Table 3.
Diagnostic performances of the functional tests for the diagnosis of HIT.
| Functional Assay | Study | Positivity Criteria of the Test | Patients Studied | Criteria for the Diagnosis of HIT | Diagnostic Performance (*) | Comparison |
|---|---|---|---|---|---|---|
| PAT | A. Greinacher et al. [59] | Platelet aggregation > 25% with LCH and no aggregation in the presence of the buffer (4 random donors) | 209 patients with suspicion of HIT | A positive reaction with platelets of 2 or more donors | Not applicable (NA) | HIPA performed with platelets from same donors. Poor agreement between HIPA and PAT. |
| B. Chong et al. [33] | Platelet aggregation > 25% with LCH | Thrombocytopenia due to other causes (n = 20) Non-thrombocytopenic patients who received heparin (n = 17) Patients with HIT (n = 17) Healthy donors (n = 23) |
Clinical diagnosis; n = 17 | Ss 39% with the least reactive donor Ss 81% with the most reactive donor Sp: 90% |
with SRA | |
| C. Pouplard et al. [60] | Platelet aggregation > 20% with LCH and with a sharp slope (5 random donors) | 100 patients with clinical suspicion of HIT | Clinical diagnosis; n = 40 | Ss: 91% Sp: 77% |
NA | |
| V. Galea et al. [62] | Maximal aggregation > 25% with LCH, no response in the presence of saline, and platelet aggregation inhibited with HCH | 200 consecutive patients with clinical suspicion of HIT | Clinical context and positive SRA; n = 21 | Ss: 76% Sp: 96% PPV: 80% NPV: 97% |
NA | |
| J. Brodard et al. [61] | Platelet aggregation > 50% with LCH and with two out of four selected platelet donors | 122 patients with clinical suspicion of HIT and positive anti-PF4/H ELISA | Clinical context and positive HIPA; n = 39 | Ss: 69% Sp: 100% |
NA | |
| Flow Cytometry Assay | A. Tomer et al. [63] | Annexin V binding. ≥6.6% platelet activation with LCH and inhibition with HCH Grey zone 6–6.6%. The number of platelet donors is not mentioned |
25 patients with clinical suspicion of HIT | Clinical diagnosis + and positive SRA; n = 19 | Ss: 95% Sp: 100% |
SRA |
| S. Poley et al. [64] | Annexin V binding. >13% platelet activation with LCH and inhibition of platelet activation with HCH. Pooled platelets of selected donors |
248 patients with clinical suspicion of HIT | Clinical diagnosis and positive HIPA; n = 17 | Ss: 95% Sp: 96% |
HIPA (4 donors) | |
| HS. Garritsen et al. [65] | Annexin V binding (HIT Alert®). ≥7.6% platelet activation in the presence of LCH and platelet activation reduced by ≥50% in the presence of HCH. One selected platelet donor |
346 patients with clinical suspicion of HIT | Clinical diagnosis; n = 17 | Ss: 88.2% Sp: 99.1% |
For IgG ELISA negative sera: 98% agreement with HIT Alert®. For IgG ELSA positive sera: 52.7% agreement with HIT Alert® |
|
| F. Mullier et al. [66] | Ratio PMP annexin V expression (LDH/HDH). One platelet donor only | 53 patients with clinical suspicion of HIT | Clinical diagnosis; n = 9 | Ss: 88.9% Sp: 100% |
NA | |
| E. Malicev et al. [67] | >10% CD62P-positive platelets at LCH and ≥50% and inhibition of platelet activation at HDH. Two platelet donors | 41 patients with clinical suspicion of HIT and positive ELISA IgG | Clinical context and positive HIPA; n = 14 | Ss: 82% Sp: 83% |
NA | |
| B. Tardy et al. [68] | P-selectin expression. >16.5% platelet activation with LCH and inhibition with HCH Two selected platelet donors |
228 patients with clinical suspicion | Expert opinion adjudication (clinical diagnosis + local laboratory results); n = 106 | Ss: 83% Sp: 97% |
NA | |
| M. Cipok et al. [69] | ≥2-fold greater P-selectin expression than that of the normal control. One platelet donor only | 63 patients with clinical suspicion | Positive SRA; n = 21 | Ss: 90.5% Sp: 95% |
NA | |
| K. Althaus et al. [70] | P-selectin expression (Emo-test HIT®). %HEPLA: >13.0%. Grey zone 9.6–13%. One unselected platelet donor only |
164 surgical or medical patients with clinical suspicion of HIT and positive EIA IgG | Positive HIPA; n = 33 | Ss: 69.7% Sp: 75.4% |
NA | |
| HIPA | A. Greinacher et al. [27] | HIPA was positive if the suspension became transparent with LCH, but not with heparin HCH (4 random donors) | 34 patients with suspicion of HIT | Not applicable | Not applicable | Excellent agreement with SRA: Kappa = 0.85 Moderate agreement with PAT: Kappa = 0.46 |
| P. Eichler et al. [71] | HIPA was positive if the suspension became transparent with LCH, but not with HDH. A sample was judged positive if positive results were obtained with test platelets of at least 2 of the 4 donors |
Workshop involving 9 laboratories with 8 samples: 2 from healthy blood donors, 5 from HIT patients (with HIT antibodies), 1 from a patient with sepsis | Not applicable | Not applicable | Expected results in 82% of cases | |
| HIMEA | M.C. Morel-Kopp et al. [34] | ISTH criteria | 181 patients with suspicion of HIT and positive EIA | Clinical context and positive SRA; n = 72 | Ss: 90.3% Sp: 89% |
HIMEA and SRA were performed with the same good responder donors |
| V. Galea et al. [62] | AUC with LCH > 267 AU with a representative shape of a platelet aggregation curve and a decrease in the AUC value with HCH > 50% | 200 consecutive patients with suspicion of HIT | Clinical context and positive SRA; n = 21 | Ss: 81% Sp: 99% NPV: 98% PPV: 89% |
NA | |
| V. Minet et al. [72] | Platelet aggregation occurred in the presence of LCH with a reduction of >80% with HCH | 116 patients with suspicion of HIT | 4Ts score and Accustar HIT; n = 2 | Ss: 100% Sp: 90% |
NA | |
| J. Jin et al. [73] | AUC > 50 with LCH and AUC = 0 or inhibition of at least 50% of the AUC obtained with HCH | 70 patients with suspicion of HIT | 4Ts score > 4 and positive EIA IgG and positive SRA; n = 7 | Ss: 85% Sp: 98% |
NA | |
| V. Galea et al. [74] | Aggregation curve at LCH was typical and AUC decreased by 50% or more with HDH | 87 patients with suspicion of HIT | Clinical context, positive SRA, and positive IgG ELISA; n = 12 | Ss: 91% Sp: 100% |
NA | |
| SRA | D. Sheridan et al. [52] | Release > 20% with LCH and < 20% with HCH. One donor | 28 patients with suspicion of HIT 573 non-HIT patients |
Clinical diagnosis: n = 6 | Ss: 100% Sp 99% |
NA |
| C. Pouplard et al. [60] | Release > 20% with LCH and < 20% with HCH. One donor | 100 patients with suspicion of HIT | Clinical diagnosis: n = 40 | Ss: 88% Sp: 100% |
NA | |
| B. Chong et al. [33] | Release > 20% with LCH and < 20% with HCH. One donor | Thrombocytopenia due to other causes (n = 20) | Clinical diagnosis: n = 17 | Ss: 65% with the least reactive donor Ss: 94% with the most reactive donor Sp: 90% |
Comparison with PAT, Kappa = 0.60 | |
| F. Mullier et al. [66] | Release > 20% with LCH and < 20% with HCH or less than 50% of that observed with LCH. | 53 patients with suspicion of HIT | Clinical diagnosis: n = 9 | Ss: 88.9% Sp: 95.5% |
NA |
Abbreviations: AU: arbitrary units; AUC: area under the curve; ELISA: enzyme linked immunosorbent assay; HCH: high concentration heparin; HIPA: heparin-induced platelet aggregation; ISTH: International Society on Thrombosis and Haemostasis; LCH: low concentration heparin; NA: not applicable; NPV: negative predictive value; PPV: positive predictive value, Sp: specificity; Ss: sensitivity; SRA: serotonin release assay, % HEPLA: heparin platelet activation index; (*) Proper determination of the diagnostic performance is crucial, but difficult (see text). Moreover, the number of patients being low, 95% confidence interval is large.