Search Results

You are looking at 1 - 9 of 9 items for

  • Author or Editor: Mitchel S. Berger x
  • Refine by Access: all x
  • By Author: Keles, G. Evren x
Clear All Modify Search
Restricted access

G. Evren Keles, Kathleen R. Lamborn, and Mitchel S. Berger

Object. The goal of this study was to perform a critical review of literature pertinent to low-grade gliomas of the cerebral hemisphere in adults and, on the basis of this review, to evaluate systematically the prognostic effect of extent of resection on survival and to determine if treatment-related guidelines could be established for patients in whom these tumors have been newly diagnosed. Quality of evidence for current treatment options, guidelines, and standards as well as methodological limitations were evaluated.

Methods. Several prognostic factors thought to affect outcome in patients with low-grade gliomas include the patient's age and neurological status, tumor volume and histological characteristics, and treatment-related variables such as timing of surgical intervention, extent of resection, postoperative tumor volume, and radiation therapy. Patient age and the histological characteristics of the lesion are generally accepted prognostic factors. Among treatment-related factors, timing and extent of resection are controversial because of the lack of randomized controlled trials addressing these issues and the difficulty in obtaining information from available studies that have methodological limitations.

All English-language studies on low-grade gliomas published between January 1970 and April 2000 were reviewed. Thirty studies that included statistical analyses were further evaluated with regard to the prognostic effect of extent of resection. Of these 30 studies, those that included pediatric patients, unless adults were analyzed separately, were excluded from further study because of the favorable outcome associated with the pediatric age group. Also excluded were studies including pilocytic and gemistocytic astrocytomas, because the natural histories of these histological subtypes are significantly different from that of low-grade gliomas. Series in which there were small numbers of patients (< 75) were also excluded. Results for oligodendrogliomas are reported separately.

Currently, for patients with low-grade glial tumors located in the cerebral hemisphere, the only management standard based on high-quality evidence is tissue diagnosis. All other treatment methods are practice options supported by evidence that is inconclusive or conflicting. The majority of published series that the authors identified had design-related limitations including a small study size, a small number of events (that is, deaths for survival studies), inclusion of pediatric patients, and/or inclusion of various histological types of tumors with different natural histories. Of the 30 series addressing the issue of timing and extent of surgery, almost all had additional design limitations. Methods used to determine the extent of resection were subjective and qualitative in almost all studies. Only five of the 30 series met the authors' criteria, and these studies are discussed in detail.

Conclusions. Management of low-grade gliomas is controversial and practice parameters are ill defined. This is caused by limited knowledge regarding the natural history of these tumors and the lack of high-quality evidence supporting various treatment options. Although a prospective randomized study seems unlikely, both retrospective matched studies and prospective observational trials will improve the clinician's ability to understand the importance of various prognostic factors.

Restricted access

Mark Bernstein

Restricted access

Edward R. Laws Jr.

Restricted access

J. Paul Elliott, G. Evren Keles, Michael Waite, Nancy Temkin, and Mitchel S. Berger

✓ The ventricular system is not infrequently entered during the course of maximum cytoreductive surgery for high-grade supratentorial gliomas. It is unclear if ventricular entry during surgery and/or proximity of the tumor to the ventricular system affects cerebrospinal fluid (CSF) tumor dissemination or the patients' overall survival rate. The authors retrospectively reviewed hospital records and neuroradiological studies of 51 patients operated on at the University of Washington between 1987 and 1991. Inclusion in this study necessitated a pathological diagnosis of malignant glioma and the availability of preoperative and postoperative computerized tomography scans or magnetic resonance images. Patients were excluded from the study if they had radiographic evidence of ventricular entry or CSF tumor dissemination prior to referral to the authors' institution. The index operation was defined as the first operation at the University of Washington or (in those patients with ventricular entry) the operation in which the ventricle was entered. Patients were followed until time of death or, in the case of survivors, until February, 1992. The effect of both ventricular entry and the proximity of the tumor to the ventricular system on CSF tumor dissemination and survival rate was assessed using statistical survival methodology.

There was no significant difference in time from diagnosis to the index operation between groups compared (Mann-Whitney U-test, p > 0.40). Cerebrospinal fluid dissemination was radiographically documented in 18 patients (35%) following the index operation. This occurrence was not significantly influenced by either ventricular entry during surgery (Mantel-Cox test, p = 0.13), the proximity of the tumor to the ventricular system (p = 0.63), or these two variables combined (p = 0.28). Survival rate following the index operation was not significantly affected by ventricular entry (p = 0.66), proximity of the tumor to the ventricular system (p = 0.61), or these two variable considered in combination (p = 0.44). However, survival rate was significantly decreased once CSF tumor dissemination had occurred (Cox model, p = 0.03).

Restricted access

G. Evren Keles, Kathleen R. Lamborn, Susan M. Chang, Michael D. Prados, and Mitchel S. Berger

Object. For patients with recurrent glioblastomas multiforme (GBMs) the prognosis is poor. Although chemotherapy may provide a survival advantage, the role of the extent of tumor resection, or the volume of the residual tumor at the time of recurrence, before instituting chemotherapy, is unclear. This study was designed to assess the response to chemotherapy based on the volume of residual disease (VRD) at the start of treatment in patients with recurrent GBMs. To accomplish this, the authors evaluated a homogeneous group of patients with recurrent GBMs who received the same chemotherapeutic agent.

Methods. One hundred nineteen adult patients with recurrent supratentorial GBMs received temozolomide chemotherapy at the time of tumor recurrence. In this cohort the authors analyzed the prognostic significance of volumetrically assessed tumor mass on time to tumor progression (TTP) and survival time (ST).

Multivariate analysis demonstrated that the VRD at the beginning of chemotherapy was a statistically significant predictor of both TTP (p < 0.0001) and ST (p < 0.006) when adjusted for the patient's age, performance score, and time from the initial diagnosis. Patients in whom the VRD was less than 10 cm3 at the start of chemotherapy had a 6-month progression-free survival rate of 32% compared with 8% for patients with a VRD between 10 and 15 cm3 and 3% for patients with a VRD larger than 15 cm3. Patients in whom the VRD was smaller than 10 cm3 had a 1-year survival rate of 37% compared with 9% for patients with a VRD between 10 and 15 cm3 and 18% for patients with a VRD larger than 15 cm3.

Conclusions. These data indicate that patients with recurrent GBMs who start chemotherapy with a smaller volume (< 10 cm3) of residual disease may have a more favorable response to chemotherapy and a more favorable outcome.

Restricted access

Duc H. Duong, Robert C. Rostomily, David R. Haynor, G. Evren Keles, and Mitchel S. Berger

✓ The authors describe a method for quantitation of the area and volume of the resection cavity in patients who have undergone surgery for brain tumors. Using a slide scanner and Image 1.27, a public domain program for the Apple Macintosh II computer, computerized tomography scans and magnetic resonance images can be digitized and analyzed for a particular region of interest, such as the area and volume of tumor on preoperative and postresection scans. Phantom scans were used to analyze the accuracy of the program and the program users. User error was estimated at 2%, program error was 4.5%. This methodology is proposed as a means of retrospectively calculating the extent of tumor resection.

Restricted access

G. Evren Keles, David A. Lundin, Kathleen R. Lamborn, Edward F. Chang, George Ojemann, and Mitchel S. Berger

Object. Intraoperative stimulation mapping of subcortical white matter tracts during the resection of gliomas has become a valuable surgical adjunct that is used to reduce morbidity associated with tumor removal. The purpose of this retrospective analysis was to assess the morbidity and functional outcome associated with this method, thus allowing the surgeon to predict the likelihood of causing a temporary or permanent motor deficit.

Methods. In this study, the authors report their experience with intraoperative stimulation mapping to locate subcortical motor pathways in 294 patients who underwent surgery for hemispheric gliomas within or adjacent to the rolandic cortex. Data were collected regarding intraoperative cortical and subcortical stimulation mapping results, along with the patient's neurological status pre- and postoperatively. For patients in whom an additional motor deficit occurred postoperatively, its evolution was examined.

Of 294 patients, an additional postoperative motor deficit occurred in 60 (20.4%). Of those 60, 23 (38%) recovered to their preoperative baseline status within the 1st postoperative week. Another 12 (20%) recovered from their postoperative motor deficit by the end of the 4th postoperative week, and 11 more recovered to their baseline status by the end of the 3rd postoperative month. Thus, 46 (76.7%) of 60 patients with postoperative motor deficits regained their baseline function within the first 90 days after surgery. The remaining 14 patients (4.8% of the entire study population of 294) had a persistent motor deficit after 3 months. Patients whose subcortical pathways were identified with stimulation mapping were more prone to develop an additional (temporary or permanent) motor deficit than those in whom subcortical pathways could not be identified (27.5% compared with 13.1%, p = 0.003). This was also true when additional (permanent) motor deficits lasted more than 3 months (7.4% when subcortical pathways were found, compared with 2.1% when they were not found; p = 0.041).

Conclusions. In patients with gliomas that are located within or adjacent to the rolandic cortex and, thus, the descending motor tracts, stimulation mapping of subcortical pathways enables the surgeon to identify these descending motor pathways during tumor removal and to achieve an acceptable rate of permanent morbidity in these high-risk functional areas.

Restricted access

Edward F. Chang, Matthew B. Potts, G. Evren Keles, Kathleen R. Lamborn, Susan M. Chang, Nicholas M. Barbaro, and Mitchel S. Berger


Seizures play an important role in the clinical presentation and postoperative quality of life of patients who undergo surgical resection of low-grade gliomas (LGGs). The aim of this study was to identify factors that influenced perioperative seizure characteristics and postoperative seizure control.


The authors performed a retrospective chart review of all cases involving adult patients who underwent initial surgery for LGGs at the University of California, San Francisco between 1997 and 2003.


Three hundred and thirty-two cases were included for analysis; 269 (81%) of the 332 patients presented with ≥ 1 seizures (generalized alone, 33%; complex partial alone, 16%; simple partial alone, 22%; and combination, 29%). Cortical location and oligodendroglioma and oligoastrocytoma subtypes were significantly more likely to be associated with seizures compared with deeper midline locations and astrocytoma, respectively (p = 0.017 and 0.001, respectively; multivariate analysis). Of the 269 patients with seizures, 132 (49%) had pharmacoresistant seizures before surgery. In these patients, seizures were more likely to be simple partial and to involve the temporal lobe, and the period from seizure onset to surgery was likely to have been longer (p = 0.0005, 0.0089, and 0.006, respectively; multivariate analysis). For the cohort of patients that presented with seizures, 12-month outcome after surgery (Engel class) was as follows: seizure free (I), 67%; rare seizures (II), 17%; meaningful seizure improvement (III), 8%; and no improvement or worsening (IV), 9%. Poor seizure control was more common in patients with longer seizure history (p < 0.001) and simple partial seizures (p = 0.004). With respect to treatment-related variables, seizure control was far more likely to be achieved after gross-total resection than after subtotal resection/biopsy alone (odds ratio 16, 95% confidence interval 2.2–124, p = 0.0064). Seizure recurrence after initial postoperative seizure control was associated with tumor progression (p = 0.001).


The majority of patients with LGG present with seizures; in approximately half of these patients, the seizures are pharmacoresistant before surgery. Postoperatively, > 90% of these patients are seizure free or have meaningful improvement. A shorter history of seizures and gross-total resection appear to be associated with a favorable prognosis for seizure control.

Restricted access

G. Evren Keles, Edward F. Chang, Kathleen R. Lamborn, Tarik Tihan, Chih-Ju Chang, Susan M. Chang, and Mitchel S. Berger


To investigate the prognostic significance of the volumetrically assessed extent of resection on time to tumor progression (TTP), overall survival (OS), and tumor recurrence patterns, the authors retrospectively analyzed preoperative and postoperative tumor volumes in 102 adult patients from the time of the initial resection of a hemispheric anaplastic astrocytoma (AA).


The quantification of tumor volumes was based on a previously described method involving computerized analysis of magnetic resonance (MR) images. Analysis of contrast-enhancing tumor volumes on T1-weighted MR images was conducted for 67 patients who had contrast-enhancing tumors. Measurements of T2 hyperintensity were obtained for all 102 patients in the study.

The presence or absence of preresection enhancement, actual volume of this enhancement, and the percentage of preoperative enhancement as it relates to the total T2 tumor volume did not have a statistically significant relationship to TTP or OS. In addition to age, the volume of residual disease measured on T2-weighted MR images was the most significant predictor of TTP (p < 0.001), and residual contrast-enhancing tumor volume was the most significant predictor of OS (p = 0.003) on multivariate analysis. In contrast to low-grade gliomas, there was no statistically significant relationship between the extent of resection and histological characteristics at the time of recurrence, that is, tumor Grade III compared with Grade IV.


Data from this retrospective analysis of a histologically uniform group of hemispheric AAs treated in the MR imaging era suggest that residual tumor volumes, as documented on postoperative imaging studies, may be a prognostic factor for TTP and OS for this patient population.