Implant stability in patients treated with platelet‐rich fibrin and bovine bone substitute for alveolar ridge preservation is associated with peripheral blood cells and coagulation factors

Abstract Aims The aim of the present study was to assess the association between dental implant stability and peripheral blood cell composition and levels of coagulation factors in patients treated with alveolar ridge preservation with platelet‐rich fibrin (PRF) and bovine bone substitute. Materials and methods Fifty patients were included between 2015 and 2017. PRF was prepared from autologous blood, in which blood cells and coagulation factor levels were measured. PRF and bovine bone were placed in the socket, followed by closure with PRF membrane. Implants were placed 14 (±2.5) weeks postextraction. The implant stability quotient was measured at t = 0, t = 10 days, t = 7 weeks, and t = 17 weeks by resonance frequency analysis. Results Erythrocyte count was inversely associated with PRF membrane length, but not with implant stability. Conversely, platelet count did not correlate with membrane size but inversely correlated with implant stability at 7 and 17 weeks. In addition, implant stability was directly correlated with levels FXIII (t = 0, p < .01), active von Willebrand factor (VWF; t = 0 and 7 weeks, p < .05), and total VWF (t = 7 weeks, p = .012). Conclusion Implant stability following alveolar ridge preservation with PRF and bovine bone substitute is associated with circulating blood cells and coagulation factors. In particular, fibrin structure, VWF, and FXIII may be important modulators of implant stability.

The use of PRF (with or without additional bone grafting material, such as deproteinized bovine bone mineral, DBBM) for ridge preservation has been reported previously. However, its efficacy has been evaluated with heterogeneous approaches, and results are inconsistent (Castro et al., 2017;Pan et al., 2019;Temmerman et al., 2016). Interestingly, it was observed that some patients have smaller and/or shorter PRF membranes than others (Mazzocca et al., 2012). This may in part be due to differences in red blood cell (RBC) content, as individuals with lower hematocrit have larger PRF membranes (Mazzocca et al., 2012;Miron et al., 2019). Indeed, the composition of the peripheral blood, including but not limited to RBC content, might influence the size and composition of PRF. Besides the cellular fraction, circulating coagulation factors may also influence the properties of the PRF and hence implant stability.
Therefore, in the current study, we investigated the association between implant stability and the peripheral blood cell composition and levels of coagulation factors closely involved in fibrin network formation and platelet incorporation.

| PRF treatment protocol
The study procedure consisted of two stages, namely, ridge preservation with PRF and implant placement, as described below. All procedures were performed under local anesthesia by the same implantologist (J. B.). Atraumatic extraction was performed to preserve as much bone as possible and avoid fracture of the buccal plate.
Periotomes and small luxators were used to remove the teeth. DBBM (on average 0.5-1 ml, Bego-Oss ® , BEGO Implant Systems GmbH & Co., Bremen, Germany), on average 0.5-1 ml, mixed with PRF (see Section 2.3) was placed into the socket until the desired vertical height was achieved. An envelope flap was created on the buccal and palatal side. A PRF membrane was placed in the envelope flap, followed by a mattress suture to fixate the membrane. No antibiotics or pain medications were prescribed before extraction and postoperatively. After a healing period of 14 ± 2.5 weeks, the implant site was prepared using a trephine drill with a diameter of 2 mm. Bone level implants (Bego RS or SC implants, BEGO) with diameters between 4.1 and 5.5 mm were placed. Insertion torque (N/cm) was determined using the Oral Implantcenter (Acteon, Bordeaux, France). All patients were followed 17 weeks after implantation.

| Analysis of blood cell and coagulation parameters
Peripheral blood samples were obtained by venipuncture (concurrently with blood for PRF preparation, prior to extraction) into two 2-ml EDTA tubes and two 2-ml 3.2% sodium citrate tubes (Vacuette, Greiner Bio-One, Kermsmunster, Austria). Blood cell counts (platelets, erythrocytes, hematocrit, and leukocytes) were measured using the hematology Cell-Dyn Sapphire analyzer (Abbot Diagnostics, Wiesbaden, Germany). Clotting times (PT and APTT) and the coagulation factors fibrinogen, FVIII and FXIII, were measured using the STA-Max analyzer (Stago, Asnières-sur-Seine, France). To measure von Willebrand factor (VWF) and active VWF, that is, VWF in its unfolded state able to bind platelets via the GPIb platelet receptor, we used inhouse sandwich enzyme-linked immunosorbent assays, as described earlier (van der Vorm et al., 2019). (Active) VWF levels are expressed as a percentage (%) of levels measured in normal pooled plasma.

| Measurement of implant stability
The stability of the implants was evaluated with resonance frequency analysis. The measurements were carried out with the Osstell device (Osstell, Göteborg, Sweden). The implant stability quotient (ISQ) has a range from 1 to 100, with a higher number indicating a more stable implant. ISQs were measured buccal/pallatinal (bp) and mesial/distal (md). Ideally, the same ISQ value is found from both directions, indicating that the bone-implant interface is the same around the implant.
However, if the bone is inhomogeneous, the implant can have different stability in different directions. Therefore, we provide both the bp and md ISQ values. ISQ measurements were performed immediately after surgery (t = 0, n = 38) and 10 days (n = 36), 7 weeks (n = 39), and Week 17 (n = 50) after implant placement. Implant stability, measured as the ISQ, was determined at several time points over the course of the study (Table 1 and  Together, these data confirm the good primary and secondary stability of the dental implants.

| Effect of peripheral blood cell composition on PRF membrane
We hypothesized that peripheral blood cell composition and coagulation factor levels may influence PRF membrane characteristics (Table 2). A significant inverse correlation between erythrocyte count and PRF membrane length (n = 41) was observed: PRF membranes prepared from patients with a higher erythrocyte count in peripheral blood (and higher hematocrit) resulted in shorter PRF membranes than patients with a lower erythrocyte count (Figure 2a,b). Conversely, there were no significant associations between platelets ( Figure 2c) and PRF membrane size parameters. FVIII and active VWF levels correlated significantly (p = .024 and p = .033, respectively) with PRF membrane area (n = 17; Figure 2d,e), whereas fibrinogen levels did not correlate with PRF membrane area (Figure 2f) or length.

| Effect of peripheral blood cell composition on implant stability
Because we found that specific cellular and coagulation parameters influence PRF characteristics, we were interested in the effect of the composition of peripheral blood on implant stability. Platelet count was significantly inversely correlated with bp ( Figure 3a) and md ( Figure 3b) ISQ scores at 7 and 17 weeks, whereas erythrocytes were not significantly correlated with implant stability. At 7 weeks postimplantation, higher implant stability was associated with lower leukocyte count (Figure 3c), but this was not observed at 17 weeks.
Significant correlations between implant stability and a number of coagulation parameters were observed, namely, between ISQ at t = 0 and FXIII ( Figure 4a) and active VWF (only bp, Figure 4b) and between ISQ at t = 7 weeks and active VWF ( Figure 4b) and total VWF (only bp, Figure 4c).

| DISCUSSION
In the present study, we analyzed the stability of nonimmediate implants placed following ridge preservation with a mixture of autologous PRF and DBBM xenograft. We found several interesting correlations between implant stability, peripheral blood cell counts, and the coagulation factors (active) VWF and FXIII.
Blood composition may have dual effects on wound healing and tissue regeneration, namely, a direct effect on the composition of the (PRF) clot and an indirect effect on the circulation of the wound area.
Conversely, the underlying condition requiring tooth extraction (e.g., periodontitis) may influence peripheral blood cell composition and coagulation status. We found several interesting correlations between blood cell counts and coagulation factor levels on the one    this effect is mediated through reduced thrombin generation (Wolberg, 2007). Although none of our patients suffered from hemophilia, it could be hypothesized that lower FVIII levels affect the fibrin structure in the PRF clot to produce a smaller membrane. Regarding active VWF, one of the key functions of VWF is to tether platelets to the exposed subendothelial collagen after vessel wall damage. Circulating VWF can only exert this function after conversion to its active conformation, which is induced among others by shear stress (Huizinga et al., 2002). In healthy individuals, only a minute amount of VWF circulates in its active conformation, but nevertheless, there is an interindividual variation of around 15% (van der Vorm et al., 2019).
During clot formation, functionally active VWF readily incorporates into the fibrin network and subsequently supports platelet adhesion (Miszta et al., 2014). Also, the presence of VWF results in the formation of a less dense fibrin network (Miszta et al., 2014). Thus, during PRF clot formation, higher active VWF levels may also facilitate binding of more platelets and may result in a more loose fibrin network and, hence, a larger membrane area. Of note, no significant correlation between membrane size and fibrinogen levels was observed, probably because fibrinogen circulates in excess amounts and hence does not limit the size of the PRF membrane.
More of clinical interest is the effect of these peripheral blood parameters on implant stability. Surprisingly, we found that peripheral platelet and leukocyte counts were inversely associated with secondary implant stability (at 7 weeks postimplantation). This finding is counterintuitive as both platelets and leukocytes are considered essential for tissue regeneration and osteogenesis by the release of growth factors and by supporting angiogenesis and lymphogenesis (Soloviev et al., 2014). Our findings challenge this view on the cellular fraction of PRF as the key elements responsible for the clinical efficacy. However, future studies quantifying cell counts as well as growth factor and inhibitor levels within the clot itself are required to support these findings. Similar to our findings on membrane size, active VWF was also associated with primary implant stability and both active VWF and total VWF levels correlated with secondary stability at 7 weeks. As mentioned, (active) VWF may contribute to both  Figure 4b). bp, buccal/palatal; md, mesial/distal; NPP, normal pooled plasma; VWF, von Willebrand factor; w, weeks platelet incorporation in the fibrin mesh and to fibrin network structure (Miszta et al., 2014). Moreover, activated VWF may indirectly affect implant stability by supporting primary wound healing (Miszta et al., 2014). Levels of FXIII, the coagulation factor responsible for fibrin crosslinking, also correlated with primary implant stability. FXIII is known to be involved in wound healing, as demonstrated by impaired wound repair in FXIII deficient mice (Inbal et al., 2005). Thus, circulating FXIII levels may both directly (through fibrin crosslinking) and indirectly (through wound healing) affect implant stability. Altogether, our data suggest that the blood composition and the fibrin structure of PRF may be critical modulators of implant stability and require more emphasis in the future, more detailed studies.
The most important limitation of our study is the lack of randomization to a PRF treatment group and a control group treated with only DBBM, or alternatively a split-mouth model with these two treat-  (Molly, 2006). However, the lack of a control group in the current study already precludes drawing conclusions on the effect of the combination graft of PRF and DBBM on bone density and implant stability; hence, in further studies, both a control group and histological analysis should be included.
In addition, we are fully aware that association is not the same as causation and that a large variety of patient-related factors can potentially confound the observed correlations between hematological parameters and implant stability. For instance, age is known to influence coagulation, and older age is associated with higher VWF levels (Favaloro et al., 2005). Although our inclusion criterium was to include all patients >18 years, the majority (43/50, 86%) of our patients was older than 50 years. This may explain why we see no significant associations between age and implant stability itself nor with any of the hematological parameters associated with implant stability. Furthermore, implant stability was not significantly different between men and women or between smokers and nonsmokers. Although the negative effects of smoking on implant stability have been described extensively in literature (Kasat & Ladda, 2012), the number of smokers in the current study is likely too small to detect a significant difference. Likewise, subgroups with comorbidities were too small to perform meaningful statistical analysis. Larger studies are required to assess possible confounding effects of these patient-related factors on implant stability.
In conclusion, implant stability, following alveolar ridge preservation with PRF and bovine bone substitute, is associated with several hematological parameters. Our results suggest that fibrin structure and levels of (active) VWF and FXIII, more than platelet and leukocyte count, may be determining factors for PRF membrane characteristics and implant stability.