Development of Duffy Null–Specific Absolute Neutrophil Count Reference Ranges

Abstract

This study establishes a Duffy null phenotype–specific absolute neutrophil count reference range to optimize care and improve health equity.

Duffy antigens are glycoproteins on erythrocytes that bind chemokines.¹ A single nucleotide variant in the DARC/ACKR1 gene prevents Duffy expression and results in the Duffy null phenotype (Fy[a−b−]), which provides selective advantage against Plasmodium vivax.¹ This phenotype is reported in 66.7% of individuals identifying as Black in the US and is associated with adequate total body neutrophils with a clinically insignificant lower peripheral neutrophil count.¹ One study found that 23.8% of healthy people with the Duffy null phenotype have an absolute neutrophil count (ANC) below 2000/μL, the lower limit of normal of many reference ranges.² With 70% of medical decisions based on laboratory values,³ reference ranges that do not reflect normal values for a significant proportion of the population may lead to iatrogenic harm. For example, increased testing and referrals⁴ and discontinuation of critical medications like azathioprine have been found among individuals with Duffy null status and ANCs below conventional reference ranges.⁵ We aimed to establish a Duffy null phenotype–specific ANC reference range at our multi-institutional academic center.

Methods

Healthy adults at a Brigham and Women’s Hospital primary care center with self-reported Black or African American race who were scheduled for a nonurgent visit were identified through electronic medical records between January 27, 2021, and January 4, 2022. Exclusion criteria are shown in the Table. Blood samples were accrued until a preset goal of 120 acceptable Duffy null data points were obtained, based on Clinical Laboratory Standards Institute (CLSI) guidelines for establishing reference ranges.⁶ Complete blood counts with differential were performed on the Sysmex XN-9000 analyzer. Phenotyping for the Duffy antigens Fy^a and Fy^b was performed by tube testing using serologic reagents from Bio-Rad and Quotient, respectively. The CLSI guidelines and EP Evaluator version 12.0.0.11 (Data Innovations) were used to establish a 95% reference interval by a nonparametric percentile method. A 1-sample, 1-sided Wilcoxon signed-rank test was used to compare ANC reference median with the Duffy null cohort median with a preset significance level of .05 (Stata version 16.1; StataCorp). Local institutional review board approval was obtained as well as verbal patient consent to collect research samples.

Table. Exclusion Criteria for Patients Presenting for Nonurgent Care Visits at a Single Primary Care Site.

	Exclusion criteria
Autoimmune	Type 1 diabetes, rheumatoid arthritis, psoriasis, multiple sclerosis, systemic lupus erythematosus, inflammatory bowel disease, Addison disease, Graves disease, Hashimoto thyroiditis, myasthenia gravis, Sjögren syndrome, pernicious anemia, autoimmune vasculitis, celiac disease, autoimmune hepatitis, vitiligo, immune thrombocytopenia, dermatomyositis
Immunodeficiency	Any primary immunodeficiency including chronic granulomatous disease, common variable immunodeficiency, Chediak-Higashi syndrome, cyclic neutropenia, leukocyte adhesion defects, or congenital neutropenia; any individual with previous assessment for or diagnosis of neutropenia of any kind was also excluded
Infectious	Hepatitis C infection, hepatitis B infection, HIV infection
Malignancy	Any malignant neoplasm or hematological cancer including ductal carcinoma in situ, intraductal papillary mucinous neoplasm, or monoclonal gammopathy of undetermined significance
Transplant	Any history of receiving organ transplant or skin graft including history of allogeneic or autologous hematopoietic stem cell transplant
Medication	Carbimazole, clozapine, dapsone, dipyrone, methimazole, penicillin G, procainamide, propylthiouracil, rituximab, sulfasalazine, ticlopidine, hydroxychloroquine, infliximab, lamotrigine, oxacillin, quinine, infliximab, trimethoprim-sulfamethoxazole; any chemotherapeutic agent or biologic

Results

Of 176 blood samples collected, 120 were Duffy null. The distribution of ANC among individuals with Duffy null status and the overlying bell curve of the current ANC reference range is shown in the Figure. Median ANC among Duffy null samples was 2820/μL (IQR, 2090/μL-3480/μL; range, 1080/μL-5950/μL). The Duffy null reference median was significantly lower than the institution’s ANC reference median (2820/μL vs 4750/μL; P < .001). The standard institutional ANC reference range is 1920/μL to 7600/μL. The Duffy null nonparametric central 95% ANC interval was 1210/μL (95% CI, 1080/μL-1390/μL) to 5390/μL (95% CI, 4640/μL-5950/μL). The data were right skewed with a smaller CI on the lower limit of normal.

Figure. Duffy Null Absolute Neutrophil Count Reference Range Compared With Current Institutional Reference Range.

One hundred twenty Duffy null samples are represented in the histogram. The central 95% reference range for the Duffy null cohort is 1210/μL to 5390/μL. The overlying bell curve represents the current institutional central 95% reference range of 1900/μL to 7600/μL. There is a significant different between the Duffy reference median and the current reference median (2820/μL vs 4750/μL; P < .001).

Discussion

This study developed a new ANC reference range for individuals with Duffy null status that is lower than the institution’s standard ANC reference ranges. Based on these data, Brigham and Women’s Hospital as well as 5 associated clinic sites have opted to add a comment below the current ANC reference indicating that the expected ANC reference range is 1210/μL to 5390/μL for individuals with Duffy null status. Limitations of this study include using a convenience sample of individuals presenting to primary care and including only individuals who self-identified as Black or African American. Individuals who identified as another race and had the Duffy null phenotype may therefore have been excluded. Reference ranges that do not reflect the diverse populations served may cause iatrogenic harm such as overtesting, ineligibility for clinical trials, and even inappropriate dose reduction or cessation of essential medications.⁵ Use of this alternative reference range is expected to provide more personalized and equitable health care as well as emphasize the full range of normal values for all people. Other institutions should examine their ANC reference ranges and consider adding a Duffy null–specific reference range as a simple way to optimize care and improve health equity.

Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.

Supplement.

Data Sharing Statement