Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession : Systematic Review and Meta-Analysis

include coronally advanced flap [3], laterally advanced flap [4], double papilla flap [5]. Others include autogenous grafts such as sub-epithelial connective tissue graft [6], free gingival graft [2], and the use of non-autologous substitutes including; acellular dermal matrix [7], Enamel Matrix Derivatives (EMD), Living Cellular Construct (LCC), and guided tissue regeneration [8]. The choice of surgical modality and material of choice is based on different factors such as degree of recession, location, width of keratinized tissue, gingival tissue biotype, esthetic demands, and patient’s preference. CTG has been considered the golden standard for the treatment of GR due to its significant outcomes in complete RC, attachment gain, keratinized tissue gain, and overall long-term treatment stability [9,10]. On the other hand, the need for a second surgical site to harvest the CTG is a main disadvantage due to the increased risk of bleeding, pain, and swelling that leads to the need for other RC alternatives [10]. As a result, approaches have been proposed and showed approximately similar results to CTG based treatment and considered an alternative depending on the clinician’s and patient’s choices; these include: Guided Tissue Regeneration (GTR), Acellular Dermal Matrix Graft (ADMT), and biological materials e.g. Platelet Rich Fibrin (PRF), Platelet Rich Plasma (PRP), Enamel Matrix Derivatives (EMD), among others [10,11].


Introduction
Gingival Recession (GR) is the apical migration of the free gingival margin, in relation to the Cemento-Enamel Junction (CEJ), due to periodontal disease, mechanical factors, underlying bone dehiscence, or post orthodontic movement, among other etiological factors [1].It has been classified into four categories; depending on the extent of gingival tissue involvement and the level of inter proximal bone [2].Many Root Coverage (RC) approaches have been proposed including; pedicle flaps; which

Abstract
The purpose was to systematically review the effect of platelet rich fibrin (PRF) in the treatment of Miller class I and II recession defects in comparison to conventional surgical procedures.Three electronic databases were searched, and hand search was performed for relevant articles, up to October 2015.All relevant articles were independently screened to specific inclusion criteria.Primary outcomes were Recession Depth (RD), Keratinized Tissue Width (KTW), and Percentage of Root Coverage (%RC).Secondary outcomes were Clinical Attachment Level (CAL), Probing Depth (PD), Healing Index (HI), and Pain.Ten randomized clinical trials met the inclusion criteria and seven were included in the meta-analysis.No statistically significant difference was found in %RC between Coronally Advanced Flap (CAF) and CAF + PRF or between CAF + Connective Tissue Graft (CAF + CTG) and CAF + PRF (p = 0.17 and p = 0.56) respectively.A borderline statistical difference was observed between CAF and CAF + PRF (p = 0.05), and no statistically significance difference between CAF + CTG and CAF + PRF (p = 0.23) in KTW.In regards to pain and healing, a significant reduction in pain during the first 5-7 days and faster healing observed in the PRF intervention when compared to the use of CTG or Enamel Matrix Derivative (EMD).In conclusion, there was no statistical or clinical difference present between PRF and CAF, CAF + CTG, or CAF + EMD for RD, %RC and KTW when treating Miller class I and II gingival recession.Significant improvement of postoperative pain and healing can be achieved, which may indicate PRF use as an alternative to conventional surgical approaches.

Search strategy
Comprehensive search strategies were established.MEDLINE (via pubmed), EMBASE and CENTRAL databases were searched from the earliest records through October 2015.Unpublished studies, thesis and reference lists were hand searched.Details regarding the search terms are: ((((gingival recession) OR gingival dehiscence) OR mucogingival defect) OR gingival defect) OR gingival undergrowth AND (((((Platelet rich fibrin) OR PRF) OR platelet derivatives) OR autologous platelet concentrate) OR platelet growth factors) OR platelet derivatives factors AND(( (((((((((((((( Relevance Databases were searched without language restrictions using MESH terms, key words and other free terms, and Boolean operators (OR, AND) were used to combine searches.
Relevant articles were screened with no language limitation.

Assessment of validity
Two independent reviewers (R.J. and H.K.) screened the titles, abstracts and full texts that were identified.Disagreement between the reviewers was resolved through discussion and consensus was reached.Authors were contacted to resolve ambiguity and to retrieve missing data from the trials.Cohen's Kappa score was used to assess inter-reviewer agreement of selection process [26].The reasons for excluding studies were recorded (Figure 1).Studies meeting the inclusion criteria underwent data extraction and validity assessment.

Data Extraction
Pre-designed extraction forms were developed to assess the following data: author name(s), publication year and place, source of funding, conflict of interest, study design, sample size, follow-up period, source, selection and description of the study population (including age, gender, race and ethnicity, and presence and characteristics of gingival recession at baseline), definition and measurement method of the intervention, controls, outcomes, results and their variation, and risk of bias.

Data Synthesis
Data synthesis was preformed through organizing data in an evidence table and a descriptive summary was created to determine study characteristics (Table 1).Descriptive statistical analysis according to the mean values was used to evaluate the outcomes of test and control groups (Table 2).

Quality assessment and risk of bias
The methodological quality of RCTs included was assessed and recorded in Table 3 according to PRISMA [27].

Results
The screening process is shown in (Figure 1).Electronic and hand searches yielded 540 articles, of which 12 were selected (2) for full-text evaluation after screening their titles and abstracts.Two articles [28,29] were further excluded and reasons are listed in (Figure 1).The k value for inter-reviewer agreement for potentially relevant articles was 0.95 for full text articles reviewing; indicating an ''almost perfect'' agreement between the two reviewers [30].

Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Copyright: © 2017 Jasser, et al.
teeth [31,32,35,40], where as two included maxillary teeth only [33,39].The remaining four studies did not mentioned which arch was treated [34,36,37,38].In regard to the type of teeth treated, six studies included anterior teeth and premolars (32,33,34,35,37,39].While, one study included all teeth [40].The remaining three studies did not mention the type of teeth treated in a detailed manner [31,36,38].All studies included Miller's class I and II GR; however the exact baseline depth of recession was mentioned in only one study [38].

Protocol for PRF preparation
All studies followed Choukroun's (2001) protocol in the preparation of PRF [12].10 mL of venous blood was collected from patients and taken to the chair-side centrifuge.A trained auxiliary staff member prepared PRF simultaneously with the surgical procedure.Centrifusion speed ranged between 2500-3000 rpm in a time frame ranged between 10-12 minutes.

Surgical technique
A full thickness periosteal flap was elevated on the buccal aspect of the tooth being treated in four trials [31,36,37,39], where one of them obtained a trapezoidal full thickness flap by the addition of two vertical incisions starting from its mesial and distal extremities extending beyond the muco-gingival junction [31].Howeve, the other three did not perform any vertical incisions [36,37,39].In six studies a split-thickness flap with two vertical incisions mesial and distal to the involved tooth was elevated instead [32,33,34,35,38,40].In Aroca (2009) trial, the surgeon performed a split thickness flap without any vertical incisions [40].In regards to graft thicknesses, PRF membrane thickness varied among studies between 0.5-2 mm.However, the thickness was only mentioned in two studies [35,39].In regards to CTG thickness, It ranged between 1-1.5 mm.Final flap position was another factor that varied among studies; In five studies final flap position was 1 mm coronal to CEJ [31,34,36,37,39] while in the remaining studies it was positioned at the CEJ.

A.
Primary Outcomes: RD, %RC, and KTW: Primary outcomes were reported in all ten studies, two studies revealed that PRF groups had statistically significant more reduction in RD [36,38], most pronounced difference was shown by Padma (2013) [36] when compared CAF + PRF to CAF alone; authors reported that the mean RD for PRF groups and CAF groups were 0.00±0.00mm vs. 1.13±0.72mm respectively at 6 months follow up (p = 0.001).Similarly, Jankovic (2010), as comparison was performed between CAF + PRF versus CAF + EMD, the mean RD for CAF + PRF and CAF + EMD groups were 1.05±0.45mm vs. 1.15±0.65 mm respectively at one year follow up (p < 0.05).In contrast, one study by Aroca (2009) reported less RD at 6 months for CAF when compared to CAF + PRF measuaring 0.2±0.4mm vs. 0.6±0.6 mm respectively (p = 0.0039) [40]. .The remaining seven studies reported no significant difference in RD.
B. Secondary outcomes: PD, CAL, HI, and pain: PD was reported in nine studies.Only one study showed a statistical significant reduction in PD favoring CAF + PRF compared to CAF + CTG at 6 months follow-up measuring 1.09±0.29 mm vs. 1.45±0.60mm respectively (p = 0.017) [35].CAL was documented in nine studies as well, from which only two reported significant difference between test and control groups.Padma (2013) [36] reported statistical significant gain in CAL in CAF + PRF group when compared to CAF alone group for a value of 1.00±0.00mm vs. 2.00±0.89mm respectively (p = 0.002).Additionally, Aroca (2009) reported a statistical significant gain favoring CAF alone (1.37±0.62 mm) when compared to CAF + PRF (1.76± 0.97 mm; p = 0.0004).HI was measured in three studies.Aleksic (2010) [39] scored the healing on a 1-5 index, based on tissue color, response to palpation, granulation tissue, incision margin, and suppuration, where 5 indicated excellent healing and 1 indicating poor healing [41].It was noted that HI was significantly different in the CAF + PRF group after the first and second weeks of 3.11±0.25 and (2) Systematic Review and Meta-Analysis.J Dent Oral Disord Ther 5(1): 1-12.

Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Copyright: © 2017 Jasser, et al. 4.25±0.25 respectively, than in the control group during the same period 2.25±0.52 and 3.05±0.40respectively (p < 0.05).Jankovic (2010) used a HI that was based on redness, granulation tissue, bleeding, suppuration and epithelialization for the first and second week post surgery and reported a statistically significant superior healing for the one-week postoperatively in CAF + PRF when compared to CAF + EMD (p < 0.05).However, this significant difference was absent at the 2 weeks follow up 4.51±0.21vs. 4.29±0.36(p > 0.05) respectively.Further, Jankovic (2012) [37] showed enhanced healing values obtained in the CAF + PRF group for the first 2 weeks after surgery in comparison with the CAF + CTG group.Results recorded in the PRF group after 1 and 2 weeks of surgery were 3.11±0.32and 4.20±0.27,respectively, while for CAF + CTG were 2.25±0.54and 3.05±0.38,respectively (p < 0.05).This statistical difference disappeared again at the three weeks follow up 4.51±0.21and 4.29±0.36,respectively (p > 0.05).
Pain was recorded on a horizontal pain scale, where 0 meant no pain, 1 intermediate pain, and 2 severe pain, before and after the procedure in three studies.Jankovic (2010) [38] assessed post-operative pain for 7 days after the surgery, the authors stated that the pain intensity was significantly different between CAF + EMD group and CAF + PRF group favoring the later in the first 5 days 0.82±0.22 vs. 0.60±0.33(p = 0.048), and at day 7 this difference was no longer significant between the groups (p = 0.143).A more recent study comparing CAF + CTG to CAF + PRF by the same author reported that all patients indicated a greater discomfort in the CTG group, where the pain intensity was statistically different between groups for the first 7 days favoring the CAF + PRF group 0.20±0.41 vs. 0.46±0.51(p < 0.05) [37].Similarly, Aleksic (2010) [39] indicated that patients treated with CAF + PRF reported statistically significant less pain than patients treated with CAF + CTG in the first 7 days postoperatively 2.09±0.46 vs. 0.46±0.51(p < 0.05) respectively.

Risk of Bias Assessment
The results of the bias assessment of the included studies are presented in table 3. None of the studies obtained the highest score in the quality analysis.Allocation concealment was clearly mentioned on only one study [32], as this can bring these studies to uncertain risk of bias [42].Blinding was not reported in three of the included studies a [36,39,40].None of the studies reported adherence to the CONSORT statement recommendations [25].

Meta Analysis Results
Meta Analysis was performed separately among studies comparing CAF to PRF + CAF and among studies comparing CAF + CTG to CAF + PRF as followed:

CAF vs. CAF + PRF
This analysis included four studies [31,33,36,40].Figure 2 depicts a forest plot with a continuous outcome variable of RD and KTW.A random effect model with confidence interval (CI) of 95%. 1 2 = 0% so studies are considered homogenous and it is safe to have confidence that the effects of the intervention being tested are accurate and can be trusted.

RD Change
When RD change was assessed from baseline (0.31 [0.08, 0.54]) to 6 months follow up (0.23 [-0.10, 0.57]), the black diamond cross the 'line of no effect', the calculated difference between the experimental and control groups is not considered as statistically significant (p = 0.17) 1 2 = 52%, which is greater than 50% and studies are very heterogeneous.
The overall studies are considered as homogeneous with 1 2 =2%, also, the black diamond (with the average effect size of 0.25) falls on the right-hand side of the graph that shows studies that received the control condition reported bigger changes than the studies that received the experiment condition.

%RC at 6 months follows up
(Figure 3) depicts a forest plot with a continuous outcome variable.We used a random effect model with CI of 95%.We compare %RC studies with 6 month, the black diamond (with the average effect size of -3.51) cross the 'line of no effect', the calculated difference between the experimental and control groups is not considered as statistically significant.(p = 0.60). 1 2 = 56% and studies are heterogeneous.

CAF + CTG Versus CAF + PRF
This analysis inlcuded three studies [34,35,37].Figure 4 depicts a forest plot with a continuous outcome variables of RD and KTW.A random effect model with CI of 95%.In terms of consistency, 1 2 =0%, and studies are regarded as highly homogeneous.
RD Change: When change in RD was compared between sites treated with CTG and PRF form baseline (0.24 [0.15, 0.32]) to 6 months follow up (0.03 [-0.08, 0.14]), the black diamond cross the 'line of no effect', the calculated difference between the two groups is not considered as statistically significant (p = 0.56). 1 2 =36% and studies are heterogeneous.KTW Change: KTW was compared in the same studies between baseline (-0.07 [-0.31, 0.18]) and 6 months follow up (-0.29 [-0.77, 0.19]), the black diamond cross the 'line of no effect', the calculated difference between the experimental and control groups is not considered as statistically significant (p = 0.23). 1 2 = 36% and studies are heterogeneous.The average effect size of the overall studies is 0.01 which shows there is no difference between control and experimental studies.

Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Copyright: © 2017 Jasser, et al.

Discussion
Free Gingival Graft (FGG) and Connective Tissue Graft (CTG) are considered to be the golden standard for the re-establishment of keratinized tissue width and root coverage respectively [43,10].But the accompanying second surgical site, post-operative pain and discomfort had led to the search for other alternatives in the current conservative era where emerging regenerative approaches, as proteins and growth factors, have gained much popularity [44].
Among the alternatives, Enamel Matrix Derivative (EMD) showed strong evidence supporting its use with CAF for root coverage, it had shown long-term (>24 months) stable results comparable to that accomplished by CTG [10].
While acellular dermal matrix (ADMT) resulted in 93% root coverage compared to 97% root coverage for CTG in short-term (13 weeks) and 66% versus 97% respectively in long-term (49 months) [45].While GTR resulted in 41% complete root coverage and 74% recession depth reduction with attachment and keratinized tissue gain [8].This review aimed to compare PRF to CTG and its known alternatives (CAF, EMD, ADMT, FGG, and GTR).Findings conclude a comparable final result of RC and gain of KTW in sites treated with PRF in contrast to similar sites treated with the aforementioned conventional surgical approaches.Thus, these findings can enhance the use of PRF as to reduce the need for second surgical site and prevent relevant risks associated.As well as, to prevent high cost associated with growth factors and

Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Copyright: © 2017 Jasser, et al.

accompanied materials used.
Two meta analysis were conducted to gain a more precise comparison of the primary outcomes between CAF and CAF + PRF and between the addition of CTG or PRF.Main results concluded that there was no statistically significant difference in RC between CAF and CAF + PRF or between CAF + CTG and CAF + PRF (p = 0.17 and p = 0.56) respectively, which was consistent with the recently published systematic review on the same topic [46].In terms of the KTW, a borderline statistical difference was observed between CAF and CAF + PRF (P = 0.05), and no statistically significance difference between CAF + CTG and CAF + PRF (P = 0.23).The lack of significance between CAF + PRF and CAF + CTG was inconsistent with the systematic review published by Moraschini (2016) [46], in which they included seven studies, where they found that the addition of CTG resulted in increased gain in KTW when compared to PRF, the result variation might be explained by the inclusion of two more studies in the present meta analysis comparing the two interventions [32,39].In these two studies, keceli (2015) found no significant difference between CAF + PRF and CAF + CTG in the gain of KTW (p = 0.077), while Aleksic (2010) found increased gain in KTW favoring CAF + CTG when compared to CAF + PRF (p = 0.013).Also, the present Meta-Analysis found that CAF + PRF did not increase %RC when compared to CAF alone (p = 0.60).In regards to healing and pain assessment, all the three studies that included HI and pain in their outcomes consistently reported a significant reduction in pain for the first 5-7 days and faster healing in the PRF intervention when compared to the use of CTG or EMD, which was consistent with numerous studies describing faster healing potential of growth factors in regenerative procedures [32,38,39,[47][48][49][50][51][52].This finding is explained by PRF characteristics, when compared to other forms of blood derived growth factors like PRP, in the slow release of growth factors over a period of 10 days which is the time needed for revascularization and CT formation in a soft tissue regeneration procedure, while PRP was known for an earlier release of growth factors in the healing cascade [53][54][55].A meticulous care should be taken in terms of proper handling of PRF, as well as immediate application to recession site after preparation [56,57].It should be noted that some of the drawbacks for the use of PRF in root coverage procedure is the need to cover the membrane completely to prevent its early resorption, which requires clinical case pre-requisite for a successful coronally advanced flap including; the presence of keratinized tissues, a recession depth not exceeding 4 mm, the presence of a vestibular depth to prevent flap tension, the placement of final flap margin coronal to CEJ, and a tissue biotype thickness of no less than 0.8 mm [58,59].On the other hand, incomplete coronally advancement of the flap is acceptable for CTG where one-third to one-half can be left exposed to prevent vestibular shortening and increased tension on the flap margins [6,60].Most of the studies followed the PRF preparation techniques described by Choukroun (2001) [12] but with some variation in time and centrifugation speed which might have affected the RC potential of the PRF membrane, as it has been shown that the increase of centrifugation time from 10 to 12 minutes increased the amount of vascular endothelial growth factor but it did not affect any of the other growth factors or enzymes in the platelet rich fibrin membrane [61].The lack of histological analysis for the evaluation of regenerative capacity and type of cells populating the previously exposed root surface is another limitation of the current study.More randomized clinical trials with a splitmouth design are needed to overcome the heterogeneity in hostresponse and tissue biotype.Additionally, long-term stability of PRF membrane for RC has to be evaluated in long-term studies.

Conclusions
This review indicated no statistical or clinical difference in the use of PRF when compared to CAF.This lack of statistical difference makes PRF a comparative alternative to CAF for soft tissue regeneration in the treatment of Miller class I and II gingival recession.The present review showed no difference between CAF+ PRF vs CAF+ CTG or CAF+ EMD.Further clinical comparative trials are needed to study the difference between these treatment modalities in order to draw a more valid comparison and conclusions.In addition, it was difficult to withdraw a strong comparison between PRF and EMD since there was only one study included in this review.Finally, the reduced post-operative pain and accelerated healing by the PRF offers an advantage of using it compared to CTG or EMD which also need to be confirmed by future clinical and histological evaluation.
12 full-text articles assessed for eligibility 2 articles excluded for the following reasons (Kappa=0.95): • PRF was not the growth factor that was utilized specifically.

Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Copyright: © 2017 Jasser, et al.
coronally positioned flap) OR coronally advanced flap) OR connective tissue graft) OR laterally positioned flap) OR sliding pedicle flap) OR laterally sliding flap) OR free gingival graft) OR sub epithelial connective tissue graft) OR autologous soft tissue grafts) OR mucogingival graft) OR allogenic soft tissue graft) OR soft tissue graft) OR mucoderm) OR alloderm) OR pedicle flap))AND(((((((((root coverage) OR attachment gain) OR keratinized tissue) OR attached tissue) OR area of coverage) OR recession depth) OR gingival biotype) OR gingival thickness) OR gingival recession width).

Figure 2 :
Figure 2: Forest Plot of RD change and KTW change at baseline and 6 months follow up for CAF+PRF vs. CAF

Figure 4 :
Figure 4: Forest Plot of RD change and KTW change at baseline and 6 months follow up for CAF+PRF vs. CAF + CT

Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis
Citation: AL Jasser R, AlKudmani H, Andreana S (2017) Platelet-Rich Fibrin as a New Approach in Treating Gingival Recession: Systematic Review and Meta-Analysis.J Dent Oral Disord Ther 5(1): 1-12.Platelet-