Towards 50 years of platelet function analyser (PFA) testing

Introduction

The platelet function analyser (PFA) is a widely popular platelet function screening instrument, and comes in two models–the original PFA-100 and the contemporary PFA-200. The PFA-100 was released into the marketplace in the mid 1990s [1], [2], [3], and so is approaching 50 years of age. The PFA-200, released in some locations, including Australia and Europe, in the mid 2010s, is destined to eventually replace the PFA-100, but is not yet available in some locations, notably in the USA. The PFA-100 was originally viewed as a replacement for the then ‘popular’ skin bleeding time, and was known for a while as the ‘in vitro bleeding time’ [4], [5], [6].

The two instruments are treated as ‘identical’ in terms of test results, being a ‘closure time’ (CT) [7]. This CT is the time in seconds (s) taken for occlusion of an aperture in the test cartridge, after whole blood anticoagulated with buffered sodium citrate comes into contact with platelet aggregating agonists coated on a membrane in the cartridge. There are three different types of test cartridge, the two main ones: collagen/epinephrine (C/Epi) and collagen/adenosine diphosphate (C/ADP), and a third “Innovance PFA P2Y” only available in certain geographical locations (again, excluding the USA).

That the two instrument outputs are treated as ‘identical’ for both PFA-100/200 by the manufacturer (Siemens Healthineers) is clearly identifiable since the normal reference ranges (NRR) provided by the manufacturer with the different test cartridges are the same for both models. The current manufacturer-cited NRR for both the PFA-100 and PFA-200, and for the C/Epi and C/ADP test cartridges are respectively 82–150 s and 62–100 s, using standard 3.2% (0.105 M) sodium citrate blood collection tubes (Table 1).

The PFA-100/200 test system, using C/Epi and C/ADP cartridges, is highly sensitive to von Willebrand disease (VWD) and to aspirin therapy (C/Epi only), but has only moderate sensitivity to defects in platelet function and/or deficiencies in platelet number [7], [8], [9], [10], [11], [12]. Accordingly, recommendations for their use as a platelet function screening instrument vary according to user experience. For example, the Platelet Physiology Subcommittee of the Scientific and Standardization Committee (SSC) of the International Society on Thrombosis and Haemostasis (ISTH) reflected in 2006 that “the test does not have sufficient sensitivity or specificity to be used as a screening tool for platelet disorders” [12]. We essentially agree with this view, and although the C/EPI and C/ADP CTs have utility for exclusion of severe platelet dysfunction, they have insufficient sensitivity to exclude mild platelet function disorders, which account for the majority of such disorders, thereby limiting the utility of both PFA-100 and PFA-200. Moreover, the P2Y cartridge, whilst marketed by the manufacturer as sensitive to clopidogrel, has in our experience poor utility for same. However, the PFA100/200 is very sensitive to von Willebrand factor (VWF) level and activity, and thus has a role in screening for von Willebrand disease (VWD) [7], [8], [9], [10], [11]. Some of us have also shown the potential utility of these instruments in monitoring certain therapies in VWD, in particular desmopressin (DDAVP) use [13, 14]. Other workers have alternatively used the instrument to assess thrombosis risk or pre-operative bleeding risk [15], [16], [17]. We therefore provide an overview of the history of the PFA, and its clinical utility.

Early historical aspects of the PFA and its use

As mentioned, the PFA-100 was released into the marketplace in the mid 1990s [1], [2], [3], as largely accompanied by two seminal papers [2, 3] published in a supplement to the journal Seminars in Thrombosis and Hemostasis (STH). One paper provided a detained description of the PFA-100 [2], and was authored by employees of the manufacturer (then Dade Diagnostics). The other paper provided early data from a field trial of PFA-100 [3], and was authored by the then Editor in Chief of that journal, Prof. Eberhard Mammen and his colleagues. A third article from Mammen and colleagues was published in STH in 1998 [18]. As noted in a separate publication [19], two of these papers [2, 18] count among the five most highly cited papers from STH from all time. Notably, the PFA-100 was developed from an earlier instrument called the Thrombostat 4,000, with descriptions of this instrument and its utility taking up the majority of the aforementioned supplement to STH. We mention this history because it reflects an interesting chronicle, which also helps define the then potential future utility of the PFA-100. Thus, the Thrombostat 4000 was identified in that STH supplement as a sensitive screening test for VWD [20], a useful method to test antiplatelet agents (then primarily aspirin) [21], and being affected by various hematological parameters [22], with all these also shown to later hold true for the PFA-100 [7], [8], [9]. The Thrombostat 4000 was also described as an in vitro bleeding time [23, 24], which was also adopted as a descriptor for the PFA-100 [4], [5], [6].

In summary, in the initial years of the PFA-100, its use developed as a replacement of the more invasive and less standardized ‘skin bleeding time’, as well as replacing the larger, less user friendly Thrombostat 4000, and for screening of platelet function and VWD, as well as a test for evaluation of antiplatelet agents, most notably aspirin.

What is the normal reference range for the PFA-100 or PFA-200?

One of the more interesting historical aspects of the PFA-100 and PFA-200 is the question of the NRR for tests performed on the instrument. As noted above, the current manufacturer-cited NRR for the C/Epi and C/ADP test cartridges are respectively 82–150 s and 62–100 s, for use on both the PFA-100 and PFA-200 using the standard 3.2% (0.105 M) sodium citrate blood collection tubes (and based on 309 normal donors). The corresponding NRR for use of 3.8% (0.129 M) citrate are 84–160 s (C/Epi) and 68–121 s (C/ADP) (based on 127 donors). However, the first description of the PFA-100, in a meeting abstract [1], actually listed the NRR as 97–188 s (C/Epi) and 63–123 s (C/ADP) (as based on 28–35 donors; citrate concentration not specified, but presumed to be 3.8%). In the separate independent field trial reported by Mammen and colleagues [3], the derived NRR based on 99 normal donors were 77–133 s (C/ADP) and 98–185 s (C/Epi), as based on 3.8% citrate. In the later updated study [18], the estimated NRR were reported as 72–120 s (C/ADP) and 94–191 s (C/Epi), again based on 3.8% citrate. These early variations are listed in Table 1.

In an early review from one of us [9], it was identified that this variation in reported NRRs continued for the next decade, and in a recent update, can be shown to continue to the current day [25]. Figure 1 shows a summary of some of this variation, which even includes variation within laboratories belonging to the same pathology network (Table 1). The manufacturer states in its product information: “As differences in donor population and other factors may affect results, it is recommended that each laboratory demonstrates transference of the reference ranges shown above (refer to CLSI guideline C28-A324 [26]). If transference of these reference range cannot be verified, the laboratory should establish its own reference ranges.” Although this is standard laboratory practice, we can highlight some problems with this approach.

Figure 1:

Normal reference range (NRR) data sets from the literature. Panels (A) and (B) respectively show some examples of published NRR data for C/ADP and C/Epi CTs, with some historical data detailed in a previous review [9] shown on the left and more contemporary data detailed in a recent publication [25] shown on the right. The Siemens product information NRR CTs are shown by the horizontal dashed lines. The historical and contemporary data show similar variations in NRRs. Most literature data, both historical and contemporary, approximate the manufacturer indicated ranges for C/Epi, but tend to be much wider than the manufacturer ranges for C/ADP.

First, in a recent NSW Health Pathology (NSWHP) Network evaluation [25], marked differences were seen in both the established NRRs, as used throughout the network, as well as those NRRs that could be generated using updated site-specific normal donor samples (Table 1 [25]). We would not expect to see such different NRRs based on different populations, since populations at each site location would be expected to be similar. Similarly, the differences would not be based on differences in collection tubes, since all sites use 3.2% citrate. Most likely, the different NRRs determined at each site just reflect the use of a different, perhaps too small in number, group of normal individuals, who may display variable effects of factors that influence the CT (such as platelet count, hematocrit value, VWF level and activity, prevalence of blood groups, and potential use of recent medication or ingestion of certain foods or supplements [9]). However, the cut-off value representing the upper limit of the NRR will generally be used to dichotomise individual CTs as ‘normal’ (below or equal to upper cut-off value) or ‘abnormal’ (above upper cut-off value). Thus, the NRR estimated by laboratories will determine whether a patient result is classified as normal or abnormal, and thus whether a patient may be identified to have a primary hemostasis defect or not. Clearly, too high a cut-off value will miss patients with hemostasis defects, and too low a cut-off value will over-call patients as suffering hemostasis defects. Thus, the different arising NRRs in our network would mean that should the same individual be assessed at different sites, some sites might classify that individual as normal, and others as not.

Many supplements and drugs may affect the CT, and even if the usual ‘culprits’ (e.g., non-steroidal anti-inflammatory drugs such as aspirin) are excluded in selected normal donors, others may not be recognised. Even certain foods can affect CTs, for example high chocolate and garlic intake [9]. In the NSWHP study [25], we were able to determine that all sites could adopt C/Epi ranges from the manufacturer, based on transference using CLSI guidance [26], and as per manufacturer advice, as based on the fact that >90% of data points fell within the manufacturer ranges. However, no site could adopt the C/ADP ranges from the manufacturer, which are much tighter with a very low comparative cut-off of 100 s. Indeed, a similar problem is found with the NRR reported in the literature (Figure 1; [25]). Thus, there is closer alignment in literature NRR data to the manufacturer C/Epi CT than to the C/ADP CT. This suggests that the manufacturer NRR for C/ADP may require a reset, as laboratories that adopt this NRR may over call patients as having an abnormal C/ADP CT.

One historical article published by one of us more than 20 years ago failed to find any significant difference of both C/Epi and C/ADP CT in pediatrics, showing just a non-significant trend towards lower values in children up to 15 years old [27]), likely reflecting developmental hemostasis and progressive increase of VWF in parallel with age [28, 29]. Moreover, there are no known differences in CT between sexes, or in older people [29], despite increasing levels of VWF as we age [30].

What is the clinical utility of the PFA-100 and PFA-200?

A recent PubMed search of “(PFA 100) OR (PFA 200) OR (platelet function analyser)” yielded 2,554 results, although not all captured papers actually reflect use of the PFA-100 or PFA-200. Indeed, in the literature, the term ‘PFA’ has also been used to abbreviate terms such as ‘pulsed-field ablation’, ‘parallel factor analyses’, ‘profunda femoris artery’, ‘parameter-free attention’, ‘patellofemoral arthroplasty’, and ‘paraformaldehyde’ to name a few. Nevertheless, the majority of captured papers do reflect use of the PFA-100 or PFA-200, with ongoing similar results over recent years, averaging around 100–140 papers/year in recent times (Figure 2), and suggesting that we will not see any expected lack of interest in this analyser in the foreseeable future.

Figure 2:

PubMed publications over the years using a search of “(PFA 100) OR (PFA 200)) OR (platelet function analyser)”.

Similar results are evident in the contemporary literature on a year by year basis, averaging around 100–140 papers/year.

The utility of the PFA-100/200 has been explored in several reviews from the lead author [7], [8], [9], with a brief summary of potential applications detailed in Table 2.

What is the evidence for ‘equivalence’ of the PFA-100 and PFA-200?

Given the age of the PFA-100 technology, the manufacturer has replaced the PFA-100 with the newer model, the PFA-200, in some geographic locations such as Australia and Europe. In theory, the mechanical elements of the two instruments are the same, with the PFA-200 essentially updating the user interface (touch screen), updating the software, and increasing quality control requirements (e.g., forced daily system checks). So, what is the evidence that the output from the PFA-100 and the PFA-200, namely the CTs, are identical? Interestingly, there are no head to head studies of PFA-100 vs. PFA-200 that we could identify from the literature, other than a study from the Westmead laboratory [25], and some external quality assessment (EQA) data which also in part comes from some of us [31].

In summary, there are a number of published evidences that the CTs from the PFA-100 and PFA-200 are ‘identical’. First, as mentioned, the manufacturer defined NRRs for the PFA-100 and PFA-200 are the same, and align instead to the test cartridges used. Accordingly, the manufacturer considers the output from the PFA-100 and PFA-200 as essentially equivalent. Second, a direct comparison of PFA-100 vs. PFA-200 CTs has been performed by the Westmead laboratory to verify their equivalence for laboratory diagnostic use, and for accreditation purposes [25]. Third, EQA data from the Royal College of Pathologists of Australasia Quality Assurance Program (RCPAQAP) has captured data related to PFA-100 and PFA-200 use by EQA participants, and sub-analysis of data for PFA-100 vs. PFA-200 did not uncover any concerning differences [31]. Finally, as the PFA-200 starts to take hold in laboratories, one can assess if there are any concerning signals in the literature related to different findings, which are not evident to our knowledge. Indeed, in both the Westmead and Verona laboratories, we continue to use the PFA-200 for the same purpose and in the same way as the PFA-100, with essentially equivalent findings.

The PFA-100/200 and CCLM

As noted elsewhere, this journal is celebrating its 60th birthday. Clinical Chemistry and Laboratory Medicine (CCLM) has been part of the landscape for about 10 years longer than the PFA-100. Nevertheless, as often happens in a journey spanning 60 years, there are intertwined histories, and several relevant papers related to the PFA-100, or to related fields of platelet function and VWD screening and testing have been published in CCLM over past decades. The most relevant papers are listed in Table 3, together with a brief description of the importance to the field.

Conclusions

The PFA-100 has been part of the laboratory landscape for nearly 50 years. The PFA-100 is gradually being replaced by a newer model called the PFA-200, but both instruments yield essentially the same test results, being a closure time, and using different test cartridges, with differing sensitivities to various parameters such as platelet count and function, VWF level and activity, anti-platelet medication, and various foods and supplements. The PFA-100/200 is used by different laboratories for potentially different reasons, but at its heart, the high sensitivity to disturbances in VWF level and function suggests its main purpose is for rapid exclusion of VWD, should normal CTs be determined with C/ADP and C/Epi cartridges [10, 11].

In turn, as CCLM is turning 60, the story of the PFA-100/200, or associated papers in CCLM (Table 3), indicates at least some shared history. The authors wish both the PFA-100/200 and CCLM the best of wishes for their respective 50- and 60-year anniversaries.