Abstract
Background and Purpose. Skin wound healing is a dynamic process driven by molecular events responsible for the morphofunctional repair of the injured tissue. In a systematic review, we analyzed the relevance of plant fractions and isolates on skin wound healing. By revising preclinical investigations with murine models, we investigated if the current evidence could support clinical trials. Methods. Studies were selected in the MEDLINE/PubMed and Scopus databases according to the PRISMA statement. All 32 identified studies were submitted to data extraction and the methodological bias was investigated according to ARRIVE strategy. Results. The studies demonstrated that plant fractions and isolates are able to modulate the inflammatory process during skin wound healing, being also effective in attenuating the oxidative tissue damage in the scar tissue and stimulating cell proliferation, neoangiogenesis, collagen synthesis, granulation tissue expansion, reepithelialization, and the wound closure rate. However, we identified serious methodological flaws in all studies, such as the high level of reporting bias and absence of standardized experimental designs, analytical methods, and outcome measures. Conclusion. Considering these limitations, the current evidence generated from flawed methodological animal studies makes it difficult to determine the relevance of herbal medicines to treat skin wounds and derails conducting clinical studies.
1. Introduction
The skin wound healing is a dynamic and complex process divided into three complementary stages: inflammatory, proliferative, and maturation. The inflammatory phase comprehends the intense leucocytes recruitment to the wound area, removal of cellular and extracellular matrix debris, and syntheses of regulatory molecules such as cytokines and chemokines [1, 2]. The proliferative phase progresses with an intense proliferation and migration of fibroblasts, endothelial cells, and keratinocytes as well as formation of the granulation tissue (rich in type III collagen) and progressive reepithelialization [1–3]. At the maturation phase, type III collagen is gradually replaced by type I collagen, which originates more thicker and resistant collagen fibers [2–4].
It has been demonstrated that flaws on the leukocyte recruitment and function can impair the healing process due to reductions in the synthesis of regulatory molecules that drives the extracellular matrix assembly [5–7] and neoangiogenesis [8]. In this way, the development of drugs and alternative treatments that favor the migration and cellular activity during the inflammatory and proliferation phases may enhance the skin wound repair.
Skin wounds represent a serious health problem worldwide frequently associated with high costs and inefficient treatments [9, 10]. The use of herbal drugs is opening a new perspective for the treatment of skin wounds, mainly in developing countries. Once herbal strategies represent a simple pharmacological option, 80% of the population uses herbal drugs in their health care [1]. Although several plant species are currently used in the popular medicine to treat skin wounds worldwide [11–14], the scientific evidence that supports this practice is scarce. Thus, determining the security and efficiency of herbal drugs is an urgent and challenging task, which is essential to develop new technologies and products potentially applied in wound care.
In general, the healing properties of plant products are related to specific secondary metabolites, especially tannins, saponins, flavonoids, and alkaloids [11, 47, 48]. Plant products present a broad spectrum of biological functions such as astringent, antimicrobial, antioxidant, and anti-inflammatory [49–54] functions, which has been systematically associated with the beneficial effects in stimulating the healing process [49, 52, 54]. Before extrapolation to the human condition, preclinical researches using animal models have been useful for testing the toxicological security and biological effects of plant fractions and isolated molecules with potential applicability in the treatment of skin wounds [11, 52].
Despite the increasing number of experimental trials in the last decade, few advances were observed in the treatment of skin wounds, especially in humans. Considering that studies using animal models are conceived to support clinical investigations, there is a clear limitation in translating the findings obtained from animal models of wound healing to the human context. Considering that herbal drugs are extensively used in the popular medicine, we still do not know where the gap is that hinders the implementation of experimental findings for the development of innovations and technologies potentially useful in the clinical management of skin wounds. Thus, we systematically revised preclinical studies with murine models that investigated the effects of plant fractions and isolated molecules on the treatment of skin wounds. Beyond determining the relevance of plant derivatives in the skin repair, we analyzed the methodological quality of all preclinical studies identified, especially considering that the quality of evidence generated from flawed methodological studies could compromise the generalizability of the findings and derail conducting clinical studies.
2. Materials and Methods
2.1. Search Strategy
Research papers that investigate the action of plant fractions and isolated molecules in murine models of skin wound healing, published until 09/04/2015 (15:05:23), were recovered and independently analyzed by three researchers (FBL, MMS, and RVG). The search strategy was constructed by four components: “animals (filter),” “injury (wounds),” “organ (skin),” and “plants extract (isolates and fractions).” The filters were developed from PubMed database according to the hierarchical distribution of Medical Subject Headings [MeSH Terms]. A standardized search filter for animal studies was applied in PubMed database [55]. The same search strategy was adapted and used to recover studies in the Scopus platform. The standard animal filter provided by Scopus was used. The complete search strategy is described in Table S1 in Supplementary Material available online at http://dx.doi.org/10.1155/2016/4916068. Language restrictions were applied to recover only articles in English, Spanish, and Portuguese.
2.2. Selection Strategy
An initial selection based on title and abstract [TIAB] was independently conducted by the researchers (FBL, MMS, and RDN). Duplicate studies were removed and only studies investigating the effect of fractions and isolated molecules from plant extracts in murine models of skin wound healing were considered. After the initial search, all relevant studies were recovered in full text and evaluated by eligibility criteria. Works containing unrefined extracts, commercial isolates, in vitro assays, humans, nontraumatic injuries, other animal models, first intention wounds, metabolic diseases associates, and secondary studies (i.e., letter to the editor, note, review, and editorial) were excluded (Figure 1).
2.3. Data Extraction
Data were extracted and tabulated in a descriptive way (Tables 1(a), 1(b), 2(a), and 2(b)). The characteristics investigated were publication characteristics (author, title, publication year, and country); research methods (control group, randomization, experimental procedures, and blind evaluation of the results); experimental model (animal, number of animals, sex, age, weight, species, acclimation period, animal’s housing, number of animals per cage and experimental groups, food supply, temperature, and light cycle); plants (plant’s species, isolates, fractions, dose, toxicity test, exotic/native plant, popular name, utilized part of the plant, and popular indication); wounds description (wound area, measurement interval, and treatment duration) (Tables 1(a), 1(b), 2(a), and 2(b)). In a comprehensive approach, ethnobotanical/ethnopharmacological aspects were also investigated as follows: plant’s species investigated (geographic distribution and existence or not of bioprospecting), popular indication, and reports of toxicity tests (Figure 2).
2.4. Analysis of Bias
The articles quality was analyzed by the criteria described on the ARRIVE platform (Animal Research: Reporting of In Vivo Experiments). These criteria are based on short descriptions that indicate essential characteristics of all studies with animal models, such as theoretical and methodological basis, research objective, refinement of the analytical methods, statistical design, sample calculations, and measure outcomes [15]. Recently there has been an increasing interest in the systematic reviews of research involving animals [16]. Considering the purpose of the systematic review on evaluating important aspects of the referenced publications, we built a table summarizing all the aspects investigated as well as their relevance, describing positive and negative characteristics of the recovered studies (Tables 2(a) and 2(b)).
3. Results
3.1. Included Studies
From the PubMed and Scopus database, 1008 articles were recovered. 164 duplicated studies and 489 with thematic inadequacy were excluded after reading the title and abstract. After recovery of 329 articles in full text, 303 studies were excluded for not meeting the eligibility criteria. Thus, 26 studies were included in the systematic review. The reference list of all included studies was carefully analyzed to ensure the identification of additional relevant studies. Thus, six studies were additionally identified and recovered, completing 32 works added to this review. From these studies, 19 studies utilized fractions, 12 studies utilized plant isolates, and 1 study used both fractions and isolates for the treatment of cutaneous wounds (Figure 1).
3.2. Qualitative Analysis
The analyzed studies were conducted in 13 different countries, especially India (40.62%, ), followed by Brazil and Turkey (12.5%, each). The most utilized animal models on the experiments were rats (75%, ), followed by mice (12.5%, ) and both (12.5%, ). Considering the animal strain, 65.7% were Wistar rats, 17.14% were Sprague Dawley rats, 11.42% were Swiss mice, and 5.71% were Hairless mice. Half of the experimental models used male animals (), 15.62% () used females, and 18.75% () used both sex. 15.62% () of all studies did not report this information. The animals’ age ranged from 2 to 5 months for rats and from 8 to 12 weeks for mice; however 71.8% () of the studies did not relate this information. The weight of rats ranged from 150 to 400 g and the mice weighted between 18 and 40 g, and only 2 studies (6.25%) did not report this data.
More than half of the studies did not describe the popular name of the plant species investigated (59.37%, ). The first treatments utilized on the control group were as follows: 25% () used ointment base (which did not have its formulation described), 15.6% () used saline solution, 9.4% () used nitrofurazone, and 6.2% () utilized distilled water. Only 3.1% () did not present the treatment for the control group. The other works utilized miconazole and nonionic cream; gentamicin; Matrigel solution; soft paraffin (85%), cetostearyl alcohol (5%), hard paraffin (5%), and wool fat (5%); framycetin ointment; PBS; sodium alginate; Vaseline; Tween 80; tragacanth; povidone iodine ointment; madecassol and ointment base; chlorocresol BP 0.1% mentioned only once, representing 40.6% of all included studies (). 62.5% () of the plant species were native and 12.5% () were exotic and 25% () of the studies did not describe this characteristic.
Investigated wound area presented a large variation (5 mm2 to 600 mm2), and 9.37% () of the studies did not describe this data. The calculations used to measure the wound area were described in only 59.37% () of the studies. All the works described the interval in which the wound area was measured, and the most common interval was daily, 31.25% (), followed by measurements taken each 4 days, 28.12% () (Tables 1(a) and 1(b)). From the 32 species of plants, 23 different families were reported, and the main ones are Asteraceae 18.75% (), Euphorbiaceae 9.37% (), Leguminosae 6.25% (), and Fabaceae 6.25% (), and the other families, Liliaceae, Boraginaceae, Scrophulariaceae, Ranunculaceae, Apiaceae, Myrsinaceae, Mimosae, Malpighiaceae, Tiliaceae, Crassulaceae, Martyniaceae, Rutaceas, Araliaceae, Piperaceae, Solanaceae, Caprifoliaceae, Dipterocarpaceae, Oleaceae, and Combretaceae, were mentioned once and represent 59.37% () of the included studies. The most used plant structures were the leaves representing 37.5% (), followed by the flowers 12.5% (), bole bark 12.5% (), and seeds 6.25% (). The fruit, the whole plant, and the latex were mentioned once, representing 3.12% () each. However, 21.87% () of the studies did not mention this information. Considering the popular indication, healing effects were described in 46.87% () of the studies, followed by anti-inflammatory effects 34.37% (), treatment of gastrointestinal diseases 28.12% (), burns 18.75% (), and antirheumatic 12.5% () and ophthalmological diseases 6.25% (). 18.75% () of the studies did not report the popular indication. Only 33.4% () of the studies report toxicity tests (Figure 2).
3.3. Bias Analysis
Among the analyzed works, 78.12% presented a title coherent to the text, 90.6% presented abstracts containing the objectives, methods, main results, and conclusions, and 75% presented an introduction with sufficient scientific base. All studies described ethical approval and no work reported a blind controlled study. Most studies (87.5%) related to the therapeutic dose administered (90.62%) reported the route of administration and (96.87%) the treatment duration. The choice of administration route was not justified in any study. Most studies (96.87%) reported the investigated animal strain. The sex and weight were reported in 84.37% and 93.75% of the works, respectively, but only 31.25% provided information about the age of the animals. 59.37% of the studies provided information about the experimental conditions (temperature, humidity, light cycles, feed, and water). A statistical analysis was conducted by all studies, but only 68.75% of the studies specified the data analyzed. 84.37% of the studies reported the number of animals in each group. No study reported mortality or modifications on the experimental protocol by adverse events. A coherent interpretation of the results and direct relationship between objectives and hypothesis were described in 75% of all included studies (Table S2).
In general, the animals treated with isolates and fractions of plants presented an elevated closure rate of the wound, representing 72.72% of the studies [17, 18, 20, 23–27, 29, 30, 33, 35–46], increase in tissue reepithelialization (30.3%) [15, 18–20, 23, 27, 35, 38, 43], increase of the traction strength on the cicatricle tissue (75.75%) [15, 17–20, 22–25, 27, 31, 33–46], greater content and organization of the extracellular matrix on fast expansion of the granulation tissue (42.42%) [17, 18, 23, 30, 32, 33, 36, 38–40, 42, 44–46], and stimulation of the activity of endogenous antioxidant enzymes (9.09%) [16, 21, 22] (Tables 2(a) and 2(b)).
4. Discussion
The use of plant based strategies is opening a new perspective for the treatment of skin wounds, mainly in developing countries, once it represents a simple, low cost, and affordable therapy [1, 7, 56–58]. There are several studies indicating beneficial effects of herbal medicines in all phases of the healing process. In fact, most of the studies included in this systematic review reported that plant fractions and isolates were able to improve the skin wound healing. Apparently, these medicines were especially favorable in controlling the cutaneous inflammatory and oxidative response and in stimulating the granulation tissue formation, collagen maturation, and reepithelialization.
In this review, we did not include studies testing crude plant extracts, since the chemical characterization of the extracts makes it difficult to determine the herbal components responsible for the effects reported. Even in case of including only studies with murine models, different animal strains were observed. This aspect makes the generalizability of the results difficult, since the biological variability directly influences the response to the treatments. In addition, among the 32 analyzed studies, there were large methodological variation and discrepancies in the measure outcomes. An evident example was the wide variation in wound area and time of wound closure. These considerations are important because they are directly associated with the tensile force experienced by the tissue, which profoundly affect the speed and quality of skin repair [59, 60]. Our findings show that 20% of the studies that utilized fractions neglected the analysis of wound closure, an essential piece of information to assess the ability of any intervention to stimulate the healing process. In addition, the interval between measurements of wound area and the used protocols for the calculations were variable, representing methodological flaws that compromise the study reproduction and generalizability of the findings [61, 62].
Considering that the reepithelization and organization of the granulation tissue are fundamental aspects to understand how chemical substances act to stimulate wound healing, only 60% of all studies analyzed the reepithelialization rate and 75% evaluated the molecular components of the extracellular matrix, especially collagen. These parameters indicate if the wound closure follows a normal process, in which the newly formed tissue gradually develops drastic structural changes to reconstitute the morphofunctional characteristics of the intact skin. Works which demonstrated the importance of these analyses assert that the type and quantity of collagen fibers deposited on the tissue can be used as a marker of tissue mechanical resistance [2, 3, 9, 58]. The connective and epithelial tissues form a support structure to promote the correct closure of the wound [57, 63], reducing the chances of opportunistic infections in the wounded area [38, 64]. During the formation of granulation tissue, there is predominance of sulfated molecules which attract water, facilitating the cellular migration, and also serve as a support structure for the first formed collagen (type III collagen) [65]. There are enough evidences that the synthesis and differentiation of cells and matrix components are crucial for a normal wound closure [60]. It is already known that the oxidative stress induces cell damage, lipid, protein, and nucleic acids oxidation [66, 67]. It is recognized that cutaneous trauma increases the tissue oxidative stress in the wounded area [66–69]. Although reactive species are able to activate cell signaling pathways and stimulate cell proliferation, differentiation, and neoangiogenesis, excessive production of these molecules inhibits the healing process, especially by inducing cell death and molecular damage in the extracellular matrix [23, 70, 71]. Thus, there is a notorious importance in analyzing the redox balance during skin repair. However, from all analyzed studies, only 15% investigated the oxidative status. This is a surprising finding, since the antioxidant effect is a pivotal mechanism indicated in several studies to support the applicability of plant extracts in the treatment of tissue damage, including skin wounds [66–69]. Another fundamental result on the cutaneous repair process is the restoration of the biomechanical properties, especially the tensile force of the newly formed tissue, which provides functional estimates on the quality of the healing process [37]. In this review, only 35% of all studies investigating plant fractions evaluated the traction resistance of the scar tissue, aspects investigated in 61.53% of the studies with plants isolates.
In our findings, we see that the majority of the studies used male animals, an aspect potentially associated with the hormonal stability, which is not observed in female animals due to the estrous cycle [72]. The use of rats as the experimental model was higher (75%), aspect potentially related to the large body area needed to perform experimental wounds (1 to 5). Thus, it is possible to construct a larger number of wounds and to use a smaller number of animals in each group. In addition, in rats it is possible to collect enough fragments in order to fully analyze the healing process. Another interesting piece of data was the age of the animals, which presented a large variation (rats, 2 to 8 weeks; mice, 5 to 12 weeks). However, 71.8% of the studies did not report this information, making it difficult to establish a temporal basis to determine the effectivity of the herbal treatments investigated. More than half of the studies (59.37%) did not describe the popular name of the plant species. The large number of works that did not describe important variables such as age of the animals and plant species represents a concerning number, once these characteristics are of great importance to ensure the study reproducibility and to allow the elaboration of broad reports with a critical review of the findings [15]. The orientation cited in the ARRIVE guideline describes the minimum information that all scientific publications using animals should include. This guide also brings items that help to understand the quality of the writing and potential methodological bias that compromise the quality of the evidence [15]. The work title should refer the readers to a brief summary of the article content, providing keywords and terms that could be researched in electronic databases [73]. Only 78.12% of the studies presented a coherent title, while 90.6% presented abstracts with clear information relative to the objectives, methods, main results, and conclusions. Furthermore, 75% presented introduction with enough scientific base, which can make it harder for the reader to understand the relevance of the study. Another observation made through ARRIVE guide refers to the health conditions of the animals during the experiment. Thus, aspects such as information about environmental conditions (temperature and humidity), mortality, feeding, randomization, and reactions indicative of systemic or local toxicity were neglected in most studies, demonstrating that the report bias is a serious limitation of these preclinical tests that compromise the reliability of the results and the quality of the evidence [74].
5. Conclusions
The current evidence indicates that fractions and isolated molecules from plant extracts stimulate the healing process in cutaneous wounds. Apparently, the main effects of these herbal medicines are associated with the stimulation of collagen synthesis, expansion of the granulation tissue, reepithelialization, modulation of the inflammatory response, and oxidative stress during tissue repair. Together, these effects promote increase of the speed of wound closure and the biomechanical resistance of newly formed tissue. However, the serious methodological flaws and report bias observed in most included studies make the current evidence fragile. Thus, the relevance of fractions and isolated molecules from plant extracts in the treatment of skin wound cannot be accurately determined. Considering these limitations, it seems impossible to use these evidences to construct a rational basis that supports clinical studies. Therefore, there is an urgent need to improve research reports in experimental studies with herbal medicines in murine models of skin wound healing. This task requires a collective effort of authors, journal editors, reviewers, and financial organisms, to ensure the reproducibility, reliability, and generalizability of the evidence, fundamental elements to determine to what extent herbal medicines are promising in the treatment of skin wounds.
Competing Interests
The authors declare that there are no competing interests regarding the publication of this paper.
Acknowledgments
This work received financial support from the Brazilian agency Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) (FAPEMIG Process: APQ 00685-14).
Supplementary Materials
The full search strategy is presented in Table S1. Keywords were obtained in the MeSH (Medical Subject Headings) database, which is the U.S. National Library of Medicine (NLM) controlled vocabulary thesaurus used for indexing manuscripts in PubMed. In the additional databases used to recover relevant articles, the keywords were adapted acording the search algoritm adopted in the search plataforms.
Table S2 shows the results of bias analysis. All criteria investigated were based on ARRIVE guideline, which states the essential elements that should be reported in In Vivo animal experiments.