Previous studies have claimed that SPM is one of the leading causes of death in BC patients. Our study found that the survival rate of female BC patients with SPM was lower than that of patients with only BC, which proved the threat of SPM to survival. Furthermore, factors that affected the OS of femaleBC patients with SPM were SPM, including TNM stage, surgery, sites, and latency. In addition, younger age at BC diagnosis predicted poor survival in these participants. Based on these risk factors, two nomograms established to predict survival probabilities after BC and SPM were diagnosed presented better discriminatory capacity, which demonstrated potential predictive value for BC management in the clinic.
Among the patients in our study, those who developed SPM had a survival rate of 75.32%, while those who did not develop SPM had a survival rate of 100% at the 5-years follow-up. In addition, the 10-years and 15-years survival rates of BC patients who developed SPM were decreased to 64.42% and 56.44%, respectively. It has been proven that the survival rate of BC is associated with socioeconomic status and culture [31]. In America, a report based on the Surveillance, Epidemiology, and End Results (SEER) database showed that the 5-years survival rate of female BC patients was 91.1% during 2008–2014 [32]. In China, the 5-years female BC survival rate was 80.4–83.3% from 2005 to 2014, which was relatively lower than that in the US [4]. In our study, the 5-years survival rate of female BC patients was 88.34% during 2002–2015, which is higher than the reported 80.4–83.3%. This discrepancy may be due to the economic advantages and advanced medical care available in Shanghai, but this needs to be further clarified.
In addition, regarding the age of BC patients, a previous study based on SEER found that in the US, 60.3% of female BC patients with SPM were 60 years old and older [6]. In our study, these patients (< 56 years old) accounted for 62.58%, which indicated that BC patients with SPM tend to be younger in China. Interestingly, it is a protective factor compared to those who were younger than 56 years old when predicting the overall survival probability of BC patients after developing SPM.
In this study, we constructed univariable and multivariable analyses as well as well-discriminatory prognostic nomogram models to predict OS in BC patients with SPM. TNM stage is an important predictor of cancer outcomes. Consistent with prior studies in other cancers [33,34,35], we found that late-stage (III + IV) SPM was associated with poor survival in BC patients with SPM. One reason for this result may be a delayed diagnosis of SPM in that patients diagnosed with late-stage SPM may not receive early detection and comprehensive treatment in a timely manner [36]. Evidence-based studies need to be further analysed. This is a reminder that early cancer screening is of importance even for a patient already diagnosed with BC.
Surgery is a vital treatment for cancer therapies. Although a study pointed out that women nursing home residents who received BC surgery had high 1-year mortality [37], our study showed that surgical treatment for primary BC had no significant effect on the OS or CSS for BC patients with SPM. However, receiving surgical treatment for SPM was shown to be protective for survival, which is similar to the findings from a study on nasopharyngeal tumours [38]. Our results may provide a reference for clinicians making treatment decisions for BC patients with SPM.
Different SPM sites were related to significant survival outcomes. In our results, SPM in the colon and rectum and thyroid predicted better survival. Colorectal cancer (CRC) is the third most common cancer worldwide [39] and the top five most common cancers in China [40]. It has been proven that the CRC screening program could improve CRC patient survival in developed and developing countries [41], which provides a reason for our result. In particular, BC patients with SPM in the thyroid had an excellent 15-years OS of 96.43%. Although the incidence of thyroid cancer has increased recently worldwide, thyroid cancer patients still have better survival rates; for instance, the 5-years survival rate of thyroid cancer patients reached 98% in America [42]. It is worth noting that a high cumulative radioiodine dose increased the incidence of SPM in thyroid cancer patients, which is a vital risk factor for death [43]. Therefore, stricter surveillance of BC patients is needed to facilitate early diagnosis and screening of SPM, and more professional and comprehensive treatment should be offered.
The risk effect of latency period on developing SPM has been studied in many cancers, but the association between latency and survival has rarely been studied. Previous studies proved that longer latency was associated with lower mortality in astrocytoma [16] and lung cancer [17]. In BC, our study first pointed out that developing an SPM with a longer latency (more than 5 years) was associated with a decreased risk of all-cause death, which is consistent with related studies of other cancers. A literature showed that adolescents and young adults who developed BC within 5 years had a 2.6-fold increased risk of death [7], however, BC was an SPM rather than a primary cancer in this study.
In the present study, two nomograms were established based on surgical treatment for SPM, SPM sites, TNM stage of SPM, latency, and age of BC diagnosis to facilitate a quantitative assessment of survival of BC patients who developed SPM once their clinical data were available. Internal validation statistically indicated that the current nomograms were accurate in their predictive abilities. Particularly, in predicting BC patients 10-years OS from SPM diagnosis, the AUC reached 0.94 and 1.00 in the training and validation cohorts, respectively. These models would enable a survival prediction of BC patients who developed an SPM, but TNM stage could not be classified in the clinic. Since the factors of latency and surgery were included, the nomogram could longitudinally evaluate overall survival and weight the risk of various management options, including surgery, at optimal time points.
Few studies have illustrated the factors influencing the survival of BC patients with SPM. Our study first identified latency, which affected the overall survival of these patients. It is worth paying attention to latency because longer latency is associated with improved survival. In addition, with respect to exploring survival risk factors for BC patients with SPM, obesity and metabolic diseases should be considered and are worthy of study, as they are important risk factors for cancers including BC [44, 45].
There are several limitations in the present study. First, although our data were collected from 50 hospitals in Shanghai during 2002–2015, the study only included 163 BC female patients with SPM, which was relatively small and affected the calibration of the nomogram models. Second, due to the long-term follow-up, some patient information about surgery for BC was not completed, which resulted in some missing bias. Third, we did not obtain health data related to obesity and metabolic diseases of participants, which may be risk factors for breast cancer [45]. Finally, we evaluated the nomogram models by internal validation, and they need to be externally validated in future real-world studies to assess the accuracy and verify their utility for clinicians.