Intirnational Endodontic Journal (1995) 28.

12-18

Periapical status of endodontically treated teeth in relation to the

technical quality of the root filling and the coronal restoration
H. A. RAY, & M. TROPE
Department of Endodontology. Temple Dental School Philadelphia. PA 19140, USA

Summary
The purpose of this sfudy was to evaluate fhe relafionship of fhe quality of the coronal restoration and of the
root canal obturation on the radiographic periapical
status of endodontically treated teeth. Full-mouth
radiographsfromrandomly selected new patient folders
at Temple University Dental School were examined. The
first 1010 endodontically treated teeth restored with a
permanent restoration were evaluated independently by
two examiners. Post and core type restorations were
excluded. According to a predetermined radiographic
standard set of criteria, the technical quality of the root
filling of each tooth was scored as either good (GE) or
poor (PE), and the quality of the coronal restoration
similarly good (GR) or poor (PR). The apical one-third of
the root and surrounding structures were then evaluated radiographically and the periradicular status
categorized as (a) absence of periradicular inflammation (API) or (b) presence of periradicular inHammation
(PPI). The rate of API for all endodontically treated teeth
was 61.07%. GR resulted in significantly more API cases
than GE, 80% versus 75.7%. PR resulted in significantly
more PPI cases than PE, 30.2% versus 48.6%. The
combination of GR and GE had the highest API rate of
91.4%, significantly higher than PR and PE with a API
rate of 18.1%.
Keywords: endodontic success, obturation, restoration.
Introduction
Many follow-up studies have been performed on
endodontically treated teeth, and it is generally accepted
that the success rate of treatment is positively correlated
with the criteria for good technical quality of the root
filling (Strindberg 1956. Grahnen & Hansson 1961.
Kerekes&Tronstad 1979. Sjogren etui. 1990).
Correspondence: Dr. Martin Trope, Department of Endodontics,
School of Dentistry. CB# 7450, University of North Carolina, Chapei
Hill. NC 27599-7450. USA.

12

Traditionally it has been assumed that a root filling of

good technical quality provides an effective sea! of the
obturated root canal, which is critical for success,
whereas leakage along afillingof poor quality will result
in failure (Dow & Ingle 1955, Swanson & Madison
1987). Thus, the leakage both apically and coronally of
obturated root canals to dyes (Simons et al. 1991).
radioisotopes (Marshall & Massler 1961), or to bacteria
(Torabinejad et al 1990) have been extensively studied
to compare the quality of the obturation performed
under different conditions. Most recently, coronal
leakage of obturated root canals has received a great
deal of attention (Madison et al. 1987, Swanson &
Madison 1987, Torabinejad et al 1990). While most of
the canals in these experiments were obturated in vitro
or in vivo under ideal conditions and almost certainly
performed to the technf caf quality required for success in
the prognosis studies, the results of these studies indicate
that coronal leakage will be consistent and extensive if
the access cavity is left unfilled and thus exposed to fluids
(Swanson & Madison 1987, Madison et al. 1987,
Torebinejad et al 1990).
The high long-term success rate of wefl-obturated root
canals in clinical studies appears contradictory to the
extensive coronal leakage in vitro of root canals
obturated under ideal conditions. Either leakage of fluids
and bacteria is not as important as has been assumed for
endodontic failure or the primary barrier to leakage is
not the obturated root canal but the seal above it, i.e.,
the seal of the coronal restoration.
The purpose of this study was to evaluate the relationship of the quality of the coronal restoration and of the
root canal obturation on the radiographic periapical
status of endodontically treated teeth.

Materials and methods

EuU-mouth radiographs from randomly selected new
patient folders from the general patient pool at Temple
University School of Dentistry were examined. Only

Factors in endodontic success

patients who reported not having had dental treatment

for at least 1 year previous to the X-rays were selected.
The first 1010 endodontically treated teeth which
were restored with a permanent restoration were evaluated independently by two examiners using a Viewscope
X-ray view (J.S. Dental Inc., Ridgefleld, CT, USA), at x2
magnification. A third independent dentist had selected
the teeth to be examined. Teeth restored with post and
core type restorations were not included in the study,,
since the remaining obturation material was too
variable in length. Teeth were categorized according to
the radiographic quality of the endodontic obturation
and coronal restoration as follows:
1. Good endodontic filling (GE): if all canals were
obturated, no voids were present and the fifl of the
main gutta-percha point was within 0 to 2mm from
the radiographic apex (Fig. 1).
2. Poor endodontic filling (PE): if one or more of the
criteria in (1) were not met (Fig. 2).
3. Good restoration (GR): any permanent restoration
that radiographicaliy appeared sealed (Ffg, 3).
4. Poor restoration (PR): any permanent restoration
with radiographic signs of overhangs: open margins
or recurrent decay (Fig. 4).
The radiographic appearance of the apical one-third of
the root and surrounding structures were then evaluated and categorized as follows:
1. Absence of periradicular Inflammation (API): if the
contours, width and structure of the periodontal
ligament were normal or slightly widened if an
excess of filling material was present (Fig. 5),

13

2. Presence of periridacular inflammation (PPf): if one

or more of the criteria of success were not fulfilled
(Fig. 6).

Results
The periradicufar status for each category of treatment quality are shown in Table 1 and when the
criteria of treatment quality were combined in
Table 2.

Table 1. Periradicular status for each categor}' of treatment quality

Group Endo

Coronal

No, teeth

PPI

API

%AP1

1
2
3
4

Anv
Any
Good (GR)
Poor (PR)

49 5.0
490.5
633.0
352,5

120.5
252.0
126,5
246.0

374, 5
238, 5
506, 5
106, 5

75,7
48,6
80,0
30.2

Good (GEl
Poor(PE)
Anv
Any

PPL presence of periradicular inflammation

API. absence of periradicular ,Lnflammation

Table 2. Periradicular status for various combinations of treatment

quality
Group EndO'

Coronal

No. teeth

PPI

API

%API

1
2
3
4

GoodlGR)
Poor (PR)
Good (GR)
Poor (PR)

330.5
164.5
302.5
188.0

28,5
92,0
98.0
154.0

302.0
72,5
,204,5
34.0

91,4
44,1
67,6
18,1

Good (GE)
Good (GE)
Poor(PE)
Poor (PE)

PPI, presence of periradicular inflammation.

AP], absence of periradicular inJlammation.

Fig,, 1. Radiograph, of an endodonttcafly treated

premolar assessed as good endodontic filling {GE}. No
voids are present and the fill of the main gutta-percha
point is 0-2mm from the radiographic apex.

14

H. A.Ray etal

Fig. 2. Radiographs of endodontically

treated teeth assessed as poor endodontic
tilling (PE) la) too short (b) too long (c)
obvious voids are present.

t
*

-.
c

**

Mi

Analysis
Logistic regression was performed using SAS PROC
CATMOD to model the effects of ENDO and RESTORATION

on the likelihood of an outcome of API. Maximum likelihood methods were used to compute the parameter
estimates and their standard errors. The likelihood
ratio %' of 1.09 (P=0.296) indicated that goodness of fit

Factors in endodontic success

15

r..

^]^.
Fig. 3. Radiographs of endodonticaily treated molar
with amalgam restoration assessed as good restoration (GR). The amaigam appears radiographicaily to
be sealed along its entire circumference.

of the model was supported and the ENDO-RESTORATION interaction was not significant (i.e., effects of
ENDO and RESTORATION were homogeneous, or
independent, relative to one another). Therefore, the
following results were based on the main effects model
including ENDO and RESTORATION.
ENDO
RESTORATION

Odds Ratio
4.32
11,12

(3.11.6.00)
(8.00.15.47)

These results indicated that the odds of API outcome

were 4.32 times greater when ENDO was present, and
11.12 times greater when RESTORATION was present.
Testing the comparison between ENDO and RESTORA-

TION gave a %- statistic of 22.83 (P<0.00]). indicating

that the effect of RESTORATION on the likelihood of
API ii'as statistically greater than the effect of ENDO
on API.
Mantel-Haenszel statistics were computed as a
supportive anatysis. This method looked at the effects of
ENDO on the outcome across the levels of RESTORATION, and similarly for RESTORATION across the levels
of ENDO. The results of the stratified contigency table
analysis are as follows:

ENDO
RESTORATION

Odds Ratio
4.30
11.07

95% CI
(3.09.5.99)

(7.96. 15.38)

Jj^

- *

Fig. 4. Radiograph of endodonticaily treated molar

assessed to be restored vv'ith poor restoration (PR).
Obviou.s defects in the distal sea] of the crown are
seen.

16

IlARmjetal.
Chi-square statistics were calculated to test the association of ENDO, collapsed over the levels of RESTORATION, with the outcome: the similarly for
RESTORATION, collapsed over the levels of ENDO with
the outcome. Both associations were statistically
signiflcant.
Z^
ENDO
RESTORATION

76.58
238.86

P-value
<0.001

Conclusions

f
Fig. 5. Radiographic appearance of roots of tooth 11 and tooth 21
categorized as successfol. The lumina dura can be tracted around the
entire length of the roots.

The presence of ENDO' was significantly associated with

an increased likelihood of a API outcome, and this
association was homogeneous across the levels of
RESTORATION as well as when collapsing over the
levels of RESTORATION. Similarly, the presence of
RESTORATION was signiiicantly associated with an
increased likelihood of API outcome, with the association being homogeneous across the levels of ENDO as
well as ivhen collapsing over the levels of RESTORATION. It was also found that the association of
RESTORATION with API outcome was significantly
greater than the association of ENDO' with successful
outcome.

Discussion
The Bresiow-Day Test for homogeneity of odds ratios
was equivalent to the likelihood ratio of 1.1)9
|p=0.296). This result supported homogeneity of the
odds ratios.

This study is based on retrospective radiographic data

and the limitations this created must be taken into
account. There was no knowledge of pre-existing conditions prior to treatment and although it was known that

Fig. 6, Radiographic appearance of root of tooth ] 4

categorized as unsuccessful. A break in the lamina
dura and an obvious apical radioluceiicy Is seen.

Factors in endodontic success

treatment took place more than 1 year prior to the

study, the exact length of time between treatment and
radiographic evaluation is unknown. Another limiting
factor is the evaluation of quality of restoration and
endodontic treatment with a two-dimensional X-ray.
Nevertheless, the fact that over 1000 endodonticalh?
treated teeth were examined overcomes the limitations
as much as is possible and overall, the teeth were representative of the appropriate groups. Examiner bias was
overcome by using two examiners working independently, A third person involved in the study selected the
teeth to be examined but did not evaluate the periradicular status ofthe teeth.
The overall rate of 61,07% of API for the endodontically treated teeth in an urban American population correlates quite closely to the success rate of
similaiiy treated teeth in Oslo, Norway (Eriksen et al
1988, Eriksen & Bjertness 1992), Since this overall
success correlates with other studies it is likely that
the further breakdown of API according to the quality
of the endodontics and coronal restoration is aiso
valid.
Previous studies have shown consistent coronal
leakage when the obturated root canal is exposed to
fluids (Swanson & Madison 1987, Madison et al. 1987,
Torabinejad 1990), These results would appear clinically invalid since the success rate for endodontic treatment reported in most prognosis studies is high
(Strindberg 1956, Grahnen & Hanssen 1961, Kerekes &
Tronstad 1979, Sjogren et ai. 1990), It should be
assumed that most of the teeth studied were adequately
restored before follow up and yet none of the prognosis
studies for nonsurgical endodontic treatment have
included the quality of the permanent restoration as a
criteria for success, A study by Safavi ft aL (1987) found
no difference in the endodontic prognosis if the placement ofthe coronal permanent restoration was delaj^ed.
They assumed that the temporary filling would leak if
left in place for a long time. However, the results of the
study do not bear out this assumption. The results ofthe
present study indicate that the coronal restoration may
be of critical importance for success and appear to
provide clinical evidence that the obturated root canal is
not an adequate barrier to leakage and to validate the
previous coronal leakage studies.
Penetration by bacteria to the apex might not be
necessary for an apical inflammatory response to occur.
In a recent study by our group (Trope et al. 1993) we
have shown that endotoxin predictably moves through
an obturated root canal in vitro. Thus, with a leaking or
absent restoration, it is conceivable that the appropriate

17

bacteria would only have to populate the coronal aspect

ofthe tooth and tbe smaller endotoxin particles, or other
bacterial products, could move to the apex stimulating
the inflammatory response.
Corroborating the results of earlier studies, the present
work has demonstrated quite clearly that an endodontic
treatment of a high quality offers a better prognosis than
a poorly performed treatment. Therefore, endodontic
treatment of the highest quality must remain an important aim for the long-term health of the attachment
apparatus of teeth. However, the present results indicate
that more emphasis should be placed on completion of
the coronal restoration as a means of securing the
results ofthe endodontic treatment.
These results should be alarming to the eododontist in
that it appears that after the initial chemo-mechanical
phase of root canal treatment, the quality ofthe work of
the restorative dentist appears most important for
periapical health of the tooth. Clearly, the current
obturation techniques do not fulfill the main stated
criteria of obturation which is to hermetically seal the
root canal space. The need for better obturation
materials is obvious. An impervious seal may be created
at the orifice after the root canal is filled (Beckham et al.
1993) or the coronal restoration should be extended
apically with a view to sealing off the root canal system.
With these methods retreatment will be difficult if not
impossible, and surgical treatment might be the only
alternative to failed treatment. Research would need to
be carried out to assess if the prognosis of treatment
would be improved to an extent to which these disadvantages are overcome.
Conclusions
In li)10 endodontically
radiographically;

treated

teeth

examined

1, Absence of periradicular pathology was present in

61,07% ofthe teeth examined;
2, The technical quality of the coronal restoration was
significantly more important than the technical
quality of the endodontic treatment for apical
periodontal health.

References
BECKHAM BM, ANDERSON RW, MORRIS CF (1993) An evaluation of

three materials as. barriers to coronal mieroleakage in endodontically treated Xeeth. jourmi of Endodontics t 9 , J 8 S - 9 1 ,
Dow PR, INGLE fl (1955) Isotope determination ofrool canal failure.
OmlSurgerii. Ord Medicine and Oral Pathology S. 110'0-4.

18

H. A. Rtii/et al.

ERIKSEN HM. BfERTNESS E {1991) Prevalence of apical periodontitis

and results of endodontic treatment in middle-aged adults in
Norway. Endodonties and Denial Traumatologti 7. 1-4.
ERIKSCN HM. BjERTNESS E, OBSTAVIK D (1988) Prevalence
and qualit5' of endodontic treatment in an urban adult population in Norway. Endodontics and Dental traumatotogij 4,
122-6.
GRAHNEN H.. EL^NSON L il961) Prognosis of pulp and root canal
therapv, A clinical and radiographic follow-up examination.
Odontologieal Revij 12. 146-65,
KEREKE.S K. TRONSTADL (1979) Long term results of endodontic treatment performed with a standardized technique. Jourmil of
Endodontics S.ii-V,.
MADLSON S. SWANSON K. CHILES SA (1987) An evaluation of coronal
microieakage in endodonticaily treated teeth. Part II. Sealer
Types. Jrainifilo/Endorfontics 13. 109-12.
MAR.SHALL FJ. M..I.SSLEB M. (19611 The sealing of pulpless teeth
evaluated with isotopes. Journal of Dental Medicine 16.
172-7.
RAPP E. BROWN CE. NEWTON CW (1991). An analysis of success
and failure of apicoectomies. Joitrnal oj Endodonties 17.
508-12.

SAFAVI KE. DOWDEN WE, LANCELANI) K (1987) Influence of delayed

coronal permanent restoration on endodontic prognosis,
Endodonties and Dental Traumatology 3. 187-91.
,Si M ONS). IBA NE2 B. FRIEDMAN S, TROPE M (1991) Leakage after lateral
condensation with finger spreaders and D1 IT spreaders, journal of
Endoibnties 17.101-4.
SjOGREN U. HAGGLUNB B, SUMJOVIST G, WiNG K (1990) Factors
affecting the long term results of endodontic treatment, jounia} of
Endodenties Id. 498-504.
STRIKIIBERG LZ (19561 The dependence of the results of pulp therapy
on certain factors. An anal^^tical study based on radiographic and
clinical follow up examinations, Aeta Odontologiea Scandinavia 14
(SuppL 21). 1-175.
SvvANSOM K, MADISON S (1987) An evaluation of corona]
microieakage in endodonticaily treated teeth. Part I. Time periods.
Journal of Endadontfcs 13. 56-9,
ToRABiNEiAi) M. IJNG B, KETTERING JD (1990) In vitro bacterial
penetration of coronally unsealed endodontically treated teeth.
Journal of Endodontics 16. 566-9.
TROPE M. CHOW E, NISSAN R (1993) In vitro endotoxin penetration of
coronally unsealed endodontically treated teeth, journal of Dental
i. 188.Abs(676).