Table 2.
Univariate analysis of clinicopathological baseline data and IHC markers derived from breast cancer liver metastases and their effect on overall survival (OS after occurrence of liver metastasis)
| Parameter | Classification | Distribution | Impact on survival Hazard ratio [95 % CI] |
p value (logrank) |
|---|---|---|---|---|
| Type of surgery | Resection (%) | 20.58 % (7/34) | Punch: 3.36 95 % CI [1.14–9.88] |
0.0207 |
| Punch (%) | 79.41 % (27/34) | |||
| Other distant metastases present at liver metastasis | Yes (%) | 44.12 % (15/34) | Present at diagnosis: 1.70 95 % CI [0.80–3.65] |
0.1662 |
| No (%) | 55.88 % (19/34) | |||
| Chemotherapy (CTx) after diagnosis of liver metastasis | Yes (%) | 86.21 % (25/29*) | Yes: 2.21 95 % CI [0.65–7.50] |
0.1911 |
| No (%) | 13.79 % (19/29*) | |||
| Estrogen receptor (ER) | Positive (%) | 55.88 % (19/34) | ER positive: 0.87 95 % CI [0.42–1.80] |
0.7003 |
| Negative (%) | 44.12 % (15/34) | |||
| Progesterone receptor (PGR) | Positive (%) | 32.35 % (11/34) | PR positive: 0.93 95 % CI [0.43–1.99] |
0.8428 |
| Negative (%) | 67.65 % (23/34) | |||
| Her2/neu | Positive (%) | 26.47 % (9/34) | HER2/neu positiv: 1.06 95 % CI [0.45–2.50] |
0.8997 |
| Negative (%) | 73.53 % (25/34) | |||
| Proliferation index Ki67 | Positive (%) | 91.18 % (31/34) | Ki67 positive: 2.79 95 % CI [0.65–11.92] |
0.1481 |
| Negative (%) | 8.82 % (3/34) | |||
| Modified proliferation index Ki67 | Positive (%) | 52.94 % (18/34) | Ki67 > 30 % positive: 2.46 95 % CI [1.11–5.44] |
0.0222 |
| Negative (%) | 47.06 % (16/34) | |||
| β-catenin dependent WNT score | High > 1 (%) | 44.12 % (15/34) | Wnt-Score high: 0.74 95 % CI [0.35–1.55] |
0.4179 |
| Low ≤ 1 (%) | 55.88 % (19/34) | |||
| β-catenin independent WNT score |
High > 5(%) | 64.71 % (22/34) | Wnt-Score high: 2.19 95 % CI [1.02–4.69] |
0.0391 |
| Low ≤ 5 (%) | 35.29 % (12/34) |
Patient cohort was characterized according to listed parameters in the first column. Type of classification and distribution within the cohort as well as impact on survival including p value (logrank) is given for each parameter
Bold and underlined p-values are meant to highlight those below 0.05
* Cases where not for all patients baseline data was available