The Clinicopathological Features and Survival Prediction of Gastrointestinal Leiomyosarcoma: A SEER Based Analysis

doi:10.21203/rs.3.rs-4673854/v1

Background

There is still lacking knowledge and awareness of patients with gastrointestinal leiomyosarcomas (GILMS). This study aimed to clarify the disease features and prognosis of GILMS based on Surveillance, Epidemiology, and End Results (SEER) database.

Methods

Clinicopathological, treatment, survival, and prognostic data of GILMS from 2000–2020 were retrieved. Appropriate statistical approaches were used to evaluate the demographic features and survival outcomes.

Results

A total of 479 GILMS cases conformed the screening criterion. The median age of patients was 64 (15–90) years. Only 7.9% of these cases were diagnosed with lymph node metastasis. 78.1% of patients were diagnosed at T2 or higher primary tumor extension. The 5-year cancer specific survival (CSS) of patients undergone surgery was 66.5%, and 5-CSS for patients without surgery was 0. Multivariate Cox proportional hazards analysis based on both overall survival (OS) and CSS showed Female was significant independent protective factor. Significant independent risk factors included: age at 65–79 years, age ≥ 80, poor differentiation, T2 and higher T stage, distant metastasis, and without surgery. Both chemotherapy and radiotherapy performed no influence on survival and prognosis.

Conclusion

GILMS was commonly diagnosed in old patients with low rate of nodal and distant metastasis. Whereas most GILMS was presented with large tumor size and high extent of tumor invasion. Complete excision was demonstrated to be the unique way to prolong survival outcomes. Chemotherapy and radiotherapy can’t bring benefits on survival and prognosis highlighting the urgent need for exploration of effective systematic treatments.

Biological sciences/Cancer/Gastrointestinal cancer

Biological sciences/Cancer/Sarcoma

Chemotherapy

Gastrointestinal leiomyosarcomas

Radiotherapy

SEER

surgery

Soft tissue sarcomas are a rare subtype of malignant tumors, with a low incidence rate of 0.69% last year in the United States. ^[1]. Leiomyosarcoma (LMS) is a relatively common subtypes of STS and represents approximately 10–20% of all STS^[2]. Common sites of origin include uterus, larger blood vessels, and retroperitoneum^[3]. Clinical behaviour and response to therapy of LMS in different primary sites are variable^{[4, 5]}. Uterus is the most vulnerable to LMS compared with other organs. Understanding behaviour of uterus LMS is feasible in large population based study^[3]. Given the rarity of LMS located in certain sites such as gastrointestinal (GI) tract, relative clinical studies remain critical to elucidate clinicopathological and prognostic information and inform clinicians valuing LMS located in GI tract.

Prior to the finding of gene mutation of KIT, gastrointestinal stromal tumors (GIST) were defined as LMS before 2000^[6]. GIST is distinguished from GILMS according to genomic alteration (FDGFRA or KIT) and immunohistochemical indicators (CD117, and/or CD 34, DOG1)^[5]. This distinction is paramount important as GILMS is rarer, and with different treatment and prognosis compared to GIST^[7]. only two retrospective studies concentrating GILMS published nowadays^{[7, 8]}. One of the studies only analyzed 11 cases and reviewed published literature mentioned on GILMS since 2000. The other study summarizes limited clinicopathological features and survival outcomes based on 47 cases in a single institution between 2000–2020. The demographic, tumor extension, and lymph and distant metastasis rate is not well described in the literature. Furthermore,

Surgical outcomes and response to chemotherapy and radiotherapy remains poorly understand.

Surveillance, Epidemiology, and End Result Analysis (SEER) database includes the demographic, treatment and prognosis data of ~ 28% of American population. We determined to clarify the clinical characteristics and survival outcomes of a large cohort of cases with GILMS managed within the SEER database to further improve the recognition of this rare subtype.

Participants

This study retrieved the SEER dataset named “the Incidence – SEER Research Data, 17 Registries, Nov 2022 Sub (2000–2020)”. Data were downloaded using SEER*Stat software (version 8.4.2) acquired from the SEER official website (https://seer.cancer.gov/data-software/). Cases with the diagnosis of gastrointestinal leiomyosarcoma were identified according to their specific histologic codes of the International Classification of Diseases for Oncology, 3rd edition (ICD - O − 3): 8890/3, 8891/3, 8896/3, and their corresponding location: Site recode ICD-O-3/World Health Organization [WHO] 2008: Esophagus, stomach, small intestine, and colorectum.

Variables

The clinicopathological variables used for analysis in this study included: gender, year of diagnosis, race, age, primary sites – labeled, grade record, regional node positive, metastasis positive, surgery, radiotherapy, chemotherapy, tumor size, primary tumor extension, vital status records, survival months. There were no recommended AJCC prognostic stage groups at this time, But the definition of primary tumor (T), reginal lymph node (N), and distant metastasis (M) was in clinical use. T was classified into T1 (organ confined), T2 (tumor extension into tissue beyond organ), T3 (invade another organ), and T4 (multifocal involvement). N was further subdivided into N0 (no lymph node involvement) and N1 (lymph node involvement present). AJCC also classified the M as M0 (no metastasis) and M1 (metastasis present).^[9]

First, 704 patients with GILM were found out from January 1 2000, to December 31 2020. Second, 12 patients without record on surgery, survival months, and race were excluded. Finally, 479 patients were extracted for our study after exclusion of 213 cases who lost information on tumor size, tumor extension, lymphatic metastasis, and distant metastasis. The screening process was detailed in Fig. 1.

Statistical analysis

Unordered variables were analyzed by chi-square test. Ranked variables were compared using rank sum test. Continuous variables were compared by student’s t - test. Kaplan - Meier plots was applied to plot survival curves. Survival months between different variables were compared by log - rank test. Cox proportional hazards model was used to pinpoint prognosis factors for overall survival (OS) and Cancer-specific survival (CSS). To achieve a true embodiment of the influence of GILM on survival, a death that was on account of GILM was defined as an event, and other reasons were censored observations.CSS was defined as the period from diagnosis to death caused by GILM or until the last follow-up. Statistical analyses were performed using version 26.0 (IBM Inc, Chicago, IL, USA). A two - tailed p - value < 0.05 was deemed as statistically significant.

The methods of this study were carried out in accordance with the research guidelines published in the SEER database. This study received prior approval by the Institutional Review Board at the Third Affiliated Hospital of Kunming Medical University.

Features of patients with GILMS

As detailed in Fig. 1, only 704 patients were diagnosed with GILMS between 2000–2020. Our line chart showed little variation in the number of this rare sarcoma from year to year between the two decades (Fig. 2A). Disease-specific information of the patients with GILMS was recorded in Table 1. The male to female ratio was close to 1, and 50.5% patients were male. The mean age was 63 ± 14 years, patients with age>50 years were the main population accounting for 82.9%. Only 3.8% of the patients had a primary site in esophagus. A total of 78.1% of patients were diagnosed at T2 or higher primary tumor extension. Rate of lymph node metastasis was 7.9% of these cases, and 55 patients (11.5%) were diagnosed with distant organ metastasis. Regarding therapeutic measures, 8.6% of GILM patients received radiotherapy, and 13.8% undergone chemotherapy. Most of the patients (93.9%) received surgical management. In addition, surgery alone (77.2%) was the most common treatment regimen (Fig. 2B).

Table 1

Characteristics of patients with GILM (N = 479).
Characteristics	GILM
Characteristics	N	%
All	479	100
Gender
Male	242	50.5
Female	237	49.5
Age at diagnosis (years old)
Median	64 years old
Mean	63 ± 14 years old
<50	82	17.1
50–64	164	34.2
65–79	166	34.7
≥80	67	14.0
Years of diagnosis
2000–2005	159	33.2
2006–2010	80	16.7
2011–2015	111	23.2
2016–2020	129	26.9
Ethnicity
Caucasian	382	79.7
African-American	60	12.5
Asian	37	7.7
Primary sites
Esophagus	18	3.8
Stomach	121	25.3
Small intestine	163	34.0
Colorectum	177	37.0
Pathological differentiation
Well	33	6.9
Moderate	67	14.0
Poor	61	12.7
Undifferentiated	111	23.2
Unknown	207	43.2
Tumor size (cm)
Median	6.0 cm
Mean	7.7 ± 5.5 cm
0–5 cm	179	37.4
5–10 cm	195	40.7
>10 cm	105	21.9
Tumor extension
T1	105	21.9
T2	265	55.3
T3	60	12.5
T4	49	10.2
Lymphatic metastasis
Yes	38	7.9
No	441	92.1
Distant metastasis
Yes	55	11.5
No	424	88.5
Surgery
Yes	450	93.9
No	29	6.1
Radiotherapy
Yes	41	8.6
No	438	91.4
Chemotherapy
Yes	66	13.8
No	413	86.2
Treatment regimens
Surgery alone	370	77.2
Chemotherapy alone	19	4.0
Radiotherapy alone	6	1.3
Surgery + chemotherapy	45	9.4
Surgery + radiotherapy	24	5.0
Chemotherapy + radiotherapy	4	0.8
Surgery + chemotherapy + radiotherapy	11	2.3

Comparison of clinicopathological information was processed in Table 2. Patients diagnosed at earlier years were more likely diagnosed with advanced T stage (p = 0.007). Patients with T3-T4 stage meant higher proportion of lymph node metastasis (12.8%, p = 0.031), distant metastasis (18.3%, p = 0.011), and undifferentiated differentiation (p = 0.043). Patients with primary site in small intestine showed a trend toward T3-T4 (42.2%), and primary site in colorectum had a higher percentage of T1-T2 (40.0%, p = 0.050). For treatment measures, patients with T3-T4 had higher proportion of chemotherapy (32.1%, p < 0.001) and lower possibility of surgery compared with T1-T2 (87.2, p = 0.001). In comparison with male, female usually had a younger onset age (p = 0.005) and lower rate of distant metastasis (8.5%, p = 0.039).

Table 2

Characteristics of GILM patients classified by tumor extension and gender respectively.
Characteristics	T1-2		T3-4		p value	Male		Female		p value
Characteristics	N	%	N	%	p value	N	%	N	%	p value
All	370	100	109	100		242	100	237	100
Gender					0.839	-	-	-	-
Male	186	50.3	56	51.4		-	-	-	-
Female	184	49.7	53	48.6
Age at diagnosis (years)					0.396					0.005
<50	64	17.3	18	16.5		32	13.2	50	21.1
50–64	124	33.5	40	36.7		79	32.6	85	35.9
65–79	125	33.8	41	37.6		101	41.7	65	27.4
≥80	57	15.4	10	9.2		30	12.4	37	15.6
Years of diagnosis					0.007					0.140
2000–2005	110	29.7	49	45.0		73	30.2	86	36.3
2006–2010	59	15.9	21	19.3		49	20.2	31	13.1
2011–2015	93	25.1	18	16.5		53	21.9	58	24.5
2016–2020	108	29.2	21	19.3		67	27.7	62	26.2
Ethnicity					0.981					0.184
Caucasian	295	79.7	87	79.8		201	83.1	181	76.4
African-American	46	12.4	14	12.8		26	10.7	34	14.3
Asian	29	7.8	8	7.3		15	6.2	22	9.3
Primary sites					0.050					0.137
Esophagus	15	4.1	3	2.8		13	5.4	5	2.1
Stomach	90	24.3	31	28.4		61	25.2	60	25.3
Small intestine	117	31.6	46	42.2		87	36.0	76	32.1
Colorectum	148	40.0	29	26.6		81	33.5	96	40.5
Pathological differentiation					0.043					0.077
Well	30	8.1	3	2.8		11	4.5	22	9.3
Moderate	53	14.3	14	12.8		31	12.8	36	15.2
Poor	52	14.1	9	8.3		29	12.0	32	13.5
Undifferentiated	77	20.8	34	31.2		53	21.9	58	24.5
Unknown	158	42.7	49	45.0		118	48.8	89	37.6
Tumor size (cm)					<0.001					0.450
0–5 cm	159	43.0	20	18.3		85	35.1	94	39.7
5–10 cm	144	38.9	51	46.8		99	40.9	96	40.5
>10 cm	67	18.1	38	34.9		58	24.0	47	19.8
Tumor extension										0.907
T1-2	-	-	-	-		186	76.9	184	77.6
T3-4	-	-	-	-		56	23.1	53	22.4
Lymphatic metastasis					0.031					0.343
Yes	24	6.5	14	12.8		220	90.9	221	93.2
No	346	93.5	95	87.2		22	9.1	16	6.8
Distant metastasis					0.011					0.039
Yes	35	9.5	20	18.3		35	14.5	20	8.4
No	335	90.5	89	81.7		207	85.5	217	91.6
Surgery					0.001					0.605
Yes	355	95.9	95	87.2		226	93.4	224	94.5
No	15	4.1	14	12.8		16	6.6	13	5.5
Radiotherapy					0.515					0.375
Yes	30	8.1	11	10.1		18	7.4	23	9.7
No	340	91.9	98	89.9		224	92.6	214	90.3
Chemotherapy					<0.001					0.078
Yes	31	8.4	35	32.1		40	16.5	26	11.0
No	339	91.6	74	67.9		202	83.5	211	89.0

Survival analysis

The 1-, 3-, and 5-y OS of patients with GILMS was 84.0%, 64.9%, and 54.6%, respectively (Fig. 3A). Meanwhile, higher 1-, 3-, and 5-y CSS compared with OS was showed in Fig. 3B, they were 87.7%, 72.1%, and 64.1%, respectively. Compared to male, female had better OS and CSS (p < 0.001) (Fig. 3C, D). Patients diagnosed with well and moderate differentiation showed significantly higher OS and CSS rate (p < 0.001) than poor and undifferentiated types (Fig. 3E, F). Early T stage performed significantly longer OS and CSS than advanced T stage (p < 0.001) (Fig. 3G, H). onset of distant metastasis hinted poorer OS and CSS (p < 0.001) (Fig. 3I, J). Undergoing surgery could bring obvious survival benefits both appeared in OS and CSS (Fig. 3K, L) (p < 0.001). Longitudinal time analysis demonstrated trend of OS and CSS improvement from 2006 to 2020, though the p value was not significant. The 5 years CSS rate of GILMS patients between 2006 and 2010 was 59.7%, which increased to 65.2 between 2016 and 2020 (Fig. 3M, N). Radiotherapy did not show prolong survival time. The 1-, 3-, and 5-y OS and CSS rates of the relative variables were detailed in Table 3.

Table 3

Overall cumulative and cancer-specific survival data of 479 patients with GILM.
Variables	OS rate			CSS rate
Variables	1-year	3-year	5-year	1-year	3-year	5-year
Gender
Male	0.834(0.787–0.881)	0.576(0.509–0.643)	0.463(0.394–0.532)	0.869(0.826–0.912)	0.657(0.59–0.724)	0.568(0.495–0.641)
Female	0.815(0.764–0.866)	0.697(0.636–0.758)	0.634(0.569–0.699)	0.86(0.815–0.905)	0.773(0.716–0.83)	0.731(0.668–0.794)
Age at diagnosis (years)
<50	0.959(0.914–1.004)	0.799(0.705–0.893)	0.681(0.567–0.795)	0.959(0.914–1.004)	0.812(0.72–0.904)	0.739(0.631–0.847)
50–64	0.872(0.819–0.925)	0.678(0.602–0.754)	0.570(0.488–0.652)	0.902(0.855–0.949)	0.765(0.694–0.836)	0.676(0.594–0.758)
65–79	0.803(0.742–0.864)	0.587(0.509–0.665)	0.482(0.400-0.564)	0.855(0.800–0.910)	0.695(0.619–0.771)	0.613(0.529–0.697)
≥80	0.607(0.489–0.725)	0.477(0.354-0.600)	0.403(0.28–0.526)	0.680(0.564–0.796)	0.532(0.403–0.661)	0.488(0.357–0.619)
Years of diagnosis
2000–2005	0.780(0.715–0.845)	0.635(0.561–0.709)	0.528(0.450–0.606)	0.825(0.764–0.886)	0.715(0.642–0.788)	0.634(0.556–0.712)
2006–2010	0.788(0.698–0.878)	0.556(0.446–0.666)	0.439(0.329–0.549)	0.822(0.738–0.906)	0.663(0.555–0.771)	0.597(0.481–0.713)
2011–2015	0.864(0.799–0.929)	0.617(0.527–0.707)	0.553(0.459–0.647)	0.907(0.852–0.962)	0.704(0.614–0.794)	0.641(0.547–0.735)
2016–2020	0.873(0.810–0.936)	0.723(0.629–0.817)	0.603(0.466–0.740)	0.906(0.851–0.961)	0.761(0.667–0.855)	0.652(0.511–0.793)
Ethnicity
Asian	0.773(0.634–0.912)	0.619(0.454–0.784)	0.583(0.405–0.761)	0.798(0.665–0.931)	0.731(0.580–0.882)	0.675(0.501–0.849)
African-American	0.828(0.732–0.924)	0.634(0.507–0.761)	0.505(0.366–0.644)	0.894(0.814–0.974)	0.744(0.622–0.866)	0.662(0.523–0.801)
Caucasian	0.829(0.790–0.868)	0.640(0.589–0.691)	0.554(0.501–0.607)	0.863(0.828–0.898)	0.704(0.655–0.753)	0.644(0.591–0.697)
Primary sites
Stomach	0.797(0.724–0.87)	0.606(0.516–0.696)	0.536(0.442–0.630)	0.811(0.740–0.882)	0.669(0.581–0.757)	0.612(0.518–0.706)
Small intestine	0.821(0.76–0.882)	0.598(0.520–0.676)	0.476(0.394–0.558)	0.888(0.837–0.939)	0.711(0.633–0.789)	0.615(0.527–0.703)
Colorectum	0.853(0.8-0.906)	0.708(0.637–0.779)	0.608(0.530–0.686)	0.893(0.846–0.940)	0.770(0.701–0.839)	0.700(0.624–0.776)
Esophagus	0.778(0.586–0.97)	0.538(0.303–0.773)	0.404(0.165–0.643)	0.778(0.586–0.970)	0.538(0.303–0.773)	0.462(0.217–0.707)
Pathological differentiation
Well	0.909(0.811–1.007)	0.846(0.723–0.969)	0.784(0.643–0.925)	0.969(0.908–1.03)	0.934(0.846–1.022)	0.898(0.788–1.008)
Moderate	0.924(0.859–0.989)	0.817(0.723–0.911)	0.707(0.595–0.819)	0.939(0.882–0.996)	0.859(0.773–0.945)	0.808(0.710–0.906)
Poor	0.754(0.646–0.862)	0.525(0.4–0.65)	0.399(0.274–0.524)	0.828(0.73–0.926)	0.605(0.476–0.734)	0.476(0.339–0.613)
Undifferentiated	0.755(0.675–0.835)	0.526(0.432–0.62)	0.421(0.327–0.515)	0.820(0.747–0.893)	0.675(0.581–0.769)	0.596(0.494–0.698)
Unknown	0.832(0.779–0.885)	0.638(0.564–0.712)	0.538(0.458–0.618)	0.857(0.806–0.908)	0.685(0.612–0.758)	0.602(0.522–0.682)
Tumor size (cm)
0–5 cm	0.866(0.815–0.917)	0.745(0.676–0.814)	0.668(0.592–0.744)	0.893(0.846–0.940)	0.801(0.738–0.864)	0.742(0.669–0.815)
5–10 cm	0.836(0.783–0.889)	0.598(0.525–0.671)	0.490(0.416–0.564)	0.870(0.821–0.919)	0.675(0.604–0.746)	0.595(0.519–0.671)
>10 cm	0.722(0.634–0.81)	0.533(0.433–0.633)	0.410(0.310–0.510)	0.792(0.710–0.874)	0.635(0.535–0.735)	0.546(0.436–0.656)
Tumor extension
T1	0.926(0.397–1.455)	0.774(0.680–0.868)	0.733(0.627–0.839)	0.945(0.898–0.992)	0.846(0.764–0.928)	0.825(0.735–0.915)
T2	0.830(0.785–0.875)	0.653(0.594–0.712)	0.546(0.483–0.609)	0.873(0.832–0.914)	0.736(0.679–0.793)	0.650(0.587–0.713)
T3	0.742(0.628–0.856)	0.482(0.351–0.613)	0.364(0.235–0.493)	0.776(0.668–0.884)	0.571(0.436–0.706)	0.497(0.356–0.638)
T4	0.667(0.534-0.800)	0.409(0.268–0.550)	0.321(0.188–0.454)	0.740(0.613–0.867)	0.487(0.334–0.640)	0.404(0.251–0.557)
Lymphatic metastasis
Yes	0.728(0.585–0.871)	0.492(0.325–0.659)	0.400(0.235–0.565)	0.772(0.633–0.911)	0.598(0.426–0.77)	0.556(0.376–0.736)
No	0.833(0.798–0.868)	0.650(0.603–0.697)	0.548(0.497–0.599)	0.872(0.841–0.903)	0.723(0.678–0.768)	0.646(0.595–0.697)
Distant metastasis
Yes	0.683(0.558–0.808)	0.318(0.191–0.445)	0.148(0.048–0.248)	0.742(0.620–0.864)	0.398(0.253–0.543)	0.258(0.113–0.403)
No	0.843(0.808–0.878)	0.679(0.632–0.726)	0.584(0.533–0.635)	0.880(0.849–0.911)	0.754(0.709–0.799)	0.678(0.627–0.729)
Surgery
Yes	0.854(0.821–0.887)	0.67(0.625–0.715)	0.565(0.516–0.614)	0.889(0.860–0.918)	0.747(0.704–0.790)	0.665(0.616–0.714)
No	0.359(0.179–0.539)	0.120(0-0.245)	0.120(0-0.245)	0.427(0.227–0.627)	0.171(0.004–0.338)	0
Radiotherapy
Yes	0.801(0.678–0.924)	0.501(0.340–0.662)	0.465(0.302–0.628)	0.801(0.678–0.924)	0.564(0.399–0.729)	0.524(0.353–0.695)
No	0.827(0.792–0.862)	0.650(0.603–0.697)	0.543(0.492–0.594)	0.871(0.838–0.904)	0.731(0.686–0.776)	0.650(0.599–0.701)
Chemotherapy
Yes	0.816(0.722–0.910)	0.461(0.338–0.584)	0.342(0.222–0.462)	0.844(0.756–0.932)	0.503(0.376–0.630)	0.439(0.308–0.570)
No	0.826(0.789–0.863)	0.668(0.621–0.715)	0.570(0.519–0.621)	0.868(0.835–0.901)	0.754(0.709–0.799)	0.671(0.620–0.722)

Prognosis analysis

The Cox proportional hazard model analysis results of patients with GILMS were depicted in Table 4. Prognosis analysis was processed both for OS and CSS. Univariate analysis hinted female, younger age, well differentiation, small tumor size, early T stage, without lymph node and distant metastasis were protective factors both for OS and CSS (p<0.05). Surgery was the beneficial therapeutic measure and chemotherapy may cause poor prognosis (p<0.05). OS related independent protective variable was female (HR = 0.711, 95%CI 0.549–0.921; p = 0.010). Female was also an independent protective factor for CSS (HR = 0.657, 95% CI 0.485–0.890; p = 0.007). CSS related multivariate analysis demonstrated age at 65–79 years (HR = 2.107, 95% CI 1.292–3.435; p = 0.003), age at ≥ 80 years (HR = 3.787, 95% CI 2.198–6.526; p<0.001), poor differentiation (HR = 2.507, 95% CI 1.215–5.175; p = 0.013), T2 (HR = 1.659, 95% CI 1.005–2.740; p = 0.048), T3 (HR = 2.706, 95% CI 1.474–4.967; p = 0.001), T4 (HR = 2.465, 95% CI 1.340–4.534; p = 0.004), distant metastasis (HR = 1.875, 95% CI 1.226–52.867; p = 0.004), no surgery (HR = 3.436, 95% CI 2.048–5.765; p<0.001). Independent prognostic analysis based on OS also observed another risk factor, tumor size>10 cm (HR = 1.771, 95% CI 1.247–2.515; p = 0.001) in addition to the variables mentioned in CSS related independent factors analysis above.

Table 4

univariate and multivariate analysis for patients with GILM (N = 479).
Variables	OS				CSS
	Univariate		Multivariate		Univariate		Multivariate
	HR (95% CI)	p	HR (95% CI)	p	HR (95% CI)	p	HR (95% CI)	p
Gender
Male	Reference		Reference		Reference		Reference
Female	0.622(0.489–0.791)	<0.001	0.711(0.549–0.921)	0.010	0.587(0.442–0.779)	<0.001	0.657(0.485–0.890)	0.007
Age at diagnosis (years)		<0.001		<0.001		<0.001		<0.001
<50	Reference		Reference		Reference		Reference
50–64	1.539(1.018–2.327)	0.041	1.432(0.930–2.205)	0.103	1.411(0.877–2.269)	0.156	1.277(0.779–2.093)	0.333
65–79	2.317(1.551–3.460)	<0.001	2.356(1.544–3.597)	<0.001	2.086(1.316–3.308)	0.002	2.107(1.292–3.435)	0.003
≥80	3.572(2.277–5.602)	<0.001	3.879(2.414–6.234)	<0.001	3.382(2.022–5.656)	<0.001	3.787(2.198–6.526)	<0.001
Years of diagnosis		0.119				0.462
2000–2005	Reference				Reference
2006–2010	1.184(0.865–1.621)	0.292			1.161(0.799–1.688)	1.161
2011–2015	0.94(0.691–1.279)	0.695			0.944(0.657–1.355)	0.944
2016–2020	0.686(0.453–1.039)	0.075			0.771(0.481–1.235)	0.771
Ethnicity		0.944				0.660
Asian	Reference				Reference
African-American	0.952(0.552–1.64)	0.952			0.787(0.403–1.538)	0.484
Caucasian	0.929(0.593–1.453)	0.929			0.973(0.573–1.651)	0.918
Primary sites		0.037		0.587		0.135
Stomach	Reference		Reference		Reference
Small intestine	1.159(0.862–1.559)	0.328	1.058(0.775–1.444)	0.723	1.005(0.708–1.426)	0.979
Colorectum	0.772(0.564–1.056)	0.106	0.855(0.609-1.2)	0.365	0.744(0.518–1.069)	0.110
Esophagus	1.28(0.71–2.306)	0.411	1.01(0.544–1.874)	0.975	1.405(0.737–2.679)	0.301
Pathological differentiation		<0.001		0.001		<0.001		0.010
Well	Reference		Reference		Reference		Reference
Moderate	1.023(0.563–1.858)	0.941	0.98(0.531–1.806)	0.948	1.084(0.513–2.29)	0.832	1.008(0.471–2.158)	0.984
Poor	2.306(1.294–4.109)	0.005	2.044(1.128–3.703)	0.018	2.868(1.411–5.831)	0.004	2.507(1.215–5.175)	0.013
Undifferentiated	2.567(1.502–4.386)	0.001	1.936(1.113–3.37)	0.019	2.638(1.343–5.184)	0.005	1.825(0.911–3.659)	0.090
Unknown	1.703(1.002–2.892)	0.049	1.325(0.767–2.29)	0.313	2.155(1.114–4.168)	0.023	1.6(0.811–3.157)	0.175
Tumor size (cm)		<0.001		0.006		0.002		0.140
0–5 cm	Reference		Reference		Reference		Reference
5–10 cm	1.697(1.279–2.252)	<0.001	1.318(0.971–1.789)	0.076	1.638(1.18–2.272)	0.003	1.295(0.918–1.827)	0.141
>10 cm	2.096(1.516–2.898)	<0.001	1.771(1.247–2.515)	0.001	1.855(1.263–2.726)	0.002	1.492(0.991–2.246)	0.056
Tumor extension		<0.001		0.014		<0.001		0.005
T1	Reference		Reference		Reference		Reference
T2	1.99(1.332–2.971)	0.001	1.563(1.03–2.372)	0.036	2.062(1.269–3.351)	0.003	1.659(1.005–2.740)	0.048
T3	2.928(1.819–4.713)	<0.001	2.228(1.334–3.723)	0.002	3.321(1.888–5.842)	<0.001	2.706(1.474–4.967)	0.001
T4	3.172(1.946–5.171)	<0.001	1.98(1.177–3.333)	0.010	3.701(2.083–6.579)	<0.001	2.465(1.34–4.534)	0.004
Lymphatic metastasis
Yes	Reference		Reference		Reference		Reference
No	1.56(1.049–2.322)	0.028	1.166(0.756–1.799)	0.487	1.611(1.014–2.558)	0.043	1.047(0.635–1.724)	0.858
Distant metastasis
Yes	Reference		Reference		Reference		Reference
No	2.57(1.862–3.545)	<0.001	1.886(1.305–2.726)	0.001	2.625(1.803–3.82)	<0.001	1.875(1.226–2.867)	0.004
Surgery
Yes	Reference		Reference		Reference		Reference
No	4.594(3.045–6.931)	<0.001	3.333(2.083–5.334)	<0.001	4.907(3.064–7.86)	<0.001	3.436(2.048–5.765)	<0.001
Radiotherapy
Yes	Reference				Reference
No	0.754(0.507–1.123)	0.165			0.661(0.424–1.031)	0.068
Chemotherapy
Yes	Reference		Reference		Reference		Reference
No	0.559(0.412–0.759)	<0.001	0.904(0.623–1.311)	0.595	0.49(0.347–0.693)	<0.001	0.77(0.505–1.174)	0.225

GILMS is a rare diagnosis. Only two available cohort of patient studies published focused on GILMS as the rarity of this tumor. A series of 11 GILMS patients published from The American University of Beirut Medical Centre (AUBMC) in 2016^[8]. The relatively small number of cases limited the analysis of histopathological feathers, adjuvant therapy, and outcome though had the largest patient number at that time. In 2021, a lager cases size study reviewed 46 patients at The Royal Marsden Hospital in UK^[7]. This study lacked the information on therapy efficiency, survival months, and prognosis. Some clinicopathological data also need lager papulation-based study to validate. Data mining of rare malignant tumor is convenient in SEER as the giant population size. In the current study, a total of 704 patients with GILMS were retrieved from the SEER database in the past two decades years. This is the largest study that firstly clarified the clinicopathological characteristics, survival, and prognostic factors and vastly eliminated knowledge gaps on behaviour of this rare subtype.

Our study coincidentally covered 20 – years started with the 2000 that distinguished the molecular subtype of LMS from GIST. Diagnosis number of this disease hold steady and rare over this period illustrated the difficulty in the validation of the etiology.

As in LMS in general, the incidence peaks at the seventh decade and overall incidence rate increased with age^[10]. Our study proposed that 68.9% of the patients with GILMS was diagnosed in old age (50–79 years). The median onset age of GILMS patients in this cohort was 64 – year, which was older than age of cases (56 years old) with GILMS in previous studies^{[8, 11]}. Old age was the risk factor of GILMS occurrence. We proposed that female patients (49.5%) had the similar proportion with male (50.5%) based on the largest population study so far. Different from prior reports, our study indicated that esophageal LMS patients were the fewest (18 cases) rather than gastric LMS.

The mean tumor size was 7.7 cm, which was congruent with previous study. Relatively large tumor size often implied high proportion of high tumor extension. Localized tumors (T1) only accounted for 21.9% of GILMS patients. This data was extremely lower compared with two prior studies. Complete excision with negative margins (R0 resection) remains the cornerstone treatment, which offers the best efficiency of cure^[12]. The obvious survival benefits of surgery seen in our cohort was an imperative support for this therapy. 93.9% of GILMS patients were performed with surgery though 78.1% of them were diagnosed in high T stage. GILMS with Large tumor size and high extension may achieving higher rate of surgical R0 resection than LMS in extraperitoneal LMS such as retroperitoneum (65%) as a result of anatomic convenience^[13].

11.5% of patients with GILMS appeared synchronous metastasis at diagnosis in our study. This was lower compared with large retrospective cohorts across STS subtypes (14-26.5%)^[14]occurrence of synchronous metastasis meant extremely low survival time (14.8% 5-y OS, and 25.8% 5-y CSS). The proposal that low likelihood of nodal metastasis was given in published literature^[15]. Our analysis agreed with this opinion as a 7.9% nodal metastasis rate in our cohort. Positive lymph node suggested a touch of survival outcomes (14.8% 5-y OS, and 25.8% 5-y CSS).

66 cases received chemotherapy in this series, the largest number for analysis of chemotherapy in clinical studies. Our data showed chemotherapy can’t bring survival improvement. Conversely, patients with chemotherapy had shorter survival time (5-y OS: 37.2% vs. 57.0%; 5-y CSS: 43.9 vs. 67.1%). This outcome was partly caused by large proportion of high tumor extension in chemotherapy (32.1% vs. 8.4%). Some studies believed that contemporary chemotherapy protocols (whether doxorubicin – based or gemcitabine – based chemotherapy) had no real benefits on survival^{[16, 17]}. There is no established best first-line chemotherapy treatment. We had tried to erase the variables differentiation by propensity – score match (PSM). The PSM result showed no significant survival difference between the two groups (data were not presented in the part of result), consistent with the published literatures. Only 8.6% of patients received radiotherapy, and same no effect as chemotherapy of survival benefits was presented in our study. Besides, radiation exposure may be considered as a putative trigger for the development of GILMS^[18]. There is no established effective radiotherapy treatment.

Pathological differentiation grade, tumor size, and extent of tumor invasion were the three common prognostic factors for STS^{[19, 20]}. In addition, this study first showed that gender, age, synchronous metastasis, and surgery also the significantly independent prognostic factors. Female patients usually had a better prognosis than male. Whereas the potential mechanisms of this results must be explored. Old age meant poor prognosis, which may be caused by the poor physical recovery and severe complications in patients with old age.

Our study had some limitations. First, signs and symptoms that may reflect the invasive range of GILMS were not recorded in the SEER database. Most of the GILMS patients (96%) had their special symptoms such as abdominal pain, bleeding, intussusception, and bowel obstruction^[7]. The presentation of these symptoms was high comparatively with iliocaval leiomyosarcoma and GIST (77–81%)^[21–23]. Discovery of common symptoms for GILMS was important for diagnosis of this sarcoma even for survival outcomes. However, summary of the symptoms was inherently difficult given their variable presentation and rarity. Second, recurrence rate of GILMS after therapy was lacking in SEER database. Recurrence rate after therapy was critical to evaluate the efficiency of various treatment regimens and survival outcome. Half of the patients with GILMS experienced recurrence even were performed with complete oncologic resection^[7]. High rate of distant metastasis with surgery supported the opinion that systematic therapy may be the real cornerstone in future for this anatomic variant of LMS. Last, the names of chemotherapy agent, chemotherapy protocols, and combination therapy regimens and treatment order for surgery and chemoradiotherapy also could not be presented because of the missing of this information in the SEER database.

This study was the largest cohort of patients with GILMS. We firstly and systematically proposed the clinicopathological characteristics, survival, and prognostic factors of GILMS. GILMS was frequently diagnosed in old patients with low rate of nodal and distant metastasis. Whereas most GILMS was presented with large tumor size and high extent of tumor invasion. Complete excision was the unique effective resort to prolong survival outcomes. Chemotherapy and radiotherapy had no benefit on prognosis highlighting the urgent need for better systematic treatment to improve the survival outcomes of patients with GILMS.

Disclosure

The authors declared no conflicts of interest.

Funding

This work was supported by Scientific Research Foundation for Doctor, the First Affiliated Hospital of Kunming Medical University (2021BS017) and Yunnan Provincial Department of Science and Technology Foundation for Youths (202301AU070195).

Author Contribution

Yanan Bao, Xudong Yang, Quanming Zhao: Writing - Original Draft, Conceptualization, Methodology. Formal analysis, Investigation, Resources, Visualization, Validation, Data curation. Wenliang Li: Writing - Review & Editing, Supervision. Visualization, Resources, Validation, Data curation.

Acknowledgement

The authors thank the National Cancer Institute for providing the Surveillance, Epidemiology, and End Results (SEER).

Data Availability

The data that support this study can be found in the SEER database and Supplementary file.

Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin 2022; 72: 7–33. https://doi.org/10.3322/caac.21708
Howlader N NA, Krapcho M, Miller D, Brest A, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin KA (eds). SEER Cancer Statistics Review, 1975–2018. National Cancer Institute April 2021; Bethesda, MD, https://seer.cancer.gov/csr/1975_2018/, based on November 2020 SEER data submission, posted to the SEER web site.
Amant F, Coosemans A, Debiec-Rychter M, Timmerman D, Vergote I. Clinical management of uterine sarcomas. Lancet Oncol 2009; 10: 1188–1198. https://doi.org/10.1016/s1470-2045(09)70226-8
Bathan AJ, Constantinidou A, Pollack SM, Jones RL. Diagnosis, prognosis, and management of leiomyosarcoma: recognition of anatomic variants. Curr Opin Oncol 2013; 25: 384–389. https://doi.org/10.1097/CCO.0b013e3283622c77
Gladdy RA, Qin L-X, Moraco N, Agaram NP, Brennan MF, Singer S. Predictors of Survival and Recurrence in Primary Leiomyosarcoma. Annals of Surgical Oncology 2013; 20: 1851–1857. https://doi.org/10.1245/s10434-013-2876-y
Katz SC, DeMatteo RP. Gastrointestinal stromal tumors and leiomyosarcomas. Journal of Surgical Oncology 2008; 97: 350–359. https://doi.org/https://doi.org/10.1002/jso.20970
Smrke A, Benson C, Strauss DC, Hayes AJ, Thway K, Hallin M, Fisher C, Messiou C, Huang PH, Jones RL, Smith MJ. Gastrointestinal leiomyosarcoma demonstrate a predilection for distant recurrence and poor response to systemic treatments. Eur J Surg Oncol 2021; 47: 2595–2601. https://doi.org/10.1016/j.ejso.2021.04.043
Hilal L, Barada K, Mukherji D, Temraz S, Shamseddine A. Gastrointestinal (GI) leiomyosarcoma (LMS) case series and review on diagnosis, management, and prognosis. Med Oncol 2016; 33: 20. https://doi.org/10.1007/s12032-016-0730-3
Mahul B. Amin SBE, Frederick L. Greene, David R. Byrd, Robert K. Brookland, Mary Kay Washington, Jeffrey E. Gershenwald, Carolyn C. Compton, Kenneth R. Hess, Daniel C. Sullivan, J. Milburn Jessup, James D. Brierley, Lauri E. Gaspar, Richard L. Schilsky, Charles M. Balch, David P. Winchester, Elliot A. Asare, Martin Madera, Donna M. Gress, Laura R. Meyer. AJCC Cancer Staging Manual, Eighth Edition.2017.
George S, Serrano C, Hensley ML, Ray-Coquard I. Soft Tissue and Uterine Leiomyosarcoma. J Clin Oncol 2018; 36: 144–150. https://doi.org/10.1200/jco.2017.75.9845
van Houdt WJ, NS IJ, Marjolein Schrijver A, Huis In 't Veld E, Thway K, Jones RL, Fotiadis N, Hayes AJ, Bruining A, Zavrakidis I, van Coevorden F, Steeghs N, Mathijssen RHJ, Strauss DC, Smith MJF. Oncological Outcome After Diagnostic Biopsies in Gastrointestinal Stromal Tumors: A Retrospective Cohort Study. Ann Surg 2021; 274: e1093-e1098. https://doi.org/10.1097/sla.0000000000003744
Anaya DA, Lev DC, Pollock RE. The role of surgical margin status in retroperitoneal sarcoma. J Surg Oncol 2008; 98: 607–610. https://doi.org/10.1002/jso.21031
Stoeckle E, Coindre JM, Bonvalot S, Kantor G, Terrier P, Bonichon F, Nguyen Bui B. Prognostic factors in retroperitoneal sarcoma: a multivariate analysis of a series of 165 patients of the French Cancer Center Federation Sarcoma Group. Cancer 2001; 92: 359–368. https://doi.org/10.1002/1097-0142(20010715)92:2<359::aid-cncr1331>3.0.co;2-y
Italiano A, Mathoulin-Pelissier S, Cesne AL, Terrier P, Bonvalot S, Collin F, Michels JJ, Blay JY, Coindre JM, Bui B. Trends in survival for patients with metastatic soft-tissue sarcoma. Cancer 2011; 117: 1049–1054. https://doi.org/10.1002/cncr.25538
Sakurai S, Fukasawa T, Chong JM, Tanaka A, Fukayama M. C-kit gene abnormalities in gastrointestinal stromal tumors (tumors of interstitial cells of Cajal. Jpn J Cancer Res 1999; 90: 1321–1328. https://doi.org/10.1111/j.1349-7006.1999.tb00715.x
Beck AH, Lee CH, Witten DM, Gleason BC, Edris B, Espinosa I, Zhu S, Li R, Montgomery KD, Marinelli RJ, Tibshirani R, Hastie T, Jablons DM, Rubin BP, Fletcher CD, West RB, van de Rijn M. Discovery of molecular subtypes in leiomyosarcoma through integrative molecular profiling. Oncogene 2010; 29: 845–854. https://doi.org/10.1038/onc.2009.381
Demetri GD, Schöffski P, Grignani G, Blay JY, Maki RG, Van Tine BA, Alcindor T, Jones RL, D'Adamo DR, Guo M, Chawla S. Activity of Eribulin in Patients With Advanced Liposarcoma Demonstrated in a Subgroup Analysis From a Randomized Phase III Study of Eribulin Versus Dacarbazine. J Clin Oncol 2017; 35: 3433–3439. https://doi.org/10.1200/jco.2016.71.6605
Patel SR. Radiation-induced sarcoma. Current Treatment Options in Oncology 2000; 1: 258–261. https://doi.org/10.1007/s11864-000-0037-6
Pisters PW, Leung DH, Woodruff J, Shi W, Brennan MF. Analysis of prognostic factors in 1,041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol 1996; 14: 1679–1689. https://doi.org/10.1200/jco.1996.14.5.1679
Coindre JM, Terrier P, Guillou L, Le Doussal V, Collin F, Ranchère D, Sastre X, Vilain MO, Bonichon F, N'Guyen Bui B. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer 2001; 91: 1914–1926. https://doi.org/10.1002/1097-0142(20010515)91:10<1914::aid-cncr1214>3.0.co;2-3
Ong CJ, Smith MJ, Snow HA, Benson C, Jones RL, Hayes AJ, Strauss DC. Primary iliocaval leiomyosarcomas: The path beyond surgery. Eur J Surg Oncol 2020; 46: 893–897. https://doi.org/10.1016/j.ejso.2020.01.034
Bümming P, Ahlman H, Andersson J, Meis-Kindblom JM, Kindblom LG, Nilsson B. Population-based study of the diagnosis and treatment of gastrointestinal stromal tumours. Br J Surg 2006; 93: 836–843. https://doi.org/10.1002/bjs.5350
Caterino S, Lorenzon L, Petrucciani N, Iannicelli E, Pilozzi E, Romiti A, Cavallini M, Ziparo V. Gastrointestinal stromal tumors: correlation between symptoms at presentation, tumor location and prognostic factors in 47 consecutive patients. World J Surg Oncol 2011; 9: 13. https://doi.org/10.1186/1477-7819-9-13

No competing interests reported.

rawdata.xlsx

The Clinicopathological Features and Survival Prediction of Gastrointestinal Leiomyosarcoma: A SEER Based Analysis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

Introduction

Methods

Participants

Variables

Statistical analysis

Results

Features of patients with GILMS

Survival analysis

Prognosis analysis

Discussion

Conclusion

Declarations

Disclosure

Funding

Author Contribution

Acknowledgement

Data Availability

References

Additional Declarations

Supplementary Files

Status:

Version 1