|
|
Prognostic factors for survival in
adult patients with recurrent glioma enrolled on New Approaches to
Brain Tumor Therapy (NABTT) CNS Consortium phase I and II clinical
trials
K. Carson,
S. A. Grossman, J. D. Fisher and E.
Shaw
Johns Hopkins Univ, Baltimore, MD;
Wake Forest Univ, Winston-Salem, NC
Background. Prognostic
factor analyses have proven useful in predicting outcome in
patients (pts) with newly diagnosed malignant glioma.
Similar analysis should be of benefit in recurrent glioma.
Methods. Between 1995 and
2002, 333 adult pts were enrolled on 10 phase I or II
clinical trials of systemic or local chemotherapy or
brachytherapy for the treatment of recurrent glioma.
The studies had similar inclusion criteria and were
conducted within the NABTT Consortium.
Survival time was calculated from start of treatment until
death or last follow-up.
308 (93%) pts have died.
Univariate Cox proportional hazards (PH) regression analysis was
performed to identify possible prognostic factors associated with
an increased risk of death.
Significant univariate predictors were included in
multivariate analysis.
Recursive partitioning analysis (RPA) was performed, and
the log-rank test was used to group RPA classes with
similar Kaplan-Meier survival estimates (KM).
Results. Factors
associated with an increased risk of death on univariate PH
were increased age, lower KPS, initial and on-study
histologies of GBM, steroid use, shorter time from original
diagnosis to recurrence, and tumor location other than frontal.
Factors that were not significant included gender, race,
number of prior therapies, and anticonvulsant use.
The final multivariate PH model included initial histology
of GBM (relative risk (RR) = 2.04; 95% confidence interval
(CI) = 1.49, 2.79), KPS < 90 (RR=1.35, 95% CI = 1.03,
1.77), 10 year increase in age (RR=1.20, 95% CI=1.06,
1.35), and steroid use (RR=1.56, 95% CI = 1.16,
2.10).
RPA resulted in 5 classes, but 3 had similar KM and were
combined.
Median survival was poorest in GBM pts, age ≥ 50, on
steroids (4.7 months (m); 95% CI=3.5, 5.3), best in pts
with initial histology other than GBM and KPS ≥ 90 (20.2 m;
95% CI=12.2, 30.3), and was 7.4 m (95% CI=6.3, 9.0) for all other
pts.
Conclusions. Initial
histology, KPS, age, and steroid use, are predictive of
survival in recurrent glioma pts.
Phase II and III clinical trials in pts with recurrent
glioma should stratify by or adjust for these factors.
|
