SEARCH

SEARCH BY CITATION

Keywords:

  • antiretroviral therapy;
  • causes;
  • death;
  • HIV-infected persons;
  • incidence

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background

Since the introduction of highly active antiretroviral therapy (HAART), the incidence of death in HIV-infected patients has dramatically decreased, and causes of death other than those related to HIV infection have increased, although it is unclear how these parameters compare with those in the age-matched general population living in the same geographical region.

Methods

Consecutive HIV-infected adults who were prescribed HAART in our hospital were prospectively followed from January 1997 to December 2004 or until death, loss to follow-up or discontinuation of HAART. Estimations of the annual incidence and causes of death in the general population of similar age in Catalonia per calendar year in the study period were obtained and compared with those in the HIV-infected cohort.

Results

There were 235 deaths among the 4471 patients on HAART (5%). The incidence of mortality decreased over time in HIV-infected patients (P<0.001; χ2 test for trend), although it has remained approximately five times higher than that for the age-matched general population. AIDS-related events were the most common cause of death (n=95; 40%), although they significantly decreased over time (P<0.001; χ2 test for trend), whereas liver diseases (P<0.001; χ2 test for trend) and non-AIDS-defining infections (P=0.008; χ2 test for trend) significantly increased over time. Infections in general (33 times higher), liver diseases (11 times higher) and non-Hodgkin lymphoma (5 times higher) were overrepresented as causes of death in the HIV-infected cohort compared with the age-matched general population.

Conclusions

Non-AIDS-defining infectious diseases, liver diseases, and non-Hodgkin lymphoma represent specific targets for efforts to further decrease mortality in HIV-infected patients receiving HAART.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The dramatic decrease in AIDS-related mortality associated with the routine introduction of highly active antiretroviral therapy (HAART) in developed countries has changed consideration of HIV infection from a rapidly fatal to a chronic manageable disease [1–4]. The clinical benefit associated with HAART seems to be largely a result of its ability to produce sustained suppression of viral replication [5], and its effectiveness has increased with time [6,7]. Recent data from the observational DAD Study cohort showed that mortality in HIV-infected patients not only from HIV-related causes but also from other causes, such as neoplasia and liver and heart disease, increased with immunological decline [8]. These data are in accordance with those from the largest randomized study on an antiretroviral treatment interruption trial, the SMART Study, in which deaths not only from HIV-related but also from non-HIV-related causes were significantly increased in the therapy discontinuation arm [9].

Knowledge concerning mortality in HIV-infected patients receiving HAART has been evolving as the amount of data available has increased, but the studies in which the data were obtained have several limitations. The incidence of mortality in HIV-infected persons receiving HAART may not be accurate because some cohort studies have used a combined endpoint of AIDS or death, as a result in part of lack of information on causes of death [5,6,10]. In addition, data on the causes of death in the era of HAART are scarce. Some cohort studies have distinguished generically between AIDS-related and non-AIDS-related causes without further details on specific causes of death [11,12], or have been limited exclusively to certain causes of death such as liver disease [13]. Other studies reporting more specifically the causes of death of HIV-infected patients have not compared them with those in the general population [14–16] or lacked combined information on the incidence of mortality [17].

As the proportion of non-HIV-related deaths has been increasing in observational cohorts, there is a need to define and standardize causes of death more effectively and to compare them with those in the general population of similar age in order that the excess mortality attributable to any one cause may be identified [12]. This information would help health policy makers and clinicians to design strategies to diminish further the mortality of HIV-infected persons receiving HAART [18]. We aimed to analyse the incidence and causes of death in HIV-infected adults treated with HAART and cared for in our hospital, and to compare the results with those for the general population of similar age living in the same geographical region.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

From January 1997 to December 2004, consecutive HIV-infected adults (at least 16 years old) who had HAART prescribed at the Infectious Diseases Unit of our hospital were prospectively followed. For the purpose of the study, we chose January 1997 as the starting date of the study because HAART was not widely available in our hospital until that time. The Institutional Review Board at Hospital Clínic approved this study. Patients initiating or changing antiretroviral therapy were visited at least 1 month after the introduction of the HAART regimen and every 3–6 months thereafter. Clinical and laboratory data for patients were recorded at each visit. This cohort has been previously described [19–21] and represents one of the largest cohorts of HIV-infected adults from a single institution in Spain.

HAART was defined as at least triple therapy including two nucleoside reverse transcriptase inhibitors (NRTIs) plus at least one protease inhibitor (PI) or one nonnucleoside reverse transcriptase inhibitor (NNRTI) or a third NRTI. Criteria for initiation or maintenance of HAART were based on the updated recommendations existing at the time of the study. Patients for whom HAART was not recommended and those for whom it was recommended but not taken up were not included in this study. Antiretroviral therapy could be modified at the discretion of the treating physician, and patients remained in the study as long as they remained on HAART. For the purpose of the study, patients who discontinued antiretroviral therapy were retained in the study if the time without HAART was less than 3 months. Reasons for censoring were death, loss to follow-up, discontinuation of HAART longer than 3 months, or use of antiretroviral therapy not included in the definition of HAART, whichever came first.

Deaths were identified either directly from physician reports for patients who died during hospital admission or indirectly through inquiries to relatives or friends for patients who missed their last scheduled visit. Systematic inquiries to the National Register Office were also made to confirm death in those patients with missing visits. Special efforts were made to ultimately ascertain the cause of death in each patient from available sources. When more than one cause of death was reported, only the primary cause of death was considered. The death was considered related to AIDS if the primary cause of death was an AIDS-defining condition. Cause of death was classified as unknown if there was no death certificate available or if the cause of death was not provided. The most recent laboratory values of CD4 cell counts and plasma HIV RNA prior to death were identified for each individual and were included in the analyses as death-related if measured within the 6 months prior to death. Early death was defined when occurring in the first 6 months after initiation of HAART because the risk of AIDS-defining events potentially leading to death in patients with advanced HIV infection is higher during this period [22].

Estimations of the annual incidence and causes of death in the population of Catalonia per calendar year in the study period were obtained from the Institut d'Estadística de Catalunya [23], in which deceased persons are grouped by age. The group in the general population aged 16–65 years at death was chosen for comparison with HIV-infected patients because it matched the median age and range of the HIV-infected patients who died in our cohort. Specific data for smaller age groups within the age group 16–65 years were not available for analysis. Annual mortality rates in the HIV-infected cohort and in the general population were calculated as the number of deaths per 100 person-years by year.

Univariate analyses were performed using the χ2 test, Fisher's exact test, Student's t-test, the Wilcoxon test or the Kruskal–Wallis test, as appropriate. Adjusted odds ratios for death at different time-points were compared using the χ2 test for linear trend. Multivariate analyses through logistic regression models were used to assess the factors associated with HIV-related death among patients who died. Variables examined included age, gender, route of HIV transmission, CD4 cell count and plasma HIV-1 RNA at the initiation of HAART, active use of illicit drugs, hepatitis C or B virus infection (defined by positive serological tests), non-HAART antiretroviral therapy prior to the use of HAART, and duration of exposure to HAART. Time-to-event analysis was used to assess the incidence of the most common causes of death in HIV-infected patients. Statistical analysis was performed using stata software (release 7.0; StataCorp, College Station, TX). Statistical significance was determined using two-tailed tests, with P<0.05 considered to be significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Characteristics of the HIV-infected patients who died

Of the 4471 patients on HAART followed in the study period, 235 patients (5%) died. The patients who died were more commonly men (n=183, 78% vs. n=3050, 72%; P=0.052, Fisher's exact test) and older at the initiation of HAART [median age (interquartile range) 40 (34, 47) vs. 34 (30, 40) years; P=0.016; Wilcoxon rank sum test] than those remaining alive (n=4236). Persons who acquired HIV infection through injecting drug use were overrepresented in the population of patients who died (53% vs. 44%; P=0.031; Fisher's exact test), but persons who acquired HIV infection through homosexual male contact (31% vs. 32%; P=0.782; Fisher's exact test), heterosexual contact (17% vs. 19%; P=0.543; Fisher's exact test) or haemophilia (2% vs. 5%; P=0.076; Fisher's exact test) were not. As expected, baseline CD4 cell counts in patients who died were lower [median (interquartile range) 114 (32, 227) cells/μL, n=230 vs. 339 (201–510) cells/μL, n=4088; P<0.001; Wilcoxon rank sum test] and baseline plasma HIV RNA levels were higher [median (interquartile range) 5.4 (4.2, 6.4) log10 copies/mL, n=177 vs. 4.9 (3.8, 5.6) log10 copies/mL, n=3974; P<0.001, Wilcoxon rank sum test] than the respective baseline values obtained for patients who remained alive.

Incidence of mortality

The number (percentage of all HIV-infected patients receiving care in the hospital) of HIV-infected patients on HAART in the cohort per year was 1127 (83%) in 1997, 1674 (92%) in 1998, 1946 (95%) in 1999, 2037 (95%) in 2000, 2112 (93%) in 2001, 2094 (88%) in 2002, 2028 (85%) in 2003, and 2053 (84%) in 2004. The number (percentage of all patients on HAART) of HIV-infected patients on HAART who died each year was 51 (5%) in 1997, 45 (3%) in 1998, 31 (2%) in 1999, 18 (1%) in 2000, 18 (1%) in 2001, 21 (1%) in 2002, 24 (1%) in 2003, and 27 (1%) in 2004. The population of Catalonia aged 16–65 years increased progressively in the years of the study, from 4148061 persons in 1997 to 4211244 persons in 2004. The number of deaths in this sector of the population tended to decrease with time, from a maximum of 9952 deaths in 1999 to a minimum of 9096 in 2003. The annual incidence of mortality in the HIV-infected cohort as compared with that in the general population aged 16–64 years in Catalonia is shown in Fig. 1. There was a significant decrease over this period for HIV-infected patients (P<0.001; χ2 test for trend), but there were no significant changes for the general population (P=0.936; χ2 test for trend). The effect of time was not linear for the HIV-infected cohort (Fig. 1). Nevertheless, the slight increase after 2000 was not found to be statistically significant in a test for nonlinear trend.

image

Figure 1.  Annual incidence of mortality in the HIV-infected cohort as compared with the general population aged 16–65 years in Catalonia. There were significant changes over this period for HIV-infected patients (P<0.001; χ2 test for trend), but not for the general population (P=0.936; χ2 test for trend).

Download figure to PowerPoint

Causes of mortality

The general causes of mortality in the HIV-infected cohort in decreasing order were AIDS-defining events (n=95; 40%), late-stage liver disease (n=53; 23%), non-AIDS-defining, nonhepatitis infections (n=33; 14%), non-AIDS-defining, nonhepatitis-related neoplasia (n=25; 11%), cardiovascular disease (n=14; 6%), drug overdose (n=6; 3%), unknown (n=4; 2%), suicide (n=3; 1%), and accidental/traumatic events (n=2; 1%) (Table 1). Seventy-five patients (32%) had a viral load <200 copies/mL at the time of death. Compared with patients dying with ≥200 copies/mL (n=102), patients dying with <200 copies/mL (n=75) died more commonly of late-stage liver disease (n=33, 44% vs. n=17, 17%), non-AIDS-defining, nonhepatitis-related neoplasia (n=14, 19% vs. n=10, 10%), and cardiovascular disease (n=8, 11% vs. n=5, 5%), and less commonly of AIDS-defining events (n=14, 19% vs. n=35, 34%), non-AIDS-defining, nonhepatitis infections (n=4, 5% vs. n=26, 25%), and drug overdose (n=1, 1% vs. n=2, 2%) (P<0.001; Fisher's exact test).

Table 1.  Major and specific causes of mortality (n=235) in the cohort of HIV-infected patients receiving highly active antiretroviral therapy (HAART)
Cause of deathn
AIDS-defining events95 (40%)
 Non-Hodgkin lymphoma22
 Cytomegalovirus disease20
 Tuberculosis19
 Pneumocystis jiroveci pneumonia15
 Mycobacterium avium intracellulare disseminated infection8
 Progressive multifocal leucoencephalopathy6
 Toxoplasmic encephalitis2
 Disseminated cryptococcosis2
 Kaposi sarcoma1
Late-stage liver disease53 (23%)
 Hepatitis C virus-related42
 Hepatitis B virus-related6
 Hepatitis C and B virus-related3
 Nonviral cirrhosis2
Non-AIDS-defining, nonhepatitis infections33 (14%)
 Pneumonia16
 Sepsis13
 Abdominal infections3
 Pulmonary aspergilosis1
Non-AIDS-defining, nonhepatitis-related neoplasia25 (11%)
 Lung10
 Colon8
 Pancreas2
 Unknown origin2
 Prostate1
 Breast1
 Thyroid1
Cardiovascular disease14 (6%)
 Myocardial infarction12
 Cerebrovascular disease2
Drug overdose6 (3%)
Accidental/traumatic events2 (1%)
 Car crash1
 Weapon injury1
Suicide3 (1%)
Unknown4 (2%)

The proportion of deaths attributable to AIDS-defining events significantly decreased over time, from 84% in 1997 to 7% in 2004 (P<0.001; χ2 test for trend) (Table 2). Among deaths attributable to non-AIDS-defining events, the proportion from late-stage liver disease (P<0.001; χ2 test for trend) and non-AIDS-defining infections (P=0.008; χ2 test for trend) significantly increased with time (Table 2), and these causes were responsible for the slight increase in the incidence of mortality after 2000. We were unable to find any significant trend with time regarding other causes of death among HIV-infected patients. In the general population of Catalonia of a similar age, the annual rates of cardiovascular diseases, accidental/traumatic events, and suicide as causes of death tended to increase, whereas neoplasia in general, liver diseases, and infections in general tended to decrease during the study period (data not shown).

Table 2.   Distribution of the number of patients dying per year according to major causes of mortality
 n
19971998199920002001200220032004
  1. Trends over time were significant for AIDS-defining events (P<0.001), late-stage liver disease (P<0.001) and non-AIDS-defining, nonhepatitis infections (P=0.008), but not for other causes of death (P>0.05) (χ2 test for linear trend).

AIDS-defining events (n=95)43 (84)20 (44)13 (42)6 (33)4 (22)5 (24)2 (8)2 (7)
Late-stage liver disease (n=53)4 (8)7 (16)3 (10)6 (33)5 (28)7 (33)10 (42)11 (41)
Non-AIDS-defining, nonhepatitis infections (n=33)1 (2)2 (4)7 (23)2 (11)1 (6)4 (19)8 (33)8 (30)
Non-AIDS-defining, nonhepatitis-related neoplasia (n=25)2 (4)7 (16)7 (23)1 (6)5 (28)1 (5)02 (7)
Cardiovascular disease (n=14)03 (7)03 (17)2 (11)2 (10)2 (8)2 (7)
Drug overdose (n=6)03 (7)1 (3)01 (6)1 (5)00
Accidental/traumatic events (n=2)01 (2)0001 (5)00
Unknown (n=4)0000002 (8)2 (7)
Total (n=235)51 (100)45 (100)31 (100)18 (100)18 (100)21 (100)24 (100)27 (100)

Infections in general [85 HIV-infected patients (36%) vs. 821 individuals in the general population (1%); 33 times higher in HIV-infected patients] and liver diseases [53 HIV-infected patients (23%) vs. 1507 individuals in the general population persons (2%); 11 times higher in HIV-infected patients] as causes of death were clearly overrepresented among HIV-infected patients, whereas neoplasia other than lymphoma [26 HIV-infected patients (11%) vs. 33 161 individuals in the general population (44%); 4 times lower in HIV-infected patients], cardiovascular diseases [14 HIV-infected patients (6%) vs. 14320 individuals in the general population (19%); 3 times lower in HIV-infected patients], and accidental/traumatic events [two HIV-infected patients (1%) vs. 9798 individuals in the general population (13%); 15 times lower in HIV-infected patients] were underrepresented compared with the general population of Catalonia of a similar age (Table 3). Non-Hodgkin lymphoma was the most common neoplasia leading to death in the cohort of HIV-infected patients (n=22; 46% of the deaths attributable to neoplasia), and this cause was overrepresented compared with the general population [22 HIV-infected patients (9%) vs. 1436 individuals in the general population (2%); 5 times higher in HIV-infected patients].

Table 3.   Deaths stratified by cause in the cohort of HIV-infected patients receiving highly active antiretroviral therapy (HAART) and in the general population of Catalonia aged 16–65 years in the period of the study
Cause of deathn (%)
HIV-infected patients receiving HAART (n=235)General population Catalonia aged 16–64 years (n=75 367)
  • For comparison purposes, the causes of death have been grouped according to the classification system used in the general population of Catalonia by the Institut d'Estadística de Catalunya [6]. There were additional HIV-infected patients who died from a drug overdose (n=6) and unknown causes (n=4) who are not included in the table because we were unable to assess the rates of mortality from these causes in the general population. There were other causes of death in the general population that are not listed here because they could not be compared with similar causes in the cohort of HIV-infected patients (e.g. neurological and renal causes).

  • *

    All cases of lymphoma leading to death in HIV-infected patients were non-Hodgkin lymphoma.

Infections with viruses, parasites, fungi, and nontuberculosis mycobacteria53 (23)467 (0.6)
Bacterial infections32 (14)354 (0.5)
Tuberculosis19 (8)143 (0.2)
Lymphoma22 (9)*1436 (2)
Neoplasia other than lymphoma26 (11)33161 (44)
Cardiovascular diseases14 (6)14320 (19)
Liver diseases53 (23)1507 (2)
Suicide3 (1)2216 (3)
Accidental/traumatic events2 (1)9798 (13)

AIDS-related vs. non-AIDS-related deaths

Ninety-five patients (40%) died as a result of AIDS-defining events. Patients dying of AIDS-defining events were younger (median, interquartile range, n: 38, 33–47 years, 95 vs. 42, 36–47 years, 140; P=0.025; Wilcoxon rank sum test), were more commonly men (n, %: 81, 85% vs. 102, 73%; P=0.026; Fisher's exact test), and had more commonly acquired HIV infection through the parenteral (rather than sexual) route (n, %: 50, 53% vs. 45, 47%; P=0.083; Fisher's exact test) than patients dying of non-AIDS-defining events. CD4 counts (cells/μL) in patients dying of AIDS-defining events at the initiation of HAART (median, interquartile range, n: 123, 48–254, 95 vs. 167, 90–246, 134; P=0.063; Wilcoxon rank sum test) and at the time of death (median, interquartile range, n: 63, 11–138, 95 vs. 154, 45–318, 135; P<0.0001; Wilcoxon rank sum test) were significantly lower than in patients dying as a result of non-AIDS-defining events. Plasma HIV RNA (log copies/mL) in patients dying of AIDS-defining events at the initiation of HAART (median, interquartile range, n: 5.5, 5.2–5.7, 37 vs. 5.2, 4.8–5.5, 95; P=0.027; Wilcoxon rank sum test) and at the time of death (median, interquartile range, n: 4.9, 2.3–5.5, 49 vs. 2.9, 2.3–5.2, 128; P=0.013; Wilcoxon rank sum test) was significantly higher than in patients dying of non-AIDS-defining events.

Multivariate analysis identified two independent risk factors for dying of an AIDS-defining event in HIV-infected patients receiving HAART: a CD4 cell count lower than 200 cells/μL at the initiation of HAART (adjusted odds ratio 5.4; 95% confidence interval 1.8–16.7; P=0.003) and an increasing plasma viral load at the initiation of HAART (per unit log copies/mL) (adjusted odds ratio 4.5; 95% confidence interval 1.7–12.3; P=0.003).

Time to death after the initiation of HAART in HIV-infected patients who died

Ninety patients (38%) died in the first 6 months after the initiation of HAART. AIDS-defining events (n=60; 63%), drug overdose (n=4; 67%), and suicide (n=2; 67%) were more concentrated in the first 6 months on HAART than other causes of death: non-AIDS-defining infections (n=9; 27%), non-AIDS-defining neoplasia (n=5; 20%), late-stage liver disease (n=8; 15%), cardiovascular disease (n=2; 14%), and accidental/traumatic events (n=0; 0%) (P<0.001; Fisher's exact test). There were no significant differences in the annual distribution of causes of early death per year (data not shown).

The Kaplan–Meier survival estimates for the four major causes of death in this cohort (AIDS-defining events, non-AIDS-defining neoplasia, late-stage liver disease and cardiovascular disease) showed significant differences among them (Fig. 2) (P<0.001; generalized log-rank test). Patients who died of an AIDS-defining event did so earlier than patients who died of non-AIDS-defining neoplasia (P<0.001; log-rank test) and cardiovascular disease (P=0.002; log-rank test). Patients who died of end-stage liver disease also died earlier than patients who died of non-AIDS-defining neoplasia (P=0.002; log-rank test) and cardiovascular disease (P=0. 015; log-rank test) (Fig. 2).

image

Figure 2.  Kaplan–Meier survival estimates by cause of death for the four most common causes of death: AIDS-defining events, non-AIDS neoplasia, liver disease and cardiovascular disease. There were significant differences in the overall comparison among the four causes (P<0.001; generalized log-rank test). The comparisons between AIDS-defining events and non-AIDS neoplasia (P<0.001; log-rank test) and between AIDS-defining events and cardiovascular disease (P=0.002; log-rank test), and the comparisons between liver disease and non-AIDS neoplasia (P=0.002; log-rank test) and between liver disease and cardiovascular disease (P=0.015; log-rank test) were also significant.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Among HIV-infected patients receiving HAART, we have confirmed that patients who acquired HIV infection through injecting drug use, and those who had lower CD4 cell counts and higher plasma viral loads at baseline had a higher risk of death. A higher risk of death in former or current injecting drug users has also been reported in several previous studies, and factors such as drug overdose, chronic viral hepatitis and end-stage liver disease, and a higher risk of bacterial infections and suicide might at least partly explain the higher risk of mortality in this population [12,13,24–27]. Data from very large cohort studies have also shown that both baseline [6] and current [8] low CD4 cell counts and baseline [6] high plasma viral load are associated with a higher risk of death. It is surprising that the higher risk of death associated with advanced immune suppression [8] and uncontrolled viral suppression in therapy interruption studies [9] seems to be a result not only of AIDS-defining events but also of non-HIV-related causes.

There was a decreasing incidence of mortality in our cohort from 1997 until 2000, with a slightly increasing plateau from 2000 until 2004. However, even in the latter period, the incidence of mortality in HIV-infected patients was approximately five times higher than that in the age-matched general population. These data are in accordance with those recently reported by others [18,24,28,29] and emphasize the need for additional measures to decrease further the incidence of mortality in HIV-infected patients despite the availability of HAART.

Although AIDS-related events have been the most common cause of death during the study period, there has been a significant decrease over time, whereas liver diseases and non-AIDS-defining infections have been significantly increasing over time. Among HIV-infected patients who died, those with lower CD4 cell counts and higher plasma viral loads were more likely to die of an AIDS-defining event than of other causes. Patients dying of an AIDS-event also died earlier than patients dying of other non-HIV-related causes such as neoplasia or cardiovascular disease. This fact may reflect the lag in the immunological and clinical benefit in highly immunesuppressed patients despite the availability of effective HAART. Almost one-third of the patients died with optimal suppression of plasma viral replication, and the causes of death in these patients were more commonly noninfectious diseases. Several studies have highlighted the increasing importance of liver-related disease as a cause of death in HIV-infected patients [14,18]. Although a specific analysis of the EuroSIDA cohort found a decreasing rate of death from liver-related disease, explained in part by increases in CD4 cell count after starting antiretroviral therapy, there was an increase in deaths from liver-related disease over time after adjustment for antiretroviral-induced CD4 cell increases [13]. Patients coinfected with hepatitis B or C virus are also well known to have an increased death rate from liver-related disease [30]. Non-AIDS-defining infections have been shown to decrease with HAART in some studies [18], but some others have shown the opposite effect [14]. In general, the causes of death in HIV-infected patients receiving HAART have been understudied, and some experts have pointed out the need to define and standardize causes of death in these patients more effectively [12].

When comparing the rates of death from specific causes, we were able to detect an excess of mortality in HIV-infected patients treated with HAART attributable to infections in general and to liver diseases. The impact of infectious causes may be attributable at least in part to AIDS-defining events, which were among the most common causes of death at the beginning of the study. However, other AIDS-defining infections that are not exclusive of HIV infection, such as tuberculosis, and bacterial infections were overrepresented in HIV-infected patients compared with the general population. The reasons for this excess mortality from non-AIDS-defining infections are not clear. Despite the controversy on whether there has been a decrease or an increase in deaths from liver-related diseases in HIV-infected patients on HAART over time [13], our data suggest that liver-related deaths are clearly overrepresented in these patients compared with the age-matched general population. The incidence of systemic non-Hodgkin lymphoma in HIV-infected patients has not clearly decreased as a result of the use of HAART [31], and the survival rate in HIV-infected patients is still lower than that in non-HIV-infected persons, despite the fact that survival has improved with effective HAART [32].

Our study has several limitations. We compared the causes of death in HIV-infected individuals with those in the general population, but the general population is not an epidemiologically relevant reference group. HIV-infected individuals are different from the general population in various respects, such as in having more comorbidities or more behavioural and psychological impairments, and the conceptual approach of comparison with the general population may be of arguable validity, particularly in the light of the limited covariates available in the general population. Although we assessed the causes of death in HIV-infected persons using a simple algorithm, it is possible that the methods used to assess the causes of death in the general population were different. The available data on the follow-up of HIV-infected patients remaining alive were too limited to enable calculation of standardized mortality ratios in the HIV-infected cohort. Finally, we were not able to take into account competing events in investigating the incidence of the different causes of death in HIV-infected individuals. Despite these limitations, we believe that our study offers a potentially valuable insight into the trends of mortality in HIV-infected persons receiving HAART.

In summary, mortality in HIV-infected patients receiving HAART has decreased over time but is still higher than that in the general population of a similar age. While AIDS-defining events have been steadily decreasing as a cause of death, there has been an increase in deaths from non-AIDS-related infections and liver-related diseases. These increasingly common causes of death, together with non-Hodgkin lymphoma, are clearly overrepresented in HIV-infected patients receiving HAART compared with the age-matched general population, and they represent specific targets for efforts to decrease further the mortality in this population.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was supported in part by a grant from the “Red Temática Cooperativa de Grupos de Investigación en Sida del Fondo de Investigación Sanitaria (FIS),” and grant ISCIII-RETIC RD06/006 from the Instituto de Salud Carlos III, Madrid (Spain). Dr. E. Martinez and Dr. J. M. Miró were recipients of a Research Grant from the “Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS)” and the “Conselleria de Salut de la Generalitat de Catalunya, Barcelona (Spain)”.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Egger M, Hirschel B, Francioli P et al. Impact of new antiretroviral combination therapies in HIV infected patients in Switzerland: prospective multicentre study. Swiss Cohort Study. Br Med J 1997; 315: 11941199.
  • 2
    Pallella FJ, Delaney KM, Moorman AC et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338: 853860.
  • 3
    Mocroft A, Vella S, Benfield TL et al. Changing patterns of mortality across Europe in patients infected with HIV-1. Lancet 1998; 352: 17251730.
  • 4
    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: 1726.
  • 5
    Ledergerber B, Egger M, Opravil M et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Lancet 1999; 353: 863868.
  • 6
    Sterne JAC, Hernán MA, Ledergerber B et al. Long-term effectiveness of potent antiretroviral therapy in preventing AIDS and death: a prospective cohort study. Lancet 2005; 366: 378384.
  • 7
    Schneider MF, Gange SJ, Williams CM et al. Patterns of the hazard of death after AIDS through the evolution of antiretroviral therapy: 1984–2004. AIDS 2005; 19: 20092018.
  • 8
    Weber R, Friis-Møller N, Sabin C et al HIV and non-HIV-related deaths and their relationship to immunodeficiency: The D.A.D. Study. XIIth Conference on Retroviruses and Opportunistic Infections. Boston, MA, February 2005 [Abstract 595].
  • 9
    El-Sadr W, Neaton J. Episodic CD4-guided use of ART is inferior to continuous therapy: results of the SMART Study. XIIIth Conference on Retroviruses and Opportunistic Infections. Denver, CO, February 2006 [Abstract 106LB].
  • 10
    Egger M, May M, Chêne G et al. Prognosis of HIV-1-infected patients starting highly active antiretroviral therapy: a collaborative analysis of prospective studies. Lancet 2002; 360: 119129.
  • 11
    Mocroft A, Brettle R, Kirk O et al. Changes in the cause of death among HIV positive subjects across Europe: results from the EuroSIDA study. AIDS 2002; 16: 16631671.
  • 12
    Mocroft A, Gatell J, Reiss P et al. Causes of death in HIV infection: the key determinant to define the clinical response to anti-HIV therapy. AIDS 2004; 18: 23332337.
  • 13
    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: 21172125.
  • 14
    Louie JK, Hsu LC, Osmond DH, Katz MH, Schwarcz SK. Trends in the causes of death among persons with acquired immunodeficiency syndrome in the era of highly active antiretroviral therapy, San Francisco 1994–98. J Infect Dis 2002; 186: 10231027.
  • 15
    Wang C, Vlahov D, Galai N et al. Mortality in HIV-seropositive versus -seronegative persons in the era of highly active antiretroviral therapy: implications for when to initiate therapy. J Infect Dis 2004; 190: 10461054.
  • 16
    Krentz HB, Kliewer G, Gill MJ. Changing mortality rates and causes of death for HIV-infected individuals living in Southern Alberta, Canada from 1984 to 2003. HIV Med 2005; 6: 99106.
  • 17
    Lewden C, Salmon D, Morlat P et al. Causes of death among human immunodeficiency virus (HIV)-infected adults in the era of potent antiretroviral therapy: emerging role of hepatitis and cancers, persistent role of AIDS. Int J Epidemiol 2005; 34: 121130.
  • 18
    Crum NF, Riffenburgh RH, Wegner S et al. Comparisons of causes of death and mortality rates among HIV-infected persons: analysis of the pre-, early, and late HAART (highly active antiretroviral therapy) eras. J Acquir Immune Defic Syndr 2006; 41: 194200.
  • 19
    Pérez-Cuevas B. Influencia del Tratamiento Antirretrovírico de Gran Actividad en la Incidencia de las Infecciones Oportunistas y Neoplasias Asociadas a la Infección por VIH-1 y en la Mortalidad. http://www.tdx.cesca.es/TESIS_UB/AVAILABLE/TDX-0213102-091238/TOL33.pdf 2006.
  • 20
    Martinez E, Mocroft A, Garcia-Viejo MA et al. Risk of lipodystrophy in HIV-1-infected patients treated with protease inhibitors: a prospective cohort study. Lancet 2001; 357: 592598.
  • 21
    Garcia F, De Lazzari E, Plana M et al. Long-term CD4+T-cell response to highly active antiretroviral therapy according to baseline CD4+T-cell count. J Acquir Immune Defic Syndr 2004; 36: 702713.
  • 22
    Michelet C, Arvieux C, Francois C et al. Opportunistic infections occurring during highly active antiretroviral treatment. AIDS 2006; 12: 18151822.
  • 23
    Iustitut d'Estadística de Catalunya. Estudis de la població. http://www.idescat.net/cat/poblacio/poblestudis.html 2007.
  • 24
    Van Sighem A, Danner S, Ghani AC et al. Mortality in patients with successful initial response to highly active antiretroviral therapy is still higher than in non-HIV-infected individuals. J Acquir Immune Defic Syndr 2005; 40: 212218.
  • 25
    Pérez-Hoyos S, Ferreros I, Del Amo J et al. Survival and progression to AIDS in a seroconverter cohort in the post-highly active antiretroviral therapy era: effectiveness goes on. AIDS 2006; 20: 289291.
  • 26
    Sabin C. on behalf of the Antiretroviral Therapy Cohort Collaboration. AIDS events among individuals initiating HAART: do some patients experience a greater benefit from HAART than others? AIDS 2005; 19: 19952000.
  • 27
    Lloyd-Smith E, Brodkin E, Wood E et al. Impact of HAART and injection drug use on life expectancy of two HIV-positive cohorts in British Columbia. AIDS 2006; 20: 445450.
  • 28
    Jaggy C, Von Overbeck J, Ledergerber B et al. Mortality in the Swiss HIV Cohort Study (SHCS) and the Swiss general population. Lancet 2003; 362: 877878.
  • 29
    Keiser O, Taffé P, Zwahlen M et al. All cause mortality in the Swiss HIV cohort study from 1999 to 2001 in comparison with the Swiss population. AIDS 2004; 18: 18351843.
  • 30
    Rockstroh J, Konopnicki D, Mocroft A et al. Hepatitis B. and hepatitis C in the EuroSIDA cohort: prevalence and effect on mortality, AIDS progression and response to HAART. XIth Conference on Retroviruses and Opportunistic Infections. San Francisco, CA, February 2004 [Abstract 799].
  • 31
    Kirk O, Pedersen C, Cozzi-Lepri A et al. Non-Hodgkin lymphoma in HIV-infected patients in the era of highly active antiretroviral therapy. Blood 2001; 98: 34063412.
  • 32
    Gerard L, Galicier L, Maillard A et al. Systemic non-Hodgkin lymphoma in HIV-infected patients with effective suppression of HIV replication: persistent occurrence but improved survival. J Acquir Immune Defic Syndr 2002; 30: 478484.