Prevalence of peri‐implantitis in a sample of HIV‐positive patients

Abstract Objectives This study aimed to assess the prevalence of peri‐implantitis in human immunodeficiency virus (HIV)‐positive patients and the presence of a possible correlation between the immunological profile and serological values, of peri‐implantitis, and of possible differences between all‐on‐4 and single crown/bridge prostheses. Subjects and methods This retrospective study included 58 adult HIV‐positive patients (222 implants) with either all‐in‐4 prostheses or single crowns/bridges on at least one dental implant loaded for more than a year who were followed for 3 year (mean follow‐up). Data pertaining to the probing pocket depth (PPD), bleeding on probing, and immunological and systemic profile were collected. Results Patients with single crown/bridge implant rehabilitation showed higher prevalence of peri‐implantitis (34%) than patients with all‐on‐4 rehabilitation (0%) (p = 0.012). Patients with all‐on‐4 rehabilitation were significantly older than those with single crowns/bridges (p = 0.004). Patients with peri‐implantitis had implants for a significantly longer duration than those without (p = 0.001), implying that the probability of peri‐implantitis increases as the age of implant increases. Conclusions The prevalence of peri‐implantitis was 26% in the HIV‐positive patients population. No correlation was found between patients' immunological and serological factors and peri‐implantitis. The most important risk factor for peri‐implantitis and mucositis was implant age.


| INTRODUCTION
Dental implants are being increasingly used for the replacement of missing teeth and have a high survival rate (> 10 years) . Despite their long-term success rate, dental implants are subject to biological complications characterized by inflammation of the soft tissues and bone in contact with the implant and its prosthetic components, caused primarily by the accumulation of bacterial plaque (Costa et al., 2012;Vignoletti et al., 2019), and bearing a pathogenic mechanism similar to periodontitis (Heitz- Mayfield & Lang, 2010). These biological complications are classified as peri-implant mucositis and peri-implantitis (Lang et al., 2011;Mombelli et al., 2012;Araujo et al., 2018, Renvert et al., 2014. Peri-implant mucositis is inflammation of the mucosa around the implant and is not associated with the loss of supporting bone. In contrast, peri-implantitis is characterized by inflammation of the periimplant mucosa associated with a progressive loss of supporting bone (Heitz-Mayfield & Salvi, 2018;Lindhe et al., 2008;Schwarz et al., 2018;Zitzmann & Berglundh, 2008).
Systemic conditions, other than diabetes, correlated with periimplantitis include cardiovascular diseases, rheumatoid arthritis, osteoporosis, osteopenia, thyroid disease, hepatitis, radiation, and/or chemotherapy (Dalago et al., 2017;de Araujo Nobre et al., 2015;Dvorak et al., 2011;Marrone et al., 2013). However, these diseases have not been proved to be correlated to peri-implantitis, and no cause-effect relationship has been found, as has been for diabetes mellitus (de Araújo Nobre & Mal o, 2017).
The presence of inflammation in peri-implant tissues is a common factor between peri-implantitis and mucositis. Hence, determination of bone loss through radiographs is essential for distinguishing between the two Schwarz et al., 2018).
However, functional bone loss between 0.5 and 2 mm within 1 year from the loading of implants is considered physiological, and occurs as a result of the natural healing process (Lindquist et al., 1996;Mombelli et al., 1987). Any other radiographic evidence of bone loss should be considered pathological .
The progressive loss of bone around dental implants can be quantified by comparing a current radiograph with one obtained at least 1 year after implant loading. In absence of initial radiographs, diagnosis of peri-implantitis is based on: bone levels > 3 mm apical to the most coronal portion of the intra-osseous part of the implant and/or probing depths of >6 mm (Fransson et al., 2005;Koldsland et al., 2010;Papantonopoulos et al., 2015;Renvert et al., 2018).
Dental implants have been widely used even in human immunodeficiency virus (HIV)-positive patients with results similar to healthy subjects, owing to the introduction of the highly active antiretroviral therapy, which has extended the average life expectancy of such patients (Capparé et al., 2019;Dios et al., 1999;Gherlone et al., 2016a;Gherlone et al., 2016b;Oliveira et al., 2011;Rubinstein et al., 2019;Sabbah et al., 2019). The analysis of the immunological profile and serological values in a population of HIV-positive patients who had undergone prosthetic rehabilitation with implants (7 years of observation) revealed a statistically significant correlation (p = 0.009) between the viral load of patients and early implant failure due to infection and lack of osseointegration (Capparé et al., 2019). Additionally, a statistically significant correlation was found between the CD4+/CD8+ ratio and late implant failure, possibly due to the presence of periimplantitis (defined as a progressive loss of bone with signs of infection around the implant) (Capparé et al., 2019). Other authors did not find a correlation between implant survival (1 year followup) and levels of CD4+ cells (CD4 < 749.5 or CD4 > 749.5) ( Gherlone et al., 2016a;Gherlone et al., 2016b). However, to date, data for understanding the interaction between immune status and success of implant therapy are not sufficient (Duttenhoefer et al., 2019).
The aim of the present study was to assess the prevalence of peri-implant disease in an HIV-positive population and to evaluate correlation between the peri-implant disease and the population immune status.

| Type and sample of the study
This retrospective monocentric observational study assessed the prevalence of peri-implant disease in HIV-positive patients treated at the department of dentistry of our hospital. At the same hospital, the HIV-positive patients were undergoing antiretroviral treatment.
This study included 58 adult patients with at least one dental implant and a total of 222 implants followed for at least 1 year. Demographic, clinical, and laboratory information of the patients was collected.

| Study design
All included patients were divided into two groups, according to the type of prosthesis inserted after implant placement (all-on-4 prosthesis vs. single crowns\bridges), to assess the differences between the two types of prostheses in the inflammatory state, and therefore, in the presence of mucositis and peri-implantitis as well as in the variables analyzed.
As shown in Table 1, the data related to the immunological and systemic profile of the patients (HIVRNA load, CD4+ level, CD8+ level, hemoglobin level, and platelet count) were collected through the database of the hospital.

| Case definition for peri-implant condition
The clinical and radiological data for the diagnosis of peri-implant disease were collected according to the criteria laid in the Periodontology World Workshop in 2017 by Renvert et al. (2018).

| Clinical examination data
• Peri-implant mucositis: It was diagnosed on the basis of the following criteria: 1. Visual inspection demonstrating the following signs of inflammation in the peri-implant region: red as opposed to pink color, swollen tissues as opposed to no swelling, and soft as opposed to firm tissue consistency.
2. Presence of profuse (line or drop) bleeding and/or suppuration on probing.
3. An increase in probing pocket depths (PPDs) compared to baseline.
• Peri-implantitis: It was diagnosed on the basis of the following criteria: 1. Evidence of visual inflammatory changes in the peri-implant soft tissues combined with bleeding on probing and/or suppuration; 2. Increasing PPDs as compared to measurements obtained at placement of the supra-structure; and 3. Progressive bone loss in relation to the radiographic bone level assessment at 1 year following the delivery of the implantsupported prosthetics reconstruction; and 4. In the absence of initial radiographs and probing depths, radio-graphic evidence of bone level ≥ 3 mm and/or probing depths ≥6 mm in conjunction with profuse bleeding represents peri-implantitis.

| Radiological examination data
Intraoral periapical radiographs are considered gold standard for radiological evaluation and were used in this study. The radiographic data were collected from the records of the patient. The position of the marginal bone was measured manually on an ultraspeed radiographic film (Ultra speed, Kodak, USA) using a dental caliper (0-10 mm). The distance to the marginal bone was measured from the mesial and distal aspects of the implants. The implant platform was used as a reference for the measurements.
Two vertical lines were drawn parallel to a vertical line passing through the centre of the implant. The largest value was considered for statistical analysis. All assessments were performed by a single investigator (LC).
Peri-implantitis was assessed by radiographic examination data as progressive bone loss in relation to the radiographic bone level assessment at 1 year following the delivery of the implant-supported prosthetics reconstruction and in the absence of initial radiographs, radiographic evidence of bone level ≥3 mm.

| Inclusion criteria
All patients (>18 years old), who were undergoing antiretroviral therapy at the same hospital, who had undergone prosthetic rehabilitation using one or more dental implants, with at least 1 year of follow-up after loading, and who had returned for a maintenance appointment were considered for inclusion.

| Exclusion criteria
Patients who discontinued antiretroviral therapy, patients not undergoing therapy, patients whose previous radiographs and data on bleeding on probing and PPD at gingival level were unknown, patients with decompensated systemic diseases (for example, decompensated diabetes mellitus), patients treated with drugs that affect the bone turnover (e.g., bisphosphonates), and patients under 18 years of age were not included in this study.

| Statistical analysis
A descriptive analysis was performed to assess all the obtained data.
Continuous quantitative variables were described as medians and interquartile ranges, and categorical qualitative variables as frequencies and percentages (% All statistical tests were applied to 2 sails, and values were considered significant if the calculated probability was <0.05. The following two types of analyses were performed: the first analyzed the variables individually in the entire study population and the other in the population divided into two groups according to the type of implantprosthetic rehabilitation performed.

| RESULTS
The prevalence of peri-implantitis in the HIV-positive patient population was 26%. Table 2 shows the comparisons of individual variables (characteristics) between the patients of the two groups. For the first variable "age," the results showed that patients with allon-4 prosthetic rehabilitation were significantly older than those with single crowns/bridges (p = 0.004). For the second variable "years of HIV infection," the results indicated that patients with allon-4 prosthetic rehabilitation were infected with HIV for significantly greater number of years than those with single crowns/bridges (p = 0.003). For the variable "number of implants," patients with all-on-4 prosthetic rehabilitation had significantly greater number of implants than those with single crowns/bridges (p = 0.0001). This finding was expected and is obvious considering that the all-on-4 protocol requires the insertion of four implants. A statistically significant difference was observed in the prevalence of "peri-implantitis" between the two groups (p = 0.012). In fact, no patient with all-onfour prosthetic rehabilitation exhibited any sign of peri-implantitis, while in patients with single crowns/bridges, the prevalence was 34%. In terms of "PPD," which is directly correlated with periimplantitis, it was significantly lower in patients with all-on-4 prosthetic rehabilitation than that in patients with single crowns/ bridges (p = 0.0001).
The variables related to the patients' immunological and serological profiles (CD4+ and CD8+ levels and viremia) did not correlate with the type of prosthetic rehabilitation performed. Table 2 shows the comparison of variables between patients with and without peri-implantitis. Statistically significant differences were found between the two groups (patients with and without periodontitis) for the variables "duration of implant function" (p = 0.001), "implants for all-on-4 prosthetic rehabilitation" (p = 0.012), and Patients with all-on-4 prosthetic rehabilitation were not affected by peri-implantitis. Therefore, statistically significant differences were observed between patients with and without peri-implantitis for the variable "implants for all-on-4 prosthetic rehabilitation" For the variable 'duration of implant function', the results showed that patients with peri-implantitis had implants for a significantly longer duration than those without (p = 0.001) or that the probability of peri-implantitis increases as the age of the implant increases.
As evident from Table 3, PPDs in patients with peri-implantitis were significantly higher than those in patients without periimplantitis (p = 0.0001). This is obvious as the diagnosis of periimplantitis is based on increase in PPDs. The variables related to the patients' immunological profile (CD4+ and CD8+ levels and viremia) did not correlate with peri-implantitis.
However, the CD4+/CD8+ ratio at the time of implant placement was correlated with the presence of mucositis. Table 3 shows the comparisons between patients with and without mucositis in individual variables. The CD4+/CD8+ ratio significantly correlated with mucositis (p = 0.037). This means that patients' immunological profiles correlated with mucositis, and as the CD4+/CD8+ ratio decreases (specifically when the value falls below 1), the probability of mucositis increases.
The variable "number of implants" significantly correlated with mucositis (p = 0.0002), implying that the probability of mucositis increases as the number of implants increases.
The variable 'implant age' also significantly correlated with mucositis (p = 0.003), implying that as the implant age increases, the probability of mucositis also increases, as shown in Table 4.      type of prosthesis. In contrast, peri-implantitis and mucositis significantly correlated with implant age, as with increasing implant age, the probability of peri-implantitis (p = 0.001) and mucositis (p = 0.003) also increases. Figure 4 shows the relationship of CD4+/CD8+ ratio with the variables "type of prosthesis" (all-on-4, single crown/bridge), "periimplantitis," and "mucositis." No correlations were observed between type of prosthesis and CD4+/CD8+ ratio, and between periimplantitis and CD4+/CD8+ ratio. However, mucositis significantly correlated with CD4+/CD8+ ratio (p = 0.037); as this ratio decreases, the probability of mucositis increases.

| DISCUSSION
The prevalence of peri-implantitis in HIV-positive patients treated at the department of dentistry of our hospital (58 patients with 222 implants) was 26% and that of mucositis was 36%. Only one study has specifically analyzed the prevalence of peri-implantitis in HIV-positive patients (Capparé et al., 2019). However, these results should be interpreted with caution, because no specific criteria were established for the diagnosis of peri-implantitis (clinical and radiographic criteria) (Capparé et al., 2019). In our study, the prevalence of peri-implantitis in patients with all-on-4 prosthetic rehabilitation was 0%. However, considering the shorter follow-up period of our study compared with that of the study by Capparè et al.
Indeed, Schwarz et al. (2018) have clarified that mucositis can progress to peri-implantitis by a mechanism similar to the progression of gingivitis to periodontitis. This process, which is influenced by oral hygiene maintenance and plaque accumulation, has not been completely understood (Costa et al., 2012). This should be evaluated through future studies conducted in the same population of HIVpositive patients included in this study.
The prevalence of peri-implantitis in individuals not affected by HIV has been evaluated previously (Derks & Tomasi, 2015). However, the results here should also be evaluated with caution due to the different methodologies used for the diagnosis of peri-implantitis (Vignoletti et al., 2019). The systematic review by Derks and Tomasi in 2015 showed that the prevalence of peri-implantitis reported in various studies ranges between 1% and 47% (Derks & Tomasi, 2015).
The findings of our study fall within this reference range (26%). Vignoletti et al. (2019) reported the prevalence of peri-implantitis and mucositis as 35% and 38%, respectively, in a population of non-HIV-positive patients (237 patients with 831 implants). The presence of bleeding on probing and/or suppuration with a radiographic bone loss <2 mm and ≥ 2 mm were the diagnostic criteria for mucositis and peri-implantitis, respectively. In contrast, the threshold for bone loss in the diagnosis of peri-implantitis and mucositis was 3 mm in our study. Furthermore, the number of enrolled patients (831 vs. 58) and follow-up period (38 months vs. 56 months) were lower in our study.
These differences in study parameters could be the reason for the differences in the prevalence of peri-implantitis (26% vs. 35%); the reported prevalence of mucositis was similar between the two studies (36% vs. 38%).
The secondary objective of this study was the evaluation of correlation between the immunological profile and peri-implant pathology. No correlations were found between the prevalence of periimplantitis and immunological and serological profiles of patients.
However, a correlation was found between the CD4+/CD8+ ratio and mucositis. Univariate analysis (Table 3) showed that CD4+/CD8+ ratio was significantly lower in patients with mucositis than that in patients without mucositis (p = 0.037), and as this value decreased, in particular as this value became less than 1, the probability of mucositis increased. This ratio is a diagnostic immunological marker used in numerous studies to evaluate the disease progression in HIV-positive patients (Antinori et al., 2018). CD4+ (T-helper) and CD8+ (T-suppressor) are two phenotypes of T-lymphocytes, characterized by distinct markers on surface and of different functions.
F I G U R E 1 Prevalence of peri-implantitis and mucositis in the two types of implant rehabilitation F I G U R E 2 Relation between patient age and type of implant and the presence of peri-implantitis and mucositis F I G U R E 3 Relation between implant age and the type of implant rehabilitation and the presence of peri-implantitis and mucositis F I G U R E 4 Relation between baseline CD4+/ CD8+ ratio and implant type and the presence of peri-implantitis and mucositis A CD4+/CD8+ ratio between 1.5 and 2.5 is usually considered normal, but heterogeneity according to sex, genetic factors, and infections is observed (McBride & Striker, 2017). In fact, a decrease in this ratio is associated with cytomegalovirus and sexually transmitted infections. HIV-positive homosexual patients are at greater risk of sexually transmitted diseases such as syphilis, gonorrhea, viral hepatitis, and chlamydia and herpes virus infections. These diseases have been correlated with a decrease in CD4+/CD8+ ratio and an increase in CD8+ lymphocytes (Freeman et al., 2016;Pope et al., 1994;Strindhall et al., 2013;Vieira et al., 2017). The results of various studies in this regard have been conflicting, with some associating a decrease in this ratio in HIV-positive patients with greater morbidity and mortality (Ferguson et al., 1995;Verboeket et al., 2020;Wikby et al., 1998)) and others reporting no correlation between the two peri-implantitis, if further follow-ups and investigations are conducted (Schwarz et al., 2018). If this occurs, the correlation between decrease in CD4+/CD8+ ratio and peri-implantitis reported in the aforementioned study would be confirmed (Capparé et al., 2019). In the peri-implant mucosal system, the host response includes both vascular and cellular elements, and T-lymphocytes are usually present in small groups in the connective tissue lateral to the epithelial attachment (Crespi et al., 2012). Mucositis is characterized by inflammatory infiltrate rich in vascular elements, plasma cells, and lymphocytes, which does not extend into the connective tissue above the bone crest and is confined to the level of the junctional epithelium (Berglundh et al., 2018). Therefore, a disorder in the lymphocytic inflammatory response may influence the prevalence of peri-implant disease.
Another secondary objective was to evaluate the differences among types of implant-prosthetic rehabilitation with respect to periimplant disease and the variables analyzed.
The patients were divided into 2 groups according to the type of  Crespi et al., 2012). This type of implantprosthetic rehabilitation involves the placement of 4 implants in the anterior maxilla or in the inter-foraminal region of the mandible ( Gherlone et al., 2016a;Gherlone et al., 2016b). Of these 4 implants, the 2 distal-most implants are inclined distally and the 2 mesial implants are straight to support a full-arch prosthesis with a molar cantilever of variable size ( Gherlone et al., 2016a;Gherlone et al., 2016b;Penarrocha-Diago et al., 2017).
In our study, of the 222 implants, 122 were rehabilitated using single crowns/bridges (43 patients) while 100 were rehabilitated using all-on-4 prostheses (15 patients).
Of the two groups, patients with single crowns/bridges showed a prevalence of peri-implantitis higher (34%) than patients restored using all-on-4 prostheses (0%), and the difference was statistically significant (p = 0.012). This finding should be interpreted carefully, due to the differences in the number of patients and implant age between the two groups (12 months), that is patients with all-on-4 prostheses had implants 1-year-younger than patients with single crowns/bridges. However, the data can also be partly explained by The findings of this study showed that patients rehabilitated using all-on-4 implant-prosthetic restorations were significantly older than those rehabilitated using single crowns/bridges (p = 0.004). This could be attributed to the facts that all-on-4 rehabilitations are performed in cases of bone atrophy and that the degree of bone atrophy is correlated with loss of teeth and precisely to patient's age Agliardi, Panigatti, et al., 2010;Penarrocha-Diago et al., 2017). The number of implants used in a patient often correlates with peri-implantitis (Derks et al., 2016;Vignoletti et al., 2019). In this study as well, this parameter appears to have the greatest correlation with peri-implantitis and mucositis, according to the multivariate analysis.
This study has several limitations. The presence of plaque and history of periodontitis are strongly associated with peri-implantitis; however, it was not possible to evaluate the presence of plaque and periodontal status of the same patient. Therefore, periodontally healthy patients could not be differentiated from periodontitis patients and patients with good oral hygiene could not be distinguished from those with poor oral hygiene.
From the available data and within the limitations of this study, we found that the prevalence of peri-implantitis was 26% in the population of HIV-positive patients and 8% in the dental implants. No correlation was found between patients' immunological and serological factors and peri-implantitis. However, mucositis was found to be associated with CD4+/CD8+ ratio, which was found to be an immunological risk factor for mucositis.
The most important risk factor for peri-implantitis and mucositis was implant age.

ACKNOWLEDGMENTS
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Authors kindly thank Prof. Raffaele Vinci and Prof. Massimo De Sanctis for their valuable teaching in Oral Surgery and Periodontology.

CONFLICT OF INTEREST
The authors declare no conflict of interest.

AUTHOR CONTRIBUTION
Luca Casula contributed to conception, design, data acquisition and interpretation, drafted and critically revised the manuscript. Andrea Poli performed all the statistical analysis. Giulia Artuso contributed to design, data acquisition and interpretation. Paolo Capparé contributed to conception, drafted and critically revised the manuscript. Enrico F.
Gherlone contributed to conception, drafted and critically revised the manuscript. All authors gave their final approval and agree to be accountable for all aspects of the work.

ETHICS APPROVAL STATEMENT
All procedures were approved by the Ethics Committee of University Vita-Salute, San Raffaele Hospital, Milan (on 15/7/2020 with number of protocol "PERIHIV2"-EC Reg. N. 133/INT/2020).

PATIENT CONSENT STATEMENT
Appropriate informed consent forms were provided to all patients and were signed if the patients agreed to participate in the study.

SOURCES
NOT Applicable because all the material is product by the author of this article in University Vita-Salute, San Raffaele Hospital.