Session Item

During head and neck radiotherapy (RT), variations such as patient weight loss, tumor regression and/ or diminution of the volume of organs at risk (OARs) are likely to occur. These variations may result in changes in the dose distribution that cannot be compensated for by a simple rigid repositioning. In order to more easily prioritize these variations according to their potential impact, an Image-Guided traffic light protocol (IG-TLP; Figure 1) has been implemented in 2018 in our department as a clinical decision support system.

The aim of this study was to evaluate how often treatment plan adaptation was done after daily CBCT analysis and which anatomical changes reported in the IG-TLP required adaptation. Secondly, we aimed to identify subgroups of patients where re-plans were often executed.

Head and Neck RT treatment information registered in our in-house software between February 2018 and September 2020 by means of our IG-TLP was analysed. Review orders from the IG-TLP resulting in a new CT as well as the patients’ characteristics were carefully evaluated to allow the identification of possible predictors.

The patient related information were weight evolution and tumor site while the treatment parameters were the treatment regimen, the reason for new CT, the week of new CT and major/minor violation of re-planning criteria assessed by the physicians (Table 1).

Among the 266 patients analyzed, our protocol resulted in 807 review orders: 87% of them did not require further action after investigation; in 77 cases (10%), a new CT scan was made; and in 24 cases (3%), plan adaptations were performed. For these 24 re-planned patients, 12 resulted from criteria in the yellow action level, 6 from the orange and 6 from the red level.

When analyzing the new CT scan orders, we realized that certain criteria or a combination of two criteria is more prone to result in plan adaptation (Figure 2).

No strong statistical associations were found between the different criteria evaluated and the need for plan adaptation (p>0.05). However, 42% of re-planning was found in oropharyngeal cancer and 41% in primary RT.

In 15/24 cases, re-planning was performed to ensure an adequate target coverage and in four cases because of a potential increased dose to the OARs. Those re-planning occurred mostly during the first (n=9) and the third week (n=6) of the treatment. Contrary to other studies, most re-planning was observed in patients with weight gain or weight loss up to 5%, rather than in patients with more substantial weight loss. 

Only few RTT-reported IG-TLP changes led to re-planning. Moreover, no strong statistical associations were found between the TLP levels and plan adaptation. Therefore, ideally, TLP should serve as screening, followed by more objective dosimetric criteria (e.g delta dose and remaining sessions), preferably using actually delivered doses, in order to define a predictive model for adaptive planning.