Does Double Centrifugation Lead to Premature Platelet Aggregation and Decreased TGF-β1 Concentrations in Equine Platelet-Rich Plasma?

Blood-derived autologous products are frequently used in both human and equine medicine to treat musculoskeletal disorders. These products, especially the platelet-rich plasma (PRP), may contain high concentrations of growth factors (GFs), and thus improve healing in several tissues. Nevertheless, the procedures for preparation of PRP are currently non-standardized. Several protocols, which are based on distinct centrifugation patterns (rotation speed and time), result in PRPs with different characteristics, concerning platelet and GFs concentrations, as well as platelet activation. The aim of the present study was to compare two different protocols for PRP preparation: protocol (A) that is based on a single-centrifugation step; protocol (B), which included two sequential centrifugation steps (double-centrifugation). The results here reported show that the double-centrifugation protocol resulted in higher platelet concentration, while leukocytes were not concentrated by this procedure. Although platelet activation and aggregation were increased in this protocol in comparison to the single-centrifugation one, the TGF-β1 concentration was also higher. Pearson’s correlation coefficients gave a significant, positive correlation between the platelet counts and TGF-β1 concentration. In conclusion, although the double-centrifugation protocol caused premature platelet aggregation, it seems to be an effective method for preparation of PRP with high platelet and TGF-β1 concentrations.


Introduction
In the last decade, hemoderivative therapeutics has become a suitable alternative to treat musculoskeletal disorders in both human and veterinary medicine. Particularly, the platelet-rich plasma (PRP) has shown beneficial effects on several joint diseases, which help to relieve pain and improve the articular structures (reviews in [1][2][3][4]).
A systematic review on the efficacy of PRP in equine and human orthopedic therapeutics [1] concluded that, although many equine studies yielded positive results, the same is not true for human clinical trials. Furthermore, beneficial results for both species were more frequently observed in studies with a high risk of bias, which lead the authors to conclude that the use of PRP has still not shown strong positive effects in clinical scenarios.
Campbell et al. [2] have demonstrated, through meta-analysis and systematic review, that many studies, including comprehensive and meticulous ones, lack detailed information regarding the procedure for PRP preparations (centrifugation steps) and their characteristics (platelet counts, growth factor concentrations, and administered volumes). three and five years, which were clinically healthy and free of any drugs for 15 days prior to the study, were included in the present study.
The sample size calculation was determined to provide an 80% statistical power mainly for TGF-β1, and platelet concentration to detect a difference of 30% between the groups, with a two-sided alpha level of 0.025 and β = 0.20 based on two-way analysis of variance.

Blood Collection
Blood was collected from jugular vein following aseptic preparation of the site. For the single centrifugation protocol (see below), blood was collected while using a 21-gauge needle and a 10 mL syringe, and transferred to a Falcon tube (15 mL) containing 1 mL of 3.8% sodium citrate as anticoagulant. For the double centrifugation protocol, a 21-gauge needle and seven vacutainer tubes containing 3.8% sodium citrate were used (4.5 mL; final dilution blood:anticoagulant was 9:1). After harvesting and homogenization of the blood, the final volume was aliquoted and transferred to three Falcon tubes (15 mL), 10.5 mL each.

Platelet Rich Plasma Preparation
Two different protocols were used for PRP preparation: a single and a double centrifugation procedure, as schematized in Figure 1. The products were analyzed. The single centrifugation protocol that was used here was adapted from Ottaiano et al. [24], in which premature platelet aggregation is avoided. In brief, blood was collected in one tube and centrifugated at 140× g for 12 min. (room temperature), and the supernatant was collected was named "PRP-0".
For the double centrifugation protocol [4], three 15 mL tubes were used, containing 10.5 mL of blood each. After collection, the blood samples were incubated at room temperature for 25 min., and then were centrifugated at 300× g for 5 min., followed by another 25 min. "rest period". It was empirically observed that less platelet activation occurred when these "rest period" were included. The supernatants were collected, and one of them was reserved and named "PPP-1", while the other two were submitted to a second 700× g, 15 min. centrifugation, followed by another "rest period" of 45 min. The supernatant from the second tube was collected, reserved, and named "PPP-2". The pellet and remaining volume were gently homogenized using a pipette and named "PRP-1". The supernatant of the third tube, named "PPP-0", was collected and cleared by centrifugation at 1000× g for 10 min., to assure minimal platelet counting in the supernatant. This was use as a sample with 100% light transmission in aggregometry testing, as described below.

Platelet and Leukocyte Counts
A complete blood count was performed on each blood sample. Leukocyte (WBC) counts were completed in the PRP-0, PRP-1, and PPP-1 sample while using a flow cytometry hematology system (ADVIA 2120i, Siemens -Erlangen, Germany).
Platelet counts were performed in whole blood (baseline values) samples and in the PRP-0, PRP-1, PPP-1, and PPP-2 samples while using a Neubauer chamber. Platelet concentrations were described as a ratio between platelet count for each sample and the platelet count for the whole blood sample.

Aggregometry Test
Platelet aggregation responses were evaluated while using an aggregometer (Chrono-Log Corporation, model 700) and monitored by the turbidimetric method developed by Born and Cross [25] using Agrolink software (Chrono-Log Corporation, Havertown, Pennsylvania, USA). Prior to this test, the PPP-0 was used as sample with 100% light transmission. To perform this test, 500 µL of PPP-0 was used as a blank. Likewise, 498 µL of each sample were placed in a 6-mm-wide siliconized cuvette with a stirring bar magnet and maintained at 37 • C. The samples were then identified with the software and baselines were adjusted to 0%. Two µL of agonist collagen type I (Chrono-Log Corporation) were added to the cuvettes, which were agitated. The aggregation curve was then observed for a 6-min. period.
This test was performed in all samples, except PPP-2, because, due to its low platelet counts, PPP-2 did not gave reliable results. The agonist alone was added as a control. All of the samples were stored at −80 • C until further analysis and TGF-β1 determination.

Determination of TGF-β1 Concentrations
TGF-β1 concentrations were quantified in all samples by ELISA, while using a commercially available human TGF-β1 kit (Quantikine ELISA, R&D Systems, Inc; Minneapolis, MN, USA, DB100B), previously validated for use in equine samples [26,27], in accordance with manufacturer instructions. All of the samples were run in duplicates.
After thawing and homogenization, TGF-β1 of each sample was activated: aliquots (20 µL) were placed in microcentrifuge tubes (200 µL), and 10 µL of 1 M HCl were added, which were then homogenized and incubated at room temperature. After 10 min., 10 µL of 1.2 M NaOH and 160 µL of RD5-53 diluent (R&D Systems, Inc; Minneapolis, MN, USA, part #895587) were added to each sample. RD5-53 diluent (50 µL) was added to all the wells of a 96-well anti-TGF-β1 antibody coated plate (TGF-β1 plate). Standards or samples (50 µL) were then added to the appropriate wells and then incubated for 2 h at room temperature with gentle agitation. The solution was discarded and the plate was washed four times with wash buffer solution (R&D Systems, Inc; Minneapolis, MN, USA, part #895126). Next, 100 µL of anti-TGF-β1 polyclonal antibody conjugated with HRP (R&D Systems, Inc; Minneapolis, MN, USA, part #893003) were added to each well, which were then covered by a film and then incubated again for 2 h at room temperature. The washing procedure was repeated, as previously described, and then 100 µL of substrate solution were added to each well. The plate was then covered with aluminum foil and incubated for 30 min. at room temperature. Finally, 100 µL of stop solution was added to each well. The wells were homogenized and absorbances were read while using a microplate reader at 450 nm.

Statistical Analyses
The data were analyzed for normality while using Kolmogorov-Smirnov tests. Afterwards, an analysis of variance (ANOVA), followed by Tukey-Kramer or Bonferroni tests, were used to compare parametric data. Non-parametric data were analyzed via the Friedman test. Correlation analyses were performed while using coefficients of determination and Pearson's correlation coefficients. Statistical analyses were performed using GraphPad Instat 3 (GraphPad Software, San Diego, CA, USA). Statistical significance was set to p < 0.05 and confidence interval (CI) of 95%.

Platelet and Leukocyte Counts and Platelet Concentration Times
The baseline platelet counts for the horses that were used in the present study was 180 × 10 3 /µL. Figure 2 shows that the platelet counts were significantly increased in PRP-0, obtained by the single-centrifugation protocol, and also in the products that were obtained by the first and the second centrifugation steps in the double-centrifugation protocol (PPP-1, and PRP-1, respectively) (p = 0.0001). For PRP-0, the mean platelet count was 494 ± 157 × 10 3 /µL (2.7 times the baseline), while for PPP-1 and PRP-1, the platelet counts were 756 ± 228 × 10 3 /µL (4.2 times) and 1371 ± 423 × 10 3 /µL (7.6 times), respectively. These data show that higher platelet counts were obtained in the products that were prepared by the double-centrifugation protocol (PPP-1 and PRP-1), as compared to the lower values that were obtained by the single centrifugation procedure (PRP-0). As expected, the platelet counts in PPP-2 (supernatant of the second centrifugation step) was lower than the baseline. In contrast, the leukocyte counts decreased in all analyzed products (PRP-0, PRP-1, and PPP-1) relative to the baseline (Figure 2), which indicated that the platelet-rich plasma was not leukocyte-rich.

Discussion
The aim of the present study was to establish a protocol to prepare PRP with high platelet counts, as well as high TGF-β1 concentration. PRPs with such characteristics are supposed to be more useful for intra-articular administration in treating joint diseases.
Two different protocols were tested, and the products thus obtained were characterized, concerning platelet counts, TGF-β1 concentrations, and premature platelet activation/aggregation.
Our results have shown that the first protocol, based on a single centrifugation step, avoided premature platelet activation/aggregation, but resulted in a preparation with relatively low platelet and TGF-β1 concentrations (PRP-0).
In contrast, the final product that was obtained by the double-centrifugation protocol-PRP-1-presented high platelet counts and high TGF-β1 concentration, but some degree of premature platelet activation occurred. Leukocyte counts were not concentrated.
PPP-1, the product of the first centrifugation step in the double centrifugation protocol, was similar to PRP-0. This product could be considered to be a "PRP", since it presents the minimum platelet concentration required to be considered a "PRP" by Fortier et al. [19] (4.2 times). Table 3 summarizes the main characteristics of our preparations and of the preparations described in the literature. TGF-β1: Transforming growth factor beta 1. *: Platelet concentration (times).
The platelet concentrations in PPP-1 and PRP-1 were higher than many other preparations for equines [7,28,29,31], and was also higher than the concentration that was obtained by Hessel et al. [30] while using the E-PET ® device.
The PRP-1, obtained after the second centrifugation step, resulted in a platelet concentration of 7.6 times the baseline. This concentration was higher than most of other studies, both in equines and human (Table 3), but it was slightly lower than those reported by Textor et al., using double centrifugation [32] that achieved weight times concentration, and Sutter et al. [8] that achieved 13 times and nine times concentrations, respectively, by apheresis/filtration and buffy coat methods. Nevertheless, apheresis only gave lower platelet concentrations (5 times) [8].
The presence of leukocytes in PRP may be useful or harmful, depending on the planned use and treatment objective [38]. In the present study, PRP was aimed for intra-articular injection. In this case, leukocytes are undesirable, because it is already known that patients who receive leukocyte-rich PRP present more joint swelling and pain [39], besides other deleterious effects [2,33]. Leukocyte were not concentrated in none of our products-PRP-0, PPP-1, and PRP-1. Nevertheless, it is important to note that the leukocytes count in PRP-1 were higher than those that were observed by Vendruscolo et al. [29] in their study of ten different protocols for double centrifugation. We also found higher leukocyte counts than those that were reported by Pereira et al. [35]. These results might be explained by differences in the plasma collection after centrifugation (although always avoiding the "buffy coat").
The leukocyte concentration in our PRP-0 was also very low, in contrast to the results that were reported by Fontenot et al. [28], who also used a single-centrifugation protocol, but at higher speed, resulting in leukocytes-rich plasma. Of course, the growth factor composition of PRP is affected by the presence of other cell types [34].
In PRP preparation, besides the concentration of platelets, its functionality is also an issue, because platelets that were prematurely activated have possibly released their growth factors. This point was evaluated here by an aggregometry test, which was induced by collagen as exogenous agonist. By the single-centrifugation protocol, the platelets were not activated, in contrast to the double-centrifugation protocol, in which 90% of platelets responded after the first centrifugation step (PPP-1), and only 66.8% responded after the second centrifugation step (PRP-1). These results suggest that increased centrifugation speed and time could lead to platelet activation, which is in agreement with other authors [40]. Argüeles et al. [7] observed higher platelet activation after the first centrifugation step, which was possibly due to the presence of red blood cells in these samples, while Kingston et al. [41] have shown that sodium citrate partially inhibits platelet aggregation.
In the present study, collagen was used as exogenous agonist. Reversible responses are obtained [42], although other agonists may be also used, such as epinephrine and ADP, in contrast to collagen that induces irreversible aggregation and release of growth factors, mainly TGF-β1 and PDGF-BB in a dose-dependent fashion, up to 10% of the estimated total platelet content [32].
The TGF-β1 concentrations were similar in PRP-0 (single-centrifugation) and PPP-1 (first centrifugation step in the double-centrifugation protocol), despite their differences in platelet counts. This could be due to the premature platelet activation and partial TGF-β1 release in the preparation of PPP-1.
The highest TGF-β1 concentrations were found in PRP-1 (double-centrifugation procedure), again demonstrating a correlation between platelet counts and TGF-β1 concentration. This result agrees with other studies on PRP in equines [8,33,34,36]. PPP-2, which is the supernatant of the double-centrifugation protocol, contained low platelet counts and high TGF-β1 concentrations, possibly due to the premature platelet activation and TGF-β1 release during the second centrifugation step. Table 3 also shows that other authors [32,37] obtained preparations with high platelet counts, but low TGF-β1 concentrations in comparison to ours. According to Textor and Tablin [37], treatment with bovine thrombin as agonist or with 10% calcium chloride increased TGF-β1 concentrations, which indicated that TGF-β1 was released by platelet activation. Argüelles et al. [7] obtained a preparation with low platelet counts, but high TGF-β1 concentrations (after 10% calcium chloride activation), and the authors concluded that single centrifugation was "more efficient", meaning that less premature platelet activation occurred. This result is similar to ours.
Pereira et al. [35] compared seven different double-centrifugation protocols, which concentrated the platelets from 4.1 to 5.4 times, and the TGF-β1 concentrations achieved our PRP-1 values in only one of the protocols ( Table 3). The values that were obtained in all of these studies, as well as in the present report, were lower than those that were reported by Sutter et al. [8] by apheresis, followed by filtration.
As platelet activation is necessary to release growth factors, the avoidance of premature activation guarantees the presence of growth factors at the time of PRP use. Most of the growth factors have short half-lives after release (minutes to a few hours). If activated platelets are not used in an adequate timeframe, the growth factors become inactivated before application to the target tissue [43]. Consequently, premature platelet activation or a delay in use can result in poor therapeutic efficacy.
It should be emphasized that the procedures described here permit the preparation of PRPs with different characteristics, which could be useful in treating different clinical joint conditions. For example, PRP-0 may be useful to treat large joints, in which the injection volume is not an issue, while PRP-1, whereas more laborious and time-consuming, may be appropriate to treat small joints.

Conclusions
The present study has compared two procedures for preparation of PRP, one that is based on a single-centrifugation step and one based on double-centrifugation. In the single-centrifugation procedure, no premature platelet activation occurred, but platelets and TGF-β1 were less concentrated. The double-centrifugation protocol resulted in high platelet and TGF-β1 concentrations, but a certain degree of activation occurred, which was evidenced by both aggregometry tests and the presence of TGF-β1 in supernatant. All of our preparations were platelet-rich but not leukocyte-rich. Taken together, these results emphasize the need to carefully characterize the PRP obtained by any procedure, to produce reliable and reproducible results.