Patterns and Predictors of Distant Failure After Stereotactic Body Radiation Therapy for Localized Prostate Cancer: A Retrospective Multi-Institutional Analysis

• Published in: International journal of radiation oncology, biology, physics Vol. 111; no. 3; pp. e292 - e293
• Main Authors: Shabsovich, D., Xiang, M., Katz, A.W., Mantz, C.A., Fuller, D.B., van Dams, R., Philipson, R., Juarez, J.E., Ma, T.M., Pepin, A., Suy, S., Stephans, K.L., Murthy, V., Nickols, N.G., Miszczyk, L., Aghdam, N., Glicksman, R., Collins, S.P., Loblaw, D.A., Kishan, A.U.
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.11.2021
• DOI: 10.1016/j.ijrobp.2021.07.928
• ISSN: 0360-3016

Stereotactic body radiation therapy (SBRT) has emerged as a promising modality for the treatment of prostate cancer (PCa) and involves the delivery of ablative doses of radiation to the prostate. The patterns of failure at the time of biochemical recurrence (BCR) after SBRT remain unknown. The purpose of this study is to explore the development of distant metastasis (DM) following the diagnosis of BCR. We queried a multi-institutional database of 2,430 patients who were prospectively treated with SBRT on single institution phase II trials or registries to identify patients who developed a BCR by the Phoenix criteria or by initiation of salvage therapy for PSA rise at a lower threshold. Patterns of distant failure were obtained for patients who had BCR, including presence and sites of DM. Multivariable analysis (MVA) using logistic regression was used to evaluate the association between multiple clinicopathologic and treatment variables and DM. 150 patients with BCR were identified (6.2% of total). Of these, 19% presented with low-risk, 42% with intermediate-risk, and 39% with high-risk disease. The median PSA at BCR was 4.56, median time to BCR was 36 months, and median follow-up after BCR was 15.5 months. Sixty-seven patients (44.7%) had DM identified at or after the time of BCR. Among those with DM, 13 had M1a disease, 50 had M1b disease, 3 had M1c disease, and 1 had an unknown distribution of disease. Time to BCR was significantly shorter in patients with DM compared to those who did not have DM (28.1 vs. 41.4 months, P = 0.002) and initial PSA (iPSA) was significantly greater (11.0 vs. 8.0, P = 0.03). On MVA, Gleason Grade Group (GG) 4 (vs. GG 1, OR 3.76, P = 0.047) and higher SBRT dose (36.25 Gy vs. 35 Gy: OR 4.03, P = 0.02; 40 Gy vs. 35 Gy: OR 3.73, P = 0.03) were significantly associated with DM. A substantial proportion of patients with BCR after SBRT for localized PCa ultimately develop DM, particularly those with higher iPSA and shorter time to BCR. The association with SBRT dose is likely reflective of the use of higher doses in more modern cohorts with higher-risk disease, while the association with GG likely reflects the more aggressive biology of these tumors at presentation. Further research into patterns of failure after SBRT is needed, especially in the context of more sensitive nuclear imaging. Early predictors of treatment failure are needed to facilitate timely treatment intensification.