Efficacy and safety of PERIOdontal treatment versus usual care for Nonalcoholic liver disease: protocol of the PERION multicenter, two-arm, open-label, randomized trial

Background We report the first protocol for a multicenter, randomized comparison study to compare the efficacies of periodontal scaling and root-planing treatment against that of tooth-brushing treatment for nonalcoholic fatty liver disease (NAFLD) (PERION: PERIOdontal treatment for NAFLD). Nonalcoholic steatohepatitis (NASH) is an advanced form of NAFLD, which can progress to cirrhosis and hepatocellular carcinoma. Increased endotoxemia is associated with the progression of NAFLD. Periodontal bacteria possess endotoxins; Porphyromonas gingivalis is well-known as a major pathogenic bacterium in periodontitis, and serum antibody levels for P. gingivalis are high in patients with periodontitis. Several reports have indicated that P. gingivalis is related to NAFLD. This study aims to investigate the effect of periodontal treatment for liver damage, P. gingivalis infection, and endotoxemia on patients with NAFLD. Methods We will include adult patients (20–85 years old) with NAFLD, alanine aminotransferase (ALT) ≥ 40 IU/L, and equivalent steatosis grade ≥ 1 (target sample size, n = 40 patients; planned number of patients with outcome data, n = 32). Participants will be randomly assigned to one of two groups: a scaling and root-planing group or tooth-brushing as the usual group. The primary outcome will be the change in ALT levels from baseline to 12 weeks; the key secondary outcome will be the change in the serum immunoglobulin G (IgG) antibody titer for P. gingivalis at 12 weeks. Discussion This study should determine whether periodontal treatment decreases liver damage, P. gingivalis infection, and endotoxemia in patients with NAFLD. Trial registration University Hospital Medical Information Network (UMIN) Clinical Trials Registry, ID: UMIN000022079.


Background
The broad spectrum of fatty liver diseases in individuals who consume little-to-no alcohol is called nonalcoholic fatty liver disease (NAFLD) and includes nonalcoholic steatohepatitis (NASH). NASH is an increasingly common cause of chronic liver disease worldwide and is associated with increased liver-related mortality and hepatocellular carcinoma [1][2][3]. NASH progresses to cirrhosis in 15-20% of the affected individuals and is a rising indication for liver transplantation [4]. However, approved therapies for NASH have not yet been established; therefore, preventive therapies to inhibit the progression of fatty liver disease to NASH are required.
Periodontal disease is an infectious disease of the gums and tissues surrounding the teeth and causes tooth loss resulting from the destruction of tooth-supporting tissues. The incidence rate of periodontitis is > 47% in adults in the USA [5]. More than 700 bacterial species or phylotypes have been detected in the oral cavity [6]. Some species/ complexes are closely associated with advanced periodontal lesions, such as Porphyromonas gingivalis, Treponema denticola, Tannerella forsythia, Prevotella intermedia, Fusobacterium nucleatum, and Aggregatibacter actinomycetemcomitans [7,8]. Among them, P. gingivalis, a gram-negative anerobic bacterium, is the major etiologic agent that contributes to periodontal disease progression and bone and tissue destruction [9,10]. The lipopolysaccharide (LPS) cell-wall component of P. gingivalis is one of the virulence factors that trigger a wide range of host responses, including the production of proinflammatory cytokines, anti-inflammatory cytokines, and chemokines [11]. These cytokines and inflammatory mediators play important roles in the progression of periodontitis at the stage where host immune and inflammatory responses lead to the destruction of periodontal tissue under the influence of multiple behavioral, environmental, and genetic factors [12].
Recently, several studies have reported the relationship between NAFLD and periodontal disease [13,14]. Yoneda et al. [15] reported that the detection frequency of P. gingivalis in the saliva of patients with NAFLD and patients with NASH was significantly higher than that in non-NAFLD control subjects. Moreover, they presented preliminary evidence to suggest that nonsurgical periodontal treatments in 10 patients with NAFLD for 3 months ameliorated the liver function parameters, such as the serum levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT). Consequently, it is thought that infection with a periodontal pathogen, mainly P. gingivalis, is associated with fibrosis severity in patients with NAFLD and that the prevention and elimination of P. gingivalis infection by periodontal treatment may have a beneficial effect on the management of NASH (Fig. 1).
Therefore, we hypothesized that the elimination of oral infection, including P. gingivalis infection, by periodontal treatment in patients with NAFLD would ameliorate NAFLD-related clinical markers. We performed a clinical study to confirm the preliminary finding under collaborative medical and dental care.
Thus, we have devised a prospective, multicenter, randomized comparison trial to evaluate periodontal treatment as a candidate for NAFLD treatment. This is the first protocol for a randomized comparison trial for periodontal treatment against NAFLD in humans.

Design
The PERION trial is designed as a prospective, multicenter, two-arm, randomized comparison study to test the efficacy of the 12-week scaling and root-planing group versus the tooth-brushing group in NAFLD with moderate periodontitis. The study will recruit 40 adults and evaluate the efficacy and safety of periodontal treatment for 60 weeks, with the primary endpoint at 12 weeks. The study design is shown in Fig. 2.

Recruitment process and allocation
The PERION trial patient population will be derived from the Kanagawa Dental University Yokohama Clinic, Kanagawa Dental University, Iwasaki Internal Medicine Clinic, and the Yokohama City University Hospital Cohort. The randomized allocation will be conducted at Yokohama City University. Eligible patients will be screened by the principal and sub-investigator (gastroenterologists and periodontists). Patient recruitment will be performed 8 h a day, 5 days a week.

Endpoint detection
In the short-term studies (Phases I and IIa), aimed primarily at detecting wasted signals to make direct decisions on further development, a sustained improvement in AST and ALT levels will be useful as the endpoint of PERION (Fig. 3) [16]. While the use of ALT as a surrogate marker for NAFLD is controversial, studies have shown that ALT reduction is associated with reduced hepatocyte damage and liver inflammation [17], but not steatosis [18]. Because there are no other commonly established noninvasive biomarkers for use in the NAFLD/NASH clinical trial, ALT reduction was selected as the primary endpoint of this trial. To assist the primary endpoint, several secondary endpoints were selected to assess the pathogenesis of NAFLD using noninvasive methods (Fig. 3). Therefore, the primary endpoint of PERION is set as a change in ALT levels from baseline after 12 weeks of intervention. Also, the PERION study will evaluate periodontal treatment as a candidate for the first treatment to improve NAFLD pathogenesis via decreasing P. gingivalis count. Therefore, the key secondary endpoint is the serum immunoglobulin G (IgG) antibody titer of P. gingivalis [19].

Periodontal disease examination
The subject will be examined to assess the amount of periodontal disease bacteria (P. gingivalis, etc.), degree of infection, and periodontal disease severity. All examinations will be performed by two periodontal disease specialists enrolled at Kanagawa Dental University Hospital and Kanagawa Dental University Yokohama Clinic. The amount of P. gingivalis in saliva is measured by quantitative polymerase chain reaction (qPCR), and the infection level of periodontal disease is examined using a serum IgG antibody titer test for P. gingivalis FDC381 by enzyme-linked immunosorbent assay (ELISA). The severity of periodontal disease will be examined by probing depths, clinical attachment levels, gingival bleedings on probing (BOP) at six sites per tooth using a calibrated periodontal probe, and the stability of the teeth.

Periodontal treatment
The primary purpose of nonsurgical periodontal treatment is to control periodontal infection of microorganisms by removing bacterial biofilm, calculus, and toxins from the root surface. According to a review of the scientific literature, mechanical nonsurgical periodontal treatment significantly reduces the inflammation level and the probing pocket depth and increases the clinical attachment level [20]. The successful treatment of plaque-induced periodontitis will recover periodontal health, but gingival recession often occurs. At first, instruction on correct brushing of the teeth will be given to patients with periodontal disease (tooth-brushing group). Then, the removal of supra-and sub-gingival bacterial plaque/ biofilm and calculus by periodontal scaling and rootplaning will be performed. Quadrant/sextant-wise instrumentation (conventional staged debridement, CSD) will be performed (scaling and root-planing group).

NAFLD treatment
We will provide standard lifestyle modification recommendations that each site can provide to the patients at the time that the informed consent document is signed. We will recommend a hypocaloric diet (daily calorie reduction of 500-1000 kcal) with reduced consumption of processed carbohydrates and fructose-containing beverages and will recommend the performance of moderateintensity exercise for 30-45 min three to four times/ week. This diet and exercise therapy will be performed without changing the prescription.

Sample size determination
Our previous pilot study showed that nonsurgical periodontal treatments on 10 NAFLD patients for 3 months ameliorated the liver function parameters, such as the serum levels of AST and ALT [15]. It showed a mean On the basis of these data, the sample size is determined to guarantee the power of the analysis of variance F test. Assuming the mean changes in ALT in the no-treatment group and active treatment groups to be 0 and − 25, respectively, with a common standard deviation of 25, the required number of patients per group with a power of 80% and a two-sided significance level of 5% was calculated to be 16. We aim to recruit a total of 40 patients (scaling and root-planing treatment, 20, tooth-brushing treatment, 20) to compensate for the dropout patients.

Eligibility criteria
Patients with NAFLD who will be recruited in this study must satisfy the inclusion and exclusion criteria presented in Tables 1 and 2. Fatty liver, steatosis grade and fibrosis stage will be assessed using noninvasive methods (ultrasound, vibration-controlled transient elastography (VCTE), and magnetic resonance imaging (MRI)).

Randomization
Before providing informed consent, patients will undergo a screening test as a first step to determine whether they meet the inclusion criteria for the study and none of the exclusion criteria. For patients determined to be eligible, the principal investigator or coinvestigator will complete a Patient Enrollment Form with the necessary information, which is sent by fax or e-mail to the Patient Enrollment Center (Yokohama City University). The Patient Enrollment Center will then confirm the eligibility of the patient based on the enrollment form, enroll and randomize the patient, and notify the principal investigator or co-investigator of the Patient ID number and the allocation number via fax or email. After an eligibility check, patients will be randomly assigned to receive periodontal treatment or no treatment at the Central Registration Center by a computer program, using a stratified block randomization method, adjusting for age (≥ 65/< 65 years) and sex (male/female). Therefore, the patient assignment is concealed from the investigator.

Study procedures
The schedule of assessments for the study procedures is summarized in Table 3.

Observation of adverse events
All adverse events (AEs) during the study will be recorded with regard to the following: date of onset and date of completion (if applicable), severity of AEs, investigator's view on the relationship to periodontal treatment, information on actions taken, information on the treatment of the AE, the cause of the event (if known), and the solution or outcome. AEs classified as serious will be recorded in the serious adverse event reporting tool and reported to the sponsor. AE intensities are graded according to the National Cancer Institute's Adverse Event Common Terminology (NCI CTCAE) version 4.03, including the AE-intensity classifications shown in Table 4.

Additional study procedures
Patients must be fasted for at least 8 h prior to the visit.

Criteria for discontinuation of study treatment
The study must be discontinued under the following conditions: drug-induced liver injury, unacceptable toxicity, acute viral hepatitis B and C, autoimmune or alcoholic hepatitis, hypoxic or ischemic liver injury, biliary tract disease, or pregnancy.

Outcome measures Evaluation of efficacy
The efficacy endpoints for this study are shown in Fig. 3.

Evaluation of safety
The safety and tolerability of periodontal treatment will be evaluated over the 60 weeks of treatment in NAFLD patients with periodontitis. This will include the evaluation of AEs, clinical laboratory tests, physical examination, and vital signs. The clinical laboratory tests will include liver and fasting metabolic parameters. Liver parameters will include alkaline phosphatase (ALP), AST, ALT, total bilirubin (T-Bil), and γ-glutamyltransferase (GGT).

Statistical analyses
The Full Analysis Set (FAS) will be the primary analysis set for efficacy. We define FAS to include any subjects who receive any amount of the study medication without a lack of information on the primary endpoint (complete case analysis). For the blood liver function test using ALT, the summary statistics (mean, standard deviation) will be calculated at baseline and 12 weeks. Details of the statistical design will be described in the Statistical Analysis Plan. In this study, statistical analysis will be Other patients who are inappropriate as participants in this research in the opinion of the principal investigator, etc.
NAFLD nonalcoholic fatty liver disease, PERION PERIOdontal treatment for NAFLD performed mainly for the following items. For the primary analysis, (ALT after treatment -ALT at baseline) will be sought for each subject and the corresponding Wilcoxon test will be performed. The significance level will be 5% on both sides. In addition to p values, we will provide point estimates with 95% confidence intervals. As a sensitivity analysis, the two groups would be compared using the analysis of covariance using the primary endpoint as an outcome, adjusting for ALT at baseline as a covariate. For the key secondary endpoint, P. gingivalis IgG antibody titer in the blood (values after treatmentvalues at baseline) will be sought for each subject and the paired Wilcoxon test will be performed. Subgroup analysis will be conducted. Stratified analysis will be performed using P. gingivalis bacterial content, proton density fat fraction (PDFF), magnetic resonance  All objectives will be compared between the periodontal scaling and root-planing treatment group and the tooth-brushing treatment group. ALP alkaline phosphatase, ALT alanine transaminase, AST aspartate transaminase, BMI body mass index, BUN blood urea nitrogen, CAP controlled attenuation parameter, CK-18 cytokeratin 18, Cr creatinine, EAA endotoxin activity assay, eGFR estimated glomerular filtration rate, FBS fasting blood sugar, GGT γ-glutamyltransferase, h-CRP high-sensitivity C-reactive protein, HDL-C high-density lipoprotein-cholesterol, HOMA-IR homeostasis model assessment of insulin resistance, HRQOL health-related quality of life, IL-6 interleukin-6, LDL-C low-density lipoprotein-cholesterol, MRE magnetic resonance elastography, MRI magnetic resonance imaging, NAFLD nonalcoholic fatty liver disease, NGS next-generation sequencer; PERION, PERIOdontal treatment for NAFLD, PDFF proton density fat fraction, qPCR quantitative polymerase chain reaction, SF-8 short form-8, T-Bil total bilirubin, T-Cho total cholesterol, TG triglycerides, TNF-α tumor necrosis factor-α elastography (MRE), and periodontal pocket as indices. Stratified analysis will be performed using above/below the median of P. gingivalis bacterial content, endotoxin, PDFF, MRE, and periodontal pocket depth. We hypotheses a larger reduction in ALT with the intervention in the patient group with high P. gingivalis bacterial content, high endotoxin levels, high PDFF, low MRE, and deep periodontal pockets at baseline. Therefore, we conducted multivariable linear regression with difference in ALT between groups as dependent variable and at least the treatment group, the respective subgroup variable (categorized), and the interaction term (treatment group×subgroup variable) as independent variables. The significance level will be 5% on both sides. In addition to p values, we will provide point estimates with 95% confidence intervals. As a sensitivity analysis, the two groups will be compared using analysis of covariance using the key secondary endpoint as an outcome adjusting for P. gingivalis IgG antibody titer at baseline as a covariate.
As the safety analysis, the incidence and severity of AEs and reactions will be calculated.

Trial Steering Committee and Data Monitoring Committee
The Trial Operation Committee is integrated and consists of three persons appointed by independent clinical and basic investigators (a general internist in primary care, a palliative care specialist and a statistician from Yokohama City University School of Medicine). They provide overall supervision and ensure that all registered trials. These investigators are anonymous and randomly selected. The Independent Data Monitoring Committee will be established with two persons from the Department of Biostatistics, Yokohama City University School of Medicine. The management team will observe progress and data monthly by phone call, mail, and/or webconferencing. If the Monitoring Committee decides that on-site monitoring is necessary, monitoring members will visit the site for face-to-face monitoring.

Discussion
This is the first study proposed to explore the effect of periodontal treatment on NAFLD patients with periodontitis. The primary endpoint used in previous studies was the liver histology, which was evaluated by a liver biopsy (Pivens, Flint, and Golden study) [21][22][23][24]. Liver histology endpoints, such as the complete resolution of NASH, are considered surrogates for preventing cirrhosis (i.e., they are thought to predict a clinical benefit but are not direct measurements of it). However, due to increased cost, possible risk, inter-and intra-observer bias, and healthcare resource utilization, an invasive liver biopsy is poorly suited as a diagnostic test for such a prevalent condition [25]. Furthermore, the histological lesions of NASH are unevenly distributed throughout the liver parenchyma; therefore, liver biopsy sampling error can result in substantial stratification and staging inaccuracies [26]. Currently, the noninvasive methods used to assess NASH progression are not robust enough to replace liver biopsy. We considered that it is important to perform noninvasive, safe, low-cost, and shortterm clinical trials as proof-of-concept studies. Many studies reported that the steatosis grade can be evaluated using controlled attenuation parameter (CAP) [27][28][29][30][31][32], which is based on the properties of ultrasonic signals acquired by VCTE, and MRI-PDFF [33,34], which is an MRI-based method for quantitatively assessing hepatic steatosis and is available from several manufacturers of MRI scanners. Moreover, VCTE and MRE have superior diagnostic ability to evaluate steatosis and fibrosis in patients with NAFLD [35,36]. Therefore, we consider that noninvasive evaluation of NAFLD pathogenesis using VCTE and MRI could replace invasive methods, such as liver biopsy, as a proof-of-concept study. Therefore, noninvasive methods, such as VCTE and MRI, to assess NASH/NAFLD progression, were included as secondary endpoints to compare against liver biopsy. Clinical research in Japanese university students has suggested that having periodontal disease in young men was significantly associated with an increased level of ALT [37]. In addition, the incidence of periodontal disease in healthy Japanese women was reported to be significantly increased with elevated serum levels of AST, ALT, and cholinesterase [38]. Furthermore, an observational study with annual workplace health check-ups at a company in Japan reported an association between periodontal condition and the combination of elevated ALT and metabolic syndrome (MetS) in men [39]. Besides, it has been suggested that more severe periodontal disease is associated with increased serum levels of GGT, a liver biochemical parameter, in Japanese adults with no alcohol-drinking habits [40]. In an in-vivo mouse study, it was demonstrated that areas of fibrosis with proliferation of hepatic stellate cells and collagen formation were observed in mice with P. gingivalis infection fed on a high-fat diet. In addition, in steatotic hepatocytes, the expression of toll-like receptor 2 (TLR2), one of the P. gingivalis-LPS receptors, was upregulated. P. gingivalis-LPS further increased messenger ribonucleic acid (mRNA) levels of palmitate-induced inflammasome and proinflammatory cytokines in steatotic hepatocytes [41]. That is to say, the dental infection of P. gingivalis exacerbated the pathological progression of NASH from simple steatohepatitis to steatohepatitis with fibrosis through the upregulation of the P. gingivalis-LPS-TLR2 pathway and activation of inflammasomes. Recently, evidence in mice has shown that disturbance of the gut microbiota composition by orally derived periodontopathic bacteria, such as P. gingivalis, may be a causal mechanism linking periodontitis and systemic disease including NAFLD [42][43][44].
In conclusion, this study should determine whether periodontal treatment to decrease endotoxin levels and treat P. gingivalis infection improves the disease status of patients with NAFLD. The PERION study is the first randomized controlled study suppressing hyper-endotoxemia in NAFLD with periodontitis. This study should allow the assessment of the efficacy and safety of periodontal treatment in a larger population of NAFLD patients with periodontitis.

Strengths
To our knowledge, no direct comparison has been made between randomized controlled groups of patients with NAFLD with periodontitis. The strength of this study is that non-invasive MRI-PDFF will be used to assess changes in hepatic fat mass, as opposed to liver biopsy. Importantly, improved laboratory tests for periodontal disease and P. gingivalis removal may correlate with improved liver fat mass and liver function.

Limitations
There are several limitations in our study. First, the small sample size and short treatment period (3 months) limit our findings. Second, it is important to differentiate NASH in NAFLD patients; however, at present, the gold standard for such differentiation is liver-tissue diagnosis by liver biopsy. In this study, NAFLD patients would not have undergone liver biopsy. Third, our protocol is open-label. Fourth, the specified primary and key secondary outcomes are only surrogate outcomes. The extent of decrease in ALT or IgG antibody titers will translate into improved quality of life (QOL) or any other more patient-relevant outcomes of NAFLD patients will not be determined in the present trial.