An EGFR and AKT Signaling Pathway was Identified with Mediation Model in Osteosarcomas Clinical Study - PubMed (original) (raw)

An EGFR and AKT Signaling Pathway was Identified with Mediation Model in Osteosarcomas Clinical Study

Huiyun Wu et al. Biomark Insights. 2007.

Abstract

Identification of correlation pattern and signal pathway among biomarkers in patients has become increasingly interesting for its potential values in diagnosis, treatment and prognosis. EGFR and p-AKT signaling in osteosarcoma (OS) patients were analyzed for its relationship with cancer cell proliferation maker, Ki-67, using causal procedures and statistical tests. A total of 69 patients were collected who present to Vanderbilt University Medical Center with newly diagnosed, previously untreated osteosarcomas during the clinical study period 1994 through 2003. Tissue microarrays were constructed for EGFR, p-AKT and Ki-67. The mediation model was constructed with structural equation model (SEM) for the causal analysis of the three biomarkers in osteosarcoma patients. The results suggested a mediating effect of p-AKT for the causal relationship between EGFR and Ki-67. The study also found significant associations between EGFR and Ki-67 (p = 0.002), EGFR and p-AKT (p = 0.027), and p-AKT and Ki-67 controlling EGFR (p = 0.004). After the impact of EGFR on Ki-67 was accounted for by p-AKT, the relation between EGFR and Ki-67 was no longer significant (p = 0.381). The mediating effect was confirmed with Sobel test (p < 0.001) and Goodman (I) test (p < 0.001). The study indicated that a mediation model could be an approach to exploring the correlation pattern of EGFR and AKT signal pathway for cancer cell proliferation in OS patients in clinical study.

Keywords: AKT; EGFR; clinical study; mediation model; osteosarcomas.

PubMed Disclaimer

Figures

Figure 1

Figure 1

Path for Model 1. A two-variable correlation model between an explanatory variable and an outcome variable. τ is the parameter of interest. ɛ is the error term.

Figure 2

Figure 2

Paths for Model 2 and Model 3. A three-variable mediation model. X, M, and Y are symbols of the three variables. α_,_ β and τ′ are the parameters of interest. ɛ is the error term.

Figure 3

Figure 3

Structural equation modeling for mediation effect. Dependent variable Ki-67 is predicted with single-headed EGFR and p-AKT. Each single-headed arrow represents a regression weights. Age and gender are considered to be associated with all the three biomarkers and are also allowed to be correlated and represented with double-headed arrow. Circles of e1, e2 and d1 are error terms with mean at 0 and variance at 1.

Figure 4

Figure 4

Expression of EGFR and p-AKT. Left panel: membrane staining in tumor cells with EGFR; right panel: nuclear and cytoplasmic staining in tumor cells with p-AKT.

Figure 5

Figure 5

Scatter plots with lowess lines. Circles are data points and lines are the lowess smoothing lines.

Similar articles

Cited by

References

    1. Anninga JK, van de Vijver MJ, Cleton-Jansen AM, Kristel PM, Taminiau AH, Nooij M, Egeler RM, Hogendoorn PC. Overexpression of the HER-2 oncogene does not play a role in high-grade osteosarcomas. Eur J Cancer. 2004;40:963–70. - PubMed
    1. Arteaga CL. Overview of epidermal growth factor receptor biology and its role as a therapeutic target in human neoplasia. Semin Oncol. 2002;29:3–9. - PubMed
    1. Beck SL. Symptom clusters: impediments and suggestions for solutions. J Natl Cancer Inst Monogr. 2004:137–8. - PubMed
    1. Beck SL, Dudley WN, Barsevick A. Pain, sleep disturbance, and fatigue in patients with cancer: using a mediation model to test a symptom cluster. Oncol Nurs Forum. 2005;32:542. - PubMed
    1. Curtis RK, Oresic M, Vidal-Puig A. Pathways to the analysis of microarray data. Trends Biotechnol. 2005;23:429–35. - PubMed

LinkOut - more resources