Abstract
BACKGROUND:
Zinc oxide eugenol (ZOE) cement is a popular dental material due mainly to its analgesic, antibacterial and anti-inflammatory effects. The formulation of ZOE cement from nano particle-sized zinc oxide (ZnO) has the potential to increase these properties as well as reduce its adverse effects to the surrounding tissues.
OBJECTIVE:
This study evaluated the subcutaneous tissue response towards nano ZOE cements (ZOE-A and ZOE-B) in comparison to conventional ZOE (ZOE-K).
METHODS:
Test materials were implanted into 15 New Zealand white rabbits. Tissue samples were obtained after 7, 14, and 30 days (n = 5 per period) for histopathological evaluation of inflammatory cell infiltrate, fibrous tissue condensation, and abscess formation.
RESULTS:
ZOE-A showed the lowest score for the variable macrophage and lymphocyte at day 7. Both ZOE-A and ZOE-B presented lower fibrous tissue condensation and abscess formation compared to conventional ZOE-K. By day 30, ZOE-A exhibited less lymphocytic and neutrophilic infiltrate compared to the other materials, while ZOE-B had the lowest score for macrophages. ZOE-K exerted higher inflammatory cell response at almost all of the experimental periods. All of the materials resulted in thin fiber condensation after 30 days.
CONCLUSIONS:
Rabbit tissue implanted with ZOE-A and ZOE-B showed better response compared to ZOE-K.
Introduction
Zinc oxide eugenol (ZOE) is commonly used in clinical dentistry as a base, liner, temporary filling, root canal sealer, gingival dressing, and impression material [1–4]. It is widely accepted for its biocompatibility with connective tissue due to its neutral pH, and good chemical properties such as its analgesic, anti-inflammatory, sedative, and antibacterial effects [1,5]. These advantages make it a popular temporary dental filling material, but unsuitable as a permanent restoration as it is notably weak, highly soluble, and cannot withstand high occlusal forces [6]. It also has the potential to irritate the dental pulp as eugenol has the ability to leach out and diffuse through the dentinal tubules [7,8], rendering it unsuitable as a direct pulp capping agent.
Application of nanotechnology in material development has the potential to improve mechanical strength, chemical stability, durability, and flexibility of the materials [9]. There is also growing application of nanotechnology in dental materials development over the years, especially in the formulation of nanocomposites, antimicrobial nanomaterials, nano bio-mineralization, and nano coatings [10]. One such particle is in the form of nano ZnO (NZnO), which popular for its antibacterial, antifungal, electrical, chemical, optical, and potentially biocompatible properties [11]. Several studies investigating the performance of nano ZOE as a root-canal sealant material have described its good biocompatibility and improved bacteriostatic activity, physicochemical properties, and sealing ability [12–16]. Inclusion of NZnO in dental luting cement has also reported improved antibacterial, mechanical, acid resistance property and biocompatibility of the material [17].
Two types of NZnO (NZnO-A and NZnO-B) have been previously synthesized by the catalyst-free combust oxidized method, using a custom-made zinc boiling furnace [18]. The furnace was heated at temperatures of 1000–1500 °C to melt and vaporize the Zn metal. The Zn vapor which is pressurized inside the crucible becomes oxidized into ZnO when exposed to ambient air. These particles were morphologically different, whereby NZnO-A is rod-like, measuring between 51 to 60 nm, while NZnO-B is plate-like, ranging from 71 to 80 nm in size [19]. Mixed with eugenol, these NZnOs produce ZOE-A and ZOE-B dental cements. These materials were reported to be less cytotoxic and have improved compressive strength when compared to conventional ZOE [19–21]. However, ZOE has the potential to leach through dentinal tubules to the dental pulp and result in pulp inflammation. Since the tissue response in connective tissues presents with similar characteristics, the implantation and assessment of test material in subcutaneous tissue may provide preliminary data for evaluation of its biocompatibility [22–24]. This study aimed to evaluate the subcutaneous tissue responses elicited by ZOE-A and ZOE-B in comparison to a commercial ZOE (ZOE-K), in an in-vivo rabbit model.
Materials and methods
Animal
This study was approved by the Animal Research and Ethics Committee of Universiti Sains Malaysia, Health Campus, Kelantan, Malaysia (No. 2015/96/634). Fifteen, 4 to 5-months-old male New Zealand white rabbits weighing 2500 g to 3000 g were divided into three experimental periods of 7, 14, and 30 days, with five rabbits randomly assigned to each group (n = 5, per group). Before the procedure, the rabbits were quarantined for 2 weeks at the Animal Research & Service Centre (ARASC), USM. They were kept separately in cages, fed a standard rabbit diet, and given water ad libitum.
Preparation of test materials
The test materials were prepared prior to the implantation procedure by combining powder NZnO with liquid eugenol derived from the commercial ZOE (ZOE-K), to form ZOE-A and ZOE-B cements. The mixture was in accordance with the manufacturer’s instructions for ZOE-K, at a powder to liquid ratio of 4:1. The mixed materials were inserted into an autoclaved metal mold measuring 6 mm in length and 3 mm in diameter, to set at room temperature. The materials were then removed and disinfected by immersing in 2% chlorhexidine for 1 minute [25], rinsed in saline, dried with sterilized gauze, and stored in a sterile universal container prior to implantation. ZOE-K, which served as the control material, was also prepared in the same manner.
Implantation procedure
Each rabbit was implanted with all of the materials at the upper left abdomen (ZOE-A), upper right abdomen (ZOE-B), and lower abdomen (ZOE-K) (Fig. 1a). The animals were anesthetized with an intramuscular injection of ketamine in combination with xylazine at doses of 35 mg/kg and 5 mg/kg body weight, respectively, prior to surgery. The incision sites were shaved, cleaned, and disinfected. Three 1-cm-long incisions spaced at least 2-cm apart were made. Two incisions were made at the upper left and right abdominal region, and one at the lower portion of the abdomen. Using blunt-tipped scissors, lateral tearing of the subcutaneous tissue was performed, providing three surgical cavities at the aforementioned sites. The specimens were inserted with pliers into the surgical area under aseptic conditions, parallel to the incision line. The incisions were sutured with 3/0 Silk (Black Braided Silk; UNIK Surgical Sutures Mfg., Taiwan). A sterile adherent, plasticized dressing was applied, and baby clothing was fitted to protect the surgical sites. An intramuscular injection of meloxicam (Melonex Injection; Intas Pharmaceuticals Ltd., India) at a dose of 0.1 mg/kg was administrated for one-day post-operatively while an intramuscular injection of enrofloxacin (Baytril 5% injectable solution; Bayern, Germany) at a dose of 0.1 ml/kg was given for 5 days. The rabbits were returned to their cages and monitored regularly for their activity, feeding, habits, color of coats, and the presence of nasal secretions. Suture removal was performed one-week post-operatively.

Illustrations depicting site of implant placement on rabbit abdomen (a), and the tissue surfaces (A–D) evaluated for the histopathological scoring (b).
At the end of each experimental period (7, 14, and 30 days), the animals were euthanized by an overdose of intravenous sodium pentobarbital (Dorminal 20% injection solution; Alfasan, Holland). The surgical area was shaved and marked, and an excisional biopsy of the implant areas was performed with a safety margin of 1 cm. The excised tissues were then fixed in 10% formaldehyde. The implanted materials were removed, and the formalin-fixed tissue specimens were bisected longitudinally. The tissue then underwent a series of tissue processing (tissue processor; Thermo Scientific, UK) and embedded into paraffin blocks. Samples were sectioned using rotary microtome HM 355 (MICROM, International GmbH, Germany) with 5 μm thickness for each section. The selected slides were dewaxed over a Lab-line slide warmer (Branstead, USA), stained with nuclear dye haematoxylin (Merck, Germany), rinsed, counter-stained with eosin (Merck, Germany), dehydrated, and mounted with DPX new non-aqueous mounting medium (Merck, Germany). The stained slides were dried and viewed using a light microscope for screening.
Histopathological evaluation
The hematoxylin and eosin (H&E) stained sections were scanned using a digital scanner (MIRAX Scanner; Carl Zeiss, Germany) and examined using slide viewer software (MIRAX viewer) in a computer by a single, blinded calibrated examiner. The slides were analyzed according to the criteria described by Figueiredo et al. [26] and adapted from previous studies [27–29]. Histopathological tissue response was assessed based on the inflammatory infiltrate, fibrous tissue condensation, and abscess formation.
The inflammatory infiltrate was determined by the presence of lymphocytes/plasma cells, macrophages, neutrophils, and multinucleated giant cells, and scored according to (1) Absent: cell absent or within vessels, (2) Mild: cells present, but sparse or in reduced clusters, (3) Moderate: cell present, yet not dominating the microscopic field, and (4) Intense: cell present in the form of infiltrate close to the material used. Fibrous tissue condensation was scored based on: (1) absence of collagen fibers, (2) presence of a thin layer of collagen fibers, and (3) presence of a thick layer of collagen fibers. Abscess formation was classified according to the following criteria: (1) absence of abscess, (2) presence of abscess in contact with the surgical cavity where the material had been inserted, and (3) presence of abscess in areas distant from the surgical cavity where the material had been inserted.
The tissue specimens mostly exhibited a rectangular central cavity formed by the implant material. The histopathological tissue responses were assessed from all connective tissue surfaces that were in contact with the implant material (Fig. 1b). These surfaces (A-D) were scored, and the mean score from all fields [(A+B+C+D)/4)] was rounded to the closest categorical parameter.
SPSS version 23 software (IBM Corp., Armonk, NY, USA) was used for data analysis. The scorings for different types of materials and experimental periods were compared using Kruskal–Wallis test followed by Mann-Whitney U test for cases where statistically significant differences were detected. The level of significance was set at p < 0.05.
Results
During the histological screening, eight samples had to be excluded due to exposure of the implant material (Table 1). Comparison of the study results between different materials and experimental periods is summarized in Fig. 2. Although tissue responses varied for the test materials, statistically significant findings were only observed for the macrophage parameter.
Summary of samples included and excluded for histopathological analysis
Summary of samples included and excluded for histopathological analysis

A comparison among the study groups and between experimental periods for variables lymphocyte/plasma cell, macrophage, neutrophil, multinucleated giant cell, fibrous tissue condensation, and abscess. Statistically significant differences were detected among materials within the same experimental period (§ p < 0.05), and among the same material at different time intervals († p < 0.05).
At day 7, a mild inflammatory cell infiltrate was observed for ZOE-A, while ZOE-B and ZOE-K elicited a mild to moderate inflammatory cell response (Fig. 3). ZOE-A also showed significantly lowest score for the variable macrophage at this experimental period (p = 0.040). ZOE-A and ZOE-B presented with lower fibrous tissue condensation and abscess scores compared to the control material. At day 14, the degree of inflammatory response was similar to day 7 (Fig. 4). However, it was noted that neutrophils, fibrous tissue condensation, and abscess scores for ZOE-K was higher than the test materials.

Photomicrographs of representative histological samples at day 7. (a) ZOE-A: mild inflammatory cell infiltrate (∗), thin fibrous tissue condensation (black arrow), and foci of abscess (+). (b) ZOE-B: moderate inflammatory cell infiltrate (∗), and thin fibrous tissue condensation (black arrow). (c, d) ZOE-K: moderate inflammatory cell infiltrate (∗), thin fibrous tissue condensation (black arrow), and abscess (+) (H&E, ×200).

Photomicrographs of representative histological samples at day 14. (a) ZOE-A: mild inflammatory cell infiltrate (∗) and thin fibrous tissue condensation (black arrow). (b) ZOE-A: moderate inflammatory cell infiltrate (∗) and abscess (
At the end of day 30, a mild to moderate inflammatory cell infiltrate was observed (Fig. 5). Macrophage score for ZOE-B was significantly lower than that of ZOE-A and ZOE-K (p = 0.026). ZOE-A exhibited significantly increased macrophage score compared to day 7 and day 14 of its implantation (p = 0.012). Although not statistically significant, lymphoplasmacytic infiltrate score was slightly decreased for all materials compared to day 14, while the reverse was observed for neutrophils. Focal presence of multinucleated giant cells was also noted in ZOE-B and ZOE-K samples. Fibrous tissue condensation was also similar between test materials, and with day 14. Abscess formation was still present by day 60 and was slightly higher in ZOE-B compared to other materials.

Photomicrographs of representative histological samples at day 30. (a) ZOE-A: moderate inflammatory cell infiltrate (∗) and abscess (
ZOE material is widely used in various dental specialties, particularly owing to its ease of use, as well as its analgesic, anti-inflammatory, and antibacterial attributes [6]. As with other dental materials, it has also been linked to genotoxicity and cytotoxicity [30–32]. Toxicity of ZOE sealers have been ascribed to the release of eugenol from the set material [31,33]. In in vitro settings, eugenol is reportedly toxic to human dental pulp fibroblasts of primary teeth, even at low concentrations [34]. Although the eugenol itself exerted cytotoxic effect on human pulp cells, it was found to be non-genotoxic in an earlier study [35]. However, later in vitro and in vivo genotoxicity works on ZOE sealers reported otherwise [30,36]. A recent cytotoxicity study on human gingival fibroblasts suggested that ZOE leaching, comprising both Zn ions and eugenol, were responsible for generating reactive oxygen species (ROS), resulting in oxidative damage to cells. This effect was also greater with more prolonged exposure [33]. Although these experimental results were unfavourable towards ZOE, it has reportedly performed well during usage tests and in clinical applications [37]. In the clinical setting for example, the cement does not necessarily come into direct contact with pulp tissue when used as a temporary restoration. The buffering ability of saliva in the oral environment may also play a role in alleviating the potential cytotoxic effects of the material [6]. Additionally, the degree of cytotoxicity decreases as the material sets [31].
This study demonstrated the response of subcutaneous tissue surrounding implanted ZOE material. This method increases the tissue contact area with the material and allows for more tissue to be assessed. However, the inflammatory response could be significantly higher as the tissue is exposed to more foreign material in comparison to polyethylene tubes usage, as described in other in vivo studies [28,29]. This could also contribute to the exposure of implant materials as seen in our study. In the previous studies utilizing polyethylene tubes, endodontic sealer material is inserted into a polyethylene tube which is then implanted subcutaneously, thus exposing only a small area of the material to the connective tissue environment. Comparison between direct submucous injection of endodontic material and subcutaneous implantation of polyethylene tube, however, has reported no significant difference between the methods [26].
The experimental periods in this study were similar with past studies which have reported various intervals of 2, 7, 14, 28, 30, 60, and 90 days [16,22,26–29,38]. In general, a minimum of 30 days observation is recommended for biomaterial testing of permanent medical devices or materials [39]. The selected time intervals enable the monitoring of the impact at passage of time for the assessed materials.
When test materials are implanted to living tissue, tissue response towards the materials will be initiated. This includes inflammation, wound healing, foreign body reaction, and fibrous encapsulation [39]. Inflammatory response can be divided into: (1) acute, which takes place within minutes to hours, and (2) chronic, which has a slower and longer onset [40]. Neutrophils are the primary mediators of acute inflammation, and their presence are followed by eosinophils, basophils, macrophages, lymphocytes, and plasma cells [41]. The presence of neutrophils together with plasma protein exudation is a characteristic of acute phase inflammation. This takes place shortly after tissue injury, lasting from minutes to a few days, before progressing to the chronic phase, which is characterized by replacement of neutrophils with macrophages and lymphocytes, vascular proliferation, and fibrosis [40]. In general, acute inflammation and chronic inflammation will be resolved after 2 weeks. Fibrous tissue condensation and abscess will occur, and foreign body giant cells can be detected after this period [40]. The inflammatory cells infiltrate observed in our study persisted from day 7 to day 30. This can be attributed to the persistent irritation from the implanted materials.
All of the materials in our study resulted in prolonged presence of neutrophils which seemed to increase within the range of mild to moderate infiltration by day 30. This could be attributed to the persistent presence of the implant materials as mentioned previously. Many forms of chronic inflammation continue to present with neutrophils as a result of persistent microbes or necrotic cells, or mediators elaborated by macrophages [40]. In the periapical inflammatory diseases for example, chronic periapical lesions may undergo acute exacerbation if the infected tooth remains untreated, with an imbalance in equilibrium between the pathogens and the host response to contain it [42]. By day 30, higher neutrophil score was observed for ZOE-K compared to the test nanomaterials, suggesting its higher potential to cause irritation.
Lymphocytes and macrophages produce co-stimulators and cytokines from their bidirectional interactions, leading to a chronic, severe inflammatory response. They can be detected from 2 days until after several weeks of the injurious event [40]. By day 30, lymphocyte score was shown to be slightly reduced for all materials. Lymphocyte score was lowest in ZOE-A, especially by day 30, with ZOE-K producing more intense lymphocytic reaction compared to nanomaterials. Others have reported a similar trend of lymphocyte score for some endodontic materials tested in their studies [27–29].
The presence of macrophage and multinucleated giant cells are consistent with the presence of non-degradable material. Macrophages play a phagocytic in the process of removing tissue debris and foreign material. Presence of persistent pathogens or foreign materials that are difficult to degrade result in fusion of activated macrophages to form multinucleated giant cells [27,39,40,43]. This was evident by the presence of increasing macrophage infiltrate at day 30 of our study and several other reports, [27,29] except for ZOE-B which showed a significantly lower score. Multinucleated giant cells were only observed in ZOE-B and ZOE-K samples at day 30, in which the latter showed greater number. Previous studies have reported that giant cells were more likely to be detected in response to tissue implanted with ZOE-based materials compared to epoxy resin- and methacrylate resin-based ones [27,29].
Eosinophils are typically associated with hypersensitive reactions or parasitic infections [27,40]. In this study, eosinophils were detected at very rare locations and were not adequate to be represented in the final data. This was consistent with previous studies [44,45], although some others were able to report their presence, albeit in reduced numbers [27,38].
There was generally thin fibrous tissue condensation in response to the materials. Although a similar score was observed for all three materials at day 30, ZOE-K exhibited slightly higher scores at day 7 and day 14. A previous study reported that nano ZOE sealers showed denser fibrous tissue condensation at experimental periods of 15 and 30 days but became less dense at longer intervals of 60 days [16]. This fibrous encapsulation is a defense response in containing the inflammatory reaction towards the foreign material, preventing further insult to the surrounding tissue [46]. Abscess formation at areas in contact with surgical site was observed from day 7 until day 30 for all materials. Although ZOE-K score at day 30 was similar to ZOE-A, the former exhibited the highest scores at day 7 and day 14. The abscess formation is consistent with the presence of neutrophilic infiltrates, and indicative of the persistent tissue irritation elicited by the test materials.
Although most of the inflammatory cell parameters showed no significant difference between the groups, a slightly higher number of chronic inflammatory cells were observed in the ZOE-K group which also presented with more prolonged acute inflammation. These findings are similar to that of another study which reported lower degree of subcutaneous connective tissue responses towards nano ZOE endodontic sealers in comparison to other ZOE-based and epoxy resin-based sealers [16]. Previous cytotoxicity studies on human gingival fibroblasts have also reported higher viability of cells treated with diluted nano ZOE in comparison to conventional ZOE [20,33].
The better cytological and tissue response towards nano ZOE may be attributed to their size and morphology. NZnO-A (51 to 60 nm, rod-like) has the smallest particle size, followed by NZnO-B (71 to 80 nm, plate-like), and ZnO (larger than 100 nm, non-homogenous) [19,20,47,48]. Therefore, NZnO provides a bigger surface-to-volume ratio and provide larger surface area for interaction when mixed with liquid eugenol. This results in a stronger form of zinc eugenolate due to the stronger interlock bond between NZnO and eugenol and reduces ZOE leachings. The larger-sized ZnO particles, however, provide smaller surface area for interaction when mixed with eugenol. This results in more leaching, contributing to ROS generation, and more cytotoxic effects [20,49,50].
Conclusion
Although none of the materials exhibited ideal tissue responses, ZOE-A and ZOE-B presented with better outcomes than ZOE-K. This supports the potential role of these nano ZOE cements as a temporary filling material that are comparable to the conventional ZOE.
Footnotes
Acknowledgements
The authors gratefully acknowledge the Research University Individual Grant (RUI) Universiti Sains Malaysia [1001/PPSG/812160], Craniofacial Science Laboratory, School of Dental Sciences and Animal Research and Service Centre (ARASC), Universiti Sains Malaysia, Health Campus.
Conflict of interest
The authors declare no competing interests.
