Treatment outcomes
In the 13 UCSD patients with PD-L1 amplification, nine different malignancies were identified, and all patients had advanced or metastatic disease and were heavily pretreated. Of the nine treated patients, five received anti-PD-1 monotherapy, one received anti-CTLA4/anti-PD-1 combination therapy, and three received a PD-1/PD-L1 inhibitor plus an investigational agent, which was immunotherapeutic, Dr. Goodman said.
The 67% overall response rate was similar to that seen in Hodgkin lymphoma, and many of the responses were durable; median overall survival was not reached.
Of note, genomic analysis in the 13 UCSD patients found to have PD-L1 amplification showed there were 143 total alterations in 70 different genes. All but one patient had amplification of PD-L1, PD-L2, and JAK2, and that one had amplification of PD-L1 and PD-L2.
Of six tumors with tissue available to test for PD-L1 expression by immunohistochemistry, four (67%) tested positive. None were microsatellite high, and tumor-infiltrating lymphocytes were present in five cases.
The tumors that tested negative for PD-L1 expression were from the patient with the rare basal cell cancer, and another with glioblastoma. Both responded to anti-PD1/PD-L1 therapy.
The glioblastoma patient was a 40-year-old man with progressive disease, who underwent standard surgical debulking followed by concurrent radiation therapy plus temozolomide. He progressed soon after completing the concurrent chemoradiation therapy, and genomic profiling revealed 12 alterations, including 9p24.1 amplification, Dr. Goodman said, adding that nivolumab therapy was initiated.
“By week 12, much of the tumor mass had started to resolve, and by week 26 it continued to decrease further. He continues to be in an ongoing partial response at 5.2 months,” he said.
Recommendations
The findings of this study demonstrate that PD-Ll amplification is rare in solid tumors.
“However, PD-L1 amplification appears to be tissue agnostic, as we have seen in over 100 tumor histologies. We also noted that PD-L1 amplification was enriched in many rare tumors with limited treatment options, including anaplastic thyroid cancer, sarcomatoid carcinoma, and some sarcomas. We believe testing for PD-L1 amplification may be warranted given the frequent responses that were durable and seemed to be independent of mutational burden,” he concluded.
Ravindra Uppaluri, MD, session chair and discussant for Dr. Goodman’s presentation, said that Dr. Goodman’s findings should be considered in the context of “the complex biology [of PD-L1/PD-L2] that has evolved over the last few years.”
He specifically mentioned the two patients without PD-L1 expression despite amplification, but with response to immune checkpoint blockade, and noted that “there are several things going on here ... and we really want to look at all these things.”
The PDJ amplicon, especially given “the ability to look at this with the targeted gene panels that many patients are getting,” is clearly contributing to biomarker stratification, said Dr. Uppaluri of Dana-Farber Cancer Institute and Brigham and Women’s Hospital, Boston.
However, it should be assessed as part of a “global biomarker” that includes tumor-infiltrating lymphocytes and tumor mutational burden, he said.
Dr. Goodman reported having no disclosures. Dr. Uppaluri has received grant/research support from NIH/NIDCR, Merck, and V Foundation, and has received honoraria from Merck.
sworcester@frontlinemedcom.com
SOURCE: Goodman A et al. ASCO-SITC, Abstract 47