HIV can enter the central nervous system (CNS) shortly after seroconversion and activate microglial and macrophage cells to release toxic factors such as excitotoxins, viral products, cytokines, and chemokines. These lead to marked inflammation, immune activation and suppression, and disruption of the blood-brain barrier (1–3) and result in neuronal damage that is visible as changes in brain white matter structure. The use of combination antiretroviral therapy (cART) has transformed HIV from a fatal disease into a chronic, manageable disease (4) and can reverse and prevent milder impairments (ANI and MND) to some extent. However, also of concern are the iatrogenic effects of the drugs on brain integrity (5, 6). For example, previous studies have demonstrated axonal dysfunction and synaptic damage in HIV patients (7, 8). Diffusion tensor imaging (DTI) is a widely used noninvasive neuroimaging technique that is very sensitive to microstructural changes (9). Previous studies have shown decreased fractional anisotropy (FA) and increased mean diffusivity (MD) throughout the brain of HIV-infected patients (10–21), even those on effective antiretroviral therapies (7, 11). However, studies have also found no significant differences between HIV-infected patients and healthy controls (22), and some investigators noted that the use of cART does not appear to prevent or reverse existing brain damage (23–25). However, most of these studies were cross-sectional and involved multiple factors, including route of infection, inclusion criteria, duration of infection, complications, different treatment regimens, and lack of patient adherence, which are difficult to control for (13, 17, 18). Because most changes occur months or even years after infection, studies to investigate early changes in the brain of HIV-infected individuals are needed. Therefore, longitudinal studies with controlled covariances should be conducted to address the above problem and shed light on the mechanisms underlying the dynamic changes that are seen with and without the use of cART.

An appropriate animal model whose pathological mechanisms are similar to those of HIV patients can be used to clarify this issue. Studies have shown that the pathology of simian immunodeficiency virus (SIV)-infected Chinese rhesus monkeys is very similar to that of HIV-infected patients, especially in terms of central nervous system (CNS) dysfunction, cognitive or behavioral deficits, and the development of AIDS (26–28). Moreover, previous DTI studies have shown SIV-infected rhesus macaques to have metric abnormalities similar to those seen in HIV patients (29, 30).

In the present study, we dynamically investigated white matter changes in SIV-mac239-infected Chinese rhesus monkeys with and without regular cART treatment and the relationship between DTI metrics and clinical tests (including plasma CD4 T cell counts, CD4/CD8 ratio, plasma viral load, and CSF viral load) at different time points.

The current study investigated longitudinal structural changes in brain white matter in 10 SIV-mac239-infected monkeys with and without regular cART treatment. We found significant white matter structural changes in these monkeys that consisted of higher mean diffusion and lower fractional anisotropy, as determined by DTI. The results showed that these effects increased with time. Furthermore, we found that significant white matter disruption occurred as early as 4 weeks after inoculation and that damage could be mitigated or even reversed with regular cART. In addition, DTI metric changes correlated significantly with clinical measures such as CD4 T cell count, CD4/CD8, plasma viral load, and CSF viral load.

The etiology underlying white matter changes in HIV-infected individuals is still unclear. Possible pathophysiological mechanisms include synaptodendritic damage, inflammation, demyelination, and even microvascular abnormalities that correlate with concomitant cardiovascular risk factors, particularly hypertension. In the context of HIV, on the basis of DTI acquisition parameters, FA reflects the axonal structural integrity aspect, which reflects the cytoskeleton and membrane integrity (34) and the structural integrity of the ricytes (35). MD has been interpreted to be related to the degree of WM microstructural density (34, 36), such as the extracellular space between WM tracts (35), which may be affected to a lesser extent by neuroinflammation, including glia involvement and myelin loss. In this case, the contribution of myelin to MD changes is questionable (35) because massive demyelination is not a common feature of HIV-related brain injury, even in patients with severe HIV encephalopathy (37). Previous studies in HIV patients and healthy controls have shown subtle but widespread white matter alterations, particularly changes in fractional anisotropy and mean diffusion (9, 11, 19, 38–45), suggesting the disorganization of white matter micro- and macrostructures (46).

Few studies have, however, investigated the initial timing of the changes, groups of regions in which white matter is first destroyed, and what happens after early cART (as early as 40 days after inoculation), which may shed light on the underlying neuropathological mechanisms and illustrate the impact of timely cART on imaging findings, immune indicators, and the relationship between them. In our study, we longitudinally examined white matter changes in rhesus macaques with and without regular cART treatment from baseline and 10 days, 4 weeks, 12 weeks, 24 weeks, and 36 weeks after inoculation. In the present study, changes were first observed 4 weeks after inoculation in the amygdala, superior temporal gyrus (STG), posterior part of the internal capsule (left), fornix, bilateral brainstem, and bilateral anterior limb of the internal capsule (ALIC-L) at FA, and the posterior limb of the internal capsule-left (PLIC-L) and right sagittal striatum (SS -R) at MD, suggesting that these regions are more susceptible to SIV infection. Studies have shown that HIV enters the central nervous system very early after seroconversion via monocytes and perivascular macrophages. The virus then replicates and causes neuronal damage mainly through neuroinflammatory mechanisms triggered by infected microglial cells. Inflammation occurs selectively in dopamine-rich areas of the brain, including subcortical areas, particularly the dopamine-rich basal ganglia, leading to subcortical dementia (47–50). Our findings that FA alternates in the amygdala and sagittal striatum, which are part of the basal ganglia, are consistent with this work. The fornix is an important hippocampal outflow tract, a structure that is critical for normal memory function and has been reported to be abnormal in AD, schizophrenia, and multiple sclerosis. Microstructural alterations to the fornix contribute to early episodic memory impairment in non-demented individuals (51–54). The internal capsule is a region that is sensitive to HIV and AIDS in individuals with cognitive impairment (55). A previous study also reported that a higher HIV RNA density of infected macrophages was found in the subcortical white matter, including IC (56). In addition, many studies also reported white matter damage in the subcortical white matter, especially the inner capsule (11, 12, 40, 57). Regarding alterations to the cerebellar peduncle and the stratum sagittale, similar changes have been found in previous studies in adolescents (58).

Twelve weeks after inoculation, the PLIC-R, GCC, BCC, EC-L, and ST-R regions showed significant changes at FA, and PLIC-R was altered at MD. Early cerebral infiltration of HIV may occur before antiretroviral drugs are administered, exposing the surrounding white matter to neurotoxic effects before effective viral suppression occurs. Periventricular white matter, such as CC, has been suggested to be particularly vulnerable to viral attack because of its proximity to the CSF, which is an HIV reservoir that can transmit the virus within 8 days of infection (59–61). Studies have also shown that the highest density of HIV-infected macrophages is found in the subcortical white matter, including the CC (56). Consistent with these early effects on the brain, white matter depletion has been found within the first 100 days of HIV transmission (62). As an important communication pathway between hemispheres, the CC is responsible for the functional integration of complex cognitive, motor, and behavioral tasks. Therefore, observations of the CC may indicate the presence of an underlying neurocognitive disorder in HIV-infected individuals. The outer capsule serves as a pathway for psychomotor functions, and many previous studies have shown the outer capsule to be altered in HIV patients (63, 64).

After cART, we found that the initial changes disappeared in BCC and FX; there was no progression in AA-WM-R, PLIC-R, or PLIC-L, slow progression in STG-WM-L, GCC, CP-R, and CP-L, and no effect on PLICR, ALIC-R, SS-R, or ST-R. In contrast, in EC-L, we found that cART damaged the structure to some extent. White matter repair by viral suppression and cART could lead to a higher degree of axonal structural integrity than the normal effects of aging. This may be particularly possible in the inner capsule because this is a region that is susceptible to white matter damage in individuals with cognitive impairment and AIDS (55). The absence of aggressive HIV-associated WM damage in the CC after cART suggests that early treatment is neuroprotective to some extent (65). On the one hand, cART may contribute to synaptic damage via oxidative stress, as demonstrated in vitro and in animal models (66). On the other hand, early and continuous antiretroviral therapy may be neuroprotective and limit white matter damage (67, 68).

Viral load levels reflect the replication of HIV virus in the body and can be used to determine disease progression and the effectiveness of antiviral treatment. A negative relationship was found between DTI parameters and viral load in SIV-mac239 infected rhesus monkeys with and without cART treatment, which is relatively consistent with a prior study that showed that myeloid cells, particularly microglia, are likely to be major reservoirs in the brain, and that more microglia exist in white matter (69). Activation of both infected and uninfected macrophages and microglia in the brain could promote inflammation and edema and affect DTI metrics. Specifically, the greater amount of inflammatory cells and surrounding debris could restrict water diffusion and lead to alterations in both FA and MD (30, 31), suggesting that the increased diffusion may reflect increased glial activation or inflammation. However, this suggestion would be made more convincing by appropriate pathologic studies.

In the clinic, current CD4 T cell counts are used as markers to monitor current immune functions and track disease progression (70). CD4/CD8 ratio is a marker of persistent inflammation and immunosenescence caused by viral infections (71). In our study, we found a decreased number of CD4 T cell counts were associated with a decreased FA value and an increased MD value in patients without cART treatment, which was also seen in a prior study (72), indicating the close relationship between CD4 T cell counts and white matter alterations, although the underlying the mechanism is not completely understood. After early and regular cART treatment, there were no significant changes in CD4 T cell counts or DTI metrics, suggesting that absolute CD4 T cell counts may not accurately reflect the risks to HIV-infected individuals because immune dysfunction persists even when CD4 counts are back to normal (73). Furthermore, we found that increased MD was associated with decreased CD4/CD8 in many brain regions in macaques with and without regular cART treatment, which is consistent with prior studies in humans and macaques (29, 60). A low CD4/CD8 ratio is an immunologic risk phenotype associated with altered immune senescence, immune function, and chronic inflammation. After inoculation with SIV, a sharp increase in inflammatory cytokine production may indirectly lead to white matter damage, suggesting that CD4/CD8 in cART eara may be a better biomarker of disease progression, treatment response, morbidity, and mortality in virally suppressed individuals (74).

Animals infected with SIV-mac239 may be ideal models for studying HIV-induced, HIV-associated brain changes. In our model, the first white matter changes appeared in a structure adjacent to the periventricular area as early as 4 weeks after inoculation. The changes progressed over time, and cART treatment may have had different effects on each region, including reversing, alleviating, and even progressive effects. These changes were closely related to the CD4/CD8 ratio and viral load even after cART. Further studies are urgently needed to elucidate the underlying mechanisms.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The animal study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) at the Institute of Laboratory Animal Science, Chinese Academy of Medical Sciences (IACUC Approval No: LHJ18001).

Conceptualization: JL. Data collection: DL, YQ, XA, SH, YG. Methodology: BN. Data analysis: JL, BN. Supervision: HL. Writing—original draft: JL, BN. Writing—review & editing: XH, HL. All authors contributed to the article and approved the submitted version.