研究背景及目的

乳腺癌已成为女性群体中最常见的恶性肿瘤之一，且已超越肺癌成为发病率最高的恶性肿瘤。在全世界范围内70岁及以上的老年女性患乳腺癌的比例也在显著上升。老年患者的肿瘤生物学行为、对治疗的耐受性、共病状况等均与年轻患者相比存在显著差异。

年龄和合并症情况是预后的重要因素，先前的研究结果表明，合并症情况和确诊年龄均与乳腺癌患者的预后显著相关。查尔森共病指数（the Charlson Comorbidity Index，CCI）是在1987年由Charlson等人首次提出，在1994年他们根据研究结果再次提出了年龄调整后的查尔森共病指数（the Age-adjusted Charlson Comorbidity Index，ACCI）。ACCI目前已在多种类型恶性肿瘤中显示出对预后较好的预测价值。但目前尚无针对中国老年性乳腺癌患者共病与预后的相关性研究，而且ACCI在乳腺癌人群患者中对预后方面的预测价值也尚未阐明。

2型糖尿病（the type 2 diabetes，T2D）在老年乳腺癌患者群体中的发病率较高，流行病学统计结果显示大约16%的老年性乳腺癌患者同时伴有2型糖尿病。2型糖尿病和乳腺癌预后的相关性研究一直是热点话题，多数学者认为2型糖尿病是乳腺癌发病和预后的独立危险因素，然而，迄今为止还没有探讨2型糖尿病对老年性乳腺癌患者预后影响的独立研究。

三阴性乳腺癌（triple negative breast cancer，TNBC）是一种特殊的乳腺癌分子分型，约占所有乳腺癌患者的13%，与激素受体阳性的乳腺癌相比，TNBC有更具侵袭性的肿瘤特征、更高的复发转移率和更差的临床预后。本研究小组先前通过对已在线发表的TNBC转录组微阵列数据进行生信分析发现Syndecan-1（Sdc1，也称作CD138）在TNBC中被认定为不良预后基因。Sdc1是Syndecan家族的重要成员，其表达情况与多种类型肿瘤的预后和治疗反应相关。在乳腺癌中Sdc1的高表达与侵袭性表型和不良临床行为相关，被看做是一种乳腺癌预后相关的生物标志物。

我们旨在开展以ACCI评分作为自变量评估合并症对中国老年性乳腺癌患者预后的影响的临床研究，并就2型糖尿病这一单病种对老年患者预后的影响做了进一步回顾性分析，同时还调查了所有伴有2型糖尿病患者的空腹血糖水平以明确不同血糖控制情况对乳腺癌患者预后的确切影响，最后我们拟通过检测患者肿瘤组织石蜡病理切片中Sdc1在肿瘤上皮细胞的表达情况探讨其在三阴性老年性乳腺癌中的预后价值。

材料及方法

根据中国老年性乳腺癌治疗共识，我们将年龄≥70岁的患者视为老年乳腺癌患者。本系列研究纳入的均为在我院乳腺外科确诊为乳腺癌并接受手术治疗的年龄≥70岁的女性患者，同时或异时（首次确诊年龄＜70岁）的乳腺癌和诊断时已有远处转移的患者均被排除在外。根据患者的死亡原因，我们分别分析了总生存期（OS）、乳腺癌特异性生存期（BCSS）和非乳腺癌特异性生存期（NBCSS）。ACCI评分是从70岁开始，年龄每增加10岁在原本的CCI评分上加1（70-79：+1，80-89：+2，≥90：+3）。我们选择了最具代表性的患者肿瘤组织蜡块样本进行重新切片及免疫组化染色，对于syndecan-1的免疫组织化学染色，我们使用了小鼠的重组单克隆抗体（ab128936，ABclonal）。肿瘤组织上皮细胞syndecan-1的表达水平一般分为四类：阴性（＜10%的细胞为阳性）、弱阳性（10-20%的细胞为阳性）、中度阳性（21-60%的细胞为阳性）和强阳性（＞60%的细胞为阳性）。考虑到本研究的样本量，我们将不同强度的阳性样本统一归类为阳性。我们通过卡方检验分析目标变量与各项临床病理学特征之间的关联。通过对数秩检验结果的 Kaplan-Meier 曲线进行单因素生存分析。我们还使用Cox比例风险模型来检验目标变量是否是不同生存结局的独立预测因素。

研究结果

第一部分研究结果显示，该研究队列中超过四分之三的患者伴有一种或多种合并症，但绝大多数患者的ACCI评分≤3。除激素受体状态以外，其他所有因素均与ACCI评分高低无显著相关性。高分组激素受体阳性患者比例较高（88.1% vs 76.2%，p=0.037）。局部复发或远处转移、全因死亡及乳腺癌特异性死亡的发生率在两组间均无明显差异（p＞0.05），但ACCI＞3组的非乳腺癌特异性死亡的发生率明显高于ACCI≤3组（19.4% vs 10.7%，p=0.032）。ACCI评分是全因死亡（HR=0.42，95% CI: 0.22-0.83，p=0.012）和非乳腺癌特异性死亡（HR=0.41，95% CI: 0.20-0.87，p=0.020）的独立影响因素，低分组患者的全因死亡和非乳腺癌特异性死亡率均优于高分组。生存分析显示ACCI评分是OS（HR=2.18，95% CI:1.22-3.92，p=0.009)和NBCSS (HR=2.04，95% CI:1.02-4.08，p=0.044)的独立预测因子。

第二部分研究结果显示，该老年队列中有将近25%的患者同时伴有2型糖尿病（n = 163，24.8%）。在这些患者中，36.8 %的FBG水平＜6.1 mmol/L，28.8 %的FBG水平≥6.1且＜7.0 mmol/L，而34.4%的FBG水平≥7.0 mmol/L。T2D组患者与非T2D组患者在激素受体状态和ACCI评分方面存在显著差异（p＜0.05）。在生存分析中，虽然两组患者的DFS（p = 0.795）、OS（p = 0.738）、BCSS（p = 0.639）以及NBCSS（p = 0.424） 等方面均没有显著差异，但在部分亚组分析中观察到了具有统计学差异的结果，如激素受体阴性、TNBC分型和未接受内分泌治疗的患者群体。在多变量Cox回归分析中2型糖尿病是OS（HR = 2.572，95 % CI：1.034-6.397，p = 0.042）和NBCSS（HR = 2.706，95 % CI：1.157-6.329，p = 0.022）的独立危险因素。关于空腹血糖水平部分的结果显示血糖控制不良并未显著影响DFS，但会显著影响OS、BCSS和NBCSS。

第三部分研究结果显示，该研究队列共包括81名老年三阴性乳腺癌患者，其中41名患者的肿瘤组织上皮细胞的Sdc1表达为不同程度的阳性，而另外40名患者的肿瘤组织上皮细胞的Sdc1表达为阴性。Kaplan-Meier生存曲线及相应的log-rank结果提示两组患者在DFS（p = 0.017）和DMFS（p = 0.002）有显著差异，而在OS（p = 0.375）、LRFS（p = 0.282）、BCSS（p = 0.143）以及NBCSS（p = 0.365） 等结局方面均没有显著差异。在多变量Cox回归分析中Sdc1阳性表达是DFS（HR = 1.85，95 % CI：1.16-2.97，p = 0.021）和DMFS（HR = 2.31，95 % CI：1.23-4.33，p = 0.009）的独立危险因素。

结论

1. ACCI评分是全因死亡和非乳腺癌特异性死亡的独立影响因素。此外，它也是总生存期和非乳腺癌特异性生存期的重要预测指标。

2. 2型糖尿病对老年性乳腺癌患者的长期预后会产生负面影响，而这种影响在某些预后较差亚组中则更为明显。2型糖尿病被证明是OS和NBCSS的独立影响因素，而空腹血糖水平的控制情况比单纯是否患有糖尿病在预后预测方面可能更有价值。

3. 亚组分析结果提示合并症包括2型糖尿病虽然均会导致接受局部扩大切除的患者比例显著升高，但并未最终影响到其乳腺癌特异性生存期，对总生存期的影响多归因于非乳腺癌特异性方面。

4. Syndecan-1在肿瘤上皮细胞的表达水平升高则是老年性乳腺癌患者的无病生存期和无远处转移生存期的独立危险因素，而在总生存期方面未观察到阳性结果。

Background and purpose:

Breast cancer has become one of the most common malignant tumors in women, and has surpassed the lung cancer as the most common malignant tumor. There is also a significant increase in breast cancer rates in elderly women aged 70 and older worldwide. There are significant differences in tumor biological behavior, tolerance to treatment, and comorbidities in elderly breast cancer patients compared with their younger counterparts.

Age and comorbidity are important prognostic factors in cancer, and previous studies have shown that both comorbidity and age at diagnosis are significantly associated with the prognosis of breast cancer patients. The Charlson Comorbidity Index (CCI) was first proposed in 1987, and then in 1994, the Age-adjusted Charlson Comorbidity Index (ACCI) was proposed based on the research results. ACCI has shown a good predictive value for prognosis in various types of malignant tumors. However, there is no research on the correlation between comorbidities and prognosis in the elderly breast cancer patients, and the predictive value of ACCI in breast cancer patients has not been elucidated.

The type 2 diabetes (T2D) has a high incidence in cancer patients, and epidemiological statistics show that about 16% of elderly breast cancer patients are also accompanied by type 2 diabetes. The research on the correlation between type 2 diabetes and breast cancer prognosis has always been a hot topic. Most scholars believe that type 2 diabetes is an independent risk factor for the incidence and prognosis of breast cancer. However, to date, no independent studies have investigated the impact of type 2 diabetes on prognosis in elderly breast cancer patients.

Triple negative breast cancer (TNBC) accounts for approximately 13% of all breast cancer patients. Compared with hormone receptor-positive breast cancer, TNBC has more aggressive tumor characteristics, higher recurrence and metastasis rates and worse prognosis. Syndecan-1 (Sdc1, also known as CD138) was identified as a poor prognostic gene in TNBC patients through bioinformatic analysis of transcriptome microarray data published online in our previous research. Sdc1 is an important member of the Syndecan family, and its expression level is associated with the prognosis and treatment response of various types of tumors. High level of Sdc1 expression in breast cancer is associated with an aggressive phenotype and poor clinical behavior, and is regarded as a biomarker associated with breast cancer prognosis.

We aimed to conduct a clinical study using ACCI score as an independent variable to evaluate the effect of comorbidities on the prognosis of Chinese elderly patients with breast cancer, and further retrospectively analyzed the effect of type 2 diabetes on the prognosis specifically. At the same time, the fasting blood glucose levels of all patients with type 2 diabetes were also investigated to clarify the exact effect of different blood glucose control status on the prognosis of breast cancer patients. Finally, we planned to detect the expression of Sdc1 in tumor epithelial cells in paraffin pathological sections of tumor tissues to explore its prognostic value in the elderly TNBC cases.

Materials and methods:

According to the Chinese consensus on the treatment of elderly breast cancer, we considered patients aged ≥ 70 years as elderly breast cancer patients. This series of studies included all female patients aged ≥70 years who were diagnosed with breast cancer in our hospital and received surgery, and had breast cancer at the same time or at a different time (the age of first diagnosis < 70 years old) and those who had been diagnosed with distant metastasis were excluded. We analyzed overall survival (OS), breast cancer-specific survival (BCSS), and non-breast cancer-specific survival (NBCSS) according to the cause of death. The ACCI score starts at the age of 70, and the original CCI score is increased by 1 for every 10 years of age (70-79: +1, 80-89: +2, ≥90: +3). We selected the most representative patient tumor tissue paraffin block samples for resectioning and immunohistochemical staining. For the immunohistochemical staining of syndecan-1, we used a recombinant mouse monoclonal antibody (ab128936, ABclonal). The expression level of Sdc1 in tumor tissue epithelial cells is generally divided into four categories: negative (<10% of cells are positive), weakly positive (10-20% of cells are positive), moderately positive (21-60% of cells are positive) and strongly positive (>60% of cells were positive). Considering the sample size of this study, we uniformly classified all positive samples with different levels as Sdc1 positive group. We analyzed the association between the target variable and various clinicopathological features by chi-square test. Univariate survival analysis was performed by Kaplan-Meier curves with the log-rank test results. We also used a Cox proportional hazards model to test whether the target variable was an independent predictor of different survival outcomes.

Results:

The results of the first part showed that more than three-quarters of the patients in this study cohort had one or more comorbidities, but the vast majority of patients had an ACCI score ≤ 3. Except for hormone receptor status, all other factors were not significantly correlated with ACCI score. The proportion of hormone receptor-positive patients in the high ACCI group was higher (88.1% vs 76.2%, p=0.037). There was no significant difference in the incidence of local recurrence or distant metastasis, all-cause death and breast cancer-specific death between the two groups (all p>0.05), but the incidence of non-breast cancer-specific death was significantly higher in the ACCI>3 group than in ACCI≤3 group (19.4% vs 10.7%, p=0.032). ACCI score was an independent factor of all-cause death (HR=0.42, 95% CI: 0.22-0.83, p=0.012) and non-breast cancer-specific death (HR=0.41, 95% CI: 0.20-0.87, p=0.020) factors, all-cause mortality and non-breast cancer-specific mortality were better in patients in the low ACCI group than in the high group. Survival analysis showed that ACCI score was an independent predictor of OS (HR=2.18, 95% CI: 1.22-3.92, p=0.009) and NBCSS (HR=2.04, 95% CI: 1.02-4.08, p=0.044).

The results of the second part showed that nearly 1/4 of the elderly cohort also had type 2 diabetes (n = 163, 24.8%). Among these patients, 36.8 % had FBG levels <6.1 mmol/L, 28.8 % had FBG levels ≥6.1 and <7.0 mmol/L, and 34.4 % had FBG levels ≥7.0 mmol/L. There were significant differences in hormone receptor status and ACCI score between T2D group and non-T2D group (p<0.05). In the survival analysis, although there were no significant differences in DFS (p=0.795), OS (p=0.738), BCSS (p=0.639) and NBCSS (p=0.424) between two groups, there were significant differences in some subgroup analyses. Statistically different outcomes were observed in patients with negative hormone receptor, TNBC subtype and patients without endocrine therapy. Type 2 diabetes was independent risk factors of OS (HR=2.572, 95 % CI: 1.034-6.397, p=0.042) and NBCSS (HR=2.706, 95 % CI: 1.157-6.329, p=0.022) in multivariate Cox regression analysis. Results in the section on fasting blood glucose levels showed that poor glycemic control did not significantly affect DFS, but significantly affected OS, BCSS, and NBCSS.

The results of the third part showed that this cohort included a total of 81 triple-negative breast cancer patients, of which 41 patients had positive Sdc1 expression in tumor tissue epithelial cells, while another 40 patients were negative. The Kaplan-Meier survival curve and the corresponding log-rank results indicated that the two groups were significantly different in DFS (p=0.017) and DMFS (p=0.002), while there were no significant differences in outcomes such as OS (p=0.375), LRFS (p=0.282), BCSS (p=0.143), and NBCSS (p=0.365). In multivariate Cox regression analysis positive Sdc1 expression was independent risk factors of DFS (HR=1.85, 95 % CI: 1.16-2.97, p=0.021) and DMFS (HR=2.31, 95 % CI: 1.23-4.33, p=0.009).

Conclusion:

1. ACCI score is an independent factor for all-cause mortality and non-breast cancer-specific mortality among the elderly breast cancer. In addition, it is also an important predictor of overall survival and non-breast cancer-specific survival.

2. Type 2 diabetes negatively affects the long-term prognosis of elderly breast cancer patients, and this effect is more obvious in some subgroups with poorer prognosis. Type 2 diabetes was shown to be an independent factor for OS and NBCSS, and the control levels of fasting blood glucose may be more effective in prognostic prediction than the presence of diabetes alone.

3. The results of subgroup analysis suggested that comorbidities including type 2 diabetes led to a significant increase in the proportion of patients undergoing extended local resection, but did not ultimately affect breast cancer specific survival, and the impact on overall survival was mostly attributable to non-breast cancer specific aspects.

4. The elevated expression level of Syndecan-1 in breast cancer epithelial cells is an independent risk factor for disease-free survival and distant metastasis-free survival in elderly breast cancer patients, but no positive results were observed in overall survival, BCSS or NBCSS.