1 s2.0 S2468785522002348 Main
1 s2.0 S2468785522002348 Main
1 s2.0 S2468785522002348 Main
Available online at
ScienceDirect
www.sciencedirect.com
Review
A R T I C L E I N F O A B S T R A C T
Article History: Purposes: To evaluate clinical usefulness of endoscope-assisted medial orbital wall fracture repair via the retro-
Received 11 August 2022 caruncular approach (rc-EAMOWFR) vs. no surgery (NS), and to perform a narrative review of relevant literature.
Accepted 16 August 2022 Methods: This was a retrospective cohort study enrolling isolated medial orbital wall fracture (IMOWF) eyes
Available online 18 August 2022
presented to two German level 1 trauma centers during a 7-year interval. The predictor variable was treat-
ment type (rc-EAMOWFR vs. NS), and the main outcomes were late enophthalmos (LE) and retrobulbar hem-
Keywords:
orrhage (RH) assessed at 9−15 posttraumatic months. Descriptive and bivariate statistics were computed at
Orbital fracture
a = 95%. Binary adjustments enabled calculation of number needed to treat (NNT), to harm (NNH), and likeli-
Endoscope
Enopthalmos
hood to be helped or harmed (LHH) for demonstrating benefit-risk tradeoffs. Moreover, a narrative review
Retrobulbar hemorrhage was also performed.
Benefit-risk Results: The sample comprised 502 patients (28.3% females; mean age, 46.5§19.2 years) with 541 IMOWF
eyes (5.9% NS; 7.2% LE; 1.3% RH). Operated eyes had significantly lower LE events than NS eyes (symptomatic
IMOWF: P < .0001; 95% confidence interval [CI], .03 to .16; NNT = 2 [95% CI, 1.1 to 6.1]; asymptomatic
IMOWF: P < .0001; 95% CI, .01 to .07; NNT = 2 [95% CI, 1.1 to 1.8]). There were 7 (1.5%) RH events following
rc-EAMOWFR (P = .99; 95% CI, .06 to 17.4; NNH = 68 [95% CI, 38.3 to 254.2]). LHH calculations posited that rc-
EAMOWFR was 34 times more likely to prevent LE than to cause RH, regardless of fracture symptoms. Our
results conformed to those of other 15 studies.
Conclusions: The results of this study suggest that all IMOWFs be treated. rc-EAMOWFR performed in every
68 IMOWFs would be at risk of one RH event, but prevent 34 eyes from LE due to untreated fractures. Nearly
72% of untreated IMOWFs develop LE after 9 months.
© 2022 Elsevier Masson SAS. All rights reserved.
Introduction
Abbreviations: CI, Confidence interval; CT, Computer tomography; EAMOWFR, Endo-
scope-assisted medial orbital wall fracture repair; IMOWF, Isolated medial orbital wall Isolated medial orbital wall fractures (IMOWFs) have been diag-
fracture; LE, Late enophthalmos; LoE, Level of Evidence; LHH, Likelihood to be helped
nosed and treated more often than earlier because of a greater inci-
or harmed; MRI, Magnetic resonance imaging; NNH, Number needed to harm; NNT,
Number needed to treat; NS, No surgery (group); OCEBM, The Oxford Centre for Evi- dence of high-energy impact orbital injuries and prodigious
dence-Based Medicine; OFF, Orbital floor fracture; OMFS, Oral-Maxillofacial surgeons/ development of imaging technology and usage [1,2]. Its incidence
surgery; rc-EAMOWFR, Endoscope-assisted medial orbital wall fracture repair via the was found to be higher than that of orbital floor fractures (OFFs)
retrocaruncular approach; RG, Recommendation Grade; RH, Retrobulbar hemorrhage;
[1−3]. Unfortunately, there remain many more doubts than certain-
RR, Relative risk; STROBE, The STrengthening the Reporting of OBservational studies in
Epidemiology (guidelines)
ties whether and how IMOWFs should be treated. The German S2e
I
This study was presented in part as an abstract at the 25th Congress of the Euro- (evidence-based, non-consensus) guideline on orbital reconstruction,
pean Association for Cranio-Maxillo-Facial Surgery (EACMFS), July 2021, Paris, France. albeit valid until 2018 (viz., its update is planned out), suggested that
* Corresponding author. orbital fractures be immediately treated in case of persistent oculo-
E-mail address: poramate.pitakarnnop@gmail.com (P. Pitak-Arnnop).
1 cardiac reflex, nausea and vomiting (indicating extraocular muscle
Equal contribution.
https://doi.org/10.1016/j.jormas.2022.08.010
2468-7855/© 2022 Elsevier Masson SAS. All rights reserved.
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
entrapment). Further indications include ocular malposition, en-/ volumetric analysis of CT data, as recommended by other authors
exophthalmos, soft tissue prolapsed into the paranasal sinus, defects (LoE 5, RG D) [2]. Patients with (1) incomplete records, (2) denial in
≥ 2 cm2, and diplopia (Level of Evidence [LoE] 3, Recommendation study participation, or (3) inadequate follow-up (< 9 months) were
Grade [RG] B) [4]. Small orbital wall fractures without symptoms suit excluded from the present study.
conservative/non-operative treatment (LoE 4, RG C) [4]. However, up The bi-institutional review board approved the project, and all
to three-fourth of untreated IMOWF patients could develop late patients signed (1) the surgical consent to prophylactic surgery
enophthalmos (LE) (LoE 4, RG C) [5]. against LE due to untreated fractures, and (2) the research consent to
In 2004, some of us (J-P.M., P.P.) pioneered endoscope-assisted their anonymous data use. The “all-rather than-none” algorithm can
medial orbital wall fracture repair via the retrocaruncular approach be cancelled at any time, if harm appears higher than benefits for
(rc-EAMOWFR), and reported promising outcomes in small cohorts patients. There was neither human contact nor tissue manipulation
treated at Pitie-Salpe^trie
re University Hospital in Paris, France (LoE 4, during this study, i.e., chart reviewing only. The Helsinki Declaration’s
RG C) [6,7]. Since then, this technique has been used widely across guidelines and the STROBE statement were strictly adhered.
the country (LoE 4, RG C) [8]. In spite of the fact that much was writ- Additionally, to accomplish the narrative review of relevant stud-
ten about the classical transcaruncular approach to the IMOWF by ies, the electronic data searches based on PubMed/Medline, Embase,
many authors including the AO Reference [2,3,9−11], a cadaver study and Cochrane review from the inception to August 12, 2022.
affirmed the superiority of the retrocaruncular approach over the
transcaruncular approach with regard to lower risks of eyelid and
nasolacrimal complications, and simpler combination with the trans-
conjunctival approach to the OFF (LoE 5, RG D) [12]. Study variables
Some German otolaryngologists, despite not main specialists
responsible for facial trauma in this country, prefer the “no surgery” The main predictor variable was treatment type, which was
(NS; i.e., referred to conservative/non-operative treatment) approach recorded as binary (rc-EAMOWFR vs. NS). Surgical techniques and
to IMOWFs because they believe that nothing would happen after instruments were detailed in our previous publications [6,7], with
the IMOWF (personal communication; LoE 5, RG D). A recent retro- some modifications (see below). In every intervention, the fracture
spective study (n = 60; follow-up, 15-45 days) from Parisian oral- was repaired with 0.25- or 0.5-thick non-porous polydioxanone
maxillofacial surgeon (OMFS)-researchers suggested operative treat- sheets (PDSÒ foil, Johnson&Johnson Medical GmBH, Ethicon, Norder-
ment only for symptomatic IMOWFs because this fracture could heal stedt, Germany).
spontaneously without any consequence (LoE 4 with descriptive sta- The main outcomes were preventing LE if the fracture was left
tistics only, RG C) [1]. On the contrary, the first author (P.P.) noticed untreated (as the main therapeutic outcome), and presence of ocular
LE in many untreated IMOWFs (unpublished data before this study compartment syndrome in form of retrobulbar hemorrhage (RH; as
began), as also described by other authors (LOE 4, RG C) [5]. She has, the main adverse event). These outcomes were examined pre- and
therefore, applied the “all-rather than-nothing” treatment concept to postoperatively twice or more by an oral surgeon or OMFS resident/
IMOWFs. This tenet, however, remains scientifically unapproved in faculty, and a staff-grade or consultant ophthalmologist. The Herter
any large cohort with long-term follow-ups (≥ 6 months), and the exophthalmometer (Block Optic Design GmbH, Dortmund, Germany)
indication of IMOWF repair has continued to be controversial hith- was used to evaluate LE according to the recommendations of the
erto (at least in France [1] and Germany, where the updated practice International Thyroid Eye Disease Society (LoE 3, RG B) [13]. Although
guideline has still been absent). 70% of RH can be diagnosed by clinical presentations [14], such as
The main purpose of this study was to answer the following clini- rush stony hardness of the globe (mostly equal to palpable hardness
cal research question: “Among IMOWF patients, does the rc- of prostate hyperplasia) before visual changes, we confirmed the
EAMOWFR, when compared to NS, provide the better benefit-risk diagnosis of RH, when possible, with immediate or fast-track CT. The
profile as measured by the metrics of number needed to treat (NNT), complete list of RH symptoms can be found in the work of other
to harm (NNH), and likelihood to be helped or harmed (LHH)?” Our authors (LoE 1, RG A) [14,15].
null hypothesis was that there would be no significant differences The secondary outcome was ophthalmologic abnormalities other
between therapeutic and adverse outcomes of rc-EAMOWFR vs. NS. than LE and RH, including eyelid emphysema/edema (especially at
The specific aims of this study were as follows: (1) to identify a cohort the medial canthal area), narrowing of the palpebral fissure, propto-
of IMOWF patients, (2) to assess and compare therapeutic and sis, ocular pain, papillary disturbance, diplopia, reduced visual acuity
adverse outcomes of rc-EAMOWFR vs. those of NS, 3) to determine and/or ocular motion (especially restriction of abduction, retraction
whether the “all-rather than-none” treatment concept for IMOWFs is syndrome, and positive forced duction test), damage to the extraocu-
scientifically acceptable, and (4) to perform a narrative review of rel- lar muscles, lacrimal sac and cornea, increased intraocular pressure,
evant literature. At the end, this study will provide the 2011 Oxford optic nerve injury, blindness, cerebrospinal fluid rhinorrhea [2
Centre for Evidence-Based Medicine (OCEBM)’s LoE “3”, and RG “B”. −4,15]. This secondary outcome was also recorded as binary (no
All LoE and RG mentioned in this paper also conform to this OCEBM’s symptom [asymptomatic] vs. ≥ 1 symptoms [symptomatic]). We
guideline. counted a secondary outcome symptom associated with LE or RH as
an LE or RH component. LE and RH events were apt to, if indicated,
Materials and methods corrective surgery (for example, orbital wall augmentation, reosteot-
omy) as fast as possible. RH or any other adverse events after correc-
Study design/population tive revision did not enter the analysis, i.e., we analyzed only the
secondary outcomes from the fracture only.
The investigators designed and implemented a retrospective Other variables were demographic, clinical or radiologic parame-
cohort and narrative review study. To reach the first aim, we enrolled ters, including binary outcomes (yes/no: i.e., patient’s gender, injury
a study sample of all IMOWF patients presented to two German level cause [assault vs. other], anticoagulant therapy, blow-in fracture,
1 trauma centers during a 7-year period of data collection. To be orbital tissue entrapment, underlying sinusitis or other antral dis-
included in the study sample, the subjects had to have a computer eases), and continuous data (patient’s age, fracture size ≥ 2 cm2). The
tomography (CT)-confirmed IMOWF (recorded in the hospital regis- cutoff value of 2 cm2 was used because orbital fractures ≥ 2 cm2 are
tries), and one-year (defined as 9-15 posttraumatic months) clinical supposed to be reconstructed with metal or bone grafts (LoE 4, RG C)
follow-up using a Hertel or Naugle exophthalmometer and/or [16,17].
2
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
Data management and statistical analysis Of 32 non-operated (NS-IMOWF) eyes, all but two were left
untreated after primary evaluation by other specialty colleagues (i.e.,
One external (non-surgeon) author (L.-K.W.) extracted and trans- otolaryngologists, plastic, ophthalmic or trauma surgeons in the
ferred blinded data from the hospital registries to a Microsoft Excel absence of an OMFS). No attempt to surgery was initiated, if the frac-
2007 sheet (Microsoft Inc., WA, USA). Four of us (P.P., C.T., K.S., N.S.) ture was asymptomatic and ≥ 2 weeks old (LoE 4, RG C) [1] and/or
analyzed the blinded data using MedCalcÒ (MedCalc Software Ltd., refusal of surgery. The other two patients were physically unfit for
Ostend, Belgium). Descriptive statistics were computed for each surgery under general anesthesia. Asymptomatic IMOWF eyes tended
study variable, and bivariate analyses aimed to assess the associa- to be left untreated (P = .006; 95% CI, 1.3 to 4.9). NS-IMOWFs were
tions between the study variables at a = 95%. associated with LE (asymptomatic IMOWFs: P < .0001; relative risk
For the benefit-risk appraisal, NNT and NNH, with their respective (RR), 0.3; 95% CI, .01 to .07; NNT = 2 [95% CI, 1.1 to 1.8]; symptomatic
95% confidence interval (95% CI), were calculated. LHH was computed IMOWFs: P < .0001; RR, 0.7; 95% CI, .03 to .16; NNT = 2 [95% CI, 1.1 to
to illustrate potential trade-offs for LE prevention and RH events. If 6.1]; overall: P < .0001; RR, .04; 95% CI, .03 to .07; NNT = 2 [95%, 1.2
the 95% CI of NNT or NNH included “infinity”, the result was consid- to 1.9]). NNH of the RH events was 68 (95% CI, 38.8 to 254.2). Hence,
ered not statistically significant at the P < .05 threshold. The terms LHH is 34, regardless of symptoms.
“statistically significant” and “not statistically significant” are used Concerning fracture size/location relative to vascular landmarks
descriptively and not inferentially [18]. To be effective in the real- (i.e., the anterior and posterior ethmoid formina at »24 and »36 mm
world clinical practice, the treatment must be efficacious and tolera- from the anterior lacrimal crest, respectively, and the optic nerve is
ble enough according to the “rule of thumb”, i.e., single-digit NNTs located on 6−7 mm posterior to the posterior ethmoid foramen [3]),
for efficacy measures suggest the intervention with very useful 291 operated eyes (or 57.2%) had the posterior extent between the
advantages, and double-digit NNHs (or higher) for adverse events anterior and posterior ethmoid vessels, 5 of which (or 1.7%) devel-
indicate the potentially tolerable treatment. LHH values > 1 are oped RH. The 14 others (or 2.8%) located beyond the posterior eth-
therefore desirable [19]. moid artery and very close to the orbital apex, and 1 of these (or
The post hoc power based on the main therapeutic outcome, i.e., 7.1%) had RH. However, there was neither optic neuropathy nor
LE, were computed using G Power 3 for Windows (HHU Du € sseldorf, sinusitis nor postoperative infection, and RH events were not associ-
Du€ sseldorf, Germany) with an effect size of .5, an a error probability ated with the fracture location in relation to vascular landmarks
of .05, and a sample size of 541. (P = .088) The interfering ethmoid vessels were intraoperatively coag-
ulated as appropriate. A subciliary or subtarsal silicone drainage was
placed, if the patient received an anticoagulant. The anatomical con-
Results cern, adequate hemostasis, and placement of prophylactic drainage
may partly explain the low occurrence of RH in this cohort.
Over the 7-year period, 541 eyes (5.9% NS; 7.2% had LE; 1.3% had All procedures lasted maximally 30 min, the retrocaruncular inci-
postoperative RH; 92.5% were operated within the third posttrau- sion closed within 48 hours without the need of suturing. Patients
matic day by the first author [P.P.]) in 502 IMOWF patients (28.3% were discharged from the hospital at postoperative day 2 to 5 with
females; mean age, 46.5§19.2 years) met the inclusion criteria. No the exception of those with polytrauma. Single shots of cefuroxime
otherwise eligible patients were excluded. There was no significant 1.5 g (or clindamycin 600 mg, if penicillin-allergy) and dexametha-
distinction of almost all the parameters (i.e., age, gender, injury cause, sone 16 mg were intravenously given during the operation, except
RH events, anticoagulant therapy, fracture size ≥ 2 cm2, blow-in frac- those with underlying sinusitis who benefited from thrice-daily
ture, and underlying sinusitis) between the two groups, reinforcing ampicillin/sulbactam 2/1 g for 5−7 days. 17 operated patients (or
the assumption that these variables have no effect on the primary 3.3%) experienced a displaced PDSÒ foil and required a secondary
analysis. Table 1 presents all study variables vs. the predictor variable, revision; 12 of those (or 70.6%) had claimed heavy sneezing. There
and absolute risk and NNT or NNH. was no posttraumatic RH; all of the seven RH events were
Table 1
Cohort characteristics grouped by the main predictor, and calculations of absolute risk, and NNT or NNH.
Parameters Overall EAMOWFR NS P value(RR; 95% CI) Absolute risk(95% CI) NNT or NNH(95% CI)
Demographics
Sample size (patient) 502 (100) 471 (93.8) 31 (6.2) N/A N/A N/A
Average age 46.5 § 19.2(18.5-90.8) 46.3 § 19.1(18.5-90.8) 50.0 § 20.1(21-86.5) .3 (N/A; -3.3 to 10.7) N/A N/A
Female gender 142 (28.3) 132 (28) 10 (32.3) .6 (1.2; 0.68 to 2) N/A N/A
Sample size (eye) 541 (100) 509 (94.1) 32 (5.9) N/A N/A N/A
Injury cause .14 (1.3; .9 to 1.9) N/A N/A
Assault 332 (61.4) 317 (62.3) 15 (46.9)
Others 209 (38.6) 192 (37.7) 17 (53.1)
Clinical presentations
Symptomatic* 287 (53) 280 (55) 7 (21.9) .006 (2.5; 1.3 to 4.9) (+)33.1 (18.2 to 48.1) 4 (2.1 to 5.5)
Asymptomatic 254 (47) 229 (45) 25 (78.1)
Symptomatic with LE 15 (5.2) 11 (3.9) 4 (57.1) < .0001 (.07; .03 to .16) (-)53.2 (16.5 to 89.9) 2 (1.1 to 6.1)
Asymptomatic with LE 24 (9.4) 5 (2.2) 19 (76) < .0001 (.03; .01 to .07) (-)73.8 (57 to 90.7) 2 (1.1 to 1.8)
RH 7 (1.3) 7 (1.5) 0 .99 (1; .06 to 17.4) (+)1.5 (.4 to 2.6) 68 (38.8 to 254.2)
Anticoagulant therapy 46 (8.5) 45 (8.8) 1 (3.1) .3 (2.8; .4 to 19.9) N/A N/A
Imaging characteristics
Fracture size ≥ 2 cm2 103 (19) 98 (19.3) 5 (15.6) .6 (1.2; .5 to 2.8) N/A N/A
Blow-in fracture 36 (6.7) 35 (6.9) 1 (3.1) .4 (2.2; .3 to 15.6) N/A N/A
Orbital tissue entrapment 158 (29.2) 156 (30.6) 2 (6.3) .02 (4.9; .3 to 18.9) N/A N/A
Underlying sinusitis 28 (5.2) 28 (5.5) 0 .35 (3.7; .2 to 59.1) N/A N/A
Note: x − median; RR − relative risk; 95% CI− 95% confidence interval; N/A − not applicable (beyond the research aim); * − having ≥ 1 symptoms; LE − late enophthalmos; RH −
retrobulbar hemorrhage. Continuous data are listed as mean § SD (range). Categorical data are presented as number (percentage). Statistically significant P-values are indicated
in bold typeface.
3
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
postoperative (P = .99; RR, 1; 95% CI, .06 to 17.4; NNH = 68 [95% CI, resonance image (MRI), e.g. with a 47-mm microscopy coil, rather
38.3 to 254.2]). than conventional multi-slice CT (LoE 3, RG B) [23,24].
Based on the occurrence of LE, the post hoc power was 100%, sug- Our results also favor rc-EAMOWFR in terms of low complication
gesting very high probability of accepting the alternative hypothesis rates (≤ 10%, e.g. pathology of the lacrimal apparatus, superior or
and rejecting the null hypothesis, when the former is true (i.e., con- inferior oblique muscles, or medial canthal tendon [25]; subconjunc-
firming 100% chance of our results with a real effect). tival erythema, hyphema, and periorbital swelling usually subside
On the narrative review part, we identified 15 further studies, within three postoperative days [25]), high tolerability, and short
which were summarized in Table 2. patient’s hospital stay. Endoscopic viewing ameliorates the visualiza-
tion, reduction, and placement of implants to reconstruct the frac-
Discussion ture. Contrary to other authors (LoE 4-5, RG C-D) [3,25], we found
that the retrocaruncular incision did not require local anesthesia
This study is novel in using benefit-risk metrics to assess the clini- before an incision (i.e., the retrocaruncular tissue is less vascularized
cal usefulness of Meningaud and Pitak-Arnnop’s rc-EAMOWFR tech- than other orbital portions; hence, a blunt and bloodless dissection in
nique [6,7], compared to NS. Our null hypothesis was that outcomes this fatty areolar tissue plane can be done until reaching the posterior
of rc-EAMOWFR did not significantly differ from those of NS. Three ethmoid foramen or even near the orbital apex [LoE 4, RG C] [20,21]),
main findings were observed and refuted the null hypothesis. First, and this incision can close spontaneously (usually within 48 hours)
the risk of RH is statistically insignificant difference in the operated without necessity of wound sutures (which may cause caruncular or
and NS eyes (7 vs. 0 event; RR = 1). In other words, rc-EAMOWFR conjunctival granuloma [LoE 4, RG C] [25]). Ophthamologic examina-
seems not to cause postoperative RH (with the exception of patients tions and postoperative care were extensively described in our recent
with an anticoagulant who received prophylactic drainage). Second, meta-narrative review on blepharoplasty approach to orbital and
one in every 2 IMOWF eyes benefited from rc-EAMOWFR with the frontal sinus fractures (LoE 1, RG A-B) [26].
PDSÒ foil to eliminate the risk of LE due to untreated fractures This rc-EAMOWFR technique has a multitude of advantages over
(NNT = 2). In comparison to OFFs, PDSÒ foils appeared strong enough other approaches. For example, the medial brow incision is limited to
for IMOWF repair. A possible explanation is that there is no gravity the anterior and superior medial orbital wall and fails to free the
effect on the medial orbital wall. The decision to select materials for entrapped medial rectus muscle at the posteromedial orbital wall
the IMOWF is therefore likely to differ from the Jaquie ry et al. (LoE 4, because of the close proximity to the optic nerve. The eyelid crease
RG C) [16,17]’s protocol for OFF repair. Third, rc-EAMOWFR was incision, despite more cosmetically acceptable results, has the same
34 times more likely to prevent LE due to untreated fractures than to limitations as the medial brow incision. The direct medial canthal
cause RH, regardless of fracture symptoms (LHH = 34). These results approach (Lynch incision) can better gain access to the medial and
broadly support the work of other authors (Table 2). One unantici- inferomedial aspects of the orbit and may be combined with the
pated finding from Table 2 is that EAMOWFR was primarily instigated transconjunctival approach to the OFF. However, it creates external
by European OMFSs [6−8,21], but many subsequent investigations cicatrices, skin webbing, and the risk of telecanthus due to iatrogenic
were from Asian plastic and ophthalmic surgeons. This may thereby detachment of the medial canthal tendon. Furthermore, another
be an interesting issue for future research into determining why advantage of the trans-/retrocaruncular approach over the Lynch
EAMOWFR remains unpopular among international OMFSs. rc- incision is unneeded ligation of the anterior ethmoid artery. The cor-
EAMOWFR is a quick and easily-learned procedure, viz., the first onal incision suits panfacial, complex nasoorbitoethmoidal or frontal
author (P.P.) assisted 12 rc-EAMOWFR performed by the last author sinus trauma, but it costs time, and causes an external scar, and
(J.-P.M.) before self-performance. Notably, treatment protocols and higher risks of alopecia and injury to supraorbital, supratrocheal and
implant materials are different, that is, J.-P.M. originally used PDSÒ facial nerves (LoE 3−5, RG B-D) [2,3,6,9,21,25,27].
foils, LactoSorbÒ (Biomet, LC Dordrecht, the Natherlands), autologous Until now, two endoscopic techniques for IMOWFs have been
calvarial bone grafts, or titanium meshes, depending on the fracture advocated. Endoscope-assisted endonasal surgery amid EAMOWFR is
size and location, and IMOWFs were treated if they exceeded 1 cm2 commonly used by otolaryngologists. Its drawbacks, however,
in size, or they were < 1 cm2 with clinical symptoms [6,7]. In the cur- include (1) poor operative field if severe bleeding encounters, (2)
rent study, LE occurred irrespective of fracture size, location, and inability to repair large or comminuted fractures, (3) necessity of
symptoms. Hence, IMOWFs appears to benefit from surgery to pre- removal of the honeycomb structure of the ethmoid air cells, and (4)
vent this complication (when the patient receives adequate informa- difficulty in insertion of implants (LoE 3-5, RG B-D) [2,3,6,21,28].
tion and share the decision to treatment). Extirpation of ethmoid air cell septa and a larger area supported per
Another modification in this study from its originality [6,7] is the septum could weaken the support for the medial orbital wall (LoE 4,
use of PDSÒ foils in every procedure because it is thin and flexible, RG C) [1,29], increasing the risk of the repeated or unhealed IMOWF.
minimizing the risks of compressive optic nerve injury and delayed Some limitations in this study merit discussion. In spite of inclu-
inflammation [6,7,20,22]. Despite less Young’s modulus (1.5 GPa) and sion of a large cohort, this study was non-randomized a priori. The
osteoconductivity (i.e., neobone formation along old fracture planes treatment decision was made on the basis of a weak-evidence tenet
rather than the implant itself), and faster loss of strength (3−6 weeks) of the “all-rather than-none” concept, making the higher risks of clin-
and resorption (6 months), a recent meta-analysis has shown that the ical and personal equipoise [30]. However, IMOWF treatments have
use of PDSÒ foils is linked to lower likelihood of LE (3.2%), diplopia infrequently been scientifically proved in a large series (n ≥ 100).
(3%), delayed inflammation (1.2%), and frank eyelid malposition (< This may be due to a limited hospital volume, lack of orbital surgeons,
1%) than other materials such as poly(glycolide) (PGA; DexonÒ , Dub- inadequate imaging diagnosis, or even medical negligence. The find-
lin, Ireland), poly(L-lactide) (PLLA), poly(D,L-lactide) (PDLLA; Resorb ings of the present study suggest that IMOWFs be treated − preferen-
xÒ and Resorb xGÒ , KLS Martin, Jackonville, FL, USA), and poly(lac- tially endoscopically assisted − within two posttraumatic weeks.
tide)/poly(glycolide) (PLGA; RapidsorbÒ , DePuySynthes, West Ches- Another limitation lies in the fact that LE and RH events in our cohort
ter, PA, USA; DeltaÒ and InionÒ , Stryker, Kalamazoo, MI, USA; were low, which may draw non-significant results (type II statistical
LactasorbÒ , ut supra; VicrylÒ and EthisorbÒ , Ethicon, Somerville, NJ, error), e.g. the non-significant difference in RH between both groups.
USA; EndotineÒ , MicroAire, Charlottesville, VA, USA; Biosorb FXÒ , Issues that were also unaddressed in this study include (1) whether
Bionx, Tampere, Finnland) (LoE 1, RG A) [22]. One shortcoming of cefuroxime and dexamethasone could improve the outcomes (viz.,
PDSÒ foils is the technical demand for postoperative control, if cor- patient’s hospital stay in our cohort is shorter than that reported by
rective surgery is indicated, using high resolution magnetic other researchers [LoE 3, RG B] [28,31]), (2) whether the rc-
4
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
Table 2
Clinical studies relative to endoscopically “transorbital” repair of isolated medial orbital wall fractures (IMOWF). Pure endoscopic “endonasal”/“transantral” studies and
single case reports were excluded.
Authors(year; specialty; Surgical approach Sample size/fracture Implant type LoE (RG)x Main results
country of origin) characteristics
Chen et al. [27] (1999; Transconjunctival with/ 4 floor with/without Rib bone graft 4 (C) EAMOWFR is an excellent
trauma and plastic sur- without retrocaruncular IMOWFs(1 of which adjunct for exposure and
gery; China, USA) EAMOWFR) complete reduction of
herniated periorbital tis-
sue and bony reconstruc-
tion of the medial orbital
wall.
Baumann and Ewers [21] Transcaruncular 5 IMOWFs PDS foil or mandibular angle 4 (C) The transcaruncular
(2000; OMFS; Austria) bone graft approach gave a rapid
entry to IMOWFs without
a visible scar.
Mun et al. [32] (2002; plastic Transcaruncular with/with- 21 floor with/without Calvarial bone graft 4 (C) The endoscopically assisted
surgery; South Korea) out transconjunctival IMOWFs(15 of which transconjunctival
EAMOWFR) approach to IMOWFs
improves exposure of the
most posterior and supe-
rior aspects of the fracture
site, enabling more accu-
rate reduction of orbital
soft tissue and placement
of bone grafts.
Meningaud et al. [6] (2005; Retrocaruncular 5 IMOWFs PDSÒ foil(alternatively, cal- 4 (C) The retro-caruncular
OMFS; France) varial bone graft, titanium approach with adjunctive
mesh) endoscopic surgery should
be the gold standard for
posttraumatic IMOWFs.
Gauthier et al. [8] (2009; Retrocaruncular 6 IMOWFs PDSÒ foil 4 (C) Retrocaruncular approach is
otolaryngology and OMFS; a safe and effective tech-
France) nique that presents the
particular advantage of
not leaving a dysesthetic
scar. Peroperative endos-
copy allows then a better
accessibility and visibility
of the posterior third of
IMOWFs.
Han et al. [31] (2009; otolar- Transcaruncular vs. 48 IMOWFs (19 of which SilasticÒ sheet (endonasal) 3 (B) Reduction rates, LE, diplopia
yngology and plastic sur- endonasal EAMOWFR) vs. MedporeÒ porous poly- correction rates, and com-
gery; South Korea) ethylene sheet plication rates were not
(EAMOWFR) significantly different.
However, endoscopically
endonasal technique was
significantly associated
with longer operating
time (132.9 min vs. 81.6
min) and hospital stay
(21.5 days vs. 14.9 days).
Wu et al. [20] (2013; oph- Transcaruncular 93 IMOWFs near orbital MedporeÒ porous polyethyl- 4 (C) The endoscopic transcarun-
thalmology; China) apex ene sheet cular approach is useful to
repair IMOWFs near the
orbital apex.
Gerbino et al. [33] (2015; Retrocaruncular with/with- 18 floor with/without Titanium mesh plates 3 (B) Titanium mesh plates and
OMFS; Italy) out transconjunctival IMOWFs(5 of which retrocaruncular approach
EAMOWFR) are a reliable method to
obtain an accurate orbital
medial wall reconstruc-
tion. The use of endo-
scopic assistance through
the surgical incisions
improves accuracy of
treatment allowing better
visualization of the surgi-
cal field. Navigation aided
surgery is a feasible tech-
nique especially for com-
plex orbital reconstruction
to improve predictability
and outcomes in orbital
repair.
Chung et al. [34] (2016; plas- Transcaruncular vs. 89 IMOWFs (77 of which SynporeÒ porous polyethyl- 3 (B) Complex IMOWFs extending
tic surgery; South Korea) endonasal EAMOWFR) ene implant over the posterior eth-
moid vessels are safely
(continued)
5
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
Table 2 (Continued)
Authors(year; specialty; Surgical approach Sample size/fracture Implant type LoE (RG)x Main results
country of origin) characteristics
treated by endoscopically
transcaruncular repair.
Chang et al. [35] (2017; oph- Transconjunctival and 160 floor with/without Titanium mesh plates and 4 (C) The endoscopic transcarun-
thalmology; South Korea) transcaruncular IMOWFs MedporÒ porous polyeth- cular and transconjuncti-
ylene sheet val approach is a useful
and promising technique
to repair combined medial
and inferior orbital wall
fractures.
Yang and Liao [36] (2019; Transconjunctival and 17 floor with/without Preformed titanium mesh 4 (C) This surgical technique is a
plastic surgery; China) transcaruncular IMOWFs plates and MedporÒ safe, accurate, and effec-
porous polyethylene sheet tive method for extensive
orbital floor and medial
wall fracture repair and
clearly optimizes func-
tional and aesthetic
outcomes.
Herrera Vivas et al. [37] Transconjunctival and/or 7 floor and/or IMOWFs Titanium mesh plates and/or 4 (C) Management of orbital frac-
(2020; otolaryngology and transcaruncular and/or MedporÒ porous polyeth- tures at the posterior third
OMFS; Columbia) endonasal ylene sheet of the medial wall or floor
of the orbit through a
combined transorbital and
endoscopic approach is a
safe and effective
technique.
Kim and Kim [28] (2020; Transcaruncular vs. 54 IMOWFs(14 of which SynporeÒ porous polyethyl- 3 (B) The endoscopically endo-
plastic surgery; South endonasal EAMOWFR) ene implant nasal technique showed
Korea) significantly shorter oper-
ative time (44.7 mm vs.
73.9 mm; P = .04)
Zhou et al. [38] (2021; oph- Transconjunctival and 84 floor with/without Polyester mesh plate 4 (C) The endoscopic transcon-
thalmology; China) precaruncular IMOWFs junctival inferior fornix
and precaruncular
approach is a promising
management technique
for combined orbital
medial wall and floor frac-
tures involving the infero-
medial strut.
Hong et al. [39] (in press; Transconjunctival with/ 337 blow-out fractures(floor MedporeÒ porous polyethyl- 4 (C) The use of an endoscope
ophthalmology; South without transcaruncular with/without IMOWFs) ene sheet during blowout fracture
Korea) surgery is an effective
approach to reduce post-
operative complications
due to endoscopy’s advan-
tages in clear visualization
of the fracture site during
operation.
Pitak-Arnnop et al. (present Retrocaruncular 541 IMOWFs (509 of which PDSÒ foil 3 (B) All IMOWFs should be
study; OMFS; Germany) EAMOWFR) treated with EAMOWFR
when feasible.
Note: x − levels of evidence and recommendation grades after the 2011 Oxford Centre for Evidence-Based Medicine (OCEBM)’s; OMFS − oral and maxillofacial surgery;
EAMOWFR − endoscope-assisted medial orbital wall fracture repair.
EAMOWFR is clinically significantly better than other surgical meth- produce one RH event, but prevent 34 eyes from LE due to untreated
ods, e.g. non-endoscopic surgery and/or via other incisions, and (3) fractures. Nearly 72% of untreated IMOWFs develop LE after posttrau-
whether these results can be applicable to those operated by other, matic month 9. PDSÒ foils appear suitable for all IMOWF repair, irre-
especially young or inexperienced, surgeons. Although findings from spective of fracture size and location. No definitive conclusions can,
previous studies showed that postoperative LE ranged from 5%−10% however, be posited regarding whether its efficacy is substantially bet-
and may result from orbital fat atrophy, reherniation of the orbital ter or worse than that of other techniques. Head-to-head randomized
contents, intraoperative implant malposition, or phthisis bulbi (LoE trials among the rc-EAMOWFR vs. other surgical methods would be
3-5, RG B-D) [2−4,6,22], the cause of LE in our study remains desirable to better understand any potential similarities and differen-
unknown. Lastly, the scope of this study was limited in terms of cost- ces in clinical outcomes in the IMOWF population.
benefit analysis.
Conclusions
Despite low LoE and RG similar to (or little better than) other stud- Financial disclosure
ies (Table 2: LoE 3-4; RG B-C), our data support that the rc-EAMOWFR
is a potentially beneficial treatment for IMOWFs whenever the surgery This research did not receive any specific grant from funding
is feasible. Regarding LHH, rc-EAMOWFR in every 68 IMOWFs will agencies in the public, commercial, or not-for-profit sectors.
6
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
7
P. Pitak-Arnnop, L.K. Witohendro, C. Tangmanee et al. Journal of Stomatology oral and Maxillofacial Surgery 124 (2023) 101267
and medial wall blowout fractures. J Craniofac Surg 2017;28:963–6. doi: 10.1097/ www.jscimedcentral.com/Otolaryngology/otolaryngology-7-1246.pdf. [Accessed
SCS.0000000000003429. 12 August 2022].
[36] Yang JR, Liao HT. Functional and aesthetic outcome of extensive orbital floor and [38] Zhou G, Tu Y, Yu B, Wu W. Endoscopic repair of combined orbital floor and medial
medial wall fracture via navigation and endoscope-assisted reconstruction. Ann wall fractures involving the inferomedial strut. Eye (Lond) 2021;35:2763–70. doi:
Plast Surg 2019;82:S77–85. doi: 10.1097/SAP.0000000000001700. 10.1038/s41433-020-01304-0.
[37] Herrera Vivas AJ, de Leo Baquero MM, Roa PF, Rodríguez-Ruiz MT. Benefits of
n O, [39] Hong S, Kim J, Baek S. Blowout fracture assessment based on computed tomogra-
endoscopically assisted orbital reconstruction of fractures of the floor and medial phy and endoscopy: the effectiveness of endoscopy for fracture repair. J Craniofac
wall of the orbit. Ann Otolaryngol Rhinol 2020;7:1246–50. Available at: https:// Surg 2022 in press. doi: 10.1097/SCS.0000000000008170.