Dental Implant Placement in Patients With a History of Medications Related to Osteonecrosis of the Jaws: A Systematic Review

Successful placement and longevity of dental implants largely depends on achieving osseointegration during wound healing. Osseointegration is a dynamic process that requires normal functioning of inherent biological activities that occur during bone remodeling, specifically, the resorption of old bone by osteoclasts and the formation of new bone by osteoblasts.¹  The formation of new immature blood capillaries (angiogenesis) is essential in fueling these activities because bone cells, like all cells in the human body, require an adequate blood supply.²  Therefore, drugs that interfere with bone remodeling and angiogenesis may compromise osseointegration and result in premature implant loss.¹ 

Antiresorptives are a class of drugs known to affect bone homeostasis by inhibiting osteoclast differentiation and function. This effect supports their use in treating bone disorders characterised by excessive bone resorption such as osteoporosis and certain skeletal malignancies.³  A patient's quality of life is significantly improved with these drugs as they can prevent fractures and limit bone pain and metastatic spread.⁴  Today, there are 4 principal classes of antiresorptive drugs in use: bisphosphonates, selective estrogen receptor modulators (SERMs), calcitonin, and monoclonal antibodies such as denosumab.⁵ 

Antiangiogenics are a class of drugs used in cancer to restrict tumor vascularization.⁴  They are considered a novel and targeted approach in cancer treatment, relying on the concept that tumours cannot grow larger than 1–2 mm³ without generating their own blood supply.⁶  Most antiangiogenic drugs are monoclonal antibodies or small-molecule inhibitors that target the vascular endothelial growth factor (VEGF) pathway, as more than half of malignant tumours express high concentrations of VEGF. Examples of antiangiogenics in clinical practice include bevacizumab, pazopanib, and everolimus.⁷ 

A delayed wound healing condition associated with the use of antiresorptive and antiangiogenic drugs is known as medication-related osteonecrosis of the jaw (MRONJ). MRONJ is characterized by exposed bone or bone that can be probed through an intraoral or extraoral fistula in the maxillofacial region that has persisted for more than 8 weeks in patients with a history of treatment with antiresorptive or antiangiogenic drugs, with no history of radiation therapy to the jaws or obvious metastatic disease of the jaws.⁸  Although rare, the effects of MRONJ can be devastating, including secondary infection, swelling, painful lesions, various dysesthesias, and pathologic bone fracture.⁹  Some cases do not respond to any form of treatment, and there is no evidence to suggest that stopping drug therapy will aid in resolution of the lesion. Hyperbaric oxygen reportedly has minimal to no effect. Antibiotics cannot enter necrotic tissue, so they are only used to manage infection in adjacent tissues. The current recommendations involve palliative care or conservative treatment in symptomatic lesions.¹⁰  When surgery is indicated, large resections and complex reconstructions are often performed with limited success and often leave patients with rather notable facial deformities.¹¹  At present, the pathogenesis of MRONJ is poorly understood. Various etiopathogenic mechanisms under investigation include suppression of bone turnover, inhibition of angiogenesis, toxic effects on soft tissue cells, and infection. One of the strongest predisposing factors is dentoalveolar surgery; however, despite this, the risk of MRONJ after the placement of dental implants is currently unknown.⁴ 

With an increasing number of patients reporting a history of antiresorptive or antiangiogenic drug therapy, general and specialist dentists will be faced with the decision of whether it is safe to place dental implants in this patient group. Although not currently contraindicated for dental implant therapy, there are biologically plausible arguments that could be made to suggest a risk of implant failure and MRONJ development in these patients.¹²  Previous systematic reviews on this topic mainly focussed on bisphosphonates and the reported effect that these drugs had in relation to implant failure and MRONJ varied.¹³  Two systematic reviews were unsuccessful in their attempt to retrieve studies on denosumab.^3,14  Furthermore, to date, no systematic reviews have included studies of patients treated with antiangiogenic drugs. Therefore, the aim of the present systematic review was to systematically research the literature to address these deficits and answer the following focus question: “When compared to placing dental implants in healthy patients, are patients with a history of antiresorptive or antiangiogenic drug therapy at increased risk of implant failure and MRONJ?”

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement.¹⁵ 

Studies were required to meet strict inclusion criteria. These included the following: (1) English language; (2) randomized controlled trials, cohort studies, case-control studies, or case series; (3) retrospective, cross-sectional, or prospective design; (4) ≥5 patients with a history of antiresorptive or antiangiogenic drug therapy before implant placement; (5) clear reporting of sufficient relevant data worthy of discussion; and (6) full version available. Studies that did not meet the inclusion criteria were automatically excluded.

An electronic search was performed in PubMed, Cochrane Central Register of Controlled Trials, Scopus, and Web of Science databases. Furthermore, a search of the OpenGrey database was used to identify any unpublished studies (gray literature); the last search was July 16, 2019.

To increase the sensitivity of the search, the search strategy only included terms concerning the population and intervention. Furthermore, no date restriction was used. The full search string used in each of the databases can be found in the registered protocol on the PROSPERO database (see Note).

Two reviewers (JS, KKA) independently began the identification phase using the abovementioned search strategy. Citations of identified articles were exported into reference managing software (EndNote X8), and duplicates were removed. The screening phase was performed by the same reviewers where the titles and abstracts of all remaining studies were independently screened for studies that potentially meet the inclusion criteria. In the eligibility phase, the full-text version of all remaining studies was independently assessed by the same reviewers for eligibility into the included phase. Any disagreements over the eligibility of studies were resolved through discussion with a third reviewer (JDL).

A standardized, prepiloted Microsoft Excel spreadsheet was used to extract data from the included studies for evidence synthesis and assessment of study quality. Two reviewers (JS, KKA) extracted the data independently, and any discrepancies were resolved through discussion with a third reviewer (JDL). Where possible, missing data were requested from study authors.

The data collected from the included studies were tabulated in the following fields: (1) study design; (2) number of cases (patients with a history of antiresorptive or antiangiogenic drug therapy) and, when available, number of controls (patients without a history of antiresorptive or antiangiogenic drug therapy); (3) number of implants in cases and controls; (4) patient characteristics (systemic diseases/age/sex/smoking status); (5) details regarding antiresorptive or antiangiogenic drug history (type of drug, indication for intake, administration route, and intake before implant placement); (6) whether patients were taking the drug at the time of implant placement and implant follow-up; (7) reported outcome parameters (implant loss/failure/success/survival and incidence of MRONJ); and (8) reported outcome. Where data was missing, the term “not specified” was used.

Assessment of methodologic and reporting quality was conducted to establish the internal validity and risk of bias of studies that met the inclusion criteria.

The methodologic and reporting quality of the included cohort and case-control studies was assessed independently by 2 reviewers (JS, KKA) applying a modified Newcastle-Ottawa scale (NOS)¹⁶  as described by Stavropoulos et al.³  In the present systematic review, a percentage <50% was considered to indicate low quality, 50–70% was moderate quality, and >75% was high quality. Furthermore, for each specific item, the percentage of positive scored studies was calculated. Where there were disagreements between the 2 reviewers, a third reviewer (JDL) was involved.

The methodologic and reporting quality of the included case series was assessed independently by 2 reviewers (JS, KKA) applying the Joanna Briggs Institute (JBI) critical appraisal checklist for case series.¹⁷  In the present systematic review, a percentage <50% was considered to indicate low quality, 50–70% was moderate quality, and >75% was high quality. Furthermore, for each specific item, the percentage of positive scored studies was calculated. Where there were disagreements between the 2 reviewers, a third reviewer (JDL) was involved.

The combinations of search terms resulted in a total of 7542 titles. Of these, 1469 were found to be duplicates; as a result, 6073 references were reviewed. In turn, 6018 studies were excluded based on the evaluation of the title and abstract, leaving 55 studies to be assessed for eligibility. Of these, 26 studies were excluded for various reasons, and 29 studies met the inclusion criteria and were thus selected for inclusion in the present systematic review (Figure 1). There were no additional studies identified through cross-referencing or by contacting study authors of retrieved publications that met the inclusion criteria. Of the included studies, 28 reported on bisphosphonates,^18–45  and 1 reported on denosumab.⁴⁶  No studies reporting on SERMs, calcitonin, or antiangiogenics were identified.

Table 1a through c presents characteristics of the included studies. Table 2 presents a summary of the outcome measures from all included studies on bisphosphonate and denosumab intake. Where there was missing information required for the interpretation, estimations were calculated on a pro rata basis or by assuming the minimum number of implants placed/failed in cases and controls.

There were five cohort studies,^{21,27,28,44,45}  6 case-control studies,^{18,22,23,30,31,39}  and 17 case series,^{19,20,24–26,29,32–38,40–43}  reporting on bisphosphonate intake included in the present systematic review. Twenty studies were retrospective,^{19,22–24,27,28,30–38,40–42,44,45}  4 were cross-sectional,^20,21,25,26  and 4 were prospective.^18,29,39,43  Most of the studies were based only on information obtained from patient records. In 5 studies,^{24–26,34,35}  bisphosphonate intake before and after implant placement was reported; the cases in which implants were placed before initiating bisphosphonate treatment were excluded. Nisi et al³⁸  included 90 patients with MRONJ caused by various reasons; only the 9 cases of MRONJ caused by implant placement were included. French et al⁴⁵  evaluated several risk factors associated with marginal bone loss and prevalence of mucositis/peri-implantitis; only information pertaining to bisphosphonate therapy was used. Among studies, cases (patients with a history of bisphosphonate drug therapy) ranged from 6 to 235, whereas the number of implants placed in cases ranged from 14 to 1267 implants. Controls (patients without a history of bisphosphonate drug therapy) when present, ranged from 12 to 2026, whereas the number of implants placed in controls ranged from 28 to 4507 implants. Six studies did not specify the number of cases/controls on the patient or implant level.^{21,24,26,31,35,38}  The follow-up period after implant placement ranged from 0.3 to 12.2 years. Collectively, 20 studies provided information regarding implant loss/failure/success/survival,^{18–23,25–31,37,39,41–45}  and 22 studies provided information regarding the incidence of MRONJ.^{18–25,27,28,32–42,44}  More detailed single-patient data was able to be extracted from 6 studies.^{20–23,28,44} 

A single case series by Watts et al⁴⁶  reporting on denosumab intake was eligible to be included in the present systematic review. This study was retrospective and based on information obtained from a questionnaire. Information on invasive oral procedures and events (dental implants, tooth extraction, natural tooth loss, scaling/root planing, and jaw surgery) in long-term/crossover patients treated with denosumab was assessed, and details of positively adjudicated MRONJ cases were presented; only the patients who received dental implants were included. The number of cases (patients with a history of denosumab drug therapy) was 212, whereas the number of implants placed was not specified. The follow-up period after implant placement was also not specified. Some type of information was provided regarding the incidence of MRONJ. More detailed single-patient data were able to be extracted from this study.

Tables 3 and 4 present the quality assessment of the included cohort and case-control studies. Cohort studies received from 1 to 5 stars (14%–71%; low-moderate quality), whereas case-control studies received from 5 to 7 stars (63%–88%; moderate-high quality). For each item, the percentage of positive scored studies ranged from 0% to 100% for cohort studies and from 50% to 100% for case-control studies.

Table 5 presents the quality assessment, based on the JBI critical appraisal checklist for case series, of the included case series. Studies received from 5 to 10 yes answers (50%–100%; moderate-high quality). The percentage of yes answered case series per question ranged from 67% to 100%.

Oral bisphosphonates were prescribed for osteoporosis treatment in most studies; only 9 studies reported intravenous administration of bisphosphonates,^{24,32,34–36,38–41}  and 6 of these studies reported bisphosphonate administration related to malignancies.^{24,32,35,36,38,40}  The most frequently prescribed bisphosphonate for osteoporosis was oral alendronate; at times oral clodronate, ibandronate, or risedronate were prescribed as an alternative or with alendronate. There were some instances where intravenous ibandronate, pamidronate, or zoledronic acid was prescribed for osteoporosis alone or with oral alendronate. For all cases of malignancy, intravenous zoledronic acid was prescribed alone or with other intravenous bisphosphonates such as ibandronate or pamidronate. The number of years of bisphosphonate intake before implant placement ranged from approximately 0.25 to 20.3 years.

Other than Kasai et al,²²  where implant success rate for cases was 85.7%, and Yajima et al,⁴⁴  where implant survival rate was 88%, there were no substantial differences observed between cases/controls, with implant success rate ranging from 92.9% to 100% for cases and 95.5% to 100% for the controls. Implant losses were more likely to occur in the posterior maxillary region and shortly after placement. When including estimations because of missing data, the overall implant failure rate when combining the included studies was 2.8% for cases and 2.1% for controls. When cohort/case-control studies and case series were separated, the percentage of failed implants in cases did not differ substantially (3.1% and 2.5%, respectively).

Several studies reported no MRONJ in relation with implant placement, whereas 8 case series reported otherwise.^{24,32–36,38,40}  From these case series, more than 33 patients developed MRONJ in the mandible and 12 in the maxilla. Most of the MRONJ lesions were diagnosed in the posterior regions. In 28 patients, implant surgery was described as the trigger of MRONJ (4 cases attributed implant removal as the triggering factor), whereas in 26 patients, the trigger was considered the presence of the implant. Furthermore, in 9 patients, the trigger could not be identified, and in 4 studies (totaling 39 patients), a triggering factor was not specified.^32–34,36  Bisphosphonate intake was indicated for osteoporosis (or related conditions) in 49 patients and malignancy in 43 patients. Although Holzinger et al³⁴  did not specify an indication for bisphosphonates, cases of MRONJ developing with both oral and intravenous bisphosphonate use were investigated. The time frame between the start of bisphosphonate drug therapy and MRONJ development ranged from 1 to 223 months, whereas the time frame between when the implant was first placed and MRONJ developing ranged from 0 to 180 months. When including estimations because of missing data, the overall incidence of MRONJ when combining the included studies was 12.3%. However, when cohort/case-control studies and case series were separated, the percentage of MRONJ cases differed substantially (0% and 17.6%, respectively).

In the case series by Watts et al,⁴⁶  denosumab was prescribed for osteoporosis; 60 mg administered every 6 months. Of 212 patients receiving dental implants, there was only 1 case of MRONJ identified (0.5% incidence). This patient had 2 implants placed in the posterior maxilla with simultaneous tooth extractions and a sinus lift and subsequently developed MRONJ related to delayed osseointegration. However, the patient continued to receive denosumab (8 doses) while being successfully treated for MRONJ and managed to also retain the implants.

Osteoclast-mediated bone resorption plays an important role during osseointegration and peri-implant bone homeostasis.³  Because bisphosphonates and denosumab interfere with osteoclast function, it is reasonable to consider that these drugs may have a negative effect on implant success in terms of osseointegration. However, most of the studies included in this systematic review indicated that patients with a history of bisphosphonates for osteoporosis treatment are not at increased risk of implant failure in terms of osseointegration compared with patients without a history of such medications. With estimations made in studies that failed to specify the exact number of implants placed/failed in each patient and irrespective of study design, only 85 of 3074 implants placed in 930 patients with a history of bisphosphonates failed. Compared with 182 implants failing of 8605 implants placed in 3331 patients without a history of bisphosphonates (2.8% vs 2.1% implant failure rate, respectively), this corresponds to a 2.8% implant failure rate in 8.3% of cases vs a 2.1% implant failure rate in 4.1% of controls. When assessing the numbers from cohort/case-control studies and case series separately, the implant failure rate did not differ substantially (3.1% and 2.5%, respectively). A recent systematic review evaluating the survival of dental implants in healthy patients found a 5.4% failure rate over an average follow-up of 13.4 years.⁴⁷  This observed failure rate is similar to that observed in patients who had used bisphosphonates in the present systematic review; therefore, it seems that implants placed in patients with a history of bisphosphonate use are not at increased risk of failing. Although most patients in the implant failure studies were taking oral bisphosphonates, one study by Siebert et al³⁹  examined implant survival in patients with osteoporosis receiving yearly infusions of intravenous zoledronic acid (5 mg). In this study, the implant success rate was found to be 100% in both groups (intravenous zoledronic acid and controls). Another study by Khoury and Hidajat⁴¹  investigated implant loss in patients with osteoporosis receiving both oral and intravenous (ibandronate) and found that, of 71 implants placed in 15 patients, only 1 immediately loaded implant failed after 5 months, and it was successfully replaced. In contrast, 2 studies by Kasai et al²²  and Yajima et al⁴⁴  cast doubt on this concept and reported that there were substantial differences in success and survival rates of dental implants between cases and controls (85.7% and 88%, respectively). In general, it seemed that there was a larger amount of early implant failures reported in bisphosphonate patients and more often in the posterior maxilla. However similar patterns to this have been observed in the general population.⁴⁸  Unfortunately, a risk assessment of implant failure in patients with a history of bisphosphonates for cancer or those with a history of denosumab could not be performed because there was insufficient data available from the studies included.

The concerns for an increased risk of implant failure in bisphosphonate users should also be seen in light of MRONJ. As mentioned previously, one of the strongest predisposing factors for MRONJ appears to be dentoalveolar surgery; therefore, it is likely that implant surgery may trigger the development of MRONJ. Furthermore, infection has long been considered an important component of MRONJ development as bone in these cases may be more vulnerable to infection because of decreased remodeling. Studies have identified a complex multiorganism biofilm consisting of bacteria (especially Actinomyces species), fungi, and viruses in biopsied specimens of necrotic bone removed in patients with MRONJ.⁸  Therefore, the presence of an implant itself may place a patient at risk of developing MRONJ. Results from this systematic review suggest there is a risk of MRONJ developing after implant placement in patients with a history of bisphosphonates, regardless of whether taken for osteoporosis or malignancy. With some estimations made in studies that failed to specify the exact number of implants placed in each patient and irrespective of study design, of a total of 830 patients exposed to bisphosphonates who underwent placement of ≥1 dental implant (2841 implants in total), 102 cases of MRONJ were diagnosed, and the approximated incidence was 12.3%. When cohort/case-control studies and case series were separated, the percentage of MRONJ cases differed substantially (0% and 17.6%, respectively). This is simply explained by the fact that all of the MRONJ cases were identified in 8 case series.^{24,32–36,38,40}  Estimates for developing MRONJ after tooth extraction among osteoporosis and cancer patients exposed to bisphosphonates ranges from 0.5 to 14.8%.⁸  Therefore, it is not surprising how a similar risk estimate for MRONJ after implant placement was found. In terms of localization, almost 3 times as many patients exposed to bisphosphonates developed MRONJ in the mandible than in the maxilla. Although these results are consistent with those reported in the literature,⁴⁹  the reason for MRONJ preferentially affecting the mandible in bisphosphonate users remains unknown. Furthermore, both implant surgery-triggered and implant presence-triggered cases of MRONJ were identified, suggesting that not only can the surgical trauma from implant placement or removal predispose to MRONJ but so can the mere presence of an implant in the oral cavity. Implant presence-triggered MRONJ could, therefore, develop in people who have dental implants placed before bisphosphonate drug treatment is initiated. This supports the potential role of infection in the development of MRONJ and the importance of long-term follow-up in patients with dental implants who later start taking bisphosphonates.

On the other hand, patients treated with denosumab for osteoporosis appear to be at a much lower risk of developing MRONJ after implant placement. In the case series by Watts et al,⁴⁶  there was only 1 case of MRONJ identified out of 212 patients receiving dental implants, suggesting a risk estimate of 0.5% patients treated with denosumab. Furthermore, the patient was fortunate enough to maintain the 2 implants while continuing denosumab treatment and being successfully treated for MRONJ. Because this study only included patients treated for osteoporosis, the risk of MRONJ in cancer patients treated with denosumab receiving dental implants could not be determined. To date, there has only been 1 other case report of MRONJ developing around a dental implant in a patient treated with denosumab for osteoporosis.⁵⁰  However, in this case, the patient had a 15-year history of bisphosphonates (alendronate) before swapping to denosumab treatment. Bisphosphonates can persist in skeletal tissue for significant periods of time, with alendronate having a half-life in bone of around 10 years. Therefore, MRONJ in this case may have been from a combination of bisphosphonate and denosumab treatment.

To assess the quality of the included cohort and case-control studies in the present systematic review, a modified NOS was used.^3,16  For the included case series, the JBI critical appraisal checklist for case series was used.¹⁷  In general, most of the included cohort studies were of moderate quality (the study of Grant et al²¹  was of low quality), most of the included case-control studies were of moderate quality (the study of Yip et al³¹  was of high quality), and most of the included case series were of high quality (studies of Giovannacci et al⁴⁰  and Khoury and Hidajat⁴¹  were of moderate quality). Of note, there were no cohort studies that reported long enough follow-ups to receive a star according to the modified NOS. Otherwise, all quality reporting items were met by more than half of the studies per item, suggesting an overall moderate quality of evidence across the studies.

Although no studies reporting antiangiogenics that met the inclusion criteria were identified, a recent case report found that MRONJ can develop around dental implants in these patients.⁵¹  In this case, the patient was being treated for renal cell carcinoma with pazopanib. After 6 months of treatment, the patient changed medications to everolimus. Seven weeks later, bone exposure was observed in both mandibular posterior regions surrounding the 35, 37, 46, and 47 implants, which were placed 6 years prior. This case highlights the potential for implant-presence triggered MRONJ in these patients, which may contraindicate their placement in the first instance.

There were several limitations in the present systematic review that should be discussed. First is the small number of studies and limited information available in the literature regarding bisphosphonates and denosumab and the lack of studies reporting on SERMs, calcitonin, and antiangiogenics. In fact, no single study available reported all the relevant data described previously, and therefore, inclusion criteria (5) was eased to allow for studies that reported enough relevant data worthy of discussion, as opposed to all relevant data. Including studies with missing data negatively affected the validity of our results as missing data meant that (1) some estimations were required for the summary calculations (Table 2) and (2) controlling for covariables such as systemic disease, age, sex, and smoking was not possible. A second limitation of this systematic review is that there were no randomized controlled trials (RCTs) available to be included. Unfortunately, only observational studies, which are considered a lower level of evidence than RCTs, were available. The absence of any RCTs means that the conclusions are based on rather limited evidence, and because most of the available studies were case series without control groups to compare the outcomes, the results lack statistical validity. Furthermore, there was only a single case series identified for denosumab, so generalizations about its lack of impact on MRONJ risk after implant placement cannot truly be drawn. In terms of study quality, most of the cohort and case-control studies were of moderate quality and despite most of the case series being of high quality, these are considered as having one of the lowest levels of evidence of all the clinical study designs. An overall moderate quality found across the studies indicates a decreased internal validity and increased risk of bias, so the conclusions herein should be interpreted with caution. The last limitation of this systematic review is that the criteria for implant success or failure varied slightly between the studies. For example, Jeffcoat¹⁸  defined implant success as “<2mm of alveolar bone loss over the three-year study period, lack of mobility, lack of infection and absence of pain, and osteonecrosis of the jaws,” whereas in Zahid et al,²⁸  the criteria for implant success was “clinical and radiographic evidence of osseointegration and bone loss <0.2 mm annually after the first year of service.” Ideally, all the studies would have followed the same criteria, allowing for greater uniformity in interpretation and discussion.

The results of the present systematic review suggest the following:

• There is a lack of data available in the literature regarding the risk of implant failure or MRONJ in patients with a history of SERMs, calcitonin, or antiangiogenics.

• Patients with a history of bisphosphonate treatment for osteoporosis are not at increased risk of implant failure compared with that of healthy patients.

• There is a lack of data available in the literature regarding the risk of implant failure in patients with a history of bisphosphonate treatment for cancer or patients with a history of denosumab.

• Patients with a history of bisphosphonate treatment are at risk of developing MRONJ after implant placement.

• Patients treated with denosumab for osteoporosis have a negligible risk of developing MRONJ after implant placement.

• There is a lack of data available in the literature regarding the risk of developing MRONJ after implant placement in cancer patients treated with denosumab.

In conclusion, the current literature still leaves gray areas in terms of the safety of placing dental implants in patients with a history of antiresorptive or antiangiogenic drug therapy. Until new studies of a higher quality become available, general and specialist dentists should carefully select patients with due consideration for medications. Importantly, all patients with a history of bisphosphonates are at risk of MRONJ, necessitating this to be included in the informed consent obtained before implant placement. Further researchers in this area should consider conducting RCTs involving patients with a history of bisphosphonate and denosumab drug therapy, as well as similar observational studies looking at patients with a history of other antiresorptives and antiangiogenics. A further systematic review on this topic is required once additional studies become available.

Abbreviations

JBI:

Joanna Briggs Institute

MRONJ:

medication-related osteonecrosis of the jaw

NOS:

Newcastle-Ottawa scale

PRISMA:

Preferred Reporting Items for Systematic Reviews and Meta-analyses

RCT:

randomized controlled trial

SERM:

selective estrogen receptor modulators

VEGF:

vascular endothelial growth factor

The authors acknowledge the James Cook University College of Medicine and Dentistry Honours program and the Australian Dental Research Foundation Colin Cormie Grant for funding this systematic review.

The authors declare no conflicts of interest. The protocol is registered in the International Prospective Register of Systematic Reviews (PROSPERO) database (PROSPERO registration number: CRD42019125619).