Table 2.

Characteristics of included studies.

Authors, year, journal	Type of study	Patients with/without CWs	Adjuvant therapy	Grade of Glioma	Molecular markers	EOR	OS results	PFS results	Side effects
De Bonis et al.,2012, Acta Neurochir (Wien) (19)	Randomized controlled trial	10/67	Adjuvant therapy with TMZ	IV	NA	Non volumetric study	Adding CWs to standard treatment did not significantly improve the outcome Multivariate analysis showed the only was resection extent (p = 0.048)	NA	The toxicity after CW use was significantly higher, both for patients with newly diagnosed and patients with recurrent glioblastoma
Pallud et al., 2015, Neuro Oncol (10)	Randomized controlled trial	354/433	Chemoradiation standard protocol	IV	NA	Surgical resection at progression whether alone or combined with CW implantation was independently associated with longer overall survival in the whole series (p = 0.0001)	The median overall survival was 20.4 months and 18.0 months in the CWs group and non CWs group respectively	The median PFS was 12.0 months and 10.0 months in the CWs group and non CWs group respectively	The higher postoperative infection rate in the implantation group did not affect survival
Roux et al., 2017, J Neurooncol (11)	Randomized controlled trial	123/217	Standard combined chemoradiotherapy	IV	NA	Volumetric estimation In CWs group and non-CWs group the Subtotal (90% and >) and total (100%) removal were achieved in 55.6 and 55.1% of cases, respectively (p = 0.887)	CWs implantation was were independently associated with longer OS (p = 0.029)	CWs implantation was were independently associated with longer PFS (p = 0.045)	CWs did not significantly increase postoperative complications, including postoperative infections (p = 0.269, and p = 0.446, respectively)
Akiyama et al., 2018, World Neurosurg (20)	Randomized controlled trial	25/29	Standard combined chemoradiotherapy	IV	Evaluation of the IDH-1/2 mutation, which has been reported as a predictive factor, was performed in only a small percentage of patients	Volumetric estimation The median EOR was 93% in CWs group vs. 96% in non CWs gruop (p = 0.129)	The median OS in the CWs group and non CWs group was 24.2 months and 15.30 respectively (p = 0.027)	The median PFS in the CWs group and non CWs group was 16.8 months and 7.30 months, respectively (p = 0.009)	The incidence of adverse events were similar between the treatment groups, except for infection that was more common in the CWs patients (3.5% vs. 0%)

CWs, Carmustine Wafers, EOR, extent of resection, NA, not applicable, PFS, Progression-free survival, OS, overall survival.