The prognostic value of systemic inflammatory response index (SIRI) and CD8 + tumor-infiltrating lymphocytes for patients with localized undifferentiated pleomorphic sarcoma of soft tissue

Abstract

Background

Inflammatory blood markers and tumor-infiltrating lymphocytes (TILs) are associated with the prognosis of various cancers. However, reports on their value as prognostic markers in soft tissue undifferentiated pleomorphic sarcoma (UPS) are limited. We aimed to clarify the predictive value of these markers for the prognosis of patients with UPS, focusing on resectable tumors of the extremities and trunk.

Methods

This retrospective analysis included data from 103 patients with localized UPS in the extremities and trunk. The neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), lymphocyte-monocyte ratio (LMR), and systemic inflammatory response index (SIRI) were calculated, and the median values were determined as cut-off values. Immunohistochemical staining of CD8 + TILs was also performed. Disease-specific overall survival (DOS) and distant metastasis-free survival (DMFS) rates were analyzed using the Kaplan–Meier method, and prognostic factors were identified using the Cox proportional hazards model.

Results

Among the inflammatory blood markers, SIRI was found to be a sensitive prognostic factor, and a high SIRI was associated with worse DOS (P = 0.16) and DMFS (P = 0.09). Although CD8 + TILs were not associated with DOS (P = 0.67), high CD8 + TILs correlated with improved DMFS (P = 0.20). Only tumor size > 10 cm was significantly associated with worse DMFS (P = 0.02) in the univariate analysis, while inflammatory blood markers and CD8 + TIL showed no correlation. The combination of SIRI and CD8 + TILs revealed that high CD8 + TILs in tumor tissue could improve DMFS (P = 0.04) and DOS (P = 0.15) in patients with high SIRI compared to those with low CD8 + TILs and high SIRI.

Conclusions

In patients with localized UPS, CD8 + TILs infiltration into the tumor tissue could improve the prognosis of patients with high SIRI.

Background

Undifferentiated pleomorphic sarcoma (UPS) is a malignant neoplasm of uncertain differentiation characterized by prominent nuclear atypia. UPS accounts for approximately 20% of all soft tissue sarcomas [1] and mainly arises in soft tissue, particularly in the extremities. The mainstay treatment for localized UPS is radical resection of the primary tumor. Standard chemotherapy for UPS is anthracycline and ifosfamide; however, the need for perioperative chemotherapy has been controversial [2–5]. The 5-year overall survival rate of UPS is reported to be 60–80% [6, 7].

Inflammation plays a crucial role in the development and progression of cancer. The systemic inflammatory response detected by blood examination has been reported to be associated with the outcomes of both early- and advanced-stage cancers and is recognized as an easily applicable indicator [8, 9]. The systemic inflammatory marker includes the neutrophil-to-lymphocyte ratio (NLR), the platelet-to-lymphocyte ratio (PLR), the lymphocyte-to-monocyte ratio (LMR), and the systemic inflammatory response index (SIRI) [8–10]. Furthermore, in soft tissue sarcoma, these indicators have been reported to be prognostic biomarkers in localized and advanced settings [11–13]. However, it is unclear whether these indices correlate with inflammatory cells in the tumor microenvironment.

As a major player in the tumor microenvironment, tumor-infiltrating lymphocytes (TILs), particularly cytotoxic CD8 + T cells, have been shown to play an important role against cancer cells [14]. The number of TILs has been reported to correlate with the prognosis of various tumors, including colorectal cancer, gastric cancer, pancreatic cancer, cholangiocarcinoma, and lung cancer [15–20]. In soft tissue sarcomas, a few studies have revealed that CD8 + TILs are prognostic factors [21–24] and that the tumor microenvironment is correlated with the response to immune checkpoint inhibitors [25]. In breast cancer, high levels of CD8 + T cell infiltration are correlated with a good response to neoadjuvant chemotherapy with anthracycline [26]. There are no reports on the relationship between TILs and the efficacy of perioperative chemotherapy for UPS.

In this study, we aimed to clarify the influence of immune markers in the peripheral blood and TILs on the prognosis and the impact of adjuvant chemotherapy on survival rates in patients with UPS with localized and resectable primary lesions.

Methods

Data source

This multicenter, retrospective, observational study was conducted using data from the Higashi-Nihon Orthopaedic and Pediatric Sarcoma (HOPES) Group. Keio University Hospital, National Cancer Center Hospital, Chiba Cancer Center Hospital, and the University of Tokyo Hospital participated in this study. This study was approved by the Institutional Review Board of each institute. For informed consent, information about the study was disclosed using an opt-out procedure, and the opportunity to refuse participation was guaranteed. Information obtained from the medical records and radiographic images was retrospectively analyzed.

Patient selection and data extraction

We included patients under 75 years of age with resectable high-grade UPS without any metastasis to the extremities or trunk wall at our institute between March 2004 and September 2019. We excluded patients with the following criteria: (1) advanced disease (i.e., metastatic at first presentation); (2) low-grade disease; (3) cases treated without radical local therapy, resection, or amputation; (4) cases with the primary anatomical location at the retroperitoneum, peritoneum, thoracic cavity, mediastinum, vertebra, head and neck, and pelvis, due to the difficulty in achieving complete resection with negative margins in these areas and the strong prognostic influence of the anatomical site and immune environment; (5) age > 75 years because perioperative chemotherapy is not usually applicable; (6) positive staining for MDM2 despite a primary diagnosis of UPS, which may indicate dedifferentiated liposarcoma; and (7) cases treated with perioperative chemotherapy other than anthracyclin-based chemotherapy. For each patient, the following data were extracted: age, sex, tumor size, tumor depth, primary tumor location, status at first visit, and therapeutic intervention. The tumor locations were grouped into the upper extremities, lower extremities, and trunk. Tumor size was classified into three categories (< 5 cm, 5–10 cm, and > 10 cm). The absolute values of circulating neutrophils, monocytes, and lymphocytes at the time of diagnosis were collected. NLR was calculated by dividing the absolute neutrophil count by the absolute lymphocyte count. PLR was calculated by dividing the absolute platelet count by the absolute lymphocyte count. LMR was calculated by dividing the absolute lymphocyte count by the absolute monocyte count, and the tumor specimens from the surgery were collected and used for immunohistochemistry. SIRI was calculated by dividing the absolute neutrophil count times the absolute monocyte count by the absolute lymphocyte count. We defined high and low NLR, PLR, LMR, and SIRI based on median values.

In total, 113 cases of undifferentiated pleomorphic sarcoma of the soft tissue were identified, and 103 cases were extracted based on the inclusion criteria (10 cases were excluded because 9 cases were positive for MDM2 and 1 case was treated with gemcitabine and docetaxel as perioperative chemotherapy).

Distant metastasis-free survival (DMFS) and disease-specific overall survival (DOS) were defined as the time from diagnosis to the detection of metastasis and UPS-related death, respectively. For the survival analysis, individuals who were alive at the final follow-up were censored at that time point.

Immunohistochemistry

Surgically resected samples, including the border area of the tumor and normal tissue in the sagittal plane, were stained through Immunohistochemistry. A rabbit anti-CD8 + monoclonal antibody (clone4B11, 1:40; Leica Biosystems, Nussloch, Germany) was used to detect TILs, and an anti-MDM2 monoclonal antibody was used to exclude dedifferentiated liposarcoma from the UPS. Immunohistochemistry was performed using the Ventana BenchMark automated immunostainer (Ventana Medical Systems, Tucson, AZ, USA) with labeled streptavidin-biotin and visualized with 3,3′-diaminobenzidine.

Evaluation of CD8 + T cell infiltration

CD8 + T cell infiltration was evaluated using quantitative digital image analysis. First, immunohistochemically stained slides were digitized using a Nanozoomer 2.0-HT virtual slide scanner (Hamamatsu Photonics, Hamamatsu, Japan). Next, the proportion of stained area within the tumor was measured objectively using Tissue Studio image analysis software (Definiens, Munich, Germany). The tumor area on the slide was circled, and the proportion of stained cells within the circled area was calculated automatically. As there were three cores for each case, the mean of the three values was defined as the density of CD8 + TILs.

Statistical analysis

Fisher’s exact test and the Mann–Whitney U test were used to identify differences between categorical and continuous variables, respectively. DMFS and DOS were estimated using the Kaplan–Meier method, and cumulative survival rates were compared using the log-rank test with P < 0.05 considered to be significant. Univariate analyses were performed using a Cox proportional hazards model to explore potential prognostic factors. The results of the analyses are reported as adjusted hazard ratios (HRs) with 95% confidence intervals and respective P-values (two-sided alpha = 0.05). All statistical analyses were performed using JMP^® 18 (SAS Institute Inc., Cary, NC, USA).

Results

Clinical characteristics and outcomes

Table 1 shows the clinical characteristics of the 103 patients. The median patient age was 66 years (range: 18–90 years), and the median tumor size was 7.5 cm (range: 2–28 cm). Perioperative chemotherapy was administered to 23% of patients, and 97% underwent limb salvage surgery for primary tumors. The median follow-up period was 48 months (IQR: 29–84 months). Local recurrence was observed in 21.4% of all patients, and 31.1% experienced distant metastasis. The 3- and 5-year overall survival (OS) rates were 89.2% and 76.7%, respectively, while the 3- and 5-year DMFS rates were 70.9% and 67.8%, respectively.

Table 1.

Patient demographics and treatment

Characteristics	Number	%
Sex
Male	60	58
Female	43	42
Age, median (range)	66 (18 − 90)
< 65	48	47
≥ 65	55	53
Size, median (range)	7.5 (2 − 28)
< 5 cm	21	20
≥ 5 cm < 10 cm	51	50
≥ 10 cm	31	30
Depth
Superficial	25	24
Deep	78	76
Site
upper extremity	23	22
lower extremity	62	60
trunk	18	18
Surgery
limb salvage	100	97
amputation	3	3
Surgical margin
R0	75	73
R1	26	25
R2	2	2
Chemotherapy	24	23
Radiotherapy	18	17

R0, ; R1, ; R2

Association between clinical characteristics and peripheral blood inflammatory markers and tumor-infiltrating CD8 + lymphocytes (CD8 + TILs)

Table 2 shows the association between clinical characteristics and peripheral blood inflammatory markers. We used the median value as a cut-off value for each peripheral blood inflammatory marker (NLR: 2.92, interquartile range [IQR]: 2.09–4.45; PLR: 16.1, IQR: 13.5–24.7; LMR: 4.03, IQR: 2.56–5.30; SIRI: 1.12, IQR: 0.70–2.08). High NLR was significantly associated with male sex and large tumor size. Although not statistically significant, a low LMR and high SIRI were associated with male sex and large tumor size.

Table 2.

Association of demographic data and peripheral blood inflammatory markers

	NLR					PLR					LMR					SIRI
	High	%	Low	%	P-value	High	%	Low	%	P-value	High	%	Low	%	P-value	High	%	Low	%	P-value
Sex					0.029					0.9					0.13					0.13
Male	27	69	17	45		22	58	22	56		19	49	25	66		25	66	19	49
Female	12	31	21	55		16	42	17	44		20	51	13	34		13	34	20	51
Age					0.13					0.5					0.43					0.43
<65	20	51	13	34		16	42	17	44		15	38	18	47		18	47	15	38
≥65	19	49	25	66		22	58	22	56		24	62	20	53		20	53	24	62
Size					0.03					0.26					0.14					0.06
<5	5	13	10	26		5	13	10	26		10	26	5	13		5	13	10	26
≥5, <10	16	41	21	55		18	47	19	49		20	51	17	45		16	42	21	54
≥10	18	46	7	18		15	39	10	26		9	23	16	42		17	45	8	21
Depth					0.44					0.83					0.83					0.38
Superficial	10	26	7	18		8	21	9	23		9	21	8	23		10	26	7	18
Deep	29	74	31	82		30	79	30	77		30	79	30	77		28	74	32	82
Site					0.44					0.79					0.23					0.79
Upper extremity	7	18	10	26		8	21	9	23		6	15	11	29		8	21	9	23
Lower extremity	27	69	21	55		25	66	23	59		25	62	23	61		25	66	23	59
Trunk	5	13	7	18		5	13	7	18		8	16	4	10		5	13	7	18
Surgery					0.54					0.54					0.54
limb salvage	38	97	36	95		36	95	38	97		38	97	36	95		37	97	37	95	0.57
amputation	1	3	2	5		2	5	1	3		1	3	2	5		1	3	2	5
Surgical margin					0.45					0.49					0.37
R0	29	74	26	68		28	74	27	69		29	74	26	68		29	76	26	67	0.37
R1	10	26	11	29		10	26	11	27		9	23	12	32		9	24	12	31
R2	0	0	1	3		0	0	1	1		1	3	0	0		0	0	1	2
CD8+TILs high or low					0.99					0.11					0.25					0.82
CD8+TILs high	19	49	18	49		15	39	22	58		21	55	16	42		18	47	19	50
CD8+TILs low	20	51	19	51		23	61	16	42		17	44	22	58		20	53	19	50
Chemotherapy					0.47					0.39					0.84					0.74
+	11	28	8	21		11	29	8	21		10	26	9	24		10	26	9	23
-	28	72	30	79		27	71	31	79		29	74	29	76		28	74	30	77

NLR Neutrophil-lymphocyte ratio, PLR, Platelet-lymphocyte ratio, LMR Lymphocyte-monocyte ratio, SIRI Systemic inflammatory response index

Table 3 shows the association between clinical characteristics and CD8 + TILs. We used the median value as a cut-off value of CD8 + TILs (median: 45.6 [IQR 10.4–110]). High CD8 + TILs (≥　45.6) tended to be associated with superficial tumor localization, while low CD8 + TILs (< 45.6) were associated with deep localization. Apart from peripheral blood inflammatory markers, CD8 + TILs were not related to sex and tumor size.

Table 3.

Association between patient characteristics and CD8 + TILs

	CD8 + TILs high	%	CD8 + TILs low	%	P-value
Sex					0.84
Male	29	58	28	56
Female	21	42	22	44
Age					0.42
< 65	25	50	21	42
≥ 65	25	50	29	58
Size					0.68
< 5	11	22	8	16
≧ 5, < 10	25	50	25	50
≧ 10	14	28	17	34
Depth					0.06
Superficial	16	32	8	16
Deep	34	68	42	84
Site					0.82
Upper extremity	10	20	12	24
Lower extremity	32	64	29	58
Trunk	8	16	9	18
Surgery					0.04
limb salvage	50	100	47	94
amputation	0	0	3	6
Surgical margin					0.9
R0	37	74	35	70
R1	12	24	14	28
R2	1	2	1	2
Chemotherapy
+	8	16	16	32	0.06
-	42	84	34	68

CD8+TILs Tumor-infiltrating CD8+ lymphocytes

Association between clinical outcomes and peripheral blood inflammatory markers and CD8 + TILs

Figure 1 shows the association between the clinical outcomes and peripheral blood inflammatory markers. Among the peripheral blood inflammatory markers, SIRI was found to be a sensitive prognostic factor for discriminating between DOS and DMFS (P = 0.16 and 0.09, respectively). Although CD8 + TILs were not associated with DOS (P = 0.67), high CD8 + TIL levels tended to be associated with better DMFS (P = 0.20) (Fig. 2a and b).

Fig. 1.

Disease-specific overall survival (DOS) and distant metastasis-free survival (DMFS) in high- and low-value groups of each peripheral blood inflammatory marker

Fig. 2.

Disease-specific overall survival (DOS) and distant metastasis-free survival (DMFS) in high- and low-value groups of CD8 + TILs (a, b) and in combination with SIRI and CD8 + TILs (c, d)

Correlation between peripheral blood inflammatory markers and CD8 + TILs

Table 2; Fig. 3 show the correlations between peripheral blood inflammatory markers and CD8 + TILs. NLR, PLR, and SIRI were not correlated with CD8 + TILs. Only the LMR had a weak positive correlation with CD8 + TILs (Fig. 3). However, peripheral blood inflammatory markers and CD8 + TILs were not significantly correlated in the high- and low-dose groups (Table 2).

Fig. 3.

Correlation between CD8 + TILs and peripheral blood inflammatory markers

Prognostic factors of patients with UPS

Table 4 shows the prognostic factors of patients with UPS. Univariate analysis of DMFS revealed that only tumor size larger than 10 cm was a significantly worse prognostic factor (P = 0.02). Deep location of the tumor (P = 0.21), high SIRI (P = 0.1), and low CD8 + TILs (P = 0.21) tended to be worse prognostic factors for DMFS. In the univariate analysis of DOS, there was no significant prognostic factor; however, tumor size larger than 10 cm (P = 0.13) and high SIRI (P = 0.18) tended to be associated with worse DOS. Perioperative chemotherapy was not associated with DOS and DMFS (P = 0.94 and 0.41, respectively).

Table 4.

Prognostic factors of disease-specific overall survival (DOS) and distant metastasis-free survival (DMFS) by univariate analysis

	Disease-specific overall survival			Distant metastasis free survival
	Univariate analysis			Univariate analysis
	HR	95% CI	P-value	HR	95% CI	P-value
Age
<65	1			1
>65	0.74	0.3−1.83	0.52	0.91	0.46−1.83	0.8
Sex
Female	1			1
Male	1.52	0.6−3.8	0.38	1.21	0.75−1.94	0.43
Primary site
Upper limb	1			1
Lower limb	2.45	0.56−10.8	0.24	1.48	0.6−3.66	0.4
Trunk	2.44	0.45−13.33	0.3	1.17	0.36−3.84	0.8
Size
<5cm	1			1
≥5cm <10cm	1.77	0.38−8.14	0.47	1.98	0.57−6.88	0.28
≥10cm	3.32	0.72−15.37	0.13	0.02	1.22−14.63	0.02
Depth
Superficial	1			1
Deep	1.95	0.57−6.69	0.29	1.84	0.71−4.78	0.21
SIRI
Low	0.45	0.14−1.43	0.18	1.99	0.88−4.51	0.1
HIgh	1			1
CD8 tumor
Low	1			1
High	0.82	0.34−2	0.67	0.63	0.3−1.29	0.21
SIRI+CD8+TILs
High+Low	1			1
High+High	0.37	0.1−1.44	0.15	0.4	0.14−1.17	0.09
Low+Low	0.24	0.05−1.16	0.08	0.34	0.17−0.97	0.04
Low+High	0.39	0.08−1.9	0.25	0.27	0.07−0.96	0.04
Chemotherapy
-	1			1
+	1.03	0.4−2.71	0.94	1.37	0.65−2.9	0.41

HR Adjusted hazard ratios, CI Confidence intervals

Combination of SIRI and CD8 + TILs

According to the analysis of prognostic factors in patients with UPS, high SIRI tended to correlate with a worse prognosis, and high CD8 + TILs correlated with a better prognosis. Additionally, there was no correlation between SIRI and CD8 + TILs. Therefore, we inferred that the combination of SIRI and CD8 + TILs may improve prognostic accuracy. As shown in Fig. 2, a high SIRI was associated with worse DOS and DMFS. Dividing the four groups by the combination of SIRI and CD8 + TILs, high SIRI and low CD8 + TILs were associated with a worse prognosis (3- and 5-year DOS of 80%, respectively, and 60%, 1- and 3-DMFS of 55% and 44%, respectively); high CD8 + TILs improved the prognosis of patients with high SIRI (3- and 5-year DOS of 94% and 81%, respectively, and 1- and 3-DMFS of 77% and 71%, respectively) comparable to that of the low SIRI groups (Fig. 4).

Fig. 4.

Association of CD8+TILs and impact of perioperative chemotherapy. Disease-specific overall survival (DOS) and distant metastasis-free survival (DMFS) in low- (a,c) and high- (b,d) value groups of CD8 + TILs

Impact of CD8 + TILs on the efficacy of perioperative chemotherapy

Perioperative chemotherapy did not improve the prognosis of UPS patients in our cohort. CD8 + TILs tended to be associated with a better OS. We analyzed the effect of CD8 + TILs on the efficacy of perioperative chemotherapy. In the low CD8 + TILs group, perioperative chemotherapy did not improve the DOS. Contrastingly, in the high CD8 + TILs group, perioperative chemotherapy improved the DOS of patients with UPS (3-, 5-, and 10-year DOS rates of 88%, 88%, and 88%, respectively, vs. 3-, 5-, and 10-year DOS rates of 88%, 80%, and 70%, respectively).

Discussion

The present study analyzed the prognostic factors of soft tissue UPS, focusing on inflammatory blood markers and the tumor immune microenvironment and limiting the cases to localized, high-grade, and anatomical sites appropriate for curative resection. The high SIRI group had a poorer prognosis, and SIRI did not reflect CD8 + TILs in the tumor tissue. It is striking that high levels of CD8 + TILs in tumor tissues could improve the dismal prognosis of the group with a high SIRI.

Our primary finding was that CD8 + TILs improved the prognosis of patients with UPS and high SIRI values. A few studies have reported an association between UPS and systemic inflammatory markers, including NLR and PLR, and these studies had different inclusion criteria: some studies included patients with distant metastasis at first presentation, while others included only patients treated with surgical resection [27, 28]. One study demonstrated that a high NLR was an independent prognostic factor for worse OS but not disease-free survival [27]. However, another report showed that high PLR and NLR were not independent worse prognostic factors of OS, although univariate analysis showed an association between high NLR and PLR and worse OS [28].

We did not find a significant association between NRL, PLR, LMR, OS, and DMFS, although a high SIRI tended to be associated with worse OS. Previous reports have shown that distant metastasis at first presentation and OS are clearly associated with a high NLR [27, 28]. In our study, we included patients with tumors in the extremities and trunk wall and no distant metastasis at the first presentation, which may be one of the reasons why inflammatory markers were not significantly associated with prognosis. CD8 + TILs in tumor tissue were reported to be associated with prognosis in several cancers [15–20]. In soft tissue sarcomas, there have been a few controversial reports on immune cell infiltration in the tumor tissue as a prognostic factor [22–24]. Even in UPS, the prognostic value of high CD8 + TIL infiltration in tumor tissue was controversial, according to a few reports [22–24]. In our study, high levels of CD8 + TILs in tumor tissues, focusing on tumor margins, tended to be associated with better OS, although this was not statistically significant. One reason for the weak prognostic value of CD8 + TILs infiltration is that the CD8 + TILs count was relatively low in our study. In other reports, the median CD8 + TILs count in the tumor tissue was >100 cells/mm² [23]; however, in our cases, it was 45.6 cells/mm². This could mean that our cases had low tumor immune activity.

Although inflammatory blood markers and CD8 + TILs infiltration into the tumor tissue are not independent prognostic markers, a combination of these factors could be a prognostic factor. A high SIRI value tended to result in worse OS; however, CD8 + TIL infiltration into the tumor tissue could improve OS. This finding of a combination of inflammatory blood markers and TILs was the first report analyzing UPS and was comparable with previous reports from studies on other cancers, including laryngeal squamous cell carcinoma, uterine leiomyosarcoma, and triple-negative breast cancer [29–31]. Recently, prognostic nomograms such as SARCULATOR [5] are increasingly utilized to predict soft tissue sarcoma, and the incorporation of inflammatory markers and immune cell infiltration, including CD8 + TILs, could improve their predictive ability.

In our cases, a high NLR was significantly associated with older age and larger tumor size, which is consistent with a previous report showing that a high NLR was associated with older age, larger tumor size, higher histologic grade, higher American Joint Committee on Cancer T stage, and metastasis at presentation [27, 28]. A high SIRI and low LMR tended to be significantly associated with older age and larger tumor size. These findings indicate that cancer-related inflammation can increase in cases with aggressive tumor features, although the underlying mechanism is not fully understood. Additionally, high levels of CD8 + TILs were associated with a superficial location, which tended to have a better prognosis. Interestingly, other reports analyzing the immune cell composition in the UPS showed that CD8 + TILs were inversely correlated with tumor size [22, 24], and high CD4 + Programmed cell death protein-1 high TILs were associated with superficial tumor location [22]. These findings are not compatible with our results; however, it is interesting to note that immune profiling in tumor tissues could change according to tumor enlargement and location.

In the high CD8 + TILs group, perioperative chemotherapy tended to prolong DOS compared to surgery alone. TILs are associated with sensitivity to perioperative chemotherapy in other cancers, including breast, rectal, and cervical cancers [32–35]. In our cases, the effect of CD8 + TILs on perioperative chemotherapy was not statistically significant, and one of the causes was that the number of patients receiving perioperative chemotherapy was too small to evaluate its efficacy. To date, only one report has been examined on the association between baseline immunity and the impact of chemotherapy in UPS [24]. It revealed that CD8 + TILs infiltration correlated with improved survival but paradoxically predicted a poorer response to neoadjuvant anthracycline-based chemotherapy, although a very small number of cases were analyzed. Since this study analyzed not only CD8 + TILs but also other immune cells, this paradox may be explained by the presence of immunosuppressive cells such as M2 macrophages, which are known to be key suppressors of T cell function [24]. In our analysis, we could not evaluate response to chemotherapy; this underscores the need for further studies with larger cohorts to evaluate the interplay between the immune environment and the efficacy of chemotherapy to identify groups that benefit from perioperative chemotherapy.

In addition, a recent phase 2 study revealed the efficacy of immune checkpoint inhibitors as neoadjuvant therapy coupled with radiotherapy for UPS [36]. Future investigations should explore the efficacy of integrating immune checkpoint inhibitors with or without perioperative chemotherapy or radiotherapy tailored to the immune environment in tumor tissue. Notably, single-cell analysis for UPS has revealed that CD8 + TILs presence alone is insufficient to predict the responses of immune checkpoint inhibitors [37]. Further analysis with a large number of cases using single-cell analysis or spatial transcriptome analysis is warranted to elucidate cell-cell interactions and improve predictive models for the efficacy of immune checkpoint inhibitors.

This study has some limitations that should be considered when interpreting the results. First, we did not analyze immune microenvironments other than CD8 + TILs. Previous reports showed that CD3 + TILs were independently associated with better OS [23], and SIRPα-positive macrophage in tumor tissue was associated with worse OS in patients with UPS [38]. Second, this study was a retrospective analysis, and a number of patients who met the inclusion criteria were excluded because of the lack of blood counts prior to treatment, metastasis at first presentation, and anatomical locations other than the extremities and trunk wall, which limited the sample size. Several factors tended to be associated with the prognosis, and a larger sample size would have increased the statistical power. Third, the indications for perioperative chemotherapy depend on the institution and physician. These limitations could be resolved in a prospective study that would allow for a more standardized approach to determine the timing of blood counts and indications for perioperative chemotherapy. Despite these limitations, this is the first report focusing on localized and resectable cases of UPS to analyze the correlation and prognostic value of inflammatory blood markers and the tumor immune microenvironment.

Conclusions

In conclusion, our analysis of localized and resectable soft tissue UPS indicated that high SIRI was most strongly associated with poorer prognosis among systemic inflammatory markers. Notably, even in patients with high SIRI, CD8 + TILs infiltration could improve prognosis. To the best of our knowledge, this is the first retrospective analysis to evaluate the prognosis of soft-tissue UPS using a combination of inflammatory blood markers and TILs. In addition, CD8 + TILs may serve as a predictive marker for the efficacy of perioperative chemotherapy in UPS, although contradictory findings have been reported. Further comprehensive analyses of the immune microenvironment, such as single-cell analysis and spatial transcriptome analysis, could improve our understanding of the mechanisms and prognostic value of immune cells and their interaction with other cells in tumor tissues in cases of soft-tissue UPS.

Abbreviations

DMFS

Distant metastasis-free survival

DOS

Disease-specific overall survival

IQR

Interquartile range

HOPES

Higashi-Nihon Orthopaedic and Pediatric Sarcoma

HRs

Hazard ratios

LMR

Lymphocyte-to-monocyte ratio

NLR

Neutrophil-to-lymphocyte ratio

OS

Overall survival

PLR

Platelet-to-lymphocyte ratio

SIRI

Systemic inflammatory response index

TILs

Tumor-infiltrating lymphocytes

UPS

Undifferentiated pleomorphic sarcoma

Authors’ contributions

HK wrote the paper and performed the literature review. RN, TU, TH, AK, TY, and ST contributed to the conception and design of the study and critically revised the manuscript. YN, SI, TM, and HK collected the clinical data. HA and NM contributed to pathological diagnosis. All the authors have read and approved the final version of this manuscript.

Funding

Not applicable.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of the University of Tokyo Hospital (protocol codes 2021263NI and November 18, 2021). Informed consent from each patient was waived in this study because data was anonymous data.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.