(P-IN-2) Automated blood group antibody titrations using gel card column agglutination: equivalent accuracy for ABO and non-ABO antibody titers between automated and manual testing

Antibody titration using two-fold serial dilutions of serum or plasma and a fixed reagent red blood cell (rRBC) is a semi-quantitative measurement of antibody concentration. Clinical applications include: 1) determination of obstetrical alloantibody levels to determine risk of HDFN, 2) identification of low-titer ABO plasma and platelets for minor ABO incompatible transfusions, 3) estimation of RBC antigen site density, and 4) monitoring antibody levels in ABO mismatched solid organ transplants. An automated 2-fold serial dilution method was developed for the Erytra Eflexis system using DG gel cards with a range up to a 1:2048 serial dilution.

Study

Design/Methods:

The aim of the study was to compare the Erytra Eflexis automated antibody titrations with titrations prepared manually and tested by the DG Gel manual method. A set of 185 plasma samples was selected containing ABO or non-ABO alloantibody (Rh, K, Kidd, Duffy, MNS). For titrations using DG Gel Coombs (Anti-IgG,-C3d) and DG Gel 8 Anti-IgG cards, approximately 50% of the titrations were ABO antibodies and 50% were non-ABO antibodies. For titrations using DG Gel 8 Neutral cards, approximately 90% of the titrations were ABO antibodies and 10% were non-ABO antibodies. The study compared each titre obtained with each DG gel card on the Erytra Eflexis with the equivalent titres obtained using the DG Gel manual method. The latter technique was performed as described in the respective IFUs of each DG Gel card (DG Gel Coombs, DG Gel 8 Anti-IgG, and DG Gel 8 Neutral). The endpoint was the last dilution to show hemagglutination, which also had to be within 1 grade of hemagglutination to be considered in agreement. The agreements were calculated according to two criteria: a difference of 1 two-fold serial dilution and of 2 two-fold serial dilutions.

Results/Findings:

The manual antibody titers ranged from 8 to 1024. The percent agreement and lower bound (95%) between automated and manual DG Gel Coombs (N = 60 determinations) was 96.7% (89.9%) when the maximum allowed difference was 1 two-fold serial dilution. DG Gel 8 IgG cards (N = 60) had a similar agreement at 95.0 % (87.6%). Similarly, DG Gel Neutral cards (N = 65) had a 100% (95.5%) agreement. When considering 2 two-fold dilution difference, there was 100% agreement for all DG cards with a 95% lower bound of 95.1%, 95.1%, and 95.5% respectively.

Conclusions:

A high level of agreement was observed for antibody titrations up to a titer of 2048 between the Erytra Eflexis and manual DG Gel method using DG Gel Coombs, DG Gel 8 Anti-IgG, and DG Gel 8 Neutral cards. The results of this study demonstrate that the automated antibody titrations are equivalent to the manual DG Gel method for both ABO and non-ABO antibodies.