Diabetes Mellitus Promotes Lymph Node Metastasis in Gastric Cancer: A 15-Year Single-Institution Experience
Gastric cancer (GC) remains a leading cause of cancer-related mortality worldwide, with lymph node (LN) metastasis serving as a critical determinant of prognosis. Emerging evidence suggests diabetes mellitus (DM) may influence cancer progression, but its role in LN metastasis remains unclear. This study comprehensively evaluates the impact of DM on LN metastasis patterns in GC patients, leveraging a 15-year dataset from a high-volume institution to elucidate clinicopathological correlations and mechanistic insights.
Study Design and Patient Selection
The study analyzed 3,213 consecutive GC patients who underwent surgery at Nanfang Hospital, Southern Medical University (Guangdong, China) between October 2004 and December 2019. After excluding patients with <15 examined LNs (ELNs), preoperative therapy, or incomplete data, 2,142 patients were included. These were stratified into DM (n=149) and non-DM (n=1,993) cohorts based on diagnostic criteria (fasting glucose ≥7.0 mmol/L, oral glucose tolerance test ≥11.1 mmol/L, or documented DM history). LN metastasis was assessed using two staging systems:
- JCGC N-classification (13th edition): Focused on anatomical spread, defining N3 stations as distant nodal groups (e.g., stations 10–14).
- UICC-TNM N-staging (8th edition): Categorized metastasis burden by MLN count, with N3b indicating >15 metastatic LNs.
Key variables included tumor depth (T-stage), ELN count, lymphatic/venous invasion, and tumor size.
Clinicopathological Characteristics
DM patients were older (mean age: 63.8 vs. 55.9 years, P<0.001) and predominantly male (77.2% vs. 65.4%, P=0.003). Tumor location differed significantly (P=0.016), with DM patients having more proximal tumors (32.2% upper third vs. 21.3% in non-DM). DM cohorts exhibited higher rates of lymphatic (34.9% vs. 25.7%, P=0.014) and venous invasion (32.9% vs. 25.5%, P=0.047). However, ELN counts (median: 40 vs. 39, P=0.735) and primary tumor depth (T-stage distribution, P=0.895) were comparable between groups, minimizing detection bias.
LN Metastasis Patterns in DM vs. Non-DM Cohorts
1. N3 Stations Metastasis (Anatomical Spread):
DM patients had significantly higher N3 involvement (26.8% vs. 19.3%, P=0.026). Multivariate analysis confirmed DM as an independent risk factor (OR=1.771, 95% CI:1.139–2.755, P=0.011), alongside advanced T-stage, lymphatic invasion, and lower-third tumor location.
2. N3b Status (Metastatic Burden):
N3b prevalence was elevated in DM patients (18.8% vs. 12.8%, P=0.039). DM independently predicted N3b (OR=1.752, 95% CI:1.061–2.893, P=0.028), with additional risks from T4 staging, venous invasion, and tumor size ≥5 cm.
3. MLN Count:
DM patients had higher median MLNs (2 vs. 2, P=0.047 adjusted for covariates). Linear regression identified DM as an independent predictor of increased MLN count (β=1.424, P=0.047), alongside ELN ≥45, advanced T-stage, and lymphatic invasion.
Subgroup Analyses
A. Stratified by Primary Tumor Depth (T-Stage):
LN metastasis escalated with advancing T-stage in DM patients:
- N3 Stations: T1–2: 2.2% (DM) vs. 4.9% (non-DM); T4b: 50.0% vs. 36.6%.
- N3b Status: T3: 16.1% vs. 5.1%; T4b: 44.0% vs. 28.0%.
- MLN Count: T3: 8.6 vs. 5.2; T4b: 17.0 vs. 12.8.
In T3–T4 subgroups, DM patients exhibited higher N3 stations metastasis (37.9% vs. 26.5%, P=0.012), N3b prevalence (27.2% vs. 18.4%, P=0.028), and MLN counts (median:7 vs.5, P=0.066).
B. Stratified by ELN Count:
Higher ELN counts improved metastasis detection, accentuating DM-associated risks:
- N3 Stations: ELN ≥45:37.7% (DM) vs.25.3% (non-DM).
- N3b Status: ELN ≥45:26.4% vs.17.3%.
- MLN Count: ELN ≥45:12.0 vs.7.7.
For ELN ≥45, DM independently predicted N3 stations (OR=2.115, P=0.033) and higher MLN counts (β=3.529, P=0.021).
Mechanistic and Clinical Implications
The study proposes DM exacerbates LN metastasis through multifactorial pathways:
- Hyperglycemia: Fuels tumor glycolysis, promoting proliferation and immune evasion.
- Hyperinsulinemia: Activates IGF-1 pathways, enhancing tumor invasiveness.
- Chronic Inflammation: Adipokines and cytokines (e.g., IL-6, TNF-α) foster pro-metastatic microenvironments.
- Microbiome Dysbiosis: Altered gut flora may modulate immune responses and metastatic potential.
Clinically, DM patients with advanced T-stage or high ELN counts warrant intensified nodal assessment and adjuvant therapy. The findings advocate personalized strategies, such as extended lymphadenectomy or tailored chemotherapy regimens, for DM-associated GC.
Limitations
- Retrospective Design: Potential selection bias despite rigorous adjustments.
- ELN Variability: While ELN ≥15 minimized staging inaccuracies, optimal ELN thresholds remain debated.
- Metastatic LN Size: Data on metastatic LN size, a potential prognostic factor, were unavailable.
Conclusion
This large-scale analysis establishes DM as an independent risk factor for LN metastasis in GC, particularly in advanced T-stages and high ELN cohorts. The findings underscore the need for vigilant nodal evaluation and tailored therapeutic approaches in DM patients, while providing a foundation for investigating molecular mechanisms linking metabolic dysregulation to GC progression.
doi.org/10.1097/CM9.0000000000001795
Was this helpful?
0 / 0