From the Data Collection on Adverse Events of Anti-HIV Drugs Study it was concluded that the risk of myocardial infarction (MI) increased progressively with the duration of HAART exposure . In the case that HAART induces a fast progressing atherosclerosis that proceeds in the post-MI period, this vascular disease might have a devastating influence on post-MI mortality in HIV-infected patients. The objective of the present study was to establish the post-MI mortality in HIV-infected patients on HAART and to compare mortality with that of MI patients without HIV infection and HIV-infected patients without MI.
The setting of the present study has been described elsewhere [2,3]. The study period was 1 January 1995 to 1 January 2005. Patient and HIV-related data were obtained from the Danish HIV Cohort Study , other demographic data from the Danish Civil Registration System , and the diagnosis of MI from the Danish National Hospital Registry, which records all discharge diagnoses according to the International Classification of Diseases .
The study was conducted as a matched cohort study. From the Danish HIV Cohort Study we identified 3953 patients who were registered in the Danish Civil Registration System, lived in Denmark on 1 January 1995 or at the date of the HIV diagnosis, and were not hospitalized with MI before HAART initiation. Of these, 44 developed an MI after the initiation of HAART. The index date was defined as the date of their MI.
In order to estimate the relative risk of death in the HIV-infected MI patients we identified for each patient four controls (176 patients) with MI and no HIV infection and four controls with HIV infection and no MI. MI patients were matched with the HIV-infected MI patients for year of birth, sex and year of MI (index date is date of MI). HIV patients were matched for year of birth (± 4 years), sex, CD4 cell count (± 60 cells/μl), and had to be alive, to be on HAART and to have no record of MI on the index date (date of MI in matched patient).
Observation time was calculated from the index date until 1 January 2005, death or loss to follow-up. We used Kaplan–Meier analysis to construct survival curves. Mortality rate ratios (MRR) for the first 90-day post-MI period and a period thereafter were calculated. Cox's proportional hazard analysis was used to adjust for AIDS-defining events before the MI, CD4 cell count (< 200 versus ≥ 200 cells/μl), hepatitis C status and co-morbidity. The co-morbidity was expressed as a modified Charlson's index derived from data of non-HIV diagnosis before the index date and including three co-morbidity levels: low (score of 0); medium (score of 1–2); high (score of > 2) [6,7]. The study was approved by the Danish Data Protection Agency. Statistical analyses were performed using SPSS software version 14.0 (SPSS Inc., Chicago, Illinois, USA).
More HIV-infected MI patients had an AIDS-defining disease before the index date (48 versus 31%) and a higher co-morbidity index, but otherwise the three patient groups were well matched (data not shown). The study had 1103 years of follow-up with 102 years in the HIV-infected MI group. Only three individuals (0.8%) were lost to follow-up.
Survival curves for the three groups of patients are shown in Figure 1. During the first 90-day post-MI period the risk of dying in the HIV-infected MI group was only slightly higher than in the MI group [MRR 1.5, 95% confidence interval (CI) 0.5–4.1; adjusted MRR 1.3, 95% CI 0.5–3.6], but substantially higher than in the HIV-infected group (MRR 10.7, 95% CI 2.1–54.9; adjusted MRR 11.7, 95% CI 1.8–78.2). In the period from 90 days post-MI the risk of dying in the HIV-infected MI group was much higher than in the MI group (MRR 7.4, 95% CI 2.1–26.3; adjusted MRR 4.5, 95% CI 1.2–16.3) but only slightly higher than in the HIV-infected group (MRR 1.4, 95% CI 0.6–3.5; adjusted MRR 1.6, 95% CI 0.6–4.3). Eleven patients died in the HIV-infected MI group, four as a result of heart diseases, five because of AIDS-defining diseases, one in a road accident and one from unknown causes. None of the patients in the HIV-infected group died as a result of heart problems.
We found that mortality in HIV-infected MI patients was almost equivalent to that of MI patients in the first 90-day post-MI period, and thereafter approximated that of HIV-infected patients without MI. The strengths of the study are its nationwide population-based design, long and nearly complete follow-up and the inclusion of two well-matched control cohorts. Although the study included all Danish HIV patients on HAART registered with an MI, the study population was small, which is also reflected in the rather broad confidence intervals.
In order to identify MI we relied on hospital registry-based discharge diagnoses, which for MI have been shown to be valid . In some patients with prehospital death, however, the MI diagnosis may not have been registered correctly.
The post-MI prognosis is affected by patient and infarction characteristics as well as treatment strategies [9–11]. It has been suggested that the young age of HIV-infected MI patients may partly explain their rather benign prognosis . In our study we matched the HIV-infected MI and control patients according to age at the MI index date, which is why this confounder should not bias our results. Also we used a modified Charlson's index to adjust for co-morbidity. Although our data did not allow us to adjust for several factors of major importance to the prognosis of MI patients, e.g. the type, site and severity of the heart attack as well as the treatment, we presume that the matched design of our study makes our result rather robust.
In conclusion, we found that mortality in HIV-infected MI patients was not significantly different to that of MI patients in the first 90 days post-MI, and thereafter was not significantly different to that of HIV patients The mortality in HIV-infected MI patients seems to be almost equivalent to what can be expected from an additive effect of MI and HIV infection.
The authors would like to thank the staff of their clinical departments for their continuous support and enthusiasm, and Lars Pedersen, Institute of Clinical Epidemiology, Aarhus University, Denmark, for help with calculation of the Charlson index.
Conflicts of interest: L.D.R., G.K., C.S.L. and G.P. have no conflicts of interest. J.G. received research funding from Abbott, Roche, Bristol-Myers Squibb, Merk Sharp and Dohme, Pharmacia, GlaxoSmithKline, Swedish Orphan and Boehringer Ingelheim. N.O. received research funding from Roche, Bristol-Myers Squibb, Merk Sharp and Dohme, GlaxoSmithKline, Abbott, and Boehringer Ingelheim.
Sponsorship: This work received financial support from the Danish AIDS Foundation, Odense University Hospital, Preben and Anna Simonsen's Foundation, the Foundation of the Danish Association of Pharmacists, and the Clinical Institute of the University of Southern Denmark.
1. Friis-Moller N, Sabin CA, Weber R, d'Arminio-Monforte A, El-Sadr WM, Reiss P, et al
, the Data Collection on Adverse Events of Anti-HIV Drugs (DAD) Study Group. Combination antiretrovial therapy and the risk of myocardial infarction. N Engl J Med 2003; 349:1993–2003.
2. Lohse N, Hansen ABE, 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 2005; 37:338–343.
3. Hansen ABE, Gerstoft J, Kronborg G, Pedersen C, Sørensen HT, Obel N. Mortality in siblings of patients co-infected with HIV and hepatitis C virus. J Infect Dis
2007; in press.
4. Frank L. Epidemiology – when an entire country is a cohort. Science 2000; 287:2398–2399.
5. Andersen TF, Madsen M, Jorgensen J, Mellemkjaer L, Olsen JH. The Danish National Hospital Register – a valuable source of data for modern health sciences. Dan Med Bull 1999; 46:263–268.
6. Charlson ME, Pompei P, Ales KL, McKenzie CR. A new method for classifying prognostic co-morbidity in longitudinal studies: development and evaluation. J Chronic Dis 1987; 40:373–383.
7. Sundararajan V, Henderson T, Perry C, Muggivan A, Quan H, Ghali WA. New ICD-10 version of the Charlson co-morbidity index predicted in-hospital mortality. J Clin Epidemiol 2004; 57:1288–1294.
8. Madsen M, Davidsen M, Rasmussen S, Abildstrom SZ, Osler M. The validity of the diagnosis of acute myocardial infarction in routine statistics: a comparison of mortality and hospital discharge data with the Danish MONICA registry. J Clin Epidemiol 2003; 56:124–130.
9. Garoufalis S, Kouvaras G, Vitsis G, Perdouris K, Markatou P, Hatzisavas J, et al
. Comparison of angiographic findings, risk factors, and long term follow-up between young and old patients with a history of myocardial infarction. Int J Cardiol 1998; 67:75–80.
10. Rosengren A, Wallentin L, Simoons M, Gitt AK, Behar S, Battler A, et al
. Age, clinical presentation, and outcome of acute coronary syndromes in the Euroheart acute coronary syndrome survey. Eur Heart J 2006; 27:789–795.
11. Boersma E, Pieper KS, Steyerberg EW, Wilcox RG, Chang WC, Lee KL, et al
. Predictors of outcome in patients with acute coronary syndromes without persistent ST-segment elevation. Results from an international trial of 9461 patients. Circulation 2000; 101:2557–2567.
12. Matetzky S, Dominingo M, Kar S, Noc M, Shah PK, Kaul S, et al
. Acute Myocardial infarction in human immunodeficiency virus-infected patients. Arch Intern Med 2003; 163:457–460.