Diagnostic accuracy of surface coil MRI in assessing cartilaginous invasion in laryngeal tumours: Do we need contrast-agent administration?

Machine learning model to preoperatively predict T2/T3 staging of laryngeal and hypopharyngeal cancer based on the CT radiomic signature

OBJECTIVES

To develop and assess a radiomics-based prediction model for distinguishing T2/T3 staging of laryngeal and hypopharyngeal squamous cell carcinoma (LHSCC) METHODS: A total of 118 patients with pathologically proven LHSCC were enrolled in this retrospective study. We performed feature processing based on 851 radiomic features derived from contrast-enhanced CT images and established multiple radiomic models by combining three feature selection methods and seven machine learning classifiers. The area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were used to assess the performance of the models. The radiomic signature obtained from the optimal model and statistically significant morphological image characteristics were incorporated into the predictive nomogram. The performance of the nomogram was assessed by calibration curve and decision curve analysis.

RESULTS

Using analysis of variance (ANOVA) feature selection and logistic regression (LR) classifier produced the best model. The AUCs of the training, validation, and test sets were 0.919, 0.857, and 0.817, respectively. A nomogram based on the model integrating the radiomic signature and a morphological imaging characteristic (suspicious thyroid cartilage invasion) exhibited C-indexes of 0.899 (95% confidence interval (CI) 0.843-0.955), fitting well in calibration curves (p > 0.05). Decision curve analysis further confirmed the clinical usefulness of the nomogram.

CONCLUSIONS

The nomogram based on the radiomics model derived from contrast-enhanced CT images had good diagnostic performance for distinguishing T2/T3 staging of LHSCC.

CLINICAL RELEVANCE STATEMENT

Accurate T2/T3 staging assessment of LHSCC aids in determining whether laryngectomy or laryngeal preservation therapy should be performed. The nomogram based on the radiomics model derived from contrast-enhanced CT images has the potential to predict the T2/T3 staging of LHSCC, which can provide a non-invasive and robust approach for guiding the optimization of clinical decision-making.

KEY POINTS

• Combining analysis of variance with logistic regression yielded the optimal radiomic model. • A nomogram based on the CT-radiomic signature has good performance for differentiating T2 from T3 staging of laryngeal and hypopharyngeal squamous cell carcinoma. • It provides a non-invasive and robust approach for guiding the optimization of clinical decision-making.

Thyroid cartilage infiltration in advanced laryngeal cancer: prognostic implications and predictive modelling

Objective

Detection of laryngeal cartilage invasion is of great importance in staging of laryngeal squamous cell carcinoma (LSCC). The role of prognosticators in locally advanced laryngeal cancer are still widely debated. This study aimed to assess the impact of volume of thyroid cartilage infiltration, as well as other histopathologic variables, on patient survival.

Materials and methods

We retrospectively analysed 74 patients affected by pT4 LSCC and treated with total laryngectomy between 2005 and 2021 at the Department of Otorhinolaryngology - Head and Neck Surgery of the University of Brescia, Italy. We considered as potential prognosticators histological grade, perineural (PNI) and lympho-vascular invasion (LVI), thyroid cartilage infiltration, and pTN staging. Pre-operative CT or MRI were analysed to quantify the volume of cartilage infiltration using 3D Slicer software.

Results

The 1-, 3-, and 5-year disease free survivals (DFS) were 76%, 66%, and 64%, respectively. Using machine learning models, we found that the volume of thyroid cartilage infiltration had high correlation with DFS. Patients with a higher volume (>670 mm3) of infiltration had a worse prognosis compared to those with a lower volume.

Conclusions

Our study confirms the essential role of LVI as prognosticator in advanced LSCC and, more innovatively, highlights the volume of thyroid cartilage infiltration as another promising prognostic factor.

CT and MR anatomy of the larynx and hypopharynx

Imaging of the larynx and hypopharynx is frequently requested to assess the extent of neoplasms beyond the field of view of endoscopic evaluation. The combination of optical and cross-sectional imaging allows tumors to be classified according to AJCC/UICC guidelines. A thorough understanding of laryngeal and hypopharyngeal anatomy is crucial to guide the radiological eye along the possible pathways of the spread of diseases and to guide differential diagnoses. Computed tomography (CT) has been the first cross-sectional imaging technique used to evaluate the larynx and hypopharynx; its spatial resolution combined with volumetric capability and the use of injectable contrast medium made CT the working horse in the assessment of neoplastic and inflammatory diseases. In the last two decades, magnetic resonance (MR) supported CT in the most challenging cases, when the optimal contrast resolution due to the multisequence portfolio is needed to assess the neoplastic involvement of laryngeal cartilages, paraglottic space(s), and extra laryngeal spread. The aim of this paper is to give a comprehensive radiological overview of larynx and hypopharynx complex anatomy, combining in vivo images, anatomical sections, and images of ex vivo specimens.

A Combined Model Integrating Radiomics and Deep Learning Based on Contrast-Enhanced CT for Preoperative Staging of Laryngeal Carcinoma

RATIONALE AND OBJECTIVES

Accurate staging of laryngeal carcinoma can inform appropriate treatment decision-making. We developed a radiomics model, a deep learning (DL) model, and a combined model (incorporating radiomics features and DL features) based on the venous-phase CT images and explored the performance of these models in stratifying patients with laryngeal carcinoma into stage I-II and stage III-IV, and also compared these models with radiologists.

MATERIALS AND METHODS

Three hundreds and nineteen patients with pathologically confirmed laryngeal carcinoma were randomly divided into a training set (n = 223) and a test set (n = 96). In the training set, the radiomics features with inter- and intraclass correlation coefficients (ICCs)> 0.75 were screened by Spearman correlation analysis and recursive feature elimination (RFE); then support vector machine (SVM) classifier was applied to develop the radiomics model. The DL model was built using ResNet 18 by the cropped 2D regions of interest (ROIs) in the maximum tumor ROI slices and the last fully connected layer of this network served as the DL feature extractor. Finally, a combined model was developed by pooling the radiomics features and extracted DL features to predict the staging.

RESULTS

The area under the curves (AUCs) for radiomics model, DL model, and combined model in the test set were 0.704 (95% confidence interval [CI]: 0.588-0.820), 0.724 (95% CI: 0.613-0.835), and 0.849 (95% CI: 0.755-0.943), respectively. The combined model outperformed the radiomics model and the DL model in discriminating stage I-II from stage III-IV (p = 0.031 and p = 0.020, respectively). Only the combined model performed significantly better than radiologists (p < 0.050 for both).

CONCLUSION

The combined model can help tailor the therapeutic strategy for laryngeal carcinoma patients by enabling more accurate preoperative staging.

Additional Diffusion-Weighted Imaging with Background Body Signal Suppression (DWIBS) Improves Pre-Therapeutical Detection of Early-Stage (pT1a) Glottic Cancer: A Feasibility and Interobserver Reliability Study

(1) Background: Early-stage glottic cancer is easily missed on magnetic resonance imaging (MRI). Diffusion-weighted imaging (DWI) may improve diagnostic accuracy. Therefore, our aim was to assess the value of adding diffusion-weighted imaging with background body signal suppression (DWIBS) to pre-therapeutic MRI staging. (2) Methods: Two radiologists with 8 and 13 years of experience, blinded to each other's findings, initially interpreted only standard MRI, later DWIBS alone, and afterward, standard MRI + DWIBS in 41 patients with histopathologically proven pT1a laryngeal cancer of the glottis. (3) Results: Detectability rates with standard MRI, DWIBS only, and standard MRI + DWIBS were 68-71%, 63-66%, and 73-76%, respectively. Moreover, interobserver reliability was calculated as good (κ = 0.712), very good (κ = 0.84), and good (κ = 0.69) for standard MRI, DWIBS only, and standard MRI + DWIBS, respectively. (4) Conclusions: Standard MRI, DWIBS alone, and standard MRI + DWIBS showed an encouraging detection rate, as well as distinct interobserver reliability in the diagnosis of early-stage laryngeal cancer when compared to the definitive histopathologic report.

Magnetic resonance imaging to assess cartilage invasion in recurrent laryngeal carcinoma after transoral laser microsurgery

Objective

To evaluate the diagnostic performance of magnetic resonance (MR) with surface coils in assessing cartilage invasion in recurrent laryngeal carcinoma after carbon dioxide transoral laser microsurgery (CO₂ TOLMS).

Methods

Two expert head and neck radiologists assessed cartilage invasion (infiltrated or non-infiltrated) in submucosal recurrences of laryngeal carcinoma after CO₂ TOLMS: results were compared with histopathological report after salvage laryngectomy.

Results

Thirty patients met the inclusion criteria and 90 cartilages were assessed. Overall sensitivity, specificity, and positive and negative predictive values for cartilage infiltration were 76, 93, 72 and 94%, respectively; for thyroid cartilage, the values were 82, 79, 69 and 88% respectively; for cricoid cartilage, all values were 100%; and for arytenoids, the values were 33, 96, 56 and 93% respectively.

Conclusions

MR with surface coils was able to detect most thyroid and cricoid infiltration in the complex setting of post-CO₂ TOLMS laryngeal carcinoma recurrence. In particular, the optimal performance in assessing cricoid invasion can be valuable in choosing the most appropriate treatment among total laryngectomy, open partial horizontal laryngectomies and non-surgical strategies.

MR Imaging of Laryngeal and Hypopharyngeal Cancer

State-of-the-art MR imaging of the larynx and hypopharynx with high-resolution surface coils, parallel imaging techniques, and DWI has several advantages over CT for assessing submucosal tumor spread, in particular neoplastic involvement of the paraglottic space, laryngeal cartilages, and extralaryngeal soft tissues. Current diagnostic MR imaging criteria based on a combination of distinct imaging features on morphologic sequences combined with DWI allow improved discrimination between tumor, peritumoral inflammation, and fibrosis and, ultimately, an increased precision for submucosal tumor delineation, which is a key prerequisite for tailored treatment options. Multiparametric MR imaging with DWI has a higher diagnostic performance than CT.

High-resolution magnetic resonance and mass spectrometry imaging of the human larynx

High-resolution, noninvasive and nondestructive imaging of the subepithelial structures of the larynx would enhance microanatomic tissue assessment and clinical decision making; similarly, in situ molecular profiling of laryngeal tissue would enhance biomarker discovery and pathology readout. Towards these goals, we assessed the capabilities of high-resolution magnetic resonance imaging (MRI) and matrix-assisted laser desorption/ionisation-mass spectrometry (MALDI-MS) imaging of rarely reported paediatric and adult cadaveric larynges that contained pathologies. The donors were a 13-month-old male, a 10-year-old female with an infraglottic mucus retention cyst and a 74-year-old female with advanced polypoid degeneration and a mucus retention cyst. MR and molecular imaging data were corroborated using whole-organ histology. Our MR protocols imaged the larynges at 45-117 μm² in-plane resolution and capably resolved microanatomic structures that have not been previously reported radiographically-such as the vocal fold superficial lamina propria, vocal ligament and macula flavae; age-related tissue features-such as intramuscular fat deposition and cartilage ossification; and the lesions. Diffusion tensor imaging characterised differences in water diffusivity, primary tissue fibre orientation, and fractional anisotropy between the intrinsic laryngeal muscles, mucosae and lesions. MALDI-MS imaging revealed peptide signatures and putative protein assignments for the polypoid degeneration lesion and the N-glycan constituents of one mucus retention cyst. These imaging approaches have immediate application in experimental research and, with ongoing technology development, potential for future clinical application.

CT-based radiomics features in the prediction of thyroid cartilage invasion from laryngeal and hypopharyngeal squamous cell carcinoma

Background

Laryngeal and hypopharyngeal squamous cell carcinoma (LHSCC) with thyroid cartilage invasion are considered T4 and need total laryngectomy. However, the accuracy of preoperative diagnosis of thyroid cartilage invasion remains lower. Therefore, the purpose of this study was to assess the potential of computed tomography (CT)-based radiomics features in the prediction of thyroid cartilage invasion from LHSCC.

Methods

A total of 265 patients with pathologically proven LHSCC were enrolled in this retrospective study (86 with thyroid cartilage invasion and 179 without invasion). Two head and neck radiologists evaluated the thyroid cartilage invasion on CT images. Radiomics features were extracted from venous phase contrast-enhanced CT images. The least absolute shrinkage and selection operator (LASSO) and logistic regression (LR) method were used for dimension reduction and model construction. In addition, the support vector machine-based synthetic minority oversampling (SVMSMOTE) algorithm was adopted to balance the dataset and a new LR-SVMSMOTE model was constructed. The performance of the radiologist and the two models were evaluated with receiver operating characteristic (ROC) curves and compared using the DeLong test.

Results

The areas under the ROC curves (AUCs) in the prediction of thyroid cartilage invasion from LHSCC for the LR-SVMSMOTE model, LR model, and radiologist were 0.905 [95% confidence interval (CI): 0.863 to 0.937)], 0.876 (95%CI: 0.830 to 0.913), and 0.721 (95%CI: 0.663–0.774), respectively. The AUCs of both models were higher than that of the radiologist assessment (all P < 0.001). There was no significant difference in predictive performance between the LR-SVMSMOTE and LR models (P = 0.05).

Conclusions

Models based on CT radiomic features can improve the accuracy of predicting thyroid cartilage invasion from LHSCC and provide a new potentially noninvasive method for preoperative prediction of thyroid cartilage invasion from LHSCC.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40644-020-00359-2.

Comparison of diagnostic performance between CT and MRI for detection of cartilage invasion for primary tumor staging in patients with laryngo-hypopharyngeal cancer: a systematic review and meta-analysis

OBJECTIVE

To compare the diagnostic performance of contrast-enhanced CT with that of MRI in the detection of cartilage invasion in patients with laryngo-hypopharyngeal cancer.

METHODS

A systematic literature search in the Ovid-MEDLINE and EMBASE databases was performed for studies reporting diagnostic accuracy of CT and/or MRI in detecting cartilage invasion from laryngo-hypopharyngeal cancer between 2000 and 2018. The pooled sensitivity and specificity, and their 95% confidence intervals were calculated for CT and MRI using bivariate random effects modeling. Subgroup and meta-regression analyses were performed. Indirect comparison was also performed by univariable meta-regression.

RESULT

Fourteen articles including 776 patients were included in the systematic review and meta-analysis: eight for CT, and six for MRI. CT and MRI showed pooled sensitivities of 66% (95% CI, 49-80%) and 88% (95% CI, 79-93%), and pooled specificities of 90% (95% CI, 82-94%) and 81% (95% CI, 76-84%), respectively. MRI showed significantly higher sensitivity than CT (p = 0.02). The specificities showed no statistically significant difference between CT and MRI (p = 0.39). The CT studies showed heterogeneity and a threshold effect, while MRI showed neither heterogeneity nor threshold effect. In the meta-regression analysis for CT, the type of cartilage analyzed (thyroid only vs. thyroid/cricoid/arytenoid, p < 0.001) was a significant factor influencing the heterogeneity in the diagnostic performance of the CT studies.

CONCLUSIONS

In conclusion, MRI has significantly higher sensitivity than CT for detecting cartilage invasion in patients with laryngo-hypopharyngeal cancer, without a significant difference in the specificity.

KEY POINTS

• MRI has significantly higher sensitivity than CT for detecting cartilage invasion in patients with laryngo-hypopharyngeal cancer.

Magnetic resonance imaging of the larynx in the pediatric population: A systematic review

BACKGROUND

Magnetic Resonance Imaging (MRI) techniques to image the larynx have evolved rapidly into a promising and safe imaging modality, without need for sedation or ionizing radiation. MRI is therefore of great interest to image pediatric laryngeal diseases. Our aim was to review MRI developments relevant for the pediatric larynx and to discuss future imaging options.

METHODS

A systematic search was conducted to identify all morphological and diagnostic studies in which MRI was used to image the pediatric larynx, laryngeal disease, or vocal cords.

RESULTS

Fourteen articles were included: three studies on anatomical imaging of the larynx, two studies on Diffusion Weighted Imaging, four studies on vocal cord imaging and five studies on the effect of anaesthesiology on the pediatric larynx. MRI has been used for pediatric laryngeal imaging since 1991. MRI provides excellent soft tissue contrast and good visualization of vascular diseases such as haemangiomas. However, visualization of cartilaginous structures, with varying ossification during childhood, and tissue differentiation remain challenging. The latter has been partly overcome with diffusion weighted imaging (DWI), differentiating between benign and malignant masses with excellent sensitivity (94-94.4%) and specificity (91.2-100%). Vocal cord imaging evolved from static images focused on vocal tract growth to dynamic images able to detect abnormal vocal cord movement.

CONCLUSION

MRI is promising to evaluate the pediatric larynx, but studies using MRI as diagnostic imaging modality are scarce. New static and dynamic MR imaging techniques could be implemented in the pediatric population. Further research on imaging of pediatric laryngeal diseases should be conducted.

Sclerosis of the arytenoid cartilage and glottic carcinoma: A clinical-pathological study

BACKGROUND

Given the relevance of any tumor invasion of the arytenoid cartilage or crico-arytenoid unit to the planning open partial horizontal laryngectomy (OPHL) for laryngeal squamous cell carcinoma (LSCC), it is important to have a reliable radiological test to assess impairments of these structures.

METHODS

We retrospectively compared the endoscopic, radiological, and pathological findings in patients with glottic LSCC who underwent OPHL.

RESULTS

The endoscopic finding of a reduced (impaired or absent) vocal cord motility proved more sensitive, with better positive and negative predictive values, but less specific than the radiological finding of complete arytenoid sclerosis in detecting histologically assessable infiltration of the arytenoid cartilage.

CONCLUSIONS

Endoscopy retains a key role in the preoperative workup for glottic LSCC. CT evidence of complete sclerosis of the arytenoid cartilage is related to a dangerous contiguity of the tumor to the cartilage.

Improvements in High Resolution Laryngeal Magnetic Resonance Imaging for Preoperative Transoral Laser Microsurgery and Radiotherapy Considerations in Early Lesions

As the benefits, limitations, and contraindications of transoral laser microsurgery (TLM) in glottic carcinoma treatments become better defined, pretreatment imaging has become more important to assess the case-specific suitability of TLM and to predict functional outcomes both for treatment consideration and patient counseling. Magnetic resonance imaging (MRI) is the preferred modality to image such laryngeal tumors, even though imaging the larynx using MRI can be difficult. The first challenge is that there are no commercial radiofrequency (RF) coils that are specifically designed for imaging the larynx, and performance in terms of coverage and signal-to-noise ratio is compromised using general-purpose RF coils. Second, motion in the neck region induced by breathing, swallowing, and vessel pulsation can induce severe image artifacts, sometimes rendering the images unusable. In this paper, we design a dedicated RF coil array, which allows high quality high-resolution imaging of the larynx. In addition, we show that introducing respiratory-triggered acquisition improves the diagnostic quality of the images by minimizing breathing and swallowing artifacts. Together, these developments enable robust, essentially artifact-free images of the full larynx with an isotropic resolution of 1 mm to be acquired within a few minutes.