From Field Study to Clinical Practice, a Personal Historical Experience Using the PFA-100 Analyzer for Platelet Function Testing

In the early 1990s, the scope of laboratory testing to assess platelet function was limited to specialized laboratories performing platelet aggregation studies or using crude surrogates of platelet function with the bleeding time (BT).[1] The BT was used to assess platelet function by making an incision on select areas of the body, especially the arm or earlobe, then dabbing the incision with filter paper and recording the time when blood has clotted by absence of blood on the filter paper. The two BT methods were the Duke method and the Ivy method, with the former preferred for pediatric populations as the incision is made on the earlobe and the latter in older patients, with incision typically made on the forearm.[2] Aside from the crudeness of the procedure, challenges were often evident in patients with poor tissue turgor, where after the incision, the blood tended to “leak” into interstitial areas instead of at the incision site. The predictability of the BT, often performed prior to surgery or interventions, was called into question, as there was no association between BT test results and risks of bleeding.[3]

In 1997, Dade International Inc (hereafter referred to as Dade) received U.S. Food and Drug Administration (FDA) premarket notification (also known as clearance) for a novel device that measures platelet function, the PFA-100.[4] The decision by the FDA was based on the studies performed in selected U.S. clinical sites, comparing the PFA-100 to a predicate platelet aggregometer and traditional coagulation studies to patients with known Glanzmann's thrombasthenia (GT), von Willebrand disease (VWD), postacetylsalicylic acid (aspirin or ASA) ingestion and normal donors.[5] In that study, the PFA demonstrated overall 90.1% agreement with platelet aggregation results.[4] In known GT (N = 5) and VWD patients (N = 44), the PFA-100 was a better predictor of either condition than the BT at 61.2% and 95.9%, respectively, especially in patients with known mild VWD.[5] Of note, the University of California, Davis Medical Center (UCDMC) was one of seven study sites, and our contribution to the study used for FDA submission included 2/5 (40%) of the known GT cases, 21/44 (47.7%) of the known VWD cases (with 14 VWD:1; 3 VWD:2A, and 4 VWD:3), 22/127 (17.3%) of the pre- and post-ASA ingestion, and 35/206 (17.0%) of the normal subjects.

The PFA-100 reports closure time results in seconds, which represents the time required to occlude the aperture within a platelet agonist coated membrane under shear flow conditions. There were two types of cartridges, one consisting of collagen/epinephrine (CEPI) coated membrane and the second consisting of collagen/ADP (CADP) coated membrane. The Dade proposed using the PFA-100 testing algorithm to screen all patients with the CEPI cartridge, and if abnormal, performing the test using the CADP cartridge. Influences for CEPI or CADP closure time prolongation included low hematocrit (<35%) and low platelet counts (<150,000/mm³). With VWD or severe thrombocytopathies, the expected response would be prolongation for one or both cartridges, with an increased CEPI with normal CADP depicting the classic pattern for recent ASA exposure. A historical review of the PFA-100 has recently been published in this journal.[6]

A distinct advantage for being a site participant was our local verification of the instrument performance, as required by regulatory authorities, was already completed when the PFA-100 was cleared by the FDA for clinical use, since we implemented this device to supplant the BT.[7] The implementation of the PFA-100 at our facility created both opportunities and problems. Our experience and knowledge about the PFA-100 was robust by field study participation but also limited because, like most studies, we were population defined to subject enrollment, and thus did not appreciate the instrument performance as a general screening tool for the wider population. That said, the possibility to explore the effect of clinical situations on platelet function was a never-before opportunity given the sample stability and testing complexities for platelet aggregation or equivalent assays.

Our first real clinical encounter was PFA testing performed on a minor, a child presenting with a nosebleed in the emergency department. The PFA-100 was collected, performed, and the test results mimicked that of ASA exposure (prolonged CEPI with normal CADP). When confronted with that result, the parents were adamant that such exposure was impossible, and were irritated with the suggestion they would expose their child to possible severe side effects associated with ASA ingestion, including Reyes' syndrome.[8] [9] In fact, we later determined that the patient had a mild form of VWD, which was demonstrable by CEPI, but not so with CADP. With those findings, we started to include the PFA-100 as a part of our platelet lumi-aggregation panel, and noted that in certain patients the CADP was prolonged with a normal CEPI, and thus decided to change our testing to concurrently measure the CEPI and CADP cartridges. One other use for the PFA-100 was the assessment of the efficacy of postplatelet transfusions in Glanzmann's thrombasthenia. Another interesting patient “encounter” was from outside our institution where a physician from a northern California hospital called after noting our facility on the publication.[5] In that situation, the patient had a prolonged BT (>12 minutes), yet presented with normal closure times for both CEPI and CADP. The clinicians were adamant the PFA-100 was insensitive to detecting significant thrombocytopathy. When we asked for the patient history and diagnosis, the patient was diagnosed with Ehler–Danlos syndrome, which can be associated with prolonged BTs with normal platelet function.[10] While those two events enlightened both internal and external sites using the PFA-100, another novel consideration was undertaken during my tenure. The most common treatment for Glanzmann's thrombasthenia is periodic platelet transfusions.[11] In two separate scenarios, we evaluated the efficacy of platelet transfusions in treated GT patients. One study was an ex vivo evaluation of a patient who was known to be refractory to platelet transfusions ([Table 1]). These data suggested the patient had some potential immune response to one unit of platelets, as compared with the second unit of platelets. The other ex vivo study was in two female siblings with GT. Both girls were transfused on the same day, and the postanalysis of each are given in [Table 2]. These data suggested that one sibling had successful transfusion and platelet function recovery, whereas results from the second sibling suggested an immune response to platelet transfusions. This simple assessment was later confirmed with antibody testing and alternative therapies, especially the use of recombinant human activated factor VII (rFVIIa), which were considered in the future.[12]

Table 1 An ex vivo evaluation of a Glanzmann's thrombasthenia patient who was known to be refractory to platelet transfusions

	CEPI, s (normal range: 94–193)	CADP, s (normal range: 71–118)
PLT unit H96930	115	40
PLT unit H97000	157	88
Patient plasma and RBCs add-back to H96930 PLT	> 300	220
Patient plasma and RBCs add-back to H97000 PLT	212	110

Abbreviations: CADP, collagen/ADP; CEPI, collagen/epinephrine; PLT, platelets; RBC, red blood cells.

Table 2 An ex vivo study in two female siblings with Glanzmann's thrombasthenia

	CEPI, s (normal range: 94–193)	CADP, s (normal range: 71–118)
Sibling A: pretransfusion	>300	>300
Sibling A: posttransfusion	211	122
Sibling B: pretransfusion	>300	>300
Sibling B: posttransfusion	>300	275

Abbreviations: CADP, collagen/ADP; CEPI, collagen/epinephrine.

In the early days of PFA-100 use, there were two processes required prior to patient testing. One was checking the instrument for appropriate vacuum and absence of pressure leaks. This was, and still is, a daily check requirement using a specialized cartridge holder to ensure no vacuum leaks in the instrument's O-ring rimmed chuck.[13] The second was the testing of a normal control, which meant the use of a citrated whole blood sample that was tested on both cartridges each day of use. The latter requirement was somewhat challenging after a while, since local staff members declined in willingness to donate a sample over time. Alas, we had thought the problem was solved when we reached out to a local blood collection/donor center used for local transfusions, which was less than a mile away from our laboratory. The blood collection center guaranteed to get a normal sample each weekday within the allowable sample stability limits (4 hours from collection). It soon became apparent that such a practice was not working, as approximately 40% of the “normal” donors that the blood collection center had prolonged PFA-100 closure times. This prompted an investigation at the blood collection center because of the potential quality assurance issue associated with blood donors and abnormal platelet function. One salient feature of blood collection centers is the encouragement for blood donors to eat prior to donating blood to minimize postdonation problems (i.e., dizziness, hypoglycemia), but typically no dietary restrictions are implemented. We investigated a series of donors that were used for PFA controls, and inquired about their breakfast meal before blood donation, then did a pre-/postassessment of platelet function using the PFA-100. We were able to demonstrate that chocolate milk was the likely culprit for prolonged PFA-100 closure times.[14] We also noted in this study, along with observations in our field study participation, that ASA use is so innocuous in our society that people tend to forget when they last took ASA or any other nonsteroidal anti-inflammatory drugs that are known to affect platelet function. Two additional collaborative interdepartmental studies were published related to these findings, indicating the influence of cocoa on platelet function using the PFA-100.[15] [16]

As previously noted, the PFA-100 allowed us to assess platelet function in clinical settings that were prohibitive using traditional aggregation techniques. I personally participated in numerous interdepartmental studies evaluating platelet function in both human and animal studies, which would not have been feasible using other platelet function methods. Although not all of the clinical studies made it to manuscript end point, numerous studies were presented in the poster format.

The largest study conducted by our group was the prospective evaluation of platelet activation using flow cytometry and platelet function using the PFA-100 after trauma.[17] In that study, we demonstrated that while trauma patients in general had shorter CEPI and CADP closure times as compared with reference interval upon admission, nonsurvivors had a marked reduction platelet function as demonstrated by prolonged CEPI closure times, and patients with head injuries (abbreviated injury score >4) had significantly decreased platelet function by CEPI as compared with non-head-injury patients. Of particular note, subjects enrolled in the trauma study mostly occurred during off-hours and weekends. Due to the limited sample stability (4 hours), this necessitated the running of samples on the PFA-100 by whomever was “on-call” for subject enrollment. As such, the PFA-100 testing was performed either by myself or the trauma residents, with the latter users confirming the ease of use for instrument operation.

In this trauma study, we often had PFA-100 closure times with “C” error, indicating an early closure suggesting a clotted sample. However, we had programed our instrument to print out results using the “research” print setting, which provided a graph of closure time changes in pressure over time. Typically, in samples where the closure time yields a “C” error, there was an abrupt change in the pressure change. However, in the trauma patients, the closure time was not as abrupt, but occurred faster than anticipated, and thus the “C” error closure time was reported.

We also demonstrated that PFA-100 CADP closure times were lower in trauma patients as compared with normal and postsurgical patients, as well as demonstrating increased platelet activation using surrogate markers p-selectin (CD62p, α granule exposure after activation) and PAC-1 (activated platelet glycoprotein IIb-IIIa for fibrinogen binding; [Table 3]).[18] [19] Given these observations, we were curious whether there was a correlation between PFA closure times and markers of platelet activation (PAC-1, CD62p) and von Willebrand factor (VWF). There was only a significant inverse correlation between platelet count, hematocrit, and VWF activity for CEPI and platelet count and VWF activity for CADP, but no correlation between PFA-100 closure times and flow cytometry measurements of platelet activation ([Table 4]).

Table 3 Median (range) PFA-100 collagen/ADP (CADP) closure times in normal, trauma, and surgical patients

	CADP, s (normal range: 71–118)	PAC-1 binding, %	CD62p, %
Normal subjects (N = 33)	110 (75–182)	1.93 (0.3–13.8)	4.94 (1.9–20.0)
Trauma patients (N = 182)	89[a] (23–300)	9.02[a] (0.4–65.1)	7.69[a] (0.8–63.1)
Surgical patients (N = 60)	92[b] (40–213)	2.13[b] (0.5–13.6)	4.83[b] (1.1–17.2)

^a p < 0.05 for normal versus trauma patients.

^b p < 0.05 for surgery patients versus trauma patients.

Table 4 PFA-100 closure times versus select blood parameters

	PFA-100 CADP cartridge		PFA-100 CEPI cartridge
	Spearman rank correlation	p-value	Spearman rank correlation	p-value
Platelet count	–0.48	<0.05	–0.28	0.04
Hematocrit	–032	0.02	–0.04	0.78
VWF activity	–0.36	0.005	–0.32	0.01
CD62p	–0.19	0.15	–0.21	0.10
PAC-1 binding	–0.21	0.10	–0.14	0.29
Platelet microparticles	–0.07	0.59	–0.05	0.70

Abbreviations: CADP, collagen/ADP; CEPI, collagen/epinephrine; VWF, von Willebrand factor.

Other published studies of our experience with PFA-100 testing have included the effect of stroma-free hemoglobin oxygen carrier (HBOC) on platelet function,[20] and comparison to a viscoelastic test, the TEG5000.[21] Additional UCDMC PFA-100 studies that were presented but did not reach full manuscript publications included effects of hormonal replacement therapy on platelet function,[22] platelet function in diabetic patients,[23] platelet function in whole blood donors,[24] and relationship between closure times and flow cytometry measurements of platelet activation.[25] More recently, a publication on provisional guidance for creating abnormal surrogate samples that may be used as part of an individualized quality control plan was described for the PFA-100.[26]

My early unpublished experiments with the UCD Veterinary School of Medicine demonstrated the limited utility of CEPI in dogs, which was later confirmed.[27] Other interdepartmental UCD studies that did not rise to the level of manuscript publication or poster presentations include effects of dobutamine, renal dialysis, phenols (throat antiseptics), homocysteine, caffeine, and polyphosphates on platelet function using the PFA-100.

To conclude, the availability of the PFA-100 radically changed our practice at UCDMC for assessing platelet function. First, we were able to replace the time-consuming and relative worthless predictive test, the BT, and, second, could provide a rapid, sensitive (albeit sometimes too much so) platelet function test that could be performed around the clock, every day of the year. While the PFA-100 is not the panacea for assessing all platelet thrombocytopathies, the ease of use for instrument operation has created a bank of knowledge about the effect of diseases, diets, and interventions in both human and animal models on platelet function.

Publication History

Article published online:
20 December 2023

• References

• 1 Bick RL. Platelet function defects associated with hemorrhage or thrombosis. Med Clin North Am 1994; 78 (03) 577-607
• 2 Quick AJ. The bleeding time as a test of hemostatic function. Am J Clin Pathol 1975; 64 (01) 87-94
• 3 Rodgers RP, Levin J. A critical reappraisal of the bleeding time. Semin Thromb Hemost 1990; 16 (01) 1-20
• 4 Food and Drug Administration. DADE PFA-100 Platelet Function Analyzer; PFA Collagen/Epinephrine test cartridge. ; PFA Collagen/ADP test cartridge K970505 Summary Decision. https://www.accessdata.fda.gov/cdrh_docs/pdf/K970505.pdf . Accessed November 14, /2023
• 5 Mammen EF, Comp PC, Gosselin R. et al. PFA-100 system: a new method for assessment of platelet dysfunction. Semin Thromb Hemost 1998; 24 (02) 195-202
• 6 Favaloro EJ. Description of an in vitro platelet function analyzer (PFA-100/PFA-200) 30 years in the making. Semin Thromb Hemost 2024; 50 (02) 320-324
• 7 Gosselin RC. From ink pens to computers: a personal look back at landmark changes during 5 decades as a clinical laboratory scientist in U.S. Hemostasis Laboratories. Semin Thromb Hemost 2023; 49 (03) 225-233
• 8 Maheady DC. Reye's syndrome: review and update. J Pediatr Health Care 1989; 3 (05) 246-250
• 9 Glasgow JF, Middleton B. Reye syndrome: insights on causation and prognosis. Arch Dis Child 2001; 85 (05) 351-353
• 10 Malfait F, De Paepe A. Bleeding in the heritable connective tissue disorders: mechanisms, diagnosis and treatment. Blood Rev 2009; 23 (05) 191-197
• 11 Nurden AT, Freson K, Seligsohn U. Inherited platelet disorders. Haemophilia 2012; 18 (Suppl. 04) 154-160
• 12 Poon MC. Clinical use of recombinant human activated factor VII (rFVIIa) in the prevention and treatment of bleeding episodes in patients with Glanzmann's thrombasthenia. Vasc Health Risk Manag 2007; 3 (05) 655-664
• 13 Siemens Healthineers. PFA-100 System Instruction Manual (US), version 1.0.1. Marburg, Germany: Siemens Healthcare Diagnostics Products GmbH;; 2017
• 14 Paglieroni TG, Janatpour K, Gosselin R. et al. Platelet function abnormalities in qualified whole-blood donors: effects of medication and recent food intake. Vox Sang 2004; 86 (01) 48-53
• 15 Rein D, Paglieroni TG, Wun T. et al. Cocoa inhibits platelet activation and function. Am J Clin Nutr 2000; 72 (01) 30-35
• 16 Pearson DA, Paglieroni TG, Rein D. et al. The effects of flavanol-rich cocoa and aspirin on ex vivo platelet function. Thromb Res 2002; 106 (4–5): 191-197
• 17 Jacoby RC, Owings JT, Holmes J, Battistella FD, Gosselin RC, Paglieroni TG. Platelet activation and function after trauma. J Trauma 2001; 51 (04) 639-647
• 18 Devine DV. Novel markers for the detection of platelet activation. Transfus Med Rev 1990; 4 (02) 115-120
• 19 Michelson AD, Barnard MR, Krueger LA, Frelinger III AL, Furman MI. Evaluation of platelet function by flow cytometry. Methods 2000; 21 (03) 259-270
• 20 Jahr JS, Liu H, Albert OK. et al. Does HBOC-201 (Hemopure) affect platelet function in orthopedic surgery: a single-site analysis from a multicenter study. Am J Ther 2010; 17 (02) 140-147
• 21 Gosselin RC, Estacio EE, Song JY, Dwyre DM. Verifying the performance characteristics of the TEG5000 thromboelastogram in the clinical laboratory. Int J Lab Hematol 2016; 38 (02) 183-192
• 22 Aspry KE, Paglieroni T, Gosselin R. et al. Effects of post-menopausal hormone replacement therapy on ex-vivo platelet activation & aggregation. J Investig Med 1999; 47 (02) 45A
• 23 Dwyre DM, Rios D, Wun T, Janatpour KA, Gosselin R, Larkin EC. Evaluation of platelet function in diabetic patients using the PFA-100™. Blood 2001; 98 (11) 59B
• 24 Paglieroni TG, Janatpour K, Gosselin R. et al. Abstract #4075 Platelet function in whole blood donors. American Society of Hematology. Blood 2001; 98 (11) 108B
• 25 Gosselin RC, Paglieroni T, Owings J. et al. Decreased PFA-100 values is associated with increased platelet activation. Hematol J 2002; ;(3, Suppl 1): 89
• 26 Gosselin RC. Preparing surrogate abnormal quality control samples for platelet function studies and viscoelastic testing. Methods Mol Biol 2023; 2663: 637-645
• 27 Callan MB, Giger U. Assessment of a point-of-care instrument for identification of primary hemostatic disorders in dogs. Am J Vet Res 2001; 62 (05) 652-658