Both PF4-H-EIA and washed platelet activation assays have approximately equal sensitivity for clinical HIT (Greinacher et al., 1994a; Warkentin et al., 2000, 2005). For serum or plasma samples that are known to be positive by one sensitive washed platelet activation assay (e.g., SRA or HIPA), the corresponding probability of the PF4-H-EIA for confirming the positive result is at least 75-90% (Greinacher et al., 1994a; Arepally et al., 1995), with more recent studies suggesting >95% sensitivity of the EIA (Warkentin et al., 2005; Juhl et al., 2006). Conversely, a similar percentage of referred samples with high clinical probability of HIT that test positive in the EIA will also test positive using a washed platelet activation assay (Greinacher et al., 1994a; Lo et al., 2006; Schenk et al., 2006, 2007). The sensitivity of both EIA and SRA was even higher (>95%) for detecting antibodies that caused HIT in prospectively studied postoperative patients (Warkentin et al., 2000, 2005a).
Although both antigen and activation assays have similarly high sensitivity for clinical HIT, there is evidence that antigen assays have greater sensitivity for detecting HIT antibodies not associated with thrombocytopenia or other clinical events (Amiral et al., 1995; Arepally et al., 1995; Bauer et al., 1997; Warkentin et al., 2000, 2005a; Juhl et al., 2006; Schenk et al., 2006, 2007) (Fig. 6). Stated another way, the SRA is more specific for clinical HIT than the antigen assay. The biological explanation for greater specificity of a sensitive activation assay for clinical HIT, compared with an antigen assay, could relate to the functional heterogeneity of HIT antibodies against antigenic determinants on PF4, only some of which activate platelets strongly (Amiral et al., 2000). Data reported by Visentin and colleagues (1994) also support a higher sensitivity of antigen assays for detecting platelet-activating anti-PF4/H antibodies. These workers studied 12 HIT plasmas that tested positive in both SRA and PF4-heparin-EIA. However, at a 1:100 sample dilution, only 2 of the 12 samples still tested positive in the activation assay. In contrast, even at a 1:200 dilution, all 12 plasmas still tested positive in the EIA. Bachelot and colleagues (1998) observed that HIT plasmas that tested only weakly positive in the PF4-H-EIA tended to give negative washed platelet SRA results when using platelets with the least reactive FcyIIa receptor phenotype, Arg131.
The difference in sensitivity for HIT antibodies between the PF4-H-EIA and aggregation studies using c-PRP is considerable. Only about 33-64% of samples that test positive in the PF4-H-EIA also test positive using c-PRP aggregation (Greinacher et al., 1994a; Nguyen et al., 1995; Rugeri et al., 1999). Although one laboratory reported a greater sensitivity using c-PRP aggregation than the EIA (Look et al., 1997), these workers did not employ a two-point method, and so may have observed false-positive results using the aggregation assay.
Was this article helpful?