Laser-Assisted aPDT Protocols in Randomized Controlled Clinical Trials in Dentistry: A Systematic Review

Background: Antimicrobial photodynamic therapy (aPDT) has been proposed as an effective alternative method for the adjunctive treatment of all classes of oral infections. The multifactorial nature of its mechanism of action correlates with various influencing factors, involving parameters concerning both the photosensitizer and the light delivery system. This study aims to critically evaluate the recorded parameters of aPDT applications that use lasers as the light source in randomized clinical trials in dentistry. Methods: PubMed and Cochrane search engines were used to identify human clinical trials of aPDT therapy in dentistry. After applying specific keywords, additional filters, inclusion and exclusion criteria, the initial number of 7744 articles was reduced to 38. Results: Almost one-half of the articles presented incomplete parameters, whilst the others had different protocols, even with the same photosensitizer and for the same field of application. Conclusions: No safe recommendation for aPDT protocols can be extrapolated for clinical use. Further research investigations should be performed with clear protocols, so that standardization for their potential dental applications can be achieved.


Introduction
The discovery of penicillin by Alexander Fleming in 1928 was one of the scientific highlights of the last century. In the 1940s, antibiotics had been introduced to the market and in the 1980s, pharmaceutical companies were declaring the "end" of infectious diseases. Unfortunately, microorganisms remained, and the extensive and inappropriate use of antibiotics gradually led to the development of pervasive antimicrobial resistance. Since the efficacies of antibiotics decreases and the end of the "antibiotic era" gets closer, efforts to discover new ways to eradicate microorganisms and eliminate multidrug resistance phenomena are evolving. Photodynamic therapy (PDT) therefore serves as a promising approach [1].
Photodynamic therapy is a non-thermal photochemical reaction that involves the excitation of a non-toxic dye (photosensitizer-PS) by light at an appropriate wavelength, to produce a long-lived triplet state that can interact with molecular oxygen to produce reactive oxygen species (ROS), including singlet oxygen ( 1 O 2 ), which can damage biomolecules, such as polyunsaturated fatty acids [2]. Each of the above-mentioned components (photosensitizer, light and oxygen) are harmless by themselves, but in and provide photodynamic therapy. Indeed, there are two types of these reactions-Type I and Type II [2]. In Type I, hydrogen and electron transfers take place between the triplet excited state of the photosensitizer and other molecules, predominantly O 2 . With these chemical reactions, reactive oxygen species (ROS) are produced, that are very active and harmful towards many target cells [13]. These ROS predominantly consist of superoxide anion (O 2 •− ), hydrogen peroxide (H 2 O 2 ), hydroxyl radical ( • OH), and singlet oxygen ( 1 O 2 ) [2]. However, the converse, Type II reaction is much simpler, and involves energy transfer between the triplet state photosensitizer and O 2 . This results in the formation of ground state photosensitizer and 1 O 2 [2]. Singlet oxygen and • OH radical can readily pass through cell membranes and are the most highly reactive ROS species. In view of this, only molecules that are closely located to their site of generation can be affected by photodynamic therapy [6]. Additionally, the lifetime of singlet oxygen ( 1 O 2 ) is very limited, depending on the surrounding solvent present [14], thus its action radius is approximately 10-55 nm [12]. Hence, the most important factor that influences the outcome of photodynamic therapy is the subcellular localisation of the photosensitizer which drives the process.
In general, the efficiency of the treatment can be affected by the following factors [6]: • As noted above, the sub-cellular localisation of the photosensitizer. Within the target cell, the photosensitizer may affect lysosomes, mitochondria, the plasma membrane, Golgi apparatus and the endoplasmic reticulum. Most of the photosensitizers localise within mitochondria, where apoptosis is provoked via mitochondrial damage; lysosomes accumulate photosensitizers with more aggregation. The photosensitizer Foscan (a chlorin named m-tetrahydroxyphenylchlorin) may target the Golgi apparatus and the endoplasmic reticulum [6]. However, the plasma membrane is rarely noted as a site of photosensitizer accumulation [10].

•
The chemical characteristics of the photosensitizer. The different physiology of Gram-positive and Gram-negative bacteria can affect the degree of binding of different photosensitizers. Indeed, Gram-positive bacteria can efficiently bind to cationic, neutral and anionic photosensitizers, while only cationic ones can bind to Gram-negative bacteria [15].

•
The concentration of the photosensitizer applied. High concentrations of photosensitizer can be naturally cytotoxic in a non-illuminated state, and obstruct light transmission into tissue target sites [16].

•
The blood serum content. The presence of serum in the medium can decrease the effectiveness of the therapy, in view of probable chemical and physicochemical interactions between such agents and selected serum biomolecules [17].

•
The incubation time, also known as equilibration time, of the photosensitizer at target sites. This should ideally commence shortly prior to illumination (of a ca. a few minutes' duration), since this favours localisation into the microorganisms, and does not allow penetration into host cells (this process requires many hours to occur) [18].

•
The phenotype of the target cell. It is known that different tissue types have differential light optical properties of light (i.e., absorption and scattering) [6].
An understanding of the mode of action of antimicrobial photodynamic therapy, and a knowledge of the structure of the target host tissue is essential. This should facilitate determination of the correct choice of photosensitizer (type, concentration, incubation time, etc.), and the correct light source (kind, power, illumination time, energy, spot size, distance from the target, technique applied, etc.) in order to produce a standardized protocol.
In the scientific literature, a variety of reports exist regarding the use of aPDT in dentistry. This technique has been tested in the treatment of periodontitis, peri-implantitis, endodontic conditions, dental caries and candida disinfection, wound healing and oral lichen planus (OLP). For the latter, photodynamic therapy has been suggested as an alternative treatment based on the inflammatory pathogenesis of OLP and the immunomodulatory effect of aPDT [19].
However, until now there is no consensus regarding the protocol to be applied. The aim of this study is to critically evaluate, by a systematic review of randomized clinical trials, the recorded parameters of laser aPDT applications in clinical dentistry and oral health.

Search Strategy
An electronic search was conducted relating to aPDT applications in all fields of dentistry from 10 March until 20 March. Databases used were PubMed and Cochrane, with the following MeSH terms, keywords and their combinations: (1) (PDT OR aPDT OR photodynamic) AND laser; and (2) photodynamic AND (periodontitis OR peri-implantitis OR endodontic OR caries OR candida OR oral lichen OR halitosis).
After applying the additional filters (published within the last 10 years, only randomized clinical trials in humans, and only English language reports), the preliminary number of 7744 articles was reduced to 390.
Titles and abstracts of the above articles were independently screened by two reviewers via application of the following criteria. In case of any disagreements arising, these were satisfactorily resolved by discussions.
Inclusion criteria: After screening and implementation of the eligibility criteria, a total of 38 articles were retained. These concerned a range of different aspects of application fields in dentistry. Specifically, the number of articles per field was found to be: In accordance with the PRISMA statement [20], details of the selection criteria are presented in Figure 1.

187
In accordance with the PRISMA statement [20], details of the selection criteria are presented in

Data Extraction
Having reached a consensus regarding the selection of included articles, the two reviewers involved subsequently extracted data regarding: • Citation (first author and publication year); • Type of study/number of samples/pocket depth (only for periodontitis and peri-implantitis articles); • Test/control groups; • Laser and photosensitizer used (PS concentration); • aPDT protocol/number of sessions involved; • Follow-up; • Outcome.

Quality Assessment
Subsequent to data extraction, articles were further evaluated by assessing their risk of bias assessment. The Cochrane Risk of Bias tool [21] was modified according to the requirements of this systematic review.
The risk of bias was determined according to the number of "yes" or "no" responses to the parameters provided below, which were allocated to each study: The classification was performed according to the total number of "yes" answers to the above questions. For the current study, the degree of bias was computed according to the score limits provided below:

Primary Outcome
The primary goal of this systematic review was to evaluate the studies explored with sufficient and reproducible parameter descriptions, and also analyse their aPDT protocols.
The parameters missing from the studies with incomplete protocols are also briefly noted.

Data Presentation
The extrapolated data evaluated for each dental research field are presented in Tables 1-7. Key: TBO-Toluidine Blue, MB-Methylene Blue, ICG-Indocyanine Green.
Overall, the mean ± standard error (SEM) Cochrane risk of bias score parameter was 7.76 ± 0.20 out of a perfect, optimal value of 10.
Apart from the correct description of the aPDT protocol, the most common negative answers concerned (a) use of a power meter, and (b) the sample size power calculation and required sampling numbers included.

Analysis of Data
Regarding the primary outcome, 22/38 articles (57.9%) presented an appropriate and sufficient description of the aPDT protocol used.

Discussion
Data analysis of the publications reviewed revealed a considerable variety in the report of parameters concerning the use of aPDT treatments in different dental fields. This is in accordance with Parker et al. [60], and points out the necessity to adopt clear information on the materials and methods. We then considered studies with an appropriate description of aPDT protocols, specifically those which indicated, or allowed us to calculate, the following parameters: power, irradiation time, total energy delivered, tip diameter or spot size at target tissue, any movement and speed of movement, the photosensitizer used, its applied concentration, its incubation time, and finally protocols available for washing it away or not prior to illumination. The ideal reporting of an aPDT protocol is indicated in Table 13. An important aspect to be considered is the use of a power meter prior to the illumination process. Indeed, the laser should be calibrated in order for investigators to obtain precise parameters to record, so that a standardised protocol can be provided [61]. In this review, only 6/38 [29,31,36,51,52,55] articles used a power meter (Table 8).
With regard to the treatment outcomes observed in the surveyed investigations, only 2/38 studies showed negative results when expressed relative to those of their corresponding control groups. The remainder of the investigations showed either positive (22/38) or indifferent (14/38) result outcomes when compared to results acquired for their corresponding control groups. This heterogeneity can be mainly attributed to the different protocols applied (i.e., either laser or photosensitizer parameters, as described above). Moreover, other factors that should be considered are the complex pocket or root canal architecture, unknown total volume irradiation of the photosensitizer, and the variable numbers of treatment sessions employed by investigators.

aPDT Components
As noted in the introduction, antimicrobial photodynamic therapy is based on the combination of three components: the photosensitizer nature, light and O 2 [2]. Basic information available on each of these considerations is further analysed below.

Photosensitizers
The vast majority of articles used methylene blue (MB) as the photosensitizer, which has an absorption band located at 660 nm. It is a cationic and hydrophilic compound, i.e., an amphipathic molecule (one that combines both polar and non-polar moieties), which has a low molecular mass [3]. In view of its charge, it can bind to the lipopolysaccharides of the outer membrane of Gram-negative bacteria, and also to the teichuronic acid residues of the outer membrane of Gram-positive bacteria [7].
Another popular photosensitizer is toluidine blue (TBO), with an absorption band centred at 635 nm [7]. It is a blue colouring agent also with amphipathic characteristics, but with a positive charge and a hydrophilic portion [62]. In view of its charge, it can bind both to Gram-positive and Gram-negative bacteria [7], as documented above.
The other photosensitizer used in studies included in this review is indocyanine green (ICG). It is a green colouring agent, with anionic charge, and also has amphiphilic characteristics; indeed, its polycyclic components are lipophilic [9]. It has an absorption band with a maximum at 810 nm (although this precise value is critically dependent on the dissolution medium employed), its concentration and extent of binding to blood plasma proteins [7]. Notably, its mechanism of action is predominately based on photothermal (80%) rather than photochemical (20%) processes [63].
The final photosensitizer included is the chlorin(e6) conjugate of polyethyleneimine (PEI-ce6). It is a polycationic macromolecule, and its treatment efficacy is dependent on the molecular size (smaller values lead to greater diffusion into cells), and the cationic charge (the higher the charge, the more effective it is). As expected, its absorption spectrum in the visible region of the electromagnetic spectrum is the same as that of the free chlorin(e6) conjugating agent with absorption maxima located at 400 and 670 nm [64,65].
Unfortunately, studies with curcumin, 5-aminolevulinic acid, rose Bengal and erythrosine used as photosensitizers have not been included, since they failed to meet the inclusion criteria of this review. To date, there are no published human clinical trials using 5-aminolevulinic acid, rose Bengal and erythrosine as photosensitizers in the dental fields. Notwithstanding, for curcumin, there are recent human clinical trials that reported using LEDs as the light source, and with promising results obtained [66][67][68][69][70][71].

Light Diffusion
Light distribution depends on the shape of the beam [72]; thus, diffusor tips, as used in the included studies [22,33,[40][41][42]46], are preferable since they lead to a three-dimensional illumination [73]. As Garcez et al. pointed out, the use of a conventional tip inside the root canal will lead to ROS generation in the middle of the canal, and not inside the dentin walls, where most of the microorganisms are located [74].
Furthermore, the optical properties of the target tissue play a crucial role regarding the diffusion of light. As noted in [72], these can be identified as (a) different refraction and scattering indexes when light passes through differing media, as previously noted for trans-gingival use [75]; (b) competitive light absorbers; and (c) unevenly distributed absorbers, since the photosensitizer can lead to local "cold spots" as far as the applied irradiance is concerned [72].
Regarding the use of trans-gingival as an aPDT, as applied in studies [36,38] evaluated here, such a therapy may be considered a novel approach, and this approach appears to be able to bypass the limitation of light in accessing complex target areas, such as root furcations or deep periodontal pockets [76,77]. It is known that the penetration depth of the 660 nm wavelength is 3-3.5 mm, while that for the range of 800-900 nm is 6-6.5 mm [76]. However, it is essential to consider that light attenuation occurs within gingival tissue. Specifically, for red light at a depth of 3 mm inside the gingival tissue, there is a 50% loss of intensity [75].
With regard to the competitive host absorbers of light, such as haemoglobin and a wide range of other proteins, it is mandatory to consider that their presence can decrease the effectiveness of the therapy applied [17,78]. Therefore, the outcome should be carefully evaluated when the aPDT technique is applied immediately after the SRP or pocket debridement, as was indeed the case in the majority of the studies included here for periodontitis and peri-implantitis treatment (13/21). Respectively, in endodontic therapy, the root canals should be dried prior to application of the photosensitizer. The photosensitizers used within a confined space, i.e., a root canal or a periodontal pocket, are investigated at a precise, pre-calculated concentration. If, for any reason, this space is not "dry", the photosensitizer may not achieve the concentration required for its optimal activity.
Higher concentrations of photosensitizer applied can lead to limitations in its ability to absorb light, either by the "photobleaching" phenomenon [79], or alternatively the "optical shielding" effect [6]. The former occurs when ROS generated chemically react with the photosensitizer, as noted above, and hence circumvents any further photosensitization process [79]. The latter refers to the blocking of light in view of high superficial absorption, and prevention of the light from reaching deeper tissue layers [74].
The above mentioned three photosensitizers (MB, TBO and ICG) can be considered to be ROS-scavenging antioxidant molecules [79].

Oxygen
Sufficient oxygenation of the target tissue is crucial for inducing and propagating the direct oxidative damage of microorganisms [80]; in deep and less oxygenated areas, such as in root canals, there is an O 2 deficiency. To surmount this hurdle, firstly ICG, with its photothermal action, can be used to enhance the elimination of microorganisms, although thermal damage to surrounding tissues should be taken into consideration [81]. Secondly, pre-treatment of root canals with H 2 O 2 has been suggested. This will enhance O 2 availability in this environment and allow an improved penetration of the photosensitizer inside microbial biofilms, a process leading to a higher level of antimicrobial effectiveness [74].

Healing
The healing of tissues is known to be improved following photodynamic therapy, rendering this treatment regimen a valuable choice for wounds or other infections. An additional consideration is that in many local infections, the photosensitizer is topically administered to the infected area, and the delivered light diffuses and scatters well beyond the actual area of interest. This light can exert a substantial secondary therapeutic beneficial effect in stimulating healing and repair within the surrounding tissues by a process known as photobiomodulation (PBM) [18]. Even if the whole of the photosensitizer dye solution cannot be activated, the benefits offered by PBM are invaluable [76].

Clinical Aspects
The most investigated and effective photosensitizer is methylene blue; indeed, it was applied in a total of 29 out of the 38 studies included in the present review applied MB as the photosensitizer. Nevertheless, ICG is a very promising agent, since it is activated by an 810 nm laser, which can penetrate deeper into tissues, and therefore, trans-tissue illumination is possible. In addition, in view of its additional photothermal actions (80%), applications inside root canals, where oxygen is limited, are preferential.
However, to date there is no ideal PS available, and hence clinicians should bear in mind the following characteristics before making their choice [13]: • Selectivity for prokaryotic cells over eukaryotes, so that collateral damage to healthy tissue is minimised; • Short incubation time, so that binding selectivity is achieved; • High quantum yields for photochemical reactions and low quantum yields for photobleaching; • High extinction coefficient, which demonstrates the ability of a molecule to absorb light at a specific wavelength (usually at the maximum absorption band) [8]; • Possess cationic charge and therefore be effective against both Gram-positive and Gram-negative microorganisms; • Ability to kill multiple kinds of microorganisms at low concentrations and at low light fluences; • Low side effects, such as photosensitivity and pain; • Low dark toxicity without applied illumination; As far as the light dose is concerned, it should be noted that high fluence irradiation will lead to the depletion of molecular oxygen into the tissue, and this will give rise to an impairment of therapy efficacy [13].
From the included studies with appropriate and sufficient description of the aPDT protocol applied (Tables 9-12), the authors suggest that the power and incubation time of PS should not exceed 200 mW and 5 min, respectively (only two studies [56,58] used a 10 min duration with coupled TBO) and that the irradiation time should not be less than 30 s.
All the above have the prerequisite that the clinician has understood the mechanism of action of photodynamic therapy and its influencing factors outlined in the introduction section, and can therefore select the correct combinations of the photosensitizers and lasers for upcoming dental treatments.

Conclusions
Photodynamic therapy has been acknowledged to effectively eliminate microorganisms and enhance tissue healing processes. The scope of this systematic review was to critically appraise the recorded aPDT protocols in current clinical trials featuring this form of therapy. Almost half of the articles presented incomplete parameters, whilst the remainder had differential protocols, even with the same photosensitizer and for the same field of application. Consequently, no safe recommendation on aPDT protocols can be extrapolated for clinical use at this point in time.
Unfortunately, light dosimetry is still not widely embraced in clinical aPDT. The main reason for this may be that the effects and benefits of photomedicine are multifactorial, and that the high levels of mathematics, physics and optical technologies are not easily incorporated into clinical practices and their research investigations.
For future directions, more research studies should be performed with clear, validated protocols, so that standardisation in a range of dental applications may be achieved.