NIH Public Access

Author Manuscript
J Dent. Author manuscript; available in PMC 2009 November 1.
Published in final edited form as:
NIH-PA Author Manuscript

J Dent. 2008 November ; 36(11): 953–958. doi:10.1016/j.jdent.2008.08.002.

Dentin tubule numerical density variations below the CEJ

T. Komabayashi*, G. Nonomura, L.G. Watanabe, G.W. Marshall Jr., and S.J. Marshall
Department of Preventive and Restorative Dental Sciences, University of California, San Francisco,
707 Parnassus Avenue, D2246, University of California, San Francisco, CA 94143-0758

Abstract
Aim—To evaluate dentin tubule numerical density variations below the CEJ.
Methodology—Three human non-carious permanent canines were sectioned parallel to the CEJ
to obtain dentin disks 1 mm thick whose surfaces were 1 mm and 2 mm below the CEJ. Each disk
was sectioned into quarters resulting in four segment locations: facial, lingual, mesial, and distal.
The outer (PDL side) and inner (pulp side) surfaces of the specimens were shaped to expose dentin
with SiC papers and polished. Numerical tubule density was determined from SEM images. All data
NIH-PA Author Manuscript

were statistically analyzed using a three-way ANOVA.

Results—The dentin tubule density (number/mm2) ranged from 13,700 to 32,300. Dentin tubule
density was relatively uniform at 1 and 2 mm below the CEJ and increased by a factor of about two
from the outer to the inner surface, which was significantly different (P<0.0001).
Conclusions—The tubule density variations at the cervical root did not present the marked
changes.

Keywords
Cemento-Enamel Junction (CEJ); Cervical lesions; Density; Dentin hypersensitivity; Dentin tubules;
Root dentin; Scanning Electron Microscopy (SEM)

Introduction
Dentin hypersensitivity is defined as a transient pain arising from exposed dentin, typically in
response to chemical, thermal, tactile, or osmotic stimuli, which cannot be explained by any
NIH-PA Author Manuscript

other dental defect or pathology [1]. Dentin exposure can occur due to trauma, gingival
recession, and various restorative procedures [2]. Dentin hypersensitivity is predominately
located on the cervical part of the buccal surface [3–5], although dentin hypersensitivity can
occur on all tooth surfaces [5]. According to Brännström, an increased number of exposed
dentin tubules can result in increased dentinal fluid movement and patient discomfort [6]. The
high patency of the tubules on the outer dentin surface in dentin hypersensitivity teeth was
demonstrated by scanning electron microscopy (SEM) [7–11].

Dentin is an anisotropic biological composite and exhibits regional differences in tubule density
[12,13]. With respect to the location from the CEJ, Mjör and Nordahl demonstrated that the

*Takashi Komabayashi DDS, MDS, PhD, Assistant Professor, Department of Endodontics, Texas A&M Health Science Center Baylor
College of Dentistry, 3302 Gaston Avenue, Dallas, TX 75246 USA, TEL: 214-828-8365, FAX: 214-874-4507, E-mail address:
tkomabayashi@bcd.tamhsc.edu ICD38719@nifty.com.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting
proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could
affect the content, and all legal disclaimers that apply to the journal pertain.
 Komabayashi et al. Page 2

mean number of dentin tubules in the middle part of the root was significantly lower than in
the middle part of the crown [14]. Harran Ponce et al. observed that the numerical density of
dentinal tubules is greater in the crown, decreasing toward the apical third of the root. In that
NIH-PA Author Manuscript

study, numerical density of dentin tubules at the external dentin surface was reported to be
21,000, 15,000, and 13,000 per mm2 at crown-root junctions, cervical-middle, and middle-
apical thirds of the root, respectively [15]. Dentin tubule numerical density in root dentin has
been studied by various investigators [14,16–20]. However, dentin tubule numerical density
information which is focused on the cervical part of the tooth, located just below the CEJ, is
scarce.

The dentin localized in the close proximity to the CEJ is an area which often develops dentin
hypersensitivity and is subject to Class V root caries lesions, sclerotic non-carious lesions, and
abfraction lesion [21–28]. It has been reported that non-carious cervical lesions occur in up to
85 per cent of individuals, with their prevalence and severity increasing with age [28]. Large
non-carious cervical lesions with exposed dentine can increase the risk of pulp exposure or
tooth fracture [21]. Fractures may occur in cervically-notched teeth, severing the crown of the
tooth in anterior teeth, whereas in the cusp in posterior teeth [24]. Canine is common for class
V root caries and cervical lesions because it retains longer than other teeth and locates at the
transition from anterior to posterior.

The aim of this study was to evaluate dentin tubule numerical densities 1mm and 2 mm below
NIH-PA Author Manuscript

the CEJ.

Materials and Methods

Specimen preparation
The specimens used in this study were prepared from permanent human non-carious, erupted,
and fully root-formed maxillary or mandibular canines. All the teeth were recently extracted
from patients needing extractions as a part of their dental treatment, as approved by the UCSF
Committee on Human Research. Teeth were sterilized by gamma irradiation and refrigerated
at 4°C in Hank's balanced salt solution (HBSS) prior to use [29]. Figure 1 shows a schematic
of the specimen preparation and study procedures. Three teeth were sectioned parallel to the
CEJ with a low speed water cooled diamond saw (Buehler, Ltd., Lake Bluff, IL) to obtain
dentin disks 1 mm thick whose surfaces were 1 mm and 2 mm below the CEJ. One disc was
obtained per tooth. Each disc was sectioned into quarters, resulting in four specimens (facial,
lingual, mesial, and distal). Twelve specimens were obtained from three teeth. The outer (PDL
side) and inner (pulp side) surfaces of the specimens were shaped to expose dentin with 800
and 1200 grit SiC papers (ALLIED, Rancho Dominguez, CA) and polished using 6, 3, 1, and
0.25 µm diamond suspension (Buehler, Ltd., Lake Bluff, IL). The specimens were
NIH-PA Author Manuscript

ultrasonicated for 5 seconds in deionized water after each polishing step to remove any
polishing debris and help insure that all tubules would be visible. To remove any remaining
smear layer, 0.5 M EDTA was applied for 30 seconds, rinsed with deionized water, and gently
dried by compressed air to remove excess water, but the surface was left moist.

SEM analysis
SEM images (2000x) were collected from 12 locations per specimen (i.e. outer/inner, left/
center/right, 1 mm/2 mm) using the CFAS “wet” mode (ISI ABT SX-40A SEM, Topcon
Instruments, Pleasanton, CA). In this mode, neither drying nor sputter coating is necessary. As
summarized in Figure 1, the orientations and intratooth positions were chosen so the tubules
would be roughly perpendicular to the surface examined, yielding circular cross-sections of
the tubules. Images were stored digitally and analyzed with image analysis software (Ultrascan
2.1.1, Soft Imaging Software, Kevex Sigma, Noran Instruments, Inc., Madison, WI).

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 3

As indicated above, a tooth yielded dentin specimens at multiple sites, varying by depth (1
mm, 2 mm), location (left, center, right), and position (outer, inner) in four specimens (facial,
lingual, mesial, and distal). Overall, 144 SEM images were obtained from 3 teeth. The number
NIH-PA Author Manuscript

of tubules was counted manually on each SEM image and converted to tubule density, the
number of tubules per square mm (mm2). Tubules that were partially in an image were counted
as a half of a tubule.

Statistical Analysis
The tubule densities by location (left, center, right) were averaged because there were no
statistically significant differences in these values. Mean and standard deviation (S.D.) of the
dentin tubule densities (tubules/ mm2) were calculated. The coefficient of variation (CV) was
calculated. All data were statistically analyzed using a three-way ANOVA comparing tubule
density as a function of depth, location, and internal/external position. (Statview5, SAS
Institute Inc, Cary, NC)

Results
Figure 2 shows composite 4 SEM images, representative of the 144 SEM images observed.
Images were classified by depth (1 mm, 2 mm) and position (outer, inner) in four regions
(facial, lingual, mesial, and distal). Tubule patterns appeared similar in the images from the
facial and lingual sections and in those from the mesial and distal sections. However, the facial
NIH-PA Author Manuscript

and lingual images differed from the mesial and distal images in terms of tubule roundness,
which was due to tubule orientation.

The dentin tubule densities and their standard deviations (tubules/mm2) are summarized in
Table 1. The dentin tubule density (number/mm2) ranged from 13,700 to 32,300. The CV
ranged from 0.08 to 0.22. Dentin tubule density was relatively uniform at 1 and 2 mm below
the CEJ. However, dentin tubule density was increased by a factor of about two from the outer
to the inner surface, which was significantly different (P<0.0001). P values for other variables
were 0.08 for region (facial, lingual, mesial, and distal) and 0.10 for depth (1 mm, 2 mm).
Accordingly, no statistical significance was found among four regions (facial, lingual, mesial,
and distal). No statistical significance was found between two depths (1 mm, 2 mm).

Discussion
In preparing our samples, we employed the polishing technique previously used by Oliveira
et al. [30], intended to render the dentin surface as devoid as possible of a smear layer.
Permanent canines were used because their size was thought to be suitable for specimen
preparation when divided into four pieces as described in Figure 1 and they allowed
NIH-PA Author Manuscript

examination of outer and inner areas in the same tooth. In addition, canine teeth commonly
develop class V root caries and cervical lesions because they are retained longer than other
teeth and are located at the transition from anterior to posterior segments. However, the
extremely small specimen size in this study made polishing very difficult, which resulted in
some remnant smear layer. To remedy this, EDTA was applied to remove the smear layer
entirely before SEM observation. EDTA was chosen because it has been reported to cause less
demineralization of the dentin structure than other methods [31]. Because the EDTA caused
an increase in the tubule diameter, it was not possible to measure tubule diameter and compare
tubule openness using such categories as plugged, partially open, and open [32]. However, the
tubules near the pulp chamber did appear to have larger diameters than those near the outer
surface, as expected. To ensure an accurate count, considerable care was taken to produce
specimens from a variety of locations with tubule orientation and surface conditions.
Distributions with CV < 1 were considered low-variance. In this regard, the number of
specimen in this study was appropriate.

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 4

The tooth is longer facial-lingually than mesial-distally, so the odontoblasts in those regions
had to travel further to reach the pulp chamber. Therefore, we expected to find greater tubule
density on facial-lingual sides. This was indeed the trend; however, it was not statistically
NIH-PA Author Manuscript

significant.

With respect to the location from the CEJ, Mjör and Nordahl demonstrated that the mean
number of dentin tubules in the middle part of the root was significantly lower than in the
middle part of the crown [14]. Harran Ponce et al. observed that the numerical density of
dentinal tubules is greater in the crown, decreasing toward the apical third of the root. In that
study, numerical density of dentin tubules at the external dentin surface was reported to be
21,000, 15,000, and 13,000 per mm2 at crown-root junctions, cervical-middle, and middle-
apical thirds of the root, respectively [15]. Therefore, we expected to find gradual changes in
the numerical density of the dentinal tubules. However, our study found no statistically
significant difference between 1 mm and 2 mm below the CEJ. This may be due to the narrow
area targeted by the study, even though this is the area where cervical root exposure is frequently
observed in relation to dentin hypersensitivity. Future research should examine distances of 3
mm and greater from the CEJ toward the root apex so as to have a wider targeted area at cervical
root dentin. This study also suggests for further study of tubule density of teeth other than the
canines, though the specimen preparation will be more difficult and complex. Such research
will provide a useful database to increase knowledge of the nature and causes of dentin
hypersensitivity.
NIH-PA Author Manuscript

Dentin tubule density increased from the outer to the inner surface by a factor of two in our
study. This result agrees with the report by Pashley et al. that tubule density increased as the
pulp chamber was approached [33]. Similarly, Harran Ponce et al. observed that the numerical
density of dentin tubules at the internal and external dentin surface at crown-root junctions was
38,000 and 21,000 per mm2, respectively [15]. In regards to comparison between root and
coronal dentin, Fogel et al. found that the ratio of the number of tubules per unit area at the
pulp and outer root surface was only about 2:1 rather than the 4:1 found in coronal dentin
[18]. Mjör and Nordahl also reported that differences in the density of tubules between the
peripheral and inner aspects were more marked in the crown than in the root [14]. In the current
study, the tubule density variations at the cervical root did not present the marked changes
observed in coronal dentin studies. The transition from coronal to root dentin around the pulp
chamber needs further study.

The dentin localized in the close proximity to the CEJ is an area which often develops dentin
hypersensitivity and is subject to Class V root caries lesions, sclerotic non-carious lesions, and
abfraction lesions [21–28]. Although the mechanisms of pain transmission across dentin are
not fully understood, dentin hypersensitivity is reduced when the dentin tubules are occluded
NIH-PA Author Manuscript

[34,35]. The findings of this study provide better understanding of the anatomy of dentin as it
pertains to dentin hypersensitivity. The results of this study provide useful information for
dentin as a substrate that needs treatment for dentin hypersensitivity, carious, and non-carious
lesions. Since dentin is a substrate whereas material is an adherent, the results of this study
provide useful data for future dental material research and development to occlude exposed
dentin tubules in dentin hypersensitivity treatment.

Conclusion
The following conclusions can be drawn from this study for canines.
1. The dentin tubule density (number/mm2) ranged from 13,700 to 32,300.
2. Dentin tubule density was relatively uniform at 1 and 2 mm below the CEJ.

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 5

3. Dentin tubule density was about two times greater at the inner surface than at the outer
surface.
4. There was no difference in tubule density with region (facial, lingual, mesial, and
NIH-PA Author Manuscript

distal) or depth for the areas studied.

5. The tubule density variations at the cervical root did not present the marked changes.

Acknowledgements
This study was supported by NIH/NIDCR DE09859, the research was conducted at UCSF while Dr. Komabayashi
was a student. The authors would like to thank Kuniko Saeki, DDS, PhD, DDS, and Larz S.W. Spångberg, DDS, PhD
for useful comments.

References
1. Addy, M.; Mostafa, P.; Absi, E.; Adams, D. Cervical dentin hypersensitivity. Etiology and management
with particular reference to dentifrices. In: Rowe, N., editor. Symposium on Hypersensitive Dentin
Origin and Management. London: E. and S. Livingston Limited; 1985. p. 147-167.
2. Liu H, Lan W, Hsieh C. Prevalence and distribution of cervical dentin hypersensitivity in a population
in Taipei, Taiwan. Journal of Endodontics 1998;24:45–47. [PubMed: 9487867]
3. Flynn J, Galloway R, Orchardson R. The incidence of 'hypersensitive' teeth in the West of Scotland.
Journal of Dentistry 1985;13:230–236. [PubMed: 3862690]
NIH-PA Author Manuscript

4. Fischer C, Fischer RG, Wennberg A. Prevalence and distribution of cervical dentine hypersensitivity
in a population in Rio de Janeiro, Brazil. Journal of Dentistry 1992;20:272–276. [PubMed: 1452864]
5. Rees JS, Addy M. A cross-sectional study of dentine hypersensitivity. Journal of Clinical
Periodontology 2002;29:997–1003. [PubMed: 12472992]
6. Brännström, M. A hydrodynamic mechanism in the transmission of pain producing stimuli through
the dentine. In: Anderson, DJ., editor. Sensory mechanism in dentine. Oxford: Pergammon Press;
1963. p. 73-79.
7. Brännström M. The surface of sensitive dentine. An experimental study using replication. Odontologisk
revy 1965;16:293–299. [PubMed: 5216250]
8. Ishikawa S. A clinico-histological study on the hypersensitivity of dentin. Kokubyo Gakkai Zasshi
1969;36:278–298. [PubMed: 5266092]
9. Absi EG, Addy M, Adams D. A study of the patency of dentinal tubules in sensitive and non sensitive
cervical dentine. Journal of Clinical Periodontology 1987;14:280–284. [PubMed: 3475295]
10. Yoshiyama M, Masada J, Uchida A, Ishida H. Scanning electron microscopic characterization of
sensitive vs. insensitive human radicular dentin. Journal of Dental Research 1989;68:1498–1502.
[PubMed: 2584516]
11. Yoshiyama M, Noiri Y, Ozaki K, Uchida A, Ishikawa Y, Ishida H. Transmission electron microscopic
characterization of hypersensitive human radicular dentin. Journal of Dental Research
NIH-PA Author Manuscript

1990;69:1293–1297. [PubMed: 2355124]

12. Marshall GW, Marshall SJ, Kinney JH, Balooch M. The dentin substrate: structure and properties
related to bonding. Journal of Dentistry 1997;25:441–458. [PubMed: 9604576]
13. Wang R, Weiner S. Human root dentin: structural anisotropy and Vickers microhardness isotropy.
Connective Tissue Research 1998;39:269–279. [PubMed: 11063007]
14. Mjör IA, Nordahl I. The density and branching of dentinal tubules in human teeth. Archives of Oral
Biology 1996;41:401–412. [PubMed: 8809302]
15. Harran Ponce E, Canalda Sahli C, Vilar Fernandez JA. Study of dentinal tubule architecture of
permanent upper premolars: evaluation by SEM. Australian Endodontic Journal 2001;27:66–72.
[PubMed: 12360689]
16. Carrigan PJ, Morse DR, Furst ML, Sinai IH. A scanning electron microscopic evaluation of human
dentinal tubules according to age and location. Journal of Endodontics 1984;10:359–363. [PubMed:
6590745]

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 6

17. Ferrari M, Mannocci F, Vichi A, Cagidiaco MC, Mjör IA. Bonding to root canal: structural
characteristics of the substrate. American Journal of Dentistry 2000;13:255–260. [PubMed:
11764112]
NIH-PA Author Manuscript

18. Fogel HM, Marshall FJ, Pashley DH. Effects of distance from the pulp and thickness on the hydraulic
conductance of human radicular dentin. Journal of Dental Research 1988;67:1381–1385. [PubMed:
3183154]
19. Mjör IA, Smith MR, Ferrari M, Mannocci F. The structure of dentine in the apical region of human
teeth. International Endodontic Journal 2001;34:346–353. [PubMed: 11482717]
20. Pashley DH. Dentin: a dynamic substrate- a review. Scanning microscopy 1989;3:161–174. [PubMed:
2662395]
21. Osborne-Smith KL, Burke FJ, Wilson NH. The aetiology of the non-carious cervical lesion.
International Dental Journal 1999;49:139–143. [PubMed: 10858746]
22. Kassab MM, Cohen RE. The etiology and prevalence of gingival recession. Journal of the American
Dental Association 2003;134:220–225. [PubMed: 12636127]
23. Absi EG, Addy M, Adams D. Dentine hypersensitivity. A study of the patency of dentinal tubules in
sensitive and non-sensitive cervical dentine. Journal of Clinical Periodontology 1987;14:280–284.
[PubMed: 3475295]
24. Staninec M, Nalla RK, Hilton JF, Ritchie RO, Watanabe LG, Nonomura G, et al. Dentin erosion
simulation by cantilever beam fatigue and pH change. Journal of Dental Research 2005;84:371–375.
[PubMed: 15790746]
25. Nguyen C, Ranjitkar S, Kaidonis JA, Townsend GC. A qualitative assessment of non-carious cervical
NIH-PA Author Manuscript

lesions in extracted human teeth. Australian Dental Journal 2008;53:46–51. [PubMed: 18304241]
26. Absi EG, Addy M, Adams D. Dentine hypersensitivity--the effect of toothbrushing and dietary
compounds on dentine in vitro: an SEM study. Journal of Oral Rehabilitation 1992;19:101–110.
[PubMed: 1517870]
27. Levitch LC, Bader JD, Shugars DA, Heymann HO. Non-carious cervical lesions. Journal of Dentistry
1994;22:195–207. [PubMed: 7962894]
28. Bartlett DW, Shah P. A critical review of non-carious cervical (wear) lesions and the role of abfraction,
erosion, and abrasion. Journal of Dental Research 2006;85:306–312. [PubMed: 16567549]
29. White JM, Goodis HE, Marshall SJ, Marshall GW. Sterilization of teeth by gamma radiation. Journal
of Dental Research 1994;73:1560–1567. [PubMed: 7929992]
30. Oliveira SS, Marshall SJ, Hilton JF, Marshall GW. Etching kinetics of a self-etching primer.
Biomaterials 2002;23:4105–4112. [PubMed: 12182312]
31. Czonstkowsky M, Wilson EG, Holstein FA. The smear layer in endodontics. Dental clinics of North
America 1990;34:13–25. [PubMed: 2403940]
32. Oliveira SS, Pugach MK, Hilton JF, Watanabe LG, Marshall SJ, Marshall GW Jr. The influence of
the dentin smear layer on adhesion: a self-etching primer vs. a total-etch system. Dental Materials
2003;19:758–767. [PubMed: 14511734]
33. Pashley D, Okabe A, Parham P. The relationship between dentin microhardness and tubule density.
NIH-PA Author Manuscript

Endodontics & Dental Traumatology 1985;1:176–179.

34. Pashley DH, Livingston MJ, Greenhill JD. Regional resistances to fluid flow in human dentine in
vitro. Archives of Oral Biology 1978;23:807–810. [PubMed: 299019]
35. Reeder OW Jr, Walton RE, Livingston MJ, Pashley DH. Dentin permeability: determinants of
hydraulic conductance. Journal of Dental Research 1978;57:187–193. [PubMed: 277512]

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 7
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 8
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 1.
Schematic of sample preparation and study procedures.
NIH-PA Author Manuscript

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 9
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 10
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 11
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

J Dent. Author manuscript; available in PMC 2009 November 1.

 Komabayashi et al. Page 12
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript

Figure 2.
SEM images

J Dent. Author manuscript; available in PMC 2009 November 1.

 NIH-PA Author Manuscript NIH-PA Author Manuscript NIH-PA Author Manuscript
Table 1
The dentin tubule densities (mean (standard deviation) tubules/ mm2, CV)

Buccal CV Lingual CV Mesial CV Distal CV

Outer 1 mm 16300(1300)a 0.08 16800(2200)a 0.13 13700(1800)a 0.13 15000(1700)a 0.11

2 mm 15200(1400)a 0.09 14700(1500)a 0.10 16500(2600)a 0.16 16200(3500)a 0.22
Inner 1 mm 31400(5100)b 0.16 28400(2800)b 0.10 25300(2900)b 0.11 28700(4200)b 0.15
2 mm 32300(2600)b 0.08 31600(4400)b 0.14 26000(3400)b 0.13 29100(4700)b 0.16
Komabayashi et al.

Groups identified by different superscript letters are significantly different (P<0.0001).

J Dent. Author manuscript; available in PMC 2009 November 1.

Page 13