Abstract
Background. Elite controllers spontaneously suppress human immunodeficiency virus (HIV) viremia but also demonstrate chronic inflammation that may increase risk of comorbid conditions. We compared hospitalization rates and causes among elite controllers to those of immunologically intact persons with medically controlled HIV.
Methods. For adults in care at 11 sites from 2005 to 2011, person-years with CD4 T-cell counts ≥350 cells/mm2 were categorized as medical control, elite control, low viremia, or high viremia. All-cause and diagnostic category-specific hospitalization rates were compared between groups using negative binomial regression.
Results. We identified 149 elite controllers (0.4%) among 34 354 persons in care. Unadjusted hospitalization rates among the medical control, elite control, low-viremia, and high-viremia groups were 10.5, 23.3, 12.6, and 16.9 per 100 person-years, respectively. After adjustment for demographic and clinical factors, elite control was associated with higher rates of all-cause (adjusted incidence rate ratio, 1.77 [95% confidence interval, 1.21–2.60]), cardiovascular (3.19 [1.50–6.79]) and psychiatric (3.98 [1.54–10.28]) hospitalization than was medical control. Non–AIDS-defining infections were the most common reason for admission overall (24.1% of hospitalizations) but were rare among elite controllers (2.7%), in whom cardiovascular hospitalizations were most common (31.1%).
Conclusions. Elite controllers are hospitalized more frequently than persons with medically controlled HIV and cardiovascular hospitalizations are an important contributor.
Keywords: HIV, elite control, antiretroviral therapy, hospitalization, inflammation, psychiatric disease
(See the editorial commentary by Karris and Haubrich on pages 1689–91.)
Elite controllers represent a small but important subset of persons living with human immunodeficiency virus (HIV) (PLWH) who suppress the virus and have delayed disease progression in the absence of antiretroviral therapy (ART) [1, 2]. Although prevalence of elite control is estimated at only 0.15%–1.5% of PLWH, study of these persons provides insights into HIV pathogenesis and potential mechanisms for new HIV therapies [3–6]. Despite spontaneous and durable control of HIV viremia, elite control is associated with chronic immune activation and low-grade inflammation that exceeds the level seen in persons who achieve viral suppression via ART [7–9].
Chronic inflammation among PLWH has been linked to complications such as cardiovascular disease, opportunistic infections, and neurologic disorders [10–13]. Persistent low-grade inflammation may therefore place elite controllers at higher risk of clinical events than persons whose HIV is controlled with ART. For example, prior studies have demonstrated a high burden of coronary atherosclerosis on radiographic screening of elite controllers, but data are lacking on any association with clinical outcomes [14, 15]. The relative rarity of elite control makes it difficult to study clinical outcomes, such as disease events or hospitalizations, in this population. We used a multisite, multistate cohort of PLWH to compare hospitalization rates among elite controllers with those of immunologically intact persons with medically controlled and uncontrolled HIV.
METHODS
Site Selection and Data Collection
The HIV Research Network (HIVRN) is a consortium that includes 12 sites providing longitudinal adult HIV care in 10 US cities. Sites abstract comprehensive data from clinical records, deidentify these data, and submit them to a data coordinating center for integration into a uniform database. Eleven of the participating sites submit hospitalization data and were able to participate in chart reviews for the purpose of this study (5 Northeast, 3 West, and 3 South). Nine of these sites have academic affiliations and 2 are community-based. Inclusion in this retrospective cohort study was restricted to persons who were in active care (defined as having ≥1 outpatient primary HIV care visit, 1 CD4 T-cell count, and 1 HIV-1 RNA level during the calendar year) at these sites between 2005 and 2011. All sites contributed data for all years, except one that was not included in 2005 because of incomplete data.
The analysis was limited to persons considered immunologically intact. Person-years were excluded if they contained 2 consecutive CD4 cell counts <350 cells/mm3 or any single count <200 cells/mm3. If 2 consecutive measurements <350 cells/mm3 spanned separate calendar years, both calendar years were excluded. Participants could contribute additional observation time to the analysis after consistent CD4 cell reconstitution to >350 cells/mm3 occurred. Nadir CD4 cell count was not considered a criterion for study participation, and participants with CD4 cell reconstitution after any CD4 cell count nadir were eligible to contribute observation time during calendar years after the first year of consistent CD4 cell reconstitution to >350 cells/mm3. All participants in this study either provided informed consent for inclusion in the HIVRN research database or had a waiver of informed consent granted by their local institutional review board. Institutional review boards at each site and at the data coordinating center at Johns Hopkins University approved the collection and use of these data for analysis and publication.
HIV Control Status
Elite control was defined by ≥3 consecutive HIV-1 RNA measurements, on separate days and spanning a period of ≥12 months, registering below the limit of detection for the assay in the absence of any ART. This definition of elite control has been used in several prior reports [3, 16, 17]. Accrual of elite control observation time began only after 1 full year of undetectable HIV-1 RNA levels to minimize misclassification due to any missing ART data. (Laboratory data before 2005 were used to establish status in 2005). During elite control, detectable HIV-1 RNA levels <1000 copies/mL were permissible as long as such episodes represented the minority of measurements during the calendar year. The calendar year during which the elite control period ends was not considered an elite control person-year. Medical records of elite controllers identified via this algorithm were manually reviewed to confirm elite control status. Individuals were excluded from the analysis if medical record review was not possible. Elite controllers were not eligible to contribute observation time to other HIV control categories.
Individuals not identified as elite controllers could contribute observation time to the following groups: medical control, low viremia, and/or high viremia. Medical control was defined by ≥3 consecutive HIV-1 RNA measurements, on separate days and spanning ≥12 months, which registered below the limit of detection for the assay, while ART was prescribed. Medical control began during the first qualifying year starting with an undetectable HIV-1 RNA level. Detectable HIV-1 RNA levels <1000 copies/mL were permissible after establishment of medical control if they represented a minority of measurements during any calendar year.
Exploratory data analysis suggested a difference in hospitalization rates above and below the HIV-1 RNA threshold of 1000 copies/mL. Therefore, low viremia was defined by all HIV-1 RNA measurements in the calendar year falling below this threshold, but not satisfying other criteria for medical control. All person-years with HIV-1 RNA measurement(s) ≥1000 copies/mL were considered high-viremia person-years. HIV control status was assessed annually and participants who were not identified as elite controllers could transition between other HIV control categories with each change in calendar year. A sensitivity analysis was performed in which participants could contribute person-time to only one HIV control status category, with data censored at the time of transition from that category.
Covariates
Age was assessed annually on 1 July. Race/ethnicity and sex were categorized based on self-report. HIV transmission risk factors were divided into mutually exclusive categories: injection drug use (IDU), men who have sex with men (MSM), heterosexual transmission, or other/unknown. Individuals who reported IDU in addition to any other risk factor were categorized as IDU. Men who reported sex with both men and women were categorized as MSM. Hepatitis B surface antigen and hepatitis C virus (HCV) antibody were used to determine hepatitis status, and this assessment was updated annually. Outpatient HIV primary care visits were tallied annually. Insurance status, CD4 cell counts, and HIV-1 RNA measurements were updated with the first available assessment for each calendar year of observation. Participants with dual eligibility for Medicaid and Medicare were included in the Medicare category.
Outcomes
The primary outcome was all-cause hospitalization, and this was ascertained using admission and discharge dates that are reported by all HIVRN sites. We also investigated cause-specific hospitalization rates within the subgroup of 9 HIVRN sites that had International Classification of Diseases, Ninth Revision. (ICD-9) diagnosis code data available for each hospitalization. Hospitalizations were assigned to one of 18 diagnostic categories using a previously published algorithm [18, 19]. First, the primary diagnostic code was identified as the first-listed ICD-9 code that did not refer to HIV (042, V08, 795.71, V01.79), chronic hepatitis B virus (HBV) (070.22, 070.23, 070.32, 070.33), chronic HCV (070.44, 070.54, 070.70, 070.71), or oral candidiasis (112.0), because these represent comorbid conditions frequently recorded for billing purposes but insufficient to justify hospitalization. Second, Clinical Classifications Software (CCS) was used to assign the primary ICD-9 code into one of 18 “first-level” CCS categories [20]. Finally, we modified the CCS diagnostic categories by reassigning infections (eg, pneumonia) from organ system categories to the infection category; combining the congenital, perinatal, and unclassified categories (together representing 1% of admissions); and reassigning specific infections and malignancies into a new AIDS-defining illness category, according to Centers for Disease Control and Prevention criteria [21].
Data Analysis
All-cause and cause-specific hospitalization rates were calculated using total number of visits as the numerator and aggregate person-time as the denominator and multiplied by 100 to obtain rates per 100 person-years. Participants could contribute <1 year of observation during a calendar year owing to death or new enrollment in care. Univariable and multivariable negative binomial regression models were used to estimate incidence rate ratios (IRRs) for hospitalization rates associated with HIV control status, age, race, sex, HIV risk factor, CD4 cell count stratum, hepatitis status, number of primary HIV care visits and insurance status. Multivariable models included indicators for clinical care site to adjust for site-specific variability and for calendar year to adjust for secular trends.
All models used generalized estimating equations, clustered on person, with unstructured working correlation, robust variance estimators, and an offset for person-time. This technique adjusts the variance to account for multiple hospitalization events by a single person, including when these events occur under different exposure categories (eg, under low viremia in 1 year and medical control in a separate year) [22]. A 2-sided type I error of 5% was considered statistically significant. All analyses were performed using Stata 12.0 software (StataCorp).
RESULTS
Of 34 354 persons actively engaged in care during the study period, 23 461 satisfied the CD4 cell count criterion for inclusion in this analysis and contributed 64 290 person-years of observation. Algorithmically, 188 persons were identified as elite control candidates in the HIVRN database. Of these, 17 were excluded from the analysis because they did not have clinical records available for review and 22 were excluded because manual record review revealed they were prescribed ART. Ultimately, 149 chart-confirmed elite controllers contributed 369 person-years of observation time. The prevalence of elite control is therefore estimated to be ≥0.43% of the full population of 34 354 persons.
Among the remaining 23 273 participants, 9226 contributed 26 176 person-years of medical control, 12 044 contributed 17 313 person-years of low viremia, and 12 847 contributed 20 414 person-years of high-viremia observation time. At study entry, persons with elite control were more likely than those with medical control to be female (50.3% vs 25.7%; P < .001) and black (58.4% vs 40.9%; P < .001) (Table 1). CD4 cell counts at study entry were higher in the elite control group (median [interquartile range], 778 [580–961] cells/mm3) than in the medical control, low-viremia, and high-viremia groups (481 [396–640], 510 [401–677], and 482 [384–634] cells/mm3, respectively; overall P < .001).
Table 1.
Demographic and Clinical Characteristics at Study Entry by HIV Control Statusa
| Characteristic | Medical Control (n = 4709) | Elite Control (n = 149) | Low Viremia (n = 7998) | High Viremia (n = 10 605) |
|---|---|---|---|---|
| Age, y | ||||
| Median (IQR)b | 45.4 (39.4–51.7) | 46.4 (40.5–53.2) | 44.2 (37.4–50.5) | 40.3 (31.0–47.4) |
| 18–34 | 605 (12.8) | 21 (14.1) | 1558 (19.5) | 3690 (34.8) |
| 35–49 | 2625 (55.7) | 76 (51.0) | 4297 (53.7) | 5064 (47.8) |
| 50–64 | 1360 (28.9) | 49 (32.9) | 1974 (24.7) | 1727 (16.3) |
| ≥65 | 119 (2.5) | 3 (2.0) | 169 (2.1) | 124 (1.2) |
| Race/ethnicity | ||||
| White | 1566 (33.3) | 35 (23.5) | 2403 (30.0) | 2740 (25.8) |
| Black | 1927 (40.9) | 87 (58.4) | 3374 (42.2) | 5331 (50.3) |
| Hispanic | 1097 (23.3) | 25 (16.8) | 1968 (24.6) | 2241 (21.1) |
| Other/unknown | 119 (2.5) | 2 (1.3) | 253 (3.2) | 293 (2.8) |
| Sex | ||||
| Male | 3498 (74.3) | 74 (49.7) | 5976 (74.7) | 7410 (69.9) |
| Female | 1211 (25.7) | 75 (50.3) | 2022 (25.3) | 3195 (30.1) |
| HIV risk factor | ||||
| Heterosexual | 1701 (36.1) | 59 (39.6) | 2763 (34.6) | 4074 (38.4) |
| MSM | 2081 (44.2) | 39 (26.2) | 3618 (45.2) | 4368 (41.2) |
| IDU | 782 (16.6) | 45 (30.2) | 1247 (15.6) | 1714 (16.2) |
| Other/unknown | 145 (3.1) | 6 (4.0) | 370 (4.6) | 449 (4.2) |
| CD4 cell count, cells/mm3c | ||||
| Median (IQR) | 481 (396–640) | 778 (580–961) | 510 (401–677) | 482 (384–634) |
| 200–350 | 347 (7.4) | 1 (0.7) | 830 (10.4) | 1650 (15.6) |
| 351–500 | 2199 (46.7) | 19 (12.8) | 3030 (37.9) | 4044 (38.1) |
| 501–750 | 1411 (30.0) | 45 (30.2) | 2679 (33.5) | 3350 (31.6) |
| >750 | 752 (16.0) | 84 (56.4) | 1459 (18.2) | 1561 (14.7) |
| HIV-1 RNA, median (IQR), copies/mLc | Undetectabled | Undetectabled | Undetectabled | 7640 (1124–36 702) |
| Hepatitis statuse | ||||
| HIV monoinfection | 2397 (50.9) | 55 (36.9) | 3637 (45.5) | 5289 (49.9) |
| HIV/HBV coinfection | 157 (3.3) | 2 (1.3) | 216 (2.7) | 240 (2.3) |
| HIV/HCV coinfection | 497 (10.6) | 34 (22.8) | 790 (9.9) | 1055 (10.0) |
| HIV/HBV/HCV triple infection | 55 (1.2) | 2 (1.3) | 105 (1.3) | 150 (1.4) |
| Unknown | 1603 (34.0) | 56 (37.6) | 3250 (40.6) | 3871 (36.5) |
| Annual outpatient HIV visits | ||||
| Median (IQR) | 4 (3–7) | 4 (2–7) | 4 (2–6) | 4 (2–6) |
| 1–2 | 793 (16.8) | 52 (34.9) | 2224 (27.8) | 3694 (34.8) |
| 3–4 | 1659 (35.2) | 36 (24.2) | 2487 (31.1) | 2932 (27.6) |
| 5–6 | 950 (20.2) | 18 (12.1) | 1498 (18.7) | 1698 (16.0) |
| ≥7 | 1307 (27.8) | 43 (28.9) | 1789 (22.4) | 2281 (21.5) |
| Insurancec | ||||
| Private | 638 (13.6) | 18 (12.1) | 1109 (13.9) | 1238 (11.7) |
| Medicaid | 1358 (28.8) | 69 (46.3) | 2864 (35.8) | 3629 (34.2) |
| Medicare/dual eligible | 995 (21.1) | 12 (8.0) | 1206 (15.1) | 1078 (10.2) |
| Ryan White/uninsured | 1497 (31.8) | 46 (30.9) | 2564 (32.1) | 4160 (39.2) |
| Unknown | 221 (4.7) | 4 (2.7) | 255 (3.2) | 50 (4.7) |
| Year of study entry | ||||
| 2005 | 1859 (39.5) | 53 (35.6) | 1501 (18.8) | 2998 (28.3) |
| 2006 | 1029 (21.8) | 26 (17.4) | 850 (10.6) | 1615 (15.2) |
| 2007 | 376 (8.0) | 14 (9.4) | 838 (10.5) | 1149 (10.8) |
| 2008 | 333 (7.1) | 12 (8.0) | 904 (11.3) | 1227 (11.6) |
| 2009 | 347 (7.4) | 16 (10.7) | 926 (11.6) | 1031 (9.7) |
| 2010 | 409 (8.7) | 14 (9.4) | 1003 (12.5) | 1182 (11.2) |
| 2011 | 356 (7.6) | 14 (9.4) | 1976 (24.7) | 1403 (13.2) |
Abbreviations: HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IDU, injection drug use; IQR, interquartile range; MSM, men who have sex with men.
a Unless otherwise specified, data represent No. (%) of study participants.
b Age was assessed on 1 July of the year of study entry.
c CD4 cell counts, HIV-1 RNA levels, and insurance data used were the first available for the calendar year of study entry.
d Undetectable HIV-1 RNA refers to a level below the limit of detection for the assay used during routine clinical care, which may have been 20, 40, 48, 50, 75, 80, or 400 copies/mL.
e HBV and HCV status were determined by hepatitis B surface antigen and hepatitis C antibody, respectively. Hepatitis status at study entry was assessed by positive tests before entry or within 6 months after entry, or by negative test result at first testing any time after study entry.
There were 8456 hospitalizations among all participants. The percentage of participants ever hospitalized during time accrued in each HIV control group was 14.6% during medical control, 25.5% during elite control, 10.6% during low viremia and 15.0% during high viremia. Overall, the elite control group had the highest all-cause hospitalization rate, 23.3 hospitalizations per 100 person-years (range, 15.4–27.9), followed by the high-viremia (16.9; range, 14.1–18.8), low-viremia (12.6; range, 12.0–13.7), and medical control (10.5; range, 8.6–12.2) groups (Figure 1).
Figure 1.
Unadjusted all-cause hospitalization rates by human immunodeficiency virus control status. Rates were calculated using total number of visits as the numerator and aggregate person-time as the denominator for each calendar year and standardized as hospitalizations per 100 person-years of follow-up.
In our multivariable model, elite control was associated with a higher hospitalization rate than medical control (adjusted IRR [aIRR], 1.77; 95% confidence interval [CI], 1.21–2.60) (Table 2). Hospitalization rates were also elevated in the high-viremia (aIRR, 1.71; 95% CI, 1.57–1.87), and low-viremia (aIRR, 1.34; 95% CI, 1.24–1.46) groups, compared with the medical control group. Other factors independently associated with hospitalization included older age, female sex, IDU, lower CD4 cell count, HIV/HCV coinfection, HIV/chronic HBV/HCV triple infection, more outpatient visits, and Medicaid or Medicare insurance (vs private insurance). In the sensitivity analysis in which participants were censored at the time of transition from their first recorded HIV control category, elite control was again associated with a higher hospitalization rate than medical control in both unadjusted (IRR, 2.69; 95% CI, 1.49–4.85) and adjusted (aIRR, 2.02; 95% CI, 1.24–3.28) models.
Table 2.
Univariable and Multivariable Analyses of Factors Associated With All-Cause Hospitalization
| Characteristic | Incidence Rate Ratio (95% CI) |
|
|---|---|---|
| Unadjusted | Adjusteda | |
| HIV control status | ||
| Medical control | 1.0 (Reference) | 1.0 (Reference) |
| Elite control | 2.08 (1.34–3.23) | 1.77 (1.21–2.60) |
| Low viremia | 1.18 (1.98–1.29) | 1.34 (1.24–1.46) |
| High viremia | 1.56 (1.43–1.69) | 1.71 (1.57–1.87) |
| Age, yb | ||
| 18–34 | 1.0 (Reference) | 1.0 (Reference) |
| 35–49 | 1.22 (1.09–1.36) | 1.02 (.91–1.14) |
| 50–64 | 1.72 (1.53–1.93) | 1.26 (1.12–1.42) |
| ≥65 | 2.85 (2.35–3.45) | 2.00 (1.64–2.44) |
| Race | ||
| White | 1.0 (Reference) | 1.0 (Reference) |
| Black | 1.31 (1.19–1.44) | 1.00 (.90–1.12) |
| Hispanic | 0.95 (.84–1.07) | 0.80 (.71–.91) |
| Other/unknown | 0.68 (.48–.94) | 0.69 (.50–.95) |
| Sex | ||
| Male | 1.0 (Reference) | 1.0 (Reference) |
| Female | 1.54 (1.42–1.68) | 1.31 (1.18–1.45) |
| HIV risk factorc | ||
| Heterosexual | 1.0 (Reference) | 1.0 (Reference) |
| MSM | 0.66 (.60–.72) | 0.88 (.78–.99) |
| IDU | 1.60 (1.44–1.78) | 1.19 (1.06–1.33) |
| Other/unknown | 1.47 (1.20–1.82) | 1.34 (1.08–1.66) |
| CD4 cell count, cells/mm3d | ||
| >750 | 1.0 (Reference) | 1.0 (Reference) |
| 501–750 | 1.17 (1.07–1.28) | 1.16 (1.06–1.27) |
| ≤500 | 1.39 (1.26–1.52) | 1.33 (1.21–1.46) |
| Hepatitis statuse | ||
| HIV monoinfection | 1.0 (Reference) | 1.0 (Reference) |
| HIV/HBV coinfection | 1.07 (.83–1.38) | 1.12 (.87–1.45) |
| HIV/HCV coinfection | 2.08 (1.86–2.32) | 1.35 (1.20–1.52) |
| HIV/HBV/HCV triple infection | 1.47 (1.03–2.09) | 1.55 (1.09–2.18) |
| Unknown | 1.15 (1.05–1.26) | 1.05 (.95–1.15) |
| Annual outpatient HIV visits | ||
| 1–2 | 1.0 (Reference) | 1.0 (Reference) |
| 3–4 | 0.96 (.88–1.06) | 0.98 (.89–1.08) |
| 5–6 | 1.28 (1.16–1.42) | 1.27 (1.15–1.42) |
| ≥7 | 2.53 (2.30–2.78) | 2.40 (2.15–2.67) |
| Insuranced | ||
| Private | 1.0 (Reference) | 1.0 (Reference) |
| Medicaid | 2.39 (2.10–2.71) | 1.89 (1.65–2.16) |
| Medicare/dual eligible | 2.40 (2.09–2.75) | 1.95 (1.69–2.25) |
| Ryan White/uninsured | 1.08 (.94–1.23) | 1.10 (.95–1.27) |
| Unknown | 0.97 (.78–1.20) | 1.05 (.83–1.34) |
| Calendar year | ||
| 2005 | 1.0 (Reference) | 1.0 (Reference) |
| 2006 | 1.00 (.89–1.12) | 1.01 (.90–1.14) |
| 2007 | 1.02 (.91–1.15) | 1.05 (.93–1.18) |
| 2008 | 1.00 (.89–1.12) | 1.05 (.94–1.18) |
| 2009 | 0.97 (.87–1.09) | 1.04 (.92–1.16) |
| 2010 | 0.89 (.90–1.00) | 0.99 (.88–1.11) |
| 2011 | 0.77 (.69–.87) | 0.89 (.79–1.00) |
Abbreviations: CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IDU, injection drug use; MSM, men who have sex with men.
a The multivariable model included all listed variables and also an indicator variable for clinical care site. Results in bold are statistically significant (P ≤ .05).
b Age was assessed on July 1 of each calendar year.
c HIV risk factors were considered mutually exclusive; subjects who reported IDU in addition to any other risk factor were categorized as IDU, men who reported sex with men and women were categorized as MSM.
d CD4 cell counts and insurance data used in this analysis were the first available measurements for each calendar year.
e HBV and HCV status were determined by hepatitis B surface antigen and hepatitis C antibody, respectively, as measured before 1 July of each calendar year. Negative test results were carried backward to the time of study entry, and positive test results were carried forward for all subsequent years.
Among the subgroup of 9 clinical care sites with available ICD-9 data, 5593 total hospitalizations were observed (Table 3). Overall, non–AIDS-defining infections were the most common reason for hospitalization, representing 24.1% of admissions. In the elite control group, however, non–AIDS-defining infections accounted for just 2.7% of admissions. Conversely, cardiovascular disease was the reason for 13.5% of admissions overall but was more common among elite controllers, accounting for 31.1% of admissions. Among elite controllers, the most common diagnoses leading to cardiovascular hospitalization were chest pain (26.1%), coronary artery disease (13.0%), and heart failure (13.0%). Pulmonary disease accounted for 4.8% of admissions overall but 21.6% of admissions in the elite control group. Among elite controllers, the most common diagnosis leading to pulmonary hospitalization was asthma/chronic obstructive pulmonary disease (87.5%). Two hospitalizations for AIDS-defining-illness occurred in the elite control group, one for Kaposi sarcoma and one for Mycobacterium tuberculosis.
Table 3.
Hospitalizations by Diagnostic Category at 9 Sites With ICD-9 Data
| Diagnostic Categorya | Hospitalizations, No. (%) |
||||
|---|---|---|---|---|---|
| Overall (n = 5593) | Medical Control (n = 1999) | Elite Control (n = 74) | Low Viremia (n = 1341) | High Viremia (n = 2179) | |
| Non–AIDS-defining infection | 1347 (24.1) | 394 (19.7) | 2 (2.7) | 318 (23.7) | 633 (29.0) |
| Cardiovascular | 756 (13.5) | 335 (16.8) | 23 (31.1) | 167 (12.4) | 231 (10.6) |
| Gastrointestinal/liver | 521 (9.3) | 243 (12.2) | 5 (6.8) | 107 (8.0) | 166 (7.6) |
| Psychiatric | 444 (7.9) | 114 (5.7) | 7 (9.5) | 92 (6.9) | 231 (10.6) |
| Endocrine | 346 (6.2) | 136 (6.8) | 3 (4.0) | 98 (7.3) | 109 (5.0) |
| Injury/poisoning | 302 (5.4) | 123 (6.2) | 2 (2.7) | 80 (6.0) | 97 (4.4) |
| Renal | 298 (5.3) | 129 (6.4) | 3 (4.0) | 62 (4.6) | 104 (4.8) |
| Pulmonary | 267 (4.8) | 84 (4.2) | 16 (21.6) | 66 (4.9) | 101 (4.6) |
| Non–AIDS-defining cancer | 247 (4.4) | 104 (5.2) | 3 (4.0) | 74 (5.5) | 66 (3.0) |
| Orthopedic | 205 (3.7) | 85 (4.2) | 0 (0) | 41 (3.1) | 79 (3.6) |
| Neurologic | 194 (3.5) | 66 (3.3) | 0 (0) | 63 (4.7) | 65 (3.0) |
| Symptom based | 154 (2.8) | 50 (2.5) | 2 (2.7) | 35 (2.6) | 67 (3.1) |
| Hematologic | 143 (2.6) | 41 (2.0) | 4 (5.4) | 41 (3.1) | 57 (2.6) |
| AIDS-defining illness | 125 (2.2) | 38 (1.9) | 2 (2.7) | 22 (1.6) | 63 (2.9) |
| Obstetric/gynecologic | 106 (1.9) | 18 (0.9) | 0 (0) | 34 (2.5) | 54 (2.5) |
| Congenital/perinatal/unclassified | 57 (1.0) | 19 (1.0) | 1 (1.4) | 18 (1.4) | 19 (0.9) |
| Missing | 46 (0.8) | 13 (0.6) | 0 (0) | 14 (1.0) | 19 (0.9) |
| Dermatologic | 35 (0.6) | 7 (0.4) | 1 (1.4) | 9 (0.7) | 18 (0.8) |
Abbreviation: ICD-9, International Classification of Diseases, Ninth Revision.
a Diagnostic categories are listed in order of frequency in the overall study population. Results are given for the 9 sites with available hospitalization diagnostic data.
Multivariable models were used to explore factors associated with all-cause and diagnostic category-specific hospitalization for the 5 most common diagnostic categories at sites with available ICD-9 data (Table 4). Inferences about factors associated with all-cause hospitalization were similar to those in the overall study cohort, including a similarly heightened rate among persons with elite control as compared with medical control (aIRR, 1.99; 95% CI, 1.29–3.06). The cardiovascular hospitalization rate was significantly higher among elite controllers (aIRR, 3.19; 95% CI, 1.50–6.79) than in the medical control group. Psychiatric hospitalization rates were higher in the elite control (aIRR, 3.98; 95% CI, 1.54–10.28), low-viremia (aIRR, 1.65; 95% CI, 1.15–2.37); and high-viremia (3.14; 2.35–4.2) groups than in the medical control group. Compared with medical control, elite control was associated with a trend toward a lower non–AIDS-defining infection hospitalization rate (aIRR, 0.32; 95% CI, .08–1.30), whereas higher rates were seen in the low-viremia (aIRR, 1.57; 95% CI, 1.31–1.88) and high-viremia (aIRR, 2.48; 95% CI, 2.10–2.93) groups.
Table 4.
Multivariable Analyses of Factors Associated With Cause-Specific Hospitalization at 9 Sites With ICD-9 Dataa
| Characteristic | Adjusted Incidence Rate Ratio (95% CI)a |
|||||
|---|---|---|---|---|---|---|
| All-Cause | Non–AIDS-Defining Infection | Cardiovascular | Gastrointestinal/Liver | Psychiatric | Endocrine | |
| HIV control status | ||||||
| Medical control | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| Elite control | 1.99 (1.29–3.06) | 0.32 (.08–1.30) | 3.19 (1.50–6.79) | 1.32 (.52–3.33) | 3.98 (1.54–10.28) | 1.14 (.32–4.06) |
| Low viremia | 1.39 (1.26–1.54) | 1.57 (1.31–1.88) | 1.11 (.87–1.42) | 0.91 (.68–1.23) | 1.65 (1.15–2.37) | 1.54 (1.14–2.08) |
| High viremia | 1.79 (1.61–1.99) | 2.48 (2.10–2.93) | 1.18 (.92–1.51) | 1.15 (.85–1.54) | 3.14 (2.35–4.21) | 1.42 (1.00–1.99) |
| Age, yb | ||||||
| 18–34 | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| 35–49 | 1.10 (.94–1.28) | 1.09 (.87–1.36) | 2.20 (1.25–3.88) | 1.23 (.81–1.87) | 1.07 (.64–1.78) | 1.72 (1.00–2.94) |
| 50–64 | 1.36 (1.15–1.59) | 1.18 (.92–1.51) | 3.63 (2.05–6.44) | 1.26 (.79–2.02) | 0.72 (.42–1.26) | 2.33 (1.38–3.93) |
| ≥65 | 2.24 (1.77–2.83) | 1.44 (1.00–2.07) | 7.86 (4.06–15.20) | 2.25 (1.25–4.08) | 0.94 (.35–2.49) | 3.95 (2.02–7.72) |
| Race | ||||||
| White | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| Black | 1.05 (.92–1.20) | 0.85 (.69–1.04) | 1.25 (.92–1.69) | 0.94 (.68–1.30) | 1.17 (.80–1.71) | 1.39 (.96–2.03) |
| Hispanic | 0.84 (.72–.97) | 0.71 (.56–.88) | 1.04 (.75–1.43) | 0.97 (.70–1.35) | 0.79 (.48–1.30) | 1.10 (.66–1.86) |
| Other/unknown | 0.75 (.52–1.09) | 0.53 (.29–.96) | 1.08 (.48–2.46) | 0.50 (.17–1.42) | 0.80 (.33–1.97) | 0.51 (.16–1.67) |
| Sex | ||||||
| Male | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| Female | 1.23 (1.09–1.39) | 1.19 (.99–1.42) | 0.98 (.76–1.26) | 1.09 (.80–1.48) | 0.93 (.66–1.33) | 1.30 (.91–1.86) |
| HIV risk factorc | ||||||
| Heterosexual | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| MSM | 0.83 (.72–.96) | 1.08 (.88–1.33) | 0.62 (.46–.85) | 0.92 (.64–1.30) | 0.96 (.63–1.46) | 0.73 (.46–1.14) |
| IDU | 1.23 (1.06–1.42) | 1.56 (1.26–1.92) | 0.71 (.51–.97) | 1.47 (1.03–2.10) | 2.15 (1.44–3.33) | 1.21 (.77–1.89) |
| Other/unknown | 1.21 (.96–1.53) | 1.23 (.87–1.74) | 0.74 (.46–1.20) | 1.20 (.71–2.04) | 1.58 (.61–4.08) | 1.79 (.93–3.48) |
| CD4 cell count, cells/mm3d | ||||||
| >750 | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| 501–750 | 1.11 (.99–1.23) | 1.08 (.90–1.28) | 0.89 (.70–1.13) | 1.11 (.83–1.50) | 1.32 (.95–1.84) | 1.18 (.82–1.71) |
| ≤500 | 1.28 (1.15–1.43) | 1.28 (1.08–1.52) | 1.00 (.78–1.30) | 1.17 (.86–1.58) | 1.16 (.82–1.63) | 1.18 (.83–1.69) |
| Hepatitis statuse | ||||||
| HIV monoinfection | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| HIV/HBV coinfection | 1.07 (.78–1.46) | 1.21 (.87–1.69) | 0.72 (.32–1.63) | 1.03 (.54–1.99) | 0.87 (.38–1.98) | 0.94 (.34–2.58) |
| HIV/HCV coinfection | 1.36 (1.17–1.57) | 1.33 (1.07–1.66) | 1.53 (1.11–2.10) | 1.33 (.94–1.90) | 1.29 (.85–1.96) | 1.40 (.90–2.16) |
| HIV/HBV/HCV triple infection | 1.40 (.95–2.08) | 1.18 (.78–1.80) | 2.41 (1.17–4.96) | 1.51 (.80–2.86) | 0.25 (.04–1.74) | 0.80 (.32–2.01) |
| Unknown | 1.06 (.94–1.20) | 0.94 (.78–1.13) | 1.27 (.98–1.65) | 1.13 (.84–1.52) | 1.48 (1.02–2.17) | 0.73 (.48–1.09) |
| Annual outpatient HIV visits | ||||||
| 1–2 | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| 3–4 | 0.93 (.82–1.06) | 0.81 (.65–1.01) | 0.87 (.58–1.30) | 1.04 (.73–1.48) | 0.77 (.54–1.09) | 1.06 (.70–1.60) |
| 5–6 | 1.30 (1.13–1.48) | 1.27 (1.01–1.59) | 1.24 (.83–1.84) | 1.52 (1.03–2.24) | 0.97 (.65–1.44) | 1.96 (1.24–3.11) |
| ≥7 | 2.68 (2.34–3.07) | 2.64 (2.11–3.31) | 3.04 (2.02–4.57) | 3.88 (2.76–5.46) | 2.01 (1.43–2.82) | 3.53 (2.32–5.36) |
| Insuranced | ||||||
| Private | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| Medicaid | 1.72 (1.47–2.02) | 1.91 (1.47–2.47) | 2.15 (1.42–3.27) | 1.21 (.82–1.78) | 2.47 (1.55–3.93) | 1.57 (.91–2.72) |
| Medicare/dual eligible | 1.81 (1.53–2.15) | 2.08 (1.59–2.72) | 2.08 (1.38–3.13) | 1.29 (.86–1.92) | 2.78 (1.72–4.48) | 2.45 (1.47–4.08) |
| Ryan White/uninsured | 1.08 (.91–1.28) | 1.26 (.96–1.67) | 1.44 (.94–2.21) | 0.70 (.45–1.09) | 1.94 (1.21–3.10) | 1.14 (.65–2.02) |
| Unknown | 0.99 (.75–1.30) | 1.05 (.64–1.72) | 1.98 (1.07–3.67) | 0.56 (.22–1.45) | 1.51 (.66–3.48) | 0.43 (.08–2.26) |
| Calendar year | ||||||
| 2005 | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) | 1.0 (Reference) |
| 2006 | 0.95 (.81–1.10) | 0.85 (.66–1.10) | 1.43 (.97–2.12) | 0.94 (.58–1.52) | 0.49 (.32–0.76) | 1.33 (.75–2.38) |
| 2007 | 1.10 (.95–1.27) | 1.01 (.78–1.31) | 1.41 (.97–2.04) | 1.06 (.67–1.69) | 0.68 (.43–1.08) | 1.68 (1.02–2.74) |
| 2008 | 1.04 (.90–1.20) | 0.94 (.74–1.20) | 1.26 (.87–1.84) | 1.18 (.75–1.86) | 0.74 (.49–1.12) | 1.42 (.83–2.44) |
| 2009 | 1.04 (.90–1.19) | 0.92 (.72–1.16) | 1.13 (.78–1.65) | 1.49 (.97–2.28) | 0.42 (.28–.64) | 1.82 (1.02–3.23) |
| 2010 | 1.01 (.88–1.16) | 0.88 (.69–1.13) | 1.33 (.92–1.93) | 1.42 (.93–2.15) | 0.74 (.51–1.10) | 1.47 (.87–2.47) |
| 2011 | 0.96 (.84–1.11) | 0.99 (.78–1.26) | 0.88 (.60–1.27) | 1.30 (.85–1.99) | 0.42 (.26–.67) | 1.55 (.92–2.62) |
Abbreviations: CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus; ICD-9, International Classification of Diseases, Ninth Revision; IDU, injection drug use; MSM, men who have sex with men.
a The multivariable models used to calculate adjusted incidence rate ratios included all listed variables and also an indicator variable for clinical care site. Results in bold are statistically significant (P ≤ .05).
b Age was assessed on July 1 of each calendar year.
c HIV risk factors were considered mutually exclusive; subjects who reported IDU in addition to any other risk factor were categorized as IDU, men who reported sex with men and women were categorized as MSM.
d CD4 cell counts and insurance data used in these analyses were the first available measurements for each calendar year.
e HBV and HCV status were determined by hepatitis B surface antigen and hepatitis C antibody, respectively, as measured before 1 July of each calendar year. Negative test results were carried backward to the time of study entry, and positive test results were carried forward for all subsequent years.
The elite control group included a single participant with 21 hospitalizations during the study period. A sensitivity analysis was performed excluding this person. In this analysis, the all-cause hospitalization rate in the elite control group was attenuated, but remained significantly greater than in the medical control reference group (aIRR 1.56; 95% CI, 1.09–2.24). Many of this person's admissions were related to asthma/chronic obstructive pulmonary disease, and in the sensitivity analysis pulmonary admissions accounted for just 3.8% of admissions in the elite control group, similar to the overall study population (4.5%). Cardiovascular disease remained the most common reason for admission among elite controllers (30.2%).
DISCUSSION
This study identified an increased rate of hospitalization among persons with elite control of HIV as compared with medical control. Furthermore, elite control was associated with a higher rate of cardiovascular hospitalization, which was the most common type of hospitalization among elite controllers. Non–AIDS-defining infections, though a common reason for admission in the overall study population, were relatively rare among elite controllers. Finally, female sex was associated with elite control.
To date, data on hospitalizations among elite controllers are scarce. Two previous studies reporting no increased hospitalization risk for elite controllers both included only 25 elite controllers and examined hospitalization as a secondary outcome [3, 23].
We are unaware of previous studies specifically evaluating cardiovascular hospitalizations among elite controllers. Prior studies have shown an increased prevalence of atherosclerotic plaques among elite controllers compared with HIV-negative controls but not compared with PLWH treated with ART [14, 15]. In addition, elite controllers have been shown to have elevated levels of activated CD8 T cells, D-dimer, soluble tissue factor, and interferon γ–induced protein 10 levels compared with both HIV-seronegative persons and persons with medically controlled HIV infection [7–9]. The cause of immune activation in elite controllers is unclear, but the presence of replication-competent virus [24], persistent low-level viremia [25], ongoing low-level viral replication [26–28], and the vigorous immune response that controls this viral replication [29] may be important factors. These pathophysiologic factors could contribute to higher hospitalization rates for cardiovascular disease among elite controllers.
If immune system activation is a contributor to the excess rate of cardiovascular and other hospitalizations among elite controllers, then ART and/or anti-inflammatory medicines, such as statins or aspirin, may be beneficial for this group. Prior studies have demonstrated that ART can decrease T-cell activation [28, 30–32] and/or increase CD4 cell counts [33, 34] among elite controllers. In our cohort, only 10 participants known to be elite controllers were ever exposed to ART, so we could not effectively evaluate differences in hospitalization. Prospectively assessing the effect of ART and/or anti-inflammatory medicines on hospitalization rates among elite controllers and understanding the potential role of chronic inflammation would inform the clinical care of elite controllers and may inform research on interventions that aim to induce an elite control state, such as some candidate HIV vaccines.
Non–AIDS-defining infections accounted for relatively few hospitalizations among elite controllers in our study (2.7%), despite being the most common reason for admission overall (24.1%). Elite control is characterized by a variety of unique immunologic parameters, including differences in class I antigen presenting molecules and CD8 T-cell populations that may also plausibly affect susceptibility to other infectious diseases [29]. Future studies should evaluate immunologic response to bacterial and other pathogens among elite controllers to clarify any protective mechanisms that contribute to the low rate of non–AIDS-defining infections in this population.
We are uncertain why elite controllers were hospitalized more frequently for psychiatric conditions. This association may warrant investigation in other cohorts.
Elite controllers in our study were more likely to be female than were persons in other HIV control categories. This is consistent with trends seen in several other studies of elite controllers, though in most studies the sex difference was not statistically significant [3, 7, 17, 35–38]. This is also consistent with prior observations that women tend to have lower HIV-1 RNA levels than men, particularly within the first year after seroconversion [39]. Few prior studies contained both an appropriate comparison group of non–elite controllers and sufficient power to detect a sex difference in the prevalence of elite control. If a sex difference is confirmed in additional studies, that may provide clues to the immunologic mechanisms underlying elite control.
Our study also provides an updated estimate of the prevalence of elite control among PLWH. Our estimate of 0.43% is within the range suggested by other studies [3–5]. It is likely that this number slightly underestimates the true prevalence, given that some potential elite controllers were excluded from the analysis because manual review of medical records was not possible.
A major strength of our study is its large sample of elite controllers. Among published cohorts, only the International HIV Controller Consortium reports a larger sample of elite controllers [17]. A limitation of our study is the potential for selection bias owing to our data being limited to persons actively engaged in HIV care. Because of no perceived need for ART, elite controllers might be less likely than other PLWH to be engaged in care, so that elite controllers with additional serious medical conditions are preferentially captured in our cohort. This could contribute to an increased hospitalization rate compared with the rate for members of the cohort whose only serious medical condition may be HIV infection. However, our study captured 93% of the elite controllers (14 of 15) otherwise known to local investigators at 2 clinical care sites with unrelated ongoing studies of elite controllers, suggesting both good sensitivity of our study definition for elite control and minimal impact of selection bias.
Duration of HIV infection may influence hospitalization rates, but this information was not available for participants in this cohort. Hospitalizations occurring outside each participant's HIV care institution may not have been completely captured, though we would not expect this to have a differential impact by HIV control status. This study was limited to person-years with high CD4 cell counts to provide meaningful comparators to the primary group of interest, elite controllers, who by nature of their disease process tend to have preserved CD4 cell counts. The medical control population does not necessarily reflect the full spectrum of patients seen routinely in HIV clinics, which may, at any given time, include a subset with low CD4 cell counts.
Another limitation is our lack of data on smoking, which are not routinely collected in this cohort but represent an important risk factor for all-cause and cardiovascular hospitalization. In general, smoking rates among PLWH are high, with 46%–67% categorized as current smokers by recent estimates [40–44]. We conducted a focused chart review to investigate smoking within our study population. All elite controllers were randomly matched with up to 4 persons with medical control, based on clinical care site, sex, race, and age in 5-year bands. Records of the 149 elite controllers and of 581 matched persons with medical control were then searched to determine ever versus never smoking status. Data were available for 134 elite controllers and 555 persons with medical control. We found that 110 of 134 elite controllers (82%) and 377 of 555 (68%) with medical control had ever smoked (P = .001). Smoking was thus common in both groups, and though the higher rate among elite controllers may have contributed to more hospitalizations, it is unlikely to explain the near doubling in adjusted all-cause hospitalization rate or the tripling in cardiovascular hospitalization rate associated with elite versus medical control.
This study demonstrates that elite control of HIV is associated with higher hospitalization rate than is medical control of HIV with ART. Cardiovascular disease seems to be a major driver of this association. Compared with other HIV-infected patients, elite controllers had relatively few admissions due to non–AIDS-defining infections. These findings could reflect clinical manifestations of ongoing immune activation in elite controllers. Further investigations are needed to evaluate the mechanisms underlying these associations and to clarify the potential benefit of ART and/or anti-inflammatory agents in the management of elite controllers.
Notes
HIVRN Participating Sites. Alameda County Medical Center, Oakland, California (H. E.); Children's Hospital of Philadelphia, Pennsylvania (R. M. R.); Trillium Health, Rochester, New York (R. B. C.); Drexel University, Philadelphia (Jeffrey Jacobson, MD; S. L. A.); Fenway Health, Boston, Massachusetts (Stephen Boswell, MD); Johns Hopkins University, Baltimore, Maryland (K. A. G.; R. D. M.; Allison Agwu, MD); Montefiore Medical Group, Bronx, New York (R. B.); Montefiore Medical Center, Bronx (L. H. H.); Oregon Health and Science University, Portland (P. T. K.); Parkland Health and Hospital System, Dallas, Texas (A. E. N.; Muhammad Akbar, MD); St Jude's Children's Hospital and University of Tennessee, Memphis (Aditya Gaur, MD); St Luke's Roosevelt Hospital Center, New York, New York (V. S., Stephen Arpadi, MD); Tampa General Health Care, Florida (C. S.); and University of California, San Diego (W. C. M.);
Sponsoring Agencies. Agency for Healthcare Research and Quality (Fred Hellinger, PhD; John Fleishman, PhD; Irene Fraser, PhD) and Health Resources and Services Administration (Robert Mills, PhD; Faye Malitz, MS), Rockville, Maryland
Data Coordinating Center. Johns Hopkins University (R. D. M.; Jeanne Keruly, CRNP; K. A. G., MD; Cindy Voss, MA; Nikki Balding, MS).
Financial support. This work was supported by the Agency for Healthcare Research and Quality (grant HHSA290201100007C), the Health Resources and Services Administration (grant HHSH250201200008C), the National Institute of Allergy and Infectious Diseases (grant K23 AI084854), and the National Center for Advancing Translational Sciences (grant KL2TR001103).
Potential conflicts of interest. K. A. G. has been a consultant to Tibotec and Bristol-Myers Squibb and received research funding from Tibotec. R. B. C. has been a consultant to Gilead and served on speakers bureaus for Gilead, Abbott, and Merck. C. S. has served on the speakers bureau for Gilead. All other authors report no potential conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.Deeks SG, Walker BD. Human immunodeficiency virus controllers: mechanisms of durable virus control in the absence of antiretroviral therapy. Immunity 2007; 27:406–16. [DOI] [PubMed] [Google Scholar]
- 2.O'Connell KA, Bailey JR, Blankson JN. Elucidating the elite: mechanisms of control in HIV-1 infection. Trends Pharmacol Sci 2009; 30:631–7. [DOI] [PubMed] [Google Scholar]
- 3.Okulicz JF, Marconi VC, Landrum ML, et al. Clinical outcomes of elite controllers, viremic controllers, and long-term nonprogressors in the US Department of Defense HIV natural history study. J Infect Dis 2009; 200:1714–23. [DOI] [PubMed] [Google Scholar]
- 4.Grabar S, Selinger-Leneman H, Abgrall S, Pialoux G, Weiss L, Costagliola D. Prevalence and comparative characteristics of long-term nonprogressors and HIV controller patients in the French Hospital Database on HIV. AIDS 2009; 23:1163–9. [DOI] [PubMed] [Google Scholar]
- 5.Sajadi MM, Constantine NT, Mann DL, et al. Epidemiologic characteristics and natural history of HIV-1 natural viral suppressors. J Acquir Immune Defic Syndr 2009; 50:403–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Okulicz JF, Lambotte O. Epidemiology and clinical characteristics of elite controllers. Curr Opin HIV AIDS 2011; 6:163–8. [DOI] [PubMed] [Google Scholar]
- 7.Krishnan S, Wilson EM, Sheikh V, et al. Evidence for innate immune system activation in HIV type 1-infected elite controllers. J Infect Dis 2014; 209:931–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Noel N, Boufassa F, Lecuroux C, et al. Elevated IP10 levels are associated with immune activation and low CD4+ T-cell counts in HIV controller patients. AIDS 2014; 28:467–76. [DOI] [PubMed] [Google Scholar]
- 9.Hunt PW, Brenchley J, Sinclair E, et al. Relationship between T cell activation and CD4+ T cell count in HIV-seropositive individuals with undetectable plasma HIV RNA levels in the absence of therapy. J Infect Dis 2008; 197:126–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Armah KA, McGinnis K, Baker J, et al. HIV status, burden of comorbid disease, and biomarkers of inflammation, altered coagulation, and monocyte activation. Clin Infect Dis 2012; 55:126–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Kuller LH, Tracy R, Belloso W, et al. Inflammatory and coagulation biomarkers and mortality in patients with HIV infection. PLoS Med 2008; 5:e203. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Rodger AJ, Fox Z, Lundgren JD, et al. Activation and coagulation biomarkers are independent predictors of the development of opportunistic disease in patients with HIV infection. J Infect Dis 2009; 200:973–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Klatt NR, Chomont N, Douek DC, Deeks SG. Immune activation and HIV persistence: implications for curative approaches to HIV infection. Immunol Rev 2013; 254:326–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pereyra F, Lo J, Triant VA, et al. Increased coronary atherosclerosis and immune activation in HIV-1 elite controllers. AIDS 2012; 26:2409–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hsue PY, Hunt PW, Schnell A, et al. Role of viral replication, antiretroviral therapy, and immunodeficiency in HIV-associated atherosclerosis. AIDS 2009; 23:1059–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Walker BD. Elite control of HIV infection: implications for vaccines and treatment. Top HIV Med 2007; 15:134–6. [PubMed] [Google Scholar]
- 17.Pereyra F, Jia X, McLaren PJ, et al. The major genetic determinants of HIV-1 control affect HLA class I peptide presentation. Science 2010; 330:1551–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Berry SA, Fleishman JA, Moore RD, Gebo KA. Trends in reasons for hospitalization in a multisite United States cohort of persons living with HIV, 2001-2008. J Acquir Immune Defic Syndr 2012; 59:368–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Gebo KA, Diener-West M, Moore RD. Hospitalization rates differ by hepatitis C status in an urban HIV cohort. J Acquir Immune Defic Syndr 2003; 34:165–73. [DOI] [PubMed] [Google Scholar]
- 20.Clinical Classifications Software (CCS), 2013. Rockville, MD: US Agency for Healthcare Research and Quality, 2013. [Google Scholar]
- 21.1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Recomm Rep 1992; 41:1–19. [PubMed] [Google Scholar]
- 22.Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics 1986; 42:121–30. [PubMed] [Google Scholar]
- 23.Lucero C, Torres B, Leon A, et al. Rate and predictors of non-AIDS events in a cohort of HIV-infected patients with a CD4 T cell count above 500 cells/mm3. AIDS Res Hum Retroviruses 2013; 29:1161–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Salgado M, Swanson MD, Pohlmeyer CW, et al. HLA-B*57 elite suppressor and chronic progressor HIV-1 isolates replicate vigorously and cause CD4+ T cell depletion in humanized BLT mice. J Virol 2014; 88:3340–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hatano H, Delwart EL, Norris PJ, et al. Evidence for persistent low-level viremia in individuals who control human immunodeficiency virus in the absence of antiretroviral therapy. J Virol 2009; 83:329–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.O'Connell KA, Brennan TP, Bailey JR, Ray SC, Siliciano RF, Blankson JN. Control of HIV-1 in elite suppressors despite ongoing replication and evolution in plasma virus. J Virol 2010; 84:7018–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Mens H, Kearney M, Wiegand A, et al. HIV-1 continues to replicate and evolve in patients with natural control of HIV infection. J Virol 2010; 84:12971–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chun TW, Shawn Justement J, Murray D, et al. Effect of antiretroviral therapy on HIV reservoirs in elite controllers. J Infect Dis 2013; 208:1443–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hersperger AR, Migueles SA, Betts MR, Connors M. Qualitative features of the HIV-specific CD8+ T-cell response associated with immunologic control. Curr Opin HIV AIDS 2011; 6:169–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Greenough TC, Sullivan JL, Desrosiers RC. Declining CD4 T-cell counts in a person infected with nef-deleted HIV-1. N Engl J Med 1999; 340:236–7. [DOI] [PubMed] [Google Scholar]
- 31.Sedaghat AR, Rastegar DA, O'Connell KA, Dinoso JB, Wilke CO, Blankson JN. T cell dynamics and the response to HAART in a cohort of HIV-1-infected elite suppressors. Clin Infect Dis 2009; 49:1763–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Hatano H, Yukl SA, Ferre AL, et al. Prospective antiretroviral treatment of asymptomatic, HIV-1 infected controllers. PLoS Pathog 2013; 9:e1003691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Boufassa F, Lechenadec J, Meyer L, et al. Blunted response to combination antiretroviral therapy in HIV elite controllers: an international HIV controller collaboration. PLoS One 2014; 9:e85516. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Okulicz JF, Grandits GA, Weintrob AC, et al. CD4 T cell count reconstitution in HIV controllers after highly active antiretroviral therapy. Clin Infect Dis 2010; 50:1187–91. [DOI] [PubMed] [Google Scholar]
- 35.Pereyra F, Palmer S, Miura T, et al. Persistent low-level viremia in HIV-1 elite controllers and relationship to immunologic parameters. J Infect Dis 2009; 200:984–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Lambotte O, Pollara J, Boufassa F, et al. High antibody-dependent cellular cytotoxicity responses are correlated with strong CD8 T cell viral suppressive activity but not with B57 status in HIV-1 elite controllers. PLoS One 2013; 8:e74855. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Madec Y, Boufassa F, Porter K, et al. Natural history of HIV-control since seroconversion. AIDS 2013; 27:2451–60. [DOI] [PubMed] [Google Scholar]
- 38.Migueles SA, Sabbaghian MS, Shupert WL, et al. HLA B*5701 is highly associated with restriction of virus replication in a subgroup of HIV-infected long term nonprogressors. Proc Natl Acad Sci U S A 2000; 97:2709–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Prins M, Meyer L, Hessol NA. Sex and the course of HIV infection in the pre- and highly active antiretroviral therapy eras. AIDS 2005; 19:357–70. [DOI] [PubMed] [Google Scholar]
- 40.Burkhalter JE, Springer CM, Chhabra R, Ostroff JS, Rapkin BD. Tobacco use and readiness to quit smoking in low-income HIV-infected persons. Nicotine Tob Res 2005; 7:511–22. [DOI] [PubMed] [Google Scholar]
- 41.Cioe PA, Crawford SL, Stein MD. Cardiovascular risk-factor knowledge and risk perception among HIV-infected adults. J Assoc Nurses AIDS Care 2014; 25:60–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Cockerham L, Scherzer R, Zolopa A, et al. Association of HIV infection, demographic and cardiovascular risk factors with all-cause mortality in the recent HAART era. J Acquir Immune Defic Syndr 2010; 53:102–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Crothers K, Goulet JL, Rodriguez-Barradas MC, et al. Impact of cigarette smoking on mortality in HIV-positive and HIV-negative veterans. AIDS Educ Prev 2009; 21:40–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Capili B, Anastasi JK, Ogedegbe O. HIV and general cardiovascular risk. J Assoc Nurses AIDS Care 2011; 22:362–75. [DOI] [PMC free article] [PubMed] [Google Scholar]