Results and Discussion
Unilateral Facial Paralysis
Dynamic Reconstruction
The type of paralysis, objective measurements, the personal patient's smile, and the
gender are key concepts to be considered before scheduling a dynamic facial paralysis
reconstruction.
Objective measurements: The quantification of the facial paralysis is one of the most important aspects
when examining patients with facial palsy. Assessing the severity of the paralysis
is useful not only for doing a precise diagnosis, but also when deciding a surgical
procedure. Subjective methods of evaluation have been described for assessing various
facial nerve disorders.[3]
[4]
[5]
[6] These methods have historically been based on scale grades to rank the severity
of the damage. In 2015, a systematic review identified 19 facial nerve grading scales.
After comparison, only the Sunnybrook Facial Grading Scale satisfied all criteria
for an ideal facial nerve grading instrument.[7] However, these subjective instruments poorly characterize dynamic facial reanimation
techniques since the surgeon needs to accurately quantify the damage to know how much
should be restored.
In 2008, we introduced the FACIAL CLIMA, an automatic, quantitative, operator-independent,
objective method that allows dynamic quantification of facial movements.[8] The FACIAL CLIMA is used both for preoperative planning and for assessing postoperative
outcomes. Using an objective analysis, it is possible to evaluate the degree of symmetry
obtained by comparing the reanimated and the normal sides. This is an important aspect
to consider when comparing different techniques of facial reanimation between different
individuals' smiles. Although the FACIAL CLIMA was described 13 years ago, it is still
one of the most precise systems to quantify the facial movements when compared with
others, as described by Gerós et al in 2016.[9]
Type of paralysis: Determining the type of paralysis is imperative before deciding a treatment. In
this sense, we divide facial paralysis in to complete or incomplete, also known as
flaccid and nonflaccid facial paralyses. This distinction is important due to the
timing, because in complete facial palsies the facial musculature does not receive
neuronal input, the mimic muscles will be irreversibly atrophied if this time exceed
2 years,[10] and a new muscle unit must be incorporated to restore motion.
Conversely, in an incomplete facial paralysis, the time of onset does not influence
because the mimic muscles are receiving axonal input which avoids atrophy. The same
techniques could be proposed for incomplete and subacute complete (3 months to 2 years)
facial paralyses, since the mimic muscles are still viable. Moreover, our experience
is that female patients with facial paralysis of up to 4 years can benefit from these
techniques, avoiding a muscle transfer. On the other hand, the acute complete (less
than 3 months) facial paralysis could be restored by direct nerve repair or providing
a new nerve source as the CNFG or the masseteric nerve.[11] In our series, when comparing the use of both nerves (n = 17 and 26, respectively), we observed that the masseteric nerve achieves more symmetry
and commissural displacement and satisfaction of spontaneity (especially in women)
than the CFNG ([Tables 1]
[2]
[3]).
Table 1
Intragroup comparison of commissure displacement and commissure contraction velocity
comparing the CFNG and the masseteric transference
|
Mean ± SD CD (mm)
|
Mean ± SD CCV (mm/s)
|
|
Group
|
Healthy
|
Reanimated
|
p-Value
|
Healthy
|
Reanimated
|
p-Value
|
|
I (CFNG)
|
8.6 ± 2.8
|
6.2 ± 3.1
|
0.001
|
33.5 ± 13.2
|
23.1 ± 12.4
|
0.019
|
|
II (masseteric)
|
9.1 ± 3.1
|
7.9 ± 2.9
|
0.52
|
35.6 ± 12.9
|
32.0 ± 14.3
|
0.35
|
Abbreviations: CCV, commissure contraction velocity; CD, commissural displacement;
CFNG, cross-facial nerve graft; SD, standard deviation.
Table 2
Intergroup comparison of commissure displacement, commissure contraction velocity,
and percentage of recovery comparing the CFNG and the masseteric transference
|
Healthy side[a]
|
Reanimated side[a]
|
Recovery (%)[a]
|
|
Group
|
CD (mm)
|
CCV (mm/s)
|
CD (mm)
|
CCV (mm/s)
|
CD
|
CCV
|
|
I (CFNG)
|
8.6 ± 2.8
|
33.5 ± 13.2
|
6.2 ± 3.1
|
23.1 ± 12.4
|
64.2 ± 26.3
|
76.8 ± 29.2
|
|
II (masseteric)
|
9.1 ± 3.1
|
35.6 ± 12.9
|
7.9 ± 2.9
|
32.0 ± 14.3
|
91.9 ± 22.2
|
95.1 ± 35.6
|
|
p
|
0.61
|
0.46
|
0.046
|
0.22
|
0.039
|
0.43
|
Abbreviations: CCV, commissure contraction velocity; CD, commissural displacement;
CFNG, cross-facial nerve graft.
a Mean ± standard deviation.
Table 3
Intergroup comparison of satisfaction and spontaneity[a] comparing the CFNG and the masseteric transference
|
Group
|
SP:NSP ratio (SP%)
|
SS:NSS ratio (% SS)
|
|
I (CFNG)
|
8:2 (80.0%)
|
8:2 (80.0%)
|
|
II (masseteric)
|
16:2 (88.8%)
|
10:8 (55.5%)
|
|
p
|
0.60
|
0.25
|
Abbreviations: CFNG, cross-facial nerve graft; NSP, nonsatisfied patients; NSS, nonspontaneous
smile; SP, satisfied patients; SS, spontaneous smile.
a Fisher's exact test shows no significant differences in satisfaction and spontaneity
between the groups.
Patient's smile: Before indicating a surgical facial reanimation, the normal side smile should be
thoroughly analyzed. The two most important factors that characterize the smile are
the vector and the strength of pull. Both factors show a great interpersonal difference
as established by Rubin, who classified the smile in three anatomical types.[12] However, he concluded that all types of smile overlap in a greater or a lesser degree,
and consequently, infinite types of smile can be found.
Vector of the smile: Although we identify it by the FACIAL CLIMA system, the vector of the smile can
easily be identified on photographs of the healthy side at rest and at maximal commissure
displacement. By superimposing both images, the resultant vector of oral commissural
movement above the horizontal can be used to reconstruct the paralyzed side ([Fig. 1]).
Fig. 1 Vector of the smile obtained after superimposing images of the oral commissure at
rest and at maximal displacement. (A) Oral commissure at rest. (A′) Maximal commissural displacement. α: angle of the vector of the smile.
This vector is very important for the muscle orientation during flap inset when reanimating
long-term complete facial paralysis. The flap should be transferred to the face and
fixed to the modiolus and zygomatic arch using the same vector that is in the healthy
side. However, the vector is not relevant in complete short-term or incomplete facial
paralysis because the facial musculature is not atrophied yet, and consequently the
patient's vector smile is preserved.
Strength of the smile: Smiles are different between individuals regarding the commissural displacement.
For example, a patient with a strong smile can have a higher value of commissural
displacement than someone with a “weak” smile, without meaning that the latter has
some form of paralysis. After retrospectively analyzing our sample and our results,
the cut-off point to determine if a smile is strong or weak has been established at
8 mm of commissural displacement. Thus, a patient with a healthy side commissural
displacement of less than 8 mm is considered to have a weak smile. Conversely, a patient
with a healthy side commissural displacement greater than or equal to 8 mm is considered
to have a strong smile. This simple smile classification has enabled us to indicate
the techniques that achieve less commissural displacement for the first group of patients,
and the most powerful techniques for the latter. Once we introduced this classification
in our practice, we have observed a greater improvement in the symmetry of our patients'
face after performing smile reconstruction. For this reason, the accurate quantitative
preoperative analysis is very important before deciding a surgical technique for facial
paralysis reanimation.
Gender: Numerous publications have shown that in different circumstances, women smile more
than men.[13]
[14] Thus, it is reasonable to think that the motivational drive toward smiling is greater
in women and therefore it is very likely that after smile reanimation, female population
develops brain plasticity earlier than men on the basis that smiling is more important
for them.
Several studies have shown a dichotomy between the sexes in nerve regeneration and
brain plasticity. Regarding the former, evidence that supports the fact that women
recover from neural injury faster than men is abundant.[15]
[16]
[17]
[18] Over the years, we have seen that women defend better from nerve injury and thus
are more resistant than men to denervation and muscle atrophy, maybe influenced by
sex hormones. These observations have led us to perform techniques indicated for palsy
of short duration in women with longstanding disease achieving good functional and
aesthetic results.[19]
[20] This factor is especially important in those cases which are between short and longstanding
facial paralyses. In our series, comparison between genders of movement recovery in
patients with CFNG performed with more than 2 years of evolution and of movement dissociation
in patients reanimated using the masseteric nerve shows that women's recovery and
the possibility of dissociation of smile are better than in men ([Table 4]).
Table 4
Comparison between genders of movement recovery in patients with CFNG performed with
more than 2 years of evolution and of movement dissociation in patients reanimated
using the masseteric nerve
|
Outcome
|
Female, n (%)
|
Male, n (%)
|
p-Value
|
|
Movement restoration
|
|
• CFNG > 2 y
|
28 (80)
|
0 (0)
|
0.02
|
|
• HNT > 2 y
|
19 (100)
|
2 (33.3)
|
0.04
|
|
Movement dissociation
|
|
• Gracilis to masseteric
|
31 (73.3)
|
13 (25)
|
0.02
|
Abbreviations: CFNG, cross-face nerve graft; HNT, hypoglossal nerve transposition.
Note: Fisher's exact test shows significantly higher recovery and dissociation in
females.
Regarding brain plasticity, our studies have shown that female patients could develop
an earlier cortical adaptation than male patients after smile reconstruction with
nerves other than the facial, developing spontaneity and dissociation at a higher
rate and earlier than male patients.[20]
[21] Although we are unable to predict which patients will achieve dissociation and spontaneity
after facial reanimation using nonfacial donor nerves, this must be taken into account
when deciding a technique to get both symmetry and potential development of spontaneity.
Surgical Techniques for Unilateral Facial Paralysis
Direct nerve repair: The most desirable scenario when reconstructing the facial nerve is a direct nerve
repair. This will only be possible if the time of onset is short (less than 3 months)
and the proximal facial nerve stump and the different distal branches are identifiable.
When planning a direct nerve repair, the length of the nerve gap and the vascularity
of the surrounding recipient bed should be taken into consideration. If the length
of the gap is less than 1 cm, the direct nerve repair between both ends could be an
option as long as there is no tension. If the gap is greater or the nerve coaptation
is under tension, nerve graft is a preferable option.
Our group most frequently employs the sural nerve as a donor autograft due to its
well-known advantages. Although the morbidity of the sural nerve is considered minimal,
it should be informed to the patients, an area of sensory loss of approximately the
size of 5 to 6 cm is to be expected, and 20 to 30% of the patients experienced minimal
levels of pain, cold sensitivity, functional impairment, and scar discomfort.[22]
After harvesting the nerve graft, it should be split in as many fascicules as the
number of sectioned branches that need reconstruction, each of which will be connected
to the transected facial nerve branches. On the other hand, the vascularization of
the surrounded tissue must be confirmed with the aim to safeguard the viability of
the nerve graft. In those cases where the sural nerve graft lays directly over a completely
avascular bed, for example, after a radiation therapy or aggressive surgical resections,
a superficial temporofascial flap should be harvested to wrap around the nerve circumferentially
without tension or compression and sutured to the surrounding tissues to avoid damage
from pulling, tearing, or retraction.[23] Regarding oncologic adjuvant therapies, our group has demonstrated that the administration
of brachytherapy during the immediate postoperative period followed by external beam
radiotherapy does not affect the functional outcomes of facial nerve repair with interpositional
grafts.[24] When the proximal stump is not available, a CFNG or a masseteric transposition can
be done.
Cross-face nerve grafting: The CFNG has been devised to help restore a smile that is apparently spontaneous
and coordinated with the contralateral side. In these techniques, motor axons from
the normal facial nerve are delivered to the contralateral paralyzed side. Despite
the main advantages of this technique, it also has some disadvantages. First, two
stages are required, although some authors have advocated a one-stage procedure. In
this sense, Frey et al recommend one-stage CFNG with end-to-side coaptation distally
on the recipient nerve of the paralyzed side, although the real advantages over traditional
techniques remain to be investigated.[25] Despite the aforementioned variation, we still prefer a two-stage approach to ensure
that the axonal load grows exclusively from the healthy side to the affected side,
especially important when reconstructing incomplete facial paralysis.[26] Once Tinel's sign is noted when tapping the paretic side, the second stage can be
performed.
Another disadvantage of CFNG is that it needs a donor nerve, which, as we have explained,
carries some morbidity. Furthermore, when a nerve graft is used, the axons must cross
two sites of coaptation, which may result in suboptimal reinnervation of the target
muscle.[27] However, this loss of axons could be beneficial when reconstructing weak smiles,
especially in the male population who develop spontaneity and dissociation at a lower
rate when using nonfacial donor nerves. For this reason, CFNG is our technique of
choice in these cases.
Masseteric-to-facial transposition: Masseteric-to-facial nerve transposition has gained increasing popularity in the
reanimation of both complete and incomplete facial paralyses[28]
[29]
[30]
[31]
[32] ([Fig. 2]; [Videos 1] and [2]).
Fig. 2 A 51-year-old woman with right facial paralysis secondary to varicella zoster infection.
Masseteric nerve transfer was performed for smile reanimation. Preoperative appearance
at rest (A) and while smiling (B). Postoperative situation after 2 years, at rest (C) and while smiling (D).
Video 1 Preoperative appearance of a 34-year-old man with incomplete right facial paralysis
after a partially recovered Bell's palsy.
Video 2 The same patient from [video 1] after facial reanimation by masseteric nerve transfer at 12 months after surgery.
The nerve of the masseter offers several advantages for facial reanimation. First,
it is a one-stage procedure, in contrast with CFNG. Second, its strength of pull allows,
on the one hand, reanimation of strong smiles and, on the other hand, acquisition
of very good symmetry at rest and when smiling. Third, the proximity between the masseteric
nerve and branches of the facial nerve eliminates the need for use a nerve graft,
and thus reduces the morbidity of the donor site and avoids the loss of power that
is attributed to the use of grafts when there are two coaptation sites. Lastly, the
possibility of developing spontaneity, as noted above, with higher rates and earlier
in women than in men.[19]
[20] Thus, when comparing the masseteric nerve versus hypoglossal nerve, we observed
better results in the masseteric group than in the hypoglossal group ([Tables 5] and [6]).
Table 5
Intergroup comparisons of age and time of evolution of paralysis comparing the masseteric
(group I) and the hypoglossal transference (group II)
|
Age (y)
|
Time (mo)
|
CD (mm)
(mean ± SD)
|
CCV (mm/s)
(mean ± SD)
|
Recovery (%)
(mean ± SD)
|
|
|
Reanimated
|
Normal
|
Reanimated
|
Normal
|
CD
|
CCV
|
Recovery (d)
|
|
Group I
|
49.5 ± 8.3
|
16.4 ± 5.3
|
7.7 ± 3.9
|
9.3 ± 3.9
|
35.3 ± 16.7
|
48.7 ± 19
|
83.2 ± 4.5
|
72.5 ± 17
|
136 ± 7.4
|
|
Group II
|
44.1 ± 7.6
|
17.1 ± 4.8
|
7.8 ± 3.1
|
8.5 ± 2.7
|
36.2 ± 10.6
|
45.4 ± 12.6
|
91.8 ± 5.1
|
79.7 ± 15.5
|
62 ± 4.6
|
|
p
|
0.52
|
0.61
|
0.54
|
0.41
|
0.5
|
0.39
|
0.11
|
0.19
|
0.013
|
Abbreviations: CCV, commissural contraction velocity; CD, commissure displacement;
SD, standard deviation.
Note: Evaluated parameters (i.e., CD and CCV), percentage of recovery, and time of
recovery are shown.
Table 6
Intragroup comparisons of CD and CCV comparing the masseteric and the hypoglossal
transference
|
CD (mm)
mean ± SD
|
CCV (mm/s)
mean ± SD
|
|
Healthy
|
Reconstructed
|
p-Value
|
Healthy
|
Reconstructed
|
p-Value
|
|
Group I
|
9.3 ± 3.9
|
7.7 ± 3.9
|
0.017
|
48.7 ± 19
|
35.3 ± 16.7
|
0.036
|
|
Group II
|
8.5 ± 2.7
|
7.8 ± 3.1
|
0.2
|
45.4 ± 12.6
|
36.2 ± 10.6
|
0.17
|
Abbreviations: CCV, commissural contraction velocity; CD, commissure displacement;
SD, standard deviation.
Note: Significant differences are observed in group I (masseteric) for both parameters.
Furthermore, our working group has demonstrated the existence of cortical overlapping
between smile and jaw-clenching cerebral areas in normal healthy volunteers.[33] On the other hand, a study published in 2017 demonstrated a strong coactivation
between the masseter muscle and the zygomaticus major, suggesting that the masseter
nerve may be preferred in smile reanimation over other nonfacial nerves such as hypoglossal
or spinal nerves.[34] For this reason, masseteric-to-facial transposition has become our first-choice
surgical technique in both incomplete and short-term complete facial paralyses for
the female population,[11]
[35] even in patients with weak smiles due to their brain plasticity and the capacity
to modulate the strength of the pull to achieve symmetry.
Gracilis transfer:Longstanding unilateral facial paralysis is best addressed with free muscle transplantation
to restore the face motion, because there is no viable facial musculature. Like other
authors, the gracilis muscle is our preferred option.[36]
[37]
[38] After harvesting the flap, the most powerful branch of the obturator nerve should
be used for the nerve coaptation. If two or more branches have the same strength,
we should include all of them in our nerve coaptation to ensure the correct reinnervation
of the flap.
The position of the muscular flap should imitate the contralateral zygomaticus major
muscle to obtain the same vector of the smile, thus the preoperative healthy side
analysis is imperative. Although some variations and refinements have been described
to improve the symmetry when using gracilis free flap,[39]
[40] from our experience, recreating the same zygomaticus-modiolus vector as in the healthy
side has optimal outcomes. Once the muscle is fixed in its position, the correct movement
and position of both oral commissure and nasolabial fold should be checked by using
a nerve stimulator. Neurotization can be obtained from the cross-facial or the masseter
nerve[41]; however, the same distinction regarding the smile (weak or strong) and gender should
be done to obtain excellent results ([Fig. 3]; [Videos 3] and [4]).
Fig. 3 A 40-year-old man with right facial paralysis secondary to acoustic neurinoma resection
8 years previously. Preoperative appearance at rest (A) and when smiling (B). A gracilis muscle transplant connected to a cross-facial nerve graft was performed.
The patient 2 years postoperatively at rest (C) and when smiling (D).
Video 3 Preoperative appearance of a 67-year-old woman with complete left facial paralysis.
Video 4 The same patient from [video 3] after facial reanimation by gracilis transfer at 12 months after surgery.
When performing a free flap is not possible or the patient rejects it, a lengthening
temporalis myoplasty is a good alternative.[42]
[43] The temporalis muscle has the same nerve source as the masseteric nerve, that is
to say the trigeminal nerve, therefore the indication for this technique would be
especially useful in women because of their higher rate of spontaneity when using
a nonfacial nerve ([Videos 5] and [6]).
Video 5 Preoperative appearance of a 69-year-old man with complete left facial paralysis.
Video 6 The same patient from [video 5] after facial reanimation by lengthening temporalis myoplasty at 12 months after
surgery.
Bilateral Facial Paralysis
Bilateral facial paralysis is a rare clinical condition with an incidence of one per
5 million and an occurrence of 0.3 to 2% in facial paralysis cases,[44]
[45] sometimes in the context of Möbius syndrome. Our preferred technique in bilateral
cases is a bilateral-staged free gracilis transfer with a period of 6 months between
the surgeries as reported by several authors.[46]
[47]
[48] The flap is transferred to the face and fixed to the modiolus and zygomatic arch
recreating the zygomaticus major muscle. In bilateral cases, there is no healthy side
from which to reference the vector; therefore, a study in normal subjects has proposed
a vector of 40° to 48° above the horizontal[49] to be used in these patients.
Regarding the nerve source, since both facial nerves are damaged in bilateral cases,
it is necessary to use other nerves such as the hypoglossal, the accessory nerve,
or the masseter branch of the trigeminal nerve. Our preferred option is the masseteric
nerve because it achieves a strong symmetrical smile, which is properly controlled
by the patient, recreating a smile similar to a normal one.[50] In addition, with practice it is possible to develop a spontaneous smile. Although
the use of hypoglossal nerve has been relegated because of the morbidity caused by
its loss, and is only used when the nerve to the masseter is not available, it should
be considered as the first option as a donor motor nerve by performing an end-to-side
coaptation with minimum repercussion in speech or swallowing.[51]
Admittedly, the same technique should be done on both sides to achieve the maximum
level of symmetry ([Videos 7] and [8]).
Video 7 Preoperative appearance of a 4-year-old girl with bilateral facial paralysis secondary
to Möbius Syndrome. The patient presented an aberrant movement in the right side which
was surgically denervated.
Video 8 The same patient from [video 7] after facial reanimation by a bilateral gracilis transfer innervated with masseteric
nerve at 12 months after surgery.
As we have explained previously, bilateral facial paralysis reconstruction requires
two surgeries with a period of 6 months between them. However, our group has reported
a Möbius syndrome case in which after unilateral reconstruction, the patient discovered
what it was to smile and how it should be done, achieving a bilateral movement.[52]
[53]
Static Facial Palsy Reconstruction
When the patient is more than 70 years old and/or has several comorbidities, a static
technique is indicated. Our preferred technique is to perform a suspension of the
nasolabial fold and the oral commissure by a plantaris sling, although fascia lata
and palmaris longus can also be used.[54] This simpler technique achieves good symmetry at rest and it also solves the drooping
of the oral commissure and enhances the oral competence ([Fig. 4]).
Fig. 4 A 74-year-old woman with right complete facial. Preoperatively (A) and postoperatively (B) after performing a static facial paralysis reconstruction.
In summary, considering the variety of techniques available, choosing the right technique
for facial palsy reanimation in each type of patient is probably one of the most challenging
aspects. To aid this decision making, we suggest a treatment algorithm given our experience
and current concepts on the topic, in an attempt to bring us closer to the goal for
reanimating unilateral, bilateral, or complete and incomplete facial palsies with
the appropriate symmetry, synchrony, and spontaneity ([Fig. 5]).
Fig. 5 Algorithm for treating facial paralysis depending on the comorbidities, timing, patient's
gender, kind of smile, and paralysis type.
