Prognosis of Liposarcoma Patients in Modern ERA: Single-Center Experience

Abstract

Objective  Liposarcomas are relatively rare tumors. Prognostic and predictive factors and treatment options are limited. We herein presented our 10-year experience with liposarcomas.

Materials and Methods  Adult patients with liposarcoma treated between 2005 and 2015 in our center were included. Demographic and clinicopathologic features of patients were retrieved from patient files.

Statistical Analyses  Outcomes in terms of disease-free survival (DFS) and overall survival (OS) were assessed along with potential prognostic factors using Kaplan–Meier analyses.

Results  A total of 88 patients were included. The median age was 52. Rates of well-differentiated (WDLS), dedifferentiated (DDLS), myxoid (MLS), and pleomorphic liposarcomas (PLS) were 42, 9.1, 37.5, and 4.5%, respectively. Only 10% of patients had high-grade tumors and 93% had localized disease. Ninety-six percent of patients ( n  = 84) underwent surgery. Adjuvant chemotherapy was delivered to 16 patients. The most common regimen was ifosfamide–doxorubicin. Recurrences were observed in 30 patients, 21 had local, and 9 had distant metastasis. Five-year DFS of patients with the localized disease was 68%. All patients with PLS had relapses and those had the highest distant relapse rates among all subtypes. Multivariate analysis showed T stage and grade were associated with DFS. Five-year OS of the entire population was 68%. Five-year OS was 79, 76, 50, and 0% in WDLS, MLS, DDLS, and PLS, respectively ( p  = 0.002).

Conclusion  Management of liposarcomas is still challenging. Surgery is the mainstay of treatment. Novel effective therapies are needed, particularly in advanced disease settings.

Introduction

Metin Demir

Liposarcomas constitute approximately 20% of all soft tissue sarcomas (STS). ¹ The estimated age-adjusted incidence rate was reported as 1.08 per 100.000 person-years. ² Four major subtypes are well-differentiated/atypical lipomatous tumor (WDLS), dedifferentiated (DDLS), myxoid (MLS), and pleomorphic liposarcomas (PLS). The frequency of histological subtypes was 33% for WDLS, 20% for DDLS, 19% for MLS, 7% for PLS, and 21% for other histologies, respectively. ² Surgery is the treatment of choice. ³ MLS is more chemosensitive and radiosensitive. ^{4 5 6 7} PLS is more aggressive and highly resistant to conventional cytotoxic therapy. ^{8 9}

Liposarcomas are rare tumors, with even more rare histological subtypes with diverse responses to chemotherapy (CT) and radiotherapy (RT). Limited data are available for the treatment options. In this study, we aimed to review our 10-year experience regarding clinicopathologic features, treatment modalities, and survival data of liposarcoma patients in a tertiary oncology center.

Methods

Histologically confirmed adult liposarcoma patients diagnosed and treated between 2005 and 2015 in Hacettepe University Cancer Institute were included in this study. Demographic characteristics, pathologic features, tumor locations, surgical approaches, CT, and RT data were retrieved from patient files. Outcomes were analyzed in terms of recurrence rates, disease-free survival (DFS), or overall survival (OS). Written informed consent was obtained from all participants for this study. Approval for the study was obtained from the independent ethics committee.

Categorical variables were compared with the chi-squared or Fisher's exact test, as appropriate. Survival estimates were calculated by Kaplan–Meier analysis and median survival times were compared by log-rank test. A p -value of less than 0.05 was considered to denote statistical significance.

The possible factors detected in univariate analyses were further entered into multivariate Cox regression analysis with enter selection to determine independent risk factors for patients. Variables with a p -value of up to 0.20 in univariate analysis or with clinical significance regardless of p-value were included in multivariate analyses. SPSS Software Version 26 (IBM, Chicago, Illinois, United States) was used for the analysis.

Results

After the exclusion of patients with incomplete follow-up data ( n  = 31), a total of 88 patients were included in the study. Table 1 shows the demographic and clinical features of patients. The median age at diagnosis was 52 years (min: 21, max: 81). Thirty-nine (44%) patients were female. WDLS was the most common histology with 37 patients (42%), followed by MLS, DDLS, and PLS. Median tumor size was 13 cm (min: 2, max: 65 cm). Fifty-five (63%) patients had low-grade tumors, and 9 patients (10%) had intermediate-high-grade tumors. The majority of the patients had localized (93%) disease, while six patients had node-positive and/or metastatic disease.

Table 1. Demographic and clinical features (all patients).

	n	%
Gender
Male	49	55.7
Female	39	44.3
Histologic subtype
Well-differentiated	37	42
Dedifferentiated	8	9.1
Myxoid	33	37.5
Pleomorphic	4	4.5
Unknown	6	6.8
Tumor grade
Low	55	62.5
High	9	10.2
Unknown	24	27.3
Stage
Localized	82	93.2
Node positive	2	2.3
Metastatic	4	4.5
Tumor location
Retroperitoneal/intra-abdominal/deep localization	16	18.2
Head neck/superficial body/extremity	30	34.1
Unknown	42	47.7
Perioperative chemotherapy a	26	23.4
Perioperative radiotherapy a	16	18.2
Relapse a	30	34.1
Relapse sites a
Local	21	70
Distant	9	30

^aAmong patients with early-stage disease.

Upfront resection was performed in 84 patients (96%), 14 of whom had R1 and 2 patients had R2 resection. Disease recurrence rates were 28.3% ( n  = 15) and 50% ( n  = 7) in R0 and R1 resection groups. Adjuvant CT was performed in 16 patients (8 with MLS, 2 with PLS, 2 with DDLS, 4 with WDLS). Ifosfamide–doxorubicin was the most commonly used regimen (12 of 16 patients). Clinical characteristics of patients treated with perioperative CT were shown in Table 2 . Adjuvant RT was utilized in 19 patients (22%).

Table 2. Clinicopathological characteristics of patients who received and did not receive perioperative chemotherapy.

	Perioperative chemotherapy (%)	No perioperative chemotherapy (%)
Age (years)(median)	58	52
Grade
1	6 (60)	46 (90.2)
2–3	4 (40)	5 (9.8)
Tumor size (cm)(median)	10	13.5
Histologic subtypes
Well-differentiated	4 (25)	30 (44.1)
Dedifferentiated	2 (12.5)	6 (8.8)
Myxoid	8 (50)	25 (36.8)
Pleomorphic	2 (12.5)	2 (2.9)
Unknown	0 (0)	5 (7.4)
Tumor location
Retroperitoneal/intra-abdominal/deep location	3 (18.8)	13 (19.1)
Head neck/superficial body/extremity	8 (50)	20 (29.4)
Unknown	5(31.3)	35 (51.5)

On follow-up, 30 patients had disease recurrence, 21 had local, and 9 had a distant recurrence. Five-year DFS was 68%. Recurrence rates and patterns varied significantly according to histology; 30% in WDLS (91% local), 30% in MLS (60% local), 50% in DDLS (100% local), and 100% in PLS (25% local). Fifteen recurrences (50%) were observed within the first 2 years of diagnosis, 10 (33%) within 2 to 5 years, and 5 recurrences (17%) were seen more than 5 years after diagnosis. Five-year DFS rates were 71 and 51% for patients younger and older than 60 years, respectively ( p  = 0.06). DFS curves among different histological subtypes were shown in Fig. 1 . Five-year DFS was 85% for T1-T2, 62% for T3, and 53% for T4 disease ( p  = 0.032). Five-year DFS according to histological grade was 71% for grade 1 disease, 44% for grade 2 to 3 disease ( p  = 0.003). In the multivariate analysis, T stage and histological grade were the only independent prognostic factors ( Table 3 ).

Fig. 1.

Relapse-free survival for early stage liposarcoma patients ( p  < 0.001, log-rank test).

Table 3. Univariate and multivariate Cox regression analyses for DFS.

		Univariate				Multivariate
		HR	95.0% CI for HR		p -Value	HR	95.0% CI for HR		p -Value
T Stage	T1 and T2 (ref.)
	T3	4.75	1.19	18.91	0.03	9.56	1.02	89.71	0.05
	T4	4.58	1.21	17.41	0.03	14.71	1.62	133.90	0.02
Age	65> (ref.)
	65≤	1.55	0.53	4.52	0.42	1.90	0.41	8.73	0.41
Grade	Grade 1 (ref.)
	Grade 2 and 3	3.60	1.44	8.96	0.01	4.49	1.16	17.43	0.03

Abbreviations: CI, confidence interval; DFS, disease-free survival; HR, hazard ratio

Adjuvant CT and RT were not associated with 5-year DFS ( p  = 0.96 and p  = 0.69; respectively) ( Fig. 2 ). Only 2 of the 19 patients who had received RT had local recurrence, while 19 of the remaining patients had local recurrence (11 vs. 29%, p  = 0.14).

Fig. 2.

Disease-free survival in patients treated with or without adjuvant chemotherapy ( p  = 0.96, log-rank test).

Five-year OS was 68% among all patients, 71% among patients with localized disease, and 25% among those with metastatic disease at the time of diagnosis. Only one patient with upfront multiple lung metastases survived more than 10 years with slowly growing metastases and without response to any CT. Five-year OS according to T stage was 82% for T1 to T2, 67% for T3, and 57% for T4 disease ( p  = 0.046). Patients who had recurrence had a significantly lower rate of 5-year OS compared with those without recurrence (88 vs. 48%, respectively, p  < 0.001). Patients with distant recurrence had significantly lower 5-year OS compared with those with local recurrence (13 vs. 63%, respectively, p  < 0.001). Five-year OS according to histology was 79% in WDLS, 76% in MLS, 50% in DDLS, and 0% in PLS ( p  = 0.002). Five-year OS according to histological grade was 83% for grade 1 disease, 34% for grade 2 to 3 disease ( p  = 0.009).

Among patients treated with salvage CT at relapse, seven patients (50%) received ifosfamide–doxorubicin (IMA) and five patients (35.7%) received ifosfamide–etoposide (IMET) regimens. Only one patient (14.3%) in IMA and two patients (40%) in the IMET group achieved partial response (PR). One patient with PR had PLS and two had WDLS ( Table 4 ). In the relapsed setting, 20 patients underwent surgery and 6 patients did not. Three of 20 patients underwent metastasectomy due to pulmonary involvement. Median OS was 64.4 and 23.7 months in relapsed patients who underwent metastasectomy and those who did not, respectively ( p  = 0.94).

Table 4. Responses to salvage chemotherapy regimens for histologic subtypes.

	N (%)
	PR	SD	PD
Well-differentiated	2 (50)	0 (0)	2 (50)
Dedifferentiated	0 (0)	1 (100)	0 (0)
Myxoid	0 (0)	2 (40)	3 (60)
Pleomorphic	1 (33.3)	0 (0)	2 (66.7)
Others	0 (0)	1 (100)	0 (0)

Abbreviations: PD, progressive disease; PR, partial response; SD, stable disease.

Discussion

Liposarcomas are relatively rare tumors and treatment modalities beyond surgery are limited. In our study, 5-year DFS and OS was 68%. Surgery was the mainstay of treatment, RT reduced local recurrence rate but no effect was observed on DFS. CT in the perioperative setting was utilized in less than 20% of the patients. The benefit of CT in advanced the setting was marginal at best. PLS portended the worst prognosis with no long-term survivors in our series. A recent large database analysis from United States and Canada also showed similar results; WDLS was the most common histology, with a high 5-year OS rate of 82%, followed by 76% for MLS, 63% for mixed tumors, 55% for round cell, 51% for PLS, and 49% for DDLS. ¹⁰ Our findings are consistent with these data, except for the extremely poor prognosis of PLS in our series but there were only four patients with PLS.

The principal treatment modality for STS is still surgery. Resection of the tumor with negative surgical margins confers a low risk of local relapse and better long-term survival advantage. ¹¹ The vast majority of our patients (96%) had undergone surgery in our study. R0 resection was achieved in 77% and R1 in 21% of the patients. Our resection rates were similar to the literature. In a clinical trial including 911 STS patients (31% of those were liposarcoma), R0 and R1 resection rates were 82 and 18%, respectively. ¹² Several trials revealed that neoadjuvant RT enables to achieve negative margins, prevent local relapse, reduce tumor diameter, and is associated with increased OS. ^{12 13 14} Neoadjuvant RT might be beneficial in whom R0 resection seems unlikely.

Adjuvant CT is still a controversial issue in the treatment of STS. In SMAC meta-Analysis, doxorubicin-based adjuvant CT improved local, distant, and overall relapse-free interval, particularly in extremity sarcomas. There was a trend toward improved OS but did not reach statistical significance. ¹⁵ The more recent 2008 meta-analysis showed significant benefit in OS (odds ratio: 0.56; 95% confidence interval [CI]: 0.36–0.85) with doxorubicin–ifosfamide combination. ¹⁶ Tumor location was not specifically assessed in this meta-analysis. In a phase 3 randomized controlled trial investigating histotype-driven therapy, standard anthracycline-based therapy was not inferior compared with histology-specific antineoplastic agents in the neo-adjuvant setting. ^{17 18} Five-year DFS was 47 versus 55% (hazard ratio [HR]: 1.23, 95% CI: 0.88–1.73) and 5-year OS was 66 versus 76%, (HR: 1.77, 95% CI: 1.10–2.83), for anthracycline-based and histotype-tailored therapy, respectively. This study also included patients with high-grade MLS and the results were also similar in this subgroup with trabectedin compared with doxorubicin. We use perioperative CT in patients with large high-grade sarcomas in fit and relatively younger patients regardless of tumor location. The most common CT regimen in our study was ifosfamide and doxorubicin (75%). A low number of patients and selection according to risk factors preclude assessment of the efficacy of CT in this retrospective study.

RT is recommended in the treatment of patients with intermediate- or high-grade tumors either of the extremities or the superficial trunk and was shown to reduce the risk of local recurrence. The role of RT in the treatment of retroperitoneal sarcomas is debatable. Neoadjuvant RT did not improve abdominal RFS in the EORTC-62092 STRASS trial in patients with primary retroperitoneal sarcoma. ¹⁹ In the liposarcoma subgroup, which consisted 75% of the trial cohort, a 10% absolute abdominal RFS benefit was observed. Given most of the recurrences are local in retroperitoneal LPS, RT may be considered in selected cases in this disease with a poor prognosis. Patients who received RT had a lower local recurrence rate (26 vs. 11%) in our study, consistent with the previous studies.

There is an established role of resection of pulmonary metastases of STS. In a trial with 3149 adult STS patients, median OS was 33 and 11 months for patients who underwent surgery and those who did not, respectively. ²⁰ A more recent study showed consistent results with a median OS of 33.2 months for STS patients who underwent surgery for pulmonary metastases. ²¹ The benefit of hepatic resection of STS metastases is controversial. A systematic literature review that screened available studies between 2000 and 2018 showed that in 62.5% of case reports and in 20.8 to 100% of original articles, STS hepatic metastases were resected. This trial showed OS of up to 44 months after diagnosis of metastases. ²² In our trial, OS of three patients who underwent pulmonary metastasectomy were 7, 13, and 28 months.

In advanced disease setting, doxorubicin with or without ifosfamide is the standard of care for liposarcomas. Among subtypes, MLS is more chemosensitive compared with others. ²³ In a phase 3 trial, trabectedin improved PFS in advanced liposarcoma and leiomyosarcoma (median PFS for trabectedin vs. dacarbazine, 4.2 vs. 1.5 months; HR: 0.55; p  < 0.001). ²⁴ The only agent that improves OS in liposarcoma is eribulin. In 2016, Schöffski et al randomized more than 450 patients and patients assigned to eribulin arm achieved a median OS of 13.5 months (95% CI: 10.9–15.6) compared with those assigned to the dacarbazine arm (median OS was 11.5 months [95% CI: 9.6–13.0]) (HR: 0.77 [95% CI: 0.62–0.95], p  = 0.0169). ²⁵ In a trial in which WDL was excluded, the best responses to first-line systemic treatment were as follows: 17% PR/complete response (CR), 25% stable disease, and 46% progressive disease. Anthracycline-based regimens provided the objective responses most frequently and patients who did not progress with CT had significantly higher OS than those who progressed (HR: 0.34 [95% CI: 0.15–0.77], p  = 0.009). ²⁶ We did not have any CR in our population. Patients treated with combined chemoregimens IMA and IMET had PR ratios of 14.3 and 40%, respectively. Our data belongs to the time period in which novel agents such as trabectedin and eribulin were not frequently used in our country. These findings were similar to the literature and supported that liposarcomas are highly chemoresistant. With the paucity of available therapeutic options, new anticancer agents are investigated. Studies on murine double minute (MDM-2) inhibitors, ²⁷ murine double minute (MDM-2) inhibitors combined with mitogen-activated protein kinase (MEK) inhibitors, ²⁸ CDK4/6 inhibitors, ²⁹ and immunotherapeutics ³⁰ have promising results in liposarcoma. Unlike other subgroups, pazopanib did not show efficacy in adipocytic STS in a phase 2 trial. ³¹

Our retrospective trial has some limitations. Retrospective design and the limited number of patients preclude effective comparison of treatment regimens. Toxicity data was lacking.

In conclusion, being a chemoresistant tumor, diagnosis in early-stage and appropriate surgery with or without perioperative treatment is very important. New nomograms such as Sarculator are promising for predicting survival on an individual basis and enabling clinicians to make adjuvant therapy decisions.

Funding Statement

Funding None.