Session Item

In the RT of prostate cancer (PCa), simultaneous irradiation of the prostate and pelvic lymph nodes (pLN) requires application of separate CTV-to-PTV margins, since the two target volumes move independently of each other. IGRT protocols aiming to correct for inter-fraction prostate displacements inevitably lead to a residual setup error (RSE) of the pLN. This error must be quantified and included in the calculation of pLN margins. In the present study, we quantified the pLN RSE resulting from a prostate based IGRT protocol and calculated the corresponding pLN margins.

580 treatment fractions from 21 PCa patients that received simultaneous irradiation of the prostate and pLN were included. These patients were initially positioned with the Catalyst surface scanning system (C-RAD, Uppsala, Sweden). Thereafter, a CBCT was acquired to derive the couch translations necessary to align the prostate in accordance with the planning CT (pCT) based on a registration of gold markers. The online registrations were performed in the XVI system (Elekta AB, Stockholm, Sweden). To estimate the pLN RSE, we calculated the difference between these couch translations and translations derived from an offline grayscale registration performed in Mosaiq (Elekta AB, Stockholm, Sweden). For the grayscale registrations, bony anatomy up to the fifth lumbar vertebra were included. For each grayscale registration, a visual assessment was performed to verify that the bony anatomy had been correctly aligned with the pCT. Rotations derived from the grayscale registration were excluded but were manually converted into translations if deemed necessary after visual assessment. The RSE in pLN position was estimated separately in each of the three orthogonal directions. The population systematic uncertainty and random uncertainty of the differences in each direction were calculated and used as input to anisotropic pLN margin calculations, for which we used van Herk’s margin formula. In addition to the pLN RSE, we considered several uncertainty components as either estimated from routine QA measurements, or based on values from the literature.

The resulting distributions of the pLN RSE are shown in Figure 1. The systematic- and random uncertainty of this data set was used together with the additional uncertainty components to calculate anisotropic pLN margins, see Table 1. The resulting margins were 8.2 mm, 10.7 mm, and 9.8 mm in the left-right (LR), anterior-posterior (AP), and superior-inferior (SI) direction, respectively.

Our current isotropic pLN margins of 13 mm for prostate based IGRT can safely be reduced. Margin reduction allows for reduction of dose to organs at risk but can also enable target dose escalation. However, margin reduction must be based on careful consideration of all uncertainties in the planning- and treatment process. Each clinic should therefore derive their own treatment margins based on locally estimated uncertainties.