15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
The Laryngoscope
© 2022 The American Laryngological,
Rhinological and Otological Society, Inc.

Radiological Findings in Laryngeal Anterior Commissure Invasion:

CT Scan Highlights

Giovanni Cristalli, MD ; Antonello Vidiri, MD; Giuseppe Mercante, MD; Fabio Ferreli, MD ;
Armando De Virgilio, PhD ; Filippo Donelli, MD; Luigi Davì, MD; Pierpaola Gasparin, MD;
Patrizia Cocco, MD; Fabiola Giudici, BAM; Francesca Boscolo Nata, MD

Objectives: Preoperative anterior commissure (AC) evaluation in glottic cancer is crucial for therapeutic decisions. Endos-
copy is often inadequate to precisely detect the presence of cancer in the AC; thus, computed tomography (CT) scan could help.
We investigated the relation between AC thickness on CT scan (in mm), AC involvement by cancer at histology, and radiologic
signs of anterior paraglottic space (PGS) infiltration.
Study Design: Retrospective observational study.
Methods: An experienced radiologist retrospectively measured AC thickness and identified signs of anterior PGS infiltra-
tion on pretreatment contrast-enhanced CT scans of 80 patients with primary glottic cancer. The gold standard to define the
presence of cancer in the AC was histology. The receiver operating characteristic (ROC) curves were used to determine the
potential cut-off values of AC thickness (Youden index method) able to maximize both sensitivity and specificity in identifying
the presence of cancer in the AC at histology and PGS infiltration on CT scan.
Results: AC was significantly thicker in patients with cancer in the AC at histology (P < .001) and in patients with PGS
infiltration on CT scan (P < .001). The cut-off values to discriminate the presence of cancer at histology and PGS infiltration on
CT scan were 3.62 and 2.6 mm, respectively. We found a substantial agreement between anterior PGS infiltration on CT scan
and the presence of cancer in the AC at histology (Cohen Kappa: P = .70).
Conclusion: AC thickness and radiologic signs of PGS infiltration on pretreatment CT scan could represent a method to
predict the presence of cancer in the AC at histology.
Key Words: Laryngeal cancer, glottic cancer, computed tomography scan, anterior commissure, diagnostic imaging.
Level of Evidence: 4
Laryngoscope, 132:2427–2433, 2022

INTRODUCTION There is no consensus on the treatment of laryngeal

Laryngeal anterior commissure (AC) has a pivotal tumors that extend to the AC: some authors recommend
role in laryngeal oncology considering that up to 20% of a first-line radiation therapy (RT),9 while others suggest
laryngeal tumors can subtly extend to the AC.1 Mucosal/ endoscopic or open neck surgery.6,10 For these reasons,
submucosal involvement of the AC in glottic cancer has the preoperative evaluation of the AC is of paramount
been reported as a significant prognostic indicator for importance.11
local control2,3 and steers the treatment option.4–6 Clinical staging is the most important aspect in the
Many articles have confirmed the difficulties in management of early-stage (I-II) glottic cancer. However,
treating primary tumors involving the AC region, as well while endoscopy can highlight mucosal extent, some
as the lower effectiveness of radiotherapy in this area.5,7,8 T1–T2 tumors with anterior commissure submucosal
involvement are often under-staged using endoscopic
examination only.12 Consequently, computed tomography
From the Otorhinolaryngology Unit (G.C., F.B.N.), Ospedali Riuniti
Padova Sud “Madre Teresa di Calcutta” Hospital, Monselice, Italy; (CT) scan is recommended even in the first stages to
Radiology and Diagnostic Imaging Department (A.V.), IRCCS Regina assess the submucosal extent of the tumor.9 Some
Elena National Cancer Institute, Rome, Italy; Department of Biomedical
Sciences (G.M., F.F., A.D.V.), Humanitas University, Pieve Emanuele, Italy; authors recommend magnetic resonance imaging (MRI)
Radiology Unit (F.D., L.D.), Ospedali Riuniti Padova Sud “Madre Teresa di as the first choice method for laryngeal cancer staging
Calcutta” Hospital, Monselice, Italy; Pathological Anatomy Unit (P.G., P.
C.), Ospedali Riuniti Padova Sud “Madre Teresa di Calcutta” Hospital,
but, even if MRI shows high sensitivity and accuracy in
Monselice, Italy; and the Unit of Biostatistics, Epidemiology and Public detecting the presence of cancer in the AC, the specificity
Health, Department of Cardiac, Thoracic, Vascular Sciences and Public is low, and this can lead to overestimation of the exten-
Health (F.G.), University of Padua, Padua, Italy.
Editor’s Note: This Manuscript was accepted for publication on Jan-
sion of the tumor to the AC.13
uary 28 2022. While there are several criteria to define cartilage
The authors have no funding, financial relationships, or conflicts of invasion (erosion or lysis of thyroid cartilage, sclerosis of
interest to disclose.
Send correspondence to Francesca Boscolo Nata, MD, Otorhinolaryn- the cricoid and arytenoid cartilages, the absence of a well-
gology Unit, Ospedali Riuniti Padova Sud “Madre Teresa di Calcutta” Hospi- defined hypoattenuating line between cartilage and
tal, 35043 Monselice, Italy. E-mail: [email protected]
tumor, and a thyroid cartilage with a Hounsfield unit
DOI: 10.1002/lary.30060 (HU) score that does not differ from that of the tumor),14

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2427
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
radiological detection of subtle tumor spread into the AC but we think this value risks to overestimate tumors,
can be challenging. After a review of the literature, we with consequent overtreatment in early stages.
found that the radiologic criteria to define mucosal/ The first aim of this article is to define the sensitivity
submucosal involvement of the AC (i.e. change in density, and specificity of AC thickness evaluated by a high-
change in shape) are less standardized. In fact, the rMAP definition CT scan to predict the presence of cancer in the
system proposed by Benazzo et al.15 implies an evalua- AC at histological examination. As secondary aims we
tion performed by an experienced head and neck radiolo- investigated the correlation between anterior paraglottic
gist because the radiologic parameter used to define the space (PGS) infiltration (located anteriorly to the “magic
mucosal involvement is not indicated. Lim et al.16 listed line” as described by Succo et al.19) and AC thickness on
CT scan, and the relationship between anterior PGS infil-
radiologic criteria of cartilage but not mucosal involve-
tration on CT scan and the presence of cancer in AC at
ment. The only article to better define AC mucosal
histological examination.
involvement is that by Barbosa et al.17: by using the
GRACI, it combines a dimensional criterion (extension to
contiguous regions: superiorly, anteriorly, and inferiorly)
to AC thickness. MATERIALS AND METHODS
In our opinion, obtaining a numerical parameter on This retrospective study was performed at the ENT Clinic
pretreatment CT scan suggestive of AC mucosal/ of Ospedali Riuniti Padova Sud in accordance with the principles
submucosal involvement could represent a simple and stated in the Declaration of Helsinki (1964). Its design was
reproducible method to obtain crucial information for approved by our Hospital Ethics Committee (4875/U6/29);
clinical assessment. There is a general consensus to con- patients signed informed consent for the use of their
anonymized data.
sider the thickness of AC over 2 mm as pathological,18
A retrospective chart review was performed to obtain clini-
cal, radiologic, and histologic data of oncologic patients treated at
our hospital between August 2015 and December 2020.
TABLE I. Inclusion criteria were patients older than 18 years with
Clinicopathological Characteristics of the Cohort. primary glottic cancer, squamous cell carcinoma (SCC) histology,
and preoperative staging (CT scan with contrast, video-
Characteristic Sample (N = 80)*
laryngoscopy, biopsy) performed at our hospital. Clinical staging
Age, median (range), years 68 (39–89)
was performed following American Joint Committee on Cancer
indications 8th edition.20
Sex
Not collaborative patients, patients with laryngeal abnor-
Male 76 (95.0%) malities (associated laryngocele, previously diagnosed vocal fold
Female 4 (5.0%) palsy, Reinke edema, laryngeal trauma, history of orotracheal
Smoking habits intubation), staged at other centers or not undergoing CT scan
(because allergic/intolerant to iodine), patients with recurrent
Never 23 (28.8%)
tumors after radiotherapy or conservative laryngeal surgery, or
Ever 57 (71.2%)
with secondary tumors, were excluded.
Drinking habits
Never 58 (72.5%)
Current 22 (27.5%) 1.0
Histologic Grading
G1 20 (25.0%)
G2 51 (63.8%) 0.8
G3 9 (11.2%)
T stage†
0.6
Sensitivity

T1 32 (40%)
T2 13 (16.3%)
T3 31 (38.6%)
0.4
T4 4 (5.0%)
†
N stage
P-value = 0.13, De-Long test
N0 58 (72.5%) 0.2
N1 3 (3.8%)
A: AUC = 0.96 (0.91-1.00)
N2 17 (21.3%)
B: AUC = 0.89 (0.80-0.97)
N3 2 (2.5%) 0.0
†
M stage
1.0 0.8 0.6 0.4 0.2 0.0
M0 78 (97.5%)
M1 2 (2.5%) Specificity

*Data are presented as the number (percentage) of patients unless Fig. 1. Continuous receiver operating characteristic (ROC) curve
otherwise indicated. (A) represents cancer in the anterior commissure at histological
† examination; dashed ROC curve (B) represents anterior paraglottic
Radiologic stage.
M = metastasis; N = node; T = tumor. space infiltration on computed tomography scan.

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2428
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. 2. Squamous cell carcinoma of the left vocal cord cT3N0M0, radiologic and histologic findings. (A) Computed tomography scan, axial
plane, showing a glottic tumor of the left vocal cord involving the anterior paraglottic space; (B) anterior commissure thickness is 5.23 mm;
(C) plane section, hematoxylin and eosin staining, 1:1 magnification: the tumor involves the anterior commissure (arrow), close relationship
between the tumor and the thyroid cartilage without its infiltration.

TABLE II.
Relation Between Anterior Paraglottic Space Infiltration on Computed Tomography (CT) and Cancer in the Anterior Commissure at Histology.
Cancer in the anterior commissure at histology
No Yes Total

Anterior paraglottic space infiltration No 23 (82.1%) 6 (11.5%) 29

Yes 5 (17.9%) 46 (88.5%) 51
Total 28 52 80

Radiologic imaging was re-evaluated by an experienced the European Laryngological Society (ELS).11,22 AC thickness
head and neck radiologist blinded on histologic findings in order was measured using the graphical ruler provided by the image
to define AC thickness (in mm) and anterior paraglottic space reading software of CT. Anterior paraglottic space infiltration
infiltration. was defined when the fat normally present in this area was rep-
The gold standard to define the presence of cancer in the laced or obscured by the neoplasia.18
AC was the histological examination obtained by target biopsy
during direct diagnostic laryngoscopy or surgical specimen (total
laryngectomy, open partial horizontal laryngectomy, or transoral Statistical Analysis
laser microsurgery) in patients undergoing surgical treatment. Descriptive statistics of patient demographics and clinical
characteristics were reported as frequencies (proportions) for cat-
egorical variables and median (range min-max) or mean (stan-
Image Mode dard deviation) for continuous variables according to data
Pretherapeutic CT imaging of the larynx was performed distribution (verified through Shapiro–Wilk test of normality).
via contrast-enhanced axial CT scans (Somatom Definition The presence of a relationship between AC thickness on CT scan
Flash Dual Source SIEMENS–Siemens Healthcare Headquar- (in mm) and AC mucosal/submucosal involvement by cancer at
ters Siemens Healthcare Gmb HHenkestr. 12791052 Erlangen histological examination was investigated using the Mann–
Germany). Whitney test. The same test was used to investigate the relation-
Both static and dynamic manoeuvres were performed.21 ship between AC thickness on CT scan (in mm) and the presence
The parameters of the CT acquisition were: tube current of anterior paraglottic space involvement on CT scan.
100 mAs, voltage 120 kV, detector collimation 128  0.625, rota- To evaluate the discriminatory ability of AC thickness
tion time 1 second, table speed: rotation time 1.0 second + pitch (mm) on CT scan in assessing AC involvement by cancer at histo-
0.8, slice thickness 1.5 mm, incrementing 1.2 mm, scan time logical examination and anterior paraglottic space involvement
1.26 second, FOV 200 mm and matrix 512  512. Nonionic con- on CT scan, receiver operating characteristics (ROC) curves and
trast material (100 ml) (Omnypaque-Ioexolo 350 G e Healthcare area under curves (AUC) were calculated. Interpretation of the
Srl) was injected with a power injector at a rate of 3 ml/s delays AUC was carried out using the Swets classification, with higher
70 seconds in each patient. values indicating better discriminatory ability23: 1) AUC = 0.5,
The native images were reformatted parallel to the vocal the test is not informative; 2) 0.5 < AUC ≤0.7, the test is poorly
cords (or C4-5/C5-6 intervertebral space) from the oral cavity to informative; 3) 0.7 < AUC ≤0.9, the test is moderately accurate;
the thoracic cavity. AC was identified in the plane crossing the 4) 0.9 < AUC < 1, the test is highly accurate; and 5) AUC = 1,
junction between cricoid and arytenoid cartilages according to the test is perfect. Optimal cut-offs able to maximize sensitivity

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2429
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. 3. Squamous cell carcinoma of the left vocal cord cT2N0M0 (impaired vocal cord mobility), radiologic, and histologic findings.
(A) Computed tomography scan, axial plane, showing a glottic tumor involving the anterior two-thirds of the left vocal cord; anterior commis-
sure thickness is 1.74 mm, anterior paraglottic space is not involved by the cancer. (B) Plane section, hematoxylin and eosin staining, 1:1
magnification: the anterior commissure (arrow) is not involved by the cancer.

(SE) and specificity (SP) were identified for AC thickness on CT was analyzed by ROC analysis (Fig. 1, curve B). The AUC
scan, and SE, SP, positive, and negative predictive values was 0.89 (95% CI: 0.80–0.97) or moderately accurate
(PPV/NPV) were calculated. The method used to compute opti- according to the Swets classification. The cut-off value to
mal cut-offs is the Youden Index (J). J can be formally defined as discriminate the presence of anterior paraglottic space
J = max_c {Se (c) + Sp (c) 1}: the cut-point that achieves this
infiltration on CT scan was 2.6 mm. The resulting sensi-
maximum is referred to as the optimal cut-point (c*) because it is
the cut-point that optimizes the biomarker’s differentiating abil-
tivity, specificity, PPV, and NPV were 96% (87%–99%),
ity when equal weight is given to sensitivity and specificity 79% (60%–92%), 89% (76%–99%), and 92% (75%–97%).
(Youden, R-package: “OptimalCutPoints”). The two ROC curves There were six FP and two FN.
were compared with the DeLong test. Thus, AC thickness significantly correlated not only
Statistical significance was set at P < .05 for all analyses. with the presence of cancer at histological examination
All statistical analyses were performed using the R software ver- but also with anterior paraglottic space infiltration on CT
sion 4.0.2 (2020). scan (AUC 0.96 (0.91–1.00) and 0.89 (0.80–0.97), respec-
tively, P = .13, De Long Test) (Fig. 1).
We found a substantial agreement between anterior
paraglottic space infiltration on CT scan and the presence
RESULTS of cancer in the AC at histological examination (Cohen
A total of 80 patients affected by histology-proven Kappa: P = .70, 95% CI: 0.54–0.86, Mc-Nemar test:
glottic primary squamous cell carcinoma met the inclu- P = 1.00). As presented in Table II, 82% (23/28) of
sion criteria and were enrolled in the study. Patients and patients without cancer in the AC at histological exami-
tumors characteristics are listed in Table I. nation did not have anterior paraglottic space infiltration
Anterior commissure was significantly thicker in on CT scan (Fig. 3); similarly, 88.5% (46/52) of patients
patients with cancer in the AC at histological examina- with cancer in the AC at histological examination had
tion (P < .001 Mann–Whitney test). The accuracy of AC anterior paraglottic space infiltration on CT scan (Fig. 2).
thickness on CT scan to predict the presence of cancer at There were 11 discordant cases: six FN and five FP.
histological examination was evaluated by ROC analysis
(Fig. 1, curve A). The AUC was 0.96 (95% CI: 0.91–1.00)
or highly accurate according to the Swets classification.
The cut-off value to discriminate the presence of cancer at DISCUSSION
histological examination was 3.62 mm. The resulting sen- Clinical staging is a critical phase in the manage-
sitivity, specificity, PPV, and NPV were 92%, (81%–98%), ment of glottic carcinoma: specifically, AC mucosal/
93% (76%–99%), 96% (86%–99%), and 87% (70%–98%). submucosal involvement is crucial because, even if it does
There were two false positive (FP) and four false nega- not modify the clinical staging, it could dramatically
tive (FN). influence the therapeutic choice and prognosis.
Patients with anterior paraglottic space infiltration Cancer under- or over-staging is one of the most fre-
on CT scan had a significantly thicker AC on CT scan quent pitfalls in laryngeal cancer.12 Video-laryngoscopy is
(P < .001, Mann–Whitney test) (Fig. 2). The accuracy of not enough in the staging workup because it cannot high-
AC thickness on CT scan to discriminate patients with or light the submucosal spreading of the tumor, cartilage
without anterior paraglottic space infiltration on CT scan infiltration, initial paraglottic, or preepiglottic extension.

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2430
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Consequently, pretreatment imaging is essential to After the finding that AC was significantly thicker in
obtain detailed information on this area even in the early patients with cancer in the AC at histological examina-
stages.24 tion (P < .001), we aimed to identify the optimal AC thick-
Thanks to its speed of image acquisition, the possi- ness to predict the presence of cancer at histological
bility to acquire and reconstruct an entire volume in dif- examination. The statistical analysis performed had the
ferent planes, and the well-standardized technique, CT is objective of optimizing both sensitivity and specificity, all-
usually preferred to study the larynx; moreover, there is owing to minimize the number of FN and FP and the risk
better patient compliance. Conversely, the high-contrast of under and overtreatment. Our experience showed that
resolution and the absence of exposure to ionizing radia- 3.62 mm represents the optimal cut-off to predict AC
tions are advantages of MRI, but the need for an expert involvement at histological examination.
user to handle the many variables influencing the image This value could appear high in comparison to the
quality and the longer acquisition times are potential fateful 2 mm previously reported but, as highlighted in
drawbacks.11,15,18 The choice of the correct axial plane of the Appendix, the more widely published value of 2 mm
acquisition on CT is crucial.22 Otherwise, the clinician allows to optimize sensitivity (which is 100% for thickness
risks misjudgment on tumor spreading. up to 2.23 mm) at the expense, however, of specificity.
We expected the normal thickness of the AC to be thin The reason for this result could be that in previous litera-
because it looks like only a mucosal covering of the poste- ture the normal AC thickness refers to the non-neoplastic
rior aspect of the thyroid cartilage. However, it was demon- larynx.29 Conversely, in the larynx affected by the tumor,
strated that this region is more complex: in fact, at the various factors could be considered, such as lymphatic
level of the upper surface of the true vocal cords, a thick and blood stagnation, desmoplastic reaction to the tumor,
layer of fibrous tissue fills the interstice between the thy- muscle hypertrophy to overcome vocal impairment, and
roid cartilage and the laryngeal mucosa. This space has associated inflammations. They all, albeit in a variable
been termed “x-space of Bagatella and Bignardi” according way, might lead to a relative thickening of the commis-
to its shape,25 or “the 0-point of Yves Guerrier.”26–28 sure. This aspect has been previously noted by Jian-Hui
According to a radiologic study on 38 adult patients Wu et al., who reported a low specificity of MRI in the AC
with unaffected larynx, the mean anteroposterior width because of the local inflammatory phenomena related to
of the AC was 1.02  0.56 mm; however, if 1.7 mm was the tumour.13 These considerations justify the high num-
considered the upper limit of a normal AC width, 92% of ber of FP and the consequent low specificity (46.43%) visi-
patients in that cohort were included. The authors under- ble in the Appendix for the traditionally reported
lined that it could be expected for some patients to have thickness of 2 mm. As previously stated, our aim was to
AC measurements approaching 2 mm in normal optimize both sensitivity and specificity to reduce the risk
conditions.29 of under and overtreatment, and a cut-off of 3.62 mm
An AC thickness of more than 2 mm has been tradi- allowed us to realise this aim (92% sensitivity, and 93%
tionally considered as a sign of mucosal/submucosal tumor specificity). Confirmative studies will demonstrate if this
involvement18 but, to the best of our knowledge, this had thicker value could represent a newer standard to define
never been supported by a strong statistical analysis. In AC tumor involvement in pretreatment CT scan.
fact, previous authors have investigated CT scan diagnostic In this study, a CT scan was used to assess AC thick-
accuracy comparing imaging and histopathologic findings, ness because, as previously reported, this is the most fre-
but the radiologic criteria to define AC involvement were quently used radiologic method in cancer staging. According
not well defined. An exception is represented by the work to Zbaren et al., there are no significant differences between
by Barbosa et al., who defined gross radiologic anterior CT and MRI in the assessment of AC.30 In the future, it
commissure invasion (GRACI), which are signs indicative would be interesting to verify radiologic criteria predictive
of tumor deeply extending to the AC.17 of AC invasion using an advanced MRI technique.31
Specifically, Benazzo et al. reported a sensitivity of For the first time in the literature, we investigated
81.25% and a specificity of 80% for CT in predicting the relationship between anterior paraglottic space infil-
tumor extension in the AC. The level of tumor extension tration on CT scan and AC thickness on CT scan: specifi-
was classified as superficial (mucosa or Broyles ligament), cally, we found that patients with anterior PGS
involvement of the thyroid cartilage, and extra-laryngeal infiltration had thicker AC (P < .001). The cut-off value to
extension.15 discriminate the presence of anterior paraglottic space
Lim et al., comparing preoperative images and surgi- infiltration on CT scan was 2.6 mm. As anticipated, no
cal specimens, highlighted a high number of false posi- similar studies have been conducted before. Thus, it is
tives (27.5%) by imaging in the AC. Radiologic sensitivity not possible to compare our findings with others.
and specificity for AC were 91% and 42%.16 Consequently, in our experience, AC thickness on
These studies demonstrated a low specificity of CT CT scan was associated not only with the presence of can-
scan in identifying AC involvement, with a consequent risk cer at histological examination but also with paraglottic
of overestimation and following overtreatment. The need to space infiltration on CT scan, as demonstrated by the
identify numeric criteria in assessing AC mucosal/ similarity of the two AUC (0.96 and 0.89, respectively;
submucosal involvement is due to the fact that the density P = .13, De Long Test).
of the tumor overlaps with that of adjacent soft tissues, From previous literature, we can assume that CT
making discrimination between normal and invaded struc- and MRI can identify with some precision PGS infiltra-
tures difficult even after contrast administration.15 tion.21,32 Considering the previously listed issues in AC

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2431
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
evaluation on imaging, we wondered whether a relation- 5. Bradley PJ, Rinaldo A, Su arez C, et al. Primary treatments of the anterior
vocal commissure squamous carcinoma. Eur Arch Otorhinolaryngol 2006;
ship between anterior PGS infiltration and the presence 263:879–888.
of cancer at histological examination in the AC could 6. Jacobi C, Freundorfer R, Reiter M. Transoral laser microsurgery in early
glottis cancer involving the anterior commissure. Eur Arch
exist. We found a concordance between anterior par- Otorhinolaryngol 2019;276:837–845.
aglottic space infiltration on CT scan and the presence of 7. Reddy SP, Mohideen N, Marra S, Marks JE. Effect of tumor bulk on local
control and survival of patients with T1 glottic cancer. Radiother Oncol
cancer in the AC at histological examination using Mc- 1998;47:161–166.
Nemar test (P = 1.00), with only 11 discordant cases: six 8. Maheshwar AA, Gaffney CC. Radiotherapy for T1 glottic carcinoma: impact
of anterior commissure involvement. J Laryngol Otol 2001;115:298–301.
FN and five FP. 9. Jones TM, De M, Foran B, Harrington K, Mortimore S. Laryngeal cancer:
In our experience, paraglottic space infiltration on United Kingdom National Multidisciplinary guidelines. J Laryngol Otol
2016;130:S75–S82.
CT scan could be considered an indirect sign of AC 10. Prades JM, Gavid M, Dumollard JM, Timoshenko AT, Karkas A, Peoc’h M.
involvement at histological examination. We believe this Anterior laryngeal commissure: histopathologic data from supracricoid
correlation is important because the detection of anterior partial laryngectomy. Eur Ann Otorhinolaryngol Head Neck Dis 2016;133:
27–30.
paraglottic space involvement on CT scan is more obvious 11. Chiesa-Estomba CM, Ravanelli M, Farina D, et al. Imaging checklist for
for the radiologist. preoperative evaluation of laryngeal tumors to be treated by transoral
microsurgery: guidelines from the European laryngological society. Eur
Despite the retrospective nature of this work and the Arch Otorhinolaryngol 2020;277:1707–1714.
limited number of patients considered, this study focused 12. Baugnon KL, Beitler JJ. Pitfalls in the staging of cancer of the laryngeal
squamous cell carcinoma. Neuroimaging Clin N Am 2013;23:81–105.
on a homogeneous cohort and allowed the definition of 13. Wu JH, Zhao J, Li ZH, et al. Comparison of CT and MRI in diagnosis of
easy to reproduce radiologic parameters that are applica- laryngeal carcinoma with anterior vocal commissure involvement. Sci Rep
2016;6:30353.
ble to multicenter studies. 14. Dadfar N, Seyyedi M, Forghani R, Curtin HD. Computed tomography
appearance of normal nonossified thyroid cartilage: implication for tumor
invasion diagnosis. J Comput Assist Tomogr 2015;39:240–243.
15. Benazzo M, Sovardi F, Preda L, et al. Imaging accuracy in preoperative
staging of T3-T4 laryngeal cancers. Cancers (Basel) 2020;12:1074.
CONCLUSION 16. Lim JY, Park IS, Park SW, Kim JW, Kim YM. Potential pitfalls and thera-
peutic implications of pretherapeutic radiologic staging in glottic cancers.
Anterior commissure evaluation is of pivotal impor- Acta Otolaryngol 2011;131:869–875.
tance in the therapeutic choice of glottic cancer, but radio- 17. Barbosa MM, Araújo VJF, Boasquevisque E, et al. Anterior vocal commis-
sure invasion in laryngeal carcinoma diagnosis. Laryngoscope 2005;115:
logic features predictive of its mucosal/submucosal 724–730.
involvement are not clear. In the present study, we would 18. Huang BY, Solle M, Weissler MC. Larynx: anatomic imaging for diagnosis
and management. Otolaryngol Clin North Am 2012;45:1325–1361.
like to offer simple and broadly applicable criteria to pre- 19. Succo G, Crosetti E, Bertolin A, et al. Treatment for T3 to T4a laryngeal
dict AC mucosal/submucosal involvement by CT scan: cancer by open partial horizontal laryngectomies: prognostic impact of dif-
ferent pathologic tumor subcategories. Head Neck 2018;40:1897–1908.
that is why we chose a numerical parameter that is easily 20. Brierley JD, Gospodarowicz MK, Wittekind C. TNM Classification of
measured and reproducible using last-generation CT that Malignant Tumours. 8th ed. Chichester, UK: John Wiley & Sons, Ltd;
2017:49–51.
is widely available. 21. Celebi I, Basak M, Ucgul A, Yildirim H, Oz A, Vural C. Functional imaging
Moreover, we believe that the relationship between of the larynx via 256-slice multi-detector computed tomography in
patients with laryngeal tumors: a faster, better and more reliable pre-
anterior paraglottic space infiltration and the presence of therapeutic evaluation. Eur J Radiol 2012;81:e541–e547.
cancer in the AC at histological examination could be 22. Hermans R. Head and Neck Cancer Imaging. Berlin Heidelberg, Germany:
Springer-Verlag; 2012:38.
interesting in consideration of the simpler assessment of 23. Swets JA. Measuring the accuracy of diagnostic systems. Science 1998;240:
PGS status on imaging. 1285–1293.
24. Hartl DM, Ferlito A, Brasnu DF, et al. Evidence-based review of treatment
Future prospective multicentric studies are advised options for patients with glottic cancer. Head Neck 2011;33:1638–1648.
to validate our results and to introduce other new radio- 25. Bagatella F, Bignardi L. Morphological study of the laryngeal anterior com-
logic criteria to help the clinician in pretreatment tumor missure with regard to the spread of cancer. Acta Otolaryngol 1981;92:
167–171.
assessment. 26. Prades JM, Peoc’h M, Petcu C, Karkas A, Dumollard JM, Gavid M. The
anterior commissure of the human larynx revisited. Surg Radiol Anat
2017;39:871–876.
27. Andrea M. Vasculature of the anterior commissure. Ann Otol 1981;90:18–20.
28. Rucci L, Gammarota L, Borghi Cirri MB. Carcinoma of the anterior commis-
sure of the larynx: I. embryological and anatomic considerations. Ann Otol
BIBLIOGRAPHY Rhinol Laryngol 1996;105:303–308.
1. Rifai M, Khattab H. Anterior commissure carcinoma: I-histopathologic 29. Kallmes DF, Phillips CD. The normal anterior commissure of the glottis.
study. Am J Otolaryngol 2000;21:294–297. AJR Am J Roentgenol 1997;168:1317–1319.
2. Hakeem AH, Tubachi J, Pradhan SA. Significance of anterior commissure 30. Zbären P, Becker M, Läng H. Pretherapeutic staging of laryngeal carci-
involvement in early glottic squamous cell carcinoma treated with trans- noma. Clinical findings, computed tomography, and magnetic resonance
oral CO2 laser microsurgery. Laryngoscope 2013;123:1912–1917. imaging compared with histopathology. Cancer 1996;77:1263–1273.
3. Tulli M, Re M, Bondi S, et al. The prognostic value of anterior commissure 31. Maroldi R, Ravanelli M, Farina D. Magnetic resonance for laryngeal cancer.
involvement in T1 glottic cancer: a systematic review and meta-analysis. Curr Opin Otolaryngol Head Neck Surg 2014;22:131–139.
Laryngoscope 2020;130:1932–1940. 32. Jaipuria B, Dosemane D, Kamath PM, Sreedharan SS, Shenoy VS. Staging of
4. Zohar Y, Rahima M, Shvili Y, Talmi YP, Lurie H. The controversial treat- laryngeal and hypopharyngeal cancer: computed tomography versus histopa-
ment of anterior commissure carcinoma of the larynx. Laryngoscope 1992; thology. Iran J Otorhinolaryngol 2018;30:189–194.
102:69–72.

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2432
 15314995, 2022, 12, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/lary.30060 by HINARI-LEBANON, Wiley Online Library on [29/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
APPENDIX commissure thickness to predict histological infiltration
by cut-off score. At a cut-point of 3.62 mm, sensitivity and
Detailed report of sensitivity, specificity, and positive specificity are maximized.
and negative likelihood ratios (LR + and LR-) of anterior

Cutpoint Sensitivity Specificity Correctly Classified LR+ LR

(≥.92) 100.00% 0.00% 65.00% 1.0000

(≥.98) 100.00% 3.57% 66.25% 1.0370 0.0000
(≥1) 100.00% 7.14% 67.50% 1.0769 0.0000
(≥1.06) 100.00% 14.29% 70.00% 1.1667 0.0000
(≥1.07) 100.00% 17.86% 71.25% 1.2174 0.0000
(≥1.1) 100.00% 21.43% 72.50% 1.2727 0.0000
(≥1.27) 100.00% 25.00% 73.75% 1.3333 0.0000
(≥1.8) 100.00% 28.57% 75.00% 1.4000 0.0000
(≥1.83) 100.00% 32.14% 76.25% 1.4737 0.0000
(≥1.88) 100.00% 35.71% 77.50% 1.5556 0.0000
(≥1.9) 100.00% 42.86% 80.00% 1.7500 0.0000
(≥2) 100.00% 46.43% 81.25% 1.8667 0.0000
(≥2.06) 100.00% 50.00% 82.50% 2.0000 0.0000
(≥2.1) 100.00% 53.57% 83.75% 2.1538 0.0000
(≥2.12) 100.00% 64.29% 87.50% 2.8000 0.0000
(≥2.2) 100.00% 67.86% 88.75% 3.1111 0.0000
(≥2.23) 100.00% 78.57% 92.50% 4.6667 0.0000
(≥2.3) 98.08% 78.57% 91.25% 4.5769 0.0245
(≥2.5) 98.08% 82.14% 92.50% 5.4923 0.0234
(≥2.6) 98.08% 85.71% 93.75% 6.8654 0.0224
(≥3.24) 96.15% 85.71% 92.50% 6.7308 0.0449
(≥3.26) 94.23% 85.71% 91.25% 6.5962 0.0673
(≥3.3) 92.31% 85.71% 90.00% 6.4615 0.0897
(≥3.4) 92.31% 89.29% 91.25% 8.6154 0.0862
(≥3.62) 92.31% 92.86% 92.50% 12.9231 0.0828

Laryngoscope 132: December 2022 Cristalli et al.: Anterior Commissure Infiltration by CT

2433