Characteristics and Outcomes of Locally Recurrent Retroperitoneal Sarcoma After First Relapse in a Single Tertiary Asian Centre and Applicability of the Sarculator

OBJECTIVE: Retroperitoneal sarcomas (RPS) comprise of 15% of soft tissue sarcomas where five-year overall survival rate is about 50%. Locoregional recurrences are observed in up to 50% of patients within the first five years following resection. Various factors have been shown to influence survival outcomes, such as histological subtype and tumour size. A nomogram for first relapse locally recurrent RPS was developed using 602 patients from 22 centres. The recurrent RPS Sarculator is available in an electronic interface and includes variables of age, size, margins of re-resection, radiotherapy, chemotherapy and histology to predict for 6-year disease-free survival (DFS) and overall survival (OS). It has not been validated externally. This study aims to validate the Sarculator recurrence nomogram in predicting the survival outcomes of recurrent RPS in an Asian population as well as examine relapse patterns. METHODS: Patients diagnosed with first recurrent RPS from 1 January 2000 to 31 December 2017 with first local relapse and eligible for curative re-resection were retrospectively analysed. The type of surgery was unique for individual patients and suggestions of adjuvant therapy were based on globally recognised standards. Patients were followed up every 3 to 4 months post-operatively for the first 2 to 3 years and 6-monthly to a year thereafter. A R0 or R1 margin is deemed as complete resection, including a microscopically negative margin (R0) and microscopically positive but macroscopically clear margin (R1). R2 is classified as an incomplete resection with tumour rupture or remaining disease. Harrell’s C concordance index was used to determine the nomogram’s discriminative ability and calibration plots were used to assess accuracy. For the calibration, the patients were divided into 3 groups. Death data was retrieved from the National Birth and Death registry for accuracy. RESULTS: There were 53 patients included in this study. Patient and tumour characteristics have been summarised in Table 1. All patients had their second resection at a single centre. 66.0% had their first resection at the same centre. The median age was 53 (range 21- 79) at diagnosis, median tumour size was 17cm (12cm to 28cm) and median follow-up duration was 44.1 months. The most commonly encountered subtypes were de-differentiated liposarcoma (DDLPS) (56.6%), well-differentiated liposarcoma (WDLPS) (20.8%) and leiomyosarcoma (LMS) (11.3%) with a majority being high-grade (75.5%). The median disease-free interval was 2.9 years (2- 5.3 years) from the first surgery. The median age at second surgery was 56 (21- 79) and all patients had a complete resection (R0/R1). Recurrence patterns differed with subtypes where 90.9% and 9.1% of WDLS, 76.7% and 16.7% of DDLPS and 83.3% and 16.7% of LMS had local and distant relapses respectively from the second surgery. 62.5% of distant relapses was in the lung followed by nodes (18.8%) and liver (12.5%). The 5-year OS from the second surgery was 66.2% (95% CI: 54.3%- 80.8%). The 1-year, 3 years and 6 years DFS were 50.2% (95% CI: 38.2% - 65.9%), 10.4% (4.26% - 25.5%) and 3.91% (0.684% - 22.4%) respectively. Overall, 32 patients (60.4%) had passed away from sarcoma. The concordance indices for 6-year OS and DFS were 0.7 and 0.65 (Figure 1) respectively which represents a fairly accurate prediction by Sarculator. CONCLUSION: Our study has shown the Sarculator nomogram for primary recurrent was applicable in our cohort and its potential application in an Asian setting. The Sarculator nomogram will be a useful tool in clinical practice to improve risk stratification and facilitate prognosis-based decision-making. Moving forward, novel therapeutic strategies are required to enhance the prognosis of patients with recurrent RPS.