J Neurooncol DOI 1.17/s11-1-2228-4 CLINIcAL STUDY Proton beam therapy with concurrent chemotherapy for glioblastoma multiforme: comparison of nimustine hydrochloride and temozolomide Masashi Mizumoto 1 Tetsuya Yamamoto 2 Eiichi Ishikawa 2 Masahide Matsuda 2 Shingo Takano 2 Hitoshi Ishikawa 1 Toshiyuki Okumura 1 Hideyuki Sakurai 1 Akira Matsumura 2 Koji Tsuboi 1 Received: 18 December 215 / Accepted: 28 July 21 Springer Science+Business Media New York 21 Abstract To evaluate the safety and efficacy of postoperative proton beam therapy (PBT) combined with nimustine hydrochloride (ACNU) or temozolomide (TMZ) for glio - blastoma multiforme (GBM). The subjects were 4 patients with GBM who were treated with high dose (9. GyE) PBT. There were 24 males and 22 females, and the median age was 58 years old (range 24 7). The Karnofsky perfor - mance status was, 7, 8, 9 and 1 in 5, 1, 12, 11 and 8 patients, respectively. Total resection, partial resec - tion, and biopsy were performed for 31, 14 and 1 patients, respectively. Photon beams were delivered to high intensity areas on T2-weighted magnetic resonance imaging (MRI) in the morning (5.4 Gy in 28 fractions). More than h later, PBT was delivered to the enhanced area plus a 1 mm margin in the first half of the protocol (23.1 GyE in 14 fractions) and to the enhanced volume in the second half (23.1 GyE in 14 fraction). Concurrent chemotherapy with ACNU during weeks 1 and 4 or daily TMZ was administered in 23 and 23 patients, respectively. The overall 1 and 2 year survival rates were 82. and 47. %, respectively. Median survival was 21.1 months (95 % CI 13.1 29.2), with no significant difference in survival between the ACNU and TMZ groups. The patient characteristics were similar in the two groups. Late radiation necrosis occurred in 11 patients (six ACNU, five TMZ), but was controlled by necrotomy Koji Tsuboi tsuboi@pmrc.tsukuba.ac.jp 1 Department of Radiation Oncology, Proton Medical Research Center, University of Tsukuba, 1-1-1, Tennoudai, Tsukuba, Ibaraki 35-8575, Japan 2 Department of Neurosurgery, University of Tsukuba, Tsukuba, Ibaraki, Japan and therapy including bevacizumab. PBT concurrent with ACNU or TMZ was tolerable and beneficial for carefully selected patients with GBM. Keywords Proton beam therapy Glioblastoma GBM TMZ Temozolomide Radiotherapy Introduction Glioblastoma multiforme (GBM) has a poor prognosis and the most aggressive therapies have achieved only mod - est outcomes. Radiotherapy after surgical resection may improve survival and time to progression [ 1 3] and radio - therapy plus chemotherapy produces a modest improvement in survival [ 4, 5]. Radiotherapy of Gy in 3 fractions plus temozolomide (TMZ) increased median survival in a phase III randomized trial compared to external beam radiotherapy alone [ 8], and recent reports have shown that bevacizumab plus radiotherapy and TMZ improve progression-free survival [ 9, 1]. Combined therapy with surgery, radiotherapy and chemotherapy extends the median survival time by about 1 months. Dose escalation has also been tried for GBM using new techniques such as intensity modulated radiation therapy (IMRT) or particle therapy. Tanaka et al. found that highdose conformal radiotherapy of 9 Gy in 45 fractions for malignant glioma had potential to improve survival [ 11]. Fitzek et al. showed that conformal protons and photons in accelerated fractionation of 9 GyE for GBM improved the median survival period to 2 months [12]. Iuchi et al. used hypofractionated high-dose IMRT with concurrent and adjuvant TMZ in a phase II trial [13]. However, dose escalation increases the risk of radiation necrosis and the benefit of this approach is unclear.
2 J Neurooncol We obtained a median survival period of 21. months for GBM using high-dose proton beam therapy (PBT) concur - rent with nimustine hydrochloride (ACNU) [14, 15]. However, the efficacy and safety of PBT concurrent with TMZ is unknown. Here, we evaluated PBT given with ACNU or TMZ. Material and methods Patients From September 21 to September 212, a total of 15 patients with GBM diagnosed histologically received postoperative radiotherapy at our institute. As conventional radiotherapy, 4 5 Gy in 2 25 fractions were first delivered to the area of T2 hyperintense MRI, with a 2. cm margin. Subsequently, we added 1 2 Gy as a boost to the area with T1 post-contrast enhancement or the tumor bed plus a 2. cm margin, giving a total dose of Gy in 3 fractions. In patients with a poor general condition, such as a Karnofsky performance status (KPS) <4 or old age, consecutive daily treatment of 3 fractions is often difficult. In such cases, the total dose and number of fractions were reduced to shorten the treatment period: typically, a dose of 45 Gy in 15 fractions to the contrast-enhanced area plus a 1 2 cm margin. Of the 15 patients, 119 received conventional radiotherapy: 89 treated with doses from to 1.2 Gy in 3 34 fractions, and 3 treated with lower doses in fewer fractions. The remaining 4 patients (Table 1) desired treatment with hyperfractionated concomitant boost PBT. Written informed consent was obtained from all patients. Before PBT, radiation oncologists and brain surgeons confirmed that predicted radiation necrosis was unlikely to be fatal in all cases. The criterion for this decision was based on the potential resectability of a lesion when brain necrosis was found in the range to be irradiated at 9. Gy (lesion + resected cavity + 5 mm). The patients comprised 24 men and 22 women, and had a median age of 58 years old (range 24 7 years old). The tumor location was the frontal lobe, temporal lobe, parietal lobe and occipital lobe in 23, 1, 3 and 4 patients, respectively. Five patients had a KPS of, 1 had KPS 7, 12 had KPS 8, 11 had KPS 9, and 8 had KPS 1. One patient had undergone biopsy, 14 had received partial resection, and 31 had undergone subtotal resection/gross total resection before radiotherapy. PBT was administered concurrently with ACNU in 23 patients and with TMZ in 23 patients. Combination chemotherapy with ACNU was performed until April 28 and all patients have received TMZ since this date. The first 2 patients were also subjects of a previously reported Phase I/II prospective study [14]. Table 1 Patient characteristics Characteristics ACNU (N = 23) TMZ (N = 23) Age (years) Treatment methods 31 7 (range) 55 (median) 24 74 Gender Male 12 12 Female 11 11 Location Frontal lobe 8 15 Temporal lobe 1 Parietal lobe 2 1 Occipital lobe 3 1 Karnofsky performance status 1 4 7 4 8 8 4 9 7 4 1 1 7 Extent of surgery Biopsy only 1 Partial resection 9 5 Total resection 14 17 RPA class Class V 7 Class IV 13 9 Class III 4 7 All 4 patients received hyperfractionated concomitant boost proton radiotherapy after surgery. Computed tomography (CT) images taken at 3 mm intervals at the treatment site were used for PBT treatment planning. Proton beams with an energy of 25 MEV were generated by a booster synchrotron at the Proton Medical Research Center (PMRT). The treatment planning system gives dose distributions and settings for the collimator configuration, bolus, and range-shifter thickness. The relative biological effectiveness (RBE) of PBT was assumed to be 1.1. Clinical target volume 1 (CTV1) was defined as the area of contrast enhancement on MRI, CTV2 as the area of contrast enhancement on MRI plus a 1 mm margin, and CTV3 as surrounding edema determined by imaging (T2-weighted or fluid-attenuated inversion recovery MRI) plus a 15 mm margin. MRI was performed immediately before and after surgery in all subjects. Conventional radiotherapy of 5.4 Gy in 28 fractions was delivered to CTV3 in the morning. Additional concomitant boost proton radiotherapy of 23.1 GyE in 14 fractions was delivered to CTV2 more than h after conventional radiotherapy, and proton radiotherapy of 23.1 GyE in 14
J Neurooncol fractions was then delivered to CTV1. The planning target volume (PTV) was defined as the CTV plus 5 mm for set up error. The exposure dose was adjusted to 5 Gy for the chiasm and Gy for the brainstem and thalamus. If the dose to an organ at risk exceeded the exposure dose, the margin of the lesion was reduced. Thus, the total doses to PTV1, PTV2 and PTV3 were 9. GyE in 5 fractions, 73.5 GyE in 42 fractions, and 5.4 Gy in 28 fractions, respectively. ACNU was given intravenously at 8 mg/m 2 for 1 day in weeks 1 and 4 of radiotherapy. From 28, concurrent chemotherapy was changed from ACNU to daily TMZ at 75 mg/m 2. Follow-up procedures and evaluation criteria Acute treatment-related toxicities were assessed weekly during treatment. After completion of PBT, the patients were evaluated by physical examinations, MRI, and blood tests every 3 months for the first 2 years and every months thereafter. The Kaplan Meier method was used for calculation of local control and survival rates, and a Log-rank test was performed for evaluation of differences between two categories. Acute and late treatment-related toxicities were assessed using the National Cancer Institute Common Criteria ver. 3. and the RTOG/EORTC late radiation morbidity scoring scheme [1]. A progressive or enhanced lesion on MRI within CTV1 (which received 9. GyE) was defined as an in-field MRI change. A similar change within CTV2 (73.5 GyE) or CTV3 (5.4 GyE) was defined as a border MRI change, and a change outside the irradiated field was defined as an extra-field MRI change. Results All patients received radiotherapy and PBT within 3 months postoperatively. The period between surgery and radiother - apy was 15 87 days (median 27 days), and radiotherapy and PBT were completed in 37 5 days (median 43 days). At the time of analysis, 13 patients were alive and 33 had died. The median follow-up period for survivors was 42.1 months (range 2. 11.3 months). Death was due to cancer in 28 patients and to a disease unrelated to tumor recurrence in five patients. The overall 1- and 2-year survival rates for all patients were 82. and 47. %, respectively, and the median survival period was 21.1 months (range 2.8 11.3 months, 95 % CI 13.1 29.2 months). The 1- and 2-year MRI change-free survival rates were 37. and 11. %, respectively, and the median MRI change-free survival period was 8.3 months (range 1.5 35. months, 95 % CI.3 1.3 months). Overall and MRI change-free survival curves for all patients are shown in Fig. 1. At the last follow-up, 42 patients had a progressive or enhanced lesion on MRI 31 of these cases were diagnosed as tumor recurrence and 11 as radiation necrosis (five in the TMZ group and six in the ACNU group). Of the 31 recur - rence cases,, 1 and 9 were in areas that received irradiation of 9., 73.5 GyE, and 5.4 GyE, respectively. In contrast, 1 of the 11 cases of radiation necrosis occurred in the area irradiated with 9. GyE and one in the area that received 73.5 GyE. The median MRI change-free survival times were.7 (range 1.5 25.7) months in recurrence cases and 13.2 (.2 35) months in necrosis cases (Fig. 2). The patients in the ACNU and TMZ groups had similar characteristics (Table 1). In the ACNU group (n = 23), the median follow-up period for survivors was 88. (7.8 11.3) months. Nineteen patients completed two planned rate Necrosis Recurrence rate Overall survival MRI change 1..8..4.2. 1..8..4.2. 11 31 4 4 12 24 3 (months) 38 1 Recurrence cases (n=31) Overall survival me (median 21.1 months) MRI-change free survival (median 8.3 months) 48 Fig. 1 Kaplan Meier estimates of overall and progression-free sur - vival for all 4 patients P =.2 Radia on necrosis (n=11) 12 24 3 (months) 2 1 Fig. 2 Kaplan Meier estimates of progression-free survival in radia - tion necrosis cases (n = 11) and recurrence cases (n = 31) 21 4 1 3
4 J Neurooncol rate TMZ ACNU 1..8..4.2. 23 23 ACNU group (n=23) MST 21. months TMZ group (n=23) MST 25.7 months 12 24 3 (months) 22 1 11 1 P =.59 48 3 7 Fig. 3 Kaplan Meier estimates of overall survival in the TMZ and ACNU groups cycles and four patients had to miss the second cycle of chemotherapy due to acute reactions. In the TMZ group (n = 23), the median follow-up period for survivors was 33.4 (2. 44.2) months. Sixteen patients completed daily TMZ and seven patients required breaks due to myelosuppression. The median overall survival times were 21. and 25.7 months in the ACNU and TMZ groups, respectively (p =.59) (Fig. 3). The median MRI change-free survival was.9 months in the ACNU group and 1.4 months in the TMZ group (p =.43). Age, gender, performance status ( 8 vs. 9 1), surgery (subtotal-gross total vs. partialbiopsy), and concurrent chemotherapy (TMZ vs. ACNU) were evaluated as potential predictive factors for overall survival, but none of these factors were significant in univariate and multivariate analyses. Non-hematologic acute toxicity was mild. Only two patients had a grade 3 non-hematologic acute reaction. Grade 2 or 3 non-hematologic acute toxicity occurred in three patients in the ACNU group and 11 patients in the TMZ group. There were 14 cases of hematologic acute toxicity that was probably caused by chemotherapy in the ACNU group: four, seven, one and two patients had grade 4 leukopenia, neutropenia, lymphopenia and thrombocytope - nia, respectively. In the TMZ group, three patients had grade 4 lymphopenia. A summary of adverse events is shown in Table 2. KPS was evaluable in 35 and 1 subjects at 1 and 2 years after irradiation, respectively. Median KPS was 8 ( 1) and (3 1) before and 1 year after irradiation, respectively, and 17 patients had recurrence of GBM, four had brain necrosis, and 14 had neither recurrence nor necrosis. The changes in median KPS were 3 ( to +2), 1 ( 2 to +1) and 1 ( to +2) in these respective groups. In the 1 subjects who were evaluable at 2 years, median KPS was 8 ( 1) and 7 (3 9) before and 2 years after irradiation, respectively; and five patients had recurrence, eight had necrosis, and three had neither. The changes in median KPS were 3 ( 5 to 1), 1 ( 4 to +2) and 2 ( 4 to +1) in the respective groups. Discussion Radiotherapy is standard treatment for high-grade glioma after surgical resection because considerable residual tumor remains in the normal brain. Standard radiotherapy of Gy in 3 fractions with TMZ is typically used [ 8]. SRS/IMRT and particle radiotherapy achieve better dose localization in the tumor volume, and several prospective studies of these new methods have shown that dose escalation is feasible and survival is not inferior to standard schedules [9, 1, 13, 17]. Fitzek et al. showed that accelerated fractionated proton/photon irradiation at 9 GyE for GBM gave median survival of 2 months [12] and we found that accelerated fractionated proton/photon irradiation at 9. GyE concur - rent with ACNU also achieved better survival [14, 15]. For Table 2 Adverse events Event ACNU (N = 23) TMZ (N = 23) Grade 2 Grade 3 Grade 4 Grade 2 Grade 3 Grade 4 Anemia 3 4 3 Leukopenia 7 4 8 1 Neutropenia 2 7 7 4 1 Lymphopenia 9 1 2 1 3 Thrombocytopenia 7 3 2 Nausea and vomiting 2 1 Dermatitis 2 5 Otitis 1 1 Seizure 2
J Neurooncol consistency with standard treatment, in 28 we changed ACNU to daily TMZ, without changing the basic PBT and radiotherapy protocols. The results of this study indicate that the change from ACNU to TMZ had no significant impact on MRI change-free and overall survival using a high dose PBT protocol. The follow-up period for survivors differed substantially between the TMZ and ACNU groups because TMZ treatment was only initiated in 28. Several patients in the ACNU group survived for 5 years or longer. Longer follow up is needed to confirm whether survivors in the TMZ group will have a similar course. Mirimanoff et al. and Li et al. showed that recursive partitioning analysis (RPA) could identify significant prognostic factors for GBM [18, 19]. In these reports, RPA III was classified as KPS 9 1 and age <5, and RPA IV was classified as KPS 8 and age <5 or 5 with surgical resection. Overall survival differed among RPA classes III and IV, with median survival of 1.3 and 11.3 months, respectively. The median survival of 21.1 months in the current study was better than that in patients of RPA class III, despite 35 of the 4 patients in our study being in RPA class IV or V. Matsuda et al. reported treatment outcomes of GBM in our hospital from 1998 to 27. Median survival was 17.7 months in 35 subjects treated with ACNU who underwent irradiation at 1.2 Gy, and multivariate analysis showed that high-dose PBT and boron neutron capture therapy were significant good prognostic factors. Local failure after standard radiotherapy of Gy in 3 fractions is common for GBM [7, 8]. However, only six of 31 patients in the current study had recurrence within the 9. GyE area of irradiation and much of this was bor - der recurrence. Fitzek et al. and our previous reports have shown similar failure patterns [12, 14, 15]. These results suggest that a dose of 9 GyE is required to control GBM. In our protocol, 9. GyE was irradiated with only a.5 cm margin for the residual tumor. This margin is clearly insufficient for GBM, but the adequate safety margin for normal brain is unknown with high dose PBT. Matsuo et al. showed that methionine PET (MET-PET) is highly sensitive in the context of brain tumor tissue [19], with residual tumor occasionally detected by MET-PET at more than 2 cm beyond the contrast enhancement on MRI. Iuchi et al. also showed that methionine uptake was significantly correlated with tumor control [2]. Use of these new diagnostic methods is likely to improve target delineation and definition of the extent of the residual tumor. Radiation necrosis is also an important problem. Necrotomy, hyperbaric oxygen therapy, anticoagulants and corti - costeroids are used for treatment of radiation brain necrosis [21], and bevacizumab may provide a new treatment for radiation necrosis [22 24]. We previously found that six of 23 patients treated with PBT concurrent with ACNU who developed radiation necrosis were well controlled by necrotomy and/or bevacizumab [15]. The improved sur - vival and prolonged follow-up period led to five patients developing new radiation necrosis. We now preferentially treat inoperable symptomatic radiation necrosis with bevacizumab. Regarding survival, KPS at 1 and 2 years after irradiation decreased by about ten in patients with brain necrosis, but by about 3 in those with recurrence. There - fore, the effect of brain necrosis on KPS was less than that of recurrence. However, we have no KPS data for irradia - tion at Gy and it will be important to examine KPS after irradiation at this dose. There may be selection bias for tumors located distal to a risk organ such as the brainstem and optic chiasm, but the results of this study support the efficacy of high dose PBT concurrent with chemotherapy. However, the optimal PBT schedule is uncertain. It is clear that 9. GyE is sufficient to control GBM, but a 5-mm margin is not sufficient to cover GBM. Therefore, the total dose may have to be decreased to extend the margins in future clinical trials. However, our results indicate that high dose PBT concurrent with TMZ or ACNU extends the median survival time to 2 months and is well tolerated in carefully selected patients with GBM. Use of MET-PET and bevacizumab may further improve outcomes, and prospective studies of new irradiation meth - ods and combined treatment are warranted. Radiation necrosis is inevitable in long-term survivors. Therefore, prior to PBT, there is a need to predict the level of radiation necrosis and determine that this will not be fatal. Acknowledgments This work was partially supported by grants-inaid for Scientific Research (B) (15H491) and Young Scientists (B) (25814) from the Ministry of Education, Culture, Sports, Science and Technology of Japan. References 1. Walker MD, Strike TA, Sheline GE (1979) An analysis of dose effect relationship in the radiotherapy of malignant gliomas. Int J Radiat Oncol Biol Phys 5:1725 1731 2. Walker MD, Green SB, Byar DP et al (198) Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 33:1323 1329 3. Keime-Guibert F, Chinot O, Taillandier L et al (27) Radiotherapy for glioblastoma in the elderly. N Engl J Med 35:1527 1535 4. Fine HA, Dear KB, Loeffler JS et al (1993) Meta-analysis of radiation therapy with and without adjuvant chemotherapy for malignant gliomas in adults. Cancer 71:2585 2597 5. Stewart LA (22) Chemotherapy in adult high-grade glioma: a systematic review and meta-analysis of individual patient data from 12 randomised trials. Lancet 359:111 118. Athanassiou H, Synodinou M, Maragoudakis E et al (25) Randomized phase II study of temozolomide and radiotherapy compared with radiotherapy alone in newly diagnosed glioblastoma multiforme. J Clin Oncol 23:2372 2377 7. Stupp R, Mason WP, van den Bent MJ et al (25) Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med 352:987 99 5
J Neurooncol 8. Stupp R, Dietrich PY, Ostermann Kraljevic S et al (22) Promising survival for patients with newly diagnosed glioblastoma multiforme treated with concomitant radiation plus temozolomide followed by adjuvant temozolomide. J Clin Oncol 2:1375 1382 9. Gilbert MR, Dignam JJ, Armstrong TS et al (214) A randomized trial of bevacizumab for newly diagnosed glioblastoma. N Engl Med 37:99 78 1. Chinot OL, Wick W, Mason W et al (214) Bevacizumab plus radiotherapy-temozolomide for newly diagnosed glioblastoma. N Engl Med 37:79 722 11. Tanaka M, Ino Y, Nakagawa K et al (25) High-dose confor - mal radiotherapy for supratentorial malignant glioma: a historical comparison. Lancet :953 9 12. Fitzek MM, Thornton AF, Rabinov JD et al (1999) Accelerated fractionated proton/photon irradiation to 9 cobalt gray equivalent for glioblastoma multiforme: results of a phase II prospective trial. J Neurosurg 91:251 2 13. Iuchi T, Hatano K, Kodama T et al (214) Phase 2 trial of hypofractionated high-dose intensity modulated radiation therapy with concurrent and adjuvant temozolomide for newly diagnosed glioblastoma. Int J Radiat Oncol Biol Phys 88:793 8 14. Mizumoto M, Tsuboi K, Igaki H et al (21) Phase I/II trial of hyperfractionated concomitant boost proton radiotherapy for supratentorial glioblastoma multiforme. Int J Radiat Oncol Biol Phys 77:98 15 15. Mizumoto M, Yamamoto T, Takano S et al (215) Long-term survival after treatment of glioblastoma multiforme with hyper - fractionated concomitant boost proton beam therapy. Pract Radiat Oncol 5:e9 e1 1. Cancer Therapy Evaluation Program (2) Common terminol - ogy criteria for adverse events v3. (CTCAE). Cancer Therapy Evaluation Program, Bethesda, MD. http://ctep.cancer.gov/protocoldevelopment/electronic_applications/ctc.htm 17. Sultanem K, Patrocinio H, Lambert C et al (24) The use of hypofractionated intensity-modulated irradiation in the treatment of glioblastoma multiforme: preliminary results of a prospective trial. Int J Radiat Oncol Biol Phys 58:247 252 18. Mirimanoff RO, Gorlia T, Mason W et al (2) Radiotherapy and temozolomide for newly diagnosed glioblastoma: recursive partitioning analysis of the EORTC 2981/22981-NCIC CE3 phase III randomized trial. J Clin Oncol 24:253 259 19. Matsuo M, Miwa K, Tanaka O et al (212) Impact of [ 11 C]methionine positron emission tomograpy for target definition of glioblastoma multiforme in radiation therapy planning. Int J Radiat Oncol Biol Phys 82:83 89 2. Iuchi T, Hatano K, Uchino Y et al (215) Methionine uptake and required dose to control glioblastoma. Int J Radiat Oncol Biol Phys 93:133 14 21. Bui QC, Lieber M, Withers HR et al (24) The efficacy of hyperbaric oxygen therapy in the treatment of radiation-induced late side effects. Int J Radiat Oncol Biol Phys :871 878 22. Yonezawa S, Miwa K, Shinoda J et al (214) Bevacizumab treatment leads to observable morphological and metabolic changes in brain radiation necrosis. J Neurooncol 119:11 19 23. Lubelski D, Abdullah KG, Weil RJ et al (213) Bevacizumab for radiation necrosis following treatment of high grade glioma: a systematic review of the literature. J Neurooncol 115:317 322 24. Furuse M, Nonoguchi N, Kawabata S et al (213) Bevacizumab treatment for symptomatic radiation necrosis diagnosed by amino acid PET. Jpn J Clin Oncol 43:337 341 25. Matsuda M, Yamamoto T, Ishikawa E et al (211) Prognostic factors in glioblastoma multiforme patients receiving high-dose particle radiotherapy or conventional radiotherapy. Br J Radiol 84:S54 S