Antiviral Therapy 2012; 17: (doi: /IMP2407)

Transcription

1 Antiviral Therapy 2012; 17: (doi: /IMP2407) Original article The clinical benefits of antiretroviral therapy in severely immunocompromised HIV-1-infected patients with and without complete viral suppression Amanda Mocroft 1 *, Wendy P Bannister 1, Ole Kirk 2,3, Justyna D Kowalska 2, Peter Reiss 4, Antonella D Arminio Monforte 5, Jose Gatell 6, Martin Fisher 7, Hanna Trocha 8, Aza Rakhmanova 9, Jens D Lundgren 2,3, the EuroSIDA Study in EuroCOORD 1 Research Department of Infection and Population Health, University College London, London, UK 2 Copenhagen HIV Programme, University of Copenhagen, Copenhagen, Denmark 3 Department of Infectious Diseases, Copenhagen University Hospital, Rigshospitalet, Copenhagen, Denmark 4 Academisch Medisch Centrum bij de Universiteit van Amsterdam, Amsterdam, the Netherlands 5 Istituto Di Clinica Malattie Infettive e Tropicale, Milan, Italy 6 Hospital Clinic i Provincial, Barcelona, Spain 7 Royal Sussex County Hospital, Brighton, UK 8 Medical University, Gdansk, Poland 9 Medical Academy Botkin Hospital, St Petersburg, Russia *Corresponding author a.mocroft@ucl.ac.uk A full list of the EuroSIDA study group can be found via Additional file 1 Background: The aim of this study was to determine whether there is a protective effect of combination antiretroviral therapy (cart) on the development of clinical events in patients with ongoing severe immunosuppression. Methods: A total of 3,780 patients from the EuroSIDA study under follow-up after 2001 with a current CD4 + T-cell count 200 cells/mm 3 were stratified into five groups: group 1, viral load (VL)<50 copies/ml on cart; group 2, VL 50 99,999 copies/ml on cart; group 3, VL 50 99,999 copies/ml off cart; group 4, VL 100,000 copies/ml on cart; and group 5, VL 100,000 copies/ml off cart. Poisson regression was used to identify the risk of (non-fatal or fatal) AIDS- and non-aids-related events considered together (AIDS/non-AIDS) or separately as AIDS or non-aids events within each group. Results: There were 428 AIDS/non-AIDS events during 3,780 person-years of follow-up. Compared with group 1, those in group 2 had a similar incidence of AIDS/ non-aids events (incidence rate ratio [IRR] 1.04; 95% CI ). Groups 3, 4 and 5 had significantly higher incidence rates of AIDS/non-AIDS events compared with group 1; incidence rates increased from group 3 (IRR 1.78; 95% CI ) to group 5 (IRR 2.36; 95% CI ), demonstrating the increased incidence of AIDS/non-AIDS events associated with increasing viraemia. After adjustment, the use of cart was associated with a 40% reduction in the incidence of AIDS/non-AIDS events in patients with VL 50 99,999 copies/ml (IRR 0.59; 95% CI ) and in those with a VL>100,000 copies/ml (IRR 0.66; 95% CI ). Similar relationships were seen for non-aids events and AIDS events when considered separately. Conclusions: In patients with ongoing severe immunosuppression, cart was associated with significant clinical benefits in patients with suboptimal virological control or virological failure. Introduction HIV-related morbidity and mortality have significantly decreased since the introduction of combination antiretroviral therapy (cart) into clinical practice; cart suppresses HIV replication and hence allows immune reconstitution [1 4]. The CD4 + T-cell count and HIV viral load (VL) are both established predictors of clinical prognosis, although there is evidence to suggest a residual benefit of treatment independent of CD4 + T-cell count and VL [5 8]. There are concerns that continuation of treatment in the 2012 International Medical Press (print) (online) 1291

2 A Mocroft et al. presence of detectable viraemia increases the likelihood of evolving drug-resistant HIV, which limits future treatment options [5,9,10]. In the clinical setting, especially in settings where viral load monitoring or the availability of alternate drugs to switch to in case of viral failure are limited, patients might be maintained on cart for extended periods of time despite viral failure [5,11,12]. In the modern cart era, patients with ongoing severe immunosuppression (CD4 + T-cell count 200 cells/mm 3 ) are in a minority [5,13,14]; severe immunosuppression might result from the late presentation of HIV or from failure to achieve a satisfactory response to cart due to non-adherence or development of antiretroviral resistance. Morbidity and mortality associated with both AIDS and non-aids events is highest in these patients [15,16]. Whether cart continues to be clinically effective in the absence of virological control remains an important clinical question that has significant implications for the design of treatment programmes, especially in the resource-constrained environment. Previous research in the early cart era among immunosuppressed patients demonstrated a consistent treatment effect (independent of latest CD4 + T-cell counts) for AIDS and death [6,17,18]. However, there have been significant treatment advances since 2001 [19,20], as well as changes in our understanding of the role of HIV viral load in disease progression and developments in viral load monitoring [21,22]. Moreover, we now have an increased appreciation of the role of non-aidsdefining events such as cancer, cardiovascular, liver and renal diseases [23 25]. The aim of the present study was to evaluate the clinical benefit of cart, measured in terms of fatal and non-fatal AIDS and non-aids events, over and above that captured by CD4 + T-cell count and VL in patients with a current CD4 + T-cell count 200 cells/mm 3 with and without viral suppression. Methods Patients The EuroSIDA study is a prospective, observational cohort of 16,597 HIV-1-infected patients in 102 centres across 33 European countries, Israel and Argentina. The study has been described in detail previously [26]. In brief, patients were enrolled into eight cohorts from May 1994 onwards. Information is collected on a standardized data collection form every 6 months, including all CD4 + T-cell counts and VLs measured since the last follow-up and starting and stopping dates of all antiretroviral drugs. Dates of diagnosis of all clinical AIDS-defining illnesses are recorded using the 1993 clinical definition of AIDS from the Centers for Disease Control [27]. All deaths were also recorded, with cause of death determined by the Coding Causes of Death in HIV (CoDe) protocol and by applying a standardized algorithm [28,29]. To ensure correct patient selection and to verify that accurate data are supplied, members of the coordinating office visit all centres to check the information provided against case notes for all patients with clinical events and 10% of randomly selected patients per year. Statistical methods Patients were included in the analysis if they had at least 1 month of prospective follow-up time in Euro- SIDA with a CD4 + T-cell count 200 cells/mm 3 after 1 January 2001, when information on non-aids events began to be routinely collected and quality assured. Follow-up time was divided into calendar months and patients contributed to the person-years of follow-up (PYFU) if the CD4 + T-cell count in that month (or the preceding 6 months) was 200 cells/ mm 3 and the VL was monitored within 28 days of the CD4 + T-cell count. PYFU were allocated to one of five groups depending on viral suppression and use of cart: group 1, VL<50 copies/ml on cart; group 2, VL 50 99,999 on cart; group 3, VL 50 99,999 off cart; group 4, VL 100,000 on cart; group 5, VL 100,000 off cart. Patients entered and exited the observation period depending on their VL and CD4 + T-cell count and could have several periods of observation, not necessarily consecutive, included in the analyses. Periods where patients were not on or off cart throughout the period, or where the VL was in a different strata at the end compared with the beginning of the period, were excluded. Follow-up was censored at the primary end point or at death/last visit. cart was defined as at least three antiretroviral drugs from any class and baseline was defined as the first time someone satisfied the inclusion criteria for the study based on treatment, VL and CD4 + T-cell count at that time point. Non-AIDS events included in this analysis are those collected in EuroSIDA and as previously published [16]; expert opinion and the steering committee guide decisions about which non-aids events EuroSIDA captures information on. These non-aids events were those considered likely to significantly contribute to morbidity and mortality, and included liver disease, non-aids-defining malignancies, cardiovascular disease, end stage renal disease and pancreatitis, as previously described [16]. Separate analyses were performed for each of the three primary end points: fatal and nonfatal AIDS or non-aids events; AIDS-related events (fatal and non-fatal); or non-aids-related events (fatal and non-fatal). Recurrences of events were excluded. Univariate and multivariate Poisson regression models were used to determine whether there was an independent effect of cart (included as on-treatment), International Medical Press

3 cart in immunocompromised patients Figure 1. Inclusion of patients EuroSIDA n=16,597 Prospective follow-up >1/1/2001 n=13,276 Current CD4 + T-cell count 200 cells/mm 3 ; VL stratified into five groups n=3,198 Group 1 VL<50 On cart Group 2 VL 50 99,999 On cart Group 3 VL 50 99,999 Off cart Group 4 VL 100,000 On cart Group 5 VL 100,000 Off cart /1, /1, /1, /1, /1, /1, / / / / / / / / /430 Numbers are events/person years of follow-up (PYFU) for 1) fatal and non-fatal AIDS and non-aids events; 2) fatal and non-fatal AIDS events; and 3) fatal and nonfatal non-aids events. Viral load (VL) strata are presented as values in copies/ml. after adjustment for other potentially confounding variables, and to test for interactions between levels of viraemia, immunosuppression (categorized as CD4 + T-cell count 50 copies/ml, cells/mm 3 ) and use of cart. Multivariate models were adjusted for age, gender, race, HIV exposure group, region of Europe, prior AIDS or non-aids events, baseline date, nadir CD4 + T-cell count, exposure to Pneumocystis jiroveci pneumonia (PCP) prophylaxis, baseline use of cart, nucleosides, protease inhibitors (PIs), boosted PIs and non-nucleoside reverse transcriptase inhibitors (NNR- TIs), whether the patient had achieved a VL<500 copies/ml before baseline, and whether the patient had achieved a CD4 + T-cell count 200, 350 or 500 cells/ mm 3 prior to baseline. Hepatitis B and C status, CD4 + T-cell count, anaemia (haemoglobin 12/ 14 mg/dl for males/females, respectively), diabetes (insulin-dependent diabetes or use of oral antidiabetic medication or insulin), hypertension (treatment with angiotensinconverting enzyme inhibitors, antihypertensive medication, diastolic/systolic blood pressure 90/ 140 mm/ Hg, respectively), smoking status and the development of chronic kidney disease (confirmed [ 3 months apart] estimated glomerular filtration rate <60 ml/min/1.73 m 2 ) were all defined in-line with previous EuroSIDA work [16] and included as time-updated covariates. SAS software version 9.2 (SAS Institute, Cary, NC, USA, ) was used for all analyses. Results Figure 1 shows how the 3,198 patients were included in the analysis, together with the PYFU for each of the end points considered. Table 1 summarizes the patient characteristics at the first time they met the inclusion criteria for this analysis. As would be expected, there was considerable heterogeneity between the groups. In general, patients in group 4 (VL 100,000 copies/ml on cart) had the most advanced disease; patients exhibited a median CD4 + T-cell count at baseline of 102 cells/mm 3 (IQR cells/mm 3 ), a median nadir CD4 + T-cell count of 50 cells/mm 3 (IQR cells/mm 3 ) and 54.4% of patients had a previous diagnosis of AIDS. Those in group 3 (VL 50 99,999 copies/ml off cart) had the least advanced disease; patients exhibited a higher CD4 + T-cell count at baseline (median 160 cells/mm 3 ; IQR cells/mm 3 ), a higher nadir CD4 + T-cell count (median 135 cells/mm 3 ; IQR cells/mm 3 ) and only 22.8% had a prior diagnosis of AIDS. A comparatively low proportion of patients in each group had a prior non-aids event, the highest proportion (12.8%) was seen in group 1 (VL<50 copies/ml on cart). Table 2 compares the five groups in terms of use of PCP prophylaxis, prior and current use of antiretrovirals and history of viral and immune suppression. Over 65% of the patients had ever taken PCP prophylaxis, although only 26.2% were using prophylaxis at baseline. As expected, Antiviral Therapy

4 A Mocroft et al. patients had received extensive prior antiretroviral treatment. The majority of patients, 83.5%, started cart before baseline, with a median time of 4.1 years since first starting cart (IQR years). In groups 1, 2 and 4 (those that include cart), the majority of patients were receiving either a boosted PI or NNRTI-based regimen. Overall, 72.4% had previously achieved viral suppression (VL<500 copies/ml) a median time of 0.5 years before baseline (IQR years). Groups 3, 4 and 5 have a lower proportion of patients with previous suppression, a greater time before baseline compared with groups 1 and 2. For example, in group 3 (VL 50 99,999 copies/ml off cart), 56.8% had previous viral suppression a median time of 1.5 years prior to baseline (IQR years). The proportion of patients with a previous CD4 + T-cell count 200, 350 or 500 cells/mm 3 was 82.7%, 57.1% and 33.2%, respectively, measured at median times of 0.4 (IQR ) years, 1.1 (IQR ) years and 1.8 (IQR ) years before baseline. Incidence rates of AIDS- and non-aids-defining events A total of 428 patients experienced any (fatal or nonfatal) AIDS or non-aids event during 3,780 PYFU, incidence rate 11.3 per 100 PYFU (95% CI ). The rate of AIDS events was 6.5 per 100 PYFU (95% CI ); there were 258 events during 3,950 PYFU. Table 1. Patient characteristics Group 1 Group 2 Group 3 Group 4 Group 5 VL<50 VL 50 99,999 VL 50 99, , ,000 Characteristic All patients on cart on cart off cart on cart off cart All, n (%) 3,198 (100) 954 (29.8) 1,167 (36.5) 444 (13.9) 283 (8.9) 349 (10.9) Gender Male, n (%) 2,632 (73.9) 742 (77.7) 861 (73.8) 289 (65.1) 217 (76.7) 253 (72.5) Female, n (%) 836 (26.1) 213 (22.3) 306 (26.2) 155 (34.9) 66 (23.3) 96 (27.5) Race a White, n (%) 2,724 (85.2) 830 (86.9) 981 (84.1) 374 (84.2) 239 (84.5) 300 (56.0) Other, n (%) 474 (14.8) 125 (13.1) 186 (15.9) 70 (15.8) 44 (15.6) 49 (14.0) HIV exposure group Homosexual, n (%) 1,169 (36.6) 401 (42.0) 385 (33.0) 134 (30.2) 111 (39.2) 138 (39.5) IDU, n (%) 829 (25.9) 219 (22.9) 311 (26.7) 137 (30.9) 61 (21.6) 101 (28.9) Heterosexual, n (%) 934 (29.2) 252 (26.4) 359 (30.8) 138 (31.1) 92 (32.5) 93 (26.6) Other, n (%) 266 (8.3) 73 (8.7) 112 (9.6) 35 (7.9) 19 (6.7) 17 (4.9) Region South, n (%) 838 (26.2) 225 (23.6) 405 (34.7) 75 (16.9) 85 (30.0) 48 (13.7) Central West, n (%) 794 (24.8) 233 (24.4) 293 (25.1) 118 (26.6) 69 (24.4) 81 (23.2) North, n (%) 876 (27.4) 324 (33.9) 249 (21.3) 122 (27.5) 63 (22.3) 118 (33.8) Central East, n (%) 322 (10.1) 103 (10.8) 103 (8.8) 57 (12.8) 31 (11.0) 28 (8.0) East, n (%) 263 (8.2) 31 (3.3) 79 (6.8) 62 (14.0) 23 (8.1) 68 (19.5) Argentina, n (%) 105 (3.3) 39 (4.1) 38 (3.3) 10 (2.2) 12 (4.2) 6 (1.7) Hepatitis B status Negative, n (%) 2,352 (79.2) 775 ( (75.3) 369 (82.9) 217 (76.7) 293 (84.0) Positive, n (%) 242 (7.6) 77 (8.1) 83 (7.1) 29 (9.5) 27 (9.5) 26 (7.5) Unknown, n (%) 424 (13.3) 103 (10.8) 205 (17.6) 47 (10.6) 39 (13.8) 30 (8.6) Hepatitis C status Negative, n (%) 1,773 (55.5) 564 (59.1) 618 (53.0) 220 (49.6) 177 (62.5) 194 (55.6) Positive, n (%) 876 (27.4) 233 (24.4) 308 (26.4) 162 (36.5) 64 (22.6) 109 (31.2) Unknown, n (%) 549 (17.2) 158 (16.5) 241 (20.7) 62 (14.0) 42 (14.8) 46 (13.2) Prior AIDS, n (%) 1,296 (40.5) 401 (42.0) 533 (45.7) 101 (22.8) 154 (54.4) 107 (30.7) Median age, years (IQR) 41.2 ( ) 43.7 ( ) 40.6 ( ) 39.1 ( ) 41.5 ( ) 39.0 ( ) Median CD4 + T-cell count 154 ( ) 161 ( ) 155 ( ) 160 ( ) 102 (47 158) 144 (84 180) at baseline, cells/mm 3 (IQR) Median nadir CD4 + T-cell 80 (30 140) 65 (27 121) 67 (27 128) 135 (79 173) 50 (16 109) 118 (53 166) count, cells/mm 3 (IQR) Median viral load at baseline, 3.4 ( ) 1.7 ( ) 3.4 ( ) 4.4 ( ) 5.4 ( ) 5.4 ( ) log 10 copies (IQR) Median baseline, mm/yy (IQR) 4/03 (6/01 6/06) 5/04 (9/01 1/08) 3/02 (4/01 3/05) 2/04 (11/01 10/06) 6/02 (3/01 7/05) 6/04 (1/02 4/07) Viral load (VL) strata are presented as values in copies/ml. All P-values < for comparison across five groups except a P=0.40. Baseline defined as the first CD4 + T-cell count 200 cells/mm 3 satisfying the inclusion criteria. cart, combination antiretroviral therapy; IDU, intravenous drug user International Medical Press

5 cart in immunocompromised patients The rate of non-aids events was 5.0 (95% CI ); there were 207 events during 4,159 PYFU. Figure 2 shows the crude incidence rates for each end point for the five groups. For AIDS and non-aids events combined, and for AIDS events alone, there was a clear increase in incidence moving from group 1 (VL<50 copies/ml on cart) to group 5 (VL 100,000 copies/ ml off cart). For example, the incidence of AIDS and non-aids events was 7.8 per 100 PYFU in group 1 (95% CI ) and 28.7 per 100 PYFU in group 5 (95% CI ). The pattern for non-aids events was different: all five groups had an incidence rate of approximately 5 per 100 PYFU (as shown in Figure 2). Adjusted incidence rates of AIDS- and non-aidsdefining events and the role of cart Figure 3A shows the unadjusted and adjusted incidence rate ratios (IRR) of the three clinical end points using group 1 (VL<50 copies/ml on cart) as the reference group. Compared with group 1, those in group 2 had a similar incidence of AIDS/non-AIDS-related events (IRR 1.04; 95% CI ). Groups 3, 4 and 5 had significantly higher incidence rates of AIDS/non-AIDS events when compared with group 1; incidence rates increased from group 3 (IRR 1.78; 95% CI ) to group 5 (IRR 2.36; 95% CI ), demonstrating the increased incidence of AIDS/non-AIDS events associated with increasing viraemia. There were fewer differences between groups 1 5 for non-aids events, illustrating a less marked relationship between increasing viraemia and non-aids events. VL was alternatively included as a categorical variable (VL<50, 50 99,999 or 100,000 copies/ml) and included in an adjusted model, together with the use of cart (shown in Figure 3B). After adjustment, the use of cart was associated with a 40% reduction in the incidence of AIDS/non-AIDS Table 2. Treatment of patients Group 1 Group 2 Group 3 Group 4 Group 5 VL<50 VL 50 99,999 VL 50 99,999 VL 100,000 VL 100,000 Characteristic All patients on cart on cart off cart on cart off cart All, n (%) 3,198 (100) 954 (29.8) 1,167 (36.5) 444 (13.9) 283 (8.9) 349 (10.9) PCP prophylaxis Ever started, n (%) 2,094 (65.5) 650 (68.1) 884 (75.8) 190 (42.8) 215 (76.0) 155 (44.4) On at baseline, n (%) 838 (26.2) 284 (29.7) 356 (30.5) 62 (14.0) 78 (27.6) 58 (16.6) ARVs used before baseline Median number of ARVs (IQR) 6 (4 9) 6 (4 6) 4 (5 10) 3 (0 7) 8 (5 11) 4 (0 8) Median number of nucs (IQR) 4 (2 5) 4 (2 5) 4 (3 5) 2 (0 4) 5 (3 6) 3 (0 4) PI, n (%) 2,444 (76.4) 788 (82.2) 999 (85.6) 220 (49.6) 239 (84.5) 198 (56.7) Boosted PI, n (%) 1,072 (33.5) 388 (40.6) 429 (36.8) 57 (12.8) 129 (45.6) 69 (19.8) NNRTI, n (%) 1,909 (59.7) 626 (65.6) 779 (66.8) 147 (33.1) 209 (73.9) 148 (42.4) cart, n (%) 2,669 (83.5) 918 (96.1) 1,071 (91.8) 234 (52.7) 245 (86.6) 201 (57.6) Median number of years 4.1 ( ) 3.9 ( ) 4.0 ( ) 4.5 ( ) 4.4 ( ) 4.6 ( ) since starting cart, (IQR) ARVs in use at baseline PI, n (%) 470 (19.5) 175 (18.3) 249 (21.3) 46 (16.3) Boosted PI, n (%) 1,048 (43.6) 383 (40.1) 507 (43.4) 158 (55.8) NNRTI, n (%) 979 (40.7) 440 (46.1) 436 (37.4) 103 (36.4) VL<500 before baseline, n (%) 2,315 (72.4) 899 (94.1) 833 (71.4) 252 (56.8) 147 (51.9) 184 (52.7) Median number of years 0.5 ( ) 0.3 ( ) 0.7 ( ) 1.5 ( ) 1.4 ( ) 1.6 ( ) since VL<500 (IQR) CD4 + T-cell count before baseline 200 cells/mm 3, n (%) 2,644 (82.7) 757 (79.3) 955 (81.8) 390 (87.8) 230 (81.3) 312 (89.4) 350 cells/mm 3, n (%) 1,825 (57.1) 501 (52.5) 578 (49.5) 321 (72.3) 166 (58.7) 259 (74.2) 500 cells/mm 3, n (%) 1,063 (33.2) 298 (31.2) 289 (24.8) 198 (44.6) 92 (32.5) 186 (53.3) Median number of years since CD4 + T-cell count (IQR) 200 cells/mm ( ) 0.3 ( ) 0.4 ( ) 0.5 ( ) 0.7 ( ) 0.5 ( ) 350 cells/mm ( ) 0.6 ( ) 1.4 ( ) 1.1 ( ) 1.5 ( ) 1.2 ( ) 500 cells/mm ( ) 0.9 ( ) 2.1 ( ) 1.9 ( ) 2.3 ( ) 1.8 ( ) Viral load (VL) strata are presented as values in copies/ml. Baseline defined as the first CD4 + T-cell count 200 cells/mm 3 satisfying the inclusion criterion. All P-values < for comparison across the groups with data. ARV, antiretroviral; cart, combination antiretroviral therapy; NNRTI, non-nucleoside reverse transcriptase inhibitor; nucs, nucleosides; PCP, Pneumocystis jiroveci pneumonia; PI, protease inhibitor. Antiviral Therapy

6 A Mocroft et al. Figure 2. Crude incidence rates of clinical events Group VL, copies/ml cart Incidence rate, events per 100 PYFU (95% CI) All 1 <50 On ,999 On 50 99,999 Off 100,000 On 100,000 Off 1.00 Any AIDS/ non-aids event AIDS events non-aids events Number of events Group number All All All cart, combination antiretroviral therapy; PYFU, person-years follow-up. events combined in patients with a VL 50 99,999 copies/ml (IRR 0.59; 95% CI ; P=0.0041) and in those with a VL>100,000 copies/ml (IRR 0.66; 95% CI ; P=0.047). Similar relationships were seen for non-aids events and AIDS events when considered separately. There was no interaction between the use of cart and level of viral suppression for any of the clinical end points (P>0.2), indicating that the beneficial effect of cart was similar for those with moderate (VL 50 99,999 copies/ml) and high levels of viral load (VL 100,000 copies/ml). There were limited data among those with extremely low CD4 + T-cell counts ( 50 cells/mm 3 ). There were 142 non-aids and AIDS events occurring in 451 PYFU (incidence 31.5 per 100 PYFU; 95% CI ), 117 AIDS events in 473 PYFU (incidence 24.7 per 100 PYFU; 95% CI ) and 47 non-aids events in 555 PYFU (incidence 8.5 per 100 PYFU; 95% CI ). After adjustment, the use of cart was associated with a greater than 40% reduction in the incidence of AIDS/non-AIDS events combined in patients with VL 50 99,999 copies/ml (IRR 0.54; 95% CI ; P=0.051) and a 20% reduction in those with a VL 100,000 copies/ml (IRR 0.81; 95% CI ; P=0.53). There were insufficient data to consider AIDS and non-aids events separately. In addition, there was no interaction between CD4 + T-cell count (categorized as 50 or cells/mm 3 ), use of cart or viral load and any clinical end points (P>0.2, all tests). This observation suggests that the beneficial effect of cart was similar in those with extremely low CD4 + T-cell counts and that the relationship between higher viral load and clinical end points was similar in those with low ( cells/mm 3 ) and extremely low CD4 + T-cell counts ( 50 cells/mm 3 ). Discussion This analysis focused on over 3,000 patients with a current CD4 + T-cell count 200 cells/mm 3 and demonstrated that the use of cart was associated with a lower rate of both AIDS and non-aids events in patients with low CD4 + T-cell counts, even without full viral suppression International Medical Press

7 cart in immunocompromised patients Figure 3. Incidence rate ratios within groups and the effect of cart within viral load strata A Unadjusted Adjusted a Group VL, copies/ml cart 1 <50 On ,999 On ,999 Off 4 100,000 On 5 100,000 Off IRR, 95% CI Any AIDS/ non-aids event AIDS events non-aids events Group B Adjusted IRR of cart, 95% CI 1.00 VL strata 50 99, , Any AIDS/ non-aids event AIDS events non-aids events P-value (A) The unadjusted and adjusted incidence rate ratios of the three clinical end points using group 1 (VL<50 copies/ml on cart) as the reference group. (B) The effect of cart within viral load strata. a Multivariate models were adjusted for age, gender, race, HIV exposure group, region of Europe, prior AIDS or non-aids events, baseline date, nadir CD4 + T-cell count, exposure to Pneumocystis jiroveci pneumonia (PCP) prophylaxis, baseline use of cart, nucleosides, protease inhibitors (PIs), boosted PIs and non-nucleoside reverse transcriptase inhibitors (NNRTIs), whether the patient had achieved a VL<500 copies/ml before baseline, and whether the patient had achieved a CD4 + T-cell count 200, 350 or 500 cells/mm 3 prior to baseline. Hepatitis B and C status, CD4 + T-cell count, anaemia (haemoglobin 12/ 14 mg/ dl for males/females, respectively), diabetes (insulin-dependent diabetes or use of oral antidiabetic medication or insulin), hypertension (treatment with angiotensinconverting enzyme inhibitors, antihypertensive medication, diastolic/systolic blood pressure 90/ 140 mm/hg, respectively), smoking status and the development of chronic kidney disease (confirmed [ 3 months apart] estimated glomerular filtration rate <60 ml/min/1.73 m 2 ) were all included as time-updated covariates. Antiviral Therapy

8 A Mocroft et al. We found a strong relationship between increasing viral load and increased risk of AIDS events, in agreement with findings from the recent Strategies for Management of Antiretroviral Therapy (SMART) randomized trial [30]. These observations suggest that continuous viral suppression is optimal for reducing the risk of both fatal and non-fatal AIDS events. These results support others suggesting a beneficial clinical effect of cart over and above that measured through CD4 + T-cell count and VL in patients with low CD4 + T-cell counts [31 34]. Maintaining viral suppression on cart did not appear to have a significant protective effect against non-aids events, although there was a marginally significant increased risk in group 5 (VL 100,000 copies/ml off cart) compared with group 1 (VL<50 copies/ml on cart). One possible explanation is that viral suppression is a less important direct risk factor for end-organ disease, as found in the SMART trial with regards to cardiovascular disease [35]. These findings add weight to the fact that immunosuppression is related to non-hiv-associated events by routes other than HIV per se. For example, there are concerns that HIV is associated with accelerated ageing [36], whereby the risk of non-aids morbidity is higher among successfully treated HIV-infected persons than appropriate age-matched controls and independent of conventional risk factors [37,38]. In addition, many abnormalities of T-cells normally associated with ageing are also observed in untreated HIV infection [39 41] and the association between HIV infection and inflammation share many similarities with ageing and inflammation [36]. Accelerated ageing is probably a multifactorial process involving an increased burden of behavioural risk factors, such as smoking [42,43], an increase in coinfections such as hepatitis B and C, human papillomavirus and Epstein Barr virus [44 48], as well as comorbidities such as diabetes and hypertension [49,50]. Other contributing factors could include antiretroviral toxicities and increased vulnerability to injuries and frailty [51]. Thus, these non-aids-related diseases might be driven by HIV, but only indirectly; for example, by inducing the activation of inflammatory and coagulation pathways, which have been linked to an increased risk of cardiovascular disease [35,52], or by causing generalized immune activation leading to increased T-cell turnover, which could be linked to conditions of the cardiovascular and digestive system [53,54], putting patients at higher risk of all-cause mortality. Previous studies from EuroSIDA have demonstrated a continuum of increasing risk of fatal and non-fatal AIDS and non-aids events with decreasing CD4 + T-cell count (this is particularly true for AIDS events) and a strong relationship between increased viral replication and both AIDS and non-aids events [16]. An analysis focused on those with high CD4 + T-cell counts (>350 cells/mm 3 ) showed an increased incidence of AIDS events and a slightly increased incidence of non-aids events in patients with uncontrolled viral suppression, although there were few differences between moderate and high viraemia and the risk of non-aids events [55]. The present study focuses on those with ongoing advanced immunosuppression (those at greatest risk of clinical disease) and suggests that cart is associated with a lower rate of both AIDS and non-aids events, even in patients who are not able to maintain viral suppression. This is consistent with findings from other studies [56 58]. There are several potential explanations for such a finding. The reduction in clinical events might reflect immunological stability, whereby drugresistant virus with reduced fitness, replicative capacity and pathogenicity is selected by antiretroviral drug pressure [56,59,60]. In addition, patients on cart might be more likely to remain under follow-up than those not taking cart and benefit from other services, such as screening for comorbidities. The limitations of this analysis should be considered. Although a large number of known, measured confounding variables were considered for this analysis, the data were from an observational study. Thus, bias from unmeasured or unknown variables that could affect the risk of many comorbidities, such as socio-economic factors, duration of HIV infection, adherence, use of alcohol, level of injecting drug use or different levels of patient care, cannot be ruled out. The frequency of VL measurements and CD4 + T-cell counts varied from patient to patient, but the analysis only included data where the most recent measurement was taken within the previous 6 months to minimize this bias. This approach limited the PYFU included in the two groups off all treatment, as these patients were more likely to have limited contact with the health system and therefore might be missing laboratory measurements. We combined fatal and nonfatal AIDS and non-aids events to increase power and because non-fatal events have a significant effect on mortality and morbidity [16,25]; however, analyses considering fatal and non-fatal events separately showed consistent results (data not shown). Although it is assumed that all clinical events and deaths are recorded on the follow-up forms, it is possible that some were missed due to loss to follow-up. Loss to follow-up in EuroSIDA was assessed in 2008 and the incidence was found to be 3.7 per 100 PYFU; incidence was highest in patients who had not yet started treatment and in those with lower CD4 + T-cell counts [61]. Therefore, the incidences found in this analysis might be underestimates, most probably in the group not receiving any treatment International Medical Press

9 cart in immunocompromised patients In conclusion, despite the inability to achieve virological suppression or in patients with virological failure there are significant clinical benefits associated with the use of cart in patients with ongoing immunosuppression. Although cost-effectiveness, toxicities and the potential emergence of drug resistance that could preclude future treatment options are important considerations, patients with limited treatment options and consistently low CD4 + T-cell counts might still benefit from cart in terms of lowering the risk of clinical disease progression even without achieving optimal viral suppression. Acknowledgements We would like to thank Henry Masur (National Institutes of Health Clinical Center) for proposing the idea for this project. OK, JDL and AM designed the study. OK, JDL, PR, AD A-M, JG, MF, HT and AR collected the data. WPB, OK, JDL and AM designed the analysis. AM and WPB carried out the statistical analysis of the data. AM, WPB, OK, JDL, PR, AD A-M, JG, MF, HT, AR and JDL wrote the manuscript. Primary support for EuroSIDA is provided by the European Commission BIOMED 1 (CT ), BIOMED 2 (CT ), the 5th Framework (QLK ) the 6th Framework (LSHP- CT ) and the 7th Framework (FP7/ , EuroCoord number ) programmes. Current support also includes unrestricted grants from Gilead, Pfizer, BMS and Merck and Co. The participation of centres from Switzerland was supported by The Swiss National Science Foundation (grant ). Disclosure statement The authors declare no competing interests. Additional file Additional file 1: A list of the members of the EuroSIDA study group can be found at com/uploads/documents/avt-12-oa-2489_mocroft_ Add_file1.pdf References 1. Mocroft A, Ledergerber B, Katlama C, et al. Decline in the AIDS and death rates in the EuroSIDA study: an observational study. Lancet 2003; 362: Murphy EL, Collier AC, Kalish LA, et al. Highly active antiretroviral therapy decreases mortality and morbidity in patients with advanced HIV disease. Ann Intern Med 2001; 135: 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: Vittinghoff E, Scheer S, O Malley P, Colfax G, Holmberg SD, Buchbinder SP. Combination antiretroviral therapy and recent declines in AIDS incidence and mortality. J Infect Dis 1999; 179: MacArthur RD, Perez G, Walmsley S, Baxter JD, Mullin CM, Neaton JD. Comparison of prognostic importance of latest CD4+ cell count and HIV RNA levels in patients with advanced HIV infection on highly active antiretroviral therapy. HIV Clin Trials 2005; 6: Miller V, Phillips AN, Clotet B, et al. Association of virus load, CD4 cell count, and treatment with clinical progression in human immunodeficiency virus-infected patients with very low CD4 cell counts. J Infect Dis 2002; 186: Sterling TR, Chaisson RE, Moore RD. HIV-1 RNA, CD4 T-lymphocytes, and clinical response to highly active antiretroviral therapy. AIDS 2001; 15: Yerly S, Perneger TV, Hirschel B, et al. A critical assessment of the prognostic value of HIV-1 RNA levels and CD4+ cell counts in HIV-infected patients. Arch Intern Med 1998; 158: Kozal MJ. Drug-resistant human immunodefiency virus. Clin Microbiol Infect 2009; 15 Suppl 1: Daar ES, Richman DD. Confronting the emergence of drug-resistant HIV type 1: impact of antiretroviral therapy on individual and population resistance. AIDS Res Hum Retroviruses 2005; 21: Aleman S, Soderbarg K, Visco-Comandini U, Sitbon G, Sonnerborg A. Drug resistance at low viraemia in HIV-1- infected patients with antiretroviral combination therapy. AIDS 2002; 16: Deeks SG. Durable HIV treatment benefit despite low-level viremia: reassessing definitions of success or failure. JAMA 2001; 286: Lampe FC, Smith CJ, Madge S, et al. Success of clinical care for human immunodeficiency virus infection according to demographic group among sexually infected patients in a routine clinic population, 1999 to Arch Intern Med 2007; 167: Sabin CA, Phillips AN, Yee TT, Griffioen A, Lee CA. Twenty five years of HIV infection in haemophilic men in Britain: an observational study. BMJ 2005; 331: Achhra AC, Amin J, Law MG, et al. Immunodeficiency and the risk of serious clinical endpoints in a well studied cohort of treated HIV-infected patients. AIDS 2010; 24: Mocroft A, Reiss P, Gasiorowski J, et al. Serious fatal and nonfatal non-aids-defining illnesses in Europe. J Acquir Immune Defic Syndr 2010; 55: Mocroft A, Katlama C, Johnson AM, et al. AIDS across Europe, : the EuroSIDA study. Lancet 2000; 356: Miller V, Mocroft A, Reiss P, et al. Relations among CD4 lymphocyte count nadir, antiretroviral therapy, and HIV-1 disease progression: results from the EuroSIDA study. Ann Intern Med 1999; 130: US Department of Health and Human Services. Guidelines for the use of antiretroviral agents in HIV-1 infected adults and adolescents. AIDSinfo. (Updated 27 March Accessed 13 August 2012.) Available from Thompson MA, Aberg JA, Cahn P, et al. Antiretroviral treatment of adult HIV infection: 2010 recommendations of the International AIDS Society-USA panel. JAMA 2010; 304: Mellors JW, Rinaldo CR, Jr., Gupta P, White RM, Todd JA, Kingsley LA. Prognosis in HIV-1 infection predicted by the quantity of virus in plasma. Science 1996; 272: Mellors JW, Munoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med 1997; 126: Lifson AR, Belloso WH, Davey RT, et al. Development of diagnostic criteria for serious non-aids events in HIV clinical trials. HIV Clin Trials 2010; 11: Marin B, Thiebaut R, Bucher HC, et al. Non-AIDS-defining deaths and immunodeficiency in the era of combination antiretroviral therapy. AIDS 2009; 23: Antiviral Therapy

10 A Mocroft et al. 25. Neuhaus J, Angus B, Kowalska JD, et al. Risk of all-cause mortality associated with nonfatal AIDS and serious non- AIDS events among adults infected with HIV. AIDS 2010; 24: Mocroft A, Vella S, Benfield TL, et al. Changing patterns of mortality across Europe in patients infected with HIV-1. EuroSIDA Study Group. Lancet 1998; 352: revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep 1992; 41: Kowalska JD, Mocroft A, Ledergerber B, et al. A standardized algorithm for determining the underlying cause of death in HIV infection as AIDS or non-aids related: results from the EuroSIDA study. HIV Clin Trials 2011; 12: Kowalska JD, Friis-Moller N, Kirk O, et al. The Coding Causes of Death in HIV (CoDe) Project: Initial Results and Evaluation of Methodology. Epidemiology 2011; 22: El-Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ countguided interruption of antiretroviral treatment. N Engl J Med 2006; 355: Guiguet M, Porter K, Phillips A, Costagliola D, Babiker A. Clinical progression rates by CD4 cell category before and after the initiation of combination antiretroviral therapy (cart). Open AIDS J 2008; 2: Sterne JA, Hernan MA, Ledergerber B, et al. Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort study. Lancet 2005; 366: Torti C, Lapadula G, Maggiolo F, et al. Predictors of AIDS-defining events among advanced naive patients after HAART. HIV Clin Trials 2007; 8: Moore RD, Gebo KA, Lucas GM, Keruly JC. Rate of comorbidities not related to HIV infection or AIDS among HIV-infected patients, by CD4 cell count and HAART use status. Clin Infect Dis 2008; 47: Phillips AN, Carr A, Neuhaus J, et al. Interruption of antiretroviral therapy and risk of cardiovascular disease in persons with HIV-1 infection: exploratory analyses from the SMART trial. Antivir Ther 2008; 13: Deeks SG. HIV infection, inflammation, immunosenescence, and aging. Annu Rev Med 2011; 62: Hogg R, Lima V, Sterne JA, et al. Life expectancy of individuals on combination antiretroviral therapy in highincome countries: a collaborative analysis of 14 cohort studies. Lancet 2008; 372: Bhaskaran K, Hamouda O, Sannes M, et al. Changes in the risk of death after HIV seroconversion compared with mortality in the general population. JAMA 2008; 300: Effros RB, Dagarag M, Spaulding C, Man J. The role of CD8+ T-cell replicative senescence in human aging. Immunol Rev 2005; 205: Appay V, Almeida JR, Sauce D, Autran B, Papagno L. Accelerated immune senescence and HIV-1 infection. Exp Gerontol 2007; 42: Desai S, Landay A. Early immune senescence in HIV disease. Curr HIV/AIDS Rep 2010; 7: Chaturvedi AK, Pfeiffer RM, Chang L, Goedert JJ, Biggar RJ, Engels EA. Elevated risk of lung cancer among people with AIDS. AIDS 2007; 21: Engels EA, Brock MV, Chen J, Hooker CM, Gillison M, Moore RD. Elevated incidence of lung cancer among HIVinfected individuals. J Clin Oncol 2006; 24: Accepted 5 March 2012; published online 26 September Alter MJ. Epidemiology of viral hepatitis and HIV co-infection. J Hepatol 2006; 44 Suppl 1:S6 S Chaturvedi AK, Madeleine MM, Biggar RJ, Engels EA. Risk of human papillomavirus-associated cancers among persons with AIDS. J Natl Cancer Inst 2009; 101: Mocroft A, Soriano V, Rockstroh J, et al. Is there evidence for an increase in the death rate from liver-related disease in patients with HIV? AIDS 2005; 19: Shibata D, Weiss LM, Hernandez AM, Nathwani BN, Bernstein L, Levine AM. Epstein-Barr virus-associated non- Hodgkin s lymphoma in patients infected with the human immunodeficiency virus. Blood 1993; 81: Thio CL, Seaberg EC, Skolasky R, Jr., et al. HIV-1, hepatitis B virus, and risk of liver-related mortality in the Multicenter Cohort Study (MACS). Lancet 2002; 360: Gæde P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med 2003; 348: Mourad JJ, Le JS. Blood pressure control, risk factors and cardiovascular prognosis in patients with diabetes: 30 years of progress. J Hypertens Suppl 2008; 26:S7 S Justice AC. HIV and aging: time for a new paradigm. Curr HIV/AIDS Rep 2010; 7: Kuller LH, Tracy R, Belloso W, et al. Inflammatory and coagulation biomarkers and mortality in patients with HIV infection. PLoS Med 2008; 5:e Riddler SA, Smit E, Cole SR, et al. Impact of HIV infection and HAART on serum lipids in men. JAMA 2003; 289: Deeks SG, Phillips AN. HIV infection, antiretroviral treatment, ageing, and non-aids related morbidity. BMJ 2009; 338:a Reekie J, Gatell J, Yust I, et al. AIDS and non-aids events in HIV-1 positive individuals with high CD4 counts according to viral load strata. AIDS 2011; 25: Deeks SG, Barbour JD, Martin JN, Swanson MS, Grant RM. Sustained CD4+ T cell response after virologic failure of protease inhibitor-based regimens in patients with human immunodeficiency virus infection. J Infect Dis 2000; 181: Lawrence J, Mayers DL, Hullsiek KH, et al. Structured treatment interruption in patients with multidrug-resistant human immunodeficiency virus. N Engl J Med 2003; 349: Ledergerber B, Lundgren JD, Walker AS, et al. Predictors of trend in CD4-positive T-cell count and mortality among HIV-1-infected individuals with virological failure to all three antiretroviral-drug classes. Lancet 2004; 364: Deeks SG, Wrin T, Liegler T, et al. Virologic and immunologic consequences of discontinuing combination antiretroviral-drug therapy in HIV-infected patients with detectable viremia. N Engl J Med 2001; 344: Prado JG, Parkin NT, Clotet B, Ruiz L, Martinez-Picado J. HIV type 1 fitness evolution in antiretroviral-experienced patients with sustained CD4+ T cell counts but persistent virologic failure. Clin Infect Dis 2005; 41: Mocroft A, Kirk O, Aldins P, et al. Loss to follow-up in an international, multicentre observational study. HIV Med 2008; 9: International Medical Press