Vol. 90 No.

3 September 2000

ORAL SURGERY
ORAL MEDICINE
ORAL PATHOLOGY

ENDODONTICS Editor: Larz Spångberg

Periapical status of root-filled teeth exposed to the oral

environment by loss of restoration or caries
Domenico Ricucci,a Kerstin Gröndahl,b and Gunnar Bergenholtz,c Cetraro, Italy,
and Göteborg, Sweden
GÖTEBORG UNIVERSITY

Objective. Studies in vitro carried out on extracted teeth have demonstrated that bacterial elements may penetrate root fillings
from the coronal to the apical end after a period of exposure to artificial saliva or bacterial culture. To address the clinical
significance of this so-called issue of coronal leakage, a retrospective cohort analysis was conducted of 55 patients with root
fillings that had been exposed to the oral environment because of caries or absent restorations.
Study design. Cases were matched 1-to-1 with regard to initial pulpal and periapical diagnosis, period after completion of
endodontic therapy, tooth type, age of the patient, and the technical quality of the root filling. Only cases with a follow-up period
of 3 years or more were included. Radiographs taken at the last follow-up examination were subjected to a masked evaluation.
Results. A total of 14 osteolytic lesions were recorded. In 43 of the 55 matched pairs (78%), there were identical periapical
conditions. In 9 pairs, a periapical lesion was present in the “open” tooth category, whereas in 3 pairs, a periapical lesion was
seen exclusively in the “intact” tooth. Though the odds-ratio for a lesion to be present in the “open group” was 3.0, this was
not a statistically significant result (P > .10).
Conclusions. Data suggest that the problem of coronal leakage may not be of such a great clinical importance as implicated
by numerous studies in vitro, provided instrumentation and root fillings are carefully performed.
(Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2000;90:354-9)

Bone lesions in the periapical region of teeth often result
from an inflammatory process arising from bacterial
infection of the root canal system. Endodontic treatment
aims to either prevent or eliminate this infection. The
procedures involve mechanical instrumentation, chem-
ical disinfection, and filling of the root canal system.
Successful treatment demands that lesions be healed by
the clinical and radiographic follow-up examination of
any periapical bone lesion that existed before the treat-
ment, or that no lesion developed if there was no lesion
at the outset. Hence, incomplete or absent bone healing
suggests that, because of insufficient instrumentation,
disinfection, or filling, bacterial organisms might remain
aPrivate
practice, Cetraro, Italy.
bDepartment of Oral Diagnostic Radiology, Faculty of Odontology,
Göteborg University, Sweden.
cDepartment of Endodontology and Oral Diagnosis, Faculty of
Odontology, Göteborg University, Sweden.
Received for publication Mar 9, 2000; returned for revision Apr 7,
2000; accepted for publication May 11, 2000.
Copyright © 2000 by Mosby, Inc.
1079-2104/2000/$12.00 + 0 7/15/108802
doi:10.1067/moe.2000.108802
and might continue to adversely affect the condition of
the periapical tissues.
Leakage of bacterial elements from the oral environ-
ment along the root filling to the periapical region has
recently been proposed as an important contributing
factor to the persistence or development of periapical
lesions in root-filled teeth. Support for this assumption
comes primarily from in vitro studies in which radioac-
tive tracer,1 dye,2-4 selected bacterial organisms,4-9
lipopolysaccharide,10 or lipopolysaccharide and bacte-
rial cells of a mixed culture11 have been observed to
penetrate root canal fillings carried out in extracted
teeth. In some of these studies, exposure to artificial
saliva after setting of the sealer seems to have compro-
mised the canal seal to the extent that leakage was
promoted.2,9,12,13 In a combined in vivo/in vitro study
with teeth of monkeys, dye leakage was noted along root
fillings regardless of whether they had been exposed to
the oral environment and salivary influences.3
A retrospective study involving examination of full-
mouth radiographs from a pool of patients with a high
overall rate of periapical lesions (39%) found that
absence of apical periodontitis was significantly more

354
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY
Volume 90, Number 3
Fig 1. Follow-up radiograph taken 5 years after completion of
endodontic treatment of the lower left cuspid in a 60-year-old
man. Original diagnosis was pulp necrosis with apical peri-
odontitis. Extensive caries now surrounds cast post. This case
was categorized in the normal periapical condition group.
common among teeth with good coronal restorations
than poor ones, and that the coronal restoration quality
was more influential than the endodontic filling quality
on the presence of apical periodontitis.14 These findings
suggest that leakage of bacteria and bacterial elements
from the oral environment along the margins of restora-
tions and root fillings is indeed possible and is poten-
tially important for the outcome of endodontic therapy.
Except for the study of Ray and Trope,14 little clinical
data support the importance of the exposure of root fill-
ings to the oral microbiota. Such information is needed
to guide clinical decisions as to the necessity for retreat-
ment of root canal-filled teeth that have lost their
coronal restoration or in which caries has exposed the
root filling material to saliva. This study was conducted
to evaluate the association between coronal restoration
integrity and presence of periapical bone lesions in
root-filled teeth several years after root canal therapy.
MATERIAL AND METHODS
Subjects in this matched, retrospective cohort study
were patients receiving initial root canal therapy and
restorative treatment in a private dental clinic operated
by one of the authors, who personally treated the
patients over a period of 14 years (1983-1996). During
that time, a total of 733 patients had received
endodontic therapy, but 210 never returned for clinical
and radiographic follow-up. For the remaining 523
patients (71%), follow-up data were recorded for 1 or
more years after completion of endodontic treatment.
Patients with a follow-up period of less than 3 years
Ricucci, Gröndahl, and Bergenholtz 355
Fig 2. Follow-up radiograph showing periapical status of
maxillary left first premolar in a woman. Endodontic treatment
had been initiated 6 years and 8 months earlier in this 16-year-
old because of carious pulp exposure. Resin-composite
restoration had been placed, which subsequently was lost. This
case was categorized in the normal periapical condition group.
were excluded from the study so that existing periapical
lesions would have enough time to heal. This decreased
the patient material to 339 individuals and 564 teeth.
From this pool, all patients at follow-up with teeth
that showed either deep caries or a missing restoration
exposing the root filling to the oral environment were
identified (“open group,” Figs 1-3). For comparison,
patients in the remaining pool with intact coronal
restorations were matched 1-to-1 (“intact group”).
Matched cases were selected randomly after
matching on the following:
1. Initial pulpal and periapical diagnosis (vital pulp,
nonvital pulp without periapical lesion, nonvital
pulp with periapical lesion)
2. Observation period after completion of endodontic
treatment (within a range of 24 months)
3. Tooth type (molar, premolar, incisor)
4. Age of patient (10-20 years, 21-40 years, 41-60
years, >60 years)
5. Quality of the root filling assessed by one of the
authors as either adequately or inadequately filled.
The criteria used by Ödesjö et al15 were adopted.
Briefly, adequate root fillings were those ending 2 mm
or less from the radiographic apex if radiopaque material
in the root canal appeared homogeneous and if there was
356 Ricucci, Gröndahl, and Bergenholtz ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY
September 2000

A C

B D
Fig 3. Series of radiographs showing endodontically treated the lower left first molar in a 54-year-old man. A,
Preoperative radiograph showing osteolytic lesions associated with both mesial and distal roots. B, After
confirmation of pulp necrosis, endodontic treatment was completed. Intracanal dressing with calcium-
hydroxide was used between first and second appointment. At 6 year, 7 month follow-up, tooth is asympto-
matic. Although reduced in size, small periapical lesion was still present at distal root. Distinct carious lesion
is at distal aspect of crown. Patient refused to have carious lesion treated but returned for renewed examina-
tion 1 year, 5 months later. During this time, bridge was lost. D, Periapical status was unchanged relative to
the previous follow-up examination (C). This case was categorized in the osteolytic lesion group.

no visible space between the material and the root canal
walls. Root fillings were classified as inadequate if they
ended more than 2 mm short of the radiographic apex, if
they extended beyond the radiographic apex, or if they
displayed voids along the root canal walls.
When possible, matching was also conducted with
regard to presence or absence of a cast or prefabricated
post at the time of the last follow-up.
One tooth per patient was selected for the study. If
patients had more than 1 tooth fitting the criteria for
inclusion, 1 was selected at random. Teeth were
excluded if the radiographic image quality was poor or
if the apical region of all roots could not be seen clearly
in both the initial and the follow-up radiographs.
Fifty-five patients fit the study inclusion criteria for
the “open” group. Records confirmed that no tooth was
included in the study that had lost a restoration in the
weeks just before the last follow-up examination.
Endodontic treatment procedure
All endodontic treatments had been performed
following a standardized protocol. Before the initiation
of treatment, at least 1 diagnostic radiograph was taken.
This radiograph and subsequent radiographs were
obtained with a film holder (Rinn Corp, Elgin, Ill) to
permit projection according to the paralleling technique.
A long-cone (Fiad, Explor-X 65 KV, Trezzano S/N, Italy)
and Kodak Ultraspeed film 31×41 (DF58) or 22×35
(DF54) (Eastman Kodak Co, Rochester, NY) were used.
Radiographs were processed manually in a dark room
according to the recommendation of the manufacturer.
A pulpal diagnosis was established after recording the
signs and symptoms. Endodontic treatments were
performed with a strict aseptic technique. Plaque and
calculus were removed from the tooth surfaces with
ultrasound scaling, curettes, or both, followed by
polishing with a prophylaxis paste in a rubber cup. After
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY
Volume 90, Number 3
rubber dam isolation, the field (tooth, rubber dam, and
clamp) was disinfected with 30% hydrogen peroxide
and 5% tincture of iodine.16 In case of substantial loss
of tooth substance, buildups were carried out with either
a copper ring, composite material, or reinforced glass
ionomer cement to exclude bacterial contamination
during the operative procedure.
After the working length was established, an effort
was made to limit the instrumentation to the confines
of the root canals. Abundant amounts of 1% sodium
hypochlorite solution were used for irrigation. After
adequate preflaring of the coronal 2⁄3 of the canals and
apical instrumentation, the canals were filled with cold,
laterally condensed gutta-percha and a sealer. Different
sealers were used over the years, including AH 26 (De
Trey Frères, SA, Zürich, Switzerland); Bioseal (Ogna,
Milano, Italy), Endomethasone (Septodont, Saint-
Maur-des-Fossés, Cedex, France), Pulp Canal Sealer
(Kerr Manufacturing Co, Romulus, Mich), Tubliseal
(Kerr Manufacturing Co), and Apexit (Vivadent Ets,
Schaan, Liechtenstein). The canals were filled in either
1 visit (usually teeth with vital pulps) or 2 or more
visits (nonvital cases) after placement of calcium
hydroxide or metacresylacetate (Cresatina, Ogna,
Milano, Italy) as an intracanal dressing.
Data collection
Patient records were examined at follow-up
regarding any occurring symptoms or spontaneous or
provoked pain suggestive of an ongoing infection.
Inspection and palpation of the mucogingival fold was
also carried out to show fistulous tracts, intraoral
swelling, or abscess formation.
From each follow-up examination, at least 1 and
sometimes 2 periapical radiographs were obtained. For
evaluation of the cases included in this study, only
those radiographs that were available from the last
follow-up examination were considered. These radio-
graphs of both “open” and “intact” teeth were
subjected to a masked evaluation of the periapical
status. Masking was carried out by shielding the
coronal portion of the tooth with an opaque tape. Two
examiners, an endodontist and a radiologist with many
years of experience in assessing periapical radio-
graphs, carried out the examinations. Therefore, these
examiners (neither of whom was the operator) had no
knowledge of whether the tooth belonged to the “open”
or “intact” tooth category.
The absence or presence of an osteolytic lesion was
determined through joint assessment between the 2
examiners. Absence of lesion was designated if the
status of the periapical region was either within normal
limits or unclear (widened apical periodontal space or
diffuse lamina dura). An osteolytic lesion was regarded
Ricucci, Gröndahl, and Bergenholtz 357
Table I. Periapical status of teeth within matched pairs
(n = 55)
Intact teeth
Periapical No periapical
Periapical status lesion lesion Total
Open teeth
Periapical Lesion 1 9* 10
No Periapical Lesion 3 42 45
Total 4 51 55
*Not statistically significant (P > .10)
as present when there was a distinct radiolucent area
associated with the root apex (Fig 3, D). If only 1 root
in a multirooted tooth presented with a lesion, the tooth
was assigned a periapical lesion. In cases with a lesion
present in the follow-up radiograph, the preoperative
radiographs were examined to determine the extent to
which the periapical lesion had been present at the time
of treatment. If so, a comparison was made with the
size of the initial lesion, and it was determined whether
the persisting lesion had clearly been reduced. The
preoperative radiographs of the other cases were not
attended to in this part of the study.
The assessment of the radiographs was carried in a
room with dimmed light. Each radiograph was placed
on a light table. After capturing the image with a
video camera, the radiographs were projected on a
television monitor screen at 20× magnification. Light,
brightness, and contrast were adjusted for optimal
viewing conditions.
Comparisons of periapical status within the 55
matched pairs of “open” and “intact” teeth were carried
out with McNemar chi-squared test.17
RESULTS
The age of the patients ranged from 11 to 70 years,
with a mean of 35.4 years for the “open” group and
30.6 years for the “intact” group. The follow-up period
from the time of completion of endodontic treatment to
the last follow-up examination ranged from 3 to 15
years with a mean of 7.2 years for the “open” group
and 7.1 years for the “intact” group. The quality of the
root fillings in general was high, and 95% of the teeth
were considered adequately filled. Matching with
regard to presence of post occurred in 14 pairs.
Because of lack of teeth with appropriate other
matching criteria, 2 “intact” teeth with posts had to be
included in which the “open” teeth had no post.
The follow-up radiographs presented with a total of
14 periapical lesions: 10 in the “open” group and 4 in
the “intact” group. Fifteen teeth (7 in the “open” group
and 8 in the “intact” group) were judged to have an
358 Ricucci, Gröndahl, and Bergenholtz
unclear periapical condition. Thus, the remaining 81
teeth were found to have a normal periapical status of
which 2 (1 per group) exhibited periapical sclerosis.
Table I shows the comparison among the 55 matched
pairs. In 1 pair, both teeth had periapical lesions; in 9
pairs, a lesion was present in the “open” tooth only; and
in 3 pairs, a lesion was seen exclusively in the “intact”
tooth. Thus, 43 of the 55 pairs (78%) had identical peri-
apical conditions. The odds ratio for a lesion to be asso-
ciated with an “open” tooth was 9 to 3 = 3.0 (95% CI =
0.75-26.02), indicating a nonstatistically significant
tendency (P > .10) for “open” teeth to be associated
with periapical lesions in comparison to “intact” teeth.
Of the 14 lesions observed at the follow-up examina-
tion, 5 lesions had developed after the completion of
treatment (3 in the “open” group and 2 in the “intact”
group). The remaining lesions (7 in the “open” group
and 2 in the “intact” group) were clearly reduced in
size in comparison with the preoperative condition (Fig
3). None of the cases presented clinical symptoms
suggesting an ongoing infection.
DISCUSSION
The results of this study indicated that exposure of
root fillings to the oral microbiota, only in a limited
number of cases, influenced the periapical status. Of
the 55 teeth examined that had experienced exposure to
the oral environment, only 10 teeth (18%) presented
with an osteolytic lesion at the last follow-up examina-
tion, whereas the remaining 45 teeth (82%) exhibited
periapical conditions without overt lesion. Among
those teeth with lesions at follow-up, 3 had lesions that
emerged after therapy; 7 lesions were present before
treatment, but they were clearly reduced in size.
Nevertheless, in all these instances, leakage of bacte-
rial elements from the oral environment may have been
in effect and might have resulted in new lesions or
prevented complete healing of the lesions present at the
outset. It is also possible that, in all or some of these
cases, insufficient instrumentation, disinfection, and
filling caused residual organisms to affect the peri-
apical condition. Still, in comparison with the “intact”
group, the “open” group had more lesions (though this
difference was not statistically significant).
We realize that this study may be underpowered. One
can calculate that approximately 92 matched pairs
would have been needed to reach statistical difference
with a matched odds-ratio of 3.0 at the 5% level of
significance. Hence, the results of the current study
should not be interpreted to show that coronal leakage
of bacterial elements is irrelevant for the outcome of
endodontic therapy. However, it does suggest that the
problem may not be of such a great clinical importance
as implicated by numerous in vitro reports, which have
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY
September 2000
shown dye and bacterial element penetration along root
fillings carried out in extracted teeth.2-4, 5-11
Many factors in vivo cannot be accounted for by a
simplified experiment in vitro. Except for the fact that
tracers, such as dyes or isotopes, do not accurately
reflect the potential for bacterial migration or flow of
bacterial macromolecules and nutrients,4,18 the lack of
control for the involvement of host defense factors is a
most obvious shortcoming. Normally, under in vivo
conditions, a multitude of bacteria of pathogenic
potential are required to produce and maintain a peri-
apical lesion that can be detected by radiographic
means. The in vitro models used in the studies referred
to, which were designed to test bacterial penetration of
root fillings,4-9 were not quantitative and could only
record turbidity in a broth on bacterial entry at the
apical end of the root filling. Although such a finding
in itself is interesting, it may reflect penetration of just
1 or a few bacterial organisms. Hence, the in vitro test
systems used so far are clearly much more sensitive
than radiographic evidence of bone lesions can ever be.
Therefore, the relevance of in vitro findings to clinical
conditions is questionable, though the susceptibility of
sealer root fillings to coronal exposure of saliva and
bacterial broth should not be dismissed and warrants
clinical concern.
The quality of endodontic care rendered to the
patients of the current study was high, in contrast to
that received by subjects in the study by Ray and
Trope,14 in which many root fillings were of inferior
quality. Furthermore, because of the design of that
study, no information was available as to the
endodontic methodology used and the length of the
time fillings had been in place. By contrast, most root
fillings in this study were within the confines of the
root canal and extended to an ideal distance from the
radiographic apex. There were no attempts to maintain
so-called apical patency, which easily may promote
overinstrumentation and less successful outcomes,
such as those demonstrated by earlier studies.19-21
Therefore, in the cases observed, bacterial elements
from the oral environment may not have had as great
an opportunity to enter the apical end of the root
canals.
The exact time for exposure to the oral environment
could not be determined in each individual case.
However, records verified that teeth with restorations
that had been lost only temporarily over a few days or
less than weeks were not included in the material.
Thus, given the extensive caries in some teeth and the
fact that some patients had not received a permanent
restoration after root canal filling, oral exposure in
terms of time was necessarily months or years.
The extent to which fillings seal instrumented root
ORAL SURGERY ORAL MEDICINE ORAL PATHOLOGY
Volume 90, Number 3
canals has been and remains a major concern in
endodontics. If there is a poor or absent root filling,
any remaining organisms in the root canal will have
space after treatment (either in the entire canal or
lateral to the root filling) to multiply into such numbers
that a periapical inflammatory lesion is initiated, main-
tained, or both. Though studies have found successful
outcomes without root fillings after bacterial elimina-
tion by instrumentation and disinfection22 or use of
strong antiseptics,23 others have reported that mechan-
ical instrumentation and use of biologically acceptable
disinfecting procedures are insufficient.24,25 Therefore,
a root filling that entirely seals the instrumented root
canal space is highly desirable but may not invariably
be possible, as suggested by a large number of in vitro
studies.18,26 Therefore, and in spite of the findings in
this study, the addition of a coronal seal for root fillings
seems well advised.
We thank Dr Dan Caplan for his valuable comments and
suggestions for improvement of this paper.
REFERENCES
1. Marshall FJ, Massler M. The sealing of pulpless teeth evaluated
by radioisotopes. J Dent Med 1961;16:172-84.
2. Swanson K, Madison S. An evaluation of coronal microleakage
in endodontically treated teeth. Part I. Time periods. J Endod
1987;13:56-9.
3. Madison S, Wilcox LR. An evaluation of coronal microleakage
in endodontically treated teeth. Part III. In vivo study. J Endod
1988;14:455-8.
4. Barthel CR, Moshonov J, Shuping G, Örstavik D. Bacterial
leakage versus dye leakage in obturated root canals. Int Endod J
1999;32:370-5.
5. Torabinejad M, Ung B, Kettering JD. In vitro bacterial penetra-
tion of coronally unsealed endodontically treated teeth. J Endod
1990;16:566-9.
6. Khayat A, Lee SJ, Torabinejad M. Human saliva penetration of
coronally unsealed obturated root canals. J Endod 1993;9:458-61.
7. Chailertvanitkul P, Saunders WP, Mackenzie D. An assessment
of microbial coronal leakage in teeth root filled with gutta-
percha and three different sealers. Int Endod J 1996;29:387-92.
8. Chailertvanitkul P, Saunders WP, Saunders EM, Mackenzie D.
An evaluation of microbial coronal leakage in the restored pulp
chamber of root-canal treated multirooted teeth. Int Endod J
1997;30:318-22.
9. Chailertvanitkul P, Saunders WP, MacKenzie D. Coronal
leakage of obturated root canals after long-term storage using a
polymicrobial marker. J Endod 1997;24:610-3.
10. Trope M, Chow E, Nissan R. In vitro endotoxin penetration of
Ricucci, Gröndahl, and Bergenholtz 359
coronally unsealed endodontically treated teeth. Endod Dent
Traumatol 1995;11:90-4.
11. Alves J, Walton R, Drake D. Coronal leakage: endotoxin pene-
tration from mixed bacterial communities through obturated,
post-prepared root canals. J Endod 1998;24:587-91.
12. Magura M, Kafrawy AH, Brown CE, Newton CV. Human saliva
coronal microleakage in obturated root canals: an in vitro study.
J Endod 1991;17:324-31.
13. Siqueira JF, Rocas IN, Lopes HP, de Uzeda M. Coronal leakage
of two root canal sealers containing calcium hydroxide after
exposure to human saliva. J Endod 1999;25:14-6.
14. Ray H, Trope M. Periapical status of endodontically treated
teeth in relation to the technical quality of the root filling and the
coronal restoration. Int Endod J 1995;28:12-8.
15. Ödesjö B, Helldén L, Salonen L, Langeland K. Prevalence of
previous endodontic treatment, technical standard and occur-
rence of periapical lesions in a randomly selected adult, general
population. Endod Dent Traumatol 1990;6:265-72.
16. Möller ÅJR. Microbiological examination of the root canals and
periapical tissues of human teeth. Methodological studies.
Odontol Tidskr 1966;74(suppl 5 and 6):12-68.
17. Norman GR, Streiner DL. Biostatistics: the bare essentials. St
Louis: Mosby; 1994.
18. Wu M-K, Wesselink PR. Endodontic leakage studies reconsid-
ered. Part I. Methodology, application and relevancy. Int Endod
J 1993;2:37-43.
19. Strindberg LZ. Dependence of the results of pulp therapy on
certain factors. An analytical study based on radiographic and clin-
ical follow-up examinations. Acta Odontol Scand 1956;14(suppl
21):1-175.
20. Grahnén H, Hansson L. The prognosis of pulp and root canal
therapy. Odontol Revy 1961;12:146-62.
21. Bergenholtz G, Lekholm U, Milthon R, Engström B. Influence
of apical overinstrumentation and overfilling on retreated root
canals. J Endod 1979;5:310-4.
22. Klevant FJ, Eggink CO. The effect of canal penetration on peri-
apical disease. Int Endod J 1983;16:68-75.
23. Wang Z, Wang J. A clinical observation on extensive periapical
lesions of posterior teeth and their treatment with resinifying
therapy. Quintessence Int 1989;20:143-7.
24. Byström A, Sundqvist G. Bacteriologic evaluation of the effect
of 0.5 percent sodium hypochlorite in endodontic therapy. Oral
Surg Oral Med Oral Pathol 1983;55:307-12.
25. Byström A, Sundqvist G. The antibacterial action of sodium
hypochlorite and EDTA in 60 cases of endodontic therapy. Int
Endod J 1985;18:35-40.
26. Saunders WP, Saunders EM. Coronal leakage as a cause of
failure in root canal therapy: a review. Endod Dent Traumatol
1994;10:105-8.
Reprint requests:
Gunnar Bergenholtz
Faculty of Odontology
Box 450
S-405 30 Göteborg, Sweden
Gunnar.Bergenholtz@odontologi.gu.se