The application of platelet‐rich plasma for skin graft enrichment: A meta‐analysis

Abstract

Existing evidence demonstrated that the role of platelet‐rich plasma (PRP) in skin graft enrichment is uncertain. The aim of this study was to evaluate the efficacy and safety of PRP for skin graft. PubMed, EMBASE, Web of Science, and Cochrane Library databases were searched for randomised controlled trials that compared outcomes of skin graft treated with PRP versus those treated with blank controls. The outcomes mainly included the rate of skin graft take, number of skin graft loss and haematoma formation, and complications. There were 11 studies involving a total of 910 cases of skin grafts. Compared with the control group, PRP group had a significantly higher rate of skin graft take (mean difference = 5.47%; 95% confidence interval [CI], 2.80%‐8.14%; P < .0001), fewer number of skin graft loss (risk ratio [RR] = 0.26; 95% CI, 0.13‐0.55; P = .0004) and fewer cases of haematoma formation (RR = 0.24; 95% CI, 0.11‐0.54; P = .0006). There was no significant difference in the incidence of complications between two groups. This meta‐analysis summarises current evidence and indicates that PRP is a safe and effective adjuvant for skin graft enrichment.

1. INTRODUCTION

Resurfacing of large skin defects remains challenging in plastic and reconstructive surgery. Now, with proper harvest techniques, skin graft, especially the split‐thickness skin graft (STSG), has transitioned into popularity as an adjuvant to promote resurfacing of refractory wounds, such as diabetic foot ulcers, ¹ venous leg ulcers, ² and burns. ³ Generally, skin graft undertakes three steps for healing: anchorage, inosculation, and maturation. ⁴ Haematoma, shearing, and infection on recipient sites of skin graft can reduce skin graft take, and even lead to skin graft loss. Some strategies can increase skin graft take, such as reducing haematoma formation, promoting instant adhesion of skin graft, increasing revascularisation, and inhibiting local infection. ⁴ , ⁵ The failure of soft tissue coverage might cause reoperation, and thus increase hospital stay, medical expenses, and socioeconomic burden. The urgent need for effective therapeutic strategies has been repeatedly emphasised.

Platelet‐rich plasma (PRP) is an autologous blood‐derived biomaterial that was first introduced in 1984 by Assoian. ⁶ PRP is enriched with a 2‐ to 6‐fold concentration of platelets in comparison to normal human platelet concentration. ⁷ Multiple growth factors and bioactive proteins are stored in α‐granules of platelets. Growth factors in PRP are mostly represented by platelet‐derived growth factor (PDGF), insulin growth factor‐1, epidermal growth factor (EGF), vascular endothelial growth factor (VEGF), and transforming growth factor. ⁸ These growth factors can promote proliferation of endothelial cells, epithelial cells as well as fibroblasts, and increase angiogenesis. ⁸ , ⁹ , ¹⁰ The potential clinical application of PRP in skin graft might rely on its haemostasis, adhesive, antibacterial and healing properties.

In recent years, some randomised controlled trials have reported the favourable effects of PRP on skin graft. However, current evidence is limited to small‐sized studies. The aim of this meta‐analysis was to search available randomised controlled trials that evaluated the efficacy and safety of PRP on skin graft enrichment, and to provide evidence for clinical practice.

2. MATERIALS AND METHODS

2.1. Search strategy

This study was conducted according to Preferred Reporting Items for Systematic Reviews and Meta‐Analyses checklist. ¹¹ PubMed, EMBASE, Web of Science, and Cochrane Library databases were searched for randomised controlled trials that evaluated the topical application of PRP in skin graft enrichment. There were no restrictions on date and language of publication. Search terms were as follows: PRP, platelet concentrates, platelet gel, platelet‐rich gel, skin graft, skin grafting, skin transplantation, and dermatoplasty. Reference lists of all eligible studies were searched for relevant studies.

2.2. Criteria for considering studies

The inclusion criteria included (a) type of studies: randomised controlled trials with more than 10 participants; (b) type of participants: patients underwent surgery of skin graft; (c) type of interventions: patients in PRP group underwent topical application of PRP on recipient sites prior to skin graft while control group did not undergo topical application of PRP; and (d) type of outcomes measures (eg, rate of skin graft take, number of complete skin graft loss, cases of haematoma formation, skin graft oedema and weeping occurrence on wound beds, histopathological outcomes, incidence of complications). The exclusion criteria included (a) studies with fewer than 10 participants; (b) retrospective studies, reviews, letters, editorials, comments, abstracts, and preclinical studies; (c) without any outcome measures of interest; and (d) no original or incomplete data.

2.3. Studies selection, data extraction, and quality assessment

Two authors (C. J. G., W. Y. Y.) scanned the titles and abstracts of all records identified through the search strategy independently. The full text of potentially relevant articles was obtained and assessed independently by two authors (C. J. G., W. Y. Y.) in accordance with inclusion and exclusion criteria. Any differences of opinion were resolved through discussion with a senior author (J. H. Y.). Data were collected using a standardised form to include first author, publication year, baseline characteristics of participants, sample sizes, follow‐up duration, type of skin graft, recipient sites and donor sites of skin graft, type of intervention, details of comparison, and preparation protocols of PRP and outcomes. Two authors (C. J. G., W. Y. Y.) independently assessed the risk of bias using the Cochrane collaboration's six‐item tool. The main criteria included the assessment of random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other bias. ¹² Any disagreement was resolved through discussion with a senior author (J. H. Y.).

2.4. Statistical analysis

Calculations were performed using the Revman 5.3 software. Risk ratios (RR) with 95% confidence interval (95% CI) were calculated for dichotomous outcomes. The mean difference (MD) with 95% CI was calculated for continuous outcomes. A value of P < .05 was considered statistically significant. We used I ² statistic to calculate heterogeneity. Substantial heterogeneity was represented by an I ² value >50%. If I ² was less than 50%, the fixed effects model was used. If I ² was greater than 50%, the random effects model was used. To explore the impact of an individual study, sensitivity analysis was implemented by deleting one study each time.

3. RESULTS

3.1. Search results

A total of 277 records were identified through database search. After duplicates were removed, 184 records remained. The titles and abstracts were screened, and 161 articles were excluded. After reviewing the full text, another 12 articles were excluded for the following reasons: without any outcome measures of interest (n = 1); lack of controlled study (n = 3); retrospective study (n = 2); PRP used in donor sites (n = 4); study with fewer than 10 participants (n = 1); and no original data (n = 1). Finally, the remaining 11 studies ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ were eligible. See Figure 1 for the flowchart of studies search and selection.

FIGURE 1.

Flow chart of study search

3.2. Major characteristics of study and quality assessment

For the characteristics of included studies, see Table 1. A total of 910 cases were included in 11 studies from 2010 to 2020, with the sample size ranging from 11 to 100 patients for per group. Of these cases, 466 cases belonged to PRP group and 444 cases belonged to control group. The type of skin graft in 10 studies was STSG ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ but one study ¹⁸ did not clarify the type of skin graft. The preparation protocols of PRP differed among the included studies. Rate of skin graft take was reported in five studies. ¹⁴ , ¹⁸ , ¹⁹ , ²⁰ , ²² Five studies ¹⁶ , ¹⁷ , ²¹ , ²² , ²³ recorded the number of skin graft loss. Three studies ¹⁶ , ¹⁷ , ²³ documented the number of haematoma formation, skin graft oedema and weeping occurrence on recipient sites. Complications were reported in four studies. ¹⁵ , ¹⁶ , ¹⁹ , ²⁰ Only one study ¹³ recorded histopathological evidence of skin graft enrichment. Risk of bias assessment is demonstrated in Figure 2.

TABLE 1.

Characteristics of included studies

References	Cases of skin graft (I/C)	Skin graft	Receipt sites	Donor sites	Intervention in wound bed topically	Follow‐up	Preparation of PRP (speed and time)
Abdelkader et al 13	30/30	STSG	Lower limb	Thigh	PRP	2 (W)	First step: 3600 rpm Second step: 2400 rpm
Adly et al 14	50/50	STSG	Upper limb, lower limb, trunk, head and neck	NR	PRP	10 (D)	NR
Dai et al 15	34/33	STSG	NR	NR	PRP	2 (W)	First: 1500 rpm for 10 min Second: 3000 rpm for 10 min
Dhua et al 16	20/20	STSG	Neck, right forearm, other sites in the lower part of neck	Thigh	PRP	3 (M)	First: 3000 rpm for 20 min Second: 1000 rpm for 10 min
Gupta et al 17	100/100	STSG	NR	NR	PRP	6 (D)	3500 rpm for 10 min
Hasibuan et al 18	15/15	Skin graft	Extremities	NR	PRP	NR	1000 rpm for 12 min
Hersant et al 19	14/13	STSG	Thigh, leg, foot, arm, forearm, ankle, perineum, scrotum, pubis, gluteal, urethra	Thigh	PRP	5 (D)	First: 1500 rpm for 5 min Second: 1500 rpm for 5 min
Marck et al 20	52/52	STSG	NR	NR	PRP	5–7 (D)	First: 3200 rpm for 15 min Second: 3200 rpm for 15 min
Sonker et al 21	20/20	STSG	Extremities, neck, chest, scalp, cheek, face, abdominal wall,	NR	PRP	6 (W)	NR
Vendramin et al 22	31/11	STSG	NR	NR	PRP	14–28 (D)	First: 400 g for 10 min Second: 800 g for 10 min
Waiker et al 23	100/100	STSG	Head, neck, chest, abdomen, upper and lower limb	NR	PRP	9 (M)	1000 rpm for 5 min

Abbreviations: C, control group; D, days; I, intervention group; M, months; NR, not reported; PRP, platelet‐rich plasma; STSG, split‐thickness skin graft; W, weeks.

FIGURE 2.

Risk of bias. Green circle = low bias risk, red circle = high bias risk, yellow circle = unclear bias risk

3.3. Meta‐analysis

3.3.1. Rate of skin graft take

A total of 294 cases were included in five studies ¹⁴ , ¹⁸ , ¹⁹ , ²⁰ , ²² in the comparison of the rate of skin graft take between two groups. Compared with the control group, the PRP group had a significantly higher rate of skin graft take (MD = 5.47%; 95% CI, 2.80‐8.14%; P < .0001), and the heterogeneity was low (I ² = 19%; P = .30). A sensitivity analysis was conducted and none of included studies could change the pooled result qualitatively (Figure 3).

FIGURE 3.

Forest plot: rate of skin graft take

3.3.2. Number of skin graft loss

Five studies ¹⁶ , ¹⁷ , ²¹ , ²² , ²³ with a total of 522 cases recorded the number of skin graft loss in two groups. Statistical analysis indicated that the number of skin graft loss was significantly lower in the PRP group than in the control group (RR = 0.26; 95% CI, 0.13‐0.55; P = .0004), and the heterogeneity was low (I ² = 0%; P = .93). We also performed sensitivity analysis by removing each included study successively, and found that none of the studies could change the results (Figure 4).

FIGURE 4.

Forest plot: number of skin graft loss

3.3.3. Cases of haematoma formation, skin graft oedema, and weeping occurrence

Three studies ¹⁶ , ¹⁷ , ²³ documented the cases of haematoma formation, skin graft oedema and weeping occurrence on wound beds. The pooled results revealed that PRP group had fewer cases of haematoma formation (RR = 0.24; 95% CI, 0.11‐0.54; P = .0006), skin graft oedema (RR = 0.22; 95% CI, 0.07‐0.66; P = .007), and weeping occurrence (RR = 0.12; 95% CI, 0.04‐0.37; P = .0003) in comparison to control group (Figure 5).

FIGURE 5.

Forest plot: cases of haematoma formation, skin graft oedema and weeping occurrence on recipient sites

3.3.4. Histopathological evidence of improved skin graft outcomes

Only one study ¹³ with a total of 60 cases of skin grafts reported the histopathological evidence of skin graft enrichment. The number of some parameters in the PRP group, such as marked keratin formation, keratinocytes differentiation, neovascularisation, collagen deposition, inflammatory cells in the dermis, and melanin deposition, was significantly higher than in the control group. Furthermore, the mean epithelialisation surface area was significantly higher in PRP group than in the control group. ¹³ However, a meta‐analysis could not be performed because of only one eligible study included.

3.3.5. Complications

A combined total of 520 cases from four studies ¹⁵ , ¹⁶ , ¹⁹ , ²⁰ documented the incidence of several complications, including skin allergic reaction, pain, inflammatory response, wound infection or sepsis, and itching. Overall, the pooled results indicated that there was no significant difference in the incidence of these complications between the PRP group and the control group (Figure 6).

FIGURE 6.

Forest plot: incidence of complications

4. DISCUSSION

To augment functional and aesthetic form for non‐healing wounds, such as diabetic foot ulcers, venous leg ulcers and burns, and skin grafting is widely utilised by plastic surgeons. ²⁴ , ²⁵ Growth factors in these wound beds may be down‐regulated because of decreasing synthesis and excessive degradation. ²⁶ Moreover, some potential risk factors such as haematoma, local infection, and shear forces will affect the survival of skin graft. ¹⁶ , ¹⁷ In recent years, PRP gels have been widely applied in regenerative medicine. In our centre, after employing PRP gels into the recipient sites of skin graft, the skin graft take was greatly augmented and patients' hospital stay was shortened.

To our knowledge, this is the first meta‐analysis of current studies regarding the adjunctive treatment of PRP for skin graft. Eleven eligible studies with a total of 910 cases of skin grafts were included. The pooled results demonstrated that the rate of skin graft take was significantly higher in PRP group than in the control group with low heterogeneity. Currently, only five studies have been included for quantitative synthesis, and some of included studies are small‐sized, which might weaken the strength of this pooled result. The number of skin graft loss was lower in PRP group than in the control group with low heterogeneity. The topical application of PRP could effectively reduce the cases of haematoma formation, skin graft oedema, and weeping occurrence on recipient sites. We also conducted sensitivity analyses and found that none of studies could reverse these pooled results. The histological evidence for skin graft enrichment indicated that PRP could effectively enhance the levels of keratin formation, keratinocytes differentiation, neovascularisation, collagen deposition, and epithelialisation. ¹³ Overall, the use of PRP may be helpful for those patients who need to receipt a large area of skin grafting and lack of autologous skin particularly. It is possible that these superior outcomes may be attributed to haemostasis, adhesive, healing, and antibacterial properties of PRP.

First, as an autologous biological blood‐derived biomaterial, PRP contains high concentration of platelets together with multiple platelet‐derived bioactive molecules and leucocytes. ²⁷ Platelets initially play a key role in haemostasis. Fibrin clots in PRP gels can promote blood clotting and wound closure. ²⁸ These procedures can reduce haematoma on recipient sites of skin graft and further promote instant stable adhesion of skin graft, which will not only increase oxygen diffusion from the recipient sites into skin graft but also prevent shearing from disrupting ingrowth of new vessels. ⁵

Second, platelets also have been introduced into regenerative medicine and have beneficial roles in promoting tissue healing. After activation, platelets release a large number of growth factors and proteins. ²⁹ Another origin of growth factors may be leucocytes trapped within PRP fibrin clots. ³⁰ Through launching signalling cascades, these growth factors will initiate multiple cellular biochemical changes. For example, VEGF in PRP mainly takes part in angiogenesis. ³¹ EGF is a mitogenic and chemotactic factor for fibroblasts and epithelial cells. ³² PDGF is also a mitogenic and chemotactic protein for fibroblasts and endothelial progenitor cells. It can induce angiogenesis and collagen synthesis. ⁸ FGF is another mitogenic and angiogenetic factor for a broad range of cells, and can promote angiogenesis and wound healing. ¹⁰ Leucocytes in PRP may interact with platelets, which will increase release of anti‐inflammatory factors and further switch inflammatory process towards a regenerative phase. ³³ , ³⁴ To sum up, it is possible that PRP improves local microenvironment of recipient sites by regulating inflammatory reactions and promoting neovascularisation.

Third, PRP may play a significant role in inhibiting infection. Platelets in PRP can interact with different bacterial strains and immune cells through their surface receptors. Platelets can form composites with neutrophils, which will cause release of reactive oxygen species or formation of neutrophil extracellular traps. ³⁵ Furthermore, some microbicidal molecules are also released from activated platelets. These bioactive molecules will provide an early protection against bacterial infection through early interaction with innate immune response. ³⁵ , ³⁶

Overall, PRP allows platelets, leucocytes, growth factors and other proteins to be delivered to recipient sites. It is, therefore, effective to reduce haematoma, promote instant stable adhesion of skin graft, regulate local inflammatory process, promote neovascularisation, and inhabit infection.

The PRP for clinical application has been approved by the Food and Drug Administration. The application of PRP has a majority of advantages such as autologous source, easiness to obtain, and convenient and non‐invasive application procedures. ³² , ³⁷ In our article, there was no significant difference in the incidence of several adverse events between two groups. However, only four studies were included for evidence synthesis, and some of these studies were small‐sized. To better evaluate the safety of PRP use, we recommend that future large and multi‐centre randomised controlled trials should thoroughly document all adverse events. Several studies reported that PRP could significantly reduce operation duration, ¹⁶ , ²³ hospital stay, ¹⁶ , ²³ post‐operative dressings, ¹⁶ , ²³ duration of wounds healing, ¹⁵ , ¹⁹ and total medical expenditure. ¹⁶ , ¹⁷ However, a meta‐analysis of these outcomes could not be performed because the mean or SD, or the original data were not provided in these eligible studies. We thus suggested that further randomised controlled trials should record complete data about these outcomes, which will help us better evaluate the efficacy and cost‐efficiency of PRP, and make clinical decisions.

The pooled results should still be interpreted cautiously given some limitations. First, the sample size in some included studies was small. Some of the outcome indicators, such as the rate of skin graft take and the complications, included too few studies to conduct group studies, which may lead to bias. Second, for the pooled result of skin graft take rate, the 95% CI lower limit omits its statistical significance and weakens the strength of this finding. Third, some included studies did not clearly indicate the allocation concealment as well as the blinding of patients and outcomes assessors, which could cause selection bias and detection bias. Fourth, there was a lack of standardised PRP product in terms of preparation protocols as well as the concentration of platelets and growth factors, which also led to bias. Fifth, there were some variations among the eligible studies, such as the types of skin graft recipient sites and donor sites, techniques for skin graft harvest, thickness of skin graft, ratio of PRP gels to skin graft and follow‐up period, which could cause some heterogeneity. Sixth, the eligible studies only evaluated the early outcomes of skin graft. The long‐term outcomes of skin graft such as the scar measurement and quality of life were little reported. Finally, the outcomes about the healing of recipient sites were not evaluated in these eligible studies. So, pooled results could not be obtained. To better evaluate the efficacy of PRP, large and multi‐centre randomised controlled trials are warranted to evaluate long‐term outcomes of skin graft and the healing of recipient sites, using a standardised preparation protocol, a homogeneous population, and an optimal ratio of PRP to skin graft.

5. CONCLUSION

This meta‐analysis provides evidence that topical application of PRP is safe and valuable in enhancing rate of skin graft take and reducing complete skin graft loss because of its haemostasis, adhesive, antibacterial, and healing properties. Further large and multi‐centre randomised controlled trials for evaluating its long‐term outcomes such as scar assessment and quality of life are warranted.

CONFLICT OF INTEREST

The authors declare no potential conflict of interest.

Chen J, Wan Y, Lin Y, Jiang H. The application of platelet‐rich plasma for skin graft enrichment: A meta‐analysis. Int Wound J. 2020;17:1650–1658. 10.1111/iwj.13445