At the cellular level, osteocytes may perform a mechanosensor role (“mechanostat”) and respond to strain in the bone by releasing cytokines. In this way the skeleton is able to adapt the bone form and size to its function in the most efficient manner. According to this hypothesis, low micro-strain is associated with bone resorption, whereas high micro-strain causes increased bone mass (Frost 1987). However, evidence to support this hypothesis has described raised cytokine levels in tissues surrounding failed knee prostheses, but this finding is related to the metal and polyethylene debris, and not to low-strain loading (Chiba et al. 1994). Based on in vitro data, one radical hypothesis suggested that the chemokine ­interleukin-8 is released into the gingival tissues by the contact of blood with the titanium implant surface and stimulates an additional inflammatory mucosal reaction which is independent of bacteria (Quabius et al. 2012). A study found no ­difference in the levels of interleukin-8 in the implant crevicular fluid of patients with peri-implantitis compared to those with healthy implants (Severino et al. 2011), but another study found increased levels (Luo et al. 2011). So far, research has not elucidated the mechanism which determines whether bone loss or formation occurs with applied load, and pharmacological manipulation of the process cannot therefore be used.

The clinical experience of many clinicians has made them cautious in implant treatment planning for patients with severe bruxism. Whether the excessive occlusal load directly or indirectly causes bone loss and implant failure is controversial. Animal studies have shown that excessive dynamic loads can cause marginal bone resorption and crater-like defects around tibial implants (Duyck et al. 2001), whereas static loads had no effect. The loads were applied 6 weeks after implant insertion, and with the implant site in a healing phase, this may have influenced the response to applied load. However, the clinical implication of this result is that the static loads introduced as a result of small casting errors may be tolerated, whereas excessive dynamic loads in chewing cause bone micro-fractures. Cyclic loading of the radius of dogs with excessive loads has been shown to cause bone micro-fractures that were significantly associated with resorption spaces (Burr et al. 1985). Micro-damage of bone can therefore initiate resorption.

Many studies have found that the nonsurgical treatment of peri-implantitis does not produce predictable and sustained clinical improvement. In Fig. 1.3 and 1.4 continuous alveolar bone loss occurred despite regular implant surface debridement and oral hygiene instruction. The difficulty arises for the dentist in cleaning a moderately roughened implant surface using traditional hand instruments. Similar difficulties occur for the patient in cleaning sub-gingival restoration ­margins. Air-abrasive devices provide short-term (6 months) improvement in probing depth and clinical attachment level when treating moderate degrees of peri-implantitis (Sahm et al. 2011). Other studies have found no improvement in probing pocket depth and bone levels with sub-gingival debridement using ultrasonic devices and curettes (Karring et al. 2005). In a small, randomized controlled trial (Heitz-Mayfield et al. 2011), nonsurgical debridement had some effect in reducing bleeding on probing around implants (healing was successful in 38 % of treated implants at 3 months), but brushing with chlorhexidine gel had no additive effect. Another short-term, 3-month study found a similar conclusion: that mechanical debridement was effective in reducing mucosal inflammation when implant pocket depths were less than 5 mm, but that chlorhexidine gel provided no additional benefit (Porras et al. 2002). There is also a tendency of chlorhexidine to cause staining and an altered taste sensation, but chlorhexidine gel does not harm the titanium surface when used as disinfecting agents (Ungvari et al. 2010).

In a 6-month oral hygiene programme, which recruited patients with mild peri-implantitis, an air-abrasive system was found to reduce the mean probing depth by about 0.5 mm (Sahm et al. 2011). These changes are not clinically significant and may result from the patient’s improved oral hygiene rather than the air-abrasive intervention. In peri-implantitis, one application of an air-abrasive device or an erbium-doped yttrium, aluminium, and garnet (Er:YAG) laser was ineffective at reducing bacterial counts after 6 months or in showing any significant improvement in clinical parameters (Persson et al. 2011).

Peri-mucositis can be prevented by twice-daily use of an antiseptic mouthwash (Listerine, Warner-Lambert Co., NJ) when used as an adjunct to toothbrushing. This was shown in a double-blind, randomized clinical study that compared the antiseptic with a similar-tasting placebo mouthwash (Ciancio et al. 1995). This measure is simple, effective and readily accepted by patients. Listerine, hydrogen peroxide and chlorhexidine all have an antibacterial effect on oral biofilm microorganisms adhering to a titanium surface (Gosau et al. 2010), but their value in treating established peri-implant disease is in doubt. A 3 % hydrogen peroxide applied for 5 min or chlorhexidine gel solution applied for 5 min does not harm the titanium surface (Ungvari et al. 2010), but high-fluoride, acidic gels do cause corrosion of titanium (Stajer et al. 2008). These antiseptic solutions are supplemental measures because plaque must be mechanically removed, as failure to do so will result in bacterial recolonization of the biofilm.

Nonsurgical therapy does not provide successful results in the treatment of severe peri-implantitis but does have a place in establishing optimal gingival tissue health prior to surgery. Residual pockets tend to remain and these usually require surgical elimination, but resective surgery can cause aesthetic problems in the anterior region of the mouth. A Cochrane review was unable to determine which technique was the most effective treatment for peri-implantitis (Esposito et al. 2006). This was because of problems in the reported clinical trials with small sample sizes, short follow-up periods, analysis based on implants and not patients and failures to report the initial severity of the peri-implantitis. In the absence of clear guidance from the available literature as to which treatment of peri-implantitis is superior, the clinician should use that intervention which has minimal side effects and cost.

In a study that undertook plaque control, pre-surgical antibiotics, surgical pocket elimination and bone recontouring, half of the patients (48 %) had no bleeding on probing after a short period of follow-up (up to 2 years) (Serino and Turri 2011). Treatment was more successful (74 %) in those implants with minor levels of initial bone loss compared with only 50 % for those implants with an initial bone loss of 5–6 mm. Treatment ­success is mainly dependent on the severity of bone loss around the implant. If bone loss around the implant is limited (at less than 2 mm), a nonsurgical treatment approach is recommended (Okayasu and Wang 2011). Peri-implant bone loss can be treated successfully with flap surgery, removal of granulation tissue and effective cleaning of the implant surface, resulting in successful repair in the majority of patients. One study reported a median defect depth reduction of 6.2 mm using this protocol (Behneke et al. 2000). However, ­re-osseointegration following surgery is unlikely. A histological study in dogs found that following guided tissue regeneration, “fine fibrillar material” formed adjacent to the implant surface (Schupbach et al. 1994). Treated implants become stabilized and are well adapted to new bone but are not osseointegrated. In a systematic review which included 17 articles, it was concluded that the most likely outcome from using bone grafting techniques and membranes is an incomplete filling of the bony defect (Sahrmann et al. 2011).

Preventing further peri-implant bone loss is challenging, especially when it is accompanied by pocket depths >5 mm. Reports have documented the successful use of antimicrobial therapies combined with growth factors and guided tissue-regenerative membrane therapies (Froum et al. 2012). Membranes are used in surgery to stabilize the bone graft around the peri-implant defect. However, incorporating a membrane inevitably increases the complication rate in regenerative surgery. One study reported a loss of the membrane in 5 cases out of 42 augmentations due to infection, with further ­complications from membrane dehiscence (Gaggl and Schultes 1999). Membrane instability, exposure of the membrane or a failure to adequately remove bacterial contamination from the implant surface at surgery may be responsible for augmentation failure.

Many studies have reported the use of an air-powder cleaning device or citric acid to clean the implant surface. Air-powder systems and resin curettes result in far less surface alteration than using steel curettes or ultrasonic devices designed for cleaning teeth (Meschenmoser et al. 1996; Brookshire et al. 1997). Short, 5-s exposures to the air-abrasive system did not cause harmful changes to the implant surface, but longer 15-s exposures did roughen the implant surfaces (Chairay et al. 1997). It is difficult to make general clinical recommendations as different studies used different operating parameters, e.g. the air pressure of the device and the nature and particle size of the powder.

The use of antibiotics as the sole treatment of peri-implantitis has little support in the literature. A variety of different antibiotic regimes have been recommended but with unpredictable outcomes. A systematic review found no evidence in favour of any particular antibiotic treatment protocol (Klinge et al. 2002). As with the treatment of chronic periodontitis, the resolution of peri-implantitis requires local debridement or curettage and regular plaque control, with antibiotics playing a less important role (Ericsson et al. 1996). Recolonization of the peri-implant pocket and further bone loss are inevitable without debridement.

It would be prudent for a clinician to start with the minimum treatment to debride the pocket and prevent further bone loss rather than with a more aggressive treatment that may be unpredictable.