Evaluation of the PDQ™ centrifuge for preparing platelet rich, platelet poor and platelet free plasma samples for light transmission aggregometry and microparticle measurement

Light transmission aggregometry (LTA) is considered the gold standard for platelet function testing [1]. This test requires the preparation of platelet rich plasma (PRP) and platelet poor plasma (PPP) from each blood sample [2]. Preparation of PRP and PPP by standard protocols is time consuming and poorly standardized (in terms of choice of g values and time of centrifugation) normally taking around 30–45 minutes [3–5]. There is now also much interest in the measurement of microparticles within platelet free plasma preparations (PFP) [6]. Again the preparation of PFP is also poorly standardized but is important for eliminating the generation of cell derived microparticles from any residual contaminating platelets or other cells during sample storage [6]. Recently, a new centrifuge (PDQ™, Bio Data Corporation, Alpha Laboratories Hampshire, UK) has been introduced which is intended for the rapid preparation of PRP, PPP, and PFP samples using fixed g values and centrifugation times. The main advantage of the PDQ™ centrifuge is its ability to prepare these samples rapidly (less than 5 minutes) in a standardized manner which could be of significant benefit to clinical and research laboratories. The aim of this study was to therefore compare and contrast the properties of PRP, PPP and PFP samples prepared by the PDQ™ with our normal laboratory sample preparations.

Whole blood samples (n = 10) from apparently healthy individuals (six males and four females) who were not taking any anti-platelet drugs were collected into four Vacutainer™ tubes (Becton Dickinson, Oxford, UK) containing 0.105 M sodium citrate (1:9 v/v) using a 21 gauge needle, an initial discard tube and minimal tourniquet. Two samples were centrifuged at 170 g for 10 minutes to prepare PRP. For the preparation of PPP, samples were further centrifuged at 2000 g for 20 minutes. PFP samples were prepared by double centrifugation at 2000 g for 20 minutes followed by a rapid short centrifugation at 13,000 g for 2 minutes [7]. All supernatants (top two thirds) were carefully removed and transferred to capped tubes. Two samples were also simultaneously centrifuged in the PDQ™ to rapidly prepare PRP, PPP, and PFP samples. The instrument is pre-programmed to spin at 8500 rpm (4400 g) and prepares PRP in 30 seconds (2220 g-min), PPP in 120 seconds (8880 g-min), and PFP in 180 seconds (13,320 g-min). Platelet counts including mean platelet volume (MPV), platelet distribution width (PDW), and platelet large cell ratio (P-LCR) were measured in all PRP, PPP and PFP samples using an impedance hematology analyzer (KX-21, Sysmex, Milton Keynes, UK). Platelet aggregation within unadjusted PRP samples was also measured in response to ADP (0.5, 1, 2.5, and 5 µM final concentrations) within a four channel platelet aggregometer (Aggram™ aggregometer, Helena Biosciences, Gateshead, UK). 100% transmission values were calibrated using PPP samples prepared by the method being tested. As the impedance counter has a lower limit of 1 × 10⁹/L a modified flow cytometric reference method was also used to determine residual platelet counts in PPP and PFP samples (FacsCalibur, Becton Dickinson, Oxford, UK) [8]. Briefly, 5 µL of samples were incubated with 5 µL of a FITC-conjugated anti-CD61 antibody (Becton Dickinson, Oxford, UK) within 40 µL of Isoton for 10–15 minutes at room temperature before final dilution in 2 mL of Isoton. After gentle mixing, 1 mL was transferred to a tube containing Trucount™ beads (Becton Dickinson, Oxford, UK) and incubated for 10 minutes.

Samples were then analyzed within a flow cytometer (Facscalibur, Becton Dickinson, Oxford UK) using CellQuestPro software (Becton Dickinson, Oxford, UK) and the platelet count derived from measuring the ratio of fluorescent platelets to the known number of Trucount beads in the sample with correction for the volumes [9]. Although the volumes of PRP prepared by the two methods were identical, there were significant differences in both platelet counts and size distributions (Table I) Furthermore, although the aggregation response to ADP were normal, the maximum aggregation responses to ADP in PDQ™ prepared PRP were significantly lower than that in samples prepared by the standard protocol (Table I). The apparent loss of the largest platelets in the PRP prepared by PDQ™ centrifuge, as reflected by the lower P-LCR, MPV and PDW probably explains these results (Table I). Large platelets are indeed thought to be more reactive to agonists than smaller platelets and this data is consistent with this [10].

Table I.  Comparison of PRP prepared by a standard centrifugation protocol (170 g for 10 minutes) and the PDQ™ centrifuge (2220 g for 30 seconds).

	Standard protocol	PDQ™
Volume of PRP obtained	1.62 mL (±0.44)	1.75 mL (±0.46)
Platelet count (×10^9/L)	389 (±72)	197 (±65)^1
MPV (fl)	8.5 (±0.5)	7.6 (±0.5)^1
PDW(fl)	10.7 (±1.2)	9.4 (±0.8)^1
P-LCR (fl)	16.4 (±3.9)	10.3 (±3.1)^1
0.5 µM ADP (%MA)	38.4 (±32.2)	1.2 (±8.8)^1
1 µM ADP (%MA)	57.3 (±33.7)	20.0 (±9.6)^1
2.5 µM ADP (%MA)	76.7 (±22.5)	59.3 (±17.1)^1
5 µM ADP (%MA)	85.3 (±5.6)	78.9 (±7.0)^1

¹p < 0.05 (paired t-test).

Results expressed as mean ± SD. MA = maximal aggregation.

Interestingly, despite the reduction in responsiveness, the overall aggregation responses were also much less variable in PRP samples prepared by the PDQ™ centrifuge. However, the differences in platelet counts between the two PRP samples may also have contributed to the differences in results, although sufficient numbers of platelets are still present in the PDQ™ prepared PRP to mediate normal aggregation. The PPP and PFP samples prepared by both methods also showed differences in terms of the number of residual platelets (Figures 1a and 1b). Samples prepared by PDQ™ centrifuge always contained higher numbers of platelets than samples prepared by our standard protocol and this was statistically significant (p < 0.05, Wilkoxon paired test) particularly when measured by flow cytometry (Figure 1b), as this method has a detection limit of at least two orders of magnitude lower than conventional counting methods. However, the majority of the counts in the PDQ™ prepared PPP and PFP samples were within the centrifuge manufacturer's specifications (<10 × 10⁹/L and <0.5 × 10⁹/L respectively) The results also suggest that impedance counters are unsuitable for detecting very low numbers of platelets in these samples. This is unsurprising given the well known inaccuracy of impedance counters in severe thrombocytopenia and their theoretical lower detection limit of 1 × 10⁹/L. However, the platelet count in the PDQ™ prepared PPP samples was still acceptably low to set the 100% transmission on the aggregometer. However, the number of residual platelets also suggests that either the PPP and PFP samples may be unsuitable for subsequent freezing, storage and microparticle analysis.

Figure 1. Box plots (median and IQR) showing the comparison of the platelet counts as measured by either impedance (a) or flow cytometry (b) in platelet poor plasma (PPP) and platelet free plasma (PFP) samples prepared by standard methods (Old) or the PDQ™ centrifuge (PDQ). There were significantly higher (p < 0.05, Wilkoxon paired test) numbers of platelets measured by flow cytometry in the PPP and PFP samples prepared by the PDQ centrifuge compared to standard methods.

The PDQ™ centrifuge is potentially able to significantly reduce the total time in the sample preparation for the measurement of platelet function by LTA. It also has the additional advantage of standardizing the preparation of PRP, PPP, and PFP samples. However, the significant differences in aggregation responses to ADP compared to conventional preparation procedures, coupled with the loss of the larger sub-population of platelets could also be viewed as a disadvantage. This may be an important consideration in measuring platelet responsiveness to ADP and other agonists in a variety of situations. The PDQ centrifuge can also be potentially useful to rapidly prepare PPP and PFP samples. Clearly a single rapid PPP high g centrifuge step using the PDQ™ does not remove as many platelets as a conventional 20 minute but lower g spin. It is therefore important to be aware that there are a higher number of residual platelets within both PPP and PFP samples prepared by this new method compared to standard methods. This may not only potentially influence assays that are affected by endogenous phospholipids and/or platelet factor 3 (e.g. endogenous thrombin potential and anti-phospholipid assays) but affect the accurate measurement of microparticle number and phenotype within stored plasma samples prepared in this manner [6], [7].