JOURNAL OF ENDODONTICS Copyright 2003 by The American Association of Endodontists

Printed in U.S.A. VOL. 29, NO. 4, APRIL 2003

The Comparative Sealing Ability of Hydroxyapatite Cement, Mineral Trioxide Aggregate, and Super Ethoxybenzoic Acid as Root-End Filling Materials
C. Mangin, DMD, C. Yesilsoy, DMD, MS, R. Nissan, DMD, and R. Stevens, DDS, MS

A comparison was made of the ability of hydroxyapatite cement, mineral trioxide aggregate, and super ethoxybenzoic acid to prevent the leakage of bacteria from root canals, when used as root-end filling materials. The materials were tested in a double-chamber device in which a root segment connects the upper (delivery) chamber and the lower (receiving) chamber. The root segment was prepared by having the root canal instrumented to a #45 file, and a 3-mm deep, root-end preparation placed at the apical foramen. The canal of each root segment was filled with gutta-percha, and the root-end preparation was filled with one of three test materials, mixed according to the manufacturers directions. Negative controls were constructed with sticky wax sealing the apical foramen. A titered suspension of radioactively (3Hthymidine)-labeled bacteria (Enterococcus fecalis) was placed into the delivery chamber, and sterile saline was placed into the receiving chamber such that the apical third of each root section was immersed. At various time points, samples were taken from the receiving chamber, and measured for 3H activity. The results indicated that (a) all the test materials leaked significantly compared with the negative controls; and (b) there was no significant difference found between the leakage rates of the three materials tested.

As endodontists, part of our obligation to our profession is to constantly search for new materials and/or techniques that would improve the quality and success of our treatment. It has come to our attention that a hydroxyapatite cement (HAC), trade name Bone Source (Leibinger, Kalamazoo, MI, U.S.A.), has been used frequently during the past several years in certain medical applications. This material, developed in 1986 (1), is applied intraoperatively as a paste, which then sets to a stable implant composed of microporous hydroxyapatite in approximately 15 to 20 min (2).
261

Its primary use in medicine has been for the reconstruction of nonstress-bearing cranial skeletal defects; however, its properties suggest a potential use as an endodontic retrofill material. A useful endodontic retrofill material will possess numerous characteristics. These include: radiopacity, ease of manipulation, biocompatibility, antibacterial activity, cost-effectiveness, and the ability to establish an effective apical seal. HAC cement is composed of tetracalcium phosphate and dicalcium phosphate reactants (3). These compounds, when mixed with water, react isothermally to form a solid implant composed of carbonated hydroxyapatite (3). Its compressive strength is greater than 60 MPa. In comparison, mineral trioxide aggregate (MTA, see below) has a compressive strength of 70 MPa (3). HAC has demonstrated excellent biocompatibility, does not cause a sustained inflammatory response or toxic reaction, and has been shown to maintain its shape and volume over time (3). HAC implants are resorbed slowly and are replaced by natural bone in an approximate 1:1 ratio in an osteoconductive manner. It is claimed that HAC limits the penetration of bacteria into the micropores of the implant, which are about 6 in size (3). For the HAC to set properly (1520 min), the surgical field must be kept relatively dry (3). However, the ability of HAC to form a bacteria-tight seal has not been evaluated. Numerous investigators have examined leakage properties of different retrofill materials using a variety of techniques. In 1994, Torabinejad et al. (4) examined the dye leakage of amalgam, super ethoxybenzoic acid (sEBA), intermediate restorative material (IRM), and MTA as root-end fillings in both the presence and absence of blood. Teeth were immersed in dye for 72 h, and the extent of dye penetration was examined. The study concluded that MTA leaked significantly less than the other materials tested both with and without blood contamination. Fischer et al. (5) studied the bacterial leakage of MTA, amalgam, IRM, and sEBA as root-end filling materials. The method used was to determine the time needed for Serratia marcescens to penetrate a 3-mm thickness of retrofill material. Samples were monitored four times per week. Amalgam, IRM, sEBA, and MTA leaked, respectively, in 10 to 63 days, 28 to 91 days, 42 to 101 days, and 49 days. Additionally, four of the MTA samples did not exhibited any leakage. Kos et al. (6) used an in vitro model system to evaluate the ability of poly-HEMA, gutta-percha, and amalgam to form a fluidtight seal of the root canal space to prevent bacterial (Proteus

262

Mangin et al.

Journal of Endodontics

vulgaris and Streptococcus salivarius) migration. They found that poly-HEMA models did not leak after 38 days; however, amalgam and gutta-percha samples leaked as early as 2 days. In 1990 King et al. (7) evaluated the seal of amalgam, guttapercha, sEBA, and glass ionomer as endodontic retrofillings. Microleakage was measured at 24 h, 1, 2, and 3 weeks and 1, 2, and 3 months by using a fluid filtration technique. These results indicated that sEBA provided an excellent apical seal that did not deteriorate over a 3-month period. Torabinejad et al. (8) studied the bacterial leakage of MTA as a root-end filling material. In this study, the time needed for Staphylococcus epidermidis to penetrate a 3-mm thickness of amalgam, sEBA, IRM, or MTA as root-end fillings was determined. MTA samples did not show any leakage throughout the 90-day period. Amalgam leaked at 28.5 days, sEBA at 34 days, and IRM at 15 days. A comparison of MTA, sEBA, composite, and amalgam as root-end fillings was made by Adamo et al. (9). This study used a bacterial microleakage model and found no statistically significant differences in the rate of microleakage among five groups tested either at 4, 8, or 12 weeks. The median leakage time for all groups was 35 days. A tricalcium phosphate cement (G-5) was examined by MacDonald et al. (10). The tricalcium phosphate cement was compared with sEBA and amalgam as root-end filling materials by using fluid filtration and dye penetration techniques. Samples were tested at 6 and 24 h, 1 week, 1 month, and 3 months. The results indicated that apatite cement provided a seal comparable to amalgam and sEBA. The purpose of this study was to determine whether HAC has a sealing ability that is comparable to commonly used, well-studied retrofill materials.

FIG 1. Apparatus used for measuring leakage of retrofill materials.

MATERIALS AND METHODS Forty, single-rooted, extracted, human, anterior teeth with straight canals were placed into a 10% formalin solution. Radiographs were taken to determine canal anatomy in a mesiodistal plane. Crowns were removed at the CEJ with a #56 carbide fissure bur in a water-cooled, high-speed handpiece. The root sections were instrumented up to a master apical size of #45 in a crowndown fashion using ISO sized nickel-titanium rotary files. The canals were irrigated with 2.5% NaOCl during instrumentation. The apical 3 mm was resected at 90 degrees to the long axis of the tooth, again using a carbide fissure bur and a high-speed with water. A 3-mm deep, root-end preparation was made using a Neosonic P5 ultrasonic unit using the long axis of the canal as a guide. The prepared root sections were placed back into the formalin solution until further use. Enterococcus fecalis 29212 was grown in 5 ml of brain-heart infusion (BHI) broth for 24 h at 37C. One-half milliliter of this culture was used to inoculate 100 ml of BHI broth containing [methyl-3H] Thymidine (2 Ci/ml). The culture was incubated for 20 h and then harvested by centrifugation (16,300 g, 15 min), washed three times with a buffered KCl solution (2 mM potassiumphosphate buffer, pH 6.8, 1 mM CaCl2, and 0.05 M KCl), and finally suspended in 50 ml of buffered KCl. Fifty microliters from the final labeled cell suspension and each wash were placed into liquid scintillation vials along with 10 ml of scintillation cocktail solution (Eco-Lite) to determine baseline 3H activity counts.

Growth and labeling conditions were taken from Applebaum et al. (11). Serial 101 dilutions of the labeled bacteria suspension were plated onto BHI agar plates. Colonies were counted after overnight incubation at 37C to determine the bacterial titer. A double-chamber device modified from Trope and Nissan (12) was constructed using 2.0-ml microcentrifuge tubes and 15-ml plastic test tubes (Fig. 1). The 2.0-ml microcentrifuge tube acted as the upper delivery chamber (Fig. 1, #1) while the 15-ml tube acted as the receiving chamber (Fig. 1, #2). The lower portion of chamber #1 was then cut with a scalpel so that a prepared root section could be placed into the tube and it would remain flush with the end of the tube. The microcentrifuge tube, containing the prepared root section, was then placed into a hole cut through the cap of the plastic test tube. The junctions throughout all devices were secured with sticky wax to prevent leakage. Thirty of the prepared root sections were divided into 3 groups of 10 teeth each. Each group was filled with one of the retrofill materials: MTA, HAC, or sEBA. The retropreparations were airdried with a syringe. Each of the materials was mixed according to manufactures directions (MTA mix powder and liquid into a thick, creamy consistency; HAC 0.25 ml sterile water/g Bone Source, sEBA 0.52 g powder/0.12 g liquid) and condensed against a premeasured gutta-percha cone fitted 3 mm from the apex. After the retrofills were condensed, the apices of all experimental root sections were placed onto a saline-soaked sponge and stored at 37C for 48 h. Five apparatuses were constructed with only a gutta-percha master cone and no retrofill, and these acted as positive controls (i.e., maximum leakage). Five apparatuses were constructed with sticky wax at the apical end to act as negative controls (i.e., no leakage). One-half milliliter of the labeled bacterial suspension was placed into each delivery chamber for all experimental and control groups. Twelve milliliters of isotonic saline was then placed into each receiving chamber so that the apical 13 of each tooth section was immersed into the solution. After slight agitation, a 50-l sample was then taken from each receiving chamber and placed into a scintillation vial, along with

Vol. 29, No. 4, April 2003

Sealing Ability of Hydroxyapatite Cement

263

FIG 2. Bacterial leakage of HAC, sEBA, and MTA. HAC (open square); sEBA (closed diamond); MTA (shaded triangle); positive control (); negative control (shaded square).

10 ml of Eco-Lite cocktail solution. Similarly, samples were taken form the receiving chamber at the following times: 24, 48, 96, 216, 264, 384, 432, and 576 h. After each 50-l sample was taken, 50 l of isotonic saline was placed back into chamber #2 so that the volume in the chamber remained constant. The radioactivity in each sample was then measured in a liquid scintillation counter. The radioactive counts were converted to colony-forming units (CFU) based upon the titer and radioactivity of the original labeled bacterial suspension. From our titration, we found 1 107 CFU per 50 l (2 108 CFU/ml) in the bacterial suspension. The 3H activity of 1 ml of the bacterial cell suspension was found to be 2.54 105 CPM. Therefore, the number of CFUs per 3H CPM was 7.87 102 (2 108 CFU/ml 2.54 105 CPM/ml). To calculate the total CFUs in the receiving chamber of the test apparatus, the 3 H activity in a 50-l sample of the saline solution in the receiving chamber was multiplied by 1.88 105. This conversion factor took into account the CFUs per 3H CPM (7.87 102) and the conversion from 50 l in the sample to 12 ml in the receiving chamber (a 240-fold increase). A Kruskal-Wallis one-way ANOVA nonparametric test was used to determine whether the mean bacterial leakage differed significantly between the different materials. At each time period, the three materials (HAC, MTA, and sEBA) were compared. RESULTS The average number of bacteria that leaked in the 10 apparatuses for each test material, at each time period, was calculated (Fig. 2). Based on ANOVA, which tests the null hypothesis that k samples come from the same population, where k is at least 3, no significant differences were found between the leakage rates of MTA, HAC, and sEBA. The significance level of 0.05 was mod-

ified using the Bonferroni method, so that at each time period, the significance levels (shown in the SPSS output) that were obtained were compared to 0.05/8 (where 8 the number of time periods). This value was 0.00625. In all eight time periods, the significance level obtained was greater than 0.00625 and therefore not significant. DISCUSSION The purpose of this experiment was to compare, in vitro, the bacterial microleakage of two commonly used root-end filling materials with that of a new hydroxyapatite cement. From our statistical analysis, we conclude that there was no significant difference between the leakage observed for the three test materials when used as retrofills. In addition to the intrinsic properties of each of the test materials, other factors may potentially influence the observed leakage rates. These included: variations in the condensation of each of the test materials and differences in canal anatomy. The radioactively labeled bacteria used in this experiment offers numerous advantages over other techniques to assess leakage. First, the use of bacteria itself is clinically relevant. As endodontists, we routinely deal with microorganisms in infected root canals, and it is for this reason that we believe that the use of microorganisms is superior to other, less relevant indicators of leakage. Secondly, we can accurately quantitate the number of bacteria that have leaked through the system. Other leakage studies using bacteria have not been quantitative or have had to rely on bacterial growth to quantitate leakage. This poses additional complications such as concerns over bacterial contamination or the effect of the time of sampling on the determination of the number of microorganisms in the sample. Finally, determining the amount

264

Mangin et al.

Journal of Endodontics
2. Constantino P, Friedman C, Jones K, Chow L, Pelzer H, Sisson G. Hydroxyapatite cement. Arch Otolaryngol Head Neck Sx 1991;117:379 89. 3. Bone Source. Osteoconductive bone substitute material: information packet. 1999. 4. Torabinejad M, Higa R, McKendry D, Pitt Ford T. Dye leakage of four root end filling materials: effects of blood contamination. J Endodon 1994; 20:159 63. 5. Fischer E, Arens D, Miller C. Bacterial leakage of MTA as compared with zinc-free amalgam, IRM, and sEBA as a root-end filling material. J Endodon 1998;24:176 9. 6. Kos W, Aulozzi D, Gerstein H. A comparative bacterial microleakage study of retrofilling materials. J Endodon 1982;8:355 8. 7. King K, Anderson R, Pashley D, Pantera E. Longitudinal evaluation of the seal of endodontic retrofillings. J Endodon 1990;16:30710. 8. Torabinejad M, Rastegar A, Kettering J, Pitt Ford T. Bacterial leakage of MTA as a root end filling material. J Endodon 1995;21:109 17. 9. Adamo HL, Buruiana R, Schertzer L, Boylan RJ. A comparison of MTA, sEBA, composite and amalgams as root-end filling materials using a bacterial microleakage model. Int Endod J 1999;32:197203. 10. MacDonald A, Moore K, Newton, Brown C. Evaluation of an apatite cement as a root end filling material. J Endodon 1994;20:598 604. 11. Applebaum B, Golub E, Holt S, Rosan B. In vitro studies of dental plaque formation: adsorption of oral streptococcus to hydroxyapatite. Infect Immun 1979;25:71728. 12. Trope M, Nissan R. In vitro endotoxin penetration of coronally unsealed endodontically treated teeth. Endod Dent Traumatol 1995;11: 90 4.

of radioactive bacteria is simple and quick. The sampling process is not complex, yet it offers accurate and quantitative results. In conclusion, our results indicate that there is no significant difference in sealing ability between HAC, MTA, or sEBA. Further studies examining HAC are needed to completely evaluate its potential role in endodontics and dentistry as a whole.
The authors thank the following people for their support, skill, and invaluable time: Dr. A. Park, Dr. S. Rollison, Dr. A. Amaro, Dr. J. Duong, and Chris Reppert. The authors also express their appreciation to Mrs. R. Gorin for the statistical analysis of their data. Drs. Yesilsoy, Nissan, and Stevens are affiliated with the Department of Endodontology, Temple University School of Dentistry, Philadelphia, PA. Dr. Mangin was a former graduate student in the Department of Endontology, Temple University School of Dentistry; he is currently in private practice. Address requests for reprints to Dr. Roy H. Stevens, Dept. of Endodontology, 3223 North Broad Street, Philadelphia, PA 19140.

References 1. Brown WE, Chow LC. A new calcium phosphate, water-setting cement. In: Brown PW, ed. Cements research progress. Westerville, OH: American Ceramic Society, 1986:35279.