RPS 2108 - Head and neck cancer imaging: what's new?

Author Block: A. h. Shah, A. Chandra, S. Sen, S. Mukhopadhyay, P. Ghosh, A. Chatterjee, A. Gehani, J. Khoda, A. Patra; Kolkata/IN
Purpose: NI-RADS has been validated as a single-time point tool for recurrence prediction. Yet in the clinic, patients return repeatedly for follow-up, and radiologists often see categories shift over time. The prognostic meaning of these longitudinal NI-RADS trajectories has never been systematically studied. We evaluated whether stability or change in NI-RADS scores across surveillance scans influences recurrence risk.
Methods or Background: We retrospectively studied 628 patients with head and neck squamous cell carcinoma treated between 2012 and 2024 who had ≥3 consecutive surveillance CT or MRI scans. Two head and neck radiologists independently assigned NI-RADS categories. Trajectories were defined as: stable low-risk (repeated NI-RADS 1–2), escalating risk (progression from NI-RADS 2 → 3/4), or stable high-risk (repeated NI-RADS 3–4). Outcomes were verified by histopathology, MDT consensus, or ≥12-month follow-up. Kaplan–Meier survival and hazard ratios (HR) for progression-free survival (PFS) were calculated.
Results or Findings: We retrospectively studied 628 patients with head and neck squamous cell carcinoma treated between 2012 and 2024 who had ≥3 consecutive surveillance CT or MRI scans. Two head and neck radiologists independently assigned NI-RADS categories. Trajectories were defined as: stable low-risk (repeated NI-RADS 1–2), escalating risk (progression from NI-RADS 2 → 3/4), or stable high-risk (repeated NI-RADS 3–4). Outcomes were verified by histopathology, MDT consensus, or ≥12-month follow-up. Kaplan–Meier survival and hazard ratios (HR) for progression-free survival (PFS) were calculated.
Conclusion: Recurrence risk is not defined by a single scan but by the trajectory of NI-RADS scores over time. Stability reassures, escalation warns, and downward movement inspires hope. Incorporating trajectory analysis into structured reports could help radiologists guide MDT decisions more confidently, aligning imaging with the lived journey of cancer survivors.
Limitations: Single-center retrospective study; modality allocation by subsite.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: A. Samad, S. S. A. Khatian; Karachi/PK
Purpose: Orbital metastases are not very common but they tend to be signify progression of malignancy in the entire body and may considerably affect the prognosis of the patient. They can have an appearance similar to primary orbital tumors or inflammatory disorders/changes in clinical practice, and radiologic identification is important.
Methods or Background: A cross-sectional study was designed as a descriptive study that was carried out at Radiology, JPMC -Karachi. 144-patients who were known to have cross-sectional imaging (CT and/or MRI) and was histologically confirmed to have orbital metastases were included. Data was analyzed based on demographic profile, primary site of tumor, clinical presentation and detailed imaging features such as location, margins, enhancement, necrosis and bone involvement.
Results or Findings: Summary N=104 (mean age of patients 55.6 ±11.8 years, 62 females and 42 males).The most common primary malignancy was breast carcinoma (n=46, 44%), lung carcinoma (n=29, 28%), renal cell carcinoma (n=15, 14%), and other primaries (n=14, 14%) such as gastrointestinal and thyroid malignancies.
Breast carcinoma metastases were usually poorly defined, infiltrating lesions of soft tissue in fat in the orbit and extraocular muscles, without extensive bone destruction.
.Lung carcinoma metastases were commonly well-defined and rapidly growing and massively enlarging the orbit with evidence of necrotic tissue.
Renal cell carcinoma metastases presented with intensely enlarging, well-defined, vascular lesions, which are often accompanied with hemorrhage and bone erosion.
Conclusion: Orbital metastases though very uncommon and has characteristic imaging appearances that depends on the nature of the underlying primary tumor. The knowledge of these radiological patterns may help radiologists to make nearly accurate diagnosis proposal, facilitate systemic assessment, and prevent unnecessary procedures. It is also especially important in oncology patients with new orbital complaints who have to be identified early.
Limitations: Nil
Funding for this study: Nil
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Nil

Author Block: U. B. Abdullaeva¹, B. Pape², J. Hirvonen³; ¹Tashkent/UZ, ²Turku, Vaasa/FI, ³Tampere/FI
Purpose: To review the diagnostic performance of MRI, CT and US for detecting vascular invasion in head and neck primary tumours or metastases, using histopathology or surgical findings as the reference standard.
Methods or Background: This review was registered in Prospero (CRD42024584016). We searched PubMed and Embase for studies in English, published from 01.01.1990 to 31.12.2024, that used CT, MRI or ultrasound to assess vascular involvement in head and neck primary tumours or metastases, confirmed by histology or surgery and provided data to calculate diagnostic accuracy. Outcomes included sensitivity, specificity, PPV, and NPV, with heterogeneity assessed using Cochran’s Q test.
Results or Findings: In the systematic review, we included 28 studies with 1687 patients (mean age 57.2): 17 retrospective, seven prospective, one mixed, and the rest unspecified. The meta-analysis included 19 studies (1041 patients), while nine studies (646 patients) were analysed qualitatively.
Overall diagnostic accuracy was 0.87 (95% CI: 0.81–0.91). Pooled sensitivity, specificity, PPV, and NPV were 0.63 (0.48–0.75), 0.81 (0.73–0.87), 0.60 (0.52–0.68), and 0.90 (0.78–0.96), respectively. US performed best (accuracy 0.95, sensitivity 0.86, specificity 0.93, NPV 0.96, PPV 0.73), MRI showed intermediate results, and CT the lowest. Moderation effect for accuracy was significant (p = 0.041), with US outperforming CT (p = 0.023). Significant heterogeneity was found for sensitivity (Q = 57.81), specificity (Q = 109.01), NPV (Q = 79.42), and accuracy (Q = 170.25) (all p < 0.0001).
Conclusion: US is the most effective and precise modality, outperforming MRI and CT, though MRI is superior to CT. Overall modest sensitivity suggests the possibility of false negative findings. Significant heterogeneity requires integration of imaging findings with intraoperative validation.
Limitations: The study is largely retrospective, with sampling bias, incomplete data, and significant heterogeneity.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: E. Gattuso, F. PASSALACQUA, F. Bencivinni, A. Lo Casto; Palermo/IT
Purpose: Human papillomavirus (HPV) status plays a key role in oropharyngeal squamous cell carcinoma (OPSCC). Our study aims to evaluate MRI features of locoregional metastatic lymph nodes and apparent diffusion coefficient (ADC) values of the primary tumor in HPV+ and HPV– OPSCC to identify distinguishing patterns.
Methods or Background: MRI scans of 87 patients with suspected OPSCC were retrospectively reviewed. Exclusion criteria were prior treatment, uncertain diagnosis, or non-squamous histology. Twenty-four patients with pre-biopsy MRI and histological confirmation, including p16 immunohistochemistry evaluation, were included. For each case, lymph node morphology and mean ADC values of the primary lesion were assessed.
Results or Findings: Among HPV+ patients (14/24), 12/14 (85.7%) had nodal involvement, mainly ipsilateral (11/12; 91.7%). Mean nodal diameter was 2.33 cm (range 1.1–4.4); cystic morphology was seen in 7/12 (58.3%). Jugular vein thrombosis occurred in 3/12 (25%), and extracapsular spread in 1/12 (8.3%). Mean primary ADC was 0.79×10⁻³ mm²/s (range 0.6–1.0).
In HPV– patients (10/24), 6/10 (60%) had nodal metastases, ipsilateral in 3/6 (50%) and bilateral in 3/6 (50%). Mean diameter was 2.53 cm (range 1.5–4.5), with solid morphology in 5/6 (83.3%). Mean primary ADC was 1.15×10⁻³ mm²/s (range 0.7–1.5).
Conclusion: HPV+ OPSCC was associated with smaller, predominantly ipsilateral cystic nodes and lower ADC values, while HPV– OPSCC showed larger, bilateral, solid nodes and higher ADC values. MRI thus offers a valuable non-invasive tool for HPV status characterization.
Limitations: This was a single-center study with a relatively small sample size, which may limit the generalizability of the findings. Furthermore, no correlation with patient outcome or survival was performed, preventing assessment of the prognostic significance of the imaging features.
Funding for this study: This study received no external funding.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: Y-C. Tsai, W-C. Lin; Kaohsiung/TW
Purpose: Patients with radioactive iodine–refractory (RAIR) metastatic or locally advanced thyroid cancer have a poor prognosis and limited treatment options. This study aimed to evaluate the efficacy of targeted therapy alone versus targeted therapy in combination with local treatment in this patient population.
Methods or Background: From 2017 to 2024, a total of 52 patients aged ≥18 years with RAIR locally advanced or metastatic thyroid cancer receiving targeted therapy were evaluated. Baseline imaging with computed tomography (CT) was performed prior to initiation of therapy. Disease progression was assessed using serial CT and/or PET/CT (Positron Emission Tomography-Computed Tomography) imaging, based on a combination of Response Evaluation Criteria in Solid Tumors (RECIST) and PET Response Criteria in Solid Tumors (PERCIST). Associations between clinical factors and disease progression, as well as treatment-related adverse events, were also analyzed.
Results or Findings: Among the 52 patients, 20 received additional local treatment. This group demonstrated significantly improved progression-free survival (PFS) compared to those receiving targeted therapy alone: median PFS of 35.2 months (95% CI: 22.95–47.42) vs. 18.9 months (95% CI: 11.23–26.74); p = 0.038. Multivariate analysis identified combination therapy as an independent factor associated with favorable PFS (HR: 0.327; 95% CI: 0.140–0.763; p = 0.010). There was no significant difference in the incidence of adverse events between the two groups. The most common adverse event was proteinuria (21 patients, 40.4%).
Conclusion: Local treatment should be actively considered in eligible RAIR locally advanced or metastatic thyroid cancer patients receiving targeted therapy, as it can significantly enhance survival outcomes with acceptable safety.
Limitations: This study was limited by its retrospective, single-center design, relatively small sample size, heterogeneous treatment and limited follow-up duration. Further investigation is needed to clarify the long-term outcome.
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Chang Gung medical foundation institutional review board

Author Block: S. Li, H. LI, L-Z. Liu; Guang Zhou/CN
Purpose: This study aimed to develop a new clinical staging system integrating detailed MRI features for non-metastatic nasopharyngeal carcinoma, to address the limitations in risk discrimination of the current 9th AJCC system.
Methods or Background: This retrospective study included 1,897 patients with non-metastatic NPC from three hospitals. MRI features for primary tumor and metastatic lymph nodes were evaluated. The primary endpoint was 5-year overall survival (OS). T, N and clinical stage were optimized using univariate and multivariate analyses. Survival curves were generated with the kaplan–meier method. The proposed system was compared with the authoritative 8th/9th AJCC/UICC staging systems using risk stratification, C-index, and AIC.
Results or Findings: Among T3 patients, those with severe skull base-foramina invasion (HR:1.99, 95% CI:1.22–3.34, P=0.006) were proposed for upstaging to T4, whereas those with only slight skull base-bone invasion (HR:0.54, 95%CI:0.34–0.85, adjusted P=0.009) were recommended for downstaging to T2 (SBFI_T). The number of MRI-positive lymph nodes was independent prognosticator, supporting an N categorization based on nodal count (LNN_N: N0, 0; N1,1-4; N2, 5-9; N3,≥9). The proposed stage, which integrates SBFI_T and LNN_N, grouped T2N0 and T1N0 as stage Ia, and T1–2N1 as stage Ib.
The proposed SBFI_T,LNN_N and stage systems outperformed the 8th/9th editions, achieving higher C-indices, lower AIC values, and significant risk stratification for OS across all subgroups (all P < 0.05).
Conclusion: The proposed staging system outperforms current mainstream systems. It also emphasizes critical MRI features that require clinical attention and highlights their role in accurate prognostic prediction.
Limitations: As only M0 cases were included, the inclusion of metastatic cases is needed in future studies to optimize the M-stage and develop a comprehensive staging framework.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The ethics committee of Sun Yat-sen university cancer center

Author Block: P. Dong; Changchun/CN
Purpose: Ferroptosis has become a new approach for antitumor treatment. However, the insufficient accumulation and poor penetration of ferroptosis inducers deep in tumors greatly limit their therapeutic effects. Herein, we aim to construct a cascade penetrating metal-polyphenol ultrasonic molecular probe, Fe3+Cur-PFP@IR780-LIP (FCIPL). Under ultrasonic stimulation, the nanoparticles can targeted cascade-penetrate deep into tumors, induce domino-ferroptosis reaction, enhance sonodynamic therapy (SDT), and realize multimodal imaging.
Methods or Background: FCIPL were based on the core of FCr consists of ferroptosis inducer curcumin and Fe3+, and co-embedded with PFP in IR780-loaded liposomes. We examined their characterization, basic properties, the safety, uptake capacity, targeting function, cytotoxicity, ferroptosis potentiating SDT effect and mechanism of action, the deep penetration and multimodal imaging capabilities of nanoparticles in vitro and in vivo.
Results or Findings: The FCIPL were regular spherical with uniform size, hydrated particle size of 218.6nm, with good dispersion, stability, and encapsulation rates. Ultrasonic excitation and tumor microenvironment can promote the disassembly of FCIPL, releasing Cur and Fe2+, jointly promotes the domino effect of lipid peroxide accumulation. Vivo and vitro experiments confirmed the good tumor targeting, cascade response and deep penetration ability of the FCIPL with good biosafety, and also verified the ferroptosis potentiating SDT efficacy of the nanoprobe and its mechanism of action, which together effectively demonstrated excellent tumor killing effects. In terms of imaging, FCIPL successfully realizing multi-modal imaging integrating ultrasound, photoacoustic, MRI and fluorescence imaging.
Conclusion: A cascade penetrating metal-polyphenol ultrasonic molecular probe FCIPL was successfully synthesised. Realized multimodal imaging in diagnosis. In terms of treatment, ferroptosis in deep tumors was achieved through targeted osmosis and cascaded response drug delivery system, enhanced sonodynamic therapy, and provide new ideas for integrating ATC diagnosis and treatment.
Limitations: Subsequent clinical transformation.
Funding for this study: Technology Department of Jilin Province (No. YDZJ202201ZYTS248)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Laboratory Animal Center, College of Life Science, Jilin University

Author Block: F. Wu, S. Li, D. Cao, H. Ya, H. LI, B. Chen; Guangzhou/CN
Purpose: To explore the prognostic value of the grading of skull base structure (including soft tissue, bone, and skull base foramina) invasion (SBSI) and their potential improvement for the T staging in nasopharyngeal carcinoma (NPC).
Methods or Background: This retrospective study enrolled M0 NPC patients from three medical institutions (1320, 432, and 249, respectively). The primary endpoint was 5-year overall survival (OS). Cluster heatmaps and network analysis were used to explore the grading of SBSI. Univariate and multivariate analyses with Kaplan–Meier method are performed to compare survival outcomes.
Results or Findings: This study included 2001 patients (median age, 46 years [IQR, 39-55]; 1469 males). Multivariate analyses identified invasion of the jugular foramen or foramen ovale (severe skull base foramina invasion, SSBFI) as an independent prognostic factor (HR: 2.24, 95% CI: 1.42-3.54, p=0.001), which conferred a poor prognosis similar to T4 disease(5-year OS: 69.7% vs 75.6%, p=0.421). Conversely, the prognosis of patients with pterygoid and sphenoid base (minor skull base invasion, MSBI) was superior to other T3 skull base involvement (5-year OS: 91% vs 82.4%, p= 0.003), was comparable to that of T2 disease (5-year OS: 91% vs 86.3%, p= 0.267). A proposed T staging(proposed_T) system was created by upstaging SSBFI to T4 and downstaging MSBI to T2. This proposed_T category outperformed the 8th/9th edition and other proposed T staging systems in risk stratification and prognostic prediction.
Conclusion: The grade of SBSI is a critical factor for risk stratification in T3 NPC patients, and its incorporation into T staging leads to a more precise prognostic stratification.
Limitations: The study design was a retrospective analysis and did not include a prospective validation cohort.
Funding for this study: National Natural Science Foundation of China ( No.82171906 )
Guangdong Basic and Applied Basic Research Foundation (2025A1515011590)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The ethical approval for this study was granted by the institutional ethics committees of the three hospitals (Approval No: B2019-222-01)