Polygenic Risk Scores Reveal Genetic Drivers of Lupus Heterogeneity, With Matthew Dapas, PhD

At the American College of Rheumatology (ACR) Convergence 2025 in Chicago, Matthew Dapas, PhD, research assistant professor at Northwestern University Feinberg School of Medicine, discussed how polygenic risk scores are helping to clarify the heterogeneity of systemic lupus erythematosus (SLE) and inform clinical understanding of disease manifestations.1

Systemic lupus erythematosus (SLE) presents a wide range of clinical manifestations, complicating diagnosis. Mayo Clinic notes symptoms such as fatigue, fever, joint pain, a butterfly-shaped facial rash, sun-sensitive skin lesions, color changes in fingers and toes, shortness of breath, chest pain, dry eyes, headaches, confusion, and memory loss.2 Clinicians use classification criteria, like those from the Systemic Lupus International Collaborating Clinics (SLICC), to guide diagnosis and management.1

“[Lupus is] a very heterogeneous disease,” Dapas told HCPLive during ACR. “[Lupus in] one individual …might look pretty different from another individual that has lupus. So how do we engineer these diagnosis classification criteria to best fit this wide variety of manifestations that can come about from this one disease? With my work, we build these genetic risk models to…see to what extent these different diagnostic criteria or classification criteria [capture] different types of disease…We can use these genetic models and their associations as a tool to basically increase our ability to pinpoint the exact disease in an individual. It's part of this push towards personalized medicine.”

Using data from a genome-wide association study of 6,748 SLE cases and 11,516 controls of European ancestry, Dapas and colleagues developed a polygenic risk score that predicted SLE with an area under the curve of 0.724.1 Within the Northwestern cohort (n = 1,061), the polygenic risk score showed strong associations with lupus diagnoses (“Lupus [localized and systemic],” P = 1.6×10⁻⁹; “systemic lupus erythematosus,” P = 3.0×10⁻⁹), and nominal associations with cardiac complications (P = 2.8×10⁻³) and impaired renal function (P = 4.6×10⁻³), suggesting genetic risk may drive distinct phenotypic patterns.

The team then analyzed the correlation between the polygenic risk score and individual SLICC classification criteria in > 32,000 eMERGE participants. The polygenic risk score was significantly associated with 5 of 6 immunologic criteria and 3 of 11 clinical criteria, including acute cutaneous lupus, thrombocytopenia, and leukopenia. The strongest association was with low complement levels (P = 7.4×10⁻³¹), where patients in the highest polygenic risk score decile were nearly twice as likely to have low complement as those in the lowest decile (55% vs. 30%).1

These findings suggest that immunologic features more closely reflect underlying genetic risk, while some clinical manifestations may be influenced by environmental factors or represent phenotypes not captured in the original genome-wide association study.1

“We want to get to a place where, if a clinician [is] seeing a new patient that they suspect may have an autoimmune disease like lupus, they can order a genetic panel in addition to the other labs that they might order to try and assess this patient. This could go for newly diagnosed lupus patients as well. If you have a genetic risk panel, that genetic risk could be specific to certain complications or classification, or just lupus itself. You can include particular HLA alleles that can essentially inform the physician with greater certainty what the individual risk profile is for that patient.

He said that a patient with a greater genetic risk score for lupus with certain HLA alleles may be at a greater risk for particular outcomes from lupus that may warrant more observation or the order of a specific test.

“Ultimately, we're trying to use genetics as an additional tool in the physician’s toolkit to better tailor their treatment to their individual patients and better understand the disease of the individual patients, because, like I said from the outset, conditions like lupus [are] very heterogeneous, and it's not a one-size-fits-all solution for everyone.”

Dapas has no relevant reported disclosures.

References

1. Forrest N, Tuteja S, Pacheco J, Mitrovic V, Ramsey-Goldman R, Walunas T, Dapas M. Polygenic risk of lupus is differentially associated with individual EHR-derived classification criteria [abstract]. Arthritis Rheumatol. 2025; 77 (suppl 9). https://acrabstracts.org/abstract/polygenic-risk-of-lupus-is-differentially-associated-with-individual-ehr-derived-classification-criteria/. Accessed November 24, 2025.

2. Mayo clinic. Lupus - Symptoms and Causes. Mayo Clinic. Published October 21, 2022. https://www.mayoclinic.org/diseases-conditions/lupus/symptoms-causes/syc-20365789