CLINICAL SURVEY Silvi Scoccinti, MD Rdioterpi, Università di Firenze, Aziend Ospedlier Universitri Creggi, Firenze, Itly Stefno Mri Mgrini, MD Istituto del Rdio O. Alberti, Diprtimento di Rdioterpi Oncologic, Università di Bresci, Bresci, Itly Umberto Ricrdi, MD Rdioterpi Oncologic, Universit di Torino, Aziend Ospedlier Universitri Sn Giovnni Bttist, Torino, Itly Betrice Detti, MD Rdioterpi, Università di Firenze, Aziend Ospedlier Universitri Creggi, Firenze, Itly Michel Buglione, MD Istituto del Rdio O. Alberti, Diprtimento di Rdioterpi Oncologic, Università di Bresci, Bresci, Itly Guido Sotti, MD Istituto Oncologico Veneto IRCCS, Pdov, Itly Mrco Krengli, MD Rdioterpi, Università del Piemonte Orientle, Novr, Itly Sergio Mlut, MD Rdioterpi Oncologic, Aziend Ospedlier di Veron, Veron, Itly Slvtore Prisi, MD Rdioterpi, IRCCS Ospedle Cs Sollievo dell Sofferenz, Sn Giovnni Rotondo, Itly Filippo Bertoni, MD Rdioterpi Oncologic AOU Policlinico di Moden, Moden, Itly Cristin Mntovni, MD Rdioterpi Oncologic, Universit di Torino, Aziend Ospedlier Universitri Sn Giovnni Bttist, Torino, Itly Vincenzo Tombolini, MD Cttedr di Rdioterpi, Fcoltà Medicin e Chirurgi, Università de L Aquil, L Aquil, Itly Costntino De Renzis, MD Rdioterpi Oncologic, Diprtimento di Scienze Rdiologiche, AOU Policlinico G. Mrtino, Messin, Itly Mrco Lioce, MD Rdioterpi Oncologic, A.O.R.N. G. Rummo, Benevento, Itly Luci Ftignte, MD Rdioterpi Oncologic, AOU Pisn, Pis, Itly Vincenzo Fusco, MD Rdioterpi, CROB-IRCCS, Rionero in Vulture, Itly Polo Muto, MD Rdioterpi, P.O. Asclesi, Nples, Itly Frnco Berti, MD Istituto Oncologico Veneto IRCCS, Pdov, Itly Giovnni Rubino, MD Rdioterpi Oncologic, Diprtimento di Ptologi umn e Oncologi, Università di Sien, Sien, Itly Smnth Cipressi, MD Rdioterpi, Università di Firenze, Aziend Ospedlier Universitri Creggi, Firenze, Itly Lur Friselli, MD Rdioterpi, Fondzione Istituto Neurologico Crlo Best, Miln, Itly Mrco Lupttelli, MD Rdioterpi Oncologic, Università di Perugi, Ospedle Snt Mri Misericordi, Perugi, Itly Riccrdo Sntoni, MD Rdioterpi, Università Tor Vergt, Rome, Itly Luigi Pirtoli, MD Rdioterpi Oncologic, Diprtimento di Ptologi umn e Oncologi, Università di Sien, Sien, Itly Ptterns of Cre nd Survivl in Retrospective Anlysis of 1059 Ptients With Glioblstom Multiforme Treted Between 2002 nd 2007: A Multicenter Study by the Centrl Nervous System Study Group of Airo (Itlin Assocition of Rdition Oncology) OBJECTIVE: To investigte the pttern of cre nd outcomes for newly dignosed glioblstom in Itly nd compre our results with the previous Itlin Ptterns of Cre study to determine whether significnt chnges occurred in clinicl prctice during the pst 10 yers. METHODS: Clinicl, pthologicl, therpeutic, nd survivl dt regrding 1059 ptients treted in 18 rdiotherpy centers between 2002 nd 2007 were collected nd retrospectively reviewed. RESULTS: Most ptients underwent both computed tomogrphy nd mgnetic resonnce imging either preopertively (62.7%) or postopertively (35.5%). Only 123 ptients (11.6%) underwent biopsy. Rdiochemotherpy with temozolomide ws the most frequent djuvnt tretment (70.7%). Most ptients (88.2%) received 3-dimensionl conforml rdiotherpy. Medin survivl ws 9.5 months. Two- nd 5-yer survivl rtes were 24.8% nd 3.9%, respectively. Multivrite nlysis showed the sttisticl significnce of ge, postopertive Krnofsky Performnce Sttus scle score, surgicl extent, use of 3-dimensionl conforml rdiotherpy, nd use of chemotherpy. Use of more ggressive pproch ws ssocited with longer survivl in elderly ptients. Compring our results with those of the subgroup of ptients included in our previous study who were treted between 1997 nd 2001, relevnt differences were found: more frequent use of mgnetic resonnce imging, surgicl removl more common thn biopsy, nd widespred use of 3-dimensionl conforml rdiotherpy + temozolomide. Furthermore, significnt improvement in terms of survivl ws noted (P <.001). CONCLUSION: Chnges in the cre of glioblstom over the pst few yers re documented. Prognosis of glioblstom ptients hs slightly but significntly improved with smll but noteworthy number of reltively long-term survivors. KEY WORDS: Chemotherpy, Dtbse, Glioblstom, Ptterns of cre, Rdiotherpy Neurosurgery 67:446-458, 2010 DOI: 10.1227/01.NEU.0000371990.86656.E8 www.neurosurgery- online.com The gol of the Centrl Nervous System Study Group of AIRO (Itlin Assocition of Rdi tion Oncology), founded in 2005, Gimpolo Biti, MD Rdioterpi, Università di Firenze, Aziend Ospedlier Universitri Creggi, Firenze, Itly Reprint requests: Silvi Scoccinti, MD, Rdioterpi, Università di Firenze, Aziend Ospedlier Universitri Creggi, Vile Morggni 85, 50134 Florence, Itly. E-mil: silvi.scoccinti@unifi.it Received, June 22, 2009. Accepted, My 1, 2010. Copyright 2010 by the Congress of Neurologicl Surgeons is to promote clinicl reserch on the mngement of brin tumors nd to develop educ- ABBREVIATIONS: 3D-CRT, 3-dimensionl conforml rdiotherpy; EORTC/NCIC, Europen Orgniztion for Reserch nd Tretment of Cncer nd the Ntionl Cncer Institute of Cnd; GBM, glioblstom multiforme; GOP, Gliom Outcomes Project; KPS, Krnofsky Performnce Sttus; RPA, recursive prtitioning nlysis; RT, rdiotherpy; TMZ, temozolomide 446 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA TABLE 1. Rdition Therpy Oncology Group Acute Rdition Morbidity Scoring Criteri for the Centrl Nervous System Grde 0 Grde 1 Grde 2 Grde 3 Grde 4 None Fully functionl sttus (ie, ble to work) with minor neurologicl findings; no mediction needed Neurologicl findings requiring hospitliztion for initil mngement Neurologicl findings present sufficient to require home cre; nursing ssistnce my be required; medictions including steroids nd ntiseizure gents my be required Serious neurologicl impirment tht includes prlysis, com, nd seizures >3 per week despite mediction; hospitliztion required tionl progrms for improving the cre of neuro-oncologicl ptients. The group collected clinicl dt on ptients with high-grde glioms treted in rdition oncology centers to perform ptterns of cre study, evluting the ntionl stndrds of prctice during the most recent yers. The clinicl presenttions, surgicl mngement, nd djuvnt therpies for 1374 ptients with highgrde glioms including glioblstom multiforme (GBM) nd nplstic strocytom treted in Itly between 2002 nd 2007 were recorded using common dtbse. The current study focuses on ptients with newly dignosed GBM. For mny yers, surgery nd djuvnt rdiotherpy (RT) hve been the minsty of therpy for GBM. After publiction of the results of severl clinicl studies nd, more recently, of lrge rndomized multicenter phse III tril performed by EORTC/NCIC (Europen Orgniztion for Reserch nd Tretment of Cncer nd the Ntionl Cncer Institute of Cnd), 1 concomitnt nd sequentil temozolomide (TMZ) becme the stndrd of cre for ptients with GBM. The medin survivl of ptients with GBM ws 14.6 months in the recently published updted results of tht tril. 2 In this study, we investigted whether ptterns of prctice in Itly hve chnged in the recent pst. We previously reported the results of the Ptterns of Cre study of 1722 dult strocytom ptients (GBM, n = 1235; nplstic strocytom, n = 309; low-grde glioms, n = 178) treted between 1985 nd 2001. 3 In the current study, second ntionl survey ws done to compre current tretment modlities nd outcomes of glioblstom ptients with dt from the erlier Itlin Ptterns of Cre study. In this retrospective dtbse-driven study of 1059 GBM ptients treted in 18 RT centers between 2002 nd 2007 we intended to provide comprison with the previous Itlin Ptterns of Cre study, clling ttention to the chnges in tretment of glioblstom over the time period in which dt on TMZ efficcy were widely disseminted. METHODS AND MATERIALS The Centrl Nervous System Study Group of AIRO designed dtbse to collect clinicl dt on ptients with GBM. Severl meetings were held with the prticipting centers to ensure the homogeneity of dt collection. The coordinting study center (University of Florence) creted nd mnged the centrl computerized dtbse. Ech prticipting center sent dt forms to the centrl coordinting center. Queries could be e-miled from prticipnts to the registry stff for ssistnce nd resolution of the problems. Moreover, periodic meetings with the investigtors were rrnged to offer ssistnce in completing the forms. Audits were performed to find omissions or inconsistencies, nd registry stff contcted ech center to resolve ny mbiguities in responses. Only completed nd checked dt were stored in the centrl dtbse. The study ws pproved by the institutionl review bord of the coordinting center nd of ech prticipting institution. All enrolled ptients gve informed consent. A totl of 1059 glioblstom ptients, treted between Jnury 2002 nd June 2007 in 18 Itlin RT centers, were collected. The contribution of ech center to the entire dt set rnged from 0.3% to 14.3%. Detiled dt regrding clinicl presenttion, surgicl procedure, pthologicl specimen, postopertive RT, djuvnt chemotherpy, tretmentrelted morbidity, nd survivl outcome were recorded in the centrl dtbse. Only biopsy-proven pure glioblstom cses, nlyzed by specilist neuropthologists, were included in the study, excluding mixed glioms with n oligodendroglil component. Toxicity during djuvnt RT nd/or chemotherpy ws scored with Common Terminology Criteri for Adverse Events v3.0 (CTCAE) [http://ctep.cncer.gov/forms/ctcaev3.pdf]. Worsening of neurologicl symptoms during RT ws scored ccording to Centrl Nervous System Acute Rdition Morbidity Scoring Criteri [www.rtog.org/members/ toxicity/cute.html], (Tble 1). Sttisticl Anlysis The Kpln-Meier method ws used to estimte overll survivl (OS). OS ws clculted from the dte of surgery until the time of deth or lst follow-up exmintion. Survivl curves were compred using the logrnk test. Multivrite nlysis ws performed with the Cox regression model. Differences between different subgroups of ptients were clculted using the χ 2 test. All nlyses were performed using STATISTICA, version 6.0, commercilly vilble softwre pckge (Sttistic for Windows, Sttsoft, Tuls, Oklhom). RESULTS Ptient Chrcteristics The medin ge of ptients ws 61 yers (rnge 20-85 yers) (Tble 2). The mjority of ptients (90.5%) hd single lesion. The most frequently involved site ws the temporl lobe (23.6% mong ptients with single lesion); in 32.2% of ptients, more thn one lobe ws involved; in less thn 2% of ptients, 3 lobes were involved. At the time of dignosis, the presence of focl deficits ws the most frequent symptom (76.3%). Incidentl dignosis, bsed on NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 447
SCOCCIANTI ET AL TABLE 2. Ptients Clinicl Fetures Feture No. (%) Age, y 50 215 (20.3) 51-60 312 (29.5) >60 532 (50.2) Sex Mle 651 (61.5) Femle 408 (38.5) No. of lesions Single 958 (90.5) Multiple 101 (9.5) Tumor site in ptients with single lesion 958 (90.5) Suprtentoril sites 948 (99.0) Cerebellum 8 (0.8) Brinstem 2 (0.2) Presenting symptoms Focl symptoms 808 (76.3) Crnil hypertension 336 (31.7) Seizure 215 (20.3) None 6 (0.6) Preopertive imging MRI 169 (16.0) CT 226 (21.3) MRI + CT 664 (62.7) Functionl imging 21 (3.9) Postopertive imging MRI 206 (19.4) CT 371 (35.1) MRI + CT 376 (35.5) None 106 (10.0) Preopertive KPS 90-100 288 (27.2) 80 220 (20.8) 70 551 (52.0) Prerdiotherpy KPS 90-100 390 (36.9) 80 334 (31.5) 70 335 (31.6) RPA clss III 121 (11.4) IV 651 (61.5) V 197 (18.6) VI 90 (8.5) MRI, mgnetic resonnce imging; CT, computed tomogrphy; KPS, Krnofsky Performnce Sttus; RPA, recursive prtitioning nlysis. imging performed without neurologicl symptoms, occurred in only 0.6% of cses. Mgnetic resonnce imging (MRI) ws frequently used both in preopertively nd postopertively (78.7% nd 54.9%, respectively). Most ptients (n = 664, 62.7%) underwent both preopertive MRI nd computed tomogrphy (CT), wheres only 35.5% of ptients underwent both imging studies postopertively before RT. The medin preopertive nd pre-rt Krnofsky Performnce Sttus (KPS) scle scores were 70 nd 80, respectively. Most of the ptients (61.5%) met the criteri for recursive prtitioning nlysis (RPA) clss IV (Tble 3). 4 Tretment Chrcteristics Surgery nd Periopertive Therpy Biopsy ws performed in only 123 ptients (11.6%), wheres most of the ptients underwent crniotomy with removl of the lesion (n = 936, 86.0%) (Tble 4). Postopertive imging ws performed in 953 ptients (90.0%), 901 of whom hd undergone surgicl removl. The scns in 710 of these ptients (75.9% of cses of surgicl removl) were dequte to ssess the extent of surgery. For the remining 226 ptients (191 with difficult-to-interpret postopertive scns nd 35 without postopertive scns), the extent of tumor erdiction ws bsed on surgicl report lone (24.1% of ptients who hd undergone surgicl tumor removl). In 45.8% of the ptients, surgery resulted in complete removl (no residul enhncement seen on postopertive CT or MRI or no residul tumor bsed on surgicl reports when no postopertive imging ws vilble); prtil removl occurred in 41%. Crmustine wfers were used in 11 ptients, nd 2 ptients underwent rdioimmunotherpy in the djuvnt setting. Rdiotherpy All 1059 ptients received djuvnt RT. Approximtely 90% (n = 934) of the ptients were treted with 3-dimensionl conforml RT (3D-CRT). A totl of 163 tretment plns (17.4% of ptients treted with 3D-CRT) were bsed on coregistered CT nd MRI. The vst mjority of ptients (n = 903, 85.3%) were treted with conventionl frctiontion (1.8 or 2 Gy per frction, 5 frctions per week), with the medin prescribed dose being 60 Gy (rnge 50-70 Gy). Among the ptients treted with hypofrctiontion (n = 156), 133 ptients received 3 Gy for dily frction with medin totl dose of 30 Gy (rnge 30-45 Gy). Adjuvnt Chemotherpy A totl of 287 ptients (27.1%) received RT lone, wheres 749 ptients (70.7%) underwent postopertive chemotherpy with TMZ. Concomitnt nd sequentil TMZ ws the most frequently dopted chemotherpy schedule (n = 542, 51.2%). In 77 ptients (7.3%), 1 cycle of TMZ (200 mg/mq) ws dministered before RT ccording to the design of n ongoing clinicl tril. Sixty-four ptients (6.0%) received TMZ only during RT, 448 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA TABLE 3. Recursive Prtitioning Anlysis Clss Definition for High-Grde Glioms 4 RPA Clss Age, y Histotype KPS Mentl or Neurologicl Sttus Surgery Symptom Onset RT I <50 AA Norml mentl sttus II 50 AA 70-100 3 mo before tretment III <50 AA Abnorml mentl sttus <50 GBM 90-100 IV <50 GBM <90 50 AA 70-100 <3 mo before tretment 50 GBM Good neurologicl function Resection V 50 GBM 70-100 Disbility to work Resection 50 GBM 70-100 Biopsy 54.4 Gy 50 GBM <70 Norml mentl sttus VI 50 GBM 70-100 Biopsy <54.4 Gy 50 GBM or AA <70 Abnorml mentl sttus RPA, recursive prtitioning nlysis; KPS, Krnofsky Performnce Sttus; RT, rdiotherpy; AA, nplstic strocytom; GBM, glioblstom multiforme. wheres 66 ptients (6.2%) received only sequentil TMZ fter RT lone. Only 2.2% of ptients received different djuvnt chemotherpy (crmustine: n = 9 [0.8%]; procrbzine + lomustine + vincristine: n = 14 [1.3% ]). A totl of 287 ptients (27.1%) did not receive chemotherpy. Among these ptients, 192 (66.9%) were elderly. The unsuitbility of chemotherpy in 24 of the remining 95 ptients ws bsed on poor prognostic fctors tht mde the im of tretment pllitive. Seventy-one ptients were treted exclusively with RT in centers where djuvnt chemotherpy with TMZ ws proposed only fter the publiction of the phse III tril by Stupp et l. 1 Chnges in Adjuvnt Tretment Over Time Since the evidence-bsed dt on efficcy of postopertive TMZ were disseminted in Mrch 2005, ptients were strtified ccording to the time of tretment (group A: from Jnury 2002 to Mrch 2005, the time of publiction of phse III tril of the efficcy of TMZ in the djuvnt tretment; group B: from April 2005 to June 2007, the end of the current study). Use of chemotherpy significntly incresed in the subgroup of ptients treted from April 2005 (χ 2 test, P <.001). In the sme group, the 3D-CRT ws more often dopted compred with ptients treted in the erlier yers of the study (χ 2 test, P <.001) (Tble 5). Age nd Tretment Older ptients (older thn 60 yers of ge) were more often treted with less ggressive pproch. Surgicl removl nd 3D- CRT were less frequently used in elderly ptients, lthough differences did not rech sttisticl significnce. At the sme time, there ws significnt difference in the choice of frctiontion nd systemic therpy cross ge groups. Hypofrctionted tretment ws given to higher proportion of ptients older thn 60 yers (20.7%) compred with younger ones (8.7%; χ 2 test, P <.001). Only 64% of ptients older thn the ge of 60 yers received djuvnt chemotherpy, wheres lrger percentge of younger ptients were treted with systemic therpy (82.0%; χ 2 test, P <.001). Slvge Tretment Slvge tretment ws recorded for 709 ptients. After progression, most of the ptients (n = 481, 67.8%) were treted with chemotherpy (TMZ: n = 203, procrbzine + lomustine + vincristine: n = 97, fotemustine: n = 78, bleomycin nd crmustine: n = 54, other: n = 49). Forty-two ptients (5.9%) underwent second surgery, wheres 10 ptients (1.4%) received rdiosurgery or hypofrctionted stereotctic RT s second RT. One hundred seventy-six ptients (24.8%) hd best supportive cre for the progressive disese. Survivl For the entire series, the medin survivl ws 9.5 months, nd cturil OS rtes t 1, 2, nd 5 yers were 62.3%, 24.8%, nd 3.9%, respectively. The medin follow-up, clculted from the end of RT to lst follow-up or deth, ws 6.5 months (follow-up ws longer thn 90 dys fter the end of RT in more thn 85% of ptients). Univrite Anlysis Vribles selected for survivl nlysis were chosen from both ptient nd tretment chrcteristics (Tble 6). Ptient chrcteristics significntly ssocited with better prognosis included younger ge t dignosis, single lesion, bsence of focl symptoms t dignosis, higher preopertive KPS scle score s well s pre-rt KPS scle score, nd lower RPA clss. Tretment chrcteristics significntly ssocited with survivl included more extensive surgery, use of 3D-CRT, 60 Gy s totl NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 449
SCOCCIANTI ET AL TABLE 4. Tretment Chrcteristics No. (%) Surgicl procedure Biopsy 123 (11.6) Crniectomy 25 (2.4) Crniotomy 911 (86.0) Extent of surgery Gross totl removl 485 (45.8) Subtotl removl 434 (41.0) Unknown 17 (1.6) Biopsy 123 (11.6) 3D-CRT Yes 934 (88.2) No 125 (11.8) RT frction size (Gy) 1.8 45 (4.3) 2 858 (81.0) 2.5 23 (2.2) 3 133 (12.5) RT totl dose (Gy) for 1.8/2 Gy per frction Totl 903 (85.3) <59.4 90 (8.5) 59.4 or 60 619 (58.4) >60 194 (18.3) Continues rdition dose in conventionlly frctionted tretments, nd the use of chemotherpy (Figures 1-3). Ptients who received both concomitnt (75 mg/m 2 /dy, 7 dys per week) nd sequentil (150/200 mg/m 2 for 5 d during ech 28-dy cycle for 6 cycles) stndrd schedule of TMZ (n = 542) hd better survivl compred with ptients who were on different TMZ regimens (n = 207). At the univrite nlysis, tretment time fter dissemintion of the TMZ efficcy dt ws relted to better outcome (logrnk test, P =.03). We performed univrite nlysis in the subgroup of ptients older thn 60 yers of ge to evlute how tretment modlities my TABLE 4. (Continued) No. (%) RT totl dose (Gy) for 3 Gy per frction Totl 133 (12.5) 30 78 (7.3) >30 55 (5.2) Stereotctic boost Yes 17 (1.6) No 1042 (98.4) Chemotherpy Yes 772 (72.9) No 287 (27.1) Drug Concomitnt + sequentil TMZ 542 (51.2) Other schedules of djuvnt TMZ 207 (19.5) Other chemotherpy 23 (2.2) Time of djuvnt tretment Jnury 2002-Mrch 2005 435 (41.1) April 2005-June 2007 624 (58.9) 3D-CRT, 3-dimensionl conforml rdiotherpy; RT, rdiotherpy; TMZ, temozolomide. ffect the survivl of elderly ptients. We found tht ggressive tretment (surgicl removl, 3D-CRT, conventionlly frctionted tretment, nd ssocition with chemotherpy) ws significntly ssocited with better survivl rte in the older ge group (Tble 7). Multivrite Anlysis All the vribles evluted for the entire group were ssessed using the Cox regression model. On multivrite nlysis, 5 vribles retined prognostic significnce: ge (P =.002), pre-rt KPS scle score (P =.01), extent of surgery (P <.001), 3D-CRT (P <.001), nd chemotherpy (P <.001). Five vribles mong those tht were significnt in the univrite nlysis filed to hve significnt impct on survivl: number of lesions; focl symptoms s presenting symptoms, preopertive KPS scle score, RPA clss, nd time of tretment (before or fter dissemintion of the TMZ efficcy dt). TABLE 5. Chnges in Adjuvnt Tretment Over Time Time of Tretment Totl No. Chemotherpy 3D-CRT χ 2 Test Yes No Yes No χ 2 Test Jnury 2002-Mrch 2005 435 268 (61.6%) 167 (38.4%) P <.001 365 (83.9%) 70 (16.1%) P <.001 April 2005-June 2007 624 504 (80.8%) 120 (19.2%) 569 (91.2%) 55 (8.8%) 3D-CRT, 3-dimensionl conforml rdiotherpy. 450 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA TABLE 6. Univrite Anlysis Log-Rnk Proportion Surviving (%) Test P Vlue 1 Yer 2 Yers 4 Yers All 62.3 24.8 6.8 Age, y <.001 50 75.4 39.5 16.4 51-60 69.0 29.1 5.2 >60 53.3 17.1 4.1 Sex.14 Mle 61.4 22.5 5.3 Femle 63.8 28.9 11.5 No. of lesions <.001 Single 63.6 26.7 7.3 Multiple 48.5 6.6 NA Presenting symptoms Crnil hypertension.25 Yes 57.4 22.2 6.8 No 64.6 26.2 6.8 Seizure.16 Yes 68.9 30.1 8.6 No 60.5 23.4 6.1 Focl symptoms.01 Yes 60.6 23.6 6.4 No 72.7 30.5 10.9 Preopertive KPS <.001 90-100 70.1 29.6 9.7 80 66.0 25.8 6.6 70 55.9 21.7 4.5 Prerdiotherpy KPS <.001 90-100 74.1 34.1 9.9 80 60.0 17.8 5.2 70 48.8 21.6 5.3 RPA clss <.001 III 80.6 46.5 18.8 IV 66.8 25.7 5.8 V 44.9 12.2 7.4 VI 40.2 17.8 0 Extent of surgery <.001 Biopsy 36.8 7.6 NA Subtotl removl 58.8 22.7 8.5 Gross totl removl 73.1 31.9 5.8 3D-CRT <.001 Yes 64.4 26.8 7.4 No 41.8 4.7 0 Continues TABLE 6. Univrite Anlysis Log-Rnk Proportion Surviving (%) Test P Vlue 1 Yer 2 Yers 4 Yers RT totl dose (Gy) for 1.8/2 Gy per frction <.001 <60 45.3 16.6 NA 60 70.7 32.9 8.5 >60 64.3 21.3 6.5 RT totl dose (Gy) for 3 Gy per frction.15 30 20.2 5.4 0 >30 41.9 NA NA Chemotherpy <.001 Yes 69.1 29.2 9.8 No 40.6 11.7 1.0 Temozolomide <.001 Concomitnt + 78.0 41.7 6.7 sequentil TMZ Other schedules of TMZ 62 20.3 10.0 Time of djuvnt tretment.03 Jnury 2002-Mrch 2005 58.3 22.7 6.2 April 2005-June 2007 65.8 26.7 7.7 NA, not vilble; KPS, Krnofsky Performnce Sttus; RPA, recursive prtitioning nlysis; 3E-CRT, 3-dimensionl conforml rdiotherpy; TMZ, temozolomide. FIGURE 1. Extent of surgery nd survivl (log-rnk test, P <.001). Toxicity A retrospective ssessment of toxicity ws performed. Among ptients who received RT lone (n = 287), 26 ptients (9.0%) hd grde 2 hedche, wheres grde 2 or higher worsening of preexisting focl neurologicl deficit ws reported in 51 ptients (17.7%) (Tble 8). NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 451
SCOCCIANTI ET AL FIGURE 2. Three-dimensionl conforml rdiotherpy (3D-conforml RT) nd survivl (log-rnk test, P <.001). A totl of 338 ptients (62.4%) did not show evidence of toxicity during concomitnt RT nd TMZ. During the concomitnt phse, 361 toxic events were recorded. Severe myelodepression (grde 3 or higher) ws recorded in 24 ptients (4.4%), nd, consequently, ll these ptients premturely discontinued TMZ tretment; in ddition, 10 ptients (1.8%) received reduced doses of TMZ for the remining weeks of tretment becuse of grde 2 hemtologicl toxicity. In 6 ptients (1.1%) TMZ tretment ws interrupted becuse of grde 3 deep venous thrombosis. There were 9 cses of grde 3 elevtion of liver trnsminses (1.7%). FIGURE 3. Postopertive chemotherpy nd survivl (log-rnk test, P <.001). A totl of 115 ptients (21.3%) hd dverse effects during sequentil TMZ tretment, for totl of 301 events. TMZ tretment ws premturely interrupted in 39 ptients (7.2%) becuse of grde 3 or higher hemtologicl toxicity (neutropeni: n = 18, thrombocytopeni: n = 19, nemi: n = 2), wheres 17 ptients (3.1%) required dose reduction of TMZ for grde 2 myelodepression. Seven ptients (1.3%) discontinued TMZ becuse of grde 3 deep venous thrombosis. In ddition, severe liver toxicity (grde 3) ws reported in 5 cses (0.9%). The most common nonhemtologicl dverse event during tretment ws nuse, reported by 103 (19.0%) nd 107 (19.7%) ptients, during concomitnt nd sequentil TMZ, respectively. TABLE 7. Survivl According to Tretment Modlity in Elderly Ptients Age >60 y, n = 532 (50.2%) Log-Rnk Proportion Surviving (%) Test P Vlue 1 Yer 2 Yers 4 Yers Surgery <.001 Biopsy 67 (12.6%) 34.6 3.0 0 Exeresis 465 (87.4%) 56.1 19.2 4.8 RT 3D-CRT.03 Yes 460 (86.5%) 54.5 18.8 4.6 No 72 (13.5%) 40.9 0 0 Frctiontion.004 Conventionl frctionted RT 422 (79.3%) 57.0 19.1 4.7 Hypofrctionted RT 110 (20.7%) 34.1 5.7 NA Chemotherpy <.001 Yes 340 (64%) 60.6 20.6 5.8 No 192 (36%) 38.3 10.3 1.4 3D-CRT, 3-dimensionl conforml rdiotherpy; RT, rdiotherpy. 452 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA TABLE 8. Toxicity of Postopertive Tretment G1 G2 G3 G4 Toxicity for rdiotherpy lone (n = 287), no. (%) Hedche 40 (13.9) 26 (9.0) Nuse 11 (3.8) Emesis 1 (0.3) Neurologicl symptoms 10 (3.5) 48 (16.7) 3 (1.0) Toxicity for concomitnt temozolomide nd rdiotherpy (n = 542), no. (%) Hedche 42 (7.7) 14 (2.6) Nuse 95 (17.5) 8 (1.5) Emesis 41 (7.6) 7 (1.3) Incresed liver 12 (2.2) 9 (1.7) trnsminses Neutropeni 8 (1.5) 11 (2.0) 3 (0.5) 6 (1.1) Thrombocytopeni 23 (4.2) 11 (2.0) 7 (1.3) 5 (1.0) Anemi 9 (1.7) 1 (0.2) 3 (0.5) Deep venous thrombosis 6 (1.1) Ftigue 28 (5.2) 12 (2.2) Toxicity for sequentil temozolomide (n = 542), no. (%) Hedche 13 (2.3) 14 (0.5) Nuse 72 (13.3) 35 (6.5) Emesis 10 (1.8) 3 (0.5) Incresed liver 17 (3.1) 5 (0.9) trnsminses Neutropeni 12 (2.2) 12 (2.2) 9 (1.7) 9 (1.7) Thrombocytopeni 18 (3.3) 26 (4.8) 12 (2.2) 7 (1.3) Anemi 1 (0.2) 4 (0.7) 2 (0.4) Deep venous thrombosis 7 (1.3) Ftigue 13 (2.4) G, grde. Comprison With Previous Itlin Ptterns of Cre Study In 2006, we reported the results of the Ptterns of Cre study bsed on the collbortion of 12 Itlin RT centers tht collected dt over 15-yer period (1985-2001) on totl of 1722 gliom ptients, the mjority of whom were ffected by glioblstom (n = 1235, 72%). 3 In the current study, the results of common dtbse tht included dt on ptients treted between 2002 nd 2007 re reported. We compred dt from the current study with those of the most recent period of the previous Itlin study (1997-2001) to determine whether significnt chnges occurred in clinicl prctice during the pst 10 yers. The differences between these 2 periods of time were ssessed with the χ 2 test (Tble 9). In both cses, most of the ptients were preopertively stged either with CT or MRI (61% vs 62.7%, P =.48), but we found significntly incresed use of postopertive MRI (30% vs 54.9%, P <.001) either s the exclusive imging modlity (14.0% vs 19.4%, P =.003) or s complementry to CT scnning (16% vs 35.5%, P <.001). The frequency of biopsies hd decresed over time (22% vs 11.6%, P <.001), with the mjority of ptients undergoing crniotomy. We did not find significnt chnge in the rte of gross totl resection (48% vs 45.8%, P =.49), but in the current series, the extent of surgery ws confirmed with postopertive MRI or CT most of the time (75.9% of cses). The mjority of ptients included in the current series underwent 3D-CRT (26% vs 88.2%), suggesting the generl cceptnce of CT-bsed RT nd, t the sme time, n incresed interest in treting glioblstom ptients with more complex techniques. Most of the ptients (85.1%) received conventionlly frctionted tretment, wheres there ws slight but significnt increse in the proportion of ptients receiving hypofrctionted RT (10% vs 14.9%, P =.009). The role of chemotherpy in clinicl prctice hs drmticlly chnged (21% vs 72.9%) since postopertive concurrent TMZ + RT followed by sequentil TMZ hd become the stndrd of cre for newly dignosed cses of glioblstom during the ccrul period. Survivl of the whole group of ptients reported in the previous Itlin study (1985-2001) significntly differs from the survivl of the ptients included in the current series (log-rnk test, P <.001) In compring the survivl of the subgroup of ptients treted between 1997 nd 2001 (medin OS: 8.1 month, 1-yer OS: 31.1%, 2-yer OS: 11.6%) with the outcomes in the current study (medin OS: 8.1 months, 1-yer OS: 62.3%, 2-yer OS: 24.8%), we found significnt improvement in terms of survivl (log-rnk test, P <.001) (Figure 4). DISCUSSION Comprison With Other Ptterns of Cre Studies This retrospective study exmined ptterns of cre for newly dignosed glioblstom ptients in Itly from 2002 to 2007. To our knowledge, it is the lrgest survey of glioblstom ptients for such short period of observtion. During this time period, the tretment for GBM significntly chnged, mostly becuse RT + TMZ hs become the stndrd of cre. Thus, the nlysis reflects n importnt period of chnge in the multidisciplinry mngement of glioblstom. In 2005, Chng et l 5 reported the results of Gliom Outcomes Project (GOP), multi-institutionl dtbse tht monitored clinicl prctice ptterns mong North Americn ptients with highgrde glioms between 1997 nd 2000. Strtifiction of ptients ccording to sex, number of lesions, nd seizure s presenting sign produced results similr to ours (Tble 10). The ge-djusted incidence ws different in the 2 series, with the proportion of ptients younger thn 40 yers being smller in the current study (9.8% vs 5.4%, P =.002). MRI ws the most commonly performed imging study in both series but ws used more frequently in North Americ thn in Itly (P <.001). In the GOP study, NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 453
SCOCCIANTI ET AL TABLE 9. Min Differences With the Previous Itlin Ptterns of Cre for Glioms 3 Previous Itlin Ptterns Previous Itlin Ptterns Comprison With of Cre Study: Glioblstom of Cre Study: All Current Study Previous Itlin Ptients in the Most Recent Glioblstom Ptients Study (1997-2001) Recruitment Period Period of observtion 1985-2001 1997-2001 2002-2007 Glioblstom ptients 1235/1722 (72%) 633/859 (74%) 1059/1059 (100%) Survivl Medin survivl 8.7 mo 8.1 mo 9.5 mo <.001 b 1-y cturil survivl rte 29.0% 31.1% 62.3% <.001 c 2-y cturil survivl rte 9.0% 11.6% 24.8% <.001 c 5-y cturil survivl rte 1.0% 2.2% 3.9% <.001 c Sex, no. (%) Mle 1070 (62) 443 (58) 651 (61.5) Femle 652 (38) 322 (42) 408 (38.5) Age, y, no. (%) 50 511 (30) 223 (26) 215 (20.3) 51-60 446 (26) 204 (24) 312 (29.5) >60 756 (44) 432 (42) 532 (50.2) Preopertive imging, no. (%) MRI 173 (10) 118 (14) 169 (16.0) CT 715 (41) 216 (25) 226 (21.3) MRI + CT 834 (49) 525 (61) 664 (62.7) Postopertive imging MRI 152 (9) 116 (14) 206 (19.4) CT 1014 (59) 472 (55) 371 (35.1) MRI + CT 195 (11%) 141 (16) 376 (35.5) Surgery, no. (%) Biopsy 437 (25) 185 (22) 123 (11.6) Subtotl resection 535 (31) 256 (31) 434 (41.0) Gross totl resection 750 (44) 409 (48) 485 (45.8) 3D-CRT, no. (%) Yes 281 (16) 219 (26) 934 (88.2) No 1441 (84) 640 (74) 125 (11.8) RT frction size, no. (%) 1.8-2 Gy 1521 (88) 751 (88) 903 (85.1) >2 Gy 155 (9) 90 (10) 156 (14.9) Adjuvnt chemotherpy, no. (%) Yes 272 (16) 177 (21) 772 (72.9) No 1450 (84) 682 (79) 287 (27.1) MRI, mgnetic resonnce imging; CT, computed tomogrphy; 3-CRT, 3-dimensionl conforml rdiotherpy; RT, rdiotherpy. b F test, P vlue. c Log-rnk test, P vlue. d Chi-squre test, P vlue..125 d.001 d.04 d <.001 d <.001 d <.001 d.009 d <.001 d biopsy ws less frequently performed (4.6% vs 11.6%, P =.002). Dt regrding the extent of tumor removl were not comprble becuse their study used surgicl reports rther thn postopertive imging. Intropertive chemotherpy with crmustine wfers were used more frequently in the North Americn study thn the current series. 454 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA FIGURE 4. Survivl of ptients in the current series compred with the survivl of ptients recruited in the most recent ccrul period of the previous Itlin Ptterns of Cre study 3 (log-rnk test, P <.001). The proportion of ptients receiving chemotherpy in the GOP survey ws much lower thn tht of the present series. Furthermore, the uthors reported tht crmustine nd lomustine were the most frequently used chemotherpy drugs since results of TMZ phse III tril 1 were published some yers fter the end of ccrul of the GOP survey. Rosenthl et l 6 recently reported n Austrlin Ptterns of Cre survey, exmining dt for 473 GBM ptients mong 828 eligible ptients with glioms. In this study, the Victorin Coopertive Oncology Group exmined ptterns of clinicl prctice for ptients with glioms (World Helth Orgniztion I-IV) over 3-yer period (1998-2000) using questionnire developed by multidisciplinry committee. Their questionnire ws designed to obtin dt regrding ptient chrcteristics nd therpeutic mngement, but it did not evlute number of lesions, presenting symptoms, KPS scle score, nd dignostic imging procedures. Furthermore, detils bout RT (technique, totl dose, frctiontion) were not reported. Compring their findings in GBM ptients with ours, gross totl removl ws performed less frequently thn in our series (31% vs 45.8%, χ 2 test, P <.001), lthough the uthors reported the extent of tumor removl by surgicl report lone. Only 86% of ptients in the Austrlin study were referred to RT center, nd only 322 ptients (68%) underwent RT. Once gin, the min difference compred with our results ws the lower rte of djuvnt chemotherpy in the Victorin Ptterns of Cre (P <.001). In more recent series, Cher et l 7 specified tht the use of concurrent systemic therpy ws extremely rre (<1%), wheres 13% of cses performed post-rt chemotherpy. Survivl nd Prognostic Fctors Medin survivl in the whole series ws 9.5 months, similr to those of other published Ptterns of Cre studies (7.4-9.5 months). 5,8 Two-yer OS ws higher thn wht ws observed in other Ptterns of Cre studies (24.8% vs 9%-11%. 3,8 Survivl hs sttisticlly improved compred with the previous Itlin series 3 nd prticulrly the ptients recruited in the lst ccrul period (1997-2001) (Figure 4). Multivrite nlysis showed sttisticl significnce for ge, postopertive KPS scle score, extent of surgery, use of 3D-CRT, nd use of chemotherpy. Age ws reported to be significnt prognostic fctor in severl Ptterns of Cre studies. 3,8,9 In ddition, postopertive performnce dt from the previous Itlin survey 3 nd from the GOP 5 were vlidted s hving prognostic vlue. Extent of resection hs lso been shown to be n importnt fctor in determining outcome in ll the Ptterns of Cre studies tht performed survivl nlysis compring subgroups of ptients treted with different surgicl modlities. 3,9,10 Brnholtz-Slon et l 10 showed n impct on survivl tht depended on tretment modlity, demonstrting n incresed risk of deth for ptients treted with biopsy only, single-modlity tretment (vs surgery + RT), or no chemotherpy. Chemotherpy-treted ptients lso hd better outcome in the previous Itlin study. 3 All the prmeters selected for multivrite nlysis, except the use of 3D-CRT, were ssocited with higher survivl rte in the recently published nlysis of the EORTC/NCIC tril 26981-22981, imed to derive nomogrms for predicting survivl of GBM ptients. 11 We observed reltively long-term survivl in subgroups of ptients with the previously mentioned fvorble prognostic fctors: 4-yer OS rnging from 5.8% to 16.4% (Tble 6). Elderly Ptients: Prognosis nd Tretment Advnced ge, s previously mentioned, ws ssocited with significntly lower survivl in other Ptterns of Cre studies. 3,8,9 Younger ge hs been vlidted s n independent prognostic fctor lso in the recently published subnlysis of the EORTC/NCIC tril 26981-22981. 11 Our results confirm tht older ptients hve worse prognosis. Age proved to be sttisticlly significnt prognostic fctor in both univrite nd multivrite nlyses (P <.001). We found significnt disprity in the djuvnt tretment prescribed cross ge groups. Ptients older thn 60 yers were treted more often with hypofrctiontion nd without chemotherpy compred with ptients 60 yers of nd younger (P <.001 for both). Severl Ptterns of Cre studies, some of which specificlly focused on elderly ptients, showed significnt difference in the use of postopertive therpy in the different ge groups. 3,5,7,9,10,12 Lowry et l 9 reported tht 18% of elderly ptients (older thn 65 yers) dignosed between 1990 nd 1995 received no tretment t ll, wheres 13% of cses underwent surgery lone. Brnholtz- Slon et l, 10 using the Surveillnce, Epidemiology nd End Results (SEER) dtbse, identified 1753 cses of elderly GBM ptients (older thn 65 yers) nd concluded tht ge ws strongly ssocited with tretment difference, even mong older ptients, with very elderly ptients (older thn 75 yers) being more likely to receive limited tretment (surgery lone or RT lone). Lutterbch et l 12 exmined lrge monoinstitutionl cohort of glioblstom NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 455
SCOCCIANTI ET AL TABLE 10. Min Differences With the Subgroup of Glioblstom Ptients Included in Gliom Outcome Project 5,8 Gliom Outcome Project Current Study Sttisticl Comprison, P Vlue Period of observtion 1997-2000 2002-2007 Glioblstom ptients 418/565 (74%) 1059/1059 (100%) Survivl NA Medin survivl 9.5 mo 9.5 mo 2-y cturil survivl rte 11.0% 24.8% Sex, no. (%) Mle 247 (60.4) 651 (61.5) Femle 162 (39.6) 408 (38.5) Age, y, no. (%) <40 39 (9.8) 57 (5.4) 40-60 165 (41.5) 470 (44.4) >60 194 (48.7) 532 (50.2) No. (%) of lesions Single 364 (88.4) 958 (90.5) Multiple 48 (11.7) 101 (9.5) Presenting symptoms nd signs Seizure 97 (23.5) 215 (20.3) Preopertive imging MRI 375 (90.8) 833 (78.7) CT 313 (75.8) 890 (84.0) Surgery, no. (%) Biopsy 19 (4.6) 123 (11.6) Subtotl resection 113 (27.4) 434 (41.0) Gross totl resection 192 (46.5) 485 (45.8) BCNU wfers, no. (%) Yes 62 (15.1%) 11 (1.0%) No 348 (84.9%) 1048 (99.0%) Postopertive KPS scle score, no. (%) <70 74 (18) 335 (31.6) 70 339 (82) 724 (68.4) Postopertive RT, no. (%) Yes 365 (89.7) 1059 (100) No 53 (10.3) 0 (0) Adjuvnt chemotherpy, no. (%) Yes 222 (54.6) 772 (72.9) No 196 (45.4) 287 (27.1) Percentges from Gliom Outcome Project re bsed on the number of ptients who nswered ech question on the questionnire, s reported by the uthors. Sttisticl comprison of outcome in terms of survivl is not fesible becuse no mesure of vribility (no stndrd errors or stndrd devitions) ws reported in the Gliom Outcome Project (thus mking it impossible to pply sttisticl tests). NA, no vilble; MRI, mgnetic resonnce imging; CT, computed tomogrphy; BCNU, bleomycin nd crmustine; KPS, Krnofsky Performnce Sttus; RT, rdiotherpy. b Chi-squre test, P vlue..67 b.01 b.23 b.22 b <.001 b <.001 b <.001 b <.001 b <.001 b <.001 b ptients (n = 430) nd found more frequent use of 3D-CRT in younger ptients. The previous Itlin study 3 lredy reported more frequent use of conventionlly frctionted RT nd chemotherpy in ptients ged 60 nd younger. Results of the GOP 5 confirmed tht ptients receiving chemotherpy tended to be younger. Cher et l 7 nlyzed the use of chemotherpy in ptients who were included in the Victorin Ptterns of Cre nd found tht only 15% of ptients older thn 60 yers received chemotherpy. 456 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com
PATTERNS OF CARE FOR GLIOBLASTOMA Our series showed tht elderly ptients undergoing ggressive tretment hd longer survivl (Tble 7), which ws in ccordnce with other studies. 9,13-15 These findings suggest tht old ge in itself should not preclude ggressive tretment when fesible. Toxicity of Adjuvnt Tretment Despite the well-known limits of retrospective toxicity ssessment, dt regrding tretment-relted side effects were collected. All dverse events were recorded nd scored ccording to vlidted toxicity criteri. RT lone ws globlly well tolerted. Grde 3 centrl nervous system toxicity (neurologicl findings requiring hospitliztion) ws observed in only 3 ptients (1.0%). During concomitnt RT + TMZ, most dverse events (n = 322) were mild or moderte (grde 2 or lower), wheres severe toxicity ws reported in 39 of 542 ptients (7.2%). Fifty-one ptients (9.4%) experienced grde 3 or 4 toxicity during sequentil chemotherpy. The toxicity profile of RT + TMZ reported in the current retrospective series (nd consequently limited by its own nture) ws similr to tht described in the EORTC/NCIC tril 1 in which grde 3 or 4 hemtologicl effects developed in 7% to 14% of ptients during concomitnt nd sequentil therpy, respectively. Limits nd Advntges of the Study Exmining 1059 biopsy-proven pure glioblstom cses, this series differs from others tht included different histologicl grdes 3,5,6 or mixed glioms with n oligodendroglil component 5 s well s ptients without histologicl dignosis. 6 Detils on imging, surgery (with emphsis on postopertive rdiologicl ssessment of surgicl extent), RT, nd chemotherpy mke this nlysis comprehensive. Moreover, it is the first survey tht includes strtifiction of ptients ccording to RPA clsses nd nlysis of tretment-relted toxicity scored ccording to vlidted criteri (while recognizing the limittions of retrospectively ssessing toxicity). The results, lthough limited by the retrospective nture of the study, my highlight tht greter dignostic nd therpeutic resources re now offered to glioblstom ptients. More frequent use of postopertive MRI, for instnce, might reflect greter ttention to ssessing the extent of surgicl resection (nd its true prognostic vlue) or to better defining RT trget volume (using MRI for coregistrtion in the tretment plnning phse). We think tht the wide diffusion of postopertive MRI my lso reflect the need to hve bseline rdiologicl study for the follow-up, especilly for more ccurte ssessment of the erly response to the tretment (ie, pseudoprogression fetures). Even though some of the prognostic fctors such s ge, KPS scle score, nd extent of surgery hve been previously reported in literture, this is the first time tht the use of 3D-CRT nd the role of concomitnt nd sequentil TMZ hs been vlidted in such lrge group of ptients. This study my be criticized for severl limittions. First, the dt re limited by the retrospective nture of the study with high number of censored ptients. Second, it included only ptients who received RT, excluding ptients who hd biopsy- proven glioblstom but were not referred for RT. Third, the lck of biologicl dt, such s O(6)-methylgunine-DNA methyltrnsferse sttus, is consistent limittion of our survey, considering the predictive nd prognostic importnce of such dt. 11,16 Moreover, the lck of uniform pthologicl review remins drwbck of this study. Finlly, we did not ssess the qulity of life of ptients with vlidted tests nor we did collect dt on supportive cre (eg, steroids, ntiepileptic drugs, thromboembolic prophylxis). CONCLUSION This present study provides benchmrk nlysis of current ptterns of clinicl prctice nd outcomes for ptients with glioblstom. Becuse of the time period covered (2002-2007), it reveled mjor chnges in GBM ptient cre, highlighting the greter dignostic nd therpeutic resources tht re now vilble to glioblstom ptients. The routine use of MRI, the decision to perform biopsy in only minority of ptients, the more frequent use of 3D-CRT, nd the widespred use of chemotherpy represent the most significnt chnges in clinicl prctice reported in our nlysis. The observed improvement in survivl, with significnt proportion of reltively long-term survivors (4-yer OS: 6.8%; 5-yer OS: 3.9%), suggests tht most glioblstom ptients would benefit from ggressive multimodlity therpy. Disclosure The uthors hve no personl finncil or institutionl interest in ny of the drugs, mterils, or devices described in this rticle. REFERENCES 1. Stupp R, Mson WP, vn den Bent MJ, et l. Europen Orgnistion for Reserch nd Tretment of Cncer Brin Tumor nd Rdiotherpy Groups; Ntionl Cncer Institute of Cnd Clinicl Trils Group. Rdiotherpy plus concomitnt nd djuvnt temozolomide for glioblstom. N Engl J Med. 2005;52(10):987-996. 2. Stupp R, Hegi ME, Mson WP, et l. Europen Orgnistion for Reserch nd Tretment of Cncer Brin Tumour nd Rdition Oncology Groups; Ntionl Cncer Institute of Cnd Clinicl Trils Group. Effects of rdiotherpy with concomitnt nd djuvnt temozolomide versus rdiotherpy lone on survivl in glioblstom in rndomised phse III study: 5 yer nlysis of the EORTC-NCIC tril. Lncet Oncol. 2009;10(5):459-466. 3. Mgrini S, Ricrdi U, Sntoni R, et l. Ptterns of prctice nd survivl in retrospective nlysis of 1722 dult strocytom ptients treted between 1985 nd 2001 in 12 Itlin rdition oncology centers. Int J Rdit Oncol Biol Phys. 2006;65(3):788-799. 4. Currn WJ Jr, Scott CB, Horton J, et l. Recursive prtitioning nlysis of prognostic fctors in three Rdition Therpy Oncology Group mlignnt gliom trils. J Ntl Cncer Inst. 1993;85(9):704-710. 5. Chng SM, Prney IF, Hung W, et l. Ptterns of cre for dults with newly dignosed mlignnt gliom. JAMA. 2005;293(5):557-564. 6. Rosenthl MA, Drummond KJ, Dlly M, et l. Mngement of gliom in Victori (1998-2000): retrospective cohort study. Med J Aust. 2006;184(6):270-273. 7. Cher L, Rosenthl MA, Drummond KJ, et l. The use of chemotherpy in ptients with glioms: ptterns of cre in Victori from 1998-2000. J Clin Neurosci. 2008;15(4):398-401. 8. Lws ER, Prney IF, Hung W, et l. Survivl following surgery nd prognostic fctors for recently dignosed mlignnt gliom: dt from the Gliom Outcomes Project. J Neurosurgery. 2003;99(3):467-473. NE UROSURGERY VOLUME 67 NUMBER 2 AUGUST 2010 457
SCOCCIANTI ET AL 9. Lowry JK, Snyder JJ, Lowry PW. Brin tumors in the elderly: recent trends in Minnesot cohort study. Arch Neurol. 1998;55(7):922-928. 10. Brnholtz-Slon JS, Willims VL, Mldondo JL, et l. Ptterns of cre nd outcomes mong elderly individuls with primry mlignnt strocytom. J Neurosurg. 2008;108(4):642-648. 11. Gorli T, vn den Bent MJ, Hegi ME, et l. Nomogrms for predicting survivl of ptients with newly dignosed glioblstom: prognostic fctor nlysis of EORTC nd NCIC tril 26981-22981/CE.3. Lncet Oncol. 2008;9(1):29-38. 12. Lutterbch J, Brtelt S, Momm F, Becker G, Frommhold H, Ostertg C. Is older ge ssocited with worse prognosis due to different ptterns of cre? A longterm study of 1346 ptients with glioblstoms or brin metstses. Cncer. 2005;103(6):1234-1244. 13. Brndes AA, Vstol F, Bsso U, et l. A prospective study on glioblstom in the elderly. Cncer. 2008;97(3):657-662. 14. Mukerji N, Rodrigues D, Hendry G, Dunlop PR, Wrburton F, Kne PJ. Treting high grde glioms in the elderly: the end of geism? J Neurooncol. 2008;86(3):329-336. 15. Shw EG. Nothing ventured, nothing gined: tretment of glioblstom multiforme in the elderly. J Clin Oncol. 2004;22(9):1540-1541. 16. Hegi ME, Diserens AC, Gorli T, et l. MGMT gene silencing nd benefit from temozolomide in glioblstom. N Engl J Med. 2005;352(10):997-1003. Acknowledgments We thnk the following collegues for their vluble collbortion in collecting dt: Brbr Grilli Leonulli, MD, (Florence), Icro Mettini, MD, (Florence), Pol Vitli, MD, (Bresci), Elen Montni, MD, (Bresci), Brbr Lzzri, MD, (Bresci), Angel Botticell, MD, (Torino), Stefno Dll Oglio, MD, (Veron), Lur Msini, MD, (Novr), Annlis Ntuno, MD, (Sn Giovnni Rotondo), Ptrizi Gicobzzi, MD, (Moden), Adele Piscopo, MD, (L Aquil), Antonio Pontoriero, MD, (Messin), Crlo Ivn Cusno, MD, (Benevento), Fbrizio Mtteucci, MD, (Pis), Vivin Brbieri, MD, (Rionero in Vulture), Roberto Mnzo, MD, (Nples), Lun Di Murro, MD, (Rome), Rebec Nicolis, MD, (Rome), Id Milnesi, MD, (Miln), Alessi Frneti, MD, (Perugi). S.S. thnks Alberto Innlfi, MD, for the helpful support during the process of building the computerized dtbse. We re lso very grteful to Dvid Biro, MD, who reviewed the English version of the mnuscript; his help ws invluble. INTERNATIONAL TRAVELING FELLOWSHIP IN PEDIATRIC NEUROSURGERY The Joint Peditric Neurosurgery Section of the Americn Assocition of Neurologicl Surgeons nd Congress of Neurologicl Surgeons hs estblished n interntionl trveling fellowship for neurosurgeons who t the time of their ppliction re either trining in residency progrm outside the United Sttes nd Cnd, or who hve completed residency trining outside the United Sttes nd Cnd within the pst five yers. The fellowship will cover the trveling nd living expenses for three-month period to be spent observing the ctivities of n estblished Peditric Neurosurgicl service of the pplicnt s choosing in the United Sttes or Cnd. One fellowship will be wrded yerly on the bsis of competitive evlution by committee of the Peditric Section t the nnul meeting in December. The mximum fellowship stipend is $5000. The ppliction must include: 1) A sttement defining the purpose of the proposed fellowship; 2) A letter of recommendtion from the pplicnt s current Neurosurgicl progrm director; 3) A letter of cceptnce from the institution where the pplicnt intends to tke the fellowship, confirming the description of the fellow s potentil ctivities during the period of the wrd; 4) The pplicnt s current Curriculum Vite. The strict dedline for ppliction submission is November 15, 2010. The completed ppliction should be sent to: Bermns J. Iskndr, M.D. Deprtment of Neurologicl Surgery University of Wisconsin School of Medicine nd Public Helth 600 Highlnd Avenue, Room K4/832 CSC Mdison, WI 53792 [or vi e-mil to: iskndr@neurosurg.wisc.edu] 458 VOLUME 67 NUMBER 2 AUGUST 2010 www.neurosurgery-online.com