Nomogram based on homogeneous and heterogeneous associated factors for predicting bone metastases in patients with different histological types of lung cancer

Lung cancer is the most common carcinoma and the leading cause of cancer death globally [1]. Patients with metastatic disease continue to exhibit a poor prognosis [2]. Bone metastasis (BM) was reported to occur in 15–40% of lung cancer patients. The median survival time of patients with BM was reported to be less than one year [3‐5].

Early diagnosis and intervention in patients with BM could significantly influence the survival rate of patients [6, 7]. However, the Lung Cancer National Comprehensive Cancer Network (NCCN) screening guidelines do not recommend performing routine assessment or continued reassessment for BM by skeletal imaging in asymptomatic patients [8]. Previous studies reported a series of associated factors for BM, which provided the basis for predicting the BM risk [9, 10]. To promptly perform metastatic screening, a predictive nomogram based on the clinicopathologic features of lung cancer is warranted.

With the development of tumor homogeneity/heterogeneity theory, different histological types of lung cancer are now recognized to exhibit distinct prevalence of BM [4, 11]. We also believe that different histological types of lung cancer are associated with different factors for BM occurrence. Further study identifying the homogeneous and heterogeneous BM associated factors could help physicians to specifically identify the BM risk for different types of lung cancer and tailor targeted preventive treatment strategies.

The purpose of the present study was to characterize the prevalence and associated factors for BM in patients with different histological types of lung cancer using the Surveillance, Epidemiology, and End Results (SEER) database. Meanwhile, the clinical factor based nomogram was built to predict the BM risk and potentially guide the BM screening.

The present study utilized the SEER database and determined the prevalence and associated factors for BM in patients with different histological types of lung cancer. Meanwhile, the quantitative prediction nomogram for different histological types of lung cancer were built.

BM was less frequent in squamous cell carcinoma. However, adenocarcinoma and SCLC exhibited an association with BM, and adenocarcinoma accounted for more than 50% of all the lung cancer patients with BM, a finding that is partly consistent with a hospital-based study [12]. Based on the large population data set provided by SEER, the present study further concluded that small cell lung cancer and lung adenocarcinoma exhibit higher risks of BM than do the other histological subtypes. Previous studies reported similar results. After evaluating 413 patients who were diagnosed with lung cancer, Oliveira et al. found that adenocarcinoma was associated with a higher risk for developing BM, whereas squamous cell cancer was associated with a lower risk of BM [9]. Another study also reported that the most common histological type accompanied by BM was adenocarcinoma [4]. A high prevalence of BM in lung cancer and different BM prevalence rates across various histological subtypes may partly reflect the homogeneity and heterogeneity of lung cancer. Additionally, a cancer registry-based study conducted in Sweden reported that the prevalence of BM in adenocarcinoma was approximately 39%, which was higher than our results. Moreover, although BM was most prevalent among adenocarcinoma, SCLC exhibited a significantly lower prevalence of BM (25%) than did the other lung cancer subtypes, a finding that was different from ours [13]. Accordingly, the homogeneity and heterogeneity of lung cancer may differ according to ethnicity.

In the present study, we also found that different lung cancer histological subtypes exhibited homogeneity and heterogeneity with respect to the factors associated with BM. Regarding homogeneity, we identified three metastatic associated factors for all types of lung cancer: male gender, extrapulmonary metastatic site and lymphatic metastasis.

Among the present cohort, male gender was an associated factor for BM, a finding that was independent of the clinical features of lung cancer. To our knowledge, this is the first report describing sex as a risk factor for BM in lung cancer patients. Lung cancer, unlike breast cancer, is a non-hormone-dependent tumor. The reason that sex may significantly influence metastasis in non-hormone-dependent tumors remains unknown. Further research to determine potential explanations is needed.

The number of metastatic sites was also reported to be one of the metastatic associated factors for all types of lung cancer. According to our analysis, an increase in the number of metastatic sites was associated with an increased prevalence of BM in lung cancer patients. At the same time, for all histological subtypes, a higher grade of lymphatic involvement was associated with an increasing prevalence of BM in lung cancer patients. The aforementioned homogeneous associated factors may help in the surveillance of BM in lung cancer patients. Thus, physicians should focus on their lung cancer patients with these associated factors.

The advantages of early metastasis detection are as follows: 1) less toxic therapy [14, 15]; 2) SREs can be prevented through timely bone-targeted therapy [16]; and 3) improved performance status after therapy [17]. Thus, prompt metastatic screening is necessary. To make an early diagnosis and improve the survival rate of cancer patients, efforts have been made to identify the optimal screening time and method. In the latest studies, bone biomarkers and breast osteoblast-like cells were identified as potential strategies for clinical metastatic screening and early diagnosis [18, 19]. However, the aforementioned strategies require extra techniques and equipment support, which may decelerate the process of clinical screening application. Imaging remains a reliable and accepted strategy for metastatic screening and early diagnosis [20]. To prevent unnecessary radiation exposure and cost, in the present study, a histological type-based prediction nomogram was constructed. A quantitative metastatic risk could be generated utilizing the patient’s sex, lymph node metastasis, extrapulmonary metastatic site number, histological differentiation, and tumor size. We concluded that physicians could perform metastatic screening for their lung cancer patients with a high BM risk.

Inevitably, the present study has several limitations. First, in the present study, only the presence/absence of BM based on the initial diagnosis was analyzed. Disease recurrence or subsequent sites of disease were not provided in the SEER database. Thus, the actual rate of BM in patients with lung cancer might be underestimated. Second, the information on performance status, smoking status, blood type, and body mass index were not provided in the SEER database, which may affect the precision of the predictive nomogram. Third, the present predictive nomogram lacked external validation, more studies are needed in future. Fourth, the site of BM was not recorded in the SEER database, thus it cannot be further identified or analyzed in the study.

Despite the limitations, the present study provided insight into the epidemiology of BM in patients with newly diagnosed lung cancer, as recorded by the SEER database. The prevalence of BM in lung cancer was 20.9% and different lung cancer histological types showed homogeneous associated factors (male gender, more metastatic sites and more lymphatic metastasis) and heterogeneous associated factors (tumor size and histology differentiation) for BM. The median survival of the bone metastatic lung cancer patients was relatively low. The histological types significantly affect the median survival time of the lung cancer patients with BM. The nomogram has good performance for predicting the BM development in different histological types of lung cancer and the imaging of the skeletal system should be considered to lung cancer patients with a high BM risk.