A prospective cohort study of light transmission platelet aggregometry for bleeding disorders: Is testing native platelet-rich plasma non-inferior to testing platelet count adjusted samples?

Abstract

Light transmission platelet aggregometry (LTA) is important to diagnose bleeding disorders. Experts recommend testing LTA with native (N) rather than platelet count adjusted (A) platelet-rich plasma (PRP), although it is unclear if this provides non-inferior, or superior, detection of bleeding disorders. Our goal was to determine if LTA with NPRP is non-inferior to LTA with APRP for bleeding disorder assessments. A prospective cohort of patients, referred for bleeding disorder testing, and healthy controls, were evaluated by LTA using common agonists, NPRP and APRP (adjusted to 250 x 10⁹ platelets/l). Recruitment continued until 40 controls and 40 patients with definite bleeding disorders were tested. Maximal aggregation (MA) data were assessed for the detection of abnormalities from bleeding disorders (all causes combined to limit bias), using sample-type specific reference intervals. Areas under receiver-operator curves (AUROC) were evaluated using pre-defined criteria (area differences: < 0.15 for non-inferiority, > 0 for superiority). Forty-four controls and 209 patients were evaluated. Chart reviews for 169 patients indicated 67 had bleeding disorders, 28 from inherited platelet secretion defects. Mean MA differences between NPRP and APRP were small for most agonists (ranges, controls: -3.3 to 5.8; patients: -3.0 to 13.7). With both samples, reduced MA with two or more agonists was associated with a bleeding disorder. AUROC differences between NPRP and APRP were small and indicated that NPRP were non-inferior to APRP for detecting bleeding disorders by LTA, whereas APRP met superiority criteria. Our study validates using either NPRP or APRP for LTA assessments of bleeding disorders.

Recommendations for the standardization of light transmission aggregometry: a consensus of the working party from the platelet physiology subcommittee of SSC/ISTH

“…Uncertainty remains over what is the best practice to follow when the platelet count in PRP exceeds about 600 9 10 9 L À1 . More recently, a study showed that both methods -native and adjusted platelet count -are valid to assess a bleeding disorder [37]. Abnormalities of platelet aggregation were more frequent using adjusted platelet count both in controls and patients [37].…”

Electron microscopy examination of platelet whole mount preparations to quantitate platelet dense granule numbers: Implications for diagnosing suspected platelet function disorders due to dense granule deficiency

“…While the finding of DGD by EM has been associated with a bleeding disorder, we decided to address the typical findings and within‐subject variability for the assay by examining 12 years of prospectively collected data on diagnostic EM tests for DGD. We postulated that this would generate useful evidence, like our studies of LTA and DG ATP release . To explore relationships between confirmed DGD, other laboratory findings, and bleeding, we also evaluated whole mount EM data for an overlapping cohort that had their bleeding history evaluated as part of a study on PFD.…”

Improving blood disorder diagnosis: reflections on the challenges

“…Our end-point was the diagnosis of a bleeding disorder by aggregometry, and the study was carried out by parallel testing of native and platelet count-adjusted samples from both patient and healthy controls [42]. This study, led by Dr. Jean Francois Castilloux, represents the first published noninferiority study of a diagnostic test performed in hematology laboratories [42].…”

“…Our end-point was the diagnosis of a bleeding disorder by aggregometry, and the study was carried out by parallel testing of native and platelet count-adjusted samples from both patient and healthy controls [42]. This study, led by Dr. Jean Francois Castilloux, represents the first published noninferiority study of a diagnostic test performed in hematology laboratories [42]. While the study validated that both sample types detect platelet function abnormalities due to bleeding disorders, platelet count-adjusted samples were superior for diagnosing a bleeding disorder [42].…”

“…This study, led by Dr. Jean Francois Castilloux, represents the first published noninferiority study of a diagnostic test performed in hematology laboratories [42]. While the study validated that both sample types detect platelet function abnormalities due to bleeding disorders, platelet count-adjusted samples were superior for diagnosing a bleeding disorder [42]. The study also confirmed, in a different cohort, that single-agonist abnormalities in LTA often represent false positives [42], an issue that many laboratories fail to recognize when interpreting the findings [29].…”