Diederik Hentenaar

145 General discussion and conclusions exerts further potential to predict/determine disease progression or to distinguish peri-implant mucositis from peri-implantitis remains to be established in longitudinal clinical studies. Our study is one of the first to compare levels of MMP-8 in peri-implantitis implants with healthy implants. Previously, MMP-8 was detected in peri-implant sites with ongoing bone loss (Arakawa et al. 2012). Classic pro-inflammatory cytokines IL-1β and TNF-α have shown to induce the synthesis and secretion of MMP-8 which in part might explain the elevated levels of MMP-8 found in our study (Siwik et al. 2000). Comparable to what is found in patients with periodontal disease, there seems moderate evidence in the literature showing upregulated levels of MMP-8 in PICF of implants with peri-implant disease (Salvi et al. 2012, Ghassib et al. 2019, Alassy et al. 2019). It therefore might be hypothesized that, in addition to IL-1β, MMP-8 could also serve a promising role to differentiate between peri-implant health and disease (Arakawa et al. 2012, Thierbach et al. 2016, Al-Majid et al. 2018, Alassy et al. 2019). Polymorphism in the promoter region of MMP-8 might be another reason for elevated levels of MMP-8 in PICF around implants with peri-implantitis. Polymorphisms might explain varied responses between different individuals with the same disease category (Ghassib et al. 2019). For individuals who are at high risk of peri-implantitis, this finding could be of great benefit considering that genetic polymorphisms are constant and can be measured before disease onset or implantation. Futures studies should however more often include this marker to further elucidate its contribution to the complex immune response around inflammated implants. No significant differences in quantitative outcomes for the other markers investigated were found in the PICF between both groups, suggesting no discriminating potential. A finding which seems in line with the current literature (Dursun et al. 2016, Duarte et al. 2016, Theodoridis et al. 2021). However, considering that the study in chapter 2 included a small sample size with a variety of implant brands, studies with greater sample size, evaluating patients with similar implant brands and designs are needed to confirm our outcomes. Moreover, considering the fact that peri-implant diseases present a cyclic evolution, the immunoinflammatory events responsible for tissue breakdown may not always be active in a cross-sectional studies design with a single moment of fluid collection. Therefore, future research should explore cytokine levels at multiple time points to evaluate the role of the different cytokines and their signaling pathways in different peri-implant conditions. In addition, when researching ‘new’ potential biomarkers it is recommended to perform recoverability experiments, before using a Luminex assay, to assure the optimal sampling method with matching elution protocol, per separate biomarker. 7