High-dose proton boost for confined type glioblastoma

4.1 Imaging biomarkers and clinical impacts

4.1.2 High-dose proton boost for confined type glioblastoma

High-dose proton boost to tumor bed was investigated for local control of glioblastoma [93]. In the further analysis by PE extent, glioblastoma patients of PE ≤ 19 mm compared with those > 19 mm showed better OS after high-dose proton boost to tumor bed (42.4 versus 18.4 months, p = 0.023) [94]. For patients with PE ≤ 19 mm, 92% tumor progression developed within 19 mm from the tumor margin. The 3-year local control rate was 38% [94].

These results suggested that intensive local treatment improves the disease control and OS for glioblastoma with relative small PE.

4.1.3 Disease classification and other treatment strategy

Wang et al. demonstrated high vascular endothelial growth factor (VEGF) expression levels with a peritumor edema extent of > 2 cm in glioblastoma [95]. Takano et al. described that the antiedema effect of bevacizumab, a monoclonal antibody that blocks angiogenesis by inhibiting VEGF, administered for the treatment of recurrent glioblastoma was marked and prolonged at 6 months [96]. The efficacy of additional bevacizumab for enhancing tumor

control at the edematous area in Group IV patients requires further investigation.

Accumulating evidence suggests that the stem cell markers, CD133 and CD44, indicate molecular subtypes in glioblastoma. In the study of Brown et al., coexpression analysis of CD133 and CD44 identified proneural and mesenchymal subtypes of glioblastoma. Patients with CD133 coexpression module signature (CD133-M) had significantly improved survival from radiotherapy treatment but no significant benefit from temozolomide [97]. In contrast, patients with CD44-M had higher survival benefit from temozolomide treatment but lower benefit from radiotherapy, compared to CD133-M patients [97]. In the study of Piccirillo et al., they observed that for glioblastoma tissues of the same patients, the tumor cells derived from the part of sub-ependymal zone were more resistant to supra-maximal chemotherapy doses of different drugs than tumor cells outside the zone [98], which suggests more intensive drug therapies are required when glioblastoma involving the SVZ.

Chaichana et al reported that patients with butterfly glioblastoma had poor prognosis but nevertheless benefited from aggressive treatments including debulking surgery, maximal safe surgical resection, temozolomide chemotherapy, and radiotherapy [31]. Glioblastoma with sSVZCC invasion is considered an early stage of butterfly glioblastoma and therefore may also benefit from comprehensive treatment strategies. For those with and without sSVZCC invasion, our study significantly determined the different prognosis and distinct progression areas, specifically at the CC, bilateral hemispheres, preoperative edema, and lateral ventricles [23]. These results imply that the therapeutic approach to patients with sSVZCC should be different from that for patients without invasion.

4.1.4 Imaging biomarkers and molecular biology

Because our IDH1 data were only available for 24 patients with one presenting mutated type, further analysis could not be performed. Wang et al. reported that in patients with the IDH1 mutation, ring-like tumor contrast enhancement and peritumoral edema (> 1 cm) indicated shorter PFS and OS [99]. Another limitation of our study is that we could not analyze the prognosis or progression patterns of MGMT promotor methylation status because of the lack of data. According to Carrillo et al., edema levels determined survival in methylated but not unmethylated tumors, meaning that patients with MGMT promotor methylation and peritumor edema of < 1 cm had favorable OS [100].

Our previous study demonstrated that the tumor progression rates at PE areas are > 40%

and < 10% for those with and without sSVZCC invasion, respectively [23]. The high progression rates at the CC and bilateral hemispheres for glioblastoma with sSVZCC suggest the potential correlation between glioblastoma stem cells and interhemispheric connections of cerebral commissures [23]. In the same study, all 3 patients with combined invasion of lateral ventricles and anterior commissure in our study exhibited progression patterns at the bilateral hemispheres that were comparable with those having sSVZCC invasion [23]. Kakita et al reported that in the neonatal rat model, progenitors within the SVZ at the angle of the lateral ventricle appeared to move in an undirected manner; however, after migration, they extensively moved radially and tangentially in the ipsilateral hemisphere of the injection side.

They also proposed that the progenitors migrate along the unmyelinated axon fiber fascicles in the CC [101], which corresponds to the progression sites of our patients with sSVZCC invasion. In the adult mammalian brain, neural progenitor cells are present and neurogenesis occurs in the SVZ on the walls of the lateral ventricles [36-38]. The migration of glial progenitors depends on their contact with the cerebral commissures, as demonstrated in a

study on neonatal rats [102]. The high rates of progression to the CC and bilateral hemispheres suggests that the migration of glial progenitors depends on their contact with NFTs [23]. These findings indicate the potential interactions between glioblastoma stem cells and the CC.

4.2 Treatment Impact and Clinical Application of Hydrogel Carboplatin

Through the comprehensive study design, we integrated the biomaterial, in vitro, and in vivo investigations to propose a novel combination of carboplatin with hydrogel and surveyed its optimized treatment combining with RT for glioma. According to the analyses of drug release profile, biocompatibility, tumor control, survival, and side effects in this study, we demonstrated hydrogel carboplatin as an effective, simplified, and safe modality for local drug delivery to combine with RT for treating mice glioma, which provides a further clinical application for brain tumor treatment.

4.2.1 Effectiveness of hydrogel carboplatin combined with RT for tumor control

We demonstrated that the hydrogel dye diffused 7 to 9 mm from the injection site after intratumoral injection, which provides adequate distribution extent of hydrogel carboplatin in brain tumors after excision clinically. Regarding to the carboplatin dose, we used different dosages of carboplatin combined with RT to investigate the synergistic treatment effect. In the first-stage in vivo experiment using low-dose carboplatin (3 µg/g), the tumor control effect of ACR and HCR treatment groups was not different from R treatment group, which suggests low-dose carboplatin is not sufficient for tumor control. In the second-stage in vivo

experiment using high-dose carboplatin (15 µg/g), the tumor control effect and survival of both HCR and ACR treatment groups were significantly better than R, HC and AC treatment groups. We designed the timing of RT on the second and third days after hydrogel carboplatin injection according to the drug release profile of hydrogel carboplatin and the tumor growth course of mice to maximize the CCRT synergistic effect. The tumor control effect of high-dose CCRT resulted in the higher complete response rate and longer survival time than carboplatin or RT alone. These results demonstrated that the synergistic effect of combining high-dose carboplatin with RT is more robust than RT or carboplatin alone. Clinically, most malignant gliomas progress or recur at the surgical cavity after tumor excision and RT [16-18]. Considering the clinical demand and our study results, the local delivery of hydrogel carboplatin at surgical cavity combined with RT is an effective treatment for malignant gliomas.

4.2.2 Convenience of hydrogel carboplatin administration to combine with RT

The administration of drug by thermogelling hydrogel carrier provides a different drug delivery option compared with wafer implant and aqueous drug via CED implant during surgery. The hydrogel form presents better conformal ability compared with wafer, and better attachment compared with aqueous form [42]. Our results indicates that once oxi-HA/ADH carboplatin was injected into in vivo condition at 37 °C, oxi-HA/ADH carboplatin formed gel in 17 seconds [67]. This gelling time of oxi-HA/ADH carboplatin is adequate for intratumoral injection and hydrogel stabilization for mice subcutaneous tumor model. For human brain tumor application, longer drug gelling time is required, which can be achieved

by increasing the concentration of ADH in the hydrogel, ranging from 2 to 3 min at 37 °C [54]. The thermogelation (thermal-dependent gelling) characteristic provides adequate time for drug injection and stabilization [68], which makes hydrogel carboplatin more adaptable to the shape of surgical cavity, easier adhesion to the tissue, and full coverage of surgical cavity compared with wafer-form and aqueous drugs [42].

Our results of low-dose and high-dose carboplatin experiments indicated that the synergistic effect with RT requires sufficient carboplatin dose. To maintain the adequate carboplatin dose takes drug via CED with daily infusion [49, 51] or slow release via drug carrier, including wafer [41] or hydrogel [55-58]. The clinical use of CED or wafer is limited due to its inconvenience and potential side effects [42, 52, 53]. Using the features of slow and steady release of hydrogel and synergistic effect of carboplatin with RT [46, 47], we compounded and applied the hydrogel carboplatin to treat the mice glioma. By comparing the effects of single injection of hydrogel carboplatin with multiple injection of aqueous carboplatin to investigate the feasibility and convenience of hydrogel carboplatin. Comparing the HC verse AC and HCR versus ACR groups respectively of the second-stage in vivo experiment, the tumor control effects and survivals of single dose hydrogel carboplatin (300 µg) were comparable with the aqueous carboplatin with daily dose (100 µg) for 3 days. The results indicate that the single injection of hydrogel carboplatin facilitates the drug administration without compromising the anti-tumor effect. When combining with RT, the single injection of hydrogel carboplatin simplified the concurrent RT course without compromising the synergistic effects.

One published study using covalent linking of platinum and hydrogel to compose hydrogel polymer–platinum, which made the platinum release lasted more than 60 days [68].

Clinically, the RT course duration for malignant glioma ranges from 3 to 6 weeks and hydrogel carboplatin is feasible for single injection during surgery rather than daily local infusion via CED. The degradation time of hydrogel ranges from 4 to 6 weeks after injection [67], which avoids interfering the neuroimaging follow-up in 8 to 12 weeks after treatment competed. Our results and published data indicates that using local and single injection of hydrogel carboplatin to combine with RT is a feasible treatment modality for malignant glioma [68, 70].

4.2.3 Safety of hydrogel carboplatin with RT

According to our in vitro study, the WST-1, LDH test, and LIVE/DEAD staining validated the biocompatibility of oxi-HA/ADH hydrogel. In the first-stage in vivo experiment, the tumor growth curves of R and HRT groups showed no significant difference, which suggests no additional toxicity of hydrogel when combining RT. In the second-stage in vivo experiment, through the white blood count, and renal and liver function tests, no systemic side effect was detected under the high-dose (15 µg/g) of either hydrogel carboplatin single dose or aqueous carboplatin daily injection for 3 days. In all treatment groups, no drug injection, radiation or CCRT-related skin ulcers were observed and the transitory weight loss was acceptable.

Due to the size scale of mice, the total drug dose for single injection was limited.

Comparing with human brain, the drug dose loaded by hydrogel is able to be higher.

According to the mice tumor growth schedule, we design the total drug release in 4 days and RT on the second and third days after first drug dose injection. The drug release time can be

prolonged up to 5 weeks technically [54] to fit the clinical CCRT course for malignant gliomas. In order to measure the tumor volume serially, we adopted the subcutaneous implant model rather than brain implant model. Our results demonstrated the intratumoral drug delivery, but the effect to normal brain tissue or its tolerance dose requires further investigation. The influences on tumor growth and treatment effect resulting from the environment differences between subcutaneous tissue and brain needs more survey.

C

HAPTER

5 C

ONCLUSIONS AND

F

UTURE

P

ROSPECT

5.1 Clinical Investigation

In the clinical investigation of this study, we analyzed the associations between these imaging factors and survival and tumor progression patterns and rates after CCRT and proposed corresponding RT target volume delineations and dose prescriptions. Our study demonstrated that EPE and sSVZCC invasion are highly associated with glioblastomas spreading extent and routes. The two imaging biomarkers provide clinical indexes to classify glioblastoma patients into different groups and in radiotherapy decision-making for individual patients, which introduces potential clinical investigation topics for the future.

5.2Basic Investigation

In basic investigation of this study, we compounded a novel drug combination of thermogelling hydrogel carboplatin to satisfy the unmet clinical need for glioma treatment.

Through the comprehensive biomaterial, cell, and animal experiment design, we significantly demonstrated that thermogelling hydrogel carboplatin simplified the drug delivery method

and frequency without compromising the synergistic effect with RT, which makes intraoperative single drug injection followed by RT a feasible and potential clinical treatment.

We chose carboplatin with well-established CCRT effects and hydrogel with well-accepted biocompatible features as a novel drug delivery combination, which helps accelerate the application process for further clinical trials. The impact of our study results is that we significantly demonstrated intratumoral injection of hydrogel carboplatin with RT as an effective, convenient, and safe treatment combination for malignant gliomas which is potential for clinical application.

Figure 1. The clinical and basic research perspectives of our glioblastoma study. We

proposed a comprehensive study, including imaging analysis and drug delivery investigation, to satisfy the clinical unmet needs for glioblastoma treatment.

Figure 2. The correlation between tumor location with edema and tumor migration.

The interactions between the anatomical factor subventricular zone (red line) and corpus callosum (green line) the pathophysiological factor preoperative edema (blue color wash) and the associated clinical impacts, including tumor migration ability and directions (black line with arrow) [64].

Figure 3. The rationale and purpose in the current basic study: by comparing the

characteristics of the intratumoral delivery modalities and drugs for malignant gliomas to propose a novel combination to satisfy the unmet clinical need [59].

Figure 4. The method of evaluating the preoperative edema extent in our clinical study.

(A) Select the image section that presents the maximum diameter of preoperative tumor. (B) Create a 2-cm expansion of tumor edge along SVZ (the lateral wall of lateral ventricles) or CC if these structures are invaded. Otherwise, draw tangential lines to the tumor edge and then use their normal lines to expand a 2-cm margin of the tumor edge. (C) Use the 2-cm expansion margin of tumor edge as a reference boundary to evaluate the PE extent. (D) Apply the 2-cm expansion margin of the preoperative tumor edge on the similar image sections after CCRT to evaluate the PD extent [64].

Figure 5. The definitions of edema extent and progression patterns. (A) Illustrations of

(A1) EPE (PE ≥ 2 cm from the tumor margin) combined with (A2) sSVZCC invasion to evaluate the rates of EPD (continuous or discrete progression of tumors spreading along the

PE areas and > 2 cm from the tumor margin), (A3) SVZEPD, and CCEPD after CCRT. (B) Illustrations of (B1) EPE− and (B2) sSVZCC+/− invasion to evaluate the rate of (B3) EPD after CCRT. (C) Illustrations of (C1) EPE+ and (C2) sSVZCC+/− invasion to evaluate the rates of (C3) EPD, SVZEPD, and CCEPD [64].

Figure 6. The workflow of our basic study design: biomaterial, in vitro, and in vivo

investigations [59].

Figure 7. Drug preparation: the procedures of (A) oxi-HA/ADH hydrogel and (B)

hydrogel carboplatin preparation. (Figures courtesy of Xue-Shi Lai, MSc.)

Figure 8. The treatment regimens and evaluation protocol of our mice study: (A)

low-dose carboplatin (first-stage experiment) and (B) high-low-dose carboplatin (second-stage experiment) [59].

15 µg/g 3 µg/g

Figure 9. Kaplan-Meier’s estimates of (A1) OS and (A2) PFS for patients with and

without EPE, (B1) OS and (B2) PFS for patients with and without sSVZCC invasion [64].

Figure 10. MRI demonstration of patients with different tumor locations. Distinct

progression patterns after CCRT in 4 patients categorized according to the different combinations of the SVZ and CC invasion. (A1, A2) The patient with synchronous left occipital horn of the SVZ (red dashed lines) and left posterior CC (green dashed lines) invasion exhibited distant progression to the left cerebellum (blue arrow). (A3) Other

progression (blue arrow) developed in the preoperative edematous areas (blue dashed lines), which migrated across the posterior CC to the contralateral hemisphere (yellow dashed arrows). (B1, B2) The patient with synchronous left frontal horn of the SVZ (red dashed lines) and left anterior CC (green dashed lines) invasion showed progression in the tumor bed of the left anterior CC (red arrows) and the distant area at the left posterior CC along the occipital horn (blue arrows). (C) The patient with left temporal horn of the SVZ (red dashed lines) but without CC invasion exhibited progression involving the tumor bed only (red arrow). (D) The patient with neither SVZ nor CC invasion showed progression involving the tumor bed only (red arrow) [23].

Figure 11. MRI demonstration of patients with different edema extents and tumor locations. Distinct progression patterns after CCRT in five patients categorized according to

the different combinations of EPE and sSVZCC invasion. A patient with (A1,2) EPE−/sSVZCC− before surgery presented with (A3) local progression without EPD. Another patient with (B1,2) EPE−/sSVZCC+ before surgery presented with (B3) local progression without EPD. Panels C1,2 demonstrate a patient with EPE+/sSVZCC− before surgery, who presented with (C3) local progression with EPD. Panels D1,2 show a patient with EPE+/sSVZCC+ and CCEPE before surgery, who presented with (D3) local progression with EPD along the CC. Panels E1,2 describe a patient with EPE+/sSVZCC+ and SVZEPE before surgery, who presented with (E3) local progression with EPD along the SVZ [64].

Figure 12. Illustrations of by FTIR analysis: (A1) the spectrum peak at 1730 cm-1 of

oxi-HA disappeared after mixing oxi-oxi-HA with ADH and the appearance of a new forming peak at 1528 cm-1 of oxi-HA/ADH and (A2) the appearance of a new forming peak at 545 cm-1 of oxi-HA/ADH hydrogel carboplatin [59].

Figure 13. The rheological properties of HA/ADH: (A1) the gelation of

HA/ADH started at temperature higher than 27.6 °C and (A2) the gelling time of oxi-HA/ADH from liquid state to gel state were 17 seconds at 37 °C (body temperature) [59].

Figure 14. Degradation properties of oxi-HA/ADH hydrogel: At 72 and 120 h,

degradation percentages for oxi-HA/ADH (Figure 4C) were 5.2% and 18.2%, respectively [59].

Figure 15. Drug release profile: The ICP-MS result demonstrates two phases of the

carboplatin release from hydrogel, including a burst release of 63.7% during the ﬁrst 24 h, followed by a steady release of 16.6% over the 24 to 96 h [59].

Figure 16. Biocompatibility of oxi-HA/ADH: (A) The WST-1 analysis demonstrated the

cell viability of 3T3 cells cultured in oxi-HA/ADH hydrogel extraction medium was not signiﬁcantly different compared with those in the control and negative control groups (p = 0.644). (B) The LDH assay indicated the cytotoxicity oxi-HA/ADH is not significantly different from the negative control group (p = 0.173) [59].

Figure 17. The LIVE/DEAD staining: Nearly all the 3T3 cells were viable in the

oxi-HA/ADH hydrogel after 3 days’ cultivation [59].

Figure 18. The IC

test of carboplatin: The in vitro concentrations of carboplatin to inhibit

50% ALCS1C1 cells to proliferation was 44.4 and 18.5 µg/mL after 1-day and 3-day treatment, respectively [59].

Figure 19. The BLIs evolution of the first-stage in vivo experiment. Comparing with the

sham group, the BLIs demonstrate the relatively delayed tumor progression in all treatment groups [59].

Figure 20. The tumor volume evolution of the first-stage in vivo experiment.

On day 24, the tumor volume analysis (excepting sham group) by one-way ANOVA showed no difference of tumor progression for RT with and without low-dose carboplatin (p = 0.787) [59].

Figure 21. The BLIs evolution of the second-stage in vivo experiment. The BLIs

demonstrate the tumor nearly complete response in HCR and ACR groups, while tumor progression in other treatment groups [59].

Figure 22. The bioluminescence signal of the second-stage in vivo experiment [59].

Figure 23. The tumor volume evolution of the second-stage in vivo experiment. In ACR

and HCR groups, the tumor volume curves demonstrate good tumor control without difference (p = 0.904) [59].

Figure 24. The survival curves of the second-stage in vivo experiment: The HCR and

ACR groups had 104-day survival rates of 50% and 66.7% without significant difference (p

= 0.648) [59].

Figure 25. The gross and histopathological findings of the second-stage in vivo

experiment: (A) The blue area illustrated the diameters of dye distribution 1 day and 3 days

after oxi-HA/ADH hydrogel dye injection were 7 and 9 mm, respectively. (B1) The H&E stain of tumor receiving RT showed both cell death (yellow rectangular area at 100× and yellow arrow at 400×) and tumor cell proliferation (red rectangular area at 100× and red arrow at 400×). The H&E stain of tumor receiving either aqueous carboplatin with RT (B2) or hydrogel carboplatin with RT (B3) showed prominent cell death. Contrastingly, H&E stain of tumor cell growth under no treatment (C1) and injection of hydrogel (C2), aqueous carboplatin (C3), and hydrogel carboplatin (C4) all showed tumor cell proliferation only [59].

Figure 26. The weight change and skin reaction of mice in the second-stage in vivo

experiment: (A) Mice weight change in the second-stage experiment. (B) No skin ulcer after

treatment of RT with high-dose (C1) hydrogel carboplatin or (C2) aqueous carboplatin [59].

Figure 27. The proposed personalized glioblastoma treatment strategies [59].

Figure 28. Progression patterns and sSVZCC invasion. The comprehensive progression

conditions for sSVZCC invasion patients (C1) after CCRT. (C2) The progression sites analysis classified by the anatomical structures. (C3) The progression patterns analysis based on the progressive tumor locations corresponding to the tumor bed and preoperative edematous areas. Abbreviations: C, corpus callosum; D, distant; E, edema; H, bilateral hemispheres; L, local; V, ventricles [23]

TABLES

REFERENCES

[1] Q.T. Ostrom, L. Bauchet, F.G. Davis, I. Deltour, J.L. Fisher, C.E. Langer, M. Pekmezci, J.A. Schwartzbaum, M.C. Turner, K.M. Walsh, M.R. Wrensch, J.S. Barnholtz-Sloan, The epidemiology of glioma in adults: a "state of the science" review, Neuro Oncol 16(7) (2014) 896-913.

[2] Cancer Registry Annual Report 2014, Health Promotion Administration, Ministry of Health and Welfare, Taiwan, 2016, p. 93.

[3] Cancer Registry Annual Report 2013, Health Promotion Administration, Ministry of Health and Welfare, Taiwan, 2015, p. 93.

[4] Cancer Registry Annual Report 2012, Health Promotion Administration, Ministry of Health and Welfare, Taiwan, 2014, p. 93.

[5] P.Y. Wen, S. Kesari, Malignant gliomas in adults, N Engl J Med 359(5) (2008) 492-507.

[6] S. Cha, Update on brain tumor imaging: from anatomy to physiology, AJNR Am J Neuroradiol 27(3) (2006) 475-87.

[7] J. Li, M. Wang, M. Won, E.G. Shaw, C. Coughlin, W.J. Curran, Jr., M.P. Mehta, Validation and simplification of the Radiation Therapy Oncology Group recursive partitioning analysis classification for glioblastoma, International journal of radiation oncology, biology, physics 81(3) (2011) 623-30.

[8] R.O. Mirimanoff, T. Gorlia, W. Mason, M.J. Van den Bent, R.D. Kortmann, B. Fisher, M. Reni, A.A. Brandes, J. Curschmann, S. Villa, G. Cairncross, A. Allgeier, D. Lacombe, R.

Stupp, Radiotherapy and temozolomide for newly diagnosed glioblastoma: recursive partitioning analysis of the EORTC 26981/22981-NCIC CE3 phase III randomized trial,

Journal of clinical oncology : official journal of the American Society of Clinical Oncology 24(16) (2006) 2563-9.

[9] H.K. Liang, C.W. Wang, H.M. Tseng, C.Y. Huang, K.H. Lan, Y.H. Chen, S.L. You, J.C.

Cheng, A.L. Cheng, S.H. Kuo, Preoperative prognostic neurologic index for glioblastoma patients receiving tumor resection, Annals of surgical oncology 21(12) (2014) 3992-8.

[10] M.E. Hegi, A.C. Diserens, T. Gorlia, M.F. Hamou, N. de Tribolet, M. Weller, J.M. Kros, J.A. Hainfellner, W. Mason, L. Mariani, J.E. Bromberg, P. Hau, R.O. Mirimanoff, J.G.

Cairncross, R.C. Janzer, R. Stupp, MGMT gene silencing and benefit from temozolomide in glioblastoma, The New England journal of medicine 352(10) (2005) 997-1003.

[11] S.E. Combs, S. Rieken, W. Wick, A. Abdollahi, A. von Deimling, J. Debus, C. Hartmann, Prognostic significance of IDH-1 and MGMT in patients with glioblastoma: one step forward, and one step back?, Radiat Oncol 6 (2011) 115.

[12] M.D. Walker, S.B. Green, D.P. Byar, E. Alexander, Jr., U. Batzdorf, W.H. Brooks, W.E.

Hunt, C.S. MacCarty, M.S. Mahaley, Jr., J. Mealey, Jr., G. Owens, J. Ransohoff, 2nd, J.T.

在文檔中膠質母細胞瘤之個人化治療: 以影像生物標記預測腫瘤進展模式與發展對應之腫瘤內藥物傳輸系統 (頁 48-0)