Relationship between HIV and immune checkpoint molecules

The therapeutic landscape for malignancies is rapidly evolving with the advent of immune checkpoint inhibitors (ICI), most notably programmed-cell death (ligand)-1 (PD-(L)1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) inhibitors. Taking into account that one of the hallmarks of cancer is its innate ability to evade the immune system, ICIs may hold transformative potential owing to their ability to block the suppressive immune signals produced by tumor cells. These agents have impressive clinical activity in a broad array of both solid and hematologic malignancies, including patients with advanced, refractory disease. To date, seven checkpoint inhibitors have been approved by the Food and Drug Administration (FDA).

Despite the encouraging results with ICIs in multiple cancer types, there is a paucity of data regarding the use of these agents in patients with HIV-associated malignancies because these patients are often excluded from clinical trials. Yet, patients living with HIV have a significantly higher incidence of cancer including non-Hodgkin lymphoma (~ 21% of cancers in HIV-infected people), Kaposi sarcoma (~ 12%), lung cancer (~ 11%), anal/cervical cancer (~ 10%), as well as other tumor types such as colorectal (~ 5%), oral/pharyngeal (4%), and others (NCCN guidelines version 2.2019 AIDS related Kaposi sarcoma (https://www.nccn.org/professionals/physician_gls/pdf/kaposi.pdf)(https://www.cancer.gov/about-cancer/causes-prevention/risk/infectious-agents/hiv-fact-sheet).

Importantly, patients with rampant HIV infections – high viral loads in the absence of being on antiretroviral therapy (ART) – have more expression of CTLA-4 on their CD4+ T cells when compared to those of healthy controls (Table 1). Additionally, CTLA-4 levels are inversely related to total CD4+ T cell population and directly related to HIV viral load and cancer progression [11]. This same patient population also expresses higher levels of PD-1 on their CD4+ T cells when compared to those of healthy controls and this is associated with T cell exhaustion; further, similar to CTLA-4, PD-1 levels are also related to HIV viral load and cancer progression [9, 10]. The dual changes in checkpoint cell surface molecules in patients who are infected with HIV and have cancer could be exploited in their treatment as is being explored in a clinical trial of ipilimumab with nivolumab in HIV-associated solid tumors and lymphoma (NCT02408861) as well as with pembrolizumab monotherapy in HIV and various cancers (NCT02595866) (clinicaltrials.gov). HIV evades the immune response by promoting a state of immune exhaustion, which is similar to the mechanism of how cancers with upregulated PD-L1/PD-1 axis and/or CTLA-4 expression elude immune eradication [9, 10, 13]. Therefore, in theory, ICIs may be beneficial to both the HIV infection and to cancer.

Virus-induced cancers: immune and mutational landscape and Neoantigen immunogenicity

Mechanistically, not only do viral infections lead to an increase in expression of the checkpoint cell surface molecules CTLA-4 and PD-1, they also subvert the DNA damage response within the host cell DNA. Indeed, DNA viruses (EBV, HHV-8, HPV, etc.) thrive by inserting their viral genome into the genome of the host cell and subsequently hijacking the host cell replicative enzymes [12]. The host DNA damage response that should be activated in response to the replicating viral DNA or the virally-stimulated cellular transition from a quiescent to mitotic/cell cycle state, induced by viruses in order to facilitate replication, is attenuated by specific proteins expressed by the DNA tumor viruses. Virally-induced cancers may also have distinct mutational portfolios and metabolic patterns that can impact immune response and prognosis. For instance, HPV-associated squamous tumors of the head and neck may harbor more PIK3CA alterations whereas non-HPV tumors may have TP53 and cyclin pathway (CDKN2A and CCND1) alterations. Finally, molecular editing mechanisms mediated by apolipoprotein B mRNA editing enzyme catalytic polypeptide-like (APOBEC), a family of evolutionarily conserved cytidine deaminases involved in DNA and mRNA editing that are upregulated with viral infection leading to inactivation of viral genomes, may be relevant. These upregulated enzymes constitute a crucial part of mammalian innate immunity and are also a major source of mutations in multiple cancer types. Relevantly, APOBEC-related mutagenesis increases neoantigen hydrophobicity, a key feature of immunogenicity [7]. Cancers with upregulated APOBEC show high levels of PD-L1 expression that presumably enable the cancer to evade the immune system and survive in light of the immunogenic mutation-related peptides induced by APOBEC [8]. Hence, not surprisingly, PD-1/PD-L1 ICIs are associated with high response rates in human cancers that bear APOBEC mutational gene expression patterns [7, 8]. The production of mutations that result in immunogenic neoantigens or presentation of the viral antigens themselves may also explain the fact that malignancies such as virus-related Merkel cell carcinoma and Kaposi sarcoma respond well to ICIs, despite a low tumor mutational burden, the latter usually being associated with a poor response to these immunotherapeutics [4, 14, 15].

Activity of immune checkpoint inhibitors among HIV-positive patients with cancers

There have been only a small number of case series on the use of ICIs in HIV-positive patients. A recent review that included all studies published (73 HIV-patients suffering from several primary tumors treated with either PD-1, CTLA-4, or both inhibitors) found a response rate of 67% for Kaposi sarcoma, 30% for non–small cell lung cancer, and 27% for melanoma; in addition, activity including complete responses was seen in Merkel cell carcinoma and in Hodgkin lymphoma [4, 16]. Importantly, only 9% of individuals reported greater than or equal to grade 3 immune-related toxicities, most of which occurred in patients who received ipilimumab as part of their regimen [16]. HIV remained suppressed in 93% of patients with available data and undetectable viral load and, overall, CD4+ counts increased [4, 16]. The low numbers of patients in the dataset reflects the practice of exclusion of HIV-positive patients from the majority of studies. However, the activity of ICIs in these reports and their lack of toxicity suggest that additional trials are needed.