higher CD4 counts at the time of treatment resumption (data not shown); these individuals then tended to experience more rapid increases in CD4 counts in the first 3 months after resumption of therapy but slower increases subsequently. Individuals who had experienced less rapid CD4 decline during TI (ie, quartiles 2, 3, and 4 compared with quartile 1) tended to experience less rapid increases in CD4 count in the first 3 months after treatment resumption. The CD8 slope during TI was positively correlated with the CD4 count at treatment resumption (data not shown) but was negatively associated with the rate of CD4 increase during the first 3 months. Subjects who resumed HAART with a higher (≥5,000 copies/mL) viral load experienced a more rapid rate of CD4 increase during the first 3 months after treatment resumption. Although initial (ie, during the first 3 months after HAART resumption) rate of CD4 increase was similar among those with sustained, initial, or nonresponse to HAART resumption, subsequent rates of CD4 increase were significantly higher in sustained responders (Table 3). It is worth noting that when we looked at the gain of >200 CD4 cells/μL, the percentage was larger in those with sustained virologic response followed by those with initial virologic response.
Patients who resumed treatment with a PI-based regimen tended to experience a more rapid rate of CD4 increase during the first 3 months after resuming therapy than patients starting other regimens; in contrast, those who reinitiated HAART with the same class of drug as they had been receiving pre-TI tended to experience a less rapid rate of CD4 increase in the first 3 months.
Despite the publication of findings from several important studies reporting the negative short- and long-term impacts of TIs, it is unlikely that all patients will manage to maintain lifelong HAART, particularly if there is a move toward earlier initiation of HAART.16 The best illustration of this was reported by SMART trial investigators, who reported that 1 year after the premature closure of the study, and against the advice of trial investigators, only 84% of the patients who had originally been randomized to the treatment discontinuation arm had restarted treatment.17
The present study was set-up with the aspiration of providing some help to clinicians when advising their patients on the possible immunological consequences of taking a TI. An increase of 200 cells/μL was chosen to define immunological response because even for patients restarting HAART with a very low CD4 count, such an increase could permit them to discontinue prophylactic regimens. Whilst over two-thirds of patients included in our study experienced a CD4 increase of at least 200 cells/μL after a median of 8 months after resuming therapy, only 45.5% of the patients reached their pre-TI CD4 level by 2 years after resuming therapy. These findings take on greater relevance when considering 2 methodological aspects of our study. First, to be eligible, all patients had to have a pre-TI CD4 count of >500 cells/μL; this restriction was included because a TI at lower CD4 counts would be associated with an unacceptable level of clinical risk,4-6,8 and also because we wanted to ensure that included patients had a strong immune system that could be reconstituted on resumption of HAART. Second, despite the high pre-TI CD4 counts of patients included in the study, the duration of TIs were shorter than those in most trials, including our own.4-8 These patients often decided to discontinue treatment by themselves, thus it could be plausible that they were nonadherent and that immunological response was a direct consequence of an uncontrolled viremia after treatment was restarted. On the contrary, more than 90% of the patients reached an undetectable viral load after resuming HAART. A possible explanation for the differences among effect of viral suppression on CD4 gain could be the CD4 values at HAART initiation among these groups of patients.
Our data have shown that the first phase of CD4 increase seems to be independent of whether one has complete virologic suppression or not, whereas in the second phase CD4 increases are higher in those with a complete virologic suppression. This finding could be explained by the fact that, as shown in naive patients, the first increase in CD4+ cells after therapy is due to the redistribution of memory cells18 that are already present in the organism and dissequestered from lymphoid organs. The second phase of increase, due to naive cells, requires the presence of an intact pool of lymphoid precursors that are generated in the thymus. Most of these cells express the CD4 molecule, can be infected by HIV and thus undergo cell death.19 It is therefore likely that even a low amount of the virus can impair the intrathymic process of T-cell maturation, and that a complete suppression of viral production is essential to have functional T-cell precursors able to generate peripheral blood CD4+ lymphocytes.
Because most of these patients restarted treatment while their CD4 count was above 350 cells/μL, it could be argued that a return to pre-TI CD4 count levels by 2 years may have minimal clinical significance as most patients would already be protected from complications. However, it is important to evaluate the capability of the immune system to reconstitute after a break in therapy. In that regard, the finding that the time taken for the CD4 count to return to pre-TI levels after resuming therapy was substantially longer than the duration of the TI itself (and therefore that the rate of CD4 loss during TI is more rapid than the rate of CD4 gain after resumption of therapy) is concerning. The factors that determine CD4 response in patients resuming therapy are only partially known, but clearly depend on both the host and the virus.20,21 Apparently, the posttreatment resumption immune response was more rapid in patients who had started HAART while ART naive. This may indicate that the immune system may be less damaged, but further studies on the quality of CD4+ T cells are urgently needed to ascertain, for example, the degree to which the CD4 counts that are restored are of memory or naive phenotype. In ARV-naive patients starting HAART for the first time, the CD4 increase is biphasic, with a far more rapid increase in CD4 count in the first 3 months. This initial increase follows a reduction in T-cell activation, and primarily consists of a release of memory CD4 cells that have been trapped in lymphoid tissue (ie, cell redistribution).22,23 It is currently unknown whether the same phenomenon occurs during treatment resumption after TI, but the fact that patients with steeper CD4 declines during TI have steeper CD4 increases in the first 3 months following resumption of treatment supports, at least in part, a similar CD4 redistribution.
In our patients, immune reconstitution was less marked in those who had experienced slower CD4 losses during TI, or who restarted treatment with a higher CD4 count. This phenomenon has also been described in ARV-naive patients starting treatment for the first time,24 although these findings are not universal.25,26
A higher viral load at treatment resumption was associated with a better immunologic response as is also seen in patients initiating HAART for first time.24,27
Interestingly, we found that patients with steeper CD8 drops during TI had faster CD4 increases after restarting HAART, even after adjusting for the CD4 slope during TI. It is interesting that this CD8 effect persists even after adjusting for the CD4 slope during TI. An important predictor of immune response is the level of immune activation which contributes to CD4+ T cell death via apoptosis. Indeed, it has been shown in both animal models28 and in ARV-naive patients that a greater increase in CD4 cells is associated with lower CD8 cell counts at baseline.29 There may be 2 different, although not mutually exclusive, explanations for this phenomenon. First, during HIV infection, CD8+ T cells have the role of controlling viral production by killing infected CD4+ T cells. However, it has been shown that they not only kill such cells, but also eliminate cells that are not infected but bound to viral peptides, or those that are undergoing excessive activation.30
We found that patients who reinitated HAART with the same class of drug that they had been receiving immediately before TI tended to experience poorer immunologic responses after resumption of therapy. It is possible that this finding may reflect the presence of resistance mutations at the time of TI which then compromise the individuals' response to the same regimen in the future, or to another regimen of the same class. Alternatively, as this strategy may not now be recommended (particularly if patients have experienced virologic failure on treatment) then rather than representing genuine structured TIs, treatment breaks in these patients may reflect periods of particularly poor adherence to treatment; responses to treatment may be poorer in these patients once they restart treatment if adherence problems remain. Unfortunately, information about the specific drugs received before TI and after treatment resumption was limited and therefore we cannot investigate these trends in more detail.
Our study has several other limitations. First of all, it was an observational study, thus the criteria for interruption was not established neither was the duration of follow-up after restarting HAART. Second, it was not possible to analyze the T-cell subsets in detail, and in particular the degree of differentiation of either CD4+ or CD8+ T cells in terms of newly produced, virgin, or memory lymphocytes. Thus, we can only give a partial explanation, quantitative and not qualitative, of immune reconstitution. Finally, we studied only one cycle of TI, thus we do not know what could happen after several TIs.
In conclusion, the immune system responds to HAART in the same manner in naive patients as in patients after 1 cycle of TI. It is, therefore, possible to predict the response to antiretroviral treatment reintroduced after TI on the basis of the previous response to HAART. Finally, data on qualitative immune reconstitution are not available, but, at least from a quantitative point of view, we can say that even 1 cycle of TI could have a detrimental effect on immune system.
No author had any conflict of interest relevant to the article.
1. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med. 1998;338:853-860.
2. Gehi AK, Ali S, Na B, et al. Self-reported medication adherence and cardiovascular events in patients with stable coronary heart disease: the heart and soul study. Arch Intern Med. 2007;167:1798-1803.
3. Bangsberg DR, Acosta EP, Gupta R, et al. Adherence-resistance relationships for protease and non-nucleoside reverse transcriptase inhibitors explained by virological fitness. AIDS. 2006;20:223-231.
4. Ananworanich J, Gayet-Ageron A, Le Braz M. Staccato Study Group; Swiss HIV Cohort Study. CD4-guided scheduled treatment interruptions compared with continuous therapy for patients infected with HIV-1: results of the Staccato randomised trial. Lancet. 2006;368:459-465.
5. Maggiolo F, Ripamonti D, Gregis G, et al. Effect of prolonged discontinuation of successful antiretroviral therapy on CD4 T cells: a controlled, prospective trial. AIDS. 2004;18:439-446.
6. Boschi A, Tinelli C, Ortolani P, et al. CD4+ cell-count-guided treatment interruptions in chronic HIV-infected patients with good response to highly active antiretroviral therapy. AIDS. 2004;18:2381-2389.
7. Tarwater PM, Parish M, Gallant JE. Prolonged treatment interruption after immunologic response to highly active antiretroviral therapy. Clin Infect Dis. 2003;37:1541-1548.
8. Mussini C, Bedini A, Borghi V, et al. International Study Group on CD4-monitored Treatment Interruptions. CD4 cell-monitored treatment interruption in patients with a CD4 cell count > 500 × 106 cells/l. AIDS. 2005;19:287-294.
9. El-Sadr WM, Lundgren JD, Neaton JD, et al. Strategies for Management of Antiretroviral Therapy (SMART) Study Group. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med. 2006;355:2283-2289.
10. Touloumi G, Pantazis N, Antoniou A, et al. Highly active antiretroviral therapy interruption: predictors and virological and immunologic consequences. J Acquir Immune Defic Syndr. 2006;42:554-561.
11. Touloumi G, Pantazis N, Stirnadel HA, et al. Rates and determinants of virologic and immunological response to HAART resumption after treatment interruption in HIV-1 clinical practice. J Acquir Immune Defic Syndr. 2008;49:492-498.
12. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496-509.
13. Panel of Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents
. Department of Health and Human Services. Updated 3 November 2008;1-139. Available at: http://www.aidsinfo.nih.gov/ContentFiles/AdultandAdolescentGL.pdf
. Accessed 22 February 2009.
14. Vittinghoff E, Malani HM, Jewell NP. Estimating patterns of CD4 lymphocyte decline using data from a prevalent cohort of HIV-1 infected individuals. Stat Med. 1994;13:1101-1108.
15. Lyles RH, Munoz A, Yamashita TE, et al. Natural history of human immunodeficiency virus type 1 viremia after seroconversion and proximal to AIDS in a large cohort of homosexual men. Multicenter AIDS Cohort Study. J Infect Dis. 2000;181:872-880.
16. Phillips AN, Gazzard BG, Clumeck N, et al. When should antiretroviral therapy for HIV be started? BMJ. 2007;334:76-78.
17. El-Sadr W. SMART Study Group. Re-initiation of ART in the CD4-guided ART Interruption Group in the SMART Study lowers risk of opportunistic disease or death [Abstract 36]. Paper presented at: 15th Conference on Retrovirus and Opportunistic Infections; 2008; Boston, MA.
18. Autran B, Carcelain G, Li TS, et al. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science. 1997;277:112-116.
19. Meissner EG, Duus KM, Loomis R, et al. HIV-1 replication and pathogenesis in the human thymus. Curr HIV Res. 2003;1:275-285.
20. Nasi M, Pinti M, Bugarini R, et al. Genetic polymorphisms of Fas (CD95) and Fas ligand (CD178) influence the rise in CD4+ T cell count after antiretroviral therapy in drug-naive HIV-positive patients. Immunogenetics. 2005;57:628-635.
21. Battegay M, Nuesch R, Hirschel B, et al. Immunological recovery and antiretroviral therapy in HIV-1 infection. Lancet Infect Dis. 2006;6:280-287.
22. Bucy RP, Hockett RD, Derdeyn CA, et al. Initial increase in blood CD4 (+) lymphocytes after HIV antiretroviral therapy reflects redistribution from lymphoid tissues. J Clin Invest. 1999;103:1391-1398.
23. Autran B, Carcelain G, Li TS, et al. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science. 1997;277:112-116.
24. Smith CJ, Sabin CA, Youle MS, et al. Factors influencing increases in CD4 cell counts in HIV-positive persons receiving long-term highly active antiretroviral therapy. J Infect Dis. 2004;190:1860-1868.
25. Staszewski S, Miller V, Sabin CA, et al. Determinants of sustainable CD4 lymphocyte count increases in response to antiretroviral therapy. AIDS. 1999;180:1050-1056.
26. Cozzi-Lepri A, Phillips AN, d'Arminio Monforte A, et al. When to start highly active antiretroviral therapy in chronically HIV-infected patients: evidence from the ICONA study. AIDS. 2001;15:983-990.
27. Landay A, da Silva BA, King MS, et al. Evidence of ongoing immune reconstitution in subjects with sustained viral suppression following 6 years of lopinavir-ritonavir treatment. Clin Infect Dis. 2007; 44;749-754.
28. Pitrak DL, Bolanos J, Hershow R, et al. Discordant CD4 T lymphocyte responses to antiretroviral therapy for HIV infection are associated with ex vivo rates of apoptosis. AIDS. 2001;15:1317-1319.
29. Kaufman GR, Perrin L, Pantaleo G, et al. CD4 T-lymphocyte recovery in individuals with advanced HIV-1 infection receiving potent antiretroviral therapy for 4 years: the Swiss HIV Cohort Study. Arch Intern Med. 2003;163:2187-2195.
30. Cossarizza A, Stent G, Mussini C, et al. Deregulation of the CD95/CD95L system in lymphocytes from patients with primary acute HIV infection. AIDS. 2000;14:345-355.
© 2009 Lippincott Williams & Wilkins, Inc.