Delayed alveolar bone repair and osteonecrosis associated with Zoledronic Acid therapy in rats: macroscopic, microscopic and molecular analysis

Abstract Objective This study aims to evaluate bone repair and the development of the medication related osteonecrosis of the jaw (MRONJ) associated with the use of zoledronic acid in Wistar rats. Methodology 48 male Wistar rats were divided into four groups: ZA, treated with intraperitoneal zoledronic acid, 0.6 mg/kg every 28 days, totaling five doses; control (C), treated with 0.9% sodium chloride; ZA-surgical (SZA) and C-surgical (SC), submitted to extraction of the right upper molars 45 days after the first application. Alveolar bone repair was evaluated by macroscopic and histological analysis. Protein expression evaluations were performed by qPCR. Results Macroscopic evaluation showed that 91.66% (11) of the animals in the SZA group and 41.66% (5) from the SC group presented solution of epithelium continuity (P<0.05). All animals in the SZA group and none in the SC group had bone sequestration. The area of osteonecrosis was higher in the SZA group than in the SC group (P<0.05). In molecular evaluation, the SZA group presented changes in the expression of markers for osteoclasts, with increased RANK and RANKL, and a decrease in OPG. Conclusion The results highlighted strong and evident interference of zoledronic acid in bone repair of the socket, causing osteonecrosis and delayed bone remodeling.


Introduction
Bisphosphonates (BF) are synthetic drugs analogous to inorganic pyrophosphate, which are most commonly used in the treatment of bone disorders such as osteoporosis, bone metastases, Paget's disease, and Multiple Myeloma. 1 The first report describing bone exposure associated with the use of bisphosphonates was published in 2003. 2 In 2014, the disease nomenclature changed to Medication-Related Osteonecrosis of the Jaw (MRONJ), as it was found that other anti-resorptive drugs and angiogenesis inhibitors may also cause bone exposure similar to that found in MRONJ, however bisphosphonates are still the major cause, mainly by the use of zoledronic acid. 3 MRONJ pathophysiology is still unknown, however, risk factors for developing this condition can be classified into three categories: risk factors related to drug intake, local, and systemic risk factors. 4 Among local risk factors, dentoalveolar surgeries, 5 anatomical differences between maxilla and mandible; and preexisting oral pathology 3 may be cited.
One of the possible pathways affected by BF is the RANK/RANKL/OPG system. 6 RANK is a transmembrane receptor expressed on the surface of osteoclasts and osteoclasts precursor cells, and it is activated by the ligand RANKL, a protein typically produced by osteoblasts and T cells. In the presence of macrophage colony stimulating factor, RANKL stimulates the differentiation of osteoclast precursor cells into osteoclasts and also stimulates its activation, thus stimulating bone resorption. As a regulatory element of this system, OPG is a soluble receptor acting as "decoy" of RANKL, preventing the interaction between RANK and RANKL and bone resorption. [7][8][9][10] Several attempts to develop animal models in rodents have already been published in the literature. 11,12,[13][14][15][16][17][18][19][20]21,22, Therefore, this study applied an animal model under zoledronic acid, to study the effects of MRONJ on bone repair, after extraction of maxillary molar using microscopy and quantitative molecular analysis of mRNA expression involved in bone repair.

Methodology
In this study 48 Wistar rats (Rattus norvegicus albinus) were divided into four groups (12 animals in each group), with 12 weeks of life. Due to hormonal variations, which may influence bone repair process, only male rats were used. Animals were feed with solid feed and water ad libitum. The study was approved The rats were randomly divided into four groups: • Control (C): animals received 0.9% sodium chloride in the same volume as the medications of the other groups and no tooth extractions.
• Surgical control (SC): animals received 0.9% sodium chloride in the same volume as other medications and underwent extraction of right maxillary molars. kg may be considered slightly elevated, note that it was administered monthly, and its safety is supported by in vivo preclinical studies. 14 As RT-PCR molecular analysis was performed in association with the histological analysis, we chose to use 12 rats in each group. Thus, we had a sufficient number of rats, six in each group for each analysis, to allow statistical analysis between the control and experimental groups.

Surgical procedures
Forty-five days after the start of the experiment, the animals in the SC and SZA groups were submitted to the extraction of the first, second, and third maxillary right molars ( Figure 2).

Molecular analysis
The right hemi-maxillae of the six animals of each group were cleaned, removing all soft tissue (gingiva, palatal mucosa, and periosteum) and they were stored in microcentrifuge tubes containing 1 mL of RNAlater Stabilization Reagent (Qiagen, Hilden, Germany).
Tubes were agitated for 30 seconds, seated rest for 5 minutes, at room temperature, then frozen at -80°C.
The process of RNA extraction followed the protocol of Kelly, et al. 23 (2014).
The concentration of total RNA in the samples was

Statistical analysis
The results of the macroscopic evaluation for the presence or absence of solution of continuity and the qualitative microscopic evaluation for the presence or absence of bone sequestration and root fragments in the region of dental socket of the SC and SZA groups were submitted to the Fischer's exact test. For the results of quantitative microscopic analysis between the SC and SZA groups, unpaired Student t-test was applied for normal distribution, and Mann-Whitney test was used when the distribution was not normal. In the molecular analysis, the variables did not present a normal distribution and Kruskal-Wallis test was applied followed by Dunn post-test. The tests were performed using a software GraphPad Prisma 5 ® , 5% of significance level.

RESULTS
All animals survived until the end of the experiment and their weights at day 0 and day 150 of the groups are shown in Table 1, and no statistically significant difference among groups and times was observed   Figure 4).

Qualitative microscopic analysis
The qualitative analysis of the maxillae in the teeth regions did not present relevant changes among the studied groups (C, ZA, SC and SZA). Most animals in the SC group presented bone repair with formation of mature bone tissue in the extraction site. All animals in SC and SZA groups presented root fragments in at least one studied region. Inflammatory infiltrate was usually found surrounding the root fragments.
In the SC group, when the site of tooth extraction presented oral mucosal epithelium continuity solution, this was, mostly, specific and related to the presence of root fragments. In the SZA group, solution of continuity was abundant and related to bone sequestration, which were frequent in this group. These characteristics are related to the findings in the macroscopic evaluation, in which the presence of a greater solution of continuity was observed in the SZA group.
The socket in the extraction sites of the SC group, in most cases, presented remodeling of the bone ridges and interradicular septa, with bone loss at the height of alveolar ridge. In the SZA group, a decrease in the loss of height of alveolar ridge, mainly due to a decrease in resorption of alveolar crests and interradicular septa was found. In some cases, bone sequestration occurred in the interradicular septa and alveolar bone crests. Bone sequestration often occurred in areas that presented aspect similar to segments of bone fracture, which may have occurred at the extraction ( Figure 5).

Quantitative microscopic analysis
In the microscopic analysis of the socket, bone sequestration was present in all animals from the SZA group (100%), whereas in the SC group no animals presented bone sequestration (0%) in the extraction   Table 2).
The results for osteonecrosis area were higher in the SZA group, with a statistically significant difference in the first and second molar regions (p=0.0227 and 0.02, respectively). Trabecular space and periosteal reaction variables did not present statistically significant difference between the groups SC and SZA. The total bone area, measured in each ROI, was higher in the animals of the SZA group. However, it presented a statistically significant difference only in the third molar region (p=0.0004) ( Figure 6).

Molecular analysis
RANK expression increased in the ZA, SC, and SZA groups, and there was a statistically significant difference between C and SZA groups (p=0.0481). In the RANKL variable, an increase in expression for the groups ZA, SC, and SZA was observed, presenting statistically significant difference between the groups C and ZA, C and SZA (p=0.0011). There was an increase  Osteonecrosis found in this study was higher than that observed in some experimental models that showed a high incidence of bone sequestration 14, 16,20 .
The areas of solution of continuity in the SZA group were greater, with large regions of bone tissue exposure and, when compared with histological findings, they were more related to areas of osteonecrosis and/or bone sequestration. These areas have been described in other studies evaluating the effects of BF on bone repair of dental socket in rats. 13,15,20 Importantly, no formation of spontaneous lesions in the mouth was observed in any of the studied groups, which shows that dental extractions were significant factors in the development of MRONJ lesions. 14,20 The sample collection was performed 105 days after extraction of the maxillary molars, therefore a complete repair of the socket would be expected, [24][25][26][27] a fact that did not occur in the SZA group.
Most animals in SC group presented oral mucosa epithelium covering, corroborating previous studies results. 20 The socket of animals in the SZA group  Regarding the molecular analysis, focused on the RANK/RANKL/OPG system, RANK and RANKL expression increased and OPG decrease in the animals in the SZA group, suggesting a greater activation of the osteoclasts. 30 These results corroborate those found by Di Nisio,et al. 31 (2015)  In this context, the activation of osteoclasts could be a protective mechanism of the bone tissue to delimit the necrotic area and eliminate infection. 31,32  In our study, RANKL/OPG ratio of the SC group returned to values close to that of C group at the end of the experiment, corroborating with previous studies that evaluated the alveolar repair in Wistar rats in drug interventions and concluded that the favorable alveolar bone healing is expected at 28 days after tooth extraction in rats, with equilibrium between RANKL and OPG expression or even the predominance of OPG in the end. 40 On the other hand, the administration of zoledronic acid increased the RANKL / OPG ratio mainly in the SZA group. 41 One limitation of our study is that several attempts to develop animal models in rodents have already been published in the literature; 11,12,21,22,[13][14][15][16][17][18][19][20]