Journal of Endodontics Research -
Root and canal morphology of Thai mandibular molars
By JofER editor
Published on 01/21/2002

K. Gulabivala, A. Opasanon, Y.-L. Ng & A. Alavi
Department of Conservative Dentistry, Eastman Dental Institute for Oral Health Care Sciences, University College London, London, UK

To investigate the root and canal morphology of 351 mandibular permanent molars collected from an indigenous Thai population.

This study suggests that Thai molars exhibit features of both Caucasian and Chinese teeth.

Introduction - Materials and methods.

The study of root and canal anatomy has endodontic (Vertucci 1984) and anthropological (Tratman 1950, Dahlberg 1965, Walker 1988a,b) significance. It is important to be familiar with variations in tooth anatomy and characteristic features in various racial groups since such knowledge can aid location and negotiation of canals, as well as their subsequent management. Additionally, a number of studies have shown different trends in shape and number of roots and canals amongst the different races (Somogyi-Csimazia & Simmons 1971, Curzon 1974, Vertucci & Williams 1974, Cooke & Cox 1979, Reichart & Metah 1981, Walker & Quackenbush 1985, Walker 1988a, Weine et al . 1988, Yang et al . 1988, Manning 1990a,b, Melton et al . 1991, Sperber & Moreau 1998, Weine & members of the Arizona Endodontic Association 1998, Haddad et al . 1999). These variations appear to be genetically determined (Curzon 1974) and are important in tracing the racial origins of populations.
One example of such variation is the mandibular first molar with three roots. This variant has a frequency of less than 5% in white people (British, Dutch, German, Finnish and other European), African (Bushmen, Bantu, Senegalese), Eurasian and Indian populations, whereas in those with Mongoloid traits, such as the Chinese, Eskimos and Native Americans, it occurs with a frequency of 5 to over 40% (Somogyi-Csimazia & Simmons 1971, Curzon 1974, Vertucci & Williams 1974, Reichart & Metah 1981, Walker & Quackenbush 1985). Another reported variation is the ‘C-shaped root and canal configuration’. Seldom found in white people, they have a relatively high prevalence in mandibular second molars of Chinese and Lebanese populations (Walker 1988b, Weine et al . 1988, Yang et al . 1988, Manning 1990a, Weine & members of the Arizona Endodontic Association 1998, Haddad et al . 1999).

Classification of root canal system types
Figure 1. Classification of root canal system types.

The population of Thailand originates from a mixture of Indian and Chinese emigrants. The aim of this study was to investigate the relative contribution of Mongoloid and Caucasian traits in root and canal morphology of Thai mandibular molars.

Materials and methods.
The sample teeth comprised 351 mature mandibular permanent molars extracted from patients attending the Oral Surgery Department of the Christian and Rajchaviti hospitals in Bangkok. All attached soft tissue and calculus was removed using an ultrasonic scaler and by soaking the teeth in 3% sodium hypochlorite. The teeth were classified into first (118), second (60) or third (173) molars based on crown morphology. Access cavities were cut and pulp tissue was removed by immersion in 3% sodium hypochlorite overnight before placement in an ultrasonic bath. The teeth were then rinsed under running tap water for 2 h and dried overnight. An endodontic irrigating syringe with gauge 27 needle (Sherwood Medical Company, St. Louis, USA) was used to inject the Indian ink (Calder Colours Ltd, Ashby-de-la-Zouch, Leicester, UK) into the root canal systems. Ink penetration was assisted by vacuum suction applied apically. The teeth were air dried and immersed in 10% nitric acid (MJ Patterson, Dunstable, UK) to decalcify for between 5 and 14 days. The acid solution was changed daily and the end-point of decalcification determined by periodic radiography. The teeth were washed under running tap water, dried and dehydrated using ascending concentrations of ethanol (70%, 95%, 100%) for 3 days. Finally, the teeth were rendered transparent by immersion in methyl salicylate (Merck Ltd, Poole, UK) for 2 days.
The following observations were made: (i) number of roots and their morphology; (ii) number of root canals (defined as root canal orifices in the pulp chamber) per tooth; (iii) number of root canals per root; (iv) root canal configuration in each root using Vertucci’s classification (1984) with additional modifications (Fig. 1); and (v) presence and location of lateral canals and intercanal communications. Lateral canals were defined as those branches of main canals that diverged at right or oblique angles to exit on the lateral aspect of the root. Inter-canal communications were defined as those complex canal ramifications that ran from and between main canals but did not communicate with the root surface.


The data for root and canal number and their morphology are presented in Tables 1–5.
Of the 118 first molars, 12.7% had three roots with the extra root lingual to the main distal root (Table 1). The majority had three (61%) or four (30%) canals (Table 2). In the three-rooted molars, 80% of the main distal root and 100% of the disto-lingual roots had type I canals. The majority (66.7%) of the root canals in the mesial root were type IV. The remainder were distributed between type III and additional types 3–1, 3–4, and 2–3 (Table 3).

Classification of Thai mandibular molars by root number and morphology
Table 1. Classification of Thai mandibular molars by root number and morphology.

Number and percentage of mandibular molars with 1, 2, 3, 4 or 5 canals per tooth
Table 2. Number and percentage of mandibular molars with 1, 2, 3, 4 or 5 canals per tooth.
Number and percentage of canal system types in mandibular first molars
Table 3. Number and percentage of canal system types in mandibular first molars.

Of the 60 second molars, 10% had a single C-shaped root, whilst the rest had two separate roots (Table 1). The majority had three (58%) or two (23%) canals (Table 2). These roots had an equal chance (33.3%) of containing one of three canal configurations: type I, type IV or the additional 3–4 or 2–3 varieties. In the two-rooted molars, the majority (70.4%) of the distal roots had a type I canal system, whilst only 14.8% of the mesial roots had this configuration. The majority (57.4%) of the mesial roots were type IV, followed by type II (Table 4).

Number and percentage of canal system types in mandibular second molars
Table 4. Number and percentage of canal system types in mandibular second molars.

The 173 third molars had the highest variation in root morphology, as well as a high prevalence of two separate roots (67%), two-fused roots (19.1%) and C-shaped roots (10.9%) (Table 1). The majority had two (61%) or three (28%) canals (Table 2). The teeth with fused roots contained a spread of canal configurations: types IV (33.3%), I (21.2%) and II (18.2%), with a few teeth having other varieties. In the C-shaped group, although the most prevalent canal configuration was type IV (36.8%), there was a wide spread of types, including I (15.8%), VIII (15.8%), II (5.3%) and additional varieties (26.4% collectively). In three-rooted teeth, all the distal canals were type I and the mesial canals were either type II (50%) or IV (50%). In the two-rooted variety, the dominant canal configuration in both mesial and distal canals was type I (53% and 93.2%, respectively). The remainder were distributed amongst several types (Table 5).

Number and percentage of canal system types in mandibular third molars
Table 5. Number and percentage of canal system types in mandibular third molars.

There was an increasing prevalence of lateral canals toward the apical part of the root for all types of molars (Table 6) with the highest being 4.1% in the apical third of mandibular third molars. Inter-canal communications were found more commonly in the first (39.8%) and second (43%) molars than the third molar (19.4%) (Table 6).

Number and percentage of roots with lateral canals and intercanal communications
Table 6. Number and percentage of roots with lateral canals and intercanal communications.


The method of canal staining and root clearing was found to be excellent for three dimensional evaluation of root canal morphology. It was anticipated that examination of the fine details (intercanal communications, lateral canals) would require adequate ink penetration, however, it was found that the quality of clearing was sufficient to visualize such details without staining. In this context, the low prevalence of lateral canals found in this study is surprising, since it is not the result of non-penetration of ink.
In common with other studies using extracted teeth, the sample must be regarded as biased because it is not a random selection and therefore may not be representative of the overall population.

Mandibular first molars with three roots.
The prevalence of three roots in Thai mandibular first molars (13%) falls within the range seen for other Mongoloid populations (Somogyi-Csimazia & Simmons 1971, Curzon 1974, Walker & Quackenbush 1985). It is slightly lower than the figure (19%) reported by Reichart & Metah (1981) for a Thai population, which may reflect variations in the sample, but considerably higher than in white populations (Vertucci & Williams 1974). The additional root was found on the lingual aspect of the main distal root and has been regarded as a genetic trait rather than a developmental anomaly (Walker 1988a). The nature of this additional root is variable, ranging from a short conical extension to a full-length root, with pulp extending into the root even when it is short (Reichart & Metah 1981); it can occur unilaterally or bilaterally (Somogyi-Csimazia & Simmons 1971). Radiographically, a third root should normally be readily evident in about 90% of cases (Walker & Quackenbush 1985) but occasionally may be difficult to see because of its slender dimensions. Additionally, a file placed in such a root may give the artifactual appearance of a perforation. In such instances, an angled view (vertically and horizontally) may be helpful (Somogyi-Csimazia & Simmons 1971).
The presence of a third root indicates that there will be two distal canals in the tooth. The disto-lingual canal will invariably have type I configuration, may exit the pulp chamber with a marked lingual orientation and may have a subtle curve in the buccal direction at its apex. The difficulties in location and negotiation of the narrow additional disto-lingual canal may be a cause of treatment failure. Thirty per cent of mandibular first molars had four root canals, a figure comparable to other studies (Pineda & Kuttler 1972, Vertucci 1984). It is, therefore worth searching for additional orifices after initial preparation and debridement of the pulp chamber and canals is complete. In teeth with two roots, 58% of the mesial roots had a type IV (2–2) configuration, which indicates that in these teeth more mesial canals end in separate foramina than in Caucasian teeth (Vertucci 1984).

Mandibular second molars with C-shaped roots.
C-shaped canals were first documented in the endodontic literature by Cooke & Cox (1979) in three case reports. Later studies of the root canal anatomy of mandibular second molars from Japanese, Chinese, and Hong Kong Chinese populations found a high incidence of C-shaped roots and canals (14–52%). In the present study, the prevalence of 10% in Thai mandibular second molars falls within the lower end of the range for a Mongoloid group (Walker 1988b, Yang et al . 1988, Manning 1990a).
Radiographically, a C-shaped root may present as a single-fused root or as two distinct roots with a communication. Haddad et al . (1999) described four typical images: roots that appear to fuse apically, roots with close proximity, a large distal canal or the blurred image of a third canal in the middle of two roots. In some cases confirmation is necessary by exploring the access cavity, because the canal orifice may present with a C-shape. Additionally, fused and C-shaped roots may present with narrow root grooves that predispose to localized periodontal disease, which may be the first diagnostic indication of this anatomical variant.
Manning (1990b) reported that C-shaped roots most frequently had three canals. Other studies have reported a more diverse range of canal systems in C-shaped roots (Yang et al . 1988, Melton et al . 1991, Haddad et al . 1999). In the present study, the configuration of the canals in one-third of the C-shaped roots was type I (one canal), another one-third had type IV canals (two canals), and the remainder had 3–4 or 2–3 configurations (Table 4).
The pulp chamber in teeth with C-shaped canals may be large in the occluso-apical dimension with a low bifurcation. Alternatively, the canal can be calcified, disguising its C-shape. At the outset, several orifices may be probed that link up on further instrumentation. Radiographically, a file placed in the centre of a C-shaped canal may mimic a perforation through the furcation, especially when there is excessive haemorrhage during a vital pulpectomy (Cooke & Cox 1979). In this situation, the use of a third generation apex locator (i.e. one able to read canal lengths in the presence of electrolytes) should help to distinguish between the two. In a true C-shaped canal, it is possible to pass an instrument from the mesial to the distal aspect without obstruction. In other canal configurations, such passage is impeded by discontinuous dentine bridges (Melton et al . 1991).
Interestingly, the most common root canal configuration in the present study was type IV (two canals, two foramina) in the mesial roots and type I (one canal, one foramen) in the distal roots. This result differs from the studies of Vertucci (1984), Weine et al . (1988) and Manning (1990b), who found the most common configuration to be type II (two canals, one foramen) in the mesial roots. This may be another Mongoloid trait, previously unreported.

Mandibular third molars.
In this study, mandibular third molars were associated with various types of root and canal morphology (Tables 1–5). The findings are, on the whole, comparable with other studies (Pineda & Kuttler 1972). Most of these teeth had two roots which were either separate (68%) or fused (19%). Other root morphologies included three roots (1%), one conical root (1%), four roots (1%) and a surprisingly high prevalence of C-shaped roots (11%). Most of these teeth were also found to have only two root canals (61%).