Phase II study of neoadjuvant checkpoint blockade in patients with surgically resectable undifferentiated pleomorphic sarcoma and dedifferentiated liposarcoma
Chapman, Thomas Andrew
MetadataShow full item record
BACKGROUND: Soft tissue sarcomas (STSs) are a diverse group of cancers that originate from mesenchymal tissue and are estimated to result in 13,130 new cases and 5,350 deaths this year. These neoplasms are hard to detect, which results in physicians struggling to treat late-stage STSs with a limited number of ineffective treatments. Currently, surgical excision is the primary treatment with radiation therapy administered when possible. However, even with optimal margins, the rate of recurrence is high, and the overall survival is low. There is a desperate need for new, more effective treatments. Immune checkpoint blockade (ICB) has recently had widespread success in treating melanoma, and in recent trials, SARC028 and Alliance A091401, have shown demonstrated activity of ICBs in STS in the neoadjuvant setting. Two histological subtypes of STS showed more promise than others: dedifferentiated liposarcoma (DDLPS) and undifferentiated pleomorphic sarcoma (UPS). Another issue plaguing the field of STS is that there is no universal indicator of response. The percentage of hyalinization found within the tumor was recently identified as a better marker of response than radiographic imaging or percent viable tumor. METHODS: This study was an investigator-initiated, single-center, randomized, open-label, phase II study (NCT03307616), in which 23 patients with either DDLPS of the retroperitoneum (RP) or UPS of the extremity/trunk (ET) were separated by disease into Cohort 1 and 2, respectively. Subjects in each cohort were randomized into two neoadjuvant treatment arms per cohort. Arm A (n=7) of Cohort 1 received nivolumab (anti-PD-1) monotherapy, while Arm B (n=7) of Cohort 1 received nivolumab and ipilimumab (anti-CTLA-4) combination therapy. Arm C (n=5) of Cohort 2 received nivolumab monotherapy and radiation therapy, whereas Arm D (n=4) of Cohort 2 received nivolumab/ipilimumab combination therapy and radiation therapy. A tumor biopsy was obtained before treatment, and another sample was taken during the primary treatment of surgical excision. These samples were processed and analyzed by a pathologist who determined the percentage of viable tumor, hyalinization, and necrosis in each sample. Radiographic imaging was also taken throughout to make RECIST 1.1 response determinations. RESULTS: The average treatment response (1 - % viable tumor) for Cohort 1 was 25 ± 23 and there was no difference between Arm A and Arm B, p=0.48. The average treatment response in Cohort 2 was higher at 85 ± 27, but there was also no significant difference between the arms, p = 0.46. The mean percent hyalinization for Cohort 1 was 13 ± 13%, and for Cohort 2 was 69 ± 35%. Again, there was no significant difference between the arms in the Cohort 1 or 2, p = 0.45 and p = 0.54, respectively. Lastly, the mean % necrosis in Cohort 1 was 13 ± 13 %, and in Cohort 2 was 17 ± 24%, and neither had significantly different results in the arms, p = 0.60 and p = 0.92. The RECIST 1.1 results were independent of the arms of the study, and the radiographic response (percent image change) did not correlate to any metric of histologic response. Those who received Ipilimumab had higher rates of adverse events. CONCLUSION: There is significant evidence that ICBs elicited a response in RP DDLPS and ET UPS, and the response of ET UPS was profound. However, there was no apparent benefit from the combination therapy compared to the monotherapy in either cohort. The higher response in ET UPS may be due to the additional radiation therapy or to the nature of UPS itself. Finally, radiographic imaging does not show the response which is apparent at the histological level, so treatment regimens and future experiments should no longer rely on radiographic imaging as a marker for response.