Margriet Kwint

Chapter 6 124 proportion of patients with more severe grade 3 AET, independent of use of IMRT, which was not addressed in the old V35 model. An update of the prediction model was therefore needed. For grade 3 AET, the V50 was shown to be the best predictor, and for grade 2 AET the V50 also showed to perform significantly better than the current V35 model. With no significant change in AET incidence compared to patients treated with 3DCRT, it is also logical to find the best predictor at a dosimetric level at which the volume of esophagus was not different between 3DCRT and IMRT (between V45 and V65, Figure 1). Werner-Wasik et al. [7] described in their review that a higher dose, even on a small part of the esophagus, might be a risk factor for AET. They described several dosimetric parameters to be predictive in univariate analysis for grade 2 and 3 AET: V20 till V80. But most at risk for AET were esophagus volume doses receiving >40-50 Gy. This data is consistent with our analysis, were V15 till V70 and D mean and D max of the esophagus were all significantly correlated with AET. The systematic review of Rose et al. [6] demonstrated that the D mean, V20, V30, V40, V45 and V50 were the most studied dosimetric predictors, showing high levels of association with AET. The dosimetric predictors of AET in Rose’s review are consistent with the most significant predictor we found, the V50. Caglar et.al . published an analysis based on 3DCRT and concurrent chemotherapy with 109 patients [4]. These patients were treated with or without induction chemotherapy followed by CCRT with different chemotherapy regimens. Radiotherapy dose varied between 50-68 Gy in 2 Gy fractions. V45 till V60 were indicated as most predictive dose-volume-parameters for AET. Besides the dose on the entire esophagus, Caglar et al. studied the region of the esophagus exposed to a high dose (esophagus infield). The V55 of the entire esophagus and esophagus infield was the most significant parameter to predict AET in multivariate analysis. They showed that when the D mean of the esophagus infield was below 50 Gy, no grade 3 occured. In our analysis we did not specify between entire esophagus dose and esophagus infield, but the dose of 50 Gy is in agreement with our V50 for predicting grade 3 AET. In the current study RT was given with 2.75 Gy fractions. Despite the increased fraction dose, the incidence of grade 3 AET was not higher compared to Caglar et al. [4] (25%), where conventional 2 Gy fractions were used. The radiotherapy dose of our study