The use of cranial US was previously limited by the probe size because it was too big for access via the opening made by a craniotomy, and the images were not very clear compared with computed tomography (CT) or MRI.[58] In the recent few decades, however, these issues have been addressed: the imaging quality has improved greatly with three-dimensional acquisition and advanced computer technology, and the probe size has become more suitable for craniotomies.[58] The most significant development in iUS was the introduction of a probe for superficial lesions (7.5–10 MHz) and deep lesions (3–5 MHz).[64] To the best of our knowledge, the most notable remaining problem is that brain shifting during surgery, which can vary from 2 to 25 mm depending on the lesion location and size,[65] cannot be avoided. The results of the meta-analysis of the single-arm studies indicated that the postoperative neurological status was worse in patients with iUS: 11.3% (range from 8% to 13%) of patients exhibited no postoperative neurological improvement and 19% were better off.[58,60,66] Furthermore, studies have shown that there were no more durable deficits when iUS was used alone than with conventional surgery.[47,67] Both iMRI and iUS can correct for errors caused by brain shift, but the limitation with US arises from the reduction in quality and resolution of the imaging because of artifacts associated with the resection procedure.[68] The sensitivity, specificity, and positive and negative predictive values of US were the greatest at the start of surgery (95%, 95%, 98%, and 90%, respectively), as surgery proceeded the four values decreased to 88%, 42%, 73%, and 67%, respectively, and 26%, 88%, 62%, and 62%, respectively, at the end of the surgery.[58] Another issue with iUS is that surgeons must be familiar with the anatomic orientation, though some strategies have been developed to mitigate this.[69] It is important to note that the benefits of MRI and US are highly dependent on the skill and experience of the surgeon. Another limitation of US is that it cannot be used in some cases depending on the size or location of the skull opening.[57] However, when we perform craniotomies, iUS is used for planning the surgical incision and identifying the important surrounding structures.[70] Direct comparisons between iUS and iMRI are not always possible. However, a few studies have been done to compare the sensitivities and specificities of these technologies. One study comparing two-dimensional iUS with low-field MRI among 26 patients showed that MRI outperformed iUS.[71] Another study comparing high-field MRI with linear-array US in 44 Grade II astrocytoma biopsy specimens showed that the specificity was 67% for both technologies, and the sensitivity of MRI was higher than that of US (83% versus 79%).[72] In another study, the sensitivity of US and MRI was 75% and 55%, respectively, and the specificity was 58% and 96%, respectively.[73] Our study reported that MRI was clearly the best tool for surgical guidance, but considering the high cost of MRI, iUS is an acceptable alternative.

Nineteen studies reported information about postoperative complications with four different technologies. The network chart is shown in Figure 5. During the surgical resection of a brain tumor, various complications might occur such as aphasia, epilepsy, and dyskinesia. On direct comparisons, MRI was better than neuronavigation and conventional surgery, and neuronavigation was better than conventional surgery [Table 1]. The heterogeneity was moderate, and there was a slight inconsistency for the MRI and conventional surgery data according to the network meta-analysis [Table 2]. The results of the consistency model showing the different comparisons are shown in Figure 6. Thus, the following ranking was derived: conventional surgery > neuronavigation > iMRI > iUS [Figure 7].

5-Aminolevulinic acid; conventional surgery; glioma; intraoperative; magnetic resonance imaging; network analysis; neuronavigation; resection; ultrasound

Gliomas are the most common type of brain tumor in the world. Surgical resection is one of the most effective therapeutic methods in terms of patient prognosis. However, it is difficult for neurosurgeons and health-care providers to select which imaging technology to best support the procedure. These technologies included intraoperative magnetic resonance imaging (iMRI), intraoperative ultrasound (iUS), fluorescence guidance with 5-aminolevulinic acid (5-ALA), and intraoperative neuronavigation. Hence, in this study, we compared the gross total resection (GTR), postoperative complications within or outside of the central nervous system, and postoperative clinical improvement by multiple meta-analyses, which allows the integration of data through direct and indirect comparisons.

The PE300BFA features high-intensity illumination - 5000 Lumens, with a power range of 180-320 Watts. It has a broad spectral range with 5900-degree Kelvin color temperature, and a lamp lifetime of 1,000 hours.

Modern technologies are highly advanced to guide surgeons during glioma resection. Our network meta-analysis elucidated the added value of different technologies for surgical guidance in patients with gliomas in terms of rGTR, postoperative complications, and clinical improvement.

It should be noted that our meta-analysis is limited by the quality of the underlying studies upon which it was based. However, a significant benefit of network meta-analyses is that large amounts of information can be aggregated, leading to higher statistical power and stronger point estimates than is possible from the measures derived from any individual study. Other limitations of our meta-analysis include the small sample size of some studies, relatively short follow-up, and the inconsistent use of various adjuvant therapies. Subgroup analyses were not performed, but future studies could investigate the effects of different tumor stages, different tumor histologies, adjuvant chemotherapy, patient age, MRI field strength, or tumor location. Another limitation was that six studies defined GTR as resection of 95% or 98% rather than 100%; these variations in the definition were permitted because there is no clinically significant difference between 100% and 98% tumor resection.[54] Although scientists, economists, and health-care policymakers would like to see prospective random studies comparing these technologies, such a study would be practically challenging because these technologies are complementary to each other and are generally used in combination. However, this analysis is of great clinical value as it provides insight into the benefit of image guidance with respect to the GTR, postoperative complications, and clinical improvement. Yet, our study and the limitations of the published literature emphasize the need for larger studies to evaluate the importance of imaging-guided technologies for treating gliomas.

The developed ranking probability table indicated that iMRI was superior in terms of the GTR and clinical improvement, while iUS was the least likely to result in postoperative complications. Hence, it was concluded that iMRI or iUS is the most advantageous imaging modality.

In terms of GTR, iMRI (odds ratio [OR] = 5.70, 95% confidence interval [CI]: 3.40–9.60), iUS (OR = 2.70, 95% CI: 1.10–6.90), 5-ALA (OR = 2.40, 95% CI: 0.64–8.90), and neuronavigation (OR = 1.90, 95% CI: 1.20–3.10) were found to be more effective than conventional surgery. In addition, iUS (OR = 0.15, 95% CI: 0.04–0.52), iMRI (OR = 0.24, 95% CI: 0.14–0.43), and neuronavigation (OR = 0.34, 95% CI: 0.18–0.56) were more found to result in fewer complications than conventional surgery. Furthermore, patients' clinical improvement was better with iMRI (OR = 8.10, 95% CI: 3.00–25.00), iUS (OR = 4.90, 95% CI: 0.76–33.00), and neuronavigation (OR = 2.60, 95% CI: 1.00–7.20) than with conventional surgery.

First, we conducted a pair-wise meta-analysis of studies by comparing the same interventions with a random-effects model[9,10] to combine the conclusions of different studies. Forest plots were used to assess the statistical heterogeneity.[11] The analysis was conducted using Review Manager, version 5.3 (RevMan 5; The Cochrane Collaboration, 2015).

In the past decade, there has been great interest in the use of iMRI because it can provide highly valuable information and real-time feedback on the extent of resection and presence of residual neoplasm during image-guided tumor resection. High-field MRI with an integrated neuronavigation system has been found to provide reliable anatomic and functional data during surgery.[62] The real-time visualization of the space between the white matter fiber tracts and the tumor helps the surgeon avoid the tract, which should reduce the risk of neurological deficits, particularly in the case of infiltrating tumors such as gliomas.[62] Napolitano et al.[22] reported opposite results demonstrating that iMRI improved the quality of resection and the GTR. Our data supported the idea that the use of iMRI could help neurosurgeons achieve the goal of extensive but safe tumor resection. However, the limitations of iMRI are its high cost and the time it adds to a surgery. Moreover, it is an offline method, meaning that its use requires pausing the surgery to assess the results.[63] iMRI scan prolongs the time of surgery, which may lead to an increase in postoperative complications. The most relevant risks were correlated with pressure- and heat-related skin damage because of the prolonged surgery time and possible heating of the patients in the iMRI scanner.

Thirty-eight studies reporting the GTR rates with the different surgical guidance techniques were included in the network meta-analysis. The network chart of the meta-analysis is shown in Figure 2. Direct comparisons suggested that the rate of GTR (rGTR) is better with iMRI than with neuronavigation, 5-ALA, and conventional surgery, and those of 5-ALA and neuronavigation were better than that of conventional surgery [Table 1]. The heterogeneity was moderate overall, though most comparisons showed high heterogeneity. A node-split analysis was used to evaluate the consistency between the direct and indirect comparisons [Table 2]. The results showed that there was a slight inconsistency in the comparisons between iMRI and neuronavigation/conventional surgery. The results of the consistency model for the network analysis are shown in Figure 3. The random model was selected because of the high heterogeneity. The posterior probabilities of the five methods of surgery were ranked as follows: MRI > iUS ≈ 5-ALA > neuronavigation > conventional surgery [Figure 4].

In addition to lamps, Excelitas Technologies manufactures Cermax® Xenon short-arc lamp power supplies, lamp holders, OEM lighting systems, and fiber optic light sources.

5-ALA is based on the fluorescence of protoporphyrin IX accumulating in glioma cells, acting as a biological tag of these cells; this makes it easy for the surgeon to locate, observe, and resect the glioma while avoiding damage to other brain tissues.[57] One of the important and undisputed values of protoporphyrin IX spectroscopy is that it can facilitate the standardization of ALA-fluorescence image-guided neurosurgery (FIGS) compared with the human eye, which has wide variations in color perception; it might be particularly useful for color-blind surgeons.[74] A single-arm meta-analysis of eight studies reported that the specificity of 5-ALA-fluorescence image guided surgical resection was 88.8%, and its sensitivity was 82.6%.[57] Thus, more and more clinical trials have been done to seek the US Food and Drug Administration approval for the use of 5-ALA in malignant glioma resection.[75] 5-ALA-based imaging is limited by heterogeneity within the tumor and signal attenuation in necrotic and malignant areas, making it difficult for surgeons to recognize the lesions. Yet, ALA-FIGS offers a great advantage in that it is easy to learn and integrate into practice.[76] Because there are no two-arm studies comparing ALA-FIGS with other technologies, unfortunately, ALA was not involved in the outcomes of postoperative complications and clinical improvement, so a comprehensive assessment could not be made.

2Department of Neurosurgery, Cancer Hospital of China Medical University, Liaoning Cancer Hospital and Institute, Shenyang, Liaoning Province, China

Conventional surgery without imaging assistance is not widely used for the sake of patient outcomes. Most of the data related to conventional surgery and other technologies included in this network meta-analysis were obtained from the China Knowledge Resource Integrated Database and the WanFang database. The data showed that the outcomes of conventional surgery were notably different from those of imaging-guided surgery. The outcomes of conventional surgery are greatly dependent on the surgeon's experience in recognizing the brain tissue, blood vessels, and nerves. This finding was supported by our meta-analysis: conventional surgery was ranked the lowest in terms of all the three metrics evaluated. Hence, we recommend the use of image guidance technology if the cost is acceptable.

The PubMed, Cochrane Library, Web of Science, Embase, China Knowledge Resource Integrated Database, and WanFang databases were searched for publications before April 2018. Randomized controlled trials, two-arm and three-arm prospective studies, and retrospective studies in patients who underwent surgical treatment for glioma were included. The most important outcome measures were the rates of GTR, postoperative complications, and clinical improvement.

One of the most important assumptions behind the network meta-analysis is coherence, which means that the direct and indirect evidence do not disagree beyond chance.[13] A node-split analysis was used to assess the consistency of the direct and indirect evidence for each loop in the network.[14] The node-split and network analyses were both run in R version 3.5.0.

The name of the first author, study design, year of publication, number of patients in each group, and major outcomes were collected from each article. The Newcastle–Ottawa Quality Assessment Scale[8] was used to evaluate the quality of each included article.

The results of our network meta-analysis suggest that image-guided surgery may be more effective than conventional surgery for tumor resection and support the use of iMRI, iUS, 5-ALA, and neuronavigation as valuable tools for glioma surgery. However, large randomized clinical trials are required to further elucidate the true value of image-guided surgery.

1Department of Medical Imaging, Cancer Hospital of China Medical University, Liaoning Cancer Hospital and Institute, Shenyang, Liaoning Province, China

The medical literature in PubMed, Cochrane Library, Web of Science, Embase, the China Knowledge Resource Integrated Database, and the WanFang database was searched by using the medical subject heading (MeSH) terms “glioma,” “intraoperative,” “MRI,” “magnetic resonance imaging,” “ultrasound,” “ALA,” “5-aminolevulinic acid,” “neuronavigation,” “conventional surgery,” “image guidance,” and “neurosurgery” in different combinations. All of the databases were searched up to April 2018. Randomized controlled trials, two-arm and three-arm prospective studies, and retrospective and prospective studies in patients with glioma who received surgical treatment were included. Single-arm studies and studies that did not completely report the outcomes of interest were excluded. Reviews, letters, comments, editorials, case reports, and personal communications were excluded. Two independent reviewers screened each article, and if there was uncertainty regarding eligibility, a third reviewer was consulted. The references in each included article were double-checked to ensure that all related articles were included.

Considering the rGTR, it has been demonstrated that the complete resection of a tumor is related to better survival.[53,54,55,56] Hence, it is wise for surgeons to use intraoperative guidance technologies to achieve maximal safe surgical resection. We also compared these guidance technologies with conventional surgery (i.e., surgery without any image guidance technology) as a control to evaluate the benefit associated with each technology. The GTR was defined based on the postoperative MRI within 1 week after surgery as described previously.[57,58] The results showed that iMRI was associated with the best GTR, followed by iUS, 5-ALA, and neuronavigation; the GTR was the worst with conventional surgery. There was a slight inconsistency with iMRI and neuronavigation and conventional surgery in our network meta-analysis. We tried to evaluate whether the slight inconsistency could influence the posterior probability ranking of the five methods. We found that the OR and 95% confidence interval (CI) in direct comparison of iMRI and neuronavigation were 3.21 and (2.34, 4.38), respectively, while those in the pooled comparison were 3.00 and (2.10, 4.50), respectively. Moreover, the 95% CIs mostly overlapped. These findings support the fact that iMRI was better than neuronavigation in terms of the GTR and does not affect the ranking derived from the network meta-analysis. In the pooled comparison, we just shortened the gap between iMRI and neuronavigation, however, the inconsistency did not affect the rank of neuronavigation. The position that neuronavigation was ranked second to last followed by conventional surgery could be verified both in direct and pooled comparisons. In the direct comparison of iMRI and conventional surgery, the OR and 95% CI were 3.11 and (1.53, 6.30), respectively, while those in the pooled comparison were 5.70 and (3.40, 9.60), respectively. In the direct comparison, it was clear that iMRI was better than conventional surgery, but the pooled comparison exaggerated the gap between them. The Cochrane Handbook suggests that direct comparisons can be regarded as accurate when both direct and indirect evidence exist.[59] It has been suggested that the outcomes of direct and indirect comparisons are not always the same.[60] If the direct comparison is not convincing, an indirect comparison can provide a useful supplement. As it is not possible to perform direct comparisons among all of these intraoperative guidance technologies in the clinic, this network meta-analysis offers the advantage of combining direct and indirect comparisons to rank the different methods.

Current-regulated or power-regulated power supplies with output ripples of less than 5% are recommended. Single-shot ignition pulses are advised because radio frequency starters may damage the lamp’s internal reflector.

Cermax® Xenon short-arc lamps from Excelitas Technologies are ideal for applications that require a high degree of illumination control.

From medical lighting to analytical instrumentation, clinical diagnostics, industrial, safety and security, and aerospace and defense applications, Excelitas Technologies is committed to enabling our customers' success in their specialty end-markets. Excelitas Technologies has approximately 3,000 employees in North America, Europe and Asia, serving customers across the world.

Eleven studies reporting clinical improvement and return to stable condition with four technologies were included [Figure 5]. From the direct comparisons, iMRI was better than neuronavigation and conventional surgery [Table 1]. The absence of obvious heterogeneity and inconsistency suggests that these data can be used in the network meta-analysis. The comparisons based on the consistency model. The ranking was determined as follows: iMRI > iUS > neuronavigation > conventional surgery [Figure 8].

This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.

The combination of image guidance modalities in glioma resection is a popular trend. These multimodal approaches theoretically facilitate complete tumor resection as well as the preservation of neurological function.[78] We hope that high-quality studies using multivariate analysis will be conducted in future to evaluate the combination of multiple technologies in glioma surgery.

Address for correspondence: Prof. Haozhe Piao, Department of Neurosurgery, Cancer Hospital of China Medical University, Liaoning Cancer Hospital and Institute, Shenyang 110042, Liaoning Province, China. E-mail: [email protected]

A total of 2842 related studies were obtained in the initial search, and 38 studies from 1980 to 2018 were included for the network meta-analysis[15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52] [Figure 1].

Then, we generated a random-effects model within a Bayesian framework using Markov chain Monte Carlo methods in R version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria).[7,12] The model incorporated the outcomes of the different interventions used in each study to assess the relations between these technologies in terms of odds ratios (ORs).[6] This model combined the direct and indirect evidence for all compared interventions. Thus, the technologies were ranked based on the OR for each outcome measure compared with over several iterations of the Markov chain. We could get the ranked probability of four technologies and conventional surgery in different outcome measures by calculating the OR for each surgical method compared with an arbitrary common control group, counting the proportion of iterations of Markov chain in which each method had the highest OR, the second highest, and so on.

Excelitas Technologies is a global technology leader focused on delivering innovative, customized solutions to meet the lighting, detection and other high-performance technology needs of OEM customers.

The impact of intraoperative technologies on the rate of postoperative complications is a valid end point in addition to GTR. Previously published study data did not provide a direct comparison among the four guidance technologies and conventional surgery for glioma resection. Because of the limited number of studies reporting postoperative complications, we included only iMRI, iUS, neuronavigation, and conventional surgery. The ranking results showed that conventional surgery had the worst performance in terms of the incidence of complications, whereas iUS was the best followed by iMRI and neuronavigation. Our data implied that iUS and iMRI may decrease the rate of complications following glioma surgery. However, there was a slight inconsistency in the network meta-analysis: in the direct comparison of iMRI and conventional surgery, the OR and 95% CI were 0.36 and (0.19, 0.65), respectively, whereas those in the pooled comparison were 0.24 and (0.14, 0.42), respectively. Based on the direct comparison, it was clear that iMRI was superior to conventional surgery, and the pooled comparison exaggerated the difference between them. Thus, the ranking results were confirmed. Based on this finding, iUS appears to be the safest tool used in the resection of gliomas as it reduces the risk of postoperative complications. iUS has the advantage of real-time scanning and offers the ability to differentiate blood in the resection cavity or marginal resection from residual tumor. In addition, iUS can help surgeons observe lesions in multiple directions and can be repeated many times during the surgery to enable accurate tumor resection, which is considered to lead to reduced postoperative complications. There were three included studies comparing iUS to other guidance technologies. Thus, the network analysis was based on direct and indirect comparisons. However, more studies are needed to verify the conclusion based on a large sample size.

Neuronavigation has been used in routine glioma surgery.[77] It helps to visualize the tumor borders and results in shorter surgical times and smaller craniotomies.[77] It is usually based on preoperative CT or MRI images, so the main limitation of neuronavigation is that the preoperative images are not as accurate once brain tissue is removed during the surgery. Some studies have shown the positive impact of neuronavigation in glioma surgery, whereas most reports provided questionable evidence.[77] Our meta-analysis supported the latter view and found that neuronavigation performed worse than other technologies.

Gliomas are the most common brain tumors in the world.[1] Surgery is one of the most effective therapeutic methods in terms of patient prognosis. It has been reported that the gross total resection (GTR) and extent of resection are closely correlated with the 5- and 10-survival rates.[2] However, according to the characteristic infiltrative nature of malignant gliomas, it can be challenging to fully resect these tumors.[1] In the past three decades, several visual guidance technologies have been used to achieve maximum resection while maintaining patient safety. These technologies include intraoperative magnetic resonance imaging (iMRI), intraoperative ultrasound (iUS), fluorescence-guided surgery with 5-aminolevulinic acid (5-ALA), and intraoperative neuronavigation.[3,4,5] There is no doubt that a prospective double-blinded randomized controlled trial would be the next ideal step to compare these technologies with conventional surgery. However, the execution of such trials is impractical because of logistical, financial, and ethical reasons. To date, there has been no uniform standard in terms of the overall benefits for patient survival and postoperative complications in the central nervous system and other systems to compare these technologies. In clinical practice, it remains difficult for neurosurgeons and health-care providers to select the appropriate imaging technology to gain the most support. Therefore, in this study, we compared different guidance techniques with conventional surgery for glioma resection in terms of GTR, postoperative complication rate, and postoperative clinical improvement, which are often considered the most important factors in the surgical resection of gliomas. We used multiple treatment meta-analyses[6] combining the evidence of direct and indirect comparisons.[7] The objective was to provide an informative conclusion that can help improve clinical practice.

The Cermax® Xenon lamp, Model PE300BFA, has an integrated parabolic reflector, enabling high-intensity, focused output of visible and infrared radiation. With their internal reflector and rugged ceramic body construction, Cermax® Xenon lamps are the safest and most compact alternative to conventional quartz xenon lamps. This makes them ideal for applications requiring a high degree of illumination control.

iMRI, iUS, neuronavigation, and conventional surgery were compared in terms of patients' clinical improvement. Our data confirm that iMRI was superior with respect to this metric, followed by iUS and neuronavigation; all were better than conventional surgery. There was consistency between the direct and indirect evidence in the node-split analysis. These data suggested that iMRI was the best technology for the clinical improvement of patients. The risk factors for negative results were prior surgery or radiotherapy.[61]

The most important outcome measures were the rates of GTR, postoperative complications, and clinical improvement. The GTR was defined as the percentage of cases resulting in ≥90% tumor removal as measured via postoperative magnetic resonance imaging (MRI) within a week after surgery. The postoperative complication rate was defined as the percentage of patients that experienced complications of the central nervous system, such as neurological deficits, epilepsy, cognitive deficits, and aphasia, and those outside the central nervous system, such as infection, thrombosis, and cerebrospinal fluid leakage. The clinical improvement rate was defined as the percentage of patients that were in stable condition during the follow-up examination. A dichotomous outcome was utilized in this network meta-analysis: we used the number of patients whose brain tumors were totally resected, who experienced postoperative complications, and who experienced significant clinical improvement or were stable at least 3 months after surgery.

The Cermax® Xenon short-arc lamp from Excelitas Technologies is an innovative lamp design in the specialty lighting industry. Cermax® Xenon lamps were first introduced in the early 1980s and are now used in diagnostic and surgical endoscopes in most major hospitals worldwide, in high-brightness projection display systems, and for a wide variety of highperformance applications.