WASHINGTON — A new preoperative grading system for adult hemispheric low-grade gliomas produced accurate prognoses for disease stage, Dr. Edward F. Chang said at the annual meeting of the American Association of Neurological Surgeons.
“There are few standards for guiding both the medical and surgical management of low-grade gliomas,” said Dr. Chang, a third-year resident in the department of neurologic surgery at the University of California, San Francisco. To construct a system, Dr. Chang reviewed 280 patients who had undergone operations for histologically confirmed grade 2 gliomas (from biopsy to maximal surgical resection) during 1989–2005 at UCSF.
Median patient age was 38 years at presentation. Most (88%) had seizures. A majority had a Karnofsky Performance Scale score of 100 or 90. Most tumors were in the frontal (72%) or temporal lobe (34%); median diameter was 4.5 cm.
In the follow-up period, patients survived for a median of 12 years; 65 died. A total of 134 events of progression or recurrence occurred. The median time of progression-free survival was 6 years.
The investigators used four variables highly predictive of lower survival in their grading. (See box.) The strongest was the presence of tumor in “eloquent” brain regions, particularly the sensorimotor cortex (specifically the pre- and postsensory gyri), perisylvian dominant language areas, insular areas, basal ganglia-internal capsule, thalamus, and hypothalamus. Tumor presence in an area of eloquence was the only significant, independent predictor of progression or recurrence.
The interobserver reliability of the grading system had a kappa value of 0.86 between a neurosurgery resident and an attending neurosurgeon blinded to the outcome of a subset of 200 random cases from the study.
When the study sample was graded, the median survival fell from 16 years for patients with low risk to nearly 11 years for those with intermediate risk and 8 years high risk. The median period of progression-free survival dropped as risk grew.
The researchers also analyzed predictors of the extent of resection. Tumors in an area of eloquence had a diameter greater than 4 cm with diffuse borders on MRI; tumors in the temporal lobe were significantly more likely to be treated with subtotal resection; parietal tumors were significantly more likely to undergo gross total resection.
The grading system currently is undergoing external validation.
Low-Grade Glioma Grading System
▸ Age 50 years or older
▸ Karnofsky Performance Scale score less than 90
▸ Maximum tumor diameter more than 4 cm
▸ Tumor located in “eloquent” area
A “yes” answer is given a score of 1 and a “no,” a score of 0. The categories for risk of death or progression were defined as:
Low = 0–1
Intermediate = 2
High = 3–4
Source: Dr. Chang