Immunotherapy and Targeted Therapies in Sarcoma: Proposed Synergy with Combination Treatment

Abstract

The combination of targeted therapies and immunotherapies for advanced and metastatic sarcomas has been proposed owing to the enhanced effect of antiangiogenic therapies on the tumor microenvironment. We found eight studies published to date assessing the effectiveness of combined multitargeted vascular endothelial growth factor (VEGF)–tyrosine kinase inhibitors with immune checkpoint inhibitors (ICIs) in sarcoma. It is difficult to draw conclusions owing to limited data and primarily single-arm studies, although initial literature appears promising and requires further study. It remains unknown which sarcoma subtypes may derive the most benefit owing to the limited literature. Benefit was seen primarily in angiosarcoma (AS) and alveolar soft part sarcoma (ASPS), as well as in other tumor subtypes, with few patients achieving complete response (CR). The patients who achieved CRs had desmoplastic small round cell tumor (DSRCT), AS, and chondrosarcoma (CS). Mixed results were found in patients with leiomyosarcoma (LMS), gastrointestinal stromal tumor (GIST), and bone sarcomas, although combination therapy appears to be less effective in these subtypes. Further studies are required to explore optimal treatment agents and dosing strategies to improve both efficacy and tolerability. Although initial results are promising in select patients, phase 3 randomized controlled trials are necessary to determine true treatment effect with combination therapy versus VEGF-inhibitor or ICI alone.

INTRODUCTION

The backbone of standard sarcoma treatment has historically included surgery, radiation, and/or chemotherapy. However, outcomes remain poor for many advanced metastatic subtypes. Owing to the rarity and heterogeneity of sarcoma with more than 175 different subtypes, additional research and development of therapies are necessary to establish new treatment options. Both targeted therapies and immunotherapies offer advantages over traditional chemotherapy with less off-target effects and generally superior tolerability. Additionally, we continue to learn that certain subtypes may respond better to targeted therapy and/or immunotherapy than to traditional chemotherapy owing to each tumor’s unique genomic landscape.

A number of targeted therapies for sarcoma have emerged during the past 2 decades, such as KIT receptor tyrosine kinase inhibitors (TKIs), anti–vascular endothelial growth factor (anti-VEGF), anti–mutant isocitrate dehydrogenase, anti–cyclin-dependent kinase (anti-CDK) 4 and 6, anti–mammalian target of rapamycin (anti-mTOR), anti–colony-stimulating factor 1 receptor (anti–CSF-1R), enhancer of zeste homolog 2, and anti–gamma secretase therapies.[1–8] Targeted treatments have become the mainstay of systemic treatment for gastrointestinal stromal tumor (GIST) with targets such as KIT and platelet-derived growth factor receptor alpha (PDGFRα).[9] The use of anti-VEGF therapies, which inhibit tumor angiogenesis, is well established in the treatment of many sarcoma subtypes. Pazopanib, an oral TKI with multiple targets including VEGF, is approved in several countries for the treatment of advanced soft tissue sarcoma (STS) after prior chemotherapy.[2] Meanwhile, immunotherapy has historically been considered ineffective for sarcoma except in patients with certain immunotherapy markers including microsatellite instability–high (MSI-H), mismatch repair deficiency (dMMR), or tumor mutational burden–high (TMB-H).[10] Several immune checkpoint inhibitors (ICIs) are now available in many countries, including programmed death-1 (PD-1), programmed death-ligand 1 (PD-L1), cytotoxic T-lymphocyte–associated antigen-4 (CTLA-4), and lymphocyte activation gene-3 inhibitors. Currently in the United States, the only US Food and Drug Administration (FDA)–approved use of immunotherapy for sarcoma regardless of immunotherapy marker status is with atezolizumab, an intravenous PD-L1 inhibitor, for alveolar soft part sarcoma (ASPS).[11] However, promising data continue to emerge with the use of immunotherapy in other sarcoma subtypes, such as angiosarcoma (AS), undifferentiated pleomorphic sarcoma (UPS), myxofibrosarcoma (MFS), dedifferentiated liposarcoma (DDLPS), and other sarcoma subtypes.[12,13]

The combination of targeted therapies with immunotherapy has been studied to aid in the significant unmet need for additional therapies after disease progression. The generally favorable tolerability of ICIs allows for the ability to study combinations with other treatments, such as chemotherapy and targeted therapies. Antiangiogenic therapy is the primary target currently studied in combination with immunotherapy for sarcoma owing to its well-established efficacy in many sarcoma subtypes. It has also been proposed that targeted therapies, including anti-VEGF, may help to sensitize cancer cells to the effects of immunotherapy and improve clinical response. The combination of axitinib, an oral multitargeted VEGF inhibitor, and pembrolizumab, an intravenous anti–PD-1 inhibitor, is now a preferred first-line treatment option for ASPS in the United States per the National Comprehensive Cancer Network (NCCN) treatment recommendations, although the combination is not currently FDA approved.[14] However, for most other sarcoma subtypes the use of combination therapy is not well studied.

A literature search using PubMed was conducted in August 2024 using the key terms sarcoma, ICI, and VEGF. Phase 2 and 3 clinical trials published before August 2024 were included. Abstracts and oral presentations presented at the international Connective Tissue Oncology Society conferences from 2020 to 2024 were also reviewed for phase 2 or 3 clinical trials including combination immunotherapy and targeted therapy.

CURRENT COMBINATION STUDIES

The use of combination targeted therapies with immunotherapies has been explored in phase 1 studies, such as with nivolumab plus pazopanib and pembrolizumab plus olaratumab, and in phase 2 studies.[14–25] No phase 3 studies have been conducted to date. We found eight published phase 2 studies assessing the combination of targeted therapies with immunotherapies (summarized in Table 1). One phase 2 study randomly assigned patients to combination treatment versus VEGF-inhibitor treatment alone, and seven were single-arm studies with all patients receiving combination therapy. One study included patients with AS only, one study included patients with osteosarcoma only, and six studies included a mix of patients with soft tissue and bone sarcomas. All studies included an oral anti-VEGF multitargeted TKI. Two studies included cabozantinib, one included pazopanib, one included regorafenib, one included axitinib, one included sunitinib, one included lenvatinib, and one included apatinib. Seven studies included a single-agent ICI with either a PD-1 or PD-L1 inhibitor and one study included dual ICIs with PD-1 and CTLA-4 inhibitors. The ICIs administered include PD-1 inhibitors pembrolizumab, nivolumab, and camrelizumab; PD-L1 inhibitors durvalumab and avelumab; and CTLA-4 inhibitor ipilimumab.[14,17–25]

Table 1.

Summary of combination targeted and immunotherapy trials

Authors (Trial Registration Number)	Study Design	Population	Treatment	Outcomes	Tumor Subtypes with Response	Grade ≥3–4 TRAEs
Cho et al[19] (NCT03798106)	Single-center, single-arm, phase 2 trial	N = 46 with LMS, MPNST, SS, MFS, DSRCT, UPS, DDLPS, clear cell sarcoma, endometrial stromal sarcoma, ASPS, AS, EHE, malignant glomus tumor, and stromal sarcoma	Pazopanib 800 mg PO daily and durvalumab 1500 mg IV q3w	ORR: 30.4% Median PFS: 7.7 mo (95% CI, 5.7–10.4) Median OS: NR (13.9–NR)	1 CR: DSRCT 13 PRs: ASPS, AS, UPS, DSRCT, MPNST, SS, endometrial stromal sarcoma	Neutropenia (19%), elevated AST (15%), elevated ALT (11%), thrombocytopenia (9%)
Cousin et al[20,21] (REGOMUNE; NCT03475953)	Multicenter, single-arm, phase 2 trial	STS: n = 49 with advanced LMS, SS, LPS, UPS, and other (unspecified) GIST: n = 46	Regorafenib 160 mg PO daily on days 1–21 of each 28-day cycle and avelumab 10 mg/kg IV q2w	ORR: 9.3% in STS and 7% in GIST Median PFS: 1.8 mo (1.7–3.5) in STS and 5.5 mo (3.6–7.5) in GIST Median OS: 15.1 mo (7.2–NR) in STS and 19.5 mo (13.7–33.5) in GIST	7 PRs: 4 STS* and 3 GIST	STS: PPE (12%), fatigue (10%), diarrhea (10%) GIST: PPE (18%), hypertension (14%), diarrhea (12%), maculopapular rash (12%)
Grilley-Olson et al[18] (Alliance A091902; NCT04339738)	Multicenter, single-arm, phase 2 trial	N = 21 with advanced AS	Cabozantinib 40 mg PO daily and nivolumab 480 mg IV q4w	ORR: 62% (38–82%) Median PFS: 9.6 mo (5.3–NR) Median OS: 20.5 mo (14.4–NR)	2 CRs and 11 PRs*: cutaneous and noncutaneous AS	Hypertension (10%)
Martin-Broto et al[23] and Palmerini et al[24] (IMMUNOSARC; NCT03277924)	Multicenter, single-arm, phase 1/2 trial	STS: n = 46 with advanced SS, UPS, clear cell sarcoma, SFT, epithelioid sarcoma, AS, EMC, ASPS, and EHE Bone: n = 40 with advanced osteosarcoma, CS, Ewing sarcoma, and bone UPS	Sunitinib 37.5 mg PO daily on days 1–14 and then 25 mg PO daily with nivolumab 3 mg/kg IV q2w starting on day 15	ORR: 13% in STS and 5% in bone Median PFS: 5.6 mo (3.0–8.1) in STS and 3.7 mo (3.4–4.0) in bone Median OS: 24 mo (NR) in STS and 14.2 mo (7.1–21.3) in bone	2 CRs: AS, CS 6 PRs: ASPS, AS, EMC, SS, osteosarcoma	STS: Transaminitis (17%), neutropenia (12%) Bone: neutropenia (10%), anemia (10%), ALT/AST increase (8%)
Movva et al[22] (NCT04784247)	Single-center, single-arm, phase 2 trial	N = 10 in each cohort: advanced LMS, UPS, AS and EHE, SS and MPNST, and osteosarcoma and CS	Lenvatinib 20 mg PO daily starting on day 1 and pembrolizumab 200 mg IV q3w starting on day 15	ORR: 0% in LMS, 33% in SS and MPNST, 10% in osteosarcoma and CS Median PFS: 4.1 mo (1.5–5.3) in LMS, 7.4 mo (1.0–11.8) in SS and MPNST, 4.8 mo (95% CI, 1.8–NR) in osteosarcoma and CS Median OS: not reported	4 PRs: SS, MPNST, osteosarcoma	Hypertension (14%), dyspnea (7%), noncardiac chest pain (7%), syncope (7%)
Van Tine et al[17] (NCT04551430)	Multicenter, randomized, phase 2 trial	N = 105 with metastatic STS (not specified)	Cabozantinib 60 mg PO daily (C) vs. cabozantinib 40 mg PO daily and ipilimumab 1 mg/kg IV q3w ×4 doses and nivolumab 3 mg/kg IV q3w ×4 doses, followed by nivolumab 480 mg IV q4w maintenance (C/N/I)	ORR: 6% with C vs. ORR 11% with C/N/I (p = not significant) Median PFS: 3.8 mo with C vs. 5.4 mo with C/N/I (p = 0.016) Median OS: not reported	C: 2 PRs: LMS C/N/I: 2 CRs and 5 PRs*: LMS, AS, epithelioid sarcoma, MFS	C (>10%): hypothyroidism, diarrhea, oral dysesthesia, fatigue, PPE, hypertension C/N/I (>10%): hypothyroidism, diarrhea, mucositis, oral dysesthesia, nausea, vomiting, elevated AST and ALT, anorexia, dysgeusia, headache, pruritus, maculopapular rash, hypertension
Wilky et al[14] (NCT02636725)	Single-center, single-arm, phase 2 trial	N = 33 with advanced ASPS, high-grade pleomorphic sarcoma, ULMS, NULMS, DDLPS, and other (MPNST, AS, DDCS, Ewing sarcoma, GIST, SS, EHE, and epithelioid sarcoma)	Axitinib 5 mg PO BID starting on day 1 and pembrolizumab 200 mg IV q3w starting on day 8	ORR: 25% (95% CI, 12.1–43.8%) Median PFS: 4.7 mo (3.0–9.4) Median OS: 18.7 mo (12.0–NR)	8 PRs: ASPS, conventional-type epithelioid sarcoma, LMS	Hypertension (15%), autoimmune toxicities (15%), nausea or vomiting (6%), seizures (6%)
Xie et al[25] (NCT03359018)	Single-center, single-arm, phase 2 trial	N = 43 with osteosarcoma	Apatinib 500 mg PO daily and camrelizumab 200 mg IV q2w	ORR: 20.9% Median PFS: 6.2 mo (95% CI, 4.0–6.9) Median OS: 11.3 mo (8.1–14.8)	9 PRs: osteosarcoma	Wound dehiscence (14%)

^*Subtypes with response not specified.

ALT: alanine transaminase; AS: angiosarcoma; ASPS: alveolar soft part sarcoma; AST: aspartate aminotransferase; BID: twice daily; C: cabozantinib; C/N/I: cabozantinib/nivolumab/ipilimumab; CR: complete response; CS: chondrosarcoma; DDCS: dedifferentiated chondrosarcoma; DDLPS: dedifferentiated liposarcoma; DSRCT: desmoplastic small round cell tumor; EHE: epithelioid hemangioendothelioma; EMC: extraskeletal myxoid chondrosarcoma; GIST: gastrointestinal stromal tumor; IV: intravenous; LMS: leiomyosarcoma; LPS: liposarcoma; MFS: myxofibrosarcoma; MPNST: malignant peripheral nerve sheath tumor; NR: not reached; NULMS: nonuterine leiomyosarcoma; ORR: overall response rate; OS: overall survival; PFS: progression-free survival; PO: by mouth; PPE: palmar-planter erythrodysesthesia; PR: partial response; q2w: every 2 weeks; q3w: every 3 weeks; q4w: every 4 weeks; SFT: solitary fibrous tumor; SS: synovial sarcoma; STS: soft tissue sarcoma; TRAE: treatment-related adverse event; ULMS: uterine leiomyosarcoma; UPS: undifferentiated pleomorphic sarcoma.

Cabozantinib, Nivolumab, and Ipilimumab

The only randomized study published to date compared cabozantinib to combination cabozantinib with nivolumab and ipilimumab. Patients with metastatic STS were included with one to two prior lines of therapy. One hundred five patients were randomly assigned 2:1 to cabozantinib 60 mg orally daily or cabozantinib 40 mg orally daily in combination with nivolumab 3 mg/kg and ipilimumab 1 mg/kg intravenously every 3 weeks for four doses then maintenance nivolumab 480 mg every 4 weeks with crossover allowed. The primary endpoint overall response rate (ORR) was 11% (two complete responses [CRs] and five partial responses [PRs]) with combination treatment versus 6% (two PRs) with cabozantinib alone, although the difference was not significant. Responses with combination treatment were seen in patients with leiomyosarcoma (LMS), AS, epithelioid sarcoma, and MFS. The median progression-free survival (PFS) was 5.4 months with combination treatment versus 3.8 months with cabozantinib alone (p = 0.016). The disease control rate was 80% with combination treatment versus 42% with cabozantinib alone (p = 0.0004). Grade 3–4 treatment-related adverse events (TRAEs) occurring in more than 10% of patients receiving cabozantinib alone included hypothyroidism, diarrhea, oral dysesthesia, fatigue, palmar-plantar erythrodysesthesia (PPE), and hypertension. Grade 3–4 TRAEs occurring in more than 10% of patients receiving combination treatment included hypothyroidism, diarrhea, mucositis, oral dysesthesia, nausea, vomiting, elevated aspartate aminotransferase (AST) and alanine transaminase (ALT) levels, anorexia, dysgeusia, headache, pruritus, maculopapular rash, and hypertension.[17]

Cabozantinib and Nivolumab

The Alliance A091902 phase 2 study included patients with locally advanced or metastatic AS receiving cabozantinib and nivolumab after prior taxane-based therapy.[18] All patients were anti-VEGF– and ICI-naïve. The median age was 66 years (range, 32–92) and 57% had cutaneous AS. Twenty-one patients received cabozantinib 40 mg orally daily with nivolumab 480 mg intravenously every 4 weeks. The ORR was 62% (95% CI, 38–82%) with 2 CRs and 11 PRs. Responses were seen in both cutaneous (ORR, 58%) and noncutaneous (ORR, 67%) disease. The median PFS was 9.6 months (95% CI, 5.3–not reached [NR]) and median overall survival (OS) was 20.5 months (95% CI, 14.4–NR). The only TRAE grade 3 or higher was hypertension occurring in 10% of patients.

Axitinib and Pembrolizumab

A phase 2 single-center and single-arm study was published involving 33 patients with advanced or metastatic sarcomas who received axitinib and pembrolizumab. This included patients with ASPS, high-grade pleomorphic sarcoma, uterine leiomyosarcoma (ULMS), nonuterine leiomyosarcoma (NULMS), DDLPS, and other sarcomas including malignant peripheral nerve sheath tumor (MPNST), AS, dedifferentiated chondrosarcoma (DDCS), Ewing sarcoma, GIST, synovial sarcoma (SS), epithelioid hemangioendothelioma (EHE), and epithelioid sarcoma. The median age was 44 years (range, 27–62). Ninety-four percent of patients had metastatic disease and 15% (all ASPS) had treated brain metastases. Thirty-three percent of patients received three or more prior lines of systemic therapy. Fifty-one percent of patients received prior TKI and 15% received prior ICI. The optimal treatment dosing was determined to be axitinib 5 mg orally twice daily starting on day 1 and pembrolizumab 200 mg intravenously every 3 weeks starting on day 8. The primary endpoint of 3-month PFS for all patients was 65.6% (95% CI, 46.6–79.3%) and median PFS was 4.7 months (95% CI, 3.0–9.4). Median OS was 18.7 months (95% CI, 12.0–NR). ORR was 25% (95% CI, 12.1–43.8%). Benefit was seen especially in those with ASPS with 3-month PFS of 72.7% (95% CI, 37.1–90.3%) and ORR of 55% (95% CI, 24.6–81.9%). The best response was PR, which occurred in six patients with ASPS, one with conventional-type epithelioid sarcoma, and one with LMS. Of note, 52% of evaluable patients had positive PD-L1 tumor expression but it was not associated with treatment response. The most common grade 3–4 TRAEs included hypertension (15%), autoimmune toxicities (15%), nausea or vomiting (6%), and seizures (6%).[14]

Pazopanib and Durvalumab

Another single-center, single-arm, phase 2 study included 46 patients with metastatic and/or recurrent STS who received pazopanib and durvalumab. Subtypes included LMS, MPNST, SS, MFS, desmoplastic small round cell tumor (DSRCT), UPS, DDLPS, clear cell sarcoma, endometrial stromal sarcoma, ASPS, AS, EHE, malignant glomus tumor, and stromal sarcoma. The median age was 51 years (range, 22–72) and 85% had received only one line of prior systemic treatment. Patients received pazopanib 800 mg orally daily with durvalumab 1500 mg intravenously once every 3 weeks. The primary endpoint ORR was 30.4% with 1 CR (2%) and 13 PRs (28%). Twenty-seven patients (59%) had stable disease (SD). The median PFS was 7.7 months (95% CI, 5.7–10.4) and median OS was NR (95% CI, 13.9–NR). High CD20+ B-cell infiltration and vessel density were associated with a longer PFS and increased ORR. There was a trend toward longer PFS and OS with PD-L1–positive tumors although not significant. One CR was seen in a patient with DSRCT, and PRs were seen in patients with ASPS, AS, UPS, MPNST, SS, and endometrial stromal sarcoma. The most common grade 3–4 TRAEs were neutropenia (19%), elevated AST levels (15%), elevated ALT levels (11%), and thrombocytopenia (9%).[19]

Regorafenib and Avelumab

The REGOMUNE single-arm, multicenter, phase 2 trial combined regorafenib and avelumab in solid tumors, including one cohort for advanced STS and another cohort for advanced or metastatic GIST. Patients received regorafenib 160 mg orally daily on days 1–21 of each 28-day cycle and avelumab 10 mg/kg intravenously every 2 weeks.[20,21]

The STS cohort included 49 patients with LMS, SS, LPS, UPS, and unspecified other subtypes. The median age was 57 years (range, 21–81). Patients received a median of 2 prior treatments (range, 1–7) and 22% received prior VEGF inhibitor. The primary endpoint ORR was 9.3% with PR as the best response. Tumor subtypes with response were not specified. Forty percent of patients had SD and 33% had tumor shrinkage. The median PFS was 1.8 months (95% CI, 1.7–3.5) and median OS was 15.1 months (95% CI, 7.2–NR). The most common grade 3–4 TRAEs were PPE (12%), fatigue (10%), and diarrhea (10%).[20]

The GIST cohort included 46 patients with 58% having exon 11 KIT mutations. The median age was 64 years (range 26-82). Patients received a median of 2 prior treatments (range, 1–4). The primary endpoint of 6-month PFS was 37% with 7% having PR and 30% having SD. Tumor shrinkage was seen in 52% of patients. The median PFS was 5.5 months (95% CI, 3.6–7.5) and median OS was 19.5 months (95% CI, 13.7–33.5). The most common grade 3–4 TRAEs were PPE (18%), hypertension (14%), diarrhea (12%), and maculopapular rash (12%).[21]

Lenvatinib and Pembrolizumab

An ongoing pilot study combining lenvatinib and pembrolizumab in advanced soft tissue and bone sarcomas has published initial results. This study includes five cohorts for patients with LMS, UPS, AS and EHE, SS and MPNST, and osteosarcoma and chondrosarcoma (CS) with 10 patients in each cohort who had received one to three prior therapies. Patients received lenvatinib 20 mg orally daily starting on day 1 and pembrolizumab 200 intravenously every 21 days starting on day 15. Published results to date include three of the five cohorts. The best response in the LMS cohort (n = 10) was SD (60%) and median PFS was 4.1 months (95% CI, 1.5–5.3). The best response in the SS and MPNST cohort (n = 9) was PR (33%) and 44% with SD. The median PFS was 7.4 months (95% CI, 1.0–11.8). The best response in the osteosarcoma and CS cohort (n = 10) was PR (10%) and 70% with SD. The median PFS was 4.8 months (95% CI, 1.8–NR). The most common grade 3–4 TRAEs were hypertension (14%), dyspnea (7%), noncardiac chest pain (7%), and syncope (7%).[22]

Sunitinib and Nivolumab

The combination of sunitinib and nivolumab is being studied in the ongoing multicenter, single-arm, phase 1/2 IMMUNOSARC trial. One cohort included patients with advanced STS and the second cohort included bone sarcomas. The optimal dose was determined to be sunitinib 37.5 mg orally daily on days 1–14 and then 25 mg orally daily thereafter with nivolumab 3 mg/kg intravenously every 2 weeks starting on day 15.[23,24]

The phase 2 STS cohort included 46 patients with SS, UPS, clear cell sarcoma, solitary fibrous tumor (SFT), epithelioid sarcoma, AS, extraskeletal myxoid chondrosarcoma (EMC), ASPS, and EHE. The median age was 43 years (range, 19–77) and 92% of patients had metastatic disease. The median number of prior treatments was 1 (range, 0–4) and 21% received prior antiangiogenic therapy. The primary endpoint of 6-month PFS was 48% (95% CI, 41–55%). The ORR was 13%. One patient achieved a CR (2%), 5 patients achieved PRs (11%), and 33 patients had SD (72%). The CR occurred in a patient with AS, and PRs were observed in patients with ASPS, AS, EMC, and SS. The median PFS was 5.6 months (95% CI, 3.0–8.1) and OS was 24 months (95% CI, not reported. RNA expression of certain genes, such as platelet-derived growth factor D (PDGFD), interleukin 16 (IL-16), PD-1, and PD-L1, was associated with increased response. The most common grade 3–4 TRAEs were transaminitis (17%) and neutropenia (12%).[23]

The phase 2 bone cohort included 40 patients with OS, CS, Ewing sarcoma, and bone UPS. The median age was 47 years (range, 21–74) and 90% had metastatic disease. The primary endpoint of 6-month PFS was 32%. The median PFS was 3.7 months (95% CI, 3.4–4.0), and median OS was 14.2 months (95% CI, 7.1–21.3). One patient with dedifferentiated CS achieved a CR (3%), 1 patient with osteosarcoma achieved a PR (3%), and 22 patients had SD (55%). The most common grade 3–4 TRAEs were neutropenia (10%), anemia (10%), and ALT/AST increase (8%).[24]

Apatinib and Camrelizumab

Another single-center, single-arm, phase 2 study included 43 patients with advanced or metastatic osteosarcoma who received apatinib and camrelizumab. All patients were antiangiogenic TKI– and anti-PD-1/PD-L1–naïve with progression after chemotherapy. The median age was 41 years (range, 11–43) and 95% had metastatic disease. Patients received apatinib 500 mg orally daily and camrelizumab 200 mg intravenously every 2 weeks. The primary endpoint of 6-month PFS was 50.9% (95% CI, 34.6–65.0%) and 6-month clinical benefit rate was 30.2% (95% CI, 17.2–40.1%). The ORR was 20.9% (all PRs). The median PFS was 6.2 months (95% CI, 4.0–6.9) and median OS was 11.3 months (95% CI, 8.1–14.8). PD-L1 tumor proportion score of 5% or greater and pulmonary metastases were associated with longer PFS. The study did not meet the prespecified target 6-month PFS of 60% or greater. The most common grade 3–4 TRAEs were wound dehiscence (14%), alkaline phosphatase increase (9%), bilirubin increase (9%), hypertriglyceridemia (7%), anorexia (7%), weight loss (7%), and pneumothorax (7%).[25]

DISCUSSION

The combination of targeted therapies with immunotherapies has been proposed to improve efficacy and survival outcomes for patients with sarcoma with historically poor outcomes. However, data to support this synergistic effect are lacking. The efficacy of combination therapy on the tumor microenvironment is well established in renal cell carcinoma and other tumor types. Multiple TKI and ICI combinations are now given as the standard-of-care treatment for renal cell carcinoma, such as cabozantinib and nivolumab with median PFS of 16.6 months (95% CI, 12.5–24.9) and ORR of 55.7%, demonstrating significant benefit when compared to sunitinib alone.[26] However, most sarcomas have been considered cold tumors unresponsive to immunotherapy. Owing to generally limited clinical benefit with ICIs in most sarcoma subtypes, ongoing research explores novel types of immunotherapies, such as with chimeric antigen receptor T cell, T-cell receptor, and tumor-infiltrating lymphocyte therapies in sarcoma.[27–29] The use of antiangiogenic therapies has well-known benefit in many sarcoma subtypes as a single agent. In addition to antitumor benefits, VEGF inhibition has been suggested to increase the efficacy of ICIs in otherwise cold tumors. VEGF promotes tumor angiogenesis and suppresses activation of antitumor immunity in the tumor microenvironment. As demonstrated in Figure 1, inhibition of VEGF can enhance effector T cells; decrease regulatory T cells (Tregs), tumor-associated macrophages, and mast cells; and inhibit myeloid-derived suppressor cells.[30]

Figure 1.

Enhanced effects of VEGF inhibition on the tumor microenvironment. Excess VEGF in the TME supports favorable conditions for tumor cell proliferation while creating an inhibitory environment for antitumor immunity. Inhibition of VEGFR signaling provides antitumor efficacy while reversing immunosuppressive effects of excess VEGF, allowing enhanced activity of immunotherapeutic approaches.

DC: dendritic cell; MDSC: myeloid-derived suppressor cell; TME: tumor microenvironment; T-reg: regulatory T cell; VEGF: vascular endothelial growth factor; VEGFR: vascular endothelial growth factor receptor.

To our knowledge, eight phase 2 studies have been published to date regarding outcomes of combination antiangiogenic targeted therapy with immunotherapy. Most studies were in heavily pretreated patients with advanced or metastatic disease. Only one of these studies was a randomized study that compared cabozantinib alone to cabozantinib with ipilimumab and nivolumab. There was no significant difference in the primary endpoint ORR, although it was higher in the combination group (11% versus 6%), and two patients who received combination therapy achieved a CR. There was also a significant PFS benefit with the median PFS being nearly 2 months longer in the combination group. Full safety data have not yet been published, but as anticipated there were more grade 3–4 TRAEs reported in the combination group.[17] Although combination therapy appears to be tolerable, caution must be taken as certain toxicities can be life-threatening and affect patient quality of life, such as colitis, hepatotoxicity, and thyroid toxicities. The other published studies are all single-arm in varying patient populations and therefore it is difficult to draw conclusions, although they have shown promising tolerability and improved efficacy in comparison to historical controls. The ORR ranged from 0% in an LMS cohort with 10 patients up to 62% in 21 patients with AS.[18,22] Interestingly, CR was achieved in seven total patients in four of the seven studies. In comparison, the ORR in the PALETTE study comparing pazopanib to placebo in previously treated patients with metastatic STS was 9% with pazopanib with all being PRs.[2] The median PFS with single-agent pazopanib was 4.6 months (95% CI, 3.7–4.8) and OS was 12.5 months (95% CI, 10.6–14.8). In this trial patients were also heavily pretreated, with 56% of patients receiving two or more prior lines of treatment and 21% receiving three or more.[2] Whereas few CRs and PRs were observed across the phase 2 studies, most clinical outcomes were SD. It appears that combination therapy may more commonly offer stabilization in most patients rather than tumor shrinkage.

It is difficult to determine the efficacy of combination treatment in different sarcoma subtypes owing to the limited data and the heterogeneity of sarcoma. It is evident that some subtypes will not respond to combination therapy, whereas other subtypes have an enhanced response likely due to the unique tumor biology. Some studies included only STS and others included both STS and bone sarcomas. Of note, the patients who achieved CRs had DSRCT, AS, and CS. Based on current limited literature it appears certain subtypes, such as AS and ASPS, may have increased response to combination therapy. Combination therapy may be less effective for patients with LMS, GIST, and bone sarcomas. However, the findings are mixed in these subtypes. It is known that both AS and ASPS respond to anti-VEGF TKIs and have increased response to ICI, compared to other subtypes, and therefore might explain why clinical benefit was seen with combination therapy.[11,13] The exact mechanism for AS and ASPS responsiveness to immunotherapy remains unknown. For AS, immunotherapy response may be related to site of origin, as it has been suggested that head and neck AS may be more sensitive to ICIs than other sites. AS may also be more likely to express immunogenic markers including TMB-H and PD-L1, although tumors that do not express these markers have also shown response to immunotherapy.[31] Similarly, it has been suggested that ASPS may be more likely to express PD-L1 and have higher tumor lymphocyte infiltration, possibly related to the ASPSCR1::TFE3 fusion gene that is characteristic of the diagnosis.[14,32] Although these two subtypes seem to have higher expression of immunogenic markers predicting immunotherapy response, it appears the tumor biology plays a vital role regardless of specific markers. Axitinib and pembrolizumab for ASPS is currently the only combination treatment recognized by the NCCN as a standard first-line treatment option in the United States, based on the results of Wilky et al.[14] Interestingly, one study noted 0% ORR for patients with LMS,[22] whereas other studies saw response in these patients.[14,17] Of note, an earlier phase 1 trial with combination dasatinib (KIT TKI) and ipilimumab (CTLA-4 inhibitor) involving primarily patients with GIST yielded negative results with no clinical responses seen. It is unclear if this was due to the population with GIST, therapy and dosing selection, or due to lack of synergy between KIT and CTLA-4 inhibitors.[33] Dalal et al[34] previously reported that combination therapy for bone sarcomas may not be effective. However, responses were seen in CS and osteosarcoma. With additional studies, combination therapy may be effective in bone sarcomas as based on prior responses seen with both TKIs and ICIs as single agents in this population. Future studies may consider immunotherapy with TKIs such as dasatinib, regorafenib, or ivosidenib for CS and regorafenib for osteosarcoma.[3,35–41] Further studies should be conducted in patients with GIST and bone sarcomas owing to the unique nature of these tumors as demonstrated in current trials with combination therapy. It is well known that certain immunogenic markers such as MSI-H, dMMR, TMB-H, and PD-1/PD-L1 are associated with increased immunotherapy response. However, it appears that additional biomarkers may be related to response in otherwise cold tumors. Current phase 2 studies reported possible associations with CD20+ B-cell infiltration, vessel density, PDGFD, and IL-16. However, testing was not consistent across studies and may vary significantly by tumor subtype.

Further studies are required to assess the preferred therapeutic agents as well as the optimal dosing and schedule of combination treatment to optimize efficacy and reduce side effects. The specific targets and molecular structure of each TKI and ICI likely contribute to slight differences in efficacy seen across clinical trials, although the optimal combination is unknown. Axitinib is considered one of the most potent anti-VEGF TKIs and inhibits only VEGF, while all other TKIs included in this review are multitargeted as depicted in Table 2. It remains unclear if inhibiting additional targets such as PDGFRα and fibroblast growth factor receptor (FGFR) confer additional efficacy, or merely additional toxicities in sarcoma. PD-1 inhibitors directly target the PD-1 receptor, whereas PD-L1 inhibitors target the ligand. However, a significant difference in treatment efficacy between PD-1 and PD-L1 inhibitors is not expected. The combination of PD-1 and CTLA-4 inhibitors is expected to be more efficacious than PD-1 or PD-L1 inhibition alone but is associated with more toxicities.

Table 2.

Oral VEGF–tyrosine kinase inhibitor targets

Treatment	Molecular Targets
	FGFR1	FGFR2	FGFR3	FGFR4	c-Kit	PDGFRα	PDGFRβ	RET	VEGFR1	VEGFR2	VEGFR3	Other
Apatinib					X			X		X		c-Src
Axitinib									X	X	X
Cabozantinib					X			X	X	X	X	AXL, FLT3, MER, MET, ROS1, TIE2, TRKB, TYRO3
Lenvatinib	X	X	X	X	X	X		X	X	X	X
Pazopanib	X		X		X	X	X		X	X	X	c-Fms, Itk, Lck
Regorafenib	X	X			X	X	X	X	X	X	X	Abl, BRAF, BRAFV600E, DDR2, EPH2A, PTK5, RAF-1, Sapk2, TIE2, Trk2A
Sunitinib	X	X			X	X	X	X	X	X	X	CSF-1R, FLT3

Abl: Abelson leukemia oncogene; c-Fms: cellular homolog of feline sarcoma virus; CSF-1R: colony-stimulating factor 1 receptor; DDR2: discoidin domain receptor 2; EPH2A: ephrin type-A receptor 2; FGFR: fibroblast growth factor receptor; FLT3: Fms-like tyrosine kinase 3; Itk: interleukin-2 inducible T-cell kinase; Lck: lymphocytespecific protein tyrosine kinase; MET: mesenchymal epithelial transition; PDGFR: platelet-derived growth factor receptor; RAF-1: rapidly accelerated fibrosarcoma-1; RET: rearranged during transfection; ROS1: ROS proto-oncogene 1; Sapk2: stress-activated protein kinase 2; TRKB: tropomyosin receptor kinase B; VEGFR: vascular endothelial growth factor receptor.

Many of the current oral anti-VEGF therapies have significant dose-limiting side effects, such as hepatotoxicity, dermatologic toxicity, and hypertension. Of note, a reduced TKI dose was given in four of the studies, and patients were unable to reach planned dose escalation in the axitinib study. A reduced dose of oral anti-VEGF therapy is likely necessary when given in combination with immunotherapy to improve tolerability. Of note, full-dose sunitinib or pazopanib given in combination with nivolumab was not tolerated and the pazopanib and nivolumab arm was closed early owing to dose-limiting side effects in a phase 1b trial in patients with renal cell carcinoma.[42] However, patients with sarcoma tend to be younger, with fewer comorbidities, and may potentially tolerate higher dosing than patients with other tumor types. Some overlapping side effects may also be enhanced with combination therapy, and it can be difficult to discern whether it is from the oral anti-VEGF therapy or immunotherapy, especially hepatotoxicity and dermatologic toxicity, and thyroid toxicity. It also remains unknown whether a lead-in period with targeted therapy prior to the start of immunotherapy is more beneficial than starting both treatments together. A lead-in period has been previously suggested to increase immunotherapy efficacy owing to the targeted therapy’s possible effects on increasing tumor environment sensitivity to immunotherapy. Three of the published studies administered a lead-in dosing strategy with TKI administration for 1–2 weeks prior to ICI start.

CONCLUSIONS

There remains a significant unmet need for new treatment strategies in advanced and metastatic sarcomas. The combination of targeted therapies and immunotherapies, specifically with anti-VEGF and ICIs, may help to improve treatment outcomes. Certain subtypes such as ASPS and AS appear to have increased response rates across different studies. However, current literature with randomized controlled studies is still lacking and further studies are urgently needed. It is also unknown which treatment combination may be most effective or if all anti-VEGF TKI and ICI combinations are equally effective. Initial results appear to be promising, with some patients achieving PRs and CRs among heavily pretreated patients with historically poor outcomes. Future larger, randomized studies are required to determine the efficacy of combination targeted therapy and immunotherapy compared to single-agent treatment and to other treatment options including chemotherapy. The optimal dosing of TKIs and ICIs remains unknown and lower dosing may be required to optimize tolerability. It also remains unclear if a lead-in dosing period of targeted therapies increases the efficiency of ICIs. All studies to date have combined oral multitargeted anti-VEGF TKIs with ICIs. In future studies it may be beneficial to combine other targeted therapies with known efficacy in certain tumor types, such as CDK 4/6 inhibitors in patients with well-differentiated LPS, with immunotherapy. Further developments should also be made to overcome resistance pathways with both TKIs and ICIs. The high cost of both targeted therapies and immunotherapies must also be taken into consideration when administering these treatments.