Intraoperative rapid molecular diagnosis aids glioma subtyping and guides precise surgical resection

Abstract

Objective

The molecular era of glioma diagnosis and treatment has arrived, and a single rapid histopathology is no longer sufficient for surgery. This study sought to present an automatic integrated gene detection system (AIGS), which enables rapid intraoperative detection of IDH/TERTp mutations.

Methods

A total of 78 patients with gliomas were included in this study. IDH/TERTp mutations were detected intraoperatively using AIGS in 41 of these patients, and they were guided to surgical resection (AIGS detection group). The remaining 37 underwent histopathology‐guided conventional surgical resection (non‐AIGS detection group). The clinical utility of this technique was evaluated by comparing the accuracy of glioma subtype diagnosis before and after TERTp mutation results were obtained by pathologists and the extent of resection (EOR) and patient prognosis for molecular pathology‐guided glioma surgery.

Results

With NGS/Sanger sequencing and chromosome detection as the gold standard, the accuracy of AIGS results was 100%. And the timing was well matched to the intraoperative rapid pathology report. After obtaining the TERTp mutation detection results, the accuracy of the glioma subtype diagnosis made by the pathologists increased by 19.51%. Molecular pathology‐guided surgical resection of gliomas significantly increased EOR (99.06% vs. 93.73%, p < 0.0001) and also improved median OS (26.77 vs. 13.47 months, p = 0.0289) and median PFS (15.90 vs. 10.57 months, p = 0.0181) in patients with glioblastoma.

Interpretation

Using AIGS intraoperatively to detect IDH/TERTp mutations to accurately diagnose glioma subtypes can help achieve maximum safe resection of gliomas, which in turn improves the survival prognosis of patients.

Introduction

Gliomas are the most common primary brain tumor, accounting for approximately 80% of all central nervous system (CNS) malignancies. ¹ , ² There are significant differences in survival after surgery for different glioma subtypes. ³ , ⁴ Despite the rapid development of neurosurgery in recent decades, the treatment of gliomas has progressed slowly and remains dominated by imaging‐guided maximum safe resection with limited benefit to patients. ⁵ In the fifth edition of the World Health Organization Classification of the Central Nervous System (WHO CNS5), molecular diagnosis was included in the staging criteria for gliomas for the first time, seeming to bring a new guiding strategy for surgical resection of gliomas. ⁶ , ⁷ The rapid acquisition of information on the patient's genetic mutation will be crucial to the intraoperative diagnosis of glioma subtypes and guide precise surgical resection. ⁸ As one of the most common mutation types in gliomas, TERTp mutations play an important role in glioma subtyping, surgical resection, and predicting prognosis. ⁹ , ¹⁰ And they occur most frequently at the c.1‐124 C>T and c.1‐146 C>T sites, also known as C228T and C250T. ¹¹ Related studies have confirmed that the detection of TERTp mutations in patients with IDH mutant gliomas predicted a better prognosis. Conversely, in patients with IDH wild‐type glioblastoma, it was associated with a poor prognosis for GBM. ¹² , ¹³ , ¹⁴

The prognosis of glioma patients is affected by several variables, including age, KPS, tumor heterogeneity, and surgical options. ¹ , ¹⁵ The key elements impacting patients' postoperative survival are the various glioma subtypes and the extent of resection (EOR). ¹⁶ , ¹⁷ Intraoperative MRI (iMRI), intraoperative ultrasound (iUS), neuronavigation, and intraoperative neurophysiological monitoring (IONM) are currently used to guide surgical resection of gliomas, resulting in further improvement in the rate of total tumor resection. ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² But there are yet no efficient ways to perform rapid intraoperative glioma subtyping. Our preliminary study demonstrated that intraoperative IDH mutation detection combined with intraoperative rapid pathology can help to accurately diagnose gliomas and can also be used to define the molecular margins of IDH‐mt gliomas intraoperatively. ²³ However, it is difficult to distinguish astrocytoma from oligodendroglioma by detecting only IDH mutations, and it has limited utility in guiding surgical resection of glioblastomas. The TERT mutation was found to be almost concomitant with the 1p/19q deletion in IDH mutant gliomas, with a 94% mutation rate in oligodendrogliomas and a 70%–80% mutation rate in IDH wild‐type GBM. ¹³ , ²⁴ Therefore, we propose the use of TERTp mutations as molecular biomarkers to assist in the precise intraoperative diagnosis of oligodendroglioma and to rapidly define the molecular margins of TERTp mutant GBM. In the present study, we further improved the AIGS to achieve rapid intraoperative detection of TERTp mutations and evaluated its practical use as an intraoperative molecular diagnostic tool to assist in the precision diagnosis and treatment of gliomas.

Methods and Materials

Sample information

This study was a prospective cohort study, and a total of 78 patients were included, all of whom were seen in the Department of Neurosurgery at Qilu Hospital of Shangdong University between August 2021 and November 2022. The inclusion criteria for this study were newly diagnosed glioma patients aged 18–69 years with preoperative KPS ≥80. Exclusion criteria were recurrent gliomas, incomplete MRI, other tumor‐related diseases, or patients who were lost to follow‐up. All diagnoses met the WHO CNS5. The study was approved by the ethics committee of Qilu Hospital of Shangdong University and registered with the National Clinical Research Centre, and all patients signed a written informed consent form.

Rapid intraoperative molecular detection (AIGS real‐time fluorescent PCR)

The AIGS uses PCR technology combined with TaqMan‐MGB probes for real‐time quantitative PCR (qRT‐PCR) for the detection of IDH/TERTp mutations (Fig. 1A). The technology combines nucleic acid extraction with real‐time gene amplification assays in the form of microfluidic cassettes to achieve fully automated processing from spike‐in to analysis of gene assay results. We use two fluorescent dual‐target designs, FAM and ROX, and Cy5 as a non‐competitive internal reference for full quality control of the assay. If the Cy5 fluorescence curve does not show an S‐shaped amplification, the kit is considered invalid. The specific primers and probes used in this study were designed by ourselves (Table S1). AIGS was used for the rapid intraoperative detection of IDH/TERTp mutations and to differentiate TERTp C228T and TERTp C250T mutations by specific primers and probe sequences, as well as IDH1 R132H, IDH1 R132L, and other mutant subtypes (Fig. 1C,D).

Figure 1.

The flow chart of AIGS intraoperative assistance for the precise diagnosis and treatment of gliomas. (A) Intraoperative rapid detection of IDH/TERTp mutations using AIGS enables intraoperative integrated diagnosis of gliomas. (B) Personalized surgical regimens based on integrated intraoperative diagnosis. (C) Results of various mutation types for AIGS detection of IDH mutations. IDH mutation negative, only a Cy5 fluorescence curve shows typical S‐shaped amplification (including S‐curve that do not reach a plateau); IDH1 R132H mutation, ROX and Cy5 curves both show typical S‐shaped amplification (including S‐curves that do not reach a plateau); IDH1 R132L mutation, FAM, ROX, and Cy5 curves all show S‐shaped amplification (including S‐curves that do not reach a plateau); IDH other type, FAM and Cy5 curves both show typical S‐shaped amplification (including S‐curves that do not reach a plateau). (D) Results of AIGS detection of various types of TERTp mutations. TERTp mutation negative, only a Cy5 fluorescence curve shows typical S‐shaped amplification (including S‐curve that do not reach a plateau); TERTp C228T mutation, FAM and Cy5 curves both show typical S‐shaped amplification (including S‐curves that do not reach a plateau); TERTp C250T mutation, ROX and Cy5 curves both show S‐shaped amplification (including S‐curves that do not reach a plateau).

Evaluating the effect of TERTp mutations on the accurate intraoperative diagnosis of glioma subtypes

Rapid histopathological and AIGS testing was performed intraoperatively in 41 patients, and the accuracy of the results was verified by NGS or Sanger sequencing after surgery. Immediately after removal of the tumor, the surgeon divided the specimen into three parts, one of which was sent to the pathology department for rapid histopathology and the other two for IDH and TERTp mutation testing using AIGS. The IDH mutation test results are first provided, and the pathologist makes a preliminary diagnosis in conjunction with rapid intraoperative histopathology; the TERTp mutation test results are then supplemented to the pathologist to assist the pathologist in making an intraoperative integrated diagnosis (Fig. 1A). The results of both groups were compared with the postoperative integrated diagnosis to assess the practical value of this technique in aiding intraoperative glioma staging diagnosis. Meanwhile, the neurosurgeon was informed of the IDH/TERTp mutation status and the intraoperative integrated diagnosis.

Surgical resection strategies for patients with different glioma subtypes

All patients underwent preoperative MRI, including T1‐weighted (T1w), contrast‐enhanced T1‐weighted (T1w‐CE), T2‐weighted (T2w), T2‐weighted fluid‐attenuated inversion recovery (T2w/Flair), and postoperative repeat MRI at 72 h. Based on the intraoperative integrated diagnosis, we developed personalized surgical regimens for patients, including STR, GTR, and SupTR (Fig. 1B). We did a GTR in the region of T1w, T2w, and T2w/Flair abnormal areas for patients who had an intraoperative integrated diagnosis of astrocytoma (IDH‐mt/TERTp‐wt). In terms of patients with oligodendroglioma (IDH‐mt/TERTp‐mt), patients with tumors located in non‐functional areas underwent GTR, and patients with tumors located in functional areas underwent STR, with the goal of improving the patients' quality of life in the postoperative period while preserving function to the greatest extent possible. The EOR was higher than 90% of the abnormal areas. As for patients with an intraoperative integrated diagnosis of glioblastoma (IDH‐wt, TERTp‐wt, or TERTp‐mt), SupTR was performed to resect the area of T1w‐CE and over 50% of the T2w/Flair area (Fig. 2A).

Figure 2.

The value of AIGS in the clinical application of intraoperative assistance for accurate subtype diagnosis of glioma. (A) The extent of STR, GTR, SupTR, and precision resection of glioblastoma. (B) Mean intraoperative rapid histopathology reporting time and AIGS detection time. (C) The NGS was used as the gold standard to test the accuracy of the AIGS results. (D) Postoperative integrated diagnosis of the patients (WHO CNS5, n = 41). (E) Proportion of astrocytomas, oligodendrogliomas, and glioblastomas in the preliminary diagnosis and intraoperative integrated diagnosis, and the accuracy of glioma subtype diagnosis in both diagnostic methods. (F) The EOR in AISG‐detected and non‐AIGS‐detected groups.

Based on the patient's preoperative and postoperative MRIs, two experienced neurologists were blinded to the Response Assessment in Neuro‐Oncology (RANO) criteria to calculate the tumor volume and residual tumor volume, glioblastoma volume by T1w‐CE, and astrocytoma and oligodendroglioma volume by T2w/Flair. ²⁵ , ²⁶ Preoperative and postoperative tumor volume quantification was performed using the 3D Slicer image computing platform version 5.2.2. Glioma resection rate is defined as imaging surgical resection volume/preoperative tumor volume. The equation was calculated as EOR = (Preoperative tumor volume − Residual tumor volume)/Preoperative tumor volume × 100%.

Statistical analysis

In this study, NGS or Sanger sequencing results were used as the gold standard. A four‐grid table was used to assess the accuracy, sensitivity, and specificity of AIGS for the detection of TERTp mutations. The statistical significance of AIGS detection of TERTp mutations for the diagnosis of glioma subtypes was assessed using the chi‐square test. Kaplan–Meier survival curves were used to analyze and express the overall survival and progression‐free survival of 78 glioma patients, respectively. All statistical analyses were performed using GraphPad Prism 9 (The GraphPad Software Company), and statistical significance was set at p < 0.05. All figures shown in this article were drawn using GraphPad Prism 9 and Adobe Illustrator (The Adobe Company).

Results

Patient clinical features

The mean age of the 41 patients in the AIGS detection group was 50.83 ± 12.42 years, of which 56.10% were male. Tumors were primarily located in the frontal lobe (53.66%) and secondarily in the temporal lobe (34.14%). The mean age of the 37 patients in the non‐AIGS detection group was 58 years, and 59.46% were female. Similarly, tumors were predominantly located in the frontal and temporal lobes, accounting for 59.46% and 32.43%, respectively. The mean preoperative KPS (86.34 ± 4.88 vs. 87.03 ± 6.18), mean tumor volume (24.93 ± 25.02 vs. 26.79 ± 23.69, cm³), mean operative time (287.56 ± 65.81 vs. 302.43 ± 50.89, min), postoperative complications (26.83% vs. 37.84%), and median follow‐up time (21.37 vs. 18.37, months) were not statistically different between the two groups (p > 0.05). There was no statistically significant difference in the occupancy of patients with IDH mutations between the two groups (p > 0.05). At the date of the last follow‐up, 21 patients had a recurrence in the AIGS‐detected group, of whom 10 died (1 astrocytoma WHO grade 4 and 9 glioblastomas WHO grade 4), and 19 patients in the non‐AIGS‐detected group had a recurrence, of whom 13 died (all of them glioblastomas WHO grade 4), all of whom had a recurrence of glioma without secondary surgery and died of brainstem failure (Table 1, Tables S2–S4).

Table 1.

Clinical characteristics of patients in the two groups.

Variable	AIGS detection group (ADG, n = 41)	Non‐AIGS detection group (NADG, n = 37)	p
Age (mean ± SD, years)	50.83 ± 12.42	51.38 ± 12.19	0.845
Sex			0.183
Male	23 (56.10%)	15 (40.54%)
Female	18 (43.90%)	22 (59.46%)
Location			0.864
F	22 (53.66%)	22 (59.46%)
T	14 (34.14%)	12 (32.43%)
I	2 (4.88%)	1 (2.70%)
P	2 (4.88%)	2 (5.41%)
O	1 (2.44%)	0 (0.00%)
Tumor volume (mean ± SD, cm3)	24.93 ± 25.02	26.79 ± 23.69	0.738
Pre‐op KPS (mean ± SD)	86.34 ± 4.88	87.03 ± 6.18	0.586
Surgery time (mean ± SD, min)	287.56 ± 65.81	302.43 ± 50.89	0.272
Post‐op complications	11 (26.83%)	14 (37.84%)	0.216
Intracranial infection a	8 (19.51%)	9 (24.32%)
Hemorrhage	0 (0.00%)	2 (5.41%)
Aphasia	1 (2.44%)	1 (2.70%)
Hemiplegia	0 (0.00%)	2 (5.41%)
Poor wound healing	2 (4.88%)	0 (0.00%)
Median follow‐up time (months)	21.37	18.30	0.688
Molecular mutation information			0.820
IDH‐mt (grade 2–4)	17 (41.46%)	17 (48.65%)	0.594
Grade 2	11 (64.71%)	12 (70.59%)
Grade 3	5 (29.41%)	5 (29.41%)
Grade 4	1 (5.88%)	0 (0.00%)
IDH‐wt (grade 4)	24 (58.53%)	20 (51.35%)
Recurrent
Yes	21 (51.22%)	19 (51.35%)
No	20 (48.78%)	18 (48.65%)
Survival
Death	10 (24.39%)	13 (35.14%)
Alive	31 (75.61%)	24 (64.86%)

F, frontal lobe; I, insula; O, occipital lobe; P, parietal lobe; T, temporal lobe.

^a Intracranial infection: temperature higher than 38.5°C, blood leukocyte count greater than 10 × 10⁹/L, and cerebrospinal fluid leukocyte count greater than 100 × 10⁶/L.

The role of AIGS in intraoperative aid to glioma subtype diagnosis

The mean detection time for AIGS was 58.06 ± 0.5 min, and the mean reporting time for intraoperative rapid histopathology was 29.46 ± 12.74 min (Fig. 2B). The AIGS results showed that 41.46% of patients had IDH mutations, all of which were IDH1 R132H mutations, and 48.78% had TERTp mutations (15 with C228T mutations and 5 with C250T mutations). Significantly, TERTp mutations were detected in all eight patients with oligodendroglioma. In comparison with the NGS/Sanger sequencing, the accuracy of the AIGS test was 100% (Fig. 2C).

According to CNS5, the postoperative integrated diagnosis of glioblastoma, IDH‐wt (58.54%) in 41 patients, followed by oligodendroglioma, IDH‐mt and 1p/19q‐codeleted (19.51%), and astrocytoma, IDH‐mt (21.95%, Fig. 2D). The pathologist could only initially distinguish between astrocytoma and glioblastoma based on the IDH mutation test results and intraoperative rapid histopathology, and the preliminary diagnosis was 41.46% for astrocytoma and 58.54% for glioblastoma. Using the postoperative integrated diagnosis as a reference standard, the accuracy of the preliminary diagnosis was only 80.49%. In contrast, in the intraoperative integrated diagnosis, the percentage of oligodendroglioma was 19.51%, while the percentage of astrocytoma decreased to 21.95%, and the accuracy was 100% (Fig. 2E and Table S5). The EOR of the patients in the AIGS detection group was 99.06%, whereas the EOR of the control group was 93.75%, with a statistically significant difference between the two groups of patients (p < 0.0001, Fig. 2F).

Intraoperative definition of the molecular margins of TERTp‐mt GBM

However, we found that even if the EOR exceeded 50% of the T2w/Flair abnormal area, some patients still showed tumor progression within a short period of time. ²⁷ TERTp mutations, one of the most common mutation types in GBM, are not present in normal brain tissue. Therefore, we propose to use the TERTp mutation as a biomarker to detect the tumor molecular margins of TERTp‐mt GBM. For patients with an intraoperative integrated diagnosis of GBM (IDH‐wt/TERTp‐mt), the tumor molecular margin is rapidly defined by multi‐point sampling of the tumor cavity to detect TERTp mutations. If all test results are negative, the resection is considered to reach the molecular margin of the tumor, and the operation is completed; if the test results are positive at any site, the tumor is considered to remain in that direction and further expansion of the resection is required to achieve a complete resection of the tumor (Fig. 3A). In this study, we assisted in determining the tumor molecular margins by detecting TERTp mutations in two patients with GBM and showed a detailed surgical procedure for one patient. We performed multi‐point sampling (D, F, T, and B) to detect the TERTp mutation at the edge of the resection cavity in a GBM patient with an intraoperative AIGS test result that was the IDH wild and the TERTp C228T mutation (Fig. 3B,C). All AIGS tests were negative, indicating that the extent of surgical resection was sufficient to reach normal brain tissue (Fig. 3D). Postoperative MRI showed a clean surgical resection (Fig. 3E), and Sanger sequencing showed that the patient had IDH‐wt and TERT C228T‐mutated (Fig. 3F).

Figure 3.

Intraoperative use of TERTp mutations to determine tumor molecular margins in a patient with GBM. (A) The flow chart for precision resection in patients with GBM. (B) Results of intraoperative IDH and TERTp mutation detection. (C) Multi‐point sampling of the tumor cavity (D: depth, F: front, T: top, B: back). (D) Results for each part of the test. (E) Preoperative and postoperative MRI comparison (Red: areas of T1w enhancement; Blue: areas of T2w/Flair abnormalities; Green: areas of surgical resection). (F) Postoperative Sanger sequencing results.

The effect of molecular pathology‐guided personalized resection on the prognosis of glioma patients

Molecular pathology‐guided extended resection of IDH‐wt gliomas significantly improves OS and PFS in patients in the AIGS detection group and non‐AIGS detection group (26.77 vs. 13.47 months, p = 0.0289, and 15.90 vs. 10.57 months, p = 0.0181, Fig. 4A,B). Further analysis revealed that in the AIGS detection group, the EOR of IDH‐wt glioblastoma patients was 98.63%, whereas in the control group, it was only 90.79% (p < 0.0001, Fig. S1A). However, for low‐grade IDH‐mt gliomas, the EOR for both groups was (99.66% vs. 96.87%, p > 0.05, Fig. S1B). But the follow‐up time did not reach their median survival, and the difference between the two groups was not statistically significant (Fig. S1C,D). Yet patients with the three glioma subtypes in the AIGS detection group had a significant difference in prognosis (Fig. S1E,F). And the OS of TERTp‐mt GBMs was 26.77 months and the PFS was 18.57 months. In contrast, the OS of TERTp‐wt GBMs was 24.7 months, and the PFS was only 14 months (Fig. 4C,D).

Figure 4.

Postoperative survival of patients. (A and B) Overall survival and progression‐free survival of IDH‐wt patients in AISG‐detected and non‐AIGS‐detected groups. (C and D) Overall survival and progression‐free survival of TERTp‐mt and TERTp‐wt GBM patients in the AISG‐detected group after surgery. (E and F) Overall survival and progression‐free survival of patients with IDH‐wt/TERTp‐mt compared to IDH‐wt/TERTp‐wt in relevant studies in the last decade.

Discussions

The EOR is one of the most significant factors affecting the prognosis of patients with gliomas. ⁵ , ²⁸ , ²⁹ Although the diagnosis and treatment of gliomas have entered the molecular era, the development of relevant detection technologies still does not meet our intraoperative needs. Currently, we can only rely on rapid intraoperative histopathology for initial histological classification, and even some patients need to spend a lot of time on repeated intraoperative sampling and sending of samples for examination, which severely restricts the neurosurgeon's ability to perform surgery. Our previous research has demonstrated that detection of IDH mutations combined with rapid histopathology can effectively improve the accuracy of the intraoperative diagnosis of gliomas, which empowers neurosurgeons to be more proactive. ²³ However, this is not enough because there are significant differences in the difficulty of surgical resection, postoperative adjuvant treatment regimens, and the prognosis of gliomas of different subtypes and sites. ³⁰ Therefore, a clear intraoperative subtype diagnosis of glioma will assist in treatment decisions, both intraoperatively in developing a surgical resection regimen and postoperatively in determining the need for adjuvant therapy in the near term. Our predecessors had found that TERTp‐mt was always accompanied by 1p/19q deletion in IDH‐mt gliomas as early as a decade ago but failed to study it in depth due to the limitation of genetic testing technology in surgery. ¹³ , ³¹ In this study, we used PCR‐modified AIGS for rapid detection of IDH/TERTp mutations with the goal of achieving subtype diagnosis of glioma. We increased the total resection rate of glioma patients by developing personalized surgical regimens for all types of gliomas.

The principle of surgical resection of gliomas is maximally safe resection. We usually consider EOR reaching the imaging margins (the margins shown on brain MRI) as a total resection of the glioma, but it is difficult to judge whether this actually achieves a complete resection of the tumor cells. Relevant studies have found that glioblastomas tend to recur after surgery, and most occur within 2 cm of the peritumor. ¹⁶ , ³² , ³³ Li et al ³⁴ found that for glioblastoma patients, resecting ≥53.21% of the abnormal T2w/Flair area significantly prolonged patient survival. Therefore, we performed SupTR for glioblastoma, especially TERTp‐mt GBM. Considering that patients with astrocytomas and oligodendrogliomas do not achieve median survival, the prognosis of patients with molecular pathology‐guided GBM resection was significantly improved when analyzed separately. However, we found that even though the EOR had exceeded 50% of the T2w/Flair abnormal area, many GBM patients still experienced tumor recurrence within a short period of time after surgery. Glioblastoma cells have an infiltrative growth pattern that results in a wide range of tumor invasions, with tumor cells often spreading to areas of peritumoral brain edema (PTBE) beyond the area of T1‐CE and even to more distant brain tissue. ³⁵ , ³⁶ In response, neurosurgeons can only perform empirical resections. The application of techniques such as iMRI, iUS, and neuronavigation has significantly improved the rate of total resection of gliomas, but the overall prognosis of patients is poor. The neuronavigation technique is established based on preoperative imaging, which is somewhat hampered by the inevitable brain drift that occurs with positional changes and cerebrospinal fluid release during surgery. ³⁷ With the use of iMRI and iUS, brain drift is corrected by real‐time localization, but iUS does not accurately identify tumor boundaries because of its low resolution, while iMRI is expensive and difficult to popularize. ³² , ³⁸ Further to the introduction of molecular pathology in gliomas, we expect to have an objective indicator to clarify the EOR of gliomas. Therefore, for TERTp‐mt GBM patients, we propose to use the TERTp mutation as a tumor marker to determine tumor molecular margins. In the present study, we determined the tumor molecular margins by multi‐point sampling of the surgical resection boundaries of 2 TERTp‐mt GBM cases and were surprised to find that the molecular margins were not only located within the edema area but in some directions even reached beyond the edema area. This may be one of the reasons for the recurrence of glioblastoma patients within a short period of time after surgery. Identifying glioma molecular margins using the AIGS assay is perhaps the best option for achieving total glioma resection.

With the publication of CNS5, an increasing number of researchers are committed to exploring effective genetic testing tools to guide the precise diagnosis and treatment of gliomas. Currently, the mainstream research directions include the following: (i) using mass spectrometry to detect the tumor metabolite 2‐hydroxyglutaric acid to determine IDH mutations, which has a short detection time but poor accuracy. (ii) PCR‐based improved genetic testing techniques can directly detect IDH/TERTp mutations or other mutations, but are mostly in the laboratory testing stage. ³¹ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ In this study, all AIGS test was completed in the operating room, taking less time and with higher sensitivity and specificity. Compared to other detection technologies, our technology has the advantages of instrument miniaturization and simultaneous detection of multiple channels. And we have completed clinical translation. Although the AIGS detection time was approximately 1 h, the surgery time in the AIGS detection group was not significantly prolonged compared with the control group, and the incidence of postoperative complications did not change significantly, indicating that the technique has high clinical utility and safety. Then, we systematically reviewed the last decade of research on IDH‐wt gliomas through the literature (Table S6). ¹³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ We found that GBMs with TERTp mutations had poorer OS and PFS (Fig. 4E,F). However, in the present study, we found no difference in OS and PFS between patients with TERTp‐mt GBM and those with TERTp‐wt GBM after surgical resection based on molecular pathology guidance. The PFS of TERTp‐mt GBM patients was 18.57 months, much longer than the 14.0 months of TERTp‐mt GBM patients. Further analyses showed that the total resection rate was 72.73% (8/11) in TERTp‐mt patients compared to 61.54% (8/13) in TERTp‐mt patients. Another point of note is that the two patients whose surgical resection reached the tumor molecular margins had no recurrence at their recent review. Therefore, we believe that if we can rapidly define the glioma subtype diagnosis intraoperatively and accurately identify the tumor molecular margins, it will revolutionize the precise resection of gliomas, especially glioblastomas. Even simply detecting the presence of TERTp mutations in glioblastoma patients intraoperatively can have a positive impact on surgery for GBM and, in turn, improve patient prognosis. Furthermore, the rapid intraoperative diagnosis of the patient's glioma subtype can shorten the waiting time for postoperative radiotherapy and chemotherapy, and has the potential to enable in situ targeted treatment of gliomas intraoperatively. ⁴⁸ In addition, this technique can be used to diagnose diseases other than gliomas, such as acute myeloid leukemia, melanoma, and meningioma. ⁴⁹ , ⁵⁰ , ⁵¹

Limitations

Although we found in our study that the personalized surgical regimen guided by intraoperative rapid molecular diagnosis contributed to an improved survival prognosis for patients with GBM, this may be limited by our small sample size, which requires more data and a longer follow‐up period to validate, and we are undertaking. For AIGS to identify the tumor molecular margins of TERTp‐mt GBM, only preliminary explorations have been performed, and more clinical cases are still needed to further validate the feasibility of this surgical protocol. Furthermore, our next‐generation detection technology has achieved accurate detection of IDH and TETRp mutations in less than 30 minutes.

Conclusions

The rapid detection of IDH/TERTp mutations by AIGS enables precise intraoperative glioma staging diagnosis. By effectively advancing postoperative integrated diagnosis to intraoperative, this technique could provide important data for neurosurgeons to develop personalized surgical strategies and precise resections of gliomas.

Funding Information

Research on the future application prospect of classical gene marker Kit (Contract No. 6010122006), Key Clinical Research Project of Clinical Research Center of Shandong University (2020SDUCRCA011), Taishan Pandeng Scholar Program of Shandong Province (No. tspd20210322), Taishan Scholar Program of Shandong Province (No. tsqn202211316), and Shandong Province Youth Creative Team Program (No. 2022KJ011).

Conflict of Interest

The authors have declared that no conflict of interest exists.

Author Contributions

Jia Li and Zhe Han initiated and designed the study and led the data collection, data analysis, and manuscript writing throughout. Gang Li, Hao Xue, and Xueen Li participated in the design and interpretation of the study. Caizhi Ma and Huizhong Chi contributed to the data collection and collation. Deze Jia, Kailiang Zhang, and Zichao Feng participated in the provision of specimens. Bo Han and Mei Qi participated in the glioma pathology diagnosis.

Supporting information

Figure S1.

Table S1.

Caption.

Funding Statement

This work was funded by Key Clinical Research Project of Clinical Research Center of Shandong University grants 2020SDUCRCA011 and 6010122006; Shandong Province Youth Creative Team grant 2022KJ011; Taishan Pandeng Scholar Program of Shandong Province grant tspd20210322; Taishan Scholar Program of Shandong Province grant tsqn202211316.

Data Availability Statement

The data that support the findings of this study are available in the Supplementary Material of this article.