临床医学论文范文沃森肿瘤与临床肿瘤医生对肺癌一致性的对比研究

沃森肿瘤与临床肿瘤医生对肺癌一致性的对比研究

摘  要

背景：人工智能（Artificial Intelligence,AI）在医疗领域的发展迅速，逐渐由实验室研究进入临床实践。IBM公司开发的沃森肿瘤（Watson for Oncology,WFO）在2017年开始应用于中国，它是AI在医疗领域的杰出代表，可以快速，准确的为肿瘤患者提出规范的诊疗方案。近几年来，WFO已经在全国十余个国家展开应用，在我国的应用也越来越广泛，相关的一致性研究也已经很多，但关于是否适用于中国患者，尤其是肺癌患者的报道却很有限。为此，我们做了一项关于WFO与我们肿瘤中心对肺癌患者诊疗方案一致性的回顾性研究。

目的：本研究旨在探索WFO在中国肺癌患者中的可行性，分析其存在的问题及优势，为将来的解决方案提供基础。

方法：我们随机选取2017年4月至10月在青岛大学附属医院肿瘤精准医学中心接受抗肿瘤治疗的肺癌患者(N=121)。其中100例符合WFO的入组标准。WFO的诊疗方案分为3类，即：“推荐”、“可考虑”和“不推荐”。当临床肿瘤科医生选择“推荐”方案或“可考虑”方案时，认为是一致，其余则认为是不一致。应用Microsoft Excel对所有患者的病例特征进行描述性统计，根据患者的病例分类，性别，年龄是否行手术等分组，对其一致性进行分析。采用SPSS17.0版本软件进行统计分析，对上述等因素进行logistic回归模型，进行概率比及95%置信区间估计，以p＜0.05为具有统计学意义。

结果：

• 在所有病例中，有21(21/121)个病例不适用于WFO。在100个适用的病例中，患者中位数年龄为61岁，男性居多占70%，女性为30%。其中做过手术的占21%，未行手术治疗的为79%。小细胞肺癌患者占19%，非小细胞肺癌患者占71%。
• 所有入组病例中，WFO提出的诊疗方案85%与肿瘤精准医学中心一致。小细胞肺癌（small cell lung cancer, SCLC）患者的一致性是48%；非小细胞肺癌（non-small cell lung cancer, NSCLC）患者的一致性为83.96%。根据肿瘤分期，83.33%的II期患者、83.33%的III期患者和85.94%的IV期患者的诊疗方案是一致的。
• 根据性别分类，WFO提出的诊疗方案与肿瘤精准医学中心的一致性对比，男性患者的一致性为57%，女性患者一致性为76.67%。
• 根据手术与否分类，WFO提出的诊疗方案与肿瘤精准医学中心的一致性对比，手术后患者的一致性为72%，未手术患者一致性为84.82%。
• 根据年龄分类，WFO提出的诊疗方案，年龄大于等于60岁的患者93%与肿瘤精准医学中心一致，年龄小于60岁的患者80.95%与我们中心一致。
• 非小细胞肺癌根据鳞癌和腺癌分类后，鳞癌占比40%，腺癌占71.60%。WFO提出的诊疗方案在86.95%的鳞癌患者及82.76%的腺癌患者中与肿瘤精准医学中心一致。腺癌根据是否基因突变分为EGFR突变组，EGFR野生型组及未行检测组。EGFR突变组的一致性为73.34%，EGFR野生组的一致性为85.71%，未行基因检测组为86.66%。
• 小细胞肺癌根据局限期，广泛期分类，WFO提出的诊疗方案78%的局限期患者与肿瘤精准医学中心一致，而广泛期则高达100%。
• 根据Logitic回归模型分析，肺癌不同分期、不同组织学类型间、性别、年龄等因素对一致性无影响。

结论:

1. 在中国，仍有较高比例的病例不符合WFO的入组标准，但是WFO针对入组病例提出的诊疗方案与临床肿瘤科医生具有较高的一致性，因此，若是合理的应用WFO，可以帮助肿瘤科医生节约工作时间，提高工作效率；
2. 从本研究可以看出，东西方针对肺癌的诊疗方案存在一定差异，首先，中国肺癌EGFR基因突变表型与西方国家差异较大；其次，中国患者无法及时从欧美国家最新开发的靶向及免疫药物中获益，可能需要滞后1-2年；最后，我们国家也会有一些自己的原研药物，或者说民族医药，是欧美国家所不具备的
3. 目前WFO面向临床，仍然存在很多问题，首先WFO的诊疗方案更多的代表是西方的诊疗方案，未根据我国人群本土化进行优化；其次，WFO的使用涉及到一些医学伦理方面的问题，有待成型的法律法规来约束。因此，WFO无法取代临床肿瘤学专家，需要加速其本土化，然后通过向医生提供帮助来提高其临床工作率。

                       

Abstract

Background：Artificial Intelligence (AI) has developed rapidly in the field of medicine, gradually moving from laboratory research to clinical practice. Watson for Oncology (WFO), developed by IBM, began to be applied in China in 2017.  It is an outstanding representative of AI in the medical field, and can quickly offer diagnosis and treatment Suggestions comparable to those of oncologists for tumor patients. WFO has been increasingly used in China in recent years, but there have been limited reports on whether it is suitable for Chinese patients, especially lung cancer. To this end, we conducted a retrospective study to examine the concordance between the treatment recommendation proposed by WFO and actual clinical decisions by oncologists in our cancer center.

Objective: This study aims to explore the feasibility of WFO in lung cancer patients in China, analyze its existing problems and advantages, and provide a basis for future solutions.

Methods: We selected lung cancer patients who received anti-tumor treatment in the oncology department of Qingdao university hospital from April to October 2017 (N=121). WFO has recommended treatment for all applicable cases (100). WFO’s recommendations fall into three categories: recommended, considered and not recommended. When the oncologist chooses the “recommended” or “considerable” option, it is considered consistent and the rest is not. Microsoft Excel was used to conduct descriptive statistics on the case characteristics of all patients, and the consistency of the patients was analyzed according to their case classification, gender, age, whether they underwent surgery or not. SPSS 17.0 software was used for statistical analysis. Logistic regression model was used to estimate the row probability ratio and 95% confidence interval of the above factors. P < 0.05 was considered statistically significant.

Result：1.Of all cases, 21(21/121) did not apply to WFO.In the 100 applicable cases, the median age of the patients was 61 years old, and the majority of the patients were men (70%) and women (30%).Of these, 21% had surgery and 79% had not.Small-cell lung cancer patients accounted for 19% and non-small-cell lung cancer patients accounted for 71%.

2.In all the enrolled cases, 85% of the treatment plan proposed by WFO was consistent with that of the tumor precision medicine center.The consistency of patients with small cell lung cancer (SCLC) was 89.48%.The consistency of patients with non-small cell lung cancer (NSCLC) was 83.96%.According to tumor stage, 83.33% of stage II patients, 83.33% of stage III patients and 85.94% of stage IV patients had the same treatment plan.

3.According to gender classification, the consistency of the diagnosis and treatment scheme proposed by WFO compared with the consistency of tumor precision medicine center shows that the consistency of male patients is 88.57%, and that of female patients is 76.67%.

4.According to the classification of whether surgery is performed or not, the consistency of the diagnosis and treatment scheme proposed by WFO and the precision medicine center of tumor shows that the consistency of patients after surgery is 85.72%, and that of patients without surgery is 84.82%.

5.According to the age classification, according to the diagnosis and treatment plan proposed by WFO, 87.93% of patients aged over or equal to 60 years old are consistent with the center for tumor precision medicine, and 80.95% of patients younger than 60 years old are consistent with our center.

6.According to the classification of squamous cell carcinoma and adenocarcinoma, the proportion of squamous cell carcinoma and adenocarcinoma accounted for 28.40% and 71.60% respectively.The WFO protocol was consistent with that of the center for accurate medicine in 86.95% of squamous cell carcinoma patients and 82.76% of adenocarcinoma patients.Adenocarcinoma was divided into the EGFR mutated group, the EGFR wild-type group and the undetected group according to whether the gene was mutated or not.The consistency was 73.34% in the EGFR mutant group, 85.71% in the EGFR wild group, and 86.66% in the untested group.

7. According to the classification of small-cell lung cancer by local deadline and broad deadline, 77.78% of patients with limited deadline proposed by WFO were consistent with our center, while the broad deadline was up to 100%.

8.According to Logitic regression model analysis, different stages, different histological types, gender, age and other factors of lung cancer had no effect on the consistency.

 

Conclutions:

1. In China, there are still a high proportion of cases that do not meet the inclusion criteria of WFO, but the treatment scheme proposed by WFO for the enrolled cases has a high consistency with the clinical oncologist. Therefore, the rational application of WFO can help the oncologist save working time and improve work efficiency.
2. It can be seen from this study that there are certain differences in the treatment schemes for lung cancer between east and west.Secondly, Chinese patients cannot benefit from the latest targeted and immunological drugs developed by European and American countries in a timely manner, which may require a lag of 1-2 years.Finally, our country will also have some of its own original research drugs, or ethnic medicine, which is not available in European and American countries
3. At present, there are still many problems in WFO for clinical practice. Firstly, WFO treatment schemes are more representative of western treatment schemes, which are not optimized according to the localization of Chinese population. Secondly, the use of WFO is involves some medical ethical issues, which need to be regulated by the laws and regulations. Therefore, WFO cannot replace clinical oncologists and needs to accelerate its localization and then improve its clinical work rate by providing assistance to doctors.

 

AI创作