ABSTRACT
PURPOSE: Paranasal anomalies, frequently identified in routine radiological screenings, exhibit diverse morphological characteristics. Due to the diversity of anomalies, supervised learning methods require large labelled dataset exhibiting diverse anomaly morphology. Self-supervised learning (SSL) can be used to learn representations from unlabelled data. However, there are no SSL methods designed for the downstream task of classifying paranasal anomalies in the maxillary sinus (MS). METHODS: Our approach uses a 3D convolutional autoencoder (CAE) trained in an unsupervised anomaly detection (UAD) framework. Initially, we train the 3D CAE to reduce reconstruction errors when reconstructing normal maxillary sinus (MS) image. Then, this CAE is applied to an unlabelled dataset to generate coarse anomaly locations by creating residual MS images. Following this, a 3D convolutional neural network (CNN) reconstructs these residual images, which forms our SSL task. Lastly, we fine-tune the encoder part of the 3D CNN on a labelled dataset of normal and anomalous MS images. RESULTS: The proposed SSL technique exhibits superior performance compared to existing generic self-supervised methods, especially in scenarios with limited annotated data. When trained on just 10% of the annotated dataset, our method achieves an area under the precision-recall curve (AUPRC) of 0.79 for the downstream classification task. This performance surpasses other methods, with BYOL attaining an AUPRC of 0.75, SimSiam at 0.74, SimCLR at 0.73 and masked autoencoding using SparK at 0.75. CONCLUSION: A self-supervised learning approach that inherently focuses on localizing paranasal anomalies proves to be advantageous, particularly when the subsequent task involves differentiating normal from anomalous maxillary sinuses. Access our code at https://github.com/mtec-tuhh/self-supervised-paranasal-anomaly .
ABSTRACT
PURPOSE: Since its release, Dupilumab has shown great results in treating severe uncontrolled CRSwNP. However, there is a lack of real-world data beyond 12 months of follow-up, and it is not clear to what extent biomarkers are appropriate for monitoring and predicting the Dupilumab therapy success. Hence, this study aims to analyze biomarkers for monitoring therapy, predicting therapy success and assess the effect of Dupilumab in real-world settings. METHODS: The follow-up was performed with 104 patients retrospectively up to 22 months, assessing SNOT-22, NPS, olfactometry, ACS, FEV-1, and blood biomarkers (total serum IgE, Eosinophils, ECP). Patients were divided into subgroups depending on their pretherapeutic biomarker levels and subsequent development was analyzed. RESULTS: There was substantially improvement in all clinical parameters up to 1 year and then continuously up to month 22. Patients with initially elevated baseline blood eosinophil counts (> 0.5 billion/L) had a trend of better SNOT-22 development after 1 year (- 12.19 points, p = 0.03). The course of total serum IgE showed moderate correlation with almost all clinical variables obtained. Therapy was well tolerated with only mild and transient adverse events. CONCLUSION: Dupilumab has considerably reduced symptoms and disease severity even beyond 1 year of treatment, supporting its role as targeted and effective treatment option for CRSwNP. Our data shows that total serum IgE is a promising biomarker for the monitoring during the treatment with Dupilumab. Elevated pre-therapeutic serum eosinophil counts may be a predictor of good subjective response to therapy. Larger cohorts and a long-term-follow-up over years are needed to further consolidate these findings.
ABSTRACT
OBJECTIVE: Computer aided diagnostics (CAD) systems can automate the differentiation of maxillary sinus (MS) with and without opacification, simplifying the typically laborious process and aiding in clinical insight discovery within large cohorts. METHODS: This study uses Hamburg City Health Study (HCHS) a large, prospective, long-term, population-based cohort study of participants between 45 and 74 years of age. We develop a CAD system using an ensemble of 3D Convolutional Neural Network (CNN) to analyze cranial MRIs, distinguishing MS with opacifications (polyps, cysts, mucosal thickening) from MS without opacifications. The system is used to find correlations of participants with and without MS opacifications with clinical data (smoking, alcohol, BMI, asthma, bronchitis, sex, age, leukocyte count, C-reactive protein, allergies). RESULTS: The evaluation metrics of CAD system (Area Under Receiver Operator Characteristic: 0.95, sensitivity: 0.85, specificity: 0.90) demonstrated the effectiveness of our approach. MS with opacification group exhibited higher alcohol consumption, higher BMI, higher incidence of intrinsic asthma and extrinsic asthma. Male sex had higher prevalence of MS opacifications. Participants with MS opacifications had higher incidence of hay fever and house dust allergy but lower incidence of bee/wasp venom allergy. CONCLUSION: The study demonstrates a 3D CNN's ability to distinguish MS with and without opacifications, improving automated diagnosis and aiding in correlating clinical data in population studies. LEVEL OF EVIDENCE: 3 Laryngoscope, 134:3927-3934, 2024.
Subject(s)
Diagnosis, Computer-Assisted , Magnetic Resonance Imaging , Maxillary Sinus , Humans , Male , Middle Aged , Female , Aged , Prospective Studies , Maxillary Sinus/diagnostic imaging , Diagnosis, Computer-Assisted/methods , Magnetic Resonance Imaging/methods , Paranasal Sinus Diseases/diagnostic imaging , Paranasal Sinus Diseases/epidemiology , Paranasal Sinus Diseases/diagnosis , Neural Networks, Computer , Sensitivity and SpecificityABSTRACT
PURPOSE: Paranasal anomalies are commonly discovered during routine radiological screenings and can present with a wide range of morphological features. This diversity can make it difficult for convolutional neural networks (CNNs) to accurately classify these anomalies, especially when working with limited datasets. Additionally, current approaches to paranasal anomaly classification are constrained to identifying a single anomaly at a time. These challenges necessitate the need for further research and development in this area. METHODS: We investigate the feasibility of using a 3D convolutional neural network (CNN) to classify healthy maxillary sinuses (MS) and MS with polyps or cysts. The task of accurately localizing the relevant MS volume within larger head and neck Magnetic Resonance Imaging (MRI) scans can be difficult, but we develop a strategy which includes the use of a novel sampling technique that not only effectively localizes the relevant MS volume, but also increases the size of the training dataset and improves classification results. Additionally, we employ a Multiple Instance Ensembling (MIE) prediction method to further boost classification performance. RESULTS: With sampling and MIE, we observe that there is consistent improvement in classification performance of all 3D ResNet and 3D DenseNet architecture with an average AUPRC percentage increase of 21.86 ± 11.92% and 4.27 ± 5.04% by sampling and 28.86 ± 12.80% and 9.85 ± 4.02% by sampling and MIE, respectively. CONCLUSION: Sampling and MIE can be effective techniques to improve the generalizability of CNNs for paranasal anomaly classification. We demonstrate the feasibility of classifying anomalies in the MS. We propose a data enlarging strategy through sampling alongside a novel MIE strategy that proves to be beneficial for paranasal anomaly classification in the MS.
Subject(s)
Maxillary Sinus , Neural Networks, Computer , Humans , Maxillary Sinus/diagnostic imaging , Magnetic Resonance Imaging , Tomography, X-Ray Computed , HeadABSTRACT
BACKGROUND AND OBJECTIVES: This paper presents an overview on nasal packing materials which are available in Germany. The current literature is analyzed whether there are robust criteria regarding use nasal packing after sinonasal surgery, whether there are fundamental and proven advantages or disadvantages of products, and what this means in clinical practice. MATERIALS AND METHODS: Selective literature analysis using the PubMed database (key words "nasal packing", "nasal tamponade", "nasal surgery", "sinonasal surgery", or "sinus surgery"), corresponding text books and resulting secondary literature. RESULTS AND CONCLUSIONS: Because of systematic methodological shortcomings, the literature does not help in the decision-making about which nasal packing should be used after which kind of sinonasal surgery. In fact, individual approaches for the many different clinical scenarios are recommended. In principle, nasal packing aims in hemostasis, should promote wound healing, and should not result in secondary morbidity. Nasal packing materials should be smooth (non-absorbable materials), inert (absorbable materials), and should not exert excessive pressure. Using non-absorbable packing entails the risk of potentially lethal aspiration and ingestion. For safety reasons inpatient control is recommended as long as this packing is in situ. With other, uncritical packing materials and in patients with special conditions, outpatient control could be justified.
Subject(s)
Nasal Surgical Procedures , Sinusitis , Humans , Sinusitis/surgery , Nose , Epistaxis/prevention & control , Epistaxis/surgery , Wound Healing , Nasal Surgical Procedures/methods , Endoscopy/methodsABSTRACT
Objectives: In head and neck squamous cell carcinoma (HNSCC), tumors negative for Human Papillomavirus (HPV) remain a difficult to treat entity and the morbidity of current multimodal treatment is high. Radiotherapy in combination with molecular targeting could represent suitable, less toxic treatment options especially for cisplatin ineligible patients. Therefore, we tested dual targeting of PARP and the intra-S/G2 checkpoint through Wee1 inhibition for its radiosensitizing capacity in radioresistant HPV-negative HNSCC cells. Materials and methods: Three radioresistant HPV-negative cell lines (HSC4, SAS, UT-SCC-60a) were treated with olaparib, adavosertib and ionizing irradiation. The impact on cell cycle, G2 arrest and replication stress was assessed through flow cytometry after DAPI, phospho-histone H3 and γH2AX staining. Long term cell survival after treatment was determined through colony formation assay and DNA double-strand break (DSB) levels were assessed through quantification of nuclear 53BP1 foci in cell lines and patient-derived HPV± tumor slice cultures. Results: Wee1 and dual targeting induced replication stress but failed to effectively inhibit radiation-induced G2 cell cycle arrest. Single as well as combined inhibition increased radiation sensitivity and residual DSB levels, with the largest effects induced through dual targeting. Dual targeting also enhanced residual DSB levels in patient-derived slice cultures from HPV-negative but not HPV+ HNSCC (5/7 vs. 1/6). Conclusion: We conclude that the combined inhibition of PARP and Wee1 results in enhanced residual DNA damage levels after irradiation and effectively sensitizes radioresistant HPV-negative HNSCC cells. Ex vivo tumor slice cultures may predict the response of individual patients with HPV-negative HNSCC to this dual targeting approach.
ABSTRACT
Background: The lack of detectable precancerous lesions poses challenges to the early detection of human papillomavirus-driven oropharyngeal cancer (HPV-OPC). Antibodies against HPV16 early proteins, especially E6, are uniquely sensitive and specific biomarkers detectable years prior to HPV-OPC diagnosis. Thus, HPV16 early protein serology warrants clinical investigation for HPV-OPC screening. Methods: Using multiplex serology, we analyzed HPV16 serum antibodies of the first 5000 participants (n=4,424 sera, recruited 2016-2017) of the Hamburg City Health Study, a population-based prospective cohort (45-74 years). Participants seropositive for HPV16 E6 and at least one additional early protein (E1, E2, E7) were considered at high risk for HPV-OPC development and invited to six-monthly non-invasive head and neck follow-up (FU) examinations (visual inspection, endoscopy, ultrasonography, performed 2019-2020). Participants with suspicious lesions were examined by magnetic resonance imaging and panendoscopy with biopsy. Histologically confirmed OPC cases were treated according to standard of care. Findings: In total, 35 out of 4,424 study participants (0·8%, 95% confidence interval (CI) 0·6-1·1%) were HPV16 E6 seropositive. Among these, eleven (0·3%, 95%CI 0·1-0·5%) were considered at high risk for HPV-OPC of which nine were successfully re-contacted and invited to regular clinical FU examinations. Two males and one female were diagnosed with stage I HPV-OPC within 1·3 years of clinical FU (3-4 years after initial blood draw), representing one diagnosis of prevalent advanced disease, one incident diagnosis of advanced disease, and one incident diagnosis of early disease. The remaining participants showed no detectable signs of cancer, and undergo regular examinations (median clinical FU: 1·0 years, median total FU from blood draw to last clinical FU visit: 4·7 years). Interpretation: HPV16 early antibodies allowed identifying three asymptomatic stage I HPV-OPC patients, out of eleven participants considered at high risk. However, two of the three cases already showed signs of advanced disease at diagnosis. Targeting multiple early proteins may considerably improve the positive predictive value of HPV16 serology and may have clinical utility for HPV-OPC screening. Funding: This work was funded by DKFZ and UKE intramural funding.
ABSTRACT
Background. Cholesteatoma is a destructive process of the middle ear resulting in erosion of the surrounding bony structures with consequent hearing loss, vestibular dysfunction, facial paralysis, or intracranial complications. The etiopathogenesis of cholesteatoma is controversial but is associated with recurrent ear infections. The role of intracellular innate immune receptors, the NOD-like receptors, and their associated signaling networks was investigated in cholesteatoma, since mutations in NOD-like receptor-related genes have been implicated in other chronic inflammatory disorders. Results. The expression of NOD2 mRNA and protein was significantly induced in cholesteatoma compared to the external auditory canal skin, mainly located in the epithelial layer of cholesteatoma. Microarray analysis showed significant upregulation for NOD2, not for NOD1, TLR2, or TLR4 in cholesteatoma. Moreover, regulation of genes in an interaction network of the NOD-adaptor molecule RIPK2 was detected. In addition to NOD2, NLRC4, and PYCARD, the downstream molecules IRAK1 and antiapoptotic regulator CFLAR showed significant upregulation, whereas SMAD3, a proapoptotic inducer, was significantly downregulated. Finally, altered regulation of inflammatory target genes of NOD signaling was detected. Conclusions. These results indicate that the interaction of innate immune signaling mediated by NLRs and their downstream target molecules is involved in the etiopathogenesis and growth of cholesteatoma.
Subject(s)
Cholesteatoma/genetics , Inflammation/genetics , Nod2 Signaling Adaptor Protein/biosynthesis , Receptor-Interacting Protein Serine-Threonine Kinase 2/biosynthesis , CARD Signaling Adaptor Proteins/biosynthesis , CARD Signaling Adaptor Proteins/genetics , Calcium-Binding Proteins/biosynthesis , Calcium-Binding Proteins/genetics , Cholesteatoma/etiology , Cholesteatoma/pathology , Cytoskeletal Proteins/biosynthesis , Cytoskeletal Proteins/genetics , Gene Expression Regulation , Gene Regulatory Networks , Humans , Inflammation/pathology , Microarray Analysis , Nod2 Signaling Adaptor Protein/genetics , RNA, Messenger/biosynthesis , Receptor-Interacting Protein Serine-Threonine Kinase 2/genetics , Signal TransductionABSTRACT
The increased knowledge regarding HPV-infections in head and neck squamous cell carcinoma (HNSCC) has unexpectedly contributed to several uncertainties related to i) prevalence diversities depending on tumour site and geographical origin of the patients, ii) proportion of HPV-driven tumours among HPV-DNA-positive cases, and iii) identification of patients with HPV-attributed survival benefit. To investigate this heterogeneity, we analysed 307 HNSCC cases (tonsillar, n=135; non-tonsillar, n=172) from eight health care centers mostly from Northern Germany and determined HPV-DNA/mRNA and p16INK4A-status and combined results with the patient outcome. Overall HPV-DNA prevalence rate was 23.5% (72/307); attributed to: 43.7% (59/135) and 7.6% (13/172) tonsillar and non-tonsillar cases, respectively. Among these, 96.6% tonsillar and 38.5% non-tonsillar SCC were HPV-mRNA-positive. Although the study cohort was composed of patients from regions of rather close proximity, prevalence rates showed diversities of up to 40% in HNSCC subsite analysis with the lowest prevalence for tonsillar SCC in metropolitan areas (22.2%) vs. 50.9% in rural areas. Survival analysis identified p16INK4A alone as strongest predictor, followed by HPV-DNA-status alone or in combination with p16INK4A. This survival benefit was shown for tonsillar and non-tonsillar cases. Smoking significantly correlated with HPV-status, however, it does not influence survival when stratified for HPV. In conclusion, the data emphasize the urge for further data on HPV-infection in HNSCC to, e.g. clarify to what extent survival benefits of p16INK4A-positive patients are truly attributed to HPV-infections.
Subject(s)
Carcinoma, Squamous Cell/complications , Carcinoma, Squamous Cell/epidemiology , Head and Neck Neoplasms/complications , Head and Neck Neoplasms/epidemiology , Papillomaviridae/genetics , Papillomavirus Infections/complications , Papillomavirus Infections/epidemiology , Adult , Aged , Aged, 80 and over , Carcinoma, Squamous Cell/pathology , Carcinoma, Squamous Cell/virology , Female , Genetic Variation , Geography , Germany/epidemiology , Head and Neck Neoplasms/pathology , Head and Neck Neoplasms/virology , Humans , Male , Middle Aged , Prevalence , Squamous Cell Carcinoma of Head and NeckABSTRACT
It has been proposed that p16(INK4A) qualifies as a surrogate marker for viral oncogene activity in head and neck cancer (HNSCC). By analyzing 78 HNSCC we sought to validate the accuracy of p16(INK4A) as a reliable marker of active HPV infections in HNSCC. To this end we determined HPV DNA (HPVD) and E6*I mRNA (HPVR) expression status and correlated these results with p16(INK4A) staining. In tonsillar SCC 12/20 were HPVD+ and 12/12 of these showed active HPV infections whereas in non-tonsillar SCC 10/58 were HPVD+ and 5/10 showed active HPV infections. Thus, we prove about 8% of non-tonsillar SCC to be also correlated with HPV-associated carcinogenesis. Strikingly, 3/14 (21.4%) of tonsillar and non-tonsillar HPVD+/HPVR+ cases did not show p16(INK4A) overexpression and these cases would have been missed when applying initial p16(INK4A) staining only. However, in 13 cases negative for HPV, DNA p16(INK4A) was overexpressed. In conclusion, our data confirm tonsillar SCC to be predominantly but not only associated with active HPV infections. Furthermore, our data show that p16(INK4A) overexpression is not evident in a subgroup of HNSCC with active HPV infection. Definitive HPV data should therefore be utilized in diagnostics and treatment modalities of HPV positive and HPV negative HNSCC patients, resulting in a paradigm shift regarding these obviously different tumor entities.