Actual quantitative attachment gain secondary to use of autologous platelet concentrates in the treatment of intrabony defects: A meta-analysis.

BACKGROUND: There are no authoritative meta-analyses and no clear quantitative assessments available estimating effects of open flap debridement (OFD) combined with platelet-rich fibrin (PRF) or platelet-rich plasma (PRP) over and above that of OFD. This study evaluated the actual quantitative mean gain for various clinical (clinical attachment level [CAL], probing pocket depth [PPD] and gingival marginal level [GML]) and radiographic (intrabony defect depth [IBD]) parameters of Platelet Concentrates- PCs (PRP/PRF) as sole grafting material along with OFD and OFD alone in the treatment of intrabony defects.
MATERIALS AND METHODS: A detailed electronic search was carried out in PUBMED/MEDLINE, COCHRANE, EBSCOHOST, and Google Scholar databases by unifying related search terms with additional hand searches in select specialty journals up to May 2017. The eligibility criteria included human randomized clinical trials, either of a parallel group or a split-mouth design with follow-up period of at least 6 months. Periodontal intrabony defects with radiographic IBD ≥3 mm with corresponding CAL ≥5 mm were included. For the meta-analysis, the inverse variance method was used in fixed- or random-effect models.
RESULTS: Actual quantitative mean gains were calculated for OFD with PRF/PRP (CAL = 1.1 mm, IBD = 1.68 mm, PPD = 0.97 mm and GML = 0.48 mm) over and above that of OFD alone.
CONCLUSION: Because of very high heterogeneity, the results may not be dependable. Apart from gains in radiographic bone fill, all other periodontal clinical parameters showed negligible gains. Using PRF technologies in periodontal intrabony defects may not be of great clinical significance over and above that of OFD alone, the effect sizes are also not large enough.

INTRODUCTION

As periodontitis is characterized by loss of tooth/teeth supporting tissues, any attempt of its therapy would be directed to partially or fully reconstitute the periodontal tissues. To restore and reconstitute lost periodontal structures to pre-disease morphophysiological levels as a definitive goal at regenerating periodontal structures, would remain ever elusive at least for near future. To understand the intricacies of post-loss redevelopment for any tissue or organ, a detailed knowledge of its embryonic and developmental biologic events is paramount.

Few biologic processes as far as growth factors are concerned have been identified and are being currently used for periodontal therapy and bone growth to achieve some regeneration or repair. Autologous growth factors being delivered through chair-side platelet concentrates may hold some promise for regeneration, but they may not represent the true sequence and concentration of required growth and differentiation factors.

Even though the release kinetics of various growth factors (platelet-derived growth factor [PDGF-AA/PDGF-AB, PDGF-BB], transforming growth factor-β1 [TGF-β1], vascular endothelial growth factor [VEGF], epidermal growth factor [EGF] and insulin-like growth factor-1 [IGF-1]) from platelet's alpha granules with varied autologous preparations are different,[1] platelet derivatives have been demonstrated to possess potential for enhanced and accelerated hard and soft tissue regeneration.[234]

Platelet-rich plasma (PRP) was available prior to platelet-rich fibrin (PRF). To differentiate them Dohan et al.[5] referred PRP as “ first generation platelet concentrate” and PRF as “second generation platelet concentrate”. He described PRF as an autologous platelet and fibrin material derived from whole blood. One of the biological differences between PRFs and PRPs is growth factor enmeshment;[6] the former intrinsic and the later extrinsic. Over a period of few decades, the platelet concentrates have evolved a lot and can be loosely classified under PRP (pure-PRP, leukocyte-PRP) and PRF (pure-PRF, leukocyte-PRF, advanced-PRF, titanium prepared-PRF, and injectable-PRF).[56] Due to simplicity of preparing Choukroun's PRF,[5] it is gaining popularity in its usage. The preparation is simple, inexpensive, and does not require any anticoagulants, bovine thrombin, or any other gelling agents.[5] The platelet concentrates are in use for various maxillofacial[34] and other[7] indications, including periodontal intrabony defects.

Periodontal intrabony osseous defects (IBD) or vertical defects, a characteristic of chronic or aggressive periodontitis, are being treated currently with various approaches including the use of PRPs and PRFs. As guided tissue regeneration (GTR) techniques, which can be considered as gold standard treatment for IBDs, are expensive and technically demanding, use of platelet concentrates as alternative therapy is increasing. Whether the use of these concentrates is beneficial or not as mono-therapy in IBDs is yet to be established reasonably well till date.

There have been various reviews and meta-analyses[89101112131415] over the years on effectiveness of PRP/PRF showing mixed results. While Rock[8] was a literature summary, Hou et al.[9] and Panda et al.[10] dealt with adjunctive effects of PRP in intrabony defects, they did not evaluate PRP/PRF as a sole treatment option. Roselló-Camps et al.'s[11] manuscript is not clear on whether included studies used PRP as an adjunct (i.e., PRP + bone graft [BG]) or PRP alone. Further analysis showed that furcation invasion studies are also included. Shah et al.[12] meta-analysis showed significant improvements in clinical parameters when PRF was used alone. None of these meta-analyses considered platelet concentrates (PRP and PRF) as a monotherapy and evaluated their benefits over open flap debridement (OFD) in certain quantitative terms. Furthermore, none of these studies evaluated actual quantitative mean gains for any parameter (AQMG). These calculations will give more insight into clinical significance of conducted studies and a more direct evidential reference to the clinicians. In an era when the statistical significance is fiercely debated,[16] and there are multiple ways of proving a difference as statistically significant, clinically significant differences should be projected and referred to.[17] Whether we are gaining “positive efficacy;” achieving an effect similar to or greater than the most effective regimen to date;[17] keeping in mind cost-benefit ratio, side-effects, patient comfort, and risk for PRF/PRP treatments needs to be ascertained beyond doubt.

The aim of the present meta-analysis is to evaluate actual quantitative mean gains of autologous platelet concentrates (PRF/PRP) in the treatment of intrabony defects in randomized controlled trials over and above that of OFD. The focused question for our meta-analysis is, “Whether there is any clinically significant advantage of using autologous platelet concentrates (PRF/PRP) along with OFD in intrabony defects, as represented by various clinical and radiographic periodontal parameters when compared to use of OFD alone?”

MATERIALS AND METHODS

Evaluation and analyses of clinical (attachment level gain, probing depth reduction, gingival marginal level [GML] changes) and radiographic (intrabony defect bone-fill) parameters in the treatment of periodontal intrabony defects treated with OFD alone or in combination with PRF/PRP was our objective.

In addition, we evaluated the actual quantitative mean gains in IBD parameters for PRF/PRP over and above that of the same obtained with OFD alone, considering baseline and follow-up mean values of all the studies. Is the IBD gain only statistically significant or clinically significant too when treated with PRF/PRP as compared to OFD alone was our primary focused question. As it was a meta-analysis, no ethical committee clearance was required, but Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA)[18] guidelines were followed. The meta-analysis is submitted for registration vide no. 114723 with International Prospective Register of Systematic Reviews in collaboration with National Institute for Health Research.

The null hypothesis for our meta-analysis and other statistics was, “there is no clinical and statistically significant difference in various clinical and radiographic periodontal parameters and actual quantitative mean gains for the management of periodontal intrabony defects irrespective of whether they are treated with OFD alone VERSUS treated with OFD plus PRF/PRP.”

Information sources and search strategy

A detailed search of electronic databases was conducted using PUBMED/MEDLINE, COCHRANE, EBSCOHOST, and Google Scholar using search terms, such as, “Autologous platelet concentrates,” “intrabony defects,” “platelet rich plasma,” “platelet rich fibrin,” “chronic periodontitis,” “aggressive periodontitis,” “vertical defects,” and “periodontal osseous defects.” The search was limited to clinical trials involving human subjects restricted to English language. A literature search was carried out for articles published up to May 2017, with the last electronic search being carried on May 28, 2017. Articles available in electronic form before their publication in printed form were also considered eligible for inclusion in this meta-analysis. One randomized control trial (RCT)[19] was also considered for inclusion which is already accepted for publication in International Journal of Periodontics and Restorative Dentistry (IJPRD), also can be referred at Clinical Trials Registry India No. REF/12/006069.

The references of all original research and review articles identified to be relevant were scanned for possible additional studies. Furthermore, hand searches were also carried out for manuscripts published up to May 28, 2017 from the Journal of Periodontology, the Journal of Clinical Periodontology, the Journal of Dental Research, the Journal of Periodontal Research, and the International Journal of Periodontics and Restorative Dentistry (IJPRD).

Inclusion criteria for manuscripts

Human randomized clinical trials, either of a parallel group or of a split-mouth design, reporting adequate and readable data from ≥10 subjects/osseous defects in the PRP or PRF group

A randomized controlled clinical trial where one of the groups received autologous PRF/PRP. The comparator group can be of any treatment modality but only OFD alone was considered for analyses

The patients included in the RCT had no systemic illness or abnormal platelet counts that could affect the clinical outcome of periodontal therapy

Periodontal intrabony defects with radiographic IBD ≥3 mm with corresponding CAL ≥5 mm were included. All the defects irrespective of mentioned number of walls (1, 2 or 3 walled defects) were included

Studies determining at least one of these variables were included: The clinical attachment levels (CALs), the depth of intrabony defect, and the probing pocket depths (PPDs) at baseline and final follow-up of at least 6 months

Studies mentioning furcation invasions of teeth were excluded

Study designs such as case series, case reports, retrospective studies, in vitro studies, animal studies, reviews, and meta-analyses were excluded.

Data extraction

The included studies were thoroughly read and the following information was assessed and tabulated: the title, authors and other bibliographic information, study design and qualitative parameters, test and control groups, age range of the patients, number of the defects treated in the test and control groups, characteristic of the defects, clinical and radiographic parameters (CALs, intrabony defect fill, gingival margin level, probing depth), and follow-up [Table 1].

Table 1

Details of design of included studies

Study	Number of defects treated test/control (total)	Age range (mean±SD)	Characteristic of defect	Intervention test	Comparison control	Clinical and radiographic parameters	Follow-up (months)
Ilgenli et al.[22] (2007)	12/16 (28)	18-60	PPD and CAL ≥6 mm, intrabony defect. One to three osseous walls and absence of furcation involvement	PRP	DFDBA + PRP	PD, CAL, IBD	18
Markou et al.[23] (2009)	12/12 (24)	40-65 (52.08±7.33)	PD at the site of the defect ≥6 mm; CAL at the site of the defect≥6 mm; radiographic (DD) ≥3 mm	PRP	DFDBA + PRP	PD, CAL, GR, IBD (%)	6
Pradeep et al.[24] (2009)	14/14 (28)	22-48 (34.28)	Paired, contralateral interproximal IBD ≥4 mm; interproximal PD ≥5 mm	PRP	PRP + ABM/P-15	PD, CAL, GML, IBD	9
Lekovic et al.[25] (2011)	17/17 (34)	44±9	Two similar interproximal IBD with PD ≥6 mm	PRF	PRF + BPBM	PD, CAL, IBD, GR	6
Thorat et al.[26] (2011)	16/16 (32)	25-45 (31.12±2.06)	Interproximal IBD ≥3 mmAlong with an interproximal PD ≥5 mm	PRF	OFD	PD, CAL, GML, IBD	9
Sharma and Pradeep (2011)	18/17 (35)	30-50 (35.34±6.45)	Three-wall IBD ≥3 mm along with an interproximal PD ≥5 mm	PRF	OFD	PD, CAL, GML, IBD	9
Pradeep et al.[29] (2012)	30/30/30 (90)	36.8	Three-wall IBD=3 mm along with an interproximal PD=5 mm	PRF, PRP	OFD	PD, CAL, GML, IBD	9
Pradeep et al. (2012)[28]	30/30/30 (90)	39.7	IBD≥3 mm along with an interproximal PD ≥5 mm	PRF, PRF + HA	OFD	PD, CAL, GML, IBD	9
Rosamma Joseph et al. (2012)[30]	15/15 (30)	17-44 (29.47±7.65)	Contra lateral interproximal infrabony defect with PPD PD ≥6 mm CAL ≥5 mm and IBD ≥4 mm	PRF	OFD	PD, CAL, IBD, recession	12
Ajwani et al.[31] (2014)	20/20 (40)	30.5	Two- and three-walled intrabony defect ≥3 mm, probing depth PD ≥5 mm	PRF	OFD	PD, CAL, GML, IBD	9
Jain Gupta et al.[32] (2014)	22/22 (44)	30-65	PPD ≥5mm; CAL ≤ PPD and DD on IOPA x-ray of ≥3 mm; predominantly three-wall interproximal intrabony defect	PRF	OFD + EMD	PPD, CAL, DD	6
Mathur et al.[33] (2015)	19/19 (38)	30-65 (39.66±5.72)	PPD ≥5 mm, radiographic DD on IOPA of≥3 mm	PRF	ABG + OFD	GR, PPD, CAL, IBD	6
Pradeep et al.[34] (2015)	30/30/30/30	30-50 (41)	IBD ≥3 mm, PD ≥5 mm	PRF, OFD + 1%MF, OFD + PRF + 1% MF	OFD	PD, RAL, GML, IBD	9
Suchetha et al.[35] (2015)	10/10 (20)	20-55	PPD ≥5 mm, radiographic defect depth ≥3 mm	PRF	PRP	PD, CAL, IBD, GR	9
Shah et al.[36] (2015)	20/20 (40)	20-55	PD ≥5 mm at 2 or more sites, 2 or 3 wall intrabony defect at 2 or more sites, radiographic evidence of IBD	PRF	OFD + DFDBA	PD, RAL, GML	6
Chatterjee et al.[37] (2016)	30/30/30 (90)	20-55	IBD, PPD ≥5 mm, and CAL ≥3 mm	PRFOFD + T-PRF	OFD	PD, CAL, IBD	9
Chandradas et al.[38] (2016)	12/12/12 (36)	35-50	CAL >4 mm, PD ≥5 mm and IBD ≥3 mm	PRF, PRF + DBM	OFD	PD, RAL, recession, IBD	9
Pradeep et al.[40] (2016)	30/30/30 (90)	25-45	PD ≥5 mm, CAL ≥3 mm and 2- or 3-walled IBD ≥3 mm	PRF, OFD + PRF + 1.2% RSV	OFD	PD, CAL, IBD	9
Kanoriya et al.[39] (2016)	30/30/30 (90)	39	3-walled IBD ≥3 mm , PD ≥5 mm	PRF, OFD + PRF + 1% ALN	OFD	PD, CAL, IBD	9
Martande et al.[41] (2016)	30/30/30 (90)	37.6	3-wall IBD ≥3 mm	PRF, PRF + 1.2% ATV	OFD	PD, RAL, IBD, GML	9
Agarwal et al.[42] (2016)	10/10/10 (30)	-	3 deep intrabony defects (3-walled) with (PD) >5 mm, radiographic defect depth >4 mm	PRP, PRP + DFDBA + OFD	OFD	PD, CAL, IBD, recession	12
Chadwick et al.[43] (2016)	17/19 (36)	18-89 (54.9±12.1)	IBD ≥3 mm, PD≥6 mm, radiographic base of defect ≥2 mm coronal to root apex	PRF + OFD	OFD + DFDBA	CAL, IBD, PD, GML	6
Galav et al.[44] (2016)	10/10 (20)	30-55	3-wall IBD ≥3 mm. PD ≥5 mm	PRF	OFD + ABG	PPD, RAL, RBF	9
Jalaluddin et al.[45] (2016)	10/10 (20)	25-45	PD ≥5 mm, radiographic evidence of IBD	PRP	OFD	PPD, GML, CAL, RBF	6
Thorat et al.[19] (2017)	15/15 (30)	25±1.5	IBD ≥3 mm, PD ≥5 mm, at least 2 contralateral interproximal intrabony defect	PRF	OFD	CAL, IBD, PD, GML	6 and 12
Total number of defects considered for further analysis				519	325	844

DD – Defect depth; PD – Pocket depth; CAL – Clinical attachment level; PPD – Probing pocket depth; GML – Gingival marginal level; IBD – Intrabony defect depth; OFD – Open flap debridement; PRF – Platelet-rich fibrin; DFDBA – Demineralized freeze-dried bone allograft; PRP – Platelet-rich plasma; ABM/P-15 – Anorganic bovine-derived matrix and peptide 15; BPBM – Bovine porous bone mineral; GR – Gingival recession; EMD – Enamel matrix derivative; ABG – Autogenous bone graft; RAL – Relative attachment level; RBF – Radiographic bone fill; IOPA – Oral periapical radiograph

Calculation of actual quantitative mean gain

For our meta-analysis, OFD was considered as control group and that of intervention with PRF/PRP considered as experimental (test) group, irrespective of the groupings in original selected RCTs. A table was prepared summarizing the mean values at the baseline and follow-ups, each for test and control groups for all the variables for all the studies, i.e., CAL, depth of intrabony defect, probing depth, and gingival margin level [Table 2]. The mean of all the mean values at baseline and follow-up each for test and control group was obtained for each variable of all the included studies. The actual quantitative mean gains were calculated for all these parameters (CAL, PD, IBD, and GML) [Table 3].

Table 2

Mean change in clinical and radiographic parameters of included studies

Study name	Mean change±SD (mm)
	CAL		IBD		PD		GMP
	Test	Control	Test	Control	Test	Control	Test	Control
Ilgenli et al.	1.5±0.7	-	0.6±1.2	-	2.1±0.5	-	-	-
Markou et al.	3.08±0.95	-	6.09±1.92	-	3.92±1.1	-	0.58±1.19	-
Pradeep et al.	3.21±0.58	-	1.97±0.15	-	3.5±0.65	-	−0.29±0.47	-
Lekovic et al.	2.18±0.70	-	2.13±0.66	-	3.29±0.70	-	1.09±0.37	-
Thorat et al.	4.13±1.63	2.13±1.71	2.12±0.69	1.24±0.69	4.69±1.45	3.56±1.09	−0.31±0.95	−1.31±1.01
Sharma et al.	3.31±1.76	2.77±1.44	−2.50±0.78	−0.09±0.11	4.55±1.87	3.21±1.64	−0.1±0.08	0.67±0.46
Pradeep et al. (PRF)	3.17±1.29	2.83±0.91	2.8±0.89	0.13±1.46	3.77±1.19	2.97±0.93	0.2±0.71	−0.27±0.58
Pradeep et al. (PRP)	2.93±1.08	-	2.7±0.79	-	3.77±1.07	-	0.1±0.61	-
Pradeep et al.	3.03±1.16	2.67±1.09	3.2±0.89	0.93±0.83	3.9±1.09	2.97±0.93	0.47±0.73	−0.17±0.53
Rosamma et al.	4.73±0.88	1.40±1.06	1.93±1.07	0.64±0.5	4.67±0.90	2.40±0.63	−0.07±0.26	1.13±0.74
Ajwani et al.	1.80±0.632	1.30±0.675	1.45±0.497	0.80±0.350	1.90±0.738	1.60±0.843	−0.30±0.483	−0.30±0.675
Jain Gupta et al.	−1.87±0.91	-	−1.67±1.17	-	−1.80±0.77	-	-	-
Mathur et al.	−2.53±1.06	-	1.73±1.62	-	−2.67±1.29	-	0.07±0.46	-
Pradeep et al.	4.03±0.18	2.96±0.18	2.53±0.30	0.49±0.27	4.00±0.18	3.00±0.18	0.27±0.07	−0.06±0.04
Suchetha et al. (PRP)	5.100±0.32	-	3.800±0.89	-	6.050±0.56	-	0.050	-
Suchetha et al. (PRF)	5.950±0.51	-	4.850±0.77	-	6.500±0.74	-	0.050	-
Shah et al.	2.97±1.56	-	-	-	3.67±0.69	-	−0.42±1.38	-
Chatterjee et al.	6.57±1.45	4.14±0.76	-	-	5.46±1.04	3.68±0.72	-	-
Chandradas et al.	3.27±0.65	2.25±0.62	2.30±0.83	1.22±0.62	3.82±0.75	3.00±1.21	−0.18±0.40	−1.33±0.78
Pradeep et al.	3.30±0.65	2.47±0.77	3.17±0.65	1.43±0.50	4.03±0.18	3.10±0.30	-	-
Kanoriya et al.	4.2±0.66	3.03±0.18	2.42±0.21	0.38±0.26	3.7±0.91	2.86±0.68	0.24±0.056	−0.06±0.07
Martande et al.	3.40±1.13	2.50±1.33	2.46±0.33	0.27±0.19	3.76±1.12	2.76±1.43	0.22±0.10	0.06±0.02
Agarwal et al.	4.10±1.47	1.27±0.89	1.83±0.46	0.33±0.08	4.86±2.12	2.69±1.37	2.04±1.15	3.31±0.54
Chadwick et al.	1.03±0.86	-	1.35±1.60	-	2.12±1.41	-	1.06±1.18	-
Galav et al.	4.5±0.61	-	1.16±0.34	-	4.1±0.63	-	-	-
Jalaluddin et al.	3.7±0.04	3.5±0.12	0.80±0.79	0.80±1.03	4.3±0.18	3.9±0.06	0.46±0.1	0.47±0
Thorat et al.	4.16±0.7	1.16±1.32	3.09	1.67	3.83±0.75	1.33±0.42	0.33±0.51	−0.16±0.40

CAL – Clinical attachment level; PD – Pocket depth; IBD – Intrabony defect depth; PRF – Platelet-rich fibrin; PRP – Platelet-rich plasma; SD – Standard deviation; GMP – Gingival margin position

Table 3

Actual quantitative mean gain considering all studies together

Serial number	Study name	CAL				IBD				PD				GML
		Test		Control		Test		Control		Test		Control		Test		Control
		Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up	Baseline	Follow-up
1	Ilgenli et al.[22]	8.4	6.9	-	-	4.7	4.1	-	-	7.5	5.4	-	-	-	-	-	-
2	Markou et al.[23]	7.8	4.1	NR	NR	6.0	3.6	-	-	6.9	3.4	-	-	-	-	-	-
3	Pradeep et al.[24]	9.7	6.5	NR	NR	5.0	2.4	-	-	7.7	4.2	-	-	2.1	2.4	-	-
4	Lekovic et al.[25]	NA	NA	NR	NR	NA	NA	NR	NR	7.8	4.5	-	-	-	-	-	-
5	Thorat et al.[26]	7.7	3.6	6.5	4.4	4.5	2.3	4.4	3.1	7.9	3.2	6.7	3.2	0.5	0.8	0.2	1.5
6	Sharma et al.[27]	6.9	3.6	6.2	3.5	5.2	2.7	5.0	4.9	8.6	4.1	8.1	4.9	0.9	0.8	0.9	1.6
7a	Pradeep et al.[29] (PRF)	6.2	3.1	6.3	3.4	5.1	2.3	4.8	4.7	7.8	4.1	7.8	4.8	1.5	1.3	1.6	1.9
7b	Pradeep et al.[29] (PRP)	6.2	3.2	6.3	3.4	4.9	2.2	4.8	4.7	7.9	4.1	7.8	4.8	1.6	1.5	1.6	1.9
8	Pradeep et al.[28]	6.4	3.4	6.6	3.9	5.6	2.4	5.8	4.9	8.2	4.3	8.1	5.1	1.6	1.2	1.8	1.9
9	Rosamma et al.[30]	8.2	3.5	7.5	6.1	5.1	3.1	4.6	3.9	7.5	2.9	7.1	4.7	0.7	0.6	0.3	1.5
10ǂ	Ajwani et al.[31]	9.5	7.7	9.2	7.9	3.9	2.5	3.7	2.9	5.9	4.0	6.2	4.6	3.6	3.9	3.4	3.7
11	Jain Gupta et al.[32]	6.8	4.9	NR	NR	4.9	3.2	NR	NR	6.2	4.4	NR	NR	NA	NA	NR	NR
12	Mathur et al.[33]	6.9	4.3	NR	NR	3.1	1.3	NR	NR	7.7	5.0	NR	NR	9.4	9.5	NR	NR
13ǂ	Pradeep et al.[34]	7.6	3.6	7.7	4.8	5.2	2.7	5.3	4.8	8.6	4.6	8.8	5.8	1.6	1.4	1.6	1.7
14a#	Suchetha et al.[35] (PRP)	8.4	3.3	NR	NR	8.0	4.2	NR	NR	10.3	4.2	NR	NR	−1.9	−0.9	NR	NR
14b#	Suchetha et al.[35] (PRF)	9.2	3.2	NR	NR	9.0	4.1	NR	NR	10.5	4.0	NR	NR	−1.3	−0.8	NR	NR
15	Shah et al.[36]	12.2	8.7	NR	NR	NA	NA	NR	NR	7.1	3.3	NR	NR	5.7	5.5	NR	NR
16	Chatterjee et al.[37]	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
17ǂ	Chandradas et al.[38]	12.1	8.8	11.9	9.7	5.3	3.0	5.1	3.9	8.3	4.4	8.3	5.3	0.7	0.9	0.5	1.8
18	Pradeep et al.[40]	5.7	2.4	5.6	3.2	6.0	2.8	5.9	4.4	7.8	3.8	7.9	4.8	NA	NA	NA	NA
19	Kanoriya et al.[39]	7.6	3.4	7.6	4.5	5.2	2.8	5.1	4.8	8.0	4.3	8.0	5.2	1.7	1.4	1.6	1.7
20ǂ	Martande et al.[41]	7.3	3.9	6.9	4.5	5.1	2.7	5.0	4.7	8.2	4.5	7.9	5.1	1.6	1.4	1.7	1.8
21	Agrawal et al.[42]	12.3	8.2	12.2	10.9	5.5	3.7	5.4	5.0	9.3	4.4	9.6	6.9	5.9	7.9	6.2	9.5
22	Chadwick et al.[43]	6.3	5.3	NR	NR	7.7	6.4	NR	NR	6.9	4.8	NR	NR	−0.6	0.5	NR	NR
23ǂ	Galav et al.[44]	12.6	8.7	NR	NR	5.8	4.6	NR	NR	7.3	3.2	NR	NR	NA	NA	NR	NR
24	Jalaluddin et al.[45]	12.7	9.0	11.7	8.2	2.9	2.1	3.1	2.3	8.0	3.7	7.3	3.4	2.4	1.9	2.5	2.1
25	Thorat et al.[19]	8.7	4.5	8.2	7.0	6.9	3.8	6.3	5.7	8.0	4.2	7.7	6.3	0.7	0.3	0.5	0.7
Mean		8.54	5.11	8.02	5.69	5.44	3.12	4.95	4.31	7.92	4.12	7.82	4.99	1.92	2.07	1.74	2.37
Actual quantitative		3.43		2.33		2.32		0.64		3.8		2.83		−0.15		−0.63
Mean gain		1.10				1.68				0.97				0.48

ǂStudies that mentioned relative attachment loss, #CAL is less than PPD. CAL – Clinical attachment level; PD – Pocket depth; IBD – Intrabony defect depth; GML – Gingival marginal level; PRF – Platelet-rich fibrin; PRP – Platelet-rich plasma; NA – Not available; NR – Not reported

For obtaining AQMG, all the baseline means for a particular parameter were added together synthesizing a new combined mean, and a difference was calculated by similarly adding follow-up means synthesizing a new overall mean for control group first. The same steps were followed for the test group for the said parameter. The difference between the overall means thus obtained from the control and test groups for the said parameter was considered as the AQMG, which was then used for the purposes of noting clinical significance.

Preparation protocol for platelet-rich plasma and platelet-rich fibrin

Although it is advisable to include all the studies following the same and standardized protocol for producing PRP or PRF, it is practically near to unattainable. The studies are coming from all over the world with differing climatic, academic, economic, and research environments, and also depending on indigenous armamentarium available. The centrifuge, the refrigerated centrifuge, various types of test tubes and variable amount of blood withdrawn, centrifugation time and gravity, etc., all make the process even more complicated. We were more interested in the effects of end-products and thus selected all the studies using them irrespective of the protocol they followed. A detailed description of the armamentarium and protocol followed by each study is available as a supplementary file [Supplementary File 1].

Quality assessment

All the included studies were analyzed for quality as per the Consolidated Standards of Reporting Trials statement.[2021] They were checked for sequence generation, allocation concealment, masking of the examiners, addressing incomplete outcome data, and free of selective outcome reporting. The selected RCTs were characterized as low risk if all the criteria met, moderate risk when only one criterion was missing, and high risk if two or more criteria were missing. We did not exclude any study for its risk of bias [Table 4].

Table 4

Quality analysis of included studies

Study	Study design				Any other factor that can affect bias	Risk of bias
	Sample randomization	Allocation concealment	Masking of examiners	Loss to follow-ups
Ilgenli et al.[22]	First 3 months 1 group then later 3 months other group	No	Yes	No loss to follow-up	Not clear	High
Markou et al.[23]	Yes	Yes	Yes	No loss to follow-up	Parallel design	Low
Pradeep et al.[24]	Yes	Yes	Yes	No loss to follow-up	Split mouth	Low
Lekovic et al.[25]	Yes	Yes	Yes	No loss to follow-up	Split mouth	Low
Thorat et al.[26]	Yes	No	Yes	No loss to follow-up	Not clear	Moderate
Sharma and Pradeep[27]	Yes	Yes	Yes	Seven patients	Parallel design	Moderate
Pradeep et al.[28]	Yes	No	Yes	Four patients	Parallel design	Moderate
Pradeep et al.[29]	Yes	No	Yes	Five patients	Parallel design	Moderate
Rosamma Joseph et al.[30]	Yes	Not clear	Not clear	No loss to follow-up	Split mouth	Moderate
Ajwani et al.[31]	Yes	Not clear	Yes	No loss to follow-up	Split-mouth	Moderate
Jain Gupta et al.[32]	Yes	No	Yes	No loss to follow-up	Not clear	Moderate
Mathur et al.[33]	Yes	No	No	No loss to follow-up	Not clear	High
Pradeep et al.[34]	Yes	Yes	Yes	6 patients	Parallel	Low
Suchetha et al.[35]	Yes	No	No	No loss to follow-up	Not clear	High
Shah et al.[36]	Yes	No	Yes	No loss to follow-up	Split mouth	Moderate
Chatterjee et al.[37]	Yes	Not clear	Yes	6 patients	Not clear	Moderate
Chandradas et al.[38]	Yes	Not clear	Yes	No loss to follow-up	Not clear	Moderate
Kanoriya et al.[39]	Yes	Yes	Yes	No loss to follow-up	Parallel	Low
Pradeep et al.[40]	Yes	Yes	Yes	6 patients	Parallel	Low
Martande et al.[41]	Yes	Yes	Yes	No loss to follow-up	Parallel	Low
Agrawal et al.[42]	Not clear	Not clear	Not clear	2 patients	Split-mouth	High
Chadwick et al.[43]	Yes	Yes	Yes	5 patients	Parallel	Low
Galav et al.[44]	Yes	Not clear	Not clear	No loss to follow-up	Not clear	High
Jalaluddin et al.[45]	Yes	Not clear	Not clear	No loss to follow-up	Not clear	High
Thorat et al.[19]	Yes	Yes	Yes	No loss to follow-up	Split mouth	Low

Statistical analysis

All the collected data were appropriately tabulated and subjected to simple statistical methods including sums and means for easy understanding. Meta-analyses and forest plots were obtained from SPSS software (Statistical Package for Social Sciences, version 19.0, IBM, Armonk, New York, USA) from a qualified statistician. Funnel plots were used to discern publication bias. Cohen's d was used for effect size estimation.

RESULTS

The search strategy was carried out according to PRISMA format. A total number of 4852 records were identified from the electronic and hand search. Records were excluded after evaluation of title and abstracts and according to inclusion/exclusion criteria and a total number of 256 records were screened. Again after excluding 158 ineligible manuscripts, a total of 98 full-text articles were assessed for final inclusion. Out of 98, 73 articles were further excluded reason being inappropriate control group, use of BGs and other materials etc. Finally a total number of 25 articles[19222324252627282930313233343536373839404142434445] were included for both quantitative and qualitative synthesis and further meta-analysis [Figure 1].

Figure 1

Search strategy according to Preferred Reporting Items for Systematic Review and Meta-Analyses format. n – Number

A total of 844 intrabony defects were analyzed from 25 included manuscripts (325 were control defects with only OFD and 519 were tests defects with OFD and PRP/PRF). The mean preoperative CAL (considering all the included studies) in the test group was 8.54 mm, which was reduced to 5.1 mm at follow-up visits giving an actual mean CAL gain of 3.43 mm, which at the outset looks quite impressive whereas the mean preoperative CAL in the control group was 8.02 mm, comparable to the same in the test group, which was reduced to 5.69 mm at follow-up visits giving an actual mean CAL gain of 2.33 mm, which again looks clinically significant. However, if we calculate actual quantitative mean CAL gain over and above that of OFD for test group (3.43 mm minus 2.33 mm), the difference is hardly 1.1 mm.

Similar results were obtained with probing depth and GML. The actual quantitative mean gain in probing depth and GML over and above that of OFD was only 0.97 mm and 0.48 mm, respectively.

Somewhat better results were noted with gain in intrabony defect fill measurements. The mean preoperative depth of intrabony defects in the test group was 5.44 mm which was reduced to 3.1 mm giving a mean bone gain of 2.32 mm whereas the mean preoperative depth of intrabony defects in the control group was 4.95 mm which was reduced to 4.31 mm giving a mean bone gain of only 0.64 mm. The actual quantitative mean intrabony depth gain over and above that of OFD was 1.68 mm.

A very high degree of heterogeneity has been noted between all the studies as depicted on forest plots for all the parameters (i2 for CAL = 97%, i2 for IBD = 98%, i2 for PPD = 94% and i2 for GML = 96%). This indicates inconsistent studies where the outcomes may be due to some reasons other than chance occurring. The overall test statistic in forest plots for CAL is 0.39 (95% confidence interval [CI], Z = 18.72, P < 0.00001) [Graph 1], IBD is 1.65 (95% CI, Z = 41.42, P < 0.00001) [Graph 2], PPD is 0.68 (95% CI, Z = 25.18, P < 0.00001) [Graph 3], and GML is 0.24 (95% CI, Z = 20.86, P < 0.00001) [Graph 4] favoring test group.

Graph 1

Meta-analysis of the mean difference in clinical attachment level gain

Graph 2

Meta-analysis of the mean difference in IBD depth reduction

Graph 3

Meta-analysis of the mean difference in pocket depth reduction

Graph 4

Meta-analysis of the mean difference of the change in gingival marginal level

Although we expected a lot of heterogeneity among studies regarding the preparation of PRP and PRF, many of the studies using PRF followed similar protocols. 50% of the studies utilizing PRP followed a uniform protocol, others followed differing protocols. 18 studies out of 20 used identical protocol for producing PRF; some were differing only in centrifugation force and timings. A standardized protocol which provides the best possible results needs to be proposed with consensus.

DISCUSSION

Why another meta-analysis on platelet concentrates required?

Till date, there are 5 meta-analyses available on the subject of PRP and/or PRF use for treating intrabony defects. Although these analyses covered some aspects of PC use, there were many questions unanswered. There are subtle differences between our meta-analysis and the ones published earlier. As far as our study goes, we were clear about our aims of calculating only the effects of PRP/PRF placed as sole grafting agent (monotherapy) over and above that of OFD alone. Hou et al.[9] analyzed use of PRP in all the possible combinations, like, PRP + HA, PRP + BG, etc., in intrabony defects. They observed no difference in outcomes between GTR versus PRP-combinations. They suggested that PRP may be beneficial as an adjunct to graft materials. Castro et al.[13] included studies incorporating leucocyte- and platelet-rich fibrin together with intrabony defects, furcation defects, and periodontal plastic surgery. They analyzed very few studies in subanalysis of intrabony defects whereas we included RCTs irrespective of the type of PRF. Del Fabbro et al.[14] evaluated platelet concentrates in the treatment of intrabony defects, furcation, and gingival recession. All the included studies evaluated adjunctive effect of PRP. This meta-analysis is similar to Hou et al.[9] with similar results. Kotsovilis et al.[15] also reported on similar lines with Hou et al.[9] and Del Fabbro et al.,[14] where they evaluated the adjunctive effect of PRP. They suggested that because of lot of heterogeneity in agents and procedures, there were diverse outcomes and it would be better to perform meta-analysis on individual agents or procedures. Our study stands clear on specific parameters.

The only meta-analysis similar to ours is done by Shah et al.[12] Their study included only 5 RCTs and they also considered a combination of procedures along with OFD as a comparator group. In contrast, we have purely considered OFD alone as control and expanded the control population by considering OFD groups from other RCTs as well. Shah et al. found “clinically significant” differences between all the parameters, on what basis they concluded this remains uncertain. Roselló-Camps et al.'s[11] meta-analysis was also focused on PRP similar to Hou et al.,[9] Del Fabbro et al.,[14] and Kotsovilis et al.[15] They compared the PRP as an adjunct only versus various grafting procedures and concluded about high heterogeneity and marginal beneficial effects. The focus of our study was use of PRP/PRF as monotherapy and its benefits over OFD.

Publication and other biases

A meta-analysis of randomized controlled trials still occupies the top spot for indicating the best evidence, but it also comes with its own sets of problems. An important problem in underevaluation of a meta-analysis is publication bias,[46] which may be because of investigators (not writing up and nonsubmission of study, fear of rejection), peer-reviewers (personal bias, geographical bias, language bias, influence bias, preconceived ideas about the study), editors (similar to peer reviewers, accepting publications from reputed authors, rejecting studies with negative or small effect sizes, poorly written manuscripts being rejected, acceptance associated with target audience), and funding bodies (influencing the whole process from investigators to publication). One of the ways to assess publication bias is funnel plot[46] (scatter plots of effect estimates against sample size) and a methodology to assess the number of studies potentially missed is capture-recapture method by Bennett et al.[47]

As the sample sizes of included studies were differing too much and few were small, we used Hedge's g statistic (instead of Cohen's d) for funnel plotting to discern publication bias. To be sure about any bias, two funnel plots representing two different variables for effect sizes, CAL and intrabony defect fill, were used. Both the funnel plots indicated publication bias. In addition, we calculated final effect sizes[48] of individual studies using Cohen's d analysis for IBDs and noted an average of 0.73 effect size. More than 60% studies showed higher than average effect sizes which are in between medium and large effect as per Cohen's d [Supplementary File 2].

Prospective registration of clinical trials may reduce publication and research dissemination bias. Our meta-analysis is based on reported means of groups of patients from individual studies. A meta-analysis based on individual patient data would be much useful than a meta-analysis based on extracted means. A subgroup analysis of a huge number of patients will be possible only if individual patient raw data are available in published RCTs. We could have done a subgroup analysis of different depths of intrabony defects treated with either of the procedures in a better way for more inclusive and dependable results, as a 3-mm IBD may respond differently to a 6-mm IBD. Hereby, we request the editors of all the journals to demand and publish individual patient raw data, may be as a supplementary file, from all the clinical trials. It would be of immense help for future advanced meta-analyses.

Study characteristics

As far as 519 defects are considered for overall analysis, which were treated with OFD and PRF/ PRP, it is a good number of patient data on which further decisions about treatment may be based with some certainty. The basic aim of our meta-analysis was to quantitatively go through all these defect parameters and come at a conclusion about its effectiveness over and above that of OFD alone. We considered all the studies irrespective of supposed bias as it may average out any erroneous or overlooked measurement(s).

The overall CAL gain obtained for OFD with PRF over and above that of OFD alone is just 1.1 mm, which by any standard looks suboptimally below than that to be considered as clinically significant. A visible and perceptible gain should be at least 2 mm.[17] It is to the astute clinician to put a prospective patient to be treated with more invasive procedure for a notional CAL gain of 1 mm or go ahead with other time-tested and more dependent procedure.

Although there is a wide difference of mean change in CAL for OFD with PRF/ PRP, ranging from 1.03 to 6.57 mm, the calculated mean considering all studies stands at 3.43 mm, which is clinically as well as statistically significant. The corresponding mean changes for OFD alone ranged from 1.16 mm to 4.14 mm with combined calculated mean of 2.33 mm, which is again significant both clinically and statistically. Individually, both the therapies are promising in CAL reduction, but the quantitative overall gain over and above than that of OFD alone is negligible.

The results for intrabony defect fill or resolution were better for OFD with PRF, which may be a positive indication for suggesting PRF technologies. A mean gain of 0.64 mm in the control group looks to represent almost negligible bone formation whereas a gain of 2.32 mm for OFD with PRF/ PRP is both statistically and clinically significant. We have observed a wide range of mean IBD change from 0.6 to 6.09 mm for test group and 0.09 to 1.67 mm for the control group, where the results for OFD alone looks more consistent as opposed to OFD with PRF/ PRP. A negligible bone fill of 0.64 mm for OFD alone may represent some observer bias. The reasons for such wide range of bone formation with PRF technologies may be varied, including but not limited to, initial defect morphology, background factors, variations in obtaining PRF/ PRP, handling of PRF/ PRP, platelet counts, experience of the surgeon, follow-up period, heterogeneity of target populations, observer bias, etc.

The PRF technology looks promising biologically rather than clinically. Although there is marginal difference in CAL gain, PPD reduction, and GML maintenance, bone gain takes precedence in OFD with PRF/ PRP. In retrospective, this indicates that the majority of healing occurring in OFD alone is the formation of long junctional epithelium or close epithelial adaptation. Such kind of healing has been reported well in the literature.[49] The bone formation in case of OFD with PRF/ PRP may indicate true periodontal regeneration, namely de novo formation of periodontal ligament, cementum, and alveolar bone, but the true histological picture is missing. Similar bone formation has been reported with other grafting technologies[5051] such as autogenous BGs, alloplasts, enamel matrix derivatives, growth factors, and GTR. The ultimate aim of a Periodontist is to regenerate lost periodontal tissues, which is reflected clinically as clinical attachment gain. CAL gain may happen secondary to healing by long junctional epithelium or true periodontal regeneration. Whether CAL gain by periodontal regeneration remains stable over a long period of time as compared to CAL gain by long junctional epithelium is yet to be elucidated. For that matter, if anybody is interested in treating intrabony defects with a final aim of reducing pocket depth and gaining attachment loss to a clinically maintainable state without bothering about alveolar bone gain, then OFD appears to be the best option with long-term maintenance program. However, if we can convincingly prove that a structured and augmented periodontium is the best bet for future occurrence and progression of periodontal disease, then we should strive for regeneration technologies at any cost. The studies on GTR around hopeless teeth appear to support regenerative techniques, as far as reduction in mobility and long-term morbidity is concerned.[52]

Out of a total of 16 estimable studies for forest plots, 5 showed nonsignificant results for CAL. Other studies although favoring test, the quantum of outcome was less than 0.5 mm, which is hardly of any consequence. Similar results were obtained for PPD (0.68) and GML (0.24).

The results for IBD on forest plots were better as compared to other parameters. Out of 14 estimable studies, only one showed nonsignificant results. All other studies were favoring the test group, and the quantum of change for overall effect was almost reaching 2.00 mm, which is clinically as well as statistically significant.

Limitations

We have studied PRP and PRF together on the premise that both are autologous platelet concentrates and both have supernatural concentrations of growth factors which are essentially the same. There may be some contrary claims, but the concentrations of growth factors (PDGF-BB, TGF-β1, IGF-1, and VEGF) did not differ in CGF, PRP, or PRF.[53] Still, the consideration that both will have similar healing potential may not be absolutely true. The PRP contains anticoagulants which may inhibit wound healing,[54] and its more liquid nature may not support sustained release of growth factors. These observations cannot be substantiated from other disciplines,[55] and a contrary effect has been noted. Considering both as autologous growth factor concentrates with comparable efficacy, we have combined them for a larger analysis. Ideally, they can be studied separately.

Considerable heterogeneity exists between studies, which is at present unavoidable. We found positive publication bias and no study reporting negative results in favor of the test group. We included all the studies irrespective of their quality analysis and risk of bias.

Effect sizes vary too much indicating unknown variables responsible for the same. A sub-group analysis of various IBDs would have immensely added to the knowledge database, but individual data are not available with any of the studies.

We found equitable international representation lacking for these studies as majority of the studies are from Indian Subcontinent.

Calculation of actual quantitative mean gains was from raw preoperative and postoperative data available with the included studies for test as well as control groups. These data can be considered consistent and dependable as per the common and uniform inclusion criteria of studies for this meta-analysis although the studies in general may be heterogeneous.

CONCLUSION

The search for a definitive and dependable periodontal regenerative material is still elusive. Biology may not be replicated unless it meets biologically identical product. The biologics regarding platelet concentrates may not meet ideal requisites as regenerative products and on their own may remain insufficient. Considering all the limitations, we conclude that, use of platelet concentrates (PRF/PRP) as sole grafting agents in periodontal intrabony defects does have an identifiable superiority over not using them during access flap surgeries in terms of only intrabony defect fill (IBD fill). The superiority in terms of clinical parameters (CAL gain, PPD reduction, and GML) is negligible. Therefore, use of PRP/PRF can be recommended, with some reservations, as sole grafting material considering its potential for bone fill irrespective of negligible CAL gain and PPD reduction. Considering overall moderate effect sizes in favor of PRP/PRF even for bone fill, the recommendation should be taken with caution. We did not find any outcome to recommend positively use of PRF/ PRP technologies for treating periodontal intrabony defects if your aim is CAL gain and PPD reduction. There is a need to identify biological cascades and other related factors which are responsible for a wide range of almost negative to highly superior results among analyzed studies. A dedicated large sample size RCT should be carried out to substantiate the findings of this meta-analysis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

Armamentarium and protocols used for platelet-rich plasma/platelet-rich fibrin preparation

Funnel plots for publication bias and effect size estimation