SAN FRANCISCO – Amplification of programmed death-ligand 1 (PD-L1), also known as cluster of differentiation 274 (CD274), is rare in most solid tumors, but findings from an analysis in which a majority of patients with the alteration experienced durable responses to PD-1/PD-L1 blockade suggest that testing for it may be warranted.
Of 117,344 deidentified cancer patient samples from a large database, only 0.7% had PD-L1 amplification, which was defined as 6 or more copy number alterations (CNAs). The CNAs were found across more than 100 tumor histologies, Aaron Goodman, MD, reported at the ASCO-SITC Clinical Immuno-Oncology.
Of a subset of 2,039 clinically annotated patients from the database, who were seen at the University of California, San Diego (UCSD) Center for Personalized Cancer Therapy, 13 (0.6%) had PD-L1 CNAs, and 9 were treated with immune checkpoint blockade, either alone or in combination with another immunotherapeutic or targeted therapy, after a median of four prior systemic therapies.
The PD-1/PD-L1 blockade response rate in those nine patients was 67%, and median progression-free survival was 15.2 months; three objective responses were ongoing for at least 15 months, said Dr. Goodman of UCSD.
The findings are notable, because in unselected patients, the rates of response to immune checkpoint blockade range from 10% to 20%.
Lessons from cHL and solid tumors
“Over the past few years, investigators have identified numerous biomarkers that can select subgroups of patients with increased likelihoods of responding to PD-1 blockade,” he said, adding that biomarkers include PD-L1 expression by immunohistochemistry, microsatellite instability – with microsatellite instability–high tumors responding extremely well to immunotherapy, tumor mutational burden measured by whole exome sequencing and next generation sequencing, and possibly PD-L1 amplification.
Of note, response rates are high in patients with classical Hodgkin lymphoma (cHL). In general, cHL patients respond well to treatment, with the majority being cured by way of multiagent chemotherapy and radiation.
“But for the subpopulation that fails to respond to chemotherapy or relapses, outcomes still remain suboptimal. Remarkably, in the relapsed/refractory population of Hodgkin lymphoma ... response rates to single agent nivolumab and pembrolizumab were 65% to 87% [in recent studies],” he said. “Long-term follow-up demonstrates that the majority of these responses were durable and lasted over a year.”
The question is why relapsed/refractory cHL patients treated with immune checkpoint blockade have such a higher response rate than is typically seen in patients with solid tumors.
One answer might lie in the recent finding that nearly 100% of cHL tumors harbor amplification of 9p24.1; the 9p24.1 amplicon encodes the genes PD-L1, PD-L2, and JAK2, (and thus is also known as the PDJ amplicon), he explained, adding that “through gene dose-dependent increased expression of PD-L1 ligand on the Hodgkin lymphoma Reed-Sternberg cells, there is also JAK-STAT mediation of further expression of PD-L1 on the Reed-Sternberg cells.
An encounter with a patient with metastatic basal cell carcinoma – a “relatively unusual situation, as the majority of patients are cured with local therapy”– led to interest in looking at 9p24.1 alterations in solid tumors.
The patient had extensive metastatic disease, and had progressed through multiple therapies. Given his limited treatment options, next generation sequencing was performed on a biopsy from his tumor, and it revealed the PTCH1 alteration typical in basal cell carcinoma, as well as amplification of 9p24.1 with PD-L1, PD-L2, and JAK2 amplification. Nivolumab monotherapy was initiated.
“Within 2 months, he had an excellent partial response to therapy, and I’m pleased to say that he’s in an ongoing complete response 2 years later,” Dr. Goodman said.
It was that case that sparked the idea for the current study.
9p24.1 alterations and checkpoint blockade
“With my interest in hematologic malignancies, I was unaware that [9p24.1] amplification could occur in solid tumors, so the first aim was to determine the prevalence of chromosome 9p24.1 alterations in solid tumors. The next was to determine if patients with solid tumors and chromosome 9p24.1 alterations respond to PD-1/PD-L1 checkpoint blockade.
“What is astounding is [that PD-L1 amplification] was found in over 100 unique tumor histologies, although rare in most histologies,” Dr. Goodman said, noting that histologies with a statistically increased prevalence of PD-L1 amplification included breast cancer, head and neck squamous cell carcinoma, lung squamous cell carcinoma, and soft tissue sarcoma.
There also were some rare histologies with increased prevalence of PD-L1 amplification, including nasopharyngeal carcinoma, renal sarcomatoid carcinoma, bladder squamous cell carcinoma, and liver mixed hepatocellular cholangiocarcinoma, he said.
Tumors with a paucity of PD-L1 amplification included colorectal cancer, pancreatic cancer, and cutaneous melanoma, although even these still harbored a few patients with amplification, he said.
A closer look at the mutational burden in amplified vs. unamplified tumors showed a median of 7.4 vs. 3.6 mut/mb, but in the PD-L1 amplified group, 85% still had a low-to intermediate mutational burden of 1-20 mut/mb.
“Microsatellite instability and PD-L1 amplification were not mutually exclusive, but a rare event. Five of the 821 cases with PD-L1 amplification were microsatellite high; these included three carcinomas of unknown origin and two cases of gastrointestinal cancer,” he noted.