Introduction: The optimal timing for initiation of highly active antiretroviral therapy (HAART) in patients with AIDS and tuberculosis (TB) is an unresolved question. To assess the effect of HAART on the survival of patients with TB, we designed this study.
Methods: We selected all HIV patients included in the COMESEM cohort with TB diagnosis after 1996. Clinical and epidemiological data were registered. We compared patients who started HAART at the diagnosis of TB [simultaneous therapy (ST)] or not. Survival was assessed by Cox analysis.
Results: Among the 6934 HIV patients included in the cohort, 1217 patients had TB, 322 of them (26.5%) after 1996. At the time of TB diagnosis, 45% of them started HAART (ST). There were no differences between groups regarding basal characteristics, except for a lower viral load in ST patients. ST therapy was associated with improved survival (hazard ratio 0.38; 95% confidence interval 0.20 to 0.72, P = 0.003). By univariate analysis, survival was also associated with no endovenous drug use and a later year of TB diagnosis. After adjusting for other prognostic variables, by Cox multivariate analysis, ST remained robustly associated with improved survival (hazard ratio 0.37; 95% confidence interval 0.17 to 0.66, P = 0.001).
Conclusions: Simultaneous HAART and TB treatment in HIV patients with TB is associated with improved survival.
From the *Infectious Diseases Section, Hospital Universitario Fundación Alcorcón, Alcorcón, Madrid, Spain; †Internal Medicine Department, Hospital Príncipe de Asturias, Alcalá de Henares, Madrid, Spain; ‡Internal Medicine Department, Hospital Universitario de Getafe, Madrid, Spain; §Infectious Diseases Department, Hospital de Móstoles, Madrid, Spain; and ‖Infectious Diseases Department, Hospital Severo Ochoa, Leganés, Madrid, Spain.
Received for publication February 19, 2008; accepted September 30, 2008.
Supported by Fundación para la Investigación y la Prevención del SIDA en España (FIPSE) 36296/02 and 36690/07 and by Instituto de Salud Pública, Comunidad de Madrid, Madrid, Spain.
Correspondence to: María Velasco, MD, PhD, Infectious Diseases Section, Internal Medicine Unit, Hospital Universitario Fundación Alcorcón, C/Budapest n°.1, 28922, Alcorcón, Madrid, Spain (e-mail: firstname.lastname@example.org).
The best time to begin antiretroviral treatment in patients with tuberculosis (TB) and HIV is controversial. Reconstitution immune syndrome,1 pharmacologic interactions, and high pill burden argue against a simultaneous beginning of HIV and TB treatments.2 In contrast, a delay in the antiretroviral treatment, especially in highly immunosuppressed patients, may be dangerous because of progression of AIDS and development of new opportunistic diseases.3,4 Whether any of these alternatives influence survival is largely unknown. Some prospective studies are evaluating this issue; specifically, 4 trials will address the unresolved question of the optimal timing for initiation of highly active antiretroviral therapy (HAART) in patients with AIDS and TB.5 Until such studies are completed, we must rely on retrospective information.
We have revised the effects on survival of simultaneous versus delayed HAART initiation in TB- and HIV-coinfected patients in a large cohort of HIV-coinfected patients in Spain.
COMESEM cohort comprises demographic and clinical data of all HIV-infected adult patients attended in 5 population-based hospitals in Madrid during more than 20 years (from 1984).6 The cohort was prospectively constituted in 2000; data before this date (since 1984-2000) were retrospectively collected.
We selected all patients included in the cohort with a diagnosis of TB of any kind from 1987 until 2004. To assess the effect of HAART on the survival of patients with TB and to avoid the bias of antiretroviral potency, we analyzed the patients with TB diagnosed after 1996.
The following data of the first diagnosis of TB was registered: date of diagnosis; type of TB; number of antiretroviral drugs (if any) used before the diagnosis of TB and at the end of follow-up; and the last count of CD4 lymphocyte and viral load before the diagnosis of TB and at the end of follow-up. We also included age and sex, country of origin, risk group for HIV transmission, time of the HIV infection, date of first visit, date of last visit, and survival status.
TB was diagnosed by the attending physician, responsible for infectious diseases/HIV care, by using microbiological or histological criteria. According to the 1993 Centers for Disease Control and Prevention AIDS classification, TB was divided into pulmonary (only lung localization) or extrapulmonary case (whatever localization outside lung). We calculated the time (years) between the diagnosis of HIV infection and TB.
HAART was defined as a combination therapy of at least 3 antiretroviral drugs and included protease inhibitor or nonnucleoside reverse transcriptase inhibitor. Simultaneous HAART at the time of TB [simultaneous therapy (ST)] was considered if the patient started HAART within 2 months after the beginning of treatment of TB. Patients receiving HAART for more than 2 months before the diagnosis of TB were excluded from the study. Patients who started HAART after 3 months of TB diagnosis were included in the analysis as nonsimultaneous group (no ST).
Mortality was registered in the cohort by the attending physician. In addition, we checked mortality at the Central AIDS Mortality Registry in the Madrid Community.
Variables are described as the mean (SD), median, and interquartile range (IQR) for nonnormal variables or percentages as appropriate. Comparison of variables was performed with the Student t test, Mann-Whitney U statistics, Wilcoxon test, or χ2 test as appropriate. Survival was assessed by the Kaplan-Meier Method (univariate) and Cox proportional hazards regression model. Variables associated with mortality (P < 0.1) were included in a multivariate Cox model. Analysis was performed with the statistical package SPSS 13th edition (SPSS Inc, 1989-2004, Chicago, IL).
Among the 6934 HIV patients included in the cohort, 1217 patients had at least 1 episode of TB (17.6%), and there were 322 (26.5%) patients with TB in the period after 1996. Nine of them were receiving HAART for more than 2 months at the moment of diagnosis of TB and were excluded from the study.
Patients with TB after 1996 were 80% male and mean age was 34.9 years (SD 7.5). Endovenous drug use was the most common risk group (75%). Extrapulmonary TB was diagnosed in 67.4%. Other simultaneous AIDS-defining conditions were present in 15% of cases. CD4 lymphocyte count at diagnosis of TB did not vary with time [mean 210 (193), median 160 (IQR 69-289)]. Viral load at the TB diagnosis was 4.7 log copies per milliliter (SD 1.6), median 5.0 (IQR 4.0-5.7). HIV and TB infections were diagnosed in the same year in the 37.4% of patients and in following 3 years in the 60% of cases. These data did not change significantly with time, although there was a tendency to a later diagnosis of TB (3 years) in the last 5 years. Forty-seven patients (15%) died during the follow-up.
Forty-five percent of patients (n = 140) received HAART at the time of TB diagnosis. There were no differences between those who started HAART at the diagnosis of TB or not regarding age, sex, risk practice for HIV infection, being immigrant, CD4 count at diagnosis of TB, presence of other AIDS-defining illnesses, and in the proportion of extrapulmonary TB (Table 1). However, viral load was somewhat lower in patients who started TB treatment and HAART simultaneously [median viral load 4.8 (IQR 3.3-5.7) versus 5.1 log copies per milliliter (IQR 4.7-5.9), P = 0.011]. At the end of follow-up, there were no significant differences in CD4 cell count, viral load, or the number of antiretroviral drugs used (3 both groups) between patients who started HAART at the diagnosis of TB or not (Table 1).
Age, sex, extrapulmonary TB, and CD4 cell count and viral load at the moment of diagnosis of TB were not significantly different in patients who died (Table 2). In contrast, patients who died were more likely to be drug users, had an earlier year of diagnosis of both TB and HIV infection, had another AIDS-defining condition, and received nonsimultaneous TB-HIV treatment (Table 2). However, by univariate Cox analysis, being drug users, year of diagnosis of TB, other AIDS-defining conditions, and nonsimultaneous treatment were the only significant (or nearly significant) variables associated with poor survival (Table 3). In Cox univariate analysis, simultaneous treatment was associated with a markedly improved survival [hazard ratio (HR) 0.38; 95% confidence interval (CI) 0.20 to 0.72, P = 0.003], (Table 3, Fig. 1). In a multivariate analysis including all nearly significant variables (P < 0.1), simultaneous treatment remained as a powerful predictor of survival (HR 0.35; 95% CI 0.18 to 0.67, P = 0.001, Table 3). Finally, after further adjusting for age, sex, CD4 cell count, and log viral load at the moment of TB diagnosis, the association of simultaneous treatment with survival remained essentially unchanged (HR 0.37; 95% CI 0.17 to 0.66, P = 0.001, Table 3).
Interestingly, a dramatic survival advantage was evident in the short term (Fig. 1). At 6 months of follow-up, HR for simultaneous use of TB-HAART was 0.15 (95% CI 0.03 to 0.586, P = 0.007). This survival effect was attenuated at 12 months (HR 0.33; 95% CI 0.14 to 0.78) and was maintained in a similar HR until the end of follow-up.
In this article, we have shown that patients who do not start HAART at the same time of TB diagnosis have a poorer survival than patients who do. Despite that, patients who survive do not have a worse immunological or virological prognosis.
Several determinants have been related to a better survival in HIV patients. Type of AIDS-defining illness may affect survival7; because all patients had TB in our series, this effect is attenuated. Increased CD4 cell count has been classically associated with better evolution and survival. In our cohort, CD4 cell count at TB diagnosis was not associated with survival, in agreement with recent reports.8 Year of diagnosis of TB was associated with better survival. Probably, this association reflects the different power of antiretroviral treatment and a better global management along the calendar periods.7 We restricted the cohort to patients with TB after year 1996, so our findings are a conservative approach because patients with “early” HAART were also included.
Differences in the basal characteristics of the patients who started HAART at the diagnosis of TB or not might explain the differences in mortality. However, we did not find any differences except for a slightly lower initial viral load in the group of treated patients. Other authors have reported an absence of correlation between basal viral load and survival in TB.9 Furthermore, the inclusion of viral load, CD4 cell count, drug use, and other AIDS-defined conditions in the multivariate analysis had virtually no effect on the robust association of ST with survival.
Some authors have reported a better survival effect in patients treated with HAART during TB treatment.10,11 Of note, we restricted our criteria to those patients who initiated HAART at the beginning of TB treatment. Besides that, a recent preliminary report has also described a higher mortality rate in patients who did not start HAART and TB treatment at the same time.12 Given the fact that results from clinical trials are pending,5 data from studies with a lower level of evidence (such cohort studies) are a very important source of information, as they are the best evidence available. In the same way, other authors have emphasized the good evolution of patients with concomitant TB treatment and HAART, yet without data about mortality. Hung et al13 showed a similar virological, immunological, and clinical response of treated patients for TB and HIV than patients without TB (who received only HAART). In addition, HAART may have beneficial effects in the evolution of TB as it was stated by Nahid et al.8 They found that use of HAART during TB treatment was associated with more rapid conversion of smears and cultures and improved survival. In addition, none of the patients in this study who received concomitant HAART relapsed.
Interestingly, our data suggest that the effect on survival seems very early and attenuates somewhat in the long run. This is not surprising because patients not treated simultaneously usually begin HAART in the last months of TB therapy or shortly after. However, early mortality may also be related to other medical conditions that have made preferable for attending clinicians to defer HAART. Despite our effort to correct for other important prognostic factors in the multivariate analysis, some other unregistered circumstances associated with delayed HAART may be a source of confusion.
Our work has other limitations. As a partially retrospective cohort, we do not have data about side effects or interactions of drugs or about immune reconstitution inflammatory syndrome (IRIS). The end point of the study, death, is hard enough to prevail over less relevant complications such as side effects not affecting survival.
IRIS may be a cause of death especially in patients with CD4 count less than 50 cells per milliliter.14 In a recent article, mortality by IRIS reached 1% in an African series, and it was associated with extremely low CD4 count and early initiation of treatment.15 In our series, patients had a higher CD4 count (median >150 cells/mL), and this could partially explain the good evolution of simultaneously treated patients.
The effect of physician preference for starting HAART may be a source of bias. Patients with suspected better adherence and lower probability of side effects could have been selected to start simultaneous treatment. Some variables to identify these patients such as other comorbidities or adherence are unknown in our study. Nonetheless, we did not find significant differences in the basal characteristics of the patients.
Causes of death were not included. Non-HIV related causes of death might change the conclusion of the study. Several works about mortality showed a shift in the causes of death toward 2000-2003.16 As we included deaths from 1997 to 2004, this effect may be attenuated.
In summary, our findings suggest improved survival by using concomitant TB therapy and HAART that seems independent of other classical factors such as CD4 count. Further research is warranted to identify the optimal time for HAART in the treatment of HIV-related TB.
1. Lawn SD, Myer L, Bekker LG, et al. Tuberculosis-associated immune reconstitution disease: incidence, risk factors and impact in an antiretroviral treatment service in South Africa. AIDS
2. Dean GL, Edwards SG, Ives NJ, et al. Treatment of tuberculosis in HIV-infected persons in the era of highly active antiretroviral therapy. AIDS
3. Kwara A, Carter EJ, Rich JD, et al. Development of opportunistic infections after diagnosis of active tuberculosis in HIV-infected patients. AIDS Patient Care STDS
4. Breen RA, Smith CJ, Cropley I, et al. Does immune reconstitution syndrome promote active tuberculosis in patients receiving highly active antiretroviral therapy? AIDS
5. Blanc FX, Havlir DV, Onyebujoh PC, et al. Treatment strategies for HIV-infected patients with tuberculosis: ongoing and planned clinical trials. J Infect Dis
6. Castilla V, Alberdi JC, Barros C, et al. Cohorte multicentrica de pacientes con infeccion VIH de la corona metropolitana sudeste de Madrid (COMESEM): fundamentos, organizacion y resultados iniciales. Rev Clin Esp
7. Couzigou C, Semaille C, Strat YL, et al. Differential improvement in survival among patients with AIDS after the introduction of HAART. AIDS Care
8. Nahid P, Gonzalez LC, Rudoy I, et al. Treatment outcomes of patients with HIV and Tuberculosis. Am J Respir Crit Care Med
9. Lawn SD, Wood R. Incidence of tuberculosis during highly active antiretroviral therapy in high-income and low-income countries. Clin Infect Dis
10. Manosuthi W, Chottanapand S, Thongyen S, et al. Survival rate and risk factors of mortality among HIV/tuberculosis-coinfected patients with and without antiretroviral therapy. J Acquir Immune Defic Syndr
11. Dheda K, Lampe FC, Johnson MA, et al. Outcome of HIV-associated tuberculosis in the era of highly active antiretroviral therapy. J Infect Dis
12. Lawn S, Myer L, Bekker LG, Wood R. Early mortality among patients with HIV-associated TB in Africa: implications for the time to initiate ART [Abstract O-126]. Presented at: 14th Conference on Retroviruses and Opportunistic Infections (CROI), 2007; Los Angeles, CA.
13. Hung CC, Chen MY, Hsiao CF, et al. Improved outcomes of HIV-1-infected adults with tuberculosis in the era of highly active antiretroviral therapy. AIDS
14. Manosuthi W, Kiertiburanakul S, Phoorisri T, et al. Immune reconstitution inflammatory syndrome of tuberculosis among HIV-infected patients receiving antituberculous and antiretroviral therapy. J Infect
15. Moore D, Liechty C, Ekwaru P, et al. Prevalence, incidence and mortality associated with tuberculosis in HIV-infected patients initiating antiretroviral therapy in rural Uganda. AIDS
16. 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