- Email: [email protected]
11C-CHO PET in optimization of target volume delineation and treatment regimens in postoperative radiotherapy for brain gliomas
Available online at www.sciencedirect.com
Nuclear Medicine and Biology 39 (2012) 437 – 442 www.elsevier.com/locate/nucmedbio
11
C-CHO PET in optimization of target volume delineation and treatment regimens in postoperative radiotherapy for brain gliomas☆
Fang-Ming Li a,⁎, Qing Nie a , Rui-Min Wang c , Susan M. Chang b , Wen-Rui Zhao a , Qi Zhu a , Ying-Kui Liang a , Ping Yang a , Jun Zhang a , Hai-Wei Jia a , Heng-Hu Fang a a
Department of Nuclear Medicine, Center of Radiation Oncology, People's Liberation Army's Navy General Hospital, Beijing 10048, China b Department of Neurological Surgery, University of California, San Francisco, CA, USA c PET/CT Center, Department of Nuclear Medicine, People's Liberation Army's Hospital No. 301, Beijing 100538, China Received 31 May 2011; received in revised form 2 October 2011; accepted 4 October 2011
Abstract Objective: We explored the clinical values of 11C-choline ( 11C-CHO) PET in optimization of target volume delineation and treatment regimens in postoperative radiotherapy for brain gliomas. Methods: Sixteen patients with the pathological confirmation of the diagnosis of gliomas prior to receiving radiotherapy (postoperative) were included, and on whom both MRI and CHO PET scans were performed at the same position for comparison of residual tumors with the two techniques. 11C-CHO was used as the tracer in the PET scan. A plain T1-weighted, T2-weighted and contrast-enhanced T1-weighted imaging scans were performed in the MRI scan sequence. The gliomas' residual tumor volume was defined as the area with CHO-PET high-affinity uptake and metabolism (VCHO) and one with MRI T1-weighted imaging high signal intensity (VGd), and was determined by a group of experienced professionals and clinicians. Results: (1) In CHO-PET images, the tumor target volume, i.e., the highly metabolic area with a high concentration of isotopes (SUV 1.016– 4.21) and the corresponding contralateral normal brain tissues (SUV0.1–0.62), was well contrasted, and the boundary between lesions and surrounding normal brain tissues was better defined compared with MRI and 18F-FDG PET images. (2) For patients with brain gliomas of WHO Grade II, the SUV was 1.016–2.5; for those with WHO Grades III and IV, SUVs were N26–4.2. (3) Both CHO PET and MRI were positive for 10 patients and negative for 2 patients. The residual tumor consistency between these two studies was 75%. Four of the 10 CHO-PET-positive patients were negative on MRI scans. The maximum distance between VGd and VCHO margins was 1.8 cm. (4) The gross tumor volumes (GTVs) and the ensuing treatment regimens were changed for 31.3% (5/16) of patients based on the CHO-PET high-affinity uptake and metabolism, in which the change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4) for patients with WHO Grade II III, and IV gliomas, respectively. Conclusion: Our data demonstrate that difference exists between CHO PET and MRI by which to judge and identify residual tumor for patients with brain gliomas. CHO PET is considered to be a supplementary diagnostic approach for MRI. Biological tumor target volume (BTV) displayed in the CHO PET images is useful in determining or delineating the radiotherapy target volume and making decisions in selecting treatment regimens. Tumor target volume may be defined more accurately and rationally when the CHO PET is combined with MRI. © 2012 Elsevier Inc. All rights reserved. Keywords: Brain glioma; Radioactive carbon-labeled choline; PET imaging; MRI; Radiotherapy target volume; Radiotherapy regimen
1. Introduction Revealing the gross tumor volume (GTV) of brain gliomas accurately and truly with imaging technology is a ☆
Conflict of interest and disclosure statement: The authors declare no competing financial interests. ⁎ Corresponding author. Tel.: +86 010 66958128, 66958478; fax: +86 010 68780892. E-mail address: [email protected] (F.-M. Li). 0969-8051/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.nucmedbio.2011.10.003
prerequisite and a key element in achieving treatment success. Prior to receiving radiotherapy, the tumor volumes for patients with gliomas have to be determined in order for clinicians to delineate the target volume. Tumors cannot be evaluated with conventional morphology imaging technology. PET imaging provides an unprecedented opportunity for noninvasive diagnosis of brain gliomas and is an essential way in undergoing navigated surgery, targeted radiotherapy and treatment outcome evaluation of chemotherapy [1–3]. PET makes it possible to delineate the biological target
438
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
Fig. 1. MRI, 18F-FDG PET and 11C-CHO PET/CT images of a high-grade glioma (female, age 70, anaplastic glioma, WHO Grade III). Comparison of MRI (VGd) and 18F-FDG PET (VFDG) , 11C-CHO PET/CT ( SUV 4.2) images identified target volumes and the exact boundary of the tumor and their relation to the surrounding edema and showed clearly the boundaries of the tumors (A and B: MRI T1-Gd and MRI T2; C: 11C-CHO PET/CT; D: 18F-FDG PET/CT).
volume (BTV) of gliomas. In this study, 11C-CHO PET function imaging and MRI morphology imaging results were compared and contrasted so that the radiotherapy target volume may be defined more accurately and treatment regimens optimized [4,5]. 2. Materials and methods 2.1. Patients During the period of January 2008 to June 2009, 16 patients with the pathological confirmation of the diagnosis of gliomas prior to receiving radiotherapy (postoperative) were included in this study. Among the patients were eight males and eight females with a median age of 45 years (range: 8–70). Sites of the lesions were as follows: in one case the lesion was in the suprasellar region; in the other 15 cases, the tumors were located in the frontal lobe, temporal lobe and parietal– occipital lobe area. The maximum diameter for the tumors ranged from 3.1 to 9.3 cm. Clinical manifestations included headache, nausea, vomiting, limb weakness/dyskinesia, mobility restriction, epilepsy, dysesthesia and neuropathological indications. All the patients chose surgery removal as
the first option of treatment, in which four patients underwent complete tumor excision under a microscope or with the guidance of ultrasound. Twelve patients underwent partial or nearly complete tumor excision. Postoperative pathology confirmed that five, seven and four cases were rated WHO Grade II, III and IV, respectively. 2.2. CHO-PET imaging A GE Discovery ST 16 PET-CT scanner (USA) was used in this study. 11C-CHO was produced automatically by a cyclotron (GE PET Tracer) through an autosynthesis module with radiochemical purity N95%. For the CHO PET imaging, patients were required to fast for at least 4 h before the imaging study. Three-dimensional brain imaging studies were then performed 20 min after the patients were injected with 11CCHO (296–370 MBq) while at rest. The emission scan mode was used to scan for 10 min, and 11C half-life was selected for scattering decay calibration during scanning. PET scan parameters were as follows: 3D mode; scope: from the skull roof to skull base; thickness, 3.75 mm; matrix, 256×256. CT scan parameters were as follows: voltage, 120–140 kV; current, 110–140 mA; thickness, 2.5 mm; scanning interval, 3.75 mm;
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
439
matrix, 512×512. Axial PET-CT images, sagittal, coronal CT images, and merged PET-CT images were obtained. In the CHO PET images, normal brain tissues in the healthy hemisphere were adjusted to background levels, and lesions displaying highly metabolic area were the positive site, which was defined as the abnormally highly metabolic volume in the CHO PET images (VCHO). Standardized uptake values (SUVs) were measured in this volume. In order to avoid pseudopositive results from acute inflammation and gliosis, CHO PET studies were usually performed 2 weeks postoperation. Both CHO PET and MRI scans were performed with the same treatment position during the same week for all patients. A GE MRI scanner (1.5 T) was used in the study. Scanning sequence included T1-weighted, T2-weighted and contrast-enhanced T1-weighted imaging.
(SUV) of the lesions and the normal cerebral tissues were computed. Consistency of the two images regarding the residual tumors as well as the difference between CHO PET highly metabolic area (VCHO) and range and the contrast enhancement region (VGd) seen on MRI was compared. Radiotherapy target volume was determined based on the CHO PET imaging results combined with the MRI data, and treatment regimens were selected accordingly.
2.3. Image analysis
(1) In CHO-PET images, the tumor target volume, i.e., the highly metabolic area with a high concentration of isotopes (SUV 1.016–4.21) and the corresponding contralateral normal brain tissues (SUV0.1–0.62), was well contrasted, and the bounders between lesions and surrounding normal brain tissues were better defined compared with MRI images (Figs. 1 and 3). For patients with brain gliomas of WHO
All images were reviewed by three experienced clinicians from MRI, PET and radiation therapy fields. With the CHOPET metabolic images, the lesions and corresponding normal brain tissues were analyzed quantitatively, the maximum (max) and average standardized uptake value
2.4. Follow-up All patients were followed for 9–30 months. Overall survival was counted from the date of operation. 3. Results
A
B
C Fig. 2. 11C-CHO PET/CT imaging (SUV 2.2 and SUV 0.9, arrow) of residual tumor and gliosis in the left temporal lobe 14 days after complete tumor excision under a microscope (31-year-old male patient, diffuse astrocytoma, WHO Grade II), without enhancement of MRI (VGd) (A and B: MRI T1-Gd and MRI T2; C: 11 C-CHO PET/CT).
440
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
A
B
C
D
Fig. 3. 11C-CHO PET/CT imaging (SUV 1.3, arrow) and 18F-FDG PET of residual tumor in the left frontal lobe 1 month after complete tumor excision under ultrasound guidance (32-year-old male patient, oligodendroglioma, WHO Grade II), without enhancement of MRI (VGd) (A and B: MRI T1-Gd and MRI T2; C: 11 C-CHO PET/CT; D: 18F-FDG PET/CT).
Grade II, the SUV was 1.016–2.5; for those with WHO Grades III and IV, SUVs were between 2.6 and 4.2. (2) Both CHO PET and MRI were positive for 10 patients and negative for 2 patients. The residual tumor consistency between these two studies was 75% and nonconsistency was 25%. Four of the 10 CHO-PET-positive patients were negative on MRI scans (Figs. 2 and 3). The maximum distance between VGd and VCHO margins was 1.8 cm (Fig. 1). (3) The diagnosis and radiotherapy target volume, and the ensuing treatment regimens were changed for 31.3% (5/16) of patients based on the CHO-PET high-affinity uptake and metabolism (Table 1), in which the change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4) for patients with WHO Grade II, III and IV gliomas, respectively.
4. Discussion Since MRI has an excellent resolution on tissues, plain and contrast-enhanced MRI is often performed to determine treatment options. However, MRI only reflects damages to
the blood–brain barrier (BBB) and is not the scope of actual tumor invasiveness, which may be misleading when branching scope was evaluated. In case that the BBB is not damaged significantly, MRI cannot clearly define either the tumor volumes or the margins of cerebral lesions with normal brain tissues, or heterogeneous proliferation within lesions. Zhang et al. [8] at Tianjin Medical University also concluded that, in clinical practice, MR images at times fail to provide iconographic images for tumor surgery, chemo- or radiotherapy in terms of targets, tumor volume and end points for efficacy. For cerebral gliomas of Grade II, some Grade III as well as a small portion of Grade IV, MR T1-weighted and T2-weighted images often fail to distinguish between tumor, edema surrounding the tumors and normal tissues. PET-CT imaging employs molecular tracers labeled with positron emitting nuclides, and it fully evaluates the metabolic function of tumors; this technology has considerable advantage in malignancy rating, determination of tumor volume and tumor proliferation activity, and heterogeneity of gliomas prior to surgery. In comparison with 18F-FDG PET, 11C-CHO PET/ 11C-MET PET shows lower uptake in normal brain
Died at 5 months
12
12
21
9
2.2/0.2 (residual tumor); 0.6–0.9 (gliosis) (Fig. 2) 1.3/0.2 (residual tumor) (Fig. 3) 4.2/0.4 (tumor recurrence, GTVs 9.3 cm) (Fig. 1)
1.05/0.2 (residual tumor)
CTV changed to CTV plus GTV CTV changed to CTV plus GTV CTV changed to CTV plus GTV CTV changed to CTV plus GTV Deterioration at 28 Gy 1.016/0.1 (residue tumor)
Oligodendroglioma (cystadenoma) Anaplastic astrocytoma
Diffuse astrocytoma
Without enhancement (postoperative change) Lower enhancement (postoperative change) Without enhancement (postoperative change) Without enhancement (postoperative change) High enhancement (tumor recurrence, GTVs 7.5 cm) Female/70
Male/32
Male/31
Female/46
Right parieto-temporal lobe glioma (Grade II); 18 days after complete tumor excision Right frontal lobe glioma (Grade II); 3 days after complete tumor excision under a microscope Left temporal lobe glioma (Grade II); 14 days after complete tumor excision under a microscope Left frontal lobe glioma (Grade II); 1 month after complete tumor excision under ultrasound guidance Right temporoparietal lobe glioma (Grade III); 4 months postoperation Female/34
Diffuse astrocytoma (cystadenoma) Oligo-astrocytoma
Follow-up (months) Regimen change CHO-PET (tumor-SUVmax / brain- SUVmax ) MRI (Gd) Pathology Medical history Gender /age (years)
Table 1 Data of 11C-CHO PET and MRI, pathology, diagnosis or gross tumor volumes (GTVs) and outcome of follow-up were reanalyzed for five patients (31.3%; 5/16) who underwent CHO PET imaging
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
441
tissues, lower background and higher uptake in lesions; therefore, the low background in normal brain tissues and high uptake areas in lesions contrast well: lesions and margins are displayed sharper and clearer [6]. Brain gliomas undergo rapid metabolism, and in these cells choline levels that are synthesized in the tumor are elevated. Information such as malignancy grade or whether there is residual tumor can be obtained judging on the extent of choline metabolism. This information is useful for clinicians to determine the tumor target volume and delineate the radiotherapy target. CHO PET imaging can be used to delineate the glioma target lesions more accurately, which has been confirmed by the pathological examination of stereotactic biopsy samples [7,8]. Pauleit et al. [9] performed pathological studies on the neuro-navigated tissue biopsies taken from lesions and found that MRI yielded a sensitivity of 96% for detection of cerebral gliomas but only with a specificity of 53% and accuracy of 68%, which apparently could not meet the requirement for targeted radiotherapy. Clinical studies [10–12] indicated that MET-PET had a sensitivity of 76– 95%, a specificity of 87–100% and an accuracy of 79% for neural gliomas. PET-guided biopsies for tumors, delineation of GTV and evaluation of chemo-radiotherapy and targeted treatment are highly sensitive and specific [13–15]. This study indicated that the gross tumor volumes (GTVs) and regimens were changed for 31.3% (5/16) of patients. For brain gliomas of WHO Grade II, III and IV, the treatment change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4), respectively. For patients with WHO Grade III–IV gliomas in this study group, MRI had a conformity rate of 90.9%, while the conformity rate for those with WHO Grade II was lower; in comparison, patients with WHO Grade II were affected considerably with CHO PET imaging. This could be due to the fact that the BBB damage might not occur in low-grade (WHO II) gliomas and that such area might be overlooked in MR images since it was not strongly enhanced. On the other hand, the advantage of CHO PET relies on the CHO metabolic enhancement, which is a biological behavior of the tumor self and thus may reflect what is happening with the tumor more accurately. Other possibilities include the fact that only some of the patients underwent FLAIR imaging (water-suppressed imaging) and that there was no enhancement with the cystic changes. As a matter of fact, FLAIR imaging cannot distinguish tumor malignancy as well, but CHO PET reveals tumor BTV considerably better. One option being considered is to perform FLAIR imaging first and then PET. It would be much better if FLAIR imaging could be performed for all the patients. As for reason for the low (0%) change rate for the WHO IV patients, low case number could be a possibility. In the current study, the maximum distance between the VCHO margin and VGd was 1.8 cm, which was in agreement with those reported by other groups (Deng et al. [16], 2 cm; Miwa et al. [17], 3 cm). Conducting a biopsy or performing a second surgery on the abnormally high metabolic area displayed in CHO PET to obtain pathological diagnosis is extremely difficult for
442
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
patients with gliomas who have just undergone an operation. Therefore, in order to evaluate the clinical utility of CHO PET in determination of the glioma tumor target volume, a feasible, noninvasive way would be to conduct a follow-up, observation and evaluation of the post-therapy tumorrecurring sites, local control rate and survival rate [16]. Miwa et al. [17] merged the MET PET and MR images of 10 patients with glioblastomas before surgery and found that tumor recurrence occurred in five patients, in whom the recurring lesions were located in the MET PET abnormally high signal areas outside the MRI contrast-enhanced area, even after a complete surgical resection of the lesion on the initial MRI intensive area. Grosu et al. [18] further confirmed that treatment plans based on merged MET PET-MRI images to delineate the radiotherapy target volume of patients with glioblastoma multiform were associated with improved survival in comparison with treatment plans using MRI alone (median survival time: 9 vs. 5 months; P=.03). In summary, CHO PET imaging is beneficial in selecting treatment regimen for brain gliomas; it delineates GTV more accurately and may avoid potential overtreatment, mistreatment, treatment errors and missing treatment, as well as unnecessary radiation for normal cerebral tissues. This could make more consistent the GTV delineations from different physicians. Indeed, false-negative and false-positive results may happen with PET imaging, and these should be taken into account and excluded in clinical practice. Since this is only a small group of cases, further studies need to be conducted. 5. Conclusion Our preliminary results demonstrate that a difference exists between CHO PET and MRI by which to judge and identify residual tumor for patients with brain gliomas. CHO PET is considered to be a supplementary diagnostic approach for MRI. Tumor target volume may be defined more accurately and rationally when CHO PET is combined with MRI. Biological tumor target volume (BTV) displayed in the CHO PET images is useful in determining or delineating the radiotherapy target volume and in making decisions regarding selecting treatment regimens. References [1] Grosu AL, Weber WA. PET for radiation treatment planning of brain tumors. Radiother Oncol 2010;96:325–7.
[2] Pirotte BJ, Levivier M, Goldman S, et al. PET-guided volumetric resection of supratentorial high-grade gliomas: a survival analysis in 66 consecutive patients. Neurosurgery 2009;64:471–81. [3] Wyss M, Hofer S, Brurhlmeier M, et al. Early metabolic responses in temozolomide treated low-grade glioma patients. J Neurooncol 2009;95:87–93. [4] Bussink J, van herpen CML, Kaanders JHAM, Oyen WJG. PET-CT for response assessment and treatment adaptation in head and neck cancer. Lancet Oncol 2010;11:661–9. [5] Ohtani T, Kurihara H, Ishiuchi S, et al. Brain tumor imaging with carbon-11 choline: comparison with FDG PET and gadoliniumenhanced MR imaging. Eur J Nucl Med 2001;28:1664–70. [6] Hara T, Kondo T, et al. Use of 18F-choline and 11C-choline as contrast agents in PET imaging-guided stereotactic biopsy sampling of gliomas. J Neurosurg 2003;99:474–9. [7] Kato T, Shinoda J, Nakayama N, et al. Metabolic assessment of gliomas using 11C-methonine, [ 18F]fluorodeoxyglucose, and 11Ccholine PET. Am J Neuroradiol 2008;29:1176–82. [8] Zhang C, Yang SY, Zhu T, et al. Improved resection for gliomas using positron emission tomography/computerized tomography neuro-navigation. Chin J Neurosurg 2007;23:83–6. [9] Pauleit D, Floeth F, Hamacher K, et al. O-(2-[ 18F]Fluoroethyl)-Ltyrosine PET combined with MRI improve the diagnostic assessment of cerebral gliomas. Brain 2005;128:678–87. [10] Kracht LW, Miletic H, Busch S, et al. Delineation of brain tumor extent with 11C-methionine PET: local comparison with stereotactic histopathology. J Clin Cancer Res 2004;10:7163–70. [11] Pirotte B, Goldman S, Massager N, et al. Combined use of [ 18F]FDG and [ 11C]MET in 45 PET-guided stereotactic brain biopsies. J Neurosurg 2004;101:476–83. [12] Pirotte B, Goldman S, Massager N, et al. Comparison of 18F-FDG and 11 C-MET PET- guided stereotactic brain biopsy in gliomas. J Nucl Med 2004;45:1293–8. [13] Levivier M, Massager N, Wikler D, et al. Use of stereotactic PET images in dosimetry planning of radio-surgery for brain tumors: clinical experience and proposed classification. J Nucl Med 2004;45: 1146–54. [14] Nestle U, Weber W, Hentschel M, et al. Biological imaging in radiation therapy: role of positron emission tomography. Phys Med Biol 2009; 54:R1–25. [15] Miwa K, Matsuo M, Shinoda J, et al. Hypo-fractionated high-dose irradiation planned by methionine PET for the treatment of glioblastoma multiform. Int J Radiat Oncol Biol Phys 2010;78: S260–1 (supplement). [16] Deng ML, Wu SX, Huang S, et al. A comparative study of 11C-MET PET with MRI for target volume delineation in post-operative radiotherapy for brain high grade gliomas. Chin J Radiat Oncol 2010;19:415–9. [17] Miwa K, Shinoda J, Yano H, et al. Discrepancy between lesion distribution on MET-PET and MRI image in patients with glioblastoma multiform: insight from a PET and MR fusion image study. J Neurol Neurosurg Psychiatry 2004;75:1457–62. [18] Grosu AL, Weber WA, Franz M, et al. Reirradiation of recurrent highgrade gliomas using amino acid PET/CT/MRI image fusion to determine gross tumor volume for stereotactic fractional radiotherapy. Int J Radiat Oncol Biol Phys 2005;63:511–9.
Nuclear Medicine and Biology 39 (2012) 437 – 442 www.elsevier.com/locate/nucmedbio
11
C-CHO PET in optimization of target volume delineation and treatment regimens in postoperative radiotherapy for brain gliomas☆
Fang-Ming Li a,⁎, Qing Nie a , Rui-Min Wang c , Susan M. Chang b , Wen-Rui Zhao a , Qi Zhu a , Ying-Kui Liang a , Ping Yang a , Jun Zhang a , Hai-Wei Jia a , Heng-Hu Fang a a
Department of Nuclear Medicine, Center of Radiation Oncology, People's Liberation Army's Navy General Hospital, Beijing 10048, China b Department of Neurological Surgery, University of California, San Francisco, CA, USA c PET/CT Center, Department of Nuclear Medicine, People's Liberation Army's Hospital No. 301, Beijing 100538, China Received 31 May 2011; received in revised form 2 October 2011; accepted 4 October 2011
Abstract Objective: We explored the clinical values of 11C-choline ( 11C-CHO) PET in optimization of target volume delineation and treatment regimens in postoperative radiotherapy for brain gliomas. Methods: Sixteen patients with the pathological confirmation of the diagnosis of gliomas prior to receiving radiotherapy (postoperative) were included, and on whom both MRI and CHO PET scans were performed at the same position for comparison of residual tumors with the two techniques. 11C-CHO was used as the tracer in the PET scan. A plain T1-weighted, T2-weighted and contrast-enhanced T1-weighted imaging scans were performed in the MRI scan sequence. The gliomas' residual tumor volume was defined as the area with CHO-PET high-affinity uptake and metabolism (VCHO) and one with MRI T1-weighted imaging high signal intensity (VGd), and was determined by a group of experienced professionals and clinicians. Results: (1) In CHO-PET images, the tumor target volume, i.e., the highly metabolic area with a high concentration of isotopes (SUV 1.016– 4.21) and the corresponding contralateral normal brain tissues (SUV0.1–0.62), was well contrasted, and the boundary between lesions and surrounding normal brain tissues was better defined compared with MRI and 18F-FDG PET images. (2) For patients with brain gliomas of WHO Grade II, the SUV was 1.016–2.5; for those with WHO Grades III and IV, SUVs were N26–4.2. (3) Both CHO PET and MRI were positive for 10 patients and negative for 2 patients. The residual tumor consistency between these two studies was 75%. Four of the 10 CHO-PET-positive patients were negative on MRI scans. The maximum distance between VGd and VCHO margins was 1.8 cm. (4) The gross tumor volumes (GTVs) and the ensuing treatment regimens were changed for 31.3% (5/16) of patients based on the CHO-PET high-affinity uptake and metabolism, in which the change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4) for patients with WHO Grade II III, and IV gliomas, respectively. Conclusion: Our data demonstrate that difference exists between CHO PET and MRI by which to judge and identify residual tumor for patients with brain gliomas. CHO PET is considered to be a supplementary diagnostic approach for MRI. Biological tumor target volume (BTV) displayed in the CHO PET images is useful in determining or delineating the radiotherapy target volume and making decisions in selecting treatment regimens. Tumor target volume may be defined more accurately and rationally when the CHO PET is combined with MRI. © 2012 Elsevier Inc. All rights reserved. Keywords: Brain glioma; Radioactive carbon-labeled choline; PET imaging; MRI; Radiotherapy target volume; Radiotherapy regimen
1. Introduction Revealing the gross tumor volume (GTV) of brain gliomas accurately and truly with imaging technology is a ☆
Conflict of interest and disclosure statement: The authors declare no competing financial interests. ⁎ Corresponding author. Tel.: +86 010 66958128, 66958478; fax: +86 010 68780892. E-mail address: [email protected] (F.-M. Li). 0969-8051/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.nucmedbio.2011.10.003
prerequisite and a key element in achieving treatment success. Prior to receiving radiotherapy, the tumor volumes for patients with gliomas have to be determined in order for clinicians to delineate the target volume. Tumors cannot be evaluated with conventional morphology imaging technology. PET imaging provides an unprecedented opportunity for noninvasive diagnosis of brain gliomas and is an essential way in undergoing navigated surgery, targeted radiotherapy and treatment outcome evaluation of chemotherapy [1–3]. PET makes it possible to delineate the biological target
438
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
Fig. 1. MRI, 18F-FDG PET and 11C-CHO PET/CT images of a high-grade glioma (female, age 70, anaplastic glioma, WHO Grade III). Comparison of MRI (VGd) and 18F-FDG PET (VFDG) , 11C-CHO PET/CT ( SUV 4.2) images identified target volumes and the exact boundary of the tumor and their relation to the surrounding edema and showed clearly the boundaries of the tumors (A and B: MRI T1-Gd and MRI T2; C: 11C-CHO PET/CT; D: 18F-FDG PET/CT).
volume (BTV) of gliomas. In this study, 11C-CHO PET function imaging and MRI morphology imaging results were compared and contrasted so that the radiotherapy target volume may be defined more accurately and treatment regimens optimized [4,5]. 2. Materials and methods 2.1. Patients During the period of January 2008 to June 2009, 16 patients with the pathological confirmation of the diagnosis of gliomas prior to receiving radiotherapy (postoperative) were included in this study. Among the patients were eight males and eight females with a median age of 45 years (range: 8–70). Sites of the lesions were as follows: in one case the lesion was in the suprasellar region; in the other 15 cases, the tumors were located in the frontal lobe, temporal lobe and parietal– occipital lobe area. The maximum diameter for the tumors ranged from 3.1 to 9.3 cm. Clinical manifestations included headache, nausea, vomiting, limb weakness/dyskinesia, mobility restriction, epilepsy, dysesthesia and neuropathological indications. All the patients chose surgery removal as
the first option of treatment, in which four patients underwent complete tumor excision under a microscope or with the guidance of ultrasound. Twelve patients underwent partial or nearly complete tumor excision. Postoperative pathology confirmed that five, seven and four cases were rated WHO Grade II, III and IV, respectively. 2.2. CHO-PET imaging A GE Discovery ST 16 PET-CT scanner (USA) was used in this study. 11C-CHO was produced automatically by a cyclotron (GE PET Tracer) through an autosynthesis module with radiochemical purity N95%. For the CHO PET imaging, patients were required to fast for at least 4 h before the imaging study. Three-dimensional brain imaging studies were then performed 20 min after the patients were injected with 11CCHO (296–370 MBq) while at rest. The emission scan mode was used to scan for 10 min, and 11C half-life was selected for scattering decay calibration during scanning. PET scan parameters were as follows: 3D mode; scope: from the skull roof to skull base; thickness, 3.75 mm; matrix, 256×256. CT scan parameters were as follows: voltage, 120–140 kV; current, 110–140 mA; thickness, 2.5 mm; scanning interval, 3.75 mm;
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
439
matrix, 512×512. Axial PET-CT images, sagittal, coronal CT images, and merged PET-CT images were obtained. In the CHO PET images, normal brain tissues in the healthy hemisphere were adjusted to background levels, and lesions displaying highly metabolic area were the positive site, which was defined as the abnormally highly metabolic volume in the CHO PET images (VCHO). Standardized uptake values (SUVs) were measured in this volume. In order to avoid pseudopositive results from acute inflammation and gliosis, CHO PET studies were usually performed 2 weeks postoperation. Both CHO PET and MRI scans were performed with the same treatment position during the same week for all patients. A GE MRI scanner (1.5 T) was used in the study. Scanning sequence included T1-weighted, T2-weighted and contrast-enhanced T1-weighted imaging.
(SUV) of the lesions and the normal cerebral tissues were computed. Consistency of the two images regarding the residual tumors as well as the difference between CHO PET highly metabolic area (VCHO) and range and the contrast enhancement region (VGd) seen on MRI was compared. Radiotherapy target volume was determined based on the CHO PET imaging results combined with the MRI data, and treatment regimens were selected accordingly.
2.3. Image analysis
(1) In CHO-PET images, the tumor target volume, i.e., the highly metabolic area with a high concentration of isotopes (SUV 1.016–4.21) and the corresponding contralateral normal brain tissues (SUV0.1–0.62), was well contrasted, and the bounders between lesions and surrounding normal brain tissues were better defined compared with MRI images (Figs. 1 and 3). For patients with brain gliomas of WHO
All images were reviewed by three experienced clinicians from MRI, PET and radiation therapy fields. With the CHOPET metabolic images, the lesions and corresponding normal brain tissues were analyzed quantitatively, the maximum (max) and average standardized uptake value
2.4. Follow-up All patients were followed for 9–30 months. Overall survival was counted from the date of operation. 3. Results
A
B
C Fig. 2. 11C-CHO PET/CT imaging (SUV 2.2 and SUV 0.9, arrow) of residual tumor and gliosis in the left temporal lobe 14 days after complete tumor excision under a microscope (31-year-old male patient, diffuse astrocytoma, WHO Grade II), without enhancement of MRI (VGd) (A and B: MRI T1-Gd and MRI T2; C: 11 C-CHO PET/CT).
440
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
A
B
C
D
Fig. 3. 11C-CHO PET/CT imaging (SUV 1.3, arrow) and 18F-FDG PET of residual tumor in the left frontal lobe 1 month after complete tumor excision under ultrasound guidance (32-year-old male patient, oligodendroglioma, WHO Grade II), without enhancement of MRI (VGd) (A and B: MRI T1-Gd and MRI T2; C: 11 C-CHO PET/CT; D: 18F-FDG PET/CT).
Grade II, the SUV was 1.016–2.5; for those with WHO Grades III and IV, SUVs were between 2.6 and 4.2. (2) Both CHO PET and MRI were positive for 10 patients and negative for 2 patients. The residual tumor consistency between these two studies was 75% and nonconsistency was 25%. Four of the 10 CHO-PET-positive patients were negative on MRI scans (Figs. 2 and 3). The maximum distance between VGd and VCHO margins was 1.8 cm (Fig. 1). (3) The diagnosis and radiotherapy target volume, and the ensuing treatment regimens were changed for 31.3% (5/16) of patients based on the CHO-PET high-affinity uptake and metabolism (Table 1), in which the change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4) for patients with WHO Grade II, III and IV gliomas, respectively.
4. Discussion Since MRI has an excellent resolution on tissues, plain and contrast-enhanced MRI is often performed to determine treatment options. However, MRI only reflects damages to
the blood–brain barrier (BBB) and is not the scope of actual tumor invasiveness, which may be misleading when branching scope was evaluated. In case that the BBB is not damaged significantly, MRI cannot clearly define either the tumor volumes or the margins of cerebral lesions with normal brain tissues, or heterogeneous proliferation within lesions. Zhang et al. [8] at Tianjin Medical University also concluded that, in clinical practice, MR images at times fail to provide iconographic images for tumor surgery, chemo- or radiotherapy in terms of targets, tumor volume and end points for efficacy. For cerebral gliomas of Grade II, some Grade III as well as a small portion of Grade IV, MR T1-weighted and T2-weighted images often fail to distinguish between tumor, edema surrounding the tumors and normal tissues. PET-CT imaging employs molecular tracers labeled with positron emitting nuclides, and it fully evaluates the metabolic function of tumors; this technology has considerable advantage in malignancy rating, determination of tumor volume and tumor proliferation activity, and heterogeneity of gliomas prior to surgery. In comparison with 18F-FDG PET, 11C-CHO PET/ 11C-MET PET shows lower uptake in normal brain
Died at 5 months
12
12
21
9
2.2/0.2 (residual tumor); 0.6–0.9 (gliosis) (Fig. 2) 1.3/0.2 (residual tumor) (Fig. 3) 4.2/0.4 (tumor recurrence, GTVs 9.3 cm) (Fig. 1)
1.05/0.2 (residual tumor)
CTV changed to CTV plus GTV CTV changed to CTV plus GTV CTV changed to CTV plus GTV CTV changed to CTV plus GTV Deterioration at 28 Gy 1.016/0.1 (residue tumor)
Oligodendroglioma (cystadenoma) Anaplastic astrocytoma
Diffuse astrocytoma
Without enhancement (postoperative change) Lower enhancement (postoperative change) Without enhancement (postoperative change) Without enhancement (postoperative change) High enhancement (tumor recurrence, GTVs 7.5 cm) Female/70
Male/32
Male/31
Female/46
Right parieto-temporal lobe glioma (Grade II); 18 days after complete tumor excision Right frontal lobe glioma (Grade II); 3 days after complete tumor excision under a microscope Left temporal lobe glioma (Grade II); 14 days after complete tumor excision under a microscope Left frontal lobe glioma (Grade II); 1 month after complete tumor excision under ultrasound guidance Right temporoparietal lobe glioma (Grade III); 4 months postoperation Female/34
Diffuse astrocytoma (cystadenoma) Oligo-astrocytoma
Follow-up (months) Regimen change CHO-PET (tumor-SUVmax / brain- SUVmax ) MRI (Gd) Pathology Medical history Gender /age (years)
Table 1 Data of 11C-CHO PET and MRI, pathology, diagnosis or gross tumor volumes (GTVs) and outcome of follow-up were reanalyzed for five patients (31.3%; 5/16) who underwent CHO PET imaging
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
441
tissues, lower background and higher uptake in lesions; therefore, the low background in normal brain tissues and high uptake areas in lesions contrast well: lesions and margins are displayed sharper and clearer [6]. Brain gliomas undergo rapid metabolism, and in these cells choline levels that are synthesized in the tumor are elevated. Information such as malignancy grade or whether there is residual tumor can be obtained judging on the extent of choline metabolism. This information is useful for clinicians to determine the tumor target volume and delineate the radiotherapy target. CHO PET imaging can be used to delineate the glioma target lesions more accurately, which has been confirmed by the pathological examination of stereotactic biopsy samples [7,8]. Pauleit et al. [9] performed pathological studies on the neuro-navigated tissue biopsies taken from lesions and found that MRI yielded a sensitivity of 96% for detection of cerebral gliomas but only with a specificity of 53% and accuracy of 68%, which apparently could not meet the requirement for targeted radiotherapy. Clinical studies [10–12] indicated that MET-PET had a sensitivity of 76– 95%, a specificity of 87–100% and an accuracy of 79% for neural gliomas. PET-guided biopsies for tumors, delineation of GTV and evaluation of chemo-radiotherapy and targeted treatment are highly sensitive and specific [13–15]. This study indicated that the gross tumor volumes (GTVs) and regimens were changed for 31.3% (5/16) of patients. For brain gliomas of WHO Grade II, III and IV, the treatment change rate was 80% (4/5), 14.3 % (1/7) and 0% (0/4), respectively. For patients with WHO Grade III–IV gliomas in this study group, MRI had a conformity rate of 90.9%, while the conformity rate for those with WHO Grade II was lower; in comparison, patients with WHO Grade II were affected considerably with CHO PET imaging. This could be due to the fact that the BBB damage might not occur in low-grade (WHO II) gliomas and that such area might be overlooked in MR images since it was not strongly enhanced. On the other hand, the advantage of CHO PET relies on the CHO metabolic enhancement, which is a biological behavior of the tumor self and thus may reflect what is happening with the tumor more accurately. Other possibilities include the fact that only some of the patients underwent FLAIR imaging (water-suppressed imaging) and that there was no enhancement with the cystic changes. As a matter of fact, FLAIR imaging cannot distinguish tumor malignancy as well, but CHO PET reveals tumor BTV considerably better. One option being considered is to perform FLAIR imaging first and then PET. It would be much better if FLAIR imaging could be performed for all the patients. As for reason for the low (0%) change rate for the WHO IV patients, low case number could be a possibility. In the current study, the maximum distance between the VCHO margin and VGd was 1.8 cm, which was in agreement with those reported by other groups (Deng et al. [16], 2 cm; Miwa et al. [17], 3 cm). Conducting a biopsy or performing a second surgery on the abnormally high metabolic area displayed in CHO PET to obtain pathological diagnosis is extremely difficult for
442
F.-M. Li et al. / Nuclear Medicine and Biology 39 (2012) 437–442
patients with gliomas who have just undergone an operation. Therefore, in order to evaluate the clinical utility of CHO PET in determination of the glioma tumor target volume, a feasible, noninvasive way would be to conduct a follow-up, observation and evaluation of the post-therapy tumorrecurring sites, local control rate and survival rate [16]. Miwa et al. [17] merged the MET PET and MR images of 10 patients with glioblastomas before surgery and found that tumor recurrence occurred in five patients, in whom the recurring lesions were located in the MET PET abnormally high signal areas outside the MRI contrast-enhanced area, even after a complete surgical resection of the lesion on the initial MRI intensive area. Grosu et al. [18] further confirmed that treatment plans based on merged MET PET-MRI images to delineate the radiotherapy target volume of patients with glioblastoma multiform were associated with improved survival in comparison with treatment plans using MRI alone (median survival time: 9 vs. 5 months; P=.03). In summary, CHO PET imaging is beneficial in selecting treatment regimen for brain gliomas; it delineates GTV more accurately and may avoid potential overtreatment, mistreatment, treatment errors and missing treatment, as well as unnecessary radiation for normal cerebral tissues. This could make more consistent the GTV delineations from different physicians. Indeed, false-negative and false-positive results may happen with PET imaging, and these should be taken into account and excluded in clinical practice. Since this is only a small group of cases, further studies need to be conducted. 5. Conclusion Our preliminary results demonstrate that a difference exists between CHO PET and MRI by which to judge and identify residual tumor for patients with brain gliomas. CHO PET is considered to be a supplementary diagnostic approach for MRI. Tumor target volume may be defined more accurately and rationally when CHO PET is combined with MRI. Biological tumor target volume (BTV) displayed in the CHO PET images is useful in determining or delineating the radiotherapy target volume and in making decisions regarding selecting treatment regimens. References [1] Grosu AL, Weber WA. PET for radiation treatment planning of brain tumors. Radiother Oncol 2010;96:325–7.
[2] Pirotte BJ, Levivier M, Goldman S, et al. PET-guided volumetric resection of supratentorial high-grade gliomas: a survival analysis in 66 consecutive patients. Neurosurgery 2009;64:471–81. [3] Wyss M, Hofer S, Brurhlmeier M, et al. Early metabolic responses in temozolomide treated low-grade glioma patients. J Neurooncol 2009;95:87–93. [4] Bussink J, van herpen CML, Kaanders JHAM, Oyen WJG. PET-CT for response assessment and treatment adaptation in head and neck cancer. Lancet Oncol 2010;11:661–9. [5] Ohtani T, Kurihara H, Ishiuchi S, et al. Brain tumor imaging with carbon-11 choline: comparison with FDG PET and gadoliniumenhanced MR imaging. Eur J Nucl Med 2001;28:1664–70. [6] Hara T, Kondo T, et al. Use of 18F-choline and 11C-choline as contrast agents in PET imaging-guided stereotactic biopsy sampling of gliomas. J Neurosurg 2003;99:474–9. [7] Kato T, Shinoda J, Nakayama N, et al. Metabolic assessment of gliomas using 11C-methonine, [ 18F]fluorodeoxyglucose, and 11Ccholine PET. Am J Neuroradiol 2008;29:1176–82. [8] Zhang C, Yang SY, Zhu T, et al. Improved resection for gliomas using positron emission tomography/computerized tomography neuro-navigation. Chin J Neurosurg 2007;23:83–6. [9] Pauleit D, Floeth F, Hamacher K, et al. O-(2-[ 18F]Fluoroethyl)-Ltyrosine PET combined with MRI improve the diagnostic assessment of cerebral gliomas. Brain 2005;128:678–87. [10] Kracht LW, Miletic H, Busch S, et al. Delineation of brain tumor extent with 11C-methionine PET: local comparison with stereotactic histopathology. J Clin Cancer Res 2004;10:7163–70. [11] Pirotte B, Goldman S, Massager N, et al. Combined use of [ 18F]FDG and [ 11C]MET in 45 PET-guided stereotactic brain biopsies. J Neurosurg 2004;101:476–83. [12] Pirotte B, Goldman S, Massager N, et al. Comparison of 18F-FDG and 11 C-MET PET- guided stereotactic brain biopsy in gliomas. J Nucl Med 2004;45:1293–8. [13] Levivier M, Massager N, Wikler D, et al. Use of stereotactic PET images in dosimetry planning of radio-surgery for brain tumors: clinical experience and proposed classification. J Nucl Med 2004;45: 1146–54. [14] Nestle U, Weber W, Hentschel M, et al. Biological imaging in radiation therapy: role of positron emission tomography. Phys Med Biol 2009; 54:R1–25. [15] Miwa K, Matsuo M, Shinoda J, et al. Hypo-fractionated high-dose irradiation planned by methionine PET for the treatment of glioblastoma multiform. Int J Radiat Oncol Biol Phys 2010;78: S260–1 (supplement). [16] Deng ML, Wu SX, Huang S, et al. A comparative study of 11C-MET PET with MRI for target volume delineation in post-operative radiotherapy for brain high grade gliomas. Chin J Radiat Oncol 2010;19:415–9. [17] Miwa K, Shinoda J, Yano H, et al. Discrepancy between lesion distribution on MET-PET and MRI image in patients with glioblastoma multiform: insight from a PET and MR fusion image study. J Neurol Neurosurg Psychiatry 2004;75:1457–62. [18] Grosu AL, Weber WA, Franz M, et al. Reirradiation of recurrent highgrade gliomas using amino acid PET/CT/MRI image fusion to determine gross tumor volume for stereotactic fractional radiotherapy. Int J Radiat Oncol Biol Phys 2005;63:511–9.