Abstract In competing risks settings, covariate effects and group comparisons are usually assessed one event at a time—through log-rank or Cox tests on the cause-specific hazards, or Gray’s test or Fine–Gray regression on a cumulative incidence function (CIF). This can obscure a clinically important quantity: the ratio between the event of interest and the competing event, since groups may differ little on the individual events yet differ sharply in their ratio. The generalized competing event (GCE) framework makes this ratio the object of inference; on the cause-specific scale the hazard ratio ω + ( t ) = λ 1 ( t ) / λ 2 ( t ) is estimated efficiently from a single stacked (Lunn–McNeil) model. We extend the framework to two scales that describe realized incidence. The subdistribution hazard ratio is estimated by a stacked, risk-set-weighted extension of the Lunn–McNeil construction; the cumulative-incidence ratio ρ ( t ) = F 1 ( t ) /F 2 ( t )—the odds that a subject’s realized event by time t is the event of interest—by jackknife pseudo-observation regression of the Aalen–Johansen estimator. We relate the three contrasts: ρ equals ω + exactly under proportional cause-specific hazards, and equals only in the small-time limit under proportional subdistribution hazards, drifting toward 1 thereafter. The orthogonality that makes ω + efficient is lost on both cumulative-incidence scales— through overlapping weighted risk sets and shared censoring weights, ρ through the shared all-cause survivor—so each carries a covariance term that must be handled and that bounds efficiency relative to the hazard-scale test. We derive the corresponding variances, study operating characteristics by simulation, illustrate on hypothetical prostate and head-and-neck cohorts, and provide an implementation in the gcemod R package.

BACKGROUND:Cisplatin remains the cornerstone radiosensitizer for definitive and adjuvant chemoradiation in patients with locally advanced squamous cell carcinoma of the head and neck. However, despite decades of clinical use, many practical aspects of cisplatin administration remain unstandardized in clinical practice and trials. METHODS:Drawing on the collective experience of the NRG Oncology Head and Neck Working Group, this consensus article provides practical guidance on the administration of cisplatin during radiation therapy. We review and propose recommendations regarding the timing of chemotherapy and radiation initiation, premedication and hydration regimens by dose, monitoring and grading of cisplatin-induced ototoxicity, and management strategies during cisplatin shortages. Our guidance is informed by clinical trial protocols, retrospective and prospective data, and multidisciplinary expert consensus. We emphasize the importance of protocol flexibility to support equitable trial accrual, minimize treatment delays, and improve patient-centered outcomes. CONCLUSIONS:This article offers a unified framework to optimize the use of cisplatin in chemoradiation protocols, improve adherence, reduce toxicity, and preserve oncologic efficacy. Our recommendations are particularly timely in the context of evolving clinical practices and recent cisplatin shortages. Insights from the NRG Oncology collaborative group aim to inform future trial designs and clinical practice guidelines, ensuring consistent and equitable care for patients with head and neck cancer.

Background:Men with aggressive, localized prostate cancer (PC) undergo definitive radiotherapy (RT) with androgen deprivation therapy (ADT). The prospective, phase II ARIEL trial evaluates a quantitative MRI biomarker, Restriction Spectrum Imaging restriction score (RSIrs), at three time points (before treatment, after ADT and after RT) for treatment response assessment. RSIrs highlights intracellular restricted diffusion and is correlated with high-grade PC. Design:Participants are men with unfavorable-intermediate-risk or high-risk localized PC undergoing definitive RT with neoadjuvant and concurrent ADT, and MRI-RSI acquisitions at three time points: before therapy, after neoadjuvant ADT but before RT, and after RT. The primary aim is to evaluate performance of RSIrs for identifying patients who will experience early biochemical recurrence. Change in RSIrs within visible tumors after ADT and RT is the primary independent variable. Results:97 patients met inclusion criteria and received ≥1 MRI. On central review, visible PI-RADS lesions were identified in 88 patients: 80 patients had one lesion, and 8 patients had two lesions. After neoadjuvant ADT, 40% of lesions were not clearly visible. Those still visible had shrank by median 55.8% (IQR: 42.8-69.0%), much more than the prostate volume decrease of 21.5% (11.9-31.6%). RSIrs maximum within visible lesions decreased from mean 329 (SD:185) pre-ADT to 209 (SD:125) pre-RT ( p <0.01), and to 107 (SD:61) post-RT ( p <0.01). Conventional apparent diffusion coefficient (ADC) changes were less consistent. Follow-up is ongoing to assess whether imaging response is related to future recurrence risk. Conclusion:ARIEL has completed accrual and preliminary results demonstrate changes in RSIrs after treatment, which may indicate tumor response. Primary results will be presented when the primary endpoint is reached. With neoadjuvant ADT, both pre- and post-ADT MRI are likely necessary for accurate focal RT boost targeting. Concurrent commencement of ADT and RT simplifies workflows and facilitates accurate gross tumor volume delineation.

6016 Background: Standard chemoradiation (CRT) for unfavorable risk p16+ head and neck squamous cell carcinoma (HNSCC) is associated with high rates of toxicity and treatment failure. We hypothesized that concurrent and adjuvant pembrolizumab (pembro) could achieve superior PFS with acceptable toxicity compared to CRT in this population. Methods: KEYCHAIN was an open-label randomized phase II trial conducted at 6 US academic centers. Eligible patients had newly diagnosed unresected p16+ HNSCC with unfavorable risk (AJCC 8 th edition stage II-III oropharynx or stage III-IVB non-oropharynx HNSCC). Patients were randomized 1:1 to RT (70 Gy in 35 fractions) plus pembro (200 mg, 2 weeks prior to RT then every 3 weeks starting day 1 of RT up to 20 total cycles) or RT plus concurrent cisplatin (cis) (100 mg/m 2 every 3 weeks). Age, ECOG status, and oropharynx site were stratification factors. The primary endpoint was PFS, defined as time from randomization to first progression or death from any cause. Primary analysis was in the modified intent-to-treat population, including all patients having ≥ 1 dose of study medication and ≥ 1 efficacy evaluation after baseline. The study design had 80% power with 1-sided α = 0.15 (log-rank test) and planned sample size of 50 analyzable subjects per arm. The study is registered with clinicaltrials.gov (NCT03383094). Results: Between Feb 2019 and Aug 2025, 108 patients were randomized (53 pembro, 55 cis), with 102 analyzable (50 pembro, 52 cis). Median follow-up was 26.5 months. 79% had minimum follow-up of 2 years. At 2 years, there were 7 PFS events and 1 death in the pembro arm, and 14 PFS events and 7 deaths in the cis arm. Two-year PFS was 84% (95% CI: 74-96%) vs. 70% (95% CI: 58-85%) in the pembro vs cis arm (HR 0.57, 95% CI: 0.25-1.31; 1-sided p = 0.09) (see Table). Two-year OS was 98% (95% CI: 94-100%) vs. 85% (95% CI: 76-96%) for pembro vs. cis (HR 0.33, 95% CI: 0.09-1.28; 1-sided p = 0.048). HRs adjusted for T3-4 and N2-3 category were 0.68 (PFS) and 0.31 (OS). As of the latest follow-up, there were 24 total PFS events and 12 deaths. Grade ≥ 3 adverse events definitely or probably related to treatment were 36% vs. 46% in the pembro vs. cis arm. The median number of cycles received in the pembro arm was 18. In the cis arm, 94% completed ≥ 2 cycles. Results by CPS status will be reported when available. Conclusions: Pembro plus RT met the pre-specified significance criterion for improvement in PFS compared to cisplatin plus RT in this phase II trial. A phase III trial of RT with pembrolizumab for unfavorable risk p16+ HNSCC is warranted. Clinical trial information: NCT03383094 . Pembro Cisplatin N 50 52 Mean age 63.5 62.5 Male 87% 89% White 85% 89% Age > 65 42% 35% ECOG 0 83% 70% Oropharynx 94% 96% T1 8% 10% T2 22% 12% T3 38% 38% T4 32% 40% N0 6% 8% N1 42% 29% N2 46% 54% N3 6% 10% 2-yr PFS 84% 70% 2-yr OS 98% 85% Completed RT w/in 56 days 96% 92% Grade ≥ 3 mucositis 18% 17% Grade ≥ 2 mucositis 72% 63% Grade ≥ 3 dysphagia 14% 6% Grade ≥ 2 dysphagia 48% 35%

The NCCN Guidelines for Head and Neck Cancers address tumors arising in the oral cavity (including mucosal lip), pharynx, larynx, and paranasal sinuses, as well as occult primary cancer, salivary gland cancer, and mucosal melanoma (MM). The specific site of disease, stage, and pathologic findings guide treatment (eg, the appropriate surgical procedure, radiation targets, dose and fractionation of radiation, indications for systemic therapy). The NCCN Head and Neck Cancers Panel meets at least annually to review comments from reviewers within their institutions, examine relevant new data from publications and abstracts, and reevaluate and update their recommendations. These NCCN Guidelines Insights summarize the panel's most recent recommendations regarding management of nasopharynx cancer and ongoing research in this area.

e18048 Background: Outcomes for patients with head and neck cancer (HNC) are affected by tumor specific features, as well as demographic characteristics and health-related comorbidities. We previously found that women have a higher risk of HNC death compared with risk of death from other causes using a General Competing Event (GCE) model. With a larger, more recent and well-annotated cohort of patients with HNC from Kaiser Permanente Northern California, we applied the GCE model to investigate sex and racial disparities in all-cause, HNC-specific, and non-HNC mortality. Methods: We identified patients with mucosal HNC treated in the curative setting between 2013-2017 on sociodemographic and clinical factors. Using GCE models adjusted for age, smoking status, Charlson comorbidity index, primary tumor site, and tumor stage, we examined if gender and race/ethnicity were associated with a higher relative hazard of HNC-specific mortality over non-HNC mortality (relative hazard ratio (RHR), ‘gcerisk’ package in R). Follow-up was measured from HNC diagnosis date to death or end of study (12/31/2022) whichever occurred earliest. Results: In our cohort of 1616 patients (mean age at diagnosis: 64.2 years), 25.7% were female. The majority were White (70.5%), followed by Asian (12.3%), Hispanic (8.0%), Black (4.8%), and other/unknown (4.3%). Women were less likely to smoke and had more oral cavity cancer and less p16+ oropharyngeal cancer than men. Patients of non-white (vs. white) race were younger, lived in more deprived neighborhoods, had less smoking and alcohol use, had more nasopharyngeal carcinoma and less p16+ oropharyngeal cancer. Over a median follow-up of 6.5 years, 274 patients died of HNC, and 279 patients died of other causes. Greater HNC-specific mortality was associated with female gender (RHR ratio 1.63; 95% confidence interval [CI] 1.10-2.41) and non-White race (RHR ratio 1.40; 95% CI 0.95-2.08). Conclusions: We observed worse HNC-specific mortality among female and nonwhite patients, which may be related to underlying disease biology and socioeconomic factors. This highlights the importance of examining both HNC-specific and non-HNC mortality to identify cancer health disparities. Mortality outcomes associated with gender and race among 1616 patients with head and neck cancer (HNC) treated in the curative setting. Variable All-Cause Mortality HR a (95% CI) Cancer Specific Mortality HR a (95% CI) Non-Cancer Mortality HR a (95% CI) Cancer Specific vs. Non-Cancer Mortality RHR* a (95% CI) Female vs. Male 1.07 (0.88, 1.29) 1.37 (1.04, 1.79) 0.84 (0.63, 1.12) 1.63 ( 1.10, 2.41 ) Non-White vs. White 0.94 (0.77, 1.14) 1.11 (0.85, 1.46) 0.79 (0.60, 1.05) 1.40 (0.95, 2.08) HR = Hazard ratio. *RHR = Relative hazard ratio for HNC-specific mortality vs. non-HNC mortality. a Adjusted for age, smoking status, Charlson comorbidity index, primary tumor site, tumor stage.

Human papillomavirus (HPV)-associated head and neck squamous cell carcinoma (HPV+ HNSCC) is one of the most rapidly increasing cancers in the United States. Although patients with HPV+ HNSCC exhibit better responses to chemoradiation compared with patients with HPV- HNSCC, a subset still develops recurrent/refractory disease. In this study, we investigated the role and mechanisms by which the HPV E5 oncogene promotes resistance to conventional chemotherapy and radiation. After treatment with chemotherapy and radiation, HNSCC cells expressing HPV E5 had decreased apoptosis, improved cell viability, enhanced DNA damage response, and increased tumor growth in murine xenograft models when compared with empty vector controls. Transcriptomic analysis showed an enrichment of signaling pathways associated with epithelial development and increased expression of stem cell markers, including SOX9 and SOX4, in HPV E5 cells compared with empty vector cells. Knockdown of SOX9 restored sensitivity to chemo- and/or radiotherapy in HPV E5-expressing cells. In patients with HNSCC, RNA sequencing analysis showed that high expression of HPV E5 was associated with enrichment of pathways involved in cellular senescence and stem cell pluripotency. Furthermore, an HPV E5-associated gene set was identified that correlates with poor patient prognosis and is significantly overexpressed in recurrent HPV+ HNSCC tumors. Taken together, these data identify HPV E5 as a viral mediator of resistance to standard-of-care therapies for HPV+ HNSCC by promoting a stem cell-like phenotype that enhances DNA damage responses and tumor cell survival. Thus, HPV E5 may contribute to chemotherapy resistance, tumor recurrence after radiation, and poor prognosis of patients with refractory or metastatic disease.Significance: HPV E5 upregulates SOX9 to induce stem cell-like characteristics in head and neck cancer that drive resistance to chemotherapy and radiation, providing rationale for developing therapeutic agents targeting HPV E5.

Differential gene expression and pathway analysis

Prognosis of HNSCC patients based on the status of E5 signature and expression distribution of E5 in single cell RNA-sequencing dataset

Background: Cervical cancer remains a major global health concern, with high recurrence rates in advanced stages. [18F]FDG PET/CT provides prognostic biomarkers such as SUV, MTV, and TLG, though these are not routinely integrated into clinical protocols. Radiomics offers quantitative analysis of tumor heterogeneity, supporting risk stratification. Purpose: To evaluate the prognostic value of clinical and radiomic features for disease-free survival (DFS) in locoregionally advanced cervical cancer using machine learning (ML). Methods: Sixty-three patients (mean age 47.9 ± 14.5 years) were diagnosed between 2015 and 2020. Radiomic features were extracted from pre-treatment PET/CT (IBSI-compliant PyRadiomics). Clinical variables included age, T-stage, Dmax, lymph node involvement, SUVmax, and TMTV. Forty-two models were built by combining six feature-selection techniques (UCI, MD, MI, VH, VH.VIMP, IBMA) with seven ML algorithms (CoxPH, CB, GLMN, GLMB, RSF, ST, EV) using nested 3-fold cross-validation with bootstrap resampling. External validation was performed on 95 patients (mean age 50.6 years, FIGO IIB–IIIB) from an independent cohort with different preprocessing protocols. Results: Recurrence occurred in 31.7% (n = 20). SUVmax of lymph nodes, lymph node involvement, and TMTV were the most predictive individual features (C-index ≤ 0.77). The highest performance was achieved by UCI + EV/GLMB on combined clinical + radiomic features (C-index = 0.80, p < 0.05). For single feature sets, IBMA + RSF performed best for clinical (C-index = 0.72), and VH.VIMP + GLMN for radiomics (C-index = 0.71). External validation confirmed moderate generalizability (best C-index = 0.64). Conclusions: UCI-based feature selection with GLMB or EV yielded the best predictive accuracy, while VH.VIMP + GLMN offered superior external generalizability for radiomics-only models. These findings support the feasibility of integrating radiomics and ML for individualized DFS risk stratification in cervical cancer.

Expression of HPV E5, E6 and E7 in different HNSCC cells and normal keratinocytes (HaCaT)

shRNA mediated knockdown of SOX9 in HNSCC cells, measurement of apoptosis and DNA damage response in SOX9 silenced EV HNSCC cells

The NCCN Guidelines for Head and Neck Cancers address tumors arising in the oral cavity (including mucosal lip), pharynx, larynx, and paranasal sinuses, as well as occult primary cancer, salivary gland cancer, and mucosal melanoma (MM). The specific site of disease, stage, and pathologic findings guide treatment (eg, the appropriate surgical procedure, radiation targets, dose and fractionation of radiation, indications for systemic therapy). The NCCN Head and Neck Cancers Panel meets at least annually to review comments from reviewers within their institutions, examine relevant new data from publications and abstracts, and reevaluate and update their recommendations. These NCCN Guidelines Insights summarize the panel's most recent recommendations regarding management of nasopharynx cancer and ongoing research in this area.

BACKGROUND:Cisplatin-based chemoradiation (CRT) plus brachytherapy for locally advanced cervical cancer (LACC) is standard. Intrinsic overexpression of ribonucleotide reductase (RNR) may enhance DNA damage repair from CRT. We report on outcomes of adding RNR inhibitor, triapine (T), to CRT. METHODS:NRG-GY006 is an open-label randomized phase III trial. FIGO 2009 LACC (stages IB2, II, IIIB or IVA) without para-aortic nodal involvement or stages II-IV vaginal cancer were eligible. Random assignment to CRT or in combination with thrice-weekly T (CRT + T) occurred. Radiation consisted of either 3D conformal (3DCRT) or image-guided intensity modulated RT (IG-IMRT) followed by intracavitary brachytherapy. Primary endpoint was overall survival (OS). Progression-free survival (PFS) was secondary. Exploratory endpoints included complete metabolic response rate on post treatment PET/CT imaging and comparative toxicity and outcomes for 3DCRT vs. IG-IMRT. FINDINGS:Four-hundred-fifty patients were randomized including 448 eligible (224 in CRT and 224 in CRT + T). Median age was 47 (range 23-85). The majority had cervical cancer (93.3 %) with squamous histology (82 %). 52 % had FIGO stage II disease. Racial/ethnic distribution included non-Hispanic white (53.8 %), black (15.2 %) and Hispanic/Latina (22.5 %). At randomization, IG-IMRT was planned in 74.3 % and HDR brachytherapy in 98.2 %. No differences in Grade 3-5 toxicities were observed: CRT: 52 % and CRT + T: 49 %, with two G5 toxicities (cardiac arrest and acidosis) in the CRT + T arm. The median patient follow-up was 28 months (IQR 15-45). HR for death was 1.018 (95 % CI 0.634-1.635) while HR for progression was 1.021 (95 % CI 0.694-1.501). INTERPRETATION:Triapine added to CRT did not improve OS.

[This corrects the article DOI: 10.1016/j.ctro.2024.100801.].

Combined immune checkpoint blockade (ICB) and chemoradiation (CRT) is approved in patients with locally advanced cervical cancer (LACC) but optimal sequencing of CRT and ICB is unknown. NRG-GY017 (NCT03738228) was a randomized phase I trial of atezolizumab (anti-PD-L1) neoadjuvant and concurrent with CRT (Arm A) vs. concurrent with CRT (Arm B) in patients with high-risk node-positive LACC. The primary endpoint was the fraction of expanded tumor-associated T-cell receptor (TCR) clones in blood at day 21 as a surrogate measure of anti-tumor immune response. Secondary objectives were safety and feasibility, 2-year disease-free survival (DFS), and predictive value of PD-L1 expression. Forty patients were randomized, 36 received treatment, and 25 were evaluable for the primary endpoint. After cycle 1, there was peripheral expansion of higher proportion of tumor-associated TCR clones in Arm A than in Arm B (p = 0.0025) that remained higher at day 21, meeting the pre-specified endpoint on two-sample T-test (p = 0.052), but not on sensitivity analysis by Wilcoxon test (p = 0.13). At the median follow up of 25.8 months, 2-year DFS was 76% in Arm A and 56% in Arm B (p = 0.28). There were no new safety signals. In conclusion, neoadjuvant ICB prior to CRT was safe and was associated with immunologically and clinically favorable outcomes, warranting larger confirmatory studies.

Measurement of apoptosis after knockdown of E5 in HNSCC cells and evaluation of DNA damage response in E5 expressed and E5 knockdown HNSCC cells

11066 Background: This study aimed to determine to what extent area-level social determinants of health (SDOH) interact with individual, institutional, and biological factors to predict outcomes in head and neck cancer (HNC) trials. Methods: Five NRG Oncology HNC trials (2635 patients receiving chemoradiation) were analyzed. Area-level SDOH coded by patient ZIP codes included rurality (rural-urban commuting area code), neighborhood socioeconomic deprivation (Area Deprivation Index [ADI] categorized as upper vs. lower quartile), and travel burden (distance and time to treatment site). Individual (demographic, cancer and treatment-related factors), institutional (accrual volume), biological (HPV+/-) factors, and outcomes (overall survival [OS], progression free survival [PFS], quality of life [QOL], and symptoms) were analyzed. Multivariable Cox proportional hazards regression and mediation analysis using logistic regression assessed associations using hazard ratio (HR) or odds ratios (OR) and 95% confidence intervals (CI). Results: Most patients were White (88%), non-Hispanic (92.7%), of mean age of 57 years, HPV+ (64.6%), and received intensity-modulated radiotherapy (95.9%) and cisplatin (94.2%). ADI and rurality were not associated with OS and PFS. OS and PFS were higher in patients with travel time <1 hour (HR=0.85, 95% CI [0.75, 0.98]; HR=0.85, 95% CI [0.73, 0.98]) and travel distance <50 miles (HR=0.84, 95% CI [0.72, 0.96]; HR=0.85, 95% CI [0.73, 0.98]). ADI, travel time, and travel distance were not associated with QOL decline. Patients treated at institutions with high rural accrual volume had worse QOL decline from baseline (OR=0.36, 95% CI [0.15, 0.85]). The impact of travel distance but not time varied by race to influence QOL decline (OR=0.38, 95% CI [0.16, 0.93]). ADI was not associated with symptoms, but patients from institutions with high rural accrual volume had worse symptoms (OR=7.83, 95% CI [1.98, 31.01]). HPV status had a significant indirect effect on the relationship between travel distance and survival at 1 year (estimate [β]=0.03, 95% CI [0.01, 0.05]) and 5 years (β=0.03, 95% CI [0.004, 0.05]), as well as a direct and total mediation effect of travel distance on QOL decline at 1 year (direct β=-0.08, 95% CI [-0.16, -0.004]; total β=-0.89, 95% CI [-0.17, -0.01]). Conclusions: This study showed the impact of area-level SDOH and their interactions with race and institutional accrual volume, which are associated with survival, QOL, and symptom changes. HPV status potentially mediated the effects of travel distance on outcomes. Our findings provide novel approaches to identify patients at risk for poor outcomes, such as those with travel burden at institutions with high rural accrual, to design community-based interventions to improve cancer outcomes.

The dominant treatment paradigm for locoregionally advanced head and neck squamous cell carcinoma (HNSCC) involves postoperative or definitive radiotherapy with concurrent cisplatin chemotherapy. Despite years of research investigating de-intensified treatment, cisplatin-based chemoradiotherapy remains the standard, yet it is associated with significant acute and chronic toxicity. However, due to shared risk factors, such as advanced age, and tobacco and alcohol use, patients with HNSCC frequently have comorbid illnesses that impact treatment tolerability, adding complexity to treatment-related decision-making. In addition, many patients have medical contraindications to cisplatin, requiring alternative treatment strategies. It is thus important to consider how well patients are likely to tolerate treatment, and how to adapt treatment in response to a patient's condition, when weighing treatment options. In this review, we aim to offer readers guidance in managing the elderly or comorbid patient with HNSCC, with particular attention to (i) approaching comorbidity and fragility assessment to make determinations on intensity of treatment, (ii) considering primary treatment modality (eg, surgery vs radiotherapy, chemo-radiotherapy vs radiotherapy alone) and (iii) choice of concurrent systemic therapy agent.