1024 - CT-Guided Online Adaptive Stereotactic Body Radiation Therapy for Liver Tumors

A. Lukez¹, E. M. Horwitz¹, T. J. Galloway¹, M. A. Hallman¹, S. S. Kumar¹, J. K. Wong¹, R. M. Shulman¹, C. M. C. Ma², A. Eldib Jr², J. Panetta², Z. Kiss¹, R. H. Freeman¹, and J. E. Meyer¹; ¹Department of Radiation Oncology, Fox Chase Cancer Center, Philadelphia, PA, ²Fox Chase Cancer Center, Philadelphia, PA

Purpose/Objective(s): We present a preliminary CT-guided online adaptive radiation therapy (CT-ART) experience in stereotactic body radiation therapy (SBRT) for liver tumors. Materials/

Methods: Patients with hepatocellular carcinoma or liver metastasis treated with CT-ART were enrolled on a retrospective registry study. Physicians were offered 2 plans, the original plan transposed on a CBCT with adapted contours (sched) and a new plan generated with updated contours (adap). Student’s t-test compared pt time in vault (TIV), planning target volume (PTV) coverage, and organs at risk (OAR) dose. ?2 test compared dose constraint violations.
Results: Thirteen pt (median age: 67, median ECOG: 1, 85% male) received SBRT 45 – 60 Gy (median: 50 Gy) in 5 fractions (fx). Median PTV 98.6 cc (range: 15.5 - 590.4 cc). Of 65 fx, 77% (50) were adapted. PTV V100%=95% criteria were not met in 35% of scheduled and 6% of adapted plans (p < 0.001). Mean PTV V100% was not significantly different among all plans (sched: 94.9%, adap: 95.7%; p = 0.088). Mean internal target volume (ITV) D100% was better for all adapted plans (sched: 8.89Gy/fx, adap: 9.31Gy/fx; p = 0.003). Mean PTV V100% coverage was better for adapted plans when PTV and OAR overlapped (sched: 93.2%, adap: 95.1%; p = 0.046). Nearest OAR maximum point dose (Dmax) was lower for adapted plans when PTV-OAR overlapped (p = 0.025). Mean stomach Dmax was lower for adapted plans (sched: 4.1 Gy/fx, adap: 3.3 Gy/fx; p = 0.001), with 5 scheduled plans not meeting criteria. Mean small bowel Dmax was lower for adapted plans (sched: 3.2Gy/fx, adap: 2.6 Gy/fx; p = 0.006), with 2 scheduled plans not meeting criteria. Mean heart Dmax was greater for adapted plans (sched: 3.1 Gy/fx, adap: 3.3 Gy/fx; p = 0.013); 1 scheduled plan did not meet criteria. TIV was greater for the first 2 pt (median: 55 min) versus the final 2 pt (median: 42 min) [p = 0.013]. While pt TIV was longer for fx 1 (median [min]: 66) than fx 5 (42), there were no differences in TIV between fx 2 (52), 3 (53), or 4 (52) versus fx 5. Addition of single breath-hold (SBH) CBCT acquisition led to a non-significant improvement in TIV (w/o SBH: 59 min, w/ SBH: 51 min; p = 0.058), non-significant improvement in mean PTV V100% (adap: 95.3%, SBH adap: 96.1%; p = 0.09), and did not affect OAR Dmax (adap: 5.2Gy/fx, SBH adap: 4.5Gy/fx; p = 0.19).
Conclusion: CT-ART SBRT for liver tumors is feasible. In this series, adapted plans increased PTV coverage and ITV coverage. Among all cases, adapted plans improved mean stomach Dmax, improved mean small bowel Dmax, and increased mean heart Dmax. Among instances of PTV-OAR overlap, adapted plans improved mean PTV coverage and decreased OAR Dmax. Table. Nearest OAR Constraints

		Scheduled	Adapted	P (95% CI)
Dmax, Mean [Gy/fx] (range)
	All (n = 13)	5.5 (1.0 - 10.9)	5.0 (0.8 - 10.2)	0.052 (-0.004, 1.0)
	PTV-OAR overlap (n = 6)	6.9 (1.0 - 10.9)	5.6 (0.8 - 7.0)	0.025 (0.2, 2.4)
	PTV-OAR 0.1 - 0.5 cm (n = 4)	5.5 (2.2 - 10.2)	4.8 (1.9 - 10.2)	0.12 (-0.2, 1.7)
	PTV-OAR > 0.5 cm (n = 3)	4.1 (2.1 - 6.4)	4.6 (2.7 - 6.7)	0.024 (-1.0, -0.1)