Int. J. Oral Maxillofac. Surg. 2007; 36: 884–889
doi:10.1016/j.ijom.2007.06.004, available online at http://www.sciencedirect.com
Leading Clinical Paper
Oral Surgery
Lingual nerve injury in third
molar surgery
I. Observations on recovery of
sensation with spontaneous
healing
S. Hillerup1,2, K. Stoltze1,2
1
Department of Oral and Maxillofacial
Surgery, Rigshospitalet, Blegdamsvej 9, DK2100 Copenhagen Ø, Denmark; 2Department
of Periodontology, Dental School, Faculty of
Health Sciences, University of Copenhagen,
Denmark
S. Hillerup, K. Stoltze: Lingual nerve injury in third molar surgery. Int. J. Oral
Maxillofac. Surg. 2007; 36: 884–889. # 2007 International Association of Oral and
Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Abstract. The aim of this study was to investigate the healing potential of damaged
lingual nerves with some remaining function at least 3 months post injury. Fortysix patients were monitored at different time intervals after injury. A simple
neurosensory examination included the perception of tactile, thermal stimuli and
location of stimulus, as well as two-point discrimination, pain and the presence of a
neuroma at the lesion site. Neurogenic signs and symptoms related to the injury and
their variation over time were registered. Females were more often referred than
males. Most lingual nerve injuries exhibited a significant potential for recovery, but
only a few patients made a full recovery with absence of neurogenic symptoms.
The recovery rate was highest during the first 6 months. Recovery was not
influenced by gender, and only slightly by age. The presence of a neuroma was
associated with a more severe injury. Patients should be monitored repeatedly for at
least 3 months, and not operated on until neurosensory function no longer
improves, and is less than what might be rendered by microsurgical repair. Through
proper training and mastery of the surgical approach, every effort should be
focused on sparing the lingual nerve, considering its proximity to the field of
surgery.
The tongue is an important and sensitive
anatomical structure that serves a range of
vital functions. Unintended iatrogenic
injury to the lingual nerve (LN) may happen during third molar surgery due to its
0901-5027/100884 + 06 $30.00/0
anatomical proximity, separated from the
cortex of the third molar region only by the
periosteum15,16. Some LN injuries cause
temporary sensory disturbances but a fraction of cases fail to resolve and result in
Key words: lingual nerve; nerve injury; recovery; third molar surgery.
Accepted for publication 28 June 2007
Available online 4 September 2007
permanent neurosensory disability, loss of
sensory function and neurogenic symptoms7.
The incidence of LN injury varies and
depends on a number of factors: the
# 2007 International Association of Oral and Maxillofacial Surgeons. Published by Elsevier Ltd. All rights reserved.
Lingual nerve injury in third molar surgery
experience of the surgeon21, difficulty of
the case, depth of impaction, presence of
overhanging ramus bone, lingual flap elevation, and operating time14, surgical
approach19 (lingual split bone technique)
and the mere focus on registration and
documentation of such injury1,20,26.
Rates of temporary effects on the LN
after third molar surgery have been
reported in the surprisingly high order of
magnitude of 15%, and permanent
damage may occur in 0.3–0.6%3,4,14.
MASON14 demonstrated anatomical factors
(depth of impaction, distal overhanging
bone, state of eruption and angulation of
tooth) and surgical factors (lingual flap
elevation, bone removal, lingual plate
splitting and operation time), all associated with a significantly increased incidence of disturbance of the LN.
Management of LN injury is a challenge to the oral and maxillofacial surgeon, and decision making in terms of
therapeutic action, micro-neurosurgical
repair versus wait and see, must be based
on evidence-based criteria6. These
include considerations related to the outcome of neurosensory examination and
timing, since an injured nerve may
recover with some regained function
within a certain time limit.
The aims of the present study were to:
demonstrate the potential for spontaneous neurosensory recovery in patients
that exhibited some nerve function
within the first 3 months after the injury
or later;
describe neurosensory malfunctions
associated with the injury and their
change over time;
investigate the possible influence of
age, gender and the presence of a neuroma on neurosensory recovery.
Patients and methods
Patients with LN injury meeting the criteria given below were drawn from a
database of 449 injuries to oral branches
of the trigeminal nerve collected consecutively during the period 1987–2005. Of
these, 261 were lingual nerve injuries of
various etiologies10.
Criterion for inclusion: patients with
iatrogenic injury to the LN nerve caused
by third molar surgery and with some
remaining sensory function at 3 months
after injury or later, depending on time of
referral. Patients seen less than 12 months
after the injury were offered one or more reexaminations. Only patients with a course
of follow up were included, n = 46. Criterion for exclusion: neurological disease,
known alcoholism, patients with bilateral
injuries and patients who had received
reconstructive micro-neurosurgery.
Follow-up examinations were intended
at 3, 6 and 12 months post injury or later.
The course of follow up was on average
7.4 months (SD = 4.0, range 2–17
months).
Neurosensory evaluation
Patient records included date and mode of
injury, an interview addressing the
patients’ subjective assessment of reduced
sensory function of the injured LN, and
neurogenic malfunctions (paraesthesia,
etc.). A simple neurosensory examination
was carried out as described previously11,12. Details of the examination
protocol have been presented in a recent
article10. Follow-up examinations were
performed with the examiner blinded to
the results of preceding examination(s).
Tactile perception of the following stimuli was assessed: (1) feather light touch
(by extruded filaments of a cotton stick),
(2) pin prick (point of dental probe), (3)
point/dull discrimination (point of dental
probe versus blunt touch with the tip of the
probe handle), (4) warmth (touch of blunt
instrument heated to 45–50 8C), (5) cold
(touch of blunt instrument cooled to 0–
20 8C), (6) point location (touch of blunt
instrument), (7) brush stroke direction
(blunt instrument moved over area to be
examined). The perception of stimuli 1–7
was rated according to a simple scale
ranging from 0 to 3: 0 = no perception
of touch, 1 = perception of touch with
no ability to differentiate (pointed/blunt,
warm/cold, localization of touch, direction
of moving touch), 2 = perception with
ability to differentiate less clear than normal, and 3 = normal perception10,12. The
level of overall neurosensory function was
characterized through the sum of perception ratings (1–7) that might range from 0
to 21: sum score 0 signifying a total loss of
nerve conductivity and sum score 21
denoting normal neurosensory function
of the nerve in question.
Two-point discrimination thresholds
(2PD) were set to 5, 10, 15, 20 mm (8).
Pain perception on pinching with a tissue
forceps was rated as present or absent (9).
An unpleasant, irradiating sensation in the
injured side of the tongue, evoked by
digital pressure to the region of suspected
injury at the medial aspect of the mandibular ramus, was interpreted as being
caused by a traumatic neuroma. The pattern and distribution of fungiform papillae
were assessed with the uninjured side as
control.
885
Patients with impaired LN function
were informed on the potential of
improvement of perception rendered by
‘sensory re-education’. They were urged
to practice targeted exercises in order to
obtain a central adaptation to a changed
pattern of afferent neurosensory input22,
thus utilizing the plasticity of the central
nervous system.
Nerve injuries causing signs and symptoms, reduced function or neurogenic malfunction more than 12 months after the
injury were considered permanent.
Statistics
Differences between categorical scores
were tested with a ‘sign test’. When appropriate, variables were described through
mean, standard deviation (SD), range or
median values. Chi-square or Kruskal–
Wallis tests were applied to test differences
between distributions. Level of significance: 5%. The software used were the
EPI6 program for DOS and SPSS for Windows version 13, and graphics were produced with the help of the SPSS and
Microsoft Office program packages.
Results
Demography
A significant over-representation of
referred female patients was found, F/M
ratio among the 46 patients being 33/13
(72%/28%), P < 0.001. The patients’
mean age at time of initial examination
was 29 years (SD = 8.9, range 15–53
years) with no difference between gender
or side of nerve injury.
The average time course from injury to
initial examination was 4.5 months (range
0–10 months). Median time between
injury and final examination was 12
months (range 3–24 months). The course
of healing from initial to final follow-up
examination was monitored in all 46
patients. Of these, 19 patients were also
examined between the initial and final
examinations.
Subjective signs and symptoms
The patients’ subjective rating of sensory
function in the injured side of the tongue at
the initial examination was classified as
anaesthesia (n = 9), hypoaesthesia (n =
36) and subjective normal sensory function
in spite of objective deficit (n = 1). At the
final examination the ratings were anaesthesia (n = 1), hypoaesthesia (n = 25) and
normal sensation (n = 14). Data for comparison were missing in six patients.
886
Hillerup and Stoltze
The patients’ initial ratings of their
subjective sensory status averaged score
1.2 according to the rating scale, range
0–3. The sum score rating at the final
examination was 1.9. Twenty-seven
patients’ scores indicated an improvement
in subjective sensory scores (recovery), 11
were unchanged, and eight patients
reported a worsening. The recovery, felt
and rated by the patients, is statistically
significant (P < 0.01).
Paraesthesia (n = 29) was the most prevalent neurogenic symptom followed by
dysaesthesia (n = 5, unpleasant, painful,
burning sensation). Eight patients with
initial paraesthesia recovered to normal.
Other neurogenic disturbances at the final
examination included taste abnormality
(n = 15) that might be an inability to
recognize taste, typically salt with incapability to adjust while cooking, thermal
hypersensitivity (n = 7), speech difficulty
(n = 5), pain on tooth brushing (n = 2), and
tongue bite (n = 1). Nine out of 11 patients
(82%) with full sensory recovery (sum
score 21) still had paraesthesia.
Neurosensory ratings
Pain perception, as an extreme of neurosensory perception, was initially absent in
four patients and finally in only one individual.
Two-point discrimination thresholds
improved considerably during follow
up. Initially 12 patients exhibited a
2PD >20 mm. The average threshold value
in the remaining patients was 9.7 mm (SD
3.6). For comparison, the 2PD in the uninjured side was 6.3 mm (SD 2.3). At the final
examination six patients scored a 2PD
threshold >20 mm, and the mean threshold
value of the remaining patients was 8.5 (SD
4.3). An average 1.4 mm improvement in
2PD thresholds over time was found
(P < 0.05).
All tested tactile and thermal qualities
(1–7) showed a significant reduction of
function compared to the uninjured side at
the initial examination. A significant
recovery took place over time in terms
of improved ratings of the sensory qualities tested (Fig. 1). Improvement was
registered in 40 patients, two were
unchanged, and in four patients a
worsening in sensory perception was
recorded (Fig. 2). Recovery was registered
mainly during the first 12 months after
injury, with the highest rate of sensory
regain during the first 6 months, and less
improvement beyond 12 months (Fig. 3).
Eleven patients (24%) reached normal
levels of sensory perception (sum score
21). In the remaining group of patients, 29
Fig. 1. Sensory recovery of tactile and thermal perception, location, and overall subjective
perception in patients with LN injury after third molar surgery. Average follow up = 8 months.
Paired observations, n = 46.
(63%) showed an increase in sum score
ranging from 1 to 19, two patients (4%)
remained unchanged and four (9%) got
worse during follow up.
No difference in severity of injury or
recovery was found related to gender. The
final ratings of neurosensory function
showed a negative correlation with age
(P < 0.01; Pearson, r = 0.42) (Fig. 4).
In terms of recovery, expressed as the
difference between final and initial sum
score, there was no statistically significant
association with age within the range
given.
Digital pressure to the region of suspected injury at the medial aspect of the
mandibular ramus at the initial examination
evoked an irradiating abnormal sensation
towards the tip of the tongue in 19 patients
(53%), a reaction that was interpreted as
produced by a traumatic neuroma. No data
were obtained in 10 patients, and no such
Fig. 2. Sensory recovery of LN after injury caused by third molar surgery in 46 patients.
Recovery is expressed in terms of difference between initial and final score ratings of tactile,
thermal and locational perception (sum score difference). Average follow up = 8 months.
Lingual nerve injury in third molar surgery
Fig. 3. Sensory recovery of LN after injury caused by third molar surgery in 46 patients. Sum
scores denote the added ratings of perception of seven qualities of stimulus. Rate of recovery is
fastest during first 6 months, thereafter fading.
reaction was experienced in the healthy
side. This feature remained essentially
unchanged, since 20 patients (53%) exhibited a similar reaction at the final examination (no data in eight patients). The
neurosensory function was more severely
affected in patients displaying a traumatic
neuroma than in patients without clinical
signs of a neuroma at the final examination
(P < 0.01; Table 1). The recovery in terms
of sensory improvement expressed as the
difference in sum scores between final and
initial examinations in 18 patients without a
neuroma averaged a median value of 5.5
(range 1–21) versus 3.5 (range 8 to 15) in
20 patients with a neuroma. This difference
was not statistically significant.
The distribution of fungiform papillae
was registered at both initial and final
examination in 35 patients. In 14 patients,
ratings were identical at the two examinations (9 with fewer fungiform papillae in
the injured side, 5 without difference
between sides), and 16 patients with a
reduced number of fungiform papillae
‘normalized’ during follow up. In five
patients with evenly distributed fungiform
papillae, the number of fungiform papillae
reduced in the injured side during the
course of follow up. No association was
found between the distribution of fungiform papillae and the presence of a neuroma or sensory recovery.
Discussion
Previous studies have focused on LN injuries and their occasional potential for
recovery3,8,14. As shown in the present
study, based on history and neurosensory
examination of each patient followed over
time, injured LNs have a considerable
potential for spontaneous recovery. A
nerve that appears anaesthetised shortly
Fig. 4. LN function after injury caused by third molar surgery correlated with age of patient.
Negative influence of age on final nerve function, r = 0.42, P < 0.01 (Pearson correlation).
Sum score stands for added ratings of perception of seven qualities of stimulus.
887
after injury may recover to full function
depending on the nature of injury and
probably the age of the patient. Conversely, follow-up examinations may show a
nerve to be injured to a degree that can
only be improved by micro-neurosurgical
repair.
Almost all LN injuries associated with
third molar surgery are ‘closed injuries’.
This means that the surgeon is not aware
of the lesion at the time of surgery, and is
unable to diagnose its true nature of origin
(compression, traction, partial or total
severance) or to classify the lesion
according to SUNDERLAND23 at the first
postoperative examination. The only
way to acquire a clue to the nature of a
LN lesion, and to make an evidence-based
decision on how to handle the injury, is
follow up with repeated and standardised
neurosensory examinations6,18. As a general rule, injured LNs in a process of
recovery, subjective or objective (sum
score improvement or similar), should
be monitored according to a follow-up
protocol. Micro-neurosurgical repair
must be considered in nerves with a
persistent, total or near total loss of function beyond 3–6 months, and where the
potential benefit of repair will justify
microsurgery.
The recovery observed in the present
study is characteristic for Sunderland
grade 2 and 3 lesions, and it may be that
the few injuries with poor recovery
belonged to grade 423. The rate of recovery proved fastest during the initial 6
months after injury. Later recovery
was slower but still possible, and the
distinction between recovery of function
rendered by healing and improvement due
to sensory re-education is difficult. The
mode of recovery was in contrast to nerve
lesions associated with the injection of
local analgesia, where no systematic trend
for improvement of sensory functions was
found12. This may explain the difference
between mechanical nerve injury and the
probable neurotoxic damage rendered by
certain local analgesics that may mirror
Sunderland grade 4 lesions5.
Whereas injury to the inferior alveolar
nerve may be unavoidable and constitute a
calculated risk to be accepted, injuries to
the LN can and should be prevented by
adaptation of the surgical procedure to the
regional anatomy. Surgeons have nothing
to do ‘on the wrong side of the periosteum’. As stated by BLACKBURN2: ‘The
lesson to be learnt is quite simple, never
let the bur enter the tissues on the lingual
side of the mandible, whether there is
a lingual flap retractor/guard in position
or not’.
888
Hillerup and Stoltze
Table 1. Association of the presence of a neuroma and neurosensory function (sum score) in
patients with LN lesions after third molar surgery
Traumatic neuroma
Median of sum score (min–max)
P*
14.5 (3–21)
20.0 (14–21)
<0.001
Present (n = 20)
Absent (n = 18)
At final examination after an average course of follow up of 7.4 months (SD = 4.0, range 2–17
months).
*
Kruskal–Wallis test for two groups (equivalent to Chi square).
If the unhappy event of an LN injury
should occur, the surgeon must deal with
the situation in a correct manner. It is
almost impossible to establish a prognosis based on only one examination of
a nerve injury patient shortly after the
damage. The prognosis of a LN lesion
may be assessed at 3–6 months after the
injury, but still with some uncertainty. If
the patient has at least pain perception at
this stage, there is hope for further recovery. As seen in Fig. 3, even patients seen
for initial examination more than 6
months after the injury showed some
regeneration of function. One patient
scheduled for surgery on the admission
day, 7 months post injury, reported a
recent improvement of sensation, and
surgery was cancelled. She recovered
to a level comparable to that which
may be obtained with reasonably successful microsurgical repair. Generally,
a patient with persistent total or near total
loss of sensation in the affected LN well
beyond 6 months will have a lesser
chance of spontaneous recovery to a
function as good as or better than can
be obtained through reconstructive
micro-neurosurgery17.
Nerve injuries, including LN injuries,
are frequent objects of litigation or malpractice suits9,13,24, and the importance
of written informed consent to the risk
has been emphasized13. Court decisions
have varied. HANDSCHEL et al.9 stated: ‘In
some former decisions only the fact of
the damage was equivalent to a lack of
care, while recently other courts had the
opinion that an injury of the lingual
nerve could also be caused in spite of
careful treatment’. Considering the fact
that LN injuries in third molar surgery
are avoidable, informed consent to such
a complication is of questionable ethical
value, and focus has to be directed
towards adequacy of training in mastering the correct atraumatic surgical
approach.
The presence of a neuroma indicates
fascicular severance with subsequent
axonal outgrowth and Schwann cell proliferation, and may be conceived as a
feature of the more severe lesions. Cases
with a traumatic neuroma are apparently
associated with a poorer prognosis than
cases with absence of a clinically recognizable neuroma (Table 1). Notably, some
patients got worse during the course of
follow up. Apart from minor fluctuations
ascribable to inaccuracies in the rating of
sensory functions, a neuroma in continuity in Sunderland grade 3 lesions23 might
constrict intact nerve fibres, establish an
intra-fascicular compartment syndrome,
or otherwise present an obstacle for nerve
conductivity22.
LN injury may severely compromise
the well-being and quality of life of
patients in a number of aspects, including
communication, chewing ability and
enjoyment of food and drinks. LN injury
should be avoided, and merely focussing
on the problem is helpful25.
References
1. Blackburn CW. A method of assessment in cases of lingual nerve injury. Br
J Oral Maxillofac Surg 1990: 28: 238–
245.
2. Blackburn CW. Experiences in lingual
nerve repair. Br J Oral Maxillofac Surg
1992: 30: 72–77.
3. Blackburn CW, Bramley PA. Lingual
nerve damage associated with the
removal of lower third molars. Br Dent
J 1989: 167: 103–107.
4. Carmichael FA, McGowan DA. Incidence of nerve damage following third
molar removal: a West of Scotland Oral
Surgery Research Group Study. Br J Oral
Maxillofac Surg 1992: 30: 78–82.
5. Cornelius CP, Roser M, Wiethölter
H, Wolburg H. Nerve injection injuries
due to local anaesthetics. Experimental
work. J Craniomaxillofac Surg 2000:
28(Suppl. 3):134–135.
6. Dodson TB, Kaban LB. Recommendations for management of trigeminal nerve
defects based on a critical appraisal of the
literature. J Oral Maxillofac Surg 1997:
55: 1380–1386.
7. Fielding AF, Rachiele DP, Frazier G.
Lingual nerve paresthesia following third
molar surgery. A retrospective clinical
study. Oral Surg 1997: 84: 345–348.
8. Gulicher D, Gerlach KL. Sensory
impairment of the lingual and inferior
alveolar nerves following removal of
impacted mandibular third molars. Int J
Oral Maxillofac Surg 2001: 30: 306–312.
9. Handschel J, Figgener L, Joos U. Die
forensische Bewertung von Verletzungen
der Nerven und des Kieferknochens nach
Weisheitszahnentfernungen im Blickwinkel der aktuellen Rechtsprechung. Mund
Kiefer Gesichtschir 2001: 5: 44–48.
10. Hillerup S. Iatrogenic injury to oral
branches of the trigeminal nerve: records
of 449 cases. Clin Oral Investig 2007: 11:
133–142.
11. Hillerup S, Hjørting-Hansen E, Reumert T. Repair of the lingual nerve after
iatrogenic injury. A follow-up study of
return of sensation and taste. J Oral Maxillofac Surg 1994: 52: 1028–1031.
12. Hillerup S, Jensen R. Nerve injury
caused by mandibular block analgesia.
Int J Oral Maxillofac Surg 2006: 35:
437–443.
13. Lydiatt DD. Litigation and the lingual
nerve. J Oral Maxillofac Surg 2003: 61:
197–200.
14. Mason DA. Lingual nerve damage following lower third molar surgery. Int J
Oral Maxillofac Surg 1988: 17: 290–
294.
15. Miloro M, Halkias LE, Slone WH,
Chakeres DW. Assessment of the lingual nerve in the third molar region using
magnetic resonance imaging. J Oral Maxillofac Surg 1997: 55: 134–137.
16. Pogrel MA, Renaut A, Schmidt B,
Ammar A. The relationship of the lingual
nerve to the mandibular third molar
region: an anatomic study. J Oral Maxillofac Surg 1995: 53: 1178–1181.
17. Robinson PP, Loescher AR, Smith KG.
A prospective, quantitative study on the
clinical outcome of lingual nerve repair.
Br J Oral Maxillofac Surg 2000: 38: 255–
263.
18. Robinson PP, Loescher AR, Yates JM,
Smith KG. Current management of
damage to the inferior alveolar and lingual nerves as a result of removal of third
molars. Br J Oral Maxillofac Surg 2004:
42: 285–292.
19. Robinson PP, Smith KG. Lingual nerve
damage during lower third molar
removal: a comparison of two surgical
methods. Br Dent J 1996: 180: 456–461.
20. Robinson PP, Smith KG, Johnson FP,
Coppins DA. Equipment and methods for
simple sensory testing. Br J Oral Maxillofac Surg 1992: 30: 387–389.
21. Sisk AL, Hammer WB, Shelton DW,
Joy ED. Complications following
removal of impacted third molars. J Oral
Maxillofac Surg 1986: 44: 855–859.
22. Sunderland S. Nerve Injuries and their
Repair. A Critical Appraisal. Edinburgh:
Churchill Livingstone 1991.
23. Sunderland S. A classification of peripheral nerve injuries producing loss of
function. Brain 1951: 74: 491–516.
24. Venta I, Lindqvist C, Ylipaavalniemi
P. Malpractice claims for permanent
nerve injuries related to third molar
removals. Acta Odont Scand 1998: 56:
193–196.
Lingual nerve injury in third molar surgery
25. Walters H. Reducing lingual nerve
damage in third molar surgery: a clinical
audit of 1350 cases. Br Dent J 1995: 25:
140–144.
26. Zuniga JR, Meyer RA, Gregg JM,
Miloro M, Davis LF. The accuracy of
clinical neurosensory testing for nerve
injury diagnosis. J Oral Maxillofac Surg
1998: 56: 2–8.
Address:
Søren Hillerup
Department of Oral and
Maxillofacial Surgery
Rigshospitalet
Blegdamsvej 9
DK-2100 Copenhagen Ø
Denmark
Tel: +45 35458399
Fax: +45 35452364
E-mail: sohi@rh.dk
889