Cannabis Use and Neuromelanin MRI Biomarkers of Dopaminergic Integrity in People With and Without HIV

Tara Najafi, BA
University of Miami

Co-authors: Tara Najafi¹, Suresh Pallikkuth¹, Lisa Reidy¹, Denise Vidot¹, Allan Rodriguez¹, Fernando M. Moradel¹, Eva Widerstrom-Noga¹, Varan Govind¹, Teddy Salan^1
1University of Miami

Background: Neuromelanin (NM), a byproduct of dopamine metabolism that accumulates in the substantia nigra (SN), can be noninvasively assessed using neuromelanin-sensitive MRI (NM-MRI), with SN-to-crus cerebri contrast ratio (CR) serving as a marker of nigral dopaminergic integrity. People with HIV (PWH) exhibit heightened dopaminergic vulnerability compared to people without HIV (PWoH), while cannabis (CB) use may initially enhance but chronically attenuate dopaminergic function. However, the combined effects of HIV infection and cannabis use on NM-MRI measures remain poorly understood.

Objectives: To examine the independent and interactive effects of HIV serostatus and cannabis use on NM-MRI signal (CR) in the SN using a 2×2 design (HIV+/HIV− × CB+/CB−), and to evaluate associations between CR and cannabis use severity (CUDIT score) and cannabinoid metabolites.

Methods: NM-MRI data were collected from four groups: HIV−CB− (n=14; 4M/10F; age=30.7±8.6), HIV+CB− (n=9; 2M/7F; age=38.8±3.8), HIV−CB+ (n=16; 9M/7F; age=36.0±8.5), and HIV+CB+ (n=17; 9M/8F; age=38.5±6.8). The SN and CC were segmented, and CR computed as (μSN−μCC)/μCC. A two-way factorial ANOVA evaluated main and interaction effects of HIV and CB status on CR, with a follow-up ANCOVA controlling for age and sex. Pearson correlations assessed CR against CUDIT score and cannabinoid metabolites within HIV+ and HIV− cannabis users; age-CR associations were also examined post hoc within HIV+CB− and HIV+CB+.

Results: Neither HIV status (p=0.31) nor cannabis use (p=0.889) was significantly associated with CR, with no HIV × CB interaction (p=0.6). Among cannabis users, the CUDIT-CR association was more positive in HIV+CB+ (r=0.23, p=0.38) than HIV−CB+ (r=−0.08, p=0.79). CR was not significantly correlated with cannabinoid metabolites (all p>0.15). After controlling for age and sex, group effects remained non-significant; however, age was significantly associated with CR (F=6.91, p=0.011, partial η²=0.12), increasing across all participants. Post hoc, this age-CR association was stronger in HIV+CB− (r=0.82, p=0.006) than HIV+CB+ (r=0.45, p=0.068).

Conclusions: Nigral NM-MRI signal did not differ significantly by HIV serostatus or cannabis use overall, nor was it associated with cannabinoid metabolites. Notably, CR increased with age across all participants, consistent with well-established age-related NM accumulation; this relationship appeared shallower among cannabis-using PWH, raising the possibility that cannabis use may attenuate HIV-related dopaminergic aging. The CUDIT-CR association was also more positive in HIV+ than HIV− cannabis users, possibly reflecting heightened dopaminergic vulnerability of the HIV brain to cannabis, consistent with reports of elevated neuromelanin signal in cannabis use disorder. Both trends are exploratory and warrant confirmation in an adequately powered sample.