Analysis of more than 100,000 patients with cancer for gene CD274 (programmed cell death ligand 1 [PD-L1]) amplification may have implications for treatment with immune checkpoint blockade. Although shown to be rare in solid tumors, copy number alterations in PD-L1 genes were present in more than 100 unique solid tumor histologies. According to data presented at the 2018 ASCO-SITC Clinical Immuno-Oncology Symposium, 67% of patients with solid tumors with PD-L1 amplification responded to programmed cell death protein 1 (PD-1)/PD-L1 blockade, with a median progression-free survival of 15.2 months.¹ What’s more, these responses appeared to be independent of tumor mutational burden, the authors noted.

Aaron M. Goodman, MD

“Testing for PD-L1 amplification is warranted, given the frequent and durable responses seen in solid tumors with PD-L1 amplification treated with checkpoint blockade,” said Aaron M. Goodman, MD, a hematologist/oncologist and Assistant Professor of Medicine at the University of California, San Diego (UCSD). “PD-L1 amplification was also found to be enriched in some rare tumors with limited treatment options, including anaplastic thyroid cancer, renal sarcomatoid carcinoma, and soft-tissue sarcoma.”

Although approximately 10% to 20% of unselected patients respond to PD-1/PD-L1 blockade, investigators have identified numerous biomarkers in recent years that can select subgroups of patients with an increased likelihood of response. These biomarkers include PD-L1 expression by immunohistochemistry, microsatellite instability (with high tumors responding to immunotherapy), tumor mutational burden, and possibly PD-L1 amplification. In classical Hodgkin lymphoma, for example, the overwhelming majority of patients (97%) have copy number alterations in the PD-L1, PD-L2, and JAK2 genes, resulting in high response rates to PD-1/PD-L1 blockade, even in refractory disease.

“Classical Hodgkin lymphoma is exquisitely sensitive to PD‑1/ PD-L1 blockade,” said Dr. Goodman, who noted the prevalence and significance of PD-L1 amplification as a predictor of response to PD-1 blockade are unknown across all solid tumors.

PD-L1 Amplification Rare but Tumor-Agnostic

For this study, Dr. Goodman and colleagues analyzed 118,187 tumor samples from the Foundation Medicine database and a subset of 2,039 clinically annotated patients from the UCSD. The researchers performed comprehensive genomic profiling on all samples, including assessment for microsatellite instability and tumor mutational burden. PD-L1 amplification was called for copy number alterations greater or equal to 6. Investigators also enumerated tumor-infiltrating lymphocytes and PD-L1 immunohistochemistry on select patient samples.

As Dr. Goodman reported, PD-L1 amplification was identified in 0.7% of 118,187 samples but was present in 121 unique solid tumor histologies. Although rare in most histologies, said Dr. Goodman, an increased prevalence was seen in breast carcinoma, head and neck squamous cell carcinoma, lung squamous cell carcinoma, and undifferentiated soft-tissue sarcoma. PD-L1 amplification was also found to be enriched in a few rare tumors with limited treatment options, including anaplastic thyroid cancer, renal sarcomatoid carcinoma, and soft-tissue sarcoma. Tumors with the lowest prevalence of PD-L1 amplification included colorectal adenocarcinoma, pancreatic cancer, and cutaneous melanoma, although these histologies still harbored a few cases, said Dr. Goodman.

The investigators then looked at tumor mutational burden and microsatellite instability in amplified versus unamplified patients. Although the median tumor mutational burden was higher in amplified patients, said Dr. Goodman, the difference was small (7.4 vs 3.6 mutations). Dr. Goodman also noted that 85% of PD-L1–amplified patients still had a low-to-intermediate tumor mutational burden. In addition, high levels of microsatellite instability and PD-L1 amplification were not mutually exclusive, but having both was a rare event. The authors noted that 5 of 821 cases with PD-L1 amplification were microsatellite instability–high.

In the UCSD cohort, 13 of 2,039 patients were found to harbor PD-L1 amplification, with 9 different malignancies identified. All 13 patients had locally advanced/metastatic disease, and all were heavily pretreated, with a median of 4 prior systemic therapies. Of these 13 patients, 9 were treated with PD-1/PD-L1 blockade (either alone or in combination with another immunotherapeutic or targeted therapy). As Dr. Goodman reported, 6 of 9 patients (67%) responded to checkpoint blockade, a rate similar to that observed in Hodgkin lymphoma. Moreover, with a median progression-free survival of 15.2 months, many of these responses were durable, and median overall survival has not yet been reached.

Finally, genomic and biomarker analysis of the UCSD cohort identified 70 genes with 143 alterations. Following immunohistochemistry staining, the investigators found four of six tumors expressed PD-L1. Interestingly, said Dr. Goodman, the two patients who tested negative both responded to immunotherapy. -Tumor-infiltrating lymphocytes were present in all five cases tested, ranging from 10% to 60% per high-powered field). 

This work was done as part of a collaboration with Razelle -Kurzrock, MD, who directs the University of California, San Diego, Center for Personalized Cancer Therapy. Drs. Conelly and Frampton from Foundation Medicine also provided valuable collaboration. ■

DISCLOSURE: Dr. Goodman reported no conflicts of interest.

REFERENCE

1. Goodman AM, Piccioni DE, Kato S, et al: Analysis of over 100,000 patients with cancer for CD274 (PD-L1) amplification: Implications for treatment with immune checkpoint blockade. 2018 ASCO-SITC Clinical Immuno-Oncology Symposium. Abstract 47. Presented January 25, 2018.