Elsevier

World Neurosurgery

Volume 97, January 2017, Pages 21-26
World Neurosurgery

Original Article
Permeability Surface Area Product Using Perfusion Computed Tomography Is a Valuable Prognostic Factor in Glioblastomas Treated with Radiotherapy Plus Concomitant and Adjuvant Temozolomide

https://doi.org/10.1016/j.wneu.2016.09.072Get rights and content

Objective

The current standard treatment protocol for patients with newly diagnosed glioblastoma (GBM) includes surgery, radiotherapy, and concomitant and adjuvant temozolomide (TMZ). We hypothesized that the permeability surface area product (PS) from a perfusion computed tomography (PCT) study is associated with sensitivity to TMZ. The aim of this study was to determine whether PS values were correlated with prognosis of GBM patients who received the standard treatment protocol.

Methods

This study included 36 patients with GBM that were newly diagnosed between October 2005 and September 2014 and who underwent preoperative PCT study and the standard treatment protocol. We measured the maximum value of relative cerebral blood volume (rCBVmax) and the maximum PS value (PSmax). We statistically examined the relationship between PSmax and prognosis using survival analysis, including other clinicopathologic factors (age, Karnofsky performance status [KPS], extent of resection, O6-methylguanine-DNA methyltransferase [MGMT] status, second-line use of bevacizumab, and rCBVmax).

Results

Log-rank tests revealed that age, KPS, MGMT status, and PSmax were significantly correlated with overall survival. Multivariate analysis using the Cox regression model showed that PSmax was the most significant prognostic factor. Receiver operating characteristic curve analysis showed that PSmax had the highest accuracy in differentiating longtime survivors (LTSs) (surviving more than 2 years) from non-LTSs. At a cutoff point of 8.26 mL/100 g/min, sensitivity and specificity were 90% and 70%, respectively.

Conclusions

PSmax from PCT study can help predict survival time in patients with GBM receiving the standard treatment protocol. Survival may be related to sensitivity to TMZ.

Introduction

Glioblastoma (GBM) is the most frequent and malignant glioma. The current standard treatment for patients with newly diagnosed GBM includes surgery, a 6-week cycle of external beam radiation therapy, and concomitant oral temozolomide (TMZ), followed by adjuvant TMZ.1 TMZ is currently the most important chemotherapy agent for controlling glioma progression. However, to date, improvement and overall success with TMZ administration remain limited. The European Organization for Research and Treatment of Cancer trial showed that the median survival in GBM patients treated with radiation plus TMZ is limited to 2.5 months longer than that in patients treated with radiation alone.1 At present, the O6-methylguanine-DNA methyltransferase (MGMT) status is considered a powerful predictor of the response to TMZ in patients with GBM.2 The search for prognostic markers, especially in vivo imaging biomarkers, continues with significant improvements in the resolution of clinically available imaging tools. One obstacle to effective chemotherapy is believed to be the heterogeneous permeability of GBMs.3

The vascularity and permeability of gliomas can be assessed by a perfusion computed tomography (PCT) study in which 2 parameters, cerebral blood volume (CBV) and permeability surface area product (PS), are calculated.3, 4, 5, 6 CBV is defined as the volume of blood flowing in the blood vessels per 100 g of brain tissue. PS characterizes diffusion of a contrast agent from vessels into the interstitial space because of a defective blood–brain barrier (BBB). Perfusion magnetic resonance imaging is currently the most commonly used perfusion imaging method for assessing gliomas.6, 7, 8, 9 However, the linear relationship between signal intensity and iodine concentration is a major advantage of PCT compared with perfusion magnetic resonance imaging. Additionally, perfusion magnetic resonance imaging is limited in absolute quantification of perfusion parameters.10 Therefore, PCT is a reliable technique that can be used for absolute quantification of CBV and PS.

To date, only 2 articles have reported the correlation between CBV and PS using PCT and prognosis in patients with high-grade gliomas (grades III and IV).11, 12 However, we consider that grades III and IV gliomas should not be assessed together because grade III gliomas include oligodendroglial tumors, which have a more favorable prognosis compared with astrocytic tumors.13, 14 Additionally, in these 2 articles, the treatments that patients received were not standardized. Therefore, in the current study, we selected only GBM patients treated with the standard treatment protocol of radiotherapy plus concomitant and adjuvant TMZ. We speculated that PS is associated with drug delivery, is an important factor for assessing chemosensitivity to TMZ, and is correlated with the prognosis of patients with GBM. Preoperative prediction of sensitivity to TMZ and overall survival (OS) time by using noninvasive imaging techniques is essential for improving the clinical management of GBMs, such as the surgical strategy. The aim of this study was to determine whether PS values from a preoperative PCT study are correlated with prognosis of GBM patients who received the standard treatment protocol of radiotherapy and TMZ administration.

This study included 36 patients with supratentorial GBM who were newly diagnosed and underwent resection between October 2005 and September 2014 and who underwent a preoperative PCT study at Hiroshima University Hospital. Histopathologic assessment was carried out according to the criteria of the World Health Organization classification. The study population comprised 18 men and 18 women who ranged in age from 31 to 78 years (mean age, 60 years). All patients underwent PCT 1–4 days before surgery. After surgery, all patients received the standard treatment protocol of radiotherapy plus concomitant and adjuvant TMZ.1 We excluded GBM patients who did not receive the standard treatment protocol or a preoperative PCT study.

This retrospective study protocol was approved by the institutional review board of Hiroshima University. Because of the study's retrospective nature, institutional review board waived the requirement for informed consent for this study. To protect patient privacy, we removed all identifiers from our records on completion of our analyses.

PCT studies were performed on a 16-slice (LightSpeed Ultra [GE Medical Systems, Milwaukee, Wisconsin, USA]) multidetector row computed tomography scanner. A low radiation dose noncontrast computed tomography head study was done to localize the region of interest (ROI) before obtaining a perfusion scan. For the perfusion scan, 40 mL of nonionic contrast (Omnipaque [Iohexol,N,N′ - Bis(2,3-dihydroxypropyl)-5-{N-(2,3-dihydroxypropyl)-acetamido}-2,4,6-triiodo-isophthalamide] 350 mg/mL) was injected at a rate of 5 mL/s through a 20-gauge line using an automatic injector. At 5 seconds into the injection, a cine (continuous) scan was initiated with the following parameters: 80 kVp, 80 mA, and 1 second per rotation. After the initial 5-second cine scan, 8 more axial images were acquired, 1 image every 2 seconds for an additional 52 seconds, therefore giving a total acquisition time of 57 seconds. Eight 5-mm-thick axial slices were acquired, resulting in a total coverage area of 4 cm. For all patients, perfusion maps of CBV and PS were generated with an Advantage Windows workstation using computed tomography perfusion 3.1 software (GE Medical Systems). We used the superior sagittal sinus as the venous output function in all patients and the artery with the greatest peak and slope on time-attenuation curves as the arterial input function. An ROI was drawn within the confines of a large vessel, and the automatic function of the software selected the pixels with greatest peak and slope on the time-attenuation curve for analysis.

Two of the authors (T. Saito and M. Ishifuro) placed multiple ROIs in areas of the tumor with the highest CBV and PS values, arriving at a consensus. Six more ROIs on the CBV map were positioned in the contralateral normal white matter. ROIs were placed taking care not to include necrotic/cystic areas and avoiding any major cortical vessels. The maximum value of PS (PSmax) was determined. The maximum CBV values of all intratumoral ROIs and the mean CBV of the contralateral ROIs were calculated. The maximum value of relative CBV (rCBVmax) values were then expressed as a ratio of the maximum CBV value within the tumor to the mean CBV value in the contralateral normal-appearing white matter. This approach has been demonstrated to provide the best inter- and intraobserver reproducibility.15

The avidin-biotin immunoperoxidase or Simple Stain MAX-peroxidase (Nichirei, Tokyo, Japan) technique was used to perform MGMT monoclonal antibody (DAKO, Glostrup, Denmark) assays according to the manufacturers' instructions in selected sections from each case. In accordance with a previous report,16 we determined the percentages of MGMT-positive tumor cells in individual cases by continuously counting over 400 nuclei in high-magnification views. For individual cases, when the proportion of labeled nuclei/all tumor nuclei was 30% or more, the case was regarded as positive, whereas a case was considered negative when the proportion was less than 30%.

Data were analyzed using the SPSS version 16.0 software package (SPSS Inc., Chicago, Illinois, USA). Correlations between OS and age (<60 vs. ≥60 years), Karnofsky performance status (KPS) (<80 vs. ≥80), extent of resection (total, partial, and biopsy), MGMT status, use of second-line bevacizumab, rCBVmax (mean, 8.89 mL/100 g; <8.89 vs. ≥8.89 mL/100 g), and PSmax (mean, 8.02 mL/100 g/min; <8.02 vs. ≥8.02 mL/100 g/min) were calculated using the log-rank test. Multivariate survival analysis was performed using the Cox proportional hazards regression model, which included the following factors: age, KPS, extent of resection, MGMT status, and PSmax. A P value <0.05 indicated a statistically significant difference in all analyses.

Additionally, we defined a patient who survived more than 2 years after the initial surgery as a longtime survivor (LTS). For evaluation of the diagnostic efficacy of the PSmax value in discrimination of whether a patient was an LTS or not, receiver operating characteristic curve analysis was done with calculation of the area under the curve. For this purpose, the Youden index was used to calculate the optimal cutoff value of PSmax.17

Section snippets

Results

Of the 36 patients with GBM who received the standard treatment protocol of radiotherapy and TMZ administration, 11 patients underwent bevacizumab treatment for recurrence or progression of the lesion. Log-rank tests revealed that age (P = 0.0418), KPS (P = 0.0011), MGMT status (P = 0.0265), and PSmax (P = 0.0003) were significantly correlated with OS, whereas the extent of resection, the use of second-line bevacizumab, and rCBVmax were not (Table 1 and Figure 1A). The differences in OS time

Discussion

We have demonstrated with a PCT study that the PSmax value was significantly correlated with OS in patients with GBM who were treated with the standard treatment protocol of radiotherapy plus concomitant and adjuvant TMZ. PSmax remained the most significant predictor of OS, even after adjusting for important prognostic factors (age, KPS, extent of resection, and MGMT status), indicating that the PSmax value could improve prediction of OS. Additionally, the threshold PSmax value of 8.26 mL/100

Conclusions

We have demonstrated that PSmax from a PCT study can help predict survival time in patients with GBM who were treated with the standard protocol of radiotherapy plus concomitant and adjuvant TMZ. This result may be related to sensitivity to TMZ. Patients with a low PSmax showed worse OS than patients with a high PSmax. Additionally, the threshold PSmax value of 8.26 mL/100 g/min may be a useful aid to differentiate LTSs from non-LTSs. Our data add new information regarding the correlation

Acknowledgments

The authors thank Masako Yoshihiro, Department of Neurosurgery, Hiroshima University, Graduate School of Biomedical and Health Science, Hiroshima, Japan, for the immunohistochemical staining.

References (24)

  • T.P. Yeung et al.

    Dynamic perfusion CT in brain tumors

    Eur J Radiol

    (2015)
  • R. Jain et al.

    Prognostic vascular imaging biomarkers in high-grade gliomas: tumor permeability as an adjunct to blood volume estimates

    Acad Radiol

    (2013)
  • H.H. Engelhard et al.

    Oligodendroglioma and anaplastic oligodendroglioma: clinical features, treatment, and prognosis

    Surg Neurol

    (2003)
  • R. Stupp et al.

    Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma

    N Engl J Med

    (2005)
  • M.E. Hegi et al.

    MGMT gene silencing and benefit from temozolomide in glioblastoma

    N Engl J Med

    (2005)
  • R. Jain et al.

    In vivo correlation of tumor blood volume and permeability with histologic and molecular angiogenic markers in gliomas

    AJNR Am J Neuroradiol

    (2011)
  • B. Ding et al.

    Comparison of cerebral blood volume and permeability in preoperative grading of intracranial glioma using CT perfusion imaging

    Neuroradiology

    (2006)
  • A.M. Di Nallo et al.

    Quantitative analysis of CT-perfusion parameters in the evaluation of brain gliomas and metastases

    J Exp Clin Cancer Res

    (2009)
  • R. Jain et al.

    Glioma angiogenesis and perfusion imaging: understanding the relationship between tumor blood volume and leakiness with increasing glioma grade

    AJNR Am J Neuroradiol

    (2015)
  • P. Kickingereder et al.

    Evaluation of microvascular permeability with dynamic contrast-enhanced MRI for the differentiation of primary CNS lymphoma and glioblastoma: radiologic-pathologic correlation

    AJNR Am J Neuroradiol

    (2014)
  • K. Deike et al.

    Prognostic value of combined visualization of MR diffusion and perfusion maps in glioblastoma

    J Neurooncol

    (2016)
  • A.K. Bag et al.

    Survival analysis in patients with newly diagnosed primary glioblastoma multiforme using pre- and post-treatment peritumoral perfusion imaging parameters

    J Neurooncol

    (2014)
  • Cited by (6)

    Conflict of interest statement: This work was supported by the Japan Society for the Promotion of Science KAKENHI (grant number 26162183).

    View full text