Patient characteristics
After data selection, 3888 BC patients with lung metastasis were eligible for the subsequent analyses, with the training group of 1944 patients and the validation group I of 1944 patients. In addition, 374 patients from the Chinese XXX database were assigned as the validation II. The demographic and clinicopathological features of the population are shown in Table1. The distribution of each subgroup in the training cohort and the validation cohort I were similar enough, verifying that two cohorts were randomly divided. In the training cohort, 5.35% (104/1944), 37.14% (722/1944), 45.99% (894/1944), and 11.52% (224/1944) of the patients aged 18-39, 40-59, 60-79 and >80 respectively. Moreover, patients with HER2-/HR-, HER2+/HR-, HER2-/HR+, and HER2+/HR+ subtype were accounted for 18.31% (356/1944), 10.24% (199/1944), 53.14% (1033/1944), 18.31% (356/1944), and only 692 patients (35.60%) had metastasis confined to the lung. The median survival time was 15 months.
For the validation cohort II, 19.52% (73/374), 58.02% (217/374), 20.86% (78/374), and 1.6% (6/374) of the patients aged 18-39, 40-59, 60-79 and >80, respectively. In addition, patients with HER2-/HR-, HER2+/HR-, HER2-/HR+, and HER2+/HR+ subtype were accounted for 21.66% (81/374), 17.38% (65/374), 43.05% (161/374), 17.91% (67/374), and only 207 patients (55.35%) had metastasis confined to the lung. The median survival time was 54 months.
Univariate and Multivariate analyses of Overall Survival
Outcomes from the univariate and multivariate analyses of the training cohort and the validation cohort are listed in Table2. The following prognostic parameters were identified determinant prognostic parameters for OS: age (40-59: HR 1.206, 95%CI 0.909-1.599; 60-79: HR 1.46, 95%CI 1.101-1.935; >80: HR 2.399, 95%CI 1.728-3.332; 18-39 as a reference), race (black, HR 1.291, 95%CI 1.109-1.503; others: HR 0.873, 95%CI 0.692-1.102; white as a reference), grade(III/IV: HR 1.696, 95%CI 1.478-1.948; I/II as a reference), subtype (HER2+/HR-: HR 0.413, 95%CI 0.325-0.524; HER2-/HR+: HR 0.390, 95%CI 0.328-0.464; HER2+/HR+: HR 0.267, 95%CI 0.216-0.331; HER2-/HR- as a reference), number of metastatic sites to liver, brain, and bone (1: HR 1.396, 95%CI 1.203-1.619; 2: HR 2.548, 95%CI 2.152-3.016; 3: HR 3.710, 95%CI 2.707-5.085; 0 as a reference), surgery (BCS: HR 0.552, 95%CI 0.436-0.699; mastectomy: HR 0.691, 95%CI 0.591-0.807; no as a reference), and chemotherapy(no/unknown: HR, 1.465, 95%CI 1.271-1.688, yes as a reference). Therefore, we selected these statistically significant prognostic parameters and established a nomogram to predict the 2-, 3-, and 5- year OS.
Breast cancer-specific survival
Assessments of the probabilities of breast cancer-specific death (BCSD) and non-breast cancer-specific death (NBCSD) among the training group are shown in Table3. Obviously, black, grade III/IV, HER2-/HR-subtype, 3 metastasis sites to the liver, bone, and brain, and not accepting surgery were only significantly correlated with higher two- and five- year cumulative incidence of BSCD (P < 0.001). Especially, 7 determinant prognostic factors for OS were also significantly correlated with the probabilities of BCSD (P < 0.05 for all outcomes), indicating that the previous outcome was not influenced by the competing events. All prognostic factors which were significantly associated with cumulative incidences of BCSD and statistically significant for OS via multivariate analysis simultaneously were selected for nomogram construction.
Construction, validation, and calibration of the nomograms
Nomograms integrated eight independent prognostic factors to predict the 2-, 3- and 5- year OS and BCSS in the training cohort (Figure2). Scores assigned for the factors in each subgroup are listed in Table4.Among all involved factors, 3 metastatic sites to the liver, brain, and bone had the highest score of 100, followed by subtype (HER2-/HR-: score 96), age (>80: score 69), the number of metastasis sites to the liver, brain, and bone (2: score 68), surgery (no: score 43), grade (III/UD: score 37), race (black: score 34), chemotherapy (no: score29). To predict the probability of 2-, 3-, and 5- year OS, we simply added up the scores of each variable to obtain the total score of individual patients and located the total score on the bottom scales.
The C-indexes for the nomograms predicting OS and BCSS in the training group were 0.707 (95%CI 0.690-0.724) and 0.698 (95%CI 0.702-0.732), suggesting acceptable discrimination performance. In the training group and the validation groups, the calibration curves of the nomograms predicting OS and BCSS (Figure3) fell on a 45° diagonal line, which indicated consistency between the nomogram prediction and recorded results for 2-year OS and 2-year BCSS in both the US SEER cohort and the Chinese XXX cohort.
Risk stratification model
Furthermore, we developed a risk stratification model according to the total score of individual patients from the nomogram and stratified them into three groups: low-risk group (791/1994, 39.67%; total score ≤150), intermediate-risk group (899/1994, 45.09%; total score 151-224), and high-risk group (284/1994, 14.24%; total score ≥225). In the training cohort, the median OS time in the low-, intermediate- and high-risk group was 46.1months, 28.7months, and 13.4months, respectively. In the validation cohort II, the median OS time in the low-, intermediate- and high-risk group was 83.5months, 50.9months, and 39.0months, respectively. The Kaplan-Meier survival plots of OS and BCSS for all cohorts, the training cohort, and the validation cohort I, and validation cohort II, achieved statistical significance (P < 0.001 for all outcomes), which illustrated the capacity of the risk stratification model to discriminate prognosis between the different risk subgroups (Figure4).