香京julia种子在线播放

We performed a retrospective, population-based research by extracting data from the Surveillance, Epidemiology, and End Results (SEER) national database. Cases were included in this research on the basis of the following inclusion and exclusion criteria.

Inclusion criteria: (1) Diagnosis of lung cancer was made pathologically between the year 2010–2014; (2) Lung cancer was the first primary malignancy; (3) Detailed information about metastatic status was complete.

Exclusion criteria: (1) Age under 18 years old; (2) Metastatic status was unknown; (3) Follow-up data was missing; (4) Information about histological type was unknown.

Descriptive statistics were used to summarize patients' demographic, clinicopathological, and therapeutic variables in different histological subgroups. We conducted χ²-tests to make comparisons of metastatic distribution among different histological types. Odds ratios were calculated to analyze co-occurrence relationships between different metastatic lesions. Kaplan–Meier methods were performed to analyze overall survival (OS) and cancer-specific survival (CSS) according to different metastatic sites were conducted to identify independent prognostic factors. Two-sided P < 0.05 were defined as statistically significance. We used GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA) and SPSS 22.0 (SPSS Inc. Chicago, IL, USA) to perform the statistical analyses.

According to the inclusion and exclusion criteria, we finally enrolled 159,241 cases diagnosed with lung cancer. Detailed selection flowchart was illustrated in Figure 1. Among the final cohort, 75,231 cases (47.2%) were adenocarcinoma, 37,179 cases (23.3%) were squamous cell carcinoma, 2,832 cases (1.8%) were large-cell lung cancer, 22,709 cases (14.3%) were small -cell lung cancer, and 21,290 cases (13.4%) were non-small cell lung cancer. The baseline demographic and clinicopathological parameters according to different metastatic lesions were shown in Table 1.

Among the final cohort, 60,580 cases (38.0%) were recorded as extrathoracic metastasis. In total, the four metastatic lesions (bone, brain, liver, and distant lymph node) accounted for 94.0% (56,933/60,580) of all extrathoracic metastatic sites. And the frequencies of bone, brain, liver and distant lymph node (DL) metastasis were 19.2% (30,620/159,241), 13.3% (21,246/159,241), 12.2% (19,380/159,241), and 6.5% (10,393/159,241), respectively.

As shown in Figure 2, incidence rate of bone metastasis was the highest in SCLC (23.3%), followed by adenocarcinoma (21.5%), NSCLC/NOS (21.3%), LCLC (18.0%), and squamous cell carcinoma (11.2%). And frequencies of brain metastasis were 15.0, 5.7, 17.3, 16.7 and 16.7% in adenocarcinoma, squamous cell carcinoma, LCLC, SCLC, and NSCLC/NOS, respectively. The incidence of brain metasiasis almost the same except squamous cell carcinoma. Also, the metastatic rate of liver was extremely high in SCLC (31.5%) and relatively low in squamous cell carcinoma (6.2%). In addition, the frequency of DL metastasis in SCLC (10.6%) was higher than LCLC (8.2%), NSCLC/NOS (7.0%), adenocarcinoma (6.4%) and squamous cell carcinoma (3.9%).

For clinicopathological features, metastatic group tended to have younger age, poorer tumor differentiation, larger tumor size and higher frequency of regional lymph node invasion (Table 1). As for therapies, advanced-stage patients received less surgery and more chemotherapy than non-metastatic patients. And patients with bone or brain metastasis received more radiation therapy than non-metastatic patients.

For further analyzing combination of metastases, we performed pie charts to investigate single-metastases and co-metastases among different histological types of lung cancer (Figure 3). It is shown that bone was the leading lesion as a single metastatic site in adenocarcinoma (28.9%), squamous cell carcinoma (29.9%) and NSCLC/NOS (24.2%). Also, brain was the leading single-metastatic lesion in LCLC (23.5%), and liver was the most frequent site in SCLC (24.4%). As for combination of metastases, bi-site pattern (adenocarcinoma: 24.9%, squamous cell carcinoma: 19.1%, LCLC: 24.8%, SCLC: 28.7%, and NSCLC/NOS: 23.5%) was significantly higher than tri-site (adenocarcinoma: 7.1%, squamous cell carcinoma: 4.4%, LCLC: 6.7%, SCLC: 8.4%, and NSCLC/NOS: 6.1%) and tetra-site pattern (Adenocarcinoma: 0.8%, Squamous cell carcinoma: 0.6%, LCLC: 0.8%, SCLC: 1.1%, and NSCLC/NOS: 0.8%).

Furthermore, we calculated odds ratios to compare each possible combination of different extrathoracic metastatic lesions (Figure 4, Supplementary Figure 1). Bone preferentially tended to co-metastasize with liver (OR: 5.287) and DL (OR: 3.013). And liver metastasis was significantly correlated with DL metastasis (OR: 3.093).

In the present study, we analyzed 1-year OS and CSS in cases with diverse extrathoracic metastatic lesions (Table 2). Univariate analyses indicated that survival differences existed between non-metastatic and metastatic patients (OS: bone 50.8 vs. 20.0%, brain 48.4 vs. 22.2%, liver 49.2 vs. 13.6%, DL 46.4 vs. 22.9%; CSS: bone 57.5 vs. 25.9%, brain 55.1 vs. 27.4%, liver 55.8 vs. 20.1%, DL 53.0 vs. 28.8%). And Kaplan–Meier curves further illustrated the survival data between non-metastatic and metastatic groups (Figure 5).

Furthermore, Cox regression models were conducted to identify independent prognostic factors (Table 3). With adjusting for histological type, gender, age, race, marital status, grade, tumor size, regional lymph node invasion and therapies, all extrathoracic metastatic lesions were independent risk factors for OS (bone: HR 1.312, 95%CI 1.302–1.321; brain: HR 1.339, 95%CI 1.328–1.351; liver: HR 1.344, 95%CI 1.333–1.355; DL: HR 1.263, 95%CI 1.235–1.290) and CSS (bone: HR 1.337, 95%CI 1.328–1.348; brain: HR 1.368, 95%CI 1.357–1.381; liver: HR 1.375, 95%CI 1.363–1.388; DL: HR 1.283, 95%CI 1.254–1.313).

Additionally, survival differences between different bi-organ metastases were analyzed (Figure 6). It is suggested in the Kaplan–Meier curves that combined metastasis resulted in worse prognostic ending than the separated single-organ metastasis. Once metastasis happens, lung cancer patients might get a worse outcome.

Lung cancer related deaths are mainly attributable to extrathoracic metastasis (16, 17). Advanced lung cancer seems to metastasize to lymph nodes and other distant organs, such as brain, bone and liver. Most metastasis could cause corresponding symptoms which is represented by the N and M staging in the TNM system. In order to understand its metastatic patterns, we elucidated the following points in this research: (1) Comparing the frequencies of different metastatic lesions in different histological types; (2) Elaborating the tendency of combined metastases; (3) Clarifying the prognostic significance of single-site and bi-site metastases. As the first comprehensive, population-based research focusing on metastatic patterns in different histological types of lung cancer, the findings may provide sufficient information for clinical decision and cancer research.

According to the reported data, bone and brain were two leading distant targets for metastasis in NSCLC (18, 19). Our results further supported these findings, suggesting that bone was the most common metastatic site, followed by brain, liver and DL in all histological sites of NSCLC (including adenocarcinoma, squamous cell carcinoma, LCLC and NSCLC/NOS). By comparing the frequency of extrathoracic metastasis in diverse histological types of NSCLC, we found that more than 30% of adenocarcinoma and LCLC patients showed distant metastasis, while squamous cell cancer had the lowest rate of distal metastasis. Moreover, among all histological types of lung cancer, SCLC had the highest frequency of extrathoracic metastasis, especially to the liver, which is consistent with the reported data in previous studies (20–22). So, according to these conclusion, adenocarcinoma and LCLC patients could be arranged serious and continual follow-up, more importantly, these patients could take cutting-edge therapies, such as combined immunotherapy, neoadjuvant chemotherapy, and so on. For SCLC patients, liver ultrasound and CT scan need to be focused on.

Notably, according to the clinicopathological features, metastatic group tended to have a poorer tumor differentiation, a larger tumor size and a higher rate of regional lymph node invasion, which indicated a more aggressive and invasive hallmark of tumor biology. Compared to non-metastatic patients, advanced-stage patients received less surgery and more chemotherapy, because they lost the chance of curative resection at the time of diagnosis. And since radiation could control tumor growth of metastatic nodules as well as alleviating symptoms, patients with bone or brain metastasis received more radiation therapy than non-metastatic patients.

But these conclusions have their own historical limitaions. With the development of immunotherapy and neoadjuvant chemotherapy, patients may benefit from these modern and fancy therapies, and they could even get the chance of surgery due to the shrinking tumors. Considering these demographic, clinicopathological and treatment variables that may have impact on survival outcomes, we further conducted multivariate analysis and found that all single-site metastases were independent prognostic factors.

To our knowledge, no previous population-based researches studied the combined metastatic patterns of lung cancer. Our results indicated that bone preferentially tended to co-metastasize with liver and distal lymph nodes. And liver metastasis was significantly correlated with distant lymph node metastasis. To our knowledge, analyzing tendency of co-metastases would be rather useful to assess potential risks and make diagnosis and treatment strategies. Once bone metastasis was found, we need to screen the liver and get an enhanced CT to detect the lymph nodes. Thus, patients may get a comprehensive system treatment. And, if liver metastasis needed to be surgical removed, doctors should note that lymph node dissection is the necessary and best choice. Moreover, we further assessed the prognostic values of bi-site metastases. As shown in Kaplan–Meier curves, combined metastasis resulted in worse prognostic ending than the separated single-organ metastasis. So patients with multi-organ metastasis may need more aggressive therapeutic regimens.

Though we seriously performed this population-based research, there may still be several potential limitations. The first limitation may be the retrospective nature of this study. We only enrolled patients with detailed distal metastasis since SEER database recorded from year 2010. Second, information of extrathoracic metastatic sites was restricted to bone, brain, liver, and DL. However, these four metastatic lesions accounted for the majority of extrathoracic metastatic sites in lung cancer. Third, the metastasis condition from SEER was synchronous when diagnosed, but in the real world, metachronous carcinoma accounts for the majority. These limitations could cause bias in some results.

In a word, we comprehensively analyzed the pattern of extrathoracic metastases in different histological types of lung cancer in this population-based study. We found that the frequency of bone metastasis was the highest in adenocarcinoma, squamous cell carcinoma, LCLC and NSCLC/NOS, while liver was the most common metastatic site in SCLC. Bi-site metastases occurred more common than tri-site and tetra-site metastases. Several metastatic sites, such as bone and liver, intended to co-metastasize preferentially. All single-site metastases were independent prognostic factors and co-metastases ended up with even worse survival outcomes. Thus, our findings would be beneficial for future research design and clinical practice.