The pulp cavity of equine maxillary and mandibular cheek teeth is subject to profound age-related changes throughout its life span . Due to continuous deposition of secondary dentine around the walls of the dental cavity, and distinct growth processes within the apical third of the tooth, the pulp chamber narrows and subsequently divides into separate pulp compartments. In the present study, variable configurations of the pulp cavity were observed even in teeth of the same age and the same Triadan position. However, major trends of pulpar segmentation were identified for mandibular and maxillary cheek teeth. Generally, the number of separate pulp compartments within one tooth increased significantly with age. Interestingly, there was a trend of premolars having less interpulpar communication in young teeth and more often solitary pulp compartments in aged teeth compared to molars. Isolated and completely filled pulp horns were found only in the third molars (pulp horns 4 and 8 in mandibular and maxillary teeth, respectively) and in mandibular Triadan 09s (pulp horn 4). Senile teeth >20 years showed rudimentary endodontic cavities due to advanced abrasion and attrition, with interpulpar communications being absent.
Common pulp chamber
The common pulp chamber was found in six (20%) of 30 maxillary teeth and in nine (26%) of 35 mandibular teeth. However, a similar study found a higher prevalence of 26% in maxillary teeth, but a lower prevalence of 14% in mandibular teeth . Presumably, small intermissions between pulp compartments in maxillary teeth might not be detected by clinical computed tomography having a resolution of 1 mm, which was used by Windley et al. (2009) . Nonetheless, Windley et al. (2009)  stated the common pulp chamber to be the most frequent pulpar configuration in maxillary cheek teeth which is in line with data of our study.
It has been shown that pulp horn coalescences are more common in younger horses [1, 4]. The mean dental age of maxillary teeth displaying the common pulp chamber was 4 years in the present study. Dacre et al. (2008)  observed pulpar communication in teeth having a mean dental age of only 2 years. Furthermore, the oldest maxillary tooth in the present study showing the common pulp chamber was 9 years, whereas Windley et al. (2009)  observed the common pulp chamber only up to 6 years post eruption. The mean dental age of mandibular teeth displaying the common pulp chamber was 5 years in the present study, this being similar to results documented by Dacre et al. (2008) , who found the mean dental age of teeth having pulpar communication to be 4.5 years. The common pulp chamber of mandibular teeth was seen up to 15 years post eruption in the present study. Our results differ greatly from studies of Kirkland et al. (1996)  and Windley et al. (2009) , who found the common pulp chamber only in teeth <6 years, and <2 years, respectively. These differing results are probably one effect of the μCT inherent higher resolution compared to clinical computed tomography (used by Kirkland et al. (1996)  and Windley et al. (2009) ), which enabled us to detect even delicate and curved endodontic cavities.
Segmented pulp cavity
Once the common pulp chamber was divided into separate pulp compartments, configurations of the dental cavities varied and appeared unpredictable . Results indicate that in mandibular teeth >5 years solitary pulp compartments are frequently present, with still a minor risk left of teeth having all pulp horns connected in middle-aged and even in old teeth. Apparently, segmentation of the maxillary pulp system into separate solitary pulp compartments increased gradually between 2.5 and 8 years post eruption. As shown, maximal segmentation of the pulp cavity was not observed even in some aged teeth. It probably would be of clinical interest to determine the factors causing delayed pulpar segmentation in cheek teeth. Despite the observed gradual beginning of pulpar segmentation, further separation of the pulp system, perhaps once the root canals have been developed, may proceed slowly. However, to fully consider this question, further studies are needed providing closely spaced age groups between 5 and 20 years of tooth age, including large numbers of specimens.
In maxillary teeth ≤16 years having a segmented pulp cavity eight different configurations were observed in the present study. Hereby, two, three or four separate pulp compartments were displayed, with pulp horns appearing variably coalesced. However, in accordance with findings of Windley et al. (2009) , two configurations were most commonly observed (in ten (50%) of 20 teeth with segmented pulp systems): a) the pulp horns 1-3-5, 2 and 4 were separately derived from root canals I-III, II and IV, respectively, and b) pulp horns 1–3, 5, 2 and 4 were solely derived from root canals I, III, II and IV, respectively. The second most common configuration was observed in three teeth (15%), whereby “I-1-3-5-III” and “II-2-4-IV” was seen. In general, all teeth had at least two coalesced pulp horns present, these being most commonly the mesial pulp horns 1 – 3 (in 19 of 20 divided pulp systems; 95%) and the mesial pulp horns 3 – 5 (in 10 of 20 divided pulp systems; 50%). Similarly, observations of sectioned teeth by Dacre et al. (2008)  and studies of donkey teeth by Du Toit et al. (2008)  revealed coalescence of pulp horns 3 – 5 to be mostly present. It is noteworthy that separation of the pulp cavity into mesially and distally located pulp compartments occurred in 16 (80%) of 20 maxillary teeth. Conversely, coalescence of a mesial and a distal pulp horn was only seen in four (20%) of 20 divided pulp systems, this always being the buccal pulp horns 1 – 2. The maxillary pulp horn 3 was missing a corresponding root canal and therefore was always connected to pulp horns 1 and/or 5. Windley et al. (2009)  aptly mentioned that pulp horn 3 was split and appeared to be derived from the mesiobuccal and palatal roots.
Within segmented mandibular pulp systems four different configurations were observed. As already documented by Kirkland et al. (1996) , only two separate pulp compartments were developed in teeth of Triadan positions 06 to 10. In teeth of Triadan position 11, configurations of the pulp system within pulp horns 1 to 5 did not display additional patterns of segmentation, but a third solitary pulp compartment was established distally, with pulp horn 7 being derived from root canal III and terminating at an additional apical foramen. Conversely, Windley et al. (2009)  observed pulp horn 7 as being always connected to pulp horns 4 and 5 with no additional distal root canal being present.
Generally, in 32 (100%) mandibular teeth ≤17 years, distinct pulp horns were always connected: The mesial pulp horns 1 and 3 were coalesced as well as the distal pulp horns 2 and 5. However, Du Toit et al. (2008)  found only 37.5% of teeth having both these communications present, whereas Dacre et al. (2008)  only saw this pattern in one tooth. In the present study both coalescences were part of solitary pulp compartments. Furthermore, pulp horn 4 (if present and not isolated) was either connected to the distal or to the mesial pulp compartment, more commonly being part of the distal pulp compartment (11 of 19 teeth; 58%) compared to the mesial pulp compartment (eight of 19 teeth; 42%). Windley et al. (2009)  even documented the distal coalescence of pulp horn 4 in 68% of mandibular cheek teeth. Interestingly, only pulp horn 4 was seen to be isolated or absent, as observed by Windley et al. (2009) , this being found in three teeth >8 years. In accordance with other studies no further interpulpar communication between the mesial and the distal pulp compartment was seen [1, 2, 4]. This finding can be explained by the pronounced buccal enamel infolding (ectoflexid) in mandibular teeth reaching further apically compared to the lingual infolding (linguaflexid). Consequently, apart from pulp horn 4, the gross anatomy of mandibular teeth does not permit any further connection between the mesial and distal pulp compartments. If pulp horn 4 in mandibular teeth is exposed, either the mesial or the distal pulp cavity might be affected, or this pulp tissue could be necrotic due to isolation of the pulp horn. In turn, if exposure of pulp horns other than number 4 is diagnosed, their corresponding root canal can be clearly determined.
Root canals can take various pathways to the apex including branches, divisions and rejoining, as has been described in great detail in human dentistry. Features describing variations of the root canals comprise accessory, lateral and furcation canals, canal orifices, apical deltas and apical foramina . Many features of the root canals were recently observed in mandibular teeth . In the present study the mesial and distal root canal of mandibular teeth dispersed into two branches of varying length, width and route with gradually increasing age. Windley et al. (2009)  found two branches of the distal root canal even in all teeth >10 years and accordingly observed the root canals to unite occlusally to the roots. Kirkland et al. (1996) , who observed teeth up to eight years, found two root canals in the mesial root and one root canal in the distal root in teeth >5 years. Westenberger (2002)  observed a varying total number of up to three root canals in mandibular cheek teeth. In general, root canals of maxillary teeth appeared shorter compared to mandibular root canals, and were not divided except for the second premolars and the third molars. These teeth are macroscopically of larger size, thus giving space for additional pulpar tissue within roots in order to maintain the nutritional function.
Occlusal pulpar exposure is associated with previous pulpar insults followed by reduced or ceased deposition of dentine [26–28]. Multiple pulpar exposure (two and more pulp horns) is considered to indicate that the entire endodontic system is affected . However, Joest (1970)  stated that due to the pulp configuration a partial pulpitis does not necessarily affect all pulp horns. Casey and Tremaine (2010)  showed multiple defective secondary dentinal areas to be more prevalent in diseased mandibular teeth compared to maxillary teeth. Assumingly, this might be due to developing solitary pulp compartments which are composed of only one pulp horn and one root canal more frequently in maxillary teeth, compared to mandibular teeth which display solitary pulp compartments comprising up to three coalesced pulp horns. Furthermore, the most commonly defective area identified by Casey and Tremaine (2010)  was pulp horn 2 in maxillary teeth. This result tallies with findings in the present study whereby the solitary pulp compartment comprising only pulp horn 2 and root canal II was found in 11 (55%) of 20 teeth having a divided pulp cavity, thus being the second most common individual pulp compartment observed in maxillary cheek teeth.
As reported in clinical studies the outcome of endodontic therapy was more successful in mandibular compared to maxillary teeth [11, 13, 32]. A study by Carmalt and Barber (2004)  even reported only one root being affected in 100% of 14 mandibular teeth having a mean age of 5.3 years. Similar to this finding, in 69% of 32 mandibular teeth of the present study, which had a mean age of 9 years, the pulp cavity was maximally divided, presumably preventing the spread of pulpar infection to adjacent pulp compartments via mesio-distal connections. However, only in 23% of 26 maxillary teeth of the present study (mean age: 9 years) was maximal segmentation present, thus pulpar infection could spread to unaffected pulps via variable connections in the majority of teeth. Despite such high degree of variable pulpar configurations in maxillary teeth, we observed a division of the pulp cavity into mesially and distally located pulp compartments similar to mandibular teeth, with the solitary pulp compartments “4-IV”, “2-II” and “5-III” being most commonly present. Presumably, the knowledge  and selected treatment of solitary pulp compartments could simplify endodontic procedures and therefore improve clinical outcomes.