M.-K.Wu, L.W. M. van der Sluis & P. R.Wesselink
Department of Cariology, Endodontology, Pedodontology, Academic Centre for Dentistry Amsterdam (ACTA), Amsterdam, the Netherlands.

Introduction.
In infected root canals, the inner layer of dentine may contain microorganisms (Peters et al. 2001). One aim of root canal instrumentation is to remove the inner layer of dentine from all aspects of the root canal wall (Walton & Torabinejad 1996). However, in many cases bacteria have penetrated deeply into the dentine (Armitage et al. 1983, Ando & Hoshino 1990, Peters et al. 2001), making it difficult to completely remove them from the dentinal tubules using instruments. Moreover, it would be more difficult to remove the entire inner layer of dentine in long oval root canals than in round (Wu & Wesselink 2001).
In many dental schools, students are taught that the apical root canal should be enlarged to three sizes larger than the first file that binds at the working length (the first binding file) (Weine1996).The aim of this procedure is to remove the entire inner layer of dentine from the apical canal wall. The first binding file is the smallest instrument that enables dentists to feel resistance at or before reaching the working length. It is thought that this file can gauge the apical diameter, so that after enlargement using three larger files, the inner layer of dentine together with the microorganisms can be removed from the entire wall.
In a recent study (Wu et al. 2002), however, it was found that at its working length the first binding file touched only one side of the wall in75%of curved canals, and failed to touch any wall in the other 25%. This indicates that after further enlargement the inner layer can be removed from one side of the wall only. Whether the inner layer can be removed from the entire canal wall remains questionable.
Large master files have been recommended in the past to scrape the entire circumference of the root canal wall (Tronstad 1991). For instance, sizes 70-90 have been recommended for all maxillary central incisors. The internal diameter of maxillary central incisors may vary from 0.19 to 0.94 mm, 2 mm from the apex (Wu et al. 2000b); at this level the diameter of a size 90 master file is 0.92 mm. Clinically, dentists do not know whether the canal is 0.19 or 0.94 mm, and using large files in all maxillary central incisors could unnecessarily result in a severe weakening of those small roots (Trope & Ray1992). Furthermore, using large files in curved roots can lead to apical lacerations and ledging (Tang & Stock 1989, Briseno & Sonnabend 1991, Nagy et al. 1997, Buchannan 2000, Wu et al. 2000a). Some textbooks say that curved canals should not be prepared apically beyond a size 20 or 25 (Ingle et al.1994,Walton & Torabinejad 1996). The use of Ni-Ti rotary instruments can reduce, but not completely prevent, the occurrence of apical transportation (Wu et al. 2000a, Hulsmann et al. 2001). This means that using large files can weaken the root and increase the risk of apical transportation.
Oval-shaped canals, most of which have long buccolingual but short mesio-distal diameters, exist in 25% of roots (Wu et al.2000b). Using a larger file in long oval canals inorder to include the entire oval canal in the preparation can result in perforation of the mesial or distal wall, as suggested by Wu & Wesselink (2001). It has been supposed that a circumferential filing technique with a small file will prevent this, while completely scraping the wall. However, several studies have shown that circumferential filing is not capable of contacting the entire canal wall (Reynolds et al. 1987, Zuolo et al. 1992, Siqueira Jr et al. 1997, Evans et al. 2001). Access cavity location and design may influence the percentage of the wall surface that is contacted by the instruments. However, Mannan et al. (2001), who used different cavity designs in maxillary anterior teeth, found that regardless of access cavity design, mechanical preparation using step back filing did not allow instrumentation of the entire wall.
The balanced force technique (Roane et al. 1985) has been used in the preparation of curved root canals (Wu & Wesselink 1995). However, it has been found that in two-thirds of oval canals use of the balanced force method left a portion of the root canal wall uninstrumented (Wu &Wesselink 2001).
Different methods have been used to evaluate the cleaning efficacy of root canal preparation. Histological cross-sections have been used and the capability of different techniques to remove predentine evaluated (Reynolds et al.1987, Zuolo et al.1992, Siqueira Jr et al. 1997, Evans et al. 2001). However, predentine was not always visible (Evans et al. 2001). Longitudinal sectioning allows for an evaluation of the entire root surface (Lumley et al.1993,Wu & Wesselink 1995). However, the root surface can be evaluated only once, after the root canal preparation. The muffle model introduced by Bramante et al. (1987) made it possible to examine the root canal both before and after instrumentation, at any level within the same canal system. Since then, various modified versions of this model have been used to evaluate the effects of root canal instrumentation (Calhoun & Montgomery 1988, McCann et al. 1990, Sydney et al. 1991, Hulsmann et al.1999, 2001).
The purpose of this study was to use a muffle model to evaluate the capability of two hand instrumentation techniques, i.e. balanced force and circumferential filing, to remove the inner layer of dentine in oval canals.