What is the best grading system in stereotactic radiosurgery for brain metastases, RTOG-RPA, SIR, BSBM, GPA or a modified RPA?





Keywords: brain metastasis, radiosurgery, gamma knife, grading system, outcome

Interactive Manuscript

Ask Questions of this Manuscript:

   



Abstract

     A new index for Recursive Partitioning Analysis (RPA) classification of patients with brain metastases proposed by Masaaki Yamamoto was applied to our patient series for comparison with previously established systems including the original RPA, score index for SRS (SIR), basic score for brain metastases (BSBM) and graded prognostic assessment (GPA), in order to ascertain which grading system is the best for gamma knife surgery (GKS).
     We evaluated our results according to different available grading systems.
     The application of radiosurgery for non-neoplastic spine disease also deserves future investigation.
     Twenty-two hundred and forty-six patients with brain metastases, treated by the first author with GKS during the 1998-2010 period at Chiba Cardiovascular Center and Tsukiji Neurological Clinic, were analyzed.
     There were 1341 men and 895 women. Median age was 65, range 7-94. Extracranial disease status was judged to be “controlled” in 281. Median tumor numbers were 3, range 1-100.  The most common primary site was the lung (1490 cases), followed by gastro-intestine (277), breast (230), uro-genital (136) and others (113). There were 133 (MST 24.8Mo) RPA class I, 1812 (7.8) II and 301 (3.6) III cases, with statistically significant differences (p<0.0001) between all pairs of classes. According to Yamamoto’s new system, MSTs were 17.1 months in class II-a (267), 10.5 in II-b (639) and 8.9 in II-c (884) with p-values <0.0001 (II-a vs II-b and II-b vs II-c). There were 367 cases in SIR 0-3 (MST 3.1 months), 1703 in SIR 4-7 (8.3), and 176 in SIR 8-10 (21.7) (all; p<0.0001). There were 327 BSBM 0 (MST 2.6 months), 881 BSBM 1 (5.1), 793 (11.4) BSBM 2 and 245 (24.4) BSBM 3 cases, with all p values <0.0001. There were 928 GPA 0-1.0 (MST 5.0 months), 1160 GPA 1.5-2.5 (9.3), 120 GPA 3.0 (20.4), and 38 GPA 3.5-4.0 (27.4) cases (GPA 3.0 vs GPA 3.5-4.0 ; ns, others ; p<0.0001).
     This was a retrospective analysis.
     Yamamoto’s new system appears to be the best grading system, showing highly statistically significant differences, with a well-balanced patient class distribution and simple routine clinical application.
     Grading systems should be tested and validated.


Acknowledgements

Project Roles: