Original Paper Acta Anatómica 1992;143:130-138
Second Department of Oral Anatomy. School of Dentistry. Showa University. Tokyo.Japan
Key Words Rat Molar tooth Accessory canal Structure Distribution Formation
Three-Dimensional Observations of Accessory Canals in Mature and Developing Rat Molar Teeth
Abstract The structure, distribution and formation of accessory canals in the developing and mature molar teeth of rat mandibular jaws were investigated with scanning electron microscopy and with three-dimensional image analysis using serial lightmicroscopic sections. In the initial stage of the accessory canal formation, most of the canals appeared in the gaps of the epithelial root sheaths formed by their approaching each other in the initial stage of the root formation. However, some of the canals appeared in the slits which may be formed by the destruction of the epithelial root sheath in the root apex regions. When the gaps and slits were in vaded by blood vessels, the regions surrounding the vessels did not mineralize but became accessory canals. Usually, an accessory canal with one blood vessel con nected the periodontal ligament to the dental pulp; however, in some cases, the canals were broken off midway following the destruction of the vessels.
Introduction Abnormal root canals of teeth except for the main root canals have been variously termed: accessory canals [ 1-13], accessory foramina [5, 9, 14, 15], lateral canals [1, 5, 10. 16], aberrant canals [6] and pulpopcriodontal canals |17|. In the present study, these abnormal canals will be generically named accessory canals. These accessory canals were reported to be more fre quent in rodent teeth than in human teeth and were located in the interradicular root surfaces at all levels and the root apexes: but in most instances the canals might be found in the root furcations ]2, 6, 18. 19], Kovacs |4] thought that the root formation was sometimes disturbed by blood ves sels, which caused the accessory canals to appear. Seltzer |5] suggested that the accessory canals in the root apex regions were formed by the destruction of the epithelial
Received: July in. 1991 Accepted: August 19. 1991
root sheath and subsequently by the invasion of the blood vessels. Scott and Symons [10] presumed that the accessory canals, the aberrant openings of roots, were caused by a localized failure in the formation of the epithelial root sheath. On the other hand. Osborn and Ten Cate [17] thought that the accessory canals were produced by the imperfect fusion in the usual multiroot formation. Pre viously [20], we reported the formation process of the ‘seams', termed by Lester and Boyde [21 ] in developing rat molars, and suggested that most of the accessory canals were formed in the ‘seams', which were the meeting points of the epithelial root sheaths during the multiroot forma tion. The present study aims to investigate the structure, dis tribution and formation of the accessory canals in rat molars using scanning electron microscopy (SEM) and the
Mic Kuroiwa Second Department of Oral Anatomy School of Dentistry. Showa University 1-5-8 Halanodai. Shinagawa-ku. Tokyo 142 (Japan)
© 1992 Karger A G . Basel 0001-5180/92/1432-0130 $ 2.75A»
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
M. Kuroiwa T. Kodaka M. Abe S. Higashi
Materials and Methods Fifteen Sprague-Dawley rats were sacrificed so as to provide a growing series aged 1-9 weeks after birth. The animals were anesthe tized with ether. The mandibular jaws were extracted, subsequently fixed in 10% neutral formaldehyde for about I week and rinsed in run ning tap water. Scanning Electron Microscopy Half of the jaws were treated with 2-5% NaOCI for 3-5 h. and the first, second and third molars were extracted |2(). 22], The specimens were rinsed in running tap water, dehydrated with a graded ethanol series and dried at a critical point with CO,. After coating with a 10- to 15-nm platinum-palladium layer by ion sputtering, the specimens were observed under a Hitachi S-430 scanning electron microscope oper ated at 20 kV. Transmitted Light Microscopy The remaining jaws were demineralized in 10% ethylene diamine tetraacetic acid at pH 7.2. They were dehydrated with a graded etha nol series and embedded in paraffin. The serial sections were sliced at intervals of 7 pm in thickness in the mesiodistal, buccolingual and transverse planes of molars. The sections were stained with hematoxy lin and cosin (HE), Azan and Van Gicson stainings on every specimen glass. These preparations were serially photographed under a Nikon light microscope. Computer Graphic Image Analysis The outlines of Hertwig’s epithelial root sheath, blood vessels and root tissues including the dentine and cementum were traced on the serial LM micrographs. The micrographs were reproduced on trans parent plastic sheets with a Ricoh reduce/enlarge ricopv FT-4510. The sheets taken from serial 14-44 sections at intervals of 7 or 35 and 70 pm were superimposed on one another so that the basal lines of the struc tures were decided. They were fed into an NEC PC-9801RX personal computer using a Nikon digitizer tablet. The production of the com puter graphic images was performed by a Nikon Cosmozone 2SA three-dimensional image-analyzing system. The images were photo graphed with a Nikon F2 camera.
Results SEM Observation Figures 1-6 are the SEM micrographs of rat molars treated with NaOCI. Figure 1 shows a fractured surface of the first molar. One accessory canal in the mineralized floor of a pulp chamber was observed. In the mandibular jaw, the mineralization of the floors of the pulp chambers in the First, second and third molars already began from 3. 4 and 4 weeks after birth, respectively. Figures 2-5 are viewed from the root furcations, and figure 6 is viewed
from the dental pulp removed with NaOCI in the second molar. Generally, the accessory canals in the mineralized root tissues started from the periodontal ligament (fig. 1-5) and opened the dental pulp (fig. 1. 6). Most of the canals were found in the ‘seams' running from the root furcation towards the root apexes (fig.2-4). This type was observed in 31 out of 34 accessory canals in 42 molars. The percent age was 91 (31/34). When the inner suface of the dentine adjacent to the dental pulp was observed, there was an indistinct band rec ognized as a loose zone of dentinal tubules (fig.6). The den tinal tubules in this zone were arranged perpendicularly to the dentine surface, whereas in the surrounding areas the tubules ran obliquely. This zone which possessed the acces sory canal and brush-like fibrous structures corresponded to the ‘seam’ (fig.6). In the root apex regions, however, some of them did not exist in the ‘seams’ (fig.4). Such a type was observed in 3 out of 34 accessory canals (9%). From their appearance seen from the periodontal liga ment, the accessory canals were classified into three types: a volcano (fig.2). a simple hole (fig.3, 5) and a slit-like shape (fig.4). The volcano-like shape was found in the root furcation, the simple hole-like shape was very common, and the slit-like shape was sometimes seen in the root apex region. The accessory canals were more or less surrounded by cellular cementum, and the diameter measured in a range from about 14 to 100 pm. The frequency of the canals in the first, second and third molars was 75% (9/12), 60% (6/10) and 100% (7/7). respectively. The maximum number in one tooth was four canals observed in the third molar. The distribution of the accessory canals in one molar tooth was roughly divided into four different regions: the root furcation (fig. 1. 2). the transitional regions from the furcation to the succeeding roots or the basal regions of the multiroots (fig.3), the multiroot regions (fig.5; see fig. 10) and the root apex regions (fig.4). Table 1 shows the distri bution patterns of the accessory canals in the four regions of the mandibular molars. In developing teeth with immature roots, the accessory canals in the root edges were not included in the root apex regions but in the multiroot regions. That is, the first, the second and the third molar completely formed the root apexes in 6,6 and 7 weeks after birth, respectively. The first molar (fig.3) had four roots such as: the mesial, the distal, the buccal and the lingual root. The second (fig.2) and the third molars (fig.4) had usually three roots: the mesiobuccal. the mesiolingual and the distal root. The normal root positions arc shown in figures 2^k in addition, abnormal roots arc rarely observed in rat molars (fig.4). Table 2 shows the frequency of the accessory canal in each
131
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
three-dimensional image analyzer with serial light-micro scopic (LM) sections.
132
Kuroiwa/Kodaka/Abe/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rai Molars
Table 1. Distribution pattern (%: numbers in parentheses) of the accessorycanal in the four parts of the root regions of rat mandibular molars
Tooth
Fr
Bs
Rt
Ap
First molar (1/12) Second molar (1/10) Third molar (1/7)
8 (8/12) 10 (0/10) 14 (1/7)
67 (3/12) 0 (5/10) 14 (7/7)
25 (1/5) 50 (4/5) l(X) (2/3)
20 12 or 5 80 10 or 5 67 7 or 3
n
F r= Root furcation: Bs = basal regions of the multiroots: Rt = multiroot regions: A p= root apex regions.
Table 2. Frequency (%: numbers in parentheses) of the accessory canal in the different roots of rat mandibular molars
Tooth
Me
Me-Li
Li
Me-Bu
Bu
Di
First molar
42 (-5/12)
—
25 (3/12)
_
8 (1/12)
8 (1/12) 10 (1/10) 14 (1/7)
Second molar Third molar
_
40 (4/10) 57 (4/7)
_
20 (2/10) 43 (3/7)
_
n
12 10 7
Me = Mesial root; Li = lingual root; Bu = buccal root: Di = distal root; Mc-Li = mesiolingual root: Me-Bu = mcsiobuccal root.
LM Observation and Computer Graphic Image Analysis Figure 7a is an HE-staincd micrograph taken from buccolingual serial sections of a tooth germ of the first molar l
Fig. 1-6. SEM micrographs of rat mandibular molar teeth treated with NaOCI: mesiodistal fractured surface of the first molar 5 weeks after birth (1), bottom of the second molar 4 weeks after birth (2). bottom of the first molar 6 weeks after birth (3). bottom of the third molar 6 weeks after birth (4). root furcation side surface of the mesiolingual root near the root apex in the third molar 6 weeks after birth (5) and dental-pulp side surface of the buccal root near the root furcation in the second molar 6 weeks after birth (6). Arrow = Acces sory canal: S = ‘seam’; F = brush-like fibrous structures; Dp = dental pulp; Me = mesial root; Li = lingual root; Bu = buccal root; Di = distal root; Me-Li = mesiolingual root: Me-Bu = mesiobuccal root; A b=abnormal root. 1 x50. 2-4 x 40. 5 X 1,000. 6 x4(X).
week after birth. Epithelial root sheaths extending from the cervical region towards the future root furcation were observed. Based on 44. 7-pm-thick sections at intervals of 35 and 70 tim. the three-dimensional image of the epithelial root sheath was reconstructed (fig.7b). The photograph is viewed from the bottom of the tooth germ. The four posi tions of the future roots are presumed. The epithelial root sheath extends to the future root furcation, although the root sheaths have not reached the future ‘seam’ positions yet. Figure 8a is a Van-Gieson-stained micrograph of the mesial root taken from transverse serial sections of the first molar 3 weeks after birth. A fine slit is seen on the distal side of the mineralized mesial root. Based on 14 serial 7pm-thick sections, the three-dimensional image of the root tissues including mineralized dentine and cementum was reconstructed (fig.8b). The photograph is seen from the mesial side. The slit or developing ‘seam’ incompletely closes in the cervical-side region. The mesial-side region of the root is still widely open. In serial sections of the first eight molars with develop ing roots 2-5 weeks after birth, five accessory canals were
133
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
root except for the root furcation. In the first molar, the fre quency decreased in the orderof the mesial, the lingual, the distal and buccal roots, whereas in the second and third molars, the frequency decreased in the order of the mesiolingual. the mesiobuccal and the distal root.
134
Kuroiwa/Kodaka/Abc/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rat Molars
Fig. 7. Tooth germ of the mandibular first molar I week titter birth, a Buccolingual section with ME staining, b Reconstruction of the epithelial root sheath based on 44 sections at intervals of 35 or 70 pm. Viewed from the bottom of the tooth germ. Me = Mesial side: Li = lingual side: Bu - buccal side: Oi = distal side: S = initial 'scam', a x32. b about x.30. Fig. 8. Mesial root of the first molar 3 weeks after birth, a Trans verse section with Van Gicson staining, b Reconstruction of the root tissues based on 14 sections ¡it intervals of 7 pm. Viewed from the dis tal side. S = Developing 'seam': Me = mesial side; Di = distal side, a x80. b about X80. Fig. 9. Floor of a pulp chamber in the first molar 2 weeks after birth, a Mesiodistal section with Azan staining, b Reconstruction based on 15 sections at intervals of 7 pm. Viewed from the dental pa pilla. F. = Epithelial root sheath; R = root tissues: Bv = blood vessel: arrow initial accessory canal; Dp = dental papilla, a x320. b about x 300.
ever, in one example of the first molar 7 weeks after birth, the accessory canal was obstructed midway. Figures 11a and b are two Van-Gieson-stained micrographs of the floor of a pulp chamber taken from the mesiodistal serial sec tions. From such LM micrographs, one accessory canal from the periodontal ligament to the dental pulp was expected in the floor, although the canal contained no blood vessels but organic remnants (fig. I la). Brush-like fibrous structures were seen on the dentine surface adja cent to the dental pulp (fig. 1lb). Based on 27 serial 7-iimthick sections, the three-dimensional image of the outline of the mineralized root tissues was created. The photo graph was viewed from the buccolingual cut plane. The accessory canal was broken off halfway.
Discussion In this study, the structure, distribution and frequency of the accessory canals of rat molar teeth were investigated by SEM. The structures have been mainly reported by using light microscopy with demineralized sections 11-3,5,6, 13, 18. 19.23], but there are a few reports using SEM [9. 15.20. 24]. The present SEM observations, which were performed after the removal of organic tissues with NaOCI, revealed that most of the accessory canals existed in the ‘seams', termed by Lester and Boyde [21). which run from the root furcation towards the root apexes. It has been suggested that the formation of the accessory canals is caused by the disturbance of blood vessels [3] and the destruction of the epithelial root sheath [5] or the local ized failure of the epithelial root sheath [ 10]. However, our previous [20] and present studies based on LM and SEM observations more or less supported the suggestion by Osborn and Ten Cate [17], That is. we strongly suggest that most of the accessory canals are initially formed by the gaps derived from the incomplete contact of the epithelial root sheaths, in other words the imperfect 'seams’, and the for mation subsequently carried out by the invasion of blood vessels into the imperfect 'seams’. If Kovacs's suggestion [4] is correct, the accessory canals would be observed ev erywhere. However, we never obtained such findings. In human teeth, it has been reported that the accessor)' canals are frequently seen in the root apex regions [ 1 ,5 1as well as the root furcation [7. 8. 11, 12], In rat molar teeth observed in this study, some of the accessoiy canals in the root apex regions were also found in the ‘seams’. Flowever. the disturbance by blood vessels [3] and the destruction [5] or failure of the epithelial root sheath [ 10) may rarely cause the formation of the accessory canals in the root apex Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
found. Figure 9a is an Azan-stained micrograph of the future floor of a pulp chamber taken from the mesiodistal serial sections of the first molar 2 weeks after birth. One blood vessel with a small quantity of mesenchymal tissue entered the gap of the epithelial root sheath and divided the root tissues with an odontoblast layer into two parts. Based on 15 serial 7-um-thick sections, the three-dimensional image of the future floor of a pulp chamber was recon structed (fig.9b). The photograph was taken from the den tal papilla. The vessel was completely surrounded by the epithelial rooth sheath or the developing 'seam'; in other words, the vessel invaded the gap. which was the initial stage of the accessory canal. The root tissues had not yet completely covered the floor. Figures 10a and b are Azan- and HE-stained micro graphs of the mesial root taken from the mesiodistal serial sections of the first molar 3 weeks after birth. One blood vessel with a relatively large quantity of mesenchymal tis sue just divides the root tissues and the epithelial root sheat in figure 10a, but in figure 10b the vessel divides the epithe lial root sheath into two parts with no root tissues. How ever. the inner and the outer epithelial cell layers adjacent to the vessel were joined intact to each other; in other words, the edge of the epithelial root sheath is unbroken. Based on 19 serial 7-um-thick sections, the three-dimen sional image of the mesial root was reconstructed (fig. 10c). The photograph was viewed from the dental pulp. The epi thelial root sheath partially surrounded the vessel. The gap region was a part of the developing ‘seam' and the initial stage of the accessory canal. In the fully formed multiroots, the accessory canals usu ally connected the periodontal ligament and the dental pulp as shown in the SEM micrographs (fig. 1-6). How-
136
Kuroiwa/Kodaka/Abe/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rai Molars
fl»
regions, because the cell activity of the epithelial root sheaths is in the final stage of the development of the sheath [17, 25. 26]. Such accessory canals are not directly con cerned with the ‘seam’ formation (fig.4) even if the canals happen to exist in the ‘seam’. According to several reports using human teeth 110. 18, 27], a single or one pair of blood vessels existed in one accessory canal. In these LM observations using rats, one accessory canal contained one blood vessel in all of the five examples (fig.9,10), although the slit-like canals in the root apex regions might contain two or more blood vessels (fig-4). In many cases, the accessory canals connected the perio dontal ligament with the dental pulp as shown in figures 1.5 and 6 |1—1, 6, 10, 13]. However. Yoshida ct al. |23] sug gested that the accessory canals did not connect them in some cases. In this study, such a case with no blood vessels was found in only one example (fig. 11). This case could be caused by the blood vessel in the developing accessory canal creating a stricture and immediately a closure during the root formation, and then the vessels finally decaying after the dentine had partially been added in the region. Such accessory canals might be found in the case of a small quantity of mesenchymal tissue surrounding the canals as shown in figure 9. In rodent teeth, it has been reported that the accessory canals frequently appear in the root furcation |2. 6. 18-20]. Certainly, this region where the ‘seams' join together is apt to form incomplete junctions. In this study, a few examples were observed (fig. 1, 2; table 1): however, the accessory canals in the mandibular first molar were mostly found in the basal regions of the multiroots. The accessory canals may be initially formed by the complicated movement of the epithelial root sheath as the sheath abruptly bends in these regions (fig.3). Finally, the formation of the canals will be carried out in the gaps of the ‘seams’.
In the second and third molars, the canals appeared in the multiroot and the root apex regions more than in the root furcation and the basal regions (table 1). In such cases, the formation of the accessory canals in the second and third molars may be caused by the rapid and rough forma tion of the multiroots compared with those in the first molar [22]. The phenomena may induce the incomplete meeting of the epithelial root sheaths and finally the gaps for the future accessory canals. The frequency of the accessory canal in each root of the molars with multiroots was investigated (table 2). From the total results including the first, second and third molars, the accessory canals appeared in the roots of the mesial and lingual sides more than in those of the distal and buccal sides. According to Fujita [19], the meeting of the epithelial root sheaths in the first molar of a mouse was seen later on the mesiolingual side than the distal side; therefore, the above-mentioned appearance may be influenced by the delay of the meeting of the epithelial root sheaths. It was reported that the accessory canals were dotted in the shallow grooves of the dentine surface adjacent to the dental pulp in human deciduous teeth [24]. In rat molars, however, the ‘seam’ region where the accessory canals opened agreed with the loose zone of the dentinal tubules with brush-like structures (fig.6, 1lb). The formation of the brush-like fibrous structures may be related to the activity of odontoblasts in the ‘seam’ region; however, the details could not be elucidated in this study.
Acknowledgments Wc thank Mr. M. Yamada. from the Second Department of Oral Anatomy. Showa University, for his continuous cooperation.
137
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Fig. 10. Mesial root on the root furcation side of the first molar 3 weeks after birth, a, b Mcsiodistal sections with Azan and HE stainings. c Reconstruction based on 19 sections at intervals of 7 pm. Viewed from the dental pulp. E = Epithelial root sheath: R = root tis sues; Bv= blood vessel; arrow initial accessory canal; Dp = dental pulp, a, b x320. c about x 1(H). Fig. 11. Floor of a pulp chamber in the first molar 7 weeks after birth, a, b Mcsiodistal sections with Van Gicson staining. An acces sory canal anil brush-like fibrous structures arc seen, c Reconstruction of the outlines of the mineralized root tissues based on 27 sections at intervals of 7 pm. Viewed from the buccolingual cut plane. Arrow = Accessory canal: E brush-like fibrous structures: double arrows = in terrupted accessory canal: Dp = dental pulp, a, b x 320. c about x 180.
1 Seltzer S. Bender IB. Ziontz M: The imerrelationship of pulp and periodontal disease. Oral Surg 1963:16:1474—1490. 2 Winter GB. Kramer IRH: Changes in perio dontal membrane and bone following exper imental pulpal injury in deciduous molar teeth in kittens. Arch Oral Biol 1965:10:279-289. 3 Rubach WC. Mitchell DF: Periodontal dis ease. accessory canals and pulp pathosis. J Periodontol 1965:36:34—38. 4 Kovacs I: Contribution to the ontogenetic mor phology of roots of human teeth. .1 Dent Res 1967:46:865-874. 5 Seltzer S: Endodontology. New York. McGraw-Hill. 1971. pp 1-32 . 417-425. 6 Ulmanskv M. Sela J. Hirschfeld Z: Accessory canals in rat molars. J Dent Res 1972:51:879. 7 Lowman .IV. Burke RS. Pelleu GB: Patent accessory canals: Incidence in molar furcation region. Oral Surg 1973:36:580-584. 8 Burch JG. I lulcn S: A study of the presence of accessory foramina and the topography of molar furcations. Oral Surg 1974:38:451-455. 9 Koenigs JF. Brilliant JD. Foreman DW: Pre liminary scanning electron microscope investi gations of accessory foramina in the furcation areas of human molar teeth. Oral Surg 1974: 38:773-782. Ill Scott JII. Symons NBB: Introduction to Den tal Anatomy. Edinburgh. Churchill Living stone. 1977.'pp 242-254.
11 Gutmann .IL: Prevalence, location, and patency of accessory canals in the furcation region of permanent molars. J Periodontol 1978:49:21-26. 12 Ringelstcin D. Scow WK: The prevalence of furcation foramina in primary molars. Pediatr Dentist 1989:11:198-202. 13 Avery JK: Pulp: in Bhaskar SN (ed): Orban's Oral Histology and Embryology, ed 10. St Louis. Mostly. 1986. pp 135-197. 14 Green D: A slereomicroscopic study of the root apices of 400 maxillary and mandibular anterior teeth. Oral Surg 1956:9:1224—1232. 15 Perlich MA. Reader A. Foreman DW: A scan ning electron microscopic investigation of accessory foramens on the pulpal floor of human molars. .1 Endodonlol 1981;7:402-406. 16 Vertueci FJ. Williams RG: Furcation canals in the human mandibular first molar. Oral Surg 1974:38.-308-314. 17 Osborn JW. Ten Cate AR: Advanced Dental Histology, ed 3. Bristol. Wright & Sons. 1976. pp 123-127. 18 Takahashi S: On the abnormal structure of the bifurcation of multirooted teeth (in Japanese). Kokubyo Z 1930:4:26-28. 19 Fujila Y: Formation of subpulpal wall in mouse maxillary first molar (in Japanese). Jpn J Oral Biol 1978:20:221-228. 20 Kuroiwa M. Kodaka T. Higashi S: Morpholog ical study on the 'seams' in the multiroot for mation of rat molar teeth. Acta Anat 1991:142: 6-14.
21 Lester KS. Bovde A: Scanning electron microscopy of developing roots of molar teeth of the laboratory rat. J Ultrastruct Res 1970: 33:80-94. 22 Kuroiwa M. Saito T. Kodaka T. Higashi S: Scanning electron microscopic study of the developing roots in rat molars (in Japanese). J Showa Univ Dent Soe 1988:8:452-456. 23 Yoshida II. Yakushiji M. Sugihara A. Tanaka K. Taguchi M. Machida Y: Accessory canals at floor of the pulp chamber of primary molars (in Japanese). Shikwa Gakuho 1975,75:580-585. 24 Goto G. Chang Y. Hosoya Y: A scanning elec tron microscopic investigation of accessory foramens on the pulpal floor of human primary molars (in Japanese). Jpn J Pediatr Dent 1990: 28:371-380. 25 Orban B. Mueller E: The development of the bifurcation of multirooted teeth. J Am Dent Assoc 1929:16:297-319. 26 Sharawv M. Bhussrv BR: Development and growth of teeth; in Bhasker SN (ed): Orban's Oral Histology and Embryology, ed 10. St Louis. Mosby. 1986. pp 24-44. 27 Kramer IRH: The vascular architecture of the human dental pulp. Arch Oral Biol 1960:2: 177-189.
138
Kuroiwa/Kodaka/Abe/Higashi
Accessory Canals in Rat Molars
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
References
Second Department of Oral Anatomy. School of Dentistry. Showa University. Tokyo.Japan
Key Words Rat Molar tooth Accessory canal Structure Distribution Formation
Three-Dimensional Observations of Accessory Canals in Mature and Developing Rat Molar Teeth
Abstract The structure, distribution and formation of accessory canals in the developing and mature molar teeth of rat mandibular jaws were investigated with scanning electron microscopy and with three-dimensional image analysis using serial lightmicroscopic sections. In the initial stage of the accessory canal formation, most of the canals appeared in the gaps of the epithelial root sheaths formed by their approaching each other in the initial stage of the root formation. However, some of the canals appeared in the slits which may be formed by the destruction of the epithelial root sheath in the root apex regions. When the gaps and slits were in vaded by blood vessels, the regions surrounding the vessels did not mineralize but became accessory canals. Usually, an accessory canal with one blood vessel con nected the periodontal ligament to the dental pulp; however, in some cases, the canals were broken off midway following the destruction of the vessels.
Introduction Abnormal root canals of teeth except for the main root canals have been variously termed: accessory canals [ 1-13], accessory foramina [5, 9, 14, 15], lateral canals [1, 5, 10. 16], aberrant canals [6] and pulpopcriodontal canals |17|. In the present study, these abnormal canals will be generically named accessory canals. These accessory canals were reported to be more fre quent in rodent teeth than in human teeth and were located in the interradicular root surfaces at all levels and the root apexes: but in most instances the canals might be found in the root furcations ]2, 6, 18. 19], Kovacs |4] thought that the root formation was sometimes disturbed by blood ves sels, which caused the accessory canals to appear. Seltzer |5] suggested that the accessory canals in the root apex regions were formed by the destruction of the epithelial
Received: July in. 1991 Accepted: August 19. 1991
root sheath and subsequently by the invasion of the blood vessels. Scott and Symons [10] presumed that the accessory canals, the aberrant openings of roots, were caused by a localized failure in the formation of the epithelial root sheath. On the other hand. Osborn and Ten Cate [17] thought that the accessory canals were produced by the imperfect fusion in the usual multiroot formation. Pre viously [20], we reported the formation process of the ‘seams', termed by Lester and Boyde [21 ] in developing rat molars, and suggested that most of the accessory canals were formed in the ‘seams', which were the meeting points of the epithelial root sheaths during the multiroot forma tion. The present study aims to investigate the structure, dis tribution and formation of the accessory canals in rat molars using scanning electron microscopy (SEM) and the
Mic Kuroiwa Second Department of Oral Anatomy School of Dentistry. Showa University 1-5-8 Halanodai. Shinagawa-ku. Tokyo 142 (Japan)
© 1992 Karger A G . Basel 0001-5180/92/1432-0130 $ 2.75A»
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
M. Kuroiwa T. Kodaka M. Abe S. Higashi
Materials and Methods Fifteen Sprague-Dawley rats were sacrificed so as to provide a growing series aged 1-9 weeks after birth. The animals were anesthe tized with ether. The mandibular jaws were extracted, subsequently fixed in 10% neutral formaldehyde for about I week and rinsed in run ning tap water. Scanning Electron Microscopy Half of the jaws were treated with 2-5% NaOCI for 3-5 h. and the first, second and third molars were extracted |2(). 22], The specimens were rinsed in running tap water, dehydrated with a graded ethanol series and dried at a critical point with CO,. After coating with a 10- to 15-nm platinum-palladium layer by ion sputtering, the specimens were observed under a Hitachi S-430 scanning electron microscope oper ated at 20 kV. Transmitted Light Microscopy The remaining jaws were demineralized in 10% ethylene diamine tetraacetic acid at pH 7.2. They were dehydrated with a graded etha nol series and embedded in paraffin. The serial sections were sliced at intervals of 7 pm in thickness in the mesiodistal, buccolingual and transverse planes of molars. The sections were stained with hematoxy lin and cosin (HE), Azan and Van Gicson stainings on every specimen glass. These preparations were serially photographed under a Nikon light microscope. Computer Graphic Image Analysis The outlines of Hertwig’s epithelial root sheath, blood vessels and root tissues including the dentine and cementum were traced on the serial LM micrographs. The micrographs were reproduced on trans parent plastic sheets with a Ricoh reduce/enlarge ricopv FT-4510. The sheets taken from serial 14-44 sections at intervals of 7 or 35 and 70 pm were superimposed on one another so that the basal lines of the struc tures were decided. They were fed into an NEC PC-9801RX personal computer using a Nikon digitizer tablet. The production of the com puter graphic images was performed by a Nikon Cosmozone 2SA three-dimensional image-analyzing system. The images were photo graphed with a Nikon F2 camera.
Results SEM Observation Figures 1-6 are the SEM micrographs of rat molars treated with NaOCI. Figure 1 shows a fractured surface of the first molar. One accessory canal in the mineralized floor of a pulp chamber was observed. In the mandibular jaw, the mineralization of the floors of the pulp chambers in the First, second and third molars already began from 3. 4 and 4 weeks after birth, respectively. Figures 2-5 are viewed from the root furcations, and figure 6 is viewed
from the dental pulp removed with NaOCI in the second molar. Generally, the accessory canals in the mineralized root tissues started from the periodontal ligament (fig. 1-5) and opened the dental pulp (fig. 1. 6). Most of the canals were found in the ‘seams' running from the root furcation towards the root apexes (fig.2-4). This type was observed in 31 out of 34 accessory canals in 42 molars. The percent age was 91 (31/34). When the inner suface of the dentine adjacent to the dental pulp was observed, there was an indistinct band rec ognized as a loose zone of dentinal tubules (fig.6). The den tinal tubules in this zone were arranged perpendicularly to the dentine surface, whereas in the surrounding areas the tubules ran obliquely. This zone which possessed the acces sory canal and brush-like fibrous structures corresponded to the ‘seam’ (fig.6). In the root apex regions, however, some of them did not exist in the ‘seams’ (fig.4). Such a type was observed in 3 out of 34 accessory canals (9%). From their appearance seen from the periodontal liga ment, the accessory canals were classified into three types: a volcano (fig.2). a simple hole (fig.3, 5) and a slit-like shape (fig.4). The volcano-like shape was found in the root furcation, the simple hole-like shape was very common, and the slit-like shape was sometimes seen in the root apex region. The accessory canals were more or less surrounded by cellular cementum, and the diameter measured in a range from about 14 to 100 pm. The frequency of the canals in the first, second and third molars was 75% (9/12), 60% (6/10) and 100% (7/7). respectively. The maximum number in one tooth was four canals observed in the third molar. The distribution of the accessory canals in one molar tooth was roughly divided into four different regions: the root furcation (fig. 1. 2). the transitional regions from the furcation to the succeeding roots or the basal regions of the multiroots (fig.3), the multiroot regions (fig.5; see fig. 10) and the root apex regions (fig.4). Table 1 shows the distri bution patterns of the accessory canals in the four regions of the mandibular molars. In developing teeth with immature roots, the accessory canals in the root edges were not included in the root apex regions but in the multiroot regions. That is, the first, the second and the third molar completely formed the root apexes in 6,6 and 7 weeks after birth, respectively. The first molar (fig.3) had four roots such as: the mesial, the distal, the buccal and the lingual root. The second (fig.2) and the third molars (fig.4) had usually three roots: the mesiobuccal. the mesiolingual and the distal root. The normal root positions arc shown in figures 2^k in addition, abnormal roots arc rarely observed in rat molars (fig.4). Table 2 shows the frequency of the accessory canal in each
131
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
three-dimensional image analyzer with serial light-micro scopic (LM) sections.
132
Kuroiwa/Kodaka/Abe/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rai Molars
Table 1. Distribution pattern (%: numbers in parentheses) of the accessorycanal in the four parts of the root regions of rat mandibular molars
Tooth
Fr
Bs
Rt
Ap
First molar (1/12) Second molar (1/10) Third molar (1/7)
8 (8/12) 10 (0/10) 14 (1/7)
67 (3/12) 0 (5/10) 14 (7/7)
25 (1/5) 50 (4/5) l(X) (2/3)
20 12 or 5 80 10 or 5 67 7 or 3
n
F r= Root furcation: Bs = basal regions of the multiroots: Rt = multiroot regions: A p= root apex regions.
Table 2. Frequency (%: numbers in parentheses) of the accessory canal in the different roots of rat mandibular molars
Tooth
Me
Me-Li
Li
Me-Bu
Bu
Di
First molar
42 (-5/12)
—
25 (3/12)
_
8 (1/12)
8 (1/12) 10 (1/10) 14 (1/7)
Second molar Third molar
_
40 (4/10) 57 (4/7)
_
20 (2/10) 43 (3/7)
_
n
12 10 7
Me = Mesial root; Li = lingual root; Bu = buccal root: Di = distal root; Mc-Li = mesiolingual root: Me-Bu = mcsiobuccal root.
LM Observation and Computer Graphic Image Analysis Figure 7a is an HE-staincd micrograph taken from buccolingual serial sections of a tooth germ of the first molar l
Fig. 1-6. SEM micrographs of rat mandibular molar teeth treated with NaOCI: mesiodistal fractured surface of the first molar 5 weeks after birth (1), bottom of the second molar 4 weeks after birth (2). bottom of the first molar 6 weeks after birth (3). bottom of the third molar 6 weeks after birth (4). root furcation side surface of the mesiolingual root near the root apex in the third molar 6 weeks after birth (5) and dental-pulp side surface of the buccal root near the root furcation in the second molar 6 weeks after birth (6). Arrow = Acces sory canal: S = ‘seam’; F = brush-like fibrous structures; Dp = dental pulp; Me = mesial root; Li = lingual root; Bu = buccal root; Di = distal root; Me-Li = mesiolingual root: Me-Bu = mesiobuccal root; A b=abnormal root. 1 x50. 2-4 x 40. 5 X 1,000. 6 x4(X).
week after birth. Epithelial root sheaths extending from the cervical region towards the future root furcation were observed. Based on 44. 7-pm-thick sections at intervals of 35 and 70 tim. the three-dimensional image of the epithelial root sheath was reconstructed (fig.7b). The photograph is viewed from the bottom of the tooth germ. The four posi tions of the future roots are presumed. The epithelial root sheath extends to the future root furcation, although the root sheaths have not reached the future ‘seam’ positions yet. Figure 8a is a Van-Gieson-stained micrograph of the mesial root taken from transverse serial sections of the first molar 3 weeks after birth. A fine slit is seen on the distal side of the mineralized mesial root. Based on 14 serial 7pm-thick sections, the three-dimensional image of the root tissues including mineralized dentine and cementum was reconstructed (fig.8b). The photograph is seen from the mesial side. The slit or developing ‘seam’ incompletely closes in the cervical-side region. The mesial-side region of the root is still widely open. In serial sections of the first eight molars with develop ing roots 2-5 weeks after birth, five accessory canals were
133
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
root except for the root furcation. In the first molar, the fre quency decreased in the orderof the mesial, the lingual, the distal and buccal roots, whereas in the second and third molars, the frequency decreased in the order of the mesiolingual. the mesiobuccal and the distal root.
134
Kuroiwa/Kodaka/Abc/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rat Molars
Fig. 7. Tooth germ of the mandibular first molar I week titter birth, a Buccolingual section with ME staining, b Reconstruction of the epithelial root sheath based on 44 sections at intervals of 35 or 70 pm. Viewed from the bottom of the tooth germ. Me = Mesial side: Li = lingual side: Bu - buccal side: Oi = distal side: S = initial 'scam', a x32. b about x.30. Fig. 8. Mesial root of the first molar 3 weeks after birth, a Trans verse section with Van Gicson staining, b Reconstruction of the root tissues based on 14 sections ¡it intervals of 7 pm. Viewed from the dis tal side. S = Developing 'seam': Me = mesial side; Di = distal side, a x80. b about X80. Fig. 9. Floor of a pulp chamber in the first molar 2 weeks after birth, a Mesiodistal section with Azan staining, b Reconstruction based on 15 sections at intervals of 7 pm. Viewed from the dental pa pilla. F. = Epithelial root sheath; R = root tissues: Bv = blood vessel: arrow initial accessory canal; Dp = dental papilla, a x320. b about x 300.
ever, in one example of the first molar 7 weeks after birth, the accessory canal was obstructed midway. Figures 11a and b are two Van-Gieson-stained micrographs of the floor of a pulp chamber taken from the mesiodistal serial sec tions. From such LM micrographs, one accessory canal from the periodontal ligament to the dental pulp was expected in the floor, although the canal contained no blood vessels but organic remnants (fig. I la). Brush-like fibrous structures were seen on the dentine surface adja cent to the dental pulp (fig. 1lb). Based on 27 serial 7-iimthick sections, the three-dimensional image of the outline of the mineralized root tissues was created. The photo graph was viewed from the buccolingual cut plane. The accessory canal was broken off halfway.
Discussion In this study, the structure, distribution and frequency of the accessory canals of rat molar teeth were investigated by SEM. The structures have been mainly reported by using light microscopy with demineralized sections 11-3,5,6, 13, 18. 19.23], but there are a few reports using SEM [9. 15.20. 24]. The present SEM observations, which were performed after the removal of organic tissues with NaOCI, revealed that most of the accessory canals existed in the ‘seams', termed by Lester and Boyde [21). which run from the root furcation towards the root apexes. It has been suggested that the formation of the accessory canals is caused by the disturbance of blood vessels [3] and the destruction of the epithelial root sheath [5] or the local ized failure of the epithelial root sheath [ 10]. However, our previous [20] and present studies based on LM and SEM observations more or less supported the suggestion by Osborn and Ten Cate [17], That is. we strongly suggest that most of the accessory canals are initially formed by the gaps derived from the incomplete contact of the epithelial root sheaths, in other words the imperfect 'seams’, and the for mation subsequently carried out by the invasion of blood vessels into the imperfect 'seams’. If Kovacs's suggestion [4] is correct, the accessory canals would be observed ev erywhere. However, we never obtained such findings. In human teeth, it has been reported that the accessor)' canals are frequently seen in the root apex regions [ 1 ,5 1as well as the root furcation [7. 8. 11, 12], In rat molar teeth observed in this study, some of the accessoiy canals in the root apex regions were also found in the ‘seams’. Flowever. the disturbance by blood vessels [3] and the destruction [5] or failure of the epithelial root sheath [ 10) may rarely cause the formation of the accessory canals in the root apex Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
found. Figure 9a is an Azan-stained micrograph of the future floor of a pulp chamber taken from the mesiodistal serial sections of the first molar 2 weeks after birth. One blood vessel with a small quantity of mesenchymal tissue entered the gap of the epithelial root sheath and divided the root tissues with an odontoblast layer into two parts. Based on 15 serial 7-um-thick sections, the three-dimensional image of the future floor of a pulp chamber was recon structed (fig.9b). The photograph was taken from the den tal papilla. The vessel was completely surrounded by the epithelial rooth sheath or the developing 'seam'; in other words, the vessel invaded the gap. which was the initial stage of the accessory canal. The root tissues had not yet completely covered the floor. Figures 10a and b are Azan- and HE-stained micro graphs of the mesial root taken from the mesiodistal serial sections of the first molar 3 weeks after birth. One blood vessel with a relatively large quantity of mesenchymal tis sue just divides the root tissues and the epithelial root sheat in figure 10a, but in figure 10b the vessel divides the epithe lial root sheath into two parts with no root tissues. How ever. the inner and the outer epithelial cell layers adjacent to the vessel were joined intact to each other; in other words, the edge of the epithelial root sheath is unbroken. Based on 19 serial 7-um-thick sections, the three-dimen sional image of the mesial root was reconstructed (fig. 10c). The photograph was viewed from the dental pulp. The epi thelial root sheath partially surrounded the vessel. The gap region was a part of the developing ‘seam' and the initial stage of the accessory canal. In the fully formed multiroots, the accessory canals usu ally connected the periodontal ligament and the dental pulp as shown in the SEM micrographs (fig. 1-6). How-
136
Kuroiwa/Kodaka/Abe/Higashi
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Accessory Canals in Rai Molars
fl»
regions, because the cell activity of the epithelial root sheaths is in the final stage of the development of the sheath [17, 25. 26]. Such accessory canals are not directly con cerned with the ‘seam’ formation (fig.4) even if the canals happen to exist in the ‘seam’. According to several reports using human teeth 110. 18, 27], a single or one pair of blood vessels existed in one accessory canal. In these LM observations using rats, one accessory canal contained one blood vessel in all of the five examples (fig.9,10), although the slit-like canals in the root apex regions might contain two or more blood vessels (fig-4). In many cases, the accessory canals connected the perio dontal ligament with the dental pulp as shown in figures 1.5 and 6 |1—1, 6, 10, 13]. However. Yoshida ct al. |23] sug gested that the accessory canals did not connect them in some cases. In this study, such a case with no blood vessels was found in only one example (fig. 11). This case could be caused by the blood vessel in the developing accessory canal creating a stricture and immediately a closure during the root formation, and then the vessels finally decaying after the dentine had partially been added in the region. Such accessory canals might be found in the case of a small quantity of mesenchymal tissue surrounding the canals as shown in figure 9. In rodent teeth, it has been reported that the accessory canals frequently appear in the root furcation |2. 6. 18-20]. Certainly, this region where the ‘seams' join together is apt to form incomplete junctions. In this study, a few examples were observed (fig. 1, 2; table 1): however, the accessory canals in the mandibular first molar were mostly found in the basal regions of the multiroots. The accessory canals may be initially formed by the complicated movement of the epithelial root sheath as the sheath abruptly bends in these regions (fig.3). Finally, the formation of the canals will be carried out in the gaps of the ‘seams’.
In the second and third molars, the canals appeared in the multiroot and the root apex regions more than in the root furcation and the basal regions (table 1). In such cases, the formation of the accessory canals in the second and third molars may be caused by the rapid and rough forma tion of the multiroots compared with those in the first molar [22]. The phenomena may induce the incomplete meeting of the epithelial root sheaths and finally the gaps for the future accessory canals. The frequency of the accessory canal in each root of the molars with multiroots was investigated (table 2). From the total results including the first, second and third molars, the accessory canals appeared in the roots of the mesial and lingual sides more than in those of the distal and buccal sides. According to Fujita [19], the meeting of the epithelial root sheaths in the first molar of a mouse was seen later on the mesiolingual side than the distal side; therefore, the above-mentioned appearance may be influenced by the delay of the meeting of the epithelial root sheaths. It was reported that the accessory canals were dotted in the shallow grooves of the dentine surface adjacent to the dental pulp in human deciduous teeth [24]. In rat molars, however, the ‘seam’ region where the accessory canals opened agreed with the loose zone of the dentinal tubules with brush-like structures (fig.6, 1lb). The formation of the brush-like fibrous structures may be related to the activity of odontoblasts in the ‘seam’ region; however, the details could not be elucidated in this study.
Acknowledgments Wc thank Mr. M. Yamada. from the Second Department of Oral Anatomy. Showa University, for his continuous cooperation.
137
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
Fig. 10. Mesial root on the root furcation side of the first molar 3 weeks after birth, a, b Mcsiodistal sections with Azan and HE stainings. c Reconstruction based on 19 sections at intervals of 7 pm. Viewed from the dental pulp. E = Epithelial root sheath: R = root tis sues; Bv= blood vessel; arrow initial accessory canal; Dp = dental pulp, a, b x320. c about x 1(H). Fig. 11. Floor of a pulp chamber in the first molar 7 weeks after birth, a, b Mcsiodistal sections with Van Gicson staining. An acces sory canal anil brush-like fibrous structures arc seen, c Reconstruction of the outlines of the mineralized root tissues based on 27 sections at intervals of 7 pm. Viewed from the buccolingual cut plane. Arrow = Accessory canal: E brush-like fibrous structures: double arrows = in terrupted accessory canal: Dp = dental pulp, a, b x 320. c about x 180.
1 Seltzer S. Bender IB. Ziontz M: The imerrelationship of pulp and periodontal disease. Oral Surg 1963:16:1474—1490. 2 Winter GB. Kramer IRH: Changes in perio dontal membrane and bone following exper imental pulpal injury in deciduous molar teeth in kittens. Arch Oral Biol 1965:10:279-289. 3 Rubach WC. Mitchell DF: Periodontal dis ease. accessory canals and pulp pathosis. J Periodontol 1965:36:34—38. 4 Kovacs I: Contribution to the ontogenetic mor phology of roots of human teeth. .1 Dent Res 1967:46:865-874. 5 Seltzer S: Endodontology. New York. McGraw-Hill. 1971. pp 1-32 . 417-425. 6 Ulmanskv M. Sela J. Hirschfeld Z: Accessory canals in rat molars. J Dent Res 1972:51:879. 7 Lowman .IV. Burke RS. Pelleu GB: Patent accessory canals: Incidence in molar furcation region. Oral Surg 1973:36:580-584. 8 Burch JG. I lulcn S: A study of the presence of accessory foramina and the topography of molar furcations. Oral Surg 1974:38:451-455. 9 Koenigs JF. Brilliant JD. Foreman DW: Pre liminary scanning electron microscope investi gations of accessory foramina in the furcation areas of human molar teeth. Oral Surg 1974: 38:773-782. Ill Scott JII. Symons NBB: Introduction to Den tal Anatomy. Edinburgh. Churchill Living stone. 1977.'pp 242-254.
11 Gutmann .IL: Prevalence, location, and patency of accessory canals in the furcation region of permanent molars. J Periodontol 1978:49:21-26. 12 Ringelstcin D. Scow WK: The prevalence of furcation foramina in primary molars. Pediatr Dentist 1989:11:198-202. 13 Avery JK: Pulp: in Bhaskar SN (ed): Orban's Oral Histology and Embryology, ed 10. St Louis. Mostly. 1986. pp 135-197. 14 Green D: A slereomicroscopic study of the root apices of 400 maxillary and mandibular anterior teeth. Oral Surg 1956:9:1224—1232. 15 Perlich MA. Reader A. Foreman DW: A scan ning electron microscopic investigation of accessory foramens on the pulpal floor of human molars. .1 Endodonlol 1981;7:402-406. 16 Vertueci FJ. Williams RG: Furcation canals in the human mandibular first molar. Oral Surg 1974:38.-308-314. 17 Osborn JW. Ten Cate AR: Advanced Dental Histology, ed 3. Bristol. Wright & Sons. 1976. pp 123-127. 18 Takahashi S: On the abnormal structure of the bifurcation of multirooted teeth (in Japanese). Kokubyo Z 1930:4:26-28. 19 Fujila Y: Formation of subpulpal wall in mouse maxillary first molar (in Japanese). Jpn J Oral Biol 1978:20:221-228. 20 Kuroiwa M. Kodaka T. Higashi S: Morpholog ical study on the 'seams' in the multiroot for mation of rat molar teeth. Acta Anat 1991:142: 6-14.
21 Lester KS. Bovde A: Scanning electron microscopy of developing roots of molar teeth of the laboratory rat. J Ultrastruct Res 1970: 33:80-94. 22 Kuroiwa M. Saito T. Kodaka T. Higashi S: Scanning electron microscopic study of the developing roots in rat molars (in Japanese). J Showa Univ Dent Soe 1988:8:452-456. 23 Yoshida II. Yakushiji M. Sugihara A. Tanaka K. Taguchi M. Machida Y: Accessory canals at floor of the pulp chamber of primary molars (in Japanese). Shikwa Gakuho 1975,75:580-585. 24 Goto G. Chang Y. Hosoya Y: A scanning elec tron microscopic investigation of accessory foramens on the pulpal floor of human primary molars (in Japanese). Jpn J Pediatr Dent 1990: 28:371-380. 25 Orban B. Mueller E: The development of the bifurcation of multirooted teeth. J Am Dent Assoc 1929:16:297-319. 26 Sharawv M. Bhussrv BR: Development and growth of teeth; in Bhasker SN (ed): Orban's Oral Histology and Embryology, ed 10. St Louis. Mosby. 1986. pp 24-44. 27 Kramer IRH: The vascular architecture of the human dental pulp. Arch Oral Biol 1960:2: 177-189.
138
Kuroiwa/Kodaka/Abe/Higashi
Accessory Canals in Rat Molars
Downloaded by: University of Connecticut 132.174.250.220 - 9/19/2017 6:42:02 AM
References
