CT Perfusion in the Prognostication of Cerebral High Grade Glioma

High grade cerebral glioma is the most common primary brain tumor in adults and accounts for about 2.5% of all cancer deaths. Brain tumor affects approximately 2300 individuals per year in Canada. Noninvasive accurate and timely diagnosis is imperative. High grade glioma is an aggressive neoplasm with median survival of 12 months, irrespective of any treatment. The prognosis of these patients can only be decided based on pathology after biopsy or surgery. Conventional imaging techniques, such as routine magnetic resonance imaging(MRI), do not accurately predict the grade of malignancy of cerebral gliomas. Computed tomography(CT) perfusion allows us to study the blood supply to the tumor at the level of capillaries. This information permits determination of aggressiveness of cerebral gliomas at the time of diagnosis. In a preliminary study of 20 patients with high grade cerebral gliomas, we have shown that CT perfusion can predict survival at the time of diagnosis irrespective of the pathological grade and the treatment received. In the present study, we would like to extend our preliminary findings in larger group of patients to ensure that this technique is indeed robust. If our hypothesis was supported by our study, we will be able to subselect patients based on initial imaging for more aggressive treatment. In patients with shorter survival, the perfusion parameters may help in identifying new therapeutic targets (e.g., anti-angiogenic agents) that may help in the treatment of these patients.

Purpose of the research: Evaluate the utility of the non-invasive imaging tool of computed tomography (CT) perfusion in preoperative prognostication of high grade cerebral gliomas. Background information: Brain tumour affects approximately 2300 individuals per year in Canada. High grade glioma is the most common primary brain tumour in adults and accounts for approximately 2.5% of all cancer deaths. The real incidence of high grade glioma in the country is not known, but it affects approximately 2 per 100,000 people per year in Nova Scotia. Non-invasive, accurate and timely diagnosis is imperative. High grade glioma is an aggressive neoplasm, and the median survival of affected patients is 12 months, irrespective of any form of treatment. The prognosis of these patients can only be determined on pathology after biopsy or surgery. Conventional imaging techniques such as as magnetic resonance imaging(MRI) alone do not help in predicting the level of malignancy of cerebral gliomas. Formation of new blood vessels plays a central role in the growth and spread of tumours. The advanced imaging tool of computed tomography(CT) perfusion can be used to study blood supply to the tumor at the level of capillaries. This information permits determination of aggressiveness of cerebral gliomas at the time of diagnosis. To direct better and aggressive care in patients with high grade cerebral glioma, it is important to identify patients whose life expectancy could be relatively longer. The two high grades of gliomas have been grouped together in the literature for both treatment and prognosis. There are only a few studies that have addressed the differentiation of the two higher grades of gliomas. Some of these are our own preliminary studies. In a preliminary study of 20 such patients,investigators have shown that CT perfusion can predict survival, irrespective of the pathological grade and treatement received. In the present study, investigators would like to extend our preliminary findings in a larger group of patients to ensure that this technique is indeed robust. In patients with shorter survival, the perfusion parameters may help in identifying new therapeutic targets (e.g., anti-angiogenic agents) that may help in the treatment of these patients. Both CT and MR perfusion studies can be used to assess the blood supply of brain tumors. MR perfusion is limited in terms of absolute quantification of the perfusion parameters due to the non-linear relationship between the concentration of contrast material and the signal intensity change. MR perfusion studies also require two bolus injections of contrast agents. CT perfusion is a widely available and reliable technique which can give absolute quantification of perfusion parameters with a single acquisition and single bolus of a contrast agent. Hypothesis- CT perfusion in high grade cerebral gliomas can predict survival irrespective of the pathological grade and irrespective of the treatment given. Specific objectives: Conduct a prospective study in a larger cohort of patients with high grade cerebral gliomas to assess the utility of CT perfusion and compare this with demographic factors, tumour grade and treatment for estimating prognosis. Research design: This prospective cohort study will be done over a period of 2 years at QE II Health Sciences Centre, Infirmary Site. All patients, presenting with a newly detected brain lesion likely to be high grade cerebral glioma either on CT or MRI of the brain, will be asked to participate in this study. Pregnant patients or those with impaired renal function will be excluded from the study. Informed consent will be obtained. Approximately 100 new patients with cerebral glioma present to our brain tumour clinic every year. Our intent is to recruit at least 50 patients with high grade cerebral glioma each year for 2 consecutive years. Each patient will undergo a single CT perfusion scan before undergoing any biopsy or surgical resection. CT perfusion technique- The CT perfusion scan will be performed on a 128 slice multi-detector CT scanner (Somatom Definition Flash, Siemens). CT perfusion imaging will be performed at the time of routine preoperative neuronavigation imaging. A non-contrast CT head scan will be acquired to localize the tumour before obtaining a perfusion scan. Four 5 mm thick slices with matrix size of 512x512 will be selected at the level of the tumour. Multiple images will be acquired at each levels starting 5 sec after the injection of 50 cc of non-ionic iodinated contrast media (Omnipaque 300) at a rate of 4 cc/sec. The acquisition parameters will be 80 kVp and 100 mA. The images will be acquired every second for a total of 110 sec. The images will be sent to a separate post-processing work station and the post- processing will be performed using the CT perfusion software (Siemens). Two CT perfusion parameters (cerebral blood volume and permeability surface area product) will be calculated using multiple regions of interest in the areas of tumour showing the highest perfusion values. Another region of interest will be placed in the contralateral normal looking white matter. The patient will undergo biopsy or surgery and histopathological examination of the tissue as per standard of care in our hospital. The perfusion parameters will be correlated with the grade of glioma on pathology. The patient will receive adjuvant treatment as per the institutional standard of care and will be followed clinically as per standard of care. The perfusion parameters will be then correlated with the survival of these patients. Data analysis- Perfusion parameters will be compared between glioma and normal brain tissue and between different grades of gliomas using a Student 2-sample test. A Cox regression analysis with proportional hazards will be used to determine the relationship between overall survival and perfusion parameters. Kaplan-Meier survival curves will be drawn to demonstrate the relationship between perfusion parameters and overall survival in patients. A p value of less than 0.05 will be considered statistically significant. Expectations and Relevance of the proposed research CT perfusion will help in preoperative characterization of high grade cerebral gliomas and their level of aggressiveness. Perfusion parameters will help in predicting the prognosis(irrespective of grade of tumour and treatment given) of the patients with high grade cerebral glioma. This may help in targeting aggressive therapy to patients with predicted longer survival. In patients with predicted shorter survival, the perfusion parameters may direct future research on new therapeutic targets. Possible pitfalls and possible solutions: Radiation exposure from CT perfusion- These patients undergo radiotherapy as part of their treatment, which will have significantly higher radiation dose compared to CT perfusion. Radiation dose from CT perfusion is equivalent to that of 2 head CT scans. Potential kidney toxicity of iodinated dye- Only 50 cc of CT dye will be used, which is smaller compared to the regular dose for enhanced CT scan. Patients with compromised renal function will be excluded from our study.

The CT perfusion scan will be performed on a multi-detector CT scanner at the time of routine preoperative neuronavigation imaging. A non-contrast CT head scan will be acquired to localize the tumour. Four 5 mm thick slices will be selected at the level of the tumour. Multiple images will be acquired at each levels starting 5 sec after the injection of 50 cc of non-ionic iodinated contrast media at a rate of 4 cc/sec. The acquisition parameters will be 80 kVp and 100 mA. The images will be acquired every second for a total of 110 sec. The post- processing will be performed and parameters will be calculated using regions of interest in the areas of tumour showing the highest perfusion values. Another region of interest will be placed in the contralateral normal looking white matter.

Survival of patients with high grade cerebral glioma based on the CT perfusion parameters irrespective of the pathological grade and irrespective of the treatment given.

Inclusion Criteria: All patients, presenting with a newly detected brain lesion likely to be high grade cerebral glioma either on CT or MRI of the brain Exclusion Criteria: Pregnant patients those with impaired renal function

Shankar JJ, Woulfe J, Silva VD, Nguyen TB. Evaluation of perfusion CT in grading and prognostication of high-grade gliomas at diagnosis: a pilot study. AJR Am J Roentgenol. 2013 May;200(5):W504-9. doi: 10.2214/AJR.12.8967.