Following tooth removal bone formation normally takes 16 weeks and may result in less than adequate volume for the necessary reconstruction. Platelet rich plasma (PRP) has been promoted as an effective method for improving bone formation. Its use is often expensive, time consuming, or not clinically convenient for the patient and/or clinician. This study examines a simple method for obtaining a “Buffy Coat”-PRP (BC-PRP) and its effect on bone healing following the removal of bilateral mandibular 3rd molars. Subtraction digital radiography and CT scan analysis were used to track changes in radiographic density at PRP treated sites in comparison to ipsilateral non-PRP treated sites. PRP treated sites demonstrated early and significant increased radiographic density over baseline measurements following tooth removal. The greatest benefit of PRP is during the initial 2-week postoperative healing time period (P < .001). During weeks 3 though 12, BC-PRP treatment resulted in significant (P < .0001) increases in bone density compared to control, but there was no significant interaction between time and treatment (P > .05). For the entire time period (0–25 weeks) PRP treatment was significant (P < .0001) and time was significant (P < .0001) but there was no significant interaction (P > .05) between the effect of PRP treatment and time. It required 6 weeks for control extraction sites to reach comparable bone density that PRP treated sites achieved at week 1. Postoperative pain, bleeding, and numbness were not significantly affected by BC-PRP application. Results suggest that this simple technique may be of value to clinicians performing oral surgery by facilitating bone regeneration following tooth extraction.
Immediately following tooth removal, a healing process begins that affects the eventual alveolar bone volume and architecture of the alveolar ridge. Satisfactory and timely healing are essential to obtain ideal functional reconstruction. Traumatic removal of a tooth, or a poor healing response, may lead to excessive bone loss delaying tooth replacement, necessitate expensive and time-consuming reconstructive surgeries, or even be impossible to correct.1 Patients and clinicians could benefit if a cost-effective, simple technique were available that decreased bone-healing time and increased the predictability of favorable results. The objective of this study was to determine whether the application of platelet rich plasma (PRP) to a tooth extraction site would facilitate bone healing.
Following tooth extraction, a cascade of inflammatory reactions immediately begins, and the extraction socket is temporarily closed by clotting blood. Epithelial tissue proliferation and migration start within the first week and tissue integrity is quickly restored. Histological evidence of active bone formation in the extraction socket is seen as early as 2 weeks following extraction. The socket is filled with newly formed bone in about 6 months.2,3 If the removed tooth is not replaced by an implant within 6–12 months, the residual ridge alveolar bone undergoes permanent catabolic remodeling.2,4,5 If a clinically acceptable method existed to consistently decrease healing time, both patients and clinicians would benefit. The objective of this study was to evaluate the effect of locally administered concentrated PRP, obtained by a “Buffy Coat” technique (BC-PRP),6 on gingival healing and bone formation following bilateral mandibular third molar extractions.
Radiographic techniques have been used to document the gross morphologic changes of the alveolar processes after extraction.1,7,8 This study used a variation of a well-established method for detection of subtle bone changes using a subtraction radiography technique for regions of interest (ROI).9–13 The ROIs were interpreted using a computer program that examined pixel number and converted to grayscale images.
Macro- and microstructural bone formation was also examined by computerized tomogram scanning technology (CT scan) that provided information on bone healing using a three-dimensional representation of the object14 and can elucidate the skeletal response to innovative therapies.15–21 Since this technology is expensive and requires more radiation exposure than digital radiography, it was used on only 3 patients in this study.
Materials and Methods
All patients (n = 6) selected for the study were between ages 18 and 40 with an unremarkable medical history and were nonsmokers. Patients were informed individually of possible complications and voluntarily signed study, surgical, and IV sedation informed consent forms. Patients received financial remuneration at the completion of the 25 week evaluation period.
All bilateral mandibular 3rd molar extractions were of teeth with similar eruptive states. Each patient had 1 extraction site treated with PRP, while the contra-lateral side did not receive PRP. Three patients received PRP on the right side and 3 received PRP on the left side. Treatment side was predetermined as follows: the first, third, and fifth patients received PRP on the left side and the second, fourth, and sixth patients received PRP on the right side. This decision was made prior to viewing patient radiographs, and position in the sequence was based on their order of scheduling.
A venous puncture using a 21 gauge 1.5 inch latex-free needle (EXELINT Int. Co, Los Angeles, Calif) with an attached vacutainer holder (EXELINT Int. Co) in the ante-cubital fossa, or the dorsal surface of the hand contra-lateral to the IV administration site was performed. Whole blood was drawn using two 4.5 mL BD Vacutainer tubes containing 0.45 mL of the anticoagulant tri-sodium citrate (9 ∶ 1) (Becton Dickinson & Co, Franklin Lakes, NJ). The Vacutainer tubes were centrifuged for 10 minutes at 1150 × g. After centrifugation, the tubes were removed and placed in a test tube rack where the red blood cell/plasma interface was allowed to set for 3 minutes. Premade labels were placed on the outside of the tubes to delineate the PRP layer to be harvested, with a center dotted line positioned at the blood cell/plasma interface, a solid line 3 mm above, and a second solid line 2 mm below (Figure 1). Figure 2 displays the label on a Vacutainer tube filled with water (for illustration purposes only).
The upper plasma layer (platelet poor plasma [PPP]) was aspirated to the 3 mm mark (top solid line) using a sterile beveled 20-gauge 1.5 inch BD Precision Glide needle (Becton Dickinson & Co) attached to a 3 mL BD latex-free syringe (Becton Dickinson & Co). The PRP was collected from between the upper 3 mm mark and the lower 2 mm mark using a separate beveled 20-gauge 1.5 inch BD Precision Glide needle attached to a 3 mL BD latex-free syringe.
Third molar surgical technique and PRP placement
All patients underwent conscious IV sedation with a combination of midazolam HCl, diphenhydramine HCl, and fentanyl citrate (Ace Surgical, Brockton, Mass) as needed to induce a state of conscious sedation and monitored with electrocardiogram, automatic blood pressure, and percent hemoglobin oxygen saturation. Once the patient was sedated, the blood draw was performed at a separate intravenous puncture site to acquire the PRP for use after the teeth were removed. Depending on allergies, patients were administered either cefazolin 2.0 gram (Baxter Healthcare Co, Deerfield, Ill) or clindamycin 600 mg (Ace Surgical) immediately preoperative via IV. Following sedation and administration of antibiotics, all patients were administered lidocaine HCl 2% with epinephrine 1 ∶ 100 000 (Cook-Waite, Rochester, NY) bilaterally as a Gow-Gates inferior alveolar block, buccal infiltration and lingual block. Incisions were made in the envelope fashion. The gingival softtissue was reflected exposing bone overlaying the impacted tooth. Bone was removed by the use of an Elcomed 100 electric handpiece (W&H Dentalwerk, Burmoos, GMbH, Austria) and a #8 round surgical bur (Henry Schein Inc, Melville, NY). Teeth were sectioned as needed with a 45° high impact handpiece with a #700 Fisher bur (Henry Schein Inc). The control extraction site was treated immediately postextraction as follows:
placement of Gelfoam (Pharmacia Corporation, Kalamazoo, Mich) only (no PRP);
primary closure was achieved with 3-0 chromic gut suture material (Henry Schein Inc); and
after removal of the tooth, a periapical digital radiograph (Dentsply Gendex, York, Pa) was taken using an XCP-DS mount to determine a baseline grayscale representing bone density at each extraction site.
The BC-PRP treatment site was treated immediately postextraction as follows:
injection of 200 µL of PRP into the extraction site;
placement of Gelfoam, moistened with 175 µL PRP;
primary closure was achieved with 3-0 chromic gut suture material; and
after removal of the tooth, a small periapical digital radiograph was taken using an XCP-DS mount to determine a baseline grayscale representing bone density at each extraction site.
To help prevent excessive postoperative inflammation, patients were administered dexamethasone sodium phosphate (Ace Surgical) 8 mg IV preoperatively and methylprednisolone suspension Depo-medrol (Pharmacia & Upjohn Co, New York, NY) suspension 40 mg IV immediately post-op. Patients were also administered ketorolac trimethamine 30 mg (Hospira, Lake Forest, Ill) IV immediately post-op. Immediately postoperative bilateral Gow-Gates mandibular block injection using 1.8 mL of bupivacaine HCl 0.5% with 1 ∶ 200 000 epinephrine (Hospira) was administered to prevent immediate post-op pain. Patients were advised to take ibuprofen 600 mg every 6 hours for 3 days and then every 6 hours as needed thereafter. They were advised to take acetaminophen 500–1000 mg every 6 hours if the ibuprofen did not control the pain.22,23 Patients were dismissed with a grade card for notations every 8 hours for 3 days, then notations every 12 hours for the following 6 days. Each surgical site was graded with a visual analogue scale (VAS) for: pain, temperature (external feeling of warmth), facial edema, bleeding, and numbness or altered sensation of tongue, face, lip, or chin.
The patients returned postoperatively for observer evaluations plus digital radiographs using the same protocol for immediate postoperative radiographs. These evaluations were repeated at the following postoperative time points: 3 days plus weeks 1, 2, 3, 4, 6, 8, 12, 16, 20, and 24. Blinded observers used a VAS to evaluate the following: dehiscence, bleeding, inflammation, facial edema, and pain.
Three of the patients had a mandibular CT scan performed by a local hospital. Each of these 3 patients had 1 CT scan taken. Patient-1 had theirs taken at 18 weeks postoperative, Patient-2 at 14 weeks postoperative, and Patient-3 at 10.5 weeks postoperative. This allowed for verification of the grayscale from digital radiography with bone density Hounsfield units (HUs) of the CT scan and percent bone fill of the extraction sites.
Radiographic technique and exposure
Radiographs were taken with a Trophy ETX Dental periapical X-ray machine (Trophy Radiologie, Vincennes, Cedex, France). X-radiation settings were 70 KV, 8 mA, for 0.05 second exposure time. The digital x-ray sensor used was a Gendex GX-S sensor (Gendex Dental System, Lake Zurich, Ill). The radiograph sensor was held by a Dentsply Sensor Holder XCP-DS posterior holder and ring (Dentsply International, York, Pa). Digital radiographs were formatted by VinWix Pro computer software (Gendex Dental System).
CT scan technique and exposure
Mandibles were scanned 1 mm inferior to the mandible to the superior surface of the mandibular teeth. CT specifics were as follows: image size: 512 × 512 pixels; gantry tilt: 0.0 degrees; scanner slice thickness: 0.5–1.0 mm; scanner step increment: 0.5–1.0 mm; and field of view: typically 150–180 mm. Images were taken in uncompressed DICOM 3.0 format. KV and MA were kept as low as possible with a table pitch 1 ∶ 1. CT scans were reformatted for quantitative measurement using Logic VIPTM Data Conversion software.
Digital radiographs were used to evaluate changes in radiographic bone density for each extraction site. Three blinded dental professionals working independently evaluated all radiographs.
Image J software (http://rsb.info.nih.gov/ij/) was used for digital radiograph analysis. The radiographs were assessed by obtaining the average density of 3 independent readings of the 3rd molar extraction socket sites. This was compared to the average of 3 density outlines of the adjacent tooth. When all radiographs for a patient were assessed, they were normalized to the original radiograph and the same untouched adjacent tooth. The baseline socket average was then subtracted from the normalized average for each tooth extraction socket at the different time points. This accounted for minor exposure and technique variation with each time point. The final, normalized socket value differences for each PRP-treated and -nontreated site for the various time points were compared.
All radiographs were taken by a blinded, certified x-ray dental assistant or hygienist. They were taken immediately post-op and at the following times after the day of surgery: 3 days, 1 week, 2 weeks, 4 weeks, 6 weeks, 12 weeks, 16 weeks, 20 weeks, and 24 weeks. Radiographs were evaluated by 3 blinded dental professionals.
CT scan analysis
CT scans were evaluated for percent bone fill, newly formed bone density (NFBD), and total socket formed bone density (TFBD). CT scans were reformatted with Logic VIPTM Data Conversion software, which permitted measurements at 1 mm increments of the extraction socket. Percent bone fill was determined by measuring the quantity of bone that had formed in the socket, dividing this value by the entire socket depth, and multiplying by 100%.
NFBD was measured in Hounsfield units (HU) with an elliptical selection. Areas of new bone formation were outlined and quantified.
TFBD was a measurement in HUs for the entire socket, regardless of the quantity of bone formed. This measurement was done in a similar manner to NFBD, the difference being that the entire height of the socket was considered, not just the areas that had bone formation. This would more accurately represent the same areas measured by digital radiographs, and allow for comparison of the two methods (not shown).
CT scans for Patients-1 and -2 were measured in 1 mm cross section planes (anterior to posterior). The CT scan for Patient-3 was measured in panoramic 1 mm planes. The cross section planes were preferred, but the topography of the bone after tooth removal for Patient-3 did not permit accurate measurements with those projections. CT scans were evaluated by 3 blinded evaluators working independently.
Statistical analysis was performed using GraphPad Prism 3.0 (GraphPad Software, Inc, La Jolla, Calif). Two-way analysis of variance (ANOVA) analysis was performed for digital radiographs taken over 0–2 weeks, 3–12 weeks, 3–25 weeks, and 0–25 weeks time-periods. Two-tailed paired t tests were performed for individual time points with digital radiographs and CT scan analysis.
Digital radiographic results
Radiographic analysis of the digital radiographs taken for the 6 patients over the 25 week period show a significant (P < .0001) increase in bone density in the socket treated with PRP. The data represent the percent change in radiographic density for each socket and each time point from corresponding baseline values immediately following tooth removal and placement of the Gelfoam and sutures. Figure 3 illustrates the mean densities for all patients. The PRP treatment had a positive effect on bone density immediately following tooth extraction, whereas the control side showed a decrease in bone density compared to baseline for the initial week which subsequently returned to baseline by week 2. Weeks 3, 4, 6, and 8 showed parallel increases in density in control and PRP treated sockets from their respective baselines. Neither weeks 16 nor 25 (P > .05) showed significant differences in bone density between treated and control sockets. The PRP treatment group had greater bone density changes over baseline immediately following extraction, but the differences between the groups were not significant at later time points with the exception of weeks 12 and 20, which may be due to radiographic variation. When examining the entire 25-week period, PRP treatment was significant (F = 37.49, P < .0001), time elapsed following surgery was significant (F = 13.75, P < .0001) but there was no interaction between PRP treatment and time (F = 0.6975, P > .05).
CT scan results
Three factors were evaluated for each CT scan: percent bone fill of the extraction socket, NFBD, and TFBD.
When comparing all patients and their respective time points together, two-way ANOVA revealed that the PRP sites did show significantly (F = 19.76, P < .05) greater percent bone fill than the control sites (Figure 4a). Time was also significant (F = 185.3, P < .01).
NFBD was greater at later time points than the earlier 10.5-week time point. Patient-2 had denser NFBD at the 14-week time point as compared with Patient-1 at the 18-week time point, but this may be due to individual patient variation or measurement variables. When comparing NFBD for all patient treatments (Figure 4b) and time points using two-way ANOVA, the PRP-treated and control sites did not differ significantly (F = 0.01475.134, P > .05). Time was not a significant (F = 13.19, P > .05) factor.
Patient-3 (10.5 weeks) (Figure 4c) with the earliest CT scan had the lowest TFBD, while Patient-2 (14 weeks) had a slightly higher value than Patient-1 (18 weeks). When comparing all of the patients together (Figure 4d) using two-way ANOVA, the PRP-treated sites did not show a significant increase (F = 0.2557, P > .05) in TFBD, but time elapsed since tooth removal was a significant (F = 96.12, P > .0001) factor.
Patients' evaluation results
Patients were asked to postoperatively rate their perception of pain, bleeding, numbness, facial edema, and temperature using a VAS (Table 1). When using a two-tailed paired t test for each specific time point there was no statistical difference (P > .05) between the PRP treatment side and the control side for any of the parameters.
Observers' evaluation results
Blinded clinical observers evaluated the degree of tissue opening (dehiscence), bleeding, inflammation, intra-oral edema, perceived pain, and facial edema using a VAS (Table 2). When using a two-tailed paired t test for each specific time point, there was no statistical difference (P > .05) between the PRP treatment side and the control side for any of the factors.
Observers' evaluation of facial edema analyzed with two-way ANOVA for the entire postoperative period did show a significant treatment effect (F = 4.416, P < .05), and time was extremely significant (F = 8.986, P < .0001). However, there was not a significant interaction between treatment and time for facial edema (F = .639, P > .05).
This prospective study offers insights into what is occurring when BC-PRP is applied to fresh extraction sites. Third molar extractions are often used as a measurement tool for comparing treatments because they are usually performed electively on a younger population that do not present with multiple confounding factors (eg, systemic pathologies, multiple medications). Therefore, this study is especially relevant for healthy 18 to 40 year old patients.
Intra-oral digital radiographs taken of the individual surgical sites revealed that the effects of PRP were significantly beneficial (P < .05) for increasing bone density following surgery (Figure 3). The increase in bone density suggests a greater volume of new bone formation with PRP treatment. Moreover, the increase in bone density (presumed as increased volume of new bone formation) was found to occur at earlier time points than non-PRP treated control sites.
Of note was the immediate increase in grayscale readings which indicates enhanced early bone formation. This corresponds with results from the Lucarelli molecular study that demonstrated PRP treatment of human mesenchymal stem cells induced early proliferation of these cells and possibly differentiation into osteoblasts.24 Accelerated bone formation is in contrast to the drop in bone density (representing bone loss) seen at the control site before bone formation began to take place. It took approximately 6 weeks for the control sites to reach the same bone density that the PRP-treated site had reached by week 1 (Figure 3). The PRP-induced acceleration in bone formation may be due to the presence of bone morphogenetic proteins (BMPs)-2 and -6 in PRP that stimulates mesenchymal stem cells to begin osteoblast differentiation and subsequent calcification.25,26 Healing at control sites would be dependent upon initial osteoclast activity to break down existing bone thereby releasing BMPs. The findings of this study correspond with the known, accepted bone repair timeline.27 The immediate start of bone formation seen with PRP treatment is of clinical relevance because it is the initial 2 weeks following bone-manipulation oral surgery that are important in preventing infection, loss of the blood clot and/or AO (dry socket) formation. There was a slight drop in grayscale density at the 1.5 to 2 week period, suggesting that the initial direct impact of BC-PRP treatment might end or decrease by 1.5 to 2 weeks.
Further examination of the digital radiographic density changes shows that subsequent changes in bone density with the PRP-treated and control sites are parallel, with no significant differences, except for the 12- and 20-week time points. The finding of parallel increases in bone density in later weeks most likely represents normal healing taking place at both sites, with greater bone density for the PRP-treated sites due to earlier, more rapid bone formation. The significant differences seen at 12 and 20 weeks may be due to bone remodeling or inherent variations in the radiographic evaluation.
The control sites required 16 weeks to reach the same degree of radiographic density for the total extraction socket as the PRP treatment achieved in 8 weeks. This is of clinical significance because it indicates more rapid healing at the PRP-treated sites. Patients who undergo complete maxillary and/or manidular extractions and are in need of immediate prostheses can benefit from PRP treatment. The patient may be given an initial temporary denture and the definitive prosthesis fabricated at the 2 to 3 month time point, returning the patient to full function in a shorter time period. The decrease in bone healing time could also apply to the placement of dental implants, which normally requires 4 to 6 months. Excepting those cases that call for immediate implant placement, PRP treatment at the time of tooth extraction may permit dental implants to be placed at a 2 to 4 month time interval, decreasing the time to implant placement by half.
This study used noninvasive radiographic techniques. Bone histology studies could have provided additional information, but that analysis technique is invasive. Digital panoramic radiographs might have been a better choice than digital periapical radiographs, as the 2 surgical sites would be represented on 1 film, eliminating the need for normalization between the sites. Normalization between the radiographic series and baseline radiographs would still have been necessary. The disadvantage of digital panoramic radiographs is that these films have a 20 to 25% distortion factor because it is an extra-oral film, as opposed to the intra-oral individual (periapical) radiographs. The periapical radiograph was chosen over the panoramic radiograph due to the lower distortion, which results from being in close proximity to the site being evaluated.
The CT scans might have exhibited greater differences between the PRP treated sites and the control sites, had they been obtained at earlier time points. Based on these findings and knowing how the effects of PRP treatment are greatest during the initial 2 weeks, it would have been beneficial to acquire CT scans at earlier time points. Future studies by the authors will have the CT scans taken at earlier time points to confirm the early effects of PRP treatment.
The adjuvant medications used perioperatively most probably reduced the postoperative pain, which may explain why the patients did not report high pain ratings. The use of bupivacaine local anesthesia would explain the higher numbness rating for the initial 24 hours. The administration of glucocorticosteroids and the patient's use of ice packs may also have contributed to the lower ratings for edema. Also, the patients' use of NSAIDs (ketorolac and ibuprofen) may have helped keep temperature elevations to a minimum.
The administration of dexamethasone, methylprednisolone, ketorolac, and ibuprofen interfere with inflammation and therefore may impact bone healing.28,29 During the early phase of bone formation there are increased concentrations of pro-inflammatory cytokines, mitogenic growth factors, and cell differentiating factors. Inflammation, which occurs at the onset of early phase injury, or surgically induced bone formation, is hypothesized to provide the initial signaling molecules for continuation of the healing cascade.30,31
Additionally, these same medications were administered for a short time period because of their analgesic and anti-inflammatory efficacy.32–36 It is thought that inflammation is necessary for normal bone healing and formation; therefore, postoperative NSAID therapy may potentially delay bone healing. However, PRP treatment may offset this inhibition. Figure 5 offers a possible mechanism for the beneficial effects of PRP that can compensate the negative bone healing effects of NSAIDs. PRP contains platelets and monocytes, which both contain the cytokine interleukin-1α (IL-1α).37,38 IL-1α is involved in the local regulation of bone homoeostasis.39 IL-1α also stimulates prostaglandin E2 (PGE2) formation.39–43 This stimulation would counter the inhibition of PGE2 synthesis produced by NSAID and glucocorticoid steroid administration. Transient release of IL-1α has also been shown to be a potent stimulator of bone formation.44,45 The platelets in PRP also contain several growth factors (platelet-derived growth factor AA [PDGF]AA, PDGF-BB, insulin-like growth factor [ILGF], transforming growth factor β1 [TGF-β1], vascular endothelial growth factor [VEGF], and osteoprotegrin [OPG]) that work in concert in bone development and repair. Therefore, PRP use would likely counter PGE2 inhibition induced by short-term use of the NSAID/glucocorticoid combination.
An alternative to glucocorticosteroids and NSAIDs for postoperative pain control would be acetaminophen and/or narcotics. Because acetaminophen is not anti-inflammatory, patients would experience greater inflammatory effects, such as edema and pain.46 Narcotics are often fraught with unwanted side effects such as decreased mental acuity, lethargy, and gastrointestinal disturbance.47 Therefore, these two analgesics have little advantage over NSAIDs.
When examining the study data, it appears that PRP accelerated bone formation and decreased the time necessary to return to full function. Other factors could have affected pain perception, bleeding, and numbness. Therefore, it is difficult to confirm PRP's effect on these factors. Blinded observers did report significantly less inflammation and facial edema (P < .05) on the sides receiving PRP. However, a larger study with fewer confounding factors would be necessary to confirm the effect of PRP on inflammation and subsequent edema.
The results of this study suggest that the use of a simple, cost-effective BC-PRP method to increase the rate of bone formation and decrease healing time in the initial 2 weeks following oral surgery may be beneficial. Treatment with PRP may not affect pain, bleeding, and/or numbness, but may decrease inflammation.
American Academy of Implant Dentistry
“Buffy Coat” technique
bone morphogenic proteins
insulin-like growth factor
newly formed bone density
platelet-derived growth factor
platelet rich plasma
red blood cell
regions of interest
total socket formed bone density
transforming growth factor β1
visual analogue scale
vascular endothelial growth factor
white blood cells
This study was reviewed and approved by the Investigational Review Board for the Protection of Human Subjects at Duquesne University, Pittsburgh, Pa. The American Academy of Implant Dentistry (AAID) Research Foundation provided funding for this study. The authors would like to express appreciation to the following individuals who helped with this clinical study: Deborah L. Rutkowski; Charles T. Merrow, ASTD; Jennifer A. Orr, CDA; Stacey L. Campbell, CDA, EFDA; Trudy Smith, RDH; Joseph Thomas, MS; Douglas M. Smith, PhD. This project would not have been possible without the financial support of the AAID Research Foundation.
Duquesne University, Graduate School of Pharmaceutical Sciences.
Duquesne University Mylan School of Pharmacy, Graduate School of Pharmaceutical Sciences.
Private practice, Norwood, Ohio.