Unproven case

Sir, in support of Dr Qualtrough's comment (BDJ 2006; 200: 329) to treat the study by Bonsor et al. (BDJ 2006; 200: 337–341) with caution, I would like to share the following observations.

• 1 The study was stated as a randomised trial but the only randomisation procedure was in the recruitment of patients, whether a patient was selected or not. The sample size was small; there were no proper controls and no statistical analysis.

• 2 Microbial sampling and culturing from the root canal are exacting procedures and were developed based mainly on Moller's¹ classical thesis. The canal sampling method used in this study has little in common with recognised protocol.

• 3 Rubber dam was only placed after access to the pulp chamber was gained despite the awareness of the risk of contamination.

• 4 O'Neil et al. (2002) were quoted in support of effective disinfection of the operating field after rubber dam placement. However, no field samples were taken for confirmation.

• 5 A chosen canal, not each canal, in a multi-rooted tooth should be a test unit.

• 6 For maximum efficacy, the emitter was placed within 4 mm of the canal length. Since no radiographic determination of canal length was performed, the emitter may be within 4 mm of the estimated but not necessarily the true canal length.

• 7 An irrigant should never be 'injected' because of the risk of extrusion.

• 8 Only sterile water but no inactivating solution eg sodium thiosulphate was used prior to taking Sample B.

• 9 The three cases pretreated with Ledermix should have been excluded.

• 10 It is common practice to leave sodium hypochlorite to soak in root canals to improve disinfection. Unlike the photosensitiser, sodium hypochlorite was not given diffusion time.

• 11 The statement (line 15) in Results about confirmation that cross-contamination did not occur is unconvincing.

• 12 The toxicity of sodium hypochlorite is increased at higher concentrations but toxicity is also related to pH. Therefore, a higher concentration does not necessarily mean greater toxicity.

• 13 In Discussion, the claim that the results of the study demonstrated the reliability of their technique is questionable.

• 14 The emitter is equivalent to ISO #40 so canals have to be prepared accordingly. Apart from the risks and difficulties in preparing to this size eg mesial roots of mandibular molars, by the time this was attained, significant disinfection was achieved (80%). PAD led to further improvement (15%) but not total bacterial eradication so an intracanal medicament is still required.

• 15 The first line in Conclusions should be qualified — 'The results of this limited study show...'.

• 16 The last two In Brief bullet points should also be qualified —

  • Conventional chemo-mechanical canal preparation techniques may not always disinfect the canals predictably and consistently.

  • PAD may offer, as an adjunct, the potential to eliminate bacteria from the root canals especially where conventional techniques may have failed to do so.

PAD may have the potential to improve root canal disinfection but from this study, the case is far from proven.

The authors, S. Bonsor and G. Pearson, respond to the above three letters: We would like to thank Drs Chong, Karunanayake and Watson for their interest and comments on the article on PAD.

May we correct some errors of fact in the letters?

• 1 The canals were prepared Profile 0.4 rather than 0.6 as suggested in one letter.

• 2 The teeth where Ledermix was used were excluded from the results in Figure 3 and this is stated in the text.

• 3 The photosensitiser is not an irrigant as indicated by respondents. It is a solution which when activated by light provides singlet oxygen which kills the bacteria. As such the volume used is that required to fill the lumen of the canal. This volume is delivered by the use of a safe ended endodontic needle passively inserted into the lumen of the canal to prevent extrusion of fluid beyond the apical foramen. This is followed by gentle agitation with a hand instrument to eliminate air bubbles. No 'active irrigation' of the photosensitiser was carried out.

• 4 The use of prolonged saturation of hypochlorite in the canal was demonstrated in that the canal preparations were carried out over periods of between 20 and 30 minutes evidenced by the recorded time at which samples were taken. During that time there was copious irrigation with the co-irrigants: hypochlorite and citric acid.

• 5 The emitter tip is of a size such that it will reach to within 3 mm of a canal which has been prepared with a size 25 instrument of the conventional 02 ISO standard and would not therefore have been over-prepared as suggested. Laboratory studies have shown that light transmission in air is unattenuated and that passage down a canal even with a curvature is not affected as little attenuation occurs even on passage through 3 mm of dentine.

With regards to the other points raised, the following should also be borne in mind:

Randomisation

In the context of this trial the patients were a random sample of those requiring endodontic therapy. No prior screening was carried out to select patients for inclusion.

Controls

This study was an initial investigation of a technique which had been evaluated in considerable depth in the laboratory.² Within this study an attempt was made to examine the effect of using the technique as an adjunct to the current therapy which, while being good, is not foolproof. The option would exist to have a sample group where the canal was left open for the period of time equivalent to the time that the photosensitiser would be in the canal ie three minutes and then sample again. This was considered and a number of preliminary samples were taken of teeth at these specific points in an endodontic procedure to explore the differences. No difference in bacterial load was found and it was therefore decided that the initial sample was a representative example of the canal as prepared by the operator at a point where it would normally be filled.

Blinding

In respect to the blinding of the trial the treatment regiment was identical in all cases and the blinding was achieved by the fact that the microbiological samples (the only aspect where bias is likely to have occurred) was carried out by a member of the Microbiological Department of the hospital with no knowledge of the coding system which was used in the clinic.

Numbers

Numbers of cases in studies are always a problem and we would be the first to acknowledge that the numbers in this study are limited. This was pointed out by the commentator Dr Qualtrough on page 329 of the journal. However, microbiological studies are regularly carried out worldwide in the endodontic arena and the numbers quoted here are similar to a large number of other studies. In fact, a substantial number of comparative studies of effectiveness of materials / medicaments in endodontics use a similar range of numbers in each test. Pinherio et al. 2004 (n = 21), Attin et al. 2002 (n = 22), Kont Cobankara et al. 2002 (n = 10 in each group) are some of the examples of this. ^{3, 4, 5}

Rubber dam placement

This was placed after access to the pulp chamber as it is a well known complication that loss of tooth angulation when dam is in situ can cause iatrogenic damage during pulp chamber access. The dam was placed before any canal cleaning and so if contamination had occurred then sample A would have shown a higher result, potentially allowing a greater fall off of bacterial numbers at sample B thus if anything, showing a greater effect of conventional techniques.

Estimation of canal length

An apex locator was used to determine the working length of the canal and then this was confirmed radiographically at the cone fit stage. This is a recognised technique for determining the working length. It is well documented that apex locators provide a precise estimate of canal length and indeed have been shown to be more accurate than the use of a diagnostic file radiograph. ^{6, 7}

Irrigation of the canal

It should also be borne in mind that in the intervening time between the sampling after hypochlorite irrigation and sampling for PAD, the canal was only filled with photosensitiser solution and agitated with a hand instrument. No irrigation in the normal endodontic meaning of the word was carried out. Hence the flow volume and change to canal wall would not affect the results as Dr Karunanayake implies. This point also relates to one of those raised by Dr Watson as the canal was not actively irrigated with PAD solution just filled unlike the use of hypochlorite where continuous flushing out is advised.

Neutralisation of the sodium hypochlorite

This is an interesting point particularly since the photosensitiser is a light blue colour which is bleached by sodium hypochlorite. In fact the concentration of bleach required to achieve this is a 0.0032% solution. Since the concentration use in the study was greatly in excess of this (2.25%), the presence of any residual hypochlorite would therefore have easily been recognised as the photosensitiser would have bleached. This phenomenon was not observed.

A canal opposed to a tooth as a test unit

It was shown with the results gained that no cross contamination of multi-rooted teeth took place. Furthermore it is well known that in multi-rooted teeth, different canals may exhibit different stages of pathology and apical pathology concurrently and hence bacterial loads. Sampling of each canal was therefore fully justified. Interestingly if the tooth is treated as the unit the results for the current study are similar with a 15% increase in canals with no culturable bacteria after PAD application.

Sampling

The use of an instrument to sample the walls has previously been reported by a number of works including Hancock et al.⁸ The method of sampling here requires that the instrument be drawn up the side of the wall as it is moved circumferentially round the canal such that swarf and debris from the canal wall would be collected. This would pick up material particularly in the apical third.

Change in canal volume

An interesting point is raised as to the comparative changes within the root canal as a result of the method of sampling and suggests that this would influence the outcome. This was a point of discussion in the preamble to the study and calculations were made as to the effect that the removal of swarf from the apical part of the canal which is primarily involved would mean in terms of volume involved. The volume change in the apical 4 mm by increasing instrument size is from 0.00034cc to 0.00044cc. The variation in solution volume is therefore 0.0001 ml. There may therefore be a slight change in canal geometry but certainly not of any significant effect.

Toxicity of sodium hypochlorite

It is well accepted that the toxicity of sodium hypochlorite is pH dependent. However at the pH commonly used in endodontics, particularly in the absence of any buffering, the higher the concentration at the specific pH the more likely to produce tissue damage as instanced by the references reported in the article.

Intra-canal medicament required Dr Chong mentions the need for intra canal medication after treatment. While this was not directly part of the study, Calcium hydroxide was placed in all canals after sampling and root canal preparation. This would not necessarily have dealt with species such as E or S faecalis even if these organisms were present after treatment since calcium hydroxide is ineffective against these organisms. ^{9, 10}

Statistical analysis

The objective of the trial was, as stated earlier, a preliminary trial to examine the possible effect of the technique as an adjunct to conventional therapy. To this end, complex statistical analysis would be inappropriate. However, without pre-empting future publications in press, statistical analysis in a larger study directly comparing differing methods of disinfection suggest that photo-activated disinfection is an effective alternative to conventional disinfectants.