Multimodal treatment of glioblastoma with multiple lesions - a multi-center retrospective analysis

Abstract

Objective: The presence of multiple localizations (ML) in glioblastoma is rare and associated with perceived poor prognosis. The aim of this study is to evaluate the impact of a multimodal treatment on progression-free survival (PFS) and overall survival (OS) in ML glioblastoma.

Methods: Patients presenting with CNS WHO grade 4 glioblastoma with ML to 2 major German Departments of Neurosurgery between January 1st, 2008, to December 31st, 2020 were included in this study. Primary outcome parameters were extent of resection (EOR) using the 2021 RANO criteria, progression free- and overall survival.

Results: A total of 483 patients with newly diagnosed glioblastoma (CNS WHO grade 4) were assessed. 134 patients presented with ML (72 multifocal (MF), 62 multicentric (MC)). The median PFS and OS did not differ among MC and MF glioblastomas. The EOR was a significant predictor of PFS and OS in ML glioblastoma. complete-, near total-, and subtotal resection significantly prolonged PFS (p < 0.0001) and OS (p < 0.0001) compared to biopsy alone. Standard radiotherapy (p = 0.045) and hypofractionated (p < 0.0001) radiotherapy and adjuvant treatment (Stupp protocol) prolonged PFS (p = 0.0012) and OS (p < 0.0001). In multivariate analysis Karnfosky performance score, EOR, and concomitant adjuvant treatment remained significant factors influencing OS. Propensity score matching of patients with ML and solitary lesion tumors showed similar PFS and OS (p = 0.08).

Conclusion: The presented data suggests that glioblastomas with multiple lesions treated with multimodal therapy equal survival rates compared to patients with solitary lesion tumors can be achieved. The results reflect the importance of an equally aggressive maximal treatment effort in this particular and often marginalized group of patients.

Keywords: Extent of resection; Glioblastoma; Multicentric; Multifocal; Multimodal therapy.

Fig. 1

MRI T1 with Gadolinium contrast enhancement (upper) and T2 FLAIR (lower). Solitary lesion glioblastoma (A), multicentric glioblastoma without FLAIR interconnection (B) and multifocal glioblastoma (C)

Fig. 2

Survival of Glioblastoma with solitary and multiple lesions: a) The mean PFS (p = 0.084) and OS (p = 0.287) are similar in SL and ML glioblastoma. b) To adjust for biological and treatment differences ML and SL tumors were matched for age, ECOG, EOR, MGMT promotor methylation and adjuvant treatment according to the Stupp protocol using propensity score matching. The median PFS and OS of patients with ML (4.86 months, 95%CI 0.67–1.4 and 10.9 months, 95%CI 0.87–1.76 respectively) was statistically not different from those with SL (4.86 months, 95%CI 0.71-1.50- and 9.11-months 95%CI 0.57–1.15 respectively) (p = 0.08). c) Surgical resection was associated with improved survival in both groups compared to biopsy (p < 0.0001)

Fig. 3

Univariate association of multiple lesions, hemisphere involvement and MGMT status with progression-free (upper) and overall survival (lower). MF and MC have similar PFS (p = 0.062) and OS (p = 0.74) (a). Involvement of both hemispheres is not associated with impaired outcome (p = 0.186), while MGMT methylation is a predictor for prolonged OS (p = 0.024) (b + c)

Fig. 4

Univariate association of surgery (EOR), radiotherapy and adjuvant treatment with progression-free (upper) and overall survival (lower). CR, NTR, and STR significantly prolonged PFS (HR: 0.28, 95% CI: 0.10–0.81, p < 0.0003; HR 0.26, 95% CI: 0.08–0.80, p < 0.0001; and HR: 0.40, 95% CI: 0.09–0.80, p = 0.0068 respectively) and OS (HR: 0.28, 95% CI: 0.14–0.56, p < 0.0001; HR 0.33, 95% CI: 0.17–0.63, p < 0.0001 and HR: 0.32, 95% CI: 0.17–0.62, p = 0.0068) compared to biopsy alone (a). cRT was associated with prolonged PFS (HR: 0.11, 95% CI: 0.004–3.17, p < 0.0001) while both, cRT and hfRT results to prolonged OS (HR: 0.41, 95% CI: 0.13–1.29, p = 0.045; HR: 0.13, 95% CI: 0.02–0.93, p < 0.0001) (b). Treatment according to the Stupp protocol was associated with prolonged PFS (HR: 0.49, 95% CI: 0.28–0.86, p = 0.0012) and OS (HR: 0.44, 95% CI: 0.26–0.73, p < 0.0001) (c)