|
|
Validation of the RTOG recursive partitioning
analysis (RPA) classification for small cell lung cancer-only brain metastases:
A single institution experience
G. M. Videtic, D. J. Adelstein, T. Mekhail, T. W. Rice, G. H. J. Stevens,
S.-Y. Lee, J. H. Suh
Cleveland Clinic Foundation, Cleveland, OH
Background.
Brain metastases clinical trials have generally excluded
small cell lung cancer (SCLC) cases.
The RTOG developed 3 prognostic classes for brain
metastases using a recursive partitioning analysis (RPA), however only 4% of
patients had an SCLC diagnosis.
Validation of these RPA classes for SCLC-only is
important for rational trial design.
Methods.
A 20-year retrospective review of 164 SCLC patients with brain
metastases treated between 4/1983 and 5/2003 was performed.
Patients who underwent surgery as part of their
treatment were excluded from this study.
This provided 154 patients for analysis.
RPA criteria used for class assignment were
Karnofsky performance status (KPS), primary tumor status (PD), presence of
extracranial system metastases (SD) and age.
Results.
Median follow-up was 4.7 months (range 0.3-40.3).
Median age was 65 (range 42-85).
Median KPS was 70 (range 40-100).
Median number of lesions was 3 (range1-30) [data on
139 patients].
Number of patients with controlled PD and no SD
was: 20 (13%) and with SD: 27 (18%); without controlled PD and no SD: 34 (22%)
and with SD: 73 (47%).
RPA class distribution was: class I- 8 (5%); class
II- 96 (62%); class III- 51 (33%).
For the whole patient cohort, median survival was
4.9 months, with 4 (2.6%) alive at analysis.
Median survival (in months) by RPA class for
SCLC-only is: class I- 8.6; class II- 5.3; class III- 2.5.
Original RTOG RPA survival (in months) by class is:
class I- 7.1; class II- 4.2; class III- 2.3.
Outcomes for SCLC-only brain metastases appear
similar to the RTOG database.
Conclusions.
This analysis suggests that the RPA classes are valid for
brain metastases from SCLC and supports the inclusion of SCLC patients in future
brain metastases trials.
This classification may also serve as a basis for
historical comparisons.
Copyright 2004 American Society of Clinical Oncology All rights
reserved worldwide.
Source: http://www.asco.org/ac/1,1003,_12-002636-00_18-0026-00_19-00211,00.asp
|
