Oculoplastic Surgery

Aesthetic Surgery Journal

Preliminary Report 2017, 1–11
© 2017 The American Society for
Aesthetic Plastic Surgery, Inc.
Efficacy of Retrobulbar Hyaluronidase Injection Reprints and permission:
journals.permissions@oup.com

for Vision Loss Resulting from Hyaluronic Acid DOI: 10.1093/asj/sjw216

www.aestheticsurgeryjournal.com

Filler Embolization

Guo-Zhang Zhu, MD, PhD; Zhong-Sheng Sun, MD; Wen-Xiong Liao, MD;

Bing Cai, MD; Chun-Lin Chen, MD; Hui-Hui Zheng, MD; Li Zeng, MD; and
Sheng-Kang Luo, MD, PhD

Abstract
Background:  Vision loss is a rare but serious complication of facial hyaluronic acid (HA) filler injection, for which there is no proven rescue therapy.
Retrobulbar hyaluronidase injection is advocated by many plastic surgeons as an emergency treatment, but has not been carefully assessed for its efficacy.
Objectives:  To evaluate the efficacy of retrobulbar hyaluronidase injection as a rescue treatment for vision loss caused by HA filler embolization.
Methods:  Patients with vision loss caused by HA filler embolization were treated with retrobulbar hyaluronidase injection. Their visual acuity and
fundoscopic images before and after treatment were analyzed for efficacy assessment.
Results:  One patient with branch retinal artery occlusion (BRAO), one patient with posterior ischemic optic neuropathy (PION), one patient with
ophthalmic artery occlusion, and one patient with both BRAO and PION were treated with one or two retrobulbar injections of 1500 or 3000 units hyalu-
ronidase. No patients demonstrated substantial retinal artery recanalization or vision acuity improvement after treatment.
Conclusions:  One or two retrobulbar injections of 1500 to 3000 IU hyaluronidase are unable to recanalize retinal artery occlusion or improve the
visual outcome of patients who presented with vision loss caused by HA filler embolization at least four hours after onset.

Level of Evidence: 4

Editorial Decision date: October 13, 2016.

Vision loss is a rare but devastating complication of facial
hyaluronic acid (HA) filler injection.1-4 The filler presum-
ably causes this complication through the mechanism of
retrograde embolization.5,6 Depending on which artery is
occluded, vision loss can be classified into six subtypes:
ophthalmic artery occlusion (OAO), generalized posterior
ciliary artery occlusion (PCAO) with relative central reti-
nal artery sparing, localized PCAO, central retinal artery
occlusion (CRAO), branch retinal artery occlusion (BRAO),
anterior ischemic optic neuropathy (AION), and posterior
ischemic optic neuropathy (PION).5-7 Twenty-nine such
cases have been reported, and seventy percent of them
developed profound vision loss,1-7 which seriously threat-
ens the patient’s quality of life.
One therapeutic option is the standard conservative
treatment protocol, which includes acetazolamide, tim-
olol, anterior chamber paracentesis, ocular massage,
aspirin, nitroglycerin, mannitol, hyperbaric oxygen, and
corticosteroids.8 However, there is no evidence that this
From the Department of Plastic and Reconstructive Surgery and
the Department of Ophthalmology, Guangdong Second Provincial
People’s Hospital, Guangzhou City, China.
Corresponding Author:
Prof. Sheng-Kang Luo, Department of Plastic and Reconstructive
Surgery, Guangdong Second Provincial People’s Hospital, 466 Middle
Xin Gang Road, Guangzhou City, Guangdong Province 510317, China.
E-mail: gdprs@sina.com
2
protocol improves visual prognosis better than predicted
by the natural disease history. Only patients with ischemic
optic neuropathy appear to benefit from very early and
aggressive corticosteroid treatment.9
Since hyaluronidase can rescue impending nasal skin
necrosis associated with HA filler injection,10 efforts have
been made to deliver hyaluronidase to the ophthalmic
circulation for the hydrolysis of HA emboli. One deliv-
ery technique is the direct injection of hyaluronidase into
ophthalmic artery through selective angiography. Seven
patients with HA filler-induced vision loss were treated
this way, and two of them achieved partial ophthalmic
artery recanalization.2,3 However, this technique demands
an interventional facility and sophisticated skills, and thus
is not widely available as a rescue treatment.
Another delivery technique is retrobulbar hyaluronidase
injection, which is strongly advocated in the literature.1,11
This technique is feasible in theory because this enzyme
has long been used safely as a spreading factor for oph-
thalmic regional blocks,12 and an ex vivo study shows that
hyaluronidase can hydrolyze HA filler transarterially.13
Hu’s group recently tried this method in one patient with
PCAO, and the visual acuity improved slightly two weeks
after treatment.4 However, they failed to report the dose of
retrobulbar hyaluronidase and the recanalization status of
PCAO after treatment. Currently, there is no evidence that
this technique is an effective rescue treatment for patients
with HA filler-induced vision loss.
Here, we report the results of four patients who were
treated with retrobulbar injections of high dose hyaluroni-
dase. We carefully analyzed the temporal changes of their
visual acuity and fundoscopic images before and after
treatment.
METHODS
The study was conducted from June 2015 to July 2016, with
approval from the Institutional Review Board of Guangdong
Second Provincial People’s Hospital (Guangzhou City,
China). All patients were clearly informed of the experi-
mental nature of the retrobulbar hyaluronidase treatment,
and each patient provided written informed consent. The
injection technique is similar to that published.11,12 Briefly,
a dedicated retrobulbar injection needle (25 gauge, 3.8 cm
long) is inserted through the skin at the junction of lateral
one-third and medial two-thirds of the infraorbital rim.
The patient is instructed to rotate the globe toward the
superior and medial direction, and the needle is advanced
1 to 2 cm along the inferior wall of the orbit. After cross-
ing the equator of the globe, the needle is directed medi-
ally and cephalad toward the apex of the orbit. Once a
negative blood aspiration is confirmed, 1500 or 3000 IU
ovine testicular hyaluronidase (Shanghai First Biochemical
Aesthetic Surgery Journal
Pharmaceutical Corporation, China) in 2.0 mL saline is
slowly injected into the retrobulbar space. The needle is
finally withdrawn, and the globe is gently compressed for
one minute.
RESULTS
We treated four referral patients who had monocular vision
loss associated with facial HA filler injections, using retro-
bulbar hyaluronidase injections. Table 1 summarizes their
disease profiles and treatment courses. All patients were
young (mean age, 27 years; range, 23-35 years) and lean
(mean body mass index, 19 kg/m2; range, 17-21 kg/m2)
females without any history of ophthalmic disease. They
all had facial filler injections shortly before developing
the vision loss. The filler injections were made with only
HA filler, which had been injected for nose augmenta-
tion in three patients and for forehead augmentation in
one patient. Three patients developed vision loss imme-
diately after filler injection, and one patient started to
experience blurred vision one hour after filler injection.
The underlying pathologies of vision loss were BRAO and
PION for patient I, BRAO for Patient II, OAO for Patient III,
and PION for Patient IV. All patients presented with some
blepheroptosis, and two patients also presented with oph-
thalmoplegia, consistent with oculomotor nerve palsy. The
blepheroptosis and ophthalmoplegia all resolved in a few
weeks after treatment. No patient had concomitant brain
infarction. The median follow-up time of the patients was
7 weeks (range, 3-8 weeks).
Although patients I to III all presented with retinal artery
occlusion with retinal opacity and edema, the underlying
pathologies of their vision losses were different. Patient
I (23-year-old female) presented with no light perception
(NLP) in her left eye (OS) 32 hours after onset of blurred
vision, which developed one hour after her nose augmen-
tation with 2.0 mL HA filler. An initial fundus photogram
revealed superonasal, superotemporal, and inferotemporal
BRAOs with retinal opacity and edema in the distributions of
these branch retinal arteries (BRA) (Figure 1A). The BRAOs
affected about 40 percent of macular perfusion, but fovea
appeared well perfused (Figure 1A), as confirmed by a sub-
sequent fundus fluorescein angiogram (FFA) (Figure 2C).
Because fovea is the structural basis of central visual acu-
ity, the preservation of fovea perfusion was not consistent
with the patient’s complete blindness. PION was therefore
suspected, in light of the facts that this patient presented
with a persistent defect of afferent pupillary light reflex,
and her ocular computed tomography (CT) with enhance-
ment suggested that the thickness of the left optic disc and
the diameter of the left intraorbital optic nerve were larger
than those of the other eye. In addition, the patient devel-
oped optic disc opacity eight weeks afterwards (Figure 3),
Zhu et al3

Table 1.  Disease Profiles and Treatment Courses for the Four Patients
Patient I Patient II Patient III Patient IV

Age (years) 23 23 35 28

Gender Female Female Female Female

Body mass index (kg/m2) 20 17 21 18

Filler injection site Nose Nose Nose Forehead

HA volume (ml) 2.0 0.8 0.25 5.0

Time to vision loss One hour Immediate Immediate Immediate

Diseased eye Left Right Left Left

Diagnosis BRAO, PION BRAO OAO PION

Blepheroptosis Mild Severe Mild Moderate

Ophthalmoplegia (-) Exotropia (-) Exotropia

Pupillary light reflex Afferent (-) Normal Afferent (-) Afferent (-)
Efferent (+) Efferent (-) Efferent (+)

Cherry red spot (-) (-) (+) (-)

Cornea edema (-) (+) (-) (-)

Brain CT/MRI (-) (-) (-) (-)

Visual field defect NA (+) NA (+)

OCT NE Macular edema NE Normal

ERG response NE Reduced NE Normal

VEP response NE Reduced NE Reduced

Optic nerve CT/MRI (+) NE NE (+)

CTA/ICA DSA NE (-) NE (-)

Hours to treatment 32 12 34 4

Hyaluronidase 1500 IU *2 1500 IU * 1 3000 IU * 2 1500 IU *2

Corticosteroids (-) (+) (+) (+)

RA recanalization Partial (-) Partial NA

Initial visual acuity NLP HM/5 cm NLP 20/200

Final visual acuity NLP 20/60 NLP NLP

Follow up (weeks) 8 6 8 3

BRAO, branch retinal artery occlusion; CTA/ICA DSA, CT angiography/internal carotid artery digital subtraction angiography; ERG, electroretinogram; HM, hand motion; Hours to treatment, hours to
retrobulbar hyaluronidase treatment; Hyaluronidase, retrobulbar injection of hyaluronidase; NA, not applicable; NE, not examined; NLP, no light perception; OAO, ophthalmic artery occlusion; OCT,
optical coherence tomography; PION, posterior ischemic optic neuropathy; VEP, visual evoked potential; RA recanalization, retinal artery recanalization.

which confirmed PION as the fundamental cause of the
complete loss of her left eye visual acuity.9
Patient II (23-year-old female) presented with tem-
poral BRAO in her right eye (OD) twelve hours after an
injection rhinoplasty with 0.8 mL HA filler and immedi-
ate vision loss (Figure 4A). Her initial visual acuity was
confined to perception of hand motion (HM) at a 5 cm
distance. She was found to have cornea edema by slit
lamp biomicroscopy (Supplemental Figure 1A, availa-
ble as Supplementary Material at www.aestheticsurgery-
journal.com), retinal hemorrhage by fundus photography
(Figure 4A), and macular edema by optical coherence
tomography (OCT) (Figure 4B). The cornea edema and ret-
inal bleeding were probably caused by prior aggressive and
incorrect ocular massage, and they blocked the patient’s
vision initially. However, the underlying cause of her right
4 Aesthetic Surgery Journal

Figure 1.  Fundus photograms of the left eye of Patient I, a 23-year-old woman (A) before and (B) eight hours after the first retrobulbar
injection of 1500 IU hyaluronidase. (A) The initial fundus photogram before retrobulbar hyaluronidase injection was characterized by
inferotemporal, superotemporal and superonasal BRAOs (arrows) and retinal opacity with edema in the distributions of these BRAs.
The BRAOs affected about 40% of macular perfusion, but the fovea appeared well perfused and did not display a cherry red spot. (B)
The BRAOs and retinal opacity remained unchanged eight hours after the first retrobulbar hyaluronidase injection.

Figure 2.  Eight hours after the first retrobulbar hyaluronidase injection, Patient I, a 23-year-old woman, was treated with a
second retrobulbar injection of 1500 IU hyaluronidase. Fundus photograms revealed that from (A) three days to (B) one week
after treatment there was a gradual recovery of the diameters and fillings of the superonasal BRA and the inferior branch of the
superotemporal BRA (black arrows), with a concomitant relief of regional retinal opacity and edema; however, inferotemporal
BRA and the superior branch of superotemporal BRA (white arrows) remained occluded. (C) FFA one week after treatment
confirmed the fundus photographic findings.
Zhu et al5
eye vision loss was the ischemic retinal injury caused by
BRAO, as confirmed by subsequent FFA (Figure 5C), and
the presences of visual field, pattern visual evoked poten-
tial (VEP) and electroretinography (ERG) defects detailed
below (Supplemental Figures 2A-4A).
Patient III (35-year-old female) presented with NLP OS
and extensive retinal opacity and edema 34 hours after a
nose augmentation with 0.25 mL HA filler and immediate
vision loss (Figure 6A). However, her initial FFA revealed
severe deficiencies of both retinal and choroidal perfusions
(Supplemental Figure 5A), suggesting the occlusion of
both the central retinal artery (CRA) and posterior ciliary
artery (PCA). We therefore diagnosed OAO as the funda-
mental pathology causing her left eye vision loss.
Like Patient I, Patient IV lost her vision primarily because
of PION. This 28-year-old female experienced immediate
vision loss in her left eye following an excessive forehead aug-
mentation with about 5.0 mL of HA filler. The visual acuity
dropped to 20/200 OS four hours afterwards when she was
referred to our center. Although the retina was completely
healthy (Figure 7A), this eye suffered from an extensive
Figure 3.  Fundus photogram of the left eye of Patient I, a
23-year-old woman, eight weeks after treatment revealed a
pale optic disc that was consistent with the diagnosis of PION.
Figure 4.  Initial fundus photogram (A) and OCT (B) of the right eye of Patient II, a 23-year old woman. (A) Although the
fundus photogram was fuzzy because of cornea edema, it demonstrated the presence of temporal retinal opacity, edema, and
retinal hemorrhage. (B) OCT identified the presence of macular edema (arrows).
visual field defect (Figure 8A), negative afferent pupillary
light reflex, and delayed flash VEP response at reduced ampli-
tudes (Supplemental Figure 6). We therefore believe that the
underlying pathology of her vision loss was PION.
We treated these patients with one or two retrobulbar
injections of 1500 or 3000 IU hyaluronidase, in addition
to the standard conservative treatment protocol.8 The
rationale for choosing 1500 IU as our starting dose was
that the highest concentration of hyaluronidase used for
peribulbular regional block is 300 IU/mL,12 and the total
amount of hyaluronidase in 4.0 mL anesthetic solution is
1200 IU. In addition, the half-life of human hyaluronidase
is about 30 min in the subcutaneous tissue of rodents,14
and thus the initial dose of retrobulbar hyaluronidase
injection should be at least 1200 IU, so that 300 IU hyaluro-
nidase would remain active one hour afterwards, the same
dose that degraded HA filler transarterially.13 We therefore
injected 1500 IU per session in three patients and doubled
the dose for Patient III, who had a more severe condition.
The initial fundus photograms of Patients I-III revealed
clear retinal artery occlusions, which made it possible for
us to observe the therapeutic effect of retrobulbar hyaluro-
nidase injection on retinal artery recanalization. The effect
of this treatment on retinal artery recanalization is very dis-
appointing. In Patient I, superonasal, superotemporal and
inferotemporal BRAOs remained unchanged eight hours
after the first retrobulbar injection of 1500 IU hyaluroni-
dase (Figure 1B, arrows). The second retrobulbar injection
of 1500 IU hyaluronidase only led to a very slow recovery
of the diameters and fillings of the superonasal BRA and
the inferior branch of the superotemporal BRA from three
days to one week after treatment (Figures 2A and 2B, black
arrows); other BRAOs remained unchanged (Figures 2A and
2B, white arrows). In Patient II, there was no relief of retinal
opacity or BRAO one week after one retrobulbar injection
of 1500 IU hyaluronidase (Figure 5A). In Patient III, only
inferotemporal BRAO appeared to be recanalized one day
after the first retrobulbar injection of 3000 IU hyaluronidase
(Figure 6B, arrows). The second retrobulbar injection of
6 Aesthetic Surgery Journal

Figure 5.  The right eye of Patient II, a 23-year-old woman, was treated with one retrobulbar injection of 1500 IU hyaluronidase.
(A) The fundus photogram one week after treatment identified a gradual absorption of retinal hemorrhage, but with no
improvement of temporal retinal opacity. (B) The fundus photogram showed that most of the temporal retinal opacity and
retinal hemorrhage resolved three weeks after treatment. (C) FFA demonstrated the presence of persistent temporal BRAOs
(arrows) and a temporal nonperfusion zone three weeks after treatment.

3000 IU hyaluronidase failed to generate additional BRAO
recanalization (Figure 6C). FFA identified little improve-
ment of retinal and choroidal perfusions one week after
treatment (Supplemental Figure 5B). Therefore, none of the
three patients demonstrated substantial recovery of retinal
artery patency and/or retinal perfusion within one day after
treatment.
None of the four patients experienced visual acuity
improvement because of retrobulbar hyaluronidase injec-
tion. In fact, the visual acuity of Patient IV deteriorated
from 20/200 to NLP OS sixteen hours after treatment
(Figure 8C). The visual acuity of Patient II improved from
HM/5 cm to 20/60 OD two weeks after treatment; however,
the treatment most likely made no contribution to this par-
tial recovery of visual acuity, because it coincided with
the resolution of cornea edema (Supplemental Figure 1B)
and the gradual absorption of retinal hemorrhage (Figures
5A and 5B), both of which blocked visual pathway ini-
tially. In addition, this eye still suffered from peripheral
and inferonasal central visual field defects two and six
weeks after treatment (Supplemental Figure 2). The dis-
eased right eye displayed delayed pattern VEP responses
at lower amplitudes than the left eye did two weeks after
treatment (Supplemental Figure 3). The right eye also gen-
erated a flatter electroretinal response than the left eye
did in response to the same visual stimuli six weeks after
treatment (Supplemental Figure 4). These data point to
ischemic retinal injury that the retrobulbar hyaluronidase
failed to salvage.
DISCUSSION
Vision loss caused by HA filler embolism is a medical
emergency for which there is no proven rescue treatment.
Although retrobulbar injection of hyaluronidase is strongly
advocated by many physicians as a potential rescue treat-
ment,1,11 its efficacy remains unknown. In this study, four
Zhu et al7

Figure 6.  Fundus photograms of the left eye of Patient III, a 35-year-old woman, (A) before, (B) one day, and (C) two days
after retrobulbar hyaluronidase treatment. (A) The initial fundus photogram identified extensive retinal artery occlusions,
retinal opacity and edema, and a cherry red spot. (B) One day after the first retrobulbar injection of 3000 IU hyaluronidase,
inferotemporal BRA appeared to be recanalized (arrows), but retinal opacity and edema remained unchanged. (C) The second
retrobulbar injection of 3000 IU hyaluronidase one day after the first injection failed to recanalize additional retinal arteries or
improve overall retinal opacity and edema.

patients who presented with HA filler-induced BRAO,
OAO, and/or PION at least four hours after the onset of
vision loss were treated with retrobulbar injection of high
dose hyaluronidase, and no patient demonstrated substan-
tial retinal artery recanalization or visional acuity improve-
ment. This questions the efficacy of this treatment among
patients with HA-induced vision loss that lasts at least
four hours.
Facial HA filler injection presumably causes vision loss
through retrograde embolism.5,6 By this mechanism, filler
particles are accidentally injected into the orbital branches
of ophthalmic artery (supratrochlear, dorsal nasal, or
supraorbital artery) or into the branches of facial or super-
ficial temporal arteries that communicate with the oph-
thalmic artery system. The high pressure injection pushes
the filler particles against the blood stream to the trunk of
the ophthalmic artery. When the injection stops, the HA
particles are carried by the blood stream to occlude the
ocular branches of the ophthalmic artery, such as the CRA,
PCA, and various arteries that nourish the posterior optic
nerve. The CRA is a terminal branch of the ophthalmic
artery that nourishes the inner layer of the retina, while
the PCA branches of the ophthalmic artery directly sup-
ply blood to the choroid, outer layer of the retina, and
the anterior part of the optic nerve.8,9,15 The posterior part
of the optic nerve derives its blood supply from various
sources other than the PCA.9 Because the retina, choroid
and optic nerve are all critical for vision, the occlusion of
their supplying arteries by the HA filler leads to various
types of vision loss.
The CRA is a terminal artery that supplies the inner
retina through its four major BRAs. There is little collat-
eral circulation for the CRA except the cilioretinal artery
found in some individuals.8,15 This makes the inner retina
particularly vulnerable to ischemia and necrosis following
retinal artery occlusion. In rhesus monkey, the retina only
8 Aesthetic Surgery Journal

Figure 7.  Patient IV, a 28-year-old woman, (A) presented with a healthy retina in her left eye, which remained intact (B) eight
hours and (C) 16 hours after two retrobulbar injections of 1500 IU hyaluronidase each, separated by one hour.

Figure 8.  Patient IV, a 28-year-old woman, (A) presented with visual acuity 20/200 OS with extensive visual field defect in her
left eye, which remained unchanged (B) eight hours after retrobulbar hyaluronidase treatment, but progressed to full blindness
(C) 16 hours after retrobulbar hyaluronidase treatment. Note that each numerical value in the visual field plot represents the
retinal sensitivity in decibels (dB) to a light stimulus at that retinal point in the visual field. The regions in the visual field
with high retinal sensitivity are marked blue, the regions with low retinal sensitivity are marked red, and the regions with
intermediate retinal sensitivity are marked green.

tolerates up to 97 min of CRA clamping without causing time window was found to be as long as 15 hours in human
detectable injury, and four hours of complete CRA occlu- patients with CRAO caused by fibrinoplatelet emboliza-
sion results in massive irreversible retinal injury.16 This tion, particularly when incomplete CRAO was involved.8,17
Zhu et al9
However, it is very likely that soft HA gel particles cause
total or subtotal CRAO and an effective rescue treatment
needs to be initiated within four hours, and must be able
to substantially recanalize HA occlusions and restore oph-
thalmic perfusion within four hours, if that treatment is to
improve the visual outcome of patients with vision loss
caused by HA filler embolization.
Hyaluronidase is an effective rescue treatment for
impending nasal skin necrosis caused by HA filler embo-
lization.10 To recanalize the HA embolization in the oph-
thalmic artery system, hyaluronidase must be delivered to
the vicinity of the HA emboli using various delivery tech-
niques. The most efficient technique is the direct infusion
of hyaluronidase into the ophthalmic artery using selec-
tive angiography.2,3 However, this interventional treatment
demands a highly specialized facility, instruments, and
skills that are not readily available to most physicians and
patients. The other potential delivery option is retrobulbar
hyaluronidase injection.1,4,11 Ophthalmologists have long
been using hyaluronidase as a spreading factor for retro-
bulbar or peribulbar regional blocks.12 With proper train-
ing, plastic surgeons and dermatologists should be able to
utilize this simple, fast, and well-established procedure for
the rescue treatment of patients with vision loss caused by
HA filler embolism.
The rationale behind retrobulbar hyaluronidase treat-
ment is that the ophthalmic artery, CRA, PCA, and the
arteries of the posterior optic nerve course through the
retrobulbar space, which is located posterior to the globe
and within the extraocular muscle cone.11,15 Because
hyaluronidase degrades HA filler transarterially,13 it is
expected that retrobulbarly injected hyaluronidase will
cross the walls of these arteries and enter the ophthal-
mic circulation. The enzyme would then be carried by
the blood to the branches of the ophthalmic artery that
are occluded by HA emboli. The direct contact of hyal-
uronidase with HA occlusion would allow the enzyme
to hydrolyze the HA emboli, and restore the perfusions
of ischemic retina, choroid or optic nerve. It is not clear
whether hyaluronidase can penetrate human sclera and
reach the choroid and retina. The feasibility of this treat-
ment was recently tested in one patient with vision loss
caused by the embolization of PCA by HA filler.4 However,
this study did not report the exact dose of retrobulbar
hyaluronidase injection and failed to follow the patient
using fundoscopic imaging techniques. There is an urgent
need to determine the effect of retrobulbar hyaluronidase
injection on retinal artery recanalization and patient’s
visual outcome.
In this study, we had a rare opportunity to treat four
patients with vision loss associated with facial HA filler
injection. The fundus photograms of three of these patients
showed clear retinal artery occlusions (Figures 1A,4A,
and 6A), which allowed us to explore the recanalization
effects of retrobulbar hyaluronidase treatment. However,
this treatment was unable to substantially recanalize ret-
inal artery occlusion for any patient within four hours.
One retrobulbar injection of 1500 IU hyaluronidase failed
to recanalize any BRAO for Patients I and II within eight
hours (Figures 1 and 5). A second retrobulbar injection
of 1500 IU hyaluronidase for Patient I only gradually and
slightly opened up some BRAOs in a few days, while the
major BRAOs remained unchanged (Figure 2). A very sim-
ilar phenomenon was also found in Patient III, for whom
one retrobulbar injection of 3000 IU hyaluronidase just
recanalized one branch of the CRA in one day, without
any apparent relief of global retinal opacity and edema
(Figure 6B). A second retrobulbar injection of 3000 IU
hyaluronidase made no further improvement of retinal per-
fusion (Figure 6C). These results suggest that one or two
retrobulbar injections of 1500 to 3000 IU hyaluronidase are
not fast and effective enough to meet the requirements of
a rescue treatment, namely recanalyzing the retinal artery
occlusion within four hours.
This recanalization failure could not be attributed to
an insufficient dose of hyaluronidase. The half-life of this
enzyme is about 30 min in the subcutaneous tissues of
rodents.14 If that is also the case for the enzyme in human
retrobulbar space, there will be at least 325 IU hyaluro-
nidase active one hour after one retrobulbar injection of
1500 IU hyaluronidase (Patients I and II). Patient III was
retrobulbarly injected with 3000 IU hyaluronidase, while
Patient IV received two retrobulbar injections, 1500 IU
hyaluronidase each, separated by one hour. Both patients
were expected to have more than 325 IU active hyaluroni-
dase two hours after treatment. This amount of enzyme
is greater than the 300 IU bovine testicular hyaluronidase
that degraded HA filler across a fresh cadaveric facial artery
within four hours.13 Therefore, our patients had more than
enough active hyaluronidase at the retrobulbar space for
greater than one or two hours, able to enter the ophthalmic
circulation to digest HA emboli.
The retinal artery recanalization likely failed because
hyaluronidase is unable to efficiently permeate into the
ocular branches of the ophthalmic artery and reach the
HA emboli. It is very likely that the ophthalmic artery and
its branches in a living patient are much less permeable
to hyaluronidase than a cadaveric facial artery.13 It is also
possible that there is blood stasis in occluded ophthalmic
artery branches, and hyaluronidase is unable to reach the
distal HA emboli once it crosses the arterial wall into the
ophthalmic artery system. There are many hyaluronidase
inhibitors in human blood, and the half-life of hyaluroni-
dase is 3.2 min in human serum.14,18 It is likely that hyal-
uronidase is quickly inactivated once it crosses the arterial
wall and enters the ophthalmic circulation. To ensure there
is sufficient amount of active hyaluronidase in the oph-
thalmic circulation to hydrolyze HA emboli, the enzyme
10
needs to be constantly delivered into the ophthalmic artery
system. Because repeated retrobulbar injections carry the
inherent risk of ocular perforation,12 retrobulbar injec-
tion may be a poor delivery technique of hyaluronidase.
Direct infusion of hyaluronidase into the ophthalmic artery
through selective angiography would be a more effective
approach.
The fourth possibility for the recanalization failure is
that secondary thrombosis might develop soon after HA
embolization, requiring both hyaluronidase and a fibrino-
lytic drug for treatment. This possibility is supported by the
observation that HA filler-induced rat inferior epigastric
artery obstruction is more effectively recanalized by the
intravenous infusion of both hyaluronidase and urokinase
than by hyaluronidase infusion alone.19 However, clinical
studies do not support this concept.2,3 Korean ophthalmol-
ogists found that the injection of high dose (9000 IU) hyal-
uronidase into the ophthalmic artery partially recanalized
OAO in one patient, while low dose (1500 IU) hyaluro-
nidase plus urokinase did not always partially recanalize
OAOs. This result suggests that hyaluronidase treatment
remains the most critical, and the dose probably needs to
be very high, since the half-life of this enzyme in blood is
very short.14,18 However, further increasing the dose and/
or frequency of retrobulbar hyaluronidase injections may
not be an option, because of the unforeseeable risk to the
optic nerve or surrounding ocular tissues.
This lack of visual acuity improvement among our
patients after retrobulbar hyaluronidase injection is not
unexpected considering the poor recanalization effect of
this treatment. The first three patients presented 32, 12,
and 34 hours, respectively, after the onset of vision loss,
much longer than the four hours of retinal survival time
identified in rhesus monkey.16 Their inner retina and/or
posterior optic nerve probably had developed irreversible
ischemic injury before any treatment could make a differ-
ence. Patient IV, who received retrobulbar hyaluronidase
treatment four hours after the onset of vision loss, could
have benefited from the reperfusions of the optic nerve;
however, the hyaluronidase treatment failed to rescue the
vision even when Patient IV had residual visual acuity
(20/200 OS). Hu’s group recently treated a patient seven
hours after HA filler-induced PCAO using one retrobulbar
injection of an unknown dose hyaluronidase.4 The visual
acuity improved from NLP to HM OD; however, this effect
was observed two weeks after treatment, suggesting that
this treatment made no contribution to the visual acuity
improvement. Therefore, retrobulbar hyaluronidase injec-
tion has not been found to be effective in improving the
visual outcome of patients who present at least four hours
after the onset of vision loss caused by HA embolism.
This study is limited by a small number of patients
without a control group, since vision loss is a rare
Aesthetic Surgery Journal
complication of facial filler injection. This study is also
limited by the fact that most of our patients received ret-
robulbar hyaluronidase treatment much longer than four
hours after the onset of vision loss, which severely limits
the effect of this treatment in rescuing the patient’s visual
outcome. This study does not rule out the possibility that
patients who receive retrobulbular hyaluronidase within
four hours after filler injection benefit visually from this
treatment. The exact answer awaits future studies. It is
worthy of note that retrobulbar injection itself and hyalu-
ronidase may cause serious complications,12 and due pre-
caution should be exercised. It is not clear why the visual
acuity of Patient IV worsened from 20/200 to NLP OS
sixteen hours after retrobulbar hyaluronidase treatment,
which does suggest that this treatment is not without risk.
CONCLUSIONS
One or two retrobulbar injections of 1500 to 3000 IU hya-
luronidase are unable to recanalize retinal artery occlusion
or improve the visual outcome of patients who presented
with vision loss caused by HA filler embolization at least
four hours after onset.
Supplementary Material
This article contains supplementary material located online at
www.aestheticsurgeryjournal.com.
Disclosures
The authors declared no potential conflicts of interest with
respect to the research, authorship, and publication of this
article.
Funding
The authors received no financial support for the research,
authorship, and publication of this article.
REFERENCES
1. Beleznay K, Carruthers JD, Humphrey S, Jones D. Avoiding
and treating blindness from fillers: a review of the world
literature. Dermatol Surg. 2015;41(10):1097-1117.
2. Oha BL, Jungb CK, Parka KH, Hongc YJ, Woo SJ. Therapeutic
intra-arterial hyaluronidase infusion for ophthalmic artery
occlusion following cosmetic facial filler (hyaluronic acid)
injection. Neuro Ophthalmol. 2014; 38(1):39-43
3. Kim YK, Jung C, Woo SJ, Park KH. Cerebral angio-
graphic findings of cosmetic facial filler-related oph-
thalmic and retinal artery occlusion. J Korean Med Sci.
2015;30(12):1847-1855.
4. Hu XZ, Hu JY, Wu PS, Yu SB, Kikkawa DO, Lu W.
Posterior ciliary artery occlusion caused by hyaluronic
acid injections into the forehead: a case report. Medicine
(Baltimore). 2016;95(11):e3124.
Zhu et al11
5. Park SW, Woo SJ, Park KH, Huh JW, Jung C, Kwon OK.
Iatrogenic retinal artery occlusion caused by cosmetic facial
filler injections. Am J Ophthalmol. 2012;154(4):653-662.e1.
6. Park KH, Kim YK, Woo SJ, et al.; Korean Retina
Society. Iatrogenic occlusion of the ophthalmic artery
after cosmetic facial filler injections: a national sur-
vey by the Korean Retina Society. JAMA Ophthalmol.
2014;132(6):714-723.
7. Chen Y, Wang W, Li J, Yu Y, Li L, Lu N. Fundus artery
occlusion caused by cosmetic facial injections. Chin Med
J (Engl). 2014;127(8):1434-1437.
8. Varma DD, Cugati S, Lee AW, Chen CS. A review of central
retinal artery occlusion: clinical presentation and man-
agement. Eye (Lond). 2013;27(6):688-697.
9. Hayreh SS. Management of ischemic optic neuropathies.
Indian J Ophthalmol. 2011;59(2):123-136.
10. Sun ZS, Zhu GZ, Wang HB, et al. Clinical outcomes of
impending nasal skin necrosis related to nose and nasol-
abial fold augmentation with hyaluronic acid fillers. Plast
Reconstr Surg. 2015;136(4):434e-441e.
11. Carruthers J, Fagien S, Dolman P. Retro or peribulbar
injection techniques to reverse visual loss after filler injec-
tions. Dermatol Surg. 2015;41 Suppl 1:S354-S357.
12. Palte HD. Ophthalmic regional blocks: management, chal-
lenges, and solutions. Local Reg Anesth. 2015;8:57-70.
13. DeLorenzi C. Transarterial degradation of hyaluronic
acid filler by hyaluronidase. Dermatol Surg. 2014;40(8):
832-841.
14. Wasserman RL. Overview of recombinant human hyalu-
ronidase-facilitated subcutaneous infusion of IgG in pri-
mary immunodeficiencies. Immunotherapy. 2014;6(5):
553-567.
15. Michalinos A, Zogana S, Kotsiomitis E, Mazarakis A,
Troupis T. Anatomy of the ophthalmic artery: a review
concerning its modern surgical and clinical applications.
Anat Res Int. 2015;2015:591961.
16. Hayreh SS, Zimmerman MB, Kimura A, Sanon A. Central
retinal artery occlusion. Retinal survival time. Exp Eye
Res. 2004;78(3):723-736.
17. Ahn SJ, Kim JM, Hong JH, et al. Efficacy and safety
of intra-arterial thrombolysis in central retinal artery
occlusion. Invest Ophthalmol Vis Sci. 2013;54(12):
7746-7755.
18. Wolf RA, Chaung LY, O’Hara D, Smith TW, Muller
JE. The serum kinetics of bovine testicular hyaluroni-
dase in dogs, rats and humans. J Pharmacol Exp Ther.
1982;222(2):331-337.
19. Chiang C, Zhou S, Liu K. Intravenous hyaluronidase with
urokinase as treatment for arterial hyaluronic acid embol-
ism. Plast Reconstr Surg. 2016;137(1):114-121.