ESTRO 2024 - Abstract Book

S4050

Physics - Inter-fraction motion management and offline adaptive radiotherapy

ESTRO 2024

1 Isala Hospital, Department of Radiation Oncology, Zwolle, Netherlands. 2 University Medical Center Groningen, Department of Radiation Oncology, Groningen, Netherlands

Purpose/Objective:

Since 2022, updated CTV-to-PTV margins are used in breast-cancer radiotherapy (RT) in several clinics in the Netherlands, i.e. 5 mm CTV-to-PTV margin for breast and 5 mm (level 1-2) or 8 mm (level 3-4) for lymph node (LN) target volumes. The goal of this study is to analyze retrospectively whether using these margins leads to adequate CTV coverage, by specifically taking into account inter-fraction variations using daily CBCTs.

Material/Methods:

Data of 33 breast-cancer patients, treated with nine varying local or loco-regional RT regimes without replanning, was analyzed. Treatment techniques included 3D-CRT and VMAT, or VMAT alone. Clinical treatment plans, planning CTs (pCTs) and 431 daily set-up CBCTs, representing the anatomy during the treatment fractions, were available. HU mass density conversion was performed using six density regions and the field-of-view was complemented with water equivalent material according to the pCT external contour, which resulted in the creation of synthetic CTs (sCTs). Target volumes included the whole breast (breast), with or without a boost volume (boost), or partial breast (PB). LN CTVs comprised levels 1-2 (L1/L2), levels 3-4 (L3/L4), internal mammary nodes (IMN), or levels 1-3 (axilla). Dose distributions were recalculated on each sCT to represent the daily fraction dose. Using DIR between pCT and sCT, focusing on the patient contour, two different methods were used to daily evaluate CTV coverage: (A) establishing daily fraction dose, based on deformed and transferred CTV contours from the pCT; and (B) establishing accumulated CTV dose progressively, by deforming each sCT dose back to the pCT and creating a weighted sum per day. In method A, breast and nodal CTVs were deformed from the pCT to the sCT using DIR, and clipped 5 mm below the patient contour. Boost volumes were rigidly transferred from the pCT to the sCTs. Target coverage, i.e. dose to 98% of the CTV volume (D98), was assessed for both methods. D98 parameters were daily-averaged over all patients per CTV, leading to the average D98 per day (method A) and per progressive accumulation (method B). For this average, interquartile range (IQR), confidence interval (CI) and correlation to the fraction number (time trend) using univariate regression analysis were determined, significant if p <0.001.

Results:

Overall, both methods A and B showed that coverage was well maintained. For all individual sCTs (A), CTV D98 was >90% for all CTVs, except for a few sCTs for L3/L4, IMN and axilla (Table 1). For all deformed sCT doses D98 (B) was >90% for all CTVs, except for a few deformed doses for breast, boost, L3/L4, IMN and axilla. Significant correlation was found between these two methods for all CTVs except for L3/L4 and axilla. Patient D98 average sCT (A) vs total accumulation (B) correlated significantly for breast only, but if a patient D98 average sCT (A) was insufficient, also the total accumulation (B) was insufficient.

Table 1. CTV coverage per individual sCT (method A) and deformed sCT dose (method B) and average per patient using both methods.

CTV