Abstract:

Objective　To analyze the relationship of preoperative frailty with postoperative cognitive impairment and prognosis in elderly patients with colorectal cancer. Methods　Ninety-one elderly patients with colorectal cancer treated at Ankang Traditional Chinese Medicine Hospital from January 2019 to January 2023 were selected as the study objects. All the patients were assessed using the Montreal Cognitive Assessment (MoCA) after operation; the patients scoring ≥26 were set as a cognitively normal group (69 cases), and those scoring <26 as a cognitively impaired group (22 cases). There were 48 males and 21 females who were (73.07±7.41) years old in the cognitively normal group. There were 15 males and 7 females who were (74.81±7.64) years old in the cognitively impaired group. The basic data, pathological features, and basic biochemical indicators were compared between the two groups. The patients' frailty before operation was evaluated using the Tilburg Frailty Indicator (TFI). The serum levels of albumin, tumor necrosis factor-α (TNF-α), brain-derived neurotrophic factor (BDNF), and carcinoembryonic antigen (CEA) were measured. Pearson correlation analysis was used to evaluate the relationship between the preoperative TFI score and postoperative MoCA score. All the patients were followed up for one year; the patients who had tumor recurrence, lesion metastasis, and new lesion detection or died were categorized as a poor prognosis group (36 cases), while the remaining patients were set as a good prognosis group (55 cases). The TFI scores before operation and levels of TNF-α, BDNF, CEA, and albumin were compared between these two groups. The logistic multivariate analysis was used to explore the risk factors of poor prognosis. The predictive value of frailty status for postoperative cognitive impairment and prognosis was assessed using the receiver operating characteristic curve (ROC). t and χ² tests were used for the statistical analysis. Results　The cognitively impaired group had higher TFI score and TNF-α level, and lower levels of albumin and BDNF than the cognitively normal group [7.26±1.53 vs. 4.61±1.22, (34.18±4.34) ng/L vs. (27.97±3.59) ng/L, (31.60±3.78) g/L vs. (40.42±4.61) g/L, and (3.48±0.72) μg/L vs. (5.22±0.91) μg/L], with statistical differences (t=8.327, 6.709, 8.135, and 8.179; all P<0.05). The Pearson correlation analysis showed a negative correlation between the preoperative TFI score and postoperative MoCA score (r=-0.543; P<0.05). The TFI score and levels of TNF-α and CEA in the poor prognosis group were higher than those in the good prognosis group [7.56±1.71 vs. 3.74±1.32, (33.55±4.42) ng/L vs. (26.80±3.60) ng/L, and (181.13±24.16) μg/L vs. (148.20±13.55) μg/L]; the level of albumin in the poor prognosis group was lower than that in the good prognosis group [(34.48±3.69) g/L vs. (40.78±4.36) g/L]; there were statistical differences (t=11.994, 7.986, 8.319, and 7.151; all P<0.05). The multivariate analysis indicated that the TFI score and levels of TNF-α, CEA, and albumin were predictors of poor prognosis (all P<0.05). The accuracies of preoperative TFI score in the prediction of postoperative cognitive dysfunction and poor prognosis was 90.3% and 94.7%, respectively. Conclusion　Preoperative frailty in elderly patients with colorectal cancer affects postoperative cognitive function and may increase the risk of poor prognosis.

Key words:  , Colorectal cancer,  , Frailty,  , Cognitive dysfunction,  , Poor prognosis,  , Risk factor prediction

摘要：

目的　分析老年结直肠癌患者术前衰弱与术后认知障碍及预后的关系。方法　选取2019年1月至2023年1月在安康市中医医院治疗的老年结直肠癌患者91例作为研究对象。所有患者术后均进行蒙特利尔认知评估量表（MocA）评分，将得分≥26分的患者设为认知正常组（69例），<26分的患者设为认知障碍组（22例）。认知正常组男48例，女21例，年龄（73.07±7.41）岁；认知障碍组男15例，女7例，年龄（74.81±7.64）岁。对比两组患者基本资料、病理特征及基本生化指标。患者术前均通过Tilburg衰弱量表（TFI）对衰弱状态进行评估，并测定术前血清白蛋白、肿瘤坏死因子-α（TNF-α）、脑源性神经营养因子（BDNF）及癌胚抗原（CEA）水平，通过Pearson相关性分析评估患者术前TFI评分和术后MocA评分的关系。对所有患者进行为期1年的随访，将肿瘤复发、病灶转移、发现新病灶或死亡患者设为预后不良组（36例），其他患者设为预后良好组（55例），对比两组患者术前TFI评分、TNF-α、BDNF、CEA和白蛋白水平，并通过logistic多因素分析探讨患者预后不良的风险因素，采用受试者操作特征曲线（ROC）评估衰弱状态对术后认知障碍和预后情况的预测价值。采用t检验、χ2检验进行统计分析。结果　认知障碍组TFI评分、TNF-α水平均高于认知正常组［（7.26±1.53）分比（4.61±1.22）分、（34.18±4.34）ng/L比（27.97±3.59）ng/L］，白蛋白、BDNF水平均低于认知正常组［（31.60±3.78）g/L比（40.42±4.61）g/L、（3.48±0.72）μg/L比（5.22±0.91）μg/L］，差异均有统计学意义（t=8.327、6.709、8.135、8.179，均P<0.05）。Pearson相关性分析结果显示，患者术前TFI评分和术后MocA评分呈负相关（r=-0.543，P<0.05）。预后不良组TFI评分、TNF-α、CEA水平均高于预后良好组［（7.56±1.71）分比（3.74±1.32）分、（33.55±4.42）ng/L比（26.80±3.60）ng/L、（181.13±24.16）μg/L比（148.20±13.55）μg/L］，白蛋白水平低于预后良好组［（34.48±3.69）g/L比（40.78±4.36）g/L］，差异均有统计学意义（t=11.994、7.986、8.319、7.151，均P<0.05）。logistic多因素分析结果显示，TFI评分、TNF-α、CEA、白蛋白水平均是患者预后不良的影响因素（均P<0.05）。术前TFI评分对患者认知功能障碍和预后不良的预测准确度分别为90.3%、94.7%。结论　老年结直肠癌患者术前衰弱状态可影响术后认知功能，增加预后不良风险。

关键词: 结直肠癌,  ,  , 衰弱状态,  ,  , 认知障碍,  ,  , 预后不良,  ,  , 风险因素预测