Review of Intraoperative Adjuncts for Maximal Safe Resection of Gliomas and Its Impact on Outcomes

Simple Summary Understanding the impact of intraoperative modalities in glioma surgery on the extent of resection (EOR), survival, and complications is vital to maximizing safe resection while preserving neurological function. A systematic literature search was performed to assess the impact of intraoperative modalities of glioma surgery, including one or a combination of the following: intraoperative magnetic resonance imaging (iMRI), awake/general anesthesia craniotomy mapping (AC/GA), fluorescence-guided imaging, or combined modalities. The heterogeneity in reporting the amount of surgical resection prevented further analysis. The studies reviewed indicated that these modalities significantly improved EOR but most often underreported Progression-free survival/overall survival (PFS/OS). Combining intraoperative modalities during the same brain glioma operation seems to have the highest effect compared to each modality alone. Abstract Maximal safe resection is the mainstay of treatment in the neurosurgical management of gliomas, and preserving functional integrity is linked to favorable outcomes. How these modalities differ in their effectiveness on the extent of resection (EOR), survival, and complications remains unknown. A systematic literature search was performed with the following inclusion criteria: published between 2005 and 2022, involving brain glioma surgery, and including one or a combination of intraoperative modalities: intraoperative magnetic resonance imaging (iMRI), awake/general anesthesia craniotomy mapping (AC/GA), fluorescence-guided imaging, or combined modalities. Of 525 articles, 464 were excluded and 61 articles were included, involving 5221 glioma patients, 7(11.4%) articles used iMRI, 21(36.8%) used cortical mapping, 15(24.5%) used 5-aminolevulinic acid (5-ALA) or fluorescein sodium, and 18(29.5%) used combined modalities. The heterogeneity in reporting the amount of surgical resection prevented further analysis. Progression-free survival/overall survival (PFS/OS) were reported in 18/61(29.5%) articles, while complications and permanent disability were reported in 38/61(62.2%) articles. The reviewed studies demonstrate that intraoperative adjuncts such as iMRI, AC/GA mapping, fluorescence-guided imaging, and a combination of these modalities improve EOR. However, PFS/OS were underreported. Combining multiple intraoperative modalities seems to have the highest effect compared to each adjunct alone.


Introduction
Gliomas are the most common primary intracranial tumors in adult patients. The age-standardized incidence rate of gliomas was 4.7 per 100,000 person-years [1]. Around 8000 new cases are diagnosed annually in the United States [2]. Gliomas originate from the glial cells of the central nervous system and are graded from I to IV according to the World Health Organization (WHO) classification. The most common histologic types in adult patients include glioblastoma (GBM) (grade IV), astrocytic tumors (grade I-III), oligodendroglioma (grade II-III), and ependymomas (grade I-III) [3]. Among gliomas, GBM holds the poorest prognosis, with 2-year survival rarely exceeding 10% [4]. Maximal safe resection is the standard of treatment in the neurosurgical management of gliomas [5]. Enhancing cytoreduction and minimizing neurologic deficits can be especially critical when gliomas involve or approach eloquent regions of the brain [6]. Achieving a high extent of resection (EOR) while preserving functional integrity has been shown to improve surgical outcomes and overall survival (OS) [5,7].
Adjuvant modalities and technological enhancements have led to improved identification and visual discrimination of the tumor-brain margin to assist neurosurgeons in the operating room [8]. Intraoperative imaging techniques such as intraoperative magnetic resonance imaging (iMRI) and fluorescence imaging using 5-aminolevulinic acid (5-ALA) and sodium fluorescein have shown an increased EOR in lesions primarily amenable to subtotal resection (STR) [8,9]. Furthermore, when operating around eloquent areas, awake or asleep surgery with intraoperative neuromonitoring aids has enhanced the detection of safe corridors for tumor access and provided accurate real-time feedback regarding critical structures such as motor and language areas [10,11].
How these modalities differ in their effectiveness in increasing the EOR, survival benefit, and complication profile remains relatively understudied. Understanding each adjuvant technological advancement and its impact on surgical outcomes is vital to guide the surgeon's choice of modality for enhanced surgical precision. The objective of this narrative review was to outline the current literature regarding the advances of intraoperative modalities in glioma surgery utilized to achieve optimal resection.

Materials and Methods
An advanced PubMed search was performed in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines ( Figure 1) [12], using the following key search terms: "glioma", "glioblastoma", "GBM", "MRI", "monitoring", "imaging", "mapping", "awake, "5-ALA", "fluorescence", "resection", "adjunct", and "removal". No institutional review board approval was required since no original data were used. Inclusion criteria were studies published between 2005 and 2022, involving glioma surgery, and including one or a combination of intraoperative modalities such as iMRI, brain mapping through an awake craniotomy (AC) or general anesthesia (GA), or fluorescenceguided imaging. Exclusion criteria were non-human studies, review articles including narrative reviews, systematic reviews and meta-analyses, non-surgical management of gliomas, pediatric patients, spinal gliomas, case reports, and non-English language articles. Considering the rapid evolution of adjuncts used in glioma surgery over the last several years the studies prior to 2005 were excluded.
The primary outcome was the EOR, which was reported either as the percentage of resection, type of resection (i.e., gross total resection (GTR), subtotal resection (STR), etc.), or the amount of residual volume (cm 3 ). Secondary outcomes were progression-free survival (PFS) and (OS), as well as postoperative complications. Postoperative transient neurologic deficits were not reported. Other collected variables included study period, study design, sample size, histologic diagnosis reported as GBM vs. lower-grade gliomas (I-II-III), and the location of tumors. The primary outcome was the EOR, which was reported either as the percentage of resection, type of resection (i.e., gross total resection (GTR), subtotal resection (STR), etc.), or the amount of residual volume (cm 3 ). Secondary outcomes were progression-free survival (PFS) and overall survival (OS), as well as postoperative complications. Postoperative transient neurologic deficits were not reported. Other collected variables included study period, study design, sample size, histologic diagnosis reported as GBM vs. lowergrade gliomas (I-II-III), and the location of tumors.

Intraoperative Magnetic Resonance Imaging
In the early 1990s, neurosurgeons endorsed MRI technology in the operating room to improve tumor detection and enhance the resection of brain tumors [13]. iMRI assists neurosurgeons in demarcating the margins of the tumor and accounts for brain shifts on neuronavigation [14].

Intraoperative Magnetic Resonance Imaging
In the early 1990s, neurosurgeons endorsed MRI technology in the operating room to improve tumor detection and enhance the resection of brain tumors [13]. iMRI assists neurosurgeons in demarcating the margins of the tumor and accounts for brain shifts on neuronavigation [14].
A total of 1170 patients were included in 4 prospective and 3 retrospective studies [15][16][17][18][19][20][21]. Moreover, 3/7 articles included lower-grade gliomas, and all 7 articles included patients with GBM. Studies that involved iMRI are summarized in Table 1. In a retrospective study of 135 patients with GBM, Kuhnt et al. [17] identified residual tumors in 65% of cases using iMRI, which allowed further resection in 19 patients. Patients with GTR had an increased OS reaching 14 months compared to 9 months in the rest of the cohort, and 1 (0.7%) patient had a permanent language deficit. In a prospective study of 141 GBM and 83 (37.1%) lower-grade gliomas (I-III), Scherer et al. [16] performed an additional resection in 70% of patients after iMRI; however, 15 (6.7%) patients had a permanent new neurological deficit. While iMRI has demonstrated a significant EOR improvement in patients with GBM and lower-grade gliomas, the impact on PFS and OS was not well established. In a randomized controlled trial of 49 patients with GBM and lower-grade gliomas, Senft et al. [18] found a longer PFS in patients undergoing glioma resection under iMRI compared to the control group, but this did not reach statistical significance (p = 0.083). In another multicenter study, Shah et al. [21] found that while iMRI increased EOR, iMRI was not an independent predictor of OS.
Despite the increased operative time and the carried risk of infection and prolonged anesthesia, iMRI was demonstrated to be cost-effective with an incremental benefit of 0.18 quality-adjusted life years in high-grade gliomas [22].

Awake vs. Asleep Cortical and Subcortical Mapping
Intraoperative stimulation mapping was developed by Penfield and Cushing in the early 1900s [23] and is considered currently the gold standard in identifying eloquent areas of the brain. Mapping allows accurate real-time detection of areas with important functionality, in addition to tumor-related epileptic foci [24,25]. Mapping of eloquent areas is performed under either GA or AC protocol [26]. In AC, patients are asked to perform motor tasks to evaluate muscle strength and coordinate subcortical stimulation, which would ensure subcortical tract integrity prior to tumor resection.
As summarized in Table 2A-C, a total of 1961 patients were included in 3 prospective, 17 retrospective studies, and 1 combined prospective and retrospective study . Of 21 articles using brain mapping, 18 articles included lower-grade gliomas, and 16 articles included patients with GBM.
GA mapping alone was used in 3 studies and included 346 patients (Table 2A). With regards to GA monitoring, Duffau et al. [38] found a higher total resection rate in patients undergoing lower-grade gliomas resection in patients with GA monitoring (25.4%) compared to patients without (6%), with no significant impact on OS.
Of 21 included studies, 9 studies included both GA and AC cohorts, involving a total of 1211 patients (Table 2C). While previous literature was in relative agreement regarding the established effect of AC and GA mapping on increased EOR, improved OS, and decreased postoperative morbidity, the superiority of one protocol against the other was hardly found in the literature, with not enough comparison of PFS, OS, and complications between patients with AC and GA monitoring [10,26,47,48]. Table 2. (A-C) Studies investigating the role of intraoperative mapping through awake and general anesthesia monitoring in improving the extent of resection: (A) general anesthesia; (B) awake craniotomy; (C) general anesthesia and awake craniotomy.

Fluorescence-Guided Imaging
Under the operating microscope, the similarity between tumor appearance and nearby brain parenchyma makes it challenging to achieve a complete tumor resection. Fluorescence-guided surgery has provided neurosurgeons with real-time distinction of not only the tumor core, but also the tumor-brain interface that defines the extent of resection [49]. Biomarkers that were mostly investigated to maximize the extent of glioma resection are 5-ALA and fluorescein sodium.
In 1947, fluorescein sodium was introduced to identify malignant gliomas. Fluorescein sodium is given intravenously, and it enters the brain in the areas of blood-brain barrier breakdown, where it resides in the extracellular space. Fluorescein sodium can be grossly perceived by the naked eye and allows real-time visualization of the areas contrasted by gadolinium in MRI [65]. Of 15 included studies, fluorescein sodium alone was used in 7 studies and included 275 patients (Table 3A). In a retrospective study of 48 patients with GBM, Catapano et al. [62] found a higher GTR (82.6%) in patients with sodium fluoresceinguided removal compared to controls (36%). Acerbi et al. [56] was the first group to initiate a phase II clinical trial in 20 patients undergoing high-grade glioma resection using sodium fluorescence and found that 80% of patients achieved a complete resection, with a sensitivity of 80% in detecting tumor tissue. In another prospective study conducted by Chen et al. [60], patients with fluorescein sodium showed a longer PFS compared to the control group (7.2 vs. 5.4 months). While previous studies investigated EOR and PFS in patients undergoing glioma removal with fluorescein sodium, the impact on OS remains understudied.
On the other hand, 5-ALA is ingested orally prior to surgery and is rapidly absorbed into the bloodstream and converted into protoporphyrin IX intracellularly. The latter molecule emits fluorescence when excited by blue-violet light conditions, facilitating the identification of tumor tissues [66]. Of 15 included studies, 5-ALA imaging was used in 7 studies and included 563 patients (Table 3B). In a phase III randomized controlled trial performed by Stummer et al. [53] involving 237 GBM and 85 lower-grade gliomas, patients who received 5-ALA had a higher GTR and PFS compared to the control group. Only one study utilized both 5-ALA and fluorescein sodium (Table 3C). When comparing 5-ALA and fluorescein sodium, Zeppa et al. [50] included 99 patients with GBM and found no significant difference in EOR or OS between 5-ALA alone, sodium fluorescein alone, and a combined 5-ALA sodium fluorescein-guided imaging.
A combination of 5-ALA and GA with mapping was also reported [81,83]. In a prospective study of 36 patients with various grade gliomas undergoing surgical resection, Feigl et al. [81] found a 64% GTR rate when combining 5-ALA and GA with mapping. In another prospective study of 31 patients with low-grade and high-grade gliomas, GTR was achieved in 74% of patients when combining 5-ALA with either AC or GA with mapping [83]. These studies concluded that the combination of different intraoperative modalities seems to have a synergistic overall effect on EOR and overall survival.   Other modalities in adjunct to the aforementioned imaging include intraoperative or contrast-enhanced ultrasound. Ultrasound is cost-effective, widely available, offers real-time visualization of gliomas, and improves the accuracy of navigation. Additionally, doppler ultrasound tracks the flow patterns through the tumor's microvessels using radiation-free images [85]. Prada et al. [86] found an increased EOR in patients undergoing GBM resection using contrast-enhanced ultrasound. Furthermore, in a retrospective study of 75 patients with GBM, Neidert et al. [87] found an increased OS and PFS with intraoperative ultrasound compared to the non-ultrasound group.
A few studies have described the use of intraoperative histology using stimulated Raman scattering microscopy, which was introduced in 2008 [8,88]. Artificial intelligence has shown high accuracy in differentiating high-and low-density tumor regions, detecting tumor subtypes, and distinguishing infiltrative cells at tumor margins through handheld probes, which showed a potential improvement in EOR and subsequently PFS and OS. [89,90] Nonetheless, the number of studies reporting the use of intraoperative histology is limited.

Discussion
Considering the plethora of advancements in neurosurgery, we conducted a systematic review to summarize the recent investigations of the intraoperative modalities that assist the surgical treatment of gliomas, including iMRI, awake and asleep cortical and subcortical mapping, fluorescence-guided imaging, and combined modalities. While a number of studies have been published focusing on each adjuvant modality, studies comparing one adjuvant intraoperative modality with another are lacking.
The use of iMRI as an adjunct was associated with an increase in EOR; however, its impact on PFS and OS was limited. Though iMRI offers a vast range of advantages and intraoperative capabilities in improving EOR, it is limited by its feasibility and cost. Furthermore, special infrastructure requirements, as well as special expertise, are required to operate the use of iMRI [91].
Patients undergoing glioma removal under AC or GA monitoring have shown an increased EOR, PFS, and OS. However, the superiority of AC over GA remains to be elucidated. When considering AC, patient selection, appropriate patient selection, and coordination between experienced anesthesia, surgery, neuromonitoring, neurology, and neuropsychology teams are required. The risk of intraoperative seizures, postoperative emotional distress, and the need for conversion from AC to GA should be considered [26,32,39,92]. Regardless of the AC or GA approach, cortical and subcortical mapping of motor areas facilitates maximal safe resection. Studies have demonstrated that motor evoked potential (MEP) in motor mapping exhibits high reliability with few to no false-negative monitoring [93]. For the speech region and insular gliomas, AC with intraoperative language mapping is associated with improved EOR [26,94]. Studies comparing AC and GA for these eloquent regions are sparse.
While more studies have already established the significant impact of 5-ALA on EOR, PFS, and OS, research involving fluorescein sodium remains minimal. With regard to possible disadvantages of fluorescence-guided images, previous evidence showed an increased fluorescence in both low-and high-grade gliomas, some studies witnessed a lack of visible fluorescence in grade I and grade II gliomas, rendering high-grade gliomas more sensitive to fluorescent biomarkers [95]. Side effects of fluorescent agents include a sensitization of the skin that lasts for 24 h after ingestion. Furthermore, reactive astrocytosis, highly vascularized areas, and inflamed tissues may emit fluorescence regardless of the presence of malignancies [67]. Moreover, due to the lack of uptake of fluorescent biomarkers in nonvascularized tissues, the necrotic portion of the tumor may not be accurately detected [61]. Other downsides include the requirement of a dark surgical field, the cost associated with fluorescent biomarkers, and the time required for 5-ALA to reach peak concentration in blood [56,96].
Combining different intraoperative modalities is associated with a significant increase in EOR, PFS, and OS while keeping a safe complication profile. Cost-effectiveness, however, may play an important role in decision making. While other combined modalities have shown a greater effect compared to each surgical adjunct alone, randomized clinical trials are still lacking.
Newer advanced visualization technologies such as exoscope, contrast-enhanced ultrasound, intraoperative histopathology, imaging probe devices, stimulated Raman scattering microscopy, probe-based microscopy, and wide-field endomicroscopy are currently being tested and implemented intraoperatively [8]. Therefore, surgeons may employ the adjuncts based on their own expertise with various technologies, tumor histology and location, and cost-effectiveness. Other intraoperative modalities include fluorescence using indocyanine green, which demonstrated an increased affinity to endothelial cells in the peritumoral tissue [97], theranostics using optical coherence tomography that improved resection at a cellular-level precision [98], and other imaging modalities such as stereotactic navigation [99], whole-brain tractography, and diffusion tensor imaging [100].
This systematic review has some limitations that warrant further discussion. First, the allocated time for the search strategy (2005-2022), though arbitrary, aimed at removing non-relevant studies and narrowing the review of recent articles regarding intraoperative modalities. Second, the modalities that were explored in the listed studies may not be mutually exclusive. For example, some studies may have focused on iMRI but also have also used GA monitoring without mentioning it. Third, the lack of consistency in reporting EOR prevented further analysis. Fourth, while the current objectives were to report the PFS and OS in studies investigating intraoperative modalities for glioma resection, these parameters were underreported, likely due to the previously established correlation between EOR and prolonged survival, and some studies may have waived the need to explore this area of research. Finally, the heterogeneity of the location of gliomas was not accounted for in most of these studies, which might have impacted the EOR, survival analysis, and operative complications. The impact of postoperative adjuvant chemotherapy and radiotherapy on PFS and OS is not well reported in the studies reviewed. Furthermore, most studies investigating intraoperative modalities comprised small sample sizes. Largescale multicenter studies are necessary to accurately establish the differences in EOR, PFS, and OS with the use of these adjuvant intraoperative modalities.

Conclusions
Intraoperative adjuncts such as iMRI, AC and GA mapping, fluorescence-guided imaging (5-ALA and fluorescein sodium), and a combination of these modalities have the potential to improve the extent of maximal safe resection for gliomas. The impact of using these adjuncts on PFS and OS is underreported. Combining multiple intraoperative modalities seems to have the highest effect compared to each adjunct alone. Future prospective studies with larger sample sizes are warranted to evaluate the differences between these modalities and their impact on short-and long-term outcomes.