Incidence, treatment and outcome of abdominal metastases in extremity soft tissue sarcoma: Results from a multi‐centre study

Abstract Background and Objectives Abdominal metastases (AM) from soft tissue sarcoma (STS) are rare and prognosis is poor. The aims of the study were to (a) identify risk factors for the development of AM and to (b) investigate the outcome of AM‐patients. Methods Seven‐hundred‐sixty‐nine STS‐patients with localised disease at diagnosis treated at three tumour centres (2000‐2016) were retrospectively included (409 males; mean age, 55.6 years [range, 8‐96 years]; median follow‐up, 4.1 years [interquartile‐range, 2.5‐6.6 years]). Results Two‐hundred‐two patients (26.3%) developed secondary metastases, and 24 of them AM (3.1%). Ten patients developed first AM (FAM) after a mean of 2.4 years and 14 patients late AM (LAM, after being diagnosed with metastases to other sites) after a mean of 2.0 years. Patients with liposarcoma had a significantly higher risk of developing AM (P = .007), irrespective of grading. There was no difference in post‐metastasis‐survival (PMS) between patients with AM at any time point and those with metastases to other sites (P = .585). Patients with LAM or FAM showed no difference in post‐abdominal‐metastasis‐survival (P = .884). Conclusions Survival in patients with AM is poor, irrespective of whether they develop secondarily to other metastases or not. Patients at high‐risk of AM (ie, liposarcoma) may be followed‐up regularly by abdominal‐ultrasound/CT.

prognosis with 2-year survival rates of 43%. 7 Diagnosis can be difficult as AM may be asymptomatic for a long time or may only cause vague discomfort. 8 Symptoms involve intestinal obstruction, abdominal pain, or gastrointestinal bleeding. 8 Since these very symptoms can also represent side effects from chemotherapy (CTX) or pain management, they can be easily misinterpreted. 6,8 There is no clear consensus how to follow-up patients with localised STS following surgical resection, with different studies describing different follow-up regimes. [9][10][11] The current ESMO guidelines 12 suggest that STS-patients should be followed-up every 3 to 4 months for the first 3 years, then bi-annually for the following 2 years and thereafter annually by chest-X-rays and/or computed tomography-scan (CT-scan) of the lungs. However, the impact of regular abdominal CT-scans in follow-up remains unclear. 12 Three studies have been published so far on the topic of AM of STS. 7,13,14 However, the number of patients with AM identified in these studies was relatively low and more importantly they did not distinguish between patients with AM as the first metastatic manifestation and those with AM developing after other metastases had occurred, therefore leaving many clinical questions unanswered.
Incidence of AM, risk factors for their development, diagnostic tools and impact on patient survival were analysed in the present retrospective multi-centre study. The aims of the study were to (a) to investigate the outcome of these patients in comparison to STS-patients with metastases to other sites and those without metastases and (b) to identify factors associated with a higher risk for developing AM.

| MATERIALS AND METHODS
We performed a retrospective analysis of the files of 769 patients who had been diagnosed with a primary localised extremity STS (G1-G3) between January 2000 and May 2016 at three tertiary tumour centres.

| Statistical analysis
All statistical analyses were performed with the Stata software version 15.1 (StataCorp, TX). Means and medians were calculated for normally and non-normally distributed data using t-tests and Mann-Whitney-U-tests, respectively. Comparisons between groups were made using χ 2 tests. Kaplan-Meier estimates and Cox-regression models were used to estimate outcome variables, providing hazard ratios (HRs), 95% confidence intervals (95%CI) and P values.
Considering that only two patients with AM underwent additional surgery, subgroup analysis to assess the effect of metastasectomy on PMS in this group of patients was not performed. Furthermore, the multivariate Cox-regression analysis was limited to two factors, in accordance with the 'one in ten rule'. 16 All P values are two-sided; a P value of <0.05 was considered statistically significant.

| RESULTS
Two-hundred-two patients (26.3%) developed SM after a median of 15 months (IQR, 10-29 months). Demographic features and tumour as well as treatment specific details of patients with and without SM are depicted in Table 1.
Taking into account all metastatic foci developing during the course of the disease, the most common location was the lung (n = 114), followed by bone (n = 32) and regional as well as distant lymph nodes (n = 27; Figure 1). Rare locations included the pericard/ endocard as well as the skin in four cases, the subcutis in three and the meninges in two cases.
Twenty-four patients developed AMs during the course of disease (3.1%), including 13 hepatic, three intestinal, two pancreatic and two peritoneal metastases. Further four patients had multiple intestinal metastases, of whom two also had metastatic foci in the retroperitoneum.
Ten patients presented with a FAM after a mean of 2.4 years (range, 7 months to 8.3 years) and 14 patients a LAM, after having developed a primary metastasis to another site. In the latter case, the meantime to LAM from the development of SM was 2.0 years (range, 1 month to 3.6 years). Moreover, hepatic FAM (n = 5) showed no tendency to develop earlier than FAM at other sites (n = 5; 17.1 vs 10.1 months; ttest P = .793). The most common symptom reported by patients at diagnosis with AM included unspecific abdominal pain in one-fifth of patients (n = 5).   in our patient cohort (Table 2).

| Risk factors for abdominal metastases
In the multivariate analysis, histological subtype only (liposarcoma vs others; P = .007) could be identified as an independent negative prognostic parameter regarding the development of AM as the first metastatic manifestation, irrespective of grading (Table 3).

| Outcome
At last follow-up, 523 patients were alive without (68.0%) and 67 alive with disease (8.7%). One-hundred-nineteen patients had died of STS Patients without SM had a significantly better OS than patients with SM (log-rank P < .0001; Figure 3). There were no differences in PMS between patients with AM at any time point and patients with SM other than AM (P = .585; Figure 4). Patients undergoing surgery for their metastasis had a significantly better PMS than patients On the other hand, we were able to include a very large number of patients with eSTS, treated according to at the time current guidelines at experienced tertiary referral sarcoma centres, factors which we believe largely offset the impact of the above limitations.

ETHICS STATEMENT
The current study has been carried out in accordance with the ethical standards of the Declaration of Helsinki and has been approved by the local ethics committee (EK-Nr. 24-573 ex 11/12).

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.