Stereotactic body radiation therapy for melanoma and renal cell carcinoma: impact of single fraction equivalent dose on local control

Abstract

Background: Melanoma and renal cell carcinoma (RCC) are traditionally considered less radioresponsive than other histologies. Whereas stereotactic body radiation therapy (SBRT) involves radiation dose intensification via escalation, we hypothesize SBRT might result in similar high local control rates as previously published on metastases of varying histologies.

Methods: The records of patients with metastatic melanoma (n = 17 patients, 28 lesions) or RCC (n = 13 patients, 25 lesions) treated with SBRT were reviewed. Local control (LC) was defined pathologically by negative biopsy or radiographically by lack of tumor enlargement on CT or stable/declining standardized uptake value (SUV) on PET scan. The SBRT dose regimen was converted to the single fraction equivalent dose (SFED) to characterize the dose-control relationship using a logistic tumor control probability (TCP) model. Additionally, the kinetics of decline in maximum SUV (SUVmax) were analyzed.

Results: The SBRT regimen was 40-50 Gy/5 fractions (n = 23) or 42-60 Gy/3 fractions (n = 30) delivered to lung (n = 39), liver (n = 11) and bone (n = 3) metastases. Median follow-up for patients alive at the time of analysis was 28.0 months (range, 4-68). The actuarial LC was 88% at 18 months. On univariate analysis, higher dose per fraction (p < 0.01) and higher SFED (p = 0.06) were correlated with better LC, as was the biologic effective dose (BED, p < 0.05). The actuarial rate of LC at 24 months was 100% for SFED ≥45 Gy v 54% for SFED <45 Gy. TCP modeling indicated that to achieve ≥90% 2 yr LC in a 3 fraction regimen, a prescription dose of at least 48 Gy is required. In 9 patients followed with PET scans, the mean pre-SBRT SUVmax was 7.9 and declined with an estimated half-life of 3.8 months to a post-treatment plateau of approximately 3.

Conclusions: An aggressive SBRT regimen with SFED ≥ 45 Gy is effective for controlling metastatic melanoma and RCC. The SFED metric appeared to be as robust as the BED in characterizing dose-response, though additional studies are needed. The LC rates achieved are comparable to those obtained with SBRT for other histologies, suggesting a dominant mechanism of in vivo tumor ablation that overrides intrinsic differences in cellular radiosensitivity between histologic subtypes.

Figure 1

Local Control. Actuarial local control for both melanoma and RCC lesions

Figure 2

Tumor Control Probability. Tumor Control Probability graphs generated from dose response relationship modeling. Doses to individual lesions were grouped into tertile bins and the x-axis value was the mean dose given in that bin, expressed as either (a) SFED or (b) BED. The y-axis value was the probability of LC at 12 months.

Figure 3

Overall survival. Actuarial overall survival of patients based on disease state. Oligometastatic disease was defined as three or less metastases in which all site of disease were treated with aggressive local therapy. Extensive disease was defined as patients with more than three sites of metastases.

Figure 4

Change in SUV for controlled lesions. The Standardized uptake values were plotted with pre-treatment PET used for planning as time 0. Follow-up PET/CT's were fused and SUV was generated for each treated lesion that was controlled. An exponential equation was generated revealing a post-treatment baseline level of activity of 2.6 at 7 months.