Clinical efficacy of chemotherapy in colorectal cancer patients over 80 years old

Colorectal cancer (CRC) is one of the most common cancers worldwide. According to data from cancer statistics, CRC ranks third in morbidity and second in mortality [1]. In addition to incidence rate, the age of onset is gradually improving. Among CRC patients in the US in 2017, 27% of new cases and 40% of deaths in women occurred in people aged 80 and over, compared with 18% and 27% in men, respectively [2]. For patients with CRC, surgery and chemotherapy are considered the preferred treatments, which significantly reduces the mortality and improves the prognosis [3, 4]. For the super elderly patients (over 80 years old), due to the comorbidity and frailty issues, there are some confusions in the choice of treatment for CRC. Different from patients in early age onset, elderly patients should consider receiving active treatment after risk assessment. In previous studies, there were many studies on the surgical methods and chemotherapy regimens of CRC patients, but few studies have shown the impact of two treatment methods, especially chemotherapy, on the prognosis of super elderly patients [5‐8]. There is no representative randomized controlled study on the antitumor treatment of CRC in super elderly patients. More importantly, even super elderly patients actively choose because of their health status, so strong selection bias needs to be considered. Due to the lack of real-world research data, the medical community currently lacks guiding opinions on super elderly patients. The huge heterogeneity of health status in super elderly patients with CRC (due to multiple comorbidities) makes it difficult to establish randomized controlled trials in these populations. The current clinical trial design may not be suitable for super elderly patients. Moreover, for CRC patients with different stages, the effects of chemotherapy and surgery are also different [9, 10]. Therefore, in the absence of clinical trial research, the purpose of this retrospective study was to investigate the efficacy of chemotherapy in super-elder patients with CRC in different TNM stages.

In the present study, we analyzed the effects of chemotherapy on OS and CSS in super elderly patients in different TNM stages. With the global aging process, there have been many studies on the prognostic factors of super elderly patients with CRC, but these studies generally define the elderly as over 65 years old. For patients over 80 years old, relevant research is also important. Because of the special situation of super elderly patients, there are few studies on the prognostic factors of CRC over 80 years old [11]. The special manifestations of super elderly patients are as follows: (1) They are often complicated with a variety of basic diseases, which has obvious obstacles to surgery and chemotherapy; (2) liver and kidney function and hematopoietic system function decline and cannot tolerate radical treatment schemes; (3) the expected survival time is short, and it is difficult to predict the future for the treatment completed at risk. Therefore, the choice of treatment for super elderly patients with CRC is often a dilemma. To the best of our knowledge, this study is the first to analyze the effects of chemotherapy on OS and CSS in super elderly patients in different TNM stages.

First of all, it is believed that for elderly patients with CRC(> 75 years), surgery treatment has an acceptable survival rate, and age should not be the reason for refusing surgery [12]. In our study, it revealed that surgery can significantly improve the prognosis of the super elderly CRC patients. Whether OS or CSS, the prognosis of patients in the operative group is better than that in the nonoperative group (P < 0.001). Therefore, based on the results of this paper, surgical treatment is recommended for super elderly patients with CRC. However, physical and nutritional status need to be fully evaluated before operation [13, 14]. Improving the surgical selection rate of elderly patients may improve the long-term prognosis, but it may be at the cost of higher 30-day mortality [15].

Second, chemotherapy plays an important role in CRC treatment. Patients with CRC after operation were divided into four groups according to TNM stage, and the effects of chemotherapy on each stage were observed. We found that some stage I super elderly patients with CRC received chemotherapy. The available data from the SEER database were lack of reasons of chemotherapy for post-surgery CRC patients, though it was well known that adjuvant therapy after surgery was not recommended in the standard therapeutic method according the NCCN guidelines and other guideline for these patients [13, 16]. Due to the lack of original records of chemotherapy, we do not know the exact reason for chemotherapy in these patients. Therefore, we can only speculate through the existing data. There may be many reasons, including positive margins, lymphovascular invasion, or SM3 invasion (submucosal invasion to the lower third of the submucosal level); at the same time, the patient refuses to operate again. Other possible causes are postoperative recurrence or nonstandard treatment. In order to figure out this problem, we may need larger sample size data and relevant research specifically for postoperative chemotherapy of stage I colon cancer, as well as detailed data on the causes of chemotherapy. There is controversy in the previous literature on patients with stage II. It is generally accepted that high-risk patients are recommended for chemotherapy. In one review article, patients with pT4, lymphovascular invasion, perineural invasion, obstruction, perforation, or high risk in gene expression signature in stage II or in stage III, should accept chemotherapy [17]. For patients with stage I and stage II CRC, chemotherapy could not improve OS (P > 0.05). Therefore, for super elderly CRC patients accepted surgical treatment, stage I and stage II patients who are not recommended for chemotherapy. Due to the limitations of the database, the relevant risk factors of stage II mentioned in the above literature cannot be analyzed.

It was found that the CSS of patients with chemotherapy in stage I/II surgical group was lower than that of patients without chemotherapy. To clarify this question, we analyzed the Cox proportional hazard regression model for the two groups of super elderly CRC patients. The P values of chemotherapy factor in the two groups were 0.303 and 0.160 respectively, and the P values of log-rank test were 0.002 and < 0.001 respectively. Then, we performed the same analysis on all remaining groups, and both were consistent. It is known that log-rank test belongs to nonparametric method, and Cox belongs to semi parametric method. When the conditions are met, the efficiency of parametric test is higher than that of nonparametric test, and Cox can prevail. Secondly, there may be a bias between chemotherapy patients and non-chemotherapy patients. Patients who choose chemotherapy may have more serious conditions, although we did not find significant differences in the included factors. If we want to understand this puzzle, we may need a larger sample size and prospective research. Finally, given that OS is widely recognized in cancer research, although the CSS results are equivocal, it can also be concluded from the results of OS that stage I/II super elderly CRC patients cannot benefit from chemotherapy.

For stage III and IV CRC, this study indicates chemotherapy can improve OS and CSS in super elderly patients (P < 0.001). A study that included four randomized, controlled trials suggested that patients with stage III CRC of all ages should be treated with XELOX or FOLFOX as standard treatment option [18]. For stage III and IV CRC patients, the median progression-free survival (PFS) of patients receiving chemotherapy was significantly longer than that of patients not receiving chemotherapy (the mean age was 61.3 years, P < 0.001) [19]. Unfortunately, there is no data on specific chemotherapy regimen in SEER database, so correlation analysis cannot be carried out. However, through previous studies, it can be seen that for super elderly CRC patients, it is also very important to choose an appropriate chemotherapy regimen, and further clinical research is also needed.

OS and CSS were improved in all TNM stages after chemotherapy in nonoperative group (P < 0.001). A phase II clinical study involving 40 patients (median age 77.3 years) suggested that 5-fluorouracil and irinotecan (FOLFIRI 1) regimen was an effective treatment for elderly subjects with advanced CRC in good clinical condition [20]. In addition, in a phase III trial of elderly patients with metastatic CRC (media age was 80 years), 71 patients were randomly divided into LV5FU2 group, 71 into simplified LV5FU2 group, 70 into LV5FU2 irinotecan group, and 70 into FOLFIRI group. It is considered that irinotecan cannot significantly improve the progression-free survival (PFS) or overall survival (OS) of patients. At the same time, compared with simplified LV5FU2, the classic LV5FU2 regimen can improve the OS of patients [21]. An open label randomized trial involving 459 patients with advanced CRC showed that frail and elderly patients could participate in randomized controlled trials through appropriate adjustments, including reducing the starting dose of chemotherapy. Overall, combination drugs including oxaliplatin were superior to single agent fluoropyrimidine, although the primary endpoint of PFS was not reached. Capecitabine did not significantly improve the quality of life of patients with CRC compared with fluorouracil [22]. Therefore, super elderly patients with colorectal cancer, especially those without surgical treatment, should not refuse chemotherapy due to age.

However, there are still many limitations in this study. First, the data studied in this paper come from a single database, Multi center research is more meaningful for the treatment choice of super elderly CRC patients. Secondly, SEER database only has the data of chemotherapy or not, but there is no specific data for the chemotherapy regimen. For the elderly CRC patients over 80 years old, the patient’s own combined diseases and the choice of chemotherapy regimen (chemotherapy drugs and dose) have a great impact on the prognosis. If these two data are recorded, it should be convincing for this research. Third, there is less data for patients in nonoperative group. Therefore, multicenter prospective research is needed.

For super elderly patients with CRC, surgery can improve OS and CSS. Chemotherapy can improve OS and CSS in patients with stage III and IV CRC, but not in patients with stage I and II CRC. For patients without surgical treatment for various reasons, chemotherapy can improve OS and CSS for patients in all TNM stages. For these patients, age should not be the only indicator to decide whether to operate or chemotherapy. Patients with different TNM stages should adopt different treatment schemes. In addition, despite the difficulties in implementation, clinical trials in super elderly patients should be considered.