Abstract
Background
Solitary fibrous tumors (SFTs) are rare fibroblastic neoplasms for which prognostication is challenging. Peripheral hematological indices such as lymphocyte-monocyte ratio (LMR) and neutrophil-lymphocyte ratio (NLR) have demonstrated prognostic value in various malignancies but remain incompletely characterized in localized SFTs.
Objectives
This study aimed to investigate clinicopathological factors and peripheral blood counts as predictors of survival outcomes in patients with SFT; and to explore their relationship with intratumoral gene expression profiles.
Design
Retrospective cohort study.
Methods
We analyzed patients with localized SFT managed at the National Cancer Centre Singapore (n = 77). Cut-offs for indices were derived from prior studies, and survival outcomes were analyzed using the Kaplan-Meier method and Cox proportional-hazards regression. Subgroup analysis with the NanoString PanCancer IO360 panel (n = 9) was done to explore associations between peripheral LMR and intratumoral immune-oncologic signals.
Results
Median age at diagnosis was 55.7 years (range: 12.7 to 86.5 years) with a median follow-up time of 10.1 years. Low LMR (≤ 2.4) was observed in 14 patients (18.2%) and was significantly associated with poorer overall survival (OS) (HR 6.73, 95% CI 1.65 – 27.4, p = 0.0078) alongside older age (> 65 years) and absence of curative surgery. Low LMR was also associated with poorer metastatic-free survival (MFS) (HR 5.11, 95% CI 1.48 – 17.70, p = 0.0101) and trended toward worse event-free survival (EFS) (HR 2.76, 95% CI 0.92 – 8.31, p = 0.071). LMR-low was significantly correlated with lower lymphocyte counts (median: 1.23 vs 1.94×109/L; p < 0.0001) and higher monocyte counts (median: 0.66 vs 0.51×109/L; p = 0.0041). Using NanoString analysis, LMR was negatively correlated with intratumoral cytokine and chemokine signaling pathway scores (rho = -0.733, p = 0.0246), mast cell scores (rho = -0.700, p = 0.0358) and CD8 T cells (rho = -0.667, p = 0.0499).
Conclusion
LMR may hold prognostic value for SFTs, though further validation in larger, multi-institutional cohorts is warranted.
Keywords: sarcoma, prognostic biomarker, gene expression profiling, medical oncology, tumor microenvironment
Plain language summary
Solitary fibrous tumors (SFTs) are rare soft tissue tumors that can behave unpredictably. Some patients do very well after surgery, while others experience recurrence or shortened survival. Doctors currently have limited tools to accurately predict which patients are at higher risk. This study examined whether simple blood test results taken before treatment could help predict outcomes in patients with localized SFT. We focused on a measure called the lymphocyte monocyte ratio (LMR), which reflects the balance between two types of white blood cells involved in immune responses. Similar blood-based markers have been shown to predict outcomes in other cancers, but they have not been well studied in SFT. We analyzed 77 patients treated at the National Cancer Centre Singapore and followed them for a median of 10 years. Patients with a low LMR had significantly worse overall survival compared to those with higher LMR. Low LMR was also associated with lower lymphocyte counts and higher monocyte counts. There was also a trend toward earlier recurrence in patients with low LMR. In a smaller subgroup of tumors, we examined gene expression patterns to better understand the biology behind these findings. Lower LMR was linked to differences in immune-related signals within the tumor, suggesting that blood test results may reflect changes in the tumor immune environment. Overall, our findings suggest that LMR, a simple and widely available blood test marker, may help identify patients with SFT who are at higher risk of poor outcomes. Larger studies are needed to confirm these results.
Introduction
Solitary fibrous tumors (SFTs) are rare soft tissue tumors of fibroblastic origin. 1 These are classified as intermediate soft tissue tumors under the 2020 World Health Organization (WHO) classification. 2 They arise across a variety of anatomical regions, most classically within the pleura but also extra-thoracically within the head and neck, abdomen, visceral organs and soft tissue.3,4 These tumors have a distinct histopathological appearance and contain characteristic NAB2-STAT6 gene fusions on chromosome 12, which underlie the basis of molecular confirmation for SFT diagnosis. 5
SFTs demonstrate a wide spectrum of clinical behavior, ranging from indolent tumors with favorable long-term outcomes to highly aggressive tumors characterized by local recurrence, metastasis, and mortality. 6 While classified as tumors with intermediate malignant potential that rarely metastasize, the risk of metastasis is reportedly up to 45% over a ten-year period. 7 On histopathological examination, tumors which appear to have benign features may paradoxically recur or metastasize.8,9 The converse is also true as some SFTs with malignant features eventually run an indolent course. Hence, current prognostic modalities for SFTs are limited in their ability to predict for clinical course and outcomes. 10 Contemporary prognostication for SFT relies largely on clinicopathological features. 11 Increased tumor size, mitotic activity, presence of tumor necrosis, nuclear atypia are several factors consistently associated with adverse clinical outcomes. Additionally, positive resection margins also significantly increase the risk of recurrence. 12
Though histological appearance of SFT alone may not reliably predict clinical behavior, risk stratification models applied at the time of diagnosis may confer prognostic relevance at the time of metastatic spread. This is supported by prospective data from clinical trials of patients with metastatic disease arising from “low-risk” SFTs showing better progression-free survival (PFS) than those stemming from “high-risk” tumors, highlighting the biological significance of baseline risk classification.13,14
Current clinical indicators predicting the risk of metastasis in extra-meningeal SFTs include the criteria by Demicco et al., incorporating age, tumor size, and mitotic count in the classification of low, intermediate and high-risk groups.15,16 While several other risk stratification models for SFTs exist, their predictive ability remains limited. Current risk stratification models are predominantly grounded on static histological and clinical features; however, these may not fully reflect host-tumor interactions that may be dynamic.11,15,17,18 Secondly, there also remains little integration of immunological or molecular biomarkers that could augment current predictive accuracy. 19 Thus, there is an unmet need for novel, reproducible and clinically accessible biomarkers to improve personalized prognostication in SFTs.
There has been increasing evidence in existing literature showing that systemic inflammatory indices such as the lymphocyte-monocyte ratio (LMR), neutrophil-lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR) are inexpensive and readily accessible prognostic biomarkers across various malignancies, reflecting the interplay between pro-tumor inflammatory activity and anti-tumor immunity. In particular, low LMR, reflecting relative lymphopenia and monocytosis, has consistently been associated with inferior survival outcomes in both solid and hematological malignancies. Previously within a cohort of patients with SFT with metastatic disease on pazopanib treatment, NLR and red cell distribution width (RDW) were found to be independent predictors of poorer outcomes. There is growing appreciation of peripheral indices of systemic inflammation in various cancers, yet their role in the prognostication of localized SFT remains relatively unexplored. 20
To this end, this study aims to bridge this gap in the risk stratification of SFTs by investigating both clinicopathological factors and peripheral hematological indices of systemic inflammation such as LMR, NLR and PLR as predictors of survival outcomes in patients with SFT. Additionally, we aim to explore associations between these markers of systemic inflammation and the intratumoral gene expression profile within a subset of these patients. Using this approach, we hope to offer a deeper insight into SFT biology and provide a basis for future prospective validation.
Materials and methods
Study design
We conducted a retrospective cohort study involving 126 patients diagnosed with SFT managed at National Cancer Centre Singapore (NCCS) from 1992 to 2023. The reporting of this study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement for cohort studies. 21 The completed STROBE checklist is provided in Supplementary Table S1. Patients with histologically confirmed SFT reviewed by certified pathologists underwent review of their electronic medical records using Sunrise Citrix Manager. Patient visits were retrieved and clinical notes were accessed to obtain information pertaining to patient demographics, clinicopathological data, pathological reports and laboratory reports as outlined further below. A total of 77 patients with localized SFT who had pre-treatment complete blood counts at the time of diagnosis – that is prior to the initiation of any treatment modality such as surgery, chemotherapy, and/or radiotherapy whether for palliative or curative intent – were included in the final cohort for analysis (Figure S1). Patients with evidence of active infection, hematological disorders, or immunosuppressive medication intake, as verified by clinical information available in the patient notes at the time of blood draw were excluded. All patients meeting inclusion criteria during the study period were included due to disease rarity. Clinicopathological information collected included sex, age, ethnicity, primary tumor site, tumor size and tumor depth at the time of diagnosis, comorbid conditions and the presence of curative surgery – defined as surgical resection done with curative intent resulting in R0/R1 resection. Conversely, the absence of curative surgery included R2 resection (e.g. tumor debulking) or no surgical treatment done. Demographic characteristics of patients (e.g. sex, ethnicity and date of birth) were verified using their National Registry Identification Card number. Primary tumor site was categorized into anatomical regions such as abdominal/pelvis, head (extracranial) and neck, intracranial, limb/trunk, and pulmonary/pleural. Tumor size was derived from the largest diameter measured from radiological imaging or resected pathological specimens. Tumor depth was dichotomized into deep and superficial with deep tumors extending beyond the superficial fascia and vice versa for superficial tumors. This study was done with the approval of the SingHealth Centralised Institutional Review Board and informed consent was obtained.
Analysis of peripheral hematological indices of inflammation
Lymphocyte-monocyte ratios (LMR), neutrophil-lymphocyte ratios (NLR) and platelet-lymphocyte ratios (PLR) were calculated by dividing the absolute lymphocyte counts by monocyte counts, absolute neutrophil counts by lymphocyte counts, and platelet counts by lymphocyte counts from pre-treatment complete blood counts respectively. Cut-offs for LMR (≤ 2.4), NLR (> 2.5) and PLR (> 182) as univariable predictors of overall survival (OS) were derived from a prior study on soft tissue sarcomas (STS), including SFT, to maintain methodological consistency and comparability. 22 Receiver operating characteristic (ROC) analyses were explored within the cohort. However, internally derived cut-offs had poor AUC values and suboptimal sensitivity and specificity and therefore were not used for primary analyses.
Gene expression profiling
The NanoString PanCancer IO360 panel (NanoString Technologies, Seattle, WA, USA) was used to assess the gene expression profiles of formalin-fixed paraffin embedded tissue of 9 patients, as per manufacturer’s protocol. Briefly, RNA was extracted from tumor tissue and analyzed with 2100 Bioanalyser (Agilent Technologies, Palo Alto, CA, USA) to ensure adequate RNA integrity and content. Samples passing quality control would be evaluated using the nSolver 4.0 Advanced Analysis module to derive differentially expressed genes, pathway scores and cell type scores.
Statistical analysis
Normality of LMR, NLR and PLR was analyzed with Kolmogorov-Smirnov test. Comparisons between categorical variables were performed with Pearson’s Chi-squared tests. Comparisons between non-normally distributed numerical variables were performed with Mann-Whitney U tests. LMR, NLR and PLR were analyzed as univariable predictors of OS, which was defined by the interval period from the date of SFT diagnosis to death of any cause. MFS and EFS were used as secondary survival endpoints and determined by the interval period between the date of SFT diagnosis to metastasis or death; and the date of SFT diagnosis to disease recurrence, metastasis or death of any cause respectively. For survivors, survival was censored at the date of last follow-up. Kaplan-Meier analyses were conducted to identify statistically significant univariable predictors of OS, EFS and MFS, with differences between groups being assessed with log-rank tests. Hazards ratios reported with 95% confidence intervals were calculated using Cox proportional hazards regression. Multivariate Cox regression via backwards procedure was used to assess the independence of statistically significant factors identified in univariate analysis. Spearman’s rank order correlation was used to analyze the relationship between LMR and both immune-oncogenic signaling pathways and cell type from NanoString analysis. All statistical analyses were performed using two-tailed tests with p = 0.05 unless stated otherwise. Statistical tests were performed with MedCalc version 23.3.7 for Windows (MedCalc Software Ltd, Belgium).
Results
Patient demographics and clinical characteristics
Of 126 identified patients, 77 met the inclusion criteria. Patients were excluded due to metastatic disease, absent blood counts or confounding hematologic conditions (Figure S1). Missing data was minimal and limited to excluded variables. The median age at diagnosis for the analyzed cohort was 55.7 years (range: 12.7 to 86.5 years) with a median follow-up time of 10.1 years. Forty were male (51.9%) and 37 were female (48.1%). There were 25 older patients above 65 years (32.5%) compared to 52 patients below or equal to 65 years (67.5%). Most of the patients were ethnic Chinese (75.3%) among other ethnicities reported such as Malays, Indians and Eurasians (24.7%). SFT localized to the limb/trunk was the most common anatomical site for the primary tumor (29.9%), with other sites including abdominal/pelvic (20.8%), pulmonary/pleura (19.5%), head (extracranial)/neck (18.2%) and intracranial (11.7%). Median tumor size was 5.55 cm (range: 1.40 to 30.0 cm) and the majority of tumors were > 5 cm (57.1%). Additionally, most patients had deep tumors (85.7%) that were located beneath the superficial fascia. Cardiovascular comorbidities were present in nearly half of the cohort (48.1%) while only 6.5% had rheumatological disease. A minority of patients (6.5%) did not undergo curative surgery due to medical (e.g. unresectable due to anatomical location) or personal reasons. Most patients (84.4%) underwent curative surgery as the only treatment modality whilst 7.8% had adjuvant radiotherapy following surgery (Supplementary Table S2). Median time elapsed between pre-treatment peripheral blood test and surgery was 6 days (range: 0 to 510 days). Within the cohort, LMR (median: 3.59; range: 0.96 to 11.44), NLR (median: 2.57; range 0.81 to 10.65) and PLR (median: 147.2; range: 48.9 to 598.1) did not follow normal distributions (p = 0.0044 for LMR and p < 0.0001 for NLR and PLR). Cut-offs for dichotomizing LMR, NLR and PLR into high versus low groups were derived from a previous study on patients of a soft tissue sarcoma cohort. 22 The cut-offs for LMR, NLR and PLR-high were > 2.4, > 2.5 and > 182 respectively.
Clinicopathological correlates
LMR-low was significantly associated with older age at diagnosis (p = 0.0006), NLR-high (p = 0.0001) and PLR-high (p = 0.0001) but not associated with sex, ethnicity, primary tumor site, size and depth, presence of cardiovascular or rheumatological comorbidities and presence of curative surgery (Table 1). Compared to LMR-high patients at diagnosis, LMR-low patients tended to have significantly lower levels of lymphocytes (median: 1.23 vs 1.94×109/L; p < 0.0001) and significantly higher levels of monocytes (median: 0.66 vs 0.51×109/L; p = 0.0041) (Figure 1). Additionally, neutrophils were significantly elevated in LMR-low patients (median: 5.25 vs 4.18 ×109/L; p = 0.0093) but there were no significant changes for platelet levels.
Table 1.
Clinical and demographic characteristics of patients (n = 77).
| Characteristic (n) | LMR at diagnosis (%) | p | |
|---|---|---|---|
| > 2.4 | ≤ 2.4 | ||
| Total (77) | 63 (81.8%) | 14 (18.2%) | |
| Sex | |||
| Male (40) | 34 (85.0%) | 6 (15.0%) | 0.4546 |
| Female (37) | 29 (78.4%) | 8 (21.6%) | |
| Age at diagnosis (years) | |||
| > 65 (25) | 15 (60.0%) | 10 (40.0%) | 0.0006 |
| ≤ 65 (52) | 48 (92.3%) | 4 (7.7%) | |
| Ethnicity | |||
| Chinese (58) | 47 (81.0%) | 11 (19.0%) | 0.7569 |
| Others (19) | 16 (84.2%) | 3 (15.8%) | |
| Primary tumor site | |||
| Abdominal/Pelvic (16) | 14 (87.5%) | 2 (12.5%) | 0.1391 |
| Head (Extracranial)/Neck (14) | 11 (78.6%) | 3 (21.4%) | |
| Intracranial (9) | 8 (88.9%) | 1 (11.1%) | |
| Limb/Trunk (23) | 21 (91.3%) | 2 (8.7%) | |
| Pulmonary/Pleura (15) | 9 (60.0%) | 6 (40.0%) | |
| Tumor size (cm) | |||
| ≥ 5 (44) | 33 (75.0%) | 11 (25.0%) | 0.0679 |
| < 5 (33) | 30 (90.9%) | 3 (9.1%) | |
| Tumor depth | |||
| Deep (66) | 53 (80.3%) | 13 (19.7%) | 0.4015 |
| Superficial (11) | 10 (90.9%) | 1 (9.1%) | |
| Cardiovascular comorbidities | |||
| Present (37) | 29 (78.4%) | 8 (21.6%) | 0.4546 |
| Absent (40) | 34 (85.0%) | 6 (15.0%) | |
| Rheumatological disease | |||
| Present (5) | 4 (80.0%) | 1 (20.0%) | 0.9138 |
| Absent (72) | 59 (81.9%) | 13 (18.1%) | |
| Curative surgery | |||
| Present (72) | 59 (81.9%) | 13 (18.1%) | 0.9138 |
| Absent (5) | 4 (80.0%) | 1 (20.0%) | |
| Chemotherapy | |||
| Present (10) | 8 (80.0%) | 2 (20.0%) | 0.8739 |
| Absent (67) | 55 (82.1%) | 12 (17.9%) | |
| Radiotherapy | |||
| Present (9) | 9 (100%) | 0 (0.0%) | 0.1349 |
| Absent (68) | 54 (79.4%) | 14 (20.6%) | |
| NLR | |||
| > 2.5 (41) | 27 (65.9%) | 14 (34.1%) | 0.0001 |
| ≤ 2.5 (36) | 36 (100.0%) | 0 (0.0%) | |
| PLR | |||
| > 182 (25) | 14 (56.0%) | 11 (44.0%) | 0.0001 |
| ≤ 182 (52) | 49 (94.2%) | 3 (5.8%) | |
Figure 1.
Correlation of pre-treatment LMR with peripheral blood counts. (a) Relative lymphopenia, (b) monocytosis, and (c) higher neutrophil counts in setting of low LMR. (d) No significant difference in platelet levels was observed. Medians of each constituent were compared across both LMR-high (LMR > 2.4) and LMR-low (LMR ≤ 2.4) groups. **p < 0.01, ****p < 0.0001 by Mann-Whitney U tests, n.s., not significant.
Survival analyses
At the point of data analysis, there were 17 patient deaths (22.1%), 21 events of metastasis or death (27.3%), and 27 events of recurrence, metastasis or death (35.1%). Across the cohort, LMR-low at the time of diagnosis was associated with worse OS (HR 6.73, 95% CI 1.65 to 27.42, p = 0.0078) (Table 2). Median OS in LMR-low patients was 10.0 years and not reached in LMR-high patients (Figure 2(a)). Other factors associated with worse OS on univariate analysis were older age (HR 4.39, 95% CI 1.53 to 12.64, p = 0.0060) for which median OS was 16.0 years and not reached for younger patients (Figure S2); and the absence of curative surgery (HR 7.36, 95% CI 1.01 to 53.51, p = 0.0487) for which median OS was 11.7 years and not reached for those who underwent curative surgery. Additionally, LMR-low was also associated with worse MFS (HR 5.11, 95% CI 1.48 to 17.7, p = 0.0101) with MFS for LMR-low patients as 16.0 years and not reached in LMR-high patients (Figure 2(b)). Older age was also associated with poorer MFS (HR 3.12, 95% CI 1.21 to 8.04, p = 0.0185). Median EFS for LMR-low patients was 16.0 years compared to 18.7 years in LMR-high patients, but the difference was not statistically significant (Figure 2(c)). Male sex (HR 2.54, 95% CI 1.16 to 5.54, p = 0.0196) and the absence of curative surgery (HR 16.17, 95% CI 2.37 to 110.34, p = 0.0045) were associated with worse EFS. Limb/trunk as the primary tumor site (HR 0.28, 95% CI 0.09 to 0.88, p = 0.0281) was associated with a better EFS (Figure S2). Neither NLR-high nor PLR-high was associated with worse OS, MFS or EFS (Table 2).
Table 2.
Univariate survival analysis of clinicopathological characteristics and peripheral indices of systemic inflammation (LMR, NLR and PLR).
| Characteristic | Overall survival | Metastatic-free survival | Event-free survival | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p | HR (95% CI) | p | HR (95% CI) | p | |
| Age (years), > 65 vs ≤ 65 | 4.39 (1.53 – 12.64) | 0.0060 | 3.12 (1.21 – 8.04) | 0.0185 | 1.74 (0.75 – 4.03) | 0.1961 |
| Sex, Male vs Female | 1.42 (0.54 – 3.76) | 0.4776 | 1.96 (0.82 – 4.68) | 0.1300 | 2.54 (1.16 – 5.54) | 0.0196 |
| Ethnicity, Chinese vs Others | 1.52 (0.43 – 5.32) | 0.5150 | 0.98 (0.32 – 2.96) | 0.9706 | 0.70 (0.26 – 1.93) | 0.4932 |
| Absence vs Presence of cardiovascular comorbidities | 1.07 (0.63 – 1.82) | 0.8011 | 0.96 (0.41 – 2.26) | 0.9212 | 0.96 (0.44 – 2.08) | 0.9107 |
| Absence vs Presence of rheumatological disease | 1.15 (0.44 – 3.03) | 0.7639 | 1.56 (0.32 – 7.58) | 0.5822 | 1.82 (0.45 – 7.36) | 0.4008 |
| Absence vs Presence of curative surgery | 7.36 (1.01 – 53.51) | 0.0487 | 4.26 (0.71 – 25.68) | 0.1143 | 16.17 (2.37 – 110.34) | 0.0045 |
| Tumor size (cm), ≥ 5 vs < 5 | 1.47 (0.56 – 3.83) | 0.4345 | 1.97 (0.83 – 4.66) | 0.1245 | 2.00 (0.92 – 4.35) | 0.0794 |
| Deep vs Superficial tumor | 1.85 (0.44 – 7.84) | 0.4028 | 1.27 (0.34 – 4.75) | 0.7206 | 1.58 (0.50 – 5.02) | 0.4388 |
| Primary tumor site | ||||||
| Abdominal/Pelvis (ref) | 1.00 | 0.6919 | 1.00 | 0.1162 | 1.00 | 0.0281 |
| Head (Extracranial)/Neck | 1.09 (0.46 – 2.61) | 0.30 (0.07 – 1.32) | 0.42 (0.12 – 1.48) | |||
| Intracranial | 1.38 (0.47 – 4.04) | 0.69 (0.13 – 3.82) | 1.41 (0.30 – 6.75) | |||
| Limb/Trunk | 1.04 (0.48 – 2.23) | 0.29 (0.08 – 1.05) | 0.28 (0.09 – 0.88) | |||
| Pulmonary/Pleura | 0.70 (0.31 – 1.57) | 0.33 (0.09 – 1.23) | 0.43 (0.13 – 1.43) | |||
| LMR ≤ 2.4 vs > 2.4 | 6.73 (1.65 - 27.42) | 0.0078 | 5.11 (1.48 – 17.70) | 0.0101 | 2.76 (0.92 – 8.31) | 0.0706 |
| NLR > 2.5 vs ≤ 2.5 | 0.73 (0.28 – 1.91) | 0.5165 | 0.73 (0.31 – 1.74) | 0.4774 | 0.78 (0.36 – 1.71) | 0.5382 |
| PLR > 182 vs ≤ 182 | 1.81 (0.63 – 5.17) | 0.2696 | 1.50 (0.59 – 3.83) | 0.3996 | 1.02 (0.44 – 2.37) | 0.9567 |
Figure 2.
Correlation of LMR levels with survival outcomes. Low LMR was associated with worse (a) overall survival (OS) and (b) metastatic-free survival (MFS) but not (c) event-free survival (EFS).
On multivariate analysis that included clinicopathological factors predicting survival, LMR-low remained an independent predictor of poor OS (HR 3.19, 95% CI 1.00 to 10.15, p = 0.0497) alongside older age at diagnosis (HR 2.94, 95% CI 1.02 to 8.47, p = 0.0457) and the absence of curative surgery (HR 5.55, 95% CI 1.41 to 21.90, p = 0.0145) (Table 3). LMR-low alone remained as an independent predictor for poorer MFS (HR 3.17, 95% CI 1.25 to 7.98, p = 0.0147). For EFS, only male sex (HR 2.50, 95% CI 1.07 to 5.84, p = 0.0337) and presence of curative surgery (HR 0.30, 95% CI 0.10 to 0.90, p = 0.0312) remained independent predictors.
Table 3.
Multivariable logistic regression analysis using backwards procedure of factors associated with OS and EFS.
| Characteristic | Overall survival | Event-free survival | Metastatic-free survival | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p | HR (95% CI) | p | HR (95% CI) | p | |
| Age (years), > 65 vs ≤ 65 | 2.94 (1.02 – 8.47) | 0.0457 | N.A. | N.A. | ||
| Sex, Male vs Female | N.A. | 2.50 (1.07 – 5.84) | 0.0337 | N.A. | ||
| Absence vs Presence of curative surgery | 5.55 (1.41 – 21.90) | 0.0145 | 3.34 (1.12 – 10.01) | 0.0312 | N.A. | |
| Primary tumor site, Intracranial vs Others | N.A. | 2.36 (0.91 – 6.11) | 0.0768 | N.A. | ||
| LMR ≤ 2.4 vs > 2.4 | 3.19 (1.00 – 10.15) | 0.0497 | N.A. | 3.17 (1.25 – 7.98) | 0.0147 | |
Intratumoral gene expression profiles
As part of an exploratory analysis, a subgroup of nine patients was analyzed. We examined the relationship between peripheral LMR levels with respect to the relative abundance of tumor-infiltrating cells and oncogenic pathways inferred from tumoral tissue transcriptomic profiling using the NanoString PanCancer IO360 Panel. Of the nine tumor samples analyzed, LMR was negatively correlated with cytokine and chemokine signaling pathway scores (rho = -0.733, p = 0.0246), mast cell scores (rho = -0.700, p = 0.0358) and CD8 T cells (rho = -0.667, p = 0.0499) (Figure 3).
Figure 3.
Correlation of peripheral LMR with intratumoral immuno-oncogenic signaling pathways and cell types via NanoString analysis. A significant negative correlation between peripheral LMR and cytokine and chemokine signaling pathway scores (rho = -0.733, p = 0.0246), mast cell scores (rho = -0.700, p = 0.0358), and CD8 T cell scores (rho = -0.667, p = 0.0499) were observed.
Discussion
In this retrospective cohort study involving patients with localized SFT, we examined clinicopathological correlates and peripheral indices of systemic inflammation as predictors of survival outcomes. Low pre-treatment LMR was independently associated with poorer OS. Additionally, older age and absence of curative surgery were also predictors of worse OS. LMR reflects the levels of circulating lymphocytes responsible for anti-tumor surveillance23–25 relative to that of circulating monocytes, which infiltrate tumor tissue and differentiate into tumor-associated macrophages (TAMs).26,27 TAMs may subsequently promote tumor progression through facilitating processes such as angiogenesis, extracellular matrix remodeling, metastasis and immune evasion. 28 Low peripheral LMR may thus reflect a systemic environment that favors tumor progression rather than regression through immune containment. While LMR has been widely studied across a variety of malignancies – including lung, breast, gastrointestinal and hematologic cancers – its role in SFTs remains unclear.29–33 More broadly, NLR and LMR have demonstrated their utility as possible prognostic predictors in soft-tissue sarcomas.22,34
Our current findings extrapolate this emerging body of evidence to SFT and suggest that LMR may serve as an adjunct to contemporary risk stratification models like the Demicco criteria. Crucially, the association between low LMR and worse OS persisted on multivariate analysis in addition to age and surgical status, implying that there might be prognostic value in LMR that was not otherwise fully captured by traditional clinicopathological factors. As a potential prognostic biomarker, LMR’s clinical utility also lies in its inexpensive, reproducible and readily available nature given the routine nature of pre-treatment laboratory investigations. This is contrary to histological features that require tumor biopsy or resection and expertise from trained pathologists. Furthermore, peripheral indices of systemic inflammation may also be assessed dynamically and possibly reflect host-tumor interactions longitudinally. To this end, LMR may serve as a useful supplement to established risk stratification models rather than a replacement for current frameworks. Given the limited discriminatory performance of ROC-derived cut-offs for this cohort, possibly due to small sample size and disease heterogeneity, we utilized previously established cut-offs from a larger STS cohort that included SFT to improve reproducibility and comparability. Additionally, STAT6 immunohistochemistry, which is currently a key diagnostic marker for SFT, was not consistently available for initial cases within the cohort. This may limit diagnostic consistency across the cohort over this long study period. Future research with larger multi-institutional cohorts and longer follow-up may provide greater refinement of SFT-specific biomarker cut-offs and better clarity on their clinical utility.
To understand the biological basis behind the relationship between LMR and survival outcomes, intratumoral gene expression profiles were evaluated using bulk transcriptomics via the NanoString Pancancer IO360 panel. The small sample size (n = 9) was a limitation, though our analyses revealed biologically plausible and coherent trends. There was a statistically significant negative correlation observed between peripheral LMR and intratumoral cytokine and chemokine signaling, mast cell and CD8 T cell abundance. These findings suggest that low LMR is associated with increased inflammatory signaling and immune cell infiltration intratumorally, aligning with recent analyses demonstrating that tumor biology in SFTs is shaped by multiple molecular and immunological features, including alterations within the tumor microenvironment driven by processes such as immune cell infiltration and cytokine/chemokine signaling.19,35,36 In previous work done on angiosarcomas, intratumoral mast cell infiltration has been implicated with poorer survival outcomes. 37 This is because the presence of mast cells may be associated with pro-tumorigenic processes such as angiogenesis, and the dysregulation of other immune-oncologic pathways such as apoptosis, DNA damage repair and cell proliferation.38,39 These findings allude to the possibility that peripheral indices of systemic inflammation may mirror pro-tumoral inflammatory signaling. Even though intratumoral abundance of CD8 T cells was also inversely correlated with LMR, a nuanced interpretation of this finding may be warranted. Rather than effective anti-tumor immunity, this may reflect immune dysfunction due to functional exhaustion and subsequently compensatory recruitment, or perhaps immune exclusion within an immunosuppressive tumor milieu. 40 Given the small sample size, these findings should be considered hypothesis-generating and warrant validation in larger cohorts.
Conclusion
To conclude, this study has shown that low peripheral LMR is an independent predictor of poorer outcomes among patients with localized SFT and correlates with an immunosuppressive immune landscape intratumorally. These findings suggest a pertinent role of peripheral indices of systemic inflammation as widely accessible and biologically relevant biomarkers to augment prognostication in patients with SFT. While future validation is needed, inclusion of such variables into existing risk assessment models may represent a step forward into more individualized management of this rare, heterogenous entity.
Supplemental material
Supplemental material for Clinical and hematological indices as prognosticators in localized solitary fibrous tumors: A retrospective study by Ariel Yoong Yi KOH, Ryan Mao Heng LIM, Jason Yongsheng CHAN in Therapeutic Advances in Medical Oncology
Supplemental material for Clinical and hematological indices as prognosticators in localized solitary fibrous tumors: A retrospective study by Ariel Yoong Yi KOH, Ryan Mao Heng LIM, Jason Yongsheng CHAN in Therapeutic Advances in Medical Oncology
Author contributions: A.Y.Y.K. analyzed the data and prepared the first draft of the manuscript; R.M.H.L. provided clinical information and contributed to data interpretation; J.Y.C. conceptualized the study, interpreted the results, had unrestricted access to all data, and revised the manuscript. All authors read and approved the final manuscript.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Singapore Ministry of Health’s National Medical Research Council Research Transition Award (TA21jun-0005), Clinician Scientist Individual Research Grant (CIRG25jan-0007), Large Collaborative Grant (OFLCG23May-0039), TETRAD II Collaborative Centre Grant (CG21APR2002), SingHealth Duke-NUS AM/ACP-Designated Philanthropic Fund Grant Award (08/FY2023/EX/27-A65), Tanoto Foundation, NCCS Cancer fund, as well as the Khoo Bridge Funding Award provided by Duke-NUS Medical School and the “Estate of Tan Sri Khoo Teck Puat” (Duke-NUS-KBrFA/2025/0090).
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental material: Supplemental material for this article is available online.
ORCID iDs
Ariel Yoong Yi KOH https://orcid.org/0009-0005-6429-3919
Jason Yongsheng CHAN https://orcid.org/0000-0002-4801-3703
Ethical considerations
Tissue collection and consent protocols were under ethics approval from the SingHealth Centralized Institution Review Board (CIRB 2018/3182).
Consent to participate
Written informed consent from patients for use of clinical data and biospecimens was obtained in accordance with the Declaration of Helsinki.
Data Availability Statement
Gene expression data have been deposited in Gene Expression Omnibus (accession number GSE322577).*
References
- 1.Kallen ME, Hornick JL. The 2020 WHO Classification: What's New in Soft Tissue Tumor Pathology? Am J Surg Pathol 2021; 45: e1–e23. 10.1097/pas.0000000000001552 [DOI] [PubMed] [Google Scholar]
- 2.WHO. Soft tissue and bone tumours . World Health Organization Classification of Tumours Editorial Board 2020. [Google Scholar]
- 3.Gao X, Zhang J, Qian Z, et al. Malignant solitary fibrous tumor of the pleura: A narrative review of clinical characteristics, diagnosis and therapeutic options. Respiratory Medicine and Research 2024; 86: 100961. 10.1016/j.resmer.2022.100961 [DOI] [Google Scholar]
- 4.Martin-Broto J, Mondaza-Hernandez JL, Moura DS, et al. A Comprehensive Review on Solitary Fibrous Tumor: New Insights for New Horizons. Cancers (Basel) 2021; 13: 20210610. 10.3390/cancers13122913 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Robinson DR, Wu Ym Fau - Kalyana-Sundaram S, Kalyana-Sundaram SFau - CX, et al. Identification of recurrent NAB2-STAT6 gene fusions in solitary fibrous tumor by integrative sequencing. Nature genetics 2013; 45: 180–185. 10.1038/ng.2509 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Thway K, Ng W, Noujaim J, et al. The Current Status of Solitary Fibrous Tumor: Diagnostic Features, Variants, and Genetics. Int J Surg Pathol 2016; 24: 281–292. 10.1177/1066896915627485 [DOI] [PubMed] [Google Scholar]
- 7.Swaminathan G, Muratori J, Rauf F, et al. A Rare Metastatic Fibrous Tumor: Review of Literature With a Case Report. JTO Clin Res Rep 2025; 6: 901–908. 10.1016/j.jtocrr.2025.100901 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cranshaw IM, Gikas PD, Fisher C, et al. Clinical outcomes of extra-thoracic solitary fibrous tumours. Eur J Surg Oncol 2009; 35: 994–998. 10.1016/j.ejso.2009.02.015 [DOI] [PubMed] [Google Scholar]
- 9.Gold JS, Antonescu CR, Hajdu C, et al. Clinicopathologic correlates of solitary fibrous tumors. Cancer 2002; 94: 1057–1068. [PubMed] [Google Scholar]
- 10.Choi JH. Solitary fibrous tumor: an updated review. J Pathol Transl Med 2026; 60: 20–46. 10.4132/jptm.2025.10.08 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Demicco EG, Wagner MJ, Maki RG, et al. Risk assessment in solitary fibrous tumors: validation and refinement of a risk stratification model. Mod Pathol 2017; 30: 1433–1442. 10.1038/modpathol.2017.54 [DOI] [PubMed] [Google Scholar]
- 12.Tolstrup J, Loya A, Aggerholm-Pedersen N, et al. Risk factors for recurrent disease after resection of solitary fibrous tumor: a systematic review. Front Surg 2024; 11: 1332421. 10.3389/fsurg.2024.1332421 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Martin-Broto J, Stacchiotti S, Lopez-Pousa A, et al. Pazopanib for treatment of advanced malignant and dedifferentiated solitary fibrous tumour: a multicentre, single-arm, phase 2 trial. Lancet Oncol 2019; 20: 134–144. 10.1016/s1470-2045(18)30676-4 [DOI] [PubMed] [Google Scholar]
- 14.Martin-Broto J, Cruz J, Penel N, et al. Pazopanib for treatment of typical solitary fibrous tumours: a multicentre, single-arm, phase 2 trial. Lancet Oncol 2020; 21: 456–466. 10.1016/s1470-2045(19)30826-5 [DOI] [PubMed] [Google Scholar]
- 15.Demicco EG, Park MS, Araujo DM, et al. Solitary fibrous tumor: a clinicopathological study of 110 cases and proposed risk assessment model. Mod Pathol 2012; 25: 1298–1306. 10.1038/modpathol.2012.83 [DOI] [PubMed] [Google Scholar]
- 16.Connor RF. Demicco EG, Meyer C. Solitary fibrous tumor. 2026; https://www.uptodate.com/contents/solitary-fibrous-tumor-epidemiology-clinical-features-and-diagnosis
- 17.Sugita S, Segawa K, Kikuchi N, et al. Prognostic usefulness of a modified risk model for solitary fibrous tumor that includes the Ki-67 labeling index. World J Surg Oncol 2022; 20: 29. 10.1186/s12957-022-02497-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Pasquali S, Gronchi A, Strauss D, et al. Resectable extra-pleural and extra-meningeal solitary fibrous tumours: A multi-centre prognostic study. Eur J Surg Oncol 2016; 42: 1064–1070. 10.1016/j.ejso.2016.01.023 [DOI] [PubMed] [Google Scholar]
- 19.Zhang R, Yang Y, Hu C, et al. Comprehensive analysis reveals potential therapeutic targets and an integrated risk stratification model for solitary fibrous tumors. Nat Commun 2023; 14: 7479. 10.1038/s41467-023-43249-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Hidalgo-Ríos S, Carrillo-García J, Moura DS, et al. Peripheral Inflammatory Indexes Neutrophil/Lymphocyte Ratio (NLR) and Red Cell Distribution Width (RDW) as Prognostic Biomarkers in Advanced Solitary Fibrous Tumour (SFT) Treated with Pazopanib. Cancers (Basel) 2022; 14: 20220829. 10.3390/cancers14174186 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.von EE, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLoS Med 2007; 4: e296. 10.1371/journal.pmed.0040296 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chan JY, Zhang Z, Chew W, et al. Biological significance and prognostic relevance of peripheral blood neutrophil-to-lymphocyte ratio in soft tissue sarcoma. Sci Rep 2018; 8: 11959. 10.1038/s41598-018-30442-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ruffell B, DeNardo DG, Affara NI, et al. Lymphocytes in cancer development: polarization towards pro-tumor immunity. Cytokine Growth Factor Rev 2010; 21: 3–10. 10.1016/j.cytogfr.2009.11.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kravtsov DS, Erbe AK, Sondel PM, et al. Roles of CD4+ T cells as mediators of antitumor immunity. Front Immunol 2022; 13: 972021. 10.3389/fimmu.2022.972021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Notarbartolo S, Abrignani S. Human T lymphocytes at tumor sites. Semin Immunopathol 2022; 44: 883–901. 10.1007/s00281-022-00970-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Biswas SK, Allavena P, Mantovani A. Tumor-associated macrophages: functional diversity, clinical significance, and open questions. Semin Immunopathol 2013; 35: 585–600. 10.1007/s00281-013-0367-7 [DOI] [PubMed] [Google Scholar]
- 27.Wang Y, Li J, Nakahata S, et al. Complex Role of Regulatory T Cells (Tregs) in the Tumor Microenvironment: Their Molecular Mechanisms and Bidirectional Effects on Cancer Progression. Int J Mol Sci 2024; 25: 20240704. 10.3390/ijms25137346 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Noy R, Pollard JW. Tumor-associated macrophages: from mechanisms to therapy. Immunity 2014; 41: 49–61. 10.1016/j.immuni.2014.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Li W, Ma G, Wu Q, et al. Prognostic value of lymphocyte-to-monocyte ratio among Asian lung cancer patients: a systematic review and meta-analysis. Oncotarget 2017; 8: 110606–110613. 10.18632/oncotarget.20574 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Zhu Y, Zhao W, Mao G. Perioperative lymphocyte-to-monocyte ratio changes plus CA199 in predicting the prognosis of patients with gastric cancer. J Gastrointest Oncol 2022; 13: 1007–1021. 10.21037/jgo-22-411 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhu L, Qiao J, Yang C, et al. Prognostic value of lymphocyte-to-monocyte ratio for breast cancer: a systematic review and meta-analysis. BMC Womens Health 2025; 25: 569. 10.1186/s12905-025-04120-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Chan JC, Chan DL, Diakos CI, et al. The Lymphocyte-to-Monocyte Ratio is a Superior Predictor of Overall Survival in Comparison to Established Biomarkers of Resectable Colorectal Cancer. Ann Surg 2017; 265: 539–546. 10.1097/sla.0000000000001743 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Vajavaara H, Leivonen SK, Jørgensen J, et al. Low lymphocyte-to-monocyte ratio predicts poor outcome in high-risk aggressive large B-cell lymphoma. EJHaem 2022; 3: 681–687. 10.1002/jha2.409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Watanabe S, Shimoi T, Nishikawa T, et al. Lymphocyte-to-monocyte ratio as a prognostic and potential tumor microenvironment indicator in advanced soft tissue sarcoma treated with first-line doxorubicin therapy. Scientific Reports 2023; 13: 10734. 10.1038/s41598-023-37616-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Georgiesh T, Namløs HM, Sharma N, et al. Clinical and molecular implications of NAB2-STAT6 fusion variants in solitary fibrous tumour. Pathology 2021; 53: 713–719. 10.1016/j.pathol.2020.11.010 [DOI] [PubMed] [Google Scholar]
- 36.Medina-Ceballos E, Machado I, Giner F, et al. Immunological Tumor Microenvironment of Solitary Fibrous Tumors-Associating Immune Infiltrate with Variables of Prognostic Significance. Cancers (Basel) 2024; 16: 20240921. 10.3390/cancers16183222 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Tai SB, Lee ECY, Lim BY, et al. Tumor-Infiltrating Mast Cells in Angiosarcoma Correlate With Immuno-Oncology Pathways and Adverse Clinical Outcomes. Lab Invest 2024; 104: 100323. 10.1016/j.labinv.2024.100323 [DOI] [PubMed] [Google Scholar]
- 38.Chimal-Ramírez GK, Espinoza-Sánchez NA, Fuentes-Pananá EM. Protumor activities of the immune response: insights in the mechanisms of immunological shift, oncotraining, and oncopromotion. J Oncol 2013; 2013: 835956. 10.1155/2013/835956 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Guo X, Sun M, Yang P, et al. Role of mast cells activation in the tumor immune microenvironment and immunotherapy of cancers. Eur J Pharmacol 2023; 960: 176103. 10.1016/j.ejphar.2023.176103 [DOI] [PubMed] [Google Scholar]
- 40.Zheng S, Wang W, Shen L, et al. Tumor battlefield within inflamed, excluded or desert immune phenotypes: the mechanisms and strategies. Exp Hematol Oncol 2024; 13: 80. 10.1186/s40164-024-00543-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material for Clinical and hematological indices as prognosticators in localized solitary fibrous tumors: A retrospective study by Ariel Yoong Yi KOH, Ryan Mao Heng LIM, Jason Yongsheng CHAN in Therapeutic Advances in Medical Oncology
Supplemental material for Clinical and hematological indices as prognosticators in localized solitary fibrous tumors: A retrospective study by Ariel Yoong Yi KOH, Ryan Mao Heng LIM, Jason Yongsheng CHAN in Therapeutic Advances in Medical Oncology
Data Availability Statement
Gene expression data have been deposited in Gene Expression Omnibus (accession number GSE322577).*