Target volume selection and delineation (T and N) for primary RT of oral cavity, oropharyngeal, hypopharyngeal and laryngeal SCC
TL;DRElective neck levels II-IVa for N0 pharyngolaryngeal SCC and Ib-V for N2a-b/N3; primary CTV is GTV + 5 and 10 mm, anatomically edited.
The nodal dose recipe is the part that transfers: SIB with 54-56 Gy to elective levels, 70 Gy in 33-35 fx to nodal GTV + 5 mm, and 10 mm into abutting muscle or parotid as expert opinion. Danish 0-10 mm margin data put relapses mainly inside the GTV, which supports tighter primary CTVs.
In N0 oropharyngeal or laryngeal SCC treated with definitive IMRT, this supports elective coverage of levels II-IVa rather than Ib-V. It does not extend to the postoperative neck, adenoid cystic histology or N2a-N3 disease.
The nodal dose recipe transfers: elective levels at 54-56 Gy, therapeutic nodal CTV of GTV + 5 mm at 70 Gy in 33-35 fx, and 10 mm into abutting muscle or parotid as expert opinion. Danish 0-10 mm margin data put relapses mainly inside the GTV, supporting tighter primary CTVs.
11 details
Narrative review of international consensus guidelines for nodal level selection, node-level delineation and primary-tumor CTV, by authors from the guideline groups. No new patient data.
Primary RT for SCC of the oral cavity, oropharynx, hypopharynx and larynx, p16-negative and p16-positive oropharynx alike. Not valid for the postoperative setting or adenoid cystic carcinoma. After induction chemo, contour from pre-chemotherapy information.
SIB: elective nodal PTV 54-56 Gy in 33-35 fx over 6.5-7 weeks. Therapeutic nodal CTV (nodal GTV + 5 mm) gets 70 Gy in 33-35 fx of 2.12-2.0 Gy. GTV + 10 mm anatomically edited becomes the low-dose CTV-P2, on 2 mm contrast-enhanced planning CT.
The consensus sits between the anatomic school (whole compartments, 2-3 cm where no barrier) and the DAHANCA geometric 5 + 5 mm rims. The authors expect consensus volumes to be smaller than anatomic volumes in all directions, and smaller than DAHANCA volumes mainly axially for T1-T2 tumors.
Support for margin reduction rests on one three-center Danish series and transoral resections in mainly T1-T2 tumors. The volume comparisons are single illustrative cases, and no systematic comparison of the three approaches had been done.
Nodal selection follows long-standing drainage-pattern data and changes little for most centers. The real shift is the smaller geometric primary CTV. Its safety depends on the geographic-miss analysis the authors describe as ongoing.
| Nodal category | Oral cavity | Oropharynx / hypopharynx / larynx |
|---|---|---|
| N0 | Levels I-III | Levels II-IVa; + VIIa for posterior pharyngeal wall |
| N1 | As N0 | As N0, except hypopharynx (high nodal propensity) |
| N2a, N2b, N3 | Ib (Ia if anterior) to V | Ib to V; + VIIa (RP) for oropharynx and hypopharynx |
| N2c | Per N category on each side | Per N category on each side |
| Approach | High-dose CTV | Low-dose CTV |
|---|---|---|
| Anatomic (Eisbruch; Grégoire/Daisne; Lapeyre) | GTV + 3-10 mm by barrier | Whole at-risk compartment; 2-3 cm where no barrier |
| Geometric (DAHANCA) | GTV + 5 mm | Additional 5 mm |
| International consensus | GTV + 5 mm, edited | GTV + 10 mm, anatomically edited (CTV-P2) |
- Geographic-miss rate with 5 + 5 mm geometric primary CTV margins
- Can elective nodal volumes shrink for p16-positive oropharynx?
📚 Sources · 📄 1 paper
The longer read
The useful part of this review is not a new recommendation but its side-by-side view of three contouring philosophies, which shows where the international consensus lands between them. The anatomic school treats microscopic spread as bounded by compartments: once a space is breached, the whole compartment is at risk, and where no barrier exists the default margin grows to 2-3 cm. The DAHANCA geometric school assumes tumor cell density falls with distance from the gross edge and stacks two 5 mm rims. The consensus takes the geometric 5 and 10 mm expansions as its default and then edits them along routes of spread and anatomic barriers. In practice the volume is geometric by default and anatomic in its exceptions, a compromise the authors frame as a balance between local recurrence and morbidity, not as a derived optimum.
The evidence for the smaller primary volume is thinner than the confidence with which it is now applied. It rests on three supports. Pathology series place most infiltration within the first 5 mm. Transoral resections with 2 to 3 mm margins, mostly in T1 and T2 tumors, show low local recurrence without adjuvant RT. And a Danish comparison across three centers using 0 to 10 mm margins found that recurrences sat mainly within the GTV and did not track margin width. That last point is the strongest argument in the paper. The authors read the few out-of-CTV failures, all still inside the 95% isodose, as pointing to biology rather than delineation, which is plausible. Still, it is a retrospective comparison of centers that likely differed in more than margin, and the formal DAHANCA geographic-miss analysis was unfinished at the time of writing. Extending the surgical-margin logic from small transorally resected tumors to bulky primaries with submucosal spread is an inference, not an observation.
The nodal recommendations stand on firmer ground because they codify decades of drainage-pattern data and broadly match existing practice. Two details deserve more attention than they usually get. The 10 mm extension into abutting muscle or parotid around a large node is stated plainly as expert opinion. And the decision not to shrink elective volumes for p16-positive oropharynx rests on the absence of data, not on evidence that the drainage risk is the same. With the newer staging, N categories no longer map cleanly onto nodal burden, which is why the authors steer readers to node number, location and laterality rather than the N label.
What the review cannot settle is whether consensus volumes improve outcome. Its closing claim that the guidelines will improve tumor control and reduce morbidity is an expectation, and no comparative outcome data are presented. The contouring examples are single cases chosen to illustrate, not to measure. A reader should treat the nodal level tables as a reasonable floor for definitive IMRT, and treat the reduced primary CTV as well supported for small primaries and still under audit for large ones.