Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Beyond the Brain: The Role of Brain-Derived Neurotrophic Factor in Viroimmune Responses to Antiretroviral Therapy among People Living with HIV with and without Alcohol Use

Beyond the Brain: The Role of Brain-Derived Neurotrophic Factor in Viroimmune Responses to... Objective: Given the emerging data suggesting the key role of brain-derived neurotrophic factor (BDNF) in the immune system, we assessed longitudinally whether BDNF depletions induced by hazardous alcohol use (HAU) would impact a response to antiretroviral therapy (ART). Methods: In a prospective single-site cohort, virological and immunological responses to ART in 200 hazardous and 200 nonhazardous users were obtained, along with plasma BDNF levels. Results: Hazardous drinkers were more likely to have BDNF levels <4000 pg/mL (odds ratio [OR] ¼ 1.6, P ¼ .01). Participants with BDNF <4000 pg/mL were less likely to have CD4 counts of more than 500 cells/mm (P ¼ .02) and to achieve viral suppression over the follow- up period (OR ¼ 1.5, P ¼ .03). Multivariate analysis confirmed the significant role of HAU and low BDNF in predicting viroimmune responses. Conclusion: Hazardous alcohol use was associated with BDNF alterations, which in turn were linked to a limited response to ART in terms of viral suppression and CD4 count improvements. Keywords HIV/AIDS, alcohol, BDNF, CD4, viral load As indicated by many studies, brain-derived neurotrophic Introduction factor (BDNF) is a neutrophil mostly known for its role in the Antiretroviral therapy (ART) has been one of the greatest 9-11 central nervous system (CNS) that is affected by HIV. advances in the fight against the HIV. When a complete viro- Brain-derived neurotrophic factor plays an important role in logic suppression (viral load [VL] <400 copies/mL by 24 a variety of brain functions, including the lasting potentiation weeks or <50 copies/mL by 48 weeks) is attained, people living of synaptic efficacy that modulates brain plasticity. Brain- with HIV (PLWH) benefit from immune reconstitution and derived neurotrophic factor is involved in learning and memory 1,2 AIDS-free survival. Unfortunately, incomplete suppression processes and has been closely associated with cognitive deficit of viral replication, which is present in a sizable proportion 9-11 in HIV, aging, and neurodegenerative diseases. Until of cases, results in immune exhaustion, the emergence of drug recently, BDNF has been linked with the immune system, but resistance, and ultimately in increased morbidity and mortal- ity. Even among patients reaching these virologic targets, sub- stantial interindividual differences in immune responses occur 1 School of Integrated Science and Humanity, Florida International University, among them, suggesting that other factors may be involved. Miami, FL, USA Department of Medicine, University of Miami School of Medicine, Miami, FL, Our group, along with others, has demonstrated that hazardous USA alcohol use (HAU) influences CD4 count recovery and clinical Departments of Epidemiology and Medicine, University of Florida, Gaines- outcomes. Our previous studies indicated that alcohol ville, FL, USA decreases the size and cellularity of the thymus, leading to low Department of Psychiatry and Behavioral Sciences, University of Miami Miller 4,5 total numbers of peripheral T cells. Others have demon- School of Medicine, Miami, FL, USA strated only small gradual increases in CD4 counts that largely 3 6-8 Corresponding Author: remained less than 500 cells/mm . However, the thymus Marı´aJose´ Miguez-Burbano, School of Integrated Science and Humanity, model only partially explains the impaired recovery of hazar- Florida International University, 13313 Kingsbury Drive Wellington, FL 33414, dous alcohol users living with HIV (HAULWH), indicating the USA. need to look for additional factors. Email: [email protected] Mı´guez-Burbano et al 455 Figure 1. Levels of brain-derived neurotrophic factor (BDNF) by amounts of alcohol consumption. several lines of evidence indicate that BDNF plays important Participants were dichotomized as hazardous and non-HAUs, 12-14 roles in proliferation, apoptosis, and T-cell survival. Yet, its based on self-reports of alcohol intakes. Alcohol consumption role in CD4 count depletion or immune replenishment under scores were computed by averaging cross products of quantity ART is unknown. Of interest is the study by Avdoshina and col- and frequency of beer/wine and hard liquor reported on the leagues who did not relate the findings to genetic mutations, Alcohol Use Disorders Identification Test, developed by the nor to platelet counts. However, their sample was small, and World Health Organization (WHO) as a simple method of information on platelet counts was not provided. Although evi- screeningfor excessive drinking andtoassistinbrief assess- 16 17,18 dence that gender can impact BDNF levels, the authors failed ment, and the Alcohol Dependence Scale responses. to analyze gender differences in BDNF levels. Then, based on the National Institute of Alcohol Abuse and In light of these findings, we examined whether a decrease in Alcoholism guidelines criteria, men who reported >14 drinks/ BDNF could explain the reduced T cell counts and the excessive week or >4 drinks in 1 day and women >7 drinks/week or >3 VL observed in PLWH (with comorbidities such as HAU and/or drinks in 1 day were classified as HAU, while those who thrombocytopenia [TCP]). Because the Platelets Mediating reported fewer drinks were categorized as non-HAU. Alcohol and HIV Damage Study (PADS) includes a gender- The group was chosen to represent relatively ‘‘pure’’ alcohol diverse sample of participants, with substantial variation in alco- users with minimal drug use and without major confounding hol use, we were afforded a valuable opportunity to explore the factors. Nonambulatory patients and those presenting major multifaceted associations among alcohol use, BDNF, gender, medical comorbidities such as major neuropsychological (ie, and immune responses to ART. active CNS opportunistic infection, tumors, and developmental disorders), immune-based (ie, malignancies and autoimmune diseases), and chronic diseases (renal failure) were excluded. In addition, individuals who had cirrhosis, active viral hepati- Methods tis, or liver enzymes and who were 2 standard deviations (SDs) Sampling above the normal values were not eligible to participate in the study. To reduce the confounding effects of illicit drug use, the The Platelets Mediating Alcohol and HIV Damage Study PADS Diagnostic and Statistical Manual of Mental Disorders (Fourth is a large, single-site multiethnic cohort, consisting of 400 Edition, Text Revision) questionnaire was applied, and those PLWH who are at least 18 years old and younger regular care participants who were dependent on drugs or injecting illicit at Miami’s primary open-access public health system. Patients psychoactive substances were also excluded. Otherwise, the were recruited via flyers, personal contact at the clinics, or calls patients were enrolled. to our office to schedule appointments. Recruitment and follow- The Platelets Mediating Alcohol and HIV Damage Study ups took place in the period between June 2010 and June 2012. was approved by the Central Governing Institutional Review Our choice of PLWH in an open-access public health system Boards at Florida International University and University of with standard treatment protocols was purposefully designed Miami. The study was conducted according to the principles to minimize social, medical, and treatment inequalities. 456 Journal of the International Association of Providers of AIDS Care 13(5) expressed in the Declaration of Helsinki. Those participants for Disease Control and Prevention [CDC] clinical staging): who provided written informed consent were consecutively complete blood counts (thrombocytopenia was defined as enrolled and followed over a period of 6 months. platelet counts of less than 150  103 cells/mm [41-42] and a biochemical profile (calcium, sodium, potassium, albumin, glucose, lipids, kidney, and liver function). HIV- Laboratory Outcomes related and not-related treatments (ie, start date and discon- Blood was drawn from fasting patients to best evaluate the tinued) were obtained and confirmed with the pharmacy and immunological, hematological, and platelet-associated factor medical records. An AIDS Clinical Trial Group (ACTG) profiles. Blood samples were collected and processed within self-reported adherence questionnaire was used at each visit. 6 hours. Isolated peripheral blood mononuclear cells were pre- Basedonthe missed dosesper week andduringthe week- pared for 4-color direct immunofluorescence procedures (Becton end, the percentage of adherence was calculated at baseline Dickinson, San Jose, California). Flow cytometry quantified the and at the follow-up visit. percentage and absolute numbers of T-lymphocyte subpopula- tions CD3/CD4 and CD3/CD8. A good immunological Statistical Analyses response was defined as having a CD4 count of more than 3 3 The data were analyzed using SAS version 8 and SPSS version 500 cells/mm or as a gain in CD4 count 50 cells/mm from 19, and P values <.05 were considered to be statistically signif- week 0 to week 24. icant. The normality of the distribution of primary outcomes of HIV viral burden was quantified using the Amplicor HIV interest was examined with a normal probability plot. Viral monitor test (Roche Diagnostic System, Indianapolis, IN). The load was log transformed to correct skewness. Following lower threshold for detection at the time of the study was 50 descriptive statistical analyses, the mean variables were com- copies/mL. Virological success was defined as achieving unde- pared using Student t test and one-way analysis of variance pro- tectable VL. Poor virological response was defined as a plasma cedures. Correlations among the main variables of interest VL >2.7 log copies/mL at week 24. were examined with Pearson coefficients. Additionally, chi- square analyses were performed to compare the proportions for Brain-Derived Neurotrophic Factor gender and race. Univariate analyses were used to calculate the odds ratios (ORs) and 95% confidence intervals (CIs). Logistic The circulating levels of BDNF were selected because prior regression analyses were used to evaluate the effects of alcohol studies have demonstrated that, although different from those (continuous or as hazardous versus nonhazardous), thrombocy- in the cerebrospinal fluid (CSF), they are correlated with CSF topenia (counts and dichotomized as thrombocytopenia <150 measures in other CNS diseases. To obtain platelet-poor 000 versus normal platelet counts), age (continuous and dichot- plasma (PPP), the blood samples were collected in EDTA- omized at age 50 years), and gender. coated tubes (plasma; BD Diagnostic Systems, New Jersey) The multivariable models aimed to predict VL and CD4 and stored on ice. Plasma was separated by centrifugation at count status at the last visit and included all covariates whose 40 C for 15 minutes at 1500g. This plasma was again recentri- likelihood ratio for P value from the univariate analyses was fuged at 10 000g and the aliquots of PPP were stored in poly- <.1. Although no significant differences were observed propylene tubes at 80 C until assayed. The BDNF levels in between groups for education level or CDC status, both vari- PPP were measured using a commercially available enzyme- ables were controlled for in the final analyses. In addition, other linked immunosorbent assay (ELISA) kit (R&D System), potential predictors (ie, gender, race/ethnicity, and body mass according to the manufacturer’s instructions. However, during index) were selected on the basis of the literature and were the standardization a sizable proportion of PLWH had a BDNF added to the model. Although the nadir CD4 count has been value of 4000 pg/mL (ceiling effect), so the samples were 20- recommended as a preferable measure of disease severity, data fold diluted. The concentration of BDNF in plasma was calcu- were not available for most participants and therefore the base- lated based on a standard curve. The minimum detectable dose line CD4 counts were used. More parsimonious models were of BDNF is typically less than 62 pg/mL. The repeatability of explored by removal of covariates, one at a time, starting with the BDNF ELISA as measured by intra-assay precision was the covariate with the largest P value, until the final full model 6%, and the reproducibility as measured by interassay precision was achieved. was 9%. Coefficient of variation (CV) was 7.9 (CV% ¼ SD/ mean  100%). Results Covariates Brain-Derived Neurotrophic Factor and Sample Upon entry into the study, data were collected at baseline and Characteristics after 24 weeks by using standardized questionnaires; socio- demographic (age, gender, income, and race/ethnicity) and med- Among the 400 PLWH, a wide concentration range of BDNF ical history information and the following covariates were was found in circulation, from 298 to >20 000 (mean 8384 obtained (ie, AIDS-defining conditions¼ yes/no and US Centers + 6366 pg/mL). Based on our prior studies, participants were Mı´guez-Burbano et al 457 Table 1. Sample Characteristics by BDNF Groups. Brain-Derived Neurotrophic Factor and T Cells P Participants had open access to antiretroviral (ARV) medicat, Variable BDNF < 4000 BDNF > 4001 Value and all received a prescription for ART. Most were receiving Truvada (44%) followed by Atripla (22%) alone or in combina- Age 43 + 5.7 42.5 +7.8 tion with Norvir (32%) or Kaletra (13%). At the follow-up visit, Gender all but 2% of the sample were not taking ART. Notably, all Men 74% 56% .01 non-HAUs were taking ART as prescribed. Adherence, as mea- Women 26% 44% Race/ethnicity sured by the ACTG questionnaire, was similar between the African American 63% 73% .1 groups and was high during the week (93%) and more limited Black Caribbean 3% 2% during weekends (83%, baseline). Adherence was also similar Hispanic 27% 20% between the BDNF groups (85% versus 87%). White 7% 5% Since little is known concerning the plausible impact of ART Education (years of 11.6 + 2.3 11.3 + 2.3 .2 on BDNF, we first compared the values between those few who school) were prescribed but were not taking their ARV medications with Albumin, mg/dL 4 + 0.6 4.2 + 0.5 .9 Liver enzymes those receiving treatment. Analyses indicated no significant dif- AST, IU/L 32.5 + 23 34.1 + 18 .6 ferences in the BDNF levels between the treatment groups ALT, IU/L 37.1 + 33 37.8 + 20 .8 (8712 + 6917 versus 7146 + 4526 pg/mL, P ¼ .2). Despite no Total number of drinks 17 +316 + 1.3 .8 differences in treatment or adherence, we the relationship per week between BDNF and the cellular immune response. Correlations CD4 counts between plasma BDNF and T-cell parameters were all signifi- Baseline 310.7 + 219 477.5 + 299 .0001 Follow-up 287.9 + 181 463.6 + 300 cant. A moderate statistically significant correlation was found CD8 counts between BDNF and CD3 counts (r ¼ .345, P ¼ .0001). The asso- Baseline 756.8 + 389 921.7 + 423 .004 ciation between BDNF levels and CD4 counts was also signifi- Follow-up 850.4 + 442 871.6 + 489 .8 cant (r ¼ .355, P ¼ .0001), but was less strong with CD8 HIV viral load log counts (r ¼ .228, P ¼ .0001). Baseline 74,078 + 35,489 26,813 + 7,015 .1 At baseline, the low-BDNF group exhibited significantly Follow-up 33,898 + 12,053 16,480 + 12,053 .001 lower CD3 counts than that of the high-BDNF group Abbreviations: BDNF, brain-derived neurotrophic factor; AST, aspartate ami- (1087.5 + 482.8 versus 1426.5 + 542.1 cells/mm ). Then, notransferases; ALT, alanine aminotransferases. analyses were focused on determining whether, during ART, BDNF deficiencies have an impact on CD4 count status. As shown in Table 1, the low-BDNF group exhibited signifi- dichotomized as more than and less than 4000 pg/mL: low- cantly lower CD4 counts (310.7 + 219 cells/mm )than BDNF group for those with values <4000 pg/mL and high- BDNF group for those with values >4001 pg/mL. To permit those in the high-BDNF group (477.5 + 299 cells/mm , a better definition, a brief characterization of the different P ¼ .0001). Significant group differences in CD8 counts groups was performed. were also observed; patients in the low-BDNF group exhib- As depicted in Table 1, the BDNF levels differed by gen- ited lower CD8 counts. Then, CD4 counts of more than 500 der, with women exhibiting the highest levels (9958.9 + 6578 cells/mm were used as a proxy of immune response for versus 7470 + 6068 pg/mL, P ¼ .001). Participants in the low- those under treatment. Univariate analyses indicated that BDNF group were twice as likely to be male (95% CI: 1.4-3.4, PLWH with low BDNF were 3 times less likely to have a P ¼ .003). Although at first glance age was not significantly CD4 count of more than 500 cells/mm (OR ¼ 3.3, 95% different, we pursued additional analyses as prior literature CI: 1.8-6.3, P ¼ .0001). indicated an age effect over BDNF. Indeed, PLWH aged Since gender differences in the BDNF levels were evident, 50 years and older had significantly higher levels than that of we pursued gender analysis. At baseline, women had a mean their younger counterparts (11 207 + 8790 versus 8369 + 6529, CD4 count of 508 + 323 cells/mm compared with 393 + P ¼ .003). Race distribution was similar, though a slightly 267 cells/mm for men (P ¼ .001). Indeed, women were twice as likely than men to have a CD4 count of more than 500 cells/ (nonsignificant) higher proportion of Hispanics were in the mm (OR: 2.3, 95% CI: 1.4-3.4, P ¼ .0001). low-BDNF group. Overall parameters, such as albumin and Lymphocyte profiles were available at the 6-month follow-up, liver enzymes, were similar between the groups. and a similar trend was observed: the low-BDNF group exhibited Notably, participants in the low-BDNF group were more likely even lower CD4 counts (287.9 + 181 cells/mm ) compared to to be (OR ¼ 1.6, 95% CI: 1-2.4; P ¼ .01) and to have thrombocy- those in the high-BDNF group (463.6 + 300 cells/mm , P ¼ topenia (OR¼ 3; 95% CI: 1-7.8, P¼ .04). While in those who had .0001). In adjusted regression models, the low-BDNF group had less than 3 drinks per day, no significant differences were greater odds of not increasing their T-cell counts during the 6 observed, beyond this the threshold differences were always sta- months of receiving ART (relative risk [RR] ¼ 11, P ¼ .02). Dif- tistically significant (see Figure 1). However, it needs to be recog- nized that not all HAUs exhibited low BDNF levels. ferences in CD8 counts were no longer significant. 458 Journal of the International Association of Providers of AIDS Care 13(5) Table 2. Multivariate Analyses Examining Predictors of CD4 Counts after 24 Weeks of ART. Coefficients Unstandardized Coefficients Standardized Coefficients Model B Std Error b t Sig 1 (Constant) 419.376 73.256 5.725 .000 BDNF .006 .003 .158 2.244 .026 Gender 123.824 40.394 .216 3.065 .003 Viral log 85.188 14.844 .398 5.739 .000 Abbreviations: ART, antiretroviral therapy; Std, standard; Sig, significance; BDNF, brain-derived neurotrophic factor; CDC, US Centers for Disease Control and Prevention; TCP, thrombocytopenia. The results presented here emerged from final multivariate regression analyses. The model was adjusted for covariates that in previous studies have been related to CD4 counts after ART (ie, age, race, gender, income, albumin, anemia, TCP, and liver enzymes), HIV factors (viral load, CD4, CDC stage, years living with HIV, ART, and adherence). The difference in CD4 count response between men and age, education, gender, race, albumin, and liver enzymes), HIV women also increased with time on ART. At the 6-month visit, factors (VL, CD4 count, CDC stage, years living with HIV, ART, men exhibited significantly lower CD4 counts when compared and adherence), and those variables that were significant in the to women (365 + 253 versus 521 + 313 cells/mm , P ¼ .0001). simple regression analyses (ie, thrombocytopenia [TCP], alcohol They also exhibited lower CD8 counts (869.0 + 437 versus use, and BDNF). After removal of covariates with no significant P 911.5 + 503 cells/mm , P ¼ .004). value, these analyses indicate that alcohol use and BDNF altera- tions are significant predictors of viral-immune status (see Tables 2and 3). Brain-Derived Neurotrophic Factor and VL The first regression model examined the predictors of CD4 The average VL burden was 36 849 + 8935 copies/mL (log ¼ counts at the last visit, and the R was .24, indicating that the 2.7 + 1.3). Notably, nearly half of the study population had a model predicted approximately 24% of the variance in the VL of less than 400 copies/mL. Viral burden was not signifi- CD4 counts. The model indicates that no demographic or cantly different between genders, but this time females exhib- disease-specific variables except for gender and VL were asso- ited slightly higher burden than males did. ciated with immunologic response. Alterations in BDNF and To further investigate the relationship between BDNF the number of drinks consumed per day were significantly levels and VL, we started by correlating the 2 variables, and associated with VL at the last visit. A trend was evident for analyses demonstrated a significant relationship between older age (P ¼ .06). plasma BDNF and HIV VLs (r ¼.225, P ¼ .0001). The The second model (see Table 3) was aimed to predict VL low-BDNF group had higher VL (74,078 + 35,489 com- status at the last visit, and the R was .25, indicating that the pared to.... 26,813 + 7,015 3.2 + 1.3 log) compared to model predicted approximately 25% of the variance in viral those in the high-BDNF group (log ¼ 2.6 + 1.3, P ¼ status. In this fully adjusted model (ie, age, education, gender, .004). In adjusted regression models, the low-BDNF group race, albumin, liver function tests, anemia, TCP, CD4 count, had greater odds of having detectable VLs compared to those in CDC stage, years living with HIV, ART, and adherence), the high-BDNF group (RR ¼ 1.35, 95% CI: 1.04-1.75; P ¼ .03). BDNF alterations and being an HAU were significantly associ- At the follow-up visit, the mean VL of the sample was sig- ated with virological status during treatment. Low CD4 counts nificantly reduced to 21 550 + 5638 copies/mL. Undetectable at baseline were also significant predictors of VL at the end of VL were achieved by 66% of the sample at the last visit. Paired the study. As depicted in the Table 3, the interaction between t-test analyses indicated that being in the low-BDNF group was HAU and BDNF was not significant. associated with less improvement in CD4 count. Additional analyses indicated that the low-BDNF group was less likely Discussion than patients in the high-BDNF group to achieve viral suppres- sion during the follow-up period (OR: 1.5, 95% CI: 1-2.2, P ¼ The present investigation began by demonstrating that the .03). This suggests that BDNF could, in part, explain the changes in the T-cell compartment are generally widely vari- incomplete viral suppression. able and confirmed that HAU are more likely to exhibit poor immunological responses. Yet, the mechanisms underpinning these differences remain to be fully elucidated. In this regard, Multivariate this article is unique in demonstrating the role of BDNF over To examine the independent effect of alcohol and BDNF levels on a wide range of key immunological parameters in PLWH CD4 count and VL, at the last visit we used a multivariate analysis receiving ART. First, our analyses support the hypothesis that to adjust for potential confounders, such as sociodemographic (ie, reduced levels of BDNF may be a risk factor for incomplete Mı´guez-Burbano et al 459 Table 3. Multivariate Analyses Examining Predictors of HIV Viral Load. Coefficients Unstandardized Coefficients Standardized Coefficients Model B Std Error b t Sig 1 (Constant) 1.444 .249 5.792 .000 BDNF 2.281E-5 .000 .172 3.276 .001 HAU .165 .092 .092 1.782 .076 CD4 .240 .111 .109 2.166 .031 BDNF  HAU 1.010E-5 .000 .090 1.680 .094 Abbreviations: ART, antiretroviral therapy; BDNF, brain-derived neurotrophic factor; HAU, hazardous alcohol use; Std, standard; Sig, significance; TCP, thrombocytopenia. This table illustrates the final multivariate regression analyses for viral load after 24 weeks of receiving ART. The results were adjusted for covariates that in prior studies were associated with viral loads after ART (ie, age, race, gender, income, albumin, anemia, and liver enzymes), HIV factors (viral load, CD4, ART, and adherence), and those variables that were significant in the simple regression analyses (TCP, age, gender, BDNF, and HAU). immunological responses to ART. Furthermore, it contributes During ART, patients in the low-BDNF group had higher new pieces of information by demonstrating that the effects are VL compared to those in the high-BDNF group. This is in line not global, as we observed significant effects on CD3, CD4, with the experimental studies in which BDNF downregulates and B cell counts, but not on CD8 counts. Second, our data also CXCR4 and 5 expressions, the 2 coreceptors implicated in 9-11,27 demonstrated that in vivo BDNF plays an important role in viral HIV infection. These findings explain why blunted control. These findings may contribute new insights into our response to HAART was observed in these individuals. understanding of the immune alterations associated with HIV. Nevertheless, our results had 2 limitations: participants were They are clinically relevant, given that a sizable proportion of restricted to those followed at University of Miami/Jackson individuals receiving highly active ART (HAART) do not fully Memorial facilities and we cannot establish causality. Nonethe- respond to treatment. In our opinion, our results are also signif- less, these results may have important research, clinical, and icant because they open a potential therapeutic avenue to be tar- therapeutic implications. They improve our understanding of geted mainly to address poor virological and immunological the neuroimmune system. By identifying those HIV-infected responses. Hence, BDNF can be seen as a potential therapeutic individuals at risk for incomplete ART responses, such as immune agent and not only neuroprotective. HAULWH and those with decreased BDNF levels, the study Data also demonstrated that women achieved better immune may permit early clinical follow-ups and interventions. The response to ART. Notably, analyses demonstrated that BDNF is findings may also guide the development of future adjuvant also significantly higher among females, suggesting that this bio- therapies. logical difference may indeed be a factor mediating gender differ- ences. The gender-based difference in immune reconstitution has Authors’ Note been inconsistently described in previous studies. Some studies The study was run at the Clinical Research Center, University of 22,23 are showing a better female immune reconstitution, whereas Miami Miller School of Medicine, Miami, FL 33136, USA. This 2 literature reviews concluded that there are no gender differences research was presented at the International Conference on HIV/AIDS, 24-26 regarding virological and immunological responses to ART. STDs and STIs during October 24-25, 2013 at Holiday Inn Orlando Our data, indicating a close relationship between BDNF and International Airport, Orlando, FL, USA. immune status, are in accord with laboratory experiments, sug- gesting that BDNF could be involved in promoting the resilience Declaration of Conflicting Interests of T cells. Given our results, and the well-known antiapoptotic The author(s) declared no potential conflicts of interest with respect to effect of the neurotrophins for T cells, we may speculate that the research, authorship, and/or publication of this article. a decrease in BDNF could be one of the pathological mechan- isms used by HIV-1 to induce the apoptosis of T cells. Although Funding our data suggest that recovery of CD4 counts in individuals The authors received no financial support for the research, authorship, receiving HAART is related to BDNF status, the study focused and/or publication of this article: The grant was funded by the NIAAA only on quantitative analyses. Functional assays will be needed R01 AA018095-01A1 and the NIAAA 1U24AA022002-01 grants. in future research to ascertain whether the quantitative recovery of the CD4 T-cell compartment that we observed by phenotypi- References cal measures also leads to improvements in T-cell function. Results from the multivariate model indicating that BDNF is 1. Panel on Antiretroviral Guidelines for Adults and Adolescents related to VL are congruent with BDNF’s known effects. Guidelines for the use of antiretroviral agents in HIV-1-infected 460 Journal of the International Association of Providers of AIDS Care 13(5) 15. Avdoshina V, Garzino-Demo A, Bachis A, et al. HIV-1 decreases adults and adolescents. Department of Health and Human Services. http://aidsinfo.nih.gov/guidelines. Accessed October 20, 2013. the levels of neurotrophins in human lymphocytes. AIDS. 2011; 2. HIV clinical resource. UPDATE: antiretroviral therapy. http://www. 25(8):1126-1128. hivguidelines.org/clinical-guidelines/adults/antiretroviral-therapy/. 16. Sun MK, Alkon DL. Differential gender-related vulnerability to Accessed October 20, 2013 depression induction and converging antidepressant responses in 3. Marconi VC, Grandits G, Okulicz JF, et al. Cumulative viral load Rats. J Pharmacol Exp Ther. 2006;316(2):926-932. and virologic decay patterns after antiretroviral therapy in HIV- 17. Babor TF, Higgins-Biddle JC, Saunders JB, et al. The Alcohol infected subjects influence CD4 recovery and AIDS. PLoS ONE Use Disorders Identification Test: Guidelines for Use in Pri- Med. 2011;6(5):e17956. mary Care, 2001; 2nd Edition, WHO Document No. WHO/ 4. Mı ´guez MJ, Lewis JE, Fishman J, Asthana D, Malow RM. The MSD/MSB/01.62. Geneva, Switzerland: World Health Organi- influence of different types of alcoholic beverages on disrupting zation; 2001. highly active antiretroviral treatment (HAART) outcome. Alcohol 18. Skinner HA, Horn JL. Alcohol Dependence Scale (ADS): Alcohol. 2009;44(4):366-371. User’s Guide. Toronto, Canada: Addiction Research Founda- 5. Mı ´guez MJ, Lewis JE, Moreno J, Fishman J. Cognitive perfor- tion; 1984. mance and the thymus among HIV infected subjects receiving 19. John PA, Veronica BW. NIH Publication No. 03–3745 Revised 2003. HAART. J Biologics Targets Ther. 2008;2(2):321-327. http://pubs.niaaa.nih.gov/publications/AssessingAlcohol/ ... /10 _ 6. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected ADS.pdf. Accessed October 20, 2013. patients starting highly active antiretroviral therapy: a collabora- 20. Sobell LC, Sobell MB. Alcohol consumption measures. In: Allen tive analysis of prospective studies. Lancet. 2002;360(9327): JP, Columbus M, eds. Assessing alcohol problems: a guide for 119-129. clinicians and researchers. Bethesda, MD: National Institute on 7. Kaufmann GR, Furrer H, Ledergerber B, et al. Characteristics, Alcohol Abuse and Alcoholism; 1995;55-73. determinants, and clinical relevance of CD4 T cell recovery to 21. Klein AB, Williamson R, Santini MA, et al. Blood BDNF concen- <500 cells/mL in HIV type 1-infected individuals receiving potent trations reflect brain-tissue BDNF levels across species. Int J Neu- antiretroviral therapy. Clin Infect Dis. 2005;41(3):361-372. ropsychopharmacol. 2011;14(3):347-353. 8. Moore RD, Keruly JC. CD4 cell count 6 years after com- 22. Gazzola L, Tincati C, Bellistri GM, Monforte Ad, Marchetti G. mencement of highly active antiretroviral therapy in persons The absence of CD4þ T cell count recovery despite receipt of vir- with sustained virologic suppression. Clin Infect Dis. 2007; ologically suppressive highly active antiretroviral therapy: clini- 44(3):441-446. cal risk, immunological gaps, and therapeutic options. Clin 9. Bachis A, Major EO, Mocchetti I. Brain-derived neurotrophic fac- Infect Dis. 2009;48(3):328-337. tor inhibits human immunodeficiency virus-1/gp120-mediated 23. Giordano TP, Wright JA, Hasan MQ, White AC Jr, Graviss cerebellar granule cell death by preventing gp120 internalization. EA, Visnegarwala F. Do sex and race/ethnicity influence J Neurosci. 2003;23(13):5715-5722. CD4 cell response in patients who achieve virologic suppres- 10. Bachis A, Avdoshina V, Zecca L, Parsadanian M, Mocchetti I. sion during antiretroviral therapy? Clin Infect Dis. 2003; Human immunodeficiency virus type 1 alters brain-derived neu- 37(3):433-437. rotrophic factor processing in neurons. J Neurosci. 2012;32(28): 24. Hunt PW, Deeks SG, Rodriguez B, et al. Continued CD4 cell 9477-9484. count increases in HIV-infected adults experiencing 4 years of 11. Nosheny RL, Amhed F, Yakovlev AG, et al. Brain-derived neuro- viral suppression on antiretroviral therapy. AIDS. 2003;17(13): trophic factor prevents the nigrostriatal degeneration induced by 1907-1915. human immunodeficiency virus-1 glycoprotein 120 in vivo. Eur 25. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1- J Neurosci. 2007;25(8):2275-2284. infected patients in the first year of antiretroviral therapy: compar- 12. De Santi L, Cantalupo L, Tassi M, Raspadori D, Cioni C, Annun- ison between low-income and high-income countries. Lancet. ziata P. Higher expression of BDNF receptor gp145trkB is asso- 2006;367(9513):817-824. ciated with lower apoptosis intensity in T cell lines in multiple 26. Maman D, Pujades-Rodriguez M, Subtil F, et al. Gender differ- sclerosis. J Neurol Sci. 2009;277(1-2):65-70. ences in immune reconstitution: a multicentric cohort analysis 13. Azoulay D, Urshansky N, Karni A. Low and dysregulated BDNF in sub-Saharan Africa. PLoS One. 2012;7(2):e31078. secretion from immune cells of MS patients is related to reduced 27. Nicastri E, Leone S, Angeletti C, et al. Sex issues in HIV-1- neuroprotection. J Neuroimmunol. 2008;195(1-2):186-193. infected persons during highly active antiretroviral therapy: a sys- 14. Maroder M, Bellavia D, Meco D, et al. Expression of trKB neu- tematic review. J Antimicrob Chemother. 2007;60(4):724-732. rotrophin receptor during T cell development. Role of brain 28. Avdoshina V, Becker J, Campbell L, et al. Neurotrophins modu- derived neurotrophic factor in immature thymocyte survival. J late the expression of chemokine receptors in the brain. J Neuro- Immunol. 1996;157(7):2864-2872. Virol. 2011;17(1):58-62. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of the International Association of Providers of AIDS Care (JIAPAC) SAGE

Beyond the Brain: The Role of Brain-Derived Neurotrophic Factor in Viroimmune Responses to Antiretroviral Therapy among People Living with HIV with and without Alcohol Use

Download PDF
7 pages

Loading next page...
 
/lp/sage/beyond-the-brain-the-role-of-brain-derived-neurotrophic-factor-in-EbZI7LfDkb

References (27)

Publisher
SAGE
Copyright
Copyright © 2022 by SAGE Publications Inc. unless otherwise noted. Manuscript content on this site is licensed under Creative Commons Licenses
ISSN
2325-9574
eISSN
2325-9582
DOI
10.1177/2325957414535253
Publisher site
See Article on Publisher Site

Abstract

Objective: Given the emerging data suggesting the key role of brain-derived neurotrophic factor (BDNF) in the immune system, we assessed longitudinally whether BDNF depletions induced by hazardous alcohol use (HAU) would impact a response to antiretroviral therapy (ART). Methods: In a prospective single-site cohort, virological and immunological responses to ART in 200 hazardous and 200 nonhazardous users were obtained, along with plasma BDNF levels. Results: Hazardous drinkers were more likely to have BDNF levels <4000 pg/mL (odds ratio [OR] ¼ 1.6, P ¼ .01). Participants with BDNF <4000 pg/mL were less likely to have CD4 counts of more than 500 cells/mm (P ¼ .02) and to achieve viral suppression over the follow- up period (OR ¼ 1.5, P ¼ .03). Multivariate analysis confirmed the significant role of HAU and low BDNF in predicting viroimmune responses. Conclusion: Hazardous alcohol use was associated with BDNF alterations, which in turn were linked to a limited response to ART in terms of viral suppression and CD4 count improvements. Keywords HIV/AIDS, alcohol, BDNF, CD4, viral load As indicated by many studies, brain-derived neurotrophic Introduction factor (BDNF) is a neutrophil mostly known for its role in the Antiretroviral therapy (ART) has been one of the greatest 9-11 central nervous system (CNS) that is affected by HIV. advances in the fight against the HIV. When a complete viro- Brain-derived neurotrophic factor plays an important role in logic suppression (viral load [VL] <400 copies/mL by 24 a variety of brain functions, including the lasting potentiation weeks or <50 copies/mL by 48 weeks) is attained, people living of synaptic efficacy that modulates brain plasticity. Brain- with HIV (PLWH) benefit from immune reconstitution and derived neurotrophic factor is involved in learning and memory 1,2 AIDS-free survival. Unfortunately, incomplete suppression processes and has been closely associated with cognitive deficit of viral replication, which is present in a sizable proportion 9-11 in HIV, aging, and neurodegenerative diseases. Until of cases, results in immune exhaustion, the emergence of drug recently, BDNF has been linked with the immune system, but resistance, and ultimately in increased morbidity and mortal- ity. Even among patients reaching these virologic targets, sub- stantial interindividual differences in immune responses occur 1 School of Integrated Science and Humanity, Florida International University, among them, suggesting that other factors may be involved. Miami, FL, USA Department of Medicine, University of Miami School of Medicine, Miami, FL, Our group, along with others, has demonstrated that hazardous USA alcohol use (HAU) influences CD4 count recovery and clinical Departments of Epidemiology and Medicine, University of Florida, Gaines- outcomes. Our previous studies indicated that alcohol ville, FL, USA decreases the size and cellularity of the thymus, leading to low Department of Psychiatry and Behavioral Sciences, University of Miami Miller 4,5 total numbers of peripheral T cells. Others have demon- School of Medicine, Miami, FL, USA strated only small gradual increases in CD4 counts that largely 3 6-8 Corresponding Author: remained less than 500 cells/mm . However, the thymus Marı´aJose´ Miguez-Burbano, School of Integrated Science and Humanity, model only partially explains the impaired recovery of hazar- Florida International University, 13313 Kingsbury Drive Wellington, FL 33414, dous alcohol users living with HIV (HAULWH), indicating the USA. need to look for additional factors. Email: [email protected] Mı´guez-Burbano et al 455 Figure 1. Levels of brain-derived neurotrophic factor (BDNF) by amounts of alcohol consumption. several lines of evidence indicate that BDNF plays important Participants were dichotomized as hazardous and non-HAUs, 12-14 roles in proliferation, apoptosis, and T-cell survival. Yet, its based on self-reports of alcohol intakes. Alcohol consumption role in CD4 count depletion or immune replenishment under scores were computed by averaging cross products of quantity ART is unknown. Of interest is the study by Avdoshina and col- and frequency of beer/wine and hard liquor reported on the leagues who did not relate the findings to genetic mutations, Alcohol Use Disorders Identification Test, developed by the nor to platelet counts. However, their sample was small, and World Health Organization (WHO) as a simple method of information on platelet counts was not provided. Although evi- screeningfor excessive drinking andtoassistinbrief assess- 16 17,18 dence that gender can impact BDNF levels, the authors failed ment, and the Alcohol Dependence Scale responses. to analyze gender differences in BDNF levels. Then, based on the National Institute of Alcohol Abuse and In light of these findings, we examined whether a decrease in Alcoholism guidelines criteria, men who reported >14 drinks/ BDNF could explain the reduced T cell counts and the excessive week or >4 drinks in 1 day and women >7 drinks/week or >3 VL observed in PLWH (with comorbidities such as HAU and/or drinks in 1 day were classified as HAU, while those who thrombocytopenia [TCP]). Because the Platelets Mediating reported fewer drinks were categorized as non-HAU. Alcohol and HIV Damage Study (PADS) includes a gender- The group was chosen to represent relatively ‘‘pure’’ alcohol diverse sample of participants, with substantial variation in alco- users with minimal drug use and without major confounding hol use, we were afforded a valuable opportunity to explore the factors. Nonambulatory patients and those presenting major multifaceted associations among alcohol use, BDNF, gender, medical comorbidities such as major neuropsychological (ie, and immune responses to ART. active CNS opportunistic infection, tumors, and developmental disorders), immune-based (ie, malignancies and autoimmune diseases), and chronic diseases (renal failure) were excluded. In addition, individuals who had cirrhosis, active viral hepati- Methods tis, or liver enzymes and who were 2 standard deviations (SDs) Sampling above the normal values were not eligible to participate in the study. To reduce the confounding effects of illicit drug use, the The Platelets Mediating Alcohol and HIV Damage Study PADS Diagnostic and Statistical Manual of Mental Disorders (Fourth is a large, single-site multiethnic cohort, consisting of 400 Edition, Text Revision) questionnaire was applied, and those PLWH who are at least 18 years old and younger regular care participants who were dependent on drugs or injecting illicit at Miami’s primary open-access public health system. Patients psychoactive substances were also excluded. Otherwise, the were recruited via flyers, personal contact at the clinics, or calls patients were enrolled. to our office to schedule appointments. Recruitment and follow- The Platelets Mediating Alcohol and HIV Damage Study ups took place in the period between June 2010 and June 2012. was approved by the Central Governing Institutional Review Our choice of PLWH in an open-access public health system Boards at Florida International University and University of with standard treatment protocols was purposefully designed Miami. The study was conducted according to the principles to minimize social, medical, and treatment inequalities. 456 Journal of the International Association of Providers of AIDS Care 13(5) expressed in the Declaration of Helsinki. Those participants for Disease Control and Prevention [CDC] clinical staging): who provided written informed consent were consecutively complete blood counts (thrombocytopenia was defined as enrolled and followed over a period of 6 months. platelet counts of less than 150  103 cells/mm [41-42] and a biochemical profile (calcium, sodium, potassium, albumin, glucose, lipids, kidney, and liver function). HIV- Laboratory Outcomes related and not-related treatments (ie, start date and discon- Blood was drawn from fasting patients to best evaluate the tinued) were obtained and confirmed with the pharmacy and immunological, hematological, and platelet-associated factor medical records. An AIDS Clinical Trial Group (ACTG) profiles. Blood samples were collected and processed within self-reported adherence questionnaire was used at each visit. 6 hours. Isolated peripheral blood mononuclear cells were pre- Basedonthe missed dosesper week andduringthe week- pared for 4-color direct immunofluorescence procedures (Becton end, the percentage of adherence was calculated at baseline Dickinson, San Jose, California). Flow cytometry quantified the and at the follow-up visit. percentage and absolute numbers of T-lymphocyte subpopula- tions CD3/CD4 and CD3/CD8. A good immunological Statistical Analyses response was defined as having a CD4 count of more than 3 3 The data were analyzed using SAS version 8 and SPSS version 500 cells/mm or as a gain in CD4 count 50 cells/mm from 19, and P values <.05 were considered to be statistically signif- week 0 to week 24. icant. The normality of the distribution of primary outcomes of HIV viral burden was quantified using the Amplicor HIV interest was examined with a normal probability plot. Viral monitor test (Roche Diagnostic System, Indianapolis, IN). The load was log transformed to correct skewness. Following lower threshold for detection at the time of the study was 50 descriptive statistical analyses, the mean variables were com- copies/mL. Virological success was defined as achieving unde- pared using Student t test and one-way analysis of variance pro- tectable VL. Poor virological response was defined as a plasma cedures. Correlations among the main variables of interest VL >2.7 log copies/mL at week 24. were examined with Pearson coefficients. Additionally, chi- square analyses were performed to compare the proportions for Brain-Derived Neurotrophic Factor gender and race. Univariate analyses were used to calculate the odds ratios (ORs) and 95% confidence intervals (CIs). Logistic The circulating levels of BDNF were selected because prior regression analyses were used to evaluate the effects of alcohol studies have demonstrated that, although different from those (continuous or as hazardous versus nonhazardous), thrombocy- in the cerebrospinal fluid (CSF), they are correlated with CSF topenia (counts and dichotomized as thrombocytopenia <150 measures in other CNS diseases. To obtain platelet-poor 000 versus normal platelet counts), age (continuous and dichot- plasma (PPP), the blood samples were collected in EDTA- omized at age 50 years), and gender. coated tubes (plasma; BD Diagnostic Systems, New Jersey) The multivariable models aimed to predict VL and CD4 and stored on ice. Plasma was separated by centrifugation at count status at the last visit and included all covariates whose 40 C for 15 minutes at 1500g. This plasma was again recentri- likelihood ratio for P value from the univariate analyses was fuged at 10 000g and the aliquots of PPP were stored in poly- <.1. Although no significant differences were observed propylene tubes at 80 C until assayed. The BDNF levels in between groups for education level or CDC status, both vari- PPP were measured using a commercially available enzyme- ables were controlled for in the final analyses. In addition, other linked immunosorbent assay (ELISA) kit (R&D System), potential predictors (ie, gender, race/ethnicity, and body mass according to the manufacturer’s instructions. However, during index) were selected on the basis of the literature and were the standardization a sizable proportion of PLWH had a BDNF added to the model. Although the nadir CD4 count has been value of 4000 pg/mL (ceiling effect), so the samples were 20- recommended as a preferable measure of disease severity, data fold diluted. The concentration of BDNF in plasma was calcu- were not available for most participants and therefore the base- lated based on a standard curve. The minimum detectable dose line CD4 counts were used. More parsimonious models were of BDNF is typically less than 62 pg/mL. The repeatability of explored by removal of covariates, one at a time, starting with the BDNF ELISA as measured by intra-assay precision was the covariate with the largest P value, until the final full model 6%, and the reproducibility as measured by interassay precision was achieved. was 9%. Coefficient of variation (CV) was 7.9 (CV% ¼ SD/ mean  100%). Results Covariates Brain-Derived Neurotrophic Factor and Sample Upon entry into the study, data were collected at baseline and Characteristics after 24 weeks by using standardized questionnaires; socio- demographic (age, gender, income, and race/ethnicity) and med- Among the 400 PLWH, a wide concentration range of BDNF ical history information and the following covariates were was found in circulation, from 298 to >20 000 (mean 8384 obtained (ie, AIDS-defining conditions¼ yes/no and US Centers + 6366 pg/mL). Based on our prior studies, participants were Mı´guez-Burbano et al 457 Table 1. Sample Characteristics by BDNF Groups. Brain-Derived Neurotrophic Factor and T Cells P Participants had open access to antiretroviral (ARV) medicat, Variable BDNF < 4000 BDNF > 4001 Value and all received a prescription for ART. Most were receiving Truvada (44%) followed by Atripla (22%) alone or in combina- Age 43 + 5.7 42.5 +7.8 tion with Norvir (32%) or Kaletra (13%). At the follow-up visit, Gender all but 2% of the sample were not taking ART. Notably, all Men 74% 56% .01 non-HAUs were taking ART as prescribed. Adherence, as mea- Women 26% 44% Race/ethnicity sured by the ACTG questionnaire, was similar between the African American 63% 73% .1 groups and was high during the week (93%) and more limited Black Caribbean 3% 2% during weekends (83%, baseline). Adherence was also similar Hispanic 27% 20% between the BDNF groups (85% versus 87%). White 7% 5% Since little is known concerning the plausible impact of ART Education (years of 11.6 + 2.3 11.3 + 2.3 .2 on BDNF, we first compared the values between those few who school) were prescribed but were not taking their ARV medications with Albumin, mg/dL 4 + 0.6 4.2 + 0.5 .9 Liver enzymes those receiving treatment. Analyses indicated no significant dif- AST, IU/L 32.5 + 23 34.1 + 18 .6 ferences in the BDNF levels between the treatment groups ALT, IU/L 37.1 + 33 37.8 + 20 .8 (8712 + 6917 versus 7146 + 4526 pg/mL, P ¼ .2). Despite no Total number of drinks 17 +316 + 1.3 .8 differences in treatment or adherence, we the relationship per week between BDNF and the cellular immune response. Correlations CD4 counts between plasma BDNF and T-cell parameters were all signifi- Baseline 310.7 + 219 477.5 + 299 .0001 Follow-up 287.9 + 181 463.6 + 300 cant. A moderate statistically significant correlation was found CD8 counts between BDNF and CD3 counts (r ¼ .345, P ¼ .0001). The asso- Baseline 756.8 + 389 921.7 + 423 .004 ciation between BDNF levels and CD4 counts was also signifi- Follow-up 850.4 + 442 871.6 + 489 .8 cant (r ¼ .355, P ¼ .0001), but was less strong with CD8 HIV viral load log counts (r ¼ .228, P ¼ .0001). Baseline 74,078 + 35,489 26,813 + 7,015 .1 At baseline, the low-BDNF group exhibited significantly Follow-up 33,898 + 12,053 16,480 + 12,053 .001 lower CD3 counts than that of the high-BDNF group Abbreviations: BDNF, brain-derived neurotrophic factor; AST, aspartate ami- (1087.5 + 482.8 versus 1426.5 + 542.1 cells/mm ). Then, notransferases; ALT, alanine aminotransferases. analyses were focused on determining whether, during ART, BDNF deficiencies have an impact on CD4 count status. As shown in Table 1, the low-BDNF group exhibited signifi- dichotomized as more than and less than 4000 pg/mL: low- cantly lower CD4 counts (310.7 + 219 cells/mm )than BDNF group for those with values <4000 pg/mL and high- BDNF group for those with values >4001 pg/mL. To permit those in the high-BDNF group (477.5 + 299 cells/mm , a better definition, a brief characterization of the different P ¼ .0001). Significant group differences in CD8 counts groups was performed. were also observed; patients in the low-BDNF group exhib- As depicted in Table 1, the BDNF levels differed by gen- ited lower CD8 counts. Then, CD4 counts of more than 500 der, with women exhibiting the highest levels (9958.9 + 6578 cells/mm were used as a proxy of immune response for versus 7470 + 6068 pg/mL, P ¼ .001). Participants in the low- those under treatment. Univariate analyses indicated that BDNF group were twice as likely to be male (95% CI: 1.4-3.4, PLWH with low BDNF were 3 times less likely to have a P ¼ .003). Although at first glance age was not significantly CD4 count of more than 500 cells/mm (OR ¼ 3.3, 95% different, we pursued additional analyses as prior literature CI: 1.8-6.3, P ¼ .0001). indicated an age effect over BDNF. Indeed, PLWH aged Since gender differences in the BDNF levels were evident, 50 years and older had significantly higher levels than that of we pursued gender analysis. At baseline, women had a mean their younger counterparts (11 207 + 8790 versus 8369 + 6529, CD4 count of 508 + 323 cells/mm compared with 393 + P ¼ .003). Race distribution was similar, though a slightly 267 cells/mm for men (P ¼ .001). Indeed, women were twice as likely than men to have a CD4 count of more than 500 cells/ (nonsignificant) higher proportion of Hispanics were in the mm (OR: 2.3, 95% CI: 1.4-3.4, P ¼ .0001). low-BDNF group. Overall parameters, such as albumin and Lymphocyte profiles were available at the 6-month follow-up, liver enzymes, were similar between the groups. and a similar trend was observed: the low-BDNF group exhibited Notably, participants in the low-BDNF group were more likely even lower CD4 counts (287.9 + 181 cells/mm ) compared to to be (OR ¼ 1.6, 95% CI: 1-2.4; P ¼ .01) and to have thrombocy- those in the high-BDNF group (463.6 + 300 cells/mm , P ¼ topenia (OR¼ 3; 95% CI: 1-7.8, P¼ .04). While in those who had .0001). In adjusted regression models, the low-BDNF group had less than 3 drinks per day, no significant differences were greater odds of not increasing their T-cell counts during the 6 observed, beyond this the threshold differences were always sta- months of receiving ART (relative risk [RR] ¼ 11, P ¼ .02). Dif- tistically significant (see Figure 1). However, it needs to be recog- nized that not all HAUs exhibited low BDNF levels. ferences in CD8 counts were no longer significant. 458 Journal of the International Association of Providers of AIDS Care 13(5) Table 2. Multivariate Analyses Examining Predictors of CD4 Counts after 24 Weeks of ART. Coefficients Unstandardized Coefficients Standardized Coefficients Model B Std Error b t Sig 1 (Constant) 419.376 73.256 5.725 .000 BDNF .006 .003 .158 2.244 .026 Gender 123.824 40.394 .216 3.065 .003 Viral log 85.188 14.844 .398 5.739 .000 Abbreviations: ART, antiretroviral therapy; Std, standard; Sig, significance; BDNF, brain-derived neurotrophic factor; CDC, US Centers for Disease Control and Prevention; TCP, thrombocytopenia. The results presented here emerged from final multivariate regression analyses. The model was adjusted for covariates that in previous studies have been related to CD4 counts after ART (ie, age, race, gender, income, albumin, anemia, TCP, and liver enzymes), HIV factors (viral load, CD4, CDC stage, years living with HIV, ART, and adherence). The difference in CD4 count response between men and age, education, gender, race, albumin, and liver enzymes), HIV women also increased with time on ART. At the 6-month visit, factors (VL, CD4 count, CDC stage, years living with HIV, ART, men exhibited significantly lower CD4 counts when compared and adherence), and those variables that were significant in the to women (365 + 253 versus 521 + 313 cells/mm , P ¼ .0001). simple regression analyses (ie, thrombocytopenia [TCP], alcohol They also exhibited lower CD8 counts (869.0 + 437 versus use, and BDNF). After removal of covariates with no significant P 911.5 + 503 cells/mm , P ¼ .004). value, these analyses indicate that alcohol use and BDNF altera- tions are significant predictors of viral-immune status (see Tables 2and 3). Brain-Derived Neurotrophic Factor and VL The first regression model examined the predictors of CD4 The average VL burden was 36 849 + 8935 copies/mL (log ¼ counts at the last visit, and the R was .24, indicating that the 2.7 + 1.3). Notably, nearly half of the study population had a model predicted approximately 24% of the variance in the VL of less than 400 copies/mL. Viral burden was not signifi- CD4 counts. The model indicates that no demographic or cantly different between genders, but this time females exhib- disease-specific variables except for gender and VL were asso- ited slightly higher burden than males did. ciated with immunologic response. Alterations in BDNF and To further investigate the relationship between BDNF the number of drinks consumed per day were significantly levels and VL, we started by correlating the 2 variables, and associated with VL at the last visit. A trend was evident for analyses demonstrated a significant relationship between older age (P ¼ .06). plasma BDNF and HIV VLs (r ¼.225, P ¼ .0001). The The second model (see Table 3) was aimed to predict VL low-BDNF group had higher VL (74,078 + 35,489 com- status at the last visit, and the R was .25, indicating that the pared to.... 26,813 + 7,015 3.2 + 1.3 log) compared to model predicted approximately 25% of the variance in viral those in the high-BDNF group (log ¼ 2.6 + 1.3, P ¼ status. In this fully adjusted model (ie, age, education, gender, .004). In adjusted regression models, the low-BDNF group race, albumin, liver function tests, anemia, TCP, CD4 count, had greater odds of having detectable VLs compared to those in CDC stage, years living with HIV, ART, and adherence), the high-BDNF group (RR ¼ 1.35, 95% CI: 1.04-1.75; P ¼ .03). BDNF alterations and being an HAU were significantly associ- At the follow-up visit, the mean VL of the sample was sig- ated with virological status during treatment. Low CD4 counts nificantly reduced to 21 550 + 5638 copies/mL. Undetectable at baseline were also significant predictors of VL at the end of VL were achieved by 66% of the sample at the last visit. Paired the study. As depicted in the Table 3, the interaction between t-test analyses indicated that being in the low-BDNF group was HAU and BDNF was not significant. associated with less improvement in CD4 count. Additional analyses indicated that the low-BDNF group was less likely Discussion than patients in the high-BDNF group to achieve viral suppres- sion during the follow-up period (OR: 1.5, 95% CI: 1-2.2, P ¼ The present investigation began by demonstrating that the .03). This suggests that BDNF could, in part, explain the changes in the T-cell compartment are generally widely vari- incomplete viral suppression. able and confirmed that HAU are more likely to exhibit poor immunological responses. Yet, the mechanisms underpinning these differences remain to be fully elucidated. In this regard, Multivariate this article is unique in demonstrating the role of BDNF over To examine the independent effect of alcohol and BDNF levels on a wide range of key immunological parameters in PLWH CD4 count and VL, at the last visit we used a multivariate analysis receiving ART. First, our analyses support the hypothesis that to adjust for potential confounders, such as sociodemographic (ie, reduced levels of BDNF may be a risk factor for incomplete Mı´guez-Burbano et al 459 Table 3. Multivariate Analyses Examining Predictors of HIV Viral Load. Coefficients Unstandardized Coefficients Standardized Coefficients Model B Std Error b t Sig 1 (Constant) 1.444 .249 5.792 .000 BDNF 2.281E-5 .000 .172 3.276 .001 HAU .165 .092 .092 1.782 .076 CD4 .240 .111 .109 2.166 .031 BDNF  HAU 1.010E-5 .000 .090 1.680 .094 Abbreviations: ART, antiretroviral therapy; BDNF, brain-derived neurotrophic factor; HAU, hazardous alcohol use; Std, standard; Sig, significance; TCP, thrombocytopenia. This table illustrates the final multivariate regression analyses for viral load after 24 weeks of receiving ART. The results were adjusted for covariates that in prior studies were associated with viral loads after ART (ie, age, race, gender, income, albumin, anemia, and liver enzymes), HIV factors (viral load, CD4, ART, and adherence), and those variables that were significant in the simple regression analyses (TCP, age, gender, BDNF, and HAU). immunological responses to ART. Furthermore, it contributes During ART, patients in the low-BDNF group had higher new pieces of information by demonstrating that the effects are VL compared to those in the high-BDNF group. This is in line not global, as we observed significant effects on CD3, CD4, with the experimental studies in which BDNF downregulates and B cell counts, but not on CD8 counts. Second, our data also CXCR4 and 5 expressions, the 2 coreceptors implicated in 9-11,27 demonstrated that in vivo BDNF plays an important role in viral HIV infection. These findings explain why blunted control. These findings may contribute new insights into our response to HAART was observed in these individuals. understanding of the immune alterations associated with HIV. Nevertheless, our results had 2 limitations: participants were They are clinically relevant, given that a sizable proportion of restricted to those followed at University of Miami/Jackson individuals receiving highly active ART (HAART) do not fully Memorial facilities and we cannot establish causality. Nonethe- respond to treatment. In our opinion, our results are also signif- less, these results may have important research, clinical, and icant because they open a potential therapeutic avenue to be tar- therapeutic implications. They improve our understanding of geted mainly to address poor virological and immunological the neuroimmune system. By identifying those HIV-infected responses. Hence, BDNF can be seen as a potential therapeutic individuals at risk for incomplete ART responses, such as immune agent and not only neuroprotective. HAULWH and those with decreased BDNF levels, the study Data also demonstrated that women achieved better immune may permit early clinical follow-ups and interventions. The response to ART. Notably, analyses demonstrated that BDNF is findings may also guide the development of future adjuvant also significantly higher among females, suggesting that this bio- therapies. logical difference may indeed be a factor mediating gender differ- ences. The gender-based difference in immune reconstitution has Authors’ Note been inconsistently described in previous studies. Some studies The study was run at the Clinical Research Center, University of 22,23 are showing a better female immune reconstitution, whereas Miami Miller School of Medicine, Miami, FL 33136, USA. This 2 literature reviews concluded that there are no gender differences research was presented at the International Conference on HIV/AIDS, 24-26 regarding virological and immunological responses to ART. STDs and STIs during October 24-25, 2013 at Holiday Inn Orlando Our data, indicating a close relationship between BDNF and International Airport, Orlando, FL, USA. immune status, are in accord with laboratory experiments, sug- gesting that BDNF could be involved in promoting the resilience Declaration of Conflicting Interests of T cells. Given our results, and the well-known antiapoptotic The author(s) declared no potential conflicts of interest with respect to effect of the neurotrophins for T cells, we may speculate that the research, authorship, and/or publication of this article. a decrease in BDNF could be one of the pathological mechan- isms used by HIV-1 to induce the apoptosis of T cells. Although Funding our data suggest that recovery of CD4 counts in individuals The authors received no financial support for the research, authorship, receiving HAART is related to BDNF status, the study focused and/or publication of this article: The grant was funded by the NIAAA only on quantitative analyses. Functional assays will be needed R01 AA018095-01A1 and the NIAAA 1U24AA022002-01 grants. in future research to ascertain whether the quantitative recovery of the CD4 T-cell compartment that we observed by phenotypi- References cal measures also leads to improvements in T-cell function. Results from the multivariate model indicating that BDNF is 1. Panel on Antiretroviral Guidelines for Adults and Adolescents related to VL are congruent with BDNF’s known effects. Guidelines for the use of antiretroviral agents in HIV-1-infected 460 Journal of the International Association of Providers of AIDS Care 13(5) 15. Avdoshina V, Garzino-Demo A, Bachis A, et al. HIV-1 decreases adults and adolescents. Department of Health and Human Services. http://aidsinfo.nih.gov/guidelines. Accessed October 20, 2013. the levels of neurotrophins in human lymphocytes. AIDS. 2011; 2. HIV clinical resource. UPDATE: antiretroviral therapy. http://www. 25(8):1126-1128. hivguidelines.org/clinical-guidelines/adults/antiretroviral-therapy/. 16. Sun MK, Alkon DL. Differential gender-related vulnerability to Accessed October 20, 2013 depression induction and converging antidepressant responses in 3. Marconi VC, Grandits G, Okulicz JF, et al. Cumulative viral load Rats. J Pharmacol Exp Ther. 2006;316(2):926-932. and virologic decay patterns after antiretroviral therapy in HIV- 17. Babor TF, Higgins-Biddle JC, Saunders JB, et al. The Alcohol infected subjects influence CD4 recovery and AIDS. PLoS ONE Use Disorders Identification Test: Guidelines for Use in Pri- Med. 2011;6(5):e17956. mary Care, 2001; 2nd Edition, WHO Document No. WHO/ 4. Mı ´guez MJ, Lewis JE, Fishman J, Asthana D, Malow RM. The MSD/MSB/01.62. Geneva, Switzerland: World Health Organi- influence of different types of alcoholic beverages on disrupting zation; 2001. highly active antiretroviral treatment (HAART) outcome. Alcohol 18. Skinner HA, Horn JL. Alcohol Dependence Scale (ADS): Alcohol. 2009;44(4):366-371. User’s Guide. Toronto, Canada: Addiction Research Founda- 5. Mı ´guez MJ, Lewis JE, Moreno J, Fishman J. Cognitive perfor- tion; 1984. mance and the thymus among HIV infected subjects receiving 19. John PA, Veronica BW. NIH Publication No. 03–3745 Revised 2003. HAART. J Biologics Targets Ther. 2008;2(2):321-327. http://pubs.niaaa.nih.gov/publications/AssessingAlcohol/ ... /10 _ 6. Egger M, May M, Chene G, et al. Prognosis of HIV-1-infected ADS.pdf. Accessed October 20, 2013. patients starting highly active antiretroviral therapy: a collabora- 20. Sobell LC, Sobell MB. Alcohol consumption measures. In: Allen tive analysis of prospective studies. Lancet. 2002;360(9327): JP, Columbus M, eds. Assessing alcohol problems: a guide for 119-129. clinicians and researchers. Bethesda, MD: National Institute on 7. Kaufmann GR, Furrer H, Ledergerber B, et al. Characteristics, Alcohol Abuse and Alcoholism; 1995;55-73. determinants, and clinical relevance of CD4 T cell recovery to 21. Klein AB, Williamson R, Santini MA, et al. Blood BDNF concen- <500 cells/mL in HIV type 1-infected individuals receiving potent trations reflect brain-tissue BDNF levels across species. Int J Neu- antiretroviral therapy. Clin Infect Dis. 2005;41(3):361-372. ropsychopharmacol. 2011;14(3):347-353. 8. Moore RD, Keruly JC. CD4 cell count 6 years after com- 22. Gazzola L, Tincati C, Bellistri GM, Monforte Ad, Marchetti G. mencement of highly active antiretroviral therapy in persons The absence of CD4þ T cell count recovery despite receipt of vir- with sustained virologic suppression. Clin Infect Dis. 2007; ologically suppressive highly active antiretroviral therapy: clini- 44(3):441-446. cal risk, immunological gaps, and therapeutic options. Clin 9. Bachis A, Major EO, Mocchetti I. Brain-derived neurotrophic fac- Infect Dis. 2009;48(3):328-337. tor inhibits human immunodeficiency virus-1/gp120-mediated 23. Giordano TP, Wright JA, Hasan MQ, White AC Jr, Graviss cerebellar granule cell death by preventing gp120 internalization. EA, Visnegarwala F. Do sex and race/ethnicity influence J Neurosci. 2003;23(13):5715-5722. CD4 cell response in patients who achieve virologic suppres- 10. Bachis A, Avdoshina V, Zecca L, Parsadanian M, Mocchetti I. sion during antiretroviral therapy? Clin Infect Dis. 2003; Human immunodeficiency virus type 1 alters brain-derived neu- 37(3):433-437. rotrophic factor processing in neurons. J Neurosci. 2012;32(28): 24. Hunt PW, Deeks SG, Rodriguez B, et al. Continued CD4 cell 9477-9484. count increases in HIV-infected adults experiencing 4 years of 11. Nosheny RL, Amhed F, Yakovlev AG, et al. Brain-derived neuro- viral suppression on antiretroviral therapy. AIDS. 2003;17(13): trophic factor prevents the nigrostriatal degeneration induced by 1907-1915. human immunodeficiency virus-1 glycoprotein 120 in vivo. Eur 25. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1- J Neurosci. 2007;25(8):2275-2284. infected patients in the first year of antiretroviral therapy: compar- 12. De Santi L, Cantalupo L, Tassi M, Raspadori D, Cioni C, Annun- ison between low-income and high-income countries. Lancet. ziata P. Higher expression of BDNF receptor gp145trkB is asso- 2006;367(9513):817-824. ciated with lower apoptosis intensity in T cell lines in multiple 26. Maman D, Pujades-Rodriguez M, Subtil F, et al. Gender differ- sclerosis. J Neurol Sci. 2009;277(1-2):65-70. ences in immune reconstitution: a multicentric cohort analysis 13. Azoulay D, Urshansky N, Karni A. Low and dysregulated BDNF in sub-Saharan Africa. PLoS One. 2012;7(2):e31078. secretion from immune cells of MS patients is related to reduced 27. Nicastri E, Leone S, Angeletti C, et al. Sex issues in HIV-1- neuroprotection. J Neuroimmunol. 2008;195(1-2):186-193. infected persons during highly active antiretroviral therapy: a sys- 14. Maroder M, Bellavia D, Meco D, et al. Expression of trKB neu- tematic review. J Antimicrob Chemother. 2007;60(4):724-732. rotrophin receptor during T cell development. Role of brain 28. Avdoshina V, Becker J, Campbell L, et al. Neurotrophins modu- derived neurotrophic factor in immature thymocyte survival. J late the expression of chemokine receptors in the brain. J Neuro- Immunol. 1996;157(7):2864-2872. Virol. 2011;17(1):58-62.

Journal

Journal of the International Association of Providers of AIDS Care (JIAPAC) – SAGE

Published: May 16, 2014

Keywords: HIV/AIDS; alcohol; BDNF; CD4; viral load

There are no references for this article.