Rt

EMa

b

c

d

e

a

ARRAA

KTERM

1

rttgsCutr

0d

Clinical Neurology and Neurosurgery 112 (2010) 662–667

Contents lists available at ScienceDirect

Clinical Neurology and Neurosurgery

journa l homepage: www.e lsev ier .com/ locate /c l ineuro

adiation induced early necrosis in patients with malignant gliomas receivingemozolomide

mel Yamana,∗, Suleyman Buyukberberb, Mustafa Beneklib, Yusuf Onerc, Ugur Coskunb,uge Akmansud, Banu Ozturkb, Ali Osman Kayab, Dogan Uncue, Ramazan Yildizb

Department of Medical Oncology, Mersin State Hospital, Mersin, TurkeyDepartment of Medical Oncology, Gazi University, Faculty of Medicine, Ankara, TurkeyDepartment of Radiology, Gazi University, Faculty of Medicine, Ankara, TurkeyDepartment of Radiation Oncology, Gazi University, Faculty of Medicine, Ankara, TurkeyDepartment of Medical Oncology, Ankara Numune Hospital, Turkey

r t i c l e i n f o

rticle history:eceived 7 September 2009eceived in revised form 4 May 2010ccepted 8 May 2010vailable online 2 June 2010

eywords:emozolomidearly necrosisadiationRI spectroscopy

a b s t r a c t

Background: Temozolomide is the major drug in the treatment of malignant gliomas. Radiation inducednecrosis can behave radiologically and clinically like a recurrent tumor. The major problem is the dif-ferentiation between recurrence and radiation injury especially in early phases of treatment. The aimof this study was to evaluate the patients receiving temozolomide showing early clinical or radiologicalprogression and impact of early necrosis on follow-up.Patients and methods: We retrospectively evaluated medical records of 67 patients with malignant gliomareceiving temozolomide. All patients received concomitant radiotherapy and temozolomide followed byadjuvant temozolomide. In case of any radiological or clinical progression, MRI spectroscopy evaluationwas used to confirm tumoral progression.Results: Radiological or clinical progression was observed in 17 (25.4%) patients. Early radiation inducednecrosis was diagnosed in 4 of 17 patients (23.5%) by surgery (n = 3) and MRI spectroscopy (n = 1). Theobserved incidence of pseudoprogression was 4 in 67 (6%) patients. Patients with diagnosis of early

radiation injury had median progression-free survival of 7 months compared to 5 months in patientswithout radiation damage (p = 0.004). However, there was no statistically significant difference in termsof overall survival between groups.Conclusion: Temozolomide can cause early radiation induced injury which can mimic progressive tumor.Although the discrimination between two entities results in the accurate evaluation of response to ther-apy and benefits those patients, it did not affect overall survival. MRI spectroscopy is a valuable tool to

rosis

define early radiation nec

. Introduction

Treatment of malignant gliomas has dramatically changed inecent years after the introduction of temozolomide. Concurrentemozolomide with radiotherapy followed by 6 courses of adjuvantemozolomide has become the new standard of care in malignantliomas. It has been shown that this strategy results in a medianurvival of 14.6 months with a 2-year survival rate of 27% [17].

onventional magnetic resonance imaging (MRI) techniques aresed in the follow-up of the brain tumors in routine daily prac-ice. Although radiation necrosis is a common manifestation ofadiotherapy with an incidence of 3–24% [7,12], it has been increas-

∗ Corresponding author. Tel.: +90 324 3370955; fax: +90 324 3360256.E-mail address: emel.yaman@gmail.com (E. Yaman).

303-8467/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.clineuro.2010.05.003

and should be further evaluated in larger prospective studies.© 2010 Elsevier B.V. All rights reserved.

ingly recognized following temozolomide treatment concurrent toradiotherapy [1–4,6,19]. Moreover, the size of the lesions appearsbigger with necrosis leading to a false perception of progression,i.e., pseudoprogression. In a recent review, Brandsma et al. [2] ele-gantly described pseudoprogression as a continuum between thesubacute radiation reaction and treatment-related necrosis. In thisreport, we described our experience with early radiation inducednecrosis in patients with malignant gliomas receiving temozolo-mide concurrent with radiotherapy.

2. Patients and methods

2.1. Patients

Medical records of all patients with diagnosis of malignantglioma between April 2005 and December 2008 were retro-

and N

srawMtAiHna5Mrhanuacues

2

a6zp11

2

czwtb

TP

E. Yaman et al. / Clinical Neurology

pectively analyzed. All patients underwent maximum surgicalesection followed by radiotherapy concomitant to temozolomide,nd adjuvant temozolomide thereafter [21]. The extent of surgeryas evaluated using the postoperative Gd-enhanced T1-weightedRI taken 1 month after operation. Gross total resection of the

umor was defined as resection with no residual enhanced tumor.ll patients had histopathologic diagnosis of malignant glioma

ncluding grade III or IV astrocytoma based on the 2000 Worldealth Organization (WHO) scheme [11] and measurable intracra-ial disease by contrast-enhanced MRI of the cranium. Patientsged 18–75 years with Karnofsky performance status better than0 and life expectancy >3 months were included in this analysis.oreover, patients were also required to have adequate bone mar-

ow (white blood cell count > 3 × 109 l−1, platelets > 100 × 109 l−1,emoglobin > 10 g/dl), liver (total bilirubin < 2 mg/dl, aspartateminotransferase or alanine aminotransferase <3× upper limit oformal) and renal (blood urea nitrogen <30 mg/dl, creatinine <1.5×pper limit of normal) functions. Written consents were taken fromll patients. Patients with prior malignancies other than basal cellarcinoma or cervical carcinoma in situ, psychiatric disorder orncontrolled infections were excluded. Corticosteroid and anti-pileptic usage were allowed to control neurological signs andymptoms.

.2. Treatment

After initial surgical resection, standard cranial radiotherapy oftotal dose of 60 Gy at 2 Gy/day fractions, 5 days in a week duringweeks of period was given to all patients. Concurrent temo-

olomide was administered as 75 mg/m2 daily during radiotherapyrogram. After 4 weeks of break, adjuvant temozolomide was given50 mg/m2/day at the first cycle and then 200 mg/m2/day on days–5 every 28 days.

.3. Response evaluation

Disease activity was evaluated 1 month after completion of

hemoradiotherapy program and every 3 cycles of adjuvant temo-olomide. Early necrosis was defined as development of necrosisithin 6 months of completion of radiotherapy and concomitant

emozolomide treatment. To overcome any investigator relatedias, blinded radiographic review process was done. In case of

able 1atient characteristics.

Characteristic All patients (n = 67) (%) Patients wit

Median age, years (range, 23–74) 45 44

GenderMale 34 (50.8) 4 (100)Female 33 (49.2) 0

SurgeryGross total resection 51 (76.1) 3 (75)Subtotal resection 15 (22.4) 0Stereotactic biopsy 1 (1.5) 1 (25)

Tumor gradeIII 9 (13.4) 1 (25)IV 58 (86.6) 3 (75)

Karnofsky performance statusMedian 70 7080–100 20 (29.8) 2 (50)60–79 47 (70.2) 2 (50)

Diagnostic method for recurrenceCranial MRI 17 4 (23.5)MRI spectroscopy 14 (82.4) 1 (25)Histopathology 3 (17.6) 3 (75)

eurosurgery 112 (2010) 662–667 663

any clinical or radiological progression detected by conventionalMRI, magnetic resonance spectroscopic imaging (MRS) was done.Findings suggesting radiological progression on conventional MRIincluded increased contrast enhancement, mass effect, edema orsignal abnormalities on T2 weighted and FLAIR images. Multi-voxel 2D proton MRS was performed prior to the administrationof intravenous contrast agent with two different echo times(TE = 35 ms and 135 ms) with a 1.5 T MRI scanner (GE, Milwaukee,WI). MR spectroscopy was performed with a multi voxel techniquewhich enabled measurements from voxels including normal brainparenchyma and tumor tissue. For the multivoxel technique, voxelpositioning is made in order to avoid any contamination from adja-cent parenchyma, and therefore enables spectral analysis of thenormal parenchyma and tumoral area during the same examina-tion. Therefore this method is superior to a single voxel technique.The volume of interest (VOI) was defined to include the tumor areaas well as normal appearing brain tissue. To prevent strong con-tribution to the spectra from subcutaneous fat signals, the VOI wascompletely positioned within the brain. The metabolite peaks wereassigned as follows: choline (Cho) 3.22 ppm, creatin (Cr) 3.02 ppm,N-acetylaspartate (NAA) 2.02 ppm and mobile lipids 0.5–1.5 ppm.Lactate was identified at 1, 33 ppm by its characteristic doublet thatinverted at moderate TE. Cho/Cr, Cho/NAA and NAA/Cr ratios wereobtained from the lesion by a neuroradiologist experienced in MRS.To ensure quality control and acceptable quality of spectroscopicdata, normal values for NAA/Cr and Cho/Cr were also obtained innormal appearing white matter in the contralateral hemisphere.

2.4. Statistical analysis

Progression-free survival (PFS) was the primary objective ofthis study; secondary objective was overall survival (OS). PFSwas defined as the period from the start of the radiotherapy andconcomitant temozolomide treatment to the first observation ofdisease progression or death from any cause. OS time was definedas the period from the first day of radiotherapy and concomitanttemozolomide treatment until the date of the last follow-up or

death. All analyses were performed with the Statistical Package forSocial Sciences software version 11.5 (SPSS). The statistical compar-isons between groups were performed by using the Kaplan–Meiersurvival curves and log-rank test analysis. Two-tailed p-values lessthan 0.05 were considered statistically significant.

h early necrosis (n = 4) (%) Patients with true early progression (n = 13) (%)

54

7 (53.8)6 (46.2)

12 (92.3)1 (7.7)0

1 (7.7)12 (92.3)

705 (38.5)8 (61.5)

13 (76.4)13 (100)

0

664 E. Yaman et al. / Clinical Neurology and Neurosurgery 112 (2010) 662–667

Tab

le2

Des

crip

tion

ofd

emog

rap

hic

sin

pat

ien

tsw

ith

rad

ioth

erap

yin

du

ced

nec

rosi

s.

Pati

ent

Age

,yea

rsSe

xG

rad

eLo

cali

zati

onSu

rger

yR

TM

RIi

nte

rval

/Dxm

dos

age,

mg

MR

Ich

ange

sco

mp

ared

top

rior

MR

ISu

rviv

al,m

onth

s

Pre-

RT

Post

-RT

Post

-RT

At

3mon

ths

At

6m

onth

s

145

M4

Left

tem

por

opar

ieta

lmas

sG

ross

tota

lres

ecti

on60

Gy

5/4

4/4

SDSD

PD12

,die

d2

53M

3Le

ftte

mp

oral

mas

sB

iop

sy60

Gy

4/16

7/36

PDSD

SD14

,die

d3

23M

4R

igh

tfr

onto

par

ieta

lmas

sG

ross

tota

lres

ecti

on60

Gy

0/8

3/36

PDPR

PR9,

aliv

e4

44M

4Le

ftfr

onta

lmas

sG

ross

tota

lres

ecti

on50

Gy

8/16

4/28

PDSD

SD15

,die

d

RT:

rad

ioth

erap

y;D

XM

:d

exam

eth

ason

e;Pr

e-R

T,in

wee

ks:

inte

rval

betw

een

pre

-RT

MR

Ian

dbe

gin

nin

gof

RT;

Post

-RT,

inw

eeks

:in

terv

albe

twee

nen

dof

RT

and

firs

tp

ost-

RT

MR

I;PD

:p

rogr

essi

ved

isea

se;

PR:

par

tial

resp

onse

;SD

:st

able

dis

ease

;su

rviv

alis

mea

sure

din

mon

ths

from

the

firs

td

ayof

rad

ioth

erap

y.

Fig. 1. (a and b) Multivoxel MR spectroscopy performed with a TE of 135 ms showedreduced main brain metabolites together with elevated lipid and lactates, which isconsistent with radiation necrosis.

3. Results

Medical records of 67 patients with diagnosis of malignantglioma were retrospectively analyzed. Efficacy and toxicity ofthe temozolomide concurrent and adjuvant to radiotherapy werereported previously in 62 patients [17]. Five more patients treatedat the same time period were identified from the medical records.Median age of the patients was 45 years (range, 23–74) and 33(49.2%) of them were females. Characteristics of the patients weresummarized in Table 1. Nine patients (13.4%) had grade III astrocy-

toma and 58 (86.6%) glioblastoma multiforme (GBM). All patientsunderwent maximum surgical resection followed by temozolo-mide concomitant and adjuvant to radiotherapy.

In 17 patients (25.4%), progression was detected in early phasesof treatment (<6 months after completion of chemoradiotherapy).

E. Yaman et al. / Clinical Neurology and N

F

Ao“pdTadip(een

7gwww

ig. 2. Progression-free survival of patients with necrosis vs without necrosis.

ll patients were evaluated by MRS to confirm progression. Four outf 17 patients (23.5%) with progression were diagnosed as havingpseudoprogression” with MRS. Observed incidence of pseudo-rogression was calculated as 6% (4 of 67 patients). Clinical andemographic characteristics of the patients were summarized inable 2. In 3 of those, shift and/or uncal herniation were detectedfter termination of chemoradiation program leading to urgentecompression surgery. Pathological examination of surgical spec-

mens revealed radiation necrosis in these patients. One moreatient was diagnosed as having early radiation necrosis by MRSFig. 1a and b) which demonstrated reduced brain metabolites andlevated lipid–lactate levels. Strikingly, 3 of 4 (75%) patients witharly radiation necrosis had grade IV glioma. Median time for diag-osis of radiation necrosis was 5.5 weeks (range, 3–20).

The patients were followed for a median of 19 months (range,–29). Survival data were last updated in March 2009. In the whole

roup (n = 67), median PFS was 10 months in GBM group, while itas not reached in grade III astrocytoma group. Similarly, in thehole group, median OS was 19 months in GBM group, while itas not reached in grade III astrocytoma group. However, median

Fig. 3. Overall survival of patients with necrosis vs without necrosis.

eurosurgery 112 (2010) 662–667 665

PFS was 5 months in 17 patients with early progression; it was7 and 5 months in patients with and without radiation necrosis,respectively (p = 0.004) (Fig. 2). Median OS has not been reachedin patients with early necrosis vs 14 months in patients withoutnecrosis; however, this difference was not statistically significant(p > 0.05) (Fig. 3).

4. Discussion

Temozolomide is a new prospect in the treatment of malignantgliomas [10,14,17,21].

Nowadays, strategies to increase both response rates andresponse duration to temozolomide treatment are being devel-oped. However, treatment failures are common. Some of thesefailures have been shown to be so-called “pseudoprogression” dueto early radiation necrosis [1–4,6,19]. Recent reports pointed outthat temozolomide concurrent with radiotherapy may cause earlyradiation necrosis resulting in misperception of tumor progression[13]. Herein, we reported our experience with temozolomide-related early radiation injury which was interpreted as tumorprogression by conventional imaging techniques.

In our study, 4 of 17 (23.5%) patients with progressive diseasewere identified as having radiation necrosis. The incidence wascalculated as 6% in the whole cohort. Three of 4 patients withearly radiation induced necrosis were diagnosed by histopatho-logic examination following urgent decompression surgery. In all3 cases, there were no viable residual tumor cells in the speci-mens. Microscopically, radiation induced disseminated necrosis,hyalinization, vascular walls thickened with hyalinization and vas-cular bed changes were observed. In one patient, the diagnosis wasestablished with 1H-MRSI which showed reduced brain metabo-lites but increased lactate–lipids levels. The patients with grosstotal resection had higher incidence of early tumor recurrence.It may be related with two factors: firstly, our study had a smallnumber of patients. Secondly, there was an unequal distributionof patients between total resection and subtotal resection groups(76.1% vs 23.9%). Unfortunately, there were no enough data consid-ering the type of surgical procedures in patients with early tumorrecurrence in reports with pseudoprogression in literature. Moredata are needed in this era. Finally, our patients with early necrosishad a statistically significant longer PFS compared to patients with-out necrosis. Median OS of both groups were similar possibly dueto second-line chemotherapy following progression [22] or smallnumber of patients.

In the literature, there is limited data concerning pseudo-progression following temozolomide concurrent with radiation.Firstly, de Wit et al. [6] reported that, lesions mimicking progres-sive disease could be observed in 28.1% of patients within 3 monthsafter radiotherapy. Approximately 30% of these patients improvedor stabilized in 6 months without any treatment. Taal et al. [19]reported that pseudoprogression was observed in 21% of patientstreated with concurrent chemoradiotherapy with temozolomide.In a recent report, the pseudoprogression incidence was reportedas 31% [1]. Moreover, it was more common in patients with higherO6-methyl guanine-DNA methyl transferase (MGMT) methylationwhich indicated more favorable response to the temozolomidetreatment [1]. There is only one recent report of 54 patients ofwhom 3 (9%) were reported to develop pseudoprogression [4]. Theincidence of 6% pseudoprogression in our series appears to be lowerthan those reported in the literature. There are several reasons

to explain this discordance. First of all, our observed progressionrate was much lower than the studies reported in the literature[17/67 patients (25%) vs. 36/85 (42%) in Taal et al. [19] and 50/103(48.5%) in Brandes et al. [1]. Moreover, we think that the incidencereported in the literature may be too high because of the differ-

6 and N

ebpl1

gTMoNp

mmtoipMgrwrepitssmatnsaoorcaans1

ratacprvpNstpafPc(pc

a

[

[

[

[

[

[

[

[

[

[

[

66 E. Yaman et al. / Clinical Neurology

nces in the definition of pseudoprogression. Our study is uniqueecause of being the only study to utilize MRS to define pseudo-rogression. Interestingly, the other study by Chaskis et al. [4] with

ower pseudoprogression rate also utilized functional imaging with8F-FDG positron emission tomography (PET) to define pseudopro-ression. The criteria to define pseudoprogression in studies byaal et al. [19] and Brandes et al. [1] only involved conventionalRI, i.e., improved or stable lesions with stable/decreasing doses

f steroids and temozolomide treatment in the follow-up MRI.ew criteria need to be introduced using MRS in the diagnosis ofseudoprogression.

Tissue necrosis is a common problem in the follow-up ofalignant gliomas. As well as chemotherapy and radiotherapy,alignant glioma itself can also cause tissue necrosis and all these

hree entities may be confused with tumor recurrence. On thether hand, radiation induced necrosis is a common phenomenonn patients receiving cranial radiation and occurs in 3–24% ofatients 3–12 months following treatment [7,12]. By conventionalRI techniques, patients may be misdiagnosed as tumor pro-

ression because damage to the blood–brain barrier induced byadiation results in leakage of gadolinium into the interstitium,hich also produces a ring-enhancing lesion that can mimic tumor

ecurrence [5]. Positron emission tomography (PET), single-photonmission computed tomography (SPECT) and 1H-MRS especiallyroton magnetic spectroscopic imaging (1H-MRSI) can all be used

n distinguishing between tumor recurrence and necrosis. In addi-ion to non-invasive nature of the MRS procedure, compared tourgery, it may also help to clarify the diagnosis of pseudoprogres-ion. Moreover, this approach will also result with avoidance fromore psychological destabilization of the patients, who have alsofear of an early recurrence. 1H-MRS is an important diagnostic

ool when the tissues are composed of either pure tumor or pureecrosis; however, spectral patterns are less definitive when tis-ues composed of varying degrees of mixed tumor and necrosisre examined. Moreover, by proton MRS (1H-MRSI), measurementf brain metabolites like choline-containing phospholipids (Cho)r Cr may be helpful for discrimination of areas of both tumorecurrence and radiation injury. Although there are some reportslaiming that the sensitivity of MRS is in a range of 60–80% withspecificity of 80–90% [9], recent reports indicated that MRS is

s useful as pathologic examination in differentiation radiationecrosis from recurrence in brain tumors with a sensitivity of 94%,pecificity of 100% and diagnostic accuracy of 96.1% [18,23,24].H-MRSI can spectroscopically differentiate normal tissue fromadiation necrotic tissue, pure tumor tissue, and mixed specimenss well as pure tumor from all other tissue types in previouslyreated patients, but it cannot reliably differentiate mixed tumornd necrosis from pure necrosis [15]. 1H-MRSI has a significantlinical impact on the posttreatment evaluation of intracranial neo-lasms by means of following therapeutic response, identifyingesidual or recurrent tumor earlier than contrast-enhanced con-entional MRI and differentiating residual or recurrent tumor fromosttreatment abnormality. Elevation of Cho with depression ofAA is a reliable indicator of tumor in MRS [8]. There is exten-

ive data in the literature showing that Cho/Cr, NAA/Cr ratio andhe presence of lipid–lactates are useful in grading tumors andredicting tumor malignancy. Increases in lipid and lactate levelsre less specific than elevated Cho levels because they may resultrom both tumor tissue necrosis and radiation-induced necrosis.ostradiation and postchemotherapy necrotic areas are usuallyharacterized by absent or significantly reduced brain metabolites

NAA, Cho, Cr) and elevated lipid–lactate levels. Eventually a largeeak between 0 ppm and 2 ppm might also be observed indicatingell necrosis product [16,20].

In conclusion, subacute effects of concurrent chemoradiother-py with temozolomide may mimic tumor progression and it is

[

eurosurgery 112 (2010) 662–667

important for the treating medical oncologists, radiation oncol-ogists and neurologists to be alert to this problem. Interruptionof treatment due to pseudoprogression will result in undertreat-ment. Moreover, enrollment of these patients in clinical trials ofthe second-line treatment of gliomas may be misleading. Finally,MRS appears to be a valuable tool to define early radiationnecrosis and should be further evaluated in larger prospectivestudies.

References

[1] Brandes AA, Franceschi E, Tosoni A, Blatt V, Pession A, Tallini G, et al.MGMT promoter methylation status can predict the incidence and outcome ofpseudoprogression after concomitant radiochemotherapy in newly diagnosedglioblastoma patients. J Clin Oncol 2008;26:2192–7.

[2] Brandsma D, Stalpers L, Taal W, Sminia P, van den Bent MJ. Clinical features,mechanisms, and management of pseudoprogression in malignant gliomas.Lancet Oncol 2008;9:453–61.

[3] Chamberlain MC, Glantz MJ, Chalmers L, Van Horn A, Sloan AE. Early necrosisfollowing concurrent Temodar and radiotherapy in patients with glioblastoma.J Neurooncol 2007;82:81–3.

[4] Chaskis C, Neyns B, Michotte A, Michotte A, De Ridder M, Everaert H. Pseudo-progression after radiotherapy with concurrent temozolomide fpr high-gradeglioma: clinical observations and working recommendations. Surg Neurol2009;72(4):423–8.

[5] Covarrubias DJ, Rosen BR, Lev MH. Dynamic magnetic resonance perfusionimaging of brain tumors. Oncologist 2004;9:528–37.

[6] De Wit MC, de Bruin HG, Eijkenboom W, Sillevis Smitt PA, van den Bent MJ.Immediate post-radiotherapy changes in malignant glioma can mimic tumorprogression. Neurology 2004;63:535–7.

[7] Giglio P, Gilbert MR. Cerebral radiation necrosis. Neurologist 2003;9:180–8[Review].

[8] Graves EE, Nelson ST, Vigneron DB, Verhey L, McDermott M, Larson D, et al.Serial proton MR spectroscopic imaging of recurrent malignant gliomas aftergamma knife radiosurgery. AJNR 2001;22:613–4.

[9] Hollingworth W, Medina LS, Lenkinski RE, Shibata DK, Bernal B, ZurakowskiD, et al. A systematic literature review of magnetic resonance spectroscopyfor the characterization of brain tumors. AJNR Am J Neuroradiol 2006;27(7):1404–11.

10] Jeon HJ, Kong DS, Park KB, Lee JI, Park K, Kim JH, et al. Clinical out-come of concomitant chemoradiotherapy followed by adjuvant temozolomidetherapy for glioblastomas: single-center experience. Clin Neurol Neurosurg2009;111(8):679–82.

11] Kleihues P, Cavenee WK. Pathology and genetics of tumors of the nervous sys-tem. In: World Health Organization Classification of the Tumors of the NervousSystem, Editorial and Consensus Conference Working Group. Lyon, France:IARC Press; 2000.

12] Kumar AJ, Leeds NE, Fuller GN, Van Tassel P, Maor MH, Sawaya RE, etal. Malignant gliomas: MR imaging spectrum of radiation therapy- andchemotherapy-induced necrosis of the brain after treatment. Radiology2000;217:377–84.

13] Peca C, Pacelli R, Elefante A, Del Basso De Caro ML, Vergara P, Mariniello G,et al. Early clinical and neuroradiological worsening after radiotherapy andconcomitant temozolomide in patients with glioblastoma: tumour progressionor radionecrosis? Clin Neurol Neurosurg 2009;111(4):331–4.

14] Rhee DJ, Kong DS, Kim WS, Park KB, Lee JI, Suh YL, Song SY, et al. Efficacy oftemozolomide as adjuvant chemotherapy after postsurgical radiotherapy alonefor glioblastomas. Clin Neurol Neurosurg 2009;111(9):748–51.

15] Rock JP, Hearshen D, Scarpace L, Croteau D, Gutierrez J, Fisher JL, et al.Correlations between magnetic resonance spectroscopy and image-guidedhistopathology, with special attention to radiation necrosis. Neurosurgery2002;51:912–9.

16] Schlemmer HP, Bachert P, Herfarth KK, Zuna I, Debus J, van Kaick G. Proton MRspectroscopic evaluation of suspicious brain lesions after stereotactic radio-therapy. AJNR 2001;22:1316–24.

17] Stupp R, Mason WP, Van den Bent MJ, Weller M, Fisher B, Taphoorn MJ, et al.Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma.N Engl J Med 2005;352:987–96.

18] Taylor JS, Langston JW, Reddick WE, Kingsley PB, Ogg RJ, Pui MH, et al. Clinicalvalue of proton magnetic resonance spectroscopy for differentiating recurrentor residual brain tumor from delayed cerebral necrosis. Int J Radiat Oncol BiolPhys 1996;36:1251–61.

19] Taal W, Brandsma D, de Bruin HG, Bromberg JE, Swaak-Kragten AT, Smitt PA,et al. Incidence of pseudoprogression in a cohort of malignant glioma patientstreated with chemoirradiation with temozolomide. Cancer 2008;113:405–10.

20] Weybright P, Sundgren PC, Maly P, Hassan DG, Nan B, Rohrer S, et al. Dif-

ferentiation between brain tumor recurrence and radiation injury using MRspectroscopy. AJR Am J Roentgenol 2005;185:1471–6.

21] Yaman E, Buyukberber S, Uner A, Coskun U, Akmansu M, Benekli M, et al.Temozolomide in newly diagnosed malignant gliomas: administered concomi-tantly with radiotherapy, and thereafter as consolidation treatment. Onkologie2008;31:309–13.

and N

[

[

E. Yaman et al. / Clinical Neurology

22] Yaman E, Buyukberber S, Uner A, Coskun U, Yamac D, Ozturk B, et al. Carbo-platin and oral cyclophosphamide combination after temozolomide failure inmalignant gliomas. Tumori 2008;94:674–80.

23] Zeng QS, Li CF, Liu H, Zhen JH, Feng DC. Distinction between recurrent gliomaand radiation injury using magnetic resonance spectroscopy in combina-

[

eurosurgery 112 (2010) 662–667 667

tion with diffusion-weighted imaging. Int J Radiat Oncol Biol Phys 2007;68:151–8.

24] Zeng QS, Li CF, Zhang K, Liu H, Kang XS, Zhen JH. Multivoxel 3D proton MRspectroscopy in the distinction of recurrent glioma from radiation injury. Neu-rooncology 2007;84:63–9.