Target volumes in adaptive treatment of NSCLC show large discrepancies among experts

Purpose or Objective
Target volume delineation is subject to inter-observer variability (IOV) and thus a major source of uncertainty in radiation treatment planning. IOV in the context of treatment adaptation is largely unknown. We analysed IOV of the primary gross tumour volume (GTV) and clinical target volume (CTV) in locally advanced NSCLC delineated by experts in the pre- and mid-treatment (PT and MT) setting.

Material and Methods
Five patients from routine clinical practice, who underwent repeat imaging, were selected such that a variety of features likely to prompt adaptation, e.g. atelectasis, central primary tumour, was included. Brief case reports and CT imaging data were sent to six observers, supplemented with FDG-PET studies in two cases at PT and three at MT. Observers received PT and MT imaging data at least one week apart and were asked to delineate the GTV and CTV in their own treatment planning system and to comment on their delineation process, i.e. on how GTV to CTV expansion and adaptation were performed. Delineations were rasterised on a 1mm³ grid and their compatibility assessed with the Generalised Conformity Index (CIgen). Differences in IOV between PT and MT were probed with Wilcoxon signed-rank tests and the correlation between GTV and CTV IOV evolution with Spearman rank correlation. While it is respectively impossible and very difficult for these tests to show two-sided α=0.05 significance at n=5, more powerful parametric alternatives cannot credibly be employed.

Results
A total of 109/120 delineations were received and analysed. Figure 1 shows an overview of IOV in terms of volume overlap. All but one case saw a reduction of CIgen when transitioning from PT to MT (p=0.125 for both GTVs and CTVs). Agreement in corresponding GTV and CTV delineations was generally comparable at individual time points, and there was a trend of a correlation between GTV and CTV CIgen changes (ρ=0.9, p=0.083). This behaviour is consistent with almost all observers adopting an isotropic GTV to CTV expansion of 5 mm with rare editing for anatomical boundaries at either time point. MT GTV delineations were only occasionally adopted from PT after image registration (9/28 analysed observers and cases), the remainder contoured anew. Post-transfer adaptation to changed anatomy was performed by one observer, and inclusion of PT GTV in MT CTV by two. Figure 2 shows the two cases demonstrating the most extreme changes in IOV (improvement and deterioration). In both cases large volumes of consolidated lung lead to increased IOV, but at different time points. Shortcomings of this in silico contouring challenge were incomplete FDG-PET availability and a lack of contrast-enhanced CT, both current clinical standard. Modern MRI would also have been welcomed by many.

Conclusion
Differing approaches of reacting to intra-therapeutic changes further increase IOV in target volume delineation in the adaptive setting. Consensus guidelines and in-depth analysis of per-treatment tumour shrinkage patterns are urgently needed.