Characteristics of the enrolled studies
The search procedure was presented in Fig. 1. Among the potential literature retrieved from databases, 170 studies were initially assessed. After removing 90 duplicate publications, the remaining 80 studies were evaluated for titles and abstracts. Of these, 32 irrelevant articles were further excluded after abstract review and only 48 studies remained for full-text verification. Moreover, 22 studies were further eliminated with a variety of reasons, including five studies not related to circRNAs or osteosarcoma, six studies did not report outcomes, two reviews and case reports, four animal studies, and five lack of extractable data. Finally, 26 studies comprising 1,652 osteosarcoma cases were enrolled in this quantitative study, with 18 on clinicopathological features, ten on diagnosis [13, 18, 19, 21, 24, 30–34] and 18 on prognosis [12–14, 16–20, 22–24, 30–32, 34–42]. Among the 1,652 eligible osteosarcoma cases, patients were separated by age, gender, or other clinicopathologic features. As for age, the included studies had divergent cut-off values, with six studies on upregulated circRNAs using ≥ 25 or < 25 years old [12, 14, 18, 19, 32–34, 36–40], whereas other two studies on downregulated circRNAs applying ≥ 18 or < 18 years old to separate enrolled patients [13, 30]. In consideration of the consistency, only these studies with common cut-off value were included in the following meta-analysis with regard to patients’ age. Besides, all cases in studies on clinicopathological feature analysis were divided into male and female. Other clinicopathologic parameters consist of tumor size (≥ 5 cm/<5 cm), Enneking stage (IIB-III/I-IIA), distant metastasis (positive/negative), and WHO grade (III/I-II) were also presented in this analysis.
The baseline characteristic features of enrolled studies were shown in detail in Table 1 and Table 2. All studies were conducted in China, and the publication year ranged from 2017–2020. All osteosarcoma cases were reliably diagnosed based on histopathology. The sample size was varied from 30 to 170. CircRNAs expression was measured by quantitative real-time polymerase chain reaction (qRT-PCR) and the reference gene was GAPDH. CircRNAs were regarded as oncogenes or tumor-suppressor in terms of their expression levels. Specifically, 21 circRNAs were identified as oncogenic circRNAs with upregulated expression pattern [12, 14, 16–23, 31–34, 36, 37, 39–43], while the other five were recognized as tumor suppressors [13, 24, 30, 35, 38] with downregulated expression, respectively. The duration for follow-up ranged from 40 to 125. For diagnostic assay, ten studies with data on sensitivity, specificity and AUC [13, 18, 19, 21, 24, 30–34].
Table 1
Main characteristics of studies for diagnosis analysis in osteosarcoma
Study | Publication year | CircRNA | Sample type | Sample size | Method | Reference gene | Regulation pattern | AUC | Ref. |
Case | Control |
Li, S et al | 2020 | circ_0000190 | serum | 60 | 60 | qRT-PCR | GAPDH | downregulated | 0.889 | [30] |
Hu, Y et al | 2019 | circ-LARP4 | tissue | 72 | 72 | qRT-PCR | GAPDH | downregulated | 0.829 | [13] |
Zheng, S et al | 2019 | circLRP6 | tissue | 50 | 50 | qRT-PCR | GAPDH | upregulated | 0.874 | [31] |
Zhu, K et al | 2019 | circ_0000885 | serum | 30 | 25 | qRT-PCR | GAPDH | upregulated | 0.783 | [32] |
Ma, X et al | 2018 | circ_HIPK3 | serum | 50 | 20 | qRT-PCR | GAPDH | downregulated | 0.783 | [24] |
Xu, B et al | 2018 | CDR1as | tissue | 38 | 18 | qRT-PCR | GAPDH | upregulated | 0.857 | [33] |
Zhou, X et al | 2018 | circ_0008717 | tissue | 45 | 45 | qRT-PCR | GAPDH | upregulated | 0.782 | [34] |
Zhu, K et al (a) | 2018 | circPVT1 | serum | 50 | 20 | qRT-PCR | GAPDH | upregulated | 0.871 | [19] |
Zhu, K et al (b) | 2018 | circ_0081001 | serum | 50 | 20 | qRT-PCR | GAPDH | upregulated | 0.898 | [18] |
Liu, X et al | 2017 | circ-NT5C2 | tissue | 52 | 52 | qRT-PCR | GAPDH | upregulated | 0.753 | [21] |
Abbreviations: AUC, area under curve; CircRNA, circular RNA; qRT-PCR, quantitative real time polymerase chain reaction; Ref, reference
Table 2
Main characteristics of the meta-analysis for prognosis association of circRNAs in osteosarcoma
Study | Publication Year | CircRNAs | CircRNA expression | Sample type | Detection assay | Regulation pattern | Survival indicators | Survival analysis | Follow-up months | Ref. |
High | Low |
Wang, L et al | 2020 | circ_0001658 | 21 | 18 | tissue | qRT-PCR | upregulated | N/A | N/A | N/A | [36] |
Li, S et al | 2020 | circ_0000190 | 30 | 30 | tissue | qRT-PCR | downregulated | N/A | N/A | N/A | [30] |
Wang, Y et al | 2020 | circTCF25 | 26 | 24 | tissue | qRT-PCR | upregulated | N/A | N/A | N/A | [37] |
Ji, X et al | 2020 | circ_001621 | 20 | 10 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [12] |
Zhu, K et al | 2019 | circ_0000885 | 25 | 25 | tissue | qRT-PCR | upregulated | OS/DFS | Multivariate | 60 | [32] |
Zheng, S et al | 2019 | circLRP6 | N/A | N/A | tissue | qRT-PCR | upregulated | OS/DFS | Multivariate | 125 | [31] |
Wang, L et al | 2019 | circ_0021347 | 35 | 35 | tissue | qRT-PCR | downregulated | OS | Univariate | 40 | [38] |
Qi, H et al | 2019 | circ_0000502 | 34 | 29 | tissue | qRT-PCR | upregulated | OS | Multivariate | 60 | [43] |
Jin, J et al | 2019 | circ_100876 | 24 | 24 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [16] |
Pan, G et al | 2019 | circMMP9 | 24 | 27 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [39] |
Li, L et al | 2019 | circ_0001721 | 28 | 24 | tissue | qRT-PCR | upregulated | OS | Multivariate | 60 | [20] |
Jin, Y et al | 2019 | circ_0102049 | 38 | 38 | tissue | qRT-PCR | upregulated | OS | Multivariate | 60 | [17] |
Hu, Y et al | 2019 | circ-LARP4 | 36 | 36 | tissue | qRT-PCR | downregulated | OS/DFS | Univariate | 48 | [13] |
Ma, X et al | 2018 | circ_HIPK3 | 37 | 45 | tissue | qRT-PCR | downregulated | OS | Univariate | 60 | [24] |
Zhu, K et al | 2018 | circPVT1 | 30 | 50 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [19] |
Zhu, K et al | 2018 | circ_0081001 | 27 | 55 | tissue | qRT-PCR | upregulated | OS | Multivariate | 60 | [18] |
Zhu, K et al | 2018 | circ_0004674 | 23 | 37 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [40] |
Zhou, X et al | 2018 | circ_0008717 | N/A | N/A | tissue | qRT-PCR | upregulated | OS/PFS | Multivariate | 80 | [34] |
Huang, L et al | 2018 | circNASP | 19 | 20 | tissue | qRT-PCR | upregulated | N/A | N/A | N/A | [14] |
Nie, W et al | 2018 | circ-NT5C2 | 86 | 84 | tissue | qRT-PCR | upregulated | OS/DFS | Multivariate | 60 | [41] |
Li, B et al | 2018 | circ_0007534 | 31 | 26 | tissue | qRT-PCR | upregulated | OS | Multivariate | 60 | [23] |
Wu, Z et al | 2018 | circ_0002052 | 54 | 54 | tissue | qRT-PCR | downregulated | OS/PFS | Univariate | 60 | [35] |
Zhang, H et al | 2018 | circ_001569 | 20 | 16 | tissue | qRT-PCR | upregulated | N/A | N/A | N/A | [22] |
Zhang, H et al | 2017 | circUBAP2 | 42 | 50 | tissue | qRT-PCR | upregulated | OS | Univariate | 60 | [42] |
Abbreviations: CircRNA, circular RNA; DFS, disease-free survival; N/A, not available; OS, overall survival; PFS, progression-free survival; qRT‐PCR, quantitative real time polymerase chain reaction; Ref, reference
The quality of studies was evaluated by QUADAS II and NOS scores. For diagnostic studies, the rating scores of QUADAS II ranged from 4 to 6. While for prognostic studies, the NOS scores were from 6 to 7, suggesting high methodological quality in all selected studies.
Expression Of Circrna With Clinicopathological Parameters In Osteosarcoma
The correlation between circRNAs and clinicopathological parameters in patients with osteosarcoma were demonstrated in Table 3. Overexpression of oncogenic circRNAs were significantly correlated with poor clinical features (tumor size: OR = 4.27, 95% CI: 2.25–8.10; Enneking stage: OR = 5.52, 95% CI: 2.79–10.94; differentiation: OR = 3.06, 95% CI: 1.72–5.45; DM: OR = 4.55, 95% CI: 2.55–8.12). In contrast, upregulated expression of tumor-suppressor circRNAs were remarkably associated with improved clinicopathological characteristics (Enneking stage: OR = 0.33, 95% CI: 0.16–0.68; DM: OR = 0.18, 95% CI: 0.08–0.40). Besides, no obvious difference was noted in terms of age and gender.
Table 3
Clinicopathological characteristics of circRNAs in osteosarcoma
Clinicopathological parameters | Upregulated circRNAs | | Downregulated circRNAs |
OR | 95% CI | p | | OR | 95% CI | p |
Age | 1.02 | 0.69–1.50 | 0.93 | | 0.94 | 0.47–1.87 | 0.86 |
Gender (male/female) | 1.24 | 0.95–1.63 | 0.12 | | 1.01 | 0.59–1.74 | 0.98 |
Tumor size (≥ 5 cm/<5 cm) | 4.27 | 2.25–8.10 | < 0.00001 | | N/A |
Enneking stage (IIB-III /I-IIA) | 5.52 | 2.79–10.94 | < 0.00001 | | 0.33 | 0.16–0.68 | 0.002 |
WHO grade (III/ I-II) | 3.06 | 1.72–5.45 | 0.0001 | | N/A |
Distant metastasis (P/N) | 4.55 | 2.55–8.12 | < 0.00001 | | 0.18 | 0.08–0.40 | < 0.0001 |
Abbreviations: CI, confidence interval; N, negative; N/A, not available; OR, odds ratio; P, positive
Diagnosis Analysis
As demonstrated in Fig. 2, the forest plots showed that the pooled diagnostic value of circRNA in separating osteosarcoma from controls were as follows: sensitivity (SENS) of 0.80 (95% CI: 0.74–0.84), specificity (SPEC) of 0.80 (95% CI: 0.75–0.84), positive likelihood ratio (PLR) of 3.95 (95% CI: 3.19–4.89), negative likelihood ratio (NLR) of 0.26 (95% CI: 0.20–0.33) and combined diagnostic odds ratio (DOR) of 15.48 (95% CI: 10.85–22.10) (Fig. 2A, 2B, and 2C). Moreover, a summary receiver operator characteristic (SROC) curve was presented in Fig. 2D, and the AUC was 0.86 (95% CI: 0.83–0.89). These results indicated that circRNAs could be ideal diagnostic biomarker for osteosarcoma.
Additionally, a small set of oncogenic circRNAs with different diagnosis efficiency were synthesized in order to explore the combination panel of certain circRNAs with even higher diagnostic potential, which may facilitate the clinical application of the circRNAs for diagnosis. Among them, the pooled analysis containing circPVT1, circ_0081001, and CDR1as, has the highest diagnostic efficiency [18, 19, 33]. As shown in Figure S1, the forest plot of the pooled DOR, SENS and SPEC of these three circRNAs panel were 32.21 (95%: 9.51–109.1), 0.85 (95% CI: 0.78–0.90), and 0.83 (95% CI: 0.71–0.91), respectively. Notably, the AUC under SROC curve was 0.9116, indicating the combination of these circRNAs may have good diagnostic performance as the full panel of included circRNAs in osteosarcoma. However, more caution should be taken in the interpretation of this result since the comparatively small study numbers and limited sample size may produce outlying outcome and thereby introduce potential bias.
Expression Of Circrnas With Prognosis In Osteosarcoma
Survival analysis showed that overexpression of oncogenic circRNAs was significantly correlated with worse OS (HR = 1.92, 95% CI: 1.68–2.19), and DFS (HR = 2.65, 95% CI: 2.02–3.49) as shown in Fig. 3A and 3B, respectively. We identified one outlier study performed by Zheng S et al. in the combined effect of oncogenic circRNAs by sensitivity analysis. Besides, elevated expression of tumor-suppressor circRNAs predicted favorable OS (HR = 0.44, 95% CI: 0.28–0.69), as depicted in Fig. 3C. The fixed-effect model was applied in these studies since no obvious heterogeneity was noted.
Sensitivity Analysis And Publication Bias
The sensitivity analysis was performed in the prognostic effect sizes by omitting the enrolled studies one by one. For pooled effects of upregulated circRNAs in osteosarcoma, one study conducted by Zheng S et al. [31] was identified as the outlier (Fig. 4A). Notably, the predictive significance of OS for upregulated circRNA did not alter after excluding the aforementioned outlier data (HR = 2.51, 95% CI: 2.08–3.02). No outliers were found in other pooled effects (Fig. 4B, 4C).
The Deeks’ funnel plot test (P = 0.27) demonstrated the absence of publication bias existed in diagnostic analysis (Fig. 5A). For prognosis analysis of upregulated circRNA on DFS and downregulated circRNA on OS, the Begg’s funnel plot also showed symmetry and revealed no evidence of publication bias among the eligible studies (Fig. 5C, 5D). Nevertheless, the funnel plot of upregulated circRNA profile on OS showed significant asymmetry, indicating a possible publication bias (Fig. 5B). Therefore, a “trim and fill” method was applied to trace the possible impacts from bias as previously described. However, the predictive value of upregulated circRNAs on OS was not altered after adjustment, suggesting a closely correlation between upregulated circRNAs and poor OS among osteosarcoma patients (Fig. 5C).