Cox proportional hazards models for pneumonia and all-cause mortality among HIV+PNA+ participants are shown in Table 3. The overall associations of the 5 CD8+ TCL categories with bacterial pneumonia and all-cause mortality were statistically significant (P = 0.017 and P < 0.0001, respectively), even after adjustment for CD4+ TCL, VL, age, and ARV duration. We defined our referent category as CD8+ TCL of 401–800 cells per cubic millimeter as it was the most common CD8+ TCL range among all groups. Relative to this referent, risk for pneumonia was highest when CD8+ TCL was ≤400 cells per cubic millimeter [hazard ratio (HR) 1.65, P = 0.017, 95% confidence interval 1.10 to 2.49]. The highest all-cause mortality was also observed when CD8+ TCLs were ≤400 compared with the referent range of 401–800 cells per cubic millimeter (HR 1.45, P = 0.04, 95% confidence interval 1.02 to 2.06). Both of these results reflect adjustment for CD4+ TCL, log10VL, age, and ARV duration, and remained statistically significant when further adjusted for smoking (HR 1.61, P = 0.024; HR 1.49, P = 0.035, for pneumonia and mortality, respectively). Upon further adjustment for African American ethnicity, pneumonia and mortality risk also remained significant and borderline significant, respectively (HR 1.58, P = 0.03; HR 1.44, P = 0.054, respectively). Hence, our data indicate that compared with a referent of 401–800 cells per cubic millimeter, a CD8+ TCL <400 was associated with a higher risk for pneumonia and all-cause mortality, adjusting for CD4+ TCL, VL, age, ARV duration, smoking, and African American ethnicity. The results were similar when the model was repeated using the ≤400 cells per cubic millimeter category as the referent. CD8+ TCL categories of 801–1200, 1201–1600, and >1600 cells per cubic millimeter were not significantly different from the referent in either the pneumonia or the mortality model. Although the Cox proportional hazards model included CD4+ TCL as a time-varying variable and thereby adjusted for the nadir CD4+ TCL, we computed nadir CD4+ TCL for each subject and replaced the time-varying CD4+ TCL by the time-independent nadir CD4+ TCL in the Cox model. Nadir CD4+ TCL was significantly associated with pneumonia but not significantly associated with all-cause mortality. The overall results concerning CD8+ TCL categories remained unchanged in the nadir CD4+ TCL model.
Covariate-adjusted survival estimates based on the fitted Cox models for time to pneumonia and death are shown in Figure 3 for participants aged 50 years, CD4+ TCL of 200 cells per cubic millimeter, log10VL of 4, and ARV duration of 500 days. These parameters were chosen to reflect a high-risk group for both pneumonia and mortality. Among HIV+PNA+ participants, 65% of those with CD8+ TCLs 401–800 cells per cubic millimeter were estimated to be pneumonia-free 5 years from study entry, compared with 50% of those with CD8+ TCLs ≤400 cells per cubic millimeter. Analysis of time to death revealed that 16% more participants were estimated to be alive at 5 years in the 401–800 than in the ≤400 cells per cubic millimeter CD8+ TCL category. Similar, though less dramatic, differences were observed for nonsmokers. The covariate survival estimate model was computed several times, with changes in age, CD4+ TCL, log10VL, smoking status, and race. Neither time to pneumonia nor mortality differences were observed between the cohorts when age was decreased to 35 years, CD4+ TCL was increased to 400 cells per cubic millimeter, log10VL was decreased to 1.7, or when including only non–African American participants.
A sensitivity analysis conducted by Kohli et al16 established that their results were similar when either all definitive, probable, and presumed cases or only definitive and probable cases were included. However, when we repeated our analyses separately for definitive versus probable and possible pneumonia, none of the CD8+ TCL categories were significant compared with the referent CD8+ TCL category 400–800 cells per cubic millimeter, whereas CD4+ TCL counts were significant in both analyses.
A total of 69 participants had more than 1 pneumonia event. CD8+ TCLs of participants in this subgroup were similar to those in Figure 2 (data not shown). Cox proportional hazards model for those with recurrent pneumonia revealed no significant difference by CD8+ TCL category, perhaps owing to a smaller sample size. However, 49% of those with recurrent pneumonia had CD8+ TCLs >800 cells per cubic millimeter within 6 months of their first pneumonia event. Mean CD8+ TCL was 1420 cells per cubic millimeter for 8 participants with >5 pneumonia events and <200 cells per cubic millimeter for 2 such participants.
CD8+ TCLs increase early in HIV infection and remain persistently elevated during the early chronic phase of HIV disease.20,21 The mean baseline CD8+ TCL of HIV-infected participants in our study was similar to that reported by other groups.22,23 Baseline CD8+ TCLs were not significantly different between HIV+PNA+ and HIV+PNA− participants in our study, but there was a statistically significant association between CD8+ TCL antedating the pneumonia event and bacterial pneumonia whereby a CD8+ TCL ≤400 cells per cubic millimeter was associated with a 1.7 times higher risk of pneumonia and 1.5 times higher risk for death compared with a referent CD8+ TCL of 401–800 cells per cubic millimeter. These findings remained statistically significant after adjusting for CD4+ TCL, VL, age, ARV use, smoking, and ethnicity.
The possibility that low total CD8+ TCL could increase the risk for an HIV-associated complication like pneumonia is not surprising given that HIV controllers have an expanded and more functional CD8+ T-cell compartment and a reduction in CD8+ TCL has been linked to HIV disease progression.4,5,23,24 The possible impact of CD8+ TCL loss and the risk for pneumonia is further underscored by preclinical models demonstrating the importance of CD8+ T cells in host defense against pulmonary pathogens. For example, in mice, CD8+ T cells can compensate for CD4+ T cells in resistance to pulmonary mycobacterial and fungal pathogens, including HIV-associated pathogens such as Pneumocystis jiroveci and Cryptococcus neoformans.9,25–27 Of relevance to the risk for bacterial pneumonia, CD8+ T cells were required for protection against S. pneumoniae in immunized and naive mice 6,7 and recruited to the lungs in surviving mice.28 Similar results have been reported for Klebsiella pneumoniae.29 Our findings suggest that the possibility that CD8+ T cells contribute to resistance to HIV-associated bacterial pneumonia deserves further study.
Our data reveal a peak in the percentage of pneumonias (19%) in the CD8+ TCL category of 1201–1600 cells per cubic millimeter, although risk for pneumonia and all-cause mortality in CD8+ TCL categories >800 cells per cubic millimeter was not statistically different from the referent group (401–800 cells/mm3) in the Cox proportional hazards analysis. Nonetheless, among 69 participants with recurrent pneumonias, almost half (34 participants or 49%) had CD8+ TCL >800 cells per cubic millimeter, with the mean level among those with >5 episodes being 1420 cells per cubic millimeter. This is intriguing in light of the damage–response framework,30 which highlights that host damage can occur in the setting of either an insufficient or a vigorous immune response. Given ample evidence for CD8+ T-cell–mediated inflammation, we wonder whether higher CD8+ TCLs could be associated with increased disease risk by inducing excessive inflammation. Along these lines, high CD8+ TCLs were inversely correlated with forced expiration volume in 1 second and implicated as mediators of smoking-associated lung injury and COPD,31,32 and among HIV-infected smokers, CD8+ TCLs were significantly higher in bronchoalveolar lavage from those with than without COPD.33 In mice with PJP, CD4 depletion led to death due to respiratory compromise with CD8+ T-cell–mediated lung injury,34 and HIV-infected patients with Pneumocystis after starting ARVs had high quantities of rapidly proliferating CD8+ T cells in their bronchoalveolar lavage fluid.35 Several studies have implicated CD8+ T cells in lung inflammation in Pneumocystis colonization and progressive pulmonary decline in patients with HIV.34,36 Hence, our data raise the possibility that, as per the damage–response framework,30 either low or high CD8+ TCLs could contribute to susceptibility to HIV-associated bacterial pneumonia.
HIV+PNA+ participants had significantly more variability in their CD8+ TCLs than HIV+PNA− participants. At present, the significance of this finding is unclear. Further studies are required to evaluate the activation status and antigen specificity of CD8+ T cells in patients with pneumonia as these parameters were not evaluated in the HERS cohort. Nonetheless, our data show that CD8+ TCLs are not uniform and that greater fluctuation could be associated with risk for HIV-associated pneumonia. The rise in mean CD8+ TCL that was observed at later times in both the HIV+PNA+ and HIV+PNA− cohorts could reflect an effect of ARV introduction, as the percentage of those having started ARV increased from approximately 36% to nearly 100% with time.
At present, our data are most relevant for HIV-infected individuals with CD4+ TCL in lower ranges who are not on ARVs, such as those in resource-limited settings. As measurement of CD8+ TCL is performed on the same sample as CD4+ TCL and both measurements are commonly obtained even in resource-limited settings,37,38 our results suggest that CD8+ TCL could serve as an adjunctive biomarker for HIV-associated pneumonia risk, in addition to other known risk factors. Our findings may also be important in non–HIV-associated and/or nonbacterial pneumonia, as CD8+ T cells can contribute to the pathogenesis of influenza.39 In light of growing recognition that influenza morbidity and mortality are driven by secondary bacterial pneumonia,40,41 it is logical to hypothesize that CD8+ T-cell involvement in host defense against influenza could alter the inflammatory milieu and predispose the host to bacterial pneumonia. This area requires further investigation.
Our study has several limitations. First, most of the pneumonias did not meet definitive criteria for bacterial pneumonia, with the caveat that the definitive category required culture-proven bacterial pneumonia, which is very difficult to obtain. Although a sensitivity analysis conducted by Kohli et al16 found similar results when all definitive, probable, and presumed cases were included, and when limited to definitive and probable cases, this was not the case in our study. The most likely explanation for this is that the effect of CD4+ TCL on pneumonia and all-cause mortality is substantially greater than that of CD8+ TCL; however, our study was not designed to address this question. Based on our findings, future studies should endeavor to examine CD8+ TCLs among patients with predefined CD4+ TCLs and pneumonia definitions. Second, HERS included only women; hence, studies are needed in men. Third, we were not able to account for certain comorbid illnesses that could have confounded our findings, such as viral infections and history of lung disease (eg, COPD). Fourth, although our data suggest that CD8+ TCLs ≤400 cells per cubic millimeter may be a useful adjunct to CD4+ TCLs for identifying patients at risk for bacterial pneumonias and death, the HERS spanned a period during which potent ARVs first became available. Thus, most participants in the pneumonia cohort were not on potent ARV therapy. Although ARV therapy is now widely available in the United States and other resourced countries, there are many HIV-infected individuals living in areas where ARVs are not readily or consistently available. Our findings are most relevant to the latter and for patients who do not adhere to their ARV regimens. Finally, the scope of our study and conclusions are limited by our inability to assess CD8+ T-cell activation status, effector, or memory CD8+ T cells, as such data were not collected in the HERS.
In summary, we evaluated CD8+ TCL and pneumonia risk in a large prospective cohort of HIV-infected women and identified an association between CD8+ TCLs ≤400 cells per cubic millimeter and risk for bacterial pneumonia and all-cause mortality relative to a referent range of 401–800 cells per cubic millimeter. These findings suggest that CD8+ TCLs ≤400 cells per cubic millimeter may be a useful adjunct to CD4+ TCLs for identifying patients at risk for pneumonias and death, particularly in individuals with low CD4+ TCLs and/or who are not on ARVs. Furthermore, our finding that CD8+ TCLs were more variable in HIV-infected participants who developed pneumonia and tended to be higher in those who had highly recurrent pneumonia suggests that the role of CD8+ T cells in the pathogenesis of pneumonia is complex and requires further study.
The authors acknowledge the HERS committee members and the Centers for Disease Control and Prevention and, particularly, Lytt Gardner, who provided the HERS data sets and statistical and editorial guidance. They also acknowledge the Center for AIDS Research at Albert Einstein College of Medicine for providing statistical support (AI051519).
1. Feikin DR, Feldman C, Schuchat A, et al.. Global strategies to prevent bacterial pneumonia in adults with HIV disease. Lancet Infect Dis. 2004;4:445–455.
2. Redd SC, Rutherford GW 3rd, Sande MA, et al.. The role of human immunodeficiency virus infection in pneumococcal bacteremia in San Francisco residents. J Infect Dis. 1990;162:1012–1017.
3. Hirschtick RE, Glassroth J, Jordan MC, et al.. Bacterial pneumonia in persons infected with the human immunodeficiency virus. Pulmonary Complications of HIV Infection Study Group. N Engl J Med. 1995;333:845–851.
4. Betts MR, Nason MC, West SM, et al.. HIV nonprogressors preferentially maintain highly functional HIV-specific CD8+ T cells. Blood. 2006;107:4781–4789.
5. Saez-Cirion A, Lacabaratz C, Lambotte O, et al.. HIV controllers exhibit potent CD8 T cell capacity to suppress HIV infection ex vivo and peculiar cytotoxic T lymphocyte activation phenotype. Proc Natl Acad Sci U S A. 2007;104:6776–6781.
6. Tian H, Groner A, Boes M, et al.. Pneumococcal capsular polysaccharide vaccine-mediated protection against serotype 3 Streptococcus pneumoniae in immunodeficient mice. Infect Immun. 2007;75:1643–1650.
7. Weber SE, Tian H, Pirofski LA. CD8+ cells enhance resistance to pulmonary serotype 3 streptococcus pneumoniae infection in mice. J Immunol. 2011;186:432–442.
8. Jones HP, Tabor L, Sun X, et al.. Depletion of CD8+ T cells exacerbates CD4+ Th cell-associated inflammatory lesions during murine mycoplasma respiratory disease. J Immunol. 2002;168:3493–3501.
9. McAllister F, Steele C, Zheng M, et al.. T cytotoxic-1 CD8+ T cells are effector cells against pneumocystis in mice. J Immunol. 2004;172:1132–1138.
10. Swain SD, Meissner NN, Harmsen AG. CD8 T cells modulate CD4 T-cell and eosinophil-mediated pulmonary pathology in pneumocystis pneumonia in B-cell-deficient mice. Am J Pathol. 2006;168:466–475.
11. Maeno T, Houghton AM, Quintero PA, et al.. CD8+ T Cells are required for inflammation and destruction in cigarette smoke-induced emphysema in mice. J Immunol. 2007;178:8090–8096.
12. van Stipdonk MJ, Hardenberg G, Bijker MS, et al.. Dynamic programming of CD8+ T lymphocyte responses. Nat Immunol. 2003;4:361–365.
13. Tzortzaki EG, Siafakas NM. A hypothesis for the initiation of COPD. Eur Respir J. 2009;34:310–315.
14. Floris-Moore M, Lo Y, Klein RS, et al.. Gender and hospitalization patterns among HIV-infected drug users before and after the availability of highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2003;34:331–337.
15. Feldman C, Glatthaar M, Morar R, et al.. Bacteremic pneumococcal pneumonia in HIV-seropositive and HIV-seronegative adults. Chest. 1999;116:107–114.
16. Kohli R, Lo Y, Homel P, et al.. Bacterial pneumonia, HIV therapy, and disease progression among HIV-infected women in the HIV epidemiologic research (HER) study. Clin Infect Dis. 2006;43:90–98.
17. Melnick SL, Sherer R, Louis TA, et al.. Survival and disease progression according to gender of patients with HIV infection. The Terry Beirn Community Programs for Clinical Research on AIDS. JAMA. 1994;272:1915–1921.
18. Smith DK, Warren DL, Vlahov D, et al.. Design and baseline participant characteristics of the Human Immunodeficiency Virus Epidemiology Research (HER) Study: a prospective cohort study of human immunodeficiency virus infection in US women. Am J Epidemiol. 1997;146:459–469.
19. Rompalo AM, Astemborski J, Schoenbaum E, et al.. Comparison of clinical manifestations of HIV infection among women by risk group, CD4+ cell count, and HIV-1 plasma viral load. HER Study Group. HIV Epidemiology Research. J Acquir Immune Defic Syndr Hum Retrovirol. 1999;20:448–454.
20. Giorgi JV, Detels R. T-cell subset alterations in HIV-infected homosexual men: NIAID Multicenter AIDS cohort study. Clin Immunol Immunopathol. 1989;52:10–18.
21. Hazenberg MD, Otto SA, van Benthem BH, et al.. Persistent immune activation in HIV-1 infection is associated with progression to AIDS. AIDS. 2003;17:1881–1888.
22. Singh YG, Dar L, Singh NG. Levels of CD4 and CD8 among the inhabitants of Manipur, India. J Commun Dis. 2000;32:201–206.
23. Roederer M, Dubs JG, Anderson MT, et al.. CD8 naive T cell counts decrease progressively in HIV-infected adults. J Clin Invest. 1995;95:2061–2066.
24. Altfeld M, Kalife ET, Qi Y, et al.. HLA alleles associated with delayed progression to AIDS contribute strongly to the initial CD8(+) T cell response against HIV-1. PLoS Med. 2006;3:e403.
25. Xing Z, Wang J, Croitoru K, et al.. Protection by CD4 or CD8 T cells against pulmonary Mycobacterium bovis bacillus Calmette-Guerin infection. Infect Immun. 1998;66:5537–5542.
26. Huffnagle GB, Lipscomb MF, Lovchik JA, et al.. The role of CD4+ and CD8+ T cells in the protective inflammatory response to a pulmonary cryptococcal infection. J Leukoc Biol. 1994;55:35–42.
27. Kolls JK, Habetz S, Shean MK, et al.. IFN-gamma and CD8+ T cells restore host defenses against Pneumocystis carinii in mice depleted of CD4+ T cells. J Immunol. 1999;162:2890–2894.
28. Coleman JR, Papamichail D, Yano M, et al.. Designed reduction of Streptococcus pneumoniae pathogenicity via synthetic changes in codon-pair bias. J Infect Dis. 2011;203:1264–1273.
29. Moore TA, Perry ML, Getsoian AG, et al.. Divergent role of gamma interferon in a murine model of pulmonary versus systemic Klebsiella pneumoniae infection. Infect Immun. 2002;70:6310–6318.
30. Casadevall A, Pirofski LA. The damage-response framework of microbial pathogenesis. Nat Rev Microbiol. 2003;1:17–24.
31. Saetta M, Di Stefano A, Turato G, et al.. CD8+ T-lymphocytes in peripheral airways of smokers with chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 1998;157(3 pt 1):822–826.
32. O'Shaughnessy TC, Ansari TW, Barnes NC, et al.. Inflammation in bronchial biopsies of subjects with chronic bronchitis: inverse relationship of CD8+ T lymphocytes with FEV1. Am J Respir Crit Care Med. 1997;155:852–857.
33. Diaz PT, King MA, Pacht ER, et al.. Increased susceptibility to pulmonary emphysema among HIV-seropositive smokers. Ann Intern Med. 2000;132:369–372.
34. Wright TW, Gigliotti F, Finkelstein JN, et al.. Immune-mediated inflammation directly impairs pulmonary function, contributing to the pathogenesis of Pneumocystis carinii pneumonia. J Clin Invest. 1999;104:1307–1317.
35. Barry SM, Lipman MC, Deery AR, et al.. Immune reconstitution pneumonitis following Pneumocystis carinii pneumonia in HIV-infected subjects. HIV Med. 2002;3:207–211.
36. Norris KA, Morris A, Patil S, et al.. Pneumocystis colonization, airway inflammation, and pulmonary function decline in acquired immunodeficiency syndrome. Immunol Res. 2006;36:175–187.
37. Bekele Y, Mengistu Y, de Wit TR, et al.. Timing of blood sampling for CD4 T-cell counting influences HAART decisions. Ethiop Med J. 2011;49:187–197.
38. Hunt PW, Cao HL, Muzoora C, et al.. Impact of CD8+ T-cell activation on CD4+ T-cell recovery and mortality in HIV-infected Ugandans initiating antiretroviral therapy. AIDS. 2011;25:2123–2131.
39. Wissinger E. Altered immune responses following influenza infection and their impact on susceptibility to bacterial pathogens. Presented at: Infectious Disease Society of America; October 22, 2010.
40. Moore M. Post-influenza secondary bacterial infections. Paper presented at: Infectious Disease Society of America; October 22, 2010; Vancouver, British Columbia, Canada.
41. Palacios G, Hornig M, Cisterna D, et al.. Streptococcus pneumoniae coinfection is correlated with the severity of H1N1 pandemic influenza. PLoS One. 2009;4:e8540.
Keywords:© 2012 Lippincott Williams & Wilkins, Inc.
pneumonia; CD8+ T cell; HIV; mortality