Laser Treatment of Peri-Implantitis: A Systematic Review of Radiographic Outcomes

(1) Background: This systematic review aimed to evaluate the effects of laser therapy on radiographic bone level (RBL) changes in peri-implantitis defects. (2) Methods: A literature search with defined inclusion criteria was performed. PubMed, Web of Science, Cochrane Library, and Google Scholar were searched through September 2020. The evaluated primary outcomes were RBL changes. In studies that reported RBL data, corresponding secondary clinical outcomes were probing depth (PD), bleeding on probing (BOP), and clinical attachment level (CAL). (3) Results: Thirteen articles were selected for data extraction and risk of bias assessment. Eight studies showed evidence of RBL gain in the laser groups compared to baseline, but did not report the statistical significance. Eight of these 13 studies reported comparisons to control. Five of the eight studies did not show RBL gain in the laser groups compared to control. In the laser groups compared to baseline, 11 of 13 reported reduced PD, and 6 of 13 reported significantly reduced BOP. Compared to the control, eight of the eight reported reduction of PD, and three of six reported significantly reduced BOP. Statistical significance was not consistently reported. (4) Conclusions: Within the limits of this systematic review, laser treatment may promote bone gain in peri-implantitis defects, may reduce BOP and PDs, and may be comparable to mechanical therapy. However, definitive conclusions can only be made with statistically significant data, which were found lacking in the currently available studies. This systematic review was registered with the National Institute for Health Research, international prospective register of systematic reviews (PROSPERO): CRD42020207972.


Introduction
The increasing usage of implants to rehabilitate the edentulous alveolar ridge has led to the higher frequency of peri-implant diseases, classified as peri-implant mucositis or peri-implantitis [1,2]. Peri-implant mucositis is a reversible inflammatory lesion that occurs in the soft tissues surrounding the endosseous dental implants [3]. Untreated peri-implant mucositis develops a radiographic progressive bone loss around the osseointegrated implant, resulting in peri-implantitis [4,5]. The progression of peri-implantitis is non-linear and accelerating; it manifests as a circumferential pattern of bone loss apical to the implant platform [5]. The weighted mean prevalence of peri-implantitis has been estimated at 22% [6]. The primary etiology of peri-implant diseases is microbial biofilm [5]. An increased risk of peri-implantitis is reported in patients with a previous history of . 2022, 10, x FOR PEER REVIEW 3 Figure 1. Search strategy.

Inclusion Criteria
• Patients diagnosed with peri-implantitis, reported as inflamed peri-implant po 4 mm or more in depth and/or loss of supporting peri-implant bone, were inclu • Clinical studies with high-intensity laser therapy of peri-implant defects we cluded. High intensity laser usage that results in ablation and removal of gin crevicular epithelium were categorized as surgical treatments.

•
Studies with sufficient radiographic data for at least five patients were included • Clinical trials reporting radiographic effects of laser treatment on human per plant diseased periodontium were included. • Non-English articles were included, but were selected for full-text analysis only English translation were available.

Inclusion Criteria
• Patients diagnosed with peri-implantitis, reported as inflamed peri-implant pockets 4 mm or more in depth and/or loss of supporting peri-implant bone, were included.

•
Clinical studies with high-intensity laser therapy of peri-implant defects were included. High intensity laser usage that results in ablation and removal of gingival crevicular epithelium were categorized as surgical treatments. • Studies with sufficient radiographic data for at least five patients were included. • Clinical trials reporting radiographic effects of laser treatment on human peri-implant diseased periodontium were included. • Non-English articles were included, but were selected for full-text analysis only if an English translation were available.

Exclusion Criteria
• All in vitro, cadaver, and animal studies were excluded. • Photodynamic therapy studies were excluded. • Non-surgical studies with low-intensity laser therapy that do not result in ablation or removal of gingival epithelium were excluded. • Conference abstracts and posters were excluded.

Screening, Selection, and Data Extraction
Three reviewers (LSA, JGS, and MT) independently screened the "Title and Abstract". Articles that did not meet the inclusion criteria were excluded. Articles were included for full-text screening if there were any doubt. The full text was then independently analyzed by the three reviewers (LSA, JGS, and MT). Data extraction of final selected articles was also independently performed by the same three reviewers with a previous pilot-tested data extraction sheet. The independently extracted data were cross-referenced among reviewers for accuracy and completeness. All disagreements pertaining to the literature screening, selection, and data extraction were resolved by discussion with a fourth reviewer (JBS). The evaluated primary outcome was RBL changes, and only studies that reported this were included. The corresponding secondary clinical outcomes PD, BOP, and CAL were also reported for these included studies.

Risk of Bias Assessment
The risk of bias (Table 1) was assessed using the risk of bias tool by the Office of Health Assessment and Translation (OHAT) [23]. The same three reviewers (LSA, JGS, and MT) independently scored the risk of bias, and disagreements were resolved through discussion with a fourth reviewer (AP).

Quality of Evidence
The risk of bias (Table 1) of the selected six randomized trials [19,22,24,25,32,33] were mostly "definitely or probably low risk of bias", and the risk of bias for the other seven non-randomized studies [14,[26][27][28][29][30][31] scored varying degrees of bias ranging from "definitely high risk to definitely low risk of bias". In these seven studies, the increase in scoring of "probably high risk of bias" was due to failure to report details of the study protocol (Table 1). Of these seven studies, four studies were at "definitely high risk of bias" for detection bias [27,31] or selective reporting bias [29,30]. As approximately two-thirds of the included studies were "definitely low risk of bias" to "probably high risk of bias", the overall level of evidence level of this systematic review is moderate to low. All selected radiographic studies utilized baseline or control for comparison. However, there was limited quantitative data to enable a meaningful meta-analysis. The selected studies with controls were too heterogenous, and these studies utilized different lasers and had different treatment protocols and follow-up periods.
For RBL compared to baseline, the Nd:YAG laser had no significant effect in one study [24] and RBL gain in another study [30] with no statistical analysis; the diode laser had significant RBL loss in one study [25], and RBL gain in another study [26] where the significance was not analyzed; the Er:YAG laser did not significantly affect RBL in two studies [19,22] and in the other studies the RBL loss (one study) [32] or gain (three studies) [27,31,33] was not statistically analyzed; and the CO 2 laser studies reported RBL gain that was not statistically analyzed [14,28,29]. Compared to control, the Nd:YAG (one study) [24] did not have a significant effect on the RBL; the diode laser had significant RBL loss in one study [25]; the Er:YAG laser did not significantly affect RBL in three studies [19,22,33], and in another study [32] the reduced RBL loss was not statistically analyzed; and the CO 2 laser showed significant RBL gain in two studies [28,29] and no significant difference in another [29].
This systematic review, parsed by laser wavelength, revealed the following: • For the two diode laser studies, one reported RBL gain compared to baseline [26], but the statistical significance was not analyzed. The other reported significant RBL loss compared to baseline and control [25].

•
For the two Nd:YAG laser investigations, one showed RBL gain [30] compared to baseline, but without analysis of statistical significance. The other [24] indicated RBL loss compared to baseline and control that was not statistically significant.

•
For the five Er:YAG laser studies, two reported RBL gain [27,31] compared to baseline but did not analyze the statistical significance of the change. One study [32] showed RBL loss compared to baseline and less RBL loss compared to control; the statistical significance of both results was not analyzed. Another reported RBL loss compared to baseline and control that was not significant [22]. One investigation reported no significant RBL change compared to either baseline or control [19]. Another study [33] reported RBL gain compared to control that was not significant, and RBL gain compared to baseline without analyzing the significance.

•
For the three CO 2 laser studies, two [28,29] reported RBL gain compared to baseline (statistical significance not analyzed) and significant RBL gain compared to control. The other study [14] reported RBL gain compared to baseline, but did not analyze the statistical significance.
Eight of 13 studies reported comparisons to control [19,22,24,25,28,29,32,33]. Of these eight studies, three showed RBL gain compared to control [28,29,33]; in two of these three studies RBL gain was statistically significant [28,29], and one was not significant [33]. The two studies [28,29] that showed statistically significant RBL gain compared to control were CO 2 laser treatments compared to air abrasives by the same research group. As for the remaining five of these eight studies, two reported RBL loss that was not statistically significant [22,24], one reported no statistically significant RBL changes [19], one reported significant RBL loss [25], and one reported less RBL loss with no statistical analysis [32].
The clinical significance of laser therapy using different lasers is described in Table 8. Laser therapy was compared to baseline or control. Control was either mechanical debridement with curettes or air-powder abrasives.
Inflammation was evaluated via BOP, sulcus bleeding index (SBI), or suppuration. Compared to baseline, the Nd:YAG laser reduced inflammation in one study [24], although the significance was not analyzed; the diode laser had no significant effect on inflammation as reported in one study [25]; the Er:YAG laser significantly reduced inflammation in four studies [19,22,32,33] and in one study the reduction was not statistically analyzed [31]; and for the CO 2 laser, inflammation was significantly reduced in one study [14], and in two studies [28,29] the increase in inflammation was not statistically analyzed. The remaining three studies did not report inflammatory parameters [26,27,30]. Compared to control, the Nd:YAG (one study) [24] and the diode laser (one study) [25] did not have a significant effect on inflammation; the Er:YAG significantly reduced inflammation in two studies [19,32], and was not statistically significant in two studies [22,33]; and for the CO 2 lasers, the increase in inflammation in one study [28] was not statistically analyzed, and in one other study [29] the increase in the residual bone group or the decrease in the augmented bone group was not statistically analyzed.
For PD compared to baseline, the Nd:YAG laser reduced PD in one study [24] with no reported statistical analysis; the diode laser significantly reduced the PD in one study [25]; the Er:YAG laser significantly reduced PD in three studies [19,27,33] and in three studies [22,31,32] the reduction was not statistically analyzed; and for the CO 2 laser, PD was significantly reduced in one study [14], and in two studies [28,29] the increase was not statistically analyzed. Compared to control, the Nd:YAG (one study) [24] and the diode laser (one study) [25] did not have a significant effect on the PD; the Er:YAG laser significantly reduced PD in one study [33], did not significantly affect PD in two studies [19,22], and in another study [32] the reduction was not statistically analyzed; and for the CO 2 laser, the reduction in PD in two studies [28,29] and the insignificant change in PD in one study [29] were not statistically analyzed.

Discussion
Periodontal regeneration, defined by the American Academy of Periodontology (AAP) and published by several investigators [58], is the restoration of lost or diminished periodontal tissues including cementum, periodontal ligament, and alveolar bone. Human histological studies are the only way to assess periodontal regeneration. Osseointegrated dental implants lack cementum and periodontal ligament, so a direct comparison between teeth and implants is not possible. Histological evaluation of regeneration has been the most accurate way to evaluate regeneration around teeth [59,60]. To date, few clinical studies have reported histological outcomes after laser treatment of peri-implantitis, and these were conducted in dogs [61,62]; therefore, RBL changes post-laser treatment may be the next available option to infer histologic changes. Radiographic evaluation of bone fill and increase in radio-opacity post-treatment may indicate regeneration or repair and may be a possible way to infer regeneration or repair when bone grafting material is not used in conjunction with the laser treatment. The selected studies in this systematic review are focused on the radiographic methodology and post-treatment changes to evaluate whether laser treatment can provide positive outcomes. A recent systematic review and meta-analysis on laser treatment of peri-implantitis reported only three studies [22,25,29] for RBL changes using high-intensity laser therapy [63]. These three studies are included in the 13 studies analyzed in this review.
Positive radiographic interpretation can be bone fill around implants after periimplantitis treatment. Radiographic determination of bone changes around implants and teeth can be limited by non-standardized radiologic methodology with inconsistent sensor angulations, position, and sensitivity [64]. In some of the included studies, efforts to standardize radiographs were not mentioned [14,19,30] or done. In addition, methods to assess bone gain or loss were different in different studies.
Clinician interpretation of radiographs can be subjective and biased. The level of expertise of the clinician when taking or interpreting radiographs may vary from radiologist, dentist, or dental student, thus affecting the accuracy and consistency of the interpretation. Computer software-assisted radiographic assessment can be reproducible and reduce operator bias and inter-operator discrepancy [64,65]. However, not all the selected studies used software. The use of software is also not without limitations. The accuracy of software is dependent on operator calibration of the computer to a fixed structure in the mouth; thus, operator errors or calibration errors while using the software can also limit the accuracy of the results.
Radiographic evaluation can be limited by inter-patient variations. Different patients may have different rates and degrees of osseous healing and radio-opacity. In addition, different patients may have different bone and tissue density that may absorb radiation differently [66]: even within the same patient, slight changes in tissue remodeling at pre-treatment and post-treatment time points may affect the exact comparison of radiographs [67]. The time points at which the radiographs were taken may also have an impact on the radio-opacity of the bone fill. When radiographic evaluation is done too early (1 to 3 months), it may provide an erroneous impression that bone fill was not significant. Moreover, documented studies on the degree of calcification of bone before it becomes radiographically apparent have reported time intervals of at least 6 months post-therapy [68]. Most studies were not clear as to which time interval would best reflect bone fill, and in some cases, non-significant results may be the result of insufficient time allocated for the bone changes to be mineralized adequately to show radiographically. In addition, most of the selected studies have inconsistent follow-up time intervals and missing radiographic evaluation at certain follow-up intervals.
The clinical effects of laser treatment at more than 6 months also show promise for radiographic outcomes, probing pocket depth changes, and control of inflammation, as most of the selected studies reported reduction in PD [14,19,22,24,25,[27][28][29][31][32][33] and inflammation [14,19,22,24,29,31,33] compared to baseline (Table 8). When compared to control, many of the selected studies with controls reported positive radiographic outcomes [28,29,33], probing depth [28,29,32,33], and inflammation reduction [19,32,33], and that laser peri-implantitis treatment was as good as or possibly better than control. However, because significance was not analyzed in most of these studies, the results can only suggest a positive outcome but cannot definitively conclude that outcome is indeed statistically significant.
The risk of bias of the included studies was variable. A quarter of the studies showed definitely or probably low risk of bias; the rest were mixed, with approximately a third of the studies showing 1-2 points at definitely high risk of bias. However, the assessment of the risk of bias alone may not be sufficient to fully assess the body of evidence. The quality of evidence can be compromised by a number of potential biases. For example, 8 of the 13 studies either did not include or report on the statistical significance of radiographic bone level changes, thus showing a level of possible reporting bias [14,19,26,27,[30][31][32][33]. Only twothirds of the six randomized controlled trials included in this systematic review calculated the number of patients required for an adequately powered trial [22,24,25,33], thus revealing a potential imprecision bias in the other two trials [19,32]. A commercial bias may apply to 10 of the 13 studies that either reported some degree of industry sponsorship [22,27,[29][30][31]33] or provided no disclosure [14,26,28,32].
A possible limitation of the review process was that the keywords used in the search may have excluded articles published in a foreign language, hence some pertinent articles may have been missed.
The evidence presented in this systematic review was also constrained by insufficient standardization of data reported in the selected articles. This shortcoming can lead to confounding factors that may influence the results of this review. In addition, variability in the detected bias among the chosen papers further limited the strength of the data synthesis. Nevertheless, this review accurately reported the variables identified in the studies in order to establish a baseline of understanding of how adjunctive laser use during treatment of peri-implantitis may affect radiographic bone level changes.
Conventional surgical therapies are demanding, technique-sensitive, and time consuming. Laser therapies may reduce clinician fatigue and stress while resulting in positive clinical outcomes. Further research studies will provide more tangible clinical data on the specific type of lasers and their associated clinical outcomes.

Recommendations for Laser Treatment Protocols
For the treatment for peri-implantitis with dental lasers, the researcher and clinician should consider laser treatment protocols that have shown evidence of the following: (1) laser reduction of infection, peri-implant bacteria, or viruses; (2) laser reduction of inflammation or inflammatory cytokines; (3) minimal tissue necrosis; (4) biostimulatory or enhanced laser-induced healing; and (5) consideration for adjunctive non-laser (mechanical debridement, air abrasives, or topical chemical agents) and laser approaches for implant rescue. To ensure safe use of the laser for patient treatment, the clinician should be well educated in dental lasers and abide by the laser guidelines and protocols of the manufacturer.

Recommendations for Future Studies
Recommendations for future research should include careful documentation of all collected data (Table 9) to facilitate meta-analyses of systematic reviews. In the conduct of a study, every attempt should be made to evaluate for statistical significance. Table 9. Recommendations for future studies.   Adjunctive treatment prior to, during, or after laser irradiation (e.g., preprocedural rinse, flap reflection, mechanical debridement (hand and/or ultrasonic instrumentation) described, air abrasive treatment, antimicrobial rinse, implantoplasty, occlusal adjustment, photobiomodulation, photodynamic therapy, bone decortication) • Hand instrumentation described (e.g., plastic or titanium curettes) • Ultrasonic instrumentation described (device, tips, irrigant) • Air abrasive instrumentation described (device, powder, flow rate) • Granulation and granulomatous tissue removed or retained • Suprastructure or superstructure removed during treatment • Clot formation • Laser parameters varied according to specific application • Biomaterials Table 9 was specifically devised as a suggested guideline to enable future investigators to: (1) consider the range of variables applicable to laser-based peri-implantitis treatment, (2) develop more consistent study designs with greater reproducibility, (3) improve standardization in data collection, (4) increase the validity of research findings, (5) reduce occurrences of bias, and (6) assure greater relevance and translation of research findings to the clinician.

Conclusions
The statistical significance of the RBL changes was not analyzed in most of the 13 studies; therefore, definitive RBL gain remains inconclusive. However, the use of dental lasers to encourage radiographic bone fill may show some promise, as most studies reported bone gain compared to baseline or control. The following conclusions about dental lasers in the treatment of peri-implantitis are within the limits of this systematic review: (1) laser treatment may enhance bone gain in peri-implantitis defects, (2) laser treatment may reduce BOP and PDs, and (3) laser peri-implantitis treatment may be as good as if not better than mechanical debridement or air abrasives. Unfortunately, definitive conclusions can only be made with proper statistical analysis of the bone level changes, which was lacking in the currently available studies. Further studies with an emphasis on supporting statistics are needed. Table 9 outlines the research data needed to aid future systematic reviews on laser treatment of peri-implantitis.