Already widely used in medicine, diagnostic devices based on the use of fluorescence are becoming more popular in dentistry particularly for the diagnosis of dental caries and oral mucosal abnormalities.This article describes recent developments extending the use of fluorescence techniques for diagnostic use to endodontics.
 By Dr Laurence J Walsh

As the population ages and more teeth are retained into advanced age, there is an increasing need for root canal treatments (RCT). The clinical outcomes of RCT have been studied for over 70 years, and from the literature it is clear that in situations where persisting infection remains in the root canals of teeth the prognosis is poor and the treatment is likely to fail. There is therefore a need for techniques that can assess whether or not infection remains once the canals have been filed to the required shape.

Traditional technologies for detecting infection in the body are based on collecting a sample which is then cultured for several days in the laboratory. Unfortunately, culture-based methods are not suitable for recovering anaerobic bacteria from root canals, because these are intolerant of atmospheric oxygen, and become non-viable during the sampling procedure. Culture-based methods are very technique-sensitive and are also time-consuming.

A more recent approach is to use molecular biological techniques for detection of specific bacteria. This is typically done by using the polymerase chain reaction (PCR). Such methods are capable of detecting bacteria that are difficult or even impossible to find using traditional culture-based methods. However, PCR is also technique-sensitive, is prone to technical errors, is time-consuming and expensive, and can only detect a limited range of organisms. Moreover, it cannot be done within a dental practice environment. Because of these problems, PCR has not entered mainstream clinical use.

An alternative diagnostic method is fluorescence. The concept of fluorescence diagnosis in dentistry is well established, and today devices such as Vizilite, Velscope and DiagnoDENT exploit this light-generated reaction.

Principles of fluorescence
In an effort to identify infected root canals in real time at relatively low cost, our research group developed an alternative approach, which provides a simple, low cost but technologically elegant solution. The basic concept was that laser energy of the appropriate wavelength could be introduced into the root canal system using an optical probe. This laser energy would in turn induce fluorescence — i.e. the emission of light of a longer wavelength — from individual bacteria and from biofilms of bacteria remaining in the root canal. By choosing appropriate wavelengths of exciting light, the emitted light from bacteria could be discriminated from any light reflected back from the probe or generated by the uninfected areas of the root canal. We knew from earlier work that short wavelengths of light in the ultraviolet, visible violet and blue regions elicited fluorescence from bacteria, but did not penetrate the structure of the tooth effectively, and could not identify all species of bacteria. On the other hand, we also knew that light of certain visible red wavelengths did penetrate tooth structures effectively and could reliably detect all the bacteria of interest.

If a fluorescence method could be made to work, it would have the advantage of being able to localise the site of the infection, since the fluorescence signal would be topographically associated with the presence of bacteria within particular regions of the canal. The laser fluorescence signal could be quantitatively measured and the value recorded and used to track improvements as further cleaning, filing or other treatments were done to disinfect the canal. The ability to identify infected root canals in real-time, before they were filled, would allow the dentist to provide any necessary additional treatment, and re-measure the situation to ensure that all bacteria were eradicated. This should increase the overall success rate of RCT in dentistry.

Design and performance of new system
The starting point was the modification of an existing commercially available laser fluorescence system designed primarily for detection of dental caries (the DiagnoDENT system). This has a diode laser emitting in the visible red region at 655 nm wavelength, and an in-built digital filtering and analysis system for detecting near-infrared fluorescence emissions from bacteria. The device was fitted with a prototype optical tip and was used to record fluorescence profiles for root canals in freshly extracted teeth that were known from clinical and radiographic signs to be infected. We also examined uninfected canals from third molar teeth, and developed a laboratory system to maintain bacteria and grow biofilms in root canals. This work demonstrated that the laser fluorescence method could detect infected canals with isolated bacteria as well as with dense biofilms, and that the signal was proportional to the extent of infection. When the infected teeth underwent root canal treatment, the specific signal from bacteria disappeared. The method demonstrated a high specificity in identifying those root canals that were uninfected. We confirmed this performance by splitting the roots and examining them in the scanning electron microscope for persisting bacterial contamination. From this initial study, it was clear that the use of a laser fluorescence approach for assessing the status of the root canal system of teeth was a workable concept.

Even though dentistry has always prized high diagnostic sensitivity (i.e. the ability to find disease), in recent decades the need for high, predictable diagnostic specificity (i.e. the ability to rule out disease) has also grown in importance. The difference in fluorescence readings between healthy and infected canals provides confidence that the cleaning of the root canal has reached a biological endpoint. Laser fluorescence threshold values for normal healthy root canals and normal healthy teeth have been determined, so that precise recommendations can be made for the interpretation of
fluorescence scores.
We have since gone on to develop flexible tips that can penetrate into furthest reaches of the root canal, as well as special modifications to these optical fibres so that the fluorescence light could be delivered and collected from a wide viewing angle. When tested on sectioned extracted teeth, the optical fibres were in most cases able to reach the apical third of the root canal space without fracturing. This was possible even when the canals were curved, with the light able to reach the end of even severely curved canals. Thinner fibres and fibres with conical rather than flat ends were shown to be better at negotiating the small spaces in root canals. Small diameter optical fibres are needed to gain entry into the apical third of the root canal. This region is where persistence of bacteria is most likely to occur after cleaning and instrumentation.

Disinfection system
A further development was to link the detection function to a laser cutting or disinfection system, so that an autopilot was generated for detecting and destroying bacteria in the root canal. Bacteria within root canals are known to resist both physical cleaning approaches and chemical agents. For this reason, we developed several different systems which could be “laser-guided”.

The first used a light sensitive, laser-activated dye which bound to and killed or inactivated bacteria, including even the most resistant species known. This system was shown to work on isolated bacteria and on dense biofilms of resistant bacteria. The second system which can be laser-guided is based on the use of shockwaves which are generated in water-based fluids by laser pulses. The formation of water vapour and its implosion cause cavitation, which in turn creates massive shear forces on the walls of the root canal, which disrupts bacterial biofilms. Different types of optical fibre tips alter the fluid dynamics involved. The laser-generated shockwaves move at over 90 km per hour and eject bacteria and debris from the root canal. We were able to show that several different laser systems could be used for both approaches, and importantly that some or all of the desired effects could also be created with very small compact diode laser devices, when operated under the right conditions.

A caveat to using feedback for the selective removal of bacteria is that endodontic irrigants and medicaments, which because of their chemical struture actually produce inherent autofluoresence themselves, must be avoided because of the clear risk of the generation of false-positive results. For example, of the materials that are commonly used, both MTAD  and Ledermix are problematic because their tetracycline component in their structure gives a false positive fluorescence signal. Fortunately, alternatives to these products exist which do not suffer from the same problem; indeed,  all common endodontic materials such as calcium hydroxide, sodium hypochlorite and chlorhexidine do not fluoresce with the technology we have used.

Conclusion
In summary, from our work, we believe that an autopilot approach would offer considerable advantages over current treatments, by increasing the effectiveness of removing debris and bacteria, and giving a firm biological endpoint for treatment. This should lead to more predicable clinical treatment and fewer treatment failures. 

Selected references for further reading
AL Sainsbury, PS Bird, LJ Walsh. DIAGNOdent laser fluorescence assessment of endodontic infection. Journal of Endodontics 2009; 35(10): 1404-1407.

R George, LJ Walsh. Performance assessment of novel side firing flexible optical fibres for dental applications. Lasers in Surgery and Medicine 2009; 41(3):214-221.

QV Ho, R George, AL Sainsbury, WA Kahler, LJ Walsh. Laser fluorescence assessment of the root canal using plain and conical optical fibres. Journal of Endodontics 2010; 36(1):119-122.

R Hmud, WA Kahler, R George, L J Walsh. Cavitational effects in aqueous endodontic irrigants generated by near infrared lasers. Journal of Endodontics 2010; 36(2):275-278.

The author
Professor Laurence J Walsh,
Professor of Dental Science
The University of Queensland School of Dentistry
Brisbane,
Queensland,
Australia