D. Rietveld

Decoding tumour phenotype by noninvasive imaging using a quantitative radiomics approach.

Human cancers exhibit strong phenotypic differences that can be visualized noninvasively by medical imaging. Radiomics refers to the comprehensive quantification of tumour phenotypes by applying a large number of quantitative image features. Here we present a radiomic analysis of 440 features quantifying tumour image intensity, shape and texture, which are extracted from computed tomography data of 1,019 patients with lung or head-and-neck cancer. We find that a large number of radiomic features have prognostic power in independent data sets of lung and head-and-neck cancer patients, many of which were not identified as significant before. Radiogenomics analysis reveals that a prognostic radiomic signature, capturing intratumour heterogeneity, is associated with underlying gene-expression patterns. These data suggest that radiomics identifies a general prognostic phenotype existing in both lung and head-and-neck cancer. This may have a clinical impact as imaging is routinely used in clinical practice, providing an unprecedented opportunity to improve decision-support in cancer treatment at low cost.

Intensity-Modulated Radiotherapy Reduces Radiation-Induced Morbidity and Improves Health-Related Quality of Life: Results of A Nonrandomized Prospective Study Using A Standardized Follow-Up Program

PURPOSE The purpose of this study was to compare intensity-modulated radiation therapy (IMRT) and three-dimensional conventional radiotherapy (3D-CRT) with regard to patient-rated xerostomia, Radiation Therapy Oncology Group (RTOG) acute and late xerostomia and health-related quality of life (HRQoL) among patients with head and neck squamous cell carcinoma (HNSCC). METHODS AND MATERIALS Included were 241 patients with HNSCC treated with bilateral irradiation +/- chemotherapy. Since 2000, all patients treated with HNSCC were included in a program, which prospectively assessed acute and late morbidity according to the RTOG and HRQoL on a routine basis at regular intervals. Before October 2004, all patients were treated with 3D-CRT (N = 150). After clinical implementation in October 2004, 91 patients received IMRT. In this study, the differences regarding RTOG toxicity, xerostomia, and other items of HRQoL were analyzed. RESULTS The use of IMRT resulted in a significant reduction of the mean dose of the parotid glands (27 Gy vs. 43 Gy (p < 0.001). During radiation, Grade 2 RTOG xerostomia was significantly less with IMRT than with 3D-CRT. At 6 months, the prevalence of patient-rated moderate to severe xerostomia and Grade 2 or higher RTOG xerostomia was significantly lower after IMRT versus 3D-CRT. Treatment with IMRT also had a positive effect on several general and head and neck cancer-specific HRQoL dimensions. CONCLUSIONS IMRT results in a significant reduction of patient- and observer-rated xerostomia, as well as other head and neck symptoms, compared with standard 3D-CRT. These differences translate into a significant improvement of the more general dimensions of HRQoL.

A predictive model for swallowing dysfunction after curative radiotherapy in head and neck cancer

INTRODUCTION Recently, we found that swallowing dysfunction after curative (chemo) radiation (CH) RT has a strong negative impact on health-related quality of life (HRQoL), even more than xerostomia. The purpose of this study was to design a predictive model for swallowing dysfunction after curative radiotherapy or chemoradiation. MATERIALS AND METHODS A prospective study was performed including 529 patients with head and neck squamous cell carcinoma (HNSCC) treated with curative (CH) RT. In all patients, acute and late radiation-induced morbidity (RTOG Acute and Late Morbidity Scoring System) was scored prospectively. To design the model univariate and multivariate logistic regression analyses were carried out with grade 2 or higher RTOG swallowing dysfunction at 6 months as the primary (SWALL(6months)) endpoint. The model was validated by comparing the predicted and observed complication rates and by testing if the model also predicted acute dysphagia and late dysphagia at later time points (12, 18 and 24 months). RESULTS After univariate and multivariate logistic regression analyses, the following factors turned out to be independent prognostic factors for SWALL(6months): T3-T4, bilateral neck irradiation, weight loss prior to radiation, oropharyngeal and nasopharyngeal tumours, accelerated radiotherapy and concomitant chemoradiation. By summation of the regression coefficients derived from the multivariate model, the Total Dysphagia Risk Score (TDRS) could be calculated. In the logistic regression model, the TDRS was significantly associated with SWALL(6months) ((p<0.001). Subsequently, we defined three risk groups based on the TDRS. The rate of SWALL(6months) was 5%, 24% and 46% in case of low-, intermediate- and high-risk patients, respectively. These observed percentages were within the 95% confidence intervals of the predicted values. The TDRS risk group classification was also significantly associated with acute dysphagia (P<0.001 at all time points) and with late swallowing dysfunction at 12, 18 and 24 months (p<0.001 at all time points). CONCLUSION The TDRS is a simple and validated measure to predict swallowing dysfunction after curative (CH) RT for HNC. This classification system enables identification of patients who may benefit from strategies aiming at prevention of swallowing dysfunction after curative (CH) RT such as preventive swallowing exercises during treatment and/or emerging IMRT techniques aiming at sparing anatomical structures that are involved in swallowing.

Radiomic feature clusters and Prognostic Signatures specific for Lung and Head & Neck cancer

Radiomics provides a comprehensive quantification of tumor phenotypes by extracting and mining large number of quantitative image features. To reduce the redundancy and compare the prognostic characteristics of radiomic features across cancer types, we investigated cancer-specific radiomic feature clusters in four independent Lung and Head & Neck (H∓N) cancer cohorts (in total 878 patients). Radiomic features were extracted from the pre-treatment computed tomography (CT) images. Consensus clustering resulted in eleven and thirteen stable radiomic feature clusters for Lung and H & N cancer, respectively. These clusters were validated in independent external validation cohorts using rand statistic (Lung RS = 0.92, p < 0.001, H & N RS = 0.92, p < 0.001). Our analysis indicated both common as well as cancer-specific clustering and clinical associations of radiomic features. Strongest associations with clinical parameters: Prognosis Lung CI = 0.60 ± 0.01, Prognosis H & N CI = 0.68 ± 0.01; Lung histology AUC = 0.56 ± 0.03, Lung stage AUC = 0.61 ± 0.01, H & N HPV AUC = 0.58 ± 0.03, H & N stage AUC = 0.77 ± 0.02. Full utilization of these cancer-specific characteristics of image features may further improve radiomic biomarkers, providing a non-invasive way of quantifying and monitoring tumor phenotypic characteristics in clinical practice.

Adenoid cystic carcinoma of the head and neck: a single-center analysis of 105 consecutive cases over a 30-year period

BACKGROUND Adenoid cystic carcinoma is a rare salivary gland malignancy with a poor disease free survival due to frequent distant metastases and late local recurrences. Previous single-center reports on outcome mostly encompass small series. In this report a relative large series of 105 cases is analyzed, all treated at the VU University Medical Center, Amsterdam, The Netherlands over a 30-year period in which treatment strategies remained unchanged. METHODS All cases of ACC of the head and neck between 1979 and 2009 at our institution were analyzed through a medical chart review. Recurrence patterns and possible prognostic factors (T-stage, N-status, age, gender, type of salivary gland involved, histological grade, surgical margins, perineural invasion (PNI) and postoperative radiotherapy (RT)) were analyzed. RESULTS One-hundred and five cases of ACC of the head and neck were identified. Five-, ten- and twenty-year survival rates for overall survival were 68%, 52% and 28%, respectively. T-stage, N-status, surgical margins, histological subtype and age were highly significant predictors for survival. PNI was not a negative prognosticator. CONCLUSIONS T-stage, N-status, surgical margins, histological grade and age are the main predictors of survival-outcome in ACC of the head and neck. Distant metastasis frequently develop, mainly in the first 5 years post treatment. Local recurrences often develop even later on, warranting long term follow up of patients treated for ACC. Grade III ACC should be considered a specific entity within the group of ACC due to its typical aggressive biological behavior and relatively poor outcome, implicating the need for an improved adjuvant treatment.

radiomic Machine-learning classifiers for Prognostic Biomarkers of head and neck cancer

Introduction “Radiomics” extracts and mines a large number of medical imaging features in a non-invasive and cost-effective way. The underlying assumption of radiomics is that these imaging features quantify phenotypic characteristics of an entire tumor. In order to enhance applicability of radiomics in clinical oncology, highly accurate and reliable machine-learning approaches are required. In this radiomic study, 13 feature selection methods and 11 machine-learning classification methods were evaluated in terms of their performance and stability for predicting overall survival in head and neck cancer patients. Methods Two independent head and neck cancer cohorts were investigated. Training cohort HN1 consisted of 101 head and neck cancer patients. Cohort HN2 (n = 95) was used for validation. A total of 440 radiomic features were extracted from the segmented tumor regions in CT images. Feature selection and classification methods were compared using an unbiased evaluation framework. Results We observed that the three feature selection methods minimum redundancy maximum relevance (AUC = 0.69, Stability = 0.66), mutual information feature selection (AUC = 0.66, Stability = 0.69), and conditional infomax feature extraction (AUC = 0.68, Stability = 0.7) had high prognostic performance and stability. The three classifiers BY (AUC = 0.67, RSD = 11.28), RF (AUC = 0.61, RSD = 7.36), and NN (AUC = 0.62, RSD = 10.52) also showed high prognostic performance and stability. Analysis investigating performance variability indicated that the choice of classification method is the major factor driving the performance variation (29.02% of total variance). Conclusion Our study identified prognostic and reliable machine-learning methods for the prediction of overall survival of head and neck cancer patients. Identification of optimal machine-learning methods for radiomics-based prognostic analyses could broaden the scope of radiomics in precision oncology and cancer care.

Development and validation of a nomogram for prediction of survival and local control in laryngeal carcinoma patients treated with radiotherapy alone: a cohort study based on 994 patients.

INTRODUCTION To advise laryngeal carcinoma patients on the most appropriate form of treatment, a tool to predict survival and local control is needed. MATERIALS AND METHODS We performed a population-based cohort study on 994 laryngeal carcinoma patients, treated with RT from 1977 until 2008. Two nomograms were developed and validated. Performance of the models is expressed as the Area Under the Curve (AUC). RESULTS Unfavorable prognostic factors for overall survival were low hemoglobin level, male sex, high T-status, nodal involvement, older age, lower EQD(2T) (total radiation dose corrected for fraction dose and overall treatment time), and non-glottic tumor. All factors except tumor location were prognostic for local control. The AUCs were 0.73 for overall survival and 0.67 for local control. External validation of the survival model yielded AUCs of 0.68, 0.74, 0.76 and 0.71 for the Leuven (n=109), the VU Amsterdam (n=178), the Manchester (n=403) and the NKI cohort (n=205), respectively, while the validation procedure for the local control model resulted in AUCs of 0.70, 0.71, 0.72 and 0.62. The resulting nomograms were made available on the website www.predictcancer.org. CONCLUSIONS For patients with a laryngeal carcinoma treated with RT alone, we have developed visual, easy-to-use nomograms for the prediction of overall survival and primary local control. These models have been successfully validated in four external centers.

Flattening Filter Free vs Flattened Beams for Breast Irradiation

PURPOSE Flattening filter free (FFF) beams offer the potential for a higher dose rate, shorter treatment time, and lower peripheral dose. To investigate their role in large-field treatments, this study compared flattened and FFF beams for breast irradiation. METHODS AND MATERIALS Ten left breast clinical plans comprising 2 tangential beams and a medially located 3-field simultaneous integrated boost (SIB) were replanned. Full intensity modulated radiotherapy (IMRT), hybrid IMRT, electronic tissue compensator (ETC), and multiple static field treatment plans were created for the elective breast volume using flattened and FFF beams, in combination with a 3-field IMRT SIB. Plan quality was assessed and delivery times were measured for all plans for 1 patient. Out-of-field doses were measured using an ionization chamber for an IMRT plan optimized on a corner of simple cubic phantom for both flattened and FFF beams. RESULTS For each technique, mean target volume metrics (planning target volume coverage, homogeneity, conformity) were typically within 3% for flattened and FFF beams. Larger mean differences in boost conformity favoring flattened hybrid (7.2%) and full IMRT (5.5%) plans may have reflected limitations in plan normalization. Calculated heart and ipsilateral lung doses were comparable; however, both flattened and FFF low-dose phantom measurements were substantially higher than calculated values, rendering the comparison of low dose in the contralateral breast uncertain. Beam delivery times were on average 31% less for FFF. CONCLUSIONS In general, target volume metrics for flattened and FFF plans were comparable. The planning system did not seem to allow for accurate peripheral dose evaluation. FFF was associated with a potentially shorter treatment time. All 4 IMRT techniques allowed FFF beams to generate acceptable plans for breast IMRT.

The course of health-related quality of life in head and neck cancer patients treated with chemoradiation: A prospective cohort study

BACKGROUND AND PURPOSE To evaluate the course of health-related quality of life (HRQOL) from diagnosis to 2 years follow-up in patients with head and neck cancer (HNSCC) treated with chemoradiation (CRT). MATERIALS AND METHODS 164 patients completed the EORTC QLQ-C30 and QLQ-H&N35 questionnaires 1 week before and 6 weeks and 6, 12, 18, and 24 months after CRT. Patients were compared to a reference group. A linear mixed-model analysis was used to assess changes in HRQOL over time, and whether this was associated with age, gender, comorbidity, and tumor sublocation. RESULTS Significant differences for the majority of HRQOL scales were observed between patient and reference group at baseline, and follow-up. The course of HRQOL was different for survivors compared to non-survivors. In survivors, improvement over time was observed (in global quality of life, physical, role, and social function, fatigue, pain, swallowing, speech, social eating, and social contacts), while in non-survivors the pattern over time was either no changes in HRQOL or a deterioration (in physical function, social eating and contacts). In both survivors and non-survivors, emotional functioning improved after treatment, but deteriorated in the longer term. Patients with comorbidity reported worse physical function, and patients with oral/oropharyngeal cancer (compared to hypopharyngeal/laryngeal cancer) reported more oral pain and sexual problems, but fewer speech problems. CONCLUSIONS The course of HRQOL of HNSCC patients during the first 2 years after CRT is different for survivors compared to non-survivors and is associated with comorbidity and tumor subsite.

Five-year quality of life results of the randomized clinical phase III (RADPLAT) trial, comparing concomitant intra-arterial versus intravenous chemoradiotherapy in locally advanced head and neck cancer

The purpose of this investigation was to present 5‐years of quality‐of‐life (QOL) results of a multicenter randomized phase III trial, assessing intra‐arterial (IA) versus standard intravenous (IV) chemoradiation for inoperable stage IV head and neck cancer.