Epidemiology and Social
Declining risk of triple-class antiretroviral drug failure in Danish HIV-infected individuals
Lohse, Nicolaia; Obel, Nielsa; Kronborg, Gitteb; Laursen, Alexc; Pedersen, Courta; Larsen, Carsten Sd; Kvinesdal, Birgite; Sørensen, Henrik Toftf,g; Gerstoft, Janh
From the aOdense University Hospital, and University of Southern Denmark, Odense
bCopenhagen University Hospital Hvidovre, Hvidovre
cAarhus University Hospital Skejby, Aarhus
dAalborg Hospital, Aalborg
eHelsingør Hospital, Helsingør
fDepartment of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark
gDepartment of Epidemiology, Boston University, Boston, Massachusetts, USA
hCopenhagen University Hospital Rigshospitalet, Denmark.
Received 27 September, 2004
Revised 30 January, 2005
Accepted 14 February, 2005
Correspondence to Nicolai Lohse, Department of Infectious Diseases, Odense University Hospital, Sdr. Boulevard 29, DK-5000 Odense C, Denmark. E-mail: firstname.lastname@example.org
Objectives: To analyse the incidence, prevalence, and predictors for development of triple-class antiretroviral drug failure (TCF) in individuals infected with HIV.
Design: Population-based observational cohort study from 1 January 1995 to 31 December 2003, focusing on all 2722 recipients of highly active antiretroviral therapy (HAART) in Denmark.
Methods: We used person-years analysis, Kaplan–Meier survival curves and Cox regression analysis. TCF was defined as a minimum of 120 days with viral load > 1000 copies/ml on treatment with each of the three major drug classes.
Results: We observed 177 TCFs, yielding a crude incidence rate (IR) of 1.8 per 100 person-years [95% confidence interval (CI), 1.6–2.1]. Seven years after initiation of HAART, 17.2% (95% CI, 14.5–20.5) of antiretroviral (ART)-experienced patients, but only 7.0% (95% CI, 4.3–11.2) of ART-naive patients were estimated to have failed. After an initial rise, the IR from the third to the sixth year of HAART declined significantly for ART-experienced patients [incidence rate ratio (IRR), 0.80 per year (95% CI, 0.66–0.97); P = 0.022], and non-significantly for ART-naive patients [IRR, 0.79 per year (95% CI, 0.53–1.18); P = 0.255]. The IR for all patients being followed each year declined from 1997 to 2003 [IRR, 0.88 (95% CI, 0.81–0.96); P = 0.002]. The prevalence of TCF remained stable at less than 7% after 2000. Predictors of TCF at commencement of HAART were a CD4 cell count below 200, a previous AIDS-defining event, previous antiretroviral exposure, earlier year of HAART initiation, and young age.
Conclusions: The risk of TCF is declining in Denmark and the prevalence remains stable.
Highly active antiretroviral therapy (HAART) effectively suppresses viral load (VL) in individuals infected with HIV . HIV-related morbidity and mortality have decreased to less than one-fifth since HAART was introduced in 1996–1997 [2–5]. However, several controlled trials with strict intention-to-treat designs have indicated failure rates of 20–55% during the first year of HAART [6–8], but also in observational cohort studies the number of patients on HAART found to have detectable VL has been substantial [9,10]. High failure rates are associated both with regimens that have low efficacy and with poor patient compliance (often due to side effects), resulting in development of drug resistance. A high prevalence of resistance has been documented in the HIV Cost and Services Utilization Study (HCSUS) , and mathematical models predict a 42% prevalence of resistant virus in San Francisco by 2005 . When failure occurs during an initial regimen (recommended to include two drug classes), general practice is to include a new drug class in the next regimen [13–16], since full or partial cross-resistance within a class is the rule. Thus patients who fail their second or third regimen often harbour multi-resistant viruses that are unresponsive to the three major drug classes: nucleoside reverse transcriptase inhibitors (NRTIs), protease inhibitors (PIs), and non-nucleoside reverse transcriptase inhibitors (NNRTIs) [17,18]. Not all patients with triple class failure (TCF) experience virological failure and progressive disease ; however the majority of patients whose disease progresses despite HAART belong to the TCF group and this group is expected to grow in coming years . It would thus be of interest to analyse the incidence and prevalence of TCF, as the development in these parameters over time would give an indication of the proportion of patients who experience disease progression despite HAART. Recently the consequences of TCF have been delineated in a large study of 2488 patients with TCF , but little is known about the prevalence of TCF. The current study defines TCF according to detectable HIV-RNA despite treatment and does not include genotypic and phenotypic assays. We examine the incidence, prevalence, and predictors for TCF development in a cohort consisting of all Danish HIV-infected patients who received HAART.
The Danish HIV cohort study
The Danish HIV Cohort Study (which started as The HIV Cohort Study in Western Denmark) is a prospective, population-based cohort study of all HIV-positive individuals seen at Danish HIV clinics since 1 January 1995. Data collection and processing have been described in detail elsewhere [21,22]. The Danish health care system provides free tax-supported medical care and antiretroviral treatment to HIV-infected individuals. Treatment for HIV infection in Denmark is restricted to five departments of infectious diseases and four departments of internal medicine with one or more infectious disease specialists. The drugs are delivered to the patients directly from these hospital departments. HIV medicine cannot be prescribed by general practitioners or bought from Danish pharmacies. Patients treated at all HIV centres in Denmark have been enrolled in the cohort, and complete data on all patients seen in any of the centres since 1 January 1995 have been collected from patient files and entered into the database. As of 1 January 2004, the cohort consisted of 3974 patients, 2722 of whom received HAART. Use of the Danish 10-digit personal identification number enables the centres to avoid multiple registrations of the same patient, and to document whether patients lost to follow-up have died or moved outside the country. The database is regularly updated with information on prophylactic and antiretroviral treatment, development of opportunistic infections and other AIDS-defining illnesses, and laboratory values including VL, CD4 cell count and blood lipids. All patients who began HAART were eligible for the study. We had information on the date when the patient started antiretroviral monotherapy or dual treatment (ART).
The following definitions were used for the purposes of this study.
At least three antiretroviral drugs, including at least one PI or NNRTI, or abacavir.
Defined using the 1993 expanded European AIDS case definition, based on the clinical definition of AIDS developed by the US Centers for Disease Control and Prevention [23,24].
Defined as a VL of more than 1000 copies/ml/ for a total of 120 days (not necessarily successive) while receiving treatment with a given drug class. Periods of treatment interruptions, whether doctor or patient initiated, did not count as time with failure. VL was considered as being more than 1000 copies/ml only in time periods between two consecutive viral load measurements above 1000 copies/ml. Failure of a drug class could occur when it was administered alone or as part of a multi-drug regimen .
Time of TCF
The date the patient met the failure requirements for three drug classes. VL and CD4 cell counts at initiation of HAART were recorded as the last available measurement before this date, or the following measurement after this date up to 14 (VL) or 90 (CD4) days after.
Defined as any antiretroviral regimen that contains one or more PIs but not ritonavir, or a regimen containing ritonavir as the only PI.
We calculated time from start of HAART (baseline) to triple-class failure and summarized failure over time. Patients were followed from initiation of HAART, and were censored at death or at last visit to the clinic. We stratified data by calendar period and constructed Kaplan–Meier survival curves for each period. The use of a model with left truncation (delayed entry) [25,26] allowed us to estimate long-term, event-free survival from baseline (initiation of HAART), based on data from a short observation period (i.e. observation years 2001–2003). Log rank test was used to test for differences between curves. We estimated TCF incidence rates (IRs) using person-years analysis, stratified according to pre-baseline ART-experience, and either according to time since HAART initiation or according to calendar time. Point prevalence of TCF was estimated as the proportion of living patients with TCF among all patients receiving HAART on 1 July each year. Poisson regression was used to estimate trends in IRs and prevalence. We performed Cox’ regression analyses to find predictors for TCF, using time since initiation of HAART as the time scale, and time zero at baseline + 18 months to ensure a cohort truly at risk of an event (due to the conditions included in our event definition). The following four variables: pre-baseline ART-experience; pre-baseline AIDS-defining event; baseline CD4 cell count; and baseline VL were mutually examined for interaction by stratified analysis. Due to hereby observed variation in the risk estimates, the adjusted analysis was stratified into ART-naive and ART-experienced patients, whereas the other three variables were kept in the model. The following five variables were then entered into the adjusted model if they caused a 10% change in the point estimates of the effect of any of the three variables included in the model above: age, gender, mode of infection, race/ethnicity, and year of HAART initiation. In ART-experienced patients we additionally tested the variable: time since starting antiretroviral drugs. The proportional hazards assumption was checked using Schoenfeld's residuals . Seventy-three of 177 patients lacked measurement of VL at baseline. These patients were followed up in the same way as the other patients from time of initiating HAART, but they could not count any failure time until their first measured VL, as stated in our definition of failure. Seventy-two of these 73 patients started HAART before mid-1997. Viral loads were measured a median of 3.98 times a year [inter-quartile range (IQR), 3.3–4.7].
Approvals and permissions
The Danish Data Protection Agency approved the establishment of the cohort study. The study was not subject to approval by the ethics committee as the collection of data did not involve direct patient contact.
Incidence and prevalence of triple-class drug failure
Of the 3974 patients in the cohort, 2722 started HAART, and 1092 were exposed to all three drug classes. We observed 177 events (TCF) among patients initiating HAART during a median observation period of 3.7 years (IQR, 1.7–5.4) following treatment initiation, yielding an incidence rate (IR) of 1.8 per 100 person-years at risk (PYR) [95% confidence interval (CI), 1.6–2.1]. The IR reached a maximum of 2.9 (95% CI, 2.2–3.8) and 3.0 (95% CI, 2.2–4.1) per 100 PYR in the third and fourth year after HAART initiation, thereafter declining to 1.2 (95% CI, 0.6–2.5) in the sixth year (Fig. 1) [trend third to sixth year: IRR, 0.84 per year (95% CI, 0.70–1.00); P = 0.047]. For patients exposed to ART before HAART, the overall IR was 2.9 per 100 PYR (95% CI, 2.4–3.4), reaching a maximum of 4.6 (95% CI, 3.3–6.4) in the fourth year following HAART initiation, declining to 1.4 (95% CI, 0.6–3.2) in the sixth year [trend from third to sixth year: IRR, 0.80 per year (95% CI, 0.66–0.97); P = 0.022]. For ART-naive patients, the overall IR was 0.8 per 100 PYR (95% CI, 0.6–1.1). After a rise to 1.6 (95% CI, 0.9–2.7) in the third year following HAART initiation, the IR declined with no significant trend [third to sixth year: IRR, 0.79 per year (95% CI, 0.53–1.18); P = 0.255].
The IR per calendar year, for patients being followed up that year, peaked in 2000 at 3.7 (95% CI, 2.9–4.8) with declines to 1.6 (95% CI, 1.1–2.3) in 2001; 0.7 (95% CI, 0.4–1.3) in 2002, and 0.4 (95% CI, 0.2–1.1) in 2003. There was a significant declining trend in the IR from 1997 to 2003 [IRR, 0.88 per year (95% CI, 0.81–0.96); P = 0.002].
Seven years after baseline, 12.4% (95% CI, 10.4–14.8) of all patients; 17.3% (95% CI, 14.5–20.5) of ART-experienced patients; and 7.0% (95% CI, 4.3–11.2) of ART-naive patients were estimated to have failed. When the observations were split into three observation periods (1997–1998, 1999–2000 and 2001–2003), a lower risk of TCF was seen after 2001 than in previous years, regardless of when patients initiated HAART (log rank test for all pair wise combinations: P = 0.000) (Fig. 2). For all patients observed in the 2001–2003 period, the estimated 7-year cumulative incidence proportion was 6.8% (95% CI, 4.8–9.6). For ART-naive patients, the estimated 7-year cumulative incidence proportion in 2001–2003 was only 5.5% (95% CI, 2.9–10.1), corresponding to an IR of 0.6 (95% CI, 0.4–1.0) per 100 PYR.
The point prevalence of patients with TCF among patients on HAART, measured at 1 July each year, increased from 1997 to 2001 [increase 1.60 per year (95% CI, 1.46–1.76); P = 0.000], with a non-significant decrease observed from 2001 to 2003 [increase 0.94 per year (95% CI, 0.83–1.07; P = 0.356)] (Fig. 3).
Patients with triple-class drug failure
Individuals who developed TCF had a lower CD4 cell count at baseline than other patients [median, 75 (IQR, 26–170) versus 190 (80–300) × 106 cells/l]. A higher log10VL at baseline was observed only in ART-experienced [median, 4.9 (IQR, 4.3–5.5) versus 4.2 (IQR, 3.0–5.0)], but not in ART-naive [median, 5.2 (IQR, 4.2–5.5) versus 5.0 (IQR, 4.5–5.7)] patients (Table 1). Patients who experienced TCF were also more likely to have had a previous AIDS-defining event (46 versus 23%), to be ART-experienced (79 versus 37%), they were younger [median age, 36.8 (IQR, 30.6–44.2) versus 38.7 (IQR, 32.7–46.5) years], and they had initiated HAART earlier (median, November 1996 versus August 1998). Patients with and without TCF did not differ with regard to gender, mode of acquisition, or race. The last drug class to fail was NNRTI in 94% of the patients and PI in 6% of the patients. Ninety-eight percent of the patients had been exposed to an unboosted PI.
Predictors for development of triple-class drug failure
In crude analyses, an earlier year of HAART initiation, having a high viral load or a CD4 count below 200 × 106 cells/l at the time of HAART initiation, having had an AIDS-defining event, being ART-experienced, and being young were associated with a higher risk of TCF (Table 2). In the multivariate model, a high VL at the time of HAART initiation was a predictor in ART-experienced, but not in ART-naive patients (Table 2). The TCF risk among ART-naive patients who initiated HAART after 1999 was one-tenth of that faced by ART-naive patients with earlier HAART initiation, adjusted for other predictors.
In this study, based on a complete nationwide cohort of 2722 HIV-patients who initiated HAART, we found a declining IR of TCF after 2001, and a stable prevalence in the period 2000–2003. Thus we observed no signs of increasing failure incidence rates over time in a population in which HAART was offered to all those who were at risk of clinical progression. The IR of TCF differed substantially according to ART exposure before HAART, both in terms of absolute risk and in timing. Among the ART-experienced, a peak was observed 3–4 years after HAART initiation. The length of time for the IR to reach the maximum stems from the TCF definition, in which the occurrence of an event requires exposure to three drug classes and a minimum time period for virological failure to occur.
The decline in IR from the fifth year of HAART onwards may have several explanations. First, patients with prior ART experience who fail their first HAART are often cross-class resistant to both NRTI and PI and are at high risk of failure in their second regimen. Assuming that these cross-class-resistant patients experience TCF in the first years after initiating HAART, the remaining patients being followed up will have a risk of failure more close to that observed among the ART-naive. Other patients with high risk of failure, due to low adherence, could potentially produce a similar peak. However, the absence of a clear peak among the treatment-naive patients suggests that pre-existing resistance from a former regimen (as opposed to behavioural factors) mainly drives early failures among patients with a pre-HAART ART experience. Furthermore, the fifth year of HAART is by definition after 2000; it is possible that failure was less likely in this period due to better regimens or better patient coaching. In fact, stratification by observation period did provide evidence that patients fared better in 2001–2003 than during earlier periods of HAART, unrelated to date of HAART initiation.
Mocroft et al.  have studied the risk of TCF in a pan-European cohort study. They found a 6-year risk of TCF of 21.4% among ART-experienced, and 11.2% among ART-naive individuals with overall IRs of 1.6 and 3.9 cases/100 PYR in the two groups, respectively, which is higher than that observed in our study. Several explanations may be given for the differences in the IR of TCF observed in the two studies. First the possibility that the IR of TCF is lower in Denmark than in other parts of Europe cannot be excluded. This could be related to structural health care issues such as a high caseload per HIV-treating physician , high representation of academic centres, or the mode of drug delivery. Second, the lower risk could be related to the infected population, although the present study failed to identify mode of transmission as a risk factor. Third, our study was a nationwide cohort including all patients with ART experience in Denmark, making it possible to minimize recruitment bias.
We used a rather lax definition of TCF, because this definition approaches our clinical impression of failure, and because the same definition was used in another study  showing poor prognostic outcome for these patients.
The statistical analyses hinged on the definition of when a person became at risk of developing TCF. In common survival analysis an individual is arguably not at risk until after 119 days of failing the third drug class, but use of this definition would have excluded all but very high-risk patients. We therefore used a clinically more relevant time scale, starting when the patient was first exposed to HAART (and thereby potential drug failure), and we restricted the regression analyses to observations after 18 months, a reasonable time period to develop TCF. When the analyses were repeated using time of HAART initiation as the starting point, or restricted to patients exposed to three drug classes for a minimum of 120 days, we obtained similar significant results.
One striking observation is the decrease in failure rate according to year of HAART initiation; after controlling for confounding factors, initiating HAART during the early years carried a six-fold higher risk than initiating after 1999. There are a number of possible explanations. First, the strategy has changed from PI-based to NNRTI-based HAART over time. Second, although boosted PI therapy has been used from 1996 , it has not totally replaced unboosted PI treatment until after 2001 (data not shown). Studies suggest that boosted PIs are superior to unboosted PIs  as they produce higher drug levels, which in turn have been associated with better outcomes [31,32]. In this study, initiating therapy with an unboosted PI was associated significantly with TCF (data not shown). However, comparisons of drugs and regimens in cohort studies should be done with utmost caution, as biases are likely to be present. Third, the increasing number of drug combinations available for in-class substitutions could induce better individual patient tolerability. However the success of regimens containing NNRTIs, a drug class with very limited in-class substitution possibilities, suggests that this is not the sole reason. A fourth explanation for the improved outcome over time could be better patient coaching in more recent years, following recognition of adherence as the key factor for success of a regimen . This observed decrease in failure rate corresponds with recent years’ advances in HIV treatment. These recent advances should be kept in mind when setting success rates (not too low) for HIV treatment rollout programs in resource-limited settings.
The strength of our study was the completeness of the Danish HIV Cohort. Although HIV-infected individuals who are not on antiretroviral treatment may visit their general practitioner without inclusion in the cohort, all patients seen in a public HIV clinic are captured. The fact that antiretroviral drugs are distributed only through these clinics (where they are provided free of charge) ensures that almost all Danish HIV patients receiving ART are included in the cohort, eliminating the risk of selection bias.
Our study had several limitations. First, we did not have resistance data. Triple-class drug failure was defined by previously established criteria , where virological failure of a given magnitude (VL > 1000 copies/ml) during a given time period (120 days) was interpreted as a surrogate marker for drug failure. Resistance does not necessarily develop to all drugs in a regimen despite failure [7,34], and resistance might have been developed to all three major drug classes in patients not meeting our definition of TCF. Individuals with TCF did have difficulty attaining full viral suppression, however, as only 19% of viral load measurements after the time of TCF were below 400 copies/ml; this contrasts with 70% of VLs in other patients in the cohort after HAART initiation. Furthermore another study has revealed a high risk of death among patients with triple class failure . Thus we do believe that the majority of disease progression-related morbidity and mortality would occur in patients within this group, although we cannot exclude that a few patients receiving HAART will experience disease progression without fulfilling our criteria for TCF, for example, due to intolerance to a whole class.
Second, we had no information about patient compliance. Some patients may not have taken any of the prescribed drugs, and may have fulfilled our criteria without developing drug resistance.
Third, although HIV-RNA assays were initiated in 1995, they only became available for all patients from mid-1997 on, and we may have missed some early HAART failures. We did not include failure prior to HAART in the analyses, but used prior ART experience as an explanatory variable. NNRTIs were not available in Denmark before 1997 so we were only concerned about PI or NRTI failure before 1997. However, 167 of our TCF-patients failed NNRTI as the last drug class, and another seven initiated therapy with HAART after mid-1997, leaving only three patients (1.7%) that could have been observed with TCF at an earlier date. Had any of these been picked up earlier it would only have strengthened our finding of a declining IR. Of the 40 patients who met the criteria for NNRTI-failure, but did not develop TCF, only seven (0.3% of all non-TCF) had received HAART before mid-1997. The temporal distribution of these NNRTI failures was similar to that for TCFs in general, thus they are unlikely to have biased our results. Only failure with the three major drug classes was investigated, as fusion inhibitors have been used in less than 15 patients in Denmark. Our study was carried out during a time when surveillance revealed a yearly prevalence of transmitted resistance in the country between 0 and 5% . If transmitted resistance increases, the IR of TCF is likely to increase as well. On the other hand the low level of transmitted resistance could be a consequence of the low IR of TCF.
In conclusion, we found a declining IR of TCF in the Danish HIV-infected population. Importantly, the study provides evidence that high drug pressure on HIV at the population level does not inevitably result in increasing rates of drug failure.
Sponsorship: The study was supported by grants from the Danish AIDS Foundation; Odense University Hospital, Denmark; Preben and Anna Simonsen's Foundation; The Foundation of the Danish Association of Pharmacists; Clinical Institute, University of Southern Denmark.
1. Ledergerber B, Egger M, Opravil M, Telenti A, Hirschel B, Battegay M, et al
. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet 1999; 353:863–868.
2. Mocroft A, Ledergerber B, Katlama C, Kirk O, Reiss P, D'Arminio MA, et al
. Decline in the AIDS and death rates in the EuroSIDA study: an observational study. Lancet 2003; 362:22–29.
3. Egger M, May M, Chene G, Phillips AN, Ledergerber B, Dabis F, et al
. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360:119–129.
4. Jensen-Fangel S, Pedersen L, Pedersen C, Larsen CS, Tauris P, Møller A, et al
. Low mortality in HIV-infected patients starting HAART in advance of immunological deterioration: a comparison with the general population. AIDS 2004; 18:89–97.
5. Mocroft A, Vella S, Benfield TL, Chiesi A, Miller V, Gargalianos P, et al
. Changing patterns of mortality across Europe in patients infected with HIV-1. EuroSIDA Study Group. Lancet 1998; 352:1725–1730.
6. Gulick RM, Ribaudo HJ, Shikuma CM, Lustgarten S, Squires KE, Meyer WA III, et al
. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med 2004; 350:1850–1861.
7. Walmsley S, Bernstein B, King M, Arribas J, Beall G, Ruane P, et al
. Lopinavir-ritonavir versus nelfinavir for the initial treatment of HIV infection. N Engl J Med 2002; 346:2039–2046.
8. Bartlett JA, DeMasi R, Quinn J, Moxham C, Rousseau F. Overview of the effectiveness of triple combination therapy in antiretroviral-naive HIV-1 infected adults. AIDS 2001; 15:1369–1377.
9. Mocroft A, Devereux H, Kinloch-de-Loes S, Wilson D, Madge S, Youle M, et al
. Immunological, virological and clinical response to highly active antiretroviral therapy treatment regimens in a complete clinic population. Royal Free Centre for HIV Medicine. AIDS 2000; 14:1545–1552.
10. Lucas GM, Chaisson RE, Moore RD. Highly active antiretroviral therapy in a large urban clinic: risk factors for virologic failure and adverse drug reactions. Ann Intern Med 1999; 131:81–87.
11. Richman DD, Bozzette S, Morton S, Chien S, Wrin T, Dawson K and Hellman N. The prevalence of antiretroviral drug resistance in the US
. 41st Interscience Conference on Antimicrobial Agents and Chemotherapy
2001 [abstract 324].
12. Blower SM, Aschenbach AN, Gershengorn HB, Kahn JO. Predicting the unpredictable: transmission of drug-resistant HIV. Nat Med 2001; 7:1016–1020.
13. Deeks SG. Treatment of antiretroviral-drug-resistant HIV-1 infection. Lancet 2003; 362:2002–2011.
14. Pozniak A, Gazzard B, Anderson J, Babiker A, Churchill D, Collins S, et al
. British HIV Association (BHIVA) guidelines for the treatment of HIV-infected adults with antiretroviral therapy. HIV Med 2003; 4(suppl 1):1–41.
15. Anon. European guidelines for the clinical management and treatment of HIV-infected adults in Europe
. AIDS 2003; 17(suppl 2)
16. Deeks SG, Hellmann NS, Grant RM, Parkin NT, Petropoulos CJ, Becker M, et al
. Novel four-drug salvage treatment regimens after failure of a human immunodeficiency virus type 1 protease inhibitor-containing regimen: antiviral activity and correlation of baseline phenotypic drug susceptibility with virologic outcome. J Infect Dis 1999; 179:1375–1381.
17. Hammer SM, Vaida F, Bennett KK, Holohan MK, Sheiner L, Eron JJ, et al
. Dual vs single protease inhibitor therapy following antiretroviral treatment failure: a randomized trial. JAMA 2002; 288:169–180.
18. Lazzarin A, Clotet B, Cooper D, Reynes J, Arasteh K, Nelson M, et al
. Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Australia. N Engl J Med 2003; 348:2186–2195.
19. Ledergerber B, Lundgren JD, Walker AS, Sabin C, Justice A, Reiss P, 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; 364:51–62.
20. LaBonte J, Lebbos J, Kirkpatrick P. Enfuvirtide. Nat Rev Drug Discov 2003; 2:345–346.
21. Jensen-Fangel S, Pedersen C, Larsen CS, Tauris P, Møller A, Obel N. Changing demographics in an HIV-infected population: results from an observational cohort study in Western Denmark. Scand J Infect Dis 2001; 33:765–770.
22. Lohse N, Hansen AB, Jensen-Fangel S, Kronborg G, Kvinesdal B, Pedersen C, et al
. Demographics of HIV-1 infection in Denmark: Results from The Danish HIV Cohort Study
. Scand J Infect Dis
23. Castro KG. 1993 Revised Classification System for HIV Infection and Expanded Surveillance Case Definition for AIDS Among Adolescents and Adults
24. Ancelle-Park R, Klein JP, Stroobant A, Smith E, Haikala O, Koch MA, et al
. Expanded European AIDS case definition. Lancet 1993; 341:441.
25. Clayton D, Hills M. Statistical Methods in Epidemiology. Oxford: Oxford University Press; 1993.
26. Cleves M, Gould W, Gutierrez RG. An Introduction to Survival Analysis using STATA. College Station, Texas: Stata Press; 2002.
27. Schoenfeld D. Partial residuals for the proportional hazards regression models. Biometrica 1982; 69:239–241.
28. Mocroft A, Ledergerber B, Viard JP, Staszewski S, Murphy M, Chiesi A, et al
. Time to virological failure of 3 classes of antiretrovirals after initiation of highly active antiretroviral therapy: results from the EuroSIDA study group. J Infect Dis 2004; 190:1947–1956.
29. Kitahata MM, Koepsell TD, Deyo RA, Maxwell CL, Dodge WT, Wagner EH. Physicians’ experience with the acquired immunodeficiency syndrome as a factor in patients’ survival. N Engl J Med 1996; 334:701–706.
30. Kirk O, Katzenstein TL, Gerstoft J, Mathiesen L, Nielsen H, Pedersen C, et al
. Combination therapy containing ritonavir plus saquinavir has superior short-term antiretroviral efficacy: a randomized trial. AIDS 1999; 13:F9–F16.
31. Burger DM, Hoetelmans RM, Hugen PW, Mulder JW, Meenhorst PL, Koopmans PP, et al
. Low plasma concentrations of indinavir are related to virological treatment failure in HIV-1-infected patients on indinavir-containing triple therapy. Antivir Ther 1998; 3:215–220.
32. Durant J, Clevenbergh P, Garraffo R, Halfon P, Icard S, Del Giudice P, et al
. Importance of protease inhibitor plasma levels in HIV-infected patients treated with genotypic-guided therapy: pharmacological data from the Viradapt Study. AIDS 2000; 14:1333–1339.
33. Dybul M, Fauci AS, Bartlett JG, Kaplan JE, Pau AK. Guidelines for using antiretroviral agents among HIV-infected adults and adolescents. Ann Intern Med 2002; 137(5 Pt 2):381–433.
34. Roge B, Barfod T, Kirk O, Katzenstein T, Obel N, Nielsen H, et al
. Resistance profiles and adherence at primary virological failure in three different highly active antiretroviral therapy regimens: analysis of failure rates in a randomized study. HIV Med 2004; 5:344–351.
35. Jorgensen LB, Christensen MB, Gerstoft J, Mathiesen LR, Obel N, Pedersen C, et al
. Prevalence of drug resistance mutations and non-B subtypes in newly diagnosed HIV-1 patients in Denmark. Scand J Infect Dis 2003; 35:800–807.
This article has been cited 19 time(s).
2012 American Control Conference (Acc)
Resistance evolution in HIV - modeling when to intervene
2012 American Control Conference (Acc), ():
Journal of Infectious DiseasesPredictors of CD4(+) T-Cell Counts of HIV Type 1-Infected Persons After Virologic Failure of All 3 Original Antiretroviral Drug ClassesJournal of Infectious Diseases
Risk of extensive virological failure to the three original antiretroviral drug classes over long-term follow-up from the start of therapy in patients with HIV infection: an observational cohort study
Sexual HealthTriple-class HIV antiretroviral therapy failure in an Australian primary care settingSexual Health
Annals of Internal Medicine
Survival of persons with and without HIV infection in Denmark, 1995-2005
Annals of Internal Medicine, 146(2):
Antiviral TherapyThe incidence rate of HIV type-1 drug resistance in patients on antiretroviral therapy: a nationwide population-based Danish cohort study 1999-2005Antiviral Therapy
International Journal of EpidemiologyCohort Profile: The Danish HIV Cohort StudyInternational Journal of Epidemiology
Archives of Internal Medicine
Triple-Class Virologic Failure in HIV-Infected Patients Undergoing Antiretroviral Therapy for Up to 10 Years
Archives of Internal Medicine, 170(5):
Regional differences in the risk of triple class failure in European patients starting combination antiretroviral therapy after 1 January 1999
Hiv Medicine, 9(1):
Journal of Infectious DiseasesIncidence, Clinical Presentation, and Outcome of Progressive Multifocal Leukoencephalopathy in HIV-Infected Patients during the Highly Active Antiretroviral Therapy Era: A Nationwide Cohort StudyJournal of Infectious Diseases
Journal of Theoretical BiologyOptimal drug treatment regimens for HIV depend on adherenceJournal of Theoretical Biology
Genotypic drug resistance and long-term mortality in patients with triple-class antiretroviral drug failure
Antiviral Therapy, 12(6):
Archives of Internal Medicine
Changes over time in risk of initial virological failure of combination antiretroviral therapy - A multicohort analysis, 1996 to 2002
Archives of Internal Medicine, 166(5):
Bulletin of Mathematical BiologyDrug-sparing regimens for HIV combination therapy: Benefits predicted for "Drug Coasting"Bulletin of Mathematical Biology
Journal of Medical VirologySustained Low-Dose Growth Hormone Therapy Optimizes Bioactive Insulin-Like Growth Factor-I Level and May Enhance CD4 T-Cell Number in HIV InfectionJournal of Medical Virology
Immunological responses during a virologically failing antiretroviral regimen are associated with in vivo synonymous mutation rates of HIV type-1 env
Antiviral Therapy, 14(3):
Declining prevalence of HIV-infected individuals at risk of transmitting drug-resistant HIV in Denmark during 1997-2004
Antiviral Therapy, 11(5):
Clinical Infectious DiseasesMortality Attributable to Smoking Among HIV-1-Infected Individuals: A Nationwide, Population-Based Cohort StudyClinical Infectious Diseases
JAIDS Journal of Acquired Immune Deficiency SyndromesEvolution of HIV Resistance Mutations in Patients Maintained on a Stable Treatment Regimen After Virologic FailureJAIDS Journal of Acquired Immune Deficiency Syndromes
HIV; AIDS; cohort studies; highly active antiretroviral therapy; treatment failure
© 2005 Lippincott Williams & Wilkins, Inc.
Highlight selected keywords in the article text.