Effects of cisplatin and tamoxifen chemotherapy on peri‐implant bone remodeling around titanium implants in ovariectomized rats: An in vivo study

Abstract

Background

The proposition of the present study was to evaluate the influence of the secondary effect of tamoxifen (TAM) associated with cisplatin (CIS) in the bone remodeling of osseointegrated titanium implants installed in rat tibiae.

Methods

One hundred female rats underwent bilateral ovariectomy (OVX) and received titanium implants in both tibiae. Six weeks later, animals were treated with tamoxifen (15 mg/kg) or saline, with further subdivisions receiving cisplatin (5 mg/kg or 2.5 mg/kg) or saline. A non‐ovariectomized group served as a negative control. Animals were euthanized at 30 and 90 days after treatment initiation. Tibiae were harvested for histometric analysis of bone‐to‐implant contact (BIC) and bone ingrowth (BIN), histology, and immunohistochemistry (tartrate‐resistant acid phosphatase [TRAP], osteocalcin [OCN], and runt‐related transcription factor [RUNX2]). Additionally, scanning electron microscopy (SEM) and energy‐dispersive X‐ray spectroscopy (EDS) were used for ultrastructural and elemental analyses.

Results

The groups that received TAM showed higher BIC and BIN, increased expression of RUNX‐2 and OCN, and a lower number of TRAP‐positive cells. At 30 and 90 days, almost all spaces were filled with vital, well‐vascularized bone tissue without inflammatory foci. TAM groups also exhibited an increased calcium/phosphate (Ca/P) ratio compared to their respective controls, and a progressive thickening of collagen fibril bundles was observed in the bone matrix.

Conclusion

Tamoxifen positively influenced bone remodeling around osseointegrated titanium implants in rats undergoing cisplatin‐based cancer therapy.

Plain language summary

Dental implant surgery in patients undergoing chemotherapy is frequently approached with caution, as these pharmacological treatments can impair bone metabolism and jeopardize the integration of the implant with the bone. This study investigated whether tamoxifen, a medication widely utilized in breast cancer therapy, could potentially enhance bone healing around titanium implants, specifically when administered alongside the chemotherapeutic agent cisplatin. Utilizing a laboratory model that simulates postmenopausal bone loss, the research demonstrated that tamoxifen significantly improved both the volume and the structural quality of the bone surrounding the implants, partially mitigating the adverse effects associated with cisplatin. These findings are of clinical importance as they suggest that a history of chemotherapy should not be considered a definitive barrier to successful dental rehabilitation. Instead, with appropriate pharmacological support, dental implant procedures may be a more viable and predictable treatment option for cancer survivors than previously recognized. This research provides a foundation for clinicians to better evaluate the feasibility of oral rehabilitation in this patient population, ultimately aiming to improve their long‐term oral health and quality of life.

1. INTRODUCTION

Osseointegration of titanium implants has been considered the gold standard for aesthetic and functional rehabilitation in cases of tooth loss for more than 5 decades. ¹ As aging and tooth loss are correlated, adults 60+ years of age are the main subjects receiving dental implants. ² Osseointegration, described as the direct anchoring of the implant by bone formation without the growth of fibrous tissue at the bone/implant interface, is a dynamic process essential for the dental implant's early and long‐term clinical success. At the later stages, the so‐called “remodeling” begins, when overlapping phases of bone resorption and novel matrix deposition are orchestrated to maintain implant stability. ³ Although the reported success rate for bone‐integrated titanium implants is high, failures are still reported and mainly attributed to different local and systemic factors. ⁴

Tamoxifen (TAM) is a selective estrogen receptor modulator used in conjunction with antineoplastic drugs, ⁵ aiming to combat cancer cells, frequently administered in a long‐term regimen of 5 to 10 years. ⁶ Cisplatin (CIS) is a drug widely used for cancer treatment. ⁷ CIS has its antitumor activity due to the interaction of its central platinum atom with nitrogenous bases, which inhibits replication, transcription, translation, and repair of DNA. ⁸ , ⁹ When combined with antineoplastic drugs, TAM can their anticancer effects. When combined with CIS, TAM synergizes its antiproliferative effect, thus increasing cytotoxicity and apoptotic effect on cancer cells without increasing adverse effects, and delaying the development of resistance to the drug. ¹⁰ This combinatorial approach results in an increased survival rate of cancer patients. Importantly, the deleterious effects of CIS on the bone repair process around implants have been reported in the literature. ¹¹ , ¹²

Interestingly, TAM exerts a partial agonist effect on estrogen receptors, reducing bone turnover and preventing bone loss in post‐menopausal women. ¹³ The literature reports that TAM mimics the effect of estrogen on bone tissue, preventing bone loss in women after menopause through direct inhibition of osteoclast formation and increased osteoclast apoptosis, positively affecting the balance between bone formation and resorption. This impact on peri‐implant bone remodeling was observed by Fiorin et al. (2022), ¹⁴ who reported that TAM increased bone‐to‐implant contact (BIC), bone ingrowth (BIN), reduced the expression of tartrate‐resistant acid phosphatase (TRAP) and increased the expression of runt‐related transcription factor (RUNX) 2 and osteocalcin (OCN) around osseointegrated titanium implants installed in the tibiae of OVX rats. On the other hand, when ovariectomy (OVX) is used to mimic osteoporotic conditions in animals, studies have demonstrated that estrogen deficiency results in a lower rate of bone renewal, reduced (BIC), compromised biomechanical competence at the interface of the bone and implant, and reduction in density of cancellous bone. ¹⁵ , ¹⁶ , ¹⁷

Given the negative impact of CIS on the peri‐implant bone ¹¹ , ¹² , ¹⁸ and the improved bone remodeling mediated by TAM, ¹⁴ the behavior of the peri‐implant bone is a setting of combinatorial approach (CIS+TAM) for treating cancer remains to be understood. Hence, the objective of the work was to assess whether the secondary agonistic effect of TAM on estrogen receptors could rebalance the harmful impact of CIS on the peri‐implant bone remodeling around osseointegrated titanium implants installed in the tibiae of OVX rats. Our Null hypothesis would be a negative impact of TAM when in association of CIS‐based chemotherapy on the peri‐implant bone around osseointegrated titanium implants.

2. MATERIALS AND METHODS

2.1. Animals

One‐hundred 3‐month old female rats (Rattus norvegicus, Albinus, Wistar), weighing approximately 250 to 300 g, were used in the present study. The animals were kept in plastic cages in groups of 4 with food and water ad libitum. Before surgical procedures, all animals were kept in an environment with a stable temperature (22 ± 2°C) throughout the experimental period. All protocols described were approved by the Animal Experimentation Ethics Committee (CEEA) of the São Paulo State University (protocol no. 00196‐2020) and conducted according to the ARRIVE Guidelines.

The determination of the sample size was calculated based in a previous study. ¹⁴ BIC was chosen as the parameter to calculate the sample size in order to achieve 0.8 power and 0.05 alpha error based on a 10% potential standard deviation and the assumption that a 10% difference would be relevant. This study was designed as a randomized, controlled, and single‐blind experiment. Simple randomization with a 1:1 allocation ratio was performed using Minitab 17 software (Minitab Inc.).^*, a group size of 10 animals per group was sufficient to reject the null hypothesis in BIC, BIN histometry, immunohistochemistry, and elemental analysis.

2.2. Primary and secondary outcomes

The primary outcome was determined as the amount of peri‐implant bone; thus, BIC (mm) was considered primary outcome. The secondary outcome was defined as the BIN (mm²), cellular events, and description of the structure of the peri‐implant tissues through histological, immunohistochemical, and elemental analysis.

2.3. Experimental groups and treatments

Eighty animals were initially assigned to OVX groups. The animals of the Groups OVX received previous bilateral OVX 22 weeks prior to the begging of the experiment. The animals were then subdivided in groups which received saline solution (SS) or TAM via gavage and SS or CIS via intraperitoneal. The groups were determined as follows: Groups OVX SS–SS (n = 20) received SS via gavage and intraperitoneally, OVX SS–CIS (n = 20) received SS via gavage and CIS intraperitoneally, OVX TAM‐SS (n = 20) received TAM via gavage and SS intraperitoneally and OVX TAM–CIS (n = 20) received TAM via gavage and CIS intraperitoneally.

A SHAM SS–SS group (n = 20) was used as a means of total negative control. The SHAM group received a procedure to simulate the bilateral OVX and received SS via gavage and intraperitoneally.

2.4. Experimental protocol

2.4.1. Anesthesia

For all surgical procedures, rats were anesthetized with ketamine^† (70 mg/kg body weight) and xylazine^‡ (6 mg/kg bodyweight), via intramuscular injection. Each animal received post‐surgical intramuscular injections of 24,000 IU of penicillin G‐benzathine^§.

2.4.2. The OVX protocol

After a septic preparation, bilateral OVX was performed in the OVX groups following the protocol of Fiorin et al. (2022). ¹⁴ In SHAM group, the ovaries were lifted and returned to the original position.

All animals underwent cytological examination to ascertain its cycle phase in the 2 weeks prior to the implants placement. ¹⁹ The vaginal cells were flushed by introducing the saline and drawing it with a modified pipette inserted in the entrance of the vaginal canal. The fluid was dropped in a slide and immediately analyzed under light microscopy at 400x magnification. The phase of the cycle was determined according to the presence of epithelial cells, cornified cells, and leukocytes in the cytological examination. ²⁰

2.4.3. Systemic treatments

The administration of either 0.5 mL of 0.9% SS or TAM occurred via gavage, daily, in the morning (15 mg/kg). The systemic treatments started 6 weeks post‐implant placement, in this way, the effect of the treatment in the bone remodeling can be assessed. ²¹ , ²² The dosage calculation was based on previous studies. ¹⁴ For CIS, the first and second doses were 5 and 2.5 mg/kg of body weight, respectively. ¹¹ , ¹⁸

2.5. Experimental subgroups and treatment regimens

2.5.1. SHAM SS–SS group

Animals underwent bilateral SHAM surgery 22 weeks before the experiment, followed by estrous cycle verification from Week 0 to 2 and implant placement at Week 2. Systemic treatment began at Week 8 with daily gavage of SS (0.5 mL of 0.9%), accompanied by intraperitoneal SS (0.5 mL of 0.9%) administered 2 and 4 days after gavage initiation. Euthanasia was conducted at Weeks 12 and 20.

2.5.2. OVX SS–SS group

Animals underwent bilateral OVX 22 weeks before the experiment, followed by estrous cycle verification from Week 0 to 2 and implant placement at Week 2. Systemic treatment began at Week 8 with daily gavage of SS (0.5 mL of 0.9%), accompanied by intraperitoneal SS (0.5 mL of 0.9%) administered 2 and 4 days after gavage initiation. Euthanasia was conducted at Weeks 12 and 20.

2.5.3. OVX SS–CIS group

Animals underwent bilateral OVX 22 weeks before the experiment, followed by estrous cycle verification from Week 0 to 2 and implant placement at Week 2. Systemic treatment began at Week 8 with daily gavage of SS (0.5 mL of 0.9%), and intraperitoneal CIS (5 mg/kg and 2.5 mg/kg) was administered 2 and 4 days after gavage initiation. Euthanasia was conducted at Weeks 12 and 20.

2.5.4. OVX TAM–SS group

Animals underwent bilateral OVX 22 weeks before the experiment, followed by estrous cycle verification from Week 0 to 2 and implant placement at Week 2. Systemic treatment began at Week 8 with daily TAM gavage (15 mg/kg, diluted in 0.5 mL of 0.9% SS), accompanied by intraperitoneal SS (0.5 mL of 0.9%) administered 2 and 4 days after gavage initiation. Euthanasia was conducted at Weeks 12 and 20.

2.5.5. OVX TAM–CIS group

Animals underwent bilateral OVX 22 weeks before the experiment, followed by estrous cycle verification from Week 0 to 2 and implant placement at Week 2. Systemic treatment began at Week 8 with daily TAM gavage (15 mg/kg, diluted in 0.5 mL of 0.9% SS), and intraperitoneal CIS (5 mg/kg and 2.5 mg/kg) was administered 2 and 4 days after gavage initiation. Euthanasia was conducted at Weeks 12 and 20.

2.6. Installation protocol for implants

For each tibiae, the surgical site received 1 titanium implant (4.0 × 2.2 mm)^‖ installed at bone level. The implant surface was treated by sandblasting and acid etching. Tissues were sutured on 2 planes with resorbable and non‐resorbable threads. ¹¹ , ¹⁴ Morphine IM (2.5 mg/kg) was administered for analgesia during 3 postoperative days, with a 24‐hour interval.

2.7. Euthanasia and sample processing

Ten animals per group/ period were euthanized with an overdose (150 mg/kg) of sodium thiopental^¶ at 30 and 90 days after the first gavage administration, thus totaling 20 tibias per group/period. The tibiae were either fixed in buffered 4% formaldehyde solution for 48 hours and further processed either by ground section processing or conventional histologic processing with demineralization and paraffin embedding.

Ground sections were obtained following a previous protocol. Five tibias per group/period underwent progressive dehydration in fuchsine‐added alcohol solutions, acetone washings, and embedding in high‐viscosity crystal resin^#. Samples were sanded and polished with fine‐grade sanding paper^** (#100, #400, #600, #1200) until reaching 50‐µm thickness. ¹⁴

Following fixation in formalin, 10 tibias per group/period were demineralized in buffered 10% ethylenediaminetetraacetic acid (EDTA). After initial embedding in paraffin, with the implant installed in the tibiae, the implants were manually retrieved with caution from the specimens and re‐embedded in paraffin. Semi‐serial sections with 4 µm thickness were then cut longitudinally along the site previously occupied by the implant. Six equidistant sections comprising the central portion of the implant site from each specimen were stained with hematoxylin and eosin (H&E) for histological and histometric analyses of bone ingrowth. Six other equidistant sections were subjected to the indirect immunoperoxidase method. Primary antibodies were then used: goat anti‐TRAP^††, rabbit anti‐RUNX2^‡‡, and goat anti‐OCN^§§.

Five tibiae were fixed in a solution of 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M buffered cacodylate (pH 7.4) at room temperature for 4 hours, and transferred to pure 0.1 M cacodylate solution. Specimens were dehydrated in solutions of increasing alcohol concentration and included in high viscosity resin. The resin blocks were cut and sanded in the longitudinal direction of the implant until they reached a thickness of 30 µm. The sections were polished with diamond suspensions with particle sizes of 6, 1, and 0.25 µm. Each polished section was acid‐etched (37% phosphoric acid) for 10s, washed in 5% sodium hypochlorite solution for 5 min. After drying, they were sputtered with gold.

2.8. Analysis of the results

Calibrated and blinded examiners performed the analyses, with no prior knowledge of the treatments. A certified histologist (E.E.), specialized in osteobiology, determined 2 regions of interest (ROIs) based on the distribution and significance of events occurring in the peri‐implant tissues—whether pre‐existing or newly formed—as a result of implant placement. ¹⁴ A figure illustrating the ROI was included (Figure 1).

FIGURE 1.

Illustration of the experimental design (A) and the delimitations of the regions of interest (B). ROI 1 is represented in yellow; the analytical areas of ROI 2 (mesial and distal) are represented in white; and the basal area used exclusively for EDS normalization is represented in green. EDS, energy‐dispersive X‐ray spectroscopy; ROI, region of interest.

ROI 1: Determined as the tissue between the threads, extending from the valley of the threads to 0.4 mm (the same depth of the thread) toward the peri‐implant tissues. Selected for the histological analysis.

ROI 2: From the head of the implant, ROI 2 comprises the second thread located in cortical bone and the first and second threads located in bone marrow and was used for BIC and BIN quantification. ROI 2 was assessed on both the mesial and distal sides of the implant bed, and the 2 measurements were averaged to obtain a single value per specimen. A third basal area, displayed in green in Figure 1B, was not included in the histometric analysis; it was used exclusively for elemental normalization in the EDS assessment.

2.8.1. Histometric analysis of direct bone / implant contact

Within ROI 2, the percentage of BIC was determined using the ImageJ software^‖‖. The implant perimeter in intimate contact with bone was measured, and the results were expressed as the percentage of BIC relative to the total implant length.

2.8.2. Analysis of the BIN percentage

Within ROI 2, the BIN percentage was determined using the ImageJ software. The area of bone within the implant threads was measured, and the results were expressed as the percentage of BIN relative to the total area between the threads.

2.8.3. Histological analysis

A morphological assessment of cellular and tissue reactions was conducted under light microscopy^¶¶, focusing on the following parameters: (1) the cellularity pattern and structural organization of the peri‐implant bone tissue; (2) the cellularity pattern and structural organization of the peri‐implant connective tissue; and (3) the peri‐implant inflammatory profile, specifically the nature and intensity of the local inflammatory response.

2.8.4. Immunohistochemical analysis

Immuno‐positive cells presented a dark brown labeling confined to the cytoplasm for TRAP, confined to the nucleus for RUNX2 and present in both for OCN. Were counted the cells in which the cytosolic compartment and nucleus could be distinguished under optical microscopy at an original magnification of 400×. For RUNX2 and OCN, all immuno‐positive cells within ROI 2 were considered. For TRAP, only the immunoreactive cells in contact with the bone were included in the quantitative analysis.

2.8.5. Chemical composition of the bone

All samples were examined using a back‐scattered electron scanning electron microscopy (SEM) system (Jeol Neoscope Electron Microscope JCM‐6000). All slides were analyzed using energy dispersive spectroscopy (EDS) to determine the elemental composition. Specifically, the calcium/phosphate (Ca/P) ratio was calculated in the bone‐implant interface of all specimens. A qualitative description of the ultrastructural characteristics of the bone‐implant interface was also conducted using SEM.

2.9. Statistical analysis

For the analysis, the statistical analysis of the collected data was performed using the software BioEstat 5.0^##. The normality of data distribution was assessed by the Shapiro–Wilk test. The data were submitted to the analysis of variance (ANOVA) test with 2 criteria and the Tukey post‐hoc test. The level of significance was p ≤ 0.05.

3. RESULTS

All animals remained healthy and complication‐free throughout the experimental period. Considering BIC, the 2‐way ANOVA revealed significant main effects of treatment (F(4,90) = 79.37, p < 0.0001) and time (F(1,90) = 8.879, p = 0.0037), as well as a significant treatment × time interaction (F(4,90) = 8.563, p < 0.0001). Statistical significance was set at α = 0.05, and post‐hoc comparisons were adjusted for multiple testing. All descriptive statistics (mean ± SD) and inferential results, including exact p‐values and indicators of statistically significant differences among groups and time points, are presented in Table 1. Detailed complementary data related to the analyses are provided in the supplementary file.

TABLE 1.

Descriptive statistics (mean ± SD) and exact p‐values indicating significant differences among groups and time points.

Groups	Time	BIC	BIN	RUNX
SHAM SS–SS	30d	93.60 ± 4.40	86.90 ± 6.41	5.40 ± 1.26
	90d	99.20 ± 1.10	81.14 ± 6.00	6.30 ± 1.49
OVX SS–SS	30d	88.60 ± 2.50	52.38 ± 8.78 #(p < 0.0001)	6.80 ± 1.31 *1(p = 0.0054)
	90d	85.40 ± 3.36 #(p < 0.001) *1(p = 0.017)	55.35 ± 9.80 #(p < 0.0001) *1(p < 0.0001)	5.30 ± 1.16 *2(p < 0.0001)
OVX TAM–SS	30d	94.51 ± 3.60	60.26 ± 3.95 #(p < 0.0001)	9.20 ± 1.03 #(p < 0.0001) *1(p = 0.0054)
	90d	96.6 ± 3.78 *1(p = 0.017)	75.50 ± 3.58 $(p < 0.0001) *1(p < 0.0001)	9.60 ± 1.26 #(p < 0.0001) *2(p < 0.0001)
OVX SS–CIS	30d	83.95 ± 3.90 #(p = 0.01) *2(p = 0.004)	64.35 ± 1.95 #(p < 0.0001) *2(p = 0.0035)	6.20 ± 1.81 *3(p < 0.0001)
	90d	80.8 ± 6.38 #(p < 0.001) *3(p = 0.001)	64.55 ± 2.63 #(p < 0.0001) *3(p < 0.0001)	4.50 ± 1.43 *4(p < 0.0001)
OVX TAM–CIS	30d	94.40 ± 3.36 *2(p = 0.004)	75.10 ± 6.20 *2(p = 0.0035)	12.30 ± 1.56 #(p < 0.0001) *3(p < 0.0001)
	90d	92.20 ± 4.14 #(p < 0.001) *3(p = 0.001)	79.95 ± 4.16 *3(p < 0.0001)	10.40 ± 0.96 #(p < 0.0001) *4(p < 0.0001)
Groups	Time	OCN	TRAP	EDS
SHAM SS–SS	30d	9.9 ± 1.20	1.00 ± 0.81	1.80 ± 0.20
	90d	7.70 ± 1.89 *1(p = 0,0062)	0.50 ± 0.70	1.64 ± 0.25
OVX SS–SS	30d	10.40 ± 2.71 *2(p < 0.0001)	3.30 ± 1.25 #(p = 0.0005) $(p < 0.0001) *1	0.97 ± 0.14 #(p < 0.0001) *1(p = 0.0040)
	90d	11.00 ± 1.56 #(p = 0,0062) *1(p = 0.0062) 3	7.00 ± 1.24 #(p < 0.0001) *2(p < 0.0001)	1.03 ± 0.13 #(p = 0.0014) *2(p = 0.0209)
OVX TAM–SS	30d	16.90 ± 1.80 #(p < 0.0001) *2(p < 0.0001)	4.50 ± 1.43 #(p < 0.0001) *1(p = 0.033)	1.53 ± 0.29 *1(p = 0.0040)
	90d	14.10 ± 0.99 #(p < 0.0001) *3(p = 0.0134)	3.40 ± 0.84 #(p < 0.0001) *2(p < 0.0001)	1.51 ± 0.13 *2(p = 0.0209)
OVX SS–CIS	30d	9.60 ± 1.71 *4(p < 0.0001)	2.70 ± 1.06 #(p = 0.0318) $(p = 0.0318)	1.17 ± 0.30 #(p = 0.0009)
	90d	12.00 ± 2.10 #(p < 0.0001) *5(p < 0.0001)	4.40 ± 1.50 #(p < 0.0001) *3(p = 0.0318)	0.92 ± 0.07 #(p = 0.0001) *3(p = 0.0051)
OVX TAM–CIS	30d	15.20 ± 2.40 #(p < 0.0001) *4(p < 0.0001)	1.70 ± 0.82	1.16 ± 0.14 #(p = 0.0007)
	90d	18.60 ± 1.77 #(p < 0.0001) *5(p < 0.0001)	2.70 ± 1.16 #(p = 0.0012) *3(p = 0.0318)	1.48 ± 0.19 *3(p = 0.0051)

Note: Symbols: # Statistically significant difference with the SHAM SS–SS group at the same time points. $ Statistically significant difference with the same group in different time points.

Abbreviations: BIC, bone‐to‐implant contact; BIN, bone ingrowth; CIS, cisplatin; OVX, ovariectomy; RUNX, runt‐related transcription factor; SS, saline solution; TAM, tamoxifen.

*Statistically significant difference between groups at the same time points.

3.1. TAM improves BIC, BIN, and histological bone quality around implants

OVX SS–SS and OVX SS–CIS presented lower BIC in all experimental periods when compared to their TAM counterparts, OVX TAM‐SS and OVX TAM–CIS. The group OVX TAM–CIS presented lower BIC when compared with the SHAM SS–SS group at the late period of 90 days (Figure 2). Regarding BIN, no statistical difference was observed when comparing the group SHAM SS–SS with OVX TAM–SS and OVX TAM–CIS at 90 days. OVX SS–SS and OVX SS–CIS presented lower BIN in all experimental periods when compared to their TAM counterparts, OVX TAM–SS and OVX TAM–CIS (Figure 3).

FIGURE 2.

BIC for each experimental group. (A) Means and standard deviations (M ± SD) of BIC for each group and period. Statistical tests: 2‐way ANOVA and Tukey. Symbols: * Statistically significant difference between groups at the same time points (p ≤ 0.05). # Statistically significant difference with the SHAM SS–SS group at the same time points (p ≤ 0.05). Photomicrographs of the BIC in the groups SHAM SS–SS (B; G), OVX SS–SS (C; H), OVX TAM–SS (D; I), OVX SS–CIS (E; J) and OVX TAM–CIS (F; K). White arrows: Areas without bone/ implant contact. Staining: Basic fuchsin. Scale bars: 200 µm. ANOVA, analysis of variance; BIC, bone‐to‐implant contact; CIS, cisplatin; OVX, ovariectomy; SS, saline solution; TAM, tamoxifen.

FIGURE 3.

Bone percentage area for each experimental group. (A) means and standard deviations (M ± SD) of the BIN for each group and period. Statistical tests: ANOVA and Tukey. Symbols: * Statistically significant difference between groups at the same time points (p ≤ 0.05). # Statistically significant difference with the SHAM SS–SS group at the same time points (p ≤ 0.05). $ Statistically significant difference with the same group in different time points (p ≤ 0.05). Photomicrographs showing the histological features of the peri‐implant tissues for each group. Photomicrographs of the BIN in the groups SHAM SS–SS (B; G), OVX SS–SS (C; H), OVX TAM–SS (D; I), OVX SS–CIS (E; J) and OVX TAM–CIS (F; K). Staining: hematoxylin and eosin. Scale bars: 100 µm. ANOVA, analysis of variance; BIN, bone ingrowth; CIS, cisplatin; OVX, ovariectomy; SS, saline solution; TAM, tamoxifen.

In SS–SS, at both 30 days and 90 days, the space between the implant threads was filled with vital bone tissue, well vascularized, with few medullary spaces. At 30 days, OVX–SS–SS and OVX‐SS–CIS presented a small amount of bone tissue filling the space between the implant threads. The medullary spaces were considerably extensive and occupied by bone marrow in these groups. At 90 days, the bone tissue located between threads began to occupy most of the area, with smaller medullary spaces. Importantly, in OVX‐SS–CIS areas of non‐vital bone tissue and few inflammatory foci were observed in some specimens.

At 30 days in OVX‐TAM–SS and OVX‐TAM–CIS, the spaces between threads were occupied by equivalent amounts of vital bone tissue and medullary spaces filled with bone marrow. At 90 days, almost all of the spaces were occupied by vital and well‐vascularized bone tissue. In these groups, empty osteocyte lacunae were rarely observed in the bone tissue, and no inflammatory foci were observed (Figure 3).

3.2. TAM rebalances bone remodeling through osteogenic stimulation and resorption control

The groups OVX TAM–SS and OVX TAM–CIS presented a higher number of RUNX2‐positive cells when compared with their respective controls, OVX SS–SS, and OVX SS–CIS, and with the SHAM SS–SS in all experimental periods (Figure 4). The groups OVX TAM–SS and OVX TAM–CIS presented a higher number of OCN‐positive cells when compared with their respective controls, OVX SS–SS, and OVX SS–CIS, and with SHAM SS–SS in all periods (Figure 4). OVX SS–SS exhibited higher number of TRAP‐positive cells when compared to other groups at 90 days. The SHAM SS–SS group presented the lower number of TRAP‐positive cells amongst all groups (Figure 5).

FIGURE 4.

Immunolabeling pattern of RUNX2 (A–K) and OCN (L–V) at the peri‐implant tissues for each group and period. Means and standard deviations (M ± SD) of the number of RUNX2‐positive (A) and OCN‐positive (L) cells. Statistical tests: ANOVA and Tukey. Symbols: * Statistically significant difference between groups at the same time points (p ≤ 0.05). # Statistically significant difference with the SHAM SS–SS group at the same time points (p ≤ 0.05). $ Statistically significant difference with the same group in different time points (p ≤ 0.05). (B–K) Photomicrographs showing the immunolabeling pattern of RUNX2‐positive cells (red arrows) in SHAM SS–SS (B; G), OVX SS–SS (C; H), OVX TAM–SS (D; I), OVX SS–CIS (E; J) and OVX TAM–CIS (F; K). (M–V) Photomicrographs showing the immunolabeling pattern of OCN‐positive cells in SHAM SS–SS (M; R), OVX SS–SS (N; S), OVX TAM–SS (O; T), OVX SS–CIS (P; U) and OVX TAM–CIS (Q; V) Counter staining: (B–K) Fast Green; (M–V) Harris' hematoxylin; Scale bars: 25 µm. ANOVA, analysis of variance; CIS, cisplatin; OCN, osteocalcin; OVX, ovariectomy; RUNX, runt‐related transcription factor; SS, saline solution; TAM, tamoxifen.

FIGURE 5.

Immunolabeling pattern of TRAP (A–K) and ultrastructural and elementary properties of the peri‐implant tissue (L–V) for each group and period. Means and standard deviations (M ± SD) of the number of TRAP‐positive cells (A) and calcium/phosphate rate in the interface of bone/implant (L). Statistical tests: ANOVA and Tukey. Symbols: * Statistically significant difference between groups at the same time points (p ≤ 0.05). # Statistically significant difference with the SHAM SS–SS group at the same time points (p ≤ 0.05). $ Statistically significant difference with the same group in different time points (p ≤ 0.05). (B–K) Photomicrographs showing the immunolabeling pattern of TRAP‐positive cells in SHAM SS–SS (B; G), OVX SS–SS (C; H), OVX TAM–SS (D; I), OVX SS–CIS (E; J) and OVX TAM–CIS (F; K). (M–V) Photomicrographs showing the ultrastructural pattern of the collagen fibers in SHAM SS–SS (M; R), OVX SS–SS (N; S), OVX TAM–SS (O; T), OVX SS–CIS (P; U) and OVX TAM–CIS (Q; V) Counter staining: (M–V) Harris' Hematoxylin; Scale bars: 25 µm. Original magnification: 2000x. Scale bars: 10 µm. ANOVA, analysis of variance; CIS, cisplatin; OVX, ovariectomy; SS, saline solution; TAM, tamoxifen TRAP, tartrate‐resistant acid phosphatase.

3.3. TAM improves bone matrix organization and mineral composition

The ultrastructural pattern of the peri‐implant tissues is shown in Figure 5(B‐K). The ultrastructural pattern was similar among all groups. In all groups, regardless of the administration of TAM or chemotherapy, a progressive thickening of collagen fibril bundles was observed in the bone matrix. The groups OVX TAM–SS and OVX TAM–CIS presented a higher Ca/P ratio when compared with their respective controls, OVX SS–SS, and OVX SS–CIS, at 90 days.

4. DISCUSSION

Our study evaluated the effect of TAM, an adjuvant drug used in association with chemotherapy for the treatment of breast cancer in the peri‐implant bone of the rat. As observed by the primary outcome measure of this study, groups that received TAM presented higher BIC when compared to their counterparts, even when associated with CIS, as seen in the OVX TAM–SS and OVX TAM–CIS. For the secondary outcomes, OVX TAM–CIS presented higher BIN, RUNX‐2, OCN, and TRAP positive cells and Ca/P rate when compared with OVX SS–CIS. Furthermore, the experimental design of the present study allows one to hypothesize that the positive influence of TAM on bone remodeling exerts a protective effect on the osseointegration of titanium implants in subjects under anticancer chemotherapy with CIS.

Peri‐implant bone remodeling was the biological focus relevant for interpreting the effects of TAM and CIS in this study. The ovariectomized rat model provided the metabolic bone environment required to evaluate these interactions under estrogen‐deficiency‐induced turnover alterations. ²³ , ²⁴ Because TAM enhances CIS‐induced DNA platination, ¹⁰ assessing their combined administration was essential to determine whether this interaction modifies peri‐implant bone remodeling dynamics in a system characterized by impaired bone metabolism.

Ideally, the full contact between the bone and the implant is considered when studying the biomechanics of the implant. However, in a clinical scenario, the contact between the threads and the bone depends on a variety of factors, as systemic conditions. Estrogen deficiency and chemotherapy impair bone remodeling and present a risk factor for implant loss. ²⁵ In our study, the group OVX CIS–SS presented lower BIC when comparing to the OVX SS–TAM. As seen in previous works of our group, CIS downregulated bone formation markers and lower BIC could be observed ¹¹ ; meanwhile, TAM increased BIC in a time dependent manner. ¹⁴ It could be theorized that an imbalance between bone deposition and resorption caused lower BIC and lower BIN, as well.

BIC was interpreted in this study within the broader context of peri‐implant bone remodeling, which ultimately governs the maintenance of bone within implant threads. Systemic inflammatory activity can modulate this process by enhancing osteoclast differentiation and resorption potential. Although osteoclast activation is not strictly dependent on inflammatory mediators, osteoclastogenesis is amplified in the presence of cytokines such as interleukin (IL) ‐1 and IL‐6, ²⁶ reinforcing the need to consider systemic inflammatory status when interpreting peri‐implant bone responses. As observed in the histological analysis, greater areas of inflammatory exudate were present in the CIS groups when compared to groups under the administration of TAM. CIS contributes, via cell death, to the progression and maintenance of the inflammation, increasing the releasing of pro‐inflammatory mediators like IL‐1B, IL‐6, IL‐8, and IL‐18. ²⁷ TAM down‐regulates IL‐8 and IL‐18 and increases the IL‐1ra/IL‐1B ratio. ²⁸ IL‐8 chemoattracts and activates neutrophils in the inflamed region. ²⁹ IL‐18 is described as an interferon inducing factor and is involved in natural killer cell activation. ³⁰ IL‐1ra is a member of the family of IL‐1 that binds to IL‐1 receptors but does not induce any response, in addition, decreases the proinflammatory effect of IL‐1B. ²⁸ Not only the TAM groups presented higher BIN when compared to the experimental groups with reduced areas of inflammation, in addition, TAM groups presented a more mature structured bone tissue, with little to none non‐vital bone

Although various authors report the association of TAM and CIS in cancer patients, the impact of their interaction in the peri‐implant bone remains unassessed. Extensive studies demonstrated the negative effect of CIS in the bone repair. ¹¹ , ³¹ Despite the evidences, the pathway in which CIS negatively affects bone remodeling is unclear. RUNX2 has a key role in bone formation and osteoblast differentiation. During intramembranous ossification, RUNX2 promotes differentiation of stromal cells in osteoblasts and in endochondral ossification, and it promotes differentiation of the cartilage in osteoblast. In our study, the groups which received TAM and CIS/SS presented higher number of RUNX2‐positive cells when compared to SS groups. RUNX2 interacts with the estrogen receptor alpha (Era) receptor, ³² a nuclear receptor with the function of proliferation of cancer cells, and promotes metastasis of breast cancer. ³³ However, when bound by TAM, ERa is no longer functional and the cancerous cells enter into apoptosis. ³⁴ TAM, in addition, has an effect in Erb, which has a cancerous cell growth‐inhibitory effect. Previous authors observed that RUNX2 is linked with the sensitivity of breast cancer cells to TAM ³⁵ and that RUNX2 upregulation was associated with the CIS resistance. ³⁶ However, the specific mechanisms of the RUNX2‐ mediated drug resistance is still unknown.

The pathway in which CIS negatively affects bone healing is still undisclosed. In our study, as confirmed in the histological analysis, in the CIS groups a persistent inflammation process could be observed in the medullar bone in between the implant threads. In the same groups, the expression of TRAP‐positive cells was observed in higher number, suggesting an inflammation‐mediated bone resorption corroborating with the findings of Matheus et al. (2021). ¹¹ Administration of TAM, with or without CIS, reduced the number of TRAP‐positive cells while decreasing the expression of pro‐inflammatory cytokines. In addition, less inflammatory infiltrate was observed in these groups, strengthen the hypothesis of an inflammatory‐mediated bone resorption

In this study, OCN was used as a functional marker of osteoblastic activity to contextualize the peri‐implant bone response under TAM and CIS exposure. In our study, in the groups in which TAM was administered, an increase in the number of OCN‐positive cells was observed. CIS has a deleterious effect in the expression of OCN, ¹² although not highlighted in this study, the group OVX TAM–CIS presented the highest number of immunoreactive cells. OCN is required for bone strength, as regulates the growth of hydroxyapatite crystals and adjust the alignment of calcium deposition parallel to the collagen fibers. ³⁷ Although TAM reduces the organization of collagen fibers, in our study, as observed in the SEM of the bone/implant interface, parallel conformation of the collagen fibers/fibrin were present in all groups.

As observed by previous studies, the long‐term success of implants is directly related to the mineralization ³⁸ (Ca/P rate) of the peri‐implantar bone, as a more mineralized bone is able to withstand higher occlusal challenge. It is noted in the literature that TAM has a mineralizing effect non‐dependent of ER. ³⁹ In a previous study, we observed a higher rate of Calcium/P in the peri‐implant bone of rats treated with TAM. In this study, we observed that the Ca/P ratio in groups that received TAM, even when associated with CIS, did not present difference with the SHAM group. CIS has a harmful effect to the mineralization of the bone, ⁴⁰ including lowering the bone mass and increasing bone resorption, although, when in combination with TAM, the Ca/P rate of the implants were in conformation with the described by literature for stable implants (Ca/P ratios 1.30‐1.67). Interesting, implants with early loading, presented higher mineralization and density when compared with unloaded implants.

TAM may mitigate CIS‐induced bone impairment through coordinated regulation of oxidative stress and osteoclast–osteoblast signaling pathways. By engaging ERα in bone cells, TAM suppresses osteoclastogenesis and inhibits NF‐κB activation, a pathway strongly upregulated by CISand central to RANKL‐ and cytokine‐driven osteoclast differentiation. ⁴¹ , ⁴² , ⁴³ This ERα‐mediated effect aligns with its ability to counter CIS‐exacerbated oxidative stress, a condition known to trigger osteoblast apoptosis and impair matrix deposition and mineralization. ⁴⁴ , ⁴⁵ Additionally, TAM has been shown to activate ERα‐dependent Wnt/β‐catenin signaling, a pathway that promotes osteoblast differentiation and limits osteoclast commitment. ⁴⁶ , ⁴⁷ Its influence on the PI3K/Akt axis—crucial for osteoblast survival and balanced bone turnover, may further contribute to preserving bone homeostasis under chemotherapeutic insult. ⁴⁸ Collectively, these mechanisms provide a biologically coherent explanation for the protective skeletal effects observed in TAM‐treated animals exposed to CIS.

Recent literature also highlights additional pharmacologic and biomaterial strategies capable of modulating osseointegration under challenging systemic conditions. Anti‐sclerostin agents, by neutralizing osteocyte‐derived sclerostin, release the inhibitory tone on the Wnt/β‐catenin pathway and enhance osteoblast activity and bone remodeling around implants. ⁴⁹ Conversely, bisphosphonates—despite inducing osteoclast apoptosis and suppressing resorption—have been associated with elevated implant failure rates, with a systematic review reporting a mean failure of 49.96% irrespective of drug generation. ⁵⁰ , ⁵¹ From a biomaterial standpoint, nanometric hydroxyapatite coatings combined with recombinant human bone morphogenetic protein‐7 (rhBMP‐7) significantly increase early bone formation and improve osseointegration dynamics compared with HA coatings alone. ⁵²

Some aspects of the experimental design should be taken into account as limitations when interpreting the results. The study employed a single cycle of CIS, a regimen frequently used in preclinical models to investigate the early skeletal effects of antineoplastic agents. ¹² , ³¹ It is worth mentioning that repeated doses of anticancer chemotherapy might be necessary to achieve cancer remission/control in a clinical scenario and recurrent chemotherapy could produce accumulative effects. ⁵³ Importantly, circulating levels of CIS were identified up to 20 years after treatment in plasma concentration of 64.9 pg/g in the study by Gietema et al. (2000), ⁵⁴ where a control group had the concentration lower than 6pg/g. Hence, the cumulative effect and long‐term post‐treatment levels of both medications need to be further accessed in the peri‐implantar bone of antineoplastic treated rats. ⁵⁴

Likewise, the tibial implant model, widely adopted for evaluating systemic influences on osseointegration ⁵⁵ and peri‐implant remodeling ¹⁴ , ⁵⁶ does not reproduce the functional loading environment normally experienced by dental implants. The experimental setting was also free of microbial challenge, as is typical of controlled investigations focused on bone healing, ⁵⁷ , ⁵⁸ which naturally limits extrapolation to clinical situations in which inflammation and biofilm play a significant role. Finally, the study focused on histological, histometric, and immunohistochemical outcomes, without biomechanical or pharmacokinetic assessments.

From a clinical perspective, the present model was intentionally designed to emulate a common scenario in oncology: patients who already have osseointegrated implants and subsequently begin chemotherapy. As peri‐implant bone remodeling remains active long after osseointegration is achieved, and is particularly dynamic within the first months due to surgical micro‐damage and the regional acceleratory phenomenon, ⁵⁹ , ⁶⁰ changes induced by systemic therapies during this phase may influence long‐term peri‐implant homeostasis. The choice of the ovariectomized rat model further strengthens this clinical parallel, as it simulates the postmenopausal estrogen‐deficient state, the primary demographic for patients undergoing TAM therapy. ⁶¹ , ⁶² , ⁶³ Since OVX in rats effectively reproduces the physiological and skeletal changes observed during human menopause, ⁶⁴ , ⁶⁵ this framework provides a robust and validated approach to investigate how CISand TAM influence bone maintenance in a scenario that closely mirrors clinical reality. This rationale follows previous studies that also evaluated systemic disturbances after osseointegration rather than during the early healing phase. ¹⁴ , ¹⁵

Within this framework, we suggest that, although CIS has a negative effect on bone metabolism, TAM has a positive effect attenuating the disruption caused by CIS, suggesting the safety of maintaining an already osseointegrated implant during treatment with CIS and TAM combined. Therefore, a history of CIS‐based chemotherapy should not be viewed as an absolute barrier to oral rehabilitation. These results indicate that dental implant placement in patients undergoing such oncological treatments is a more viable option than previously recognized, provided that the patient's systemic status and drug interactions are carefully managed. However, we strongly encourage further studies assessing the functionality of the titanium implant and the response to bacterial challenges in the remodeling phase of the peri‐implant bone.

5. CONCLUSION

Based on the results of this experiment, our results do not support the hypothesis that TAM negatively affects peri‐implant bone when administered together with CIS. Within the study's limits, these findings support the clinical feasibility of dental implants in patients undergoing such therapies, suggesting that TAM may serve as a protective agent that enhances the predictability of osseointegration in oncological scenarios.

AUTHOR CONTRIBUTIONS

All the authors contributed to the conception and design of the study. Conceptualization: J.M., C.D., and E.E.; data curation: L.G.F., H.R., F.A.S., G.C.S., and R.D.B.; technical processing: E.O.S., G.C.S., and L.G.F.; writing—original draft: L.G.F. and H.R.; writing—review and editing: J.M., E.E., F.A.S., and C.D.; supervision: J.M. and C.D. All authors have read and agreed to the published version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.