Volume 83 • Number 2

Cervical Enamel Projections

and Associated Pouch-Like Opening
in Mandibular Furcations
Steven B. Blanchard,* Gay M. Derderian,† Thomas R. Averitt,‡ Vanchit John,*
and Donald H. Newell*

Background: Cervical enamel projections (CEPs) have

been listed among the tooth-associated risk factors that could
lead to isolated furcation defects around molar teeth. This is
more likely when Grade III CEPs are present. However, the his-
tologic aspects of the CEP interface with regard to the pres-
ence or absence of cementum over the enamel projection

N
within the furcation have not been well described. This study umerous predisposing factors in
was initially undertaken to evaluate this relationship. the etiology of periodontal dis-
Methods: Thirty-five mandibular molars with Grade III CEPs ease have been identified and
were selected from a collection of extracted teeth and evalu- examined as to their role in the disease
ated for the presence of cementum covering these areas by process. Developmental anomalies of
stereomicroscopy, light microscopy, and scanning electron tooth morphology, such as the cervical
microscopy (SEM). enamel projection (CEP), have been
Results: Sixteen of the 35 teeth (45.7%) with Grade III CEPs examined and correlated with destruc-
appeared completely covered by cementum. SEM evaluation tive periodontal lesions in the furcation
showed the presence of a narrow pouch-like opening between regions of molar teeth.1-5 CEP is gener-
cementum and enamel in 15 of 16 teeth (93.8%). Light mi- ally considered to be a factor in the
croscopy evaluation confirmed the presence of the pouch development of isolated furcation inva-
along with some indication of residual degenerated mesen- sions (FI) by disallowing the formation
chymal tissue within the defects. SEM evaluation showed the of a connective tissue attachment to the
presence of globular bodies in this pouch. radicular enamel.1,3,5 The lack of a con-
Conclusions: This study reports on the presence of a nective attachment to this enamel ex-
pouch-like opening between the enamel and cementum in tension may increase the potential for
mandibular molars with Grade III CEPs. The clinical signifi- deep, rapid bone loss within the furca-
cance of these cementum pouches has yet to be determined tion.1
but bacterial contamination of these pouch-like structures in Although the existence of radicular
areas of furcation periodontal breakdown may serve as a nidus enamel projections was initially reported
for recolonization and disease recurrence or for the develop- >150 years ago,6,7 Masters and Hoskins1
ment of furcation caries if exposed to the oral cavity after sur- were the first to classify CEPs by the
gical procedures. J Periodontol 2012;83:198-203. length of their extension toward the fur-
cation and noted their high association
KEY WORDS
with isolated FIs. Although no attempt
Dental cementum; developmental biology; histology; risk was made to determine a correlation be-
factors; tooth cervix abnormalities. tween CEPs and FI, they suggested the
implications that the presence of a CEP
* Department of Periodontics and Allied Programs, School of Dentistry, Indiana University, could have on the furcal area. Grewe
Indianapolis, IN.
† The Department of General Dental Sciences/Comprehensive Patient Management Groups, et al.2 and Bissada and Abdelmalek3 re-
School of Dentistry, Marquette University, Milwaukee, WI. ported significant correlations between
‡ Retired, private practice, Sarasota, FL.
the incidence of CEP and the presence
of FI, whereas Leib et al.8 found no signif-
icant correlation in their study. Swan and
Hurt4 identified a statistically significant

doi: 10.1902/jop.2011.110088

198
J Periodontol • February 2012
correlation between Grade II and III CEPs and FIs.
However, Grade I CEPs lacked significance likely be-
cause Grade I is the least severe of the three grades
with greater proximity away from most furcations.
Some reports have focused on the histologic as-
pects of the CEP interface. Suzuki9 described the
radicular enamel as frequently being covered by ce-
mentum that, when removed, revealed a high inci-
dence of enamel anomalies. Chan et al.10 found that
radicular enamel was covered by cementum only near
the apical junction or in root separation areas.
Moskow’s11 histologic examination of human block
sections demonstrated cemental structures cover-
ing CEPs in or directly adjacent to the furcation,
and revealed enamel that was covered by an irregular
cementum-like structure into which vitreous-like
corpuscular structures were often incorporated.
If Grade III CEPs are commonly covered by cemen-
tum within the confines of the furcation, there could be
connective-tissue fiber attachment despite the under-
lying presence of enamel. This situation would prob-
ably reduce the potential for development of an
isolated FI. Therefore, the initial aim of this study
was to determine the prevalence of cementum cov-
ering Grade III CEPs within the furcations and to eval-
uate the cementum-covered CEPs via light and
scanning electron microscopy (SEM).
MATERIALS AND METHODS
Thirty-five mandibular first and second molars with
Grade III CEPs were selected from a bulk tooth collec-
tion of one of the authors (DHN) that had been stored
in formalin. Grade III CEPs were defined as enamel
projections that extended horizontally into the furca-
tion.1 Teeth were excluded from the study if they
had large carious lesions in the immediate vicinity
of their CEPs, deep internal caries that would compro-
mise histologic sectioning, or showed evidence of
mechanical trauma in the furcation area from the ex-
traction procedure. The root surfaces, including the
furcations, were cleaned of all soft tissue attachment
with periodontal curets taking care not to disturb the
cemental and enamel surfaces. The CEPs were exam-
ined to their full extent within the furcations with a dis-
secting microscope at ·10 power and photographed
at ·10 and ·20 magnification. Those teeth whose
CEPs were completely covered by cementum within
the furcation were then prepared for SEM and light
microscopic examination.
Initially, the roots of the teeth were removed by
cross-sectional cuts !3 mm from the fornices of the
furcations. The teeth selected for light microscopic
examination were demineralized in a mixture of hy-
drochloride, deionized water, and EDTA for 2 weeks.
Radiographs were taken to detect the presence of re-
sidual mineralization and, if present, the teeth were
Blanchard, Derderian, Averitt, John, Newell
Figure 1.
Stereoscopic view of the relationship between the cementum and the
enamel of a Grade III CEP (original magnification ·25).
then placed back into the demineralizing solution
for an additional 2 days. After clearing and dessicating
the teeth, they were embedded in paraffin, cut into
5-mm sections bucco-lingually and mesio-distally,
and stained with hematoxylin and eosin.
A subset of the teeth with CEPs covered by cemen-
tum was selected for SEM evaluation. These speci-
mens were left intact, dried, sputter-coated with a
60/40 mix of gold/palladium, and the CEPs and fu-
rcations examined by SEM.§
RESULTS
Stereomicroscopy
Sixteen (45.7%) of 35 teeth with Grade III CEPs were
completely covered by cementum within the confines
of the furcation. With unaided vision, there consis-
tently appeared to be direct contact between the
enamel of the CEPs and the overlying cementum.
However, when viewed under a stereomicroscope at
·10 power, a small space resembling a narrow pouch
was noted between the cementum and the enamel in
15 of the 16 teeth (93.8%) whose Grade III CEPs were
covered by cementum within the furcation (Fig. 1).
Light Microscopy
Eight of 15 teeth with CEPs covered by cementum
were examined by light microscopy. Buccolingual
sections generally did not register the spaces between
the CEPs and the overlying cementum clearly to their
full extent because of the difficulty of aligning the cuts
accurately within the confines of the narrow projec-
tions of enamel. However, mesio-distal sections gave
clear cross-sectional views of the CEP spaces and
the contents of the pouch-like spaces between the
enamel and cementum (Fig. 2A). Residual extracellu-
lar matrix adjacent to the CEPs could be easily
§ ISI SS-40, Topcon Industrial Products, Paramus, NJ.
199
Histologic Observations of Cervical Enamel Projections
Figure 2.
Mesio-distal sections of the furcation area. A) Cross-section of the CEP
space between the enamel and cementum (hematoxylin and eosin;
original magnification ·30). B) Higher magnification of the framed area
in A. Residual enamel matrix of the CEP and degenerated mesenchymal
tissue found within the adjacent pouch-like area (hematoxylin and eosin;
original magnification ·150).
distinguished from degenerated mesenchymal tissue
within the confines of the adjacent pouch-like areas
(Fig. 2B).
SEM
Seven teeth with cementum-covered CEPs were
examined by SEM. The pouch-like area was more
clearly demonstrated in the SEM images (Figs. 3
and 4). At higher magnifications, globular bodies were
noted within the confines of the furcation in these
pouch-like regions (Fig. 5). These were sometimes in-
dividual globules but more often were coalesced into
multiple globules that at times appeared to be at-
tached to both the enamel and cementum (Fig. 6).
DISCUSSION
CEPs are often associated with isolated FIs. It has
been postulated that these radicular enamel exten-
sions are associated with a junctional epithelial at-
tachment and lack a connective tissue attachment
200
Volume 83 • Number 2
Figure 3.
A SEM of the cervical enamel junction of a Grade III CEP. An opening is
clearly seen subjacent to the enamel projection (original magnification
·25).
Figure 4.
Greater detail of the opening is seen on higher magnification of the region
between the cementum and the enamel projection (original
magnification ·75).
and, as such, may be more prone to periodontal
breakdown. The overall prevalence of CEPs in mandi-
ble molars has been reported to range from 10.4% to
33.7%1-4,8 in whites, but to have a much higher prev-
alence in Asian populations.5,12,13 The prevalence of
CEPs may be underestimated because Suzuki9 re-
ported that these enamel projections were often cov-
ered by cementum and not readily visible unless the
overlying cementum was removed. The reported
prevalence of Grade III CEPs in mandibular molars
ranges from 4.3% to 8% in whites1,2,4 and as high as
J Periodontol • February 2012
Figure 5.
Globular bodies noted within the pouch-like region (original magnification
·300).
45% in Asians.5,12 Of the 35 mandibular molars with
Grade III CEPs examined in this study, 16 (45.7%)
were covered with cementum in the furcation region
and may have exhibited a connective tissue attach-
ment in vivo. It is conceivable that molar teeth exhib-
iting CEPs that are covered with a layer of cementum
possess a connective tissue attachment and may be
no more susceptible to periodontal breakdown than
molar teeth with a normal cervical enamel junction
morphology.
Globular-like bodies similar to those reported by
Moskow11 were observed in several SEM specimens.
The nature of these bodies is still undetermined and
Moskow speculated that the structures are formed un-
der the influence of displaced ameloblasts and appear
to be initially formed within the periodontal ligament
but ultimately become incorporated into the root with
continued cementum deposition and calcification.11
In some of our sections, these globular bodies ap-
peared to be attached to areas of the cementum or
enamel of the CEP. It is unknown what influence these
irregular globular formations may play on periodontal
breakdown in the furcation region. More recently,
Bosshardt and Nanci14 examined the immunocyto-
Blanchard, Derderian, Averitt, John, Newell
Figure 6.
Multiple coalesced globules appear to be attached to both the enamel
and cementum (original magnification ·900).
chemical characterization of these cementicle-like
formations. They found these structures to demon-
strate concentric lamellations, be devoid of collagen
fibrils, and contain osteopontin and bone sialoprotein
detected by antibody labeling. Both osteopontin and
bone sialoprotein are non-collagenous proteins typi-
cally found in bone and cementum that play a role
in the mineralization process. These globular struc-
tures have been consistently found with radicular
enamel formations, including both CEPs and enamel
pearls.11,14-16 This suggests a causal connection and
that alterations of Hertwig’s epithelial root sheath dur-
ing root formation can result in both ectopic enamel
formation and alterations in the development of the
adjacent periodontal ligament.14 Further studies are
needed to determine the origin and implications of
these globular bodies on the integrity of the peri-
odontal attachment in the furcation area.
Of interest was the detection of a pouch-like space
between the enamel and cementum in most of the
CEPs covered by cementum. To the best of our knowl-
edge, the existence of this pouch-like opening be-
tween the enamel and cementum in mandibular
molars with Grade III CEPs has not previously been
201
Histologic Observations of Cervical Enamel Projections
reported. Fifteen (93.8%) of 16 CEPs evaluated and
covered by a layer of cementum demonstrated this
pouch-like opening in the furcation region. Residual
extracellular matrix (consistent with enamel matrix
when viewed under higher power) was noted on the
enamel surface of the CEP and the pouch-like space
between the enamel and cementum contained degen-
erated mesenchymal tissue. The observation of resid-
ual enamel matrix and degenerated mesenchymal
tissue within these confines suggests that these
spaces existed in vivo and were not fixation artifacts.
The specimens were obtained from a large bulk col-
lection of extracted teeth and, although they were
stored in formalin, many of the specimens were poorly
fixed in the region of the furcation pouch possibly be-
cause of the small dimensions of the pouch-like open-
ing and the inability of the fixative to adequately reach
the inner aspects of the pouch-like area through cap-
illary action. For this reason, it was difficult to make
more definitive light microscopic observations from
the specimens. One limitation of this study relates
to the source of the selected mandibular molars. Al-
though we were careful only to select intact teeth with
Grade III CEPs that were not affected by caries or ex-
traction damage, there was no way to determine the
reason for tooth extraction, how long ago the teeth
were extracted, or if periodontal disease was present
or absent on the selected teeth. Consequently, the
teeth used in this study may or may not be a represen-
tative sample of Grade III CEPs in mandibular molars.
These pouch-like openings were more clearly dem-
onstrated in the SEM images. It is unknown what influ-
ences, if any, these pouch-like structures may have in
a periodontally intact furcation. However, in areas
where periodontal breakdown has occurred, these
openings are likely to be colonized by periodonto-
pathic microorganisms. Non-surgical and surgical
debridement of the furcation region is extremely diffi-
cult and these pouch-like regions can serve as a nidus
for bacterial recolonization and disease recurrence.
Incomplete debridement in the interfurcal regions
containing these contaminated pouch-like recesses
could lead to failure of regeneration procedures
in the furcation regions. Although infrequently per-
formed today, caries in the furcation region after tun-
neling procedures may be the result of post-surgical
colonization of these pouch-like regions with cario-
genic bacteria because it would be impossible to ad-
equately clean the cemental pouches with any type of
mechanical brush or device. It is arguable that fluoride
gels delivered on interdental brushes would be unable
to penetrate these small spaces to provide increased
caries resistance. The high incidence of cementum
pouch-like openings associated with CEPs (15 of 35
specimens) should alert the clinician of the antici-
pated need for thorough debridement (if not complete
202
Volume 83 • Number 2
removal) of the CEP with rotary instruments or dia-
mond-impregnated ultrasonic tips whenever fur-
cation regenerative procedures are contemplated. If
resective surgical approaches are to be used on teeth
with CEPs, attempts should be made to remove any
cementum over the CEP to eliminate any potential
pouch-like defect that may be present.
CONCLUSIONS
Nearly 46% of Grade III CEPs in this sample of ex-
tracted mandibular molars were found to be covered
with a layer of cementum within the confines of the fur-
cation. On further inspection, 15 of 16 CEPs covered
with cementum contained a small pouch-like opening
between the cementum and enamel. Globular bodies
previously associated with ectopic enamel formations
were frequently found in these pouches. The clinical
significance of these cementum pouches has yet to
be determined, but bacterial contamination of these
pouch-like structures in areas of furcation periodontal
breakdown may serve as a nidus for recolonization
and disease recurrence or for the development of fur-
cation caries if exposed to the oral cavity secondary to
surgical procedures or marginal soft-tissue recession.
ACKNOWLEDGMENTS
The authors thank Dr. Charles Tomich, DDS, MSD,
professor emeritus, Department of Oral Pathology,
Medicine, and Radiology, Indiana University School
of Dentistry, Indianapolis, Indiana, for his valuable in-
put on the interpretation and evaluation of the light
microscopy histologic sections. The authors report
no conflicts of interest related to this study.
REFERENCES
1. Masters DH, Hoskins SW. Projection of cervical
enamel into molar furcations. J Periodontol 1964;35:
49-53.
2. Grewe JM, Meskin LH, Miller T. Cervical enamel
projections: Prevalence, location, and extent; with asso-
ciated periodontal implications. J Periodontol 1965;36:
460-465.
3. Bissada NF, Abdelmalek RG. Incidence of cervical
enamel projections and its relationship to furcation
involvement in Egyptian skulls. J Periodontol 1973;
44:583-585.
4. Swan RH, Hurt WC. Cervical enamel projections as an
etiologic factor in furcation involvement. J Am Dent
Assoc 1976;93:342-345.
5. Hou G-L, Tsai C-C. Relationship between periodontal
furcation involvement and molar cervical enamel pro-
jections. J Periodontol 1987;58:715-721.
6. Owens R. Odontography, or, a Treatise of the Compar-
ative Anatomy of the Teeth, vol. 2. London: Hippolyte
Bailliere; 1840:118-119.
7. Linderer J, Linderer CJ. Manual of Dentistry (in
German). Berlin: Schlesinger; 1842;2:174.
8. Leib AM, Berdon JK, Sabes WR. Furcation involve-
ments correlated with enamel projections from the
J Periodontol • February 2012
cementoenamel junction. J Periodontol 1967;38:330-
334.
9. Suzuki T. The enamel in the inter-radical space portion
of human multi-rooted teeth (in Japanese). Kokubyo
Gakkai Zasshi 1958;25:273-280.
10. Chan CP, Cassingham RJ, Yukna RA, Meffert RM,
Kasten FH. Morphologic and histometric characteris-
tics of cervical enamel projections in molars of
Chinese. Int J Periodontics Restorative Dent 1988;
8(6):52-65.
11. Moskow BS. Some observations on radicular enamel.
J Periodontol 1971;42:92-96.
12. Zee KY, Chiu ML, Holmgren CJ, Walker RT, Corbet
EF. Cervical enamel projections in Chinese first per-
manent molars. Aus Dent J 1991;36:356-360.
13. Kawaski K, Hasegawa M, Hara K. Cervical enamel
projections in human molars: Location, extent, with
associated furcation involvement (in Japanese). Jpn
J Conserv Dent 1976;19:139.
Blanchard, Derderian, Averitt, John, Newell
14. Bosshardt DD, Nanci A. Immunocytochemical char-
acterization of ectopic enamel deposits and cemen-
ticles in human teeth. Eur J Oral Sci 2003;111:51-59.
15. Kerebel B, Dard M, Le Cabellec MT, Kerebel LM.
Dental enamel pearls: Histopathological study (in
French). J Biol Buccale 1986;14:239-248.
16. Bohne W, Pouëzat JA, Kerebel B. Structural, ultra-
structural and microanalytical study of multiple
enamel pearls (in French). Bull Group Int Rech Sci
Stomatol Odontol 1989;32:53-61.
Correspondence: Dr. Steven B. Blanchard, Department of
Periodontics and Allied Programs, School of Dentistry,
Indiana University, 1121 W. Michigan St., Room 422A,
Indianapolis, IN. Fax: 317/274-1363; e-mail: stblanch@
iupui.edu.
Submitted February 11, 2011; accepted for publication
May 19, 2011.
203