Page 113 - GPD-4-1

Gene & Protein in Disease Prognostic role of SIRT1 expression in cancer

Figure 2. Forest plot illustrating the individual effect of each study. Among the 15 studies, Shuang et al. (P = 0.43) show a notably distinct effect with a
11
higher P-value compared to others. The meta-analysis, as a whole, indicates the overall significance of the collective findings.

Table 2. Meta‑regression analysis across various moderators selected moderators do not significantly impact the overall
to further confirm the absence of heterogeneity consistency of the meta-analysis results, thereby providing
confidence in the robustness of the observed prognostic
Moderators Test of residual Residual heterogeneity
homogeneity value of SIRT1 expression.
Q‑stat df Sig. Tau 2 I 2 H 2 R 2 3.3. Subgroup analyses
Country/location 1.225 13 0.999 0 0 1 0 Further subgroup analyses revealed that the prognostic
Sample size 1.637 13 0.999 0 0 1 0 value of SIRT1 in cancer was significant (Table 3) across
Cancer type 1.836 13 0.999 0 0 1 0 different geographical regions, with consistent effect sizes
Age groups 1.756 13 0.999 0 0 1 0 in both Asian (overall effect = 1.32, 95% CI [0.608 – 2.029],
Sex 1.852 13 0.999 0 0 1 0 P < 0.001) and non-Asian (overall effect = 1.82, 95%
Tumor stage 1.547 13 0.999 0 0 1 0 CI [0.8 – 2.839], P < 0.001) populations. Specifically,
SIRT1 expression level 1.475 13 0.999 0 0 1 0 the subgroup analysis for China (overall effect = 1.43,
Abbreviations: df: Degrees of freedom; Sig: Significance; SIRT1: Sirtuin 95% CI [0.415 – 2.441]) indicated a significant effect,
1; stat: Statistics. although with slightly higher heterogeneity (P = 0.006)
compared to the overall Asia and non-Asia subgroups.
expression (Figure 3F). However, the results consistently The prognostic value of SIRT1 appeared significant across
demonstrated no statistically significant heterogeneity different study sizes. Small, medium, and large studies
for any of these moderators (P ≈ 1.000), indicating all showed statistically significant results (P = 0.001,
homogeneity. In each case, the Q-statistics showed no P = 0.002, P = 0.021, respectively), with minimal to no
2
2
2
heterogeneity (P = 1.000). observed heterogeneity (I = 0.0001, I = 0.0001, I = 0.000,
respectively). These consistent effect sizes across various
The absence of heterogeneity across these moderators
implies that the prognostic value of SIRT1 remains study sizes indicate robust findings.
stable and unaffected by the specific countries, sample Studies on ovarian cancer (n = 3) showed a statistically
sizes, cancer types, age groups, tumor stages, or SIRT1 significant effect (P = 0.041) and a moderate level of
expression levels considered. The findings suggest that the observed heterogeneity (I = 0.04). Similarly, studies on
2

Volume 4 Issue 1 (2025) 5 doi: 10.36922/gpd.4294