Influence of scalp block on oncological outcomes of high-grade glioma in adult patients with and without isocitrate dehydrogenase-1 mutation

High-grade gliomas are notorious for a high recurrence rate even after curative resection surgery. Studies regarding the influence of scalp block on high-grade gliomas have been inconclusive, possibly because the condition’s most important genetic mutation profile, namely the isocitrate dehydrogenase 1 (IDH1) mutation, had not been analyzed. Therefore, we conducted a single-center study including patients with high-grade glioma who underwent tumor resection between January 2014 and December 2019. Kaplan–Meier survival analysis revealed that scalp block was associated with longer progression-free survival (PFS; 15.17 vs. 10.77 months, p = 0.0018), as was the IDH1 mutation (37.37 vs. 10.90 months, p = 0.0149). Multivariate Cox regression analysis revealed that scalp block (hazard ratio: 0.436, 95% confidence interval: 0.236–0.807, p = 0.0082), gross total resection (hazard ratio: 0.405, 95% confidence interval: 0.227–0.721, p = 0.0021), and IDH1 mutation (hazard ratio: 0.304, 95% confidence interval: 0.118–0.784, p = 0.0138) were associated with better PFS. Our results demonstrate that application of scalp block, regardless of IDH1 profile, is an independent factor associated with longer PFS for patients with high-grade glioma.

High-grade gliomas are notorious for a high recurrence rate even after gross curative surgical resection, possibly because intense stress and inflammatory response may occur during surgery, inducing immunosuppression that promotes locoregional cancer recurrence 1 . In addition, anesthesia-related factors such as the use of inhalation anesthesia and administration of opioids also induce detrimental immunosuppression 2 . By contrast, regional anesthesia effectively relives surgical stress, reduces inhalational anesthetic and opioid consumption, and has been reported to be beneficial to patients undergoing cancer surgery 1,3 . Therefore, there is increasing interest in investigations of the potential oncological benefits of regional anesthesia for cancer surgery 4 . To date, previous reports have indicated conflicting results regarding the influence of regional anesthesia, namely the scalp block, on postoperative glioma recurrence [5][6][7] . For instance, we previous reported that scalp block is associated with a prolonged postoperative glioma recurrence 5 . By contrast, another larger cohort revealed negative results 6 . This may be because a tumor's genetic profile plays a decisive role in the oncological outcomes of patients with glioma 8 , but this has not been analyzed and compared in previous reports of scalp block for glioma resection.
The revised 2016 World Health Organization classification of central nervous system tumors reclassified gliomas on the basis of molecular marker diagnostics combined with classical histological diagnosis 8 . Possibly the foremost molecular markers are mutations in isocitrate dehydrogenase (IDH) 9 . Mutations in IDH isozyme 1 (IDH1) not only play a crucial role in early tumorigenesis of astrocytomas and oligodendrogliomas but are also the decisive genetic signposts of secondary glioblastoma 10 . Most IDH1 mutations reported are located at the first or second base of codon 132 and the most frequent is R132H (IDH1 R132H ) 10 . Patients with glioma and IDH1 R132H glioblastoma have twice the median overall survival (OS) and more favorable progression-free survival (PFS) than those with the IDH1 wild type 11,12 . However, occurrence of IDH1 R132H has not yet been analyzed in the literature www.nature.com/scientificreports/ regarding scalp block and oncological outcomes. Therefore, in this study, we investigated the influence of scalp block on oncological outcomes in patients undergoing resection surgery for high-grade glioma; a complete IDH1 R132H genetic profile analysis was performed. To our best knowledge, this presents the first literature of scalp block and the oncological outcomes of high grade glioma with the analysis of IDH1 R132H mutation.

Methods
After obtaining the local ethic committee approval, a prospectively maintained database including patients of high-grade glioma undergoing craniotomy for glioma curative resection with complete IDH R132 mutation information was used for analysis. Exclusion criteria were (1) age younger than 20 years, because molecular characteristics and prognosis may differ between pediatric and adult patients with high grade glioma 13 , (2) undergoing an awake craniotomy, and (3) receiving chemotherapy or radiotherapy before surgery. Data collected included demographic characteristics, comorbidities, tumor location, tumor size, perioperative red blood transfusion, amount of opioid consumed, presence of the IDH1 R132H mutation, postoperative adjuvant therapies, and anesthetic techniques such as the use of intravenous or inhalation anesthesia and scalp block.
Surgery and anesthesia. Each patient received general anesthesia that was maintained either intravenously or by inhalation at the attending anesthesiologist's discretion. Blood transfusion was performed when the patient's hemoglobin level was less than 9 g/dL. In patients receiving inhalational anesthetics, sevoflurane was used and was kept at 0.5 times the minimum alveolar concentration to minimize interference with intraoperative electrophysiological monitoring 14 . Perioperative steroids were not routinely given because brain lymphoma sometimes resembles glioblastoma on preoperative magnetic resonance images 15 , and steroids administered perioperatively would interfere with pathological examination of the tumor. Application of scalp block was at the anesthesiologist's discretion and was performed immediately after induction of general anesthesia and before head pinning with a Mayfield head holder; the scalp block regimen was 10 mL of 0.5% levobupivacaine and 1:200,000 epinephrine mixture for each side of the scalp. The effect of the scalp block was achieved by means of local infiltration around targeted nerves, which were the supraorbital, supratrochlear, zygomaticotemporal, auriculotemporal, greater occipital, lesser occipital, and least occipital nerves. Following surgery, all patients were sent to the same neurosurgical intensive care unit for postoperative care. Postoperative analgesia was achieved with mild to moderate opioids such as tramadol or nalbuphine; nonsteroidal anti-inflammatory drugs were used once the risk of bleeding was considered minimal by the surgeons. Statistical analysis. PFS was defined as the interval between the date of surgery and the date of first evidence of tumor recurrence, which was based on postoperative magnetic resonance imaging once every 3 months, as is recommended by guidelines followed by many hospitals worldwide 16 . OS was defined as the interval between the date of surgery and date of death or loss of follow-up. To identify the targeted PFS improvement for at least one follow-up interval (3 months), 90 patients would be needed for statistical analysis with a power of 0.8 and α = 0.05.
Fisher's exact test or the chi-square test was employed to analyze dichotomous data, Student's t-test was used for normally distributed continuous data, and the Mann-Whitney U test was used for nonparametric ordinal data. PFS and OS are presented as the median (95% confidence interval [CI]) and were calculated using Kaplan-Meier survival analysis with the log-rank test to compare survival curves between groups. Univariate and multivariate Cox proportional hazards regression analyses were used to determine the effect of several risk factors on PFS and OS. A p < 0.05 was considered significant.
In our institute, the complete information regarding the IDH R132 mutation was routinely reported in postoperative pathological records since 2014. Because the incidence of high grade glioma was approximately only one per 100,000 17 , it is difficult for patient enrollment. Therefore, this study included high grade glioma patients enrolled in our previous study, which include our institutional glioma patients during 2010 to 2017 18 , to achieve a sufficient study power. Accordingly, a parallel multivariate Cox regression analysis without overlapping was also performed to investigate the influence of scalp blocks on the oncological outcomes of patients with high grade glioma. Statistical analysis was conducted using IBM SPSS Statistics for Windows, version 21.0 (IBM Corp., Armonk, NY, USA) and MedCalc Statistical Software version 18.2.1 (MedCalc Software bvba, Ostend, Belgium).
Ethics statement. This study was approved by the Research Ethics Committee of National Taiwan University Hospital (approval No.201911066RIND, approval date: December 27, 2019), and the informed consent requirement was waived by the committee. The patients' identifying information was omitted during analysis. This study adheres to the applicable EQUATOR guidelines.

Discussion
The major discovery of this study was that scalp block is an independent prognostic factor for recurrence profile in patients with high-grade glioma, regardless of differences in IDH1 R132H mutation profile.  www.nature.com/scientificreports/ To the best of our knowledge, the major difference between previous reports and the present study is that presence of the IDH1 R132H mutation was for the first time included in an analysis of the influence of anesthetic technique on high-grade glioma oncological outcomes. A multigroup collaboration in 2008 sequenced 20,661 genes in 22 glioblastoma multiforme tumor samples and identified a common point mutation in the metabolic gene IDH1 (R132H) that was present in 12% of patients with glioblastoma multiforme 19 . The proportion of IDH1 R132H in the present study (15.2%) is comparable to that reported previously. The study results are also compatible with those of previous studies indicating that patients with glioma carrying the IDH1 R132H mutation were younger and had longer PFS 12,20 . IDH1 is a key enzyme in Krebs cycle functions dependent on NADP + . Wild-type IDH1 messenger RNA and protein are commonly overexpressed in primary glioblastomas, which indicates that IDH1 activity is important to metabolic adaptation of high-grade gliomas 21 . In addition, IDH1 regulates hypoxiainducible factors related to tumor angiogenesis and invasiveness 22 . This is relevant to glioma recurrence because glioma cells exhibiting invasive characteristics after resection surgery are more highly competitive for nutrients than are normal neuronal tissue. The beneficial outcomes observed among IDH1 R132H mutant gliomas may occur through several mechanisms. First, IDH1 R132H mutations result in neomorphic enzyme activity, catalyzing the NADPH-dependent reduction of α-ketoglutarate to R(-)-2-hydroxyglutarate 23 , which increases DNA methylation of the glioma. Data from a large cohort of 1,122 diffuse grade II-IV gliomas revealed that IDH-mutant gliomas with high levels of DNA methylation had more favorable clinical outcomes than those with low levels 24 . Second, IDH1 R132H mutations inhibit glial cell proliferation through inhibition of Bcl-xL, which induces more apoptosis than wild-type IDH 25 . Third, IDH1 R132H mitigates expression of hypoxia-induced factor-1α, which interferes with glioma angiogenesis 22 . Although the present study indicated improved PFS among patients carrying IDH1 R132H , improved OS was not observed. This may be because the survival of patients with high-grade glioma has markedly improved in the last decade owing to a rapid increase in the use of adjuvant and concomitant chemotherapy and radiotherapy, earlier diagnosis, and advances in multimodal health care 26 . In addition, survival is influenced by more complex factors such as age, baseline performance status 27 , long-term care capability 28 , and gross total resection rate 29 . Therefore, the influence of IDH1 R132H on OS may have been obscured in our study.
The potential mechanisms of the protective effects of scalp blocks against high grade glioma recurrence are organized into three aspects, namely the prevention of a surgical stress response, the systemic effects of local anesthetic, and the sparing effects of general anesthetic. Accumulating evidence reveals that surgical stress and the subsequent neuroendocrine and inflammatory responses may negatively effects tumor recurrence after curative surgery 1 . First, the surgical stress response activates the sympathetic nervous system and the hypothalamic-pituitary-adrenal axis, thus promoting tumor-associated angiogenesis 30,31 . Second, surgical stress increases catecholamine circulation, which may impair immune function through mechanisms including diminished cytotoxicity of natural killer (NK) cells 32 , reduced dendritic cell maturity, and a decreased Th1/Th2 ratio, which suppresses antimetastatic cell-mediated immunity 33 . Furthermore, glioma cells express beta-adrenergic receptors, and catecholamine may thus interact with the glioblastoma proliferation 34 . Because scalp blocks effectively reduce surgical neuroendocrine stress, with concomitant reductions in plasma cortisol and catecholamine levels during craniotomy 35 , the aforementioned negative effects may be attenuated. In addition to the attenuation effects against stress responses, local anesthetics may prompt systemic protective effects against glioma proliferation through various mechanisms. First, local anesthetic blocks N-methyl-d-aspartate-type glutamate receptors from mediating rapid excitatory neurotransmission in the central nervous system 36,37 . Extracellular glutamate in the cerebral cortex negates IDH1 R132H suppression of gliomagenesis, thereby driving anchorage-independent growth and glioma progression 38 . Accordingly, negative effects of glutamate may be inhibited by the local anesthetics in scalp blocks. Second, transient receptor potential (TRP) channels regulate cell proliferation and death 39 . Local anesthetic (e.g., lidocaine) upregulates the TRPV1 channels and suppresses the TRPM7 channels, and both of these mechanisms protect against glioma cell proliferation 40,41 . Third, local anesthetic was reported to induce protective autophagy in a rat C6 glioma cell line 42 . Moreover, it weakened ZDHHC15 transcripts and reduced GP130 palmitoylation levels and membrane localization, thus impairing the growth and self-renewal of glioblastoma stem cells 43 .
General anesthetic agents may elicit multiple negative effects during the surgical resection of glioma. First, inhalational anesthetics were reported to profoundly suppress the effects of NK cells. A clinical study reported that sevoflurane induced a significant decrease in NK cells during cranial surgery 44 . Sevoflurane was also reported to attenuate NK cell-mediated cytotoxicity by inhibiting adhesion molecule leukocyte function-associated antigen-1 in an experimental study 45 . The importance of NK cells for glioma outcomes has been recently addressed; the IDH1 R132 mutation promotes recruitment of NK cells in the brain and is associated with improved glioma prognosis 46 . In addition, novel therapies based on augmentation of NK cell recruitment and function are emerging in neuro-oncology 47 . Therefore, the negative effects of inhalational anesthetics on NK cells may lead to an immunosuppressant tumor microenvironment in patients undergoing glioma resection surgery. Second, the inhalational anesthetics may induce the negative gene expression effects of human glioma cells. Isoflurane and sevoflurane were reported to increase the migration, invasion, and colony-forming ability of human glioblastoma cells in vitro, and their tumor volume and invasion/migration were increased in vivo through increases in the expression of cell surface protein 44 48 . Third, both inhalational anesthesia and propofol were reported to increase the amount of circulating extracellular glutamate in patients undergoing neurosurgery 49 , which may promote glioma proliferation. Therefore, scalp blocks may attenuate the aforementioned negative effects through strong anesthetic-sparing effects 50 .
Intravenous anesthesia has also been proposed to result in less intraoperative immunosuppression than inhalational anesthesia during surgery 1 , and a recent meta-analysis revealed beneficial oncological outcomes of intravenous anesthesia in patients with nonglioma cancer. However, the current literature on glioma is limited and reveals nonsignificant effects 18 www.nature.com/scientificreports/ because of the unique characteristics of the glioma immune microenvironment. First, glioma cells may express cell surface proteins to sequester T cells in the brain 53 . Second, the blood-brain barrier acts as a physical barrier against the exchange of immune cells from extraneural tissues and the brain 54 . Third, natural killer cells are the least abundant immune cell infiltrating glioma cells (representing only 2.11% of all infiltrating immune cells 55 ; and glioma cells secrete soluble immunosuppressive factors to suppress natural killer cell activity 56 . Because the protective immune effects of intravenous anesthesia are mainly exerted in the peripheral circulation, particularly through the activation of natural killer cells 57,58 , intravenous anesthesia may be less influential on the oncological outcomes of patients with glioma. There are several limitations in this study. First, the study is single-centered, retrospective in nature, and with a relatively small sample. Second, despite the IDH1 R132H mutation being the most influential genetic factor for postoperative oncological outcomes of gliomas, information concerning other potential factors such as the chromosomal 1p/19q codeletion, IDH2 or MGMT mutation, and ATRX mutation was not considered. Third, the number of patients with infratentorial tumors was much lower than those with supratentorial tumors. Hence, caution is needed when extrapolating the results to patients with infratentorial tumors. Fourth, steroids are not administered for patients with primary glioma in our institute; thus, the present results should be interpreted cautiously for patients receiving perioperative steroids.
In conclusion, our results indicate that scalp block is associated with favorable recurrence profiles in patients with high-grade glioma. However, given the limitations of this study and the complexity of the pathophysiology of glioma recurrence, our results should be considered exploratory rather than as a basis for practice change.

Data availability
Datasets are available on request.