Clinical Implications of the Duffy-Null Genotype in Patients on Azathioprine or Methotrexate in the All of Us Cohort

BACKGROUND: The Duffy null variant (CC genotype) is associated with lower absolute neutrophil count (ANC) and increased risk of early discontinuation or ineligibility for certain treatments. Our primary objective was to evaluate the change in ANC over time among individuals on azathioprine or methotrexate. We also evaluated normality status (i.e., a designation of abnormal vs not) of ANC under genotype-adjusted and unadjusted thresholds. METHODS: We used the UK Biobank to calculate genotype-adjusted and unadjusted thresholds. We included participants reporting azathioprine or methotrexate use in All of Us to examine the primary and secondary objectives. The primary outcome was the log-mean difference in ANC between 1 year before and 2 years post medication initiation. The secondary outcome was the odds of normality of ANC under genotype-adjusted vs unadjusted thresholds. RESULTS: We found that in azathioprine users, those with the CC genotype had nonstatistically significant lower ANC (log-mean difference, -0.017; 95% CI, -0.072 to 0.038). For methotrexate, those with the CC genotype had a statistically significant lower ANC (log-mean difference, -0.042; 95% CI, -0.074 to -0.009). Evaluations of unadjusted ANC thresholds in those on azathioprine revealed that CC individuals were more likely to be abnormal (odds ratio [OR], 7.51; 95% CI, 1.64 to 34.43). Findings were similar for methotrexate (OR, 4.35; 95% CI, 1.48 to 12.79). However, under Duffy-adjusted thresholds, there were no significant differences between both genotypes in either treatment group: azathioprine (P = .491) and methotrexate (P = .269). CONCLUSION: We conclude that Duffy-adjusted ANC reference ranges reduce the proportion of individuals classified as abnormal, with several implications for Duffy-null individuals.