1 INTRODUCTION

The most common biological complications associated with dental implants are inflammatory conditions of the surrounding soft and hard tissues secondary to the accumulation of bacterial biofilm. Such conditions are referred to as peri-implant disease, which in turn can be divided into peri-implant mucositis and peri-implantitis.1 Peri-implantitis is a plaque-associated pathological condition characterized by the presence of an inflammatory infiltrate in the connective tissue of the mucosa that progresses apically and induces progressive loss of supporting bone.2

Globally, it is difficult to estimate the real magnitude of peri-implant diseases, due to the great variability in the reported prevalences. In effect, the reported prevalence of peri-implantitis ranges from 1% to 47% at patient level.3 This wide range can be explained by the different definitions used and great heterogeneity regarding follow-up and the studied populations.4 For this reason, it has been recommended that a threshold of ≥3 mm of bone loss, together with clinical signs of inflammation, should be used in epidemiological studies to define peri-implantitis.2 Considering this recommendation, a recent study carried out in Spain in 275 patients found a prevalence of peri-implantitis of 14% at patient level.5

For many years, no specific criteria were established to assess treatment success or to define specific therapeutic protocols in the treatment of peri-implantitis.6 Treatment success should be based on a composite outcome defined as disease resolution, consisting of a peri-implant pocket probing depth (PPD) ≤5 mm, the absence of bleeding on probing and suppuration on probing (BoP/SoP), and no further bone loss >0.5 mm.7, 8

Different protocols have been proposed for the treatment of peri-implantitis, including nonsurgical and surgical approaches. However, the most recent evidence based on in vitro, animal and clinical studies, has found no superiority of any specific decontamination method or treatment modality.9, 10 Nevertheless, some treatment approaches seem to be able to stop or delay the peri-implant destruction process. For example, a 10-year follow-up study of patients undergoing resective therapy together with a strict supportive program revealed that 84% of the implants that achieved disease resolution after surgery remained without signs of peri-implant disease over the study period.11

The decision-making process in the treatment of peri-implantitis may be based on the configuration of the peri-implant bone defect. In this regard, a classification has been published, confirming that most of the defects exhibit a supra- and intra-bony component with or without buccal dehiscence.12 In these cases, a combined approach has been proposed, consisting of implantoplasty of the supra-osseous component of the defect (also in the presence of dehiscence) and reconstruction of the intra-bony component using guided bone regeneration (GBR).12-14

In order to improve the final outcome, the use of antibiotics has also been evaluated as an adjunct to reconstructive surgery.15-17 Antibiotic resistances constitute a major challenge for health professionals worldwide, with the potential to create important problems for health care.18 However, the local application of antibiotics as an adjunct to mechanical decontamination methods has been proposed in the treatment of peri-implantitis, seeking to avoid the undesirable effects of systemic antibiotics. In this respect, the present prospective study was carried out to assess the effectiveness of a new surgical protocol combining implantoplasty and reconstructive therapy together with an antibiotic solution of piperacillin/tazobactam in patients diagnosed with peri-implantitis.

2 MATERIAL AND METHODS 2.1 Study design and participants

A prospective case series study was conducted following the ethical principles of the Declaration of Helsinki. Written informed consent was obtained from all eligible patients prior to enrolment, and ethical approval was obtained from the Institutional Review Board of San Carlos Clinical Hospital (Madrid, Spain) (Ref.: 18/006-E). The study was conducted in compliance with the CARE guidelines.

Participants were consecutively screened for inclusion from among those visiting the Department of Dental Clinical Specialties (Faculty of Dentistry, Complutense University, Madrid, Spain) or Virgen de La Paloma Hospital (Madrid, Spain). The following inclusion criteria were considered: (i) patients ≥18 years of age; (ii) at least one dental implant diagnosed with peri-implantitis (PPD > 5 mm, BoP and/or SoP, and radiographic bone loss ≥3 mm)2; (iii) American Society of Anesthesiologists score I or II; and (iv) presence of a type 1b (infra-osseous defect together with dehiscence), type 3b (2–3 walls defect plus horizontal bone loss), or type 3c defect (circumferential defect plus horizontal bone loss), based on the classification proposed by Monje and colleagues.19

Patients were excluded if one or more of the following criteria was present: (i) untreated periodontitis; (ii) pregnant or breastfeeding women; (iii) immunosuppression and/or treatment with corticosteroids within the last 12 months; (iv) treatment with anticoagulants or acetylsalicylic acid; (v) treatment with bisphosphonates; (vi) signs of dysplasia or precancerous lesions; (vii) allergy to betalactam antibiotics; (viii) previous surgical treatment of peri-implantitis on the included implant(s); and (ix) presence of implant mobility. If during surgery, and after raising the flap and removing the granulation tissue, the defect morphology failed to match the inclusion criteria, the patient was excluded from the study.

2.2 Intervention

All surgeries were performed by the same surgeon with more than 35 years of experience (JMMG). All patients underwent supra- and subgingival debridement using steel curettes (Hu-Friedy, Chicago, IL, USA) and an ultrasonic scaler (Cavitron, Dentsply, NY, USA), and received oral hygiene instructions 1 month before surgery. The prosthesis was removed, and a healing abutment was screwed into place. Patients were required to present full mouth plaque and full mouth bleeding scores of ≤20% before surgery. One week before surgery, subgingival irrigation with a solution of piperacillin/tazobactam 100/12.5 was applied inside the peri-implant pocket.

At the time of surgery, full-thickness flaps were raised using intrasulcular and vertical releasing incisions to expose the buccal and lingual portion of the implant. Granulation tissue was removed with Teflon curettes, and the implant surface was debrided with an ultrasonic scaler (Cavitron, Dentsply, NY, USA). Implantoplasty was performed at the supra-osseous component of the defect and at the buccal and/or lingual dehiscences using large, medium, and fine diamond drills (Italmed, Firenze, Italy) (Figure 1A). The chemical decontamination of the exposed implant surface was performed with a kit containing orthophosphoric acid, chlorhexidine, and the antibiotic solution (Implacure® [MedTech Dental AG, Switzerland]). Briefly, the walls of the defect were protected with sterile gauzes, and the implant surface was decontaminated with 37% orthophosphoric acid and 2% chlorhexidine using a dual syringe containing both products. After 2 min, the implant surface was washed out with sterile saline solution, and the implant surface was scrubbed with gauze impregnated with piperacillin/tazobactam for 1 min (Figure 1B).

Different phases of the combined surgical therapy of a peri-implantitis defect. (A) Implantoplasty of the buccal dehiscence. (B) Dressing impregnated with 37% orthophosphoric acid and 2% chlorhexidine digluconate. (C) Synthetic hydroxyapatite bone substitute hydrated with piperacillin/tazobactam 100/12.5 mg and compacted into the defect. (D) Resorbable collagen membrane hydrated with piperacillin/tazobactam solution placed over the graft

Following decontamination of the implant surface, GBR was performed using a synthetic hydroxyapatite bone substitute with a particle size of 250–1000 μm (Osbone®, Curasan, Kleinostheim, Germany) that was hydrated with piperacillin/tazobactam 100/12.5 mg and compacted into the defect (Figure 1C). A resorbable collagen membrane (Osgide®, Curasan, Kleinostheim, Germany) likewise hydrated with the piperacillin/tazobactam solution was placed over the graft and fixed with titanium tacks (Curasan, Kleinostheim, Germany) (Figure 1D). Flaps were sutured with synthetic 4/0 suture (Supramid®, Laboratorio Aragó, Barcelona, Spain) with the aim of securing closure by primary intention.

Anti-inflammatory (dexketoprofen tromethamol 25 mg, Enantyum®, Menarini, Badalona, Spain) and analgesic medication (magnesium metamizol 575 mg, Nolotil®, Boehringer Ingelheim, Barcelona, Spain) was used every 8 h as needed. No systemic antibiotics were prescribed. Two weeks after surgery, the sutures were removed, and the prosthesis was placed back.

2.3 Study visits and outcomes 2.3.1 Follow-up visits

At the screening visit, a full-mouth periodontal evaluation was carried out and periapical radiographs of all implants were taken. All the clinical and radiographic outcomes were registered at baseline and 1 year after surgery (Figure 2). Patients were scheduled 15 days after surgery and then at 1, 3, 6, and 12 months. At 3, 6, and 12 months of follow-up, the patients underwent supragingival plaque removal using Teflon curettes.

Periapical radiograph at baseline (A) and 1 year after surgery (B) showing a complete radiographic defect fill

2.3.2 Clinical parameters The following clinical parameters were evaluated: Pocket probing depth (mm) at four sites per implant using a manual periodontal probe (CP-12, Hu-Friedy, Chicago, IL, USA). Bleeding on probing at four sites per implant based on a dichotomous (0/1) scale using a manual periodontal probe.20 Suppuration on probing at four sites per implant based on a dichotomous (0/1) scale using a manual periodontal probe. Disease resolution defined as a composite outcome that included the absence of BoP and/or SoP, PPD ≤5 mm, and no bone loss >0.5 mm 1 year after surgery.7 Type of peri-implantitis defects according to their morphology: Class I (infra-osseous/vertical defect), Class II (supracrestal/horizontal defect), or Class III (combined defect).19

One examiner in each center recorded all the clinical outcomes. The calibration session consisted of repeated examinations of 10 implants in 10 patients, spaced 1 week apart before initiation of the study. The intra- and inter-examiner reproducibility achieved ±0.5 mm for PPD in 89% and 85% of the cases, respectively.

2.3.3 Radiographic assessment

Standardized periapical radiographs (Kodak 5100 radiovisiographic system, Kodak Dental System, Atlanta, GA, USA) using the parallel technique (Rinn® system, Dentsply, Weybridge, UK) were used to evaluate the changes in radiographic peri-implant marginal bone level (MBL). Furthermore, PPD was used to represent the peri-implantitis defects according to their morphology combined with periapical radiography.16 Scanned images were measured both at the mesial and distal sites of the selected implant using as landmarks the implant shoulder and the first bone-implant contact. The implant length was used to scale measurements by means of image analysis software (Image-J, National Institutes of Health, MD, USA). An experienced investigator (CBD) performed all the radiographic measurements, with an intra-examiner agreement (correlation coefficient) of 0.985, as determined by means of a calibrating session in which 20 random radiographs were measured twice by the same examiner.

2.4 Statistical analysis

The primary outcome variable was disease resolution (treatment success). Secondary outcomes included mean changes in PPD, BoP, SoP, and MBL. One dental implant per patient was included in the analysis, so each variable was analyzed at patient level. If a patient presented more than one implant meeting the inclusion criteria, all the implants were treated with the studied protocol, but only one was randomly selected for the analysis.

Data were expressed as means, standard deviations, medians, and ranges. The Kolmogorov–Smirnov test and Shapiro–Wilk test were used to assess normal data distribution. Inferential statistical processing was carried out with the 95% confidence interval (95% CI). A logistic regression analysis was performed to determine the factors associated with disease resolution, using as dependent variable the composite outcome for treatment success proposed by Carcuac and colleagues7 (ie, absence of BoP and/or SoP, PPD ≤5 mm and bone loss ≤0.5 mm 1 year after surgery), and as independent variables patient age, gender, smoking (smokers vs never smokers), position (anterior/premolar/molar), arch (maxilla or mandible), time in function, oral hygiene, defect configuration (Class I vs Class III defects), mean PPD at baseline, deepest PPD at baseline, presence/absence of suppuration, and MBL at baseline. The results of the logistic regression analyses were reported as odds ratios (ORs) in univariate associations and in a multivariate model in which a “change-in-estimate” approach (a change in adjusted OR for a covariate of ≥10% compared to the crude OR) was used. Results were considered statistically significant at p < 0.05. Software packages (SPSS® version 23.0, IBM, Armonk, NY, USA; and STATA13.1, StataCorp, College Station, TX, USA) were used for all data analyses.

3 RESULTS 3.1 Study sample

Seventy-three consecutive patients were screened, of whom 43 fulfilled the inclusion criteria and were recruited in the study (12 from Virgen de La Paloma Hospital and 31 from the Complutense University, Madrid). The baseline demographic, clinical, and radiographical parameters are depicted in Table 1. All patients presented bone level implants, 38 with an 0.3-mm machined collar at the level of the implant shoulder (BioHorizons RBT; Internal Implants, Birmingham, AL, USA) and 5 with an acid-etched surface up to the implant-abutment interface (Phibo TSA™, Phibo Dental Solutions).

TABLE 1. Mean values (SDs) or percentages (n, %) for demographic, clinical, and radiographic parameters Variable All patients/implants (n = 43) Disease resolution Yes (n = 37) No (n = 6) p value* Age (years) 60.2 (9.3) 60.7 (9.5) 57.0 (7.8) 0.370 Gender Male 15 (34.9%) 13 (35.1%) 2 (33.3%) 0.932 Female 28 (65.1%) 24 (64.9%) 4 (66.7%) Smoking Never or former smokers 21 (48.8%) 19 (51.4%) 2 (33.3%) 0.413 Smokers 22 (51.2%) 18 (48.7%) 4 (66.7%) Position Anterior (I/C) 16 (37.2%) 13 (35.1%) 3 (50.0%) 0.783 Premolar 9 (20.9%) 8 (21.6%) 1 (16.7%) Molar 18 (41.9%) 16 (43.2%) 1 (33.3%) Arch Maxilla 16 (37.2%) 13 (35.1%) 3 (50.0%) 0.485 Mandible 27 (62.8%) 24 (64.9%) 3 (50.0%) Time in function (years) 7.8 (3.7) 8.4 (3.6) 4.0 (1.7) 0.006 Oral hygiene Brushing 1–2 times/day 10 (23.2%) 7 (18.9%) 3 (50.0%) 0.208 Brushing 3 times/day 6 (14.0%) 5 (13.5%) 1 (16.7%) Brushing 3 times/day + interdental hygiene 27 (62.8%) 25 (67.6%) 2 (33.3%) Defect configuration Class I defects 31(72.1%) 29 (78.4%) 2 (33.3%) 0.153 Class III defects 12 (27.9%) 8 (21.6%) 4 (66.7%) PD (mm) at baseline 6.4 (2.1) 6.6 (2.1) 5.2 (1.7) 0.135 BoP at baseline 43 (100%) 37 (100%) 6 (100%) - SoP at baseline 21 (48.8%) 19 (51.3%) 2 (33.3%) 0.413 MBL (mm) at baseline 5.8 (2.1) 6.0 (2.1) 4.9 (1.4) 0.256 PD (mm) at 12 months 3.2 (1.0)** 3.1 (0.9)** 3.9 (1.3) 0.067 BoP at 12 months 6 (14.0%)** 0 (0%)** 6 (100%) <0.001 SoP at 12 months 0 (0%)** 0 (0%)** 0 (0%)** - MBL (mm) at 12 months 3.2 (2.2)** 3.1 (2.2)** 3.3 (2.3) 0.903 Abbreviations: BoP, bleeding on probing; I/C, incisors and canines; MBL, marginal bone loss; n, number of patients/implants; PD, probing depth; SD, standard deviation; SoP, suppuration on probing. * p value for the comparison between successfully treated implants and not resolved peri-implantitis. ** Statistically significant difference when compared to baseline (p < 0.001).

All patients completed the 1-year follow-up period. There were no intra- or postoperative complications, no implants were lost, and no implant fractures were reported.

3.2 Clinical and radiographic outcomes

All patients presented screw-retained restorations (30 with single-crown restorations, 5 with two implant-abutment prosthesis, and 1 with three implant-abutment prosthesis). At 12 months, 86% (n = 37) of the peri-implantitis lesions showed disease resolution. Among the unresolved cases, all presented BoP, one out of 6 implants (2.3% of the total sample) had mean PPD > 5 mm at 12 months, and none presented further bone loss >0.5 mm. Moreover, four cases out of six with unsuccessful results presented a radiographic defect fill >0.5 mm. The differences in the clinical outcomes between successful and unsuccessful implants are shown in Table 1.

Bleeding on probing was present in 100% of the patients at baseline. However, 1 year after surgery, BoP showed a reduction of 86% at patient level (p < 0.001). SoP was present in 48.8% of the patients at baseline and was completely eradicated 1 year after surgery (p < 0.001). At 12 months, the mean PPD was seen to have decreased from 6.4 ± 2.1 mm to 3.2 ± 1.0 mm, representing a mean reduction of 3.2 ± 2.0 mm (p < 0.001). The frequency distribution of PPD at baseline and 12 months after treatment is depicted in Table 2.

TABLE 2. Number of patients/implants (%) with various probing depth (PD) and marginal bone level (MBL) categories at baseline and 12 months after treatment Baseline 12 months Mean PD PD < 4 mm 3 (7.0%) 30 (69.8%) 4 mm ≤ PD ≤ 5 mm 7 (16.2%) 11 (25.6%) 5 mm ≤ PD ≤ 6 mm 6 (14.0%) 1 (2.3%) PD ≥ 6 mm 27 (62.8%) 1 (2.3%) Mean MBL MBL ≤ 3 mm 2 (4.6%) 25 (58.1%) 3 mm < MBL ≤ 5 mm 12 (27.9%) 11 (25.6%) 5 mm < MBL ≤ 7 mm 22 (51.2%) 6 (14.0%) MBL > 7 mm 7 (16.3%) 1 (2.3%)

At baseline, the mean MBL was 5.8 ± 2.1 mm. One year after surgery the mean value was 3.2 ± 2.2 mm, representing a mean radiographic defect fill of 2.6 ± 1.5 mm (p < 0.001). The frequency distribution of MBL at baseline and 12 months after treatment is depicted in Table 2. The percentage of subjects with mean MBL > 5 mm at baseline was 67.5% (29 subjects), while the percentage at 12 months decreased to 16.3% (7 subjects). The frequency distribution of the bone level changes is shown in Table 3. Twelve months after treatment, 55.8% of the implants presented a radiographic defect fill >3 mm.

TABLE 3. Frequency distribution of marginal bone level (MBL) changes (number [%]) after treatment MBL change All cases Disease resolution Unsuccessful cases <−0.5 mm 0 (0%) 0 (0%) 0 (0%) −0.5 to 0.5 mm 4 (9.3%) 2 (5.4%) 2 (33.3%) 0.5 to 2 mm 8 (18.6%) 6 (16.2%) 2 (33.3%) 2 to 3 mm 7 (16.3%) 7 (18.9%) 0 (0%) >3 mm 24 (55.8%) 22 (59.5%) 2 (33.3%) Note: Negative values (−) indicate radiographic bone loss. Positive values indicate radiographic defect fill. 3.3 Factors associated with disease resolution

The results of the univariate regression analysis indicated that the odds of achieving a successful outcome were greater among those implants with ≥5 years in function (OR = 0.2; 95% CI [0.0; 1.0]; p = 0.050) (Table 4). On the other hand, it was harder to achieve disease resolution in implants presenting combined defects as compared to implants presenting just an infra-osseous defect (OR = 7.3; 95% CI [1.1; 47.0]; p = 0.038). Those implants presenting a site with a probing depth >8 mm at baseline showed a tendency to fail with the composite outcome for disease resolution (OR = 5.1; 95% CI [0.9; 34.2]; p = 0.073). When combined in a multivariate regression analysis, no risk or protective factors were significantly associated with disease resolution—though combined defects showed a tendency to correlate with unsuccessful outcomes (Class III defects, OR = 5.7; 95% CI [0.0; 1.6]; p = 0.080).

TABLE 4. Factors associated with disease resolution at 12 months Crude model Adjusted multivariate model OR 95% CI p value OR 95% CI p value Age (years) <50 Reference ≥50 1.0 0.1–9.8 0.978 Gender Male Reference Female 1.1 0.2–6.7 0.932 Smoking Never or former smokers Reference Smokers 2.1 0.3–13.0 0.420 Position Anterior (I/C) Reference Premolar 0.5 0.0–6.1 0.621 Molar 0.5 0.1–3.7 0.534 Arch Maxilla Reference Mandible 0.5 0.1–3.1 0.489 Time in function (years) <5 Reference Reference ≥5 0.2 0.0–1.0 0.050 0.2 0.0–1.6 0.129 Oral hygiene Brushing 1–2 times/day Reference Brushing 3 times/day 0.5 0.0–5.9 0.556 Brushing 3 times/day + interdental hygiene 0.2 0.0–1.3 0.096 Defect configuration Class I defect Reference Reference Class III defects 7.3 1.1–47.0 0.038 5.7