This research describes a reliable method for producing platelet concentrates and consequently for concentrating GF, such as TGF-β1 from canine blood. Some manual (tube) protocols [19–22] and a semi-automated method  have been described for producing PC in dogs. Manual protocols were performed with either sodium citrate [19–21] or ACD-A  as anticoagulant. Those protocols included simple and double centrifugation steps for concentrating between 400 X 103 to 1300 X 103 PLT/μL. However, platelets were counted manually (light microscopy) and did not report additional hematologic features of the resulting PC.
Platelet concentration reached in the protocol described here was slightly lower than a semi-automated double centrifugation method evaluated in dogs, which presented a median concentration of 1336 PLT X 103/μL . However, this method is limited because it could only be used in medium at large breed dogs, since it requires 60 mL of blood for PC preparation. The protocol described here presents the advantage that PC is easily obtained by using one centrifugation step with a small volume of blood. This last situation is important when pediatric patients or small breed dogs are treated.
The size and weight of the blood cells, the relative centrifugation forces (g) and time are factors that determine the cellular and molecular characteristics of a PC. This concept is necessary for comparing the results of the research described here with other published studies in human beings [12–15] and horses [16–18]. The protocol described here permitted obtaining two kinds of different PC. Platelet concentrate-A presented a higher concentration of PLT/μL, WBC/μL and TGF-β1/mL (independently of the activating substance used) in comparison with PC-B. From a comparative point of view both canine and human blood present the same trend when centrifuged for PC preparation. However, equine blood requires a double centrifugation  for obtaining a PC with an acceptable quantity of PLT.
There are a lot of controversies about the ideal number of concentrated platelets in a PC for its clinical use in human beings and horses. Some researches consider the highest number of platelets concentrated in PRP to yield the best clinical results. This assumption could emerge from experimental results observed in rabbits where higher platelet concentrations were better for osteointegration, than lower platelet concentration . However, excellent clinical results have been observed in human beings  and horses  using PC with 300–400 X 103 PLT/μL.
Another fact that generates controversy is the presence of leukocytes in PC. Some researches consider that WBC are a contaminant of PRP and possibly deleterious for tissues when treated, especially at high concentrations. However, others believe that WBC are important regulatory cells contained in PRP and necessary for wound healing . To date, there are no studies that definitively elucidate the significance of leukocytes in PRP. However, when manual methods are used for producing PC in human beings and horses, leukocyte concentrations are comparatively lower when semi-automated methods are used. Unfortunately, there is no data about WBC concentration in canine PC obtained by manual or semi-automated methods. However, it is important to note that platelet concentration of the protocol described here was not correlated with WBC concentration. This situation is different for PC derived from equine blood .
Both PC obtained in this study permitted concentrate two (PC-B) and three (PC-A) fold the concentration of TGF-β1 respect to the basal concentration of this protein in plasma. These findings suggest that plasma platelets were scarcely activated at the moment of blood extraction, since the MPV remained lower in whole blood in comparison with the same parameter in PC-A. However, although MPV value was statistically higher in PC-A in comparison with PC-B and whole blood, this platelet activation parameter remained between normal range values for canine platelets . Maybe, the reason, which MPV was higher in PC-A, is related to the large number of concentrated PLT and WBC. A single centrifugation process produces cellular friction that could be most active toward the platelet fraction near to the erythrocyte package. This situation has also been observed in equine PC obtained by simple and double centrifugation tube methods .
Plasma TGF-β1 concentrations of this study were quite similar to the values described for this protein from serum of two dogs (16.5 and 19.9 ng/mL) . However, other research described plasma TGF-β1 concentration ranging from 0.193-0.598 ng/mL. That study included 29 canine blood samples collected with EDTA by using a double centrifugation protocol . Plasma or serum TGF-β1 concentrations described in these two last studies [29, 30] could be influenced by methodological aspects such as the use or not of anticoagulant and the type of antibody used for TGF-β1 measurement. To note, this protein was measured in the present study with a specific canine antibody for TGF-β1, whereas the studies mentioned [29, 30] used a human TGF-β1 antibody. However, this is only an assumption and further studies are necessary to validate the actual utility of human or canine ELISA kits for canine TGF-β1 measurement.
Plasma and both PC of this study presented higher concentrations of TGF-β1 in comparison with the results of the same protein from autologous conditioned plasma (ACP) and plasma obtained with ACD-A from blood of dogs . In that research , a TGF-β1 mean concentration of 1.24 ± 0.59 ng/mL was obtained from ACP with PLT counts ranging from 277–293.5 X 103/μL. However, the ELISA human kit used did not detect plasma TGF-β1 concentration. In addition, no statistical correlation was found between the number of PLT concentrated and the TGF-β1 concentrations . This last result was different from the findings of the study described here; since strong correlations (70%) were noticed between PLT counts and TGF-β1 concentration. The difference between the results obtained in that study  and the findings of the research presented here could be related to the specificity of the antibody used for TGF-β1 detection and because they used no activating substances for stimulating the release of growth factors from platelets.
There is some controversy about the need of adding activating substances to induce the release of GF contained in PLT. Some researchers think that activating substances are not necessary when PC will be used as an injection for the treatment of tendon and ligament lesions or arthropaties . They argue that connective tissues are rich in collagen and that this autologous protein is enough to induce platelet activation. Other researchers believe that the use of activating substances is necessary to stimulate the massive release of growth factors in the foci of the lesion and thus increasing the healing process of the affected tissue . However, to date there is no scientific information to determine if activating substances should be used before PC injection.
Platelet activating substances are a necessary pre-requisite for producing platelet gel from PC. Platelet gels are used for covering large skin defects or for filling bone defects either alone or combined with other biomaterials. Platelet gels could be produced from PC activated with proteic and non-proteic substances or by combination of both. Classically, bovine thrombin (either alone or in combination with a calcium salt) has been used for platelet gel production. Thrombin induces fibrin polymerization by removing fibrinopeptides A and B from the fibrin molecule. This protein also produces platelet activation and massive release of growth factors. Some clinicians prefer not to use bovine thrombin for platelet gel formation because this substance induces cross-reacting antibodies against coagulation factors V and XI .
Batroxobin induces fibrin polymerization by removing only the fibrinopeptide A. This substance does not induce platelet activation or immunological reactions against coagulation factors . The manufacturer recommends the use of batroxobin plus calcium gluconate for platelet gel formation. In the study presented here no differences were noticed on the release of TGF-β1 from both PC activated with CG or batroxobin plus CG. Macroscopically no differences were noted about the quality of the platelet gel formed or the time required for clot formation. However, one limitation of this study is that kinetics of gelation was not performed .
Results of this study corroborate that batroxobin induce negligible platelet activation and the addition of CG was necessary to induce the TGF-β1 release. This phenomenon could be explained without the need of using an experimental group of PC activated only with batroxobin, since TGF-β1 concentration was statistically similar for both PC, independently of the activating substance used, CG or batroxobin plus CG. This study is limited since TGF-β1 release only was measured at once. Further studies are necessary for knowing if the release kinetics of this growth factor is time dependent or is related to an activating substance in particular.
Collection and concentration efficiencies are two important aspects related with the capacity of and protocol or device for concentrating the most possible number of platelets and growth factors from a whole blood sample. The protocol described here for producing PC and consequently concentrating TGF-β1 was better than the cellular and molecular results obtained for canine ACP protocol  and the platelet concentration efficiency described for a semi-automated method for producing canine PRP .