Contouring variation affects estimates of normal tissue complication probability for breast fibrosis after radiotherapy

Abstract

Background: Normal tissue complication probability (NTCP) models can be useful to estimate the risk of fibrosis after breast-conserving surgery (BCS) and radiotherapy (RT) to the breast. However, they are subject to uncertainties. We present the impact of contouring variation on the prediction of fibrosis.

Materials and methods: 280 breast cancer patients treated BCS-RT were included. Nine Clinical Target Volume (CTV) contours were created for each patient: i) CTV_crop (reference), cropped 5 mm from the skin and ii) CTV_skin, uncropped and including the skin, iii) segmenting the 95% isodose (Iso_95%) and iv) 3 different auto-contouring atlases generating uncropped and cropped contours (Atlas_skin/Atlas_crop). To illustrate the impact of contour variation on NTCP estimates, we applied two equations predicting fibrosis grade ≥ 2 at 5 years, based on Lyman-Kutcher-Burman (LKB) and Relative Seriality (RS) models, respectively, to each contour. Differences were evaluated using repeated-measures ANOVA. For completeness, the association between observed fibrosis events and NTCP estimates was also evaluated using logistic regression.

Results: There were minimal differences between contours when the same contouring approach was followed (cropped and uncropped). CTV_skin and Atlas_skin contours had lower NTCP estimates (-3.92%, IQR 4.00, p < 0.05) compared to CTV_crop. No significant difference was observed for Atlas_crop and Iso_95% contours compared to CTV_crop. For the whole cohort, NTCP estimates varied between 5.3% and 49.5% (LKB) or 2.2% and 49.6% (RS) depending on the choice of contours. NTCP estimates for individual patients varied by up to a factor of 4. Estimates from "skin" contours showed higher agreement with observed events.

Conclusion: Contour variations can lead to significantly different NTCP estimates for breast fibrosis, highlighting the importance of standardising breast contours before developing and/or applying NTCP models.

Keywords: Breast cancer; Fibrosis; Inter-observer variation; Late effects; NTCP modelling; Radiotherapy.

Fig. 1

Example of all breast contours used in this study. (A) CTV_crop and CTV_skin, (B) atlas generated contours including skin and (C) excluding skin, and (D) 95% isodose level of the prescribed dose (Iso_95%).

Fig. 2

Variation in contours across the cohort of 280 patients. (A) Breast volume of each different contour generation method extracted from Raystation, (B) The volume difference of simulated contours compared with CTV_crop (cm³), (C) the DSC, (D) the MDA (in cm), and (E) the HD (in cm) for simulated contours compared with CTV_crop.

Fig. 3

An example of contour similarity between CTV_crop and Iso_95% when generated from (A) 3D-CRT and (B) Volumetric Arc Radiotherapy, and 3D dose distribution from (C) 3DCRT and (D) Volumetric Arc Radiotherapy planning technique.

Fig. 4

The estimation of NTCP for grade 1+ breast fibrosis at 5 years after radiotherapy from (A) LKB model and (B) RS model. (C) The difference percentage of NTCP estimation of simulated contours compared with CTV_crop for LKB model, and (D) for RS model. The relative difference was calculated by $\frac{{NTCP}_{simulatedcontour}-{NTCP}_{CTV\_crop}}{{NTCP}_{CTV-crop}}\times 100$. (A) and (B) give an overview of the range of NTCP over the whole patient cohort, while (C) and (D) represents the range of the impact of contour variation on NTCP estimates for individual patients.

Fig. 5

The correlation of NTCP between contours including skin (CTV skin and all Atlas skin contours) and excluding skin (CTV crop and all Atlas crop contours) of (A) LKB model and (B) RS model.

Fig. 6

The NTCP estimates of difference fractionation scheme from LKB (A) and RS (B) models for including (skin) and excluding skin (Crop) contours of non-fibrosis cohort and fibrosis grade ≥1 cohort (see alsoAppendix B).