If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Genetic Pathology Evaluation Centre, University of British Columbia, Vancouver, British Columbia, CanadaDepartment of Pathology, Vancouver General Hospital, British Columbia, Canada
Karyotypically complex pleomorphic sarcomas are the most likely to be immune hot.
•
Alveolar soft part sarcoma show response to immune checkpoint inhibitors.
•
Multiplexed and spatially resolved technologies can aid immune biomarker discovery.
•
Sarcoma subtype and/or biomarker specific Phase 3 trials are required.
Abstract
Sarcomas are a heterogenous group of mesenchymal cancers comprising over 100 subtypes. Current chemotherapy for all but a very few subtypes has limited efficacy, resulting in 5-year relative survival rates of 16% for metastatic patients. While sarcomas have often been regarded as an “immune cold” tumor category, recent biomarker studies have confirmed a great deal of immune heterogeneity across sarcoma subtypes. Reports from the first generation of clinical trials treating sarcomas with immunotherapy demonstrate a few positive responses, supporting efforts to stratify patients to optimize response rates. This review summarizes recent advances in knowledge around immune biomarker expression in sarcomas, the potential use of new technologies to complement these study results, and clinical trials particularly of immune checkpoint inhibitor therapy in sarcomas.
Each of the immune biomarkers assessed was reviewed for subtype-specific expression patterns and correlation with prognosis. Overall, there is extensive heterogeneity of immune biomarker presence across sarcoma subtypes, and no consensus on the prognostic effect of these biomarkers. New technologies such as multiplex immunohistochemistry and high plex in situ profiling may offer more insights into the sarcoma microenvironment. To date, clinical trials using immune checkpoint inhibitor monotherapy have not shown compelling clinical benefits. Combination therapy with dual checkpoint inhibitors or in combinations with other agents has yielded more promising results in dedifferentiated liposarcoma, undifferentiated pleomorphic sarcoma, angiosarcoma and alveolar soft-part sarcoma. Better understanding of the sarcoma immune status through biomarkers may help decipher the reasons behind differential responses to immunotherapy.
Soft Tissue and Bone Tumours: WHO Classification of Tumours 5th ed. Volume 3. International Agency for Research on Cancer Publications,
Lyon, France2020
]. These broad tumor categories are further divided based on distinct morphology and genetic changes into more than 100 subtypes as defined by the World Health Organization [
Soft Tissue and Bone Tumours: WHO Classification of Tumours 5th ed. Volume 3. International Agency for Research on Cancer Publications,
Lyon, France2020
]. While most sarcomas are detected as local tumors and treated with excisional surgery and radiation, metastasis will occur in up to 50% within the first five years (depending on the site, grade and subtype) [
]. Only a few sarcoma subtypes have experienced increase cure rates from specific systemic therapies, most prominently tyrosine kinase inhibitors for gastrointestinal stromal tumors (GIST) [
First-line anthracycline-based chemotherapy for angiosarcoma and other soft tissue sarcoma subtypes: Pooled analysis of eleven european organisation for research and treatment of cancer soft tissue and bone sarcoma group trials.
]. The majority of metastatic sarcomas are restricted to using chemotherapy (typically doxorubicin and/or ifosphamide), which remains of limited and controversial value [
Historically most sarcomas have been viewed as “immune cold” tumors, despite research beginning in the 1800s that documented anecdotally-successful treatment through infection-induced antitumor immune responses [
]. More recently, it has become recognized that at the level of DNA structure, sarcomas can be broadly categorized into two main groups: (1) sarcomas with simple and specific chromosomal changes (often translocations) and a low mutational burden, and (2) karyotypically-complex sarcomas with numerous copy number aberrations and moderate mutational burdens [
Existing cancer immunotherapy modalities include chimeric antigen receptor T-cell therapy, oncolytic viruses, cancer vaccines, and immune checkpoint inhibitors (ICI). In general, while pre-clinical data suggest there are some responses of sarcomas to immunotherapy, compared to cancers such as melanoma and non-small cell lung cancer [
], clinical results in sarcoma have not shown dramatic improvements and trials have been slower in development. The positive, albeit few responses to immunotherapy, among patients with sarcoma suggest that more effort should be used to identify which patients are most likely to respond.
Immune biomarkers have been applied to many cancers to inform prognosis, and more importantly to predict response to immunotherapy. PD-L1 tumor expression has been used in non-small cell lung cancer [
Pembrolizumab versus chemotherapy for previously untreated, PD-L1-expressing, locally advanced or metastatic non-small-cell lung cancer (KEYNOTE-042): a randomised, open-label, controlled, phase 3 trial.
Safety and efficacy of pembrolizumab monotherapy in patients With previously treated advanced gastric and gastroesophageal junction cancer: Phase 2 clinical KEYNOTE-059 Trial.
]. Multiple studies have attempted to identify the immune biomarkers expressed in sarcomas that are relevant to clinical management. This review will summarize recent advances in knowledge of sarcoma immune biomarker expression, new technologies for their investigation, and ICI clinical trials focussing on sarcoma subtypes for which effective targeted therapies do not yet exist. Immune augmentation strategies like CAR-T cell therapy and cancer vaccines will not be covered in this review, as others have recently reviewed the clinical development of such modalities in sarcoma [
Intraepithelial CD8 super(+) tumor-infiltrating lymphocytes and a high CD8 super(+)/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer.
] are associated with improved survival and response to therapy due to their roles in an immune active environment (Fig. 1). On the other hand, an increase in FOXP3+ regulatory T cells may reflect an immunosuppressed environment, or the resolution of a robust anti-tumor response [
Fig. 1Immune cells in the tumor environment can exhibit anti-tumor response or have immune-suppressive activity. Dendritic cells activate CD8+ cytotoxic T cells through CD8/ MHC I/ T cell receptor (TCR or CD3) complex and helper T cells through the CD4/ MHC II/ TCR complex. Helper T cells interact with and activate CD20+ B cells which induces plasma cell differentiation and antibody class switching. Macrophages can be pro-inflammatory and anti-tumor CD68+ “M1-like”, or anti-inflammatory and pro-tumor CD163+ M2 macrophages which make up most of the tumor-associated macrophages (TAM). In addition to M2 TAMs, FOXP3+ Regulatory T cells are also highly immunosuppressive. Immune checkpoints such as PD-1, CTLA-4, TIM-3, and LAG-3 can be found on the surface of a variety of immune cells. The PD-1/PD-L1 axis and CD47/SIRPα axis are two immune checkpoint pathways that interact with tumor cells, and can lead to an immunosuppressive response if activated.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]. In rhabdomyosarcomas, the copy number- and mutation-driven embryonal subtype has more CD3+ and CD8+ T cell infiltration than its translocation-associated counterpart, alveolar rhabdomyosarcoma [
However, increased T cell infiltration among karyotypically complex subtypes is inconsistent across studies, with one publication reporting almost no T cells in leiomyosarcomas [
], lower than what might be expected in a pleomorphic subtype. Studies on synovial sarcoma – a translocation-associated subtype – also lack consensus, with one reporting CD8+ T cell infiltration in 45% (10/22) of cases [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]. This large apparent discrepancy however is easily explained by different methodologies and definitions of what constitutes a positive case, as the former study was based on 1–2 mm tissue microarray cores with a cutoff of ≥10 CD8+ cells [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]). Immune infiltrates can only be reliably compared among sarcoma types when they are described within the same study employing the same definitions of positive; tissue biomarker publications ideally should report results per mm2 of examined tissue to facilitate comparisons between studies[
] there were no significant prognostic associations with T cell infiltrates (Table 1). However, studies on individual subtypes suggest that CD8+ and FOXP3+ infiltrates may be prognostic. Newly diagnosed soft-tissue sarcomas have higher numbers of CD8+ T cells than recurrent tumors, and subsequent decreases in CD8+ T cell count indicate poor survival among patients with multiple recurrences [
PD-L1 and PD-1 and characterization of tumor-infiltrating lymphocytes in high grade sarcomas of soft tissue - prognostic implications and rationale for immunotherapy.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]. In contrast, concurrent infiltration of CD8+ and FOXP3+ T cells in myxoid liposarcoma was correlated with worse overall survival than single CD8+ infiltration [
Table 1Summary of IHC-based immune biomarker studies in sarcoma. Boxes represent prognostic effect of corresponding biomarker: positive (solid green), negative (striped red), no association (white).
*Immune biomarkers associated with prognosis as single markers and as co-infiltrates.
**Immune biomarkers associated with prognosis only as co-infiltrates.
Thus to date, publications on the T cell infiltration in specific histologic subtypes are difficult to compare due to the varying definitions of what constitutes a positive case [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]. There is a trend for copy-number driven subtypes (DDLPS, MPNST, and UPS) to have positive association between T cell infiltration and prognosis, and translocation-associated subtypes (synovial sarcoma and myxoid liposarcoma) to have mixed results [
]. The studies that reported no prognostic association with T cell infiltration had either merged dissimilar histologic subtypes or very few patients, suggesting subtype-dependent prognostic significance needs to be addressed with larger subtype-specific cohorts.
B cell infiltration: CD20
B cells play a crucial role in tumor-related immunity, and have recently garnered increased attention [
]. B cells contribute to tumor clearance through cancer-targeting antibodies that mediate antibody-dependent cell cytotoxicity and phagocytosis, and form immune complexes [
]. One of the earliest studies applying the B-cell biomarker CD20 in soft-tissue sarcoma specimens demonstrated B cell infiltration to associate with improved prognosis (Table 1) [
]. Correspondingly, a smaller soft-tissue cohort found B cells to associate with improved metastasis-free and overall survival, with results validated in The Cancer Genome Atlas cohort based on expression levels of MS4A1 (encoding for CD20) [
However, it should be noted that most studies using immunohistochemistry (IHC) have reported sparse staining of CD20 in sarcomas, and no prognostic association [
]. In retroperitoneal liposarcomas (well-differentiated, dedifferentiated, myxoid/round cell and pleomorphic liposarcoma), CD20+ B cells had the lowest frequency among all analyzed lymphocyte biomarkers (including CD8, CD4 and FOXP3) [
]. Using the microenvironment cell populations-counter method extracted from gene expression data, the authors found that a specific immune-high group comprising 18% of tested sarcomas has high expression of B cell lineage signatures [
]. Protein expression of B cell biomarkers was validated using multispectral fluorescence, showing CD20+ cells existing in high densities in tumors with tertiary lymphoid structures [
], may explain seemingly contradicting observations in other studies. B cells are mostly found in high densities in tumors with tertiary lymphoid structures, which are only identified in 12–16% of sarcoma tumor specimens [
]. Therefore, while B cell infiltration in sarcoma is mostly rare, the presence of B cells may be an important biomarker for an antitumor immune response amenable to reactivation by checkpoint inhibitor therapy.
Macrophage infiltration: CD163, CD68, CD204
A part of the innate immune system, macrophages are usually the most abundant leukocyte in tumors [
]. Tumor-associated macrophages (TAMs) mostly display an M2 phenotype, and are associated with poor survival in carcinomas of the lung, breast, and pancreas [
]. Phenotypes can be partially distinguished by IHC, as CD163 preferentially stains M2-like macrophages, whereas CD68 detects both but is often more strongly expressed in M1 macrophages [
]. Similar to observations on T cell infiltration, there are more macrophages found among copy number-driven sarcomas compared to translocation-associated sarcomas, which display macrophage density levels comparable with benign mesenchymal tumors [
], in which increased CD163+ macrophages were associated with improved survival. Despite research attempting to dichotomize macrophage phenotypes, the inconsistent correlations with survival suggest that macrophage polarization is not a definitive characterization [
]. Only one sarcoma study to date has investigated SIRPα, which found DDLPS, angiosarcoma, chordoma, and well-differentiated liposarcoma to have the highest levels of SIRPα+ macrophages [
], future studies should investigate SIRPα as potential predictive biomarker.
Immune checkpoints
PD-1
The programmed cell death-1 (PD-1) checkpoint pathway has been one of the most researched targets in immunotherapy. PD-1 is expressed mostly on immune cells such as activated T cells, B cells, natural killer cells, activated monocytes, and dendritic cells [
] (Fig. 1). Tumor cells can express the ligand PD-L1 to interact with PD-1 and thereby inhibit immune cell activity, resulting in an effective immune-evasion strategy [
Across three sarcoma subtypes (DDLPS, leiomyosarcoma, UPS), tumors with high immunogenic gene signatures have high expression levels for the genes encoding both PD-1 and PD-L1 (PDCD1 and CD274, respectively) [
]. Independent cohorts of specific subtypes have yielded supportive findings, as more than half of pleomorphic dermal sarcoma specimens have positive PD-1 expression on immune cells [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Spurny C, Kailayangiri S, Jamitzky S, Altvater B, Wardelmann E, Dirksen U, et al. Programmed cell death ligand 1 (PD‐L1) expression is not a predominant feature in Ewing sarcomas. Pediatric Blood Cancer 2018;65:e26719,n/a.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Spurny C, Kailayangiri S, Jamitzky S, Altvater B, Wardelmann E, Dirksen U, et al. Programmed cell death ligand 1 (PD‐L1) expression is not a predominant feature in Ewing sarcomas. Pediatric Blood Cancer 2018;65:e26719,n/a.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
PD-L1 and PD-1 and characterization of tumor-infiltrating lymphocytes in high grade sarcomas of soft tissue - prognostic implications and rationale for immunotherapy.
]. Consequently, efforts have also been made to address other components of the PD-1/PD-L1 pathway in sarcomas.
PD-L1
PD-L1 is expressed on stromal, malignant, and immune cells, and has been shown in other cancers to be predictive for checkpoint inhibitor therapy response and patient survival [
PD-L1 (CD274) copy number gain, expression, and immune cell infiltration as candidate predictors for response to immune checkpoint inhibitors in soft-tissue sarcoma.
Spurny C, Kailayangiri S, Jamitzky S, Altvater B, Wardelmann E, Dirksen U, et al. Programmed cell death ligand 1 (PD‐L1) expression is not a predominant feature in Ewing sarcomas. Pediatric Blood Cancer 2018;65:e26719,n/a.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Wedekind MF, Haworth KB, Arnold M, Stanek JR, Lee D, Cripe TP. Immune profiles of desmoplastic small round cell tumor and synovial sarcoma suggest different immunotherapeutic susceptibility upfront compared to relapse specimens. Pediatric Blood Cancer 2018;65:e27313,n/a.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
Wedekind MF, Haworth KB, Arnold M, Stanek JR, Lee D, Cripe TP. Immune profiles of desmoplastic small round cell tumor and synovial sarcoma suggest different immunotherapeutic susceptibility upfront compared to relapse specimens. Pediatric Blood Cancer 2018;65:e27313,n/a.
] report less than half the cases show PD-L1 expression on the sarcoma cells. In comparison, PD-L1 was expressed on the neoplastic cells in more than half of leiomyosarcoma [
The prognostic value of PD-L1 expression in sarcomas also varies with histologic subtype. PD-L1 expression measured by IHC was not significantly associated with survival in leiomyosarcoma [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
] was associated with different effects on survival in osteosarcoma and UPS. Such conflicting results highlight the need for more studies to validate published findings and suggest a nonlinear relationship between PD-L1 gene and protein expression in sarcoma.
Finally, it should be noted that not only is PD-L1 expressed on malignant cells, it is also expressed on immune cells present in leiomyosarcoma [
]. The role of PD-L1 expression on immune cells infiltrating sarcomas has not been extensively explored and may have important implications on prognosis as reported in carcinomas [
In addition to PD-1/PD-L1, other targetable immune checkpoints including CTLA-4, IDO, LAG-3, and TIM-3 offer additional directions for immune-oncology research. As expected, expression of these checkpoints is significantly associated with other immune infiltrates and checkpoint biomarker expression in sarcomas [
]. The immune-hot groups of soft-tissue sarcoma identified by immune cell gene signatures display high expression of CTLA4, HAVCR2, and LAG3 encoding CTLA-4, TIM-3, and LAG-3 respectively [
]. Furthermore, pretreatment specimens from a phase 2 clinical trial showed that two pleomorphic sarcoma subtypes, leiomyosarcoma and UPS, had around 70% cases with IDO-1 expression, while only 29% of other soft-tissue sarcomas met the same criteria for IDO-1 [
]. Nonetheless, given the close relationship between sarcoma immune status and these checkpoints, multiplex approaches to biomarker detection may be required to better understand the role of immune checkpoints in sarcoma.
Tumor mutation burden and microsatellite instability
Tumor mutation burden (TMB) and microsatellite instability are two other biomarkers that positively correlate with immunotherapy responses [
] as are seen in melanomas, a tumor type responding to checkpoint inhibitors. Preliminary data from two patients with angiosarcoma demonstrated clinical benefit from pembrolizumab treatment [
], suggesting this may be a sarcoma worthy of assessment in clinical trials.
New technologies
Lymphocyte counts on hematoxylin and eosin stained slides are the simplest and most established method to assess the immune microenvironment in sarcomas. Given the heterogeneity of immune cell composition, even within the category of lymphocytes, more specific approaches are needed to identify immune responses with different implications for prognosis and therapy [
]. Despite the specificity of conventional IHC, the fundamental limitation of interpreting a complex multicellular reaction by measuring a single biomarker per slide means that more sophisticated approaches are needed to properly characterize the sarcoma tumor immune microenvironment [
Multiplex IHC/Immunofluorescence approaches are designed to detect multiple biomarkers on a single tissue section. Such techniques identify biomarker co-expression on individual cells, reveal the spatial distribution of immune, malignant, and stromal cells and reduce consumption of precious tissue specimens. Beyond employing conventional chromogenic detection systems or fluorescently labelled antibodies, newer techniques that assess multiple biomarkers in situ take advantage of DNA barcoded [
]. New studies using multiplex approaches are identifying specific immune cells in the tumor microenvironment of sarcomas with predictive and prognostic value [
At the RNA level, whole transcriptome sequencing has been used to discover novel immune biomarkers and has already found significant clinical applications [
]. Limitations of bulk RNA sequencing, which does not attribute expression to particular cells in the tumor microenvironment, are addressed by single-cell RNAseq [
Lastly, Digital Spatial Profiling enables multiplexed measurement of proteins or RNAs within microscopically-identified regions of interest, at the resolution of a single cell type, by utilizing UV-cleavable oligonucleotide conjugated antibodies or RNA probes. Analysing immune cells in this sarcoma microenvironment using this technology has already been used to characterize mechanisms of doxorubicin resistance [
]. Going forward, these new technologies will increasingly be used to characterize the complex sarcoma tumor microenvironment to guide development of immuno-oncology strategies.
Immune checkpoint inhibitor clinical trials
Mono and dual immune checkpoint inhibitor therapy
While monoclonal ICI have demonstrated tremendous success in cancer types such as melanoma and non-small cell lung cancer [
], clinical trials involving patients with sarcoma have largely trailed behind. A single-centre phase II study using ipilimumab in adult patients with synovial sarcoma was one of the first ICI trials in sarcoma, but was terminated early due to slow accrual and lack of response: all six patients had progressive disease after the first cycle of therapy [
]. However, the authors concluded that the high frequency of immune-related adverse events (transaminitis, colitis) in the pediatric subjects precluded ipilimumab's effective use as a monotherapy [
]. Similarly, a phase II study of nivolumab (NCT02428192) in advanced uterine leiomyosarcoma demonstrated lack of response in all 12 enrolled patients [
]. Out of the 40 patients with soft-tissue sarcoma (ten each of soft tissue leiomyosarcoma, DDLPS, UPS, and synovial sarcoma), there was one complete response (UPS) and six partial responses (three UPS, two DDLPS, and one synovial sarcoma) using RECIST criteria (the primary endpoint) [
]. In the 40 patients with bone sarcoma (13 Ewing, 22 osteosarcoma, and 10 dedifferentiated or mesenchymal chondrosarcomas), only two patients (osteosarcoma and chondrosarcoma) had a partial response [
]. While neither cohort met the primary endpoint, the soft-tissue sarcoma arm met criteria to be considered an active therapy with a 12-week progression survival of 55% [
Clinical activity of pembrolizumab (P) in undifferentiated pleomorphic sarcoma (UPS) and dedifferentiated/pleomorphic liposarcoma (LPS): Final results of SARC028 expansion cohorts.
]. The UPS cohort successfully met its primary endpoints with an objective response rate of 23%, as reported at the 2019 American Society of Clinical Oncology (ASCO) meeting [
Clinical activity of pembrolizumab (P) in undifferentiated pleomorphic sarcoma (UPS) and dedifferentiated/pleomorphic liposarcoma (LPS): Final results of SARC028 expansion cohorts.
Going beyond ICI monotherapy, the phase II trial Alliance A091401 trial (NCT02500797) demonstrated that combination therapy using two ICIs can be effective in metastatic sarcoma [
]. In this study, the nivolumab monotherapy arm did not meet its pre-defined primary endpoint, as only three (ASPS, non-uterine leiomyosarcoma, and sarcoma NOS) of 38 patients achieved a partial response [
]. On the other hand, the combination therapy arm using nivolumab and ipilimumab had six of 38 patients achieve confirmed objective responses, including two complete responses (myxofibrosarcoma and uterine leiomyosarcoma) [
]. However, since this trial was not set up to directly compare the efficacy of the two arms, the authors could only conclude that the combination arm is comparable to standard sarcoma chemotherapy [
]. The potential of using combination ICI therapy as first or second line treatment for sarcoma led the Alliance cooperative group to expand the trial. At the 2020 ASCO meeting, preliminary results show that both the UPS and DDLPS expansion cohort reached the primary response endpoint in the combination ICI arm, each with two confirmed responses out of twelve patients [
Given the limited response rates observed with established ICIs, currently active monotherapy clinical trials are focusing on exploring novel agents, including anti-PD-L1 antibodies ZKAB001 (NCT03676985) and atezolizumab (NCT03141684; NCT03474094), and anti-PD-1 antibody TSR-042 (NCT04274023) (Table 2). Other trials are adopting combination strategies to involve dual ICI regimens (NCT02500797; NCT02815995; NCT03365791; NCT04095208), or an ICI with cancer vaccination therapy (NCT02609984; NCT03069378) or cytokines (NCT03063632; NCT03282344) (Table 3).
Table 3Clinical trials of immune checkpoint inhibitor in combination therapy for patients with sarcoma.
Immune checkpoint inhibitor
Other agents
NCT trial number
Phase
Status
Participants
Sarcoma subtype(s)
Includes pediatric patients
Immune checkpoint inhibitor in combination with another immunotherapy agent
Atezolizumab (anti-PD-L1)
CMB305 (dendritic cell-targeting viral vector expressing the NY-ESO-1 gene and G305 [NY-ESO-1 recombinant protein plus GLA-SE])
Immune checkpoint inhibitors in combination with other agents
Recent sarcoma clinical trials are combining ICI with chemotherapy agents and/or radiation therapy, a logical approach as the standard therapy for most sarcoma subtypes includes chemotherapy and radiation. Furthermore, tumor irradiation is associated with increased immune cell infiltration [
], suggesting radiation to be an immune activating procedure (the abscopal effect).
PEMBROSARC is a phase II study (NCT02406781) that used pembrolizumab and cyclophosphamide to treat four cohorts of soft tissue sarcoma (leiomyosarcoma, UPS, GIST, and others) [
]. While these trial results are unimpressive, the authors observed an increase in kynurenine to tryptophan ratio between cycles one and three of pembrolizumab [
]. Given the close association of kynurenine with IDO, the authors speculate that there may remain a role for combining IDO-1 inhibitors with pembrolizumab in sarcoma [
In the same trial, pembrolizumab was added to cyclophosphamide in an osteosarcoma cohort of 17 patients. Four experienced tumor shrinkage, including one confirmed partial response [
]. Interestingly, none of the three patients with tumor shrinkage and available tumor material had expression of PD-L1 on either sarcoma or immune cells [
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
]. Therefore, it is unclear whether the lack of PD-L1 expression among responders has any significance. Overall, results of the PEMBROSARC trial have been underwhelming and the authors concluded that more investigation is needed to understand mechanisms of resistance to immunotherapy in sarcoma.
One of the most successful ICI trials in sarcoma to date is a phase II study (NCT02636725), combining pembrolizumab with VEGF receptor tyrosine kinase inhibitor axitinib [
]. This study included 30 patients with advanced sarcoma, among whom twelve had ASPS; the rest included high grade pleomorphic sarcoma, soft tissue leiomyosarcoma, and DDLPS [
]. Findings were encouraging in the ASPS cohort, as six of eleven patients achieved partial response and an additional two achieved stable disease. Among patients with other histologic subtypes, there were two partial responses (epithelioid sarcoma and leiomyosarcoma) and seven with stable disease [
]. The primary endpoint of three-month progression-free survival in patients with ASPS was 72.7%, while the three-month progression-free survival for patients with other sarcomas was 61.9%. This success of ICI in ASPS was preceded by case reports [
]. However, these exciting results for ASPS have some caveats. Firstly, soft tissue sarcomas have been shown to respond to different tyrosine kinase inhibitors with median progression-free survival similar to that of this study [
Activity and safety of crizotinib in patients with alveolar soft part sarcoma with rearrangement of TFE3: European Organization for Research and Treatment of Cancer (EORTC) phase II trial 90101 'CREATE'.
], so the additional benefit of adding pembrolizumab is unclear. Secondly, the clinical course of ASPS is generally more indolent than other sarcomas in the first few years (although long-term prognosis is poor) [
]. Short term progression-free survival may be an unreliable primary endpoint in this disease. Instead, tumor size or a longer period of progression-free survival could be a more relevant primary endpoint as employed by other phase II trials on patients with ASPS [
Activity and safety of crizotinib in patients with alveolar soft part sarcoma with rearrangement of TFE3: European Organization for Research and Treatment of Cancer (EORTC) phase II trial 90101 'CREATE'.
Nevertheless, therapies combining ICI with other agents have demonstrated promising effects in select patients, leading to many more trials to investigate combination therapies (Table 3). Current ongoing trials include single ICI with chemotherapy (NCT02888665; NCT03074318; NCT03085225; NCT03123276; NCT03590210) or radiation therapy (NCT03092323; NCT03338959; NCT03548428), single ICI with tyrosine kinase inhibitors ((NCT01643278; NCT02298959; NCT02636725; NCT03899805; NCT03946943; NCT04044378; NCT04118166), and dual ICI with chemotherapy (NCT03138161; NCT04028063) or radiation therapy (NCT03116529; NCT03307616; NCT03463408).
Limitations of published studies
Several important limitations exist in the current sarcoma immune biomarker literature. Due to the comparative rarity of sarcomas relative to other cancer types, most studies have small sample sizes. Of the studies cited in this review, most had fewer than 100 patient samples (Table 1). Hence it is more difficult to identify significant correlations with other biomarkers or with patient outcomes for sarcomas, and there may be more apparent inconsistencies between studies. Many studies have consequently lumped sarcoma subtypes together to increase apparent power, but in doing so lose the ability to recognize effects specific to (biologically disparate) histologic subtypes. Immune status also varies within most histologic subtypes, which could represent an even more important impediment on identifying significant differences [
]. In the same vein, the small sample sizes and response rates in existing phase II sarcoma immunotherapy trials render it difficult to investigate the predictive value of immune biomarkers. To date, there is no data showing that any specific immune biomarker can predict response to ICI in sarcomas.
Tissue microarrays are one way to obtain higher through-put results on larger tumor datasets. However, the arrayed tissue cores are generally extracted from tumor centres whereas immune responses may be most prominent at the (unsampled) periphery. In DDLPS, no PD-L1 expression was found using tissue microarrays, whereas 22% of whole section cases were positive [
]. Apparent rates of positivity may also be hard to compare between IHC studies due to differences between antibody clones produced by different companies, staining platforms, staining conditions, interpretation and cutpoints for defining a case as positive [
], clone 22C3 (Dako) showed less staining than clone SP142 and clone 28-8 (Abcam #ab205921), respectively. Centralized protocols for quantification of PD-L1 expression should be considered in sarcoma, similar to those established in lung cancer [
], but standardized methods for staining and scoring for even commonly-used biomarkers such as CD8 and CD20 are not so well developed.
In terms of the sarcoma clinical trial literature, the elephant in the room is a lack of Phase III studies. The recent sarcoma literature contains multiple examples of agents that were promising in Phase II, even to the point of provisional approval, but which did not hold up in Phase III trials, the most recent and prominent example being olaratumab [
Effect of doxorubicin plus olaratumab vs doxorubicin plus placebo on survival in patients with advanced soft tissue sarcomas: The ANNOUNCE randomized clinical trial.
]. ICI or other immunotherapy strategies must ultimately reach this standard to prove their real value in sarcoma therapy.
Conclusion
Based on the described findings from immune biomarker studies, and the observed pattern of responses in ICI trials, dedifferentiated liposarcoma, embryonal rhabdomyosarcoma, uterine and non-uterine leiomyosarcoma, osteosarcoma, and undifferentiated pleomorphic sarcoma are the most commonly immune-hot sarcoma subtypes. These karyotypically complex subtypes are the most likely to have high levels of immune infiltration and checkpoint expression, and to exhibit responses to ICI therapy. Of all sarcomas, Ewing sarcoma is the most consistently immune-cold tumor that hence would possibly need immune augmentation strategies instead. Interestingly, other translocation-associated subtypes such as synovial sarcoma and solitary fibrous tumor demonstrate conflicting results in immunogenicity, and may need more investigation to reach consensus. Patients with alveolar soft part sarcoma have shown promising responses to ICIs for which confirmatory trials are underway.
Sarcomas have not seen the same breakthroughs in immunotherapy as seen in other malignancies. More subtype and/or biomarker specific studies are required, as published studies still inappropriately combine biologically different sarcoma subtypes. More emphasis should be given to B cell markers as they have recently demonstrated potentially stronger prognostic and predictive values than the T-cell biomarkers that have been the focus of most studies to date. New technologies can be used to discover novel immune biomarker relationships and to enable more complete understanding of the sarcoma microenvironment by employing multiplexed, spatially resolved RNA and protein techniques. Perhaps these technologies will be able to identify predictive sarcoma biomarkers for the ICI regimens with promising early results, including pembrolizumab monotherapy, nivolumab with ipilimumab, and axitinib with pembrolizumab. Subtypes such as alveolar soft part sarcoma, angiosarcoma, dedifferentiated liposarcoma and undifferentiated pleomorphic sarcoma have shown some impressive responses to checkpoint inhibitors and are particularly worthy of extra investigation in more advanced clinical trials linked to correlative science efforts employing new biomarker technologies.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors would like to thank Drs. Robin Jones, William Tap, David Thomas and Shane Zaidi for helpful discussions about active clinical trials. This research was supported by grants from the Canadian Cancer Society (705615) and the Terry Fox Research Institute (1082).
References
WHO Classification of Tumours Editorial Board
Soft Tissue and Bone Tumours: WHO Classification of Tumours 5th ed. Volume 3. International Agency for Research on Cancer Publications,
Lyon, France2020
First-line anthracycline-based chemotherapy for angiosarcoma and other soft tissue sarcoma subtypes: Pooled analysis of eleven european organisation for research and treatment of cancer soft tissue and bone sarcoma group trials.
Pembrolizumab versus chemotherapy for previously untreated, PD-L1-expressing, locally advanced or metastatic non-small-cell lung cancer (KEYNOTE-042): a randomised, open-label, controlled, phase 3 trial.
Safety and efficacy of pembrolizumab monotherapy in patients With previously treated advanced gastric and gastroesophageal junction cancer: Phase 2 clinical KEYNOTE-059 Trial.
Intraepithelial CD8 super(+) tumor-infiltrating lymphocytes and a high CD8 super(+)/regulatory T cell ratio are associated with favorable prognosis in ovarian cancer.
Expression and clinical association of programmed cell death-1, programmed death-ligand-1 and CD8(+) lymphocytes in primary sarcomas is subtype dependent.
PD-L1 and PD-1 and characterization of tumor-infiltrating lymphocytes in high grade sarcomas of soft tissue - prognostic implications and rationale for immunotherapy.
Spurny C, Kailayangiri S, Jamitzky S, Altvater B, Wardelmann E, Dirksen U, et al. Programmed cell death ligand 1 (PD‐L1) expression is not a predominant feature in Ewing sarcomas. Pediatric Blood Cancer 2018;65:e26719,n/a.
PD-L1 (CD274) copy number gain, expression, and immune cell infiltration as candidate predictors for response to immune checkpoint inhibitors in soft-tissue sarcoma.
Wedekind MF, Haworth KB, Arnold M, Stanek JR, Lee D, Cripe TP. Immune profiles of desmoplastic small round cell tumor and synovial sarcoma suggest different immunotherapeutic susceptibility upfront compared to relapse specimens. Pediatric Blood Cancer 2018;65:e27313,n/a.
Clinical activity of pembrolizumab (P) in undifferentiated pleomorphic sarcoma (UPS) and dedifferentiated/pleomorphic liposarcoma (LPS): Final results of SARC028 expansion cohorts.
30153-5&id=gr1.jpg){kind=link}
