Clinical efficacy of titanium prepared platelet rich fibrin in periodontal regeneration: A systematic review and meta-analysis

Description of the condition

Plasma fractions with higher platelet concentrations, such as platelet-rich plasma (PRP) and platelet rich fibrin (PRF) are crucial to regeneration.¹ Their use has been successfully linked to connective tissue and bone repair.² PRP has been documented and used in a variety of applications with what appears to be clinical success. It is an autologous modification of fibrin glue that is created by techniques that concentrate autologous platelets.¹ PRP supports bioactivity but exhibits no stimulant properties for activation of regenerative cells.³ Additionally, its preparation is technique sensitive.⁴ PRP preparation, on the other hand, takes time therefore various efforts have been made to modify and enhance the PRP’s characteristics and quality which include Leukocyte platelet-rich fibrin (L-PRF). It was first described by Choukroun as cited by Dohan DM et al. (2006).⁵ It is regarded as a second-generation platelet concentrate and has been used to speed up wound healing during several surgical operations. Choukroun et al. first created PRF in France in 2001 as an autologous biomaterial that includes leucocytes.⁶ The method does not require bovine thrombin or an anticoagulant, in contrast to other platelet-rich products (or any other gelling agent).^5,7–9 Although PRF has been shown to produce positive therapeutic outcomes, some clinicians are concerned that the method’s use of glass-evacuated collection tubes for blood with silica activators may pose a health risk.^10,11 O’Connell provided a description of the inevitable silica interaction.¹² Although the silica in the tube is thick enough to settle with the red blood cells, some of the particles are still colloidally suspended in the buffy coat, fibrin, and platelet-poor layers of plasma. Therefore, when the product is used for therapy, the particles could come in contact with the patient. One of them is titanium platelet rich fibrin (T-PRF).¹³ In the glass vacuum containers used to prepare “Platelet Rich Fibrin,” silica is known to have a hazardous impact that this third generation platelet concentrate is known to eliminate (PRF).

Description of the intervention

Nonsurgical periodontal therapy, such as scaling and root planing (SRP) alone or SRP plus systemic or local anti-inflammatory or antibacterial drugs, to surgical flap debridement, are the approaches for treatment of periodontitis cases.¹⁴ However, in cases of soft and hard tissue defects, surgical correction of the defects is of prime importance. For correction of bony defects, the use of autografts and allografts have been advocated.¹⁵ Autologous platelet concentrates when used in conjunction with bone grafts results in better improvement. A study by Olgun et al. (2018) reported clinical, radiographic and histological outcomes of bone that accelerated to four months compared to six months for allografts when combined with T-PRF.¹⁶ The desired outcome was seen in four months as compared to six months. Tunalı et al. in the year 2012¹⁷ discovered a mature fibrin network when T-PRF clots were examined. In the T-PRF membrane, they discovered islets of bone tissue and newly developing connective tissue. These findings demonstrate that T-PRF could, within 30 days of treatment, cause the development of new bone and connective tissue in an experimental rabbit model of wound healing.

How the intervention might work

As mentioned above, the novel product termed T-PRF was prepared by Tunali et al. (2014)¹³ utilising a modified L-PRF process. Authors took the advantage of Dohan Ehrenfest’s taxonomy, which divides platelet-rich products into four main groups to prevent any misunderstanding about how to obtain these products. These four categories are determined according to the number of leukocytes and fibrin they contain, platelet-rich products: just platelet-rich plasma products, platelet- and leukocyte rich plasma products, platelet-rich fibrin, and platelet and leukocyte rich fibrin. The creation of T-PRF, a new platelet concentration, was motivated by the idea that titanium tubes would be more potent at activating platelets than the glass tubes used in Chouckron’s approach. The distinguishing properties of T-PRF, particularly its improved biocompatibility, are produced by platelet activation with titanium as opposed to activation with silica particles.¹⁸ T-PRF fibrin network covers a larger area than L-PRF fibrin network, also fibrin seemed thicker in the T-PRF samples.¹³

Why it is important to do this review

Since the introduction of T-PRF by Tunali et al.,¹⁸ it has been assessed for various properties including its structural and biochemical characteristics. Tunalı et al. (2014)¹⁹ described the structural properties of T-PRF, and compared it to L-PRF. T-PRF samples appeared to have a highly ordered network with continuous integrity. The fibrin network in T-PRF samples appeared to be thicker and to cover a wider area than the network in L-PRF samples, according to histomorphometric analyses. T-PRF is a naturally occurring leukocyte- and PRF product that has been defined for the first time in a human investigation. Titanium platelet activation appears to provide T-PRF with certain high features. Since then T-PRF has been used for various periodontal procedures including treatment of various infrabony defects, gingival recession defects, in open flap debridement cases, periimplantitis.^{13,16,18,20–22}

Systematic evaluation of the studies to assess the clinical efficacy have not been done to date. This systematic review focuses to provide the best possible evidence for the use of T-PRF in periodontal regeneration.

Data collection and analysis

Selection of studies

An open source online screening tool Rayyan²³ (RRID:SCR_017584) was used by review writers. They independently screened the search results (RO, PD, PB, KB). By reading through the abstracts of each study that the search produced, the eligibility of each research was established. The review writers excluded studies that did not distinctly meet the inclusion criteria. For all included studies’ complete texts were acquired. Primary reviewers independently screened the complete texts of these studies to pick the most pertinent studies (RO, PD, PB, KB). The respective authors were reached by phone or email to get the required clarification for any missing data or information in the studies that had an impact on the study selection criteria. In cases of disagreement or dispute, a fifth author was asked for a judgement (KD). Prior to evaluation, the trials were not anonymized. Any language restrictions in the study selection process were not taken into consideration as a constraint for carrying out this evaluation. The complete review includes a PRISMA flow chart³² that shows the precise status of all identified studies in accordance with the recommendation found in “Part 2, Section 11.2.1 of the Cochrane Guidelines for Systematic Reviews of Interventions”.²⁴ Irrespective of the reporting of outcome data, studies were included in this review as shown in Figure 1.

Figure 1. PRISMA 2009 flow diagram.

Data extraction and management

The data extraction from “included studies” was performed by four reviewers (RO, CS, KB, and KD) and given in the “characteristics of studies table” using a pre-defined data extraction form (Table 1). Data were extracted based on the nature of research, participant information, intervention information, and reported results. The disagreement between the main reviewers was resolved by the third reviewer (PB). The fourth reviewer successfully addressed the “risk of bias evaluation” discrepancy (KB). We have included a total of nine studies for qualitative analysis and have been included in the characteristics of studies table (Table 1). All the studies were matched for type of intervention, type of defect and participant details. Three studies were included and matched for meta-analysis followed by quantitative analysis.^20,22,25

Assessment of risk of bias in included studies

Four reviewers (RO, PD, CS, and SH) from each included study independently evaluated the risk of bias (RoB) using the Cochrane domain-based, two-part tool as outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions.²⁴ The fourth reviewer was able to resolve the disagreement between the main reviewers (KB). Sequence generation, allocation concealment, participant and staff blinding, blinding of outcome evaluation, incomplete outcome data, biased outcome reporting, and other bias, such as baseline imbalance, were all areas in which we evaluated the RoB.

Measures of treatment effect

In parallel-group RCTs, the unit of analysis was the individual subject. The Elbourne-recommended method is used to integrate the cross-over designed trials into the meta-analysis.²⁶ Measurements from experimental intervention periods and control intervention periods, respectively, were taken for these trials, and they were analysed under the assumption that it was a parallel group study of intervention versus control.

Assessment of heterogeneity

The Chi² test (P value set at 0.10 for statistical significance) was used to examine clinical heterogeneity, and the I² statistic was utilised to quantify heterogeneity in the outcomes of the included studies using RevMan manager version 5.0.²⁷ Significant heterogeneity is defined as I² over 75%; substantial heterogeneity is defined as I² between 50% and 90%; moderate heterogeneity is defined as I² between 30% and 60%; and mild heterogeneity is defined as I² less than 40%. If statistical heterogeneity with I² more than or equal to 50% is found, relevant causes were investigated using predefined subgroup analysis, and a random-effects model was used and reported.

Data synthesis

Only when it was determined that the participants, interventions, comparisons, and results of the included studies were adequately comparable to yield a conclusion with clinical relevance and significance was a meta-analysis performed. We intended to carry out the meta-analysis using the Cochrane Collaboration’s open source RevMan 2014 statistical software (RRID: SCR_003581). If statistical heterogeneity with I² higher than or equal to 50% is found, the sources of the heterogeneity were found, and a random-effects model meta-analysis was then carried out. In a meta-analysis, four papers were used. Table 2 presents data from all three investigations (Data Entry).

Forest plots

The analysis of the PI, PPD, CAL, DF was done separately for the interventional (OFD+T-PRF) and control groups (OFD alone). The details of the studies included for meta-analysis are given in Figures 2-5. All the included studies reported the use T-PRF combined with OFD. Three articles reported the use of T-PRF in infrabony defects,^20,22,25 one study reported use of T-PRF combined with allograft for patients with chronic periodontitis.²⁸ Another study reported its use in sinus lift procedures.¹⁶ Two studies reported its use in gingival recession^21,29 and other in healing of extraction socket.²⁹ Based on the overall similarities in participants, outcome and data three studies were found to be eligible for meta-analysis.^20,22,25 The conclusion regarding the overall effect size estimates were made based on the meta-analysis.

Plaque index

All the included studies reported PI at baseline and after nine months.^20,22,25 However, two studies reported data in the form of mean±standard deviation. All the studies showed statistically significant reduction in plaque index at nine months follow up period. The overall meta-analysis showed marginally significant reduction in plaque scores after nine months. (Mean Difference -0.52; 95% CI -1.30 to 0.27; p=0.03; I² 78%; two studies; 33 participants). However, in view of significant heterogeneity (I²=78%), the results need to be interpreted with caution because of increased heterogeneity as shown in Table 3 and Figure 2.

Figure 2. Forest plot of Plaque index.

Probing pocket depth

All the three included studies reported data on PPD at baseline and after nine months. All the studies reported significant reduction in PPD. The overall meta-analysis showed marginally significant reduction in PPD scores after nine months. (MD -0.85; 95% CI -1.26 to -0.45; p=0.01; four studies; 62 participants). However, in view of significant heterogeneity (I²=93%), the results need to be interpreted with caution as shown in Table 4, Figure 3.

Figure 3. Forest plot of probing pocket depth.

Clinical attachment loss

All the included studies reported clinical attachment loss (CAL) gain at baseline and after nine months. All the studies reported significant improvement in CAL gain. The overall meta-analysis showed marginally significant increase in CAL gain after nine months. (MD -0.12; 95% CI - 0.59 to -0.35; p<0.0001; three studies; 62 participants). However the percentage of variation (I² – 55%) was found to be minimum in all the three included studies, therefore making the interpretation of result favourable as shown in Table 5, Figure 4.

Figure 4. Forest plot of clinical attachment loss.

Defect fill

All the included studies reported defect fill at nine months. However quantitative assessment was only possible for two studies because of difference in reporting of outcomes.^20,22 Significant defect fill was observed in both the studies. The overall meta-analysis showed a marginally significant increase in defect fill after nine months (MD - 0.07; 95% CI -0.17 to -0.03; p=0.09; two studies; 47 participants). Overall defect fill was not statistically significant. However in spite of improvement in defect fill, in view of significant heterogeneity found the results needs to be interpreted with caution as shown in Table 6, Figure 5.

Figure 5. Forest plot of defect fill.

Risk of bias (ROB) in the studies

Allocation: One study reported use of the coin toss method for allocation hence was graded as low risk,²² the other two did not report the method and hence were graded as unclear.^20,25 Blinding: all the studies were double blinded and were kept in low-risk bias.^20,22,25 Missing result information: discretionary reporting In terms of reporting, we classified all three studies as low risk. The bias in the behaviour or observations: since all procedures and observations were carried out by a single trained professional in each research, we classified the study as low risk in terms of performance or observational bias. Figures 6 and 7 show, respectively, a summary of the risk of bias in the included studies and its graphical depiction.

Figure 6. Risk of bias graph.

Figure 7. Risk of bias summary.