Journal of Cranio-MaxillofitcialSurgery (1996) 24, 237-244

© 1996European Association for Cranio-Maxillofacial Surgery

Endosteal implants in the irradiated lower jaw

F. Watzinger, R. Ewers, A. Henninger, G. Sudasch, A. Babka*, G. Woelfl**

University Clinicfor Maxillofaeial Surgery, (Head." Prof. Dr. Dr. Rolf Ewers), Medical School, * University
Clinic for Dentistry, (Head." Prof. Dr. Rudolf Slavicek), Medical School, **Institute of Medical Statistics
(Head." Prof. Dr. Peter Bauer), University of Vienna, Vienna, Austria

SUMMARY. Since 1990 Endosteal implants have been inserted in the irradiated lower jaw at our clinic.
1MZ implants have been used for dental rehabilitation in 26 patients (21 male, 5 female) suffering from
squamous cell carcinomas stage T2-T4136. The implants were either placed in local bone and soft tissue (group 1,
n = 6 0 implants), or in local bone after marginal mandibulectomy and transplanted soft tissue (group 2, n--26
implants), or in transplanted bone and soft tissue (group 3, n=52 implants). Life-table analysis according to
Kaplan-Meier demonstrated a 3-year implant survival rate of 87.8% in Group 1, 69.1% in Group 2 and 58.3%
in Group 3. There was no statistical significant difference in the amount of marginal bone loss and the degree of
marginal infection between the three groups (P > 0.29).
Major complications: A mandibular fracture passing through an empty implant socket 8 months after implant
loss (Group 2) was caused by postradiation-osteonecrosis; implant removal and bone resection was mandatory.
The poor results of the bone graft group may be explained by two patients, in whom simultaneous placement of
implants in nonvascularized bone grafts was carried out, intraoral tissue breakdown led to graft failure and loss
of the implants (n = 10).

INTRODUCTION restoration in patients having had mandibular resec-

Radiotherapy is an essential part of the treatment
modalities in patients suffering from squamous cell
carcinoma of the oral cavity. Denture retention is
frequently compromised by inadequate saliva pro-
duction following irradiation (Larsen et al., 1993).
Moreover, many patients have surgical defects which
cannot be adequately reconstructed without implants.
In these patients endosteal implants may be placed
either in irradiated local bone or in bone grafts.
It is well known that irradiation causes tissue
damage to structures adjacent to the tumour.
Hypovascular, hypocellular and hypoxaemic tissue
results (Marx, 1983; Marx and Johnson, 1987). Risks
that must be discussed include postradiation-
osteonecrosis (PRON), fracture, development or
recurrence of tumour with unknown risks of treat-
ment, nonintegration of implants (early or delayed)
and soft tissue problems (McKenzie et al., 1993).
Accidental or surgical trauma leads to delayed wound
healing and may include the risk of PRON. The risk
of developing PRON depends on the radiation dose
(Murray et al., 1980; Beumer et al., 1982), as well as
on fractionation and volume of irradiated tissue.
Nevertheless, some authors present good results of
endosteal implants placed in irradiated bone
(Albrektsson, 1988; Parel and TjellstrOm, 1991;
Granstroem et al., 1992; Taylor and Worthington,
1993).
Endosteal implants may also serve for prosthetic
237
tion and bone graft (Urken et al., 1991; Martin et al.,
1992; Zlotow et al., 1992; Sclaroff et al., 1994;
Donovan et al., 1994; Moscoso et al., 1994). Komisar
(1990) asserted that restoration of mandibular conti-
nuity alone does not enhance function in the majority
of head and neck cancer patients. The degree to
which mastication is affected depends on the amount
of mandible removed, but the status of tongue func-
tion is equally important. The use of osseointegrated
implants can minimize masticatory limitations as well
as improve facial contour.
MATERIAL AND METHODS
Since 1990, endosteal implants have been placed in
the irradiated lower jaw at our clinic. In 26 patients
suffering from squamous cell carcinomas stage
T2-T4, 136 IMZ implants have been used for dental
rehabilitation. Twenty one patients were male and 5
female, their ages ranged between 41 and 79 (mean
62) years.
After tumour staging including tattooing, preoper-
ative radiochemotherapy was applied. Fractionated
radiotherapy with a total dosage of 50 Gray was
given. Surgery followed 4 weeks-3 months after
irradiation. Radical tumour resection was carried out
as a marginal mandibulectomy in five cases and
segmental resection in 12 cases.
Soft tissue reconstruction was carried out by a
238 Journal of Cranio-Maxillofacial Surgery

Table 1 - Distribution of the implants concerning function and

GINGIVALINDEX $ULCU$ BLEEDING INDEX
implant loss
• ,, °o
Number of implants All groups Group 1 Group 2 Group 3 3~ 3

251- 25

In function 78 38 15 25 2-1 . . . . 2
Without function 16 8 0 8 15 -- 15
Lost
1- -- 1 ,.°°
No osseointegration 10 0 0 10
Marginal bone loss 16 7 5 4 05 - 05

Recurrent tu., death 16 7 4 5 0. . . . 0

Sum 136 60 24 52

Fig. 3 - Gingival index (GI) according to Loe and Silness, 0-3;

(0 = no inflammation, 3 = distinct inflammation; simplified).
Survival rate
Sulcus bleeding index (BI) according to Muehlemann, 0-3; (0 = no
bleeding on probing, 3 =distinct bleeding; simplified). Median
values per patient (.); Median values per group (-).
60 j. L_ 58.3%
50

i
40 ~ group 1
30 -- group 2
20 -- group 3

100 I I I I I I 1 I mths
o 5 10 15 20 25 30 35 40

Fig. 1 - Implant survival rate according to Kaplan-Meier.

M A R G I N A L BONE LOSS C R E V I C U L A R PROBING DEPTH

-" .I. --
3 ¸

2. ,*
!: q •
=•

Fig. 4 - Implants inserted in the area of marginal resection of the

mandible (Group 2).
Fig. 2 - Marginal bone loss and crevicular probing depth (mm),
mean values per patient (.); mean values per group (-).

Group 2: The implant socket consisted of

irradiated local bone in which marginal mandibulec-
local flap in twelve cases, by a myocutaneous pec-
tomy had been carried out; the implant was sur-
toralis muscle flap in five cases, by a microvascular
rounded by transplanted soft tissue (Fig. 4).
jejunal flap in seven cases, and by a microvascular
Group 3: The implant socket consisted of trans-
osteomyocutaneous iliac graft in two cases.
planted bone (Fig. 5) and soft tissue, the transplant
Bone reconstruction was performed by a microvas-
recipient region had been irradiated preoperatively.
cular iliac bone graft in three cases, by a free iliac
This group is rather heterogenous, since it contains
crest bone graft in five cases, by a rib graft in two
microvascutar anastomosed bone grafts and different
cases and by a cancellous iliac bone graft in Titanium
types of avascular bony reconstruction, primary and
mesh in two patients.
second step implant placement as well. These were
Prosthetic rehabilitation was carried out by using
summarized so as not to have too many small groups
endosteal implants, the minimum interval between
and to obtain a general view of implant placement in
irradiation and implant insertion was 12 months.
these patients. However, Table 2 presents the distri-
We differentiate between three groups (Table 1,
bution of the implants and the rate of implant loss
Figs. 1, 2, 3): according to different types of bone grafting. In the
Group 1: The implant socket consisted of event of two-step surgery the implants were placed
irradiated local bone and soft tissue. at least 6 months after bony reconstruction.

Fig. 5 - Implants inserted in bone grafts after segmental resection
(Group 3).
Table 2 - Distribution of implants in Group 3 according to the
variable extent of resection and the various type of bone grafting
Mean resection Number of
Type of bone grafting (cm) implants
Microvascular 6.5 19
anastomosed bone
graft and second
stage implant
placement
Avascular bone graft 5.5 18
and simultaneous
implant placement
Avascular bone graft 5 15
and second stage
implant placement
We used 2-6 IMZ cylinder implants (Friatec,
Mannheim, Germany). Uncovering of the implants
took place at least 6 months after implantation.
Prosthetic rehabilitation followed using a rigid bar
screw--retained on the implants and a removable
overdenture. The patients had a routine follow-up
every 3 months during the first year following
implantation and every 6 months thereafter.
Radiographic studies included an orthopantomogram
and dental films, using a standard filmholder and a
rectangular technique for standard documentation.
The marginal bone loss was taken from the orthopan-
tomogram and dental films, the distances from the
superior border of the implants without abutments
to the apical border of the radiolucencies surrounding
the implants were measured mesially and distally to
the implants; geometric distortion was corrected by
the known length of the implants. Apart from radio-
graphic signs, osseointegration was proved by pal-
pable stability. Peri-implant crevicular probing was
measured mesial, distal, buccal and lingual to the
implant. Further clinical parameters were established
Endosteal implants in the irradiated lower jaw 239
to estimate marginal infection: gingival index (L6e
and Silness, 1963) and sulcus bleeding index
(Miihlemann, 1978). The function of denture was
evaluated regarding retention of the implant-
supported overdenture and occlusion. The patients
were questioned whether they felt satisfied with their
masticatory function and whether speech function
was satisfactory.
Statistical methods
Kaplan-Meier analysis of implant survival per
implant was estimated, ignoring the dependence
between implants in the same patient. Comparison
of implant survival times between the three groups
was performed on minimal survival time of the
implants per patient. Since the most critical phase of
implant survival is the time of osseointegration, the
generalized Wilcoxon test for censored survival times
was applied as it attributes more weight to the early
phase as compared with the log-rank test.
Marginal bone loss, crevicular probing depth, gin-
gival index and sulcus bleeding index in patients 2
years after implantation were examined and analysed
Lost
implants statistically (Group 1 : 1 0 patients, Group 2 : 4
patients, Group 3:4 patients).
1 Concerning marginal bone loss and crevicular pro-
bing depth, mean values for implants per patient
were established; concerning gingival index and sulcus
bleeding index medians of the implants per patient
10
were used. Patients, not single implants, were used as
the sample unit for the statistical analysis. The
3 Kruskal-Wallis test was used to assess the indepen-
dent effect of Groups 1-3 classification on the out-
come of marginal bone loss, crevicular probing depth,
gingival index and sulcus bleeding index. All P-values
are results of two-sided tests. The SAS statistical
software system (SAS Institute, Cary, NC) was used
for calculation. P-values <0.05 were considered as
statistically significant.
The aim of the study is to evaluate differences of
osseointegration, condition of peri-implant bone and
soft tissue, and implant survival rate, depending on
the site of bone and soft tissue in which the implants
had been placed.
RESULTS
The results are presented in the Table 1 and Figures
1-3.
Table 1 deals with the results of the implants having
been placed (in total 136), the corresponding data of
the single groups are presented in the Table 1.
Figure 1 presents the Kaplan-Meier analysis
(Kaplan and Meier 1958) of implant survival, showing
a 3 years survival rate of 87.8% in Group 1, 69.1%
in Group 2 and 58.3% in Group 3.
240 Journal of Cranio-Maxillofacial Surgery

The comparison of minimal implant survival times
per patient gave a hint on possible differences between
the groups (P < 0.10): Implants in transplanted bone
showed a slightly lower minimum survival time.
Figure 2 presents the results of marginal bone loss
and probing depth: mean values for implants in each
patient and mean values for each group are presented.
There was no statistically significant difference
between the groups (P>0.29).
Figure 3 presents the results of gingival index and
sulcus bleeding index: median values of the implants
in each patient and median values of the patients in
each group are presented. There was no statistically
significant difference between the groups ( P > 0.68).
The results of clinical investigation of denture
retention and occlusion, and of the patients' satisfac-
tion with masticatory and speech function are pre-
sented in Table 3.
DISCUSSION
In patients who have been treated for squamous cell
carcinoma of the oral cavity, conventional prosthetic
restoration is often impossible. Scars, defects of soft
tissue and bone result in an inadequate denture base
for a mucosally-worn prosthesis. The placement of
endosteal implants in irradiated bone is discussed
controversially. Some authors present good results
(Albrektsson, 1988; Parel and TjellstrOrn, 1991;
Granstroem et al., 1992; Taylor and Worthington,
1993), others refuse to employ implant placement,
considering the risk of postradiation osteonecrosis
overshadowing the possible benefit of making pros-
thetic restoration (Fischer-Brandies, 1990). Most
authors (Granstroem et al., 1992; Taylor and
Worthington, 1993; Esser, 1994) agree that a long
interval between radiotherapy and implant placement
is preferable. Waechter and Stoll (1994) proved maxi-
mum bone damage to be at 6 months after irradiation.
Therefore the implants were placed at least 1 year
after irradiation. The uncovering of the implants was
carried out 6 months after placement, since it is
known that increased integration time also increases
the amount of osseointegration in irradiated bone
(Larsen et al., 1993).
Regarding the results of implant survival according
to Kaplan-Meier, the survival rate in Group 1 was
best (87.8%), no severe complication occurred in
these patients. The survival rate in Group 2 was
69.1%. Four primarily osseointegrated implants were
lost in one patient due to PRON. This patient had
had a marginal resection of the anterior mandible
and soft tissue reconstruction with a microvascular
jejunal flap. One of four implants was lost 18 months
after placement, caused by progressive loss of osseoin-
tegration (Fig. 6). The patient had very bad oral
hygiene and after a further 8 months the rigid bar
connecting the implants was removed to treat massive
peri-implant inflammation. Two months later, the
patient suffered from a mandibular fracture passing
through the empty implant socket (Fig. 7). The rest
of the implants had to be removed and a mandibular
resection (Fig. 8) was carried out, histological evalu-
ation proved PRON. It is known that primary radio-
therapy of tumours, adjacent to the mandible or
clinically invading it, as in the patient presented, is
more likely to result in PRON (Murray et al., 1980).
Therefore implant placement in the area of a marginal
mandibular resection (Group 1) may be somewhat
questionable. Probably, it would be more favourable
to carry out segmental resection and bone grafting in
these patients, if implant placement is considered
later on. It has been reported that PRON occurs
most frequently within the first year following
irradiation but may also present many years later
(Murray et al., 1980; Epstein et al., 1987). Therefore,
whenever implant loss happens in irradiated bone,
very careful follow-up is mandatory not to overlook
development of PRON. We are concerned about this
problem occurring in our patient, nevertheless in the
other patients there has been no evidence of PRON
after implant loss until now.
It is remarkable that in the bone graft group
(Group 3) there was a considerable amount of
implant loss within the first months. Deficient osseo-
integration of all implants (10 in total) occurred in
two patients within 6 months. In the other patients
of Group 3, only four implants have been lost.
Therefore the low survival rate (58.3%) in this group
results from the failure in two patients. In one patient
a bicortical iliac crest bone graft was transplanted
and implants were placed simultaneously. Due to
intraoral tissue breakdown, graft failure occurred and

Table 3 - Function of prostheses in patients wearing implant-supported overdenture (n = 17)

Denture retention, occlusion Masticatory function Speech function

Group I Satisfactory 3 2 3
Unsatisfactory 1 2 1
Group II Satisfactory 7 6 7
Unsatisfactory 3 4 3
Group III Satisfactory 4 3 3
Unsatisfactory 0 1 1
All groups Satisfactory 14 11 13
Unsatisfactory 4 7 5
 Endosteal implants in the irradiated lower jaw 241

Fig. 6 - Orthopantomogram,demonstratingthe loss of one implant (arrow) 18 months after placementin a patient of Group 2.

Fig. 7 - Fracture of the mandible passing through the empty implant socket (arrow), histologyproved postradiation-osteonecrosis
(PRON) 10 months later.

all implants were lost. In the other patient, an iliac
crest cancellous bone graft in Titanmesh ® was incor-
porated, implants were placed simultaneously, too.
Postoperative wound dehiscence led to the intraoral
exposure of the Titanmesh ® and grafted bone, graft
and implants were lost. Dumbach et al. (1994) also
presented poor results with the last mentioned method
in irradiated patients. Since then, we no longer use
these methods together with simultaneous implant
placement. If ever simultaneous implant placement is
contemplated, we think that it should be carried out
only in vascularized grafts. They do not depend on
the blood supply from the surrounding tissue
(Riediger, 1987), primary osseointegration of endos-
teal implants may be expected (Donovan et al., 1994).
Concerning marginal bone loss, the best results
have been observed in Group 3. In Group 2 the
results were slightly better than in Group 1 as a
considerable amount of marginal bone loss (mean
9 ram) was observed in one individual in the latter
group. However, since these four implants are clini-
cally stable, they are not presented as implant loss in
the life table analysis. The crevicular probing depth
presented higher mean values in Group 2 and Group 3
than in Group 1. The deeper pockets in Group 2 and
3 may be explained by the missing attached gingiva
242 Journal of Cranio-MaxillofacialSurgery

Fig. 8 - The implants were removedand segmental resectionof the mandible was carried out.

Fig. 9 - Orthopantomogramof a patient of Group 3: bone graftingwith a free iliac crest graft 1 year after tumour resection.Four IMZ
implants placed in a secondstage procedure 8 months later.

compared with Group 1, in which the implants are
surrounded by local mucosa. The results of gingival
index and sulcus bleeding index, indicating the degree
of marginal infection, demonstrated inflammation of
the pen-implant soft tissue in the majority of the
implants. Therefore the median values are relatively
high in all groups. On the one hand, the patients'
bad oral hygiene is responsible for the marginal
infection, on the other hand it is known that chronic
soft tissue complications are relatively common in
implants placed in transplanted soft tissue. (Mitchell
et al., 1990). All of the above mentioned parameters
may only be regarded as tendencies in our small
groups of patients, none of these values showed
statistically significant differences between the three
groups.
In all groups, removable implant-supported pros-
theses have been used, since these give patients the
ability to remove their prostheses to maintain meticu-
lous hygiene, which is considered to be critical for
long-term success (Sclaroffet al., 1994). Nevertheless,
as presented above, good oral hygiene was seldom
achieved in our patients. Moreover, in some patients
masticatory function could not be reconstructed by
incorporating an implant-supported denture pros-
thesis (Table3). The impaired tongue function

Fig. 10 - Intraoral view of the patient after a further 8 months and the placement of a rigid bar screw--retained by the implants.
Fig. 11 - Intraoral view of the prosthetic restoration.
(Komisar et al., 1990) and an extreme postirradiation
xerostomia was to a certain degree responsible for
this problem. However, depending on the patient's
compliance and efforts, the masticatory function
gradually improved with time and the aesthetic
improvement was a benefit to all the patients (Figs.
9, 10, 11).
References
Albrektsson, T.: A multicenter report on osseointegrated oral
implants. J. Prosthet. Dent. 60 (1988) 337-340
Beumer, J., R. Harrison, B. Sanders, M. Kurrasch: Postradiation
dental extractions: a review of literature and a report of 72
episodes. Head Neck Surg. 6 (1982) 581-586
Donovan, M. G., N. C. Dickerson, L. J.. Hanson, R. B. Gustafson:
Maxillary and mandibular reconstruction using calvarial bone
Endosteal implants in the irradiated lower jaw 243
grafts and Branemark implants: a preliminary report. J. Oral
Maxillofac. Surg. 52 (1994) 588-594
Dumbach, J., H. Rodemer, W.. J. Spitzer, E. Bender: Grenzen der
knoechernen Rekonstruktion des Unterkiefers mit autogener
Spongiosa, Hydroxylapatitgranulat und Titangitter,
insbesondere nach Strahlentherapie. Fortschr. Kiefer
Gesichtschir. 39 (1994) 93-95
Epstein, J. B., F L. Wong, P. Stevenson Moore:
Osteoradionecrosis: Clinical experience and a proposal for
classification. J. Oral Maxillofac. Surg. 45 (1987) 104-110
Esser, E.: Enossale Implantate im radiogen belasteten Kiefer nach
Tumorradikaloperation. Z. Zahnaerztl. Implantol. 10 (1994)
167-170
Fischer-Brandies, E.: Das Risiko enossaler Implantationen nach
Radiatio. Quintessenz 5 (1990)873-876
Granstroem, G., M. Jacobsson, A. Tjellstroem: Titanium implants
in irradiated tissue: Benefits from hyperbaric oxygen. Int. J.
Oral Maxillofac. Implants 7 (1992) 15-25
Kaplan, E. L., P. Meier: Non-parametric estimation from
incomplete observations. J. Am. Statistic. Assoc. 53 (1958)
457-65
244 Journal of Cranio-Maxillofacial Surgery
Komisar, A.: The functional result of mandible reconstruction.
Laryngoscope 100 (1990) 364-374
Larsen, P. E., M. J. Stronczek, F. M. Beck, M. Rohrer:
Osteointegration of implants in radiated bone with and
without hyperbaric oxygen. J. Oral Maxillofae. Surg. 51
(1993) 280-287
Loe H., J. Silness: Periodontal disease in pregnancy. I. Prevalence
and severity. Acta Odontol. Scand. 22 (1963) 533-36
Martin, I. C., J. I. Cawood, E. D. Vaughan, N. Barnard."
Endosseous implants in the irradiated composite radial
forearm free flap. Int. J. Oral Maxillofac. Surg. 21 (5) (1992)
266-270
Marx, R. E.: Osteoradionecrosis: A new concept of its
pathophysiology. Int. J. Oral Maxillofac. Surg. 41 (1983)
283-288
Marx, R. E., R. P. Johnson." Studies in the radiobiology of
osteoradionecrosis and their clinical significance. Oral Surg.
Oral Med. Oral Pathol. 64 (1987) 379-390
MeKenzie, M. R., F. L. Wong, J. B. Epstein, M. Lepawsky:
Hyperbaric oxygen and postradiation osteonecrosis of the
mandible. Oral Oncol. Eur. J. Cancer 29B (1993) 201-207
Mitchell, D., S. Synnott, J. Van Dercreek: Tissue reaction involving
an intraoral skin graft and CP titanium abutments: a clinical
report. Int. J. Oral Maxillofac. Implants 5 (1990) 79-84
Moseoso, J. F., J. Keller, E. Genden, H. Weinberg, H. F. Biller,
D. Buchbinder, M. L. Urken: Vascularized bone flaps in
oromandibular reconstruction. A comparative study of bone
stock from various donor sites to assess suitability for
enosseous dental implants. Arch. Otolaryngol. Head Neck
Surg. 120 (1994) 36-43
Muehlemann, H. R.: Patientenmotivation mit individuellem
Intensivprogramm fuer orale Gesundheit. In: Peters, S., (ed),
Prophylaxe. Berlin: Quintessenz 1978, 137-49
Murray, C. G., J. Herson, Z E. Daly, S. Zimmermann: Radiation
necrosis of the mandible: A 10-year study. I. Factors
influencing the onset of necrosis. Int. J. Radiat. Oncol. Biol.
Phys. 6 (1980) 543-548
Parel, S., A. Tjellstr6m: The United States and Swedish experience
with osseointegration and facial prosthesis. Int. J. Oral
Maxillofac. Implants 6 (1991) 75-79
Riediger, D.: Restoration of masticatory function by
microsurgically revascularized iliac crest bone grafts using
enosseous implants. Plast. Reconstr. Surg. 81 (1987) 861-876
Selaroff, A., B. Haughey, W. D. Gay, R. Paniello: Immediate
mandibular reconstruction and placement of dental implants.
At the time of ablative surgery. Oral Surg. Oral Med. Oral
Pathol. 78 (1994) 711-717
Taylor, T. D., P. Worthington: Osseointegrated implant
rehabilitation of the previously irradiated mandible: Results
of a limited trial at 3 7 years. J. Prosthet. Dent. 69 (1993)
60-69
Urken, M. L., D. Buchbinder, H. Weinberg, C. Viekery, A. Sheiner,
R. Parker, J. Schaefer, P. Som, A. Shapiro, W. Lawson:
Functional evaluation following microvascular
oromandibular reconstruction of the oral cancer patient: A
comparative study of reconstructed and nonreconstructed
patients. Laryngoscope 101 ( 1991) 935-950
Waeehter, R., P. Stoll: Moeglichkeiten und Grenzen enossaler
Implantate bei der oralen Rehabilitation yon Tumorpatienten
nach Bestrahlung. Z. Zahnaerztl. Implantol. 10 (1994)
171-176
Zlotow, I. M., Y. M. Huryn, Y. D. Piro, E. Lenchewski, D. A.
Hidalgo: Osseointegrated implants and functional prosthetic
rehabilitation in microvascular fibula free flap reconstructed
mandibles. Am. J. Surg. 164 (6) (1992) 677-681
Dr Franz Watzinger
Universit~tsklinik ft~r Kiefer-und Gesichtschirurgie
Neue Universit~tskliniken Wien
W~thringer Gt~rtel 18-20
A - 1090 Wien
Austria
Paper received 15 December 1995
Accepted 20 May 1996