Room to Grow: The Need for Cancer Site–Specific Research Into Biomarkers of Aging and Immunity in People With HIV

HIV and cancer have been inextricably linked since the beginning of the HIV epidemic in the United States. The unexpected clustering of Kaposi sarcoma (KS) diagnoses was one of the first pieces of evidence that a new epidemic was emerging (1). Immunosuppression resulting from HIV-mediated depletion of host CD4+ T cells impairs the ability of people with HIV (PWH) to control infections, resulting in higher rates of virus-associated cancers such as KS (2). The current cancer profile of PWH is also influenced by the fact that antiretroviral therapy (ART) has improved survival and dramatically changed the age structure of the HIV population, with more than one-fifth of PWH projected to be aged 65 years and older by 2030 (3). This longevity has translated into an increasing cancer burden that now includes age-related tumors not linked to infection. One of the important questions that follows from this changing epidemiology is whether biomarkers can be identified to predict which PWH will develop cancer.

A substantive body of research has been conducted to understand the degree to which variability in CD4+ T cell counts among PWH is associated with cancer risk (4,5). However, emerging research is evaluating additional metrics of immunity in relation to chronic diseases among PWH. This includes the ratio of CD4+ to CD8+ T cells. In this issue of the Journal, Castilho and colleagues (6) cite the association of the CD4/CD8 ratio with immune activation, immune senescence, and immune exhaustion as a strong rationale for evaluating this ratio in relation to cancer. They hypothesize that a lower CD4/CD8 ratio is linked to higher cancer risk among PWH, a hypothesis supported by data from the Veterans Aging Cohort Study (7) and the Swiss HIV Cohort Study (8), both of which indicated that a lower CD4/CD8 ratio, indicative of poorer immune function, was associated with lung cancer risk. However, those studies were able to include 277 and 68 lung cancers, respectively. This study by Castliho and colleagues (6) is able to include 5628 cancers occurring among 83 893 PWH from the North American AIDS Cohort Collaboration on Research and Design, of which 755 are lung cancers, an illustration of the expansive scope of this study.

Castilho and colleagues (6) use the observed distribution of the CD4/CD8 ratio in their study population to split individuals into approximate quartiles for interpretable presentation of findings, and the results are reported as the comparison of a CD4/CD8 ratio of 0.30 (lowest quartile) vs 0.80 (highest quartile). PWH with a CD4/CD8 ratio of 0.30 are approximately 25% more likely to develop cancer (hazard ratio [HR] = 1.24, 95% confidence interval [CI] = 1.14 to 1.35). Practically, this means that PWH with a CD4/CD8 ratio that is imbalanced (ie, further from 1.0) are more likely to be diagnosed with cancer. The authors leverage the large sample size to report site-specific effect estimates for 11 cancer types. A low CD4/CD8 ratio is statistically significantly associated with the risk of non-Hodgkin lymphoma (HR = 2.51, 95% CI = 1.82 to 3.47), KS (HR = 1.83, 95% CI = 1.26 to 2.64), and cancers of the lung (HR = 1.66, 95% CI = 1.32 to 2.10) and anus (HR = 2.33, 95% CI = 1.71 to 3.18). Site-specific cancer risk patterns differed when Castilho et al. considered the CD4/CD8 ratio at varying time points, ranging from 6 to 24 months, before cancer diagnosis, outlined in their Supplementary Figure 4 (6). In HIV oncology research, limited sample sizes often preclude a site-specific approach. Historically, results are often reported for either AIDS-defining cancers (KS, non-Hodgkin lymphoma, cervical cancer) or non–AIDS-defining cancers (all other cancers), but the differences by cancer site reported by Castilho and colleagues highlight why a site-specific approach is preferable.

However, it must be noted that this impressive research effort leaves some points open to further study. First, although Castilho and colleagues (6) accounted for detectable HIV viremia as a surrogate for ART adherence in their analyses, the impact of ART timing, and the specific HIV medications administered, deserve further research. Work published last year from North American AIDS Cohort Collaboration on Research and Design demonstrated lower cancer risk associated with earlier (<6-month lag) ART initiation after PLWH reached a CD4 count of 350-500 cells/µL (9). The impact of initiating ART at varying CD4/CD8 thresholds on future cancer risk is also of interest and could be used to guide patient management from the time of entering HIV care. Second, although the CD4/CD8 ratio is reported as the key metric in this study, this generally reflected elevated CD8 counts because CD4 was adjusted for in regression models; with CD4 held constant, higher CD8 led to lower CD4/CD8 ratios, as illustrated in Castilho and colleagues’ Supplementary Figure 5 (6). The authors supported their inclusion of CD4 and HIV viremia in regression models because of potential between-marker correlations, but separate investigation into CD4/CD8 ratio as a single biomarker in future studies will be key to understanding its predictive ability relative to other HIV measures. For example, a recently published North American AIDS Cohort Collaboration on Research and Design study directly contrasted the performance of CD4+ T cells vs HIV viremia in separate models using prediction metrics (eg, Akaike information criterion) to determine which were most accurate at identifying PWH who developed anal cancer (10).

The association between the elevated CD8+ T cell counts reflected in low CD4/CD8 ratios and cancer risk does raise the important issue of the link between aging and cancer in PWH. As cited by Castilho and colleagues (6), aberrant CD8 counts reflect immune exhaustion and senescence that are hallmarks of aging. Understanding how age-related biomarkers are associated with chronic disease outcomes is important because cancer is comprising an ever-larger proportion of mortality in PWH (11). Accumulating evidence (12) suggests that epigenetic aging, as measured through patterns of DNA methylation, is accelerated in PWH, but these aging biomarkers have not been explored alongside metrics of immunosuppression in the setting of HIV and cancer.

In summary, this study indicates elevated risk of cancer in PWH with an altered CD4/CD8 ratio. Cancer-screening rubrics tailored to PWH based on immune biomarkers do not currently exist, but these data should prompt such evaluations in the future. The work by Castilho and colleagues (6) also demonstrates that risk patterns are not uniform across cancer type. Perhaps the time has come to move beyond grouping together heterogeneous cancers to ensuring that our research examines the impact of biomarkers of aging and immunity on individual cancers, because we know that tumors differ broadly in both etiology and management.

Funding

No funding was used for this editorial.

Notes

Role of the funder: This is not applicable.

Disclosures: Authors report no conflicts of interest.

Author contributions: Conceptualization (AEC and BLD), Writing—original draft (AEC and BLD), Writing—Review & Editing (AEC), and Supervision (AEC).

Data Availability

Data sharing is not applicable to this editorial as no new data were created or analyzed.