3.1 Aim 1

3.2 Aim 2

3.3 Publications excluded from the qualitative analysis (Aim 2) of the scoping review

Two excluded publications consisted of a systematic review (Aghaloo et al., 2019), and a systematic review with meta-analysis (Chappuis et al., 2018), that referred to articles that were separately described in the present analysis. The primary aim of the systematic review by Aghaloo et al. was to review the effects of systemic diseases and medications on implant osseointegration. The primary aim for the systematic review and meta-analysis by Chappuis et al. was to review medication-related dental implant failure.

A summary of a systematic review and meta-analysis (Cheng, 2020) and an abstract (Chappuis, 2019) on the impact of medications on dental implant failure, also were excluded from the present review as those publications only summarized the review by Chappuis et al. (2018) that was itself excluded as described above.

A systematic review and meta-analysis (Vinnakota & Kamatham, 2020) were excluded from the present study as the authors included three retrospective studies (Altay et al., 2019; Chrcanovic et al., 2017; Wu et al., 2017) that were individually critiqued in the present study.

A review by Mahri et al. (2021) was excluded from the present analysis. The primary aim of that paper was to systematically map osseointegration pharmacology using artificial intelligence. The authors identified three articles that associated PPIs with a high risk of implant failure, two of which were already individually considered in the present analysis. The third study did not meet the inclusion criteria as it measured the effect of postoperative systemic administration of omeprazole on bone healing and implant osseointegration in rat tibiae versus alveolar bone (Al Subaie et al., 2016).

A review article (Surabhi, 2019) describing the effect of medications on dental prostheses was excluded. The primary aim of that study was to review the association between multiple systemic medications and dental implants. The author did not indicate which component studies were considered to conclude that PPIs alter the success of implant therapy. In fact, there was no mention of any study that was reviewed to make any conclusions.

A narrative review (Mester et al., 2019) describing the impact of proton pump inhibitors on bone regeneration and implant osseointegration also was excluded from further analysis. The primary aim of that study was to assess the effect of PPIs on dental and orthopedic implants. The orthopedic implant component did not meet the inclusion criteria of the present review, and the dental implant component included two retrospective cohort studies that already met the study criteria and were individually evaluated in Aim 1 and 2 (Chrcanovic et al., 2017; Wu et al., 2017).

A consensus report by Jung et al. (2018), describing the influence of implant length, design, and medications on clinical and patient-reported implant outcomes also was excluded. In that report, there was only one reference to the effect of medications on dental implants, the systematic review and meta-analysis by Chappuis et al. (2018), previously discussed, in which PPIs were included among the medications that were considered. Consequently, the report by Jung et al. was excluded from the qualitative analysis since the component studies were individually analyzed in the present review.

An invited commentary (Tamimi & Wu, 2017) on osseointegration pharmacology was excluded. The authors discussed four groups of drugs that could affect osseointegration. One of the studies referenced in the commentary did not meet the inclusion criteria of the present review because the authors measured only the effect of limited postoperative systemic administration of omeprazole on bone healing and implant osseointegration in rat tibiae, not in alveolar bone (Al Subaie et al., 2016). Another study included in that commentary already was considered in the present study (Wu et al., 2017). Consequently, the invited commentary was excluded, and the component study that met the inclusion criteria was separately considered. Those authors also produced a thesis (Wu, 2016) that met the inclusion criteria of the present scoping review. However, the thesis subsequently was excluded from further qualitative analysis since the section pertinent to PPIs and periodontal tissue attachment was later published (Wu et al., 2017) and separately included in our review.

An abstract by Nag (2018), which summarized a scoping review on the influence of PPIs on dental implant failure, was excluded from Aim 2. The extent to which the publication addressed the inclusion and exclusion criteria of this study could not be independently ascertained, and the authors did not reference the component studies that were included. A full-length follow-up manuscript never was published.

An abstract of the 2021 review by Mahri and Tamimi (2019) that was previously discussed in the current review, was excluded from the present study. In this abstract, the authors evaluated 247 publications. The primary objective was to assess the effect of various medications on bone-implant osseointegration, with PPIs as one of those drugs. The abstract did not describe the studies or various medications that were evaluated in their review, making it unclear if it met the inclusion criteria of the present study.

A 2021 thesis (Chawla & Yerke, 2021) that was comprised of two subsequently submitted manuscripts was excluded from the qualitative analysis: one manuscript was a preliminary version of this scoping review. The second manuscript was a cross-sectional study published after completion of the thesis (Chawla et al., 2022), and is critiqued separately in the qualitative analysis, below.

3.4 Publications included in the qualitative analysis of the scoping review

Wu et al. (2017), performed a retrospective cohort study that included a total of 1773 osseointegrated dental implants in 799 patients (133 implants in 58 PPI users and 1640 in 741 nonusers) who were treated between January 2007 and September 2015. The failure rate among PPI users was 6.8%, and 3.2% in nonusers, supporting the authors' conclusions that PPI use is associated with a higher rate of dental implant failure. The strengths of this study included large sample size, the use of implants obtained from the same manufacturer, and all procedures or surgeries performed by a single clinician. Limitations included the retrospective nature of their study and the lack of statistical analysis of the differences between PPI users and nonusers at the patient level. There also was a statistically significant difference between the number of PPI users taking NSAIDs compared with non-users. Since PPIs have been used in co-therapy to prevent the occurrence of NSAIDs-induced peptic ulcers, and NSAIDs have been shown to have a protective effect on bone when used in combination with PPIs, concurrent NSAID use might have had a confounding influence on bone around dental implants among PPI users. Since the outcome measurement of this study was dental implant failure, the direct incremental effect of PPI on bone loss was not determined. Within the noted limitations, the authors found a relationship between PPIs and dental implant failure.

Chrcanovic et al. (2017), in a retrospective cohort study, studied patients treated with implant-supported/retained prostheses at a private specialty practice who were or were not taking PPIs. A total of 3559 implants were placed in 999 patients, with 178 implants reported as failures. The implant failure rates were 12% (30/250) for PPI users and 4.5% (148/3,309) for nonusers. The authors concluded that the use of PPIs might be associated with a statistically significant negative effect on the implant survival rate. Although a wide time range resulted in correspondingly large sample size, the variability caused by various clinicians and implant systems used over the course of the study was a study limitation. Additionally, the specific criteria for implant failure were not defined. The outcome of this study was implant failure, so the mechanism through which PPIs affected bone loss could not be identified. That outcome could be observed only when sufficient bone loss had occurred, resulting in loss of osseointegration. Finally, the authors did not perform statistical analysis on differences between PPI users and nonusers at the patient level.

Altay et al. (2019), in a retrospective cohort study involving 1918 dental implants in 592 patients (69 implants in 24 PPI users and 1849 implants in 568 nonusers) evaluated the effect of PPI intake on the osseointegration of dental implants using patient- and implant-level models. There were statistically significant differences between PPI users and nonusers at the implant level, but the authors failed to show any significance at the patient level. The implant failure rates were 4.60 times greater among PPI users versus nonusers. The authors concluded that PPI use might be associated with an increased risk of early dental implant failure. This appeared to be a well-designed human study that investigated the effects of PPI intake on osseointegration failure at both patient and implant levels. One limitation included a relatively small sample size (24 PPI patients). Second, PPI users received a significantly greater number of implants in the maxilla than in the mandible, which might have influenced the outcome. However, the focus of this study was early dental implant failure, and any direct effect of PPI on bone levels was not determined. The effect of PPIs could be seen only when bone loss exceeded the threshold required to maintain osseointegration, resulting in implant failure.

Ursomanno et al. (2020), in a retrospective record review, assessed the medical and dental histories and radiographic data of 635 patients receiving 1,480 dental implants from 2010 to 2017. The authors found a significantly greater number of exposed threads from the PPI patient group than from PPI nonusers. Patients taking PPIs experienced a 5.5% implant failure rate, while patients not taking such medications had a 2.0% implant failure rate. The strengths of this study included a large patient population with appropriate controls and statistical analysis, with the same examiner responsible for all study-related radiographic measurements blinded to the patient PPI status. That study was unique since bone loss around implants was quantified by measuring bone loss in millimeters, as well as through quantitation of the number of exposed implant threads, which might serve as a more sensitive assessment of the effect of PPI on implant-bone level, compared to implant failure. Limitations included the retrospective nature of the study and lack of information on dose and length of time each patient was taking PPIs.

A cross-sectional study by Romandini et al. (2021), reported on the prevalence, risk factors, and protective indicators of peri-implant disease. Ninety-nine patients with 458 dental implants placed between September 2000 to July 2017 at a university clinic were included. Risk/protective indicators were tested individually by using peri-implantitis as the dependent variable for each potential indicator to test for significance. Using uni- and multivariant logistic regression analyses, the authors found peri-implantitis to be inversely associated with use of PPIs, one of two studies in this scoping review to report an inverse association between PPIs and implant failure. Although a statistically significant association was found between PPIs and peri-implantitis, the number of implant failures and the number of patients using PPIs was not reported. Furthermore, that report included only implants that had been successfully placed and loaded for a minimum of 1 year. Since early implant failures often occur due to unsuccessful osseointegration or impaired bone healing within the first year if placement (Mohajerani et al., 2017), many implant failures might not have been included in their analysis. This could explain why PPIs were identified as protective indicators for peri-implant disease.

An abstract by Yerke et al. (2019), described patients with generalized, chronic, moderate-to-severe periodontitis using, or not using, PPIs. They found an inverse relationship between the use of PPI and the severity of periodontal disease. The use of PPIs was associated with a decreased proportion of elevated periodontal pocket depths, implying less severe periodontal disease. Strengths included a relatively large sample size (581 patients) with appropriate controls and appropriate statistical analyses.

An abstract by Yerke and Cohen (2019) reported the impact of PPIs on bone levels around teeth in patients with moderate-to-severe periodontitis. The authors evaluated patients after adjusting for systemic factors and habits, such as inflammatory bowel disease, smoking, diabetes, use of systemic steroids, peri-menopausal hormone replacement therapy, hypothyroidism, and another autoimmune disease. Patients using PPIs were associated with reduced percentages of teeth with elevated periodontal probing depths when compared to non-users. The authors used a comparatively large sample size (518 patients) with controls and statistical analyses. This study expanded upon the previously published abstract to show that a relationship existed between PPIs and decreased probing depths among patients having or not having the potentially confounding conditions noted above. Consequently, the studies described in both abstracts analyzed different subsets of patients that were obtained from the same patient population.

A 2022 cross-sectional study retrospectively analyzed patients from a university-based faculty periodontics practice (Chawla et al., 2022). The proportion of probing depths ≥ 5 mm or ≥ 6 mm was measured for patients diagnosed with generalized periodontitis, stage III to IV, grade B to C, among PPI users compared to non-users. A 31% decrease in probing depths ≥ 5 mm, and a 42% decrease in probing depths ≥ 6 mm, was noted among patients taking PPIs. Those decreases in probing depths remained statistically significant even after excluding smokers and diabetics, as well as patients with systemic disease and taking potentially confounding medications, such as prednisone. In addition, there were no statistically significant differences in oral hygiene between groups, suggesting that plaque control was not a confounding factor. This study had a relatively large sample size (N = 1093) and used two probing depth minimum thresholds to assess periodontal severity.

An abstract from the same research group used a similar protocol to evaluate the association between severity of periodontal disease and PPI use among patients seeking care at a private periodontal practice from 2016 to 2019 (Yerke et al., 2021). In a comparatively large population (N = 1513), the prevalence of probing depths ≥ 5 mm was 28.1% in PPI users, versus 55.8% in nonusers (95% confidence interval = 14.2%–41.5%, p < .001). A similar decrease in the prevalence of probing depths ≥6 mm was observed among PPI users. No differences in plaque control were evident between the PPI and non-PPI groups.

The proportion of probing depths ≥ 6 mm also was used to assess periodontal severity in another abstract from this group (N = 1017) (Herrmann et al., 2022b). The study population included patients seeking care from a university postgraduate periodontics program during 2010–2017. Statistically significant decreases in the proportion of probing depths ≥6mm among non-PPI users again were noted, with no difference between plaque control between PPI users and nonusers. The studies from this study group were retrospective and therefore not able to account for the dose or duration of PPI use. However, the consistent decrease of periodontal disease severity among PPI users in three different populations suggests that this association is not spurious.

Ursomanno et al. (2019), examined patients receiving dental implants at a university clinic between 2010 and 2017. In an abstract, the authors reported the extent of radiographic bone loss around implants and enumerated the number of exposed threads secondary to crestal bone loss among patients using, or not using, PPIs. The authors concluded that the use of PPIs was associated with greater bone loss around dental implants. Similar results were found after adjusting for confounding factors. A comparatively large sample size (655 patients; 1480 implants) with appropriate controls and statistical analyses were used.

Rogoszinski et al. (2020) evaluated dental implants placed in patients at an unspecified practice setting between 2006 and 2013. The authors evaluated radiographic bone loss, bleeding on probing, suppuration, and periodontal probing depth among patients using, or not using, PPIs. In this study, only dental implants that survived for a minimum of 5 years were considered, yielding 881 implants in 284 patients that were evaluated. In that study, the authors found a 29.7% decreased risk of peri-implantitis among patients taking PPI medication, which was statistically significant. This publication and Romandini et al. (2021) are the only studies that suggested a protective role for PPIs on bone and tissue attachment levels around dental implants. In any event, one important limitation of the Rogoszinski et al. study was that implants that failed to integrate, or were lost before the five-year minimum survival criterion, were omitted from their analysis.

A recent study by Rogoszinski et al. (2022) evaluated dental implants at the Philadelphia Veterans Affairs Medical Center during the same time period as his previous study, 2006 to 2013. They evaluated radiographic bone loss, bleeding on probing, suppuration, and periodontal probing depth among those using, or not using, PPIs. Only implants that had a minimum of 5 years of follow-up were included. The authors assessed the effect of potential confounders, such as smoking, diabetes, age, sex, smoking, implant location, implant type, and whether the site received prior bone grafting. Univariate analysis identified statistically significant associations with confounders, such as smoking, diabetes, and age, with long-term implant failure and peri-implantitis. When multivariate analysis was performed, PPI use was not associated with long-term implant failure or peri-implantitis. Limitations included the omission of implants lost before 5 years of follow-up and a lack of information regarding the dose and duration of PPIs. However, this study did require a 5-year minimum duration of PPI use and required initiation of PPI medication to have occurred before implant placement. A disadvantage is the lack of independence within the implant data: 933 implants were placed in 284 individuals, implying that most individuals had multiple implants that were analyzed separately. Oral hygiene as a risk factor was not assessed.

4.1 Possible mechanisms affecting tissue attachment around teeth and dental implants

There are a number of explanations describing how PPIs might influence periodontal and peri-implant tissues. PPI use has been linked to changes in the gut microbiota (Jackson et al., 2016). Freedberg et al. (2015) found a greater than 10-fold increase in Streptococcaceae in the upper GI tract. An increase in gastric pH can result in a significant increase in oral bacteria, such as Peptostreptococcus stomatitis, Streptococcus anginosus, Parvimonas micra, Slackia exigua, and Dialister pneumosintes in the GI tract (Bruno et al., 2019). Mishiro et al. (2018) found a significant difference in salivary bacteria when comparing PPI users versus nonusers. As a result, it is conceivable that a shift in microbial diversity with PPI use can affect the number and distribution of periodontopathic organisms, which also might affect the extent or severity of periodontal disease, and thereby affect attachment levels around teeth.

PPIs also have been shown to increase the rate of orthodontic tooth movement, suggesting that an alteration in bone metabolism occurs that can be characterized as an increase in bone resorption relative to bone formation (Makrygiannakis et al., 2018). Similarly, Shirazi et al. (2014), using a rat model system, found a significant increase in optical density of the mandibular bone apical to the first molar after the sixth week of pantoprazole administration, suggesting the development of bone density alterations leading to an increased rate of orthodontic tooth movement, possibly due to decreased calcium absorption.

PPIs affect bone metabolism by impairing intestinal calcium absorption, leading to decreased calcium availability in circulation (Wu et al., 2017). Schinke et al. demonstrated that impaired gastric acidification negatively affects calcium homeostasis, triggering either hyperparathyroidism-induced bone loss or, when combined with diseases involving osteoclast dysfunction, impaired skeletal mineralization (Schinke et al., 2009). Osteoclasts, like gastric parietal cells, contain acidic vesicles that can be inhibited by PPIs. In vitro studies show that omeprazole inhibits the vacuolar H⁺-ATPase (V-ATPase) in bone-derived membrane vesicles, although at higher levels than those needed for inhibition of gastric H⁺/K⁺ ATPase (Costa-Rodrigues et al., 2013). The V-ATPase proton pump is located in the bone-apposed osteoclast plasma membrane, and proton pump inhibition prevents the secretion of hydrogen ions, decreasing their bone resorption ability (Costa-Rodrigues et al., 2013). However, PPIs might be a desirable situation in the context of bone tissue disorders, such as periodontal disease, where a decrease in osteoclastic activity might result in a significant improvement in overall bone strength and integrity (Costa-Rodrigues et al., 2013). That possibility was proposed by Yerke et al. (2019) who found an inverse relationship between use of PPI and severity of periodontal disease.

Some studies have reported a two- to four-fold increased risk of B₁₂ deficiency associated with PPI therapy (Freedberg et al., 2017). By altering intragastric pH levels, PPIs decrease the absorption of vitamin B₁₂. Vitamin B₁₂ is an essential protein-bound nutrient where the presence of gastric acid is required for pancreatic proteases to cleave vitamin B₁₂ from protein, allowing its reassociation with intrinsic factor and eventual absorption in the terminal ileum (Ito & Jensen, 2010). Short-term studies have suggested that acid suppressants such as PPIs decrease the absorption of vitamin B₁₂ (Makrygiannakis et al., 2018). A few cross-sectional studies have shown low serum vitamin B₁₂ levels to be associated with decreased levels of markers of bone formation (Stone et al., 2004). Therefore, it is possible that PPIs might have a potential negative effect on tissue attachment around teeth or dental implants by decreasing B₁₂ availability.

Chronic iron deficiency has been suggested to negatively affect bone (Toxqui & Vaquero, 2015). Dietary iron is present in food as either nonheme (66%) or heme iron (32%), with absorption of nonheme iron being markedly improved by gastric acid. Gastric acid facilitates the dissociation of nutrients from iron salts, which allows the formation of complexes with ascorbate, sugars, and amines. Numerous clinical conditions associated with achlorhydria/hypochlorhydria have been shown to be associated with decreased iron absorption and/or iron-deficiency anemia (Ito & Jensen, 2010). Sarzynski et al. (2011) in a retrospective cohort study, found significantly decreased hematologic indices, including hemoglobin amongst PPI users when compared with matched controls (Sarzynski et al., 2011). Therefore, PPI-induced iron deficiency might have a deleterious effect on bone and tissue attachment around teeth or dental implants.

Increasing evidence suggests that PPIs might disrupt skeletal integrity through PPI-induced hypochlorhydria, which impairs calcium absorption and reduces calcium bioavailability for incorporation into bone, resulting in compensatory and potentially chronic hyperparathyroidism that increases bone turnover (Hinson et al., 2015). Hinson et al. (2015) in a retrospective chart review of individuals 60 years and older, found statistically significant and pathologically higher PTH levels among chronic PPI users. The persistent elevation of PTH levels might lead to loss of bone strength and quality. Collectively, those findings might provide additional insight regarding the mechanism through which PPIs influence bone metabolism around teeth or implants.

The present results also have applicability to clinical practice. Alternative drugs to PPIs might be explored for patients at a higher risk for dental implant failure (Wu et al., 2017). Histamine 2-receptor blocker or over-the-counter antacids could be considered before prescribing the most potent available treatment (Benmassaoud et al., 2016). Mitigation of potential PPI risks could be attempted by PPI reduction or by giving risk-specific supplements (Freedberg et al., 2017). However, the literature regarding the use of supplements to ameliorate potential PPI risks also is limited (Freedberg et al., 2017). Probiotics show a modest benefit in preventing antibiotic-associated diarrhea but have not been tested to prevent infections in long-term users of PPIs (Freedberg et al., 2017). Supplementation of calcium and vitamin D does not conclusively decrease risk for fracture (Freedberg et al., 2017). Indeed, it has been recommended that patients taking PPIs should not use them for indefinite period, but rather attempt to taper once symptoms are controlled (Benmassaoud et al., 2016). Since there appears to be an association between PPI use and implant failure, dentists might consider PPIs to be a risk factor when planning an implant surgery for their patients. Consequently, clinicians should consider whether more frequent maintenance visits and home care reinforcements might be indicated for implant patients taking PPIs.

On the other hand, preliminary results suggest that PPIs appear to be beneficial for tissue attachment around teeth and could potentially be prescribed short-term, for patients at a risk for periodontitis. However, further well-controlled prospective studies will be required to corroborate the relationship between PPIs and tissue attachment around teeth before making treatment recommendations regarding adjunctive use of PPI in periodontal therapy. The precise mechanisms through which PPIs affect the supporting structures around teeth or dental implants also remain to be fully elucidated.