Periodontal and Peri-Implant Diseases and Systemically Administered Statins: A Systematic Review

Hyperlipidemia is a well-recognized risk factor for cardiovascular disease, which, in turn, acts as a contributory factor in periodontitis development. Periodontitis has been associated with benign tumors and cancers and withseveral disorders, including hyperlipidemia. Correspondingly, periodontal treatment may exert a positive effect on lipid metabolism, although opposite evidence has also been reported. As a counterpart, the therapy for hyperlipidemia, conventionally based on statins, has been proposed to positively affect periodontal conditions, mainly due to statin pleiotropic effects, reducing periodontal inflammation and promoting osseointegration. Therefore, the present systematic review aimed to evaluate, in subjects with untreated periodontitis and peri-implant disease (Population), the effect of routine systemically administered statins (Intervention), compared to non-statin use (Comparison), on periodontal parameters around natural teeth and implants (Outcome). Discordant results were found in periodontal parameters, and the current lack of such data related to peri-implant tissues and to alveolar bone loss highlights the need for further studies on the topic, potentially paving the way for a more comprehensive approach to periodontitis and peri-implantitis management. Indeed, the validation of the beneficial effect provided by systemically delivered statins on periodontal and peri-implant tissues may direct recall scheduling, predict response to therapy and, therefore, guide treatment strategies of periodontal and peri-implant treatments in statin users.


Introduction
Hyperlipidemia is characterized by an increase in triglycerides (TGs), total cholesterol and low-density lipoprotein (LDL) serum levels, coupled with a decrease in blood concentrations of high-density lipoprotein cholesterol (HDL) [1]. Hyperlipidemia is a wellrecognized risk factor for cardiovascular disease [2], which, in turn, acts as a contributory factor in periodontitis development [1].
Periodontitis is a chronic inflammatory disease initiated by periodontal pathogens in the bacterial biofilm that is subsequently sustained by periodontal tissues inflammation, leading to the disruption of the anatomical structures supporting the teeth, alveolar bone loss and, eventually, tooth loss [3][4][5][6][7]. Periodontal disease has been associated with several benign tumors and colorectal, breast and prostate [8,9] cancers, as well as a multitude of systemic disorders, including diabetes [10], inflammatory bowel disease [11], atherosclerosis, rheumatoid arthritis [12,13], obesity [14] and hyperlipidemia [1].
Correspondingly, periodontal treatment in hyperlipidemic periodontal subjects has been reported to improve serum lipid concentrations, decreasing total cholesterol and LDL and increasing HDL [15,16], and to reduce proinflammatory cytokines levels, such as C Reactive Protein (CRP), Tumor Necrosis Factor-a (TNF-a) and Interleukin-1 (IL-1) [17]. However, contrasting evidence does not support such a positive effect of periodontal non-surgical therapy on lipid metabolism [18].
OR CAL (acronym of Clinical Attachment Level); 6.

Study Selection Process
Study selection was conducted by two independent reviewers (FDS, FDA), and any discrepancy was solved by discussion.
Title and abstract screening were accomplished for all the records identified through database search.
Full-text reading was performed for abstracts reporting ambiguous methods and for articles considered eligible for the present systematic review, based on the inclusion/exclusion criteria, reported in Table 1. Table 1. Study eligibility: inclusion and non-inclusion criteria related to study sources, characteristics, population, intervention, comparison and outcome(s). References reported in the eligible articles were checked, and an additional literature search was carried out. Neither manual search nor contact with the corresponding author for online unavailable full-texts were performed.

Data Collection Process and Items
Data extraction and collection were performed twice by two independent authors (FDS, FDA) in a dedicated form, developed from those proposed for intervention reviews on RCTs and non-RCTs [34], also available online: https://dplp.cochrane.org/data-extraction-forms (accessed on 20 July 2021). No additional processes were carried out to obtain nor confirm data from investigators.
The following variables were recorded for each selected study: source (first author, year and journal of publication, complete article citation, fundings); study design; participants (sample size, age and gender, smoking habit, comorbidities, treatments and procedures); intervention (statin type, duration, dosing); comparison (no statine use, placebo); periodontal outcomes (clinical, radiographic and other periodontal parameters); consideration(s) and conclusion(s).

Risk of Bias Assessment for Individual Studies
Risk of bias in individual studies was assessed at the study level by two independent reviewers (FDS, FDA), and discrepancies were solved by discussion.
The Cochrane Collaboration's tool for assessing risk of bias [34] was employed to analyze the following six domains of bias and the related items in randomized clinical trials: • Selection (random sequence generation, allocation concealment); • Performance (blinding of investigators and participants); • Detection (blinding of outcome assessment); • Attrition (incomplete outcome data); • Reporting (selective outcome reporting); • Others.
The risk of bias was previously recorded for all of the items, as extensively shown in the risk of bias table. For each item, in case of sufficient data, the risk of bias was defined as "low" (-), therefore unlikely to seriously alter the results; for missing data, the risk of bias was declared as "high" (+), consequently, capable of serious alteration of the results; for insufficient information, the risk of bias was considered as "unclear" (?), raising doubts on study results.
Subsequently, the overall risk of bias within a trial was assessed, and it was defined as "low" if all items and domains were already defined as at low risk of bias and as "unclear" or "high" with at least one item formerly judged as unclear or high, respectively.
The ROBINS-I tool was employed, instead, for assessing the following seven domains of bias of non-randomized studies of the effects of interventions [35,36]

Study Selection
The study selection flow-chart, illustrated in Figure 1, shows that, through the electronic search, a total of 277 potentially relevant titles/abstracts were initially retrieved: specifically, 205 from PubMed and 72 from Scopus databases, respectively. Subsequently to duplication elimination, 213 potentially pertinent title/abstracts were identified, comprising 202 articles concerning periodontitis and 9 about peri-implantitis. After title/abstract screening, 59 records were excluded because they were not pertinent, along with 1 book chapter and 2 abstracts only;therefore, 149 full-texts were assessed. According to the eligibility criteria for sources and study characteristics (Table 1), 1 commentary, 24 reviews, 19 animal and 18 in vivo/preclinical studies, besides 5 articles written in Chinese and Japanese languages, were not included. In addition, based on the inclusion/exclusion criteria on Population, Intervention, Comparison and Outcome(s), (Table 1), 9 studies involving participants with endodontic lesions, 9 including subjects with comorbidities and 12 comprising patients undergone periodontal treatment were also excluded; moreover, 44 studies, including interventions with local statins and 1 with high-dose systemically delivered statins, were not considered.
Finally, 7 studies were included in the present systematic review, all of them concerning periodontitis and none regarding peri-implantitis.

Study Characteristics
Each of the 7 included studies comprised a minimum of 70 periodontal subjects, at least 30 years old, without comorbidities potentially affecting their periodontal/peri-implant status, not under drugs affecting lipid nor bone metabolism, antibiotics and corticosteroids and not undergone periodontal treatment within the last 3 months.
Statin intake was compared with no statin use in all the studies. Moreover, in four studies, the intervention was also compared with non-pharmacologic therapy, such as diet, combined or not with exercise, in hyperlipidemic subjects [38,[40][41][42]. Statins were not compared with placebo in any of the selected studies.

Study Characteristics
Each of the 7 included studies comprised a minimum of 70 periodontal subjects, at least 30 years old, without comorbidities potentially affecting their periodontal/peri-implant status, not under drugs affecting lipid nor bone metabolism, antibiotics and corticosteroids and not undergone periodontal treatment within the last 3 months.
Statin intake was compared with no statin use in all the studies. Moreover, in four studies, the intervention was also compared with non-pharmacologic therapy, such as diet, combined or not with exercise, in hyperlipidemic subjects [38,[40][41][42]. Statins were not compared with placebo in any of the selected studies. Clinical periodontal parameters around natural teeth were reported in all selected studies. Plaque Index (PI), or a similar parameter recording the presence of plaque [25], was evaluated in all the selected studies. Clinical Attachment Level (CAL) [38][39][40][41][42] and Periodontal Probing Depth (PPD) [25,38,[40][41][42] were both measured in five studies, Gingival Index (GI) and Bleeding on Probing (BoP) were assessed in five [37][38][39][40] and three [25,41,42] studies, respectively, while the Community Periodontal Index was computed in one study [37]. Only two studies evaluated other periodontal parameters, such as gingival crevicular inflammatory mediators, specifically Interleukin (IL-)1B [39,42], IL-10 [42]and Myeloperoxidase (MPO) [42], and no studies considered either radiographic outcomes (bone loss), tooth loss due to periodontitis, nor the number of residual teeth. None of the included studies assessed periodontal parameters around implants.
Of the seven clinical studies included in the present systematic review, three were case-control and four were cross-sectional studies; no longitudinal nor randomized clinical trials were found to be compliant with the eligibility criteria. A complete description of the selected studies, regarding the source, participants, intervention, comparison andout-comes and consideration(s) and conclusion(s) focusing on the PICO question, is reported in Table 2. Table 2. Data extracted and collected from the studies included in the present systematic review. General information: first author, year and journal of publication; reference number; fundings. Characteristics of the study: design; participants (sample size, age and gender, smoking habit, comorbidities, treatments and procedures); intervention (statin type, dosing, duration and indication); comparison (no statine use, placebo); outcomes (clinical and radiographic periodontal parameters and Gingival crevicular Inflammatory mediators, assessment timepoints); consideration(s) and conclusion(s). Abbreviations: hyperlipidemics, "HL"; normolipidemics, "NL"; statin users, "S"; non-statin users, "NS"; diet, "D"; exercise, "E"; periodontally healthy, "Ph"; with gingivitis, "G"; with periodontitis, "P"; gingival crevicular, "GC"; relative risk, "RR"; p-value, "p". Only statistically significant data concerning the outcomes and focusing on the PICO question have been reported.  Dental Hypothesis Cross-sectional study [37] No fundings

Risk of Bias within Studies
Since the seven studies included in the present systematic review were all nonrandomized studies, only the ROBINS-1 tool was currently employed, and the risk of bias within studies has been reported in Table 3. Table 3. Risk of bias of non-randomized clinical trials, assessed through the ROBINS-I tool [36], is designated as "Low", "High" and "Unclear".

Selection of the Reported Result
Cicek

Results of individual studies.
Results concerning all outcomes considered in the current systematic review have been synthesized in a discursive way in Table 4.

Discussion
Discordant results were found in the literature on the putative positive effect exerted by systemically administered statins on clinical periodontal parameters around natural teeth and on gingival crevicular inflammatory mediators; no data, instead, were retrieved concerning alveolar bone loss nor clinical and radiographic periodontal parameters around dental implants.
The hypothesized bi-directional relationship between periodontitis and peri-implantitis, on the one side, and hyperlipidemia, on the other side, may rely on the shared etiologic factors, comprising genetic susceptibility, smoking, stress, altered immune and inflammatory response, and on the common pathogenic pathways [1]. Specifically, periodontal pathogens may gain access to the systemic circulation, colonize atheromatous plaques [43], and, along with their toxins and the proinflammatory cytokines from inflamed periodontal tissues, determine the so-called "systemic inflammation", which is considered to be the key associating link between periodontitis and several systemic inflammatory and neoplastic diseases [1,8,[10][11][12][13][14]44,45]. Consequently, the causal treatment of such disorders may positively affect both periodontal conditions around teeth and implants, as well as the lipidemic status [17,19]. In particular, systemic statins may be variously beneficial for periodontitis and peri-implantitis due to their main and pleiotropic effects. Indeed, combined with the hypolipidemic effects, statins have also shown anti-inflammatory and antioxidant activities, on the one side, and stimulating properties affecting endothelial function, angiogenesis and bone formation on the other side [46]. Noteworthy, in periodontal subjects, statins have been found able to reduce osteoclasts activity and bone resorption [39], to increase Interleukin-(IL-)10 in gingival crevicular fluid while decreasing IL-1B [42], together with IL-6 and IL-8 [31] and to inhibit the release of matrix metalloproteinase-(MMP-)1, MMP-2, MMP-3 and MMP-9 [25,47], finally lowering periodontal tissues inflammation and destruction.
The present systematic review, unlike the previous ones including studies on locally administered statins and/or on systemic statins as adjuncts to periodontal treatment [20,28,31], aimed to primarily evaluate the putative positive effect on untreated periodontitis and peri-implantitis of systemic statin therapy (≥1 month) alone. For such reasons, and due to the fact that periodontitis is linked to aging [48], studies involving participants younger than 30 years of age were excluded from the current analysis; similarly, studies including periodontally healthy subjects, and those reporting periodontal treatment, within the last 3 months or as an adjunct to statin administration, were not considered. In addition, studies comprising smokers and periodontal subjects with comorbidities were also excluded to eliminate the potential confounding due to the so-called periodontitis grade modifiers [49], specifically smoke and diabetes, as well as disorders and medications affecting lipid/bone metabolism and periodontal disease [50,51].
Periodontal parameters around natural teeth and dental implants were compared between statin users vs. non-statin users; no placebo use was reported in comparison with statin use in any of the included studies. Included studies described primarily simvastatin and secondarily atorvastatin intake, in compliance with the findings by Gu et al. showing that simvastatin represents 42% and atorvastatin 20% of all cholesterol-lowering medications, respectively [22].
Clinical periodontal parameters around natural teeth were recorded in all selected studies, although periodontal charting was variously performed, complicating results comparison. In addition, periodontal parameters were incompletely reported in many studies, with a lack of data describing and comparing periodontal status before vs. during/after statin use and in statins vs. non-statin users. Similarly, heterogeneous periodontal case definitions were applied in the included studies, different from the one introduced by the 2017 classification of periodontal and peri-implant diseases and conditions [7], and, consequently, staging and grading of periodontitis were not performed. In particular, the Community Periodontal Index, employedby Kadhim et al. [37] and by Fernández et al. [40], is considered to be able to detect the prevalence but not the severity of periodontal dis-ease [40]. However, lower CAL values were reported in statin vs. non-statin users by Sayar et al. [41], who attributed such results to the anti-inflammatory effect exerted by statins, and by Cicek Ari et al., [42], although not reaching the statistical significance; conversely, opposite results were found by Poston et al. [52]. PPD values were improved by systemically delivered statins, as reported by Lindy et al. [24], Cicek et al. [42], Sayar et al. [41] and even by Poston et al. [50] but only in diabetic subjects, and by Fentoglu et al. [53] but only after more than 3 months of continuous statin intake; in addition, Lindy et al. [24] showed that also PIBI, combining the data on PPD, was 40% smaller in statin users when compared to non users. Many Authors [41,42,47,53]) detected a significant reduction in BOP values, likely linked to the anti-inflammatory effect exerted by statins, although Fentoglu et al. [53] described higher BOP values in the statin users than in non-users. Contrasting results were also reported for GI and PI values, resulted significantly lower in statin users in theSangwan et al. study [38], but not in the Cicek one [42]; Saxlin described a negative association between statinand periodontal statusamong subjects withplaque, although opposite results were found among subjects with no plaque [25].
Gingival crevicular IL-1 wassignificantly lower in statin compared to non-statin users [42,50], similarly to MPO [42] and IL-6 [51] levels, opposite to IL-10 ones [42]. Tooth loss may benefit from systemically delivered statins, which may exert a protective effect in periodontal subjects [19,25]. Radiographic periodontal parameters around natural teeth were not recorded nor reported in any of the included studies, even if bone loss may be positively affected by statin use, which has been hypothesized to reduce the risk of fracture and increase bone density [40].
Moreover, no study evaluating statins' putative effect on peri-implant tissues was found, highlighting the need for a more comprehensive evaluation of the potential beneficial effect of systemic statins. Indeed, although peri-implantitis mainly shares ethiopathogenic pathways with periodontitis [54], a certain association between peri-implant disease and hyperlipidemia has not been established, probably due to the paucity of studies investigating such a potential relationship [54].
In conclusion, discordant results were found in periodontal parameters, and the current lack of such data related to peri-implant tissues and to alveolar bone loss highlights the need for further studies on the topic, potentially paving the way for a more comprehensive approach to periodontitis and peri-implantitis management. Indeed, the validation of the beneficial effect provided by systemically delivered statins on periodontal and peri-implant tissues may direct recall scheduling, predict response to therapy and, therefore, guide treatment strategies of periodontal and peri-implant treatments in statin users.