Portal de Periódicos da CAPES

The Platelet Function Analyser (PFA)-100 and von Willebrand disease: a story well over 16 years in the making

2015; Wiley; Volume: 21; Issue: 5 Linguagem: Inglês

10.1111/hae.12710

1365-2516

Emmanuel J. Favaloro,

Heparin-Induced Thrombocytopenia and Thrombosis

Many laboratories and clinicians struggle with the diagnosis of von Willebrand disease (VWD) 1, 2. VWD arises from deficiency and/or defects of von Willebrand factor (VWF), an adhesive plasma protein essential for effective primary haemostasis. VWF possesses many functional properties 3, which essentially explains the heterogeneity in clinical symptoms and bleeding risk in VWD-affected patients, as well as diagnostic challenges. Although a comprehensive panel of tests is required to diagnose VWD, and to assign a VWD type 1, 2, laboratory evaluation for the presence or absence of VWD (and other bleeding disorders) using simple screening tools has value in many situations, including urgent pre-operative testing. The Platelet Function Analyser (PFA)-100 (Siemens, Marburg, Germany) is a screening tool for primary haemostasis 4. The instrument is very easy to use, uses whole blood, and only requires 5 min per test, meaning that an assessment of a patient's primary haemostasis can be performed within less than 30 min from blood collection. The instrument measures occlusion of blood flow [so-called 'closure time' (CT)] in a shear flow environment after flowing blood is exposed to various agonists on a membrane in a cartridge device. There are currently three test cartridges in use: (i) one that contains collagen and epinephrine (C/Epi), is most sensitive to aspirin, haematocrit, platelet count, VWD and platelet dysfunction; (ii) another cartridge, which contains collagen and adenosine diphosphate (C/ADP), is relatively insensitive to aspirin, and less sensitive to VWD and platelet dysfunction; and (iii) another cartridge, which contains ADP and prostaglandin E1, is moderately sensitive to platelet P2Y12 receptor antagonists such as clopidogrel. The latter is only available in some geographical locations, and not applicable to assessment of primary haemostasis. In total, the results of the first two cartridges (C/Epi, C/ADP), being differentially sensitive to haematocrit, platelet count, anti-platelet medication, platelet function and VWF level and activity, have diagnostic implications for identification (or exclusion) of platelet defects and VWD. Indeed, sensitivity of the PFA-100 to VWD was identified 21 years ago in its first publication in 1995 5. Subsequent early publications compared the utility of the PFA-100 against the more invasive skin bleeding time (SBT), and most studies clearly identified that the PFA-100 had better sensitivity to VWD than the SBT, but perhaps less sensitivity to platelet disorders than the SBT 6-12. In a review published in 2008 4, I estimated a sensitivity of ~90% overall for VWD, with data on >700 patients, but notably, the sensitivity of the PFA-100 for severe forms of VWD was much higher. For type 3 VWD, reflecting an absence of VWF 13, sensitivity was 100%, meaning that no case of type 3 VWD was reported to yield a normal PFA-100 CT 4. Sensitivity for type 2A VWD (reflecting loss of the most functional forms of VWF, namely high molecular weight, VWF) 13 was also 100% 4. Sensitivity to other qualitative VWF defects or forms of VWD, namely 2B (reflecting hyper-adhesive VWF) 13 and 2M (reflecting dysfunctional VWF) 13, were possibly lower at around 97%, although it is also feasible that some of the published 'negative' findings instead reflected VWD 'misdiagnosis'. What cannot be argued is that the PFA-100 has reduced sensitivity to type 1 VWD 4, reflecting a quantitative deficiency of VWF 13. In these patients, the sensitivity of the PFA-100 to VWD is influenced by the level of VWF deficiency. Thus, overall sensitivity to type 1 VWD is around 80%, but sensitivity is closer to 100% for moderate to severe deficiencies (say <25 U dL−1 VWF). On the other hand, several workers have discounted the utility of the PFA-100. For example, Quiroga et al. 14 concluded both SBT and PFA-100 to be relatively insensitive to both VWD and platelet disorders, yielding respective sensitivities for the PFA of only 61.5% and 24% respectively. However, all VWD patients in this study were identified as 'type 1', and diagnostic and laboratory test data were incompletely reported. It is likely that many of these cases of 'VWD' only showed 'mild reduction' in VWF levels 15. Importantly, the current expert recommendation is to perhaps not label such patients (those with only mild reduction of VWF and with levels above ~30 U dL−1) as having VWD 16. Similarly, the platelet disorders identified by Quiroga et al. 14 were all platelet secretion defects, potentially considered among the 'mildest' of platelet disorders. There were a smattering of other papers published in the literature also questioning the clinical utility of the PFA-100 for identification of VWD. It is not my intention here to question the work of Quiroga et al. and others, but only to highlight that there appears to be a dichotomy of views – one 'camp' appears to applaud the PFA-100 and the other 'camp' laments its use. It almost seems as if we are talking about two different instruments – one 'good' and the other 'not so good'. No doubt that individual views depend on individual experiences, both with respect to the instrument used, as well as to the patient population investigated. Perhaps an analogy would be a particular make of car driven on particular roads; not all of us like the same car, and not all of us drive on the same roads. The PFA-100 celebrates its 21st birthday in 2015, with data on its use first appearing in press in 1995 5, 6. However, its history dates back a little further, having been developed from an instrument called the Thrombostat-4000. Interestingly, the first publication on the Thrombostat-4000 only appeared two years earlier, in 1993 17. However, the interest in the PFA-100 has far outweighed that of its predecessor. A recent Medline search of 'Thrombostat-4000' yielded only 23 papers; these were published in the period 1993–2001 (a total of 9 years). In contrast, the Medline search of 'PFA-100 OR PFA100' yielded 779 papers. While not all of these papers actually relate to this 'in vitro bleeding time device' [for example, the Medline search also captured phosphonoformic acid (PFA) 100 mg!], most do, and these were published in the period 1995–2015 (Fig. 1). As this represents a time span of 21 years, the PFA-100 has perhaps finally come of age, although the publication thrust does seem to be waning. It is therefore fitting that the current issue of Haemophilia includes a contribution 18 on the PFA-100 from a French group who first wrote of their initial experience with this instrument in 1998 7. The more recent publication 18 reports on 16 years of experience with the PFA-100 in the context of VWD. The authors retrospectively analysed the results (n = 6431) of 4027 patients referred to their centre between 1997 and mid-2013 in whom PFA-100 CT and VWF ristocetin cofactor (VWF:RCo) activity had been evaluated. Critically, in their experience, the PFA-100 was more effective in screening for VWF deficiency than the VWF:RCo! The negative predictive value (NPV), the positive predictive value, the sensitivity and the specificity of the PFA-100 for VWD were respectively 0.98, 0.51, 0.98 and 0.40. In other words, a normal PFA-100 CT was very effective (98%) in ruling out VWD, but a prolonged PFA-100 CT was not as useful for diagnosing VWD. The latter should not be considered a real barrier to use of the PFA-100 to screen for possible VWD. As the PFA-100 is a global test of primary haemostasis, it is (variably) sensitive to many things (as indicated before, namely: haematocrit, platelet count, anti-platelet medication, platelet function and VWF level and activity; viz. VWD). So, if a prolonged PFA-100 CT is identified, VWD is just one possibility. Indeed, in the real world, most cases of prolonged PFA-100 CT are due to low haematocrit, low platelet count, or anti-platelet medication such as aspirin. Accordingly, one should not expect any utility of the PFA-100 to rule in VWD. There are many other similar analogies in haemostasis. For example, the activated partial thromboplastin time (APTT) is sensitive to factor VIII deficiency, and thus will be abnormal in severe to moderate haemophilia A. However, the APTT is also sensitive to many other factor deficiencies (including fibrinogen, IX, XI and XII), as well as to lupus anticoagulant and also to many clinical therapy anticoagulants. Accordingly, a normal APTT may exclude a severe FVIII deficiency, but an abnormal APTT cannot be used to diagnose haemophilia A. Another analogous example is the D-dimer, where a negative D-dimer can be used (with ~98% confidence), to exclude a deep vein thrombosis (DVT); however, D-dimers are elevated in a wide variety of disease states (e.g., cancers, disseminated intravascular coagulation), as well as post surgery. Thus, a positive D-dimer has no role in 'diagnosing' a DVT. Finally, it can also be argued that the PFA-100 is similar to bleeding scores in this regard. A low bleeding score can help exclude VWD (or other bleeding disorder), but a high bleeding score cannot be used to diagnose VWD, as the diagnosis may instead be a platelet function disorder, and this diagnosis would require laboratory testing to identify. The findings reported by Ardillon et al. 18 require additional commentary. The finding that the PFA-100 CT seems better able to identify VWF deficiency (or more correctly to exclude VWF deficiency) than the VWF:RCo seems incongruous, as the former is a global assay of primary haemostasis, and the latter is a specific test of VWF activity. However, it also has to be remembered that the VWF:RCo assay is the most highly varied of the VWF assays, both in terms of methods in use (in-house vs. commercial reagents/methods, platelet aggregometry vs. automated instrument, different instruments in use) as well as in terms of reported test results 2. For example, using local external quality assessment (EQA) data, the inter-laboratory coefficient of variation (CV) for VWF:RCo averaged ~50% for samples tested in the last decade 2. In contrast, using EQA data from the same provider, inter-laboratory CVs for the PFA-100 seem to be lower, at around 20% 19. For the PFA-100, there is only the one instrument, and generally only two cartridges used for VWD testing. Although these CVs reflect inter-laboratory variation, and the data from Ardillon et al. 18 report on a single laboratory, one could propose that intra-laboratory variation, plus intra-individual variation, may also be either comparable, or perhaps even favourable for the PFA-100, depending on the method in use for the VWF:RCo. The findings reported by Ardillon et al. 18 actually closely align to our own experience, which we last reported in this journal in 2001 20. We have continued to gather prospective data on this, and plan to update our experience in press later this year. Overall, though, we would concur with the French group that in experienced hands, a reliable PFA-100 has very high NPV for exclusion of VWD (particularly with normal C/Epi). Our use of the instrument is a little different to that of the French group, and our experience also includes the supplementary use of the VWF collagen binding assay (VWF:CB) for VWD diagnosis and typing. An algorithm summarizing our current use of the PFA-100 is shown in Fig. 2. The PFA-100 is now 21 years old. Although in human terms this means it is now an adult, in instrument terms, where technology changes rapidly, this means it is relatively old, and so it has recently been 'upgraded' to the PFA-200. The PFA-200 uses essentially the same 'internal mechanics' as the PFA-100 but with (theoretically) 'enhanced' electronics (newer software, bigger screen, touch screen, etc.). However, whether the PFA-200 proves to be as effective as the PFA-100 requires separate validation. In conclusion, a good and reliable PFA instrument, in experienced hands, can be a useful screening tool of primary haemostasis. For VWD, normal PFA CTs can be used with a high degree of confidence to exclude VWD (akin to use of the D-dimer to exclude DVT). Although in the past we would have been confident to perform just a C/Epi CT for this purpose, occasional issues in 'reliability' have caused us to revise our practice to always also include a C/ADP CT. This acts in part to confirm the finding from the C/Epi, as well as providing additional information on possible defects and severity. The C/Epi is typically more sensitive to VWD than the C/ADP, and therefore VWD typically presents with prolonged C/Epi and prolonged ('moderate/severe' VWD) or normal ('mild' VWD) C/ADP. Normal C/Epi plus normal C/ADP is inconsistent with type 2A, 2B, 2M and 3 VWD, and unlikely for type 1 VWD with levels of VWF below 25–30 U dL−1. A prolonged C/ADP with normal C/Epi is unusual, and would normally lead us to repeat CT testing for confirmation. Also, like D-dimer testing for DVT, some pre-test clinical probability assessment is required to make best use of the PFA (Fig. 2). The author stated that he had no interests which might be perceived as posing a conflict or bias.