Modern IMRT for head and neck cancer: a paradoxical paradigm of simultaneous dose-escalation and de-escalation

Advanced dose distributions and high dose gradients defined the superiority of intensity-modulated radiotherapy (IMRT) in organ-function sparing over conventional and 3-dimentional (3D)-conformal radiotherapy and encouraged treatment intensification for head and neck cancer. Various combinations of higher radiation doses, altered fractionation and chemotherapy have been associated with an increase in frequency, severity and duration of treatment-induced toxicity, that affected negatively patient’s quality of life and thus questioned the benefits of the slightly improved disease control. The demand for improvement of the therapeutic ratio – increase in disease control at minimized treatment-induced toxicity – has brought the abilities of IMRT into the challenge to suggest novel strategies. Since the late nineties of the last century, IMRT has been a standard treatment for head and neck cancer at the Ghent University Hospital. Originally, IMRT treatment protocols related heavily to conventional radiotherapy and used multiple dose levels, ranging from 50-70 Gy to the irradiated neck. The increase in toxicity associated with higher elective neck doses and the substantial improvement in nodal staging in the last decade, cast doubts on the existing practice and made us shift from multiple high-dose to single intermediate-dose prescription to the elective neck in primary and post-operative IMRT for head and neck cancer (Publication 1). Decrease in dose to the elective neck did not change the rates of regional control (87%; multiple dose prescription levels versus 97%; single dose prescription level) - at a still small number of isolated regional relapses (3%; multiple dose prescription levels versus 0; single dose prescription level) suggesting a role for this moderate dose de-escalation to the elective neck. Introducing IMRT for recurrent and second primary head and neck cancer made us reconsider the dose prescription to the tumor and we tried to reach doses as high as for primary cancer, combined with multiple dose levels to the elective neck (Publication 2). Given the increased cumulative risk of treatment-induced toxicity, we tested another toxicity-minimizing strategy: reducing the non-tumor tissue volumes by omitting the elective neck from irradiation. We did not observe any acute grade 5 toxicity but 20% and 50% grade ≥ 3 late toxicity at 2 and 5 years, respectively, including 2 fatal carotid ruptures. Five-year locoregional control and overall survival in our non-selected patient population were 40% and 20%, respectively. High-dose curative IMRT re-irradiation offers a chance of cure to the patients if they are stringently selected. Insights in tumor biology have stimulated the introduction of novel targeted therapy in treatment for head and neck cancer, in the first place anti-epidermal growth factor receptor (EGFR) agents. Based on superior treatment outcome of combined radiotherapy and cetuximab, an anti-EGFR monoclonal antibody, the latter has been recommended in case of contraindication of cisplatinum, a standard conventional chemotherapy agent for the application of chemoradiation. The anticipated lower rates of severe acute mucosal toxicity and swallowing disturbances associated with cisplatinum have indeed been observed with cetuximab, while another, unusually severe skin toxicity has been reported: 49% of patients treated with concomitant radiotherapy and cetuximab developed grade III/IV radiation dermatitis (Publication 3). Whether severe radiation dermatitis is a new dose-limiting toxicity or a sign of drug efficacy is to be found out. It is becoming evident that effectiveness of toxicity-minimizing strategies is conditional on patient-specific toxicity prediction. A few population-based models were proposed to predict risk of treatment-induced swallowing disturbances (dysphagia). We have developed a two-component predictive model incorporating dose/volume parameters of IMRT treatment plans and single nucleotide polymorphisms (SNPs) in DNA repair genes (Publication 4). The model predicts risk of acute grade ≥ 3 dysphagia in patients treated with IMRT with 79% sensitivity and 78% specificity. Further investigation of SNPs in an extended spectrum of genes, as well as other biological parameters of individual susceptibility to radiation-induced damage, is expected to enhance the predictive value of the model. Conventional dose escalation solutions go hand-in-hand with the higher rates of treatment-induced toxicity. A strategy of biological image-guided dose escalation – dose painting – presents a new paradigm of treatment intensification at unchanged or minimized toxicity. Clipping a dose prescription to signal intensity of a biological image might allow dose escalation at tolerable toxicity that might be further minimized if treatment was adapted to the tumor and non-tumor tissue changes detected with per-treatment biological imaging. We performed an adaptive treatment for head and neck cancer using dose painting by numbers (Publication 5). In a phase I trial we demonstrated feasibility of adaptive [18-F]fluoro-2-deoxy-D-glucose positron emission tomography (18F-FDG-PET)-voxel intensity-based IMRT where two dose prescription levels were tested: a median dose of 80.9 Gy to the high-dose clinical target volume (dose level I) and a median dose of 85.9 Gy to the gross tumor volume (dose level II). Because the gross tumor volume was much smaller than the high-dose clinical volume and because target adaptation resulted in reduced target volumes, greater dose escalation was possible with dose level II even though acute severe toxicity was smaller than with dose level I. Optimization of treatment-intensifying strategies should be considered together with toxicity-predicting and toxicity-minimizing strategies. Search for biomarkers of the tumor’s biology and the patient’s genetic and epigenetic variability would facilitate the development of multi-component models predicting an individual risk of treatment-induced toxicity and treatment efficacy to take toxicity-preventive measures and perform pre- and per-treatment adaptations including dose de-escalation, dose escalation or their combination.

Jaar van publicatie:2010

Toegankelijkheid:Closed