A Single-Arm Phase II Trial of Sitravatinib in Advanced Well-Differentiated/Dedifferentiated Liposarcoma

This article is featured in Highlights of This Issue, p. 987

Soft tissue sarcomas (STS) are a heterogeneous group of uncommon solid tumors of mesenchymal origin and more than 150 subtypes have been defined. Liposarcoma (LPS), a sarcoma of adipocytic origin, accounts for approximately 20% of STS diagnoses (1). Reflecting the complexity of STS, LPS is itself subclassified into several biologically distinct subtypes: well-differentiated (WD), dedifferentiated (DD), myxoid/round-cell (MRC), and pleomorphic. At the genomic level, both WD and DD LPS are characterized by amplification of cyclin dependent kinase 4 (CDK4) and murine double minute 2 (MDM2) and represent a spectrum of one disease entity. WD and DD LPS most commonly arise in the extremities and retroperitoneum of middle and older age adults with equal incidence among males and females. There are approximately 3,000 cases diagnosed annually in the United States. WDLPS is a tumor of mature adipocytes characterized by frequent locoregional recurrence but no significant potential for metastatic spread (2). In contrast, DDLPS exhibits a highly cellular and mitotically active spindle cell component and behaves more aggressively with considerable risk for distant metastases, most commonly to lung (3). DDLPS may arise de novo, and up to 40% of WDLPS manifest as DDLPS at the time of recurrence (4). Five-year disease-specific survival (DSS) is 93% for WDLPS and 44% for DDLPS, reflecting the more aggressive clinical behavior of DDLPS (5).

For patients with unresectable or metastatic WD/DD LPS, effective therapeutic options are limited. In a retrospective analysis, chemotherapy, primarily with anthracycline-based regimens, provided objective response rates of 0% for WDLPS and 12% for DDLPS with short progression-free survival (PFS; ref. 6). Trabectedin is an alkaloid cytotoxic agent that binds to the minor groove of DNA and may slow tumor progression. In the LPS subset analysis of a randomized Phase III trial, trabectedin significantly improved median PFS as compared with dacarbazine (3.0 months vs. 1.5 months) resulting in FDA approval; however, trabectedin appears most active in MRC LPS and failed to show a survival benefit (7). Eribulin, a microtubule inhibitor, is approved for advanced LPS on the basis of a randomized Phase III study demonstrating an improvement in overall survival (OS) as compared with dacarbazine (15.6 months vs. 8.4 months) but did not show an improvement in PFS (2.0 months vs. 2.1 months for DDLPS; ref. 8). Recognizing the clinical and molecular heterogeneity of LPS, there is increasing emphasis on evaluating novel agents in subtype-specific trials. The CDK4/6 inhibitors palbociclib and abemaciclib and the MDM2 inhibitor BI-907828 have demonstrated activity specifically in WD/DD LPS in single-arm studies and randomized trials are ongoing (9).

Receptor tyrosine kinases (RTK) play a critical role in transducing extracellular signals into the cell to elicit various biologic responses. In cancer, RTKs may be dysregulated by gene amplification, point mutation, increased transcription or translation, and autocrine growth factor secretion. Dysregulated RTKs play an important role in tumor development by influencing cell growth, differentiation, migration, angiogenesis, and apoptosis. Pazopanib, an RTK inhibitor primarily active on the angiogenic receptors VEGFR and PDGFRα/ß, was approved for the treatment of advanced non-adipocytic sarcomas in 2012. LPS was excluded from the registration-directed Phase III study of pazopanib after demonstrating an unfavorable progression-free rate at 12 weeks (PFR₁₂) in the preceding Phase II trial (10). Development of an effective RTK inhibitor for WD/DD LPS is of significant therapeutic interest.

We conducted preclinical investigations characterizing activated RTKs and signaling pathways in sarcoma and these results provide insight into the lack of clinical activity associated with pazopanib and other previously studied RTK inhibitors in WD/DD LPS, including regorafenib and sorafenib (11–13). We found that numerous RTKs, including IGF-1R, MET, and PDGFRα/ß, contribute to liposarcomagenesis, and several of these are not effectively suppressed by RTK inhibitors that are primarily active on angiogenic targets. Sitravatinib is an orally available, potent, small molecule inhibitor of several related RTKs at low nanomolar concentrations including DDR2 (0.5 nmol/L), AXL (1.5 nmol/L), VEGFR1–3 (2–6 nmol/L), KIT (6 nmol/L), FLT3 (8 nmol/L), MET (20 nmol/L), PDGFRα (30 nmol/L), RET (44 nmol/L), and others, including IGF1-R.

In preclinical studies, the DDLPS cell line DDLS expressed phosphorylated PDGFRa/ß, FGFR2, MET, ALK, AXL, and EphA4, whereas another DDLPS cell line, LS141, expressed phosphorylated PDGFRa/ß and IGF1-R (11). Thus, multiple phospho-RTKs were expressed at baseline. In LS141, siRNA knockdown of both IGF1-R and PDGFRß reduced cell viability by 80% and was significantly more effective than knockdown of either RTK alone. Sitravatinib was more effective at inhibiting proliferation of DDLPS cell lines as compared with the more narrowly targeted RTK inhibitors pazopanib, imatinib, and crizotinib, with greater suppression of phosphorylated PDGFRß, IGFR1-R, KIT, and downstream signaling pathway components AKT and S6. The drug also effectively suppressed tumor growth in a DDLPS mouse xenograft model as compared with no treatment and other tyrosine kinase inhibitors (TKI).

Other preclinical studies have implicated similar RTKs in liposarcomagenesis and reinforce these observations. Peng and colleagues found that MET, AXL, and IGF1-R were overexpressed in DDLPS cell lines as compared with adipocytes and pre-adipocytes (14). In preclinical studies reported by Pollack and colleagues, phospho-MET was highly expressed in DDLPS and shRNA knockdown reduced cellular proliferation, migration, and invasion (15). EMD1214063, a MET-specific TKI, significantly reduced tumor growth in DDLPS mouse xenograft models. Collectively, these observations highlight the complexity of activated RTKs and signaling pathways in WD/DD LPS and suggest that multiple RTKs contribute to tumor growth thus implicating a role for more broadly-acting TKIs, such as sitravatinib, that are also active on non-angiogenic targets. In a Phase I study of sitravatinib in advanced solid tumors, the recommended Phase II dose was 150 mg orally daily and common toxicities included fatigue, diarrhea, and hypertension (16). The Phase II study reported here was designed to evaluate the efficacy and safety of sitravatinib in patients with WD/DD LPS and to explore whether basal expression of phospho-RTKs and signaling pathway proteins could be used as a predictive marker of response.

Patients 18 years or older with histologically confirmed WD or DD LPS that was unresectable or metastatic with measurable disease per RECIST criteria version 1.1 were eligible. Patients must have received prior treatment with at least one systemic regimen (in the neoadjuvant, adjuvant, or metastatic setting) and have radiographic evidence of disease progression within 12 weeks of enrollment as determined by the treating investigator. There was no upper limit on the number of prior lines of therapy. An Eastern Cooperative Oncology Group performance status (ECOG PS) ≤1 and adequate organ and bone marrow function as defined by creatinine clearance >45 mL/min, total bilirubin ≤1.5 × upper limit of normal (ULN), aspartate aminotransferase (AST) ≤1.5 × ULN, alanine aminotransferase (ALT) ≤1.5 × ULN, absolute neutrophil count ≥1,500/mm³, and platelets ≥100,000/mm³ was required. In addition, patients were required to have a blood pressure ≤150/100 mmHg and normal left ventricular systolic function on transthoracic echocardiography during the screening period.

Patients may not have received anticancer treatment within 28 days of initiating sitravatinib except for small molecule-targeted agents, for which the washout period was 14 days or 7 half-lives, whichever was longer. Toxic effects from prior therapy must have resolved to grade 1 or baseline. Concurrent treatment with proton pump inhibitors, H2 blockers, or medications known to cause QTc prolongation was prohibited. The protocol was approved by the Institutional Review Board of the participating centers and conducted in accordance with an assurance filed with and approved by the Department of Health and Human Services, International Conference on Harmonization requirements for Good Clinical Practice and ethical principles defined in the Declaration of Helsinki. All patients provided written informed consent.

This Phase 2, single-arm, multi-center clinical trial (NCT02978859) used a Simon 2 stage design and recruited patients from three academic medical centers in the United States: Columbia University Irving Medical Center, Siteman Cancer Center at Washington University in St. Louis, and Massachusetts General Hospital Cancer Center. The primary endpoint was the PFR₁₂. Secondary endpoints were objective response, PFS, OS, and safety/tolerability.

Patients were treated with sitravatinib 150 mg orally daily in continuous 21-day cycles. In May 2018, after enrollment of 8 patients, the starting dose of sitravatinib was reduced to 120 mg orally daily in continuous 21-day cycles based on emerging tolerability data from the sitravatinib drug development program and at the recommendation of the pharmaceutical company, Mirati Therapeutics. Patients already enrolled on the study were allowed to continue at 150 mg daily or reduce to 120 mg daily at the discretion of the treating investigator. Dose reductions to 100 and 80 mg were allowed for toxicity. If further dose reduction was needed, the drug was discontinued.

Patients continued treatment until disease progression, unacceptable adverse events, withdrawal of consent, or intercurrent illness that prevented further administration of sitravatinib. Clinical examinations and laboratory testing were performed at screening, weekly during cycle 1, at the start of each subsequent cycle, and at the end of treatment. Disease status was assessed using CT imaging performed every 2 cycles (6 weeks) for the first 12 cycles (36 weeks) of study treatment and then every 3 cycles (9 weeks) thereafter, regardless of dose delays. Toxicity was assessed at all clinical visits and graded according to NCI CTCAE version 4.0. Patients were followed for information on disease status, subsequent anticancer therapy, and survival every 6 months following the end of treatment until death or 3 years, whichever came first.

Paired tumor biopsies were to be obtained from the first 10 patients accrued to the study. Biopsies were obtained within 14 days of starting treatment and again on cycle 2 day 15. The tumor tissue was used for RPPA performed by Theralink Technologies for phospho-RTK and phospho-protein expression/activation levels. Samples were received as formalin-fixed paraffin-embedded (FFPE) tissue blocks and sections were prepared on uncharged glass slides. Tissue blocks were sectioned at 8 μm. One tissue section from each sample was hematoxylin and eosin stained and annotated in order to guide laser capture microdissection (LCM)-based enrichment of tumor cells from heterogeneous tissue biopsy sample input. LCM was attempted on each sample.

RPPA were printed according to a previously described protocol (17). Sample lysates were printed in technical triplicates on nitrocellulose backed slides (ONCYTE Avid; Grace Bio-Labs) alongside analyte controls, total protein standards, and matrix blanks using an Aushon 2470 arrayer (Quanterix) equipped with 185 μmol/L pins. Nitrocellulose slides were stored at −20 °C prior to antibody staining. Slides were incubated with primary antibodies for 30 minutes at room temperature and each staining run included a single-negative control slide (only antibody diluent; no primary antibody). RPPA, image capture, and post-processing were performed as described previously (18).

The primary endpoint was the PFR₁₂, which corresponds to the number of patients alive and without evidence of disease progression per RECIST criteria v1.1 at the 12-week scan after starting treatment out of all evaluable patients. Evaluable patients were those who took at least one dose of sitravatinib. On the basis of historical controls at the time this study was designed, PFR₁₂ ≥40% was considered potentially active and worthy of further study, whereas PFR₁₂ ≤20% was considered inactive (19). A Simon 2-stage design was used for comparison of the PFR₁₂ versus these historical controls. In stage one, 13 patients were enrolled. If 3 or more patients met the PFR₁₂ endpoint, the study would proceed to the second stage and enroll an additional 16 patients. If 9 or more patients from the total population of 29 met the PFR₁₂ endpoint, sitravatinib would be considered active and promising for further study in WD/DD LPS. The design provided 85% power with α = 0.10.

The ORR was defined as the number of patients having a best objective tumor status of complete or partial response per RECIST criteria lasting at least 4 weeks divided by the number of evaluable patients. PFS was defined from the date of enrollment to the date of progression or death, the last progression-free scan for patients who withdrew consent or had unexpected adverse events, or the last follow-up for patients who withdrew consent prior to the first scan, whichever occurred first. OS was defined from the date of enrollment to the date of death or last follow-up, whichever occurred first. PFS and OS were estimated using the Kaplan–Meier method. Adverse events were categorized as counts and percentages per adverse event by grade. Results from RPPA were evaluated in an exploratory fashion without adjustment for multiple comparisons. The mean protein fluorescence intensity values were normalized to the total amount of protein present in each sample and were compared between patients with clinical benefit (meeting the PFR₁₂ endpoint) versus those without clinical benefit using a two-tailed two sample t test.

The data generated in this study are available upon request from the corresponding author.

Between March 2017 and August 2020, 29 patients from three centers were enrolled and treated with sitravatinib. The baseline characteristics of these patients are listed in Table 1 and information on the representativeness of this population is provided in Supplementary Table S1. The median age was 62 years (range: 28–88), 16 (55%) were male, and 22 (76%) were Caucasian. Twenty-seven of 29 (93%) patients had DDLPS or mixed WD/DD LPS, whereas 2 of 29 (7%) had purely WDLPS. All patients had evidence of investigator-assessed disease progression within 12 weeks prior to starting treatment. All patients had received at least one prior line of systemic therapy (range: 1–4) and 31% had received three or more prior lines. The most common prior therapies were CDK4/6 inhibitors (41%), anthracycline-based (41%) or gemcitabine-based (34%) chemotherapy, eribulin (25%), MDM2 inhibitors (17%), or immunotherapy agents (17%). As of the May 31, 2022 data cut-off, all 29 patients had discontinued treatment. Reasons for discontinuation included disease progression (24 patients, 83%), adverse events (3 patients, 10%), and withdrawal of consent (2 patients, 7%).

In the first stage of the Simon two-stage design, 5 of 13 patients (38%) met the PFR₁₂ endpoint. Therefore, the study proceeded to full accrual and enrolled an additional 16 patients with WD/DD LPS in stage 2. Overall, 12 of 29 patients (41%) were progression free on imaging performed at 12 weeks and the study met the prespecified primary efficacy endpoint. Among the 8 patients treated with a starting dose of 150 mg, 4 of 8 (50%) met the PFR₁₂ endpoint compared with 8 of 21 (38%) for those who started at 120 mg. Among the 2 patients with purely WDLPS, both (100%) met the PFR₁₂ endpoint, whereas 10 of 27 patients (37%) with purely DDLPS or mixed tumors were progression free at the 12-week scan. The confirmed ORR per RECIST v1.1 criteria was 0%. One patient achieved a partial response but treatment was interrupted due to concurrent COVID-19 infection and the confirmatory scan could not be obtained. Overall, 8 of 29 patients (28%) had some evidence of tumor regression while on sitravatinib treatment. The waterfall and swimmer's plots are shown in Fig. 1. Five patients (all with DDLPS) had stable disease for more than 30 weeks. Interestingly, genomic sequencing results performed for routine clinical care were available for 2 of these 5 patients and both showed amplification of fibroblast growth factor receptor substrate 2 (FRS2) in addition to characteristic CDK4 and MDM2 amplification. Three patients (all with DDLPS) had stable disease lasting more than 1 year.

The median PFS for the entire population was 11.7 weeks [95% confidence interval (CI), 5.9–35.9 weeks]. The median PFS excluding the 2 patients with purely WDLPS was 10.9 weeks (95% CI, 5.9–35.9 weeks). The PFS for the 2 patients with purely WDLPS was 17.7 and 17.0 weeks, respectively. Median OS for the entire population was 31.7 weeks (95% CI, 18.1–90.1 weeks). The Kaplan–Meier plots for PFS and OS are shown in Fig. 2.

Patient 001–06 had metastatic DDLPS and initiated treatment with sitravatinib in August 2019 after progression on the CDK4/6 inhibitor abemaciclib. He achieved a partial response per RECIST criteria in March 2020 (Fig. 3). He then developed symptomatic COVID-19 infection at the onset of the pandemic. In the absence of formal guidelines and without knowledge of the impact of sitravatinib on active COVID-19 infection, we elected to interrupt treatment for 5 weeks and confirmatory scans could not be obtained. CT imaging in June 2020, immediately after his recovery from COVID-19, demonstrated progressive disease. Because the patient had experienced clinical benefit on sitravatinib and after discussion with the Sponsor, we elected to resume study treatment with approval from the Institutional Review Board. He then experienced prolonged stable disease until imaging in July 2021 demonstrated further progression and sitravatinib was discontinued. For the purposes of the statistical analysis, he was considered to have progressed at the time of the first scan showing progression in June 2020.

Sitravatinib was generally well tolerated. Initially, the protocol specified a starting dose of 150 mg daily. After 8 patients were enrolled, the study was amended to use a starting dose of 120 mg based on tolerability data from the sitravatinib program and the recommendation of Mirati Therapeutics. Although 4 of the 8 patients who started at 150 mg required dose reduction at some point during treatment, only 2 of the 21 patients who started at 120 mg required dose reduction.

Treatment-related adverse events (TRAE) are listed in Table 2. The most common all-grade TRAEs included diarrhea (59%), hypertension (52%), fatigue (45%), hoarseness (41%), mucositis (31%), and nausea (31%). Hypertension, although common, was manageable with anti-hypertensive medications.

Thirteen of 29 patients (45%) experienced grade 3 or higher TRAEs, which included hypertension (24%), fatigue (7%), hyponatremia (7%), oral mucositis (3%), anemia (3%), elevated lipase (3%), and decrease in left ventricular ejection function (LVEF; 3%). The patient with left ventricular dysfunction was treated with sitravatinib 150 mg. A decrease in LVEF was observed from 60% to 45% which recovered within 1 month of interrupting sitravatinib. Two of 29 patients (7%) experienced grade 4 TRAEs. The grade 4 events were hypertension and posterior reversible leukoencephalopathy syndrome (PRES). In the patient experiencing PRES who was treated with sitravatinib 120 mg daily, symptoms resolved and the drug was permanently discontinued.

When examined by starting dose, among the 8 patients starting at 150 mg, 5 (62%) experienced a grade 3 or higher TRAE including 3 (37%) with grade 3 or higher hypertension, and 1 with grade 3 mucositis. Among the 21 patients treated with a starting dose of 120 mg, 8 (38%) experienced grade 3 or higher TRAEs with the most frequent event being hypertension (19%).

A subset of patients underwent paired tumor biopsies that were analyzed using RPPA to quantify and measure levels of RTK activation and downstream signaling architecture. Because of limitations in tumor cell content from these biopsies and high background, we elected to focus our RPPA studies on specimens that successfully underwent LCM where we were most confident in the fidelity and accuracy of the phosphoprotein measurements obtained (20, 21). LCM specimens were available for 6 patients at the pretreatment timepoint; however, only 1 patient had an on-treatment sample for LCM. Of the 6 patients with pretreatment LCM specimens, 2 met the PFR₁₂ endpoint (had clinical benefit) and 6 failed to meet the PFR₁₂ endpoint (did not have clinical benefit). The 1 patient with the paired specimen did not meet the PFR₁₂ endpoint.

RPPA was performed to evaluate the expression and activation levels of total and phosphorylated RTKs including AXL, EGFR, IGF1-R, KIT, MET, PDFGRα/ß, and VEGFR2, as well as activation of downstream AKT/mTOR and MAPK signaling pathway proteins. We observed measurable, varying levels of baseline activation/phosphorylation of all RTKs in all six pretreatment samples. We compared the average level of baseline activation in the pretreatment samples between the 2 patients with clinical benefit and the 4 patients without clinical benefit (Fig. 4). Baseline activation/phosphorylation of RTKs did not appear to be significantly different between patients with clinical benefit and those without clinical benefit with the exception of PDGFRα, where lower baseline activation was observed in patients with clinical benefit, a finding associated with borderline statistical significance (P = 0.052). Among the proteins involved in downstream AKT-mTOR and MAPK signaling, higher baseline levels of activation in the mTOR pathway protein p706K (T39) and activated MEK (S217/S221) and ERK (T202/Y204) were observed in pretreatment tumor biopsies from patients with clinical benefit as compared to those without clinical benefit, but these differences were not statistically significant.

One patient, who did not have clinical benefit from sitravatinib, had paired biopsy samples that underwent LCM and RPPA. For this patient, two samples from the pre-treatment timepoint, and one sample from the on-treatment timepoint, were available. We observed decreased activation of phospho-PDGFRα (Y754) between the pretreatment and on-treatment samples; however, none of the other phospho-RTKs or signaling pathway proteins were suppressed, and several components of the Akt–mTOR pathway (including p706K T39) appeared to be activated on treatment (Fig. 5). Without LCM paired biopsies from patients with clinical benefit, we were unable to determine which RTKs were effectively suppressed by sitravatinib and might have been relevant to mechanism.

In this prospective, Phase II, single-arm, multicenter clinical trial, sitravatinib met the prespecified primary efficacy endpoint with 41% of patients with WD/DD LPS remaining alive and without evidence of disease progression after 12 weeks of treatment. Among 2 patients with purely WDLPS and 27 with DDLPS or mixed tumors, 100% and 37% met the PFR₁₂ endpoint, respectively. All patients had unresectable tumors and demonstrated evidence of disease progression prior to starting sitravatinib. Although purely WDLPS exhibits a more indolent natural history than DDLPS, these tumors still cause considerable morbidity and require systemic treatment when surgical options are exhausted. For this reason, patients with progressive, unresectable WDLPS were included on this study. Sitravatinib was generally well tolerated and most adverse events were low grade and consistent with the known side effect profile of the drug. Hypertension was common and required careful monitoring but was easily managed with anti-hypertensive medications. One event of PRES in the setting of hypertension was observed but was reversible.

Gastrointestinal stromal tumor (GIST), dermatofibrosarcoma protuberans (DFSP), and tenosynovial giant cell tumor (TCGT) represent sarcoma subtypes with well-characterized molecular alterations resulting in oncogenic dependency upon one RTK. In other subtypes of soft tissue sarcoma, multiple RTKs may contribute to sarcomagenesis in the absence of a single recurrent activating mutation, gene fusion, or amplification. Indeed, highly recurrent genomic alterations in RTKs appear uncommon in WD/DD LPS. In one report, amplification of RTK genes was observed in 18 of 56 (32%) of WD/DD LPS patient samples and included DDR2, ERBB3, NTRK1, FGFR3, ROS1, and IGF1-R; however, none of these alterations were recurrent (22). In LPS and other STS subtypes, multiple RTKs may cooperate in support of the malignant phenotype with heterogeneity among patients and even sites of disease within a single patient. Indeed, we observed variable expression of multiple phospho-RTKs in pretreatment biopsy specimens obtained from the patients treated on this study.

Previously completed studies with other TKIs including pazopanib, regorafenib, and sorafenib showed limited activity in LPS (median PFS of 2.6, 1.9, and 2.0 months, respectively; refs. 10, 12, 13). Encouraging activity has been observed with catequetinib, another broadly-acting TKI, which shares overlapping targets with sitravatinib. Catequetinib is active upon VEGFR2–3, PDGFRα/ß, FGFR1–4, KIT, RET, and CSF1-R. In a Phase II single-arm study conducted in China, among 13 patients with LPS, the median PFS and PFR₁₂ were 5.6 months and 63%, respectively (23). However, this study did not report the proportion of patients with WD versus DD LPS, progression was not required prior to enrollment, generalizability to U.S. and European populations is uncertain, and a subsequent randomized trial appeared to show lesser efficacy.

Although sitravatinib met the prespecified efficacy endpoint, provided prolonged disease control in a subset of patients, and, acknowledging the limitations of cross-trial comparisons, was potentially superior to the previously reported activity of several other TKIs in WD/DD LPS, objective responses did not occur and median PFS was limited. There are several possible explanations for the lack of efficacy observed in some patients. Although sitravatinib inhibits a broad range of RTKs, compensatory activation of alternative RTKs and signaling pathways likely still occurs. The spectrum of activated RTKs may be divergent across different WD/DD LPS tumors raising challenges for this drug development approach, as some of the bypass RTKs and signaling pathways may not be targets of sitravatinib. Finally, CDK4 amplification, which acts downstream of RTKs and the Akt-mTOR and MAPK pathways, may render WD/DD LPS less sensitive to the mitogenic effects of RTKs in comparison with sarcoma subtypes without cell-cycle alterations. Indeed, this observation may underlie findings from multi-histology studies, where WD/DD LPS has consistently shown lesser sensitivity to TKIs than other sarcoma subtypes (10, 13). Therefore, combination strategies, including those with CDK4/6 inhibitors, may be worthy of evaluation.

To explore the association of the functional activation of the signaling architecture of RTK targets of sitravatinib with clinical benefit, we utilized an LCM-RPPA workflow with tumor biopsies from a subset of patients. We observed that tumor cells from all patients had activation of all phospho-RTKs to varying extents in baseline samples reflecting the heterogeneity of activated RTKs present in WD/DD LPS. There was no evident correlation between pretreatment activation levels of a particular phospho-RTK and clinical benefit from sitravatinib, aside from phospho-PDGFRα, where lower levels of expression were noted among those with clinical benefit, a finding that was of borderline statistical significance (P = 0.052). We observed that expression of phosphoproteins in the AKT-mTOR and MAPK pathways (specifically p70S6K T39, pMEK S217/S221, and pERK T202/Y204) were higher in patients with clinical benefit versus those without, although these observations were not statistically significant. These proteins may serve as downstream “functional integrators” of sitravatinib's RTK targets and higher baseline levels of activation/phosphorylation may reflect greater overall drug target/RTK activation and predict sensitivity to the drug. However, given the small number of patients with biopsies, these findings are highly exploratory and hypothesis-generating. Unfortunately, a paired sample was only available from one patient who failed to derive clinical benefit from sitravatinib. Apart from phospho-PDGFRa (Y754), we did not observe suppression of phospho-RTKs or signaling pathway components on treatment in this patient and indeed some components of the Akt-mTOR pathway appeared induced, perhaps due to release of negative feedback, including the activation of PDGFRß. Although these observations may explain the absence of benefit in this particular patient, the lack of paired samples from patients with clinical benefit limited our ability to determine whether sitravatinib was effective at suppressing phospho-RTK targets and whether suppression of a particular phospho-RTK correlated with clinical outcomes.

Limitations of this study include the need to modify the sitravatinib dose during the conduct of the study and the lack of other outcome measures such as the growth modulation index and quality of life indicators which are relevant to this patient population. Strengths of the study include the requirement for disease progression within 12 weeks of starting treatment and our specific focus on WD/DD LPS, considering the biologic differences between WD/DD LPS and the other LPS subtypes (myxoid and pleomorphic).

In summary, we report the clinical activity of sitravatinib, a multi-receptor TKI which was selected for further study in WD/DD LPS on the basis of supportive preclinical investigation. Sitravatinib provided a PFR₁₂ of 41%, median PFS of 11.7 weeks, and was well tolerated, although patients required careful monitoring for hypertension. Sitravatinib is presently under evaluation in a pivotal Phase III clinical trial in combination with nivolumab for patients with non–small cell lung cancer who have received prior chemoimmunotherapy. Exploratory clinical trials with sitravatinib are ongoing in renal cell carcinoma, urothelial, breast, ovarian, and other cancer types alone or in combination with immunotherapy. Further clinical study of sitravatinib as monotherapy or in combination with other agents in liposarcoma is presently being evaluated.

M. Ingham reports grants from Mirati Therapeutics during the conduct of the study; M. Ingham also reports grants from Apexigen, PTC Therapeutics, Merck, Codiak Biosciences, Inhibrx, GSK Plc, Deciphera, Bioatla, Intensity Therapeutics, Astellas, and APIM Therapeutics, as well as personal fees from Caris Life Sciences, Epizyme, PTC Therapeutics, and Apexigen outside the submitted work. S. Lee reports grants from Mirati Therapeutics during the conduct of the study, as well as personal fees from PTC Therapeutics outside the submitted work. B.A. Van Tine reports other support from Accuronix Therapeutics and Polaris, as well as personal fees from Epizyme, ADRx, Ayala Pharmaceuticals, Bayer, Daiichi Sankyo, Targeted Oncology, Intellisphere LLC, Bionest Partners, Adaptimmune, Apexigen Inc., Deciphera Pharmaceuticals, Inc., PTC Therapuetics, Boehringer Ingelheim, Agenus, and Regeneron Pharmaceuticals outside the submitted work. E. Choy reports personal fees from Bayer, Epizyme, Daiichi Sankyo, and Adaptimmune outside the submitted work. P. Oppelt reports other support from Merck outside the submitted work. G. Cote reports other support from Mirati during the conduct of the study. In addition, G. Cote reports advisory board fees, compound, and support paid to institution for the conduct of clinical trials from PharmaMar and Eisai; compound and support paid to institution for the conduct of clinical trials from Merck KGaA / EMD Serono Research and Development Institute, Bayer, and Jazz Pharmaceuticals; support paid to institution for the conduct of clinical trials from Servier Pharmaceuticals, Macrogenics, Bavarian-Nordic, SpringWorks Therapeutics, Repare Therapeutics, SMP Oncology, RAIN Therapeutics, BioAtla, and Inhibrx; and advisory board fees and support paid to institution for the conduct of clinical trials from Epizyme, Foghorn, Ikena, and C4 Therapeutics. E. Petricoin reports personal fees from Theralink Technologies, Inc., Ceres Nanosciences, Inc., and Perthera, Inc. outside the submitted work. No disclosures were reported by the other authors.

M. Ingham: Conceptualization, formal analysis, investigation, writing–original draft, writing–review and editing. S. Lee: Formal analysis, investigation, writing–original draft, writing–review and editing. B.A. Van Tine: Formal analysis, investigation, writing–original draft, writing–review and editing. E. Choy: Formal analysis, investigation, writing–original draft, writing–review and editing. J. Oza: Investigation, writing–original draft, writing–review and editing. S. Doshi: Investigation, writing–original draft, writing–review and editing. L. Ge: Formal analysis, investigation, writing–original draft, writing–review and editing. P. Oppelt: Formal analysis, investigation, writing–original draft, writing–review and editing. G. Cote: Formal analysis, investigation, writing–original draft, writing–review and editing. B. Corgiat: Formal analysis, investigation, writing–original draft, writing–review and editing. N. Sender: Data curation, project administration. S. Sta Ana: Data curation, project administration. L. Panchalingam: Data curation, project administration. E. Petricoin: Formal analysis, writing–original draft, writing–review and editing. G.K. Schwartz: Conceptualization, formal analysis, supervision, investigation, writing–original draft, writing–review and editing.

Funding to support this clinical trial was provided to Columbia University Irving Medical Center by Mirati Therapeutics.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.