A phase II study of temozolomide vs. procarbazine in patients with glioblastoma multiforme at first relapse.

A randomized, multicentre, open-label, phase II study compared temozolomide (TMZ), an oral second-generation alkylating agent, and procarbazine (PCB) in 225 patients with glioblastoma multiforme at first relapse. Primary objectives were to determine progression-free survival (PFS) at 6 months and safety for TMZ and PCB in adult patients who failed conventional treatment. Secondary objectives were to assess overall survival and health-related quality of life (HRQL). TMZ was given orally at 200 mg/m(2)/day or 150 mg/m(2)/day (prior chemotherapy) for 5 days, repeated every 28 days. PCB was given orally at 150 mg/m(2)/day or 125 mg/m(2)/day (prior chemotherapy) for 28 days, repeated every 56 days. HRQL was assessed using the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-C30 [+3]) and the Brain Cancer Module 20 (BCM20). The 6-month PFS rate for patients who received TMZ was 21%, which met the protocol objective. The 6-month PFS rate for those who received PCB was 8% (P = 0.008, for the comparison). Overall PFS significantly improved with TMZ, with a median PFS of 12.4 weeks in the TMZ group and 8.32 weeks in the PCB group (P = 0.0063). The 6-month overall survival rate for TMZ patients was 60% vs. 44% for PCB patients (P = 0.019). Freedom from disease progression was associated with maintenance of HRQL, regardless of treatment received. TMZ had an acceptable safety profile; most adverse events were mild or moderate in severity.

Permanent iodine-125 interstitial implants for the treatment of recurrent glioblastoma multiforme.

OBJECTIVE: Brachytherapy with temporary implants may prolong survival in patients with recurrent glioblastoma multiforme (GBM), but it is associated with relatively high costs and morbidity. This study reports the time to progression and survival after permanent implantation of iodine-125 seeds for recurrent GBM and examines factors predictive of outcome. METHODS: Forty patients with recurrent GBM were treated with maximal resection plus permanent placement of iodine-125 seeds into the tumor bed. A total dose of 120 to 160 Gy was administered, and patients were followed up with magnetic resonance imaging scans every 2 to 3 months. RESULTS: Actuarial survival from the time of implantation was 47 weeks, with 7 of 40 patients still alive at a median of 59 weeks after implantation. Survival was significantly better for patients younger than 60 years, and a trend for longer survival was demonstrated with gross total resection and tumors with a low MIB-1 (a nuclear antigen present in all cell cycles of proliferating cells) staining index. Median time to progression was 25 weeks and, on multivariate analysis, was favorably influenced by gross total resection and patient age younger than 60 years. After implantation, 27 of 30 patients with failure had a local component to the failure. No patient developed symptoms attributable to radiation necrosis or injury. CONCLUSION: Permanent iodine-125 implants for recurrent GBM result in survival comparable with that described in previous reports on temporary implants, but with less morbidity. Results are most favorable for patients who are younger than 60 years, and who undergo gross total resection. Despite this aggressive treatment, most patients die as a consequence of locally recurrent disease.

Preoperative anaplastic glioma tumor volume effects on patient survival.

BACKGROUND AND OBJECTIVES: The relationship between preoperative tumor volume and patient survival has long been studied, but the results have been inconsistent. Since geometric measurement of tumor volume was used in these studies, the aim of this study was to ascertain whether the inconsistency of the study results is due to less accurate geometric measurement. METHODS: Prognostic tumor volume effects were compared between the planimetry method and the geometric method using survival analysis, performed for 99 patients diagnosed with anaplastic glioma tumor. RESULTS: A significant correlation was found between planimetry tumor volume and patient survival, but there was no correlation between geometric tumor volume and patient survival. The larger planimetry tumor volume was significantly associated with shorter survival. CONCLUSIONS: The study indicated that in brain tumor research the preoperative tumor volume measured by the geometric method may not be prognostically important. The more accurate measurement, i.e., the planimetry method (based on either computed tomography or magnetic resonance imaging), is needed in brain tumor clinical research and prognostic diagnosis.

Microcomputer-based technique for 3-D reconstruction and volume measurement of computed tomographic images. Comparison of geometric and planimetry post-operative tumor volume effects on patient survival.

The relationship between brain tumor size and survival has been studied since the advent of CT and magnetic resonance imaging (MRI), however, all published studies are based on tumor sizes measured by the less accurate geometric method, with results that are inconsistent. For example, the influence of the extent of tumor resection on patient survival has been debated. Jelsma and Bucy advocated extensive resection based on their analysis of patients with glioblastoma multiforme, however, Green et al. produced conflicting results, showing that the extent of resection was not statistically significant when all variables were considered simultaneously. The present study investigates whether or not the study inconsistency is largely due to the use of the less accurate geometric tumor volume measurement. The study demonstrates that the geometric tumor volume and the planimetry tumor volume have significantly different prognostic effects.

Stereotactic radiosurgery for the treatment of brain metastases. Results of a single institution series.

BACKGROUND: Stereotactic radiosurgery is being used with increasing frequency for the treatment of brain metastases. Optimal patient selection and treatment factors continue to be defined. This study provides outcome data from a single institutional experience with radiosurgery and identifies parameters that may be useful for the proper selection and treatment of patients. METHODS: Eighty-four patients underwent stereotactic radiosurgery for brain metastases between September 1989 and November 1995. Seventy-nine patients (93%) were treated at recurrence after previous whole brain radiotherapy. Patients had between 1 and 6 lesions treated with a median minimum tumor dose of 1600 centigrays (cGy). Thirty-eight patients (45%) had active extracranial disease at the time of radiosurgery. RESULTS: Median survival for the entire group was 43 weeks from the date of radiosurgery and 71 weeks from the original diagnosis of brain metastases. Patients with 1 or 2 metastases had significantly improved survival compared with patients with > or = 3 metastases (P = 0.02), and patients without active extracranial tumor survived longer than those with extracranial disease (P = 0.03). Median time to failure for 145 evaluable lesions was 35 weeks. Local control was significantly improved for radiosurgery doses of > 1800 cGy, and for melanoma histology. CONCLUSIONS: These results are comparable to reports of patients treated with resection and significantly superior to results observed after whole brain radiotherapy. The authors conclude that stereotactic radiosurgery is an effective, low risk treatment for extending the survival of patients with recurrent brain metastasis. Although survival is best for patients with < or = two lesions and no active extracranial disease, selected patients with > two lesions or active extracranial tumor may benefit as well.

Cholesteryl ester transfer protein in human brain.

The evidence that apolipoproteins are found in the cerebrospinal fluid and low-density lipoprotein receptor is found in the brain suggests that the brain may have an active lipid transport system. In plasma, cholesteryl ester transfer protein mediates the exchange and net transfer of cholesteryl ester and triglycerides among lipoproteins. Cholesteryl ester transfer activity was measured in the cerebrospinal fluid and plasma of ten neurologically normal subjects. Cholesteryl ester transfer activity was readily detectable in cerebrospinal fluid (7.4 +/- 13% cholesteryl ester was transferred per 20 microliters), and this activity was completely abolished with specific antibody against the plasma cholesteryl ester transfer protein. The concentration of cholesteryl ester transfer activity in the cerebrospinal fluid was about 12% of that found in plasma, whereas the concentration of albumin in cerebrospinal fluid was only about 0.6% of that in plasma, suggesting direct synthesis of cholesteryl ester transfer protein within the brain. Cholesteryl ester transfer activity was found in conditioned medium from human neuroblastoma and neuroglioma cells and sheep choroid plexus. The data suggest that cholesteryl ester transfer protein is synthesized and secreted in the brain. This protein could play an important role in the transport and redistribution of lipids within the central nervous system.

Issues in cerebrospinal fluid management. CSF Venereal Disease Research Laboratory testing.

Three policies for decreasing unnecessary cerebrospinal fluid (CSF) management Venereal Disease Research Laboratory (VDRL) tests were compared. The first policy attempted to educate physicians about the use of serologic tests for diagnosing neurosyphilis but allowed the CSF VDRL to be performed either as a screening test or as a retrospective test. The second policy required that the CSF VDRL be performed as a retrospective test without regard to the patient's serologic status. The third policy required that a patient be seropositive by either rapid plasma reagin (RPR) or fluorescent treponemal antibody absorbance (FTA-ABS) before a CSF VDRL could be performed. Before these policies were instituted, VDRL testing was performed on 18.2% of all CSF samples. The optional and required retrospective policies decreased the CSF VDRL rate to 13.0% and 8.5%, respectively, but the percentages of seropositive patients for whom these procedures were performed were only 7.3% and 12.9%. The third policy decreased the CSF VDRL test rate to 1.8% (P less than 0.001) with seropositivity improving to 90%. To assure serologic tests are obtained in the evaluation of neurosyphilis, requirement for seropositivity must be implemented with the use of retrospective CSF VDRL testing.

Issues in cerebrospinal fluid management. Acid-fast bacillus smear and culture.

Meningeal tuberculosis is an uncommon disease in the United States with an annual incidence of fewer than 200 cases. This study evaluates three approaches to improving the use of the cerebrospinal (CSF) acid-fast bacillus (AFB) smear and culture procedure: (1) education alone; (2) optional screening by which physicians can select to have the AFB analysis stopped if the initial CSF findings are unremarkable; and (3) mandatory screening before the performance of all CSF AFB analyses. With education alone, the CSF AFB culture rate decreased from 20.6% of all CSF acquisitions to 15.7% (P less than 0.001); however, the effect may have been related to a decrease in all types of AFB testing. Optional screening had no impact on the AFB testing rate. Mandatory screening significantly decreased the CSF AFB rate to 6.7% (P less than 0.001), unrelated to changes in other types of AFB testing. Laboratories that employ mandatory screening should report the screening results immediately and have a mechanism whereby physicians can bypass the screen, providing CSF AFB analysis on unremarkable fluid from high-risk patients.

Amplification of HIV-1 provirus from cerebrospinal fluid and its correlation with neurologic disease.

The polymerase chain reaction (PCR) was used to detect human immunodeficiency virus type 1 (HIV-1) proviral sequences (gag and env) in nucleated cells from the cerebrospinal fluid (CSF) of 31 HIV-1-positive patients, and the results were compared with clinical and radiologic evidence of neurologic disease. Provirus was detected in 21 patients, of whom 20 had neurologic abnormalities. Provirus was not detected in another 6, all of whom were neurologically normal. No neurologic disease has developed in 4 of these 6 patients for whom 12.8 months of follow-up is available. PCR of CSF nucleated cells from HIV-positive patients provides early, rapid, direct evidence of neurologic involvement.

Treatment of patients with recurrent gliomas with cyclophosphamide and vincristine.

Seventeen patients with recurrent gliomas were treated with the combination of cyclophosphamide and vincristine. All but one had previously received and failed chemotherapy. Cyclophosphamide was administered at doses ranging from 250 to 1000 mg/sq m by intravenous infusion on Days 1 and 2, and vincristine was given at a dose of 1.0 mg/sq m (2 mg maximal dose) intravenously on Day 1; cycles were repeated every 4 weeks. Clinical and radiographic improvement was observed in eight of 16 evaluable patients, and four other patients had stabilization of previously progressive disease. Four patients are alive and off treatment without evidence of recurrence for a median period of 37 months; these included an adult with a cerebral anaplastic astrocytoma now more than 51 months after therapy. Toxicity included moderately severe myelosuppression that required hospitalization in seven patients. These results indicate that the combination of cyclophosphamide and vincristine has activity in the treatment of recurrent gliomas, and warrant the use of these drugs in larger controlled studies, particularly if they can be used in conjunction with hematopoietic growth factors.

Management of cerebrospinal fluid.

In traditional practice, doctors order tests and the laboratory performs them. A series of tests are requested before the results of any of the tests are known. The authors of this paper examine the interface between the clinicians and the laboratory. They produce a cost effective and clinically useful routine for handling spinal fluid. They bring to their practice an excellent example of quality assurance which is genuine, improves practice and is not "busy work." The editor heard Dr. Albright present this material and urged him to make it available to North Carolina doctors.

CSF TRAP: a procedure to improve laboratory testing with myelography.

CSF TRAP (Transport and Rapid Accessioning for Additional Procedures) is a procedure that provides storage of and rapid access to cerebrospinal fluid (CSF) specimens and allows clinicians to review initial findings before ordering low-yield CSF studies. The cost-effectiveness of routinely using the CSF TRAP procedure with myelography is examined in a study group of 819 patients, 74% with disc diseases, spinal stenosis, spondylolisthesis, or pain syndromes, 10% with cancer, and 16% with neuropathies and miscellaneous conditions. Routine studies on CSF obtained during myelography provided little additional clinical information, except for patients with cancer (of 80 patients with cancer, the results of cytological examination of the CSF were positive in 12) and patients with multiple sclerosis, for whom oligoclonal band and IgG analysis provided supportive diagnostic data. The utilization of the CSF TRAP procedure with elimination of unnecessary culture and cytological studies on patients with disc diseases, spinal stenosis, spondylolisthesis, and pain syndromes, reduces myelographic CSF procedures by 20%, for a savings exceeding $14,000. The CSF TRAP procedure allows for a more cost-efficient analysis of CSF obtained using myelography, while providing fluid for analysis in patients with unexpected findings.

Microcomputer-based technique for 3-D reconstruction and volume measurement of computer tomographic images. Part 1: Phantom studies.

This paper presents a microcomputer-based technique that accurately quantifies volumes from computed tomographic (CT) scans of irregularly shaped objects as well as displaying 3-D reconstructions. The method uses standard CT film, allowing analysis of previous or outside CT studies. The planimetry method showed less than 5% error in measuring irregular 2-D areas larger than 6 mm2. The method is demonstrated to be significantly more accurate than spherical, ellipsoid, or rectangular geometric models in quantifying object volume by CT (P less than .001). With a single gantry angle, planimetry showed a two standard deviation error under 10% in measuring the volume of irregular objects compared with an error over 30% for ellipsoid models. The inaccuracy of the spherical model (80% error) and the rectangular prism model (192% error) renders them impractical to provide quantitative object volume. Microcomputer planimetry provides an accurate and versatile means to measure the volume and produce 3-D reconstructions of objects scanned with CT, and it has potential application in quantifying tumor response with CT and magnetic resonance imaging.

Cerebrospinal fluid (CSF) TRAP. A method to improve CSF laboratory efficiency.

Establishment of a procedure termed cerebrospinal fluid (CSF) TRAP ("Transport and Rapid Accessioning for Additional Procedures") allows clinicians to appropriately store, at -75 degrees C, and rapidly access CSF specimens. The CSF TRAP enhances patient care by decreasing the need for repeat lumbar punctures and providing reserve fluid for the following: (1) further CSF testing; (2) repeating questionable test results; and (3) laboratory accidents. The CSF TRAP has been approved for third-party payment because it promotes efficient laboratory utilization by encouraging clinicians to review initial CSF findings before ordering low-yield CSF assays such as the venereal disease research laboratory (VDRL) and cryptococcal antigen latex agglutination tests. Currently, CSF TRAP samples are being obtained with 40% of all CSF acquisitions at the Duke University Medical Center. The availability of the CSF TRAP was associated with a significant decrease in the ordering of CSF VDRL and cryptococcal antigen assays (P less than 0.05); however, there was no significant change in the proportion of those studies being performed on normal CSF. The CSF TRAP procedure provides the framework for an overall restructuring of CSF testing that is being investigated.

Microcomputer-based technique for 3-D reconstruction and volume measurement of computed tomographic images. Part 2: Anaplastic primary brain tumors.

Serial computed tomography (CT) plays an integral part in monitoring effects of therapy for primary anaplastic brain tumors. Despite advances in CT technology, clinicians often cannot obtain accurate quantitative volume information to complement the qualitative assessment of tumor change. This paper presents a microcomputer-based method that provides both quantitative volume measurements and 3-D reconstructions of primary anaplastic brain tumors based on their hard copy CT or magnetic resonance imaging studies. The findings of this study demonstrate that planimetry is feasible for routine clinical use and is superior in accuracy to the spherical geometric model, which is shown to significantly overestimate tumor volume. The findings of 62 quantitative tumor studies (17 patients) showed a direct relationship between the total tumor volume and the volume of the hypodense intratumor core. There was no evidence of a relationship between the total tumor volume and the amount of peritumor low density (edema).

Comparison of [125I]HIPDm and [125I]iodoantipyrine in quantifying regional cerebral blood flow in rats.

We determined regional cerebral blood flow (rCBF) using [125I]HIPDm [N,N,N'-trimethyl-N'-(2-hydroxy-3-methyl-5-iodobenzyl)-1,3-propanediamin e] and [125I]iodoantipyrine autoradiography under control and pathologic conditions (hypercapnia [acidosis], hypocapnia [alkalosis], and disrupted blood-brain barrier) conditions in 35 rats. In control rats, HIPDm rCBF (indicator fractionation method, n = 5) was lower than the corresponding IAP rCBF (diffusible indicator method, n = 4), most notably in the infratentorial regions and subcortical nuclei. In hypercapnia, rCBF increased by 100% and 37% in the HIPDm (n = 5) and IAP (n = 5) groups, respectively. In hypocapnia, IAP rCBF (n = 4) decreased 34% but HIPDm rCBF (n = 4) did not change. Following disruption of the blood-brain barrier by intracarotid infusion of mannitol in eight rats, both radiotracers (HIPDm n = 4, IAP n = 4) showed decreased rCBF to regions of disruption as defined by trypan blue extravasation. Our work indicates that modeling HIPDm uptake to quantify rCBF using the indicator fractionation method will underestimate blood flow and that HIPDm kinetics are influenced by compartmental pH dynamics that will limit the accuracy of this method in quantifying rCBF in pathologic conditions.

Establishment of a CSF bank.

A bank of well-characterized CSF has been established by collecting and storing (-70 degrees C) CSF samples remaining after completion of routine clinical studies. Over 1,700 individual patient samples were collected during a 12-month period. A database derived largely from information down-loaded from existing hospital-based systems includes the results of individual CSF laboratory studies, in addition to the patient age, primary diagnoses, and details of any malignancy. CSF control material is used to verify storage conditions. The CSF bank supplies investigators with CSF handled in a standardized manner for more precise investigation of CNS disease.