Molecular Markers in Low-Grade Gliomas: Predictive or ... · kers in patients with low-grade...
Transcript of Molecular Markers in Low-Grade Gliomas: Predictive or ... · kers in patients with low-grade...
Predictive Biomarkers and Personalized Medicine
Molecular Markers in Low-Grade Gliomas: Predictive or Prognostic?
Christian Hartmann1, Bettina Hentschel2, Marcos Tatagiba3, Johannes Schramm4, Oliver Schnell6,Clemens Seidel7, Robert Stein2, Guido Reifenberger8, Torsten Pietsch5, Andreas von Deimling1,Markus Loeffler2, and Michael Weller8 for the German Glioma Network
AbstractPurpose: To investigate whether TP53 mutation, 1p/19q codeletions, O6-methylguanylmethyltransfer-
ase (MGMT) promoter methylation, and isocitrate dehydrogenase 1 (IDH1) mutation predict natural
course of disease or response to radiotherapy or chemotherapy or both in low-grade glioma patients.
Experimental Design: Cohort A consisted of 89 patients with diffuse astrocytoma World Health
Organization (WHO) grade II (n ¼ 40), oligoastrocytoma (n ¼ 23), or oligodendroglioma (n ¼ 26) who
did not receive radiotherapy or chemotherapy after first operation and were monitored until progression
[progressive disease (PD); n¼ 59] and beyond or until the end of follow-up (n¼ 30). Cohort B consisted of
50 patients with WHO grade II gliomas who received radiotherapy or chemotherapy at diagnosis. Tumors
were analyzed for TP53 mutations, 1p/19q codeletions, MGMT promoter methylation, and IDH1
mutations.
Results: Median progression-free survival (PFS) in cohort A was 4.1 years (95% CI: 3.1–5.1). No
molecular marker was prognostic for PFS after surgery alone, using multivariate adjustment for histology,
age, and extent of resection. IDH1mutations were associated with prolonged survival from the diagnosis of
PD in oligoastrocytomas (OA II)/oligodendrogliomas (O II) and with overall survival (OS) in all tumors.
1p/19q codeletion and IDH1 mutation were prognostic for PFS and OS in cohort B.
Conclusions:Noneof theparametersaresensitiveprognosticbiomarkers inWHOgradeIIgliomapatients
whodonot receive radiotherapyorchemotherapyafter surgery.Limitationsof this study include theselection
of patientswith favorable outcome, the nonrandomized allocationof treatment, and the insufficient sample
size to distinguish between effects of radiotherapy versus chemotherapy. Regardless of histology, IDH1
mutation status is the strongest prognostic marker for OS. Clin Cancer Res; 17(13); 4588–99.�2011 AACR.
Introduction
Several molecular markers have been proposed as poten-tial predictors of outcome in patients with World HealthOrganization (WHO) grade II gliomas, which comprise
diffuse astrocytomas (A II), mixed oligoastrocytomas (OAII) and oligodendrogliomas (O II). The outcome has tra-ditionally been considered to bemore favorable in O II andless favorable in A II, whereas OA II has been attributed asan intermediate prognosis (1, 2). Some molecular aberra-tions are linked to histologic subtypes of WHO grade IIgliomas and may therefore be of diagnostic value, forexample, TP53 mutations are more common in A II,whereas combined deletions of 1p and 19q (1p/19q dele-tion) are more common in OA II and O II.
The clinical relevance of these molecular changes hasremained controversial. Thus, is has remained unclearwhether it is the 1p/19q deletion or the oligodendroglialmorphology that confers a less aggressive course of diseasethan A II. Moreover, within one histologic subtype ofWHOgrade II glioma, the clinical relevance of the molecularmarkers has remained controversial. Finally, it has provendifficult to distinguish prognostic significance, defined asoverall better outcome irrespective of management, frompredictive significance, defined as a better outcome provideda specific treatment is administered.
For instance, 1p/19q deletion resulting from an unba-lanced translocation (3, 4), preferentially in oligodendro-glial tumors, were first associated with sensitivity to
Authors' Affiliations: 1Clinical Cooperation Unit Neuropathology, GermanCancer Center & Department of Neuropathology, Institute of Pathology,Ruprecht-Karls University Heidelberg, Heidelberg; 2Institute of MedicinalInformatics, Statistics and Epidemiology, University of Leipzig, Leipzig;3Department of Neurosurgery, University of T€ubingen, T€ubingen; 4Depart-ment of Neurosurgery and 5Institute of Neuropathology, University ofBonn, Bonn; 6Department of Neurosurgery, University of Munich, Munich;7Department of Neuro-Oncology, University Hospital Heidelberg, Heidel-berg; 8Department of Neuropathology, Heinrich Heine UniversityD€usseldorf, D€usseldorf, Germany; and 8Department of Neurology, Uni-versity Hospital Zurich, Zurich, Switzerland
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
Corresponding Author: Christian Hartmann, Clinical Cooperation UnitNeuropathology, German Cancer Center & Ruprecht-Karls-Universit€atHeidelberg, Department of Neuropathology, Institute of Pathology,Im Neuenheimer Feld 220/221, D-69120 Heidelberg, Germany.Phone: 49-(0)6221-56-37884; Fax: 49-(0)6221-56-4566; E-mail:[email protected]
doi: 10.1158/1078-0432.CCR-10-3194
�2011 American Association for Cancer Research.
ClinicalCancer
Research
Clin Cancer Res; 17(13) July 1, 20114588
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
alkylating agent chemotherapy (5) and later to sensitivityto radiotherapy as well (6). Eventually, we reported that the1p/19q deletion loses its powerful prognostic impact ifpatients (with WHO grade II gliomas) receive no furtherradiotherapy or chemotherapy after surgery (7), makingthe 1p/19q deletion a candidate predictive marker forprolonged progression-free survival (PFS) and overall sur-vival (OS) in response to DNA-damaging treatments ingeneral. Yet, neuroradiologic monitoring neverthelessrevealed a slower spontaneous growth rate in untreated1p/19q-deleted low-grade gliomas than in untreated 1p/19q-nondeleted tumors (8).In contrast, the prognostic or predictive role for TP53
mutations has remained controversial and no consistentassociation with response to therapy or overall outcomehas been reported (9–12). Similarly, in contrast to thestrong predictive and prognostic role for methylation ofthe promoter region of the O6-methylguanylmethyltrans-ferase (MGMT) gene in newly diagnosed glioblastomapatients treated with temozolomide, no such role hasbeen defined in other gliomas (13). MGMT promotermethylation predicted a favorable outcome in WHOgrade III anaplastic gliomas treated with either alkylatingagent chemotherapy or radiotherapy (14, 15), and itssignificance in WHO grade II gliomas remains unclear(16).The identification of isocitrate dehydrogenase 1
(IDH1) mutations in a minority of glioblastomas andpilocytic astrocytomas, but in the majority of WHO gradeII and III gliomas (17–20), has dramatically altered ourconcepts of the origin and malignant progression ofgliomas. In particular, it has become clear that mostglioblastomas do not evolve from clinically silent low-grade gliomas; second, IDH1 mutations have been con-firmed as early mutations in low-grade gliomas (21),
which may facilitate gliomagenesis along different mor-phologic and molecular pathways, including astrocyto-mas more associated with TP53 mutations andoligodendrogliomas more associated with 1p/19q dele-tion. Moreover, IDH1 mutations are associated withyounger age and better outcome within each gliomaentity (22–25).
Here, we sought to evaluate the prognostic versuspredictive relevance of molecular markers in WHO gradeII gliomas and studied the 4 most prominent molecularmarkers thought to be involved in the development,progression, or both, of gliomas, TP53 mutation, 1p/19q deletion, MGMT promoter methylation, and IDH1mutation in the tumor tissue of patients who receivedneither radiotherapy nor chemotherapy after first surgery(cohort A). We asked whether the status of these markerscorrelated with the time to the diagnosis of progressivedisease (PD) and/or the first therapeutic reintervention,or with the survival from that reintervention. Results incohort A were compared with findings in a second setof patients (cohort B) who received radiotherapy orchemotherapy immediately after the first surgicalintervention.
Materials and Methods
Patient selectionThe centers of the German Glioma Network (GGN)
identified 89 patients with A II (n ¼ 40), OA II (n ¼ 23),or O II (n ¼ 26) operated on from 1991 to 2006 who hadnot received radiotherapy or chemotherapy after theirfirst operation (cohort A). These patients were monitoreduntil the end of follow-up or until therapeutic reinterven-tion, which was based on the local diagnosis of PDdocumented by neuroradiologic criteria that conformedto Macdonald criteria (26) for contrast-enhancing lesionsor similar adapted criteria for nonenhancing lesions andbeyond. For comparison and possible validation, westudied a second group of 50 patients with A II (n ¼38), OA II (n ¼ 7), or O II (n ¼ 5) who were operated onfrom 1991 to 2009 and were treated with radiotherapyalone (n ¼ 25), chemotherapy with alkylating agentsalone (n ¼ 21), or radiotherapy and alkylating agents(n ¼ 4) immediately after the establishment of the diag-nosis (cohort B). Individual data for both cohorts areprovided in Supplementary Table S1.
Central reference pathologyFormalin-fixed, paraffin-embedded material was sub-
mitted from the local (neuro)pathologists of the GGNcenters for an independent histopathologic review to theBrain Tumor Reference Center of the German Society ofNeuropathology and Neuroanatomy in Bonn. All tumorswere classified according to the WHO classification oftumors of the central nervous system (27). In case ofdifferences between local and central reference diagnoses,the reference diagnosis overruled the local diagnosis fordata analysis.
Translational Relevance
Numerous studies have tried to assess the prognosticvalue of the molecular markers O6-methylguanyl-methyltransferase (MGMT) promoter methylation,combined 1p/19q deletions, TP53 mutation, and iso-citrate dehydrogenase 1 (IDH1) mutation in patientswith low-grade gliomas. Currently, the marker statusdoes not influence the choice of established treatmentmodalities for patients with low-grade gliomas. More-over, it has remained unclear whether these markersreflect the natural course of disease or whether theypredict response to radio- or chemotherapy. Here, wereport that none of them are sensitive prognostic mar-kers in patients with low-grade gliomas who do notreceive adjuvant genotoxic treatment after surgery.IDH1 mutations assume a prognostic role when geno-toxic treatments are administered. IDH1 mutation and1p/19q deletion status allow to define 3 prognosticallydistinct subgroups of low-grade gliomas.
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4589
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Molecular studiesPrior to the extraction of DNA from tumor tissues by
standard methods, all tumor samples were examined by 2experienced neuropathologists (C.H., A.v.D.) to ensure anestimated tumor cell content of 80% or more. The techni-ques to determine TP53 mutations by single-strand con-firmation polymorphism analysis followed by directsequencing (28), MGMT promoter methylation by one-stage methylation-specific PCR (ref. 29; detection of a bandon agarose gel in the lane containing PCR products gen-erated by methylation-specific primers was scored as posi-tive), IDH1 or IDH2 mutations by direct sequencing (30),and 1p/19q deletion by multiplex ligation–dependentprobe amplification (7) have all been described elsewhere.Cohort B patients were not analyzed for IDH2mutations. Ifformalin-fixed, paraffin-embedded tissue was available,FISH was alternatively applied to determine 1p/19q dele-tion. The 2-color FISH assay was done on 5-mm-thicksections, using mixed 1p36/1q25 and 19p13/19q13dual-color probe sets (catalogue no. 32-231004; Vysis,Inc., Applied Biosystems). For slide pretreatment, probehybridization and posthybridization processing, the His-tology Accessory FISH Kit (Dakopatts) was used. Samplesshowing sufficient FISH efficiency (�90% nuclei withsignals) were evaluated. Signals were scored in at least200 nonoverlapping, intact nuclei. Deletions of 1p and19q were defined as 50% of tumor nuclei containing 1signal (31).
Statistical analysisThe association of clinical data and molecular markers
was analyzed by the c2 test and Fisher’s exact test. Becauseof the descriptive manner of these analyses, P values werenot adjusted for multiple testing. PFS, OS, and survivalfrom the first reintervention in cohort A were analyzed bylog-rank test. PFS was calculated from the day of firstsurgery until tumor progression, death, or end of follow-up. OS was calculated from the day of first surgery untildeath or end of follow-up. Survival from first reinterven-tion in cohort A was calculated from the diagnosis of PDuntil death or end of follow-up. Cox regression modelswere fitted to assess the independent impact of the mole-cular markers adjusting for age (>40 vs. �40), diagnosis(A II vs. pooled OA II and O II), and extent of resection(total vs. no total). Karnofsky performance score was notincluded for too many missing values. Data were analyzedby PASW Statistics 18 (version 18.0.0; SPSS, Inc.).
Results
Cohort A consisted of 60 males and 29 females. Themedian follow-up was 6.3 years. PD was diagnosed in 59patients, and 17 patients have died. The estimated medianPFS without radiotherapy or chemotherapy was 4.1 years(95% CI: 3.1–5.1). The median follow-up was 8.6 years forA II, 5.5 for OA II, and 6.1 for O II. PD was documented in75.0% of A II patients (30 of 40), 65.2% of OA II patients(15 of 23), and 53.8% of O II patients (14 of 26). After
diagnosis of PD, 5 patients had second surgery, 8 patientsreceived radiotherapy alone, 25 patients received radio-therapy and alkylating agents, and 9 patients receivedalkylating agents alone; 4 patients died shortly after PD,4 patients had follow-up of less than 3 months after PD,and 4 had no further treatment at a follow-up of 1.5 to 8.5years. The median OS was 15.5 years for all patients, 17.8years for pooled OA II and O II, and undetermined for A IIbecause the probability of OS was greater than 50% (Sup-plementary Fig. S1).
Cohort B consisted of 29 males and 21 females. Themedian follow-up was 3.8 years. The initial treatment wasradiotherapy alone in 25 patients, alkylating agent che-motherapy alone in 21 patients, and combined radioche-motherapy in 4 patients. PD was documented in 22patients, and 8 patients died. The estimated median PFSwith initial treatment was 6.1 years (95% CI: 2.1–10.1).The median follow-up was 3.1 years for A II patients and4.4 years for patients with OA II or O II. PD was docu-mented in 50% of A II patients (19 of 38) and 25% ofpatients with OA II or O II (3 of 12). The median OS was13.6 years for all patients, 8.1 years for A II patients, and13.6 years for OA II or O II patients.
Molecular dataThe demographic data and clinical features pooled by
both histology for cohort A patients and molecular markerstatus are summarized in Table 1. Three patients with IDH1mutations did not have the classical R132Hmutations: Wefound 3 R132Cmutations, 2 in OA II, and 1 in A II. Only 1patient with OA II in cohort A had an IDH2 mutation. Hisdata were analyzed within the group of patients with IDH1mutant tumors. Patients with TP53 and IDH1 mutationswere younger than patients without TP53 mutations. Theother molecular markers were not linked to age. No mole-cular marker was linked to gender. Oligodendroglial his-tology was associated with 1p/19q deletion and less so withMGMT promoter methylation, whereas A II was associatedwith TP53 mutation. Patients with TP53 mutant tumorshad more aggressive surgery than patients with TP53 wild-type tumors, an observation that is difficult to explain.Supplementary Table S2 shows these data for cohort B. Thetrends were similar to those for cohort A, with lowersignificance, presumably because of lower patient num-bers, except for the association of the 1p/19q deletion withOA II and O II.
All 4 molecular markers were informative in 81 of 89cohort A patients. Table 2 shows the interrelations betweenspecific molecular markers. None of 4 alterations werefound in 7 patients, 5 A II, and 2 OA II patients. IDH1wild-type tumors were found in 15 patients: 9 A II, 4 OA II,2 O II. Only one OA II patient had both TP53mutation and1p/19q deletion. A similar pattern based on larger sub-groups emerged when cohorts A and B and OA II and O IIwere pooled (Supplementary Table S3). IDH mutanttumors had more often 1p/19q codeletions (43.5% vs.26.9%, P ¼ 0.13), as well as more often TP53 mutations(29.4% vs. 15.4%, P ¼ 0.154) and MGMT promoter
Hartmann et al.
Clin Cancer Res; 17(13) July 1, 2011 Clinical Cancer Research4590
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Tab
le1.
Molec
ular
marke
rsbyag
e,ge
nder,histolog
y,an
dex
tent
ofrese
ctionin
coho
rtA
All
TP53
mutation
NoTP53
mutation
P1p
/19q
loss
No
1p/19q
loss
PMGMT
promoter
methy
lation
NoMGMT
promoter
methy
lation
PID
H1
mutation
NoID
H1
mutation
P
Age
,y
0.00
30.47
90.15
30.03
0
�40
54(60.7%
)21
(87.5%
)33
(52.4%
)20
(57.1%
)33
(64.7%
)19
(51.4%
)32
(66.7%
)48
(66.7%
)6(37.5%
)
>40
35(39.3%
)3(12.5%
)30
(47.6%
)15
(42.9%
)18
(35.3%
)18
(48.6%
)16
(33.3%
)24
(33.3%
)10
(62.5%
)
89(100
%)
24(100
%)
63(100
%)
35(100
%)
51(100
%)
37(100
%)
48(100
%)
72(100
%)
16(100
%)
Med
ian(ra
nge)
36.7
(17.4–
75.7)
33.3
(21.1–
51.2)
39.8
(17.4–
75.0)
36.2
(17.4–
66.9)
36.7
(19.1–
75.7)
39.4
(19.1–
65.1)
36.95(17.4–
75.0)
35.55(17.4–
75.0)
43.7
(25.3–
74.0)
Gen
der
0.44
80.10
90.24
31.00
0
Male
60(67.4%
)18
(75.0%
)41
(65.1%
)20
(57.1%
)38
(74.5%
)28
(75.7%
)30
(62.5%
)48
(66.7%
)11
(68.8%
)
Female
29(32.6%
)6(25.0%
)22
(34.9%
)15
(42.9%
)13
(25.5%
)9(24.3%
)18
(37.5%
)24
(33.3%
)5(31.3%
)
89(100
%)
24(100
%)
63(100
%)
35(100
%)
51(100
%)
37(100
%)
48(100
%)
72(100
%)
16(100
%)
Histologic
diagn
osis
0.03
2<0
.000
10.03
10.18
6
AII
40(44.9%
)16
(66.7%
)23
(36.5%
)5(14.3%
)32
(62.7%
)11
(29.7%
)27
(56.3%
)29
(40.3%
)10
(62.5%
)
OAII
23(25.8%
)5(20.8%
)18
(28.6%
)12
(34.3%
)11
(21.6%
)14
(37.8%
)8(16.7%
)19
(26.4%
)4(25.0%
)
OII
26(29.2%
)3(12.5%
)22
(34.9%
)18
(51.4%
)8(15.7%
)12
(32.4%
)13
(27.1%
)24
(33.3%
)2(12.5%
)
89(100
%)
24(100
%)
63(100
%)
35(100
%)
51(100
%)
37(100
%)
50(100
%)
72(100
%)
16(100
%)
Exten
tof
rese
ction
0.01
3a0.04
3a0.57
4a0.47
4a
Biopsy
only
8(9.0%
)0(0.0%
)7(11.1%
)3(8.6%
)4(7.8%
)2(5.4%
)5(10.4%
)5(6.9%
)2(12.5%
)
Partia
l
(<50
%)
10(11.2%
)2(8.3%
)8(12.7%
)5(14.3%
)4(7.8%
)6(16.2%
)4(8.3%
)7(9.7%
)3(18.8%
)
Sub
total
(50%
–99
%)
25(28.1%
)3(12.5%
)22
(34.9%
)15
(42.9%
)10
(19.6%
)11
(29.7%
)12
(25.0%
)22
(30.6%
)3(18.8%
)
Total
38(42.7%
)16
(66.7%
)21
(33.3%
)10
(28.6%
)28
(54.9%
)15
(40.5%
)22
(45.8%
)31
(43.1%
)7(43.8%
)
Not
available
8(9.0%
)3(12.5%
)5(7.9%
)2(5.7%
)5(9.8%
)3(8.1%
)5(10.4%
)7(9.7%
)1(6.3%
)
89(100
%)
24(100
%)
63(100
%)
35(100
%)
51(100
%)
37(100
%)
48(100
%)
72(100
%)
16(100
%)
Abbreviations
:AII,
astroc
ytom
aWHO
grad
eII;
OAII,
oligoa
strocy
tomaWHO
grad
eII;
OII,
oligod
endrogliomaWHO
grad
eII.
aPva
lues
wereca
lculated
onthebas
isof
evalua
ble
case
s.
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4591
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
methylation status (47.1% vs. 30.8%, P ¼ 0.142), than didIDH wild-type tumors. None of these associations reachedstatistical significance.
Survival analysesFigure 1 and Table 3 (part A) show PFS of cohort A
patients by histology and molecular markers. None ofthe molecular markers were prognostic for PFS for alltumors pooled or for single histologic entities. Therewere interesting trends for prolonged PFS in O II andOA II favoring the absence of MGMT promoter methyla-tion, but not the presence of 1p/19q deletion, and forprolonged PFS with IDH1 mutations in A II. Table 3(part B) shows that IDH1 mutations were associated withlonger survival from first PD or reintervention in thepooled group of patients with OA II or O II. Moreover,there was a trend for better survival with 1p/19q deletionin the entire cohort A and specifically in OA II/O IIpatients. Finally, the unfavorable trend of MGMT pro-moter methylation persisted. Sample sizes became toosmall to allow estimates on the role of specific treatmentsadministered at reintervention.
We also analyzed OS from the initial diagnosis as sum-marized in Table 3 (part C). Given the natural course ofWHO grade II gliomas, the number of events is still small.
However, IDH1 mutation emerged as the most powerfulparameter of outcome when comparing 5-year survivalrates.
For comparison and possible validation of the observa-tions in cohort A patients summarized in Table 3 (parts Band C), we also analyzed PFS and OS in cohort B patientswho were not observed after the first surgical interventionbut treated with radiotherapy, chemotherapy, or both.Similar to cohort A [Table 3 (part B)], both 1p/19q deletionand IDH1 mutation were associated with longer PFS inresponse to intervention [Table 4 (part A)]. Furthermore,1p/19q deletion and IDH1 mutation were associated withlonger OS [Table 4 (part B)]. Again, sample sizes were toosmall to compare specific treatments.
Multivariate analysesAn overall analysis of the distribution of molecular
changes and outcomes in our pooled sample of cohortsA and B allowed the distinction of 3 groups of patients:patients with IDH1 mutation and 1p/19q deletion, themajority of whom had OA II or O II; patients with IDH1mutation, but no 1p/19q deletion, the majority of whomhad A II; and patients without IDH1 mutation whosetumors corresponded to different histologies and whoshare a less favorable outcome (Fig. 2A–C).
Table 2. Patterns of molecular changes in cohort A WHO grade II gliomas by histology
Diagnosis All tumors
A II O AII O II
No molecular marker altered 5 2 0 714.3% 9.1% 0% 8.6%
. . ./IDH1/. . ./. . . 8 1 1 1022.9% 4.5% 4.2% 12.3%
. . ./IDH1/TP53/. . . 10 2 3 1528.6% 9.1% 12.5% 18.5%
. . ./. . ./. . ./MGMT 1 1 0 22.9% 4.5% 0% 2.5%
. . ./IDH1/. . ./MGMT 1 3 3 72.9% 13.6% 12.5% 8.6%
. . ./. . ./TP53/MGMT 1 0 0 12.9% 0% 0% 1.2%
. . ./IDH1/TP53/MGMT 4 2 0 611.4% 9.1% 0% 7.4%
1p/19q/. . ./. . ./. . . 1 0 2 32.9% 0% 8.3% 3.7%
1p/19q/IDH1/. . ./. . . 1 3 7 112.9% 13.6% 29.2% 13.6%
1p/19q/. . ./. . ./MGMT 1 1 0 22.9% 4.5% 0% 2.5%
1p/19q/IDH1/. . ./MGMT 2 6 8 165.7% 27.3% 33.3% 19.8%
1p/19q/IDH1/TP53/MGMT 0 1 0 10% 4.5% 0% 1.2%
All tumors 35 22 24 81
Hartmann et al.
Clin Cancer Res; 17(13) July 1, 2011 Clinical Cancer Research4592
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Figure 1. PFS in cohort A byhistology and molecular markers.A, AII versus OA II versus O II; B,age >40 versus �40; C, totalversus no total resection; D–F,PFS by TP53 mutation in A II, OAII/O II, and all tumors; G–I, PFS by1p/19q deletion in A II, OA II/O II,and all tumors; J–L, PFS byMGMTpromoter methylation in A II, OA II/O II, and all tumors; M–O, PFS byIDH1 mutations in A II, OA II/O II,and all tumors.
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
A II(n = 40)OA II(n = 23)O II(n = 26)
P = 0.345
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Age ≤ ≤ 40(n = 54)Age >40(n = 35)
P = 0.459
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
No total resection(n = 43)total resection(n = 38)
P = 0.275
A B C
D E F
G H I
J K L
M N O
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
TP53 wt(n = 23)TP53 mut(n = 16)
P = 0.373
A II OA II/O II All tumors
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
TP53 wt(n = 40)TP53 mut(n = 8)
P = 0.201
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
TP53 wt(n = 63)TP53 mut(n = 24)
P = 0.280
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
No Del 1p/19q(n = 32)Del 1p/19q(n = 5)
P = 0.171
A II OA II/O II All tumors
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
No Del 1p/19q(n = 19)Del 1p/19q(n = 30)
P = 0.652
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
No Del 1p/19q(n = 51)Del 1p/19q(n = 35)
P = 0.350
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
MGMT meth–(n = 27)MGMT meth+(n = 11)
P = 0.199
A II OA II/O II All tumors
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
MGMT meth–(n = 21)MGMT meth+(n = 26)
P = 0.132
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
MGMT meth–(n = 48)MGMT meth+(n = 37)
P = 0.947
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
IDH1 wt(n = 10)IDH1 mut(n = 29)
P = 0.067
A II OA II/O II All tumors
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
IDH1 wt(n = 6)IDH1 mut(n = 43)
P = 0.945
Years
0 1 2 3 4 5 6 7 8 9 10 11 12 130
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
IDH1 wt(n = 16)IDH1 mut(n = 72)
P = 0.127
PF
SP
FS
PF
SP
FS
PF
S
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4593
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Tab
le3.
Surviva
lana
lyse
sin
coho
rtA
TP53
1p/19q
MGMT
IDH1
Patients,
nEve
nts,
nMed
ian
PFS
(95%
CI)
TP53
mutation
Med
ian
PFS
,y
95%
CI
P1p
/19q
loss
Med
ian
PFS
,y
95%
CI
PMGMT
promoter
methy
lation
Med
ian
PFS
,y
95%
CI
PID
H1
mutation
Med
ian
PFS
,y
95%
CI
P
A.PFS
with
obse
rvationalon
e
AII
4030
3.6(2.5–4.6)
Yes
(12/16
)3.6
2.1–
5.0
0.37
3Yes
(2/5)
––
0.17
1Yes
(7/11)
4.6
2.6–
6.6
0.19
9Yes
(21/29
)4.3
2.6–
6.0
0.06
7
No(17/23
)3.6
1.0–
0.6.1
No(26/32
)4.4
3.4–
5.3
No(21/27
)3.5
1.7–
5.3
No(8/10)
2.2
2.0–
2.4
OAII
2315
4.1(3.9–4.3)
Yes
(3/5)
6.7
–0.73
0Yes
(7/12)
4.1
3.2–
5.1
0.88
0Yes
(10/14
)4.1
2.7–
5.5
0.94
3Yes
(12/19
)4.1
0.4–
7.9
0.31
0
No(12/18
)4.1
3.1–
5.1
No(8/11)
4.1
1.0–
7.3
No(5/8)
6.7
1.3–
12.1
No(3/4)
3.1
–
OII
2614
5.1(1.9–8.2)
Yes
(1/3)
––
0.06
1Yes
(10/18
)5.1
1.2–
9.0
0.39
7Yes
(9/12)
2.5
1.9–
3.1
0.07
2Yes
(14/24
)4.6
0.4–
8.9
0.29
5
No(12/22
)4.7
1.7–
7.6
No(4/8)
12.9
–No(5/13)
12.9
–No(0/2)
––
OAIIþ
OII
4929
4.6(2.2–7.1)
Yes
(4/8)
6.7
0.0–
13.3
0.20
1Yes
(17/30
)4.7
3.3–
6.0
0.65
2Yes
(19/26
)4.0
2.6–
5.4
0.13
2Yes
(26/43
)5.1
2.3–
7.8
0.94
5
No(24/40
)4.1
2.4–
5.9
No(12/19
)6.7
0.9–
12.5
No(10/21
)6.7
3.4–
9.9
No(3/6)
4.0
2.0–
6.0
Alltumors
8959
4.1(3.1–5.1)
Yes
(16/24
)4.6
2.1–
7.1
0.28
0Yes
(19/35
)4.7
3.4–
5.9
0.35
0Yes
(26/37
)4.1
2.4–
5.8
0.94
7Yes
(47/72
)4.5
4.0–
5.1
0.12
7
No(41/63
)4.1
2.9–
5.3
No(38/51
)3.6
2.4–
4.7
No(31/48
)4.2
3.0–
5.5
No(11/16
)2.5
1.1–
4.0
TP53
1p/19q
MGMT
IDH1
Patients,
n
Eve
nts,
n
Med
ian
survival
(95%
CI)
TP53
mutation
Med
ian,
y
95%
CI
P1p
/19q
loss
Med
ian,
y
95%
CI
PMGMT
promoter
methy
lation
Med
ian,
y
95%
CI
PID
H1
mutation
Med
ian,
y
95%
CI
P
B.Surviva
lfrom
first
PD
orreinterven
tion
AII
3010
8.9(–)
Yes
(3/12)
8.9
–0.35
8Yes
(1/2)
––
–Yes
(1/7)
5.3
–0.48
2Yes
(6/21)
––
0.27
6
No(6/17)
––
No(8/26)
8.9
–No(7/21)
––
No(3/8)
––
OAII
153
6.8(1.9–11
.8)
Yes
(1/3)
––
0.15
0Yes
(0/7)
––
–Yes
(1/10)
6.8
––
Yes
(2/12)
6.8
1.9–
11.8
–
No(2/12)
6.8
0.0–
15.2
No(3/8)
––
No(2/5)
––
No(1/3)
–
OII
144
10.5
(2.2–18
.8)
Yes
(1/1)
––
–Yes
(3/10)
10.5
––
Yes
(2/9)
––
–Yes
(4/14)
10.5
2.2–
18.8
–
No(3/12)
10.5
–No(1/4)
––
No(2/5)
––
No(–)
––
OAIIþ
OII
297
10.5
(4.2–16
.7)
Yes
(2/4)
3.4
–0.06
0Yes
(3/17)
10.5
1.9–
19.1
0.33
4Yes
(3/19)
––
0.14
5Yes
(6/26)
10.5
3.3–
17.6
0.02
4
No(5/24)
10.5
4.9–
16.0
No(4/12)
6.8
3.6–
10.0
No(4/10)
4.9
1.9–
8.0
No(1/3)
0.9
–
Alltumors
5917
10.5
(5.1–15
.8)
Yes
(5/16)
8.9
1.8–
15.9
0.94
5Yes
(4/19)
10.5
1.8–
19.1
0.74
1Yes
(4/26)
6.8
–0.19
0Yes
(12/47
)10
.55.1–
15.9
0.05
8
No(11/41
)10
.52.8–
18.2
No(12/38
)8.6
3.3–
14.4
No(11/31
)10
.53.5–
17.4
No(4/11)
––
TP53
1p/19q
MGMT
IDH1
Patients,
n
Eve
nts,
n
5-yOS
rate,%
,
(95%
CI)
TP53
mutation
5-yOS
rate,%
95%
CI
P1p
/19q
loss
5-yOS
rate,%
95%
CI
PMGMT
promoter
methy
lation
5-yOS
rate,%
95%
CI
PID
H1
mutation
5-yOS
rate,%
95%
CI
P
C.OS
AII
4010
84.7
(73.4–
96)
Yes
(3/16)
93.3
80.7–10
00.27
2Yes
(1/5)
80.0
44.9–10
00.70
1Yes
(1/11)
100.0
–0.33
8Yes
(6/29)
92.8
83.3–10
00.07
1
No(6/23)
82.6
67.1–98
.1No(8/32)
87.1
75.3–98
.9No(7/27)
80.8
65.6–96
No(3/10)
70.0
41.6–98
.4
OAIIþ
OII
497
92.0
(83.3–
100)
Yes
(2/8)
83.3
53.5–10
00.70
2Yes
(3/30)
95.0
85.4–10
00.89
0Yes
(3/26)
95.5
86.8–10
00.78
8Yes
(6/43)
93.6
85.1–10
00.26
7
No(5/40)
93.7
85.2–10
0No(4/19)
87.8
72–10
0No(4/21)
86.9
69.8–10
0No(1/6)
80.0
44.9–10
0
Alltumors
8917
88.9
(81.9–
95.8)Yes
(5/24)
90.5
77.9–10
00.49
9Yes
(4/35)
92.5
82.5–10
00.57
2Yes
(4/37)
97.0
91.1–10
00.28
9Yes
(12/72
)93
.487
.2–99
.70.02
4
No(11/63
)89
.781
.8–97
.5No(12/51
)87
.678
.2–96
.9No(11/48
)83
.672
.4–94
.8No(4/16)
74.0
52.1–96
.0
Hartmann et al.
Clin Cancer Res; 17(13) July 1, 2011 Clinical Cancer Research4594
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Tab
le4.
Surviva
lana
lyse
sin
coho
rtB
TP53
1p/19q
MGMT
IDH1
Patients,
nEve
nts,
nMed
ian
PFS
(95%
CI)
TP53
mutation
Med
ian
PFS
,y
95%
CI
P1p
/19q
loss
Med
ian
PFS
,y
95%
CI
PMGMT
promoter
methy
lation
Med
ian
PFS
,y
95%
CI
PID
H1
mutation
Med
ian
PFS
,y
95%
CI
P
A.PFS
AII
3819
5.1(1.8–8.4)
Yes
(3/6)
5.1
0.7–
9.4
0.68
2Yes
(1/8)
––
0.17
7Yes
(6/15)
6.7
–0.11
1Yes
(4/14)
6.7
–0.00
1
No(11/18
)3.1
0.2–
5.9
No(15/27
)3.1
0.9–
5.2
No(10/15
)1.3
–No(10/10
)1.0
0.3–
1.7
OAIIþ
OII
123
12.4
(–)
Yes
(0/1)
––
–Yes
(2/8)
6.1
–0.57
7Yes
(3/7)
––
–Yes
(1/6)
––
–
No(2/8)
––
No(1/4)
––
–No(0/3)
––
–No(2/4)
––
–
Alltumors
5022
6.1(2.1–10
.1)
Yes
(3/7)
5.1
0.8–
9.3
0.90
7Yes
(3/16)
6.1
0.0–
13.9
0.01
8Yes
(9/22)
6.1
0.7–
11.5
0.07
9Yes
(5/20)
6.7
1.6–
11.7
0.00
1
No(13/26
)6.4
0.5–
12.2
No(16/31
)3.1
0.6–
5.6
No(10/18
)1.9
0.0–
4.5
No(12/14
)1.2
0.0–
2.4
TP53
1p/19q
MGMT
IDH1
Patients,
n
Eve
nts,
n
5-yOSrate,
%(95%
CI)
TP53
mutation
5-yOS
rate,%
95%
CI
P1p
/19q
loss
5-yOS
rate,%
95%
CI
PMGMT
promoter
methy
lation
5-yOS
rate,%
95%
CI
PID
H1
mutation
5-yOS
rate,%
95%
CI
P
B.OS
AII
387
85.0
(72.5–
97.4)Yes
(2/6)
83.3
60.2–10
00.52
0Yes
(0/8)
100
–0.27
8Yes
(3/15)
92.3
77.8–10
00.16
8Yes
(1/14)
100
–0.05
4
No(4/18)
80.5
–No(6/27)
84.7
70.9–98
.5No(4/15)
70.0
44.5–95
.5No(5/10)
56.0
22.6–89
.4
OAIIþ
OII
121
100
Yes
(0/1)
100
––
Yes
(1/8)
100
––
Yes
(1/7)
100
––
Yes
(0/6)
100
––
No(0/8)
100
–No(0/4)
100
–No(0/3)
100
–No(1/4)
100
–
Alltumors
508
88.7
(79.2–
98.2)Yes
(2/7)
85.7
59.8–10
00.26
4Yes
(1/16)
50.0
0–10
00.04
3Yes
(4/22)
95.0
85.4–10
00.16
2Yes
(1/20)
50.0
0–10
00.03
4
No(4/26)
86.7
72.5–10
0No(6/31)
86.5
74.1–98
.8No(4/18)
74.7
52.6–96
.8No(6/14)
68.8
42.7–94
.8
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4595
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
Several models were built to assess prognostic factorsfor the time to PD, reintervention, or both, in cohort A.Each molecular marker was analyzed in separate modelsadjusted for age, diagnosis, and extent of resection. Nomolecular marker, not even IDH1 mutation, showed asignificant interrelation with time to PD or reinterven-tion (Fig. 2D). In further separate models for eachmolecular marker, OS was also analyzed in cohort A(Fig. 2E).
On the basis of the observed 3 groups of the combina-tion of IDH1 mutation and 1p/19q status, an indicatorvariable was built and analyzed in further multivariatemodels for time to PD or reintervention [Table 5 (partA)], survival from first PD or reintervention [Table 5 (partB)], and OS [Table 5 (part C)] in cohort A. Multivariateanalyses for cohort B are summarized in SupplementaryTable S4. Cohort A patients with IDH1 mutation and 1p/19q codeletion were considered as the reference group.Age and extent of resection were included in the model, aswell therapy after PD (yes vs. no) for the analysis ofsurvival from first PD or reintervention. Age older than40 years was associated with a relevant increased risk for
death regarding survival from first PD or reintervention[relative risk (RR) ¼ 2.5, P ¼ 0.155] and OS (RR ¼ 2.5, p¼ 0.147) but was not associated with time to PD (RR ¼1.0, P ¼ 0.891). Total resection showed a strong positiveeffect on survival from first PD or reintervention (RR ¼0.4, P ¼ 0.163) and OS (RR ¼ 0.3, P ¼ 0.090), and therewas a positive trend for total resection and time to PD,too. Radiotherapy and/or chemotherapy at PD wereassociated with a decreased risk of death (RR ¼ 0.4, P¼ 0.217).
The risk for PD was doubled (RR ¼ 2.1, P ¼ 0.067) forIDH1 wild-type status compared with IDH1 mutant statusand 1p/19q codeletion, and it increased moderately to 1.4(P ¼ 0.362) for IDH1 mutant status and no 1p/19qcodeletion. Compared with the reference category (IDH1mutant and 1p/19q codeletion), IDH1 wild-type had thestrongest effect on survival from PD or reintervention (RR¼ 3.0, P ¼ 0.183), followed by IDH1 mutant, but no 1p/19q codeletion (RR ¼ 1.4, P ¼ 0.694). Similar results wereobserved for analyses of OS. The risk of death stronglyincreased in relation to IDH1 mutation and 1p/19q code-letion for IDH1 wild-type status (RR ¼ 4.1, P ¼ 0.081) and
Cohort A – OS Cohort A – PFSA B C
D E FYears
0 2 4 6 8 10 12 14 16 18 200
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1 IDH1 mut and Del 1p/19q(n = 30)IDH1 mut and no Del 1p/19q(n = 40)IDH1wt(n = 16)
P = 0.292
Years
OS
0 2 4 6 8 10 12 14 16 18 200
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
IDH1 mut and Del 1p/19q(n = 39)IDH1 mut and no Del 1p/19q(n = 51)IDH1 wt(n = 30)
P = 0.001
Years
OS
0 2 4 6 8 10 12 14 16 18 200
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
IDH1 mut and Del 1p/19q(n = 30)IDH1 mut and no Del 1p/19q(n = 40)IDH1 wt(n = 16)
P = 0.038
TP53 mutation
1p/19q codeletion
MGMT methylation
IDH1 mutation
0 1 2 3
RR with 95% CI
4 5 6 0 1 2 3
RR with 95% CI
4 5 6 0 1 2 3
RR with 95% CI
4 5 6
PF
S
Cohort A + B – OS
Figure 2. Multivariate analysis of outcome. PFS (A) and OS (B) in cohort A and OS in pooled cohorts A þ B (C) in 3 low-grade glioma groups defined by IDH1mutation and 1p/19 codeletion status. D–F, forest plots for the corresponding 3 analyses of A–C.
Hartmann et al.
Clin Cancer Res; 17(13) July 1, 2011 Clinical Cancer Research4596
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
only moderately for IDH1 mutation and no 1p/19q code-letion (RR ¼ 1.4, P ¼ 0.648).In the pooled cohorts A þ B, each molecular marker was
also analyzed with adjustment for age, diagnosis, extent ofresection, and adjuvant therapy after first surgery (yes vs.no). MGMT, 1p/19q codeletion, and TP53 each showed anonsignificant association with OS. Only IDH1 mutationswere significantly associated with OS (RR¼ 0.3, P¼ 0.022;
Fig. 2F). Analysis of OS, including the combination ofIDH1 and 1p/19q status as indicator variables and adjustedfor age and extent of resection, showed a strong effect ofIDH1 wild-type status compared with IDH1 mutation and1p/19q codeletion (RR ¼ 8.0, P ¼ 0.004). For the compar-ison of IDH1 mutation, but no 1p/19q codeletion, withIDH1mutation and 1p/19q codeletion, an increased risk ofdeath (RR= 2.3, P¼ 0.260) was observed (data not shown).
Discussion
The number of clinical, histopathologic, and molecularprognostic markers to estimate the outcome of patientswith various types of gliomas, including low-grade gliomas,is steadily increasing (2, 32). In contrast, few studies havetried to distinguish markers that characterize the naturalcourse of disease from markers that predict PFS and OS inresponse to specific therapeutic measures. The observationuntil first PD of surgically treated patients followed withoutadjuvant radiotherapy, or chemotherapy is the only way todetermine whether a marker predicts outcome in theabsence of adjuvant DNA-damaging treatment and is thusa prognostic marker independent of radiotherapy andchemotherapy. For instance, 1p/19q deletion is stronglypredictive for prolonged PFS and OS in patients withanaplastic oligodendroglial tumors (WHO grade III)who are treated with radiotherapy or radiotherapy plusnitrosourea-based chemotherapy or temozolomide alone(14, 33, 34). Yet, 1p/19q deletion did not predict PFS inpatients, mostly with WHO grade II oligodendroglialtumors, who were treated with surgery alone (7), suggest-ing a link between this molecular marker and response togenotoxic therapies.
We here extend this observation and report that TP53mutation, 1p/19q deletion,MGMT promoter methylation,or IDH1 mutation is not a sensitive prognostic marker forPFS in patients with WHO grade II gliomas treated withsurgery alone [Table 3 (part A) and Fig. 1]. In contrast, 1p/19q deletion and IDH1 mutation assumed prognosticrelevance after reintervention in cohort A [Table 3 (partsB and C)] and were prognostic for PFS in cohort B patientswho were treated with radiotherapy, chemotherapy, orboth, already at diagnosis (Table 4).
IDH1 mutations have previously been linked toimproved OS but not to response to temozolomide atprogression after radiotherapy, in patients with low-gradeastrocytomas in a Dutch study (35), whereas a French studyobserved better response to temozolomide and better OS inpatients with IDH1mutant tumors (24).We find that IDH1mutant A II may have a less aggressive spontaneous beha-vior and observed that the differential outcome of IDH1mutant versus IDH1wild-type tumors becomesmuchmorevisible once tumor-specific treatment has been initiated[Tables 3 (parts B and C), 4, and 5, Supplementary Fig. S1].Moreover, we confirm that IDH1 wild-type low-grade glio-mas can be viewed as a distinct prognostic entity withinferior outcome (25). In contrast, the prognostic role ofIDH1 mutation in low-grade glioma patients has not been
Table 5.Multivariate analyses of survival to firstPD, survival from first PD or reintervention, andOS in cohort A
RR 95% CI P
A. Time to first PDAge
�40 1>40 1.0 0.5–1.8 0.891
ResectionNo total 1Total 0.7 0.4–1.3 0.298
Molecular markersIDH1 mutant and1p/19q loss
1
IDH1 mutant andno 1p/19q loss
1.4 0.7–2.6 0.362
IDH1 wild-type 2.1 0.9–4.8 0.067B. Time from first PDAge
�40 1 0.7–8.7 0.155>40 2.5
TherapyNo 1 0.1–1.8 0.217Yes 0.4
ResectionNo total 1 0.1–1.4 0.163Total 0.4
Molecular markersIDH1 mutant and1p/19q loss
1
IDH1 mutant and no1p/19q loss
1.4 0.3–6.3 0.694
IDH1 wild-type 3.0 0.6–14.5 0.183C. OSAge
�40 1 0.7–8.5 0.147>40 2.5
ResectionNo total 1 0.1–1.2 0.090Total 0.3
Molecular markersIDH1 mutant and1p/19q loss
1 0.3–6.3 0.648
IDH1 mutant and no1p/19q loss
1.4 0.8–20.2 0.081
IDH1 wild-type 4.1
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4597
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
confirmed in all series. In fact, the largest series publishedso far reported 2 unexpected findings; first, an associationof IDH1 mutations with older age and, second, a lack ofcorrelation with outcome (36).
The unexpected suggestion of possibly shortened PFS inpatients with WHO grade II oligodendroglial tumors withMGMT promoter methylation [Table 3 (part A)] has aprecedent in that a similar observation was made in asmall cohort of diffuse astrocytoma patients from Japan(16) and is unexpected in view of the favorable prognosticrole ofMGMT promoter methylation inWHO grade III andIV gliomas (13).
We acknowledge that our study has weaknesses. Thesample size for each entity was small, the design was inpart retrospective, and choice of treatment was not stan-dardized. There was no central neuroradiologic review toconfirm PD, and there were also no standardized criteria tomonitor disease progression in these tumors until recently(37). Cohort A is a selected group of patients with low-grade gliomas because the treating physicians consideredobservation after surgery a reasonable strategy, but we triedto compensate for that by including a cohort B for com-parison. Yet, this introduces a bias for the comparison ofboth groups because patients in cohort B were consideredby their physicians to require adjuvant radiotherapy orchemotherapy after surgery. On the other hand, OS waslong and comparable in both cohorts, indicating that theextent of bias between both cohorts is limited but that bothcohorts represent favorable groups of patients with low-grade gliomas.
Table 2 and Supplementary Table S3 illustrate the likelymolecular pathogenesis of low-grade gliomas. IDH1muta-tions are early and therefore presumably important lesions.They give rise to the formation of oligodendroglial tumorsin the face of 1p/19q codeletions and otherwise lead to A II.IDH1wild-type tumors are less well understood and share aless favorable prognosis, irrespective of histology (Table 5).Yet, our data indicate that none of the biomarkers studiedhere are sensitive predictors of PFS for gliomapatients in theabsence of genotoxic treatment whereas the profoundimpact of IDH1 mutations on OS was confirmed. In thisregard, our study may provide so far the strongest evidencefor differential responsiveness to genotoxic therapy of IDH1mutant versus IDH1 wild-type low-grade gliomas.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgments
The authors acknowledge the support of the GGN teams in all participat-ing centers.
Grant Support
The German Glioma Network is supported by the Deutsche Krebshilfe.The costs of publicationof this articlewere defrayed inpart by thepayment
of page charges. This article must therefore be hereby marked advertisementin accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received December 5, 2010; revised March 29, 2011; accepted April 27,2011; published OnlineFirst May 10, 2011.
References1. Karim AB, Maat B, Hatlevoll R, Menten J, Rutten EH, Thomas DG,
et al. A randomized trial on dose-response in radiation therapy of low-grade cerebral glioma: European Organization for Research andTreatment of Cancer (EORTC) Study 22844. Int J Radiat Oncol BiolPhys 1996;36:549–56.
2. Pignatti F, Van Den Bent M, Curran D, Debruyne C, Sylvester R,Therasse P, et al. Prognostic factors for survival in adult patients withcerebral low-grade glioma. J Clin Oncol 2002;20:2076–84.
3. Jenkins RB, Blair H, Ballman KV, Giannini C, Arusell RM, Law M, et al.A t(1;19)(q10;p10) mediates the combined deletions of 1p and 19q andpredicts a better prognosis of patients with oligodendroglioma. Can-cer Res 2006;66:9852–61.
4. Griffin CA, Burger P, Morsberger L, Yonescu R, Swierczynski S,Weingart JD, et al. Identification of der(1;19)(q10;p10) in five oligo-dendrogliomas suggests mechanism of concurrent 1p and 19q loss. JNeuropathol Exp Neurol 2006;65:988–94.
5. Cairncross JG, Ueki K, Zlatescu MC, Lisle DK, Finkelstein DM,Hammond RR, et al. Specific genetic predictors of chemotherapeuticresponse and survival in patients with anaplastic oligodendrogliomas.J Natl Cancer Inst 1998;90:1473–9.
6. Bauman GS, Ino Y, Ueki K, Zlatescu MC, Fisher BJ, Macdonald DR,et al. Allelic loss of chromosome 1p and radiotherapy plus chemother-apy in patients with oligodendrogliomas. Int J Radiat Oncol Biol Phys2000;48:825–30.
7. Weller M, Berger H, Hartmann C, Schramm J, Westphal M, Simon M,et al. Combined 1p/19q loss in oligodendroglial tumors: predictive orprognostic biomarker?Clin Cancer Res 2007;13:6933–7.
8. Ricard D, Kaloshi G, Amiel-Benouaich A, Lejeune J, Marie Y, Man-donnet E, et al. Dynamic history of low-grade gliomas before and aftertemozolomide treatment. Ann Neurol 2007;61:484–90.
9. Okamoto Y, Di Patre PL, Burkhard C, Horstmann S, Jourde B, FaheyM, et al. Population-based study on incidence, survival rates, andgenetic alterations of low-grade diffuse astrocytomas and oligoden-drogliomas. Acta Neuropathol 2004;108:49–56.
10. Peraud A, Kreth FW, Wiestler OD, Kleihues P, Reulen HJ. Prognosticimpact of TP53 mutations and P53 protein overexpression in supra-tentorial WHO grade II astrocytomas and oligoastrocytomas. ClinCancer Res 2002;8:1117–24.
11. Stander M, Peraud A, Leroch B, Kreth FW. Prognostic impact of TP53mutation status for adult patients with supratentorial World HealthOrganization Grade II astrocytoma or oligoastrocytoma: a long-termanalysis. Cancer 2004;101:1028–35.
12. Watanabe K, Sato K, BiernatW, Tachibana O, von AmmonK, Ogata N,et al. Incidence and timing of p53 mutations during astrocytomaprogression in patients with multiple biopsies. Clin Cancer Res1997;3:523–30.
13. Weller M, Stupp R, Reifenberger G, Brandes AA, Van Den Bent MJ,Wick W, et al. MGMT promoter methylation in malignant gliomas:ready for personalized medicine? Nat Rev Neurol 2010;6:39–51.
14. Wick W, Hartmann C, Engel C, Stoffels M, Felsberg J, StockhammerF, et al. NOA-04 randomized phase III trial of sequential radio-chemotherapy of anaplastic glioma with procarbazine, lomustine,and vincristine or temozolomide. J Clin Oncol 2009;27:5874–80.
15. Van Den Bent MJ, Dubbink HJ, Sanson M, Van Der Lee-Haarloo CR,Hegi M, Jeuken JW, et al. MGMT promoter methylation is prognosticbut not predictive for outcome to adjuvant PCV chemotherapy inanaplastic oligodendroglial tumors: a report from EORTC Brain TumorGroup Study 26951. J Clin Oncol 2009;27:5881–6.
Hartmann et al.
Clin Cancer Res; 17(13) July 1, 2011 Clinical Cancer Research4598
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
16. Komine C, Watanabe T, Katayama Y, Yoshino A, Yokoyama T,Fukushima T. Promoter hypermethylation of the DNA repair geneO6-methylguanine-DNA methyltransferase is an independent predic-tor of shortened progression free survival in patients with low-gradediffuse astrocytomas. Brain Pathol 2003;13:176–84.
17. Balss J, Meyer J, Mueller W, Korshunov A, Hartmann C, von DeimlingA. Analysis of the IDH1 codon 132 mutation in brain tumors. ActaNeuropathol 2008;116:597–602.
18. Parsons DW, Jones S, Zhang X, Lin JC, Leary RJ, Angenendt P, et al.An integrated genomic analysis of human glioblastoma multiforme.Science 2008;321:1807–12.
19. Weller M, Felsberg J, Hartmann C, Berger H, Steinbach JP, SchrammJ, et al. Molecular predictors of progression-free and overall survival inpatients with newly diagnosed glioblastoma: a prospective transla-tional study of the German Glioma Network. J Clin Oncol2009;27:5743–50.
20. Korshunov A, Meyer J, Capper D, Christians A, RemkeM,Witt H, et al.Combined molecular analysis of BRAF and IDH1 distinguishes pilo-cytic astrocytoma from diffuse astrocytoma. Acta Neuropathol2009;118:401–5.
21. Watanabe T, Nobusawa S, Kleihues P, Ohgaki H. IDH1 mutations areearly events in the development of astrocytomas and oligodendro-gliomas. Am J Pathol 2009;174:1149–53.
22. Sanson M, Marie Y, Paris S, Idbaih A, Laffaire J, Ducray F, et al.Isocitrate dehydrogenase 1 codon 132 mutation is an importantprognostic biomarker in gliomas. J Clin Oncol 2009;27:4150–4.
23. Yan H, Parsons DW, Jin G, McLendon R, Rasheed BA, Yuan W, et al.IDH1 and IDH2 mutations in gliomas. N Engl J Med 2009;360:765–73.
24. Houillier C, Wang X, Kaloshi G, Mokhtari K, Guillevin R, Laffaire J, et al.IDH1 or IDH2 mutations predict longer survival and response totemozolomide in low-grade gliomas. Neurology 2010;75:1560–6.
25. Metellus P, Coulibaly B, Colin C, de Paula AM, Vasiljevic A, Taieb D,et al. Absence of IDH mutation identifies a novel radiologic andmolecular subtype of WHO grade II gliomas with dismal prognosis.Acta Neuropathol 2010;120:719–29.
26. Macdonald DR, Cascino TL, Schold SC Jr., Cairncross JG. Responsecriteria for phase II studies of supratentorial malignant glioma. J ClinOncol 1990;8:1277–80.
27. Louis DN. International Agency for Research on Cancer WHO classi-fication of tumours of the central nervous system. 4th ed. Lyon,France: International Agency for Research on Cancer; 2007.
28. MuellerW, HartmannC, Hoffmann A, LankschW, Kiwit J, Tonn J, et al.Genetic signature of oligoastrocytomas correlates with tumor locationand denotes distinct molecular subsets. Am J Pathol 2002;161:313–9.
29. Esteller M, Garcia-Foncillas J, Andion E, Goodman SN, Hidalgo OF,Vanaclocha V, et al. Inactivation of the DNA-repair gene MGMT andthe clinical response of gliomas to alkylating agents. N Engl J Med2000;343:1350–4.
30. Hartmann C, Meyer J, Balss J, Capper D, Mueller W, Christians A,et al. Type and frequency of IDH1 and IDH2 mutations are related toastrocytic and oligodendroglial differentiation and age: a study of1,010 diffuse gliomas. Acta Neuropathol 2009;118:469–74.
31. Seiz M, Tuettenberg J, Meyer J, Essig M, Schmieder K, Mawrin C,et al. Detection of IDH1 mutations in gliomatosis cerebri, but only intumors with additional solid component: evidence for molecularsubtypes. Acta Neuropathol 2010;120:261–7.
32. Riemenschneider MJ, Jeuken JW, Wesseling P, Reifenberger G.Molecular diagnostics of gliomas: state of the art. Acta Neuropathol2010;120:567–84.
33. Cairncross G, Berkey B, Shaw E, Jenkins R, Scheithauer B, BrachmanD, et al. Phase III trial of chemotherapy plus radiotherapy comparedwith radiotherapy alone for pure and mixed anaplastic oligodendro-glioma: Intergroup Radiation Therapy Oncology Group Trial 9402. JClin Oncol 2006;24:2707–14.
34. Van Den Bent MJ, Carpentier AF, Brandes AA, Sanson M,Taphoorn MJ, Bernsen HJ, et al. Adjuvant procarbazine, lomustine,and vincristine improves progression-free survival but not overallsurvival in newly diagnosed anaplastic oligodendrogliomas andoligoastrocytomas: a randomized European Organisation forResearch and Treatment of Cancer phase III trial. J Clin Oncol2006;24:2715–22.
35. Dubbink HJ, Taal W, van Marion R, Kros JM, van Heuvel I, BrombergJE, et al. IDH1 mutations in low-grade astrocytomas predict survivalbut not response to temozolomide. Neurology 2009;73:1792–5.
36. Kim YH, Nobusawa S, Mittelbronn M, Paulus W, Brokinkel B, KeyvaniK, et al. Molecular classification of low-grade diffuse gliomas. Am JPathol 2010:177:2708–14.
37. Wen PY, Macdonald DR, Reardon DA, Cloughesy TF, Sorensen AG,Galanis E, et al. Updated response assessment criteria for high-gradegliomas: response assessment in Neuro-Oncology Working Group. JClin Oncol 2010;28:1963–72.
Molecular Markers in Low-Grade Gliomas
www.aacrjournals.org Clin Cancer Res; 17(13) July 1, 2011 4599
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194
2011;17:4588-4599. Published OnlineFirst May 10, 2011.Clin Cancer Res Christian Hartmann, Bettina Hentschel, Marcos Tatagiba, et al. Prognostic?Molecular Markers in Low-Grade Gliomas: Predictive or
Updated version
10.1158/1078-0432.CCR-10-3194doi:
Access the most recent version of this article at:
Material
Supplementary
http://clincancerres.aacrjournals.org/content/suppl/2013/05/15/1078-0432.CCR-10-3194.DC1Access the most recent supplemental material at:
Cited articles
http://clincancerres.aacrjournals.org/content/17/13/4588.full#ref-list-1
This article cites 36 articles, 16 of which you can access for free at:
Citing articles
http://clincancerres.aacrjournals.org/content/17/13/4588.full#related-urls
This article has been cited by 6 HighWire-hosted articles. Access the articles at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
SubscriptionsReprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's
.http://clincancerres.aacrjournals.org/content/17/13/4588To request permission to re-use all or part of this article, use this link
Research. on June 22, 2020. © 2011 American Association for Cancerclincancerres.aacrjournals.org Downloaded from
Published OnlineFirst May 10, 2011; DOI: 10.1158/1078-0432.CCR-10-3194