Date of Award

January 2020

Document Type


Degree Name

Medical Doctor (MD)



First Advisor

Amy C. Justice


Chronic Human Immunodeficiency Virus (HIV) infection results in multimorbidity and shortened lifespan. Inflammation and immune dysfunction contribute to these outcomes. The aim of this thesis is (1) to investigate the utility of two inflammatory biomarkers evaluated in other settings (neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR)) for prognostication in patients living with HIV (PLWH) and comorbid inflammatory states; and (2) to compare the frequency and severity with which rheumatologic disease states occur in PLWH to demographically and behaviorally similar uninfected individuals (“controls”).

Three cohorts were created from the Veterans Aging Cohort Study. In each cohort, patients were stratified by presence or absence of HIV and (1) hepatitis C virus infection (HCV cohort); (2) incident sarcoidosis (SARC) and (3) incident rheumatoid arthritis (RA). Incident and prevalent sarcoidosis and RA were determined via chart review of patients meeting criteria including the presence of an appropriate diagnostic code.

NLR and PLR provided no discriminatory benefit when added to the VACS Index in the HCV cohort. In SARC, NLR and PLR were independently associated with development of sarcoidosis in controls, but not in PLWH. In RA, higher NLR was associated with risk for RA in PLWH while higher PLR was associated with risk of RA in controls.

Sarcoidosis and RA both occurred less frequently in PLWH (incidence rate ratio (IRR) for SARC in PLWH vs controls, 0.61 (95% CI, 0.46-0.81); IRR for RA, 0.29 (0.19-0.48)). At diagnosis, incident cases lacked critically low CD4 levels (SARC: median 409 cells/mm3, RA: median 487) and had suppressed viral loads (SARC: 77% <500 copies/mL, RA: 81%).

Incidence of sarcoidosis varied by control of HIV viremia over time; high levels of viral suppression were associated with higher rates of sarcoidosis. However, we found that sarcoidosis may have features of an immune reconstitution phenomenon: patients with lower CD4 counts at cohort entry were at higher risk of sarcoidosis than those with CD4 counts >500 at baseline (CD4 50-199, HR 3.2, p=0.02; CD4 200-349, HR=2.4, p=0.06).

Presentation and severity of sarcoidosis and RA were similar in PLWH and controls, except that autoantibody titers in PLWH were lower than in those without HIV (5% of PLWH, compared to 41% of controls, had two high-titer autoantibodies).

In both SARC and RA cohorts, non-glucocorticoid immunosuppressant therapy was prescribed less frequently to PLWH (SARC 5% vs 10%, RA 71% vs 94%). In RA, we did not find that PLWH receiving DMARD therapy were at increased risk for serious infections.

In summary, two novel inflammatory biomarkers, the NLR and PLR, did not add discriminatory power to an existing risk index. NLR and PLR were less strongly associated with most outcomes in PLWH than in persons without HIV. Both the presence and severity of HIV infection modulated the onset of rheumatologic conditions. The presentation of RA and sarcoidosis was mostly similar between PLWH and those without HIV. Although non-glucocorticoid immunosuppression was prescribed less frequently to PLWH in these real-world samples, we did not find evidence of increased rates of serious events in clinically select PLWH receiving DMARDs.


This thesis is restricted to Yale network users only. It will be made publicly available on 09/10/2022