Abstract
Management of low-grade gliomas continues to be a challenging task, because CT and MRI do not always differentiate from nontumoral lesions. Furthermore, tumor extent and aggressiveness often remain unclear because of a lack of contrast enhancement. Previous studies indicated that large neutral amino acid tracers accumulate in most brain tumors, including low-grade gliomas, probably because of changes of endothelial and blood-brain barrier function. We describe 11C-methionine uptake measured with PET in a series of 196 consecutive patients, most of whom were studied because of suspected low-grade gliomas. Uptake in the most active lesion area, relative to contralateral side, was significantly different among high-grade gliomas, low-grade gliomas, and chronic or subacute nontumoral lesions, and this difference was independent from contrast enhancement in CT or MRI. Corticosteroids had no significant effect on methionine uptake in low-grade gliomas but reduced uptake moderately in high-grade gliomas. Differentiation between gliomas and nontumoral lesions by a simple threshold was correct in 79%. Recurrent or residual tumors had a higher uptake than primary gliomas. In conclusion, the high sensitivity of 11C-methionine uptake for functional endothelial or blood-brain barrier changes suggests that this tracer is particularly useful for evaluation and follow-up of low-grade gliomas.
In the general population, gliomas are relatively rare, with an incidence of approximately 4 in 100,000. Among them, the proportion of low-grade gliomas is less than 50%.1 Yet the latter often affect relatively young and otherwise healthy people. Low-grade gliomas cause several specific difficulties for diagnosis and therapy. Often, lack of contrast enhancement on CT and MRI does not allow clear differentiation from nontumoral lesions. Tumor borders may be poorly delineated, often corresponding to infiltrating growth and local edema. Tumors may be inhomogeneous with regard to histologic grading and biological aggressiveness; thus, even in the case of image-guided and computer-assisted treatment planning,2 a tissue sample from a stereotactic biopsy may not be representative of the whole tumor.3 Even very slowly growing tumors and late recurrences have a considerable impact on the quality of life and reduce the otherwise long life expectancy of these younger patients considerably. Treatment, including macroscopic in toto resection, is generally considered as palliative in tumors of grade 2 and higher because of frequent, albeit sometimes late, recurrence.
There has not yet been a prospective study that could demonstrate a benefit in terms of survival or quality of life by early resection. Yet a retrospective study with volumetric CT/MRI-based assessment of preoperative and postoperative tumor volumes indicates that complete resection of grade 2 gliomas, which is more easily achieved in small tumors (volume <10 cm3), may reduce the probability of recurrence and malignant transformation.4 Tumors with a diameter of approximately 4 cm or less may also be treated successfully by interstitial radiotherapy.5 On the other hand, surgery probably does not improve prognosis if only part of the tumor is removed or irradiated, and it carries a risk of neurologic damage, which is particularly significant because these tumors at an early stage often do not cause functional neurologic impairment. The effect of external radiation therapy is limited, and long-term side effects with cognitive impairment may occur, arguing against regular use of this treatment modality early in the course of the disease.6 Thus, techniques for further improvement of preoperative estimation of malignancy and early detection of recurrence could play an important role in patient management.
Previous studies with PET in selected series showed that the amino acid tracer 11C-L-methionine accumulates in gliomas,7-10 including low-grade gliomas, even if contrast enhancement is absent and glucose consumption as measured by18 F-2-fluoro-2-deoxy-glucose is low, suggesting that the use of this tracer could improve diagnosis of low-grade gliomas. We studied a large series of consecutive patients to determine whether this technique adds information about tumor extent and recurrence, grading, and histologic type to morphologic imaging by CT and MRI.
Methods. From 1992 until 1996, 196 patients (108 men and 88 women; mean age, 44.8 ± 14.0 years) with suspected brain tumors were studied at our institution. Diagnosis was verified by histologic examination in 170 patients (in approximately 43% of them by stereotactic biopsy) or was established by clinical follow-up and additional investigations(essentially CSF examination and follow-up MRI or CT) in the remainder. During the same time, another 15 patients were studied but not included because of inconclusive clinical data. Most patients were examined because a low-grade glioma was suspected or considered in differential diagnosis. Clinical subgroups relevant for data analysis are listed intable 1.
Table 1 Clinical subgroups
Several patients had follow-up scans (5 patients had three follow-up scans, 5 had two, and 18 had one), resulting in 239 PET scans with11 C-methionine. Time intervals to follow-up after the initial scans were 307 ± 234 days for the first, 605 ± 286 for the second, and 921 ± 329 for the third scans. Treatment included complete or partial resection, external radiation, implantation of radioactive iodine seeds, or chemotherapy. Most follow-up scans were performed because residual or recurrent tumor was suspected on MRI or CT or because of progression of clinical symptoms. Histologic proof of recurrent tumor was obtained in 23 patients.
PETs were performed after IV slow bolus injection of 20 mCi (740 MBq)11 C-methionine, synthesized according to the method of Berger et al.11 Tracer accumulation was recorded in 47 transaxial slices from the entire brain with either an ECAT EXACT or an ECAT EXACT HR(Siemens-CTI; Knoxville, TN).12,13 In most scans (n = 208), data acquisition in conventional two-dimensional mode lasted 30 minutes, and in 31 scans, three-dimensional data acquisition with septa removed14 and extended measurement time (60 minutes) was used. Data were corrected for scatter and attenuation, and images were reconstructed from counts accumulated over 0 to 30 or 20 to 60 minutes, respectively.
For quantitative analysis, a circular region of interest of 8-mm diameter was placed in the hottest area of each lesion or in its center as located by CT or MRI if increased 11C-methionine uptake was not present. A mirror region of the same size was placed as a reference in contralateral tissue. In cases of bilateral tumors, the reference region was placed in unaffected remote tissue. A methionine uptake index was calculated by dividing lesion by reference activity.
All patients had CTs or MRIs within few weeks of each PET examination. Scans after administration of contrast agents were available for evaluation in 154 patients. Contrast enhancement was graded visually on a four-point (1= absent, 2 = questionable, 3 = moderate, 4 = intense). Because, due to the selection criteria, only three untreated gliomas showed contrast enhancement of grade 4, they were included in contrast enhancement group 3 for data analysis.
Quantitative values are reported as arithmetic means ± SD. For statistical analysis, a logarithmic transformation was necessary to approximate a normal distribution, because the untransformed distribution of methionin uptake indexes was significantly skewed (average, 2.15; median, 1.83). Comparisons between two groups were analyzed by t-test and among three groups or more by ANOVA with Tukey's multiple means comparison. Influence of tumor grade was analyzed by multiple regression. Diagnostic accuracy and the optimum threshold for differentiation among groups was determined by analysis of the receiver operating characteristic. The SAS software package (Release 6.11, SAS Institute Inc., Cary, NC) was used for all calculations.
Results. Of 99 untreated patients with suspected glioma and histologic diagnosis (see table 1, Group IA), 83 actually had an astrocytoma, oligoastrocytoma, or oligodendroglioma that could be graded according to the World Health Organization (WHO) classification.15 Although glioblastoma is regarded as a separate entity in the WHO classification, it is listed as the only astrocytic tumor of grade 4, and we therefore included it in the present study in the astrocytoma group. Methionine accumulation in these patients is listed in table 2 (Group u). Within grade 2 tumors, significantly higher methionine uptake was found in oligodendrogliomas than in astrocytomas (p = 0.013). Oligoastrocytomas did not differ from astrocytomas and thus were included into the astrocytoma group in subsequent analyses. Malignant astrocytomas of grade 3 or 4 had a significantly higher methionine uptake (n = 28, 3.0 ± 1.1) than low-grade astrocytomas (n = 47, 1.7 ± 0.6), and these had significantly higher uptake than nontumoral lesions (n = 24, 1.3 ± 0.3; p = 0.0001).
Table 2 11C-Methionine uptake in patients with histologically verified astrocytoma, oligoastrocytoma, or oligodendroglioma
Information on corticosteroid medication the day before and the day of the PET study was available for 108 patients of Group I. Sixty-seven patients were off corticosteroids, 39 were on dexamethasone (daily doses ranging from 2 to 24 mg; average, 13 ± 5.5), and 2 received 100 or 500 mg methylprednisolone, respectively, per day. There was no significant effect of corticosteroids on methionine uptake in low-grade gliomas, but uptake was moderately, albeit significantly, reduced in glioblastomas(table 3). High-grade gliomas with corticosteroids still had a higher uptake than low-grade gliomas.
Table 3 Effect of corticosteroid uptake on methionine uptake in untreated gliomas
CT or MR images for assessment of contrast enhancement were available for 84 patients of Group I. As expected, moderate or intense contrast enhancement was seen more frequently in high-grade gliomas (10/21, 48%) than in low-grade gliomas (6/41, 15%) and was present in 9 of 22 nontumoral lesions (41%). Two-way ANOVA with factors contrast enhancement and lesion type revealed no significant effect of contrast enhancement on methionine uptake and confirmed the already known effect of lesion type (p = 0.001,figure 1).
Figure 1. Box plots of methionine uptake versus contrast enhancement (grade 1 = absent, 2 = questionable, 3 = moderate or intense) for low-grade and high-grade gliomas and nontumoral lesions separately. It illustrates the significant effect of tumor grade that is independent from contrast enhancement. The extent of the boxes indicates the 25th and 75th percentiles of the column and the line inside the box marks the value of the 50th percentile. Capped bars indicate the 10th and 90th percentiles, and symbols mark all data outside of this interval.
With regard to discrimination between nontumoral and tumor lesions in patients with suspected glioma (Group I), the best discrimination between tumoral and nontumoral lesions was achieved at a threshold of 1.47. As illustrated in figure 2, 76% of tumors were above that threshold (sensitivity) and 87% of nontumoral lesions were below(specificity), resulting in 79% correct classifications. Exclusion of malignant lesions (grade 3 or higher) reduced sensitivity to 67%, and 72% of lesions were classified correctly. We observed three astrocytomas of grade 2 with less methionine uptake than in corresponding contralateral tissue, which was also less than in any nontumoral lesion. Low-grade astrocytomas with high uptake often affected the cortex and were occasionally barely visible on MRI(figure 3).
Figure 2. Receiver operating characteristic curve for differentiation between tumor and nontumoral lesions by methionine uptake index. The curve is generated by varying the threshold for methionine uptake over the whole range of values and plotting the proportion of nontumoral lesions (false positives) versus the proportion of tumors (true positives) with uptake above the respective threshold.
Figure 3. Small parasagittal astrocytoma grade 2, confirmed by biopsy, showing minor signal alterations on MRI (left, T1-weighted; middle, T2-weighted) but intense 11C-methionine uptake (right, uptake index 2.43) in the corresponding slice.
Methionine uptake in the whole group of treated gliomas (seetable 1, Group III) did not differ significantly from untreated gliomas of the same histologic type and grade and showed the same differences between histologic types (astrocytoma including glioblastoma, oligoastrocytoma, and oligodendroglioma) and grades (p = 0.0007 and p = 0.0001, respectively, in three-way ANOVA) (seetable 2, Group t). Yet in univariate t-test, methionine uptake in recurrent astrocytoma grade 2 was significantly higher than in untreated patients (p = 0.04, uncorrected for multiple comparisons), possibly reflecting malignant transformation. Higher uptake in recurrent tumors was most obvious and significant in three-way ANOVA(p = 0.028) when the analysis was limited to those 23 lesions in which recurrent or residual tumor was proven by resection or follow-up (seetable 2, Group r), eliminating the influence of potentially successful palliative treatment on the data. One scan of Group III had been performed postoperatively in a patient in whom only a hematoma had been found histologically, although a low-grade glioma had been suspected preoperatively (preoperative uptake, 1.7, a false-positive case). This follow-up scan showed a reduced uptake of 1.36 (below the threshold of 1.47), consistent with the histologic finding of absence of tumor.
Methionine uptake in lesions other than WHO-graded gliomas is listed intable 4. Most nontumoral lesions were studied because a low-grade glioma was considered in the differential diagnosis (seetable 1, Group I). Most nongliomatous tumors (Group II) were recurrent after previous treatment. The highest uptake was seen in malignant gliosarcoma, and a high uptake ≥2 was present also in pituitary adenomas, meningiomas, and ependymomas, even at low histologic grades, probably because of the absence of a blood-tumor barrier or high vascularization.
Table 4 Methionine uptake in other tumors and nontumoral lesions*
Discussion. The mechanism and biological significance of increased uptake of 11C-methionine in gliomas are not yet understood completely. Although 11C-methionine is incorporated into proteins,16 its uptake is probably not a measure of protein synthesis. Large neutral amino acids (LNAAs) are mainly transported across the blood-brain barrier (BBB) by a carrier system.17 Kinetic studies with an LNAA analogue tracer,18 F-fluorotyrosine, which is also incorporated into proteins, showed an increase of transport-related rate-constant K1 but not of metabolism-related rate-constant k3 in glioma.18 Uptake in brain tumors has the same stereospecificity as in normal brain19 and can be inhibited by competition with unlabeled amino acids.20 Increased uptake in gliomas is also observed with123 I-iodo-alphamethyl tyrosine21 that is not incorporated into proteins. These data suggest that increased uptake is mainly due to activation of carrier-mediated transport at the BBB.
The present study also indicates that methionine uptake is not mainly due to increased diffusion. Even in two glioblastomas without contrast enhancement, which is a rare atypical feature,22 high methionine uptake as typical for this grade of malignancy was observed(figure 4). Yet there was little methionine uptake in subacute ischemic lesions that often show a typical pattern of gyral contrast enhancement,23 which is probably due to BBB damage in necrotic tissue. Thus, significant uptake of methionine seems to occur only in viable tissue with intact carrier mechanisms. The independence of methionine uptake from corticosteroid medication in low-grade tumors is also consistent with an activated carrier process, whereas agents that depend on diffusion across the BBB typically show a reduction of uptake by corticosteroids.24 That effect was seen with methionine only in malignant gliomas, where diffusion may contribute to total uptake.25
Figure 4. Subacute ischemic infarct with typical gyral contrast enhancement (Gd-DTPA) but little methionine uptake (A) compared with glioblastoma without contrast enhancement but very high methionine uptake (B).
A significant correlation between tumor grade and methionine uptake has already been described in previous studies, also with a more pronounced difference between grades 2 and 3 than between grades 3 and 4.8,10 This does not seem to be specific for methionine or other amino acid tracers but is also true for glucose metabolism26 and potassium analogue tracers82 Rb and 201TI.25,27 Thus, we cannot demonstrate directly that 11C-methionine is superior in that respect to these other tracers, but its increased accumulation already at grade 2 and the complete absence of nonaccumulating malignant gliomas in this series suggest an improved sensitivity for detection of malignant degeneration.
Increased uptake in grade 2 gliomas seems to be unique to11 C-methionine and other LNAAs, such as L-[1-11C]tyrosine28 and L-[2-18F]-fluoro-tyrosine,18 whereas contrast enhancement is usually absent and glucose consumption is low, similar to normal white matter.26 It has been reported that uptake of 82Rb is also higher in grade 2 astrocytomas than in contralateral brain, but uptake rates in both tissue types are nearly 10-fold lower than for methionine,29 resulting in images with an unfavorable signal-to-noise ratio. The increase of carrier-mediated LNAA transport could possibly be due to beginning angiogenesis, because upregulation of vascular endothelial growth factor (VEGF) expression and VEGF receptor type 1 induction have been observed already in low-grade gliomas.30 In the present study, only 3 of 31 astrocytomas of grade 2 exhibited less methionine uptake than normal contralateral tissue. With conventional histologic methods, we were not yet able to identify specific morphologic features. These tumors were not particularly small; thus, increased methionine uptake would have been detected with the resolution of our PET scanners.
The increased methionine uptake even in most low-grade gliomas provided valuable information for differentiation from nontumoral lesions with an accuracy of 79%. A similar observation was reported in eight patients for differentiation of neoplastic from non-neoplastic cerebral hematoma.31 In the present study, nontumoral lesions mainly included chronic or subacute ischemic or inflammatory demyelinating lesions that had been examined because tumors could not be excluded by the initial CT or MRI studies. Patients with acute cerebral infarcts or active inflammatory processes with high CSF cell counts were not included. For these processes, high uptake of 11C-methionine was described,32 and they cannot be distinguished reliably from gliomas with 11C-methionine PET. Thus, 11C-methionine contributes significantly to differentiation of gliomas from nontumoral lesions but is not a tumor-specific tracer.
Early detection of recurrent or residual tumor is of particular interest. Because contrast enhancement is often missing in low-grade gliomas, MRI and CT often cannot differentiate between tumor and nonspecific postoperative changes, unless a mass effect or distinct bloating of cortex or other gray matter structures is seen. The distinction may be important because radiation therapy may improve outcome in incompletely resected gliomas.33 Yet, with regard to potential long-term side effects of radiation therapy on cognitive functions,6 one should restrict this treatment to patients with evidence of residual or recurrent tumor. Most proven recurrent tumors had an even higher uptake of11 C-methionine than tumors before treatment (seetable 2). This was due to malignant transformation in three of them but was also observed without change of the histologic grade. A tendency toward increased uptake was also noted in patients in whom recurrent or residual tumor was suspected clinically. A decline of methionine uptake during efficient therapy with radioiodine seeds was demonstrated previously,34 suggesting that increased methionine uptake is probably not due to a nonspecific treatment effect. Thus, the present findings suggest a high sensitivity of methionine PET to detect residual or recurrent tumor and the potential to identify patients who could benefit most from additional radiation therapy.
We chose a rather simple way of quantifying tumor methionine uptake relative to contralateral tissue that is practical also in the clinical situation. Our previous finding of a K1-related effect that is linear over the measurement duration18 corresponds to our present finding of a very high correlation of r = 0.98 between 0- and 30-minute uptake ratios (average values, 2.37 ± 1.06) and between 20- and 60-minute uptake ratios (average values, 2.26 ± 1.10; not significant in paired t-test), which were assessed in a subset of 13 scans obtained on the HR scanner in three-dimensional data acquisition mode. This independence of uptake from measurement time is fortunate with respect to clinical practice. Because of the possibility that a very early measurement could be biased by significant intravascular activity, we now favor the 20- to 60-minute interval. With a dose of 20 mCi(740 MBq), 11C-methionine and three-dimensional data acquisition total noise-equivalent counts in the order of 50 × 106 can be collected during that interval, providing images with very low background noise. Therefore, it was not necessary to use more complex procedures with quantitative modeling of methionine uptake and metabolism, which are demanding with regard to plasma metabolite analysis and have not been successful because of the complex metabolic pathways of methionine.16
Footnotes
-
Received March 7, 1997. Accepted in final form December 17, 1997.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan