Int. J. Oral Surg.

1982: 11: 156--165

(Key words: [lslIIla. o'omu'al: jlap. buccal: surger)'. llYol)

Closure of oroantral fistula with buccal

flap: Rehrmann versus Moczar
NINA VON WOWERN

Department of Oral Surgery, Royal Dental College and Department of Oral Surgery and Oral Medicine.
University Hospital (Rigshospiwlet). Copenhagell, Denmark

ABSTRACT - A prospective follow-up study of 90 patients with oroantral fistula,

persisting for:?: 10 days, closed with a buccal flap and with systemic penicillin
therapy has been carried out. The purpose was: (1) to compare the results lIfter
closure with a Rehrmann- and a M6c.::cir-buccal flap; (2) to elucidate the reasons
for failures; (3) to clarify when and why a pcrmanent reduction of the vestibular
height may occur after the use of buccal flaps. The material was divided into two
groups: preoperative preparation (PP) group: 52 cases, closed with Rehrmann
(n == 32) and M 6czar flap (n == 20). PP included: a rhino-Iaryngologic examination
and an antral irrigation. Follow-up period: 6--12 months. No PP group: 38
Rehrmann flap cases without PP, and followed for 2 months. The analysis shows:
(1) the result ofclosure is independent oflhe choice offlap; (2) the risk offailure is
negligible by careful regimens (elimination of active maxillary sinusitis
before/after closure, avoiding sneezing); (3) uneventful repair for 2 months
means pennanent repair; (4) flattening of vestibulum for 2 months results in a
persistent condition; (5) pennanent flattening of the vestibulum will occur in
about half of the Re/1I'Inalln flap cases, independent of the state of dentition or
buccal alveolar bone loss; the risk of this complication after a M6czar flap is
minimal; (6) scar tissue at the anterior incision of a M oczar flap is a frequent
finding; (7) why a Rehrmanll flap is the treatment of choice for patients with
natural teeth and a M8czar flap for edentulous patients.

(Received for publication 24 July. accepted 10 December 1981)

Several recent studies have dealt with the
clinical and the radiographical aspects of
oroantral communications\-~9. Experimental
studies lO •22 have confirmed the clinical finding
that a maxillary sinusitis is present when an
untreated oroantral communication has per-
sisted for more than 48 h (i.e. an oroantral
fistula). The inflammation in the antral mucosa
in most cases is localized to the floor of the
maxillary sinus and in relation to the
fistula 204 •9.23 . This condition has to be treated
before closure of the oroantral fistula. The
treatment of choice for closure of oroantral
fistulae appears to be a buccal flapI-3,5,6,9-'9.~3'2'.
However, there are still some controversies
concerning the preoperative management.
The purpose of the present follow-up study
is: (I) to compare the results after clos1.lre with a
buccal flap according to the techniques sug-
gested by RehrmclI1ll and by MVezar ; (2) to

0300-9785/82/030156--10$02.50;0 © 1982 Munksgaard, Copenhagen

 OROANTRAL FISTULA
elucidate the reasons for possible failure after
closure of the oroantral fistulae, and (3) to
clarify when and why a permanent flattening of
the vestibulum may occur after the use of a
buccal flap.
Material and methods
The material consisted of all 90 patients with an
oroantral fistula received in a previously appointed
period of time. Any cases of dental root displacement
in the maxillary sinus was previously eliminated. The
material included 26 females and 64 males between 15
and 75 years of age without systemic disease, chronic
maxillary sinusitis or any kind of previous treatment
of the maxillary sinus, and without symptoms of an
acute maxillary sinusitis in relation to the oroantral
fistula. The regional distribution of the fistulae was:
M-region 94% (85/90), P,-region 6% (5/90). All
fistulae were localized at the alveolar crest. All
patients underwent an identical clinical routine
examination, and radiographical examination de-
scribed previously'·.
Fig. 1. Trapezoid buccal muco-periosteal flap ad
l11udum Relmnann for closure of oroantral fistula. An
excision of attaching muscles was performed.
157
Fig. 2. Btlccal muco-periosteal flap ad l11odu111 M6czar
for closure of oroantral fistula. The flap is displaced
one tooth-width distally" after the attaching muscles
are cut off.
Division into groups according to treatment. The
material consisted of two main groups according to
the preoperative treatment. In the first 38 patients the
oroantral fistula W!lS closed immediately on the day of
admission to the department without a preoperative
preparation (no-PP group). A Relmnann flap was
used for closure, i.e. an extension of a buccal broad
trapezoid mucoperiosteal flap by splitting the peri-
osteum at the nap's base, fixing the end of the flap
above the disepithelialized palatal mucosa, and
closing the remaining wound with single sutures (Fig.
I). The remaining 52 patients underwent a preopera-
tive preparation (PP group), consisting of a rhino-
laryngologic routine examination, including a
nasoantral irrigation of the involved maxillary sinus
(at the Department of Oto-Rhino-Laryngology,
Rigshospitalet). This irrigation was repeated until
clear rinsing water was obtained. Then the oroantral
fistula was closed with a Rehrmanl1 flap (Fig. I) in the
first 32 cases. In the remaining 20 cases a Moczal' flap
was used (Fig. 2), Le. a marginal incision was made
along the teeth (or at the alveolar crest in edentulous
cases) from the tuberosity to the mesial of the canine.
Here a relaxing incision was carried obliquely
upwards and mesially in the vestibulum. The buccal
flap was reflected subperiosteally, and a relieving
parallel incision was made in the periosteum at the
158 VON WOW ERN
flap base to aid in the mobilization of the flap. The
buccal flap was then positioned with its gingival part
moved one tooth width distally, resulLjng in the
transposition of the mesial papilla to the distal of the
teeth to which they belong. The buccal flap was fixed
above the disepithelialized palatal mucosa with single
sutures as was the remaining wound.
None of the patients in the PP group revealed
pathological conditions of the nose and throat (with
the exception of the involved maxillary sinus), which
required treatment before surgical closure.
All the surgical operations in the two groups were
performed with local anesthesia, combined with an 8-
days systemic penicillin therapy starting just before
the surgical operation, Possible areas of pathological
antral mucosa were removed at the surgical oper-
ation. There was no indication of the need for a
LucCaldwell operation in any of the cases. The
patients were discharged on the day of closure with
instructions to avoid sneezing and nose-blowing. The
sutures were removed on the 8th postoperative day.
The histological findings on examination of the
excised tissue from the maxillary sinus and the
oroantral fistula were in accordance with the clinical
diagnosis in all cases.
Follow-up period. The 90 patients were followed up
for 2 months postoperatively. In the PP group, the 52
patients were further examined clinically between t-I
year after the surgical operation.
Special recordings. In the PP group the following
specific conditions were recorded: (I) during the
surgical operation: the position of the buccal alveolar
crest at the oroantral defect in relation to the
mucogingival junction; (2) before the surgical oper-
ation and at the follow-up visits: the vestibular height
at the operative site was compared with that in the
neighbouring areas and in the opposite side and
measured in mm; this examination was performed
while the patient was occluding, and the cheeks were
held slightly aside with dental mirrors in both sides;
the vestibular height was classified as normal, if the
height in mm was >1 of the original height; (3) in the
11 dentate cases of the 20 M6czar flaps, the marginal
gingiva at the anterior incision of the flap was
examined for the possible occurrence of scar tissue.
To secure a uniform treatment of the patients, all
examinations, recordings, surgical operations, and
follow-ups were performed by the investigator
specially trained for this trial 25 - 27 • Note. In all cases a
careful excision of the attaching muscles at the inner
side of the buccal flap was performed in an attempt to
prevent a permanent reduction of the vestibular
heigh t (Figs. 1, 2).
Statistical analysis. These analyses consisted of the
x'-test and Fisher's exact test for two independent
samples.
Results
The statistical analysis showed that no signific-
ant differences existed between the no-PP group
and the PP group regarding the distributions for
age, sex, location of oroantral fistula, number of
dentate and edentulous cases or cases with or
without clinical signs and symptoms of maxil-
lary sinusitis at the day of admission to the
department (Table I), as P > 0.05 in all cases.
The results of treatment in the no-PP group and
the PP group, could therefore be compared. All
24 cases from the PP group without signs and
symptoms of maxillary sinusitis at the primary
clinical examination (Table 1) showed normal
conditions at the oto-rhino-laryngological ex-
amination, including antral irrigation.
Closure of the oroantral fistula
The results ofthe 2 months follow-up are seen in
Tables 2, 3. No cases of later breakdown or
development of maxillary sinusitis were found
in the PP group, which was followed for !-l
year after the surgical closure. Table 2 shows
that one of the 32 Rehrmal1l1 nap cases in the PP
group was a failure. In this patient the fortnight
follow-up showed primary uneventful repair,
but the next day powerful sneezing caused
breakdown and reoccurrence of the fistula, but
without development of a maxillary sinusitis.
None of the 20 Moczar flap cases were failures.
The statistical analysis showed that there was
no significant difference between the Rehnnann
flap and MOczar flap groups regarding the
occurrence of failure (P> 0.05) (Table 2).
Table 3 shows that 8 of the 38 patients (21%)
in the no-PP group failed to heal and developed
an active maxillary sinusitis. The statistical
analysis showed that the frequency of failure
was significantly higher in the no-PP group than
in the PP group (P<O.OOI) (Table 3).
In the 9 cases of failure, the preoperative
preparation (used in the PP group) was
performed, succeeded by a reclosure with a
Rehrmann Dap. The 2 months follow-up re-
vealed primary, uneventful repair in all these
cases.
 OROANTRAL FISTULA 159
Table 1. Comparison of the distributions for sex, age, location of the oro-antral fistula, number of dentate and
edentulous patients and patients without and with signs and symptoms of maxillary sinusitis at the day of
admission between the group without preoperative procedure (no-PP group) and the group with preoperative
procedure (PP group); number of patients (n), level of significance (P), second premolar (P,)

Group Variable n P

Sex females males total

No-PP 7 31 38
pp 19 33 52 >0.05

Total 26 64 90

Age in 15-44 45-75 total

years
No-PP 14 24 38
>0.05
PP 25 27 52

Total 39 51 90

Location of molar
oro-antral P,-region region total
fistula
No-PP 2 36 38
>0.05
PP 3 49 52

Total 5 85 90

Dentition dentate edentulous total

No-PP 14 24 38
>0.05
PP 30 22 52

Total 44 46 90

Maxillary without with total

sinusitis
No-PP 16 22 38
24 >0.05
PP 28 52

Total 40 50 90

Changes in the vestibular height
The comparative statistical analysis (Table 4)
showed that there were no significant dif-
ferences in the distribution of dentate-
/edentulous cases and cases with and without
buccal bone loss between the Rehrmann and the
M6czar flap cases in the PP group, as P>O.05 in
both comparisons. The findings regarding the
occurrence of a permanent reduction of the
vestibular height after the two types of flap
could therefore be compared.
Table 5 shows the findings at the t-l year
follow-up of the 52 cases in the PP group. These
findings were identical with the findings at the 2
160 VON WOWERN

Table 2. Results after surgical closure of 52 oro-antral fistulae by the Rehnnann and M6czar techniques in the
group with preoperative procedure; numher of patients (n), level of significance (P)

Uneventful
primary
Surgical repair 95% confidence Breakdown
technique n (~~) limits n (~~) Total P

Rehrmann 31 97 84-100 I 3 32
>0.05
M6czar 20 100 83-100 0 0 20

Total 51 52

Table 3. Results after surgical closure of 90 oro-antral fistulae in the group without preoperative procedure
(no-PP) and the group with preoperative procedure (PP); number of patients (n), level of significance (P)

Uneventful
primary
repair 95~~ confidence Breakdown
Group n (%) limits n (%) Total P

No-PP 30 79 63-90 8* 21 38
1** <0.001
PP 51 98 90-100 2 52

Total 81 9 90

* No primary repair
** Primary repair, but breakdown after powerful sneezing after a fortnight

Table 4. Comparison of the distribution for the state of dentition and the position of the vestibular bone crest in
relation to the muco-gingivaljunction at the operative site, between the Rehl'llwl1Il and the Moczar flap cases in
the PP group; number of patients (n), level of significance (P)

Dentition
Surgical Dentate Edentulous Total
technique II 11 II P

Rehrmann 21 II 32
Moczar 9 20 >0.05
II

Total 32 20 52

Position of vestibular bone crest in relation to muco-gingival junction

Above Below Total P

11 II II

Rehrmann 7 25 32
>0.05
M6czar 5 15 20

Total 12 40 52
 OROANTRAL FISTULA 161

Table 5. Reduction of the vestibular heigh t after surgical closure of 52 oro-antral fistulae by the Rehrmann and
M6czilr techniques at the !-I year follow-up; these results were identical with the findings at the 2 months
follow-up; number of patients (II), level of significance (P)

The vestibular height

Surgical Normalized Reduced 95~;' confidence
technique 11 eo) 11 e~) limits Total P

Rehmlann 15 47 17 53 35-7\ 32
<0.00\
M6czilr 20 100 0 a 0-17 20

Total 35 17 52

months follow-up. All the cases showed a
reduction of the vestibular height immediately
after the surgical closure, which was more
pronounced after the RehrlllClfltl than the
M6czltr flap. At the 2 months follow-up, 15 of
the Rehrll1C1flfl flap cases (47~-;;) showed a
normalized vestibular height, while 17 cases
(53~~) showed a permanent reduction of the
vestibular height, corresponding to the exten-
sion of the flap (Fig 3). After the Mocza!" flap,
the vestibular height was already normalized at
the 4-week follow-up in all the 20 cases (100%).
The statistical analysis showed that the
number of cases with a permanent reduction of
the vestibular height was significantly higher
after the Relll'lnatln flap than after the MtJc::cir
flap (P<O.OOI).
In the Rehrll1{l11/l flap cases (PP group) the
position of the vestibular bone crest in relation
to the mucogingival junction was found to be
independent of the state of dentition (Table 6, P
> 0.05). Therefore. the comparative analysis of
the relation between the occurrence of vesti-
bular height reduction and (1) the state of
dentition and (2) the extension of the loss of the
buccal crest could be performed. This analysis
(Table 7) showed that the occurrence of
vestibular height reduction after use of the
Rell/"ll1ol1n flap was independent of the state of
dentition and the position of the vestibular bone
crest in relation to the mucogingival junction
(P> 0.05).

Fig. 3. Reduction of the vestibulur height after closure
of an oroantral fistula with a Rehrlll(/Illl flap, even
though a careful excision of the attaching muscles at
the inner side of the flap W~\s performed before
closure.
Marginal g;llgh'(f {(iter Moc::;(rr flClp
In all the 11 dentate cases of the lv!oc:::lt,. flap
(l OO~/:') a permanent occurrence of scar tissue
was observed in the marginal gingiva cor-
162 VON WOWERN

Table 6. Comparison of the position of the vestibular bone crest in relation to the muco-gingivaljunction and the
state of dentition in the Relzmumn flap cases of the PP group; number of patients (n), level of significance (P)

Position of vestibular bone crest in relation

State of to mllco-gingival junction
dentition above below total P

Dentate 4 17 21
:> 0.05
Edentulous 3 8 II

Total 7 25 32

Table 7. Analysis of the 32 Rehrmann flap cases of.the PP group: comparison of the occurrence of permanent
reduction of the vestibular height with (I) the state .of dentition, and (2) the position or the vestibular bone ridge
in relation to the muco-gingivaljunction at the site of the flap; number of patients (n), level of significance (P)

Dentition
Vestibular dentate edentulous
height n n Total P

Reduced 9 8 17
:> 0.05
Normalized 12 3 15

Total 21 II 32

Position of vestibular bone ridge in relation to

muco-gingival junction

above below
/l n Total P

Reduced 4 13 17
:>0.05
Normalized 3 12 15

Total 7 25 32

responding to the anterior incision of the flap;

95% confidence limits: 62-100~'~ (Fig 4).

Fig. 4. Scar tissue in the anterior part or the marginal
gingiva after closure of an oroantral fistula with a
M6czar flap.
Discussion
In the present study the sex and age distri-
butions of the patients and the location of the
oroantral fistulae corresponded to those found
in previous studies 2 ,J,.,I,.12.27. The present ma-
terial did not include acute or severe cases of
maxillary sinusitis as these cases (most likely
primary) are sent to the ear, nose and throat
specialist instead.
 OROANTRAL FISTULA
A previous study has shown that the occur-
rence of failure after closure of oroantral
listulae with a buccal flap is independent of the
presence of neighbouring teeth or the size of the
osseous defect".
The present analysis shows that the failures
may occur independently of the surgical tech-
niques, Rehrmal1ll versus M6c:ar, applied in the
PP group (Table 2, P>O.05). In order to avoid
failures, it is essential that the patients observe
the rules of careful postoperative regimen. This
is illustrated by the results in the PP group
(Table 2), where the only case of failure was
caused by tremendous sneezing after primary
uneventful repair had been observed.
Further, the present analysis shows that the
frequency of failure was significantly higher in
the no-PP group than in the PP group (Table 3).
This finding indicates that a preoperative
elimination of an active maxillary sinusitis is
essential for the prognosis of the surgical
closure.
The follow-up study of the 9 cases of failure
(Table 3) which underwent the described
preoperative preparation before rec!osure with
a Relzrmann flap, revealed primary uneventful
repair in all cases. This finding also underlines
the importance of a careful preoperative
preparation.
A previous radiological study29 (comprising
the PP group) showed that clinical examination,
including antral irrigation, gives very reliable
information concerning the presence or absence
of maxillary sinusitis and is superior to
radiographical examination. The present study
shows that cases of oroantral fistulae without
signs or symptoms of maxillary sinusitis on the
day of admission also show normal conditions
at the oto-rhino-Iaryngological examination
and at the antral irrigation, as the 24 cases from
the PP group without previous signs and
symptoms of maxillary sinusitis showed nOlmal
conditions at the preoperative preparation in all
cases. This indicates that the preoperative
preparation described is only required in cases
with clinical signs and symptoms of maxillary
163
sinusitis before or after the occurrence of the
oroantral fistula, to secure that the maxillary
ostium is in function and to eliminate an active
maxillary sinusitis.
In conclusion, the present analysis, cor-
roborated by previous studiesll-14.2J, underlines
that the risk of failure after closure of oroantral
fistulae with a buccal flap is very small, if a
careful pre-, per- and postoperative regimen is
observed. The present 1-1 year follow-up study
indicates that an uneventful, primary repair for
2 months can be interpreted as a permanen t
repair, and that a flattening of the vestibulum at
the operated site at the 2-month follow-up is a
persistent condition.
Further, this analysis (Table 5) shows that a
permanent reduction of the vestibular height
may occur in abollt half ohhe cases after thc use
of the Rehl'man/l flap, while the risk of this
complication after the use of the Mlie.ar flap is
negligible. These findings are corroborated by
the findings ill previous studies of the
RehrnWI1I1 27 and the Moc::.al'28 flaps.
Subsequently, the 32 Rehrmmm flap cases in
the PP group were analysed in order to clarify
whether there was a correlation between the
occurrence of a vestibular height reduction ancl
the presence of neighbouring teeth or a large
bone loss of the vestibular crest. These analyses
showed that a pronounced vestibular bone loss
may occur in relation to an oro antral fistula,
independent of the state of dentition (Table 6);
the occurrence of a vestibular height reduction
is independent of the state of dentition and the
position of the mucogingival junction in re-
lation to the vestibular bone crest (Table 7), i.e.
it is not possible to anticipate when a RehrmclIln
flap may cause a permanent reduction of the
vestibular height.
Finally, the follow-up analysis of the PP
group revealed that all the present 11 dentate
cases of the M Oczar flap showed a permanent
scar corresponding to the marginal part of the
anterior incision of the flap. This finding
indicates that the risk of scar tissue in the
marginal periodontium is high after the use of a
164 VON WOWERN
M6czar flap. From this point of view, the
Mik=ar flap does not seem to be well-suited for
dentate cases.
In conclusion, the ReJll'lnal1ll flap seems to be
best suited for dentate cases, as it is essential to
reduce marginal gingiva damage to a minimum.
However, the Moear flap may be used in
edentulous cases, as it is essential to avoid a
reduction of the vestibular height in denture-
wearing individuals.
References
1. CHRISTENSEN, O. & GILHuus-MoE, O. : Surgical
treatment of chronic hyperplastic sinusitis and
maxillary sinus empyema of oral/dental origin.
lilt. J. Oral Surg. 1979: 8: 276-283.
2. EHRL, P. A.: Oroantral communication. Int. J.
Oral Surg. 1980: 9: 351-358.
3. ERICSON, S., FINNE, K. & PERSSON, G.: A clinical-
radiographic review of treated oro-antral com-
munications.ln/. J. Oral Surg. 1973: 2: 185-195.
4. ERICSON, S., FINNE, K. & PERSSON, G.: Results of
apicoectomy of maxillary canines, premolars and
molars with special reference to oroantral com-
munication as a prognostic factor. Int. J. Oral
Surg. 1974: 3: 386-393.
5. GROSSENHELLEFORTH, A. & DOKER, J.: Paro-
dontalverhiiltnisse nach Kieferhohlenoperalion
im Fistelbereich. Dtsch. Zalmaer/z/I. Z. 1976: 31:
244-245.
6. HACKMANN, G., MACHTENS, E. & BONING, K.:
Die Deckung des facialen Kieferhohlenfensters
mit Iyophilisierter Dura. Dtsch, Zahnaerzll. Z.
1976: 31: 265-267.
7. HARRISON, D. F. N.: Diseases of the maxillary
sinus. In KAY, 1. W. (ed.): Oral Surgery,
Transactions of the IVth International
Conference on Oral Surgery 1971. Munksgaard,
Copenhagen 1973, pp. 13-18.
8. HJORTlNG-HANSEN, E. & WOWERN, N. v.:
Acquired defects of the antrum. In KAY, L. W.
(ed.): Oral Surgery, Transactions of the rVth
International Conference on Oral Surgery 1971.
Munksgaard, Copenhagen 1973, pp. 19-24.
9. HAANAES, H. R.: A radiographic and clinical
follow-up study of 150 oro-antral communi-
cations. Inl. J. Oral Surg. 1974: 3: 412-421.
10. HAANAES. H. R. & GIUlUUS-MoE, 0.:
Experimental oro-paranasal communications.
Ac/a Odontol. Scand. 1972: 30: 151-165.
II. HAANAES, H. R. & GILHuus-MoE, 0.: A
histologic study of experimental oro-paranasal
communications in monkeys. Int. J. Oral SlIrg.
1972: I: 250-257.
12. HAANAES, H. R. & PETERSEN, K. N.: Treatment
of oroantra! communications. Int. J. Oral Surg.
1974: 3: 124-132.
13. HAANAES, H. R., PETERSEN, K. N. & AAS. E.:
Oral antrostomi. Inl. J. 01'01 Surg. [975: 4: 55-60.
14. KILLEY, H. C. & KAY, L. W.: The maxillary sinus
alld its demalimplicalions.Wright&Sons.Bristol
1975, pp. I-53.
15. JUSELlUS, H. & KALTIOKALLlO, K.: Closure of
antroalveolul' fistulae. J. Laryngol. 1971: 85:
387-393.
16. LEE, F. M. S.: Management of the displaced root
in the maxillary sinus. Int. J. Oral Surg. 1978: 7:
374-379.
17. NEUMAN, K. &PAESSLER, L.: Das Vestibulum oris
nach Verscgluss oroantraler Kommunikationen
durch die Rehmannsche Plastik. Zallll-. MUlld-
Kieferlleilk. 1974: 62: 674-677.
18. NORMAN, J. E. DE B. & CRAIG, G.: Oro-antral
fistula. Oral Surg. 1971: 31: 734-744.
19. PLATZ, H.: Uber eine neue Methode des primiiren
Verschlusses von Mund-Antrum-Verbindungen.
Oesterr. Z. Sio/lla/ol. 1974: 71: 20-29.
20. ROLFFS, J., SCHMELZLE, R .. SCHWENZER, N. &
NEUMANN, V.: Zur chirurgischen Therapie der
odontogenen Sinuitis maxillaris -Ein Bericht
iiber 397 Fiille. Disch. zulll/aerlzl. Z. 1979: 34: 3D-
34.
21. SACHS, S. A., KAY, S. A., SPECTER, 1. & STERN,
M.: Treatment of a persistent oro-antral fistula
with a posteriorly based lateral tongue flap. Inl. J.
Oral Surg. 1979:8: 225-228.
22. VALDERHAUG, J.: Reaction of mucous mem-
branes of the maxillary sinus and the nasal cavity
to experimental periapical inflammation in mon-
keys. fill. J. Oral Surg. 1973: 2: 107-114.
23. WEIJERMAN, 1. E.: Odontogene sinusitis maxil-
laris. N. V. Boekdrukkerij Dijkstra Niemeyer,
Groningen 1972, pp. 1-97.
24. WINKLER, T., WOWERN, N. v. & BITTMANN, S.:
Retrieval of an upper third molar from the
infratemporal space. J. Oral Surg. 1977: 35: 130-
132.
25. WOWERN, N. v. : Frequency of oro-antral fistulae
after perforation to the max.illary sinus. Scand. J.
Dell/. Res. 1970: 78: 394-396.
26. WOWERN, N. v.: Oroantral communications and
displacements of roots into the maxillary sinus. J.
01'01 Surg. 1971: 29: 622-627.
27. WOWERN. N. v.: Treatment arora-antral listulae.
Arch.OlolarY/lgol. 1972: 96: 99-104.
28. WOWERN, N. v.: Correlation between the de-
 OROANTRAL FISTULA 165

velopment of an oroantral fistula and the size of Address:

the corresponding bony defect. J. Oral Surg.
Department of Oral Surgery
1973: 31: 98-102. Royal Demal College
29. WOWERN, N. v.: Clinical and radiographic 4, Ulli\'ersitelsparken
findings in maxillary sinus with oro-antral fistula. DK-2100 Copenhagen 0
Transactions of the VII International Conference Denmark
on Oral Surgery, 1980. Int. J. Oral Surg. 1981: 10:
Supp!. I: 138-142.