Peri-implantitis: bleeding around an implant, and what it means

The 2017 World Workshop set out case definitions for the tissues around implants. Health is the absence of redness, bleeding on probing, swelling and suppuration. Mucositis is inflammation of the soft tissue, whose main sign is bleeding on gentle probing, often with an apparent increase in probing depth because the tissue is swollen. Peri-implantitis is the same inflammation having progressed into loss of the bone supporting the implant, with bleeding, deeper probing, sometimes recession of the margin, and bone loss visible on a radiograph.^[1]

The practical consequence is that the two need different treatment and only one of them has taken something that does not come back. Mucositis is arrested by removing the biofilm and keeping it removed. Peri-implantitis needs that too, and then a decision about the bone defect that has already formed.

Around a natural tooth, a probing depth has a normal range you can compare against. Around an implant it does not. The case definitions say so directly: it is not possible to define a range of probing depths compatible with health, and peri-implant health can exist around an implant that already has reduced bone support.^[1] What that leaves is change over time, which only exists as a measurement if somebody took the first one.

The same document recommends obtaining baseline radiographic and probing measurements once the implant-supported prosthesis is complete.^[1] If that was never done, the first job is to start the record now, because a year from today this examination becomes the baseline the next one is read against.

Biofilm is the aetiological factor, and the evidence on that is strong.^[1] It does not follow that the patient's cleaning is the variable at fault. An implant restoration that cannot be cleaned underneath, a margin buried deep in tissue, residual cement from the day it was fitted, or a component that has loosened all produce a site where biofilm accumulates faster than any brushing removes it. Several of those are correctable, and correcting them changes the outcome more than instruction does.

The aim is the same as in periodontal disease: stop the process, then hold it. Non-surgical treatment means disrupting the biofilm on the implant surface, which is harder than on a root because the surface is threaded and textured. A 2025 meta-analysis of six randomised trials found adjunctive antibiotics gave an additional 0.66 mm of probing depth reduction over mechanical treatment alone, with systemic administration performing better than local, while rating the evidence limited and heterogeneous.^[2]

What none of it recovers is the bone already lost. That is the reason the honest conversation is about arresting the process and keeping the implant, not about returning it to how it was. An implant with reduced but stable bone support, cleaned properly and checked on a schedule, can serve for a long time. The same implant left to progress cannot.