RLMS is a very rare mesenchymal malignant that originates in the retroperitoneal space. RLMS is fatal but rare disease, and metastasis, multifocal lesions and resection of multiple organs are usually contraindications to the surgical resection [14]. So it is very important to analyze its long-term survival rate. In addition, the potential beneficial effects of surgical resection are not always obvious, resulting in very complicated decision-making. In other words, It is extremely difficult for clinically determine whether the patients with recurred RLMS could be treated by surgical resection. The nomogram established in this study is conductive for such decision or determination. For example, if the preoperative assessment shows no need for combined organ resection, the tumor diameter is smaller than 5 cm, FNCLCC shows grade 1 and single center disease, and even if it is a multiple recurrence disease, the postoperative 5-year OS rate may exceed 90%, so that patients have a high probability of benefiting from surgical resection. On the contrary, if the combined resection of multiple organs is needed, the tumor diameter is larger than 5cm, the FNCLCC showed grade 3 and multifocal disease, and the nomogram score was greater than 200 points, the postoperative 2-year OS is estimated to be less than 30%, so it has to be carefully considered to take surgical resection.
Since the prediction model can be simplified into a continuous numerical estimate tailored to individual patient conditions, the role of the nomogram prediction model has gradually become prominent with the help of precision medicine. Some nomogram prediction models have been incorporated into the staging system of the American Joint Committee on Cancer (AJCC) to improve the estimation accuracy of the prognosis for patients with subtype pathology. In the eighth edition of the AJCC manual, the nomogram by Trans-Atlantic Retroperitoneal Sarcoma Working Group for RPS [15] was included as a model that met all AJCC quality criteria. In his study, 523 patients were included to provide basis for establishing the nomogram prediction model, and 135 patients were used for foreign language validation. The c-index values for OS in the training set and validation set were 0.74 and 0.68, respectively. However, LMS only accounted for 17.6% in the training set, and its nomogram prediction model could only predict the OS of patients at the seventh year after the surgical resection, while the postoperative 5-year OS was only about 50%, showing limited practicality. In 2010, Anaya et al. established a nomogram model for predicting the postoperative 3-year and 5-year OS of patients by using the clinicopathological characteristics of 343 RPS patients, and its concordance index was 0.73. However, this model defined the pathological type of LMS as "others", which was not listed separately [8]. In 2019, Chandrajit et al. established a nomogram model for predicting the 6-year OS in RPS patients with first local recurrence based on the data of 602 patients from 22 sarcoma centers. Likewise, due to the rarity of RLMS, only 12.1% of LMS patients were included in the cohort [5].
In addition to nomogram prediction models for retroperitoneal sarcoma, there are also some LMS nomogramsfor the whole body. MingFeng et al. used the SEER database to establish a nomogram prediction model for predicting extremity LMS OS and cancer specific survival, with c-index of 0.776 and 0.835, respectively [16]. And the LMS nomogram prediction model for the 5-year OS of urinary system established by Oliver et al. with the c-index of 0.67 [17]. Although the primary tumor site has a significant impact on the prognosis of LMS, there is currently no nomogram prediction model for LMS originating in the retroperitoneum. Therefore, this study developed and internally validated a novel RLMS-specific nomogram for predicting the 1-, 2-, and 5-year OS of patients with RLMS, and the c-index of the model was 0.779 (95% CI, 0.659–0.898). The calibration plots showed that the predicted OS rate was perfectly match with the actual value. In addition, to enhance the clinical utility, patients were further rolled into high-risk and low-risk groups based on the scores of the nomogram prediction model. The median OS of patients in the high-risk and low-risk groups was 116.5 (95% CI, NA) and 24.0 (95% CI, 22.9–25.0) months, respectively. The TNM staging system of the AJCC is the most commonly used method for staging STS, including RLMS. Staging is based on pathological findings, including tumor size, lymph node status, metastasis, and tumor grade based on the FNCLCC system [18]. The RLMS nomogram in this study had a c-index of 0.779 (95% CI, 0.659–0.898) after internal validation, which was superior to the TNM staging system (c-index = 0.697). A more accurate prognosis could lead to a better postoperative counseling for patients, so it is possible to monitor the high-risk patients more appropriately.
The results of survival analysis suggested that the FNCLCC grade and multifocal lesion were the independent risk factors for postoperative OS. The FNCLCC system is a commonly used histological grading system for STS, and it is one of the best indicators for predicting the metastasis-free survival and OS [19]. This study was similar to a previous study reported by Qian Li et al., for RLMS patients with recurrent or metastatic disease who had a higher FNCLCC grade experienced worse prognosis [20]. Consistent with previous reports on RPS, patients with multifocal lesions accounted for 23% of all subjects in this study; in addition, the mortality risk of patients with multifocal lesion increased by twice compared with non-multifocal lesion [21]. Although there are some reports about the role of multifocal lesions in the prognosis of RPS, this study was the first to report it as an independent prognostic factor in RLMS.
There were some limitations for this study. First, this is a retrospective cohort study, and selection bias is inevitable. Second, the median follow-up time in this study was only 31 months, and further follow-up was needed to improve the reliability of the study. Finally, although the RLMS nomogram prediction model was internally validated, the cohort of this study was from an Asian tertiary hospital and was not externally validated, so its scope of use may be limited.