Abstract Body

Background:

Antiretroviral therapy has translated into increased life expectancy for people living with HIV (PWH), resulting in a growing number of patients surviving to older ages when non-AIDS-defining cancer (NADC) are more common. Our prior work has demonstrated that patients with HIV have higher cancer-specific mortality compared to patients without HIV. A novel approach to identify factors that drive poor cancer outcomes in PWH is needed; one compelling hypothesis is that HIV influences the molecular profile of cancers that develop in PWH.

Methods:

Tissue microarrays (TMAs) from PWH diagnosed with either prostate or anal cancer were obtained from the AIDS Cancer Specimen Resource. Comparison TMAs were created from HIV-uninfected prostate and anal cancer patients selected from the biorepositories at Moffitt Cancer Center and Huntsman Cancer Institute. In addition, one TMA was created at Moffitt to include tumors from PWH diagnosed with a range of NADCs and matched tumors from HIV-uninfected patients diagnosed with the same cancer type. Slides were stained using the Akoya Biosciences’ OPALTM 7-Color Automation IHC kit on the BOND RX autostainer (Leica Biosystems). After staining, slides were imaged using the Vectra®3 Automated Quantitative Pathology Imaging System, and multi-layer images were exported to the HALO Image Analysis Platform. We compared marker positivity by HIV status, adjusted for tumor site and patient age and race, using beta-binomial regression models.

Results:

Multiplex immunofluorescence staining of tissues from 45 PWH (prostate cancer=22; anal cancer=14; other=9) and 238 HIV-uninfected patients (prostate cancer=216; anal cancer=5; other=17) demonstrated HIV-related differences in the tumor immune microenvironment. We stained for 15 different markers, and the abundance of six of these markers was significantly (P< 0.05) higher in tumors from PWH compared to tumors from patients without HIV, after adjustment for age, race, and tumor site. This included differences in infiltration of T-cells (CD8+ [OR=1.50] and delta-gamma [OR=1.81]), B-cells (CD20+ OR=1.67), and macrophages (CD163+ OR=1.98), as well as differences in expression of clinically targetable immune checkpoint molecules (PD-L1 [OR=3.83] and TIM3 [OR=1.76]). The T-regulatory cell phenotype (CD3+CD8+FOXP3+) was also statistically significantly more abundant in tumors from PWH (OR=2.19).

Conclusions:

Our data indicate that NADCs developing in the setting of HIV are immunologically distinct.