The efficacy of platelet-rich fibrin (PRF) in post-extraction hard and soft tissue healing and associated complications: a systematic review and meta-analysis of split-mouth randomized clinical trials

Abstract

Background

Tooth loss is often followed by potential complications, including restricted mouth opening, edema, regional pain, and alveolar osteitis (AO), all of which can impair socket healing and quality of life, compromising the ability to socialize and speak post-extraction. One of the approaches currently used to minimize adverse effects is to use platelet-rich fibrin (PRF) immediately following tooth extraction to reduce these complications. This study aims to evaluate the effectiveness of platelet-rich fibrin in promoting soft and hard tissue healing, and its effects on post-extraction complications.

Method

A systematic review and meta-analysis were conducted, following an electronic and manual search of PubMed, EMBASE, Web of Science, and Scopus databases. A total of 455 articles were screened, of which 41 articles underwent systematic review, and 21 studies (701 patients) were included for meta-analysis. Statistical analyses were performed using Comprehensive Meta-Analysis software. The risk of bias was evaluated by three assessors based on five domains. To evaluate PRF effectiveness, the focus was on split-mouth designs to minimize inter-patient variability.

Result

PRF significantly improved soft tissue healing (eight studies, p < 0.05), reduced AO incidence (four studies, p < 0.05), and lowered post-operative pain (eleven studies, p < 0.05). However, PRF did not demonstrate statistically significant improvement in bone healing (three studies, p > 0.05).

Conclusion

The analysis showed that PRF significantly improved soft tissue healing and reduced the incidence of AO. While PRF helped reduce pain, the results were inconsistent and influenced by the timing of post-operative assessments. The impact on hard tissue healing was controversial, and future research should explore alternative methods to evaluate PRF’s effects on bone regeneration.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12903-025-06238-1.

Introduction

Teeth generally are extracted to avoid various complications; such as pain, infection, and pericoronitis, or to gain space in orthodontic treatment plans. Tooth loss is often followed by potential complications, including restricted mouth opening, edema, regional pain, and alveolar osteitis (AO), all of which can impair socket healing and quality of life, compromising the ability to socialize and speak post-extraction [1].

Socket healing is a complex biological process influenced by multiple factors, including prolonged inflammation (remaining inflammatory neutrophils can degenerate proteins essential for healing e.g., cytokines, clotting factors, and immunoglobulins, by releasing proteases and free oxygen radicals) [2], flap design (it is essential to be well-structured for primary wound closure and optimal perfusion) [3], medication use(e.g., anticoagulants which prevent clot formation) [4], local factors(oxygenation which is crucial to sustaining the healing, and bacterial infection) [5, 6], and systemic conditions(e.g., smoking, alcohol use, diabetes). Studies have demonstrated that nicotine has vasoconstrictive effects which can lead to decreased wound perfusion and harms healing [7, 8]. Alcohol consumption can suppress healing by disrupting angiogenesis, collagen production, and wound closure. Diabetes impairs the healing process by reducing wound perfusion, which leads to hypoxia [9, 10].

PRF, this autologous blood-derived biomaterial, can be produced simply by centrifuging a patient’s blood without any additives. It comprises a fibrin meshwork in which a higher amount of platelets and leukocytes are trapped. Likewise, the increased contents of cytokines and various growth factors are embedded in the fibrin matrix, allowing release in a long-lasting period [11, 12].

PRF as a second-generation method for concentrating the platelets can be widely used to promote healing. There are various types of PRF. Firstly leukocyte platelet-rich fibrin (L-PRF) was introduced by Choukroun et al. [13]. Later advancements in PRF such as titanium platelet-rich fibrin (T-PRF), advanced platelet-rich fibrin (A-PRF), albumin platelet-rich fibrin (Albumin-PRF), and injectable platelet-rich fibrin (I-PRF) created by making changes in preparation methods through the years. Preparation methods vary based on factors such as centrifugation speed, centrifugation time, tube type, and the state of matter. For example, L-PRF and T-PRF are similar, in terms of preparation method (centrifugation speed: 2700 rpm, duration: 12 min, and solid-state) but they differ only in tube type. L-PRF is prepared in a glass tube, but T-PRF is prepared in a titanium tube [14]. Likewise, A-PRF and Alb-PRF have a similar preparation method (centrifugation speed: 1300 rpm, and solid nature) but they differ in terms of centrifugation time. 14 min for A-PRF and 8 min for Alb-PRF [15, 16]. I-PRF was created by reducing the speed and duration in the previous centrifugation method, resulting in a liquid state, which allows for injection into soft tissue, skin, or mucous membrane [17].

PRF not only results in faster soft tissue regeneration, proliferation of osteoblasts, and accelerating wound healing but also reduces pain and postoperative inflammatory complications [18, 19]. Nevertheless, some studies have reported no statistically significant difference between PRF and the control group and the efficacy of using PRF is still under debate [20–22].

Despite numerous studies evaluating the effectiveness of PRF compared to other biomaterials, few systematic reviews specifically contrast PRF with spontaneous healing. This gap limits our understanding of PRF’s true efficacy in enhancing post-extraction healing. To address this, we conducted a thorough review with a large sample size to evaluate PRF’s clinical utility in extraction sockets.

Notably, prior reviews have not focused on split-mouth studies, which offer a unique advantage. Split-mouth designs enable direct intra-patient comparison, reducing inter-individual variability. This minimizes confounding factors like systemic diseases or lifestyle differences between patients. By using the same patient as both control and experimental subjects, bias is significantly reduced. Such methodology enhances the internal validity and reliability of outcome assessments. Our study specifically targets split-mouth RCTs to clarify the biological impact of PRF. This approach offers stronger evidence for the clinical adoption of PRF in socket healing.

Methods

Protocol

This systematic review was conducted following PRISMA guidelines [23], ensuring adherence to standards.

PICO question

Employing the PICO framework facilitates the formulation of the focused question, from which the inclusion criteria should be derived:

• Patient/Problem (P): Patients undergoing tooth extraction.

• Intervention (I): Application of PRF in socket.

• Comparison (C): Spontaneous healing)without using PRF and other biomaterials)

• Outcome (O): Evaluation of wound healing by outcome measures included soft and hard tissue healing, post-operative pain, and alveolar osteitis (AO).

• Study design (S): Randomized clinical trials (RCTs) and split-mouth studies.

Focused question

Does the application of PRF offer advantages in the healing and complications of post-extraction sockets compared to spontaneous healing?

Inclusion criteria

Studies were included if they met the following criteria:

• Studies comparing PRF with spontaneous healing or suturing as a control group were included to isolate PRF’s effect. This approach was chosen to avoid confounding results from comparing two different treatments (PRF group and non-PRF group), allowing for a more accurate assessment of PRF’s efficacy.

• Randomized controlled trials with split-mouth design.

Exclusion criteria

Studies were excluded if they met the following criteria:

• Inaccessible studies.

• Non-English articles.

Data sources and search strategy

Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Cochrane Handbook, a literature search strategy was developed with input from the research team members. The PICO method was employed to facilitate the creation of the search strategy, which was tailored for each database.

From inception to June 2024, four databases—PubMed, Web of Science, Scopus, and the Excerpta Medica database (EMBASE)—were searched to identify relevant publications. These four databases were selected for their comprehensive coverage of biomedical, clinical, and multidisciplinary research. They collectively reduce publication bias and ensure the inclusion of high-quality, peer-reviewed studies. Their combined use enhances the transparency and reliability of the literature search process. Details of the search strategy used for each database are provided in Supplementary Table 1.

Data collection

Three independent reviewers utilized the Rayyan software, a web-based tool designed to streamline systematic reviews [24], to screen studies based on a range of criteria including the title, abstract, author’s name, study design, journal, and publication date. This method allowed for efficient and organized filtering of relevant studies. In cases where discrepancies arose between the reviewers regarding the risk of bias assessment using the Cochrane Risk of Bias tool (ROB-2), these were resolved through a third-party review or consensus discussion, ensuring transparency and consistency in the assessment process. This approach fostered clear communication and helped in reaching a unified decision, minimizing any subjective judgment.

To further enhance objectivity, the Rayyan application facilitated the blinding process, preventing reviewers from being influenced by one another’s decisions, thus maintaining the integrity of the screening process. Data extraction was conducted systematically for each included study, using a standardized form to ensure consistency and accuracy across the review. The following variables were meticulously collected from each study and summarized in a table for clarity and comparison: author information, country of origin, sample size, mean age of participants, gender distribution, health condition of the subjects, assessment tools used, follow-up sessions and duration, attrition rate, and details regarding the preparation of PRF by centrifuge, including centrifugation speed and duration. This comprehensive approach ensured that all relevant data were captured to support the subsequent analysis and synthesis of findings.

Risk of bias

Three independent reviewers assessed the risk of bias using the Cochrane ROB-2 tool [25], across five domains: Randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported results. Also, it contains an overall biased judgment based on an algorithm. (Figures 1 and 2)

Fig. 1.

Summary of bias risk in each article

Fig. 2.

The graph of bias risk in detail

Data synthesis and analysis

All statistical analyses were conducted using Comprehensive Meta-Analysis version 3.3.070 utilizing a random-effects model. For continuous outcomes, we calculated the mean difference (MD) along with 95% confidence intervals (CI). An I² statistic value greater than 50% was deemed to signify significant heterogeneity [26]. When heterogeneity is detected, publication bias analysis is performed using funnel plots, and Egger’s regression and publication bias are assessed.

Study selection

After the initial search, which included both electronic and manual searches, 6284 articles were identified, with an additional 10 articles sourced from citation searches. The EndNote software was used to remove 1324 duplicate articles. In addition, 4515 articles were deleted because they were irrelevant in the Endnote search, leaving 455 articles for screening. In the first stage of screening, based on the titles and abstracts, 403 articles were excluded, and 52 articles were indicated relevant for further evaluation. One article was excluded due to lack of access to the full text. In the second stage, following the full-text review, 10 articles were excluded because they were unrelated to the definite outcomes of the current study. For instance, they assessed implant stability after using PRF which was inconsistent with ours [27–36]. This two-stage screening process ensured that only the most relevant studies were included in the final analysis. Afterward, 41 articles ultimately were included to assess the risk of bias [19, 21, 22, 37–74] (Figs. 1 and 2). Of these, 21 articles were eligible for quantitative analysis and included in the meta-analysis [19, 21, 37, 39, 40, 42, 45, 46, 49, 51, 53, 61, 63–67, 70, 72–74]. (Fig. 3)

Fig. 3.

Flowchart of included studies based on the PRISMA guidelines

Study characteristics

A total of 41 studies published between 2010 and 2024 were included. These studies were conducted in the following countries: 12 studies in India [21, 41, 45, 53, 54, 56, 57, 64–67, 74], 10 studies in Turkey [37, 42, 43, 51, 52, 55, 60, 61, 70, 72], three studies in Brazil [19, 62, 63], two studies in Iran [49, 68], two studies in Italy [50, 58], two studies in Belgium [44, 71], one study in Cyprus [73], one study in Serbia [69], one study in Thailand [22], one study in Romania [47], one study in Lithuania [46], one study in South Africa [48], one study in Iraq [40], one study in Chile [39], one study in Syria [38], and one study in Poland [59] (Table 1).

Table 1.

Key characteristics of studies

Author (Year)	Country	Sample Size	Mean Age (years)	Gender (Male/Female)	Health Status	Outcome(s)	Assessment Tool(s)	Follow – Up	Attrition Rate	PRF Preparation
Akpınar (2024)	Turkey	35	19.97 ± 0.59	9/26	Healthy	- Pain - Edema - Maximum Mouth Opening - Changes in quality of life	- VAS - Ruler - Measured in millimeter - Postoperative Symptom Severity scale	0, 6, 24 (hours) 2, 3, 4,5,6,7 (days)	None	700 rpm 3 min
Al Kassar (2023)	Syria	20	NR	NR	Healthy	- Dimensional changes of the socket	- CBCT	4 (months)	None	2700 rpm, 12 min
Aravena (2021)	Chile	19 (16)	24.75 ± 3.53	6/10	Healthy	- Alveolar wound closing - Radiographic depth measurements	- Clinical Evaluation of Alveolar Socket - Standard Periapical Radiography	7 (days) 3 (months)	16%	2700 rpm, 12 min
Asmael (2018)	Iraq	20	44.2	20/0	Smoker	- Soft tissue healing - Degree of epithelization of Socket - Pain - Socket healing (AO, Acutely infected alveolus)	- Healing index of Landry - Measurement with dental caliper - VAS - Clinical signs of AO	24, 48, 72 (hours) 7,14,21 (days)	None	3000 rpm, 10 min
Asoka (2022)	India	104(100)	NR	54/46	Diabetic	- Hard‑tissue healing - Soft‑tissue healing - Pain	- X‑ray - Modified Landry’s index - VAS	1,3,6 (weeks)	3.84%	3000 rpm, 12 min
Asutay (2017)	Turkey	30	20.32	6/24	Healthy	- Pain - Trismus - Swelling - Dry Socket	- VAS - Manual calipers - Three-dimensional facial imaging system - Clinical signs of dry socket	6,12(hours) 1,2,3,4,5,6,7(days)	None	2700 rpm 12 min
Baslarli (2015)	Turkey	20	23.9	7/13	Healthy	- osteoblastic activity - Periodontal pocket.	- panoramic image/bone scintigrams - Michigan periodontal sond	30,90 (days)	None	3000 rpm 10 min
Castro (2021)	Belgium	21(20)	NR	6/15	NR	- Bone structural healing	- CBCT / bone volume fraction	0 (days) 3 (months)	4.76%	L-PRF: 2700 rpm, 12 min / A-PRF: 1300 rpm, 8 min
da Silva (2021)	Brazil	20	23	6/14	Healthy	- Pain - Soft Tissue Healing - Color of tissues - Growth Factors - Cytokines	- VAS - Landry - Clinical evaluation - Biochemistry test - Biochemistry test	1,2,3,7,14 (days)	None	2700 rpm, 12 min
Dar (2018)	India	30	23.6 ± 4.38	13/17	Healthy	- Soft tissue healing - Swelling - Periodontal health - Bone healing - Bone density - Lamina dura formation	- Landry - Modified method of Schultze - Clinical evaluation - IOPAR/ Orthopantomographie - Bone density score - Lamina dura score	1, 3, 7, 14 (days) 1, 3 (months)	None	1300 rpm, 13 min
Daugela (2018)	Lithuania	34	22.76 ± 2.02	14/20	Healthy	- Alveolus healing - soft tissue healing - Swelling	- Clinical signs of AO - Modified healing index Landry - Clinical evaluation (non-expandable tape)	1, 3, 7, 14 (days)	None	2800 rpm, 12 min
Dimofte (2017)	Romania	63	NR	21/42	NR	- Bone density measurement	- CBCT	7,90 (days)	None	1500 rpm 14 min
Du Toit (2016)	South Africa	4	39.5	3/1	Healthy	- Bone healing	- Histomorphometric evaluation by staining shades and colors	90 (days)	None	3000 rpm, 10 min
Eshghpour (2014)	Iran	85participants (2*78 teeth)	25.09 ± 4.25	33/45	Healthy	- Frequency of AO	- Clinical signs of AO	2,7 (days)	None	3000 rpm 10 min
Gasparro (2020)	Italy	18	23.3 ± 4.73	10/8	Healthy	- Plaque index - Bleeding index - Clinical attachment level - Probing depth - Gingival recession	- Full-mouth plaque score - Full-mouth bleeding score - Clinical evaluation - Clinical evaluation - Clinical evaluation	2–4 (weeks) 3–6 (months)	None	408 g, 12 min
Gülşen (2017)	Turkey	30	20.03	21/9	Healthy	- Postoperative edema - Postoperative pain	- flexible tape - VAS and VRS	6,12(hours) 1,2,3,4,5,6,7(days)	None	3000 rpm 10 min
Gürbüzer (2010)	Turkey	20(14)	Control:23 ± 2 Prf:24 ± 2.1	7/7	Healthy	- Osteoblastic Activity	- Scintigraphic Evaluation	28 (days)	30%	2030 rpm 10 min
Kalyani (2023)	India	25	NR	NR	Healthy	- Pain - Swelling	- VAS - Dimensional measurement for swelling	1,3,7 (days)	None	2800 rpm, 12 min
Kapse (2019)	India	30	25.47 ± 0.90	13/17	Healthy	- Swelling - Bone healing	- Method of Schultze modified by Ogundipe - Digital intraoral periapical radiograph	1, 3, 7, 14 (days) 8, 16 (weeks)	None	2700 rpm, 12 min
Konuk (2022)	turkey	30	20.36 ± 2.85	8/22	13smoker, 17 Nonsmoker	- Size of Edema	- Measurement defined by Gabka and Matsumura	1,2,7 (days)	None	3000 rpm, 10 min
Kumar)2016)	India	42	NR	NR	Healthy	- Bone healing - Pain - Soft tissue healing	- Periapical Radiography - pain score - fractal analysis	1,3,7,28 (days) 2, 4, 6 (months)	None	NR
Malhotra (2020)	India	50	Male: 22.9 Female: 24.6	NR	Healthy	- Alveolar bone height - Alveolar bone density - Swelling - Periodontal pocket depth - Wound margins	- Radiological evaluation - Radiological evaluation - Clinical observation - Williams’ periodontal probe - Clinical evaluation	0, 1, 2, 4 (months)	None	3000 rpm, 10 min
Marenzi (2015)	Italy	26 participants (54 pairs of sockets)	53 ± 4	9/17	Healthy	- Pain - Soft tissue healing	- VAS - Modified Healing Index	24-48-72-96)hours( 3,7,14,21 (days)	None	2700 rpm 12 min
Niedzielska (2022)	Poland	50	35.68	28/22	Healthy	- Soft tissue healing - Grayscale value - Alveolar height measurement	- ARSTH scoring scale - CBCT radiography - Koehler 6401 caliper	10 (days) 6 (months)	None	2700 rpm NR
Öncü (2019)	Turkey	26	40.2 ± 11.5	16/10	Healthy	- Gingival margins - Bone loss - Implant stability	- Williams probe (Hu-Friedy) - Periapical radiographs - Implant stability quotient	7 (days) 1, 3, 12 (months)	None	2700 rpm, 12 min
Ozgul (2015(	Turkey	56	NR	23/33	Healthy	- Postoperative Swelling - Postoperative pain	- Vertical and horizontal measurements - Assessment of pain with VAS	1,3,7 (days)	None	3000 rpm 10 min
Ritto (2019)	Brazil	20	21.8	12/8	NR	- Soft tissue healing - Bone density	- Modified healing index of Landry - CBCT	0, 1, 3, 7 (days) 3 (month)	None	2700 rpm, 12 min
Rodrigues (2023)	Brazil	24(22)	2.41 ± 2.74	7/15	Healthy	- Bone volume - Pain - Swelling - Trabecular thickness - Trabecular distance - Wound healing - Probing pocket depth	- CBCT - VAS - Comparing the values of three facial lines - CBCT - CBCT - Landry index - Pocket probing	30(min) 2,4,6,8,12,16,24, 48, 72 (hours) 7, 90 (days)	8.33%	2700 rpm, 12 min
Sharma (2020)	India	30	23.90	14/16	Healthy	- Soft tissue healing - Bone regeneration	- Index of Landry - Digital panoramic radiograph	3, 7, 14 (days) 16 (weeks)	None	3000 rpm, 10 min
Sharma (2021)	India	60	NR	24/36	Healthy	- Hard‑tissue healing - Soft‑tissue healing - Pain	- IOPA radiography/ trephine biopsy - Pain score - Pain score	1 day 1 (week) 1,2,4,5,6 (month)	None	NR
Singh (2012)	India	20	32	10/10	Healthy	- Post-operative pain - Soft tissue healing - The trabecular pattern in healing bone.	- VAS - Landry - Radiographic IOPA	1,3,7,30,60,90 (days)	None	3000 rpm 10 min
Srinivas (2018)	India	30	NR	NR	With chronic periodontal conditions	- Wound healing - Bone density	- Landry, Turnbull, and Howley index - CBCT	24 (hours) 7 (days) 3 (months)	None	3000 rpm, 10 min
Suttapreyasri (2013)	Thailand	8 participant (10 pairs of sockets)	22.62 ± 2.44	3/5	Healthy	- Soft-Tissue Healing - Alveolar ridge contour changes - Crestal bone resorption - Dimensions of the socket orifice	- Clinical measurement - Evaluation of Alveolar Ridge by study cast - PA radiographs for bone resorption - UNC-15 periodontal probe	0, 1, 2, 4, 6, 8 (weeks)	None	3000 rpm 10 min
Sybil (2020)	India	25	32.3	14/11	Healthy	- Tenderness - Sensitivity - Edema - Plaque Index - SBI (Sulcus Bleeding Index) - Clinical Attachment Level - Probing depth - Bone height	- VAS - Sensitivity scale - Edema scale - Clinical evaluation - Clinical evaluation - Clinical evaluation - Clinical evaluation - IOPARs	0, 1, 3 (days) 1 (weeks) 1, 3, 6 (months)	None	3000 rpm, 10 min
Tabrizi (2018)	Iran	20	39.60 ± 6.74	9/11	Healthy	- Implant stability	- Resonance Frequency Analysis by Osstell device	2, 4, 6 (weeks)	None	28,000 rpm, 12 min
Tadic (2023)	Serbia	32(30)	23.33	15/15	Healthy	- Bone repair.	- CBCT /reconstruction of 3D model	7 (days)	6.25%	2700 rpm, 12 min
Tanan Karaca (2023)	Turkey	58(48)	24.5 ± 4.5	17/31	Healthy	- Pain - Swelling - Trismus	- VAS - Swelling was measured in millimeters - Trismus was measured in millimeters	0,2,7(days)	17.2%	2700 rpm, 12 min
Temmerman (2016)	Belgium	22	54 ± 11	15/7	Healthy	- Changes in horizontal width at crest-1,3,5 mm levels - Pain	- Radiography CBCT - VAS	7,90 (days)	None	2700 rpm 12 min
Unsal (2018)	Turkey	50	23.96	17/33	Smoker	- Alveolus healing - Periodontal Probing Depth	- Clinical signs of AO - Clinical evaluation	7 (days) 3 (months)	None	NR
Uyanık (2015)	Cyprus	10	22.65	4/6	Healthy	- Pain - Trismus - Swelling	- VAS - Measuring the distances - Tape measure method	1,2,3,7 (days)	None	3000 rpm 10 min
Varghese (2017)	India	30	NR	NR	Healthy	- Osseous regeneration - Soft tissue healing	- Periapical radiograph - Healing index of Landry	7,28,112 (days)	None	3000 rpm 10 min

Abbreviations: IOPAR: Intraoral periapical Radiography, VAS: Visual Analogue Scale, NR: Not Reported, CBCT: Cone Beam Computed Tomography, AO: alveolar osteitis, VRS: Verbal Rating Scale, ARSTH: Assessment Rate of Soft Tissues Healing

Soft tissue healing

In the review of 15 articles reporting on soft tissue status, four studies [19, 45, 63, 64] that employed the Landry index and reported recovery outcomes on the seventh day were included in the analysis. The remaining studies were excluded to minimize heterogeneity due to alternative indices such as the Modified Landry index, Assessment Rate of Soft Tissues Healing (ARSTH), or pain scores. These four studies were analyzed with a total of 102 subjects in the PRF cohort and 102 subjects in the Control cohort. Based on the analysis performed using the random effects model with the Inverse variance method to compare the standardized mean difference (SMD), there was a statistical difference between the cohorts, the summarized standardized mean difference (SMD) was 0.84 with a 95% confidence interval of 0.13–1.56. The test for overall effect showed a significance at p < 0.05. A significant heterogeneity was detected (p < 0.01), suggesting inconsistent effects in magnitude and/or direction. The I² value indicated that 79% of the variability among studies arises from heterogeneity rather than random chance. The findings of this review indicate a statistically significant enhancement in healing associated with the application of PRF (Fig. 4).

Fig. 4.

Forest plot and funnel plot for soft tissue healing assessment

Hard tissue healing

Among the 18 included articles, three studies [21, 39, 67] were selected to minimize heterogeneity in statistical metrics and time intervals. These studies evaluated changes in vertical ridge bone height three months after surgery. All together three studies were analyzed with a total of 71 subjects in the Experimental cohort and 71 subjects in the Control cohort. Based on the analysis performed using the random-effects model with the Inverse variance method to compare the summarized SMD there was no statistically significant difference, the summarized SMD was 0.15 with a 95% confidence interval of -0.18–0.48. The test for the overall effect did not show a significant effect. Notable variability was not detected, indicating that the effect sizes across cohorts were uniform in both size and direction. The results indicated no statistically significant association between Platelet-Rich Fibrin use and the observed changes in bone height (Fig. 5).

Fig. 5.

Forest plot and funnel plot for hard tissue healing assessment

Pain assessment

Out of the 15 articles focusing on pain, five were excluded due to insufficient statistical reporting. Analysis of the remaining 10 studies [19, 42, 51, 53, 54, 61, 63, 65, 73, 75], which assessed pain by VAS score at 6-hours, 12-hours 1-day, 2-day, 3-day, and 1-week follow-ups, revealed statistically significant reductions in postoperative pain (p < 0.01). However, subgroup analyses did not confirm this effect, despite significant intergroup differences (Figs. 6 and 7). A total of 33 time points of measurement were analyzed with a total of 1089 subjects in the PRF cohort and 1089 subjects in the Control cohort. Based on the analysis performed using a random-effects model with an Inverse variance method to compare the SMD, revealed a statistically significant difference between the two cohorts, the summarized SMD is -1.38 with a 95% confidence interval of -2.31 - -0.45. The test for overall effect showed a significance at p < 0.05. A significant heterogeneity was detected (p < 0.01), suggesting inconsistent effects in magnitude and/or direction. The I² value indicated that 94% of the variability among studies arises from heterogeneity rather than random chance.

Fig. 6.

Forest plot for pain assessment at different time points

Fig. 7.

Funnel plot for pain assessment at different time points

Alveolar osteitis incidence

Four studies [42, 46, 49, 72] investigated the incidence of alveolar osteitis(dry socket) during the first week, based on symptoms such as severe pain that is not relieved by analgesics and worsens during the first three days. All together 4 studies were analyzed with a total of 188 subjects in the PRF cohort and 188 subjects in the Control cohort. Based on the analysis performed using the random-effects model with the Mantel-Haenszel method to compare the risk ratio, a statistically significant difference was observed between two cohorts, the overall risk ratio was 0.4 with a 95% confidence interval of 0.22–0.75. The test for overall effect showed a significance at p < 0.05. Significant heterogeneity was not observed, indicating that the effect sizes across cohorts were uniform in both magnitude and direction. Results indicated that PRF placement led to a lower incidence of AO compared to the control group (Fig. 8).

Fig. 8.

Forest plot and funnel plot for alveolar osteitis incidence assessment

Discussion

Wound healing begins with hemostasis, which includes the fibrin matrix formation and the activation of an inflammatory response. Neutrophils and monocytes are recruited to the site as part of the defense. Platelets contribute to both innate and specific immune responses [76]. After platelet activation, they change morphology and release granules containing bioactive molecules, such as clotting factors, platelet-derived growth factor (PDGF), vascular endothelial growth factor (VEGF), cytokines, and chemokines. Injectable platelet-rich fibrin (I-PRF) demonstrated the potential to release higher concentrations of various growth factors. It also promotes fibroblast migration and increases the expression of PDGF, transforming growth factor-beta (TGF-β), and type I collagen. These molecules enhance cell proliferation, angiogenesis, and tissue regeneration during wound healing [77–80].

This meta-analysis demonstrated that PRF significantly enhances soft tissue healing following tooth extraction, particularly as measured by the Landry wound healing index. Although the analysis focused solely on studies utilizing the Landry index due to its frequency in the included literature, it is important to acknowledge that other indices—such as the Modified Healing Index and Healing Assessment Score—were also used in primary studies. These alternative indices similarly reported improvements in soft tissue healing with PRF application, suggesting consistency across evaluation methods. Nevertheless, the Landry index is a subjective, clinician-based assessment, which may lack sensitivity in detecting subtle histologic or biochemical changes. Therefore, future studies should consider incorporating more objective tools, such as imaging techniques or cytokine profiling, to provide a more comprehensive evaluation of healing.

While our findings are in line with individual studies such as Sharma et al. [64], who observed similar improvements in a sample of 30 participants, the comparison with existing literature remains limited. There is a lack of discussion of other systematic reviews or meta-analyses on this topic, which could provide broader context and help readers understand the relative magnitude of PRF’s effects. For instance, a review by Miron et al. (2020) reported that PRF significantly improved soft tissue outcomes in various oral surgical procedures, echoing our conclusions. Including such comparisons can strengthen the interpretation of the clinical relevance of PRF [78].

The variability in reported pain outcomes across studies suggests that multiple biological and clinical factors influence postoperative pain. Although our analysis showed a statistically significant reduction in pain on postoperative days 1, 3, and 7, no significant difference was observed at 6 and 12 h or on day 2. This inconsistency may be due to the slow release of growth factors and cytokines—such as IL-10, PDGF, VEGF, and TGF-β—resulting from gradual fibrin polymerization in PRF. These molecules modulate inflammation and promote healing but may require time to exert clinically detectable effects. Furthermore, differences in pain perception, surgical technique, and PRF preparation protocols (e.g., centrifugation speed, type of collection tubes) likely contribute to these discrepancies. For example, the study by Kapse et al. [54] supports our findings of pain reduction, whereas Gulsen et al. [51] reported no significant difference during the same time points. Such variations underscore the need for standardized PRF protocols and larger, well-powered studies to clarify PRF’s analgesic effects.

Regarding hard tissue healing, our findings suggest that PRF does not significantly reduce vertical bone resorption within the first three months post-extraction. These results are consistent with studies by Srinivas et al. [67] and Aravena et al. [39], which also found no significant differences between PRF and control groups in terms of bone preservation. This may be attributed to the limited follow-up periods in most studies, which are insufficient to observe long-term bone remodeling. Moreover, bone healing is a complex process influenced by systemic factors such as age, diabetes, and bone metabolic disorders. The absence of bone grafting materials in most included studies may also limit PRF’s efficacy in maintaining bone height. Further studies with longer follow-up durations and stratified analyses based on systemic health or adjunctive therapies are warranted to determine PRF’s potential in bone preservation.

In contrast, PRF showed a clear benefit in reducing the incidence of dry sockets. This aligns with the findings of Eshghpour et al. [49], who demonstrated a significant reduction in alveolar osteitis across 156 extraction sockets treated with PRF. The biological basis for this effect lies in PRF’s hemostatic capacity and its role as a biological membrane that stabilizes the clot and prevents mechanical dislodgement. Additionally, the anti-inflammatory and regenerative properties of PRF help protect the surgical site, promoting an optimal healing environment.

The cellular mechanisms underlying these clinical outcomes begin with the hemostasis phase, during which PRF contributes to fibrin matrix formation and recruits immune cells like neutrophils and monocytes. Upon activation, platelets release growth factors and cytokines that mediate inflammation and initiate tissue regeneration. Injectable PRF (I-PRF), in particular, has been shown to release higher concentrations of growth factors, promote fibroblast migration, and enhance the expression of PDGF, TGF-β, and type I collagen—molecules that drive angiogenesis and cellular proliferation [76–79, 81].

Although PRF’s clinical application has been extensively studied in extraction sockets, its therapeutic potential extends beyond this indication. For instance, in medication-related osteonecrosis of the jaw (MRONJ), L-PRF has demonstrated the ability to accelerate tissue regeneration and wound closure [82]. Furthermore, PRF’s immunomodulatory effects suggest broader applicability in regenerative medicine and immune-related tissue repair [83].

It is noteworthy that in terms of scales used in this study, 10 different scales were applied. Asmael et al. [40], Da Silva et al. [19], Dar et al. [45], Rodrigues et al. [63], Sharma et al. [64], Singh et al. [66], Srinivas et al. [67], and Varghese et al. [74] used the Landry Scale [84] whereas Asoka et al. [41], Daugela et al. [46], Ritto et al. [62] used the modified Landry Index [85]. In addition, Niedzielska et al. [59] employed ARSTH based on the authors’ ideas. VAS [86] was used by Akpınar et al. [37], Asmael et al. [40], Asoka et al. [41], Asutay et al. [42], Da Silva et al. [19], Gülşen et al. [51], Kalyani et al. [53], Marenzi et al. [58], Ozgul et al. [61], Rodrigues et al. [63], Singh et al. [66], Sybil et al. [21], Tanan Karaca et al. [70], Temmerman et al. [71], and Uyanık et al. [73]. The Verbal Rating Scale [87] was used by Gülşen et al. [51]. Additionally, Akpınar et al. [37] employed the Postoperative Symptom Severity (PoSSe) Scale [88] and Sybil et al. [21] used the Sulcus Bleeding Scale [89]. To evaluate the swelling and edema Dar et al. [45] and Kapse et al. [54] used the modified Schultze’s method [90]. Furthermore, Sybil et al. [21] designated a 0–3 scoring range for sensitivity and edema.

Limitations

In the discussion section, it is important to address several limitations that might influence the interpretation of the results. One limitation is the small sample sizes and the heterogeneity in the preparation methods of PRF. While we specifically chose split-mouth studies to mitigate the influence of systemic healing factors on socket healing after tooth extraction, this decision potentially led to the exclusion of some relevant articles. Additionally, the split-mouth design limited our ability to assess the effect of PRF on trismus, as this outcome could not be evaluated directly due to the design’s constraints. Another limitation involves the use of various evaluation scales for hard tissue healing, along with assessments conducted on different follow-up days. To minimize heterogeneity, we excluded studies that did not meet consistent evaluation parameters, which may have impacted the generalizability of our findings.

Clinical implications

For patients with low pain tolerance, the application of PRF in the post-extraction socket may be a rational choice due to its significant reduction in the incidence of dry socket and its ability to decrease pain during the first week postoperatively.

Conclusion

The analysis of the included studies demonstrated that PRF significantly enhanced soft tissue healing in extracted tooth sockets and reduced the incidence of AO. While PRF was effective in pain reduction, the results were inconsistent and influenced by the timing of post-operative pain assessment. Regarding the impact on hard tissue healing, the findings were controversial, and we could not conclusively confirm its efficacy for this purpose. To reduce heterogeneity, the present study focused only on studies that reported changes in vertical ridge height at a single time point within the tooth socket. Future research should consider using alternative methods to assess PRF’s effects on hard tissue regeneration to provide more comprehensive data for evaluation.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

PRF

Platelet-rich fibrin

AO

Alveolar osteitis

IOPAR

Intraoral periapical Radiography

VAS

Visual Analogue Scale

NR

Not Reported

CBCT

Cone Beam Computed Tomography

AO

Alveolar osteitis

VRS

Verbal Rating Scale

ARSTH

Assessment Rate of Soft Tissues Healing

PoSSe

Postoperative Symptom Severity

Author contributions

Pedram hajibagheri wrote the main text and performed the statistical analyses.pedram hajibagheri and Zahra Tabari and Arash Asadi-Arya reviewed the data for Figs. 1 and 2 and also Table 1.Zahra Tabari-Khomeiran conducted the data base search.Pedram Hajibagheri and Maryam Basirat contrubiuted to the drafting and revision of the manuscript and reviewed the scientific content of the article. All authors reviewed the manuscript.

Funding

This study did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The data that support the findings of this study are availableon request from the corresponding author.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.