br first fractions are calculated using

first 15 fractions are calculated using the original treatment plan and the remaining 20 fractions using the adapted treatment plan. 5: All doses are deformed to the planning
CT using the registrations created in step 1. 6: The individual doses are summed resulting in the estimated actually given dose distribution.
from it. In addition, all plans were evaluated by NTCP values for xerostomia [20], grade 2–4 dysphagia [20], and tube feeding depen-dence [21].
Results
The difference in EAGD between a PTV-plan and a CMRO-plan without plan SB 203580 for a typical case is shown in Fig. 2. This case illustrates a higher dose to the target areas and a dose reduc-tion to the organs at risk, particularly in the areas with high density gradients such as near bone and the esophagus. Similar dose differ-ences were observed in the other cases.
Robust optimization without plan adaptation
The nominal dose distribution of the CMRO-plan showed signif-icantly higher dose to the primary and prophylactic CTVs and sig-nificantly lower dose to the ipsilateral parotid gland, the inferior and superior PCM and the oral cavity compared to the PTV-plans. The NTCP values derived from the nominal dose were within 
0.5% of the values calculated from the EAGD for both the CMRO-plan and PTV-plan.
Robust optimization with plan adaptation
CTV underdosage was not observed in the studied patients. However, the RNTCP of 4/10 patients exceeded the 2.5% threshold after three weeks of treatment. Table 2 shows the results based on the EAGD derived from the adapted and non-adapted PTV-based and CMRO-based treatment simulations. In these four patients, the mean dose in the OARs was on average 1.1 (range: 1.1 to 3.1) Gy lower in the CMRO-plans than the PTV-based approach (Table 2). The plan adaptation resulted into an additional 0.7 (range: 0.5 to 1.2) Gy dose reduction to the OARs. For these patients, the estimated NTCPs were on average reduced by 5.8% and 7.9% using an adapted PTV-plan and adapted CMRO plan, respectively.
CBCT-based dose accumulation validation
The absolute difference of the D98% of both CTVs derived from a set of weekly CBCTs and verification CTs was on average 0.5 Gy and the differences in NTCP values was on average 0.6% (Supplement Fig. S.1). Comparing the dose from two different sets of weekly CBCTs, this difference was 0.1 Gy for the CTVs and 0.3% for the NTCP values.
Discussion
We evaluated CMRO in VMAT of head and neck cancer patients, and assessed plan robustness using the estimated actual given dose. It was demonstrated that CMRO resulted in approximately 2 Gy lower mean dose to several OARs with similar or improved CTV coverage as compared to conventional PTV-based plans. On average, CMRO led to around a 2.3 Gy average dose decrease in the parotids, pharyngeal muscle superior, cricopharyngeal muscle and the supraglottic larynx and a 1.7 Gy increase of the D98% of the primary and elective CTVs.
74 Robust optimization of VMAT
Fig. 2. Differences in estimated actually given dose for a PTV optimized and a robustly optimized VMAT treatment plan: typical case (a) Difference in estimated actually given dose distribution. (b) Dose–volume histogram of the estimated actually given dose for both treatment plans.
Other studies, albeit on different treatment sites, show similar results with CMRO in terms of OAR dose sparing. It was shown for CMRO in breast cancer patients that target coverage improved at identical or reduced OAR dose [6,23]. The agreement between the planned nominal dose and dose evaluated on the 4DCT was better for the CMRO-plans than the PTV-plans [6]. Archibald-Heeren et al. compared CMRO VMAT of lung cancer on a thorax phantom [24]. The authors found fewer maximum and minimum dose variations compared to other current treatment techniques such as internal target volume based planning as evaluated on a 4DCT [24]. Zhang et al. investigated the benefit of CMRO for five prostate cancer patients in which the CTV and OAR delineations were shifted inside the patient to create different scenarios [25]. Their method did not include a dose recalculation per scenario and accounted for internal target motion rather than patient posi-tioning errors. They reported a mean dose reduction of 6.4% and 19.7% for the rectal and bladder walls, respectively. Previous stud-ies on CMRO of proton therapy found larger benefits for CMRO, which was expected due to the inherent lack of dose invariance of the treatment modality [4,13,26].