Polygenic Variation Underlying Neutrophil Counts Modifies the Penetrance of Duffy‐Null Neutropenia

The Duffy-null phenotype is caused by a genotype that is common among individuals of African ancestry and leads to benign reductions in absolute neutrophil counts (ANC) [1]. Because clinical reference ranges are typically not calibrated to the normal ANC distribution observed in those with the Duffy-null genotype, individuals with this genotype are more likely to fall below the reference range in the absence of disease with a 4- to 14-fold increased risk of neutropenia (ANC < 1.5 × 103/μL) [2-4]. This misalignment predisposes these individuals to undergo unnecessary diagnostic procedures or experience treatment modifications due to provider concerns about a lower ANC [5]. This has prompted efforts to construct clinical reference intervals for ANC for those with the Duffy-null genotype [3]. Even among individuals with the Duffy-null genotype, ANC levels vary widely, and 83%–95% do not meet criteria for neutropenia in cross-sectional studies [2]. Recent studies have demonstrated that polygenic scores—quantitative measures of an individual's burden of common genetic variants associated with a trait—can capture benign variation in white blood cell (WBC) counts and thus, measure a person's polygenic predisposition to benign leukopenia [6]. We have shown that patients with an increased polygenic predisposition to leukopenia are more likely to undergo the same diagnostic escalations that have been associated with Duffy-null associated neutropenia [6]. However, it remains unclear whether polygenic variation underlying ANC modifies the penetrance of neutropenia among individuals with the Duffy-null genotype, or whether such scores can enable more precise reference intervals for ANC. In this study, we evaluated whether a polygenic predisposition to lower ANC modifies the risk of neutropenia and influences reference intervals among individuals with the Duffy-null genotype. By leveraging a large cohort of African ancestry individuals with whole genome sequencing and longitudinal clinical data, we aimed to quantify the contribution of polygenic background to ANC variation and assess its potential utility in refining diagnostic thresholds. Our findings have implications for improving the precision of hematologic care and reducing disparities in clinical decision-making. Among 2498 adults of African ancestry (65% female, median age 44 years), 1581 (63.2%) participants carried the Duffy-null genotype (Table 1, Figure S1). At baseline, 126 (8.0%) of Duffy-null participants had neutropenia (ANC < 1.5 × 103/μL), compared to 1 (0.1%) participant without the genotype. To characterize other genetic causes of benign ANC variation, we tested the association between a polygenic score for ANC (PGSANC) and clinical outcomes in Duffy-null individuals. For each standard deviation increase in PGSANC, ANC increased by a modest 0.12 × 103/μL (95% CI: 0.06–0.19, p = 3.0 × 10−4) (Figure S2). However, lower PGSANC was associated with a 26% increased risk of prevalent neutropenia (odds ratio per SD decrease in PGSANC, 0.74 [0.55–0.93]; p = 0.002). Among the 1581 participants with the Duffy-null genotype, 1035 (65.4%) had a repeat ANC measure within 5 years of their initial visit. Within this longitudinal cohort (n = 1035), there were 217 incident neutropenia events over the follow-up period, of which 88 (40.5%) were among individuals who also had neutropenia at baseline (Table S1). The risk of developing neutropenia was significantly influenced by polygenic background, with a 23% increased risk of incident neutropenia per SD decrease in PGSANC (hazard ratio 0.77 [0.60–0.94]; p = 0.002) (Figure 1A). Stratifying Duffy-null individuals by PGSANC revealed that the lower (2.5%) reference limit of ANC at baseline varied by polygenic risk. Specifically, these limits were 0.99 (0.91–1.06), 1.15 (1.06–1.22), and 1.37 (1.24–1.55) × 103/μL for the low, middle, and high PGSANC groups, respectively (Figure 1B). Notably, the lower reference limit for those with an average PGSANC (1.15 × 103/μL) was similar to a previous study of Duffy null individuals that predicted a reference limit of 1.21 × 103/μL [3]. Individuals with lower polygenic scores (i.e., stronger predisposition to neutropenia) had a reference threshold below the current cut-off for moderate neutropenia (1.0 × 103/μL). In this cohort of adults, we show that common genetic variants outside of the Duffy-null genotype contribute to variation in ANC and modify the risk of Duffy-null associated neutropenia. This study expands on the growing body of literature demonstrating the substantial heritability of white blood cell traits, particularly neutrophil counts [6, 7]. While the Duffy-null genotype remains the dominant determinant of reduced ANC, fluctuations in ANC and penetrance of neutropenia are further influenced by polygenic background. Importantly, we derived the polygenic score using genetic variants associated with neutrophil count but not neutropenia-associated diseases, ensuring that our score captures benign genetic influences rather than pathological ones. This strengthens the clinical relevance of our findings, as it allows us to isolate the contribution of inherited, non-disease-associated variation to ANC levels. Furthermore, the longitudinal nature of our dataset enabled us to observe fluctuations in ANC over time, revealing that many individuals with the Duffy-null genotype intermittently cross diagnostic thresholds. Given the known consequences of a neutropenia diagnosis—including unnecessary clinical interventions like bone marrow biopsies or premature chemotherapy discontinuation—these results underscore the need to refine ANC thresholds in a genetically informed manner [5, 8, 9]. Conceptualization and methodology: J.P.S., M.S., J.D.M. Data curation: J.P.S., J.D.M., M.S. Formal analysis: J.P.S., M.S. Software: J.P.S., J.D.M., M.S. Supervision: J.D.M. Writing-original draft: J.P.S. Writing – review and editing: J.P.S., M.S., C.P.C., L.B., J.D.M. This work was supported by the National Institutes of Health: R01GM130791 (J.D.M.), U01HG011181 (J.D.M.), U01HG013771 (J.D.M.), U01HG011166 (J.D.M.), R01GM126535 (C.P.C.), and T32GM007347 (J.P.S.). Vanderbilt University Medical Center's BioVU projects are supported by numerous sources: institutional funding, private agencies, and federal grants. These include NIH funded Shared Instrumentation Grant S10OD017985, S10RR025141, and S10OD025092; CTSA grants UL1TR002243, UL1TR000445, and UL1RR024975. Genomic data are also supported by investigator-led projects that include U01HG004798, R01NS032830, RC2GM092618, P50GM115305, U01HG006378, U19HL065962, R01HD074711; and the following additional funding sources: K07CA172294, 14GRNT20460090, P01DK038226, R24DK96527, R01LM010685, R01EY012118, K12HD043483, R01DK078616, RC2GM092618, APP1064524, R01CA162433, P01HL056693, P50GM115305, U01HG006378, U19HL065962, U01HG004603, P50CA09813, R01HD074711, R03HD078567, R01DK080007, P50HL081009. The sequencing of 250,000 individuals from BioVU, including the 35,022 described here, has been funded by the Alliance for Genomic Discovery which consists of NashBio, Illumina and industry partners Amgen, AbbVie, AstraZeneca, Bayer, BMS, GSK, Merck, and Novo. DNA sequencing was performed at deCODE genetics using Illumina's NovaSeq 6000. This study was evaluated by the Vanderbilt Institutional Review Board (IRB #180267) and determined to be non-human subjects research as data are fully de-identified. All research conformed to the principles of the Helsinki Declaration. The authors declare no conflicts of interest. Subject-level access to BioVU clinical and genetic data is controlled by the BioVU data repository (https://victr.vumc.org/biovu-description/). Upon publication, data sets of subject-level phenotype data and corresponding data dictionaries to replicate the primary findings presented here for research purposes will be made available upon request from the repository ([email protected]). Vetting for use of BioVU data includes institutional IRB approval, data use agreements, and administrative and scientific reviews. Figure S1: Identification of healthy participants with a CBC in a primary care setting. Figure S2: Distribution of absolute neutrophil counts by genotype. Table S1: Descriptive statistics for the longitudinal cohort of Duffy-null participants with > 1 labs within 5 years of the initial encounter. Table S2: ICD codes used to identify participants with acute or chronic conditions affecting neutrophil counts. Table S3: Keywords used to identify medications affecting neutrophil counts. Table S4: Phenotypes from GWAS catalog used to identify and exclude pleiotropic SNPs. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.