BACKGROUND Exposure to polychlorinated biphenyls (PCBs) has been shown to influence expression of some biomarkers that are predictive/prognostic for breast cancer. Therefore, our study was conducted to further investigating associations of different functional PCBs in adipose tissue with breast cancer prognostic biomarkers. METHODS Two hundred and five breast cancer patients were recruited in Shantou, China. Breast adipose tissues were collected during their resection surgery and levels of 7 PCB congeners were analyzed by gas chromatography-mass spectrometry (GC-MS). The PCB congeners were divided into 4 groups according to structure-activity. Socio-demographic, clinical and pathological information were obtained from questionnaire and digital medical records. Odds ratios (ORs) for associations between prognostic biomarkers and PCB levels (tertile 3 [T3], tertile 2 [T2] vs. tertile 1) were estimated from logistic regression models. RESULTS Most PCB congeners were detectable, with a highest level (22.06 ng/g lipid) of PCB153. As for estrogenic PCBs, increased PCB52 exposure was positively associated with PR expression (ORT2 = 2.36, Ptrend = 0.054), but higher PCB101 level was negatively associated with HER-2 (ORT3 = 0.24, Ptrend = 0.029) and tumor size (OR = 0.43). Limited dioxin-like PCB138 exposure was positively associated with ER (ORT2 = 3.23, ORT3 = 3.77, Ptrend = 0.047) but negatively with Top-IIα expression (ORT2 = 0.35, ORT3 = 0.28, Ptrend = 0.080). Higher PCB153 (CYP inducer) level was negatively associated with ER (ORT2 = 0.32, ORT3 = 0.19, Ptrend = 0.038) but positively with Ki-67 expression (ORT2 = 1.43, ORT3 = 3.60, Ptrend = 0.055). Higher neurotoxic PCB28 was positively associated with HER-2 (ORT3 = 5.43, Ptrend = 0.006) and tumor size (OR = 2.37). Moreover, total PCBs exposure was positively associated with VEGF-C (ORT2 = 76.91, ORT3 = 97.96, Ptrend = 0.041) and tumor metastasis (OR = 2.25). CONCLUSIONS Different functional PCB congeners have different associations (both positive and negative) with breast cancer prognostic biomarkers, as well as tumor classification stage. Therefore, the development and aggressiveness of breast cancer may depend upon exposure to specific structure-activity of PCBs.

中文翻译：

---

脂肪组织中功能性多氯联苯与乳腺癌患者的预后生物标志物之间的关联。

背景技术已经表明，暴露于多氯联苯（PCBs）会影响一些可预测/预后乳腺癌的生物标志物的表达。因此，我们进行了研究，以进一步调查脂肪组织中不同功能PCB与乳腺癌预后生物标志物的关联。方法招募了汕头市的255例乳腺癌患者。在切除手术期间收集乳房脂肪组织，并通过气相色谱-质谱法（GC-MS）分析7种多氯联苯同源物的水平。根据结构活性将PCB同类物分为4组。社会人口统计学，临床和病理学信息来自问卷和数字病历。根据逻辑回归模型估算了预后生物标志物与PCB水平之间关联的几率（OR）（三分位数3 [T3]，三分位数2 [T2]与三分位数1）。结果大多数PCB同源物都是可检测到的，PCB153的含量最高（22.06 ng / g脂质）。至于雌激素PCBs，增加的PCB52暴露与PR表达呈正相关（ORT2 = 2.36，Ptrend = 0.054），但较高的PCB101含量与HER-2（ORT3 = 0.24，Ptrend = 0.029）和肿瘤大小（OR = 0.43）。有限的二恶英样PCB138暴露与ER正相关（ORT2 = 3.23，ORT3 = 3.77，Ptrend = 0.047），但与Top-IIα表达呈负相关（ORT2 = 0.35，ORT3 = 0.28，Ptrend = 0.080）。较高的PCB153（CYP诱导剂）水平与ER呈负相关（ORT2 = 0.32，ORT3 = 0.19，Ptrend = 0。038），但与Ki-67表达呈正相关（ORT2 = 1.43，ORT3 = 3.60，Ptrend = 0.055）。较高的神经毒性PCB28与HER-2（ORT3 = 5.43，Ptrend = 0.006）和肿瘤大小（OR = 2.37）呈正相关。此外，PCBs暴露总量与VEGF-C呈正相关（ORT2 = 76.91，ORT3 = 97.96，Ptrend = 0.041）和肿瘤转移（OR = 2.25）。结论不同的功能性PCB同系物与乳腺癌的预后生物标志物以及肿瘤分类阶段具有不同的关联（阳性和阴性）。因此，乳腺癌的发展和侵袭性可能取决于暴露于PCBs的特定结构活性。多氯联苯的总暴露与VEGF-C（ORT2 = 76.91，ORT3 = 97.96，Ptrend = 0.041）和肿瘤转移（OR = 2.25）呈正相关。结论不同的功能性PCB同系物与乳腺癌的预后生物标志物以及肿瘤分类阶段具有不同的关联（阳性和阴性）。因此，乳腺癌的发展和侵袭性可能取决于暴露于PCBs的特定结构活性。多氯联苯的总暴露与VEGF-C（ORT2 = 76.91，ORT3 = 97.96，Ptrend = 0.041）和肿瘤转移（OR = 2.25）呈正相关。结论不同的功能性PCB同系物与乳腺癌的预后生物标志物以及肿瘤分类阶段具有不同的关联（阳性和阴性）。因此，乳腺癌的发展和侵袭性可能取决于暴露于PCBs的特定结构活性。