Heat Analysis of Different Devices for Thermo-explantation of Dental Implants: A Numeric Analysis and Preclinical In Vitro Model

Each year, more than 15 million dental implants are inserted worldwide.^1,2  Osseointegration is defined as a connection between vital bone cells and blood components with the implant surface.³  If no relevant influences occur, this connection is permanent. There is no technically predictable way to reverse the growth of bone cells on the implant surface.⁴  In cases of peri-implantitis and subsequent implant explantation, surgery is usually related to severe bone loss.⁵  In some cases, implants cannot be easily extracted.⁵  Implants with severe bone resorption are usually much simpler to remove than implants that are completely integrated into the bone, such as implants with incorrect positioning.⁶  Current ratchet techniques use very high forces, and in many cases, surgeons have to use burs.⁷  Traumatic implant removal further complicates new implant rehabilitation.

Thermal bone damage during surgical interventions is an undesirable side effect. It has been reported that high-frequency surgical equipment and dental lasers lead to unwanted thermal damage of peri-implant bone and may result in implant loss.^8,9  Furthermore, such uncontrolled heating may lead to severe inflammation and jaw necrosis.⁹  However, there are several publications on the successful loosening of osseointegrated implants using high temperatures with, for example, ultra-high-frequency surgical devices.^10,11  Thermal treatments may reverse the osseointegration of implants and preserve valuable bone tissue. The authors of one study used a CO₂ laser as a thermal device for implant removal.¹²  In all cases, the heat input was intentional but uncontrolled and uneven. The risk of bone necrosis seemed high as the implants were heated without any temperature control.^13–15  Therefore, these experimental procedures are currently neither suitable nor approved for clinical applications. The authors asked themselves the question whether a controlled and calculated heating of dental implants up to a threshold value is possible. Ideally, for thermo-explantation, the implant surface should show a homogeneously increased temperature distribution along the entire implant body. According to a systematic review for warm temperatures that ranged between 47°C and 55°C for 1 minute and cold stimuli, no specific threshold value for bone necrosis is available.^16,17  There is a lack of literature related to whether it is possible to heat an implant to an even threshold level; there is also lack of information regarding which specific devices represent the best thermal source. Dental implants today are generally cylindrical or conical threaded screws inserted into the jawbone. The authors hypothesized that there could be a device that leads to an even and controlled heating process. The aim of this preclinical pilot study was to lay a foundation for future in vitro and in vivo studies whose authors may explore the potential of thermal implant removals. Furthermore, different thermal devices and their effects on predictable and uniform distributed heating up to the implant tips were evaluated. The primary aim was to identify suitable sources for implant heating. An additional aim was to assess parameters that affect the homogenous implant temperature and to investigate a controlled heating process up to 47°C.

This empirical in vitro study standardized infrared (IR) measurements were conducted to evaluate the surface temperature distribution along the implant axis using 4 different thermal sources. The methodology of this in vitro investigation was reviewed by an independent statistician.

Here, m is the mass of the implant, c_p is the specific heat capacity, T₀ is the initial temperature of the implant, is the heat output, and are the heat losses.

For the case of 1-dimensional transient heat conduction, error functions can be used to calculate the necessary heating power and heating time to reach a specific temperature (Figure 1). During a low-heating power, the heating time increases disproportionately because of heat losses of the implant.

Because the manufacturer sets the dimensions and materials of the implants, only the parameters of implant temperature, heating power and duration of heating, could be varied. The heat losses of the implant are difficult to estimate in advance because of the different jaw and blood circulation structures. Furthermore, the heating method is of significant interest. In this project, 4 different heating methods (a water-based device, a monopolar surgical device, a laser, and an electrical joule heater) were investigated.

At first, measurements were used to evaluate the temperature distribution along the implant axis. The investigation included an implant (Conelog screw line, length 9 mm and diameter 3.5 mm; Camlog) that was drilled in a polytetrafluoroethylene block (S-Polytech Gmbh). Polytetrafluoroethylene was considered to show a comparable thermal conductivity (l) value to fresh human bone, with l = 0.25 − 0.32 W/mK.^18,19  For the initial temperature analysis, the drilled holes were modified to enable a close-up view of the IR camera on the implant surface. This temperature measurement has been previously published.^20,21  For this, one side of the drilled hole was removed, and a slit along the implant surface was created (Figure 2a and b; ImageIR 5325, spectral range 3.7–4.8 mm, 1-Hz sampling rate, 25-mm/pixel resolution, microscope optics 300 mm; Infratec). The camera was placed at a 90° angle to the implant surface and 300-mm away from the test block for maximum spatial resolution (Figure 2a). The IR thermal camera was mounted on a metal rig with horizonal and vertical adjustment screws with a millimeter scale for standardized measurements. Before the investigation, the temperature information of the IR camera was verified and calibrated with a thermocouple inside the field of view. Implants were heated with 4 different thermal devices. Individual device settings are presented in Table 1. A water-based device with a double-barreled luminal cannula was fabricated with running water through a thermostat (bath temperature: 83°C; Figure 3). Second, a monopolar surgical device (MD 62, 50 W; KLS Martin) was used for implant heating. Because the electrical circuit flowed through the implant, the circuit was grounded via the implant tip using aluminum foil. Furthermore, an argon-ion laser (514-nm wavelength, 82 mW, continuous wave; Dantec) and an electric joule heater (3 W) with an inserted thermostat integrated in an implant healing cap were implemented (Figure 3). The temperatures were measured in real time. The laboratory temperature was maintained at 20°C using thermostatic temperature control. Thermal image data were simultaneously transferred to a PC. The IR measurements of the first setup were carried out along the implant axis using the IRBIS 3 thermography analysis software (Infratec).

A second improved setup was established using the 3 most promising heat devices for thermo-explantation (Figure 3; water, laser, and electric joule heater devices). Individual device settings are presented in Table 1. Here, the implants were inserted in a standardized closed polytetrafluoroethylene block with thermocouples at each implant device (Figure 3; mircocone implant, length 11 mm and diameter 4.5 mm; Medentika). The thermocouples were used for verification of the IR image. A standardized block thickness of 0.5 mm was fabricated on the side of the measurements (Figure 3) to simulate the peri-implant layer. In addition, the polytetrafluoroethylene block was set to an initial temperature of 33°C using a temperature-controlled water circle for simulating the temperature of the gingiva.²²  Individual aluminum connection devices were used for an improved heat transfer (Medentica). In this setup, the laser application was performed with a new diode laser (SiroLaserBlue, Sirona Dental Systems GmbH). Furthermore, a wide-angle lens (lens 50 mm; Infratec) was used to detect all 3 implants at the same time (Figure 4b). The aim was to heat all implants simultaneously up to 47°C using different interval settings. In the literature, a threshold level of 47°C leading to thermal bone injury was described.^9,23  The IR measurements and numerical calculations were carried out at 47°C along the implant axis and along the peri-implant area at the axial plane. Furthermore, the time interval for heating from 33°C to 47°C was measured at each of the implant tips. For comparable temperature analysis, all temperature values were transferred into a standardized scale ranging from 0 to 1 (T/T_max).

Numerical analyses were used to interpret the experimental results. These were performed using COMSOL (COMSOL Multiphysics GmbH, COMSOL Multiphysics software). The individual thermo-physical properties are presented in Table 2.^14,22  The thermal recordings of both setups were performed 3 times for each device using the mean value for statistical analysis.

Power analyses were performed with the G Power software (Wilcoxon Mann-Whitney test for 2 groups) to determine the power of 0.99 (primary study aim for 60 seconds), based on the sample size (group 1 water: 61; group 2 electric: 61) using an effect size of 0.97 and an α of .05 (mean 1: 44.1, standard deviation 1: 6.39; mean 2: 37.85, standard deviation 2: 6.49).

Analyses were performed using the Prism 8 software for Mac OS X (GraphPad) running on Apple OS X. Variables were analyzed using the D'Agostino and Pearson omnibus normality test. An unpaired Mann-Whitney test was used to identify the difference between the means of the subgroups. We assessed any effect in the statistical model as significant if the corresponding P value fell below the 5% margin.

Mean temperature changes along the implant length following heating are presented in Figure 2b. All heating devices showed significant differences (P < .01). The laser device presented the most even thermal change along the implant length, followed by the water-based setup. The electrical joule heater, on the other hand, led to the highest temperature gradient along the implant length next to the monopolar source (T/T_max). The monopolar electric device was not found to be suitable for implant heating because of an uncontrolled heating procedure with unpredictable electric and heat current flows (Figure 4a).

Because of the heat transfer from the inside screw thread of the implant, water and electrical joule heating led to higher temperature values at the implant shoulder (Figures 4b and 5). Thermal analysis along the implant length showed identical thermal distributions without any significant differences for either source (P > .05). The laser device, on the other hand, produced the temperature maximum in the center and at the tip of the implant, with significant differences between the water device and electrical joule heater at intervals of 60 seconds and 600 seconds (Figure 5; P < .01). The most homogenous temperature distribution was found at 600-second intervals for all devices, with temperature differences of almost 2°C. Similarly, at the shortest interval of 60 seconds, the measurement results demonstrated the maximum thermal deviations along the implant body (Figure 5a), with differences of almost 5°C.

Figure 5b shows the heating process of the implant from 33°C to 47°C over time at the central point of the lowest implant tip. Statistically significant differences were noticeable between all devices at heating intervals of 180 seconds and 600 seconds (P < .05). The water device showed the fastest approach to reach the maximum temperature level (T_max). The water- and laser-based heater presented an digressive temperature curve, whereas the electrical joule heater led to an progressive curve shape s-curve over 600 seconds (Figure 5B; P < .05). The electric joule heater required the largest time span to heat the implant up to the target temperature. Furthermore, after the laser was switched off, there was an immediate, detectable temperature reduction. On the other hand, the water and laser sources showed an afterglow and a further temperature increase after switch off.

An important aspect when comparing different heating devices between power and heating times is the penetration in the bone in a radial direction. Figure 6 shows the experimentally determined heat penetration depth in the radial direction for 3 heating devices (water and laser devices and the electrical joule heater), different time intervals, and heating powers. The shorter heat-up time with a higher power leads to a lower penetration and thus to a lower thermal affection (and bone necrosis) in all investigated heating devices. This is in agreement with the numerically simulated implant heat distributions for 3 heat devices (water and laser devices and the electrical joule heater), different time intervals, and heating powers (Figure 7), which indicated that a short and powerful heating procedure up to 47°C may lead to a minimal temperature change in the peri-implant layer; nevertheless, internal implant temperatures have risen to almost 53°C and higher. On the other hand, a slow implant heating might avoid a high temperature in the implant body, whereas the temperature penetration might affect a greater radius. The results were reviewed by an independent statistician.

The primary aim of this study was to identify the suitable sources for implant heating. The use of monopolar electric or laser devices around titanium implants has been described as generating thermal injury to the surrounding bone, with the subsequent loss of osseointegration.^8,10–12  Wilcox et al⁸  suggested that the use of monopolar devices for thermo-explantation should be avoided. The findings from our study are in alignment with those of Wilcox et al. The monopolar electric device led to an uncontrolled heating procedure with unpredictable current flows.

Nevertheless, another study group concluded during a finite element analysis that using a monopolar electric source at a power of 5 W, the temperature increase of the implants happened to be manageable at durations of <1 seconds.²⁴  However, the assessment has not been used to investigate a homogenous temperature along the implant axis. Furthermore, Wilcox et al⁸  evaluated a Nd:YAG laser for thermo-explantation. Worni et al,¹²  on the other hand, used a CO₂ laser. In our study, the first setup included an argon-ion laser; in the second setup, a more portable diode laser was used. Our results suggested that water and laser devices and the electrical joule heater might be suitable for clinical thermo-explantation. Controlled implant heating with an ideally homogeneous temperature along the implant axis may be an important basic requirement regarding thermo-explantation. A critical reflection on this study revealed that several clinical variables involved in the process of thermo-explantation were not considered in this in vitro setup. For example, the bone composition within and between patients is highly heterogeneous. Furthermore, in this investigation, the influence of the individual blood flow around the implant bodies has not been considered, as well as the fact that in case of thermo-explantation, implants are osseointegrated into the bone. To overcome these limitations, future preclinical in vivo studies are necessary.

Thermal analysis along the implant length indicated identical temperature distributions in the upper part of the implant shoulder for water and electrical joule sources; on the other hand, the laser device produced the temperature maximum in the center and at the tip of the implant. This might be due to the heat transfer of the first 2 methods via the screw thread in the shoulder area, whereas the laser hits the inner floor directly in the center of the implant. In most cases of peri-implantitis, the implant presents crestal bone loss. In relation to this fact, the laser source may show clinical advantages in the future. Furthermore, after the laser was switched off, there was an immediate visible temperature reduction without any afterglow. Further investigations are needed that vary the degrees of bone encapsulation along the length of the implant to simulate the percentage of bone loss related to peri-implantitis and the possible effects on heat transfer along the implant.

In addition, the aim of this study was to evaluate parameters that affect homogenous implant temperature and to investigate a controlled heating process up to 47°C. The water device presented the fastest approach to the 47°C level according to an exponential heating curve. Nevertheless, future individual device programming of the electrical joule heater may improve its heating curve. In an ideal scenario, only a small area of a few cell layers in the peri-implant bone would be thermally affected in order to avoid severe bone necrosis. It is assumed that the soft tissue reacts differently than bone does to thermal changes. The increased blood flow would allow the temperature to be better dissipated in the soft tissue. This effect could subsequently reduce the planned temperature depending on the degree of osseointegration of the implant body and should be considered in potential future clinical use.

Gungormus et al²⁴  concluded that low-power settings must be used for a safe thermal necrosis-aided implant removal. Numerical analyses may suggest that neither a short and powerful heating due to very high temperatures (<60 seconds) nor a slow and weak power setting (>600 seconds) due to an expanded thermal tissue penetration might be used. The ideal interval to reach a temperature of 47°C for an even and clinically safe heating procedure may be in the range of 120 seconds. To verify that these reported results are uniform in different diameters and lengths in clinical use, further studies are suggested that test different implant diameters (narrow-body and wide-body implants) and implants of shorter and longer lengths than those tested in the current study.

According to Fajardo et al,²⁵  in bony gaps filled with fluid (such as blood), the thermal conductivity of trabecular bone increases, and on the other hand, the presence of fatty tissue or red marrow has practically no effect on either the maximum temperature or the position of the isotherms.

Atraumatic thermo-explantation could bring benefits to the patient in the future. Numerical analyses might play an important role, as many different implant dimensions are available on the market, and the temperature input affects the heating process of the individual implant. This experiment also serves to lay the foundation for the planned preclinical animal experiments. In addition, it is important to limit the enormous sample size caused by different sources and many temperature and time intervals for the planned animal experiment.

Before clinical use, further research should be concentrated on bone morphology. The bony structure shows great morphological differences depending on species and regions.²⁵  For further research based on the findings of this study, we recommend evaluating water or laser temperature-controlled systems. These might have the advantage that water shows the fastest approach to reach the maximum temperature and the laser avoids an afterglow effect. Disadvantages were noticeable with electric systems because of the very high temperatures in the abutment area, and the monopolar electric device was not found to be suitable for implant heating because of an uncontrolled heating procedure with unpredictable electric and heat current flows.

Thermal analysis along the implant length suggested almost identical thermal distributions for water and electrical joule sources with maximum temperature values in the upper part of the implant shoulder. The laser device, on the other hand, indicated the temperature maximum in the center and at the tip of the implant without any afterglow effect. In future research, the water device might present the fastest approach to achieve the maximum temperature level according to an exponential heating curve. The numerical and experimental analyses may suggest a heating time of approximately 120 seconds to 180 seconds to reach a homogenous implant temperature and a tolerable penetration in the peri-implant bone.

Dr. med. Thien An Duong Dinh (diploma of medical statistics, University RWTH Aachen, Pauwelsstraße 30, Germany) independently reviewed the methodology.

This study was funded by the German Federal Ministry for Economic Affairs and Energy (BMWi) within the Promotion of Joint Industrial Research Program (IGF) due to the decision of the German Bundestag. It was part of research project 20302N by the Association for Research and Precision Mechanics, Optics and Medical Technology (FOM) under the auspices of the German Federation of Industrial Research Associations (AIF).