Surgical Oncology 44 (2022) 101837

Contents lists available at ScienceDirect

Surgical Oncology
journal homepage: www.elsevier.com/locate/suronc

Benefit-risk appraisal of lip-split mandibular “swing” vs. transoral

approaches to posterior oral/oropharyngeal carcinomas using number
needed to treat, to harm, and likelihood to be helped or harmed
Poramate Pitak-Arnnop a, *, 1, Levyn Kay Witohendro b, 1, Chatpong Tangmanee c, 1,
Keskanya Subbalekha d, 2, Nattapong Sirintawat e, 2, Prim Auychai f, 2, Jean-Paul Meningaud g, 3,
Andreas Neff a, 3
a
Department of Oral and Craniomaxillofacial Plastic Surgery, UKGM GmbH, University Hospital Marburg, Faculty of Medicine, Philipps-University of Marburg,
Marburg, Germany
b
Faculty of Medicine, Goethe University, Frankfurt am Main, Germany
c
Department of Statistics, Chulalongkorn University Business School, Bangkok, Thailand
d
Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, Chulalongkorn University, Bangkok, Thailand
e
Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, Mahidol University, Bangkok, Thailand
f
Department of Pediatric Dentistry, Faculty of Dentistry, Chulalongkorn University, Bangkok, Thailand
g
Department of Plastic, Reconstructive, Esthetic and Maxillofacial Surgery, Henri Mondor University Hospital, AP-HP, Faculty of Medicine, University Paris-Est Créteil
Val de Marne (Paris XII), Créteil, France

A R T I C L E I N F O A B S T R A C T

Keywords: Purpose: s: To evaluate benefit-risk profiles of lip-split mandibular “swing” vs. transoral approaches (LS-MSA;
Surgical approach TOA) to the American Joint Committee on Cancer (AJCC) stage I-III posterior oral/oropharyngeal carcinomas
Oral cancer (PO/OPC).
Oropharyngeal cancer
Methods: Using a retrospective double-cohort study design, we enrolled stage I-III PO/OPC patients treated in two
Benefit-risk profile
German medical centers during a 4-year interval. The predictor variable was surgical technique (LS-MSA/TOA),
and main outcomes were complete resection with R0 margins (CR-R0), 5-year overall survival and recurrence
(OS5; R5), and adverse events (AEs). Descriptive and bivariate statistics were computed with α = 95%. Benefit-
risk profiles were investigated using number needed to treat (NNT), to harm (NNH), and likelihood to be helped
or harmed (LLH).
Results: At 5-year follow-ups of 202 subjects, LS-MSA caused significantly better CR-R0 (P = 0.001; NNT: 4) and
fewer R5 (P = 0.003; NNT: 5), but more risks of wound dehiscence ([WD]; P = 0.01; NNH = 8), and orocutaneous
fistula ([OCF]; P = 0.01; NNH: 10). LLH calculations demonstrated that LS-MSA was 2 and 1.6 times more likely
to result in CR-R0 and fewer R5 than an incident of WD. There was no significant difference in OS5, postoperative
infections (within 30 postoperative days) and AE domains according to the University of Washington Quality of
Life questionnaire version 4 (UW-QoLv4) between the surgical approach groups.
Conclusions: Compared to TOA, LS-MSA is an efficacious and tolerable intervention for inspecting and eradicating
stage I-III PO/OPCs, and reducing recurrences at 5-year follow-ups. Post-LS-MSA WD and OCF require meticulous
concerns and more investigations.

1. Introduction

Optimal management of oral/oropharyngeal carcinomas continues

* Corresponding author. Klinik für MKG-Chirurgie, Universitätsklinikum Marburg, UKGM, Baldingerstr, 35043, Marburg, Germany.
E-mail address: [email protected] (P. Pitak-Arnnop).
1
Equal contribution.
2
Equal contribution.
3
Equal contribution.

https://doi.org/10.1016/j.suronc.2022.101837
Received 9 May 2022; Received in revised form 18 July 2022; Accepted 8 August 2022
Available online 15 August 2022
0960-7404/© 2022 Elsevier Ltd. All rights reserved.
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

to challenge clinicians. Issues to be addressed include complete tumor 2. Materials and methods
resection, patients’ survival and quality of life, and the risk of devel­

Abbreviations OS5 5-year overall survival

PI Postoperative infection
AE Adverse event PO/OPC Posterior oral/oropharyngeal carcinomas
AJCC The American Joint Committee on Cancer pTNM Pathological tumor-node-metastasis staging
CI Confidence interval RR Relative risk
CR-R0 Complete resection with R0 margins in one session R0cm Excision with close margin
FDA The US Food and Drug Administration R0hr Excision with high-risk margin
FSB Frozen section biopsy R1 Incomplete excision
HPV Human papillomavirus R5 5-year recurrence
LHH Likelihood to be helped or harmed STROBE The STrengthening the Reporting of OBservational studies
LS-MSA Lip-split mandibular swing approach in Epidemiology (guidelines)
ND Neck dissection TOA Transoral approach
NNT Number needed to treat TORS Transoral robotic surgery
NNH Number needed to harm UW-QoLv4 The University of Washington Quality of Life
OCEBM The Oxford Centre for Evidence-Based Medicine questionnaire version 4
OCF Orocutaneous fistula WD Wound dehiscence

oping adverse events (AEs). Upfront surgery is the mainstay of treatment 2.1. Study design and sample description
for posterior oral/oropharyngeal carcinomas (PO/OPCs) [1–3], with the
exception of unresectable tumors such as those with gross extension to To address research objectives, the authors designed and imple­
the superior nasopharynx, skull base, prevertebral fascia, cervical mented a retrospective cohort study. The study cohort was derived from
vertebrae, mediastinum, or common or internal carotid artery, presence the populations of patients presenting to the Oral and Craniomax­
of subdermal metastasis, or involvement of the pterygoid muscles illofacial Unit in two German academic teaching hospitals (where the
associated with trismus or cranial nerve neuropathy [3]. Oral and Craniomaxillofacial Unit is the only department responsible for
Since Roux [4] first described the lip-split mandibular “swing” head and neck cancer patient care) during a 4-year period for evaluation
approach (LS-MSA) to PO/OPCs in 1836, reports consisting of a larger and management of PO/OPCs. This chart reviewing projects were
cohort and/or technical modifications have incrementally been docu­ approved by the bi-institutional review boards. The Helsinki Declara­
mented in the literature. On the other side, “inside-out” transoral robotic tion’s ethical guidelines and the STROBE statement were adhered
surgery (TORS, e.g. da Vinci®, Intuitive Surgical Inc., CA, USA) throughout the study. All subjects gave written consent to the use of
commenced in 2005 has become popular among otolaryngologists their anonymous data in future research.
hitherto. The US Food and Drug Administration (FDA), however, Subjects aged ≥18 years were eligible for study inclusion if they had
approved TORS only for resection of T1-T2 cancers of the oropharynx, a PO/OPC classified as the American Joint Committee on Cancer (AJCC)
larynx, and hypopharynx [5]. Traditional “outside-in” approaches such stage I-III. Patients were excluded if they 1) denied in participation, 2)
as transoral approach (TOA), LS-MSA, and Weber-Ferguson incision were treated in alio loco (and came only for follow-ups), 3) received
remains an essential step for oral oncologic surgery [1,2,6]. segmental mandibulectomy, myotomy of masticator muscles, coronoi­
Evidence-based data suggest that LS-MSA be appropriate for PO/OPCs dectomy, or temporomandibular joint disarticulation, 4) presented with
due to its excellent exposure to the tumor, and subsequently resulting in uncontrolled comorbid disease and/or metastatic disease with short life
resection with clear surgical margins. Nevertheless, mandibulotomy expectancy, or 5) had records unavailable for review.
may cause many AEs, such as hardware exposure, osteonecrosis, oro­
cutaneous fistula (OCF), and physical disfigurement [1]. To the best of
our knowledge, the benefit-risk profile of LS-MSA to PO/OPCs has never 2.2. Study variables
been investigated before.
This study was set out to answer the following clinical research The predictor variable was technique group (LS-MSA vs. TOA).
question: “Among PO/OPC patients, is LS-MSA better than TOA in terms Management of all PO/OPCs in this cohort were surgically dominant,
of complete resection with R0 margins (CR-R0), 5-year overall survival regardless of presence of human papillomavirus (HPV) because this
and recurrence (OS5; R5), and AEs?” The authors tested the null hy­ virus had no effect on treatment outcomes in some German populations
pothesis that in patients treated for PO/OPCs, there would be no dif­ [7]. Nonetheless, the decision to choose the surgical technique depends
ferences in CR-R0, OS5, R5 and AEs in patients approached via LS-MSA mainly on operator and patient factors, e.g. LS-MSA for large-sized
and those received TOA. The study’s specific aims were to 1) quantify and/or more posterior lesions under patient agreement. After
CR-R0, OS5, R5 and AEs following treatment options for each PO/OPC, 2) tumor-nodes-metastasis staging and panendoscopic evaluation, treat­
identify the relation between clinical variables and three AEs: wound ment planning for an individual patient was accomplished via an
dehiscence (WD), OCF, and postoperative infections (PIs), and 3) interdisciplinary tumor board, which (in Germany) includes max­
compare the findings using bivariate statistics and the metrics of number illofacial/head and neck surgeons, otolaryngologists, diagnostic radiol­
needed to treat (NNT), to harm (NNH), and likelihood to be helped or ogists, pathologists, medical oncologists, radiotherapists, and if
harmed (LHH). indicated, speech therapists, and oncology social workers and study
nurses.
As described by Holsinger et al. [8], and Na et al. [9], LS-MSA (or
“transmandibular” or “(para)median labio-mandibulo-glossotomy”
approach) in this study comprised 1) McGregor and McDonald [10]’s
technique for lower lip splitting, i.e. straight midline chin-contour

2
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

incision, before going to connect with the cervical incision of neck guideline [14,15]), OS5, R5, and ≥50% of each AE item according to the
dissection (ND), 2) precontouring 3D miniplates (or standard mini­ University of Washington Quality of Life questionnaire version 4
plates; see the Discussion below) and preparing screw holes along the (UW-QoLv4) for head and neck cancer patients (except [ND-associated]
Champy’s line of mandibular osteosynthesis [11], to hasten bone reap­ shoulder movement) [16] plus four more items: WD, OCF, PIs, and
proximation and fixation later, 3) stair-step (if possible, “paramedian” to paresthesia ≥6 months. All AEs including WD and OCF were observed
avoid releasing the suprahyoid and extrinsic tongue musculature from up to 6 postoperative months, while the term “PI” has been defined by
the genial tubercles and digastrics fossa of the mandible in order to the US Centers for Disease Control and Prevention (CDC) as an infection
preserve swallowing function [1,12]) “cut-through” osteotomy using at the incision site and/or deeper underlying tissue spaces and organs
piezoelectric cutting instruments [13], or less favorably, a thin-blade within 30 days of a surgical procedure [17].
oscillator, 4) lateralization of both mandibular segments, and 5) To stratify the resection margins (CR-R0: yes/no), the term “non-CR-
extending the incision to the tumor (see a case example in Fig. 1). TOA R0” referred to 1) close margin (R0cm; complete excision with a clear
(or “extended radical tonsillectomy” approach) is analogue to the margin between 1 and 4 mm), 2) high-risk margin (R0hr; complete
LS-MSA, but oral tissue would be retracted or only pulled out without lip excision with <1 mm margins), and 3) incomplete excision (R1) with
split and mandibulotomy [8]. microscopical tumor infiltration. All of these three resection margin
Frozen section biopsies (FSBs) were transorally obtained before ND, types (R0cm, R0hr, and R1) have been found to carry the same recurrence
and the FSB results had to be all negative margins prior to tumor risk [15].
resection. The resection requires minimal distance of 1 cm from the AEs related to adjunct procedures, such as tracheostomy (e.g.
palpable tumor margin because formalin fixation and slide preparation dysphonia, dysphagia, tracheoesophageal or tracheocutaneous fistula,
reduce the mucous margin by approximately 30%–50%. Hence, a final airway obstruction due to stenosis or mucous plugging, pneumonia)
pathological tumor margin of ca. 5 mm should be considered as surgeon- [12], or ND (e.g. accessory nerve injury, seroma, chylous fistula, carotid
measured margins of 1 cm [1,14]. The concept of prophylactic ND was sinus syndrome or rupture, pneumothorax) [18] were excluded.
applied as indicated and as preoperatively recommended by the tumor Other study variables were grouped into 1) demographic (age;
board. gender), and 2) clinical (comorbid disease relevant to OS5 and/or wound
A nasogastric tube was inserted for a 7-day course of “nil per os” complications; preoperative serum albumin <4 g/dL) categories. We did
liquid or purée diets. The patient was tracheostomized, if the composite not use the body-mass index as a study variable because of a number of
resection involved a large area of the oral floor (especially between missing data with regard to patients’ body weight monitoring in the
mental foramina) and/or bilateral ND and flap reconstruction [12]. To hospital registry. On the contrary, preoperative serum albumin was used
control the edema in the operated neck and subsequent airway instead. Albumin is a prototypical indicator for both synthetic liver
compromise, bilateral ND patients received dexamethasone 8 mg pre­ function and overall protein nutritional status. It has a longer half-life
operative [during anesthetic induction] and immediately post­ (as compared to other acute phase proteins such as transferring, pre­
operatively and every 6 h for a maximum 72 h from the first dose), albumin, and C-reactive protein) and the value is less affected by non-
together with a suction drain at each dissected neck side. This schema nutritional preoperative issues such as inflammatory- or cancer-related
was also proved by other authors [12]. changes [19]. In a recent study with an adjusted model, oral cancer
The outcome variables represented various dichotomous outcome patients with preoperative albumin level of <4 g/dL had 3.2 times
measures: CR-R0 in pathological report on the resected specimens (i.e. higher odds of WD (95% confidence interval [CI]: 1.4 to 7.4) in com­
microscopically tumor-free margins of ≥5 mm according to the German parison with those with albumin ≥4 g/dL [19]. We, therefore, used the

Fig. 1. Photographs showing LS-MSA to a left tonsillar and oropharyngeal carcinoma (pT3N0M0): (A) preoperative marking (violet arrow: left mandibular angle; red
arrow: inferior mandibular border; green arrow: cervical incision for ND; yellow arrow: paramedian McGregor and McDonald [10]’s incision line; pink star: The
paramedian incision joins the cervical incision.), (B) after lip splitting (blue star: stepped osteotomy line; yellow arrow: mental foramen with the dissected mental
nerve; The precurved miniplates and their screw holes should be performed before the osteotomy. In this case, the tumor located deeply on the left oropharyngeal
side; hence, the paramedian osteomy was moved to 4–5 mm anterior to the mental foramen to gain a wider access to the left oropharyngeal side and to avoid nerve
injury because up to 90% of humans have the anterior loop of mental foramen. [37]), (C) during tumor resection (S: tumor specimen; M: mandibular segments), (D)
before completing immediate reconstruction with a facial artery musculomucosal (FAMM) flap [38] (F: FAMM flap; T: tongue; M: mandible), (E) Intraoral suturing
should start from the flap to the osteotomy site, or from posteriorly to anteriorly, before osteosynthesis of the mandible at the osteotomy line (blue arrow: mental
foramen; The mental nerve was relocated back to its place. One important remark is that bone cutting with an oscillator often creates a small bone gap, which could
be minimized by using piezosurgery. [13]), and (F) immediate postoperative condition (D: a negative pressure drain).

3
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

albumin cut-off value at 4 g/dL for patients’ malnourished conditions in R5 (at 5-year follow-up), and AEs (within 6-month follow-ups, except PIs
this study. until 30 postoperative days).
It is beyond the scope of this study to examine deep dimensions of
dental complications (e.g. tooth loss, and periodontal disease associated
2.3. Data management and statistical analysis
with LS-MSA), survival analysis and effects of radio(chemo)therapy (i.e.
late sequelae such as osteoradionecrosis) with different pathological risk
To avoid study bias due to the primary author (P.P.) as the main
factors (see the Discussion below). This work was mainly engaged with
operator, the second author (L-K.W.) extracted most data from the
the benefit-risk profile of LS-MSA by using TOA as the reference with a
hospital database, and recorded it anonymously in a collection form of
special focus on CR-R0 (immediate outcome from the operating theater),
Microsoft Excel 2007 (Microsoft Inc., WA, USA). All analyses were

Table 1
Cohort characteristics and analyses grouped by surgical approaches to posterior oral/oropharyngeal carcinomas at 5 years of follow-up.
Parameter Overall LS-MSA TOA P value (RR; 95% CI) AR in % (95% CI) NNT or NNH (95%
CI)

Demographic
Sample size 202 (100) 71 (35.1) 131 (64.9) N/A N/A N/A
Average age 58.6 ± 10.7 57.9 ± 12.3 52.8 ± 20.4 0.52 (N/A; − 2.24 to N/A N/A
(27–86) (27–79) (41–86) 4.44)
Female gender 59 (29.2) 22 (31) 37 (28.2) 0.96 (1.01; 0.65 to 1.58) N/A N/A
Clinical
Smoking 72 (35.6) 27 (38) 45 (34.4) 0.6 (1.1; 0.76 to 1.62) N/A N/A
History of smoking 118 (58.4) 43 (60.6) 75 (57.3) 0.64 (1.06; 0.83 to 1.34) N/A N/A
Alcohol addition 17 (8.4) 8 (11.3) 9 (6.9) 0.29 (1.64; 0.66 to 4.06) N/A N/A
Diabetes mellitus 25 (12.4) 10 (14.1) 15 (11.5) 0.59 (1.23; 0.58 to 2.59) N/A N/A
Cardiovascular disease including 63 (31.2) 29 (40.8) 34 (26) 0.03 (1.57; 1.05 to N/A N/A
hypertension 2.35)
Pulmonary disease 6 (3) 1 (1.4) 3 (2.3) 0.67 (0.62; 0.07 to 5.8) N/A N/A
Preoperative serum albumin <4 d/dL 41 (20.3) 16 (22.5) 25 (19.1) 0.56 (1.18; 0.68 to 2.06) N/A N/A
T3 N0 88 (43.6) 49 (69) 39 (29.8) < 0.0001 (2.32; 1.71 to N/A N/A
3.15)
Any T N1 51 (25.2) 19 (26.8) 32 (24.4) 0.71 (1.1; 0.67 to 1.79) N/A N/A
Outcome:
CR-R0 161 (79.7) 69 (97.2) 92 (70.2) 0.001 (0.095; 0.024 to (− )26.95 4 (2.8–5.5)
0.38) (18.23–35.68)
OS5 152 (75.2) 54 (76.1) 98 (74.8) 0.84 (0.95; 0.57 to 1.58) (+)1.25 (− 11.15 to 81 (7.3–9.0)
13.65)
R5 43 (21.3) 8 (11.3) 43 (32.8) 0.0026 (0.34; 0.17 to (− )21.56 5 (3.1–9.4)
0.69) (10.66–32.45)
AEs ≥50%
Pain (VAS ≥5/10, or ≥ 50%) 17 (8.4) 7 (9.9) 10 (7.6) 0.59 (1.29; 0.51 to 3.24) (+)2.23 (− 6.07 to 45 (9.5–16.5)
10.52)
Physical appearance <50% 4 (2) 3 (4.2) 1 (0.8) 0.14 (5.54; 0.59 to (+)3.46 (− 1.45 to 29 (11.9–69)
52.24) 8.37)
Physical activities <50% 7 (3.5) 5 (7) 2 (1.5) 0.7 (0.83; 0.3 to 2.26) (+)5.52 (− 0.8 to 19 (8.5–125.7)
11.83)
Recreation <50% 23 (11.4) 8 (11.3) 15 (11.5) 0.97 (0.98; 0.44 to 2.21) (− )0.18 (− 8.97 to 548 (10.7–11.1)
9.34)
Swallowing <50% 42 (20.8) 18 (25.4) 24 (18.3) 0.24 (1.38; 0.81 to 2.37) (+)7.03 (− 5.06 to 15 (5.2–19.8)
19.13)
Chewing <50% 37 (18.3) 12 (16.9) 15 (11.5) 0.28 (1.48; 0.73 to 2.98) (+)5.45 (− 4.83 to 19 (6.4–20.7)
15.73)
Speech <50% 27 (13.4) 10 (14.1) 17 (13) 0.83 (1.09; 0.53 to 2.24) (+)1.11 (− 8.82 to 91 (9.1–11.3)
11.04)
Taste <50% 39 (19.3) 17 (23.9) 22 (16.8) 0.22 (1.42; 0.81 to 2.5) (+)7.15 (− 4.66 to 14 (5.3–21.5)
18.96)
Salivary secretion <50% 25 (12.4) 9 (12.7) 16 (12.2) 0.92 (1.04; 0.48 to 2.23) (+)0.46 (− 9.09 to 217 (10–11)
10.2)
Mood <50% 18 (8.9) 8 (11.3) 10 (7.6) 0.39 (1.48; 0.61 to 3.57) (+)3.63 (− 5.01 to 28 (8.1–19.9)
12.28)
Anxiety ≥50% 22 (10.9) 8 (11.3) 14 (10.7) 0.9 (1.05; 0.46 to 2.39) (+)0.58 (− 8.48 to 173 (10.4–11.8)
9.64)
WD 31 (15.3) 17 (23.9) 14 (10.7) 0.01 (2.24; 1.17 to (+)13.26 (2.01–24.5) 8 (4.1–49.8)
4.27)
OCF 12 (5.9) 9 (12.7) 3 (2.3) 0.0085 (5.54; 1.55 to (+)10.39 (2.23–18.54) 10 (5.4–44.8)
19.8)
PIs 32 (15.8) 10 (14.1) 12 (9.2) 0.28 (1.54; 0.7 to 3.38) (+)4.92 (− 4.56 to 21 (6.9–21.9)
14.4)
Lip paresthesia ≥6 months 3 (1.5) 3 (4.2) 0 0.09 (12.83; 0.67 to (+)4.23 (− 0.45 to 8.9) 24 (11.2–220.3)
245.02)

Note: LS-MSA – lip split with mandibulotomy swing approach; TOA – transoral approach; RR – relative risk; 95% CI– 95% confidence interval; AR – absolute risk ((+) –
absolute risk increase; (− ) – absolute risk reduction): NNT – number needed to treat; NNH – number needed to harm; N/A – not applicable; CR-R0 – complete resection
with R0–margin; OS5 – 5-year overall survival; R5 – 5-year recurrence; AEs – adverse events; VAS – visual analogue scale; WD – wound dehiscence; OCF – orocutaneous
fistula; PI – postoperative infections.
Continuous data are listed as mean ± SD (range). Categorical data are presented as number (percentage). Statistically significant P-values are indicated in bold
typeface.

4
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

blinded and performed using MedCalc® (MedCalc Software Ltd., suggest the use of TOA only for small, anterior-locating, and easily
Ostend, Belgium). Frequency and descriptive statistics were computed accessible tumors, while LS-MSA suits advanced or deeply infiltrating
to describe the sample characteristics of each study variable. We and/or very posterior-locating PO/OPCs, and/or in patients with
calculated the metrics of NNT, NNH, and LHH based on the observed trismus and/or dentition interfering access to the tumor [1,3]. One
outcomes in relation to the predictor (LS-MSA vs. TOA). Bivariate ana­ recent systematic review of 54 studies (n = 3872) concluded that
lyses were performed to assess the association between clinical param­ LS-MSA was safe and its complication rates were acceptable [20].
eters with outcomes of interest, and confirmed by a multiple logistic Conversely, another meta-analysis of six studies (n = 309) discarded
regression model. In all analyses, significance was defined as P < 0.05 differences in surgical margins, overall survival, total and local re­
with one-sided hypothesis testing. The post hoc power were computed currences, and functional outcomes including speech between LS-MSA
using G Power 3 for Windows (HHU Düsseldorf, Düsseldorf, Germany) and the mandibular-sparing visor flap, but LS-MSA owned a signifi­
with an effect size of 0.5, α error probability of .05, and a sample size of cantly higher rate of OCF [21]. In the primary author (P.P.)’s practice,
202. the visor incision was not favored because ND is performed after FSB and
before tumor resection via LS-MSA or TOA, i.e. specimens of FSB, ND,
3. Results and the tumor are separately collected. Time span during ND is in most
cases enough for obtaining FSB findings from the pathologist. Besides,
The study sample included 202 subjects (29.2% females; 100% previous evidence showed that patients in the LS-MSA group scored
squamous cell carcinomas; 35.1% with LS-MSA; 79.7% treated by the significantly better speech, swallowing, and chewing than those with the
primary author [P.P.]) who met the inclusion criteria for the study visor approach [20]. Neither those abovementioned guidelines [1,3] nor
cohort. The cohort’s average age was 58.6 ± 10.7 years (range, 27–86). both review papers [20,21] included benefit-risk appraisal (i.e. relying
More than 80% of the subjects in each treatment arm had an ASA only on pooled percentages of good vs. bad outcomes). Our study
(American Society of Anesthesiologists) status of level I or II. Neither therefore sought to apply the metrics of NNT, NNH, and LHH for eval­
postoperative intermaxillary fixation nor secondary procedures was uating benefits and risks of LS-MSA to PO/OPCs, compared to TOA. We
necessary. There was no significant difference in radio(chemo)therapy examined the null hypothesis that there would be no differences in
between both groups (LS-MSA vs. TOA: 91.5% vs. 85.5; P = 0.18; 95% CR-R0, OS5, R5 and AEs when using LS-MSA vs. TOA to PO/OPCs.
CI: 0.97 to 1.18). The results of this study rejected the null hypothesis. The following
Among demographic and clinical parameters, cardiovascular disease key findings can be drawn from the present study. First, LS-MSA at 5
including hypertension (P = 0.03; relative risk [RR]: 1.6; 95% CI: 1.1 to follow-up years resulted in statistically significantly higher CR-R0 and
2.4) and T3 tumor size (P < 0.0001; RR: 2.3; 95% CI: 1.7 to 3.2) were lower R5, albeit substantially risky to WD and OCF. It is desirable that
significantly different between the two groups. Binary analyses further NNH be larger than NNT, so that benefits are encountered more
demonstrated statistically significant differences in four other parame­ frequently than harms [22]. All analyses on AEs (except WD) revealed
ters between the two groups: CR-R0 (P = 0.001; RR: 0.1; 95% CI: 0.02 to >9 NNH, meaning that no more than a 10% disadvantage of LS-MSA
0.4), R5 (P = 0.003; RR: 0.3; 95% CI: 0.2 to 0.7), and developing WD (P with respect to potential harms. LHH calculations demonstrated that
= 0.01; RR: 2.2; 95% CI: 1.2 to 4.3) and OCF (P = 0.09; RR: 5.5; 95% CI: LS-MSA was 2 and 1.6 times more likely to result in CR-R0 and fewer R5
1.6 to 19.8). At 5 years of follow-up, one of every four patients benefited than an incident of WD. These outcomes are consistent with a recent
from LS-MSA because of CR-R0 at the first surgery (NNT = 4; 95% CI: 2.8 large series [23]’s results (n = 224). Another recent study (n = 753) by
to 5.5). LS-MSA was helpful for preventing R5 in comparison to TOA Hakim et al. [15] reported a significant correlation between oral cancer
(NNT = 5; 95% CI: 3.1 to 9.4), but it was harmful because of more risks localization and the resection-free margin. In other words, local control
of PIs (NNH = 21; 95% CI: 6.9 to 21.9). One of every 8 and 10 LS-MSA of disease is compromised when TOA is used. Posterior and floor of the
patients would experience WD (NNH = 8; 95% CI: 4.1 to 49.8) and OCF mouth ends of the tumor cannot be easily accessed especially when the
(NNH = 10; 95% CI: 5.4 to 44.8), respectively. LHH calculations showed dentition is intact [23].
that patients undergoing LS-MSA had a 2- and 1.6-time greater chance of Second, binary analyses excluded significant differences in the “UW-
CR-R0 and fewer R5 than likelihood to develop WD. Table 1 presents the QoLv4” AE domains between treatment arms. There are two likely
overview of statistical analyses. causes for futility of this questionnaire in this study: 1) small discrep­
Insignificant differences in UW-QoLv4 between treatment arms ancies in complications, which would required the sample size of
could reject the poorer tolerability of LS-MSA, when compared to TOA >1000/experiment arm and consequently rend such a trial extremely
(P > 0.05). In other words, this result raises the possibility that UW- difficult or practically impossible [20], and 2) our treatment goals, i.e.
QoLv4 should not be used to differentiate AEs after LS-MSA vs. TOA. structural restoration, functional recreation, and esthetic improvement
Most of the LS-MSA patients had temporary paresthesia around the (in agreement with other authors [1,24]), are achieved as fast as possible
mentum for a few postoperative months; 6 of those (or 8.5%) experi­ such as immediate reconstruction with locoregional flaps or free issue
enced mental paresthesia more than 6 months, which spontaneously transfer. This also includes our use of the McGregor and McDonald
resolved within the first postoperative year (P = 0.09; RR: 12.8; 95% CI: [10]’s straight midline chin-contour incision, which modifies the Roux
0.7 to 245). In this cohort, neither damages to motor (i.e. facial and [4]’s incision by following the labiomental groove to hide the scar, and
hypoglossal) nerves nor salivary gland injuries were observed. thereby reduce postsurgical physical disfigurement and prolonged
The post hoc analysis showed a power of 78.6%, suggesting high paresthesia [25]. The absence of hardware exposure and osteonecrosis
probability of accepting the alternative hypothesis and rejecting the null in our patients is explained by the exclusion criteria (i.e. segmental
hypothesis, when the former is true. mandibulectomies were excluded, and follow-ups of AEs were limited to
a 6-month interval). Both complications are often exacerbated, if a
4. Discussion mandibular rim resection is anticipated [20]. In this cohort, hardware
was removed at 6–12 postoperative months. Hardware removal in the
Upfront surgery remains the workhorse of oral cancer care (at least in atrophy mandible can be performed in local anesthesia, if surgery under
German-speaking countries), as stated by the German S3-guideline on general anesthesia posed a high risk. It should also be borne in mind that
oral cancer management (version 3.01, 2019; valid until 2023) that the discussions on post-LS-MSA bony complications in the literature are
“HPV-positive and/or p16-positive oral carcinomas should be treated in often overrated, especially when a study or review includes archaic
a similar fashion to alcohol- and nicotine-associated oral carcinomas” papers explaining wiring for postresection mandibular fixation [20,21,
[14]. There is still uncertainty, however, whether we should continue 26].
using LS-MSA to PO/OPCs. Practice guidelines from various countries When taken together, it is therefore no longer a moot point whether

5
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

tolerability of LS-MSA and TOA is comparable, and LS-MSA is highly reduce masticatory functions much more than in trauma patients [31].
associated with WD and subsequent OCF, i.e. both doctrines are true. In Besides, the average ages of patients in the meta-analyses were 25–35
contrast to the Fahmy et al. [19]’s study, however, no association be­ years, but our cohort’s patients had a mean age of ~59 years. With age,
tween preoperative serum albumin and WD was detected in our study. the chewing force appears reduced [32,33]. It is, therefore, noteworthy
This result could provide support for the hypothesis that a strong link that – in spite of using load-sharing miniplates in most cases (Fig. 1E) –
exists between serum albumin and any surgical approaches other than the evidence of hardware failure and malocclusion was nil in our cohort,
LS-MSA (Fahmy et al. [19] did not mention the surgical approach they while postoperative trismus was unstudied because it was intensified by
used). Mehanna et al. [27] believed that the vertical lip incision through postoperative radiotherapy. This may also imply that posttraumatic data
the mentalis ensures that minimum length of muscle is divided, which from the recent meta-analyses [29,30] supporting the use of 3D mini­
reduces the inherent amount of muscle trauma and long-term fibrosis. plates may not be applicable to post-LS-MSA oncologic patients (i.e.
They also found a relationship between the straight lip incision and a weak “external validity” or “generalizability”), and that miniplates
serious breakdown of the wound or formation of an OCF. The LS-MSA might be used safely in LS-MSA patients, in case of 1) treatment by a
requires the placement of a trifurcation incision in the oral vestibule. surgeon with good background on dental occlusion and oral/jaw anat­
Our results corroborate the idea of Cilento et al. [28] who suggested that omy, 2) precontouring the plates and preparation of the screw holes, 3)
(almost) all of the fistulae developed at this trifurcation and more often osteotomy using a piezotome (or a very thin-blade saw), and most
in post-irradiated tissues due to poor blood circulation and wound preferably, 4) no postoperative irradiation. Our literature search in
healing. One of our tricks is that the trifurcation and the jaw bone PubMed/Medline, Embase, Cochrane Library, and Google Scholar until
osteotomy site should not be at the same place. May 4, 2022 cannot detect any relevant study on the use of 3D mini­
Third, what is also clear from our data is that 97.2% and 85.5% of plates vs. miniplates in LS-MSA patients. Further studies on this matter
patients in the LS-MSA and TOA groups achieved CR-R0 with clear FSB need to be undertaken.
margins. In other words, FSBs could almost 100% predict CR-R0, The findings from this study make several contributions to the cur­
regardless of surgical methods (posterior probability of positive test: rent surgical oncologic literature. Although survival is the primary end
100% [95% CI: 90%–100%], and 100% [95% CI: 94%–100%]; posterior point of oral/oropharyngeal cancer patient care, data such as symptoms,
probability of negative test: 3% [95% CI: 1%–11%] and 7% [95% CI: psychological adjustment, a degree of functional deficits, and post­
4%–12%]; accuracy: 98.6% [95% CI: 95%–99.8%] and 96.2% [95% CI: operative physical appearance (which was recognized as a more com­
93.1%–98.2%] in LS-MSA and TOA groups, respectively) (Table 2), fa­ plete picture of patient outcomes) were analyzed in this investigation
voring the use of LS-MSA in real-life situations. This outcome is contrary [28]. The relatively large sample size and long follow-ups, and homo­
to some practice guidelines [1,3] that could not make a conclusion on geneity of treatments (mostly by a single surgeon [P.P.]) represent the
the role of FSBs in oral cancer surgery. strengths of this study. An interesting concept, as recently reported by
Fourth, concerning the use of osteosynthesis materials at man­ Tay et al. [34] and Bestourous et al. [36], is that once the PO/OPC
dibulotomy sites (using e.g. piezotomes or saws), it needs to be deeply invades the pharynx or even larynx, the LS-MSA can be used
considered that any osteosynthesis stabilizing the osteotomy site is together with TORS in the same operation. TOR alone appears ineffec­
primarily load-bearing and that the use of 3D (preferably angular stable tive or otherwise contraindicated in patients with anatomical barriers
systems) or reconstruction plates is strongly recommended. In cranio­ such as retrognathic mandible, macroglossia, microstomia, and (post­
maxillofacial traumatology, where bony interfragment contact usually radiation) trismus. A combination of LS-MSA and TOR can better afford
allows for a load-sharing osteosynthesis, recent meta-analyses revealed the surgeon a wider access, while sparing patients the esthetic and
the significant superiority of 3D miniplates over standard miniplates in functional morbidities associated with completely open procedures [34,
terms of hardware failure, malocclusion, and postoperative trismus, 35].
despite no significant differences in the incidence of postoperative The shortcomings of the present study, nevertheless, include its
infection, wound dehiscence, non-/malunion, and paresthesia [29,30]. retrospective study design with the lack of randomization, which can be
However, such meta-analyses had included only fracture repair studies deviated by selection and observation biases (e.g. due to that fact that
that patients were mostly injured only at the fracture site, while subjects ~80% of the cases were operated by the first author) and/or susceptible
in our study were operated on several oral and cervical regions. In to measurement error (including complications of LS-MSA after post­
oncologic cases, postoperative pain and scarring (such as due to exten­ operative month 6). A solution we used is the data collection by an
sive mucoperiosteal stripping, and huge incisions) are very likely to uninvolved (external) author (L-K.W.). Patient follow-ups were per­
formed by residents/fellows, making the inter-observer disagreement
Table 2 possible. Last but not least, our study can be at high risk of systematic
Comparison of measurement of diagnostic test. bias in assessment and judgment of the tumor borders and margins (e.g.
postresection shrinkage, processing, and inadequate orientation of the
Diagnostic test Frozen section biopsy Frozen section biopsy
before LS-MSA (value; 95% before TOA (value; 95% tumor specimen) [15]. This bias type may emerge from irregular and
CI) CI) complex shape and localization of the tumor [15]. Our findings are also
Sensitivity 97.18% (90.19%–99.66%) 92.37% (86.41%–
somewhat limited by pathological risk conditions such as pTNM, infil­
96.28%) tration depth, perineural invasion, or precancerous changes, which can
Specificity 100% (94.94%–100%) 100% (97.22%–100%) all worsen the local tumor control and survival outcomes. Because most
Negative Likelihood 0.03 (0.01–0.11) 0.08 (0.04–0.14) of the patients in this study received additional radio(chemo)therapy,
Ratio
the effect of radio(chemo)therapy on AEs (including WD, OCF, and PIs)
Positive Predictive 100% (N/A) 100% (N/A)
Value (PPV) was unjustified in this study (i.e. the sample size of
Negative Predictive 97.26% (90.05%–99.29%) 92.91% (87.84%– non-irradiated/non-chemotherapy patients is too small to overcome the
Value (NPV) 95.96%) β-error, and thereby, could mislead readers about the insignificance).
Accuracy 98.59% (95.00%–99.83%) 96.18% (93.09%– Apart from the role of reconstructive surgery and radio(chemo)therapy,
98.15%)
Posterior probability 100% (90% to 100%) 100% (94% to 100%)
it is unclear which component of these procedures is most significant or
of positive test whether the combination of the surgical procedures predisposes the
Posterior probability 3% (1% to 11%) 7% (4% to 12%) patient to AEs, i.e. the exact cause and mechanism of AEs especially high
of negative test WD and OCF following LS-MSA, are intriguing and could be usefully
Note: LS-MSA – lip split with mandibulotomy swing approach; TOA – transoral explored in future researches.
approach; 95% CI– 95% confidence interval; N/A – not applicable.

6
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

5. Conclusions [8] F.C. Holsinger, O. Laccourreye, R.S. Weber, Surgical approaches for cancer of the
oropharynx, Operat. Tech. Otolaryngol. Head Neck Surg. 16 (2005) 40–48, https://
doi.org/10.1016/j.otot.2005.02.001.
This study was the first look to identify NNT, NNH, and LHH of LS- [9] H.Y. Na, E.J. Choi, E.C. Choi, et al., Modified mandibulotomy technique to reduce
MSA. Our findings demonstrated a favorable benefit-risk profile of LS- postoperative complications: 5-year results, Yonsei Med. J. 54 (2013) 1248–1252,
MSA based on CR-R0 and fewer R5, whose LHH values consistently far https://doi.org/10.3349/ymj.2013.54.5.1248.
[10] I.A. McGregor, D.G. MacDonald, Mandibular osteotomy in the surgical approach to
exceeded one. Tolerability of both surgical approaches is comparable. the oral cavity, Head Neck Surg. 5 (1983) 457–462, https://doi.org/10.1002/
Taken together, LS-MSA merits the further use in PO/OPC patients, even hed.2890050512.
though additional studies should be performed to investigate causes and [11] M. Champy, J.P. Loddé, R. Schmitt, et al., Mandibular osteosynthesis by miniature
screwed plates via a buccal approach, J. Maxillofac. Surg. 6 (1978) 14–21, https://
preventive measures against postoperative WD and OCF. Large ran­ doi.org/10.1016/s0301-0503(78)80062-9.
domized controlled trials could provide a more precise conclusion. [12] N.J. Ledderhof, E.R. Carlson, R.E. Heidel, et al., Are tracheotomies required for
However, the evidence-based medicine requires the use of the “best, patients undergoing composite mandibular resections for oral cancer? J. Oral
Maxillofac. Surg. 78 (2020) 1427–1435, https://doi.org/10.1016/j.
current” evidence for decision making in individual patient care [36]. joms.2020.03.027.
The level of evidence of this study appears to range high among previous [13] C. McGuire, C. Boudreau, N. Prabhu, et al., Piezosurgery versus conventional
publications (except the systematic review [20] and meta-analysis cutting techniques in craniofacial surgery: a systematic review and meta-analysis,
Plast. Reconstr. Surg. 149 (2022) 183–195, https://doi.org/10.1097/
[21]). PRS.0000000000008645.
[14] Leitlinienprogramm Onkologie der Arbeitsgemeinschaft der Wissenschaftlichen
Disclosure of potential conflicts of interest Medizinischen Fachgesellschaften (AWMF), Deutschen Krebsgesellschaft (DKG)
und Deutschen Krebshilfe (DKH): konsultationsfassung S3-Leitlinie Diagnostik und
Therapie des Mundhöhlenkarzinoms, Available at: https://www.leitlinienprogra
All of the authors indicate full freedom of investigation and manu­ mm-onkologie.de/fileadmin/user_upload/Downloads/Leitlinien/Mundhoehlenka
script preparation without potential conflict of interest as regards this rzinom/Version_3/LL_Mundhoehlenkarzinom_Langversion_Konsultationsfassun
study. g_3.01.pdf. (Accessed 13 April 2022).
[15] S.G. Hakim, R. von Bialy, M. Falougy, et al., Impact of stratified resection margin
classification on local tumor control and survival in patients with oral squamous
Financial disclosure cell carcinoma, J. Surg. Oncol. 124 (2021) 1284–1295, https://doi.org/10.1002/
jso.26655.
[16] S.N. Rogers, J.P. O’donnell, S. Williams-Hewitt, et al., Health-related quality of life
There was no funding, grants or financial incentive for any author in measured by the UW-QoL—reference values from a general dental practice, Oral
this paper. Oncol. 42 (2006) 281–287, https://doi.org/10.1016/j.oraloncology.2005.08.002.
[17] R.A. Borchardt, D. Tzizik, Update on surgical site infections: the new CDC
guidelines, J. Am. Acad. Physician Assistants 31 (2018) 52–54, https://doi.org/
Availability of data and material 10.1097/01.JAA.0000531052.82007.42.
[18] M.S. Malgonde, M. Kumar, Complications after neck dissection, Med J DY Patil
Deidentified individual participant data are not available. The Univ 8 (2015) 458–462, https://doi.org/10.4103/0975-2870.160785.
[19] M.D. Fahmy, B.L. Hechler, E.R. Carlson, et al., Preoperative serum albumin
datasets generated and analyzed during this study are available from the predicts wound dehiscence but not infection after surgery for oral squamous cell
first author (P.P.) upon reasonable request. carcinoma, J. Oral Maxillofac. Surg. 79 (2021) 1355–1363, https://doi.org/
10.1016/j.joms.2020.12.021.
[20] F. Hedayat, K.K. Jerry Htwe, L.V. Vassiliou, P. Kyzas, Morbidity related to the lip-
Authorship disclosure
split mandibulotomy approach: a systematic and narrative review, Br. J. Oral
Maxillofac. Surg. 60 (2022) 430–436, https://doi.org/10.1016/j.
Conception and design: P.P., L-K.W., C.T., K.S., N.S., P.A., J-P.M., A.N. bjoms.2021.09.019.
[21] P. Pang, R.W. Li, J.P. Shi, et al., A comparison of mandible preservation method
Acquisition, analysis and interpretation of (blinded) data: P.P., L-K.W.,
and mandibulotomy approach in oral and oropharyngeal cancer: a meta-analysis,
C.T., K.S., N.S., P.A. Oral Oncol. 63 (2016) 52–60, https://doi.org/10.1016/j.
Drafting and revising the work: P.P., L-K.W., C.T., K.S., N.S., P.A., J-P. oraloncology.2016.11.004.
M., A.N. [22] L. Citrome, T.A. Ketter, When does a difference make a difference? Interpretation
of number needed to treat, number needed to harm, and likelihood to be helped or
Final approval of the work: P.P., L-K.W., C.T., K.S., N.S., P.A., J-P.M., harmed, Int. J. Clin. Pract. 67 (2013) 407–411, https://doi.org/10.1111/
A.N. ijcp.12142.
Agreement to all aspects of the work: P.P., L-K.W., C.T., K.S., N.S., P.A., [23] P. Stathopoulos, M.E.D. Kaplani, N. Kostis, et al., The impact of lip-split
mandibulotomy on patients treated for pT2 oral tongue squamous cell carcinoma: a
J-P.M., A.N. study of 224 patients, Oral Maxillofac. Surg. 25 (2021) 313–318, https://doi.org/
10.1007/s10006-020-00918-8.
References [24] C.K. Tsao, A. Kudva, D.C.F. Cheong, S.Y. Hung, Current concepts for reconstruction
of oral tissues following ablative Surgery: a comprehensive approach, J Maxillofac
Oral Surg 21 (2022) 358–368, https://doi.org/10.1007/s12663-022-01704-7.
[1] Y.H. Joo, J.K. Cho, B.S. Koo, et al., Guidelines for the surgical management of oral
[25] A.D. Rapidis, S. Valsamis, D.A. Anterriotis, C.A. Skouteris, Functional and aesthetic
cancer: Korean society of thyroid-head and neck surgery, Clin Exp
results of various lip-splitting incisions: a clinical analysis of 60 cases, J. Oral
Otorhinolaryngol 12 (2019) 107–144, https://doi.org/10.21053/ceo.2018.01816.
Maxillofac. Surg. 59 (2001) 1292–1296, https://doi.org/10.1053/
[2] P. Pitak-Arnnop, K. Subbalekha, J.P. Meningaud, et al., Factors associated with
joms.2001.27517.
epiphora following orbital-sparing maxillectomy via modified Weber-Ferguson
[26] R.H. Spiro, F.P. Gerold, E.W. Strong, Mandibular “swing” approach for oral and
incision with lower blepharoplasty, J. Surg. Oncol. 123 (2021) 1246–1252,
oropharyngeal tumors, Head Neck Surg. 3 (1981) 371–378, https://doi.org/
https://doi.org/10.1002/jso.26408.
10.1002/hed.2890030505.
[3] H.F. Köhler, H. Mehanna, J.P. Shah, et al., Comparison of different guidelines for
[27] P. Mehanna, J. Devine, J. McMahon, Lip split and mandibulotomy modifications,
oral cancer, Eur. Arch. Oto-Rhino-Laryngol. 278 (2021) 2961–2973, https://doi.
Br. J. Oral Maxillofac. Surg. 48 (2010) 314–315, https://doi.org/10.1016/j.
org/10.1007/s00405-020-06423-8.
bjoms.2009.06.231.
[4] P.J. Roux, H.T. Butlin, G.J. Spencer (Eds.), Diseases of the Tongue. Clinical
[28] B.W. Cilento, M. Izzard, E.A. Weymuller, N. Futran, Comparison of approaches for
Manuals for Practitioners and Students of Medicine, Cassell & Co., London, 1885,
oral cavity cancer resection: lip-split versus visor flap, Otolaryngol. Head Neck
pp. 331–359.
Surg. 137 (2007) 428–432, https://doi.org/10.1016/j.otohns.2007.05.006.
[5] R.K. Orosco, A. Arora, J.P. Jeannon, F.C. Holsinger, Next-generation robotic head
[29] P. Wusiman, A. Yarbag, G. Wurouzi, et al., Three dimensional versus standard
and neck surgery, ORL J Otorhinolaryngol Relat Spec 80 (2018) 213–219, https://
miniplate fixation in management of mandibular fractures: a systematic review and
doi.org/10.1159/000490599.
meta-analysis, J. Cranio-Maxillo-Fac. Surg. 44 (2016) 1646–1654, https://doi.org/
[6] P. Pitak-Arnnop, K. Subbalekha, N. Sirintawat, et al., A retrospective cohort study
10.1016/j.jcms.2016.07.027.
on predictors associated with skull base invasion of maxillary ameloblastomas,
[30] Y. Liu, B. Wei, Y. Li, et al., The 3-dimensional miniplate is more effective than the
J Stomatol Oral Maxillofac Surg (2022), https://doi.org/10.1016/j.
standard miniplate for the management of mandibular fractures: a meta-analysis,
jormas.2022.03.015 (in press).
Eur. J. Med. Res. 22 (2017) 5, https://doi.org/10.1186/s40001-017-0244-2.
[7] P. Pitak-Arnnop, L.K. Witohendro, J.P. Meningaud, et al., Which characteristics can
[31] J.A. Vermaire, C.P.J. Raaijmakers, E.M. Monninkhof, et al., Factors associated with
be expected from p16+-squamous cell carcinomas of the posterior oral cavity and
masticatory function as measured with the mixing ability test in patients with head
oropharynx? – distinctive results from Central Germany, J Stomatol Oral
and neck cancer before and after treatment: a prospective cohort study, Support.
Maxillofac Surg 121 (2020) 213–218, https://doi.org/10.1016/j.
Care Cancer 30 (2022) 4429–4436, https://doi.org/10.1007/s00520-022-06867-0.
jormas.2019.10.013.

7
P. Pitak-Arnnop et al. Surgical Oncology 44 (2022) 101837

[32] S. Kim, R.M. Doh, L. Yoo, et al., Assessment of age-related changes on masticatory retrognathic and microstomic patient, BMJ Case Rep. 13 (2020), e236010, https://
function in a population with normal dentition, Int. J. Environ. Res. Publ. Health doi.org/10.1136/bcr-2020-236010.
18 (2021) 6899, https://doi.org/10.3390/ijerph18136899. [36] P. Pitak-Arnnop, A. Hemprich, N.C. Pausch, Evidence-based oral and maxillofacial
[33] T. Kosaka, T. Ono, M. Kida, et al., A prediction model of masticatory performance surgery: some pitfalls and limitations, J. Oral Maxillofac. Surg. 69 (2011) 252–257,
change in 50- to 70-year-old Japanese: the Suita study, J. Dent. 104 (2021), https://doi.org/10.1016/j.joms.2010.07.082.
103535, https://doi.org/10.1016/j.jdent.2020.103535. [37] W. Sakdajeyont, T. Chaiyasamut, A. Arayasantiparb, N. Wongsirichat, The mental
[34] G. Tay, J. Ferrell, P. Andersen, Use of a midline mandibular osteotomy to improve foramen in panoramic versus cone beam computed tomogram, M. Dent. J. 38
surgical access for transoral robotic resection of the base of tongue in a patient with (2018) 159–167. Available at: https://dt.mahidol.ac.th/th/wp-content/uplo
trismus, Head Neck 39 (2017) E92–E95, https://doi.org/10.1002/hed.24851. ads/2018/09/08-The-mental-foramen-in-panoramic27-9-61.pdf. (Accessed 13
[35] D. Bestourous, M. Michel, C. Badger, et al., Intraoral midline mandibulotomy to April 2022).
improve access for transoral robotic surgery (TORS) base of tongue resection in a [38] T. Ayad, Facial artery musculomucosal (FAMM) flap, Operat. Tech. Otolaryngol.
Head Neck Surg. 30 (2019) 112–119, https://doi.org/10.1016/j.otot.2019.04.005.