Current guidelines for initiation of antiretroviral therapy (ARV) suggest it be started when peripheral blood (PB) CD4 T cells decrease to below 350 cells per microliter1,2 when plasma HIV RNA is high or when major symptoms associated with an AIDS-defining illness are present. This strategy focuses on preventing disease progression.
There are no validated markers to differentiate those who will have a more robust CD4 response to ARV from those who will not. Pathogenesis-based studies have shown that viral replication in lymphatic tissues (LTs) causes inflammation and scarring that is associated with impaired reconstitution of CD4 T cells in LTs and PB.3-5 The subset most affected are naive CD4 T cells,6 and pretreatment levels of naive CD4 T cells have been shown to correlate to the rise in CD4 T cells with ARV.7-9 We, therefore, reasoned that it should be possible to use naive CD4 T cells in PB as a surrogate marker for damage to LTs and identify people most likely to benefit from earlier initiation of therapy. We studied patients whose PB CD4 T-cell count was in the range where most patients consider initiating ARV, that is, 200-500 cells per microliter and show that the size of the PB-naive CD4 population was the strongest predictor of treatment associated changes in PB total CD4 count, even more so than total CD4 count.
We collected data from patients in AIDS Clinical Trials Group treatment studies 384, 388, A5014, A5095, and A5001.10-15 We studied all patients who were ARV naive, had pretreatment measurements of baseline PB-naive (CD45RA+/62L+) CD4 T cells and serial measurements of CD4 T-cell counts and plasma HIV RNA (HIV-1 Monitor assay 1.0 or 1.5) through 24 months, and had baseline CD4 T-cell counts 200-500 cells per microliter. A follow-up CD4 measurement was selected for each 4-month interval; patients selected for analysis were further required to have at least 4 postbaseline measurements including 1 in the second year after initiation of ARV. All regimens met the standard for HAART.
Our goal was to model the ability of naive CD4 percent in PB to predict clinically significant changes in PB CD4 T cells defined as a confirmed 100 or 200 cell increase in the first 24 months of ARV. A Confirmed increase was defined as 2 consecutive measures showing a 100 or 200 cell increase over baseline. We used an analysis cohort (n = 348) for the predictive model and tested the model's validity in a separate confirmation cohort (n = 100), randomly selected from the largest contributing study (AIDS Clinical Trials Group 384). We used sex, age, race/ethnicity, baseline CD4 T-cell count, plasma HIV RNA, and protease inhibitor versus nonnucleoside reverse transcriptase inhibitor regimen as predictive variables in addition to naive CD4 percent in logistic regression models. We further examined the effect of baseline naive CD4 percent to predict postbaseline CD4 T-cell counts above a clinically relevant threshold, using a generalized estimating equations (GEE) model.
We identified 348 persons who had 24 months of follow-up and baseline measures of naive CD4 percent. Most were male (87%), and the median age was 36 years. The median CD4 T-cell count at ARV initiation was 321 cells per microliter (interquartile range: 264-398 cells/μL). Overall, 278 of 348 individuals (80%) had a confirmed 100-cell increase and 192 of 348 (55%) a confirmed 200-cell increase in 24 months after initiation of ARV. Persons with higher measures of naive CD4 T cells at baseline had proportionately greater increases in PB CD4 T cells (Fig. 1). We used regression analysis to determine if PB-naive CD4 T cells can predict ARV-associated changes in PB CD4 T cells and included in the model baseline CD4 T-cell count and plasma HIV RNA. Baseline naive CD4 percent was highly significant for predicting both 100 and 200 CD4 T-cell increases (P < 0.001). Baseline viral load (VL) was also highly significant for predicting 100-cell increases (P < 0.001) but did not reach significance for 200-cell increases (P = 0.07). Importantly, baseline PB CD4 did not predict 100-cell and 200-cell increases in PB CD4 T-cell count after 24 months of ARV in this population (P = 0.59 and 0.66, respectively, from multivariate models including baseline VL and naive CD4 percent; P = 0.39 and P = 0.28 in univariate models). Thus, baseline CD4 and VL were not as reliable as naive CD4 percent at predicting immune reconstitution. Similar results were seen in 306 subjects with good virologic suppression over months 4-24. In additional modeling, age did not significantly predict 200-cell increases but was significant for predicting 100-cell increases in both univariate (P = 0.004) in multivariable models (P = 0.026). Thus, in agreement with previous studies, age did correlate to a statistically significant increase in CD4 T-cell count but did not correlate to levels of reconstitution that would bring PB CD4 T-cell counts closer to a normal level. No other factors evaluated (sex, race/ethnicity, protease inhibitor versus nonnucleoside reverse transcriptase inhibitor regimen) were additionally significant (P > 0.05). The positive effect of higher baseline naive CD4 percent remained consistent when models were evaluated among 139 subjects with baseline CD4 350-500 and for 213 subjects with baseline CD4 300-500.
We next constructed models for 100 and 200 CD4 T-cell increases and confirmed the models using a subset of patients set aside for this purpose (Table 1). Predictions of the model were generally confirmed in a subset of patients reserved to validate the model. For patients with naive CD4 percent at the 25th percentile (which corresponded with a naive CD4 percent of 25%), the probabilities of 100-cell and 200-cell increases were 0.75 and 0.49, respectively. Thus, to achieve a 200-cell increase in PB CD4 T-cell count after 24 months of ARV, the probability is <50% if the naive CD4 percent was at (or below) 25%.
Recent studies have shown that time spent at lower CD4 T-cell counts is associated with higher rates of opportunistic infection and mortality. We compared the proportional amount of time spent with CD4 T cells >350 and 500 cells per microliter when stratified by naive CD4 T-cell measures at baseline (Fig. 2). Higher baseline naive CD4 percent significantly predicted the likelihood that a postbaseline CD4 T-cell count was >350 cells per microliter (P = 0.001), in a GEE model that also adjusted for baseline CD4 T-cell count (P < 0.001), log10 HIV RNA (P = 0.12), and age (P = 0.13). In a similar multivariate model, baseline naive CD4 percent was also highly predictive for T-cell counts >500 cells per microliter during follow-up (P < 0.001).
We demonstrate that measuring naive CD4 T cells in PB can identify individuals in the chronic stage of disease who are most likely to achieve significant CD4 T-cell reconstitution from ARV and experience a greater proportion of time with less risk of morbidity and mortality. Naive CD4 percent was significantly better than PB CD4 or HIV RNA for predicting magnitude of CD4 cell rise. We focused on individuals with PB CD4 T cells between 200 and 500 cells per microliter, the range where most would consider initiating ARV and noted, moreover, that findings remained consistent when analyses were limited to subsets with the highest baseline CD4 cell counts (350-500, 300-500). The Data Collection on Adverse Events (D:A:D) and Flexible Initial Retrovirus Suppressive Therapies Trial (FIRST) studies showed the risk of non-AIDS-related morbidity and mortality among patients treated with ARV were more likely to occur at lower CD4 T-cell counts.16,17 We showed that patients most likely to have diminished CD4 immune reconstitution with ARV are those with smaller populations of naive CD4 T cells when ARV is begun.
These data have important clinical implications for timing of ARV as individuals with lower naive CD4 cells might better preserve the existing CD4 population and those with higher levels of PB-naive CD4 T cells might shift the risk/benefit ratio toward a decision to delay treatment.
The idea of using a measure of PB-naive CD4 T cells as a threshold marker to identify those least and most likely to reconstitute PB CD4 T cells was suggested by recent pathogenesis-based studies that demonstrate naive CD4 T cells are preferentially depleted in LTs6,18 and by clinical trials illustrating the predictive ability of PB-naive cells for CD4 cell reconstitution after 6 months of ARV.9 The present study considerably expands the latter work and extends observations to a 2-year treatment period. Early after initiation of ARV, effector memory cells are the first to repopulate the PB pool, likely reflecting, in part, redistribution of cells from tissues sites, whereas late reconstitution is mainly from naive CD4 T cells.19 This demonstrates that overall CD4 T-cell reconstitution is dependent on the integrity of the naive CD4 T-cell population in LTs and may be the underlying mechanism for decreased mortality seen in those who initiate ARV with higher CD4 counts compared with those deferring ARV until CD4 counts are lower.20,21 This is important because naive CD4 cells are the critical reservoir supplying the central memory and ultimately effector memory population of CD4 T cells.
In conclusion, we show that patients with relatively high measures of naive CD4 T cells are less likely to have failure of immune reconstitution with ARV and will spend relatively less time at risk for opportunistic infection or mortality from any cause. In particular, we demonstrate that this measure could be used to delineate patients at higher PB CD4 T-cell counts who are less likely to reconstitute immunity, which could be problematic if ARV treatment is delayed until the patient has lower CD4 T-cell counts. Conversely, this measure might also identify those who could expect higher levels of immune reconstitution even when treatment is delayed. Large prospective studies should be undertaken to more precisely define the role of measurement of naive CD4 T cells in the management of HIV infection.
The authors wish to acknowledge Dr. Susan Plaeger for her encouragement to develop this project and Rui Wang and James Neaton for input on study design.
1. Hammer SM, Eron Jr, JJ, Reiss P, et al. Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel. JAMA
. 2008. 300:555-570.
2. Guidelines for the Use of Antiretroviral Therapy in HIV-1 Infected Adults and Adolescents. Washington, DC: Panel on Clinical Practices for Treatment of HIV Infection Convened by the Department of Health and Human Services; 2008.
3. Schacker TW, Nguyen PL, Beilman GJ, et al. Collagen deposition in HIV-1 infected lymphatic tissues and T cell homeostasis. J Clin Invest
4. Estes J, Baker JV, Brenchley JM, et al. Collagen deposition limits immune reconstitution in the gut. J Infect Dis
5. Schacker TW, Reilly C, Beilman GJ, et al. Amount of lymphatic tissue fibrosis in HIV infection predicts magnitude of HAART-associated change in peripheral CD4 cell count. AIDS
6. Schacker TW, Brenchley JM, Beilman, GJ, et al. Lymphatic tissue fibrosis is associated with reduced numbers of naive CD4+ T cells in human immunodeficiency virus type 1 infection. Clin Vaccine Immunol
7. Notermans DW, Pakker NG, Hamann D, et al. Immune reconstitution after 2 years of successful potent antiretroviral therapy in previously untreated human immunodeficiency virus type 1-infected adults. J Infect Dis
8. Hengel RL, Jones BM, Kennedy MS, et al. Lymphocyte kinetics and precursor frequency-dependent recovery of CD4(+)CD45RA(+)CD62L(+) naive T cells following triple-drug therapy for HIV type 1 infection. AIDS Res Hum Retroviruses
9. Mildvan D, Bosch RJ, Kim RS, et al. Immunophenotypic markers and antiretroviral therapy (IMART): T cell activation and maturation help predict treatment response. J Infect Dis
10. Bosch RJ, Bennett K, Collier AC, et al. Pretreatment factors associated with 3-year (144-week) virologic and immunologic responses to potent antiretroviral therapy. J Acquir Immune Defic Syndr
11. Fischl MA, Ribaudo HJ, Collier AC, et al. A randomized trial of 2 different 4-drug antiretroviral regimens versus a 3-drug regimen, in advanced human immunodeficiency virus disease. J Infect Dis
12. Gulick RM, Ribaudo HJ, Shikuma CM, et al. Three- vs four-drug antiretroviral regimens for the initial treatment of HIV-1 infection: a randomized controlled trial. JAMA
13. Landay AL, Spritzler J, Kessler H, et al. Immune reconstitution is comparable in antiretroviral-naive subjects after 1 year of successful therapy with a nucleoside reverse-transcriptase inhibitor- or protease inhibitor-containing antiretroviral regimen. J Infect Dis
14. Shafer RW, Smeaton LM, Robbins GK, et al. Comparison of four-drug regimens and pairs of sequential three-drug regimens as initial therapy for HIV-1 infection. N Engl J Med
15. Robbins GK, De Gruttola V, Shafer RW, et al. Comparison of sequential three-drug regimens as initial therapy for HIV-1 infection. N Engl J Med
16. Baker JV, Peng G, Rapkin J, et al. CD4+ count and risk of non-AIDS diseases following initial treatment for HIV infection. AIDS
17. Weber R, Sabin CA, Friis-Moller N, et al. Liver-related deaths in persons infected with the human immunodeficiency virus: the D:A:D study. Arch Intern Med
18. Estes JD, Haase AT, Schacker, TW. The role of collagen deposition in depleting CD4+ T cells and limiting reconstitution in HIV-1 and SIV infections through damage to the secondary lymphoid organ niche. Semin Immunol
19. Autran B, Carcelain G, Li TS, et al. Positive effects of combined antiretroviral therapy on CD4+ T cell homeostasis and function in advanced HIV disease. Science
20. Sterne JA, May M, Costagliola D, et al. Timing of initiation of antiretroviral therapy in AIDS-free HIV-1-infected patients: a collaborative analysis of 18 HIV cohort studies. Lancet
21. Kitahata MM, Gange SJ, Abraham AG, et al. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med