Characteristics of articled included in network meta-analysis.
Abstract
This study aimed to compare the non-acid nucleic blood multi-factor panels together and with mammography in terms of sensitivity, specificity, and accuracy in primary breast cancer detection (I, II, III, and IV). We systematically reviewed studies assessing non-acid nucleic blood tumor markers panels’ diagnostic value in both healthy women and patients (before any anticancer treatment) for the detection of primary breast cancer. Out of the 2358 titles initially identified, 12 studies and 9 panels were included in the network meta-analysis. Panels I (MSA + B2m) and J (GATA3 + E-cadherin) had the highest sensitivity in all stages of primary breast cancer but had no significant difference with mammography. Panels L (MSA + CA15–3) and B (M-CSF + CA15–3) had the highest specificity in all stages compared to other panels but no remarkable difference with mammography. Panels J (GATA3 + E-cadherin) and I (MSA + B2m) respectively had the highest accuracy in primary breast cancer detection but no considerable difference with mammography in terms of accuracy. Panel J, including GATA3 + E-cadherin, demonstrated a higher diagnostic value for primary breast cancer detection (I, II, III, and IV) than the rest of the panels.
Keywords
- primary breast cancer
- blood tumor markers
- timely diagnosis
- sensitivity and specificity
- multi-factor panels
- network meta-analysis
1. Introduction
Based on our previous study [1], the necessity of a noninvasive, accessible, cost-effective, and reliable method for breast cancer detection based on blood factors was proved. Furthermore, blood multi-factor panels can be the best choice for such a method thanks to improving the sensitivity and specificity of cancer detection considerably compared to the individual state. In that study [1], we had determined the best non-acid nucleic blood multi-factor panels for breast cancer detection in early stages and locoregional breast cancer (I, II, and III). In this brief study, however, we compared the best non-acid nucleic blood multi-factor panels in primary breast cancer detection (I, II, III, and IV) by conducting a network meta-analysis. In fact, this study aimed to offer new insight into the diagnostic value of the best panels of non-acid nucleic blood tumor markers to detect primary breast cancer along all stages not only in early stages. The breast malignancy that emerges and can be diagnosed for the first time is named primary breast cancer, and if it recurs after primary treatment including surgery, chemotherapy, hormone therapy, and radiotherapy individually or collectively, it will be named secondary breast cancer [2]. Primary breast cancer comprises locoregional (I, II, III) and metastatic stages (IV) [3].
2. Materials and method
The systematic reviews of the observational studies were conducted based on PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-analysis) [4]. Eligibility criteria, search strategy (supplementary material 1 B), databases, study selection, data extraction, and statistical analysis conformed to our former study [1]. The difference is that, in this brief study, we systematically reviewed the studies that have simultaneously assessed several tumor markers in the form of a panel to diagnose and detect breast cancer in all stages of primary breast cancer (I, II, III, and IV).
All these panels were made based on simultaneous measurement of two or three blood tumor markers in patients and healthy people using a compatible linear combination method [5]. Panels (B, C, D, E, F, G) were assessed in more than one study (multiple studies), and panels (I, J, L) were only assessed in one study (single study). We conducted direct and indirect paired comparisons of the sensitivity, specificity, and accuracy of the included blood tumor markers panels for diagnosing primary breast cancer in all stages. All the investigations were conducted in comparison to mammography (M) as the gold standard [6, 7, 8], like our previous study (Figure 1) [1].
3. Results and discussion
3.1 Study selection
Study selection conformed to our former study [1]. However, in this brief study, among the 54 studies relevant to our research question which contained 86 unique blood tumor markers panels (supplementary material 2) conforming to our eligibility criteria, only 12 studies and 9 panels presented enough data for estimating sensitivity and specificity in all stages (I, II, III, and IV) of primary breast cancer and could be included in the systematic review and network meta-analysis. These 12 studies were similar in terms of pre-analytical procedures and analytical methods (Table 1).
First author and year | Country | Study design | Sample size and population | Clinical stages | Panel | Number of panel components | Sensitivity | Specificity | Accuracy | Method of chemical evaluation | Type of sample | Score |
---|---|---|---|---|---|---|---|---|---|---|---|---|
S Zajkowska, M.. 2016 [9] | Poland | Case–control | 240 Bc:120 B:60 H:60 Median age (range) 54 (34–72) | I:29 II:30 III:31 IV:30 | VEGF + CA 15–3 M-CSF + CA 15–3 VEGF + M-CSF + CA 15–3 VEGF+ M-CSF* | 2 2 3 2 VEGF M-CSF CA 15–3 | 96.25 91.25 96.25 90 76.25 60 83.75 | 65 67.5 57.5 76 85 90 75 | 80.6 79.3 76.8 83 | ELISA CMIA | plasma | 11 |
Sacks, N. P. 1987 [10] | Australia. | Case–control | 131 | I/ II:34 III/ IV:38 | MSA + CA15–3 | 2 | 84 | 100 | 89.2 | ELISA | serum | 11 |
Liu, Y. 2017 [11] | China | Case–control | 248 | I/ II:57 III/ IV:45 | *p16+ c-MYC + TP53 | 3 p16 c-MYC TP53 | 30 27.5 11.8 24.5 | 90 90 90 90 | 65.3 | ELISA | serum | 10 |
Molina, Rafael 1998[12] | Spain | Case–control | 292 | I/ II:118 III/ IV:68 | Ca15.3 + CEA | 2 Ca15.3 CEA | 29.2 15.6 18.3 | 90 | 51.2 | ELISA | serum | 9.5 |
Ławicki, Sławomir 2016 [13] | Poland | Case–control | 200 | I/ II/ III:77 IV:23(with metastases) | VEGF +CA15–3 | 2 VEGF CA 15–3 | 84 61 65 | 90 96 96 | 87 | ELISA | plasma | 11 |
Lawicki, S. 2017 [14] | Poland | Case–control | 200 | I/ II/ III:77 IV:23 | VEGF+ CA 15–3 | 2 VEGF CA 15–3 | 83 60 64 | 90 95 95 | 86.5 | ELISA | plasma | 11.5 |
Tjandra, JJ 1988 [15] | Australia | Case–control | 161 | I:32 II:24 III/ IV:53 | MSA + B2m | 2 MSA B2m | 93 88 39 | 90 95 90 | 90.3 | Radioimmunoassay + ELISA | serum | 12 |
Ławicki, S 2013 [16] | Poland | Case–control | 190 | I:25 II:35 III:25 IV:25(with metastases) | M-CSF + CA15–3 | 2 M-CSF CA 15–3 | 85 60 53 | 90 95 95 | 87.1 | ELISA | plasma | 11 |
Ławicki, Sławomir 2013 [17] | Poland | Case–control | 200 | I/ II/ III:75 IV:25(with metastases) | VEGF+ CA 15–3 M-CSF+ CA 15–3 VEGF+ M-CSF VEGF+ M-CSF+ CA 15–3 | 2 2 2 3 VEGF M-CSF CA 15–3 | 61 67 63 75 44 53 36 | 86 86 86 78 92 94 92 | 73.5 76.5 74.5 76.5 | ELISA | plasma | 11 |
Luo, M. 2019 [18] | China | Case–control | 200 | I/ II:47 III/ IV:73 | GATA3 + E-cadherin GATA3 E-cadherin | 2 | 90 87.5 82.5 | 91.7 73.3 87.5 | 90.6 | ELISA | serum | 11.5 |
Looi, Koksun 2006 [19] | China | Case–control | 123 | p16 + c-MYC + P53 | 3 | 43.9 | 97.6 | 79.6 | ELISA | serum | 8 | |
Guadagni, Fiorella 2001 [20] | Italy | Case–control | 2191 | I:392 II:562 III:153 IV:48 Metastatic:240 Local recurrence:58 | CEA + CA 15.3 | 2 CA 15.3 CEA | 39 33 16.7 | 85 | 54 | RIA kit | serum | 9.5 |
All the included and excluded studies are presented in Figure 2.
Association between diagnosis of primary breast cancer and blood tumor markers panels:
Panels I (MSA + B2m) and J (GATA3 + E-cadherin) had the highest sensitivity in primary breast cancer but did not have noticeable differences with mammography. Panels G (CA15–3 + CEA) and F (p16+ c-MYC+ P53) had the lowest sensitivity than the rest of the panels and mammography as mammography exhibited a remarkably better function than them, with OR = 0.13 and 95% CL (0.04–0.46) and OR = 0.15 and 95% CL (0.04–0.52) (Figure 3a, Table 2). In diagnostic tests, sensitivity had a vital role in screening diseases [21]. As a result, we can claim that the panels which had the highest sensitivity can be promising diagnostic tests in primary breast cancer screening, which included panels I and J in all stages of primary breast cancer. Panels L (MSA + CA15–3) and B (M-CSF + CA15–3) had the highest specificity but did not have remarkable differences with mammography. Panels G (CA15–3 + CEA) and D (VEGF + M-CSF + CA 15–3) had the lowest specificity as mammography demonstrated a superior function in specificity, with OR = 0.06 and 95% CL (0.01–0.39) and OR = 0.06 and 95% CL (0.02–0.19) (Figure 3b, Table 3). Mammography had a better function in specificity than a large number of panels, since it exhibited the highest specificity after panel L with OR = 2.54 and 95% CL (0.1–177.46) in diagnosing primary breast cancer. Panels J (GATA3 + E-cadherin) and I (MSA + B2m) possessed the highest accuracy in primary breast cancer but did not show significant differences with mammography. Panel L (MSA + CA15–3) did not demonstrate considerable differences with panel I; therefore, we could consider them approximately similar regarding accuracy. Panels G (CA15–3 + CEA) and F (p16+ c-MYC+ P53) possessed the lowest accuracy in primary breast cancer as mammography exhibited a considerably superior function in accuracy, with OR = 0.15 and 95% CL (0.07–0.3) and OR = 0.37 and 95% CL (0.17–0.74) (Figure 3c, Table 4).
B | C | D | E | F | G | I | J | L | M | |
---|---|---|---|---|---|---|---|---|---|---|
B | 1 | |||||||||
C | 1.18 (0.42–2.96) | 1 | ||||||||
D | 0.83 (0.28–2.29) | 0.71 (0.25–2) | 1 | |||||||
E | 1.73 (0.62–4.83) | 1.47 (0.56–4.24) | 2.09 (0.72–6.32) | 1 | ||||||
F | 1 | |||||||||
G | 1.12 (0.19–6.78) | 1 | ||||||||
I | 0.44 (0.05–3.44) | 0.37 (0.05–2.95) | 0.52 (0.06–4.22) | 0.25 (0.03–2.01) | 1 | |||||
J | 0.64 (0.09–4.78) | 0.54 (0.08–4.16) | 0.77 (0.1–6.26) | 0.37 (0.05–2.95) | 1.46 (0.11–22.19) | 1 | ||||
L | 1.13 (0.16–7.97) | 0.98 (0.14–6.79) | 1.37 (0.19–10.07) | 0.65 (0.09–4.78) | 2.6 (0.18–36.26) | 1.79 (0.13–22.86) | 1 | |||
M | 1.59 (0.69–3.53) | 1.35 (0.65–2.94) | 1.91 (0.78–4.88) | 0.91 (0.38–2.26) | 3.63 (0.55–25.11) | 2.47 (0.37–15.96) | 1.4 (0.24–8.3) | 1 |
B | C | D | E | F | G | I | J | L | M | |
---|---|---|---|---|---|---|---|---|---|---|
B | 1 | |||||||||
C | 1 | |||||||||
D | 1.46 (0.42–5.27) | 1 | ||||||||
E | 1.08 (0.31–3.99) | 0.75 (0.2–2.74) | 1 | |||||||
F | 2.8 (0.33–24.19) | 0.58 (0.07–4.87) | 0.4 (0.04–3.57) | 0.53 (0.06–4.86) | 1 | |||||
G | 6.82 (0.76–63.76) | 1.43 (0.16–12.82) | 1 (0.11–9.29) | 1.33 (0.14–12.7) | 2.62 (0.17–35.24) | 1 | ||||
I | 2.74 (0.16–61.57) | 0.57 (0.03–12.81) | 0.39 (0.02–8.77) | 0.53 (0.03–12.6) | 1 (0.04–30.25) | 0.39 (0.02–13.06) | 1 | |||
J | 2.4 (0.14–48.49) | 0.51 (0.03–9.19) | 0.34 (0.02–6.66) | 0.46 (0.03–9.64) | 0.85 (0.03–22.94) | 0.34 (0.01–9.86) | 0.87 (0.02–42.75) | 1 | ||
L | 0.16 (0.01–4.92) | 0.03 (0.01–1.04) | 0.06 (0.01–2.68) | 0.06 (0.01–3.68) | 0.07 (0.01–4.26) | 1 | ||||
M | 0.42 (0.14–1.2) | 0.16 (0.01–2.05) | 0.18 (0.01–2.3) | 2.54 (0.1–177.46) | 1 |
B | C | D | E | F | G | I | J | L | M | |
---|---|---|---|---|---|---|---|---|---|---|
B | 1 | |||||||||
C | 1.11 (0.65–1.9) | 1 | ||||||||
D | 1.24 (0.7–2.18) | 1.11 (0.65–1.97) | 1 | |||||||
E | 1.03 (0.59–1.81) | 0.92 (0.54–1.64) | 0.83 (0.46–1.46) | 1 | ||||||
F | 1.8 (0.77–4.42) | 1.61 (0.7–4) | 1.45 (0.61–3.77) | 1.78 (0.73–4.23) | 1 | |||||
G | 2.45 (0.85–6.76) | 1 | ||||||||
I | 0.54 (0.15–1.81) | 0.48 (0.14–1.62) | 0.44 (0.12–1.48) | 0.53 (0.15–1.84) | 0.3 (0.07–1.15) | 1 | ||||
J | 0.48 (0.13–1.74) | 0.43 (0.12–1.55) | 0.39 (0.1–1.43) | 0.46 (0.13–1.84) | 0.26 (0.06–1.1) | 0.91 (0.17–4.57) | 1 | |||
L | 0.58 (0.17–2.09) | 0.52 (0.16–1.92) | 0.47 (0.13–1.71) | 0.57 (0.16–2.1) | 0.32 (0.08–1.35) | 1.09 (0.22–5.35) | 1.22 (0.25–6.06) | 1 | ||
M | 0.66 (0.42–1.05) | 0.64 (0.38–1.06) | 1.22 (0.4–3.84) | 1.38 (0.42–4.41) | 1.1 (0.34–3.45) | 1 |
The best panels based on total function:
In diagnosing primary breast cancer, panels J and I exhibited the highest accuracy and total function compared to other panels. Overall, we recommend panel J because it had an even better function in accuracy than panel I, despite being minor (Table 1) and its study had a larger sample size (200). Panel J was made of GATA3 and E-cadherin. GATA3 is a transcription factor that plays a crucial role in the development and progression of breast cancer and can reverse the epithelial-mesenchymal transition. It also regulates the proliferation, differentiation, and development of cells. E-cadherin is a member of the cadherin family mainly expressed in epithelial cells. E-cadherin mediates the adhesion of allogeneic epithelial cells and plays a key role in epithelial cell aggregation and adhesion. Studies have demonstrated that the expression of cadherin is closely related to the invasion of breast cancer [18].
4. Conclusion
In conclusion, panel J including GATA3 + E-cadherin with a sensitivity of 90 and specificity of 91.7 demonstrated a higher diagnostic value for primary breast cancer than the rest of the panels as it exhibited higher function in accuracy than mammography, with OR = 1.38 and 95% CL (0.42–4.41), although it was not remarkable. After panel J, panel I (MSA + B2m) with a sensitivity of 90 and specificity of 90.3 and panel L (MSA + CA15–3) with a sensitivity of 84 and specificity of 100 had the best function in primary breast cancer detection than the rest of the panels. However, more experimental studies are required with larger samples, on different populations, and using other chemical measurement methods to verify these results.
Acknowledgments
This study was supported by Research Institute and Cancer Prevention Research Center, Isfahan University of Medical Sciences.
Author contributions
Vahid Raja had the idea for the research. The literature search was performed by Vahid Raja, Mohammad Sadegh Aboutalebi, and Rasool Nouri. The data analysis was performed by Marjan Mansourian and Khojaste Ghasemi. The article was drafted by Vahid Raja and Ziba Farajzadegan. The article was critically revised by Vahid Raja, Ziba Farajzadegan, and Fariborz Mokarian.
Data availability
The data used to support the findings of this study are available from the corresponding author upon request.
Ethical statement
Our study did not require an ethical board approval because it did not contain human or animal trials.
Supplementary material
Including traditional meta-analysis of all panels, nod-splitting analysis of inconsistency for sensitivity, specificity and accuracy, ranking of different panels in sensitivity, specificity and accuracy, the search strategy for each data base, and 54 studies were identified relevant to our research question.
References
- 1.
Raja V, Farajzadegan Z, Mansourian M, Ghasemi K, Aboutalebi MS, Nouri R, et al. Diagnostic value of nonacid nucleic blood tumor marker panels in early diagnosing breast cancer: A systematic review and network meta-analysis. Disease Markers. 2022; 2022 :15 - 2.
Sadler C, Goldfarb M. Comparison of primary and secondary breast cancers in adolescents and young adults. Cancer. 2015; 121 (8):1295-1302 - 3.
Zackrisson S, Cardoso F, Guidelines E. Clinical practice guidelines primary breast cancer: ESMO clinical practice Guidelines for diagnosis, treatment and follow-up† clinical practice guidelines. Annals of Oncology. 2015; 26 (5):8-30 - 4.
Moher D, Liberati A, Tetzlaff J, Altman DG, Group* P. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Annals of internal medicine. 2009; 151 (4):264-269 - 5.
Pepe MS, Thompson ML. Combining diagnostic test results to increase accuracy. Biostatistics. 2000; 1 (2):123-140 - 6.
Yankaskas BC, Haneuse S, Kapp JM, Kerlikowske K, Geller B, Buist DS, et al. Performance of first mammography examination in women younger than 40 years. JNCI: Journal of the National Cancer Institute. 2010; 102 (10):692-701 - 7.
Sinclair N, Littenberg B, Geller B, Muss H. Accuracy of screening mammography in older women. American Journal of Roentgenology. 2011; 197 (5):1268-1273 - 8.
Ontario HQ. Screening mammography for women aged 40 to 49 years at average risk for breast cancer: An evidence-based analysis. Ont Health Technol Assess Ser. 2007; 7 (1):1-32 - 9.
Zajkowska M, Głażewska EK, Będkowska GE, Chorąży P, Szmitkowski M, Ławicki S. Diagnostic power of vascular endothelial growth factor and macrophage colony-stimulating factor in breast cancer patients based on ROC analysis. Mediators of Inflammation. 2016; 2016 :8 - 10.
Sacks N, Stacker S, Thompson C, Collins J, Russell I, Sullivan J, et al. Comparison of mammary serum antigen (MSA) and CA15-3 levels in the serum of patients with breast cancer. British journal of cancer. 1987; 56 (6):820-824 - 11.
Liu Y, Liao Y, Xiang L, Jiang K, Li S, Huangfu M, et al. A panel of autoantibodies as potential early diagnostic serum biomarkers in patients with breast cancer. International Journal of Clinical Oncology. 2017; 22 (2):291-296 - 12.
Molina R, Jo J, Filella X, Zanon G, Pahisa J, Muñoz M, et al. c-erbB-2 oncoprotein, CEA, and CA 15.3 in patients with breast cancer: Prognostic value. Breast cancer research and treatment. 1998; 51 (2):109-119 - 13.
Ławicki S, Zajkowska M, Głażewska EK, Będkowska GE, Szmitkowski M. Plasma levels and diagnostic utility of VEGF, MMP-9, and TIMP-1 in the diagnosis of patients with breast cancer. Onco Targets and therapy. 2016; 9 :911 - 14.
Ławicki S, Zajkowska M, Głażewska EK, Będkowska GE, Szmitkowski M. Plasma levels and diagnostic utility of VEGF, MMP-2 and TIMP-2 in the diagnostics of breast cancer patients. Biomarkers. 2017; 22 (2):157-164 - 15.
Tjandra J, McLaughlin P, Russell I, Collins J, McKenzie I. Comparison of mammary serum antigen (MSA) with β2-microglobulin (β2M) and carcinoembryonic antigen (CEA) assays in patients with breast cancer. European Journal of Cancer and Clinical Oncology. 1988; 24 (10):1633-1640 - 16.
Ławicki S, Będkowska G, Wojtukiewicz M, Szmitkowski M. Hematopoietic cytokines as tumor markers in breast malignancies. A multivariate analysis with ROC curve in breast cancer patients. Advances in Medical Sciences. 2013; 58 (2):207-215 - 17.
Ławicki S, Będkowska GE, Szmitkowski M. VEGF, M-CSF and CA 15-3 as a new tumor marker panel in breast malignancies: A multivariate analysis with ROC curve. Growth Factors. 2013; 31 (3):98-105 - 18.
Luo M, Huang Y, Huang J, Huang S, Wei L, Zhang Y, et al. Evaluation of the value of GATA3 combined with E-cadherin in the diagnosis of breast cancer. Journal of BU ON: Official Journal of the Balkan Union of Oncology. 2019; 24 (3):1038-1044 - 19.
Looi K, Megliorino R, Shi F-D, Peng X-X, Chen Y, Zhang J-Y. Humoral immune response to p16, a cyclin-dependent kinase inhibitor in human malignancies. Oncology Reports. 2006; 16 (5):1105-1110 - 20.
Guadagni F, Ferroni P, Carlini S, Mariotti S, Spila A, Aloe S, et al. A re-evaluation of carcinoembryonic antigen (CEA) as a serum marker for breast cancer: A prospective longitudinal study. Clinical Cancer Research. 2001; 7 (8):2357-2362 - 21.
Stellman SD. Book review: Epidemiology. In: Gordis L, editor. Gordis Epidemiology, 6th ed. Philadelphia, PA, US: Saunders; 2009. p. 2010. ISBN: 978-0-323-55229-5. Copyright © 2019 by Elsevier, Inc. All rights reserved