Architectural dysplasia in surgical margins and the risk of local relapse in oral cancer.

BACKGROUND: A major challenge in the clinical management of oral cavity squamous cell carcinoma is local relapse. Even when surgical margins are tumor-free, local relapses occur frequently, and relapse prediction by histology remains suboptimal. In leukoplakia, an oral potentially malignant disorder, the presence of architectural dysplasia is a critical risk factor for malignant transformation. This study aimed to investigate whether the presence of architectural dysplasia in oral cavity squamous cell carcinoma surgical margins is a risk factor for local relapse. METHODS: Hematoxylin and eosin-stained slides of resection margins from a consecutive cohort of surgically treated patients diagnosed with stage I-IV oral cavity squamous cell carcinoma between 2008 and 2014 were assessed for the presence of architectural dysplasia (N = 311). Five-year local relapse-free survival rates of oral cavity squamous cell carcinoma with architectural dysplasia were compared to those of oral cavity squamous cell carcinoma without architectural dysplasia. RESULTS: In total, 92 of 311 (29.6%) of oral cavity squamous cell carcinoma displayed architectural dysplasia in the margins. The presence of architectural dysplasia was associated with higher patient age, female sex, less pack years, lower cT-stage, and a cohesive tumor growth pattern. In oral cavity squamous cell carcinomas with architectural dysplasia, postoperative (chemo)radiotherapy was less often indicated compared with oral cavity squamous cell carcinoma without architectural dysplasia (19.5% vs. 36.1%, p = 0.009). Five-year local relapse-free survival was significantly lower in oral cavity squamous cell carcinoma with architectural dysplasia than in oral cavity squamous cell carcinoma without architectural dysplasia (83.1% vs. 94.9%, p = 0.017). CONCLUSIONS: Oral cavity squamous cell carcinoma arising in the background of architectural dysplasia displays relatively favorable clinical and histopathological characteristics. Nonetheless, the presence of architectural dysplasia in oral cavity squamous cell carcinoma surgical margins is associated with a higher risk of local relapse, indicating its clinical relevance.

Prognostic radiomic signature for head and neck cancer: Development and validation on a multi-centric MRI dataset.

BACKGROUND AND PURPOSE: Prognosis in locally advanced head and neck cancer (HNC) is currently based on TNM staging system and tumor subsite. However, quantitative imaging features (i.e., radiomic features) from magnetic resonance imaging (MRI) may provide additional prognostic info. The aim of this work is to develop and validate an MRI-based prognostic radiomic signature for locally advanced HNC. MATERIALS AND METHODS: Radiomic features were extracted from T1- and T2-weighted MRI (T1w and T2w) using the segmentation of the primary tumor as mask. In total 1072 features (536 per image type) were extracted for each tumor. A retrospective multi-centric dataset (n = 285) was used for features selection and model training. The selected features were used to fit a Cox proportional hazard regression model for overall survival (OS) that outputs the radiomic signature. The signature was then validated on a prospective multi-centric dataset (n = 234). Prognostic performance for OS and disease-free survival (DFS) was evaluated using C-index. Additional prognostic value of the radiomic signature was explored. RESULTS: The radiomic signature had C-index = 0.64 for OS and C-index = 0.60 for DFS in the validation set. The addition of the radiomic signature to other clinical features (TNM staging and tumor subsite) increased prognostic ability for both OS (HPV- C-index 0.63 to 0.65; HPV+ C-index 0.75 to 0.80) and DFS (HPV- C-index 0.58 to 0.61; HPV+ C-index 0.64 to 0.65). CONCLUSION: An MRI-based prognostic radiomic signature was developed and prospectively validated. Such signature can successfully integrate clinical factors in both HPV+ and HPV- tumors.

Accuracy of p16 IHC in Classifying HPV-Driven OPSCC in Different Populations.

High-risk human papillomavirus (HPV) infection is a defined etiopathogenetic factor in oropharyngeal carcinogenesis with a clear prognostic value. The P16 IHC (immunohistochemistry) is a widely accepted marker for HPV-driven carcinogenesis in oropharyngeal squamous cell carcinoma (OPSCC); in the present paper, we discuss its reliability as a standalone marker in different populations. The literature suggests that rates of p16 IHC false positive results are inversely correlated with the prevalence of HPV-driven carcinogenesis in a population. We propose a formula that can calculate such a false positive rate while knowing the real prevalence of HPV-driven OPSCCs in a given population. As it has been demonstrated that p16 positive/HPV negative cases (i.e., false positives at p16 IHC) have the same prognosis as p16 negative OPSCC, we conclude that despite the valuable prognostic value of p16 IHC, relying only on a p16 IHC positive result to recommend treatment de-intensification could be risky. For this aim, confirmation with an HPV nucleic acid detection system, especially in areas with a low prevalence of HPV-related OPSCCs, should be pursued.

Clinical Validity of a Prognostic Gene Expression Cluster-Based Model in Human Papillomavirus-Positive Oropharyngeal Carcinoma.

Under common therapeutic regimens, the prognosis of human papillomavirus (HPV)-positive squamous oropharyngeal carcinomas (OPCs) is more favorable than HPV-negative OPCs. However, the prognosis of some tumors is dismal, and validated prognostic factors are missing in clinical practice. The present work aimed to validate the prognostic significance of our published three-cluster model and to compare its prognostic value with those of the 8th edition of the tumor-node-metastasis staging system (TNM8) and published signatures and clustering models. METHODS: Patients with HPV DNA-positive OPCs with locoregionally advanced nonmetastatic disease treated with curative intent (BD2Decide observational study, NCT02832102) were considered as validation cohort. Patients were treated in seven European centers, with expertise in the multidisciplinary management of patients with head and neck cancer. The median follow-up was 46.2 months (95% CI, 41.2 to 50), and data collection was concluded in September 2019. The primary end point of this study was overall survival (OS). Three-clustering models and seven prognostic signatures were compared with our three-cluster model. RESULTS: The study population consisted of 235 patients. The three-cluster model confirmed its prognostic value. Two-year OS in each cluster was 100% in the low-risk cluster, 96.6% in the intermediate-risk cluster, and 86.3% in the high-risk cluster (P = .00074). For the high-risk cluster, we observed an area under the curve = 0.832 for 2-year OS, significantly outperforming TNM 8th edition (area under the curve = 0.596), and functional and biological differences were identified for each cluster. CONCLUSION: The rigorous clinical selection of the cases included in this study confirmed the robustness of our three-cluster model in HPV-positive OPCs. The prognostic value was found to be independent and superior compared with TNM8. The next step includes the translation of the three-cluster model in clinical practice. This could open the way to future exploration of already available therapies in HPV-positive OPCs tailoring de-escalation or intensification according to the three-cluster model.

A Prospectively Validated Prognostic Model for Patients with Locally Advanced Squamous Cell Carcinoma of the Head and Neck Based on Radiomics of Computed Tomography Images.

BACKGROUND: Locoregionally advanced head and neck squamous cell carcinoma (HNSCC) patients have high relapse and mortality rates. Imaging-based decision support may improve outcomes by optimising personalised treatment, and support patient risk stratification. We propose a multifactorial prognostic model including radiomics features to improve risk stratification for advanced HNSCC, compared to TNM eighth edition, the gold standard. PATIENT AND METHODS: Data of 666 retrospective- and 143 prospective-stage III-IVA/B HNSCC patients were collected. A multivariable Cox proportional-hazards model was trained to predict overall survival (OS) using diagnostic CT-based radiomics features extracted from the primary tumour. Separate analyses were performed using TNM8, tumour volume, clinical and biological variables, and combinations thereof with radiomics features. Patient risk stratification in three groups was assessed through Kaplan-Meier (KM) curves. A log-rank test was performed for significance (p-value < 0.05). The prognostic accuracy was reported through the concordance index (CI). RESULTS: A model combining an 11-feature radiomics signature, clinical and biological variables, TNM8, and volume could significantly stratify the validation cohort into three risk groups (p < 0∙01, CI of 0.79 as validation). CONCLUSION: A combination of radiomics features with other predictors can predict OS very accurately for advanced HNSCC patients and improves on the current gold standard of TNM8.

The important role of cisplatin in the treatment of HPV-positive oropharyngeal cancer assessed by real-world data analysis.

OBJECTIVES: The prognostic advantage of human papillomavirus (HPV)-positive oropharyngeal squamous cell carcinoma (OPSCC) resulted in the initiation of treatment de-intensification studies. Two randomized controlled trials (RCTs) reported inferior survival of HPV-positive OPSCC treated with radiotherapy plus cetuximab compared to standard of care radiotherapy plus cisplatin. In this study we investigated whether the important role of cisplatin in the treatment of HPV-positive OPSCCs would also emerge from causal inference analyses of real-world data. MATERIAL AND METHODS: A retrospective cohort of 263 advanced-stage OPSCC-patients from 5 European clinics was studied, treated with radiotherapy (RT) alone or cisplatin-based chemoradiotherapy (CRT) based on standard clinical indications. Causal inference was applied to adjust for treatment assignment, thereby simulating a randomized setting. Average treatment effect of concurrent cisplatin on overall survival (OS) probability was estimated using Bayesian Additive Regression Trees (BART) and Bayesian logistic regression. RESULTS: Significantly better survival probabilities were found for HPV-positive OPSCC treated with CRT compared to RT alone (3-year OS probability 0.961 versus 0.798, p = 0.008). CONCLUSION: This study using causal inference of retrospective patient data confirms the important role of cisplatin in the treatment of HPV-positive OPSCC. Causal inference analyses of real-world data complements the evidence from the published RCTs.

The unveiled reality of human papillomavirus as risk factor for oral cavity squamous cell carcinoma.

The prognostic impact of human papillomavirus (HPV) in oropharyngeal cancer is generally acknowledged, and HPV-status is assessed routinely in clinical practice. Paradoxically, while the oral cavity seems the predilection site for productive HPV-infections, figures on HPV-attribution in oral cavity squamous cell carcinoma (OCSCC) differ widely, and prognostic impact is uncertain. Major obstacles are the lack of reproducible assays to detect HPV in nonoropharyngeal cancers, the relatively small cohorts studied and consequently the shortfall of convincing data. In our study, we used a validated, nucleic acid-based workflow to assess HPV-prevalence in a consecutive cohort of 1016 OCSCCs, and investigated its prognostic impact. In parallel, we analyzed p16-immunohistochemistry (p16-IHC) as surrogate marker for transforming HPV-infection and independent prognosticator. All OCSCC-patients diagnosed between 2008 and 2014 at two Dutch university medical centers were included (N = 1069). Formalin-fixed, paraffin-embedded (FFPE)-samples of 1016 OCSCCs could be retrieved. Punch biopsies were taken from the tumor area in the FFPE-blocks and tested for HPV. P16-IHC was performed on 580 OCSCCs, including all HPV-positive tumors. From 940 samples (92.5%), nucleic acids were of sufficient quality for HPV-testing. In total, 21 (2.2%) OCSCCs were HPV DNA-positive. All HPV DNA-positive tumors were E6 mRNA-positive and considered as true HPV-positive. There was no difference in survival between HPV-positive and HPV-negative OCSCCs. In total, 46 of 580 (7.9%) OCSCCs were p16-immunopositive, including all HPV-positive tumors. Survival was comparable in p16-positive and p16-negative OCSCCs. To conclude, HPV-prevalence is very low in OCSCC and neither HPV-status nor p16-status affects outcome. Based on these data, determining HPV-status in OCSCC seems irrelevant for clinical management.

Development of a multiomics database for personalized prognostic forecasting in head and neck cancer: The Big Data to Decide EU Project.

BACKGROUND: Despite advances in treatments, 30% to 50% of stage III-IV head and neck squamous cell carcinoma (HNSCC) patients relapse within 2 years after treatment. The Big Data to Decide (BD2Decide) project aimed to build a database for prognostic prediction modeling. METHODS: Stage III-IV HNSCC patients with locoregionally advanced HNSCC treated with curative intent (1537) were included. Whole transcriptomics and radiomics analyses were performed using pretreatment tumor samples and computed tomography/magnetic resonance imaging scans, respectively. RESULTS: The entire cohort was composed of 71% male (1097)and 29% female (440): oral cavity (429, 28%), oropharynx (624, 41%), larynx (314, 20%), and hypopharynx (170, 11%); median follow-up 50.5 months. Transcriptomics and imaging data were available for 1284 (83%) and 1239 (80%) cases, respectively; 1047 (68%) patients shared both. CONCLUSIONS: This annotated database represents the HNSCC largest available repository and will enable to develop/validate a decision support system integrating multiscale data to explore through classical and machine learning models their prognostic role.

The Immune Microenvironment in Head and Neck Squamous Cell Carcinoma: on Subsets and Subsites.

PURPOSE: To understand why some patients respond to immunotherapy but many do not, a clear picture of the tumor microenvironment (TME) of head and neck squamous cell carcinoma (HNSCC) is key. Here we review the current understanding on the immune composition per HNSCC subsite, the importance of the tumor's etiology and the prognostic power of specific immune cells. RECENT FINDINGS: Large cohort data are mostly based on deconvolution of transcriptional databases. Studies focusing on infiltrate localization often entail small cohorts, a mixture of HNSCC subsites, or focus on a single immune marker rather than the interaction between cells within the TME. Conclusions on the prognostic impact of specific immune cells in HNSCC are hampered by the use of heterogeneous or small cohorts. To move forward, the field should focus on deciphering the immune composition per HNSCC subsite, in powered cohorts and considering the molecular diversity in this disease.

Correction: Computed tomography-derived radiomic signature of head and neck squamous cell carcinoma (peri)tumoral tissue for the prediction of locoregional recurrence and distant metastasis after concurrent chemo-radiotherapy.

[This corrects the article DOI: 10.1371/journal.pone.0232639.].

Computed tomography-derived radiomic signature of head and neck squamous cell carcinoma (peri)tumoral tissue for the prediction of locoregional recurrence and distant metastasis after concurrent chemo-radiotherapy.

INTRODUCTION: In this study, we investigate the role of radiomics for prediction of overall survival (OS), locoregional recurrence (LRR) and distant metastases (DM) in stage III and IV HNSCC patients treated by chemoradiotherapy. We hypothesize that radiomic analysis of (peri-)tumoral tissue may detect invasion of surrounding tissues indicating a higher chance of locoregional recurrence and distant metastasis. METHODS: Two comprehensive data sources were used: the Dutch Cancer Society Database (Alp 7072, DESIGN) and "Big Data To Decide" (BD2Decide). The gross tumor volumes (GTV) were delineated on contrast-enhanced CT. Radiomic features were extracted using the RadiomiX Discovery Toolbox (OncoRadiomics, Liege, Belgium). Clinical patient features such as age, gender, performance status etc. were collected. Two machine learning methods were chosen for their ability to handle censored data: Cox proportional hazards regression and random survival forest (RSF). Multivariable clinical and radiomic Cox/ RSF models were generated based on significance in univariable cox regression/ RSF analyses on the held out data in the training dataset. Features were selected according to a decreasing hazard ratio for Cox and relative importance for RSF. RESULTS: A total of 444 patients with radiotherapy planning CT-scans were included in this study: 301 head and neck squamous cell carcinoma (HNSCC) patients in the training cohort (DESIGN) and 143 patients in the validation cohort (BD2DECIDE). We found that the highest performing model was a clinical model that was able to predict distant metastasis in oropharyngeal cancer cases with an external validation C-index of 0.74 and 0.65 with the RSF and Cox models respectively. Peritumoral radiomics based prediction models performed poorly in the external validation, with C-index values ranging from 0.32 to 0.61 utilizing both feature selection and model generation methods. CONCLUSION: Our results suggest that radiomic features from the peritumoral regions are not useful for the prediction of time to OS, LR and DM.

The role of tumour morphology in assigning HPV status in oropharyngeal squamous cell carcinoma.

OBJECTIVES: There is no consensus on the optimal approach to human papilloma virus (HPV) testing in oropharyngeal squamous cell carcinoma (OPSCC). Our objective was to classify OPSCC as HPV positive or negative based on (1) morphology alone, (2) p16 status alone, (3) combined morphology and p16 status with additional HPV testing in discordant cases in keeping with 2012 College of American Pathologists (CAP) guidelines (combined approach), and to evaluate and compare survival outcomes. MATERIALS AND METHODS: Retrospective review of 168 patients, 146 with OPSCC and 22 with cervical SCC of unknown primary site (SCCUP). Morphology was classified as keratinizing or non-keratinizing, p16 immunohistochemistry (IHC) performed and additional HPV DNA PCR testing undertaken in cases in which morphology and p16 status were discordant. Survival statistics were evaluated and compared for the 3 different approaches to classification. RESULTS: On univariate analysis all 3 classification methods significantly predicted for overall survival (OS). Both p16 status and the combined approach also predicted for disease specific survival (DSS), whereas morphology fell just outside significance (p = 0.06). On multivariate analysis only the combined approach retained significance for both OS and DSS, whilst morphology was also significant for DSS. CONCLUSIONS: Our findings confirm that tumour morphology significantly predicts for survival in OPSCC. However, we found combined tumour morphology and p16 IHC, with additional testing for discordant cases to be superior to either morphology or p16 IHC alone. Further study is required to establish the optimal testing method for HPV in OPSCC particularly in low prevalence populations.

Privacy-preserving distributed learning of radiomics to predict overall survival and HPV status in head and neck cancer.

A major challenge in radiomics is assembling data from multiple centers. Sharing data between hospitals is restricted by legal and ethical regulations. Distributed learning is a technique, enabling training models on multicenter data without data leaving the hospitals ("privacy-preserving" distributed learning). This study tested feasibility of distributed learning of radiomics data for prediction of two year overall survival and HPV status in head and neck cancer (HNC) patients. Pretreatment CT images were collected from 1174 HNC patients in 6 different cohorts. 981 radiomic features were extracted using Z-Rad software implementation. Hierarchical clustering was performed to preselect features. Classification was done using logistic regression. In the validation dataset, the receiver operating characteristics (ROC) were compared between the models trained in the centralized and distributed manner. No difference in ROC was observed with respect to feature selection. The logistic regression coefficients were identical between the methods (absolute difference <10-7). In comparison of the full workflow (feature selection and classification), no significant difference in ROC was found between centralized and distributed models for both studied endpoints (DeLong p > 0.05). In conclusion, both feature selection and classification are feasible in a distributed manner using radiomics data, which opens new possibility for training more reliable radiomics models.