onc brain

The longer read

This document belongs with contouring infrastructure, not with evidence that moves an outcome. What makes it useful is consensus across seven cooperative groups spanning Europe, Asia, Australasia and North America. Once DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG and TROG share one set of level boundaries, a level IVa in one trial protocol means the same volume as a level IVa in another. That matters more for interpreting patterns-of-failure and elective-nodal trials than for any single patient, because a failure counted as out-of-field in one dataset could have been in-field under a different atlas.

The most consequential choices are boundary reassignments. Robbins-based surgical teaching originally set the top of level V where the sternocleidomastoid meets the trapezius. Here the panel moves it to the hyoid bone, a landmark that reads easily on axial CT, and gives the uppermost posterior nodes to occipital level Xb. The lower neck is split the same way. IVa stops 2 cm above the sternal manubrium, the panel noting that a neck dissection rarely reaches the sternoclavicular joint. Below that, IVb (medial) and Vc (lateral) are named supraclavicular groups. The effect is to separate what a surgeon removes from what a radiation oncologist may have to cover, which is the correct basis for a CTV atlas.

Several definitions encode clinical judgment rather than pure anatomy. For example, VIa is reserved for lower lip and advanced gingivo-mandibular primaries, and VIb's upper limit drops to the hyoid body only for anterior floor of mouth, tongue tip and lower lip primaries. VIIa is limited to the lateral retropharyngeal nodes. Each of these narrows the default volume, and each depends on the panel's reading of drainage patterns rather than on outcome data presented in this paper. A reader who includes the medial retropharyngeal nodes, or who covers VIa more widely, is not departing from evidence shown here. They are departing from the panel's position.

The main methodological limits are that the boundaries hold only in a neutral position and that the source reports no measure of agreement within the panel or between users. Hyperextension, which some centres use to reduce moist desquamation in skin folds, changes the Ib/II relationship and the skull-base landmarks, and the panel explicitly leaves that adaptation to clinical judgment. The document also states plainly that level selection and dose are outside its scope. It defines the boxes and leaves open which boxes to fill. Pairing it with site- and stage-specific selection guidance is what turns these level boundaries into a treatment volume.