Radiation pneumonitis is the most important dose-limiting toxicity in the treatment of thoracic malignancies amendable for high-dose radiotherapy such as lung or oesophageal cancer.
Several patient-specific factors (e.g. age, smoking history, pre-existing inflammatory lung disease, tumour location and performance score) as well as treatment-related factors (e.g. radiation dose and volume, chemotherapy, hormonal therapy) have been studied as potential predictors of the risk of radiation pneumonitis. The most robust parameters that correlate with radiation pneumonitis are Dose Volume Histogram-related, such as the mean lung dose, the percentage of a volume receiving a certain dose such as the V20 and more complex models. All of these show a low overall accuracy with an area under the receiver-operator curve of about 0.65, although they might be still clinically useful by virtue of their high negative predictive value.
Besides research in the underlying genetics of radiation pneumonitis, the interaction between radiotherapy and most targeted agents has not been elucidated.
At present, validated Dose Volume Histogram parameters can be used in clinical practice. Drugs administered concurrently with irradiation of the lungs should only be carried out in combinations with proven safety in prospective trials.

(BELG J MED ONCOL 2013;7(4):105–10)