Antiviral Therapy 2013; 18: (doi: /IMP2436)

Transcription

1 Antiviral Therapy 2013; 18: (doi: /IMP2436) Original article Decreasing rate of multiple treatment modifications among individuals who initiated antiretroviral therapy in in the Danish HIV Cohort Study Marie Helleberg 1 *, Gitte Kronborg 2, Carsten S Larsen 3, Gitte Pedersen 4, Court Pedersen 5, Lars Nielsen 6, Alex L Laursen 3, Niels Obel 1, Jan Gerstoft 1 1 Department of Infectious Diseases, Copenhagen University Hospital, Rigshospitalet, Denmark 2 Department of Infectious Diseases, Copenhagen University Hospital, Hvidovre, Denmark 3 Department of Infectious Diseases, Aarhus University Hospital, Aarhus, Denmark 4 Department of Infectious Diseases, Aalborg University Hospital, Aalborg, Denmark 5 Department of Infectious Diseases, Odense University Hospital, Odense, Denmark 6 Department of Infectious Diseases, Hillerød Hospital, Hillerød, Denmark *Corresponding author mhelleberg@sund.ku.dk Background: We hypothesized that rates and reasons for treatment modifications have changed since the implementation of combination antiretroviral therapy (cart) due to improvements in therapy. Methods: From a nationwide population-based cohort study we identified all HIV-1-infected adults who initiated cart in Denmark and were followed 1 year. Incidence rate ratios (IRRs) and reasons for treatment modifications were estimated and compared between patients, who initiated treatment in , and Rates of discontinuation of individual antiretroviral drugs (ARVs) were evaluated. Results: A total of 3,107 patients were followed for a median of 7.3 years (IQR ). Rates of first treatment modification 1 year after cart initiation did not change (IRR 0.88 [95% CI 0.78, 1.01] and 1.03 [95% CI 0.90, 1.18] in and , respectively, compared with ). Rates of multiple modifications decreased markedly ( IRR 0.60 [95% CI 0.53, 0.67] and IRR 0.38 [95% CI 0.32, 0.46]). Rates of treatment modifications due to virological failure, toxicity and other/unknown reasons decreased (IRR 0.25 [95% CI 0.14, 0.45], 0.69 [95% CI 0.56, 0.83] and 0.45 [95% CI 0.36, 0.57], respectively, in compared with ), whereas the rate of modifications with the aim of simplification increased (IRR 1.85 [95% CI 1.52, 2.25]). Conclusions: Rates of first treatment modification 1 year after cart initiation have not changed since the early cart era, whereas the risk of multiple modifications has decreased markedly. Modifications due to virological failure and toxicity have decreased substantially, whereas rates of switch to simpler and less toxic regimens have increased. Introduction HIV infection is often acquired at a young age and the life expectancy of individuals initiating combination antiretroviral therapy (cart) has approached that of the background population [1]. It is therefore important to find a treatment regimen with long-term durability in terms of efficacy, toxicity and acceptability. Antiretroviral therapy (ART) needs to be individualized according to comorbidities, resistance patterns, genetics, potential drug drug interactions and convenience to assure tolerability and adherence [2]. With the introduction of combination therapy, suppression of viral replication could be achieved, but many of the antiretroviral drugs (ARVs) used in the early ART era, caused debilitating short- and long-term side effects, such as nausea, diarrhoea, lipodystrophy and neuropathy. Additionally the large pill burden and frequent dosing was a challenge for optimal adherence. A significant proportion of patients, who initiated combination therapy in the early cart era had previously received mono- or dual therapy and developed resistance. In recent years several ARVs of different classes have been introduced, which offer high efficacy, less toxicity 2013 International Medical Press (print) (online) 345

2 M Helleberg et al. and ease administration in terms of reduced pill burden and dosing frequency [3]. Despite these improvements previous studies have shown no decrease in rates of treatment modification the first year after starting ART [4,5]. We hypothesized that, although early treatment modifications may be common, these changes usually lead to a well-tolerated and efficient regimen and further alterations in the regimen have likely decreased as treatment has become both better tolerated and more effective. To test this hypothesis we assessed changes in rates of and reasons for treatment modifications over a 13-year period in a nationwide cohort where HIV treatment has been successful, resulting in low incidence of drug resistance [6] and long life expectancy of HIV patients [1,7]. Trends in modifications of ARVs in initial regimen in the period were evaluated. Methods In a nationwide, population-based cohort study we estimated incidences of first, third and all treatment modifications among HIV patients who initiated cart in the period Third treatment modification was used as a surrogate marker of multiple modifications. Risk factors and reasons for first treatment modification were evaluated. Incidence rates (IRs) of discontinuation of ARVs in first regimen were compared between individuals who initiated treatment in versus and We evaluated modifications of first-line ARV regimens used in by estimating the rates of discontinuation of individual ARVs in each calendar year. Setting Denmark has a population of 5.5 million, with an estimated HIV prevalence of 0.09% among adults. HIV patients were during the study period seen as outpatients at intended intervals of 12 weeks. There are eight specialized HIV care centres in Denmark. Antiretroviral treatment is provided at the centres free of charge and ARVs approved by the European Medicines Agency are available in all regions of the country. Fixed-dose combinations of zidovudine, lamivudine and abacavir were approved in year 2000, of abacavir and lamivudine in 2004, of tenofovir and emtricitabine in 2005 and of tenofovir, emtricitabine and efavirenz in The national criteria for initiating ART have been described previously [8]. The national guidelines for recommended first-line regimens of ARVs and adherence to these have been evaluated by Petersen et al. [9]. Structured treatment interruptions have generally not been recommended in Denmark. Data sources The Danish HIV Cohort study, described in detail elsewhere [8], is a population-based prospective nationwide cohort study of all HIV-infected individuals who are treated at Danish HIV centres after 1 January Individuals are consecutively enrolled. Data are updated yearly and includes demographics, date of HIV infection, AIDS-defining events and date and cause of death. CD4 + T-cell counts and HIV RNA measurements are extracted electronically from laboratory data files. Dates of initiation and modifications of ARVs as well as reasons for treatment modifications are recorded. Only information on generic drug name and not on formulation is recorded. Study population We included all HIV-1-infected individuals in the Danish HIV Cohort Study who were 16 years at treatment initiation, were resident in Denmark when initiating cart in and were followed for at least 1 year after cart initiation. Definitions cart was defined as a treatment regimen of at least three ARVs that included a non-nucleoside reverse transcriptase inhibitor (NNRTI), a protease inhibitor (PI), three nucleoside reverse transcriptase inhibitors (NRTIs) or a two-drug regimen with a combination of an NNRTI and a boosted PI. Treatment modification was defined as discontinuation or switch of one or more ARVs. If several ARVs were switched at the same time it was considered as one modification. Reasons for treatment modification were categorized into four groups: virological failure, toxicity, simplification or other/unknown. Classification of virological failure and toxicity were based on medical files. If failure or toxicity were not stated as the reason and modification led to reduction in pill burden or decrease in dosing frequency, the reason was classified as simplification. Statistics We calculated IRs of first and third treatment modification per 100 person-years (PY). IRs were compared between individuals initiating treatment in versus and IRs and IR ratios (IRRs) were estimated using Poisson regression analyses. The following variables were included in multivariate analyses: gender, age (time-updated variable with 5-year intervals), origin, route of HIV transmission, body weight, CD4 + T-cell count (<200 cells/ml, cells/ml or >350 cells/ml), viral load, prior or current AIDS-defining illness at time of cart initiation and receipt of ART before cart. In analyses of first and third treatment modification we calculated time from date of cart initiation until date of first and third modification, respectively. In analyses of incidence of modifications by time after cart initiation time was International Medical Press

3 Decreasing rate of treatment modifications in an HIV cohort study calculated from start of cart until date of last visit in the HIV clinic. In analyses of reasons for first modification, follow-up time was censored when treatment was modified for reasons other than the one under current analysis. In analyses of time to modification of ARV class, time was computed from cart initiation until discontinuation or switch to a different class regardless of switches within class during that time interval. In analyses of time to modification of individual ARVs in first regimen, time was computed from cart initiation until discontinuation of the drug under study regardless of other modifications within that time interval. In all analyses follow-up time was censored at the date of last visit in the HIV clinic or 31 December 2010, whichever came first. IRs of discontinuation of individual ARVs were calculated for each calendar year. Only calendar years with observation time >25 PY were analysed. SPSS statistical software version 15.0 (Norusis; SPSS Inc., Chicago, IL, USA) and Stata 8.0 (Stata Corporation, College Station, TX, USA) were used for data analyses. Results We included 3,107 individuals, who were followed for 22,549 PY with a median follow-up time of 7.3 years (IQR ). The majority of study participants were male (74.1%), of Danish origin (71.7%) and infected through homosexual (44.9%) or heterosexual contact (40.6%; Table 1). The proportion of individuals with CD4 + T-cell count <200 cells/ml at cart initiation decreased from 43.8% in to 32.4% in The proportion of individuals who had received ART before cart was 41.4% in versus 1.6% in Among individuals initiating cart in , 1,057 (86.7%) received a PI and 22 (1.8%) received an NNRTI, whereas the corresponding numbers were 152 (18.7%) and 624 (77.0%) in First and third treatment modifications Overall 1,456 (46.7%) modified treatment within 1 year of cart initiation. There were no significant differences in time to first treatment modification within 1 year of cart initiation between individuals initiating treatment in the three periods (Figure 1A). In adjusted analyses IRRs were 0.88 (95% CI 0.78, 1.01) and 1.03 (95% CI 0.90, 1.18) in and , respectively, compared with More than 1 year after cart initiation the rate of first modification was significantly lower among individuals initiating cart in and compared with (IRR 0.90 [95% CI 0.79, 1.00] and 0.42 [95% CI 0.33, 0.53]). The risk of third treatment modification also decreased markedly (Figure 1B). IRRs of third modification were 0.64 (95% CI 0.56, 0.72) in and 0.41 (95% CI 0.34, 0.51) in Treatment modifications in the first and following years after cart initiation The rate of all modifications decreased both with calendar year of and time since cart initiation. The number of treatment modifications per 100 PY within 1 year of cart initiation decreased only modestly from IR 84.7 (95% CI 79.5, 90.1) to 76.5 (95% CI 71.2, 82.2) and 74.1 (95% CI 68.0, 80.7) among individuals initiating treatment in , and , respectively. In the period 1 3 to years after cart initiation the IRs decreased substantially to 60.9 (95% CI 57.8, 64.1), 42.7 (95% CI 39.9, 45.7) and 27.1 (95% CI 24.2, 30.4) in , and , respectively. More than 3 years after cart initiation the IRs were 47.5 (95% CI 46.1, 48.9), 37.1 (95% CI 35.3, 38.9) and 28.6 (95% CI 23.6, 34.6) in , and , respectively. Reasons for treatment modification The overall IRs per 100 PY of first treatment modification due to virological failure, toxicity, simplification and other/unknown reasons were 3.0 (95% CI 2.8, 3.2), 17.7 (95% CI 16.6, 18.9), 15.4 (95% CI 14.4, 16.5) and 13.4 (95% CI 12.5, 14.4), respectively. IRs by year of cart initiation are summarized in Additional file 1. The risk of first treatment modification due to virological failure was substantially lower among individuals initiating treatment in and compared with (Table 2). The risk of first treatment modification due to toxicity only decreased in the most recent period, whereas modifications with the aim of simplification increased. Rates of first modification due to other/unknown reasons decreased both among individuals initiating treatment in and Factors associated with treatment modifications Young age, baseline CD4 + T-cell count <200 cells/ml and ART before HAART were associated with increased risk of modification due to treatment failure (Table 2). The risk of modification due to toxicity was increased among females and those with low body weight. Baseline CD4 + T-cell count <200 cells/ml was associated with increased likelihood of modification with the aim of simplification, whereas young age and injection drug use were associated with increased risk of modification of other/unknown reasons. Modifications of ARVs in initial regimen by year of cart initiation Compared with individuals receiving an initial regimen of boosted PI+NRTI, rates of treatment modifications were Antiviral Therapy

4 M Helleberg et al. lower among those receiving NNRTI+NRTI and higher among those receiving unboosted PI+NRTI (Table 3). NRTI backbones of stavudine + didanosine and stavudine + lamivudine were discontinued at significantly higher rates than other NRTIs. The likelihood of discontinuation of zidovudine + lamivudine increased substantially Table 1. Characteristics of the study population and ARVs in initial regimen by year of cart initiation Year of cart initiation Characteristic All Total, n 1,219 1, ,107 Person-years 12,598 7,594 2,357 22,549 Treatment modifications, n 6,804 3,355 1,020 11,179 Median observation time, years (IQR) 11.4 ( ) 7.3 ( ) 2.8 ( ) 7.3 ( ) Male, n 938 (77.0) 758 (70.3) 605 (74.7) 2,301 (74.1) Median age at baseline, years (IQR) 38 (32 46) 38 (32 45) 39 (32 47) 38 (32 45) Origin Denmark 925 (75.9) 732 (68.1) 567 (70.0) 2,224 (71.7) Africa 142 (11.7) 186 (17.3) 109 (13.5) 437 (14.1) Asia 42 (3.5) 78 (7.3) 56 (6.9) 176 (5.7) Other 110 (9.0) 79 (7.4) 78 (9.6) 267 (8.6) Route of transmission Homosexual 594 (48.7) 403 (37.4) 397 (49.0) 1,394 (44.9) Heterosexual 431 (35.4) 506 (46.9) 325 (40.1) 1,262 (40.6) Injection drug use 140 (11.5) 101 (9.4) 52 (6.4) 293 (9.4) Other 54 (4.4) 68 (6.3) 36 (4.4) 158 (5.1) Median CD4 + T-cell count at baseline, 219 ( ) 207 ( ) 240 ( ) 220 ( ) cells/ml (IQR) CD4 + T-cell count <200 cells/ml 534 (43.8) 503 (46.7) 262 (32.4) 1,299 (41.8) Median viral load at baseline, 4.6 ( ) 4.9 ( ) 4.7 ( ) 4.8 ( ) log 10 copies/ml (IQR) AIDS at baseline 249 (20.4) 195 (18.1) 108 (13.3) 552 (17.8) ART before cart 505 (41.4) 58 (5.4) 13 (1.6) 576 (18.5) ARVs in initial regimen ARV class Boosted PI + NRTI 212 (17.4) 167 (15.5) 150 (18.5) 529 (17.0) Unboosted PI + NRTI 845 (69.3) 64 (5.9) 2 (0.2) 697 (22.4) NNRTI + NRTI 22 (1.8) 698 (64.7) 624 (77.0) 1,344 (43.3) PI + NNRTI + NRTI 93 (7.6) 42 (3.9) 0 (0.0) 135 (4.3) Triple NRTI 46 (3.8) 80 (7.4) 1 (0.1) 127 (4.1) NRTI backbone Zidovudine + 3TC 905 (74.2) 909 (84.3) 301 (37.2) 2,115 (68.1) Stavudine + 3TC 183 (15.0) 22 (2.0) 2 (0.2) 207 (6.7) Stavudine + didanosine 51 (4.2) 48 (4.5) 0 (0.0) 99 (3.2) Abacavir + 3TC 15 (1.2) 109 (10.1) 142 (17.5) 266 (8.6) Tenofovir + 3TC/FTC 0 (0.0) 11 (1.0) 347 (42.8) 358 (11.5) Other 1 (0.1) 27 (2.5) 33 (4.1) 61 (2.0) Third ARV Ritonavir 214 (17.4) 1 (0.1) 0 (0.0) 215 (6.9) Indinavir 367 (30.1) 10 (9.3) 1 (0.1) 378 (12.2) Indinavir + ritonavir 8 (0.7) 42 (3.9) 0 (0.0) 50 (1.6) Saquinavir 169 (13.9) 1 (0.1) 0 (0.0) 170 (5.5) Saquinavir + ritonavir 206 (16.9) 85 (7.9) 1 (0.1) 292 (9.4) Nelfinavir 189 (15.5) 88 (8.2) 0 (0.0) 277 (8.9) Lopinavir + ritonavir 7 (0.6) 66 (6.1) 89 (11.0) 162 (5.2) Atazanavir + ritonavir 0 (0.0) 7 (0.7) 71 (8.8) 78 (2.5) Efavirenz 16 (1.3) 691 (64.1) 593 (73.2) 1,300 (41.8) Nevirapine 99 (8.1) 76 (7.1) 46 (5.7) 221 (7.1) Data are n (%) unless indicated otherwise. ART, antiretroviral therapy; ARV, antiretroviral drug; FTC, emtricitabine; cart, combination antiretroviral therapy; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; 3TC, lamivudine International Medical Press

5 Decreasing rate of treatment modifications in an HIV cohort study Figure 1. Kaplan Meier curves showing time to first and third treatment modification Cumulative probability of first treatment modification A B Years after HAART initiation Years after HAART initiation Cumulative probability of third treatment modification Kaplan Meier curves showing time to (A) first and (B) third treatment modification, stratified by year of combination antiretroviral therapy initiation ( , light grey; , dark grey; , black). Table 2. Incidence rate ratios of first treatment modification stratified by reason a IRR (95% CI) Failure Toxicity Simplification Other/unknown Male Female 1.03 (0.61, 1.73) 1.25 (1.03, 1.51) 1.10 (0.90, 1.35) 1.18 (0.95, 1.45) Age, per 5 year increase 0.98 (0.97, 1.00) 1.00 (0.99, 1.01) 1.00 (0.99, 1.01) 0.98 (0.97, 0.99) Origin Denmark Africa 1.36 (0.75, 2.47) 0.89 (0.71, 1.12) 1.00 (0.79, 1.27) 0.92 (0.71, 1.20) Asia 1.56 (0.75, 3.28) 0.90 (0.67, 1.21) 0.79 (0.56, 1.12) 0.71 (0.49, 1.05) Other 0.60 (0.31, 1.20) 0.92 (0.72, 1.17) 1.03 (0.80, 1.33) 0.87 (0.66, 1.14) Route of transmission Homosexual Heterosexual 0.64 (0.40, 1.02) 0.93 (0.78, 1.10) 0.92 (0.77, 1.11) 0.96 (0.78, 1.18) Injection drug use 0.79 (0.44, 1.43) 0.78 (0.60, 1.02) 0.77 (0.56, 1.05) 2.34 (1.86, 2.94) Other 0.48 (0.19, 1.24) 0.93 (0.69, 1.25) 0.85 (0.61, 1.18) 0.96 (0.65, 1.41) CD4 + T-cell count at baseline <200 cells/ml 1.52 (1.05, 2.19) 1.03 (0.89, 1.20) 1.23 (1.04, 1.44) 0.97 (0.81, 1.15) cells/ml >350 cells/ml 1.09 (0.67, 1.77) 1.16 (0.97, 1.39) 1.17 (0.96, 1.44) 1.20 (0.97, 1.48) Viral load, per log increase 1.04 (0.92, 1.18) 0.99 (0.94, 1.05) 1.00 (0.94, 1.05) 1.02 (0.96, 1.08) AIDS at baseline 0.92 (0.61, 1.38) 1.16 (0.98, 1.38) 0.83 (0.69, 1.01) 1.08 (0.88, 1.32) ART before cart 3.45 (2.30, 5.16) 0.92 (0.76, 1.11) 1.01 (0.80, 1.26) 0.77 (0.63, 0.96) Body weight, per kg increase 1.00 (0.99, 1.02) 0.99 (0.98, 0.99) 1.00 (0.99, 1.01) 0.99 (0.98, 1.00) Year of cart initiation (0.27, 0.67) 1.06 (0.91, 1.24) 1.17 (0.99, 1.40) 0.55 (0.46, 0.65) (0.22, 0.78) 0.66 (0.54, 0.82) 1.85 (1.50, 2.28) 0.41 (0.33, 0.53) a Multivariate Poisson regression analyses. ART, antiretroviral therapy; cart, combination antiretroviral therapy; IRR, incidence rate ratio. Antiviral Therapy

6 M Helleberg et al. Table 3. Incidence rates and incidence rate ratios of discontinuation of ARVs in initial regimen stratified by year of cart initiation a Year of cart initiation, IRR (95% CI) IR (95% CI) per 100 PY ARV class Boosted PI + NRTI 24.8 (22.5, 27.5) (1.1, 1.8) 1.9 (1.4, 2.5) Unboosted PI + NRTI 37.2 (34.5, 40.1) 2.2 (1.8, 2.6) 4.4 (3.2, 6.0) NA NNRTI + NRTI 11.0 (10.1, 12.0) 0.4 (0.2, 0.7) 0.4 (0.4, 0.5) 1.2 (1.0, 1.5) PI + NNRTI + NRTI 34.4 (29.0, 40.8) 1.8 (1.4, 2.3) 2.3 (1.7, 3.3) NA Triple NRTI 26.6 (22.1, 32.1) 1.2 (0.9, 1.7) 1.3 (0.9, 1.7) NA NRTI backbone Zidovudine + 3TC 27.7 (26.5, 29.0) (2.1, 2.6) 23.2 (20.1, 26.8) Stavudine + 3TC 29.8 (25.9, 34.2) 1.8 (1.5, 2.2) 16.2 (10.5, 24.8) NA Stavudine + didanosine 43.7 (24.8, 76.9) 4.9 (3.7, 6.7) 5.8 (4.3, 7.8) NA Abacavir + 3TC 15.9 (13.4, 18.9) 0.6 (0.4, 1.1) 1.0 (0.8, 1.3) 1.8 (1.3, 2.5) Tenofovir + 3TC/FTC 4.9 (3.6, 6.7) NA 0.8 (0.3, 1.7) 0.6 (0.4, 0.9) Third ARV Efavirenz 14.8 (13.6, 16.0) (0.7, 2.5) 2.5 (1.3, 4.7) Nevirapine 24.8 (21.4, 28.9) 3.0 (1.6, 5.7) 2.0 (1.0, 3.9) 1.9 (1.0, 4.11) Indinavir 31.9 (29.0, 35.2) 3.1 (1.6, 5.6) 6.7 (3.4, 13.4) NA Saquinavir + ritonavir 37.5 (34.1, 41.2) 3.8 (2.0, 7.1) 4.8 (2.6, 9.1) NA Nelfinavir 43.0 (36.6, 50.6) 3.6 (1.9, 7.0) 8.0 (4.0, 16.0) NA Lopinavir + ritonavir 37.1 (30.1, 45.8) 1.6 (0.5, 4.7) 3.2 (1.5, 6.6) 5.8 (2.9, 11.5) Atazanavir + ritonavir 13.2 (8.9, 19.7) NA 0.8 (0.2, 3.6) 1.7 (0.8, 3.7) a Multivariate Poisson regression analyses. ARV, antiretroviral drug; cart, combination antiretroviral therapy; FTC, emtricitabine; IR, incidence ratio; IRR, incidence rate ratio; NA, not applicable due to limited observation time; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; PY, person-years; 3TC, lamivudine. among individuals initiating treatment in compared with The IRR of discontinuation of abacavir + lamivudine also increased in , whereas that of tenofovir + lamivudine/emtricitabine was low throughout the study period. Switch from zidovudine + lamivudine or abacavir + lamivudine to coformulated tenofovir + emtricitabine accounted for 72% of switches of lamivudine. Rates of discontinuation of efavirenz were lower than those of nevirapine and PIs other than atazanavir, but increased among individuals initiating treatment in compared with Modifications of ARVs by calendar year Rates of modification of regimens based on boosted PI+NRTI and NNRTI+NRTI decreased significantly over calendar time, whereas those of unboosted PI+NRTI increased (Figure 2A). First-line regimens with an NRTI backbone of stavudine + lamivudine and zidovudine + lamivudine were largely abandoned from 2001 and 2004, respectively (Figure 2B). The IR of discontinuation of abacavir + lamivudine was stable during the study period, whereas that of tenofovir + lamivudine/emtricitabine decreased to a very low level. The IRs of discontinuation of lopinavir + ritonavir, efavirenz and nevirapine decreased during the study period, whereas those of unboosted saquinavir, nelfinavir and indinavir increased significantly (Figure 2C). IRs of discontinuation of atazanavir were similar to efavirenz and nevirapine (Additional file 1). Discussion We found that although the rate of treatment modifications within the first year of cart initiation has not changed, the risks of later and multiple treatment modifications have decreased markedly. The rates of modifications due to virological failure, toxicity and other reasons have dropped substantially, whereas the rate of simplifications has increased. We believe this is the first study demonstrating decreasing rates of treatment modifications since the early cart era. However, individuals who initiated treatment in the most recent period on average still modified treatment once every fourth year in the period more than 1 year after cart initiation. Several studies have evaluated the proportions, rates of and reasons for treatment modifications the first year after cart initiation [4,5,10 15]. In these studies the proportion of individuals, who switched or discontinued ARVs within the first year of treatment were 30 50%, which is in-line with our results. Only few have analysed the long-term durability of ARVs in daily clinical practice or estimated the rate of multiple modifications. Toxicity was the main reason for treatment modification and female gender and low body weight were associated with increased risk of treatment modification due International Medical Press

7 Decreasing rate of treatment modifications in an HIV cohort study Figure 2. Incidence rates of discontinuation of ARV class, NRTI backbone and third ARV in initial regimen by calendar year IR/100 PY A Boosted PI Unboosted PI NNRTI Triple NRTI to toxicity, which has also been demonstrated previously [5,16]. During the study period the risk of treatment modifications due to toxicity decreased, presumably reflecting the favourable safety profile of newer ARVs. We found no association between CD4 + T-cell count at cart initiation and risk of modification due to toxicity. In the Swiss HIV Cohort, higher CD4 + T-cell counts were associated with increased risk of modifications due to all reasons as well as toxicity [4,15,17], whereas others found no association [10,11,14]. In the present study, low CD4 + T-cell counts at cart initiation was associated with increased risk of modification due to treatment failure, which is in agreement with findings from other studies [5,12,18]. As we have previously shown, receiving ART before cart was strongly associated with increased risk of treatment modification due to failure [19]. Our finding of a markedly decreasing risk of treatment failure has also been documented in previous cohort studies [20,21]. C IR/100 PY B IR/100 PY Zidovudine + 3TC Stavudine + 3TC Stavudine + ddi Abacavir + 3TC Tenofovir + 3TC/FTC Ritonavir Saquinavir + rtv Indinavir Atazanavir + rtv Nevirapine Saquinavir Nelfinavir Lopinavir + rtv Efavirenz Incidence rates (IRs) of discontinuation of (A) antiretroviral drug (ARV) class, (B) nucleoside reverse transcriptase inhibitor (NRTI) backbone and (C) third ARV in initial regimen by calendar year. ddi, didanosine; FTC, emtricitabine; NNRTI, non-nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; PY, person-years; rtv, ritonavir; 3TC, lamivudine. The question of when to start ART is controversial and arguments for deferral of therapy in individuals with high CD4 + T-cell counts include concerns of long-term toxicity of ART, non-adherence and development of resistance. Findings in this study point to the increased risk of failure when treatment is initiated at low CD4 + T-cell count and to the favourable durability of newer ARVs. One single adjustment of first-line cart regimen is still needed up front for many individuals in order to find a regimen that is tolerated but multiple modifications are increasingly infrequent as well as modifications due to virological failure. ARVs available early in the study period caused considerable side effects, and rates of discontinuation of unboosted PIs, ritonavir and stavudine were high. A substantial increase in discontinuation rates of these drugs coincided with the introduction of ARVs with fewer side effects. With the development of new classes of ARVs and potent drugs, regimens that are Antiviral Therapy

8 M Helleberg et al. acceptable and efficacious even in the presence of drug resistance have become widely accessible. The convenience of coformulated tablets and drugs that can be administered once daily reduces suboptimal adherence [22,23]. Coformulated tablets of NRTIs were introduced in 2000 (zidovudine + lamivudine + abacavir), 2004 (abacavir + lamivudine) and 2005 (tenofovir + emtricitabine). Concurrently with these improvements in convenience and efficacy of antiretroviral therapy the rate of modifications with the aim of simplification has increased while the rate of modifications due to virological failure has decreased. In many resource-replete countries drug resistance testing has become routine before ART initiation to reduce the risk of virological failure [24], but since the prevalence of transmitted drug resistance among HIV patients in the Danish HIV Cohort is only approximately 6% [25], this has not had major impact on the overall rate of early treatment modifications. Physicians may decide to make proactive switches to prevent development of side effects. Emerging evidence for long-term toxicity, for example, lipoatrophy associated with stavudine and zidovudine may partly explain the increasing rates of discontinuation of these drugs in the first years after the turn of the millennium [26]. Data from the ACTG 5202 study, disseminated in 2008, showed inferior virological response to abacavir compared with tenofovir in individuals with pretreatment HIV RNA>100,000 copies/ml [27]. The D:A:D study, also published in 2008, linked abacavir to increased risk of cardiovascular disease [28,29]. Although the latter association has been debated [30,31], these studies might have had some influence on the use of abacavir in the recent period and partly explain the slightly increasing rate of discontinuations among individuals initiating cart in and that the IR of discontinuations of abacavir + lamivudine was higher than that of tenofovir + lamivudine/ emtricitabine in 2009, although both drugs cause few side effects on the short term. However rates of discontinuation of abacavir were still low compared with those of zidovudine, we therefore posit that the above mentioned studies did not have a significant impact on the use of abacavir in Denmark. Most replacements of lamivudine were a consequence of switch to coformulated tenofovir + emtricitabine. There are few side effects of lamivudine and even In the presence of the M184 resistance mutation it may be beneficial to continue treatment because the mutation lowers HIV replication-fitness [32]. No drug has proved more efficacious than efavirenz for suppression of viral load in treatment-naive individuals, but in recent years several alternative drugs have been introduced, which are non-inferior to efavirenz regarding virological response [33 36]. The immediate central nervous system adverse effects of efavirenz were evident early after introduction of the drug [37, 38], but persistence of subtle psychiatric and neurological disturbances related to efavirenz may only have been fully recognized in recent years [39,40]. Even minor side effects are not acceptable when alternatives are available and the introduction of PIs and integrase inhibitors with few side effects may explain the increased rate of discontinuations of efavirenz among individuals initiating cart in Nevirapine was used in initial regimen for a low proportion of patients throughout the study period. Nevirapine has shown non-inferior virological activity compared with efavirenz [41] and atazanavir [42], but is not widely used due to the risk of severe hypersensitivity reactions [43]. Major strengths of the study are the nationwide, population-based design, which make the results generalizable to resource-replete settings. The study population was rather large, follow-up time was long and included a wide span of calendar time, which allowed us to analyse changing trends over time since the early HAART era. The study was conducted in a setting where HIV care is well-organized, ARVs are widely available and treatment outcomes are favourable regarding retention in care [44], drug resistance [6] and survival [1,7]. There are certain limitations of the study. Reasons for treatment modifications were determined by chart review and were not clearly stated in all cases. We classified the reason as simplification if no other reason was documented and the change led to a reduction in pill burden or dosing frequency. We only had information on generic drug name and not the formulation and thus we assumed that coformulated tablets were used after the date, when they were introduced. We may thereby have overestimated the proportion of modifications with the aim of simplification. Some treatment modifications may have been proactive to prevent future toxicity associated with long-term exposure to ARVs. We were not able to discern proactive switches from the category of other reasons. Due to limited observation time we could not evaluate the durability of amprenavir, fosamprenavir, tipranavir and maraviroc, which have never been used in any considerable amount in Denmark, and darunavir, etravirine and raltegravir, which were usually used for salvage therapy during the study period and have only recently been offered to treatment-naive patients. In analyses of individual ARVs we only evaluated durability of the initial regimen as the risk of modification of firstand second-line regimens are unlikely to be comparable. The study was observational and thus our results reflect the mixed effects of changes in guidelines, physician s opinions, characteristics of patients as well as properties of available ARVs International Medical Press

9 Decreasing rate of treatment modifications in an HIV cohort study We conclude that although rates of first treatment modification within 1 year of cart initiation have not changed since the early cart era, the risks of later and multiple treatment modifications have decreased substantially. Rates of modifications due to virological failure and toxicity have decreased markedly, whereas rates of switch to simpler regimens have increased. In the recent period where potent ARVs with few side effects have become available, identifying a treatment regimen with long-term durability is feasible but may involve an early switch. Acknowledgements We thank the staff of our clinical departments for their continuous support and enthusiasm. All of the authors contributed to the conception and design of the study and/or the analyses and interpretation of the data. The manuscript was drafted by MH, JG and NO and was critically reviewed and subsequently approved by all authors. The centres in the Danish HIV Cohort Study are: the Departments of Infectious Diseases at Copenhagen University Hospitals, Rigshospitalet (JG and NO) and Hvidovre (GK), Odense University Hospital (CP), Aarhus University Hospitals, Skejby (CSL) and Aalborg (GP), Herning Hospital (ALL), Hillerød Hospital (LN) and Kolding Hospital (Janne Jenson). Disclosure statement NO has received research funding from Roche, Bristol Myers Squibb, Merck Sharp & Dohme, GlaxoSmithKline, Abbott, Boehringer Ingelheim, Janssen Cilag and Swedish Orphan. CP has received research funding from Abbott, Merck Sharp & Dohme, GlaxoSmithKline and Janssen Pharma/Tibotec. JG has received research funding from Abbott, Roche, Bristol Myers Squibb, Merck Sharp & Dohme, ViiV, Swedish Orphan, Janssen Pharma/Tibotec and Gilead. ALL is on the advisory board for Bristol Myers Squibb, GlaxoSmithKline and Janssen Pharma/Tibotec and has received research funding from Abbott. MH, GK, CSL, GP and LN declare no competing interests. No funding sources were involved in study design, data collection, analysis, report writing or the decision to submit the paper for publication. Additional file Additional file 1: Rates of discontinuation (IR/100 person- years [95% CIs]) in initial regimen can be found at uploads/documents/avt-12-oa-2611_helleberg_ Add_file1.pdf References 1. Lohse N, Hansen AB, Pedersen G, et al. Survival of persons with and without HIV infection in Denmark, Ann Intern Med 2007; 146: Panel on 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. October 14, 2011; (Accessed 18 October 2011.) Available from aidsinfo.nih.gov/contentfiles/adultandadolescentgl.pdf 3. Camacho R, Teofilo E. Antiretroviral therapy in treatmentnaive patients with HIV infection. Curr Opin HIV AIDS 2011; 6 Suppl 1:S3 S Vo TT, Ledergerber B, Keiser O, et al. Durability and outcome of initial antiretroviral treatments received during by patients in the Swiss HIV Cohort Study. J Infect Dis 2008; 197: Cicconi P, Cozzi-Lepri A, Castagna A, et al. Insights into reasons for discontinuation according to year of starting first regimen of highly active antiretroviral therapy in a cohort of antiretroviral-naive patients. HIV Med 2010; 11: Audelin AM, Lohse N, Obel N, Gerstoft J, Jorgensen LB. The incidence rate of HIV type-1 drug resistance in patients on antiretroviral therapy: a nationwide population-based Danish cohort study Antivir Ther 2009; 14: Obel N, Omland LH, Kronborg G, et al. Impact of non-hiv and HIV risk factors on survival in HIV-infected patients on HAART: a population-based nationwide cohort study. PLoS ONE 2011; 6:e Obel N, Engsig FN, Rasmussen LD, Larsen MV, Omland LH, Sorensen HT. Cohort profile: the Danish HIV cohort study. Int J Epidemiol 2009; 38: Petersen TS, Andersen SE, Gerstoft J, et al. Adherence to national guidelines for initiation of antiretroviral regimens in HIV patients: a Danish nationwide study. Br J Clin Pharmacol 2011; 72: Mocroft A, Youle M, Moore A, et al. Reasons for modification and discontinuation of antiretrovirals: results from a single treatment centre. AIDS 2001; 15: Mocroft A, Phillips AN, Soriano V, et al. Reasons for stopping antiretrovirals used in an initial highly active antiretroviral regimen: increased incidence of stopping due to toxicity or patient/physician choice in patients with hepatitis C coinfection. AIDS Res Hum Retroviruses 2005; 21: O Brien ME, Clark RA, Besch CL, Myers L, Kissinger P. Patterns and correlates of discontinuation of the initial HAART regimen in an urban outpatient cohort. J Acquir Immune Defic Syndr 2003; 34: Hart E, Curtis H, Wilkins E, Johnson M. National review of first treatment change after starting highly active antiretroviral therapy in antiretroviral-naive patients. HIV Med 2007; 8: Braithwaite RS, Kozal MJ, Chang CC, et al. Adherence, virological and immunological outcomes for HIV-infected veterans starting combination antiretroviral therapies. AIDS 2007; 21: Elzi L, Marzolini C, Furrer H, et al. Treatment modification in human immunodeficiency virus-infected individuals starting combination antiretroviral therapy between 2005 and Arch Intern Med 2010; 170: Kirk O, Gerstoft J, Pedersen C, et al. Low body weight and type of protease inhibitor predict discontinuation and treatment-limiting adverse drug reactions among HIV-infected patients starting a protease inhibitor regimen: consistent results from a randomized trial and an observational cohort. HIV Med 2001; 2: Hänsel A, Bucher HC, Nuesch R, Battegay M. Reasons for discontinuation of first highly active antiretroviral therapy in a cohort of proteinase inhibitor-naive HIV-infected patients. J Acquir Immune Defic Syndr 2001; 26: Antiviral Therapy

10 M Helleberg et al. 18. Braithwaite RS, Kozal MJ, Chang CC, et al. Adherence, virological and immunological outcomes for HIV-infected veterans starting combination antiretroviral therapies. AIDS 2007; 21: Lohse N, Obel N, Kronborg G, et al. Declining risk of triple-class antiretroviral drug failure in Danish HIVinfected individuals. AIDS 2005; 19: Phillips AN, Leen C, Wilson A, et al. Risk of extensive virological failure to the three original antiretroviral drugs classes over long-term follow-up from the start of therapy in patients with HIV infection: an observational cohort study. Lancet 2007; 370: PLATO II project team. COHERE Group. Trends in virological and clinical outcomes in individuals with HIV-1 infection and virological failure of drugs from three antiretroviral drug classes: a cohort study. Lancet Infect Dis 2012; 12: Juday T, Gupta S, Grimm K, Wagner S, Kim E. Factors associated with complete adherence to HIV combination antiretroviral therapy. HIV Clin Trials 2011; 12: Boyle BA, Jayaweera D, Witt MD, Grimm K, Maa JF, Seekins DW. Randomization to once-daily stavudine extended release/lamivudine/efavirenz versus a more frequent regimen improves adherence while maintaining viral suppression. HIV Clin Trials 2008; 9: Wittkop L, Gunthard HF, de Wolf F, et al. Effect of transmitted drug resistance on virological and immunological response to initial combination antiretroviral therapy for HIV (EuroCoord-CHAIN joint project): a European multicohort study. Lancet Infect Dis 2011; 11: Audelin AM, Gerstoft J, Obel N, et al. Molecular phylogenetics of transmitted drug resistance in newly diagnosed HIV Type 1 individuals in Denmark: a nation-wide study. AIDS Res Hum Retroviruses 2011; 27: van der Valk M, Casula M, Weverlingz GJ, et al. Prevalence of lipoatrophy and mitochondrial DNA content of blood and subcutaneous fat in HIV-1-infected patients randomly allocated to zidovudine- or stavudine-based therapy. Antivir Ther 2004; 9: Sax P, Tierney C, Collier A, et al. ACTG 5202: shorter time to virologic failure (VF) with abacavir/lamivudine (ABC/3TC) than tenofovir/emtricitabine (TDF/FTC) as part of combination therapy in treatment-naive subjects with screening HIV RNA 100,000 c/ml. 17th International AIDS Conference. 3 8 August 2008, Mexico City, Mexico. Abstract THAB Sabin C, Worm S, Weber R. Do thymidine analogues, abacavir, didanosine and lamivudine contribute to the risk of myocardial infarction? The D:A:D study. 15th Conference on Retroviruses and Opportunistic Infections, 3 6 February 2008, Boston, MA, USA. Abstract 957C. 29. Sabin CA, Worm SW, Weber R, et al. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D:A:D study: a multi-cohort collaboration. Lancet 2008; 371: Brothers CH, Hernandez JE, Cutrell AG, et al. Risk of myocardial infarction and abacavir therapy: no increased risk across 52 GlaxoSmithKline-sponsored clinical trials in adult subjects. J Acquir Immune Defic Syndr 2009; 51: Cruciani M, Zanichelli V, Serpelloni G, et al. Abacavir use and cardiovascular disease events: a meta-analysis of published and unpublished data. AIDS 2011; 25: Turner D, Brenner BG, Routy JP, Petrella M, Wainberg MA. Rationale for maintenance of the M184v resistance mutation in human immunodeficiency virus type 1 reverse transcriptase in treatment experienced patients. New Microbiol 2004; 27 Suppl 1: Vieira MC, Kumar RN, Jansen JP. Comparative effectiveness of efavirenz, protease inhibitors, and raltegravir-based regimens as first-line treatment for HIVinfected adults: a mixed treatment comparison. HIV Clin Trials 2011; 12: Squires K, Lazzarin A, Gatell JM, et al. Comparison of once-daily atazanavir with efavirenz, each in combination with fixed-dose zidovudine and lamivudine, as initial therapy for patients infected with HIV. J Acquir Immune Defic Syndr 2004; 36: Markowitz M, Nguyen BY, Gotuzzo E, et al. Rapid and durable antiretroviral effect of the HIV-1 Integrase inhibitor raltegravir as part of combination therapy in treatmentnaive patients with HIV-1 infection: results of a 48-week controlled study. J Acquir Immune Defic Syndr 2007; 46: Ortiz R, DeJesus E, Khanlou H, et al. Efficacy and safety of once-daily darunavir/ritonavir versus lopinavir/ritonavir in treatment-naive HIV-1-infected patients at week 48. AIDS 2008; 22: Molina JM, Ferchal F, Rancinan C, et al. Once-daily combination therapy with emtricitabine, didanosine, and efavirenz in human immunodeficiency virus-infected patients. J Infect Dis 2000; 182: Adkins JC, Noble S. Efavirenz. Drugs 1998; 56: Clifford DB, Evans S, Yang Y, et al. Impact of efavirenz on neuropsychological performance and symptoms in HIVinfected individuals. Ann Intern Med 2005; 143: Maggiolo F. Efavirenz: a decade of clinical experience in the treatment of HIV. J Antimicrob Chemother 2009; 64: van Leth F, Phanuphak P, Ruxrungtham K, et al. Comparison of first-line antiretroviral therapy with regimens including nevirapine, efavirenz, or both drugs, plus stavudine and lamivudine: a randomised open-label trial, the 2NN Study. Lancet 2004; 363: Soriano V, Arasteh K, Migrone H, et al. Nevirapine versus atazanavir/ritonavir, each combined with tenofovir disoproxil fumarate/emtricitabine, in antiretroviral-naive HIV-1 patients: the ARTEN Trial. Antivir Ther 2011; 16: Baylor MS, Johann-Liang R. Hepatotoxicity associated with nevirapine use. J Acquir Immune Defic Syndr 2004; 35: Helleberg M, Engsig FN, Kronborg G, et al. Retention in a public health care system with free access to treatment: a Danish nationwide HIV cohort study. AIDS 2012; 26: Accepted 29 September 2012; published online 16 October International Medical Press