Ten-Fraction Stereotactic Radiosurgery With Different Gross Tumor Doses and Inhomogeneities for Brain Metastasis of >10 cc: Treatment Responses Suggesting Suitable Biological Effective Dose Formula for Single and 10 Fractions.

Stereotactic radiosurgery (SRS) with >5 fractions (fr) has been increasingly adopted to improve local control and safety for brain metastases (BM) of >10 cm3, given the limited brain tolerance of SRS with ≤5 fr. However, the optimal indication and treatment design, including the prescribed dose and distribution for 10 fr SRS, remains uncertain. A single fr of 24 Gy provides approximately 95% of the one-year local tumor control probability. The potential SRS doses in 10 fr that is clinically equivalent to a single fr of 24 Gy regarding anti-tumor effect range from 48.4 to 81.6 Gy as biological effective doses (BED) as a function of the BED model formulas along with the alpha/beta ratios. The most appropriate BED formula in conjunction with an alpha/beta ratio to estimate similar anti-BM effects for single and 10 fr remains controversial. Herein, we describe four cases of symptomatic radiation-naïve BM >10 cm3 (range, 11 to 26 cm3), treated with 10 fr SRS with a standard prescribed dose of 42 Gy, for which modified dynamic conformal arcs were used with forward planning to improve dose conformity. In the first two cases with gross tumor volumes (GTV) of 15.3 and 10.9 cm3, 42 Gy was prescribed to 70%-80% isodose, normalized to 100% at the isocenter, which encompasses the boundary of the planning target volume: GTV + isotropic 1 mm margin. The tumor responses were initially marked regression followed by regrowth within three months in case 1 and no shrinkage with subsequent progression within three months in case 2. In the remaining two cases with larger GTVs of 19.1 and 26.2 cm3, the GTV boundary and 2-3 mm margin-added object volume was covered by 80% and 56% isodoses with 53 Gy and 37 Gy, respectively, to further increase the marginal and internal doses of GTV and to ensure moderate dose spillage outside the GTV, while >1-1.5 mm outside the GTV was covered by 42 Gy with 63% isodose. According to the BED based on the linear-quadratic (LQ) model with an alpha/beta ratio of 10 (BED10), 53 Gy corresponds to approximately 81 Gy in BED10 and 24 Gy in a single fr. Excellent initial maximum tumor response and subsequently sustained tumor regression (STR) were achieved in both cases. Subsequently, enlarging nodules that could not exclude the possibility of tumor regrowth were disclosed within two years, while late adverse radiation effects remained moderate. These dose-effect relationships suggest that a GTV marginal dose of ≥53 Gy with ≤80% isodose would be preferred to effect ≥1-year STR and that further dose escalation of both marginal and internal GTV may be necessary to achieve ≥2-year STR, while GTV of >25 cm3 may be unsuitable for 10 fr SRS in terms of long-term brain tolerance. Among LQ, LQ-cubic, and LQ-linear model formulas and alpha/beta ratios of 10-20, BED10 may be clinically most suitable to estimate a 10 fr SRS dose that provides anti-BM efficacy similar to that for a single fr.