Prognostic significance of oligodendrocyte transcription factor 2 expression in glioma patients: A systematic review and meta-analysis

Abstract

BACKGROUND

Gliomas are the most common primary central nervous system neoplasm. Despite recent advances in the diagnosis and treatment of gliomas, patient prognosis remains dismal. Therefore, it is imperative to identify novel diagnostic biomarkers and therapeutic targets of glioma to effectively improve treatment outcomes.

AIM

To investigate the association between oligodendrocyte transcription factor 2 (Olig2) expression and the outcomes of glioma patients.

METHODS

The PubMed, Embase, Cochrane Library, and China National Knowledge Infrastructure databases were searched for studies (published up to October 2023) that investigated the relationship between Olig2 expression and prognosis of glioma patients. The quality of the studies was assessed using the Newcastle Ottawa Scale. Data analyses were performed using Stata Version 12.0 software.

RESULTS

A total of 1205 glioma patients from six studies were included in the meta-analysis. High Olig2 expression was associated with better outcomes in glioma patients [hazard ratio (HR): 0.81; 95% (confidence interval) CI: 0.51-1.27; P = 0.000]. Furthermore, the results of subgroup meta-analysis showed that high expression of Olig2 was associated with poor overall survival in European patients (HR: 1.34; 95%CI: 0.79-2.27) and better prognosis in Asian patients (HR: 0.43; 95%CI: 0.22-0.84). The sensitivity analysis showed that no single study had a significant effect on pooled HR, and there was also no indication of publication bias according to the Egger’s and Begger’s P value test or funnel plot test.

CONCLUSION

High Olig2 expression may have a positive impact on the prognosis of glioma patients, and should be investigated further as a prognostic biomarker and therapeutic target for glioma.

Key Words: Glioma, Oligodendrocyte transcription factor 2, Prognosis, Biomarker, Meta-analysis

Core Tip: The overall prognostic value of oligodendrocyte transcription factor 2 (Olig2) for glioma remains unclear. We performed the first meta-analysis of previously published studies related to the potential prognostic value of Olig2 in glioma patients to evaluate the association between Olig2 expression and the outcomes of glioma patients. Our findings suggest that high Olig2 expression may have a positive impact on the prognosis of glioma patients, and should be investigated further as a prognostic biomarker and therapeutic target for glioma.

INTRODUCTION

Gliomas are primary tumors of the central nervous system (CNS) that originate from glial cells. Despite advances in the diagnosis and treatment of gliomas, patient prognosis remains dismal, and the general survival duration of the patients is 12-14 months[1,2]. Therefore, it is critical to identify reliable biomarkers for glioma in order to predict patient prognosis, improve clinical outcomes, and devise suitable individualized treatment plans[3]. In recent years, the identification of early diagnostic molecular biomarkers has played a vital role in predicting overall survival and guiding personalized treatments for patients with glioma[4]. Oligodendrocyte transcription factor 2 (Olig2) in particular is significantly associated with mortality[5,6].

Olig2 is a basic helix-loop-helix neural transcription factor that is ubiquitously expressed in diffuse glioma cells with oligodendroglial and astrocytic morphologies and plays an important role in the development of gliomas[7-9]. It is also a specific marker of glioblastoma (GBM) progenitor-like cells with stem-cell properties that initiate gliomagenesis and reside in the subventricular zone[6,10]. Inhibition of the Olig2 pathway delays glioma growth and sensitizes tumor cells to radiation therapy[11]. Hence, Olig2 expression or activity is a promising therapeutic target for the discovery of novel candidate drugs against glioma[12].

In recent years, numerous studies have analyzed the relationship between Olig2 expression and prognosis in glioma patients. However, findings of a single study were often difficult to generalize due to differences in research methods, sample sizes, and populations. Therefore, the overall prognostic value of Olig2 for glioma remains unclear. Here, a meta-analysis systematically assessed the literature on the relationship between Olig2 expression and prognosis in glioma patients. To our knowledge, this is the first meta-analysis to evaluate the prognostic value of Olig2 in patients with glioma.

MATERIALS AND METHODS

Literature search

The articles were searched from the PubMed, Cochrane Library, Embase, and China National Knowledge Infrastructure databases using the following keywords and text words: (“glioma” or “glioblastoma” or “CNS cancer” or “glial cell tumor” or “GMB”) and (“oligodendrocyte transcription factor 2” or “Olig2”) and (“survival” or “mortality” or “prognosis” or “follow-up studies” or “incidence” or “clinical significance”). The final search was updated in October 2023. Furthermore, the reference lists of the selected research articles and reviews were manually searched for additional relevant publications.

Study selection

The inclusion criteria for the studies were as follows: (1) Diagnosed with glioma on the basis of pathological examination, and classified according to the World Health Organization (WHO) guidelines; (2) Investigated the prognostic relevance of Olig2 in glioma; and (3) Provided direct estimation of hazard ratio (HR) and included 95% (confidence interval) CI, or presented Kaplan-Meier curves. If the same glioma patient population was analyzed in more than one study, then the report with the largest sample size was kept.

The exclusion criteria were as follows: (1) Repetitious study; (2) Studies with sample size less than 30; (3) Letters, case reports, comments, meeting records, and reviews; (4) Animal or laboratory studies or other types of CNS malignancies studies; and (5) Studies lacking sufficient data and contact information of the authors.

Data extraction

Two independent reviewers (Li PC and Huang QY) independently extracted the data from the selected studies. In case of any discrepancies, the final decision was made by consulting with the third author (Yun DB). The following data were extracted: First author, year of publication, country, sample size, proportion of male patients, median age, WHO grade, technique used to assay Olig2 levels, cut-off level, follow-up time, and Newcastle Ottawa Scale (NOS) scores. The HR for Olig2 expression and the corresponding 95%CI for the Olig2^low/negative (control) and Olig2^{high/positive} (experimental) groups were directly extracted or obtained from the Kaplan-Meier curves. In case both univariate and multivariate analyses had been performed for the overall survival (OS), the multivariate analysis was preferentially extracted. If HR were not mentioned, the logarithm-transformed HR and variance were estimated on the basis of Kaplan-Meier curves using Engauge Digitizer version 12.1[13].

Risk of bias

Two reviewers (Li PC and Huang QY) independently assessed the quality of the primary studies using the NOS, with the full score of 9. Studies with NOS scores ≥ 6 were defined as high-quality. Any disagreements were resolved by joint discussion.

Statistical analysis

Stata Version 12.0 (Stata Corporation, College Station, TX, United States) was used for all statistical analyses. OS was compared on the basis of HR and 95%CI. HR greater than 1 indicated poor prognosis for patients with high/positive Olig2 expression. The heterogeneity of the studies was evaluated using the I² metric. I²> 50% indicated significant heterogeneity, and the analysis was performed using a random-effect model. I²< 50% was indicative of non-significant heterogeneity, and the fixed-effect model was used[14]. Subgroup analyses were performed to investigate the potential causes of heterogeneity according to the country, sample size, median age, WHO grade, detection method of Olig2 expression, source, cut-off level, and follow-up time. P < 0.05 was considered statistically significant.

Sensitivity analysis was performed to evaluate the stability and reliability of the pooled effect size. if the results after exclusion were significantly different compared to the original, the sensitivity was high and the stability of the results was low. In contrast, the sensitivity was low and the results were considered stable and reliable. The publication bias was qualitatively tested by funnel plots and quantified by Begg’s and Egger’s tests.

RESULTS

Study selection

The process for literature retrieval and the results are summarized in Figure 1. A total of 162 potentially relevant articles were initially selected, of which 110 articles were excluded on the basis of the title and abstract: 11 studies were duplicates, 8 were reviews, 54 concerned a carcinoma not related to gliomas or GBM, and non-related to Olig2 (n = 37). The remaining 52 articles were further examined, and 46 irrelevant studies were excluded, including 30 without data, 6 incomplete data, 8 involving non-human subjects, and 2 reviews. Six articles were finally included in this meta-analysis, including three from Europe[15-17], and three from Asia[18-20].

Figure 1 Flow diagram of the literature search and study selection protocols.

Characteristics of studies and quality assessment

The general characteristics of the included studies are summarized in Table 1. The studies were published from 2016 to 2023, and three studies were conducted in Europe and three in Asia. A total of 1205 patients were included in these studies, and their age ranged from 49.95 years to 64 years. The minimum sample size was 73, and the maximum sample size was 471. Two studies evaluated grade II-IV gliomas and four examined grade IV glioma. The HR and 95%CI for OS could be directly extracted from two studies; it was calculated from the Kaplan-Meier curves in the remaining studies. In all studies, the Olig2 expression in the tumor tissues had been evaluated by immunohistochemistry (IHC). The percentage of patients with Olig2-positive gliomas ranged from 18.42% to 95.75%. The follow-up time ranged from 30 months to 60 months. All included studies were high-quality with NOS scores ≥ 6.

Table 1 Characteristics of included studies.

Ref.	Year	Country	Male/All sex	Positive males	Positive females	Median age	Grade I+II/III+IV	Positive I+II/III+IV	Method	HR and 95%CI	HR and 95%CI source	Cut-off	Follow-up time in months	NOS score
Batista et al[15]	2016	Spain	79/152	13	58	-	0+77/75+0	0+12/16+0	IHC	0.98 (0.84-1.15)	Reported	-	50	7
Schäfer et al[16]	2018	Germany	282/471	268	183	50.6	0+111/141+219	0+106/137+208	IHC	2.37 (1.37-4.01)	Reported	10	-	6
Behling et al[17]	2020	Germany	67/113	-	-	64	0/0+113	0/0+89	IHC	1.16 (0.66-2.03)	K-M curve	10	37	7
Zhong et al[18]	2020	China	167/277	60	60	51	0/0+277	0/0+216	IHC	0.66 (0.46-0.94)	K-M curve	10	60	6
Zhou et al[19]	2023	China	49/73	29	13	49.95	0/0+73	0/0+42	IHC	0.11 (0.05-0.65)	K-M curve	30	30	6
Mei et al[20]	2023	China	78/119	-	-	51.1	0/0+119	0/0+74	IHC	0.49 (0.28-0.82)	K-M curve	10	60	7

HR: Hazard ratio; IHC: Immunohistochemistry; K-M: Kaplan-Meier; NOS: Newcastle Ottawa Score.

Correlation between OS and Olig2 expression in gliomas

The pooled HR was 0.81 (95%CI: 0.51-1.27, P = 0.000), suggesting that high Olig2 expression is a protective factor in glioma patients. Figure 2 shows a forest plot of the association between Olig2 expression and OS. Due to the significant heterogeneity among the studies (Cochran’s Q, I² = 84.5%, P = 0.000), we used a random model for the meta-analysis.

Figure 2 Forest plot illustrating the relationship between oligodendrocyte transcription factor 2 expression and overall survival in glioma patients. Each study is accompanied by a point estimate of its hazard ratio (HR) and 95% confidence interval (CI) (extending lines). The diamonds represent the estimated pooled effect (labelled ‘total’).

Sub-group analysis

For heterogeneity values > 50%, meta-regression and subgroup analysis were performed to explore its potential source, and the results are summarized in Table 2. We stratified the patients based on the geographical area, sample size, median age, WHO grade, data extraction method, cut-off level, and follow up period. The results of subgroup analysis indicate that the geographical area may be related to heterogeneity. As shown in Figure 3, high expression of Olig2 was associated with better OS in patients of Asian descent (HR = 0.43, 95%CI: 0.22-0.84, P = 0.027) and with poor OS in European patients (HR = 1.34, 95%CI: 0.79-2.27, P = 0.008). No significant difference was observed among the other subgroups.

Figure 3 Forest plot for the association between oligodendrocyte transcription factor 2 expression and overall survival in terms of subgroup analysis of the testing index. HR: Hazard ratio.

Table 2 Subgroup analysis of the relationships between oligodendrocyte transcription factor 2 expression and overall survival.

Comparison variables	Number of studies	Number of people	HR and 95%CI	P value	Heterogeneity by I2 statistics
Total	6	1205	0.81 (0.51-1.27)	0.000	84.5
Country
Europe	3	736	1.34 (0.79-2.27)	0.008	79.3
Asian	3	469	0.43 (0.22-0.84)	0.027	72.3
number
> 150	3	900	1.10 (0.65-1.86)	0.001	86.7
≤ 50	3	305	0.46 (0.17-1.29)	0.002	84.0
Median age
> 50	4	980	0.96 (0.50-1.84)	0.000	85.8
≤ 50	1	73	0.11 (0.03-0.40)	-	-
No referred	1	152	0.98 (0.84-1.15)	-	-
WHO grade
II-III	2	404	1.47 (0.62-3.47)	0.002	89.6
IV	4	801	0.72 (0.36-1.45)	0.000	87.0
Method
Reported	2	623	1.47 (0.62-3.47)	0.002	89.6
K-M curve	4	582	0.56 (0.31-1.01)	0.006	76.0
Cut-off
> 10	1	73	0.11 (0.03-0.40)	0.000	85.8
≤ 10	4	980	0.96 (0.50-1.84)	-	-
Follow up time
> 50	2	396	0.60 (0.45-0.81)	0.366	0.0
≤ 50	3	338	0.67 (0.31-1.46)	0.003	82.6

HR: Hazard ratio; K-M: Kaplan-Meier; WHO: World Health Organization.

Sensitivity analysis and publication bias

The sensitivity analysis results showed that they are reliable (Figure 4). The funnel chart did not show any signs of publication bias, and Begger’s test was applied to provide statistical evidence for funnel plot symmetry. As expected, the P value of Begger’s test was 0.260 (Figure 5) and the P value of Egger’s test was 0.532 (Figure 6). Thereby, the possibility of publication bias can be ruled out in this study.

Figure 4 Sensitivity analysis of the relationship between oligodendrocyte transcription factor 2 and overall survival.

Figure 5 Begger’s funnel plot for the publication bias test of the oligodendrocyte transcription factor 2 expression and overall survival of gliomas.

Figure 6 Egger’s funnel plot for the publication bias test of the oligodendrocyte transcription factor 2 expression and overall survival of gliomas.

DISCUSSION

The considerable biological and genetic heterogeneity among glioma cells may account for the poor overall survival of glioma patients[21]. Therefore, it is imperative to identify novel diagnostic biomarkers and therapeutic targets of glioma to effectively improve treatment outcomes[22]. The interaction between inflammatory factors and tumor cells and other components of the tumor microenvironment may influence prognosis[23-25]. For example, Rahbar et al[26] found that higher numbers of neutrophils and macrophages in and around glioma tumors correlated with a shorter time to tumor progression. This overall poor prognosis may in some part be attributed to the immunosuppressive mediators within the tumor environment. However, the changes of these serological markers are not restricted to specific tumors and may be indicative of cancer growth in general. The current research focus for gliomas is the identification of novel targeted molecular markers. Previous studies have shown that Olig2 is a CNS-restricted transcription factor that is overexpressed in all grades of diffuse gliomas and plays a critical role in glial progenitor proliferation[27]. Trépant et al[6] identified Olig2 as the most specific GBM stem cell marker. In addition, several integrative studies have demonstrated that Olig2 directly activates regulators of cell cycle and oncogenic factors in tumors, and accelerates proliferation during tumorigenesis. In fact, deletion of Olig2 has been shown to delay glioma growth and improve survival[28-30]. However, some research groups have reported that higher Olig2 expression correlates with better overall survival of glioma patients[31-34].

To further ascertain the prognostic role of Olig2, we performed a systematic meta-analysis. Six eligible studies with 1205 patients were included, and the pooled HR with 95%CI were calculated. The final results showed that Olig2^{high/positive} expression was associated with better outcomes in glioma patients (HR = 0.81). We did not find any evidence of publication bias among the included studies. Based on these results, we can conclude that our results are reliable, and high expression of Olig2 has a predictive value for glioma patients.

Interestingly, the results of meta-regression analysis indicate that high expression of Olig2 had opposite effects on the OS of European and Asian patients (age, WHO grade, etc may not be potential sources of heterogeneity). Olig2 was an indicator of favorable prognosis in the Asian cohorts (P < 0.05), and a similar trend was observed with the survival data retrieved from the Cancer Genome Atlas database. However, European glioma patients with lower Olig2 expression (10% cut-off) survived longer compared to the Olig2^high patients (P = 0.008). These findings might be attributed to the heterogeneous clinical outcomes of glioma due to other prognostic markers or the complex tumor micro-environment, which should be assessed in future research. The present study was performed on a heterogeneous population of glioma patients, all of whom underwent resection, subtotal resection or biopsy followed by the same adjuvant therapy (chemotherapy and radiotherapy). The extent of tumor resection is an established prognostic factor for glioma[35]. However, surgical habits may vary across the different regions of Europe and Asia. In addition, previous studies have shown that Olig2 is not significantly expressed in glioma type II and III[36,37]. However, all WHO graded II-III glioma cases in the studies included in this meta-analysis were from Europe. This may be one of the reasons that affected the results of our subgroup analysis.

However, this study has several limitations that ought to be considered. Despite our efforts to collect all relevant data, the number of cases was too few, resulting in potential publication bias. Furthermore, some studies did not provide complete results and data, which may reduce the reliability of Olig2 expression as a prognostic indicator of glioma. In all included studies, Olig2 expression was detected by IHC. However, the results of IHC were affected by the primary antibody clones and concentrations. We were unable to perform a subgroup analysis of different antibodies to evaluate the underlying bias of the detection method on the pooled HR. In addition, all six studies were single center retrospective analyses with different cut-off values, which can also lead to bias. Furthermore, the HR with 95%CI for some studies were obtained from the Kaplan-Meier curve and not directly extracted, which may have increased the potential for bias. The articles included in this meta-analysis were also limited to those published in English and Chinese, resulting in publication bias due to the possible exclusion of relevant studies in other languages. In addition, differences in the polyclonal anti-Olig2 antibody testing kits and tumor sample preparation, and possible experimental errors may have introduced information bias, thereby affecting the measurement of Olig2 levels. It is important to note that we focused on the expression and not the activity of Olig2, which may or may not be correlated to each other. Thus, every factor that may have affected the analysis should be fully considered. In addition, multi-center cohort studies with larger sample sizes are needed to validate our findings, along with a deeper exploration of the functional mechanisms.

CONCLUSION

High Olig2 expression may have a positive impact on the prognosis of glioma patients, and Olig2 is a potential biomarker and therapeutic target that warrants further investigation. In light of the limitations of the current analysis, well-designed prospective clinical studies are required to elucidate the role of Olig2 in glioma progression to further evaluate its predictive value for glioma patients.