PV QA 4 - Poster Viewing Q&A 4
Purpose/Objective(s): Stereotactic Body Radiation (SBRT) utilizes a variety of immobilization, planning, and delivery techniques and relies upon the accurate delivery of very high-dose radiation with sharp gradients to small, moving targets in low density background. We investigated the impact of SBRT dose-delivery factors on local failure (LF) by surveying our 12 year experience with e-NSCLC.
Materials/Methods: We identified 1,085 consecutive patients (pts) from a prospective registry that were treated with definitive SBRT for e-NSCLC between 2005 and 2016. Given the technological evolution over this interval, pts were planned with either pencil beam (PB) or collapsed cone convolution (CCC) dose calculation algorithms (DCA), using open (dynamic arcs) or modulated beams (IMRT or VMAT), immobilized by abdominal compression or automatic breathing control (ABC, frequently selected for patients with greater expected tumor motion and able to tolerate ABC), and treated with or without available CBCT (aligned to external fiducials and KV x-rays to bone if no CBCT, PTV margins were not altered based on availability of CBCT). We limited our analysis to standard fractionation regimens, [60 Gy/3, 48 Gy/4, 50 Gy/5, & 30-34 Gy/1) chosen at the discretion of the treating physician in a risk-adapted approach relative to tumor size and location. The effect of these technical variables along with known patient and tumor factors on LF was analyzed using Fine and Gray univariate regression, with significant predictors selected for a forward step-wise multivariate regression model.
Results: At mean follow-up time of 25.6 months the cumulative incidence of LF at 1, 2, & 5 years was 3.0, 8.3, and 9.8% respectively. Overall survival (1, 2, 5 years) was 83, 62, & 28%. Univariate correlates with LF were PB TPS (HR 2.87, p=0.0004), modulated beam (HR 2.3, p=0.005), lack of CBCT (HR 2.69, p=0.0004), SBRT dose relative to 60 Gy/3 (HR 5.2, p=0.0001 for 4-5 fx; HR 2.7, p=0.051 for 1 fx), tumor size (HR 1.2 per cm, p=0.0009), PET SUV (HR 1.04 per SUV, p=0.0039), and squamous histology (HR 1.8, p=0.0051). Immobilization with ABC (n=96) versus abdominal compression (n=989) did not correlate with LF (p=0.99). On multivariate analysis PET SUV, modulated beam, and use of CBCT were no longer significant correlates with LF, while TPS (HR 2.62, p=0.0019), SBRT dose (HR 4.1, p=0.0009 for 4-5 fx relative to 60 Gy/3)& HR 2.9, p=0.039 for 1 fx versus 60 Gy/3), tumor size (HR 1.2 per cm, p=0.042) and squamous histology (HR 1.7, p=0.027) remained statistically significant.
Conclusion: While the use of PB versus CCC DCA was associated with higher rates of LF after SBRT, the use of abdominal compression vs ABC (univariate), open vs modulated beam, and CBCT vs bony alignment (multivariate) were not correlated with higher rates of LF after SBRT in e-NSCLC.
The asset you are trying to access is locked. Please enter your access key to unlock.