The data are early and the sample size small, but pembrolizumab (Keytruda, Merck & Co) used in the neoadjuvant and adjuvant setting showed antitumor activity in locally advanced head and neck squamous cell carcinoma (HNSCC), according to new findings.
No serious drug-related adverse events or unexpected surgical delays or complications occurred, explained lead author, Ravindra Uppaluri, MD, PhD, director, Head and Neck Surgical Oncology at the Dana Farber Cancer Institute, Boston, Massachusetts, who presented the findings here at the American Society of Clinical Oncology (ASCO) 2017 Annual Meeting.
"We also have an early signal of potential clinical efficacy," he said. "The numbers are small, but there were no local regional relapses or distant metastases for 14 patients at or more than one year after surgery, in this very high risk subgroup of patients."
"But most intriguingly, we’ve seen a pathologic response in 42% of patients with just a single dose," he added. "This was a finding we were surprised to see."
The Food and Drug Administration approved pembrolizumab in 2016 for patients with recurrent or metastatic HNSCC who experienced disease progression during or after platinum-containing chemotherapy. It has also been approved for use in melanoma and non-small cell lung cancer.
But outside the metastatic setting, Dr Uppaluri noted that previous studies have provided level 1 supporting the use of adjuvant chemoradiotherapy in patients with resected, high-risk, locally advanced HNSCC. But despite this, 35% of patients develop recurrent and metastatic disease in high-risk HNSCC within 1 year after postoperative adjuvant chemoradiotherapy.
"Given that, the question is if the addition of pembrolizumab can improve these outcomes, Dr Uppaluri said.
In the current phase 2 trial, patients receive a single dose of neoadjuvant pembrolizumab and then undergo surgery 2 to 3 weeks later.
Postoperative treatment largely depended on the patient. Those who were low risk or intermediate risk, such as those with negative extracapsular extension (ECE) margins received standard of care, while patients with high-risk features, including positive ECE margins, underwent adjuvant chemoradiotherapy followed by maintenance pembrolizumab.
The primary endpoint of the study is to reduce the rate of local regional relapse and distant metastasis from 35% to 15%. Secondary endpoints are safety and correlative biomarkers and genomic assessment in the pre-/post-treatment blood and tumor tissue.
The trial is ongoing and thus far has accrued 25 patients. All patients were originally treated at Washington University, St Louis, Missouri, but the trial now also includes the Dana Farber Institute.
Most patients had tumors in the oral cavity (n = 18), followed by the larynx/hypopharynx (n = 5) and oropharynx (n = 2).
The cohort in this trial is human papillomavirus negative and predominantly male (n = 18), with "significant histories of tobacco and alcohol abuse," Dr Uppaluri said.
Except for one patient with stage III disease, all patients had stage IV tumors. In addition, 92% had clinical stage T4 disease, "which is consistent with this high-risk subgroup of patients."
Dr Uppaluri noted that none of the 14 patients have had any recurrence to date, and all of them had T4 disease.
"We define tumor response as a change in the tumor on exam or on CT scan or pathology," he said. "And 50% of patients had some sort of evidence of a response in the neoadjuvant phase."
For example, patient 20, who had stage IV disease, has since been downstaged to stage I disease (pT1NO).
Another finding was the extent of the "treatment effect," which is defined as tumor necrosis and/or giant cell/histiocytic reaction to keratinous debris in more than 10% of tumor area. "Overall, 42% of patients had some sort of a treatment effect either in the primary tumor or lymph nodes," he noted. "In addition, 25% had a major effect of more than 50%."
There was also a significant correlation between baseline programmed cell death ligand 1 (PD-L1) expression on tumor cells and pathologic treatment effect in the tumor (correlation coefficient = 0.72; P = .0005).
A preliminary analysis also showed a correlation between treatment effect and CD8 cells (P = .0051), PD-L1 + CD8 T cells (P = .022), and CD4 T cells (P = .048).
Dr Uppaluri pointed out that the patient numbers are small and it is early data. "But there are significant implications with the signals we have seen," he said, "including reduced need for postoperative adjuvant chemoradiation, potential to reduce the extent of surgery, and reduction in relapse risk."
It's All About Timing
In a discussion of the paper, Antonio Jimeno, MD, PhD, professor in the Division of Medical Oncology at the University of Colorado, Denver, explained that one of the key questions about integrating immunotherapy in the treatment regimen is the timing.
"There are factors that favor preoperative immune modulations, such as the potential for downstaging and de-intensification of radiation therapy and chemoradiation therapy," Dr Jimeno said. "But there are also factors favoring use in the postoperative setting. Debulking has been advocated as a favorable factor for immune modulation."
But several ongoing studies are investigating peri- or postsurgery immune-directed therapy, he explained. "In the absence of strong rationale, randomized studies are needed to identify the optimal sequence."
Overall, the addition of immune modulation appears to be feasible in contemporary HNSCC. "With the caveat of being nonrandomized, it added minimal toxicity, there were favorable comparisons with historical outcomes data, and pathologic indications of an antitumor effect," said Dr Jimeno.
"Importantly, it did not lead to delays in curative intent surgery," he added. "But robust translational investigations are needed to develop rational patient selection strategies and to continue investigating optimal timing of immune modulation."
The study was funded by Merck, V Foundation, and National Institutes of Health. Dr Uppaluri has disclosed a financial relationship with Merck. Dr Jimeno has disclosed relationships with Merimack andSuvica, as well as the Gates, Mordecai, Grant, and Karsh Family Foundations.
American Society of Clinical Oncology (ASCO) 2017 Annual Meeting. Abstract 6012. Presented June 6, 2017.