Table 9.
Studies on the Correlations between Tumour HLA-G Expression and Clinicopathological Factors and Clinical Outcome of Ovarian Carcinoma Patients.
| First Author [Ref.] | mAb and Included Patient Cohort | HLA-G Quantification Method | HLA-G+ Samples (%) | Association with Tumour HLA-G Expression | |
|---|---|---|---|---|---|
| Clinico-Pathological Parameters with p-Values ≤ 0.05 | Clinical Outcome Associated with HLA-G (p-Value) | ||||
| * Rutten [53] |
4H84 Type II, high grade ovarian carcinoma patients |
Representation of percentage and intensity; Normal expression (<3) vs. upregulated expression (>3) |
81/152 (53) | More residual tumour after debulking surgery; Increased platinum sensitivity. |
KM analysis: Prolonged OS (0.001); Prolonged CSS (0.008); Prolonged PFS (0.036). Cox univariate analysis: Prolonged CSS, HR = 1.69 (0.009) Cox multivariate analysis: Prolonged CSS, HR = 1.62 (0.020) KM analysis, in patients with tissue collected prior to chemotherapy, n = 108: Prolonged CSS (0.011); Prolonged PFS (0.027). KM analysis, in patients with downregulated HLA-A, n = 137: CSS (ns) |
|
† Babay [51] |
4H84 Ovarian carcinoma patients, random cohort |
<1% staining (0) vs. 1–5% (1), 6–25% (2), 26–50% (3) or >50% (4) staining | Percentage of HLA-G-positive samples within specific staining groups was unattainable Cohort with available follow-up: 36/51 (71) |
No significant correlations | KM analysis: OS (ns) Multivariate binomial logistic regression analysis: Recurrence, HR = 4.115 (ns) |
| Jung [52] |
4H84 Ovarian carcinoma patients, random cohort |
Mild (0–25%; 1+), moderate (25–50%; 2+) and strong (>50%; 3+) staining Optimal cut-off value was determined by ROC curve analysis at 17% for HLA-G detection |
1+: 22/40 (55) 2+: 8/40 (20) 3+: 10/40 (25) > 17%: 24/40 (60) |
Advanced disease stage | KM analysis: Shorter OS (0.04); PFS (ns). Cox univariate analysis: Shorter OS, HR = 3.00 (0.04) Cox multivariate analysis: OS (ns) KM analysis, in patients with specific cancer stages: OS (ns) |
|
‡ 4H84 Ovarian carcinoma patients, random cohort |
Optimal cut-off value was determined by ROC curve analysis at DV 1.14 for HLA-G detection | DV >1.14: 18/40 (45) |
Advanced disease stage; CA125 over-expression. |
KM analysis: OS (ns); PFS (ns). Cox univariate analysis: OS, HR = 1.48 (ns) |
|
| Andersson [50] |
MEM-G/1 Advanced stage III/IV, serous ovarian adeno-carcinoma patients |
No staining (0) vs. 1–25% (1), 26–50% (2), 51–75% (3) or >75% (4) staining | Percentage of HLA-G-positive samples within specific staining groups was unattainable Total cohort: 14/72 (20) |
Absence of infiltrating Tregs and CD8+ T cells | KM analysis: OS (ns) KM analysis, in patients with HLA-A*02, without CD8+ cells and with HLA-G expression vs. patients with HLA-A otherwise, presence of CD8+ cells and without HLA-G expression, n = 42: Shorter OS (0.006) |
| Zhang [54] |
5A6G7 Ovarian carcinoma patients, random cohort |
Any staining >5% was considered as positive | Cohort with available follow-up: 14/17 (82) |
No significant correlations | KM analysis: OS (ns) Cox univariate analysis: OS, HR = 0.58 (ns) Cox multivariate analysis: OS, HR = 0.48 (ns) |
The technique used for HLA-G detection was immunohistochemistry with formalin-fixed paraffin-embedded tissue unless indicated otherwise. p-Values ≤ 0.05 were considered statistically significant. Abbreviations; CSS, Cancer-Specific Survival; DFS, Disease-Free Survival; DV, Densitometer Value; HR, Hazard Ratio; KM, Kaplan–Meier; n/a, not applicable; ns, not significant; OS, Overall Survival; PFS, Progression-Free Survival; RFP, Recurrence-Free Period. * Tissue microarrays were used to determine the percentage of HLA-G expression in the tumour samples. † The follow-up data of a limited amount of patients was available for the survival analyses. ‡ Western Blot analysis was used to determine the percentage of HLA-G expression in the tumour samples.