Despite advances in molecular oncology, head and neck cancer diagnosis still relies on anatomical site and histopathology. In up to 5% of metastatic cases, no identifiable primary tumor is found, defining carcinoma of unknown primary (CUP) and posing significant diagnostic and therapeutic challenges. We conducted a narrative review of the literature focusing on current diagnostic and treatment strategies for head and neck CUP, including imaging, viral biomarkers (HPV, EBV), minimally invasive procedures, and transoral robotic surgery (TORS). While TORS plays a dual role in diagnosis and treatment, a multidisciplinary approach of neck dissection, radiotherapy, and chemotherapy in high-risk patients, remains the cornerstone of management. Personalization is increasingly crucial in this comprehensive context, mainly involving treatment tailoring and radiotherapy planning, where limited evidence on field definition is available. Ongoing trials, such as UK FIND study, highlight the potential of targeted surgery to tailor and possibly reduce radiotherapy fields for better patient outcomes and improved quality of life.

6120 Background: Sinonasal carcinomas (SNCs) are rare malignancies with poor prognosis. Multimodal treatments including induction chemotherapy (ICT), surgery and radiotherapy (RT) - modulated by histology and response to ICT - are often used, aiming to improve oncological outcome in terms of local control and survival. Two phase II clinical studies published in 2023 assessed the role of ICT in SNCs, SINTART1 and 2 for resectable and unresectable tumors, respectively [PMIDs: 37164774, 37163806]. The current work aims at reporting the long-term follow-up (FUP) data for both trials. Methods: The FUP was updated as of January 2026 for patients enrolled in both clinical studies. Median FUP was estimated with reverse Kaplan-Meier method. The following survival times were analyzed with Kaplan-Meier method in each cohort: overall survival (OS), disease-free survival (DFS), loco-regional-free survival (LRFS), distant metastasis-free survival (DMFS). Results: The updated median (m) survival times (in months) with their 95% confidence intervals (CI) and the event rates are detailed in Table 1. Pooling together the 2 studies, 30% of patients (18/60) were alive at last follow-up. Conclusions: SINTART 1 and 2 represent, to date, the largest prospective cohorts with long-term survival data in SNCs. With a median FUP exceeding 9 years, these results provide a robust benchmark for ICT-based multimodal strategies in this setting. Survival outcomes remain consistent with the initial reports, indicating that prognosis is largely determined within the first 2-3 years, when most deaths occur (approximately 70%). The near-overlap of mDFS and mLRFS identifies loco-regional failure as the predominant pattern of recurrence. Moreover, the short interval between relapse and death underscores the limited opportunity for salvage. These findings collectively emphasize the need for more effective local-intensification approaches and novel systemic agents. Additional analyses are underway to identify prognostic and predictive factors. Clinical trial information: NCT02099175 ; NCT02099188 . Resectable (SINTART1)N=35 Unresectable (SINTART2)N=25 mFUP (95% CI) 108.26 (97.37-129.84) 103.22 (91.68-NR) mOS (95% CI)Events 37.53 (21.97-98.36)66% 27.07 (11.28-65.03)76% mDFS (95% CI)Events 26.25 (15.43-80.43)71% 17.1 (7.89-37.76)80% mLRFS (95% CI)Events 26.25 (15.43-80.43)71% 18.95 (7.89-39.05)80% mDMFS (95% CI)Events 34.47 (21.84-96.32)69% 26.88 (9.31-46.78)76%

Figure S1. Study flow diagram and FFPE vs frozen tumor biopsies comparisons

Abstract Squamous carcinomas of the oral cavity and head and neck share molecular, microbial, and immune features. Dysbiosis and intratumoral microbiota are increasingly implicated in epithelial transformation, with taxa such as Fusobacterium nucleatum linked to inflammation, hypoxia, immune evasion, and metastasis across squamous sites. Defining how host-microbe-immune interactions evolve from oral premalignant lesions to invasive tumors could reveal early biomarkers and modifiable microbial targets for prevention, risk stratification, and intervention. We analyzed samples from 66 oral potentially malignant diseases (OPMD) and 74 head and neck squamous cell carcinoma (HNSCC) using a multi-omics approach. Cohorts were profiled via RNA-seq, shotgun metagenomics, and multiplex cytokine assays. Epigenetic profiling of 20 HNC samples was performed using Nanopore sequencing to assess genome-wide DNA methylation. In OPMDs, a high-risk hypoxic subtype showed an elevated incidence of malignant transformation (p < 0.0001), reduced microbial diversity, and significantly elevated abundance of Fusobacterium nucleatum and Leptotrichia wadei (Log2FC > 1, FDR < 0.05). F. nucleatum activity correlated strongly (R>0.7) with keratinocytes and pericytes. While HNSCC alpha diversity remained stable, community structure shifted (PERMANOVA p=0.001), marked by an enrichment of Lactobacillus spp. driving palmitate metabolism, whereas OPMD favored starch degradation and nitrate reduction. Immune profiling revealed a profound dysregulation within head and neck tumors. While OPMD was characterized by high levels of pro-inflammatory mediators (TNF-α, IL-6), these cytokines were markedly downregulated in HNSCC. This suggests a functional shift from an acute inflammatory response in OPMDs toward a chronic, immune-suppressive state in established carcinoma. Finally, our epigenetic analysis linked oral dysbiosis with host metabolism, revealing significant differential methylation in overweight HNC patients. Our analysis revealed significant hypomethylation affecting key oncogenic and immunomodulatory genes, including NOTCH1, STAT3, and the antigen-presenting gene HLA-DRB1. Our findings demonstrate a conserved microbial trajectory in the progression of head/neck carcinomas, characterized by the enrichment of Fusobacterium and a functional switch from a pro-inflammatory to an immune-evasive tumor microenvironment. These results point to a shared microbial etiology in squamous cancers, and the ongoing epigenetic analysis will provide deeper mechanistic insights into how these host-microbe interactions drive malignant transformation. Citation Format: Armando G. Licata, Cristina Gurizzan, Deborah Lenoci, Marta Lucchetta, Carlo Resteghini, Luigi Lorini, Rosalba Miceli, Paolo Bossi, Loris De Cecco. Host-microbe-epigenetic crosstalk: Dissecting the role of microbiome, immune dysregulation, and DNA methylation in squamous cell carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2866.

Figure S5. Patients with high-B2M tumors have a better survival than patients with B2M-low tumors when treated with PD(L)1 inhibitors, related to figure 5

Figure S7. Genomic alterations associated with substance abuse and primary disease location.

Table S1. List of copy number alterations analyzed. Table S2. List of oncogenic pathways analyzed. Table S3. List of genes present in each module. Table S4. mIF panels. Table S5. mIF phenotypes. Table S6. Frequencies of genes alterations in HPV-positive R/M SCCHN and in HPV-negative R/M SCCHN, related to Figure 1A-E. Table S7. Frequencies of genes alterations in smoker and/or drinker HPV-negative patients and in non-smoker and non-drinker HPV-negative patients, related to Figure S5A. Table S8. Frequencies of genes alterations in HPV-negative R/M SCCHN according to primary disease location related to Figure S5B. Table S9. Frequencies of genes alterations in locoregional recurrence only and in distant metastatic disease SCCHN, related to Figure 2A-B. Table S10. Frequencies of genes alterations according to number of R/M treatment lines prior biopsy, related to Figure 3B and Figure S7A-B.

Human papillomavirus-positive (HPV+) oropharyngeal squamous cell carcinoma (OPSCC) is a peculiar entity, with distinct patient, tumor, and biological characteristics, and a different prognosis compared to HPV-negative (HPV-) OPSCC. Due to the rising incidence, there is a need to develop novel therapeutic approaches, especially considering the long-term morbidities of traditional treatments such as surgery and chemoradiotherapy. In this regard, therapeutic vaccines targeting HPV epitopes have been put at the forefront of the immunotherapeutic strategies for HPV+ OPSCC. Multiple clinical trials are investigating their efficacy and safety in the advanced setting, more frequently as a combination therapy. As for the early setting, HPV therapeutic vaccines could represent a strategy to further deepen responses and to facilitate de-escalation of standard treatment. This review aims to provide the clinician with useful and up-to-date information on the current advances in this field. To that end, we will provide the results of the main ongoing/completed clinical trials including patients with OPSCC, focusing on immunogenicity and clinical benefit, in both early and advanced setting. We will also map the challenges and limitations in this area to guide future research.

Figure S3. Three multiplex immunofluorescent staining of SCCHN samples with 3 different 6-plex panels and their deconvolution

Background: Microscopic tumor spread beyond radiologically evident margins poses a major challenge in the management of sinonasal and nasopharyngeal malignancies (SNNM). The biological behavior underlying this phenomenon, referred to as tumor dispersion (TD), remains poorly characterized, particularly regarding its spatial extent and histology-dependent patterns.Methods: This prospective study included patients undergoing surgical resection of SNNM with intraoperative surgical navigation (SN). Intraoperative biopsies were systematically mapped and spatially correlated with preoperative imaging. Multivariable logistic regression models evaluated the association between positive biopsy probability, distance from the gross tumor surface, and histologic characteristics.Results: Fifty-four patients and 569 biopsies were analyzed; 78 (13.7%) were positive. A significant inverse association between distance from the gross tumor surface and positive biopsy probability was observed, with an approximately 5.1% reduction in odds per millimeter (p-value<0.001). Histologic type remained independently associated with positive biopsy probability after distance adjustment (p-value<0.001). Adenoid cystic carcinoma (ACC), squamous cell carcinoma (SCC) with perineural invasion (PNI), other poorly differentiated carcinomas (OPDC), and mesenchymal malignancies showed wider dispersion, whereas SCC without PNI, intestinal-type adenocarcinoma (ITAC), and olfactory neuroblastoma (ONB) demonstrated limited TD. Tissue-specific effects were generally limited, although neural tissue was associated with increased positive biopsy probability in ACC.Conclusions: TD in SNNM follows a distance-dependent pattern with the distance-dispersion relationship varying according to histologic subtype. SN-based mapping represents a spatially informed strategy to enhance margin assessment and guide histology-driven surgical and radiotherapy planning.

OBJECTIVES:The objective of this study is to determine whether radiologic tumor volume adds prognostic value for overall survival (OS) and disease-free survival (DFS) in laryngeal squamous cell carcinoma (LSCC) and to evaluate associations with T stage, nodal status, histologic grade, subsite, and treatment modality. METHODS:Patients with cT2-T4 LSCC treated with curative intent between 2018 and 2024 were included. Tumor volumes derived from CT, MRI, or radiotherapy planning contours were analyzed as continuous and binary variables using a threshold identified by log-rank maximization. Survival was assessed with Kaplan-Meier methods and Cox proportional hazards models. RESULTS:Eighty-eight patients were included (39.8% surgical and 60.2% nonsurgical). Median tumor volume was 3.7 cm3 (range 0.03-51.3 cm3). Volume threshold of 22.498 cm3 was associated with a 3.6-fold increase in mortality (60.0% vs. 16.7%; p < 0.001), with significantly worse OS and DFS in the high-volume group (both p < 0.001). On multivariable analysis, high tumor volume remained independently associated with OS after adjustment for age, T stage, and nodal status (HR = 9.66, 95% CI: 2.56-36.39). Tumor volume correlated with T stage (Spearman ρ = 0.569, p < 0.001), and node-positive tumors were larger than node-negative tumors (median 5.6 cm3 vs. 2.6 cm3; p = 0.017). Supraglottic and transglottic tumors had larger volumes than glottic and higher-grade tumors were associated with larger volumes. Nonsurgical patients had smaller tumors than surgical patients. CONCLUSION:Tumor volume was independently associated with survival outcomes in LSCC and may improve risk stratification beyond TNM classification alone. A threshold of 22.498 cm3 identifies a high-risk subgroup and warrants prospective validation for integration into staging and treatment planning. LEVEL OF EVIDENCE: 3:

9540 Background: Sonidegib is an efficacious treatment of LA laBCC but is associated with a high risk of treatment-related adverse events (TRAEs) causing treatment discontinuation. Following a CR, treatment discontinuation rate reaches up to 60% after a year, leading to 3 years relapse free survival rate of 35%. We aimed at evaluating sonidegib tailored schedule (TS) after CR to increase treatment duration by reducing TRAEs, thus allowing CR maintenance. Methods: We conducted a multicenter, open-label, single-arm phase II study enrolling adult patients (pts) with laBCC who obtained a CR to sonidegib regardless of the tumor’s subtype and burden. Eligible pts received TS1 with sonidegib 14 days on and 14 days off. Pts on TS1 who experienced grade 2-3 toxicity (except alopecia) lasting >28 days moved to TS2 (7 days on and 21 days off). Treatment continued until progression or unacceptable toxicity. Primary endpoint was the rate of pts maintaining sonidegib 12 months after study enrolment (H0 31%, H1 60%). Evaluable pts were defined as all pts who were either on treatment or suspended treatment for reasons other than treatment-unrelated adverse events or death. Secondary endpoints were safety, treatment compliance, rate of disease relapse at 1 and 2 years, overall survival, quality of life, use of concomitant medications and of medical resources, and translational analysis. Results: Between Jan 2021 and Dec 2023, 22 pts from 10 Italian centers were enrolled; the data cut-off was Jan 2025. Pts characteristics are reported in table 1. The median follow-up was 22 months (range 2-33). Three disease and treatment-unrelated deaths occurred before completing 1 year of TS and therefore 19 pts were evaluable. At data cut-off, 12 pts had discontinued treatment: 26% (5) due to disease progression, 11% (2) to sonidegib’s unacceptable toxicity, the remaining either to personal or physician’s choice. Twelve out of 19 evaluable pts (63%) were still on treatment after 1 year from TS start (median duration 20 months, range 2-29), meeting the primary endpoint. The most common TRAEs episodes were muscle cramps (13), alopecia (6), dysgeusia (5) with overall TRAEs grade G1 (29), G2 (11), G3 (3). Twelve pts had dose reduction to TS2. Conclusions: Tailored maintenance schedule with pulsed sonidegib allows for longer treatment duration and fewer relapses in CR laBCC pts. Study follow up to evaluate secondary endpoints outcome and translational analysis are ongoing. Clinical trial information: 2020-002613-17 . Patient characteristics. Characteristic N (%) Male : Female 14 (64) : 8 (36) Median age 76y (range 56-93y) ECOG PS 0-1 15 (68): 7 (32) Histology sub-type Nodular 8 (36) Superficial 2 (9) Infiltrative 7 (32) Mixed 1 (5) Other 4 (18)

6099 Background: Cachexia is a challenging complication of cancer and its treatment, associated with an increased risk of death. Growth differentiation factor 15 (GDF15) is a known driver of cachexia in lung, pancreatic and colorectal cancer. However, we lack information about GDF15 role and temporal dynamics in patients with head and neck squamous cell carcinoma (HNSCC) receiving curative treatments. Here, we aimed to quantify circulating GDF15 in HNC patients and determine the relationship with indicators of cachexia and other prominent clinical correlates, namely oral mucositis (OM). Methods: Adults diagnosed with HNSCC, scheduled to receive definitive platinum based chemoradiotherapy (CRT), were recruited from four tertiary hospitals in Australia and Italy. Cachexia was determined using a variety of clinical parameters, including weight, grip strength, and upper arm circumference (UAC). OM was assessed using the World Health Organization (WHO) mucositis scale (grade 0-4). GDF15 was quantified in serially-collected serum using a commercial ELISA. All measures were assessed at baseline, week 3, end of treatment (EOT) and 3 months after treatment end. Results: Fifty-nine patients (pts) were enrolled, predominantly male (71%) with a mean age of 64+/-8 years. Main subsite of disease was oropharynx (44 cases, 75%), of whom 68% were HPV positive. Treatment was delivered with definitive intent in 90% of the cases. Median dose of RT was 70 Gy (66-70). At EOT, 77.3% of the pts met the diagnostic criteria for cachexia, with an average weight loss of 8.6% over the course of treatment. This was accompanied by a 7.7% and 5.6% average reduction in grip strength and UAC, respectively. Notably, these reductions continued beyond treatment cessation, with an average baseline weight loss of 9.8% after 3 months. Aligning with these outcomes, serum GDF15 levels were highly elevated following CRT compared to baseline (2.6 fold, P<0.0001), and correlated with reductions in weight (R 2 =0.168, P=0.0002), UAC (R 2 =0.1516, P<0.0001), grip strength (R 2 =0.1332, P<0.0001), as well as OM severity (R 2 =0.1013, P<0.0009). Conclusions: These data reveal that highly elevated GDF15 production correlates with OM and cachexia, providing strong clinical rationale to explore the biological basis of this new symptom cluster, as well as identifying a new clinical cohort that may benefit from GDF15 targeting agents.

e13777 Background: The G8 screening assessment tool has been validated to assess frailty in older patients with cancer. A score of 14 or less has commonly been considered to identify patients at risk of frailty. The G8 score has been implemented for all European Organisation for Research and treatment of Cancer (EORTC) clinical trials that included patients 70 years or older since 2017. Here we assessed the prognostic implications of the G8 score in older patients with newly diagnosed glioblastoma who were enrolled in EORTC clinical trials. Methods: Data from 99 patients aged 70 or older enrolled in the 1608 STEAM trial on the multikinase inhibitor zotiraciclib (NCT 03224104) (n = 23) or the 1709 MIRAGE trial on the proteasome inhibitor marizomib (NCT 03345095) (n = 76) were analyzed. We defined a G8 low group (<14) and a G8 high group (15-17). Results: The median age was 73 years, 21 patients (48%) were women and 23 patients (52%) were men. The Karnofsky performance status was 90 or more in 48 patients (52%). Steroids were taken at baseline by 55 patients (59%). The mean G8 at baseline was 14. Forty-four patients were assigned to the G8 low group whereas 49 patients were assigned to the G8 high group. Patients with a Karnofsky performance status of 90 or 100 were more often in the G8 high group (HR 5.44, 95% CI 2.17-14.46). Patients in the G8 high group had higher means for global health status scale and functioning scores at baseline than patients in the G8 low group. Treatment delivery measured by the relative dose intensity was better in patients with a high G8 than in patients with a low G8 score (HR 1.22, (95% CI 0.91-5.96). Patients in the G8 low group tended to experience more toxicity than patients in the G8 high group. Patients in the G8 high group tended to have a longer progression-free (HR 0.66, 95% CI 0.42-1.03, p = 0.066) and overall survival (HR 0.67, 95% CI 0.42-1.07, p = 0.096). A statistically significant effect was however noted when using a pre-specified adjustment model showing longer progression-free survival and overall survival in patients in the G8 high group. Conclusions: The G8 score may be a powerful tool for prognostic assessment and for patient stratification in old and frail patients with glioblastoma. It may help to identify patients in need of early involvement of a palliative care team.

Immune checkpoint inhibitors (ICIs) have revolutionized cancer treatment, but predictive biomarkers of adverse events and outcomes remain limited. Recently, nutritional and inflammatory markers have emerged as potentially predictors of ICI toxicity and efficacy. In this retrospective, monocentric study we evaluated the impact of baseline nutritional and inflammatory markers on ICI safety, activity and efficacy in patients affected by solid tumors treated with mono-immunotherapy. Baseline clinical, nutritional, and inflammatory parameters - including BMI, weight loss, albumin, prognostic nutritional index (PNI), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) - were analyzed in relation to immune-related adverse events (irAEs), best response, progression-free survival (PFS), and overall survival (OS). Between 2015 and 2021, 297 patients were included. Median age was 67 (range: 32–91), with 200 (67.3

Figure S6. B2M expression is complementary to other biomarkers of survival under PD(L)1 inhibitors, related to figure 5