Comparisons of metastatic patterns of colorectal cancer among patients by age group: a population-based study

Population-based evaluations of the incidence of metastatic colorectal cancer at diagnosis among different age groups are lacking. Therefore, we investigated the effects of age at diagnosis on metastatic colorectal cancer and patients’ prognoses. The Surveillance, Epidemiology, and End Results database was used to identify patients diagnosed with metastatic colorectal cancer. Multivariate Cox regression analyses were performed to identify factors associated with poor survival. The Kaplan–Meier analysis was used to estimate survival differences between the subgroups. We identified 30,333 adult patients diagnosed with metastatic colorectal cancer between 2010 and 2014. The younger and middle-aged groups had better survival than the older group when brain metastasis was not involved. The liver was the most common site of metastasis followed by the liver and lung combined. Age at diagnosis was an independent factor in patients’ survival. Survival differences between two and three-sites of metastases were found in the middle-aged and older groups but not in the younger group. No survival differences between three and four sites of metastases were found in any of the age groups. Therefore, the incidence and prognosis of metastatic sites for metastatic colorectal cancer varied by age group.

Overall, 30,333 patients diagnosed with mCRC were included in our study, among which 4,309 (14.2%) were younger than 50 years old, 14,383 (47.2%) were between 50 and 69 years old, and 11,641 (38.4%) were older than 69 years old, they were defined as the younger (<50), middle-aged (50-69), and older groups (>69) respectively. There were 19,717 patients died at the end of the study and included 14,911 patients died from colorectal cancer. The percentage of deaths was 51.4% (2,215/4,309), 59.9% (8,615/14,383), and 76.3% (8,887/11,641) in the younger, middle-aged, and older groups, respectively. More detailed information about the age categories are presented in Table 1. AGING tended to be diagnosed with metastasis at a younger age (15.9%, 16.9%, and 8.3% in the younger, middle aged, and older groups, respectively).

Different metastatic patterns of colorectal cancer in patients by age group
Many patients developed metastatic diseases in more than one site. Table 2 summarizes all possible combinations of four sites of metastases. We found the liver was the most common site of metastasis for colorectal cancer and accounted for more than half of all the sites in the three age groups ( Table 2).
Patients with multiple organ metastases had fewer treatment options and tended to have poorer outcomes. Unfortunately, at least 19.8% of all the cases had multiple organ metastases. The most common combination of organs with metastases was the liver and lung, which comprised 14.8% of patients with the multiple organ metastases. Only 31(0.1%) patients had metastases to all four sites ( Table 2).
Significant differences in the rates of one site metastasis were found between the three age groups (P< 0.001). The older group tended to have more single lung metastasis (16.1%, 5.7%, and 5.1% in the older, middleaged, and younger groups, respectively) and the youn-ger group tended to have more single liver metastasis (44.1%, 54.8 %, and 55.8% in the older, middle-aged, and younger groups, respectively). However, no differences were found in the rate of metastases to two (P = 0.158) or three sites (P = 0.526).

Comparisons of OS
Substantive differences were found in overall survival (OS) (P< 0.001) between the three age groups ( Figure  1). The older group had the worst survival with a median survival time (MST) of 6 months. The results of the survival analysis of the subgroups by tumor location (Figure 1) showed that the prognoses of patients with left-sided colon cancer (LCC) and recto-sigmoid cancer (RSC) worsened with increased age. The middle-aged and the younger groups with RCC had longer MSTs than the older group. However, the younger group's prognosis was not as good as the middle-aged patients in the RCC group were (P >0.05).
We also found that patients who underwent surgery of the primary site or radiotherapy had better survival, indicating potential benefits from regional treatment for metastatic patients (

AGING
The results of the univariate analysis and multivariate Cox regression, which were conducted to evaluate the independent factors for OS are presented in Table 3.
The univariate analysis showed that all the factors included in the study were associated with OS. The significant factors were then included in the multivariate analysis and all of the factors, except for marital status (P = 0.061), were associated with OS (P< 0.05). As shown in Table 3, age at diagnosis was an independent prognostic factor for patients with mCRC. Compared to the younger patients, the middle-aged patients had worse OS (HR: 1.389, 95% CI: 1.230-1.351, P < 0.001) and the older patients had the worst OS (HR: 2.141, 95% CI: 2.041-2.247, P < 0.001).

Comparisons of survival between patients with single and multi-site metastatic colorectal cancer
We compared the effects of CRC metastases to single and multiple distant organs on the MST of the study po-pulation. The results indicated that there were significant differences in OS among the patients with different specific metastatic sites ( Figure 3). No survival difference was found between patients with metastases to three or four sites (P = 0.335, Figure 3A). This survival difference between subgroups is shown in Figure 3B-D. However, no survival difference was found between patients with metastases to two or three sites in the younger group (P = 0.061). Among the patients in the middle-aged group, no survival difference was found between those with single site metastasis and metastasis to other sites (P = 0.516).
Among the patients with single site metastasis, those with lung metastasis only, had a significantly longer survival compared to the other metastatic patients (MST = 18 months, Table 4 However, the survival of the patients with other combinations of sites was very poor and no significant differences were found.
The prognoses of patients in the three age groups with the same metastatic patterns were analyzed and the results showed that their prognoses worsened with increased age among the patients with single site metastasis. Among the patients with multiple metastases, those in the younger and middle-aged groups had better survival than the patients in the older group when brain metastasis was not involved. In other words, when patients with multiple metastases had brain metastasis, there were no survival differences among the three age groups.

DISCUSSION
This large-scale study provided more in-depth knowledge and a better understanding of the heterogeneity of colorectal cancer among different age groups. We found that a younger age at diagnosis was associated with being White, having advanced N stage, more lymph node involvement, and a tendency to have LCC. Patients in the younger age group were more likely to have single liver metastasis, but less likely to have single lung metastasis compared with patients in the older age group, and the younger patients had better survival when brain metastasis was not involved.
The most common site of CRC metastasis is the liver, which is consistent with the results of a previous study [14], and the mechanism is thought to be multifaceted. First, the "seed and soil hypothesis" may partially account for the phenomenon of different metastatic patterns [15]. It seems that tumor cells from different AGING subpopulations have favored microenvironments in distant organs, which facilitate their optimal invasion and proliferation in these organs. Second, a study revealed that neutrophils contribute to the colonization of breast cancer cells in the lung [16]. The existence of a similar mechanism for circulating tumor cells in the metastasis of CRC to the liver should be studied further.
Third, previous evidence indicates that venous drainage of the colon to the portal system might be a potential mechanism underlying the metastatic pattern of CRC to the liver, first, and then to the lungs [17,18]. The underlying molecular mechanisms need further investigation to yield findings that may be used for cancer prevention. AGING Consistent with previously published reports [19][20][21], this study revealed a significant difference in OS between patients in the older age group and their younger counterparts. Age at diagnosis was one of the independent prognostic factors in the study population. The MST was 23 months for the younger patients and 17 and 6 months for the middle-aged and older groups, respectively, which may have a multi-factorial explanation. Many age-related factors, such as lower immune response [22], and higher levels of chronic inflammation [23] may influence survival of metastasis. Hu et al., reported that patients younger than 50 years have a better chance than those older than 50 years of receiving chemotherapy, radiation therapy, and surgery [24,25]. However, elderly patients are less likely to receive optimal treatment because of age-related increases in the deterioration of organ function or comorbidities [26]. Attenuation of the immune system has also been reported to influence (i.e., worsen) survival in older adults with mCRC.
Interestingly, the survival time of the younger group of patients with RCC was similar to that of the middleaged patients. Other studies have suggested differences in terms of genetics, biology, and demographics between tumor locations of CRC; RCC was more common for both the older groups [27,28]. A recent molecular subtype analysis of CRC patients showed that the "type 5" group (MSI-high, BRAF-and KRASmutation negative, non-CpG isl and methylator phenotype) was present in a significant portion of patients aged < 40 years and 40-49 years (10% and 20%, respectively) [29]. AGING Additionally, younger patients presented with metastatic disease at the time of diagnosis more often than did older patients. This may be because RCC is more likely to be classified as a mucinous adenocarcinoma (MC) in older adults [27,28], and MC has been reported to be more frequent than adenocarcinoma (AC) in presentations of multi-metastatic diseases [5]. The poor prognosis of the RCC patients with metastatic disease might have been due the fact that curative surgery is often limited to patients with liver metastasis [11,30]. This might have led to the similar survival times of the younger and the middle-aged groups with RCC. Nevertheless, there might have been other underlying factors. More importantly, we found that patients with different metastatic patterns had different survival outcomes. Specifically, the liver metastasis only group had the longest MST compared to the other patients with metastases, whereas the brain metastasis only group and the multi-site metastases group had the poor-est outcomes. Despite their large numbers, neither clinical trials nor germinate immunotherapy-based treatments have shown significant improvements in patients with metastases [31,32]. Unfortunately, therapies are limited for patients with brain metastasis (mainly because of the blood-brain barrier) and multisite metastases [33]. These results call for greater efforts to improve precision in medicine for the prevention and treatment of CRC metastasis on an individual basis.
This study has some limitations. We could not collect detailed information about the patients' treatment, such as surgical procedures, chemotherapy regimens, or radiation methods from the SEER database, which may be a confounding factor in the results. Several other factors, other than age at diagnosis, might have also influenced the survival time of patients with mCRC. Therefore, the results need further validation in future studies. AGING CONCLUSION Our research summarized the tumor characteristics and survival outcomes of patients in three age groups with mCRC from a large sample of the population. Age was a robust prognostic factor and patients in the younger age group were more likely to have single liver metastasis, but less likely to have single lung metastasis compared with the patients in the older age group. The younger patients had better survival when their cancer did not involve brain metastasis. To determine more appropriate healthcare for aging patients with mCRC, further investigations of biochemical and molecular changes with aging are required.

MATERIALS AND METHODS
Data were retrieved from the SEER database between 2010 and 2014. The datasets, which are available in the SEER dataset repository at: https://seer.cancer.gov/, represented 30% of the United States population.
Pathology was classified as adenocarcinoma (AC), mucinous adenocarcinoma (MC), or other. Grade was defined as well differentiated, moderately differentiated, and poorly differentiated, or undifferentiated. Tumor location and their anatomical components, including RCC were classified as follows: RCC (cecum, ascending colon, hepatic flexure, and transverse colon), LCC (splenic flexure, descending colon, and sigmoid colon), RSC (recto-sigmoid junction and rectum), and appendix cancer [34]. Race/ethnicity was categorized as previously described [35]. The SEER 18 dataset categorized ethnicity as White, African-American, Native American/Alaskan Native, Asian/Pacific Islander, and unknown. The presence of bone, lung, liver, and brain metastases at diagnosis were available in the SEER database and were categorized as the number of metastases among the patients in our study. Patients were observed after the first diagnosis of CRC until the last follow-up, death, or end of the study, whichever occurred first.