The Use of Chromogenic Factor VIII Assay Changes Treatment Approach in a Portion of Mild Hemophilia A Patients with Factor VIII Assay Discrepancy

Introduction Hemophilia A is an X-linked inherited bleeding disorder characterized by a deficiency of coagulation factor VIII (FVIII). FVIII assays used for diagnosis and classification of hemophilia are the widely used one-stage assay (OSA) and the chromogenic assay (CSA), a more costly and less commonly used test. The CSA has the advantages of improved precision and insensitivity to pre-activation effects of FVIII that can lead to erroneous results. In approximately one-third of mild hemophilia A cases, reliance on OSA can lead to overestimation of factor level in comparison to CSA, and this 'discrepancy' phenomenon is closely associated with the FVIII molecular abnormality. In mild or moderate hemophilia A, assay discrepancy can lead to misclassification or inappropriate treatment. The objectives of this retrospective study were to determine the proportion of patients with OSA/CSA discrepancies defined by OSA to CSA ratio of >1.5 or <0.5, to identify previously reported and new F8 gene mutations and to classify the impact of assay discrepancy on clinical treatment approach. Also, the study aimed to re-evaluate published discrepant OSA/CSA ratios to identify a ratio with the highest sensitivity to indicate that a change of treatment approach or intensity may be required. Methods We reviewed the charts of adult (>18 years old) patients with mild or moderate hemophilia A followed at the Adult British Columbia Hemophilia clinic with concomitant OSA and CSA results between January 2013 and March 2019. Data collected included patient age, disease severity classification, baseline OSA FVIII:C, and F8 gene mutation. Bleeding phenotype and impact of discrepancy on treatment approach was evaluated and determined independently by two individuals and was correlated with concomitant OSA and CSA. Gene mutation data was collected to determine whether OSA/CSA discrepancy has been previously reported in the international hemophilia A mutation database (www.factorviii-db.org). Descriptive data was reported as medians and ranges. Statistical analyses using Spearman's correlation, logistic regression and receiver operating characteristics (ROC) curve were performed using IBM SPSS v. 22.0. Results 98 patients were included in the study with median age of 53 years (18-88). 75 patients were classified by OSA at baseline as mild and 23 patients as moderate severity. Median FVIII:C by OSA was 15% (6%-40%) in the mild group and 3% (1%-5%) in the moderate severity group. Median FVIII:C using CSA was 13% (2%-66%) in the mild group and 4% (1%-7%) in the moderate group. Based on ratio alone, OSA/CSA discrepancy was detected in 51 (52%) patients out of which 3 (3%) were reverse discrepancies (CSA higher than OSA). Treatment approach was changed as a result of CSA introduction in 14 (27%) patients with OSA/CSA discrepancy (14% of the entire study population). OSA had overall good correlation with CSA (r = 0.84) and as expected, there was no correlation between age and OSA to CSA ratio (r = 0.03). In the patients with OSA/CSA discrepancies, neither age nor OSA/CSA predicted a change of treatment approach (OR 1.02 for age; 95% CI 0.99-1.06; p value 0.213 and OR 1.48 for OSA/CSA; 95% CI 0.96-2.28; p value 0.08). The OSA to CSA ratio of 1.8 to 3.5 demonstrated the highest area under the ROC curve and sensitivity for identification of requirement for treatment change in patients with OSA/CSA discrepancy (AUC 0.75; sensitivity 71%, negative predictive value 75%). Six F8 gene mutations with OSA/CSA discrepancies previously not reported in the database were identified. Conclusions OSA/CSA discrepancies were detected in 52% of the patients with mild or moderate hemophilia A and resulted in change of treatment approach in 27% of discrepant patients. Use of an adjusted OSA/CSA ratio between 1.8 to 3.5 was more likely to detect cases with a requirement for change of treatment approach and use of this ratio to indicate discrepancy may be more clinically meaningful. Six gene mutations associated with OSA/CSA discrepancies previously not described in the international hemophilia A mutation database were identified in the study population, highlighting the need for genotyping discrepant cases. Disclosures No relevant conflicts of interest to declare.