Human immunodeficiency Virus (HIV) infection continues to be a major global public health issue. More than 95% of all HIV-infected people now live in developing world with an estimated 2390000 people live in India as of 2009 (1). Involvement of central nervous system (CNS) in HIV-infected patients is commonly seen. Early involvement is characterized by impaired abstracting ability, learning difficulties and slow speed of information processing. Late involvement of CNS can be either neoplasms or opportunistic infections like Mycobacterium tuberculosis, Toxoplasmosis and Progressive multifocal leucoencephalopathy (PML). In the present study we investigated for the first time the effect of HIV infection status on brain metabolites in HIV patients with CNS involvement and asymptomatic HIV patients using magnetic resonance spectroscopy (MRS) at 3.0 Tesla.71 subjects were studied based on their HIV infection status i.e. seropositive HIV patients (n=21, age = 37.29 ± 10.60 yrs, male = 20, female = 1) with CNS involvement apparent on imaging (MRI, CT, PET), seropositive asymptomatic HIV patients (n=20, age = 36.95 ± 13.13 yrs, male = 11, female = 9) with no CNS involvement with CD4 count <500/µl and normal neurological, neuropsychological examination and healthy controls (n=30, age = 34.33 ± 12.49 yrs, male = 19, female = 11). Fasting blood glucose levels, routine complete blood count, lipid profile, liver function and kidney function tests were obtained for all the three groups. All subjects underwent MRI and MRS examination at 3.0 Tesla MR scanner (Philips Achieva/Ingenia, Philips Healthcare, Best, Netherlands). T1-weighted axial, T2 weighted axial, coronal and sagittal and FLAIR axial MRI were acquired in all subjects. Localized single voxel MRS was carried out using PRESS acquisition pulse sequence with following parameters: TR = 2000 ms, TE = 35 ms and number of scans = 128 and 16, with water suppression and without water suppression, respectively, for each selected voxel. MRS voxels to acquire spectrum were localized in three brain regions; left frontal lobe white matter (FWM), left basal ganglia caudate head nucleus (BG) and lesion apparent on MRI in case of patients. MRS data was processed using LCModel to determine metabolite concentrations.

Four types of lesion were identified on MRI in HIV patients with CNS involvement; 14 tuberculomas, 4 PML, 2 cryptococcal meningitis and 1 intracranial malignancy. CD4 count was significantly lower in HIV patients with CNS lesion (113.9 ± 70.9) compared to asymptomatic HIV patients (257.4 ± 129.0). Spectra from lesion (Fig.1C) showed lactate in 3 patients with tuberculomas (out of 14) and in 2 patients with PML (out of 4). The concentration of metabolites determined from MRS analysis (Fig. 1A and B) are summarized in Tables 1 and 2. A significant increase in concentration of creatine (Cr) in frontal region was observed in HIV patients with CNS involvement and asymptomatic HIV patients. Glx (Glu+Gln) was significantly increased in both groups of HIV patients in frontal region compared to controls. However, Cr and Glx concentrations were not significantly different in BG. The concentration of N-acetylaspartate (NAA) in basal ganglia region showed a significant decrease in HIV patients with CNS involvement compared to symptomatic HIV patients and healthy controls. However, there was no significant difference in NAA concentration in frontal region among three groups studied. Hemoglobin and serum calcium, sodium, albumin and globulin levels were significantly lower in HIV patients with CNS lesion and asymptomatic HIV patients compared to healthy controls.

In the present study we carried out absolute quantitation of brain metabolites at 3T in HIV patients with CNS involvement in a large cohort. The results showed regional variation in brain metabolite levels between HIV infected patients and healthy controls. Results of the study revealed increased levels of Glx in left FWM in HIV patients with CNS involvement and asymptomatic HIV patients compared to healthy controls. Most of the MRS studies on HIV patients have been carried out at 1.5T and analyzed metabolite ratios (2). There are few studies on measurement of Glx in HIV infected patients and these reported decrease in Glx levels in HIV patients (3). Glutamate excitotoxicity is one of the mechanism by which HIV exerts neurotoxicity resulting into neurocognitive disorders (4). Increased levels of Glutamate has been shown in CSF and plasma of HIV infected patients (5). Increased Cr in FWM and decreased NAA in BG was also observed which indicates regional variation in brain metabolism. Determination of concentration of metabolites in brain of HIV infected patients may be a useful indicator of metabolic changes related to neuronal loss affecting neurocognitive abilities.